id stringlengths 3 9 | source stringclasses 1 value | version stringclasses 1 value | text stringlengths 1.54k 298k | added stringdate 1993-11-25 05:05:38 2024-09-20 15:30:25 | created stringdate 1-01-01 00:00:00 2024-07-31 00:00:00 | metadata dict |
|---|---|---|---|---|---|---|
18810832 | pes2o/s2orc | v3-fos-license | Much Ado About Nothing? A Real-World Study of Patients with Type 2 Diabetes Switching Basal Insulin Analogs
Introduction Type-2 diabetes mellitus (T2DM) is a progressive disease, and many patients eventually require insulin therapy. This study examined real-world outcomes of switching basal insulin analogs among patients with T2DM. Methods Using two large United States administrative claims databases (IMPACT® and Humana®), this longitudinal retrospective study examined two cohorts of adult patients with T2DM. Previously on insulin glargine, Cohort 1 either continued insulin glargine (GLA-C) or switched to insulin detemir (DET-S), while Cohort 2 was previously on insulin detemir, and either continued insulin detemir (DET-C) or switched to insulin glargine (GLA-S). One-year follow-up treatment persistence and adherence, glycated hemoglobin (HbA1c), hypoglycemia events, healthcare utilization and costs were assessed. Selection bias was minimized by propensity score matching between treatment groups within each cohort. Results A total of 5,921 patients (mean age 60 years, female 50.0%, HbA1c 8.6%) were included in the analysis (Cohort 1: IMPACT®: n = 536 DET-S matched to n = 2,668 GLA-C; Humana®: n = 256 DET-S matched to n = 1,262 GLA-C; Cohort 2: n = 419 GLA-S matched to n = 780 DET-C), with similar baseline characteristics between treatment groups in each cohort. During 1-year follow-up, in Cohort 1, DET-S patients, when compared with GLA-C patients, had lower treatment persistence/adherence with 33–40% restarting insulin glargine, higher rapid-acting insulin use, worse HbA1c outcomes, significantly higher diabetes drug costs, and similar hypoglycemia rates, health care utilization and total costs. However, in Cohort 2 overall opposite outcomes were observed and only 19.8% GLA-S patients restarted insulin detemir. Conclusions This study showed contrasting clinical and economic outcomes when patients with T2DM switched basal insulin analogs, with worse outcomes observed for patients switching from insulin glargine to insulin detemir and improved outcomes when switching from insulin detemir to insulin glargine. Further investigation into the therapeutic interchangeability of insulin glargine and insulin detemir in the real-world setting is needed. Electronic supplementary material The online version of this article (doi:10.1007/s12325-014-0120-1) contains supplementary material, which is available to authorized users.
INTRODUCTION
Guidelines from the American Diabetes Association (ADA) and European Association for the Study of Diabetes (EASD) note that most patients with long-standing type 2 diabetes mellitus (T2DM) will need insulin at some point during their treatment [1]. Although the long-acting insulin analogs glargine and detemir improve glycemic control with a low risk of hypoglycemia in patients with T2DM [2], initiation of insulin treatment is often delayed [3].
Many studies have been conducted to examine the initiation of insulin glargine and insulin detemir among T2DM patients. A 52-week, open-label study compared outcomes in patients with T2DM treated with insulin glargine or insulin detemir as part of a basalbolus regimen (n = 319) [4]. Both insulin glargine and insulin detemir improved glycemic control, with similar overall effects in blood sugar control and low incidence of hypoglycemia for both. A Cochrane Database Review comparing insulin glargine and insulin detemir in randomized clinical trials likewise concluded that there were no differences in glycemic control between these two insulins, although twice-daily dosing and higher doses were more often needed with insulin detemir than with insulin glargine [5]. In addition, the ''Effect of Insulin Detemir and Insulin Glargine on Blood Glucose Control in Subjects With Type 2 Diabetes'' trial (EFFICACY: NCT00909480)-investigating the once-daily dosing of basal insulin as add-on to metformin over 26 weeks-showed that, while both improved glycemic control when added to metformin, the use of insulin glargine resulted in greater reductions in glycated hemoglobin (HbA 1c ) as compared with insulin detemir [6].
In addition to these clinical studies, a few studies have also been conducted to examine the effects of these long-acting insulin analogs in a real-world setting. These real-world studies have reported somewhat conflicting data. Xie et al., for example, reported that patients initiating insulin glargine were more likely to persist with and adhere to treatment, and to show better glycemic control and similar overall hypoglycemia rate with no increase in healthcare costs [7] or weight change [8], compared with those initiating insulin detemir. In contrast, a study of 306 patients enrolled in a large United States (US) managed care organization found that glycemic control after 180 days was similar for insulin glargine and insulin detemir [9]. In addition, Variability Evaluation'' (PREDICTIVE: NCT00659295) study was an observational study designed to evaluate switching from insulin glargine to insulin detemir (primarily with respect to adverse events) in patients with type 1 or T2DM [12]. In a European cohort of patients (n = 777) in the PREDICTIVE study, there were significant improvements in HbA 1c with fewer hypoglycemic events after switching to insulin detemir [13]. Using the framework of costeffectiveness analysis, an Asian study based on data from a Korean cohort of the PREDICTIVE study reported reduced total diabetes care costs and increased life expectancy of 0.06 years after switching from insulin glargine to insulin detemir [14]. A small retrospective study recently evaluated outcomes in patients (n = 10 with type 1 diabetes and n = 21 with T2DM) switching from insulin glargine to insulin detemir due to changes in Medicaid formulary coverage [15] Two cohorts were identified for the analyses in this study. Cohort 1 included patients who were treated with insulin glargine and subsequently either switched to insulin detemir (DET-S) or remained on insulin glargine (GLA-C). For the DET-S group, the index date is the initial switching date to detemir. The GLA-C group consisted of patients with C3 prescription drug claims for insulin glargine and no claims for insulin detemir, and their index date was chosen as a randomly selected date between the third and last insulin glargine claims dates. For both groups, patients were required to have at least one pharmacy claim for insulin glargine in each quarter during the baseline period, but not for premix or other basal insulin. Similarly, Cohort 2 included patients treated with insulin detemir and who subsequently either switched to insulin glargine (GLA-S) or remained on insulin detemir (DET-C). For the GLA-S group, the index date is the initial switching date to glargine. The DET-C group consisted of patients with C3 prescription drug claims for insulin detemir and no claims for insulin glargine, and their index date was chosen as a randomly selected date between the third and last insulin detemir claims dates. For both groups, patients were required to have at least one insulin detemir claim with each quarter, but no pharmacy claims for premix or other basal insulin during the baseline period.
Assessments
Baseline demographic and clinical characteristics were examined for all patients, using data recorded within 6 months before the index date. These variables included age at the index date, gender, region, HbA 1c level, oral anti-diabetes drugs (OAD) and rapid/regular insulin use, comorbidities, hypoglycemia rates, types of insurance coverage, copay of the index drug, device type (vial vs pen) of the baseline insulin therapy and of the index insulin therapy, and total and diabetes-related healthcare utilization and costs. Baseline HbA 1c data were from HbA 1c test level dated between 6 months before the index date and
Data Analysis
To address potential selection bias in the realworld setting, patients in both cohorts were matched by propensity score matching (PSM) to balance baseline demographic, clinical, and economic characteristics. The variables included in the PSM model were age, gender, copay, health plan, geographic region, initial year, baseline comorbidities, anti-diabetes drug, concomitant medications, baseline HbA 1c categories, all-cause and diabetes-related utilizations and costs, diabetes education and initiation device. The nearest neighbor PSM technique was used. Propensity scores were estimated by unconditional logistic regression analyses that incorporated potential predictors of therapy as independent variables in the regression and group status (e.g., DET-S vs GLA-C) as the outcome. In Cohort 1, a 1 up to 5 ratio was applied to DET-S and GLA-C patients. This was done separately for the IMPACT Ò and Humana Ò databases. For Cohort 2, a 1 up to 2 ratio was applied to DET-C and GLA-S patients, with the two independent databases being pooled together. These ratios and matchings were chosen because our preliminary analysis revealed a prescription imbalance, with significantly more eligible patients in the insulin glargine group when compared with the insulin detemir group, and a much lower number of patients in the switcher group as compared to the continuers group.
Additionally, one-to-many matching has previously been validated as a method to increase precision in cohort studies when compared with one-to-one matching, and has also been supported in a recent review assessing the quality of statistical methodologies in matched case-control studies [25,26]. Matching was implemented without replacement and any patient without at least one match was excluded from the analysis.
Between-group covariate balance was evaluated using descriptive t tests and v 2 tests (with corresponding P values) and standardized differences, where a standardized difference \10 indicated adequate balance [27].
All analyses were performed using the intention-to-treat (ITT) approach on the matched patients. Due to the study design, DET-C or GLA-C patients could not, respectively, switch to insulin glargine or Table 1 Post-PSM baseline characteristics in Cohort 1 and Cohort 2 Table 1).
Insulin Utilization
During the 1-year follow-up period, in Cohort 1, patients switching from insulin glargine to insulin detemir, compared with patients continuing on insulin glargine, showed significantly lower rates of index basal insulin treatment persistence, shorter duration of persistence (Fig. 2a, b) and lower rate of adherence (Fig. 3a, b). These findings were consistent in both commercially insured IMPACT Ò population, and Medicare-insured Humana Ò elderly population. In contrast, in Cohort 2, GLA-S patients, as compared with DET-C patients, showed numerically, but not statistically significantly, higher index basal insulin treatment persistence (Fig. 2a, b), and significantly higher rates of treatment adherence (Fig. 3a, b).
The DACON was similar for DET-S and GLA-C patients (Fig. 4). During the follow-up, RAI utilization was significantly lower among GLA-
Clinical Outcomes
At the end of the follow-up year, in Cohort 1, mean HbA 1c was significantly higher among Table 1 continued HbA 1c glycated hemoglobin, OAD oral anti-diabetes drug, RAI rapid acting insulin, SD standard deviation DET-S patients compared with GLA-C patients (Fig. 6a), and the mean change in HbA 1c from baseline was lower in the DET-S group than GLA-C group (only statistically significant in the IMPACT Ò database) (Fig. 6b). Significantly more GLA-C patients achieved a target HbA 1c \7.0% in the Humana Ò group and a target HbA 1c \8.0% in both the IMPACT Ò and Humana Ò groups (Fig. 6c).
In Cohort 2, although mean HbA 1c was not significantly different at the end of follow-up (Fig. 6a), a greater reduction from baseline in . 6b). Additionally, in the GLA-S group, at the end of 1-year follow-up, compared with those patients who remained on insulin glargine, those who restarted insulin detemir had higher HbA 1c (8.86% vs 8.23%, P = 0.0539) and lower HbA 1c reduction from baseline (-0.10% vs -0.81%; P = 0.0172). More patients in the GLA-S group achieved HbA 1c goals during follow-up than did those in the DET-C group (Fig. 6c).
Although
for Cohort 1 overall hypoglycemia prevalence rates were significantly higher for GLA-C in the Humana Ò database, overall both prevalence and event rates for Cohort 1 and 2 were similar between switchers and continuers, as were severe hospital/ED-related hypoglycemia rates (Table 2).
Economic Outcomes
Overall healthcare costs were similar across all four treatment cohorts (Table 3). Although healthcare utilization did not differ according to treatment in Cohort 1, DET-S patients had significantly higher diabetes drug (P\0.0001) and diabetes supply costs (P = 0.0006) than GLA-C patients in the IMPACT Ò cohort; while in the Humana Ò cohort only diabetes drug costs were significantly higher for DET-S patients than for GLA-C patients (P = 0.0368; Table 3).
Total healthcare expenditure was $21,845 in the DET-C cohort and $24,417 in the GLA-S cohort and total expenditure for diabetesrelated healthcare was $10,293 and $10,619, respectively. These differences were not statistically significant.
Sensitivity Analysis
For index basal insulin treatment persistence, both 75th and 95th percentiles of the refill time were used to estimate length of persistence and yielded similar results. For follow-up HbA 1c analysis, generalized linear regression was used to estimate the changes in HbA 1c from baseline, and the LOCF approach was also employed. Both approaches yielded similar results on both HbA 1c levels and proportions of patients achieving glycemic targets as the primary analysis. Finally, PSM was also conducted using 1:1 ratio and overall similar results were observed (data not shown).
DISCUSSION
This real-world US study investigated the effects of switching from insulin glargine to insulin detemir (Cohort 1) or from insulin detemir to insulin glargine (Cohort 2), compared with not switching and remaining on baseline treatment. Although hypoglycemia prevalence rates were higher for GLA-C patients from the Humana Ò group, when compared with DET-S patients, rates were similar overall. Rates of more severe inpatient/ED-related hypoglycemia were low, and similar, in both groups. Healthcare utilization and total costs were also similar in both groups, but the DET-S group had higher diabetes drug and supply costs than did the GLA-C group. Similar results were generally found in the younger population of commercially insured patients (IMPACT Ò cohort) and the older Medicare population (Humana Ò cohort).
In Cohort 2, contrasting results were observed, GLA-S patients were more adherent during 1 year of follow-up, had greater HbA 1c reduction from baseline, and had a significant increase in likelihood of achieving HbA 1c goal \8% when compared with DET-C patients. The improved outcomes observed in the GLA-S group were achieved with similar rates of hypoglycemia, heath care utilization, and costs as in the DET-C group. Table 3 continued
Rx prescription drug
These data support previously published studies that show a low incidence of hypoglycemia among patients with T2DM treated with either insulin glargine or insulin detemir [4,5,15]. Similar to the PREDICTIVE trial [13], the prevalence rate of hypoglycemia was lower after switching from insulin glargine to insulin detemir in the Medicare cohort in our study; however, the event rates were similar in both groups. No difference in hypoglycemia was observed in the commercially insured IMPACT Ò population. In contrast to previously published studies showing that insulin dose was typically higher with insulin detemir than insulin glargine [5,15,28], it was similar among the matched GLA-C and DET-S patients based on the DACON estimate from the pharmacy claims data. In our study, however, DET-S patients had a much higher rate of RAI use than did GLA-C patients. Although dosing frequency information was not available for our data analysis, existing literature suggests that insulin detemir is more likely to be dosed more frequently than insulin glargine [15,29]. Both higher twice-daily use and RAI use may explain the lower persistence and adherence rates and higher diabetes drug and supply costs in the DET-S group.
Data from Cohort 1 and Cohort 2 suggest that continued use of insulin glargine, as opposed to switching to insulin detemir, or switching from insulin detemir to insulin glargine, is associated with improved glycemic control, despite similar baseline HbA 1c levels in both groups and higher RAI use in the DET-S group during the follow-up period. Although statistically significant, the difference between the basal insulins with regard to HbA 1c reduction from baseline was 0.32% for GLA-S vs DET-C and 0.2% for GLA-C vs DET-S. The clinical relevance of differences of such magnitudes is unclear. Nonetheless, up to 8.7% more GLA-S patients achieved a target HbA 1c \8% during follow-up and as many as 8.0% more GLA-C patients achieved an HbA 1c \7% compared with DET-S. When one considers the enormity of the T2DM pandemic, such increases could represent a large number of patients. Furthermore, reductions in HbA 1c of 0.9% are associated with a significant reduction in the risk of microvascular complications associated with T2DM and a reduction in HbA 1c of just 0.6% leads to a reduction in risk of myocardial infarction and of diabetes-related and all-cause mortality [1].
The improvements in glycemic control observed in the current study conflict with results from the PREDICTIVE trial [13,14] but are consistent with those from two US studies [15,16]. In our study, the treatment persistence rate was significantly higher in the GLA-C group and medication adherence was significantly higher among GLA-S patients.
Importantly, a positive association between HbA 1c reduction and treatment persistence among T2DM patients receiving insulin therapy has previously been shown [7]. Also, similar to the results from the RELISH study [16], a significant portion of DET-S patients restarted insulin glargine after switching to insulin detemir. In contrast, 19 Our study examined issues similar to those of previously conducted small-scale US studies of patients switching from insulin glargine to insulin detemir [15,16]. However, the major strengths of our study compared to previous studies included that we examined a much larger study population from two independent cohorts, and that stringent PSM was used to balance the baseline characteristics of patients, thereby reducing potential confounders when interpreting results. Overall, the findings of this study support the use of insulin glargine, whether through continuation or switching, and may call for a careful re-examination of the therapeutic interchangeability of the two basal insulin analogs in the real-world setting. Indeed, these findings suggest that switching patients from insulin glargine to insulin detemir is more disruptive than switching patients from insulin detemir to insulin glargine. However, caution also needs to be exercised when interpreting these results due to the retrospective nature of this analysis.
CONCLUSION
In summary, the results from this study suggest that, in a real-world US managed care setting, switching patients with T2DM, including the elderly, from insulin detemir to insulin glargine, or maintaining patients on insulin glargine rather than switching to insulin detemir, may improve treatment persistence/ adherence and enhance glycemic outcomes.
This is achieved without implications regarding hypoglycemia or healthcare costs. These results suggest that, in the real-world setting, the long-acting basal insulin formulations, insulin glargine and insulin detemir, may not be therapeutically interchangeable. Further prospective studies, such as randomized pragmatic trials with a cross-over design, are needed to confirm the therapeutic non-interchangeability of insulin glargine and insulin detemir.
ACKNOWLEDGMENTS
Sponsorship and article processing charges for this study were funded by Sanofi US, Inc.
(Bridgewater, USA). The authors received writing/editorial support in the preparation of this manuscript, provided by Pim Dekker, PhD, of Excerpta Medica (Amsterdam, The Netherlands), funded by Sanofi US, Inc.
All authors had full access to all of the data in this study and take complete responsibility for the integrity of the data and accuracy of the data analysis.
All named authors meet the ICMJE criteria for authorship for this manuscript, take responsibility for the integrity of the work as a whole, and have given final approval to the version to be published.
Wenhui Wei, Steve Zhou, Onur Baser, and Jasvinder Gill contributed to the concept and design of the study, interpretation of the data, and drafting of the manuscript. Raymond Miao, Chunshen Pan, and Lin Xie assisted in the design of the study and data collection and contributed to the writing of the manuscript.
All authors read and approved the final manuscript. | 2018-04-03T05:22:35.055Z | 2014-05-16T00:00:00.000 | {
"year": 2014,
"sha1": "365de43612392e110b545a5f41dab77010773a97",
"oa_license": "CCBYNC",
"oa_url": "https://link.springer.com/content/pdf/10.1007/s12325-014-0120-1.pdf",
"oa_status": "HYBRID",
"pdf_src": "PubMedCentral",
"pdf_hash": "365de43612392e110b545a5f41dab77010773a97",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
11223552 | pes2o/s2orc | v3-fos-license | HIV-1 and HIV-2 exhibit similar mutation frequencies and spectra in the absence of G-to-A hypermutation
Background Human immunodeficiency virus type 2 (HIV-2) is often distinguished clinically by lower viral loads, reduced transmissibility, and longer asymptomatic periods than for human immunodeficiency virus type 1 (HIV-1). Differences in the mutation frequencies of HIV-1 and HIV-2 have been hypothesized to contribute to the attenuated progression of HIV-2 observed clinically. Results To address this hypothesis, we performed Illumina sequencing of multiple amplicons prepared from cells infected with HIV-1 or HIV-2, resulting in ~4.7 million read pairs and the identification of ~200,000 mutations after data processing. We observed that: (1) HIV-2 displayed significantly lower total mutation, substitution, and transition mutation frequencies than that of HIV-1, along with a mutation spectrum markedly less biased toward G-to-A transitions, (2) G-to-A hypermutation consistent with the activity of APOBEC3 proteins was observed for both HIV-1 and HIV-2 despite the presence of Vif, (3) G-to-A hypermutation was significantly higher for HIV-1 than for HIV-2, and (4) HIV-1 and HIV-2 total mutation frequencies were not significantly different in the absence of G-to-A hypermutants. Conclusions Taken together, these data demonstrate that HIV-2 exhibits a distinct mutational spectrum and a lower mutation frequency relative to HIV-1. However, the observed differences were primarily due to reduced levels of G-to-A hypermutation for HIV-2. These findings suggest that HIV-2 may be less susceptible than HIV-1 to APOBEC3-mediated hypermutation, but that the fidelities of other mutational sources (such as reverse transcriptase) are relatively similar for HIV-1 and HIV-2. Overall, these data imply that differences in replication fidelity are likely not a major contributing factor to the unique clinical features of HIV-2 infection. Electronic supplementary material The online version of this article (doi:10.1186/s12977-015-0180-6) contains supplementary material, which is available to authorized users.
Background
Human immunodeficiency virus type 1 (HIV-1) infects approximately 35 million individuals worldwide and has resulted in about 39 million deaths since the onset of the AIDS pandemic (http://www.unaids.org). Within infected individuals, HIV-1 undergoes rapid genetic diversification, promoting the acquisition of drug resistance, evasion of the host immune response, and alterations in cell tropism. Genetic diversification is, in turn, driven by large population sizes and high rates of replication, recombination, and mutation. HIV-1 has been found to mutate on the order of 10 −4 to 10 −5 mutations/base pair/cycle (m/bp/c), corresponding to ~0.1-1 mutations per genome synthesized [1][2][3][4][5]. HIV-1 thus mutates about 10,000-100,000 times faster than eukaryotic genomic DNA [6]. Multiple viral and cellular factors influence both the rates and types of mutations produced during viral replication. Reverse transcriptase (RT) is likely a major generator of mutations in vivo, as it is tremendously error-prone in vitro (mutation rates >10 −4 m/bp/c), primarily due to a lack of proofreading activity [7,8]. RNA polymerase II can also introduce mutations when transcribing the viral RNA genome, but less than that of RT [3]. Cellular DNA polymerases can introduce mutations when replicating the proviral DNA integrated into the cellular genome; however, these enzymes have been assumed to minimally contribute to HIV-1 variation due to the high fidelity of genomic DNA replication [6]. In addition, apolipoprotein B mRNA-editing enzyme catalytic polypeptide-like 3 (APOBEC3) proteins can induce G-to-A hypermutation, particularly in the absence of Vif [9,10]. APOBEC3-mediated hypermutation is often lethal to virus replication, but accumulating evidence suggests that APOBEC3 proteins can in some cases promote genetic diversification and the evolution of new variants conferring drug resistance or altered cell tropism [11][12][13][14][15]. Other factors, such as dNTP pool levels [16], other viral proteins (Vpr) [17,18], and cell type [19] have been shown to influence HIV-1 mutagenesis as well.
Relative to HIV-1, human immunodeficiency virus type 2 (HIV-2) infection is often marked clinically by lower viral loads, reduced transmissibility, and longer asymptomatic periods in infected individuals [20,21]. Differences in the mutation frequencies of HIV-1 and HIV-2 have been hypothesized to contribute to the attenuated progression of HIV-2 observed clinically. Specifically, a lower mutation rate for HIV-2 would be expected to limit genetic diversification, which could in turn reduce viral fitness and/or attenuate viral pathogenicity. Consistent with this hypothesis, viral variants with increased replication fidelity have been shown to result in impaired viral fitness and virulence in many other RNA viruses [22][23][24][25][26]. In further support of this hypothesis, HIV-2 has been found to evolve less quickly than HIV-1 [27,28], though one report has found the opposite [29]. However, it should be noted that evolutionary rates depend on a wide variety of factors besides the mutation rate, such as replication rate, population size, and selective pressures. In addition, the HIV-2 RT contains a highly conserved V75I polymorphism, a drug resistance-associated mutation in HIV-1 that improves RT fidelity [30,31]. Lastly, HIV-2 has been shown to be less sensitive than HIV-1 to APOBEC3G activity, potentially diminishing the contribution of APOBEC3G to viral mutagenesis [32].
In order to compare mutation frequencies and spectra between HIV-1 and 2, we performed Illumina highthroughput sequencing of proviral DNA from cells infected with HIV-1 or 2. We found that HIV-2 displayed lower total, substitution, and transition mutation frequencies than HIV-1, particularly due to reduced levels of G-to-A transition mutations. We also observed lowlevel G-to-A hypermutation in both HIV-1 and HIV-2 that was consistent with the activity of APOBEC3 proteins. Intriguingly, HIV-2 demonstrated significantly lower levels of G-to-A hypermutation than HIV-1. After exclusion of G-to-A hypermutants, total mutation frequencies were not significantly different between HIV-1 and HIV-2. Together, these data support the conclusion that differences in general replication fidelity are likely not a primary driver of the unique clinical features of HIV-2 infection.
Characterization of background errors induced by PCR and Illumina sequencing
In order to compare mutation frequencies and spectra between HIV-1 and HIV-2, single cycle infections with HIV-1 or HIV-2 were performed at a high MOI (1 million U373-MAGI-X4 cells per replicate infected at an MOI of 1.0, see "Methods"; Figure 1). In this assay, producer cells cannot be re-infected due to a lack of the appropriate receptor and co-receptor, and target cells likewise cannot be re-infected due to disruption of the env genes in the HIV-1 and HIV-2 vectors. Genomic DNA was purified from infected cells and first subjected to quantitative PCR (qPCR) in order to determine the level of plasmid carryover from transfections. Plasmid carryover was quantified either by: (1) determining the plasmid backbone copy number (by measuring the ampicillin resistance gene) and dividing by the proviral copy number, or (2) determining the proviral copy number from heatinactivated viral infections and dividing by the proviral copy number from un-treated infections (see "Methods"). We found that the level of plasmid carryover for HIV-1 was 0.2% when measured by either method, while the level of carryover was 2.8 or 1.4% for HIV-2, depending on the approach used (Additional file 1: Table S1). The significantly higher level of plasmid carryover for HIV-2 likely reflects the reduced infectivity of HIV-2 viral stocks, which resulted in larger volumes of viral stocks being used during infection. These results are comparable to those obtained in another study [5] and are too low to significantly impact measured mutation frequencies. Next, amplicons were prepared from proviral DNA for Illumina sequencing. In total, 12 samples were analyzed-three experimental replicates each of HIV-1, HIV-2, and HIV-1 and HIV-2 plasmid amplifications as controls to determine levels of background errors. Further, for each sample, five amplicons were prepared (Gag, Vif, HSA, EGFP-1, and EGFP-2), representing a mixture of viral (Gag, Vif ) and marker (HSA, EGFP-1, EGFP-2) gene targets. Libraries were prepared individually from samples in order to prevent inter-sample recombination during library construction. Following this, all libraries were pooled and subjected to 2 × 150 paired-end sequencing on the Illumina MiSeq, resulting in ~4.7 million total read pairs after processing, or an average of ~79,000 read pairs/amplicon/sample (Additional file 2: Table S2). After stringent filtering of Illumina data, the mutation frequencies (expressed as mutations per base pair, or m/bp) were determined for all samples, both in terms of total mutations and every possible subdivision (i.e. substitutions, transitions, transversions, etc.). Mutation counts, frequencies, and relative percentages are listed in Additional file 3: Dataset S1, both combined across all five amplicons and separated by amplicon.
After sequencing, the first objective was to utilize the plasmid controls to characterize the frequencies and spectra of background errors (i.e. errors from PCR or sequencing) in order to determine the extent to which biological mutations could be detected above the level of the background. The average mutation frequencies of the plasmid controls were 2.8 (HIV-1) and 2.6 (HIV-2) ×10 −4 m/bp (Figure 2a; Additional file 3: Dataset S1), consistent with a recent investigation into background error during amplicon sequencing on the Illumina MiSeq [33]. Most of the background errors observed were substitutions, with insertions and deletions comprising only 4.3-4.9% of total mutations. Of the substitutions, transversions and transitions were observed at similar frequencies: 1.2-1.4 × 10 −4 m/bp for transversions and ~1.2 × 10 −4 m/bp for transitions, such that each category composed ~half of all mutations (Figure 2c). However, as described in "Methods", plasmid error hotspots (i.e. positions highly prone to background error) were masked prior to this analysis, and in the absence of such masking transversions occurred ~2.5-fold more often than transitions. Thus, our findings are consistent with previous reports that have observed a bias of Illumina sequencing (using a variety of different platforms and library preparation methods) toward transversion types [33][34][35][36]. Interestingly, among the eight possible transversion types, background transversion errors were non-randomly distributed (Additional file 4: Figure S1). Specifically, background transversions were strongly biased toward C-to-A and G-to-T transversions, which composed 74% (HIV-1) or 73% (HIV-2) of all transversions. C-to-A and G-to-T are reciprocal mutational types, and these mutations may be due to oxidative conversion of guanine to 8-oxoguanine or due to Illumina imaging artifacts (see "Discussion"). Background errors were distributed quite evenly across the five amplicons (Figure 2b), demonstrating that no individual amplicon was particularly prone to background errors.
In addition to analyzing background error frequencies and spectra, the frequency of intra-sample recombination due to PCR was also examined. While PCR-mediated recombination is not known to be mutagenic, recombination could affect associations between mutations. We attempted to minimize recombination from PCR by stopping reactions after 30 cycles of amplification, corresponding to the ~end of the log-linear phase of amplification, after which recombination occurs at a much higher rate due to saturation of dNTPs and primers [37][38][39]. Intra-sample recombination frequencies were determined using genetic markers that were incorporated into the plasmid control amplifications (see "Methods"). Under these amplification conditions, intra-sample Experimental strategy for investigating HIV-1 and HIV-2 mutagenesis by Illumina DNA sequencing. Vector virus stocks were produced by co-transfecting 293T cells with HIV-1 or HIV-2 Env-deficient vectors and HIV-1 or HIV-2 CXCR4-tropic Env expression constructs. Virus stocks were concentrated, DNase I-treated to reduce plasmid carryover, and titered in U373-MAGI cells. To prepare samples for Illumina sequencing, 1 × 10 6 U373-MAGI cells were infected at an MOI of 1.0, generating approximately 1 × 10 6 proviruses per experimental replicate. This assay represents a single round of viral replication, as producer cells and target cells cannot be re-infected, due to a lack of receptor or Env expression, respectively. Polymerase chain reaction (PCR) of five amplicons (Gag, Vif, HSA, EGFP-1, and EGFP-2) was performed from the proviral DNA. Amplicons from the HIV-1 and HIV-2 proviral DNAs were either identical (HSA, EGFP-1 and 2) or homologous (Gag and Vif ) in sequence. The EGFP-1 and EGFP-2 amplicons represent non-overlapping segments of the egfp gene. Sequencing libraries were prepared from the amplicons, pooled in an equimolar fashion to normalize coverage, and subjected to 2 ×150 bp sequencing on the Illumina MiSeq, generating approximately 4.7 million read pairs after data processing. recombination was observed at frequencies of ~2-3% of the maximum observable in the assay, demonstrating that recombination from PCR was relatively rare (Additional file 5: Table S3).
We next compared the mutation frequencies and spectra of the HIV-1 and 2 biological samples to that of the plasmid controls in order to determine which types of mutations could be detected above background levels. HIV-1 and HIV-2 exhibited average mutation frequencies of 6.9 (HIV-1) and 3.1 (HIV-2) × 10 −4 m/bp (Figure 2a; Additional file 3: Dataset S1). The total mutation frequency of HIV-1 was significantly higher than the corresponding plasmid control (p < 0.001). In contrast, the HIV-2 total mutation frequency was not significantly higher than the plasmid control (p = 0.40). Upon separating out the major classes of mutations, we found that this was primarily due to high levels of transversions in the plasmid controls. Transversions occurred at frequencies of 1.4 (HIV-1) or 1.2 (HIV-2) × 10 −4 m/ bp in the plasmid controls (Figure 2a), values that were not significantly different from that of the biological samples. Similar to transversions, insertion frequencies did not significantly vary between biological samples and plasmid controls, and deletion frequencies were only significantly higher than plasmid for HIV-2 (p = 0.02, Additional file 3: Dataset S1). However, insertions and deletions had little impact on overall mutation frequencies because they were much less frequent (3-6% of total mutations) than substitutions ( Figure 2c). Notably, although the HIV-2 total mutation frequency was not significantly higher than its corresponding plasmid control, transition frequencies [5.8 (HIV-1) and 2.0 (HIV-2) × 10 −4 m/bp] were significantly higher than the plasmid controls for both viruses (HIV-1: p < 0.001; HIV-2: p = 0.038). These differences were also reflected in the mutation spectra, as transitions comprised 84% (HIV-1) Figure 2 HIV-2 has a lower mutation frequency and distinct mutation spectrum relative to HIV-1. a Mutation frequency analysis. Mutation frequencies were calculated by dividing the number of mutations by the number of reference bases (mutations + wild-type bases) and are expressed as mutations/bp, or m/bp. Mutation frequencies were determined for HIV-1 and HIV-2, as well as for plasmid controls to assess background error levels. b Transition frequency analysis. Transition frequencies were compared across the five different amplicons subjected to Illumina DNA sequencing. c Mutation spectra analysis. Mutation spectra were determined by dividing the frequency of each type of mutation by the total mutation frequency, with the results expressed as a percentage of total mutations. Data in all panels represent the mean of three experimental replicates, with error bars indicating the standard deviation. Asterisks denote statistically significant differences between HIV-1 and 2 (*p < 0.05, ***p < 0.001, N.S. not significant). The actual numbers of read pairs, mutations, and reference bases are listed in Additional file 3: Dataset S1. or 66% (HIV-2) of total mutations in the biological samples but only 44% (HIV-1) or 48% (HIV-2) in the plasmid controls ( Figure 2c). Importantly, many previous reports have demonstrated that transitions predominate during the replication of HIV-1, generally comprising 70-90% of all substitutions [1-4, 17, 19, 40]. Thus, considering that transitions are more relevant to HIV-1 replication and that they were detected at levels significantly higher than the background, most downstream analyses focused on transition mutational types.
HIV-2 exhibits a lower mutation frequency and an altered mutation spectrum relative to that of HIV-1
We next compared HIV-1 and HIV-2 mutation frequencies in order to test our initial hypothesis that HIV-2 would display a lower mutation frequency than HIV-1. We found that HIV-1 had a significantly higher total mutation frequency, as well as higher frequencies of substitutions and transitions, than HIV-2 (relative differences of 2.3, 2.3, and 2.9-fold, respectively; all p values <0.001) (Additional file 3: Dataset S1; Figure 2a). In contrast, the levels of transversion mutations were not significantly different between HIV-1 and HIV-2 (p = 0.32). The observed differences in transition frequencies were primarily due to an approximately 6.9-fold higher frequency (p < 0.001) of G-to-A transition mutations with HIV-1 than with HIV-2 (Additional file 3: Dataset S1). Less striking, but statistically significant, differences were observed for several other types of transitions as well. HIV-1 displayed a 1.1-fold higher frequency of A-to-G transitions (p = 0.049) than HIV-2, while HIV-2 displayed a 1.5-fold higher frequency of C-to-T transitions than HIV-1 (p = 0.0019). The frequency of T-to-C transitions was not significantly different between HIV-1 and HIV-2. Consistent with these results, the mutation spectrum of HIV-1 was much more heavily biased toward G-to-A transition mutations than for HIV-2 (66 vs. 22% of total mutations, Figure 2c). Overall, these data demonstrate that the HIV-1 mutation frequency is significantly higher than that of HIV-2, predominantly due to substantially higher levels of G-to-A transition mutations.
Mutation frequencies vary across amplicons for HIV-1 and HIV-2
We hypothesized that specific amplicons might be more or less error-prone than others due to features of the primary sequence (such as homopolymeric runs), secondary structures, or the relative positioning of the amplicon within the viral genome. To address this, transition frequencies were compared across the five amplicons analyzed in this study. As previously indicated, transitions in the plasmid controls were distributed relatively evenly across the five amplicons, demonstrating that no individual amplicon was particularly prone to background transition mutations ( Figure 2b). In contrast, for HIV-1, it was found that most transitions were concentrated in the EGFP-1 amplicon and, to a lesser extent, the Vif amplicon ( Figure 2b). The EGFP-1 and Vif amplicons together accounted for about 74% of all transitions in HIV-1. The relative order of transition frequencies among the amplicons was found to be EGFP-1 > Vif > EGFP-2 ≈ HSA ≈ Gag, with all indicated differences between amplicon pairs being statistically significant (p < 0.001).
Since G-to-A transitions predominated for HIV-1, the observed differences could potentially be explained by varying nucleoside content between amplicons. However, the amplicons contained relatively similar frequencies of deoxyguanosine, ranging from 19.6 to 31.2%. Further, the Vif amplicon actually contained the lowest deoxyguanosine content of the five amplicons, despite having the second highest transition frequency. For HIV-2, transition frequencies were much more evenly distributed between amplicons than for HIV-1, with a maximal difference of ~1.3-fold (between Gag and EGFP-2). We also compared transition frequencies between HIV-1 and HIV-2 at the level of individual amplicons. HIV-1 was found to have a higher frequency of transitions than HIV-2 in all amplicons except Gag (p = 0.12 for Gag; p < 0.001 for all other amplicons); however, the greatest differences were observed in the EGFP-1 and Vif amplicons (7.2 and 2.9fold differences, respectively).
Detection of G-to-A hypermutation in HIV-1 and HIV-2
Considering the high levels of G-to-A transition mutations (particularly for HIV-1) that we observed (Figure 2c), we next investigated whether HIV-1 or 2 displayed evidence of G-to-A hypermutation. G-to-A hypermutation could potentially arise from the activity of APOBEC3 proteins, though relatively little APOBEC3 activity was expected in this particular experimental system (see "Discussion"). In these analyses, hypermutants were defined as read pairs (~120 bp in length) that contained two or more mutations of the same type within the read pair. In addition to analyzing G-to-A hypermutants, we determined frequencies of other possible types of transition hypermutants (A-to-G, C-to-T, and T-to-C) as well, as HIV-1 A-to-G hypermutants have occasionally been reported in the literature [4,41]. All hypermutant counts and frequencies (defined as hypermutant read pairs/all read pairs) are tabulated in Additional file 6: Dataset S2, either combined across all amplicons or separated by amplicon.
In HIV-1, the frequency of G-to-A hypermutants was approximately 7.9 × 10 −3 (or ~1 of 127 read pairs), while in HIV-2 the frequency was much lower, approximately 2.8 × 10 −4 (or 1 of 3,623 read pairs) (Figure 3a). The average G-to-A hypermutation frequencies were about 1,060-fold higher for HIV-1 or about 17-fold higher for HIV-2 as compared to their corresponding plasmid controls, respectively. HIV-1 and HIV-2 G-to-A hypermutant frequencies were significantly higher than the plasmid controls in all five amplicons examined (p-values ranging from <0.001 to 0.01). The G-to-A hypermutation frequency was much higher (i.e., about 28-fold) in HIV-1 than observed in HIV-2 (p < 0.001), demonstrating that HIV-2 was less susceptible to G-to-A hypermutation in this experimental system. This trend was significant in all five amplicons examined (p < 0.001). Other types of transition hypermutants were also observed ( Figure 3a), but they were far less frequent than G-to-A hypermutants and were not observed at levels significantly higher than that of the plasmid controls. In HIV-1, G-to-A hypermutants comprised about 98% of all transition hypermutants, indicating a clear dominance over other possible types of transition hypermutants. When subdivided by amplicon, the distribution of HIV-1 G-to-A Figure 3 HIV-1 demonstrates higher G-to-A hypermutant frequencies than HIV-2. a The frequencies of each type of transition hypermutant were compared between HIV-1, HIV-2, and the plasmid controls. For this analysis, hypermutants were defined as read pairs containing two or more mutations of the indicated type within an individual read pair (approximately 120 bp in length). The frequency of hypermutation was then calculated by dividing the number of hypermutant read pairs by all read pairs. b The frequency of G-to-A hypermutation was compared across all five amplicons examined by Illumina DNA sequencing. c The degree of G-to-A hypermutation was analyzed by determining the numbers of G-to-A mutations within hypermutant read pairs. d The dinucleotide context of G-to-A mutations from G-to-A hypermutants was determined and expressed as a percentage of the total. Data in panels a, b, and d represent the mean of three experimental replicates, with error bars representing standard deviation, while data in panel c represent the total (i.e. compiled) data. Asterisks denote statistically significant differences between HIV-1 and 2 (*p < 0.05, ***p < 0.001, N.S. not significant). The actual numbers of G-to-A hypermutant read pairs and reference read pairs are listed in Additional file 6: Dataset S2.
hypermutation was found to resemble that of all transitions (Figure 2b), with most G-to-A hypermutation concentrated in the EGFP-1 and Vif amplicons (Figure 3b). In fact, after removal of G-to-A hypermutants, transition frequencies were very consistent across all five amplicons, demonstrating that G-to-A hypermutants were responsible for the elevated transition frequencies in EGFP-1 and Vif (Additional file 7: Figure S2). The overall observed trend of G-to-A hypermutation among amplicons in HIV-1 was EGFP-1 > Vif > EGFP-2 > HSA > G ag (all p values < 0.001). Despite being much less frequent, HIV-2 G-to-A hypermutants were also concentrated mainly in the EGFP-1 and Vif amplicons, although differences between amplicons did not reach statistical significance. These observations indicate that G-to-A hypermutation can significantly vary between different genes and even between different regions of the same gene (i.e. EGFP-1 vs. EGFP-2).
G-to-A hypermutation was further analyzed by determining the number of G-to-A mutations per hypermutant read pair. Most G-to-A hypermutants contained low numbers of G-to-A mutations (medians of 4 for both viruses), as expected for these short (~120 bp) read pairs ( Figure 3c). However, a minority of G-to-A hypermutants contained high numbers of G-to-A mutations (maxima of 22 and 21 for HIV-1 and HIV-2, respectively). In contrast to the biological samples, the ultra-rare G-to-A hypermutants observed in the plasmid controls nearly all contained only two G-to-A mutations (one triple G-to-A mutant was observed for the HIV-2 plasmid). Next, the dinucleotide contexts of G-to-A mutations in hypermutants were analyzed, as APOBEC3 proteins are strongly associated with GG or GA dinucleotide contexts [42][43][44][45], depending on the specific APOBEC3 family member. Most G-to-A mutations were found within the GA dinucleotide context for both viruses, and, strikingly, only 1% (HIV-1) or 4% (HIV-2) were found to occur in either the GT or GC sequence contexts (Figure 3d). These dinucleotide contexts were markedly different from the contexts of the rare G-to-A hypermutants identified in the plasmid controls ( Figure 3d) as well as from the contexts of single G-to-A mutants (Additional file 8: Figure S3). Although the dinucleotide context appears to differ between the HIV-1 and HIV-2 plasmid controls, it should be noted that these contexts are based on extremely low numbers of hypermutants for the plasmids (10 for HIV-1 and 17 for HIV-2). The observed bias in the virological samples was not caused by a higher prevalence of the GA dinucleotide (relative to GG, GC, and GT) within the amplicon sequences, as GA dinucleotides only accounted for 24% of all guanine-containing dinucleotides. Taken together, these findings support the conclusion that the low level of G-to-A hypermutation observed was primarily caused by one or more APOBEC3 family members, despite the presence of Vif in our experimental system.
The patterns of G-to-A hypermutation were further characterized by investigating whether G-to-A mutations in hypermutants occurred at hotspots and, if so, whether sequence preferences (beyond the dinucleotide context) could be identified. In this analysis, we defined G-to-A hypermutation hotspots as statistical upper outliers using the 1.5× interquartile range (IQR) rule. Using this criterion, 18 (HIV-1) or 16 (HIV-2) hotspots for G-to-A hypermutation were identified, all of which were located in the Vif or EGFP-1 amplicons (Additional file 9: Table S4). Of these, 14 (13 in EGFP-1 and 1 in Vif ) were shared between HIV-1 and HIV-2, suggesting a common mechanism of mutation. Because specific APOBEC3 proteins have been shown to exhibit additional sequence preferences at the −1 and +2 positions (relative to the mutated guanine), sequence logos for total G-to-A hypermutation hotspots, GA context hotspots, and GG context hotspots were generated. However, statistically significant preferences at other positions were not observed for HIV-1 or HIV-2 in any of these categories. Furthermore, in order to gauge the potential effects of hypermutation on protein expression and/or function, the coding changes introduced at G-to-A hypermutation hotspots were examined (Additional file 9: Table S4). G-to-A hypermutation did not result in the introduction of premature stop codons at any of the examined hotspots. Missense mutations were typically R-to-K (semi-conservative), D-to-N (semi-conservative), or E-to-K (non-conservative); however, other types were occasionally observed. Interestingly, several of the Vif mutations occurring in G-to-A hypermutants at these hotspots have been previously characterized (see "Discussion") [15].
HIV-1 and HIV-2 mutation frequencies and spectra are not significantly different in the absence of G-to-A hypermutants
The HIV-1 and HIV-2 mutational data was analyzed with and without G-to-A hypermutants in order to: (1) estimate the extent to which G-to-A hypermutation contributed to the total mutational data, and (2) determine whether G-to-A hypermutation was primarily responsible for observed differences between HIV-1 and HIV-2 mutation frequencies and spectra. While removal of G-to-A hypermutants had little effect on HIV-2 mutation frequencies (reducing the total mutation frequency by only 5%), their removal reduced the overall HIV-1 mutation frequency by 53% and the G-to-A transition mutation frequency by 81% (Figure 4a). Thus, G-to-A hypermutation was responsible for approximately half of all mutations observed in HIV-1, despite being a relatively rare event (~1 of 127 read pairs were G-to-A hypermutants). In the absence of G-to-A hypermutants, HIV-1 and HIV-2 total mutation frequencies were found to be 3.2 or 2.9 × 10 −4 m/bp, respectively, a 1.1-fold difference that was not statistically significant (p = 0.14; Figure 4a). However, as noted earlier, HIV-2 displayed an ~1.1-fold lower frequency of A-to-G transitions (p = 0.049) and an ~1.5-fold higher frequency of C-to-T transitions (p = 0.0019) than HIV-1, and these findings were not altered by the removal of G-to-A hypermutants. Further, HIV-1 still demonstrated an ~1.8-fold higher frequency of G-to-A transitions (p < 0.001) than HIV-2 after removal of G-to-A hypermutants. As expected, the mutation spectra for HIV-1 and HIV-2 were also much more comparable in the absence of the G-to-A hypermutation data (compare Figures 4b, 2c), although HIV-1 displayed a somewhat higher percentage of G-to-A (27 vs. 17%) and lower percentage of C-to-T (14 vs. 23%) transitions than HIV-2. Taken together, these observations suggest that most of the observed differences in viral mutation patterns between HIV-1 and 2 were due to the highly reduced levels of G-to-A hypermutants for HIV-2. These findings imply that the fidelities of other mutational sources, such as RT, are relatively similar for HIV-1 and HIV-2.
Discussion
Detection of rare mutations is often difficult with nextgeneration sequencing technologies due to high error rates, which for Illumina platforms typically range from ~10 −2 m/bp for unfiltered data and ~10 −3 to 10 −4 m/bp for stringently filtered data [33][34][35][46][47][48][49]. We adopted numerous measures to minimize background errors due Figure 4 HIV-1 and HIV-2 mutation frequencies and spectra are similar in the absence of G-to-A hypermutation. a Analysis of mutation frequency in the absence of G-to-A hypermutation. HIV-1 and HIV-2 mutation frequencies were determined either including or excluding G-to-A hypermutants, with the results superimposed. The relative percentage of the total data that can be attributed (or not attributed) to G-to-A hypermutation is indicated within the bars. b Analysis of HIV-1 and HIV-2 mutation spectra in the absence of G-to-A hypermutation. HIV-1 and HIV-2 mutation spectra were examined after excluding all G-to-A hypermutants. Mutation spectra were determined by dividing the frequency of each type of mutation by the total mutation frequency, with the results expressed as a percentage. Data in both panels represent the mean of three experimental replicates.
to PCR and sequencing, as well as to minimize PCRmediated recombination (see "Methods"). We also performed control amplifications from plasmids in which we matched conditions as closely as possible with the biological samples. After stringent filtering, we obtained error frequencies of 2.8 (HIV-1) or 2.6 (HIV-2) × 10 −4 m/bp for the plasmid controls (Figure 2a). The frequencies and distribution of background errors were nearly identical between the HIV-1 and 2 plasmid controls (Figure 2a-c), as expected considering that the amplicons were either identical (HSA, EGFP-1, EGFP-2) or ~60% homologous (Gag, Vif ) in sequence. Consistent with previous reports on Illumina background errors [33][34][35][36]49], the spectra of background errors were more heavily biased toward transversions than biological samples (Figure 2c), specifically C-to-A and G-to-T transversions (Additional file 4: Figure S1). This trend was observed despite prior masking of plasmid error hotspots, which were also mostly (~83%) C-to-A and G-to-T transversions (see "Methods" and Additional file 10: Table S5). Two mechanisms for these transversion types have been suggested in the literature [33][34][35][36]50]: (1) oxidative conversion of guanine to 8-oxoguanine during sample processing, which would lead to fixation of C-to-A and G-to-T mutations during PCR, and (2) imaging errors that result from crosstalk between the C and A channels or between the G and T channels, as the fluorophores attached to these bases are excited by the same lasers and then distinguished using different filters. Due to these issues, we unfortunately were not able to make a detailed comparison of transversion frequencies between HIV-1 and 2. However, transition frequencies [5.8 (HIV-1) and 2.0 (HIV-2) × 10 −4 m/ bp] were significantly higher than the plasmid controls for both viruses (HIV-1: p < 0.001; HIV-2: p = 0.038), and many previous reports have established transitions as the dominant type of mutations that occur during HIV-1 replication [1-4, 17, 19, 40]. Thus, further analyses focused primarily on these mutational types.
Relative to previous estimates of the HIV-1 mutation rate (which range from 1.4 to 8.5 × 10 −5 mutations/bp/ cycle) [1-5, 17, 41, 51], we obtained a somewhat higher mutation frequency for HIV-1 of 6.9 × 10 −4 m/bp (or 5.8 × 10 −4 m/bp if considering only transitions). However, these discrepancies may be due to several key differences between the assays used and the ways in which the data were analyzed. First, the level of background error in previous studies was likely much lower. Most previous estimates of the HIV-1 mutation rate were determined using the LacZα assay [1,2,4,17,41], in which a proviral lacZα sequence is purified (using the Lac repressor or Hirt extraction), directly transformed into E. coli (without using PCR), and subjected to blue/ white color screening, followed by Sanger sequencing of mutants. While this assay results in little background, it is low-throughput and can only detect and measure mutations within the lacZα marker gene. Second, previously used marker gene assays (such as the LacZα assay) necessarily underestimate true mutation rates because they only detect mutations leading to a phenotypic change (silent mutations are not detected unless co-occurring with other mutations). For the LacZα assay, measured mutation rates are thought to underestimate actual mutation rates by ~2 to 3-fold [4]. In contrast, the assay we used based on direct PCR and Illumina sequencing of multiple amplicons should detect all mutations, provided that they do not prevent amplification. Third, most previous studies used a second marker gene (typically an antibiotic resistance gene) to select for infected cells or to select for transformed E. coli [1-4, 17, 41, 51]. This approach will not detect heavily mutated sequences (such as hypermutants) in which both marker genes are mutated and functionally disrupted. Fourth, actual mutation frequencies may vary between the genes analyzed here (gag, vif, hsa, and egfp) and mutational targets from previous studies. Fifth, many previous studies scored mutants harboring multiple mutations as single mutants for the purposes of calculating mutation rates [1,2,4,17,41,51]. Also, in some cases background error frequencies were subtracted from the mutation frequencies of the biological samples in order to calculate mutation rates [5].
Unexpectedly, we identified G-to-A hypermutants for both HIV-1 and 2, though the frequency was much higher (~28-fold) for HIV-1 than HIV-2 ( Figure 3a; p < 0.001). The presence of G-to-A hypermutants could not be attributed to background errors, as their frequencies were much higher in the biological samples than in the plasmid controls. Also, the patterns of G-to-A hypermutation (in terms of dinucleotide context preferences and hypermutation hotspots) were markedly distinct between the biological samples and plasmid controls (Figure 3d and data not shown). Unlike single G-to-A mutants from the biological samples, G-to-A hypermutants exhibited a strong bias toward GA dinucleotide contexts (Figure 3d; Additional file 8: Figure S3). In sum, these findings support the idea that the G-to-A hypermutants were caused by lowlevel activity of one or more APOBEC3 proteins. However, APOBEC3G cannot be primarily responsible, due to its strong preference for GG dinucleotides [42][43][44]52]. Further experiments in which specific APOBEC3 proteins are down-regulated (through knockout or knockdown) are required to prove that the G-to-A hypermutants we observed are APOBEC3-mediated, as well as to identify the specific APOBEC3 protein(s) responsible.
The observation of G-to-A hypermutation was somewhat surprising because: (1) the HIV-1 and HIV-2 vectors used in these studies encoded for a functional Vif protein, (2) the expression levels of most APOBEC3 proteins are relatively low in 293T (i.e., the cells that produced the HIV-1 and HIV-2 vector viruses) [19], and (3) Vif-proficient and Vif-deficient viruses exhibit similar infectivities when produced in 293T cells [13,15,32,53]. Upon observing G-to-A hypermutation, we confirmed via Sanger sequencing that the vif genes in the HIV-1 and HIV-2 vectors were intact (data not shown).
Notably, previous studies have demonstrated that Vif does not always fully neutralize the activity of APOBEC3 proteins [13,42,[54][55][56]. In addition, Vif from different subtypes or containing naturally occurring polymorphisms have been shown to vary widely in their neutralization capacities [55][56][57]. Further, multiple APOBEC3 proteins (including B, C, D, and F) have been detected at the mRNA level in 293T cells and could be involved in the observed G-to-A hypermutation [19,58]. Previous studies by our group have identified rare G-to-A hypermutants occurring at GA dinucleotides by Sanger sequencing of clones (in which case the virus stocks were also generated in 293T cells) [19,59]. Taken together, the G-to-A hypermutation observed here was likely due to the failure of HIV-1 or HIV-2 Vif to fully neutralize low levels of APOBEC3 proteins present in 293T producer cells. Consistent with this hypothesis, G-to-A hypermutation was observed very infrequently even for HIV-1 (~1 of 127 read pairs were G-to-A hypermutants), despite contributing to about half of all mutations (Figure 4a). Although HIV-2 was less susceptible than HIV-1 to G-to-A hypermutation in this experimental system, further investigation will be required to determine whether these trends hold true under conditions of higher APOBEC3 expression. Intriguingly, HIV-2 ∆vif has been reported to be less sensitive than HIV-1 ∆vif to APOBEC3G [32], but susceptibilities to other APOBEC3 proteins (in the presence or absence of Vif ) have not yet been compared. Further, HIV-1 and HIV-2 Vif were recently shown to interact with APOBEC3F and APOBEC3G through completely separate sequence determinants, and differences in HIV-1 and HIV-2 Vif-induced degradation of specific APOBEC3 proteins were also noted [60].
Most of the G-to-A hypermutants occurred within the EGFP-1 and Vif amplicons (Figure 3b). This may be due in part to higher frequencies of GA dinucleotides in these particular amplicons. For HIV-1, the frequencies of GA dinucleotides (relative to GG, GT, and GC) were 13% (Gag), 32% (Vif ), 9% (HSA), 32% (EGFP-1), and 30% (EGFP-2). Thus, the GA dinucleotide frequency varied maximally by 3.6-fold, whereas G-to-A hypermutation frequencies varied up to 26.3-fold (EGFP-1 vs Gag). However, APOBEC3-mediated hypermutation can also be influenced by broader sequence contexts and secondary structures [14,15,42,44,[61][62][63]. Further, APOBEC3 activity follows a twin gradient along the HIV-1 genome corresponding to the amount of time the minus strand viral DNA remains single-stranded, such that the positioning of the amplicon could affect hypermutant frequencies [64,65]. Thus, one or more of these other features may have favored hypermutation in the EGFP-1 and Vif amplicons. Full genome sequencing will be required to address the distribution of G-to-A hypermutation in more detail. G-to-A hypermutation hotspots were also identified (Additional file 9: Table S4), which all occurred in the EGFP-1 and Vif amplicons. G-to-A hotspots did not result in any nonsense mutations, consistent with the notion that GA dinucleotidebiased hypermutation generates fewer stop codons than GG-biased hypermutation. However, the G-to-A hotspots did introduce a number of missense mutations, particularly D-to-N, E-to-K, and R-to-K. Surprisingly, all of the mutations resulting from the four hotspots in HIV-1 Vif were recently identified in another study in which humanized mice were infected with an HIV-1 variant that cannot neutralize APOBEC3D or F [15]. One of these Vif mutants (E134K) lost the ability to neutralize APOBEC3G, raising the possibility that hypermutation itself can influence susceptibility to hypermutation in further rounds of replication.
We initially hypothesized that HIV-2 would exhibit a lower mutation frequency than HIV-1 due to its attenuated pathogenicity, reduced evolutionary rate [27,28], a conserved fidelity-improving V75I polymorphism in RT [30,31], and reduced susceptibility to APOBEC3G [32]. More specifically, HIV-2 Env has been found to diversify more slowly than HIV-1 Env when compared to HIV-1-infected individuals with high viral loads [28]. However, HIV-1 and HIV-2 diversification rates were similar when compared to HIV-1-infected individuals with low viral loads. In the same report, HIV-2 Env was found to evolve more slowly than HIV-1 Env and to be subject to strong purifying selection. Indeed, HIV-2 infection appears to elicit broad and potent neutralizing antibody responses against Env more frequently than for HIV-1 [66]. In another report, HIV-2 Env was found to exhibit a lower rate of synonymous substitutions than HIV-1, implying reduced viral mutation and/or replication rates [27], though another group has reported opposing findings [29]. Unfortunately, it is difficult to compare the results of our analyses to these published reports due to these contradicting results, and larger studies of HIV-2 intra-patient diversification and evolution are clearly warranted. Nonetheless, we found that HIV-1 and HIV-2 exhibited similar total mutation frequencies in the absence of G-to-A hypermutants, suggesting that differences in replication fidelity do not have a major impact on differences in evolutionary or synonymous substitution rates between HIV-1 and HIV-2.
Conclusions
In sum, we have found that HIV-2 exhibited significantly lower total and transition mutation frequencies than HIV-1, as well as a mutation spectrum less biased toward G-to-A transitions. However, these differences were mostly due to a significantly higher G-to-A hypermutation frequency for HIV-1 than HIV-2. These findings raise the intriguing possibility that HIV-2 might be less sensitive than HIV-1 to APOBEC3-mediated hypermutation, consistent with a previous report [32], but additional experiments in other cell types will be required to fully address this question. After removal of all G-to-A hypermutants, HIV-1 and HIV-2 total mutation frequencies were not significantly different, although small but significant differences were still observed in the frequencies of G-to-A and C-to-T transitions. Overall, these results suggest that the fidelities of other mutagenic processes (such as reverse transcription) are relatively similar between the two viruses. Nevertheless, we cannot rule out the possibility that HIV-1 and HIV-2 exhibit more minor differences in mutation frequencies or spectra that we were not able to detect or that differences would be observed in other cell types. Overall, these data imply that differences in replication fidelity are likely not a major contributing factor to the unique clinical features of HIV-2 infection.
Plasmids, cell lines, and reagents
The HIV-1 vector, pNL4-3 HIG, has been previously described [67]. This vector contains a cassette encoding heat stable antigen (HSA), an internal ribosomal entry site (IRES), and enhanced green fluorescent protein (EGFP). The HIV-2 vector, pROD HIG, was created from pHIV-2 H0G [68], a kind gift from Dr. Wei-Shau Hu (HIV Drug Resistance Program, Frederick National Laboratory for Cancer Research, Frederick, MD, USA). The pHIV-2 H0G vector does not express Vpr or EGFP due to multiple point mutations. In order to construct pROD HIG, the vpr and egfp genes were restored by site-directed mutagenesis using the QuikChange II XL kit (Agilent Technologies, Inc.; Santa Clara, CA, USA), and the result-
Virus production and titering
Virus was produced by co-transfecting pNL4-3 HIG or pROD HIG with pNL4-3 Env or pROD10 Env, respectively, into 293T cells via the PEI method [71]. PEI stocks were prepared at 1 mg/mL by dissolving PEI in water, adjusting the pH to 7.0, and filtering through a 0.2 µm filter. 24 h before transfection, 4 million 293T cells/plate were seeded onto 10 cm plates pre-coated for 5 min with poly-l-lysine. For each plate, 10 µg of pNL4-3 HIG or pROD HIG + 5 µg Env expression plasmid + 45 µL 1 mg/ mL PEI were combined with serum-free DMEM to a final volume of 1 mL. After 20 min of incubation, the medium on the 293T cells was replaced and the DNA-PEI mixture was added. The medium was replaced 16 h post-transfection, and viral stocks were collected 48 h post-transfection by filtering the supernatants through a 0.2 µm filter. For each viral stock, five plates were transfected, and the resulting supernatants were pooled and concentrated (~tenfold) using 100,000 MWCO filtration columns (Vivaproducts; Littleton, MA, USA). Viral stocks were then treated with 10 U/mL of DNase I (New England Biolabs; Ipswich, MA, USA) for 2 h at 37°C to degrade residual plasmid DNA from transfections. Viral stocks were then divided into 1.0 mL aliquots and frozen at −80°C.
Prior to large-scale infections, viral stocks were first titered in U373-MAGI cells based on EGFP expression. The day before infection, 31,250 cells/well were plated in 24-well plates. After 24 h, the media was replaced and varying volumes of virus ranging from 15.625 to 500 µL (twofold dilution series) were added. To improve infectivity, the cells were infected by spinoculation (1,200 × g for 2 h at 24°C). The media was replaced again 24 h post-infection and cells were collected at 72 h postinfection for analysis of EGFP expression by flow cytometry. The cells were treated with trypsin, transferred to 96-well plates, pelleted at 500 × g for 5 min, and resuspended in 200 µL Dulbecco's phosphate-buffered saline (DPBS) + 2% FC3/well. EGFP expression from at least 10,000 gated cells was analyzed using a BD LSR II flow cytometer (BD Biosciences; San Jose, CA, USA). EGFP was excited with a blue 488-nm laser and emission detected using 505LP and 525/50 filters. Virus titers (expressed as infectious units/mL) were calculated based upon EGFP expression at low infectivities (<40%) as previously described [72].
Infections for Illumina sequencing
In order to prepare samples for Illumina sequencing, 1 × 10 6 U373-MAGI cells were infected at a multiplicity of infection (MOI) of 1.0. These infections were performed in 24-well plates (31,250 cells/well) in order to avoid any potential effect of the plate format on infectious titer. Uninfected cells and cells infected with heatinactivated viruses (i.e. virus stocks that were incubated at 65°C for 1 h) were included as negative controls. The cells were infected by spinoculation, and the medium was replaced 24 h post-infection. Cells were collected for genomic DNA extraction at 72 h post-infection by treating with trypsin, pelleting, and washing three times with DPBS to further reduce plasmid carryover. Extra wells of infected cells were analyzed by flow cytometry to verify infectivity. In this assay, all proviruses result from a single cycle of infection, as neither producer cells nor target cells can be re-infected, due to a lack of the appropriate receptor and co-receptor or to a lack of envelope expression, respectively ( Figure 1).
Genomic DNA extraction and quantification of plasmid carryover
Genomic DNA was extracted from all collected cells using the High Pure PCR Template Preparation Kit (Roche; Basel, Switzerland) following the manufacturer's instructions and eluted in 150 μL buffer. Genomic DNA was treated with DpnI for 1 h at 37°C to further reduce plasmid carryover from the transfections, after which DpnI was heat-inactivated at 80°C for 20 min. In order to quantify any residual plasmid carryover, two approaches based on quantitative PCR (qPCR) were adopted: (1) the ampicillin resistance gene copy number was determined and divided by the proviral copy number (as measured using the HSA Illumina amplicon), or (2) the proviral copy number from heat-inactivated virus infections was determined and divided by the proviral copy number from the corresponding un-treated infections. For both approaches, qPCR was performed using 4 μL water, 6.25 μL 2× Power SYBR Green Master Mix (Life Technologies), 0.625 μL each primer (500 nM final concentration), and 1 μL template. The cycling conditions used were an initial denaturation of 95°C 10 m, 40 cycles of 95°C 15 s/55°C 15 s/72°C 30 s, and a final extension of 72°C 7 min. The sequences of the primers to the ampicillin resistance gene were 5′-ACTTTATCCGCCTCCATC CAGTC-3′ and 5′-GAGCGTGACACCACGATGC-3′. Absolute standard curve series (from 10 1 to 10 6 copies/ μL) were constructed by quantifying the pNL4-3 HIG and pROD HIG plasmids with the Qubit dsDNA HS Assay Kit (Life Technologies). We found that the level of plasmid carryover for HIV-1 was 0.2% when measured by either method, while the level of carryover was 2.8 or 1.4% for HIV-2, depending on the approach used (Additional file 1: Table S1). These results are comparable to those obtained in a similar study of HIV-1 mutagenesis [5] and are too low to significantly affect measured mutation frequencies.
Amplifications for Illumina sequencing
PCR was performed on five small (~150-170 bp) amplicons (Gag, Vif, HSA, EGFP-1, EGFP-2) for each sample, with HIV-1 and 2 amplicons positioned in homologous locations. All forward primers contained 5-base barcodes (differing by at least two bases) to allow demultiplexing of pooled PCR products. All primer and barcode sequences are listed in Additional file 11: Table S6. Barcodes were randomly generated using a program written by Luca Comai and Tyson Howell (http://comailab.genomecenter. ucdavis.edu/index.php/Barcode_generator) [73]. PCR was performed using the Phusion Hot Start II High-Fidelity DNA Polymerase (Fisher Scientific; Pittsburgh, PA, USA). PCR reactions were performed with 8 µL of genomic DNA (~50,000 target copies), 500 nM of each primer, and a final volume of 50 µL/reaction. The cycling conditions used were an initial denaturation of 98°C 30 s, 30 cycles of 98°C 10 s/56°C 30 s/72°C 15 s, and a final extension of 72°C 10 min. All PCR reactions were performed in triplicate to reduce the risk of clonal amplification, and the products were pooled after amplification. Negative control reactions were performed lacking template or with genomic DNA from uninfected cells.
In order to assess the degree of background error due to PCR and Illumina sequencing, control amplifications were performed from pNL4-3 HIG or pROD HIG plasmids using the same cycling conditions. For these reactions, 50,000 copies of plasmid per 50 µL reaction was used, a target level found via qPCR to be similar to that of the other samples. Genomic DNA (8 µL/reaction) from uninfected cells was added to the plasmid PCR reactions to account for any potential impact of genomic DNA on amplification. Further, the degree of intra-sample PCRmediated recombination was investigated in the plasmid controls, as recombination could affect the association between mutations and thus alter hypermutant frequencies. To determine recombination frequencies, EGFP-1 was amplified from a 1:1 mixture (i.e. 25,000 copies each) of wild-type and mutant EGFP-1 sequences. The mutant EGFP-1 sequence contained two genetic markers positioned ~50 bp apart, and each marker consisted of two mutations to facilitate distinction from PCR and sequencing-induced errors. Intra-sample recombinants were detected by identifying read pairs from the plasmid controls with a single marker set, rather than the expected zero or two marker sets present in the wild-type and mutant EGFP-1 sequences, respectively. Importantly, all mutations utilized as recombination markers were masked before mutational analysis.
Library preparation and Illumina sequencing
Amplicons were gel-purified using the Promega SV Gel Extraction Kit (Promega Corp.; Madison, WI, USA). After gel purification, all amplicons were quantified using the Qubit dsDNA HS Assay Kit (Life Technologies) and a Qubit 2.0 fluorometer. For each sample, all amplicons were pooled together in an equimolar fashion to normalize coverage between amplicons. Next, 100 ng of each sample (12 samples in total) were submitted to the University of Minnesota Genomics Center for library preparation. The libraries were constructed with the TruSeq Nano DNA Sample Preparation Kit following the manufacturer's instructions but using AMPure XP beads (Beckman Coulter, Inc.; Indianapolis, IN, USA) at a bead to sample ratio of 1.8. The 15 libraries were again quantified using the Qubit dsDNA BR Assay Kit, size-confirmed with Agilent DNA 1000 chips (Agilent Technologies; Santa Clara, CA, USA), and pooled in an equimolar fashion to normalize coverage between libraries. Samples were pooled after library preparation, rather than before, because it had been previously determined that significant inter-sample recombination can occur during the 10 cycles of PCR typically utilized in library construction (data not shown). In order to improve sequence diversity and quality, a PhiX library was added in at ~25% of total mass. Next, 2 μL of the 10 nM library pool was denatured and diluted to a final concentration of 4 pM for DNA sequencing. Sequencing of proviral DNA was conducted by using the Illumina MiSeq with 2 × 150 paired-end sequencing. All Illumina sequencing data supporting the results of this manuscript have been deposited into the NCBI Sequence Read Archive (SRA) under accession code BioProject PRJNA287455.
Processing of proviral DNA sequencing data
First, paired-end reads from the Illumina MiSeq reaction were demultiplexed based on perfect barcode matches, and barcode sequences were trimmed off during the process. Second, poor quality reads were filtered out using quality criteria found to reduce Illumina background error rates [35]. Specifically, read pairs were discarded in which either read contained a B-tail (i.e. one or more low quality bases at the end of the read), contained at least one uncalled base, had less than two-thirds of bases with Q-score ≥30 in the first half of the read, or had an average Q-score <30 in the first 30% of the read. Third, reads from each of the 12 samples were mapped to the appropriate reference sequence (pNL4-3 HIG or pROD HIG) with GSNAP [74] using default parameters. Finally, a small number of read pairs (~35,000) that were aligned either partially or fully outside of the appropriate amplicon regions were excluded. For each sample, the numbers of initial read pairs, read pairs lost during mapping or filtering, and final read pairs are listed in Additional file 2: Table S2. We obtained ~4.7 million total read pairs after all processing steps, which removed primarily PhiX read pairs or HIV read pairs with imperfect barcodes or poor quality. About 319,000-461,000 read pairs were obtained per sample (average of 395,000 read pairs/sample), or 46,000-111,000 read pairs per amplicon per sample (average of 79,000 read pairs/amplicon/sample).
In order to identify mutations present in read pairs passing the above processing steps, a custom algorithm was developed to compile mutation frequency data for each sample, built using the Genome Analysis Toolkit (GATK) walker framework [75]. This algorithm determines both the frequency of total mutations as well as of specific mutational types (substitutions, transitions, transversions, every type of transition or transversion, insertions, deletions, etc.). In order to minimize background error rates, mutations were required to be identified on both sequences in a read pair, which was possible because forward and reverse reads were mostly overlapping due to small amplicon sizes (~150-170 bp). Furthermore, substitutions and insertions were only counted if they had a Q-score ≥30 for the relevant base(s) on both reads. Wild-type bases had to meet the same criteria as mutations (i.e. identified as wild-type and Q-score ≥30 on both sequences of a read pair). Non-overlapping segments of read pairs as well as single reads were excluded from mutational analyses. All primer sequences were also excluded from mutational analysis, as errors within primers would not represent biologically meaningful mutations. We also examined the distribution of all background errors (from PCR and sequencing) in the plasmid controls and identified numerous plasmid error hotspots (defined as upper outliers within the distribution of frequencies of individual mutations based on the 1.5 × IQR rule). Most plasmid error hotspots (~83%) were G-to-T or C-to-A transversions. Within identical amplicons (HSA/EGFP-1/EGFP-2), many common plasmid error hotspots were identified in the HIV-1 and HIV-2 plasmid controls (~88% overlap), whereas the degree of overlap was much lower for the ~60% homologous viral amplicons (~49% overlap in Gag and ~31% overlap in Vif ). Plasmid error hotspots that were shared between the HIV-1 and HIV-2 plasmid controls (Additional file 10: Table S5) were masked before further downstream analysis, as the presence of these mutations in the biological samples would most likely represent background errors. Rather than masking all mutational types at error hotspots, only the problematic type(s) (e.g. G-to-T) were masked at the indicated positions. Insertions and deletions were scored as single events regardless of the number of bases inserted or deleted. Mutation frequencies (defined as mutations/bp) were calculated by dividing the number of mutations passing filters by all reference bases (mutations + wild-type bases) passing filters.
In addition to examining mutation frequencies and spectra, hypermutants were identified and characterized within the Illumina sequencing data. Using a custom GATK walker, hypermutant counts for each type of transition were collected (G-to-A, A-to-G, T-to-C, and C-to-T). Hypermutants were defined as individual read pairs containing two or more of the same type of transition. All transitions had to be identified on both sequences with Q-scores ≥30, and a single read pair could theoretically count as two different types of hypermutants if it contained multiple instances of two different transition types. Hypermutant frequencies were calculated by dividing the number of hypermutant read pairs by all (hypermutant + non-hypermutant) read pairs that passed the processing steps. In order to examine G-to-A hypermutation hotspots, ranked lists of G-to-A mutation frequencies were generated at individual bases within G-to-A hypermutants. Hypermutation hotspots were then defined as statistical upper outliers within the distribution of frequencies using the 1.5 × IQR rule.
Biostatistical analysis of Illumina DNA sequencing data
To test for factors that may influence mutation frequencies, generalized linear mixed effects models were applied to the data that came from our Illumina data processing pipeline. The raw counts for each type of mutation (e.g. transitions) were modeled as Poisson random variables with an offset given by the total number of reference bases. The type of sample (i.e. HIV-1, HIV-2, and plasmid controls for HIV-1 and HIV-2), the type of amplicon, and their interactions were treated as fixed effects while the replicate was treated as a random effect. The logarithmic link was used, as is standard for Poisson outcomes, and penalized quasilikelihood was used to estimate the model parameters [76]. These computations were conducted using R v 3. | 2016-05-04T20:20:58.661Z | 2015-07-10T00:00:00.000 | {
"year": 2015,
"sha1": "6a679ede606602b8868e82e5c39b2628752e5317",
"oa_license": "CCBY",
"oa_url": "https://retrovirology.biomedcentral.com/track/pdf/10.1186/s12977-015-0180-6",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "6a679ede606602b8868e82e5c39b2628752e5317",
"s2fieldsofstudy": [
"Biology",
"Medicine"
],
"extfieldsofstudy": [
"Biology",
"Medicine"
]
} |
244933428 | pes2o/s2orc | v3-fos-license | Effect of Kernel Size and Its Potential Interaction with Genotype on Key Quality Traits of Durum Wheat
This study was conducted to evaluate the influence of kernel size and its potential interaction with genotype on durum wheat quality with emphases on kernel physical characteristics, milling performance, and color-related quality parameters. Wheat samples of seven genotypes, selected from the 2018 Canadian durum variety registration trial, were segregated into large (LK), medium (MK), and small-sized kernels (SK). In general, the kernel size greatly affected the durum wheat milling performance. Within a given size fraction, a strong impact of genotype was shown on the test weight of SK and the milling yields of MK and LK. Particularly, the MK fraction, segregated from the genotypes with superior milling quality, had a higher semolina yield than LK from the genotypes of inferior milling quality, inferring the importance of intrinsic physicochemical properties of durum kernels in affecting milling quality. SK exhibited inferior milling quality regardless of the genotypes selected. A strong impact of genotype was shown for the total yellow pigment (TYP) content and yellowness of semolina, while the kernel size had a significant impact on the brightness and redness of the semolina and pasta. Despite SK possessing much higher TYP, the semolina and pasta prepared from SK were lower in brightness and yellowness but with elevated redness.
Introduction
Durum wheat physical properties are very important in determining its commercial value. Strong associations have been reported between kernel physical characteristics and durum wheat milling performance, semolina composition, and pasta processing quality [1][2][3][4][5][6]. Emphasis has been on unveiling the relationship between test weight (TWT) to durum wheat milling potential by evaluating samples with a wide range of TWT, protein content, and kernel size distribution (KSD) [3][4][5]. Recent study in our laboratory has shown that kernel size is more effective than TWT in predicting the milling performance of durum wheat by assessing Canadian durum samples with a wide range of TWT and KSD [5].
In general, with the decrease of kernel size from large to medium, the semolina and total milling yields of durum wheat reduced gradually. A drastic decrease in milling quality was observed for small kernels passing through the no. 6 slotted sieve (2.38 mm aperture) [4,5] with much reduced milling yields coupled with elevated ash content. Baasandorj, Ohm, Manthey, and Simsek (2015) studied the impact of kernel size and mill type on the milling and baking quality of hard red spring wheat [7]. Compared with large-sized kernels, the small-sized kernels had a much lower flour yield because of the lower proportion of starchy endosperm to bran.
The kernel size of durum wheat can significantly affect not only the milling performance but also the semolina and pasta quality [3,5]. Semolina milled from SK exhibited higher protein content, finer granulation, and was higher in TYP but less bright in color with elevated ash [3,5]. Cooked pasta made from durum samples with a high proportion of SK had higher firmness but was duller in color.
While the impact of kernel size on semolina and pasta quality is well-documented, limited information is available on the response of genotype to the general relationships between kernel size and the key durum wheat quality parameters. Due to the variation in intrinsic quality, the degree of impact of kernel size on quality could be genotype dependent. Using milling performance as an example, it is not clear if the genotypes with superior milling quality would be less susceptible to kernel size variation than those of inferior milling quality, or vice versa. Genotypes with different intrinsic quality could respond differently to variations in kernel size.
On the other hand, differences in quality among genotypes could be affected by the variation in kernel size. Although TKW was shown to be highly correlated with semolina yield across four different durum varieties (R 2 = 0.92) evaluated by Wang and Fu (2020), greater variation in semolina yield was seen for larger kernels than for smaller ones [5]. The fact that the genotypic variation in durum milling performance was related to kernel size suggests a potentially greater role of genotype in the milling quality of large kernels than that of the small ones.
With the prevalence of hot and dry growing conditions on Canadian prairies in the last few years, some durum samples, although graded as No.1 or No. 2 Canada Western Amber Durum (CWAD), showed relatively wide range of KSD and milling quality [5]. To optimize the commercial value of durum wheat of different KSD and understand how quality parameters respond to kernel size variations, a thorough investigation is required to further elucidate the combined effect of kernel size and genotype on key durum wheat quality parameters.
Therefore, the objective of this study was to evaluate the influence of kernel size and its potential interaction with genotype on key durum wheat quality traits with emphases on the wheat physical properties, milling performance, and color-related quality attributes.
Wheat Samples
Seven genotypes were selected from the 2018 Canadian durum wheat variety registration trial based on their intrinsic differences in milling and color-related quality parameters. A composite of each genotype was prepared from wheat samples grown at nine locations across western Canada. Based on availability and grading information of wheat samples from the nine locations, a recipe was developed for the preparation of the wheat composites. All composites were graded as No.1 CWAD. Each of these variety composites was segregated into three size fractions using a Carter dockage tester (Simon-Day Ltd., Winnipeg, MB, USA) equipped with no. 6 (2.38 mm × 19.05 mm) and no. 7 (2.78 mm × 19.05 mm) slotted sieves. The segregated kernel size fractions were categorized as small-sized kernels (SK, through no.6 slotted sieve), medium-sized kernels (MK, passing no.7 but remained above no.6 slotted sieve), and large-sized kernels (LK, remained above no.7 slotted sieve).
Wheat Physical Properties
To accommodate the small sample size, the test weight (TWT) was measured using a 0.5 L container equipped with a cox funnel following the standard procedure described by the Canadian Grain Commission [8]. The value in gram per half liter was converted to kg per hectoliter using the test weight conversion chart for amber durum wheat. TKW was determined with an electronic seed counter (Model 750, The Old Mill Company, Savage, Maryland) using a 20 g sample of wheat of which all broken kernels were manually removed. KSD was determined on a series of slotted sieves (i.e., no. 6, 7, and 8). One hundred grams of wheat was subsampled and manually shaken for 30 s, after which the four fractions separated by the sieves were collected and weighted individually. All wheat physical tests were conducted in duplicate.
Standard Durum Milling Procedure
Following the mill flow previously described by Dexter et al. (1990) [9], original unsorted wheat samples were milled into semolina in duplicates of 2.3 kg lots with a four stand Allis-Chalmers laboratory mill (West Allis, WI, USA) in conjunction with a laboratory purifier. The mill room was controlled at 21 • C and 60% relative humidity. Semolina is defined as having less than 3% pass through a 149 µm sieve. The total milling yield is the combination of semolina and flour. Both the total and semolina yields are reported as a percentage of the cleaned wheat on a constant moisture basis. Semolina granules were prepared by adding the most refined flour stream(s) to semolina until 70% extraction was reached for quality analysis.
Micro-Milling and Purification Protocol
Wheat samples of various size fractions were milled to predict semolina and total milling yields following the micro-milling procedure previously developed by Wang et al. (2019) [10]. After tempering to a moisture content of 16% overnight, 200 g of wheat sample was ground with a Quadruma Junior (QJ)-II-G mill-semolina version (C.W. Brabender Instruments, Inc., South Hackensack, NJ, USA) with the original sifter removed. The resulting wholemeal was sifted through a universal laboratory sifter (Bühler MLUA GM sieve, Bühler AG) equipped with a bottom screen of 180 µm to remove the flour and a top screen of 630 µm to retain the bran-rich fraction. The unpurified semolina fraction (SY1) between the two screens was collected. Based on the prediction models developed by Wang et al. (2019) [10], semolina yield and total milling yield were calculated according to the amount of SY1 and bran-rich fraction. Formulas (1) and (2) are as follows: Semolina Yield (%) = 1.02 × Bran-rich fraction + 1.80 × SY1 − 73.17. (1) Total Milling Yield (%) = 0.62 × SY1 + 39.42 (2) To prepare refined semolina for analysis and pasta processing, the original purification steps described by Dexter et al. [9] were modified to accommodate the small semolina sample size with three purification and two sizing passages. A detailed description of the micro-milling and purification steps is illustrated in Figure 1. In a typical experiment, SY1 obtained from QJ semolina mill was passed over a laboratory purifier (Namad, Rome, Italy) equipped with four different sizing sieves (335, 425, 570, and 670µm). After the first purification (P1), large semolina granules collected in tray 4 and 5 were reduced with the first sizing roll (S1). The reduced semolina was sifted through a box sifter equipped with a 180 µm sieve for 30 s to remove the flour. The resulting fraction retained above the 180 µm sieve together with the semolina collected in tray 3 at P1 were subject to a second purification (P2). After P2, the semolina granules which remained in tray 4 and 5 were subject to a second sizing step (S2). The reduced fraction was sifted with a box sifter for 30 s to remove bran/shorts (>425 µm) and flour (<180 µm). The semolina fraction between 180 and 425 µm was combined with the semolina collected in tray 3 at P2 and transferred to the third purification (P3). Refined semolina was collected as tray 1 and 2 in P1, tray 1 and 2 in P2 and tray 1, 2, and 3 in P3. Tray 4 and 5 in P3 were defined as Feeds.
Semolina Quality Testing
The protein content of the whole wheat and semolina were measured following the method previously described by Williams et al. [11] with a LECO Truspec N CNA (combustion nitrogen analysis) analyzer (Saint Joseph, MI). Ground wheat meal was prepared using a Retsch ZM 200 mill (Retsch GmbH, Haan, Germany) equipped with a 0.5 mm screen (Trapezoid holes) at a speed of 14,000 rpm. Ash content, wet gluten, and gluten index were determined using AACC International approved methods 76-31.01 and 38-12.02, respectively [12]. Semolina color was measured with a Minolta colorimeter CR-410 (Konica Minolta Sensing, Inc., Tokyo, Japan) with a D65 illuminant. Color readings are expressed on the CIELAB color space system with L*, a* and b* parameters representing brightness, redness, and yellowness values, respectively. A micro scale rapid extraction procedure as described by Fu et al. [13] was used for the determination of the total yellow pigment (TYP) content of the semolina.
Spaghetti Processing and Color Measurement
Spaghetti were produced from semolina using a customized micro-extruder (Randcastle Extrusion Systems Inc., Cedar Grove, NJ, USA) following the method of Fu et al. [6]. Semolina was first mixed with water in a high-speed asymmetric centrifugal mixer (DAC 400 FVZ SpeedMixer, FlackTec, Landum, SC, USA) at water absorption of 31-32% to maintain a constant extrusion pressure of about 100 psi. Vacuum was applied to eliminate introduction of air bubbles and minimize oxidative degradation of the yellow pigment, after which the dough crumbs were extruded through a four-hole Teflon coated spaghetti die (1.8 mm). The fresh pasta was subsequently dried in a pilot pasta dryer (Bühler, Uzwil, Switzerland) coupled with a 325 min drying cycle and a maximum temperature of 85 °C. To measure spaghetti color, 6.5 cm bands of spaghetti strands were mounted on a white mat board with minimum interspace. Spaghetti color was determined using a Minolta colorimeter (CR-410) as described above.
Semolina Quality Testing
The protein content of the whole wheat and semolina were measured following the method previously described by Williams et al. [11] with a LECO Truspec N CNA (combustion nitrogen analysis) analyzer (Saint Joseph, MI). Ground wheat meal was prepared using a Retsch ZM 200 mill (Retsch GmbH, Haan, Germany) equipped with a 0.5 mm screen (Trapezoid holes) at a speed of 14,000 rpm. Ash content, wet gluten, and gluten index were determined using AACC International approved methods 76-31.01 and 38-12.02, respectively [12]. Semolina color was measured with a Minolta colorimeter CR-410 (Konica Minolta Sensing, Inc., Tokyo, Japan) with a D65 illuminant. Color readings are expressed on the CIELAB color space system with L*, a* and b* parameters representing brightness, redness, and yellowness values, respectively. A micro scale rapid extraction procedure as described by Fu et al. [13] was used for the determination of the total yellow pigment (TYP) content of the semolina.
Spaghetti Processing and Color Measurement
Spaghetti were produced from semolina using a customized micro-extruder (Randcastle Extrusion Systems Inc., Cedar Grove, NJ, USA) following the method of Fu et al. [6]. Semolina was first mixed with water in a high-speed asymmetric centrifugal mixer (DAC 400 FVZ SpeedMixer, FlackTec, Landum, SC, USA) at water absorption of 31-32% to maintain a constant extrusion pressure of about 100 psi. Vacuum was applied to eliminate introduction of air bubbles and minimize oxidative degradation of the yellow pigment, after which the dough crumbs were extruded through a four-hole Teflon coated spaghetti die (1.8 mm). The fresh pasta was subsequently dried in a pilot pasta dryer (Bühler, Uzwil, Switzerland) coupled with a 325 min drying cycle and a maximum temperature of 85 • C. To measure spaghetti color, 6.5 cm bands of spaghetti strands were mounted on a white mat board with minimum interspace. Spaghetti color was determined using a Minolta colorimeter (CR-410) as described above.
Statistical Analysis
All data were analyzed with Microsoft Excel and SAS 9.4 Software (SAS Institute Inc., Gary, NC, USA). A 3 × 7 factorial experiment was applied to evaluate the impact of kernel size and genotype on key durum wheat quality characteristics by including 3 levels of kernel size (small, medium, and large) and 7 different genotypes (A to G) representing the major source of variations. Each segregated kernel size fraction from a selected genotype was treated as an independent sample. Significance of each factor as indicated by F values and percentage of variability assignable to each factor as measured by the ratio of sum of square to the total sum of squares was calculated. Tukey's test, which followed the analysis of variance, indicated significant differences with a level of p < 0.05.
Influence of Kernel Size and Genotype on Physical Properties of Durum Wheat
To understand the impact of kernel size, genotype, and their interactions on major durum wheat quality parameters, seven durum genotypes with variation in milling and color related quality attributes were segregated into three kernel size fractions using a Carter dockage tester. The wheat and semolina quality parameters of the unsorted samples are summarized in Table 1. The selected genotypes differed greatly in semolina and total milling yields, TYP, and gluten index, but with less variation in wheat physical properties (i.e., HVK, TWT, TKW, KSD), wheat protein, and ash contents. The semolina and total milling yields from the micro-milling procedure were comparable to those of standard laboratory milling except genotype D which showed higher semolina and total milling yields in the micro-milling process.
The significance of kernel size, genotype, and their interactions on major durum wheat quality parameters, as measured by the F value and percentage of variability assignable to each factor and their interactions, are summarized in Table 2. Significant impact was found for kernel size, genotype, and their interactions on all wheat quality parameters examined (p < 0.001). In terms of wheat physical properties, kernel size accounted for more than 80% of the variability in TWT and TKW with minor influences shown for genotypes and their interactions. Table 3 summarizes the impact of genotype on key quality parameters in relation to kernel size. TKW reduced drastically from 51.0 ± 1.8 g of LK to 36.1 ± 0.9 g of MK, but was only accompanied by a small decrease of TWT from 83.7 ± 0.7 kg/hL to 82.2 ± 0.6 kg/hL. Further decrease of kernel size from MK to SK led to a much greater reduction in average TWT from 82.2 kg/hL to 77.6 kg/hL, suggesting SK (TKW of 23.9 ± 0.4 g) was much less dense than the corresponding larger ones. A similar decrease in TKW and TWT was reported when a bulk CWAD cargo aggregate was fractioned into five different kernel sizes [5]. Wang and Fu reported that TWT is less effective than TKW in distinguishing the difference in kernel size [5].
Interestingly, the impact of genotype on TWT was greater for SK than for both MK and LK (Table 3). Although there was no significant difference in TKW of the SK fractions, SK possessed much greater variability in TWT, ranging from 74.5 to 80.6 kg/hL (F value = 465.6, p < 0.001) as compared to MK (81.1-82.6 kg/hL, F value = 73.3, p < 0.001) and LK (82.3 to 84.5 kg/hL, F value = 130, p < 0.001). On the other hand, greater variation in TKW among genotypes was shown for LK (48.3 to 52.8 g, F value = 23.81, p < 0.001) in comparison to MK (34.8-37.0 g, F value = 4.5, p < 0.05) and SK (23.2 to 24.4 g, F value = 1.9, ns). TWT can be affected by wheat moisture, kernel density, kernel shape, and packing factors, which were not directly associated with milling yield [14][15][16][17][18]. Simmons and Meredith attributed the difference in TWT to bran surface roughness, distribution of kernel size, shape, volume, and kernel density [19]. Troccoli and di Fonzo found that kernel shape such as rectangular aspect ratio (kernel width/kernel length) and circularity shape factor (4π × area/perimeter 2 ) were positively related to TWT [20]. More recently, Wang and Fu reported that durum wheat with a high proportion of SK could exhibit TWT comparable to the wheat samples of larger kernel size but exhibited much lower milling yields [5]. The relationship appears to be genotype dependent. The great variation in TWT of the SK fraction could likely be attributable to large differences in kernel shape and packing density. Due to the potential strong impact of genotype, TWT can vary widely for small-sized kernels. Therefore, TWT may not be reliable as a direct indicator of the milling potential of durum wheat when SK is predominantly present. It is critical to monitor the KSD when a larger proportion of SK is present. Wang and Fu (2020) demonstrated that by accounting for the difference in KSD, greater relationships were found for TKW (R 2 > 0.91, p < 0.001) or the proportion of kernels passing the no.6 slotted sieve with milling yields than TWT alone (R 2 = 0.75, p < 0.001) by studying 21 wheat composites of four major CWAD varieties [5].
Influence of Kernel Size and Genotype on Milling Quality of Durum Wheat
From Table 2, a significant impact of kernel size, genotype and their interactions was found on durum milling performance (semolina and total milling yields and semolina ash content). Based on the ANOVA test, more than 80% of variation in milling yields was attributed to kernel size alone, with a greater impact of kernel size being noted for semolina yield than total milling yield (F value: 13177.7 vs. 7392.8). Figure 2 demonstrates the semolina and total milling yields in relation to TKW and TWT as affected by kernel size. Regardless of genotype selected, decrease of kernel size significantly reduced semolina and total milling yields. A drastic reduction of milling yields was evident for kernels passing no.6 slotted sieve (Table 3). On average, LK (68.0 ± 0.9%) had 1.3% higher SY than MK (66.7 ± 0.7%), and the latter was about 3.1% higher in SY than that of SK (63.6 ± 0.7%). Kernel size is clearly a better indicator of average milling yields for SK than the TWT (Figure 2). For LK and MK; however, both TWT and TKW provided strong indication of average milling quality. A similar adverse effect of SK on durum milling quality was reported by Wang and Fu (2020) and Dexter et al. (2007) by examining durum composites with a wide variation in kernel sizes [4,5].
From Figure 2, considering the response of genotype to the relationship between kernel size and milling quality, genotypes A and B appeared to be more susceptible to kernel size variations showing a greater decrease (~4.9%) in semolina yield from 68.9 to 64.0% than those of the inferior ones (e.g., G) from 66.2 to 62.9% (vs. 3.3%). A similar trend was found for total milling yield (3.5% vs. 2.7%). There were significant differences in semolina and total milling yields among the genotypes at all three kernel size fractions (Figure 2a,b). The difference in milling yields was greater for LK (2.7%) than MK (1.8%) and SK (1.3%) among the selected genotypes (Table 3).
When comparing milling quality of all kernel size fractions (Figure 2), semolina and total milling yields of MK segregated from genotypes with superior milling quality (A and C) were comparable or superior to the LK from genotypes of inferior or moderate milling quality (E, F, and G) despite the TKW of those MK (34.8 to 37.0 g) being significantly lower than LK counterparts (48.3 to 52.8 g). In addition, LK from genotypes with inferior milling quality showed lower milling yields. SK exhibited inferior milling quality to both MK and LK regardless of the genotypes selected (Table 3). SK is very detrimental to the overall milling quality but usually represents only a small proportion in commercial durum shipments. Analysis of variance by excluding SK revealed that genotype accounted for 52.0% of variation in semolina yield, followed by kernel size of 44.3% and their interaction of 3.4%. These results strongly suggest that the intrinsic kernel properties could play an important role in determining the milling quality of durum wheat. Selection of genotypes with superior milling quality could compensate the negative impact of SK which is usually present in higher percentage in dry and hot growing seasons. When a large proportion of small kernels was present; however, milling quality could be poor regardless of the genotypes selected.
Influence of Kernel Size and Genotype on Semolina and Pasta Color Parameters
Both genotype and kernel size significantly affected semolina TYP (Table 2). Figure 3 presents the semolina TYP of three kernel size fractions segregated from the selected genotypes. The decrease of kernel size led to significant increase in semolina TYP for all genotypes. Alvarez et al. (1999) reported a similar negative relationship between kernel weight and yellow pigment concentration [27]. A greater difference in TYP was shown between MK and LK (1.0-1.6 ppm) than between small and medium ones (0.2-0.9 ppm). The degree of increase in semolina TYP as shown in Figure 3 was comparable to the level previously reported by Wang and Fu, who found that semolina TYP of SK was about 1.5 ppm higher than that of LK segregated from a bulk CWAD cargo composite [5]. Large genetic variations in semolina TYP from 2.3 to 3.0 ppm were noted for the genotypes used in this study across three different kernel sizes. In addition to the milling yields, ash content is an important part of overall milling quality. The ash contents of wheat and semolina increased with the decrease of kernel size (Table 3). Coupled with the lower semolina yield of SK, its high semolina ash could further decrease the wheat milling potential when a constant degree of semolina refinement is required.
Milling quality of durum wheat is a complicated trait [10]. From Figure 2, a cooperative effect between kernel size and genotype on durum milling quality was evident when considering both MK and LK. The average milling yields of SK were lower and the impact of genotype was much less (Table 3). While the impact of some common kernel physical parameters (e.g., vitreousness, TWT, and KSD) on milling quality has been extensively investigated, the work on the intrinsic properties that contribute to varietal differences in milling quality of durum wheat are scarce [19,[21][22][23][24]. Both kernel morphological parameters (e.g., length, width, thickness, size, shape, etc.) and kernel physical properties (e.g., hardness, vitreousness, TWT) could affect milling quality. Simmons and Meredith (1979) summarized three major factors that contribute to the difference in milling quality: the amount of endosperm contained in the grain (endosperm-to-bran ratio); the separability of the endosperm from the aleurone and bran layers (structure dissociates on fracture and milling); and endosperm hardness, which determines how the kernel fragments during the milling process [19]. Novaro et al. (2001) reported ellipsoidal volume was the best predictor of semolina yield among other grain morphological parameters evaluated [25]. Haraszi et al. found that the rheological phenotype phases of an average crush response profile obtained from a single kernel characterization system provided good predictions of the laboratory milling potential of durum wheats [26].
Due to the relatively large kernel size of the original unsorted samples (Table 1) and the similar TKW of the segregated kernel fractions (Table 3), the varietal differences in milling quality among selected genotypes could be attributed to their intrinsic kernel properties. Information on hardness, endosperm-to-bran ratio, and kernel fracture behavior could shed some light on the genotypic variation in milling quality. A study is currently being conducted in our laboratory to investigate the underlying factors, which could affect the milling quality of durum genotypes with a similar size of wheat kernels.
Influence of Kernel Size and Genotype on Semolina and Pasta Color Parameters
Both genotype and kernel size significantly affected semolina TYP (Table 2). Figure 3 presents the semolina TYP of three kernel size fractions segregated from the selected genotypes. The decrease of kernel size led to significant increase in semolina TYP for all genotypes. Alvarez et al. (1999) reported a similar negative relationship between kernel weight and yellow pigment concentration [27]. A greater difference in TYP was shown between MK and LK (1.0-1.6 ppm) than between small and medium ones (0.2-0.9 ppm). The degree of increase in semolina TYP as shown in Figure 3 was comparable to the level previously reported by Wang and Fu, who found that semolina TYP of SK was about 1.5 ppm higher than that of LK segregated from a bulk CWAD cargo composite [5]. Large genetic variations in semolina TYP from 2.3 to 3.0 ppm were noted for the genotypes used in this study across three different kernel sizes. The colour of semolina and pasta made from the size fractions are summarized in Figure 4. Brightness and redness of semolina were greatly influenced by kernel size, while the genotype had a large impact on semolina yellowness ( Table 2). In general, semolina The colour of semolina and pasta made from the size fractions are summarized in Figure 4. Brightness and redness of semolina were greatly influenced by kernel size, while the genotype had a large impact on semolina yellowness (Table 2). In general, semolina prepared from MK and LK was much brighter (Figure 4a) and less dull (Figure 4c) compared to that prepared from SK. Much greater variation in brightness and redness was also shown for SK fractions than MK and LK ones (Figure 2 and Table 3).
With the decrease of kernel size from LK to MK, significant increases in semolina TYP and yellowness were shown (Figure 4e). However, except for genotypes D and G, reduction of kernel size from MK to SK did not lead to further increase in semolina yellowness despite the TYP being significantly higher in SK. The drastic decrease in semolina brightness and increase in redness for small kernels might mask semolina yellowness. Table 2 showed a large impact of kernel size on pasta color. The decrease in kernel size led to a significant reduction in pasta brightness (Figure 4b) and an increase in pasta redness (Figure 4d). Superior yellowness was seen for pasta prepared from medium and large kernel fractions. However, a drastic decrease in pasta yellowness of about 7 units was noticed for SK despite its semolina TYP being significantly higher (Figure 4f). By plotting semolina yellowness against TYP for three different kernel size fractions of the selected genotypes, it was shown that semolina b* linearly increased about 1.2 units with each ppm increase in TYP (Figure 5a). The degree of increase in semolina yellowness in relation to TYP was similar for all three size fractions. For a given TYP, however, semolina prepared from LK and MK consistently showed superior yellowness than that of SK, inferring the negative impact of SK on semolina yellowness. This negative impact was much more profound for pasta yellowness (Figure 5b). As far as SK fraction is concerned, the increase in semolina TYP resulted in little increase in pasta yellowness. This is in contrast to the MK and LK fractions evaluated in this study.
Pasta brightness and yellowness decrease with the increase of semolina ash content [28,29]. Although SK have lower semolina and total milling yields, the higher ash content suggests inclusion of a greater proportion of external tissues, which could lead to pasta browning due to high enzymatic activities [28]. Maillard reaction between amino acid and reducing sugars could lead to the undesirable reddish color of pasta dried at high temperature [30,31]. Although the protein content was not significantly higher for SK as compared with MK and LK, pasta prepared from SK was much redder (6.2-7.3 in a*) than that made from LK (2.7-3.7 in a*), suggesting other underlying factors such as amino acid composition or reducing sugar content may favor the development of the reddish coloration of pasta prepared from small kernels. Joubert et al. revisited the role of particle size, ash, and protein on pasta color and viscoelasticity [32]. By combining the milling fractions of five durum wheat patches, a series of formulated mixes of semolina/flour were prepared so that the effect of protein, ash, and particle size distribution (PSD) could be evaluated in an unbiased manner. The authors found that pasta brightness and yellowness decreased while redness increased with the increase of semolina ash content regardless of protein content and PSD. The authors attributed the increase in pasta redness to the elevation of reducing sugars accompanied by the high ash content in the semolina. A significant correlation was found between the ash content and total arabinoxylans in semolina, which were known to concentrate in the outer layers of the grain [33]. The extrusion process can significantly increase the reducing sugars due to shearing stress [34]. It is likely that the SK contains a high level of arabinoxylan, which could result in a high level of reducing sugar during extrusion and increase the potential of Maillard reactions [32]. The elevated redness/brownness and decrease in pasta brightness could subsequently mask pasta yellowness. Wang and Fu proposed that the drastic elevation in pasta redness due to the Maillard reaction under high-temperature (85 • C) drying conditions could adversely impact pasta yellowness regardless of the level of TYP [5].
Conclusions
By segregating durum samples of selected genotypes into three kernel size fractions, the impact and relative importance of kernel size, genotype, and their interaction on major quality parameters were characterized in this study. For LK and MK fractions, TWT and kernel size are closely related. However, a greater influence of genotype on TWT of SK was evident. Regardless of the genotype, the SK fraction is detrimental to durum milling performance as shown by low semolina yield, high semolina ash content, and poor semolina color. The degree of impact of genotype on the durum milling performance appears to be related to kernel size. A greater impact was shown for LK than MK and SK, based on seven genotypes evaluated in this study. When the SK fraction is excluded, the genotype or intrinsic property of the durum kernel played an important role in contributing to overall milling quality. Genotype is a dominant factor in determining semolina TYP and yellowness despite TYP increases with the decrease of kernel size. Semolina and pasta
Conclusions
By segregating durum samples of selected genotypes into three kernel size fractions, the impact and relative importance of kernel size, genotype, and their interaction on major quality parameters were characterized in this study. For LK and MK fractions, TWT and kernel size are closely related. However, a greater influence of genotype on TWT of SK was evident. Regardless of the genotype, the SK fraction is detrimental to durum milling performance as shown by low semolina yield, high semolina ash content, and poor semolina color. The degree of impact of genotype on the durum milling performance appears to be related to kernel size. A greater impact was shown for LK than MK and SK, based on seven genotypes evaluated in this study. When the SK fraction is excluded, the genotype or intrinsic property of the durum kernel played an important role in contributing to overall milling quality. Genotype is a dominant factor in determining semolina TYP and yellowness despite TYP increases with the decrease of kernel size. Semolina and pasta prepared from MK and LK fractions were much brighter and less dull than those made from SK. Regardless of the genotype, the SK fraction exerted a strong detrimental effect on pasta yellowness, despite the higher level of TYP in SK. To meet the milling and end-product quality expectation of domestic and international durum buyers, it is critical to monitor the presence of SK (through a no.6 slotted sieve) in commercial durum samples, particularly in hot and dry growing seasons. More research is needed to confirm the potential interactions between genotype and kernel size and their effects on durum quality by using wheat samples from various genotypes and different growing conditions. | 2021-12-08T16:04:50.702Z | 2021-12-01T00:00:00.000 | {
"year": 2021,
"sha1": "260932712a11db88493cc259a6ffff03d35d9487",
"oa_license": "CCBY",
"oa_url": "https://www.mdpi.com/2304-8158/10/12/2992/pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "fec71693b7ab72d5f60b7436a34fcdffcae2957c",
"s2fieldsofstudy": [
"Agricultural and Food Sciences"
],
"extfieldsofstudy": [
"Medicine"
]
} |
59493544 | pes2o/s2orc | v3-fos-license | Comparison of the anxiety and satisfaction level of the patients under fracture plating treatment type c 1 and c 2 of distal femur in the two groups of combination drugs ( gabapentin , celecoxib , and acetaminophen ) and placebo
Introduction: Prevention through pain killer drugs can be very useful compared to the treatment of the pain. Aims: Therefore, this study sought to investigate the effect of the three compound prodrugs (celecoxib, acetaminophen, and gabapentin) on the postoperation anxiety and satisfaction level of the patients under distal femur fracture plating treatment. Methods: This study is a clinical trial conducted on 100 patients in which two groups are compared respecting the levels of anxiety, satisfaction, time, morphine doses, and side effects. One hour before surgery, the case patients were given three drugs celecoxib, acetaminophen, and gabapentin with doses of 200, 500, and 400 mg, respectively. The control group patients were given the same placebos. 15 minutes before the operation, the patients were gone under spinal anesthesia through injecting epinephrine 0.2 mg and 5%-lidocaine 100 mg. After entering the ward at 1, 2, 4, 8, 12, and 00:00 O’clock following the operation, the tranquilizer intake and the first time for requesting analgesia was recorded after the elimination of spinal anesthesia. Result: There was a significant difference between the two groups respecting the comparison of the time of consuming the first morphine and its dose, anxiety level, and satisfaction. Other comparisons were not significant. Conclusion: The compound of celecoxib, acetaminophen, and gabapentin is an appropriate one so as to reduce orthopedic surgery post-operation pain. The results of this study can confirm the effect of non-drug palliative drugs on the reduction of the anxiety level and satisfaction increase.
INTRODUCTION
The thigh is a part of the lower limb that has been located between the pelvis and the knee.This part of lower limb includes a long bone in its center that is named thigh bone of Femur.This bone is very powerful and strong and contributes at upside in the formation of thigh joint and at downside in the formation of the knee joint.The thigh bone or Femur is the biggest and longest bone of the body.The long of thigh bone of a mature human is about 0.5 meter and its diameter in the center is about one inch.This bone can bear the weight 30 times the weight of a human.The distal femur fractures consist of severe and complex damages which can be resulted in the long-term inabilities and are of special importance because of being near to the knee joint (1).From 100 thousand members, the distal femur fractures are appeared in about 37 members in a year (2).These fractures include Metaphysis Distal Femur (Supracondylar) and Joint Level Distal Femur (Intracondylar) that is created in the young individuals by the mechanism of trauma with higher energy and in the old women by the mechanism of trauma with lower energy.This fracture can cause a long-term inability in the patients especially in the cases of severe damages of soft tissue, hard damage of articular cartilage and considerable crushing of bones (3).The treatment of Distal Femur Fracture is one of the most important treatment challenges because of the high level of its complications and almost all specialists emphasize the surgery to achieve favorable results (4).In these kinds of fractures, because of being near to knee joint (for the reason of the exorbitant use of this joint and also bearing the weight on that), the pain derived the surgery should be treated very soon.As we can get an acceptable result, the time of hospitalization would be reduced and the patient would begin to move earlier.So, the use of different anti-pain methods in these kinds of surgeries are of special importance.The triple therapy which is applied nowadays by some orthopedists to reduce the pain after TKA surgeries have had valuable results.The Cyclooxygenase -II such as Celecoxib don't have the side effects of Non-Steroidal Anti-inflammatory Drugs (NSAIDs).Of course, it should be noticed that the gastrointestinal, hemostasis effects and bad effects on the bone healing is reduced by the consumption of COX-2 Inhibitor in comparison to the usual NSAIDs but there is no difference in renal effects (5).The incident effects of this drug include tiredness and stress, depression, anger, paresthesia, nausea, anorexia, vomit, constipation, and xerostomia.The potential of Gabapentin in controlling the pain and its effect on the procedure of feeling the pain in the central nerve system caused that the clinical use of this drug would increase (6).The incident effects include drowsiness and dizziness and also swelling limbs.These effects often have a severity from light to average and may be disappeared by the passage of time.The other effects include nausea and vomit, diarrhea, cough, and tiredness.Acetaminophen or Paracetamol is an anti-pain and anti-fever drug and it seems that its effect is for the reason of reducing the production of Prostaglandins by the body.The prostaglandins are the materials which cause inflammation and fever.In spite of the other anti-inflammatory drugs that cause reducing the production of prostaglandins in all parts of the body, Acetaminophen reduces the production of this material just in the central nerve system (brain and spine).In a research which was done by Karvonen et al. in 2008, the effect of edible Acetaminophen and Ketoprofen in the management of pain, reduction of consumption of iPod and side effects of narcotics after the big orthopedic surgery was studied.In this study, after the surgery, the level of pain (measured by Score VAS), respiration number, oxygen saturation, level of the environmental artery, heartbeat, blood pressure and side effects were registered every 4 hours during 20 hours.In a 20 hours period, the case group consumed 0.22 Fentanyl less than the placebo group in average (p>0.05).The frequency of side effects didn't reduce by the use of non-narcotic downers (7).Niruthisard et al. in 2013 studied the effect of using separately and also the combination of two placeboes, Pregabalin (150 mg) and Ceceloxib (400 mg) in the improvement of anti-pain effect of intra-spine Morphine in the patients under Total Knee Arthroplasty (TKA).In this study, 100 patients were randomly categorized into four groups (pregabalin, celecoxib, pregabalin-celecoxib and placebo.The pain score when moving at 2 hours, in comparison to the placebo group, was meaningfully lower in pregabalin-celecoxib group (p<0.05).Furthermore, a considerable reduction was observed in anxiety scores of this group at 2 hours after the surgery and also in celecoxib group at 6 and 24 hours after the surgery (p<0.05).The side effects and also the effects after surgery, except spo2 in pregabalin group, were similar in all groups at the end of surgery.The patient's satisfaction was similar in all groups at 24 and 48 hours after the surgery (8).In the other study which was done in 2007 by Reuben et al. , the efficiency of celecoxib in the reduction of pain of patients who were under the arthroscopic therapy of Anterior Cruciate Ligament, was studied.They categorized 200 patients into two groups.One group consumed Acetaminophen (1000mg) and celecoxib (400 mg) and the other group used Acetaminophen and placebo.This regime began from one or two hours before the surgery and kept until 14 hours after the surgery.They found that the patients of celecoxib group felt less pain after the surgery (p>0.01),needed less amount of iPod to reduce their pain (p>0.001),released earlier from the hospital and felt less nausea and vomit (p>0.05).Furthermore, the patients of celecoxib-Acetaminophen experienced less pain in the move and rest at home and used less amount of Oxycodone (anti-pain drug given them (9).Therefore, with regard to the different results and studies about the effect of multidrug treatment model on the prevention of pain in different methods of orthopedic surgery, the present research aims to study the effect of consumption of three prodrugs (CAG) on the amount of stress and satisfaction after the surgery in the patients under the of plate of distal femur fracture according to Multi Analgesia Model.
METHOD AND MATERIALS
The present research is a clinical and quasi-experimental trial.From the beginning of 2013, 100 patients who were the voluntary of surgery of distal femur plate with the fracture of C1 and C2 types under the spinal anesthesia that the criteria of entrance to the study, like the age between 25 and 45, is the patient's satisfaction with the entrance to ASA (American Society of Anesthesiologists class) I and II.The patients were randomly categorized into two groups of Mord (the patients to whom three edible drugs of celecoxib 200mg, Acetaminophen 500mg and gabapentin 400 mg were given) and control group (the patients to whom the similar placebo have been given).The patient, after receiving 500cc Normal saline in the sitting situation, was put under the spinal anesthesia at one turn by transfusing 0.2mg of Adrenaline and 100mg of Lidocaine %5 by Whitaker Spinal Syringe No. 24 in the L4-L5 intervertebral space.For the purpose of all patients' having less pain after the surgery, all of them were gone under the surgery without a tourniquet.The Titanium Locking Plate was done in all patients by lateral cutting of distal femur and knee joint.After being sure of proper locking and the control by fluoroscope and proper draining and washing, all the layers of the wound were stitched and then splinted.After the entrance of patients into the department and reduction of spinal anesthesia to the level lower than T10 after the surgery, amount of pain, consumed tranquillizer and the first time of requesting the analgesic drug after the removal of spinal anesthesia were registered at 1, 2, 4, 8, 12 and 24 hours after the surgery.Every 6 hours, 5mg Morphine IM was transfused into both groups.The quality of patients' satisfaction with the anesthesia during 24 hours after the surgery was evaluated in excellent, good, average and improper criteria.Ultimately, the questionnaire of evaluation of the amount of patients' stress was given to them.This questionnaire includes 40 questions which consist of self-evaluation separate indexes for measuring the obvious and hidden stress.The obvious stress index (y-1 form of STAI) consists of 20 sentences which evaluate the individual's feelings "in this moment and the time of answering".Also, the hidden stress index (y-2 form of STAI) includes 20 sentences that evaluate the general and usual feelings of the individuals.In testees' answering to the obvious stress index, a number of choices have been presented for each sentence that the testees should choose the choice which expresses his/ her feeling in a better way.These choices include very little, little, much, so much.In answer to the hidden stress index, the testees should choose choice which is the indicative of usual and natural feeling of them, in a four-choices indexes as following: almost never, often, most of the times, almost always.The criteria except the study included: existence of background diseases (such as diabetes) or organic regurgitation, drug and alcoholism addiction, vascular problems or apoplexy and heart-failure, sensitivity to iPods and gabapentin and celecoxib, renal failure, multiple trauma, fatness (BMI<35), pathologic fractures, liver diseases, emotional disorders, pregnancy, lactation, IBD, consumption of Aspirin or Warfarin and the individuals who consume permanently analgesic drugs for any reason.Data were analyzed by the use of SPSS-21 software.The Independent T test was used in a normality form for the purpose of comparing the amount of consumption of Morphine and Paracetamol and the time of anesthesia in two groups and otherwise, the non-parametric Mann Whitney test was used for the purpose of determining the amount of effect of combinative drugs on the amount of pain after the surgery by controlling the effects of background variables and the new changed models (Glz) in a GEE method for studying the consumption of Morphine and Paracetamol.The level of significance of the tests in the present study was evaluated with p<0.05 and the tests were studied in a two-sided form.
RESULTS
In this research, 50 members in the group of combinative drugs (celecoxib, Acetaminophen, and gabapentin) and 50 members as the placebo group were studied in the regard of the level of stress and satisfaction.The findings o present research revealed that the percentage of man to woman in the group of combinative drugs was %86 to %14 and in the placebo group was %88 to %12 that this difference has not been statistically meaningful with regard to Chi Square Test (p= 0/766).The average of relative standard deviation in the group of combinative drugs has been 9/27 ± 9/3 years and in the placebo group was 1/28 ± 9/3 that the age of two studied groups was similar and has no meaningful difference (p= 0/915).Also, there was no meaningful difference in the body mass index of two groups (p= 0/973).With regard to the time of surgery, the group of combinative drugs from the minimum time of 95 to 125 minutes had 107/5 ± 7/6 minutes average and standard deviation and in the placebo group, from minimum surgery time of 90 and 120 minutes, the average and standard deviation have been 104/4 ± 7/9 minutes.Although the surgery time of two groups has been meaningful (p=0/0046), but this difference (3 min) is not clinically important in Table 1.
Table 2 shows the comparison of first morphine use time and its dose use.With regard to Mann Whitney U Test, the table data shows a meaningful difference in the time of beginning of morphine use (p<0/0001) and also in its dose use (p<0/0001) that in both measured variable the first morphine use time in the group of combinative drugs (235/7 ± 11/6) is more than the time in the placebo group (170/9 ± 13/5) and also in the regard of dose of consumption, the group of combinative group (7/6 ± 2/5) is less than the placebo group (46 ± 7/9).The comparison of the first morphine consumption and the amount of dose is shown in Figure 1.In the comparative study of severity of stress with regard to anxiety score, Table 3 shows a meaningful difference in the anxiety severity of both two groups (combinative and placebo) (p<0/0001) in a way that the percentage of relatively severe and severe and very high anxiety in the group of combinative drugs (39) has been less than the placebo group (62) (Figure 2).The data of Table 3 shows a meaningful difference in the regard of the level of satisfaction (p<0/0001), in such a way that the percentage of level of good and excellent in the group of combinative drugs has been almost %60 but in the placebo group has been %28.The ranking mean of the level of satisfaction in the group of combinative drugs has been almost more than 22 ranks (63/37 versus 37/63) (Figure 3).Also, in the regard of the amount of effects, the table data reveals that the percentage of effects in the group of combinative drugs has been %22 and in the placebo group %28; but the amount of effects has had no meaningful difference statistically (p=0/488) (Figure 4).In the comparison of effect of the combinative drugs with the placebo group on the amount of pain by the moderation of effects of morphine use time and also dose of morphine and surgery time with regard to the statistic method of Variance Analysis ANCOVA of moderation of data revealed that the effect of combinative drugs in comparison to the effect of placebo on the amount of pain (p=0/518) by controlling the amount of morphine use and its use time has not been statistically meaningful; but the used morphine dose (p=0/008) and morphine use time (p=0/013) have been of effective factors in the amount of pain.Also, the moderation between two groups studied and the used morphine dose is meaningful (p=0/044).The other comparisons and the surgery time have not been meaningful (p>0/05).
DISCUSSION AND CONCLUSION
According to the results in the regard of surgery time, the group of combinative CAG drugs had minimum time of 95 min to 125 min with the mean and standard deviation of 107/5 ± 7/6 and in the placebo group, the minimum time of surgery has been 90 min to 120 min with the mean and standard deviation of 104/4 ± 7/9, that the surgery time of two groups has been statistically meaningful.But, this difference (3 min) is not clinically important.Perhaps, the reason for this difference is that the use of analgesic drugs has reduced the need to anesthesia and analgesic drugs during the surgery and finally has reduced the time of surgery (10).Gabapentin is the structural analog of Gama Amino butyric acid (GABA).This drug at first for the treatment of seizure, then it was used effectively in alleviating the neuropathic pains.The mechanism of its action is by applying an inhibitive effect on the entrance of calcium by the activated channels with high voltage that includes subunit alpha-2 delta-1 and causes the reduction of release of Neurotransmitter and postsynaptic excitation (11) and it is recently used as an effective drug to reduce the pain after surgery (12).Celecoxib which is of the group of alternative inhibitive cyclooxygenase enzyme-2, has been expressed as a substitute for the use of non-alternative Non-Steroidal Anti-inflammatory drugs; because, in comparison to the classic Non-Steroidal Antiinflammatory drugs, it has anti-pain effects and has no side effect on the platelet, gastrointestinal mucus system and renal function (13).On one side, Acetaminophen is one of the relative new drugs which is used to reduce the pain of patients and acts as anti-pain and anti-fever and also has anti-inflammatory tinge.Mechanism of action is accompanied by the inhibition of santes prostaglandins.The Acetaminophen has been used in an edible and rectal form and recently in the transfusing form.Acetaminophen is relatively safe and has fewer side effects and also has limited contraindication; also, it has not obvious interactions with the other drugs.With regard to being safe of the aforesaid drugs, it seems that the combinative use of them can help reduce the pain, even during many hours after the surgery.According to the obtained results, the first morphine use time in the group of combinative drugs is more than the placebo group and also, the morphine use dose in the group of combinative drugs has been less than the placebo group.The results of Imani and his colleagues' study revealed that the mean of morphine use during the first 24 hours in the Gabapentin group was less than control group that there was statistically no meaningful difference between two groups (14).The results of Poornajafian and his colleagues' study which compares the painlessness after surgery in the patients under the surgery of shin fracture with the prescription of celecoxib as placebo at different times before the surgery, showed that the mean of time of requesting downer from the termination of spinal anesthesia in the group of celecoxib 400mg was 196/5 ± 121 minutes that was meaningfully more than the group of celecoxib 200 mg (127/5 ± 68) and group of control (122/7 ± 44); but in the regard of dose of received downer, there observed no statistic meaningful relationship between three groups.In the study done by Niruthisard et al. about morphine use first time and its dose, there observed no meaningful difference between the placebo group and the group received pregabalin-celecoxib that this result contradicts with the present study.The control and reduction of severe pains of the patients are so important and significant, the use of anti-pain narcotics as a prodrug is current in the general anesthesia and is generally used during the surgery as fixing the hemodynamic of the patients, too.But for the reason of side effects of these combinations such as nausea, vomit, respiratory weakness, urinary retention and …, the similar non-narcotics are recently used for controlling the pain of patients.It seems that the combination of aforesaid drugs can reduce the need for the narcotics for the control and reduction of anxiety and ultimately, the side effects of the narcotics is reduced, too.Also, the results of present study showed that there was a meaningful difference in the anxiety severity of two combinative drugs and placebo groups, in such a way that the percentage of relatively severe, severe and very high anxiety in the group of combinative drugs has been meaningfully less than the placebo group (39 versus 62).The results of the study done by Niruthisard and his colleagues revealed that the anxiety score in the patients received celecoxib and pregabalin, was reduced two hours after the surgery.Talebi et al. in a research titled study of effect of educational-combinative interference on level of anxiety in patients voluntary in orthopedic surgery, revealed that there was a statistic meaningful difference in the amount of anxiety of two groups of interference and control at the first post-test (before surgery).
Ghaneii and his colleagues in a research titled study of the relationship of anxiety before surgery with pain after cesarean showed that there observed a meaningful statistic relationship between the obvious and hidden anxiety before surgery with the pain after surgery (15).The anxiety is very usual in the patients voluntary for surgery and can cause problems at the time of surgery and after surgery.The results of the present research show that the use of a combination of downers before the surgery has resulted in the peace and reduction of anxiety in the patients.It seems that inasmuch as the pain is an effective factor in the creation of anxiety; so, the reduction of pain reduces the anxiety in the patients.In the regard of the level of satisfaction, there was a meaningful difference between two groups.In such a way that the percentage of good and excellent satisfaction in the group of combinative drugs was about %60 and in the placebo group was about %28.The ranking mean of the level of satisfaction in the group of combinative drugs has been more than 22 ranks (63/37 versus 37/63).The results of the study done by Tavakkoli et al., revealed that 70/55 percent of the patients were completely dissatisfied with the reduction of their pain and 29/45 percent have weak and average satisfaction, too (16) that the results of this study contradict with the present study.The results of Bamshaki and colleagues' study titled amount of severe pain after surgery and patients' satisfaction in Laparotomy, Cholecystectomy, and Herniorrhaphy, revealed that %921/3 of patients had relative to perfect satisfaction with the amount of controlling their own pain.And only %7/7 of patients expressed levels of dissatisfaction (17).It seems that if the pain of patient be controlled better and be reduced, it would result in the increase of satisfaction and tranquility of the patients.In the study of the effect of combinative drugs and placebo group, there was a meaningful difference between two groups in the comparison of anxiety severity and satisfaction.Although the moderation between two groups studied and the amount of morphine used was meaningful.The other comparisons and the surgery time were not meaningful.The present study shows that the combination of three drugs (Gabapentin-Ceceloxib-Acetaminophen) is a proper combination for the reduction of anxiety and increase of satisfaction after orthopedic surgery.The results of present research can be used as a guide in the next studies for further verification of effect of non-narcotic downer drugs in the reduction of pain.
Figure 1 :Figure 2 :
Figure 1: The Mean of the Post-operation Satisfaction Score based on VAS in the Two Groups
Figure 3 :
Figure 3: The Mean of the Post-operation Satisfaction Score based on VAS in the Two Groups
Figure 4 :
Figure 4: The Amount of Complications Caused by Surgery in the Two Groups
Table 1 :
Comparison of the First Morphine Consumption and the Amount of Dose
Table 3 :
Comparison of the Anxiety Level Based on Score Stae | 2019-01-01T23:12:12.569Z | 2018-04-05T00:00:00.000 | {
"year": 2018,
"sha1": "8f7b171cb1ef22af9780c522c7a8a876457f7976",
"oa_license": "CCBY",
"oa_url": "http://www.ejgm.co.uk/pdf-85678-24539?filename=Comparison%20of%20the%20anxiety.pdf",
"oa_status": "GOLD",
"pdf_src": "Anansi",
"pdf_hash": "8f7b171cb1ef22af9780c522c7a8a876457f7976",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
59331524 | pes2o/s2orc | v3-fos-license | Genetic parameters for test day milk yield of first lactation Holstein cows estimated by random regression using Legendre polynomials
Data comprising 263,390 test-day (TD) records of 32,448 first parity cows calving in 467 herds between 1991 and 2001 from the Brazilian Holstein Association were used to estimate genetic and permanent environmental variance components in a random regression animal model using Legendre polynomials (LP) of order three to five by REML. Residual variance was assumed to be constant in all or in some classes of lactation periods for each LP. Estimates of genetic and permanent environmental variances did not show any trend due to the increase in the LP order. Residual variance decreased as the order of LP increased when it was assumed constant, and it was highest at the beginning of lactation and relatively constant in mid lactation when assumed to vary between classes. The range for the estimates of heritability (0.27 0.42) was similar for all models and was higher in mid lactation. There were only slight differences between the models in both genetic and permanent environmental correlations. Genetic correlations decreased for near unity between adjacent days to values as low as 0.24 between early and late lactation. A five parameter LP to model both genetic and permanent environmental effects and assuming a homogeneous residual variance would be a parsimonious option to fit TD yields of Holstein cows in Brazil.
Introduction
The use of test day (TD) measurements instead of 305 days lactation records has been stimulated by the possibility to improve breeding value estimation of dairy cattle (Jensen, 2001).Different approaches have been used for the analysis of TD records.The simplest approach considers TD records as repeated measures of the same trait assuming a genetic correlation of unity between yields along lactation.Another approach considers TD yields as different traits and a multivariate model is used for the analysis.An intermediate approach, more refined than the repeatability model, less parameterized and probably computationally less expensive than the multivariate model, is the random regression model (Schaeffer & Dekkers, 1994).An autoregressive TD model (AR) is an alternative approach by assuming that test day yield is the product of the expression of the same set of genes throughout the productive life of the female (Carvalheira et al., 2002).
The random regression (RR) approach allows to fit sub models for adjusting the lactation curve, assumes a structure for genetic and environmental variation specific to individual TD yields and variable correlation between TD yields (Jamrozik & Schaeffer, 1997;Meyer, 1998b;Swalve, 2000).Since Kirkpatrick et al. (1994) demonstrated the use of orthogonal Legendre Polynomials (LP) in modeling the covariance structure of TD records of dairy cows, RR have been used to model milk production along lactation (Olori et al., 1999;Brotherstone et al., 2000;Pool et al., 2000).
The choice of sub model for fitting additive genetic and permanent environment effects is the focus in finding an optimal RR model.Orthogonal polynomials are most appropriate than parametric lactating curve functions for the covariates in RRM.Orthogonal polynomials of time have much lower correlations among the coefficients and provide estimates of the covariance matrices that tend do be more robust over different data sets (Schaeffer, 2004).
The use of TD records greatly increases the amount of data to be analyzed, thus requiring the selection of a more parsimonious model with respect to the number of parameters to be estimated.Olori et al. (1999) concluded that critical issues in fitting a random regression model include the order of the polynomial used to model the lactation curve at both the fixed and random level, and the grouping of observations into residual variance subclasses.
This study aimed to investigate the use of LP for modeling the lactation curve of Holstein cows in Brazil and to determine the most appropriate genetic covariance structure among daily milk yields required for genetic evaluation.
Material and Methods
Data provided by ABCBRH (Brazilian Holstein Association) comprised 263,930 milk yield TD records of first lactation from 32,449 cows calving in 467 herds between 1991 and 2001.TD yields from 5 to 305 days of lactation, time interval between successive tests less than 45 days and age of calving between 18 and 48 months were required.Contemporary groups were defined as herd-year-test month (HTM) and were required to have at least four records.Average and standard deviation for test day milk yield were 22.53 and 6.87 kg, respectively.Cows with records were daughters of 1,955 sires.The pedigree file including parental identifications defined the relationship matrix A.
Variance components were estimated by animal model REML using the "DXMRR" program (Meyer, 1998c).Nine analyses were performed using orthogonal Legendre polynomials (LP) of order 3, 4 and 5 as sub models to fit additive genetic and permanent environment effects.Two methods of accounting for residual effect (measurement error = ME) were analyzed for each RR model.First, it was assumed that the residual variance was constant (homogeneous) throughout the lactation or ME = 1.Alternatively the residual variance was assumed to be constant for TD records within, but different (heterogeneous) between the followings groups of lactation period: [5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28][29][30] for ME = 29.The models were designated by LPmMEr, where m is the order of the additive genetic and permanent environment effects and r is the number of ME classes.Then, LP5ME1 is a model with LP of order 5 for additive genetic and permanent environment effects and one ME class.
The random regression equation, assuming the same sub model to fit fixed, genetic and permanent environment effects, was: where Y ijzl is the l th observation on animal z, at days in milk (DIM) t, age at calving X of animal z, in months, as both linear and quadratic covariate (p=1,2), in season j and HTM i; β m are the fixed regressions coefficients for an average population curve; α zm the random genetic and γ zm the random environmental regression coefficients for animal z, respectively, m the order of the polynomial fit, Φ m (t) the m-th orthogonal polynomial of DIM t (standardized in the range -1 to +1 representing days 5 to 305), f and k are respectively the number of coefficients of the fixed and random effects of the polynomial and (e ijz ) r is the temporary measurement error associated with the specific test day record, where r is the number r of ME classes (r=1, 4 or 29).
Details of the log likelihood, assumptions and estimation of variance components modeled with orthogonal polynomial regressions have been presented (Meyer & Hill, 1997;Meyer, 1998a).The convergence criterion was set to 10 -4 .Fit of different models was compared by examining estimated variances, maximum likelihood and Akaike Information Criterion (Meyer, 2001) for each analysis.Variance components and heritability were also compared with estimates obtained using a repeatability model and bivariate analyses of TD milk yields (Melo et al., 2005).
Results and Discussion
Values of Log-likelihood functions, number of parameters and Akaike information criterion (AIC) obtained from fitting LP assuming both homogeneous and heterogeneous residual variances are shown in Table 1.Log-likelihood functions increased slightly with increasing order of polynomials.However differences between loglikelihood functions due to increasing number of ME within each polynomial fit decreased from 391.93 to 81.94 as the order of LP increased.In terms of AIC, the model with the best fit was the polynomial of order five with 29 ME classes (Table 1).
Estimates of residual variance decreased as the order of the model increased and were estimated as 5.41, 4.79 and 4.41 kg 2 respectively for orders 3 to 5 when it was assumed constant during lactation.Generally residual variance was larger in the beginning and decreased along lactation when assumed to vary between classes of lactation period or ME=4 (Table 2).
Looking at the heterogeneous residual variances models for LP5, differences were not large taking into account the expressive increase in number of parameters.Also, the range of residual variance (RV) estimates along lactation did not differ between ME4 and ME29 in the first half of the lactation period (Figure 1).In the second half, variation in RV was not continuous for ME29.RV estimates for ME4 were in the range (3.93-4.79) of values observed for ME29.According to Lopez Romero et al (2003) the heterogeneity of RV is related to the lactation stage and it is larger at the extremes of lactation.This is likely due a set of no specified factors in the model equation (days open, pregnancy status, body condition at calving, etc.) that make the temporary ME larger and highly variable at the beginning and at the end of the lactation.The inclusion of all these effects in the model equation might be difficult, mainly because the lack of information.
The RV estimates in this study were significantly smaller than that obtained by fitting a repeatability animal model for TD yields (RV=8.55kg 2 ) or than those from fitting individual TD yields with univariate models (between 17.86 and 15.57kg 2 ) as reported by Melo et al. (2005).A summary of the variance components estimated by different models is shown in Table 3. Estimates of the different components of variance depended on the lactation stage and the order of the LP fitted, but differences were not large.Except for the extremes of the lactation, estimates of genetic variances were smallest for LP3ME4 (Figure 2).Similarly, estimates of permanent environmental variances were smaller for LP3 than for the other models.Similar values and trends were observed for LP4 and LP5.Small differences were observed at the beginning and end of the lactation period.Meyer (1999) observed that some problems may occur with the fit of random regressions at the extremes of the ages in the data.In part that can be explained by small numbers of records and sampling variance in partitioning of total variance.Furthermore, this is likely to be due to the large effect that the values furthest from the mean have in a regression analysis.
Genetic variances from this study were slightly larger than those obtained by Melo et al. (2005).Their reported estimates were 7.21 kg 2 for a repeatability TD model and ranged between 4.8 and 7.88 with univariate models fitting TD milk yields.Estimates with TD in mid lactation were the most similar between these studies.Except for higher estimates at the extremes, the same trend was observed in both studies for phenotypic variances.
Similar estimates of genetic and permanent environmental variances were obtained for LP5 for models ME29, ME4 and ME1.Olori et al. (1999) found that different models to fit RV resulted in different estimates but variation on genetic variances was small.They observed that accurate identification of lactation stages with similar error variance (into the same ME classes) was more important than the number of ME classes specified in the RR model.Variance components estimated by fitting the model LP5ME1 are shown in Figure 3.
No specific trend was observed for estimates of covariance components and correlation for random regression coefficients.Genetic correlation estimates were in general not large and ranged from -0.54 (a 1 ,a 2 ) to 0.34 (a 3 ,a 4 ) for LP5ME1 (Table 4).All models gave similar range for the estimates of heritability (0.27-0.42), which were higher about mid lactation.Estimates for models LP3-5 and ME1 are on Figure 4 and for LP5 and ME1-29 are on Figure 5. Heritability estimates along the lactation trajectory showed similar shapes, but were less extreme at the beginning and end of lactation, because of larger permanent environmental variances.These trends were similar to results of Pool et al. (2000), but contrary to those reported by Olori et al. (1999) and Brotherstone et al. (2000).Overall LP heritabilities were larger than the estimates of individual TD yields from univariate models reported by Melo et al. (2005), which varied from 0.23 to 0.33.This is due to larger values for GV and smaller values for RV obtained by the LP models fitted in this study.
Estimates of genetic and permanent environmental covariances did not show any trend due to increase in the order of LP.The pattern of genetic correlation estimates between TD was consistent over all models.Estimates for LP5ME1 are given in Table 5.The genetic effects at the beginning of the lactation were less correlated than those at the end of the lactation.The genetic effects for the TD yields in the middle of lactation were highly correlated with correlations higher or equal 0.90 between the genetic effect at DIM 165 and genetic effects of a large part of the lactation (from 125 to 285 DIM).Genetic correlation was equal to 0.25 for the extreme parts of the lactation.
Permanent environmental correlation estimates followed the same pattern of genetic correlation estimates, declining from near unity between adjacent yields to values as low as 0.11 between the beginning and the end of lactation.A graphic illustration of genetic and permanent environmental correlations between test day milk yields for LP5ME1are shown in Figure 6.
Genetic correlation estimates are in agreement with those reported by Brotherstone et al. (2000) and Olori et al. (1999).Jamrozik & Schaeffer (1997), Kettunen et al. (2000) and Rekaya et al. (1999) reported negative correlation estimates for the extreme parts of the lactation.López-Romero & Carabaño (2003) investigated 22 alternative RR models using Holstein data in Spain.Differences in variances among models were more mostly observed at the extremes of lactation.All the models assumed homogeneous RV along lactation.The model of choice was the most complex one, a LP of order six.However, they pointed out that submodels with a lower order polynomial for the genetic than for the permanent environmental component would be sufficient to account for the variability of these effects.In addition, in a following study, Lopez-Romero et al. (2003) observed that increasing the order of regression of permanent environmental effect resulted in correction for RV.In conclusion, the assumption of homogeneity of RV was the most plausible specification when the number of random regression coefficients was set to five.Similar results regarding the higher order of LP for the permanent environmental than for the genetic effect led Liu et al. (2006) to conclude that the currently Legendre polynomial of order five (for both additive genetic and permanent environment effects), after no consensus on the best model among different evaluation criteria, was the optimal model for multiple trait genetic evaluation for production of Holstein cattle in Canada.
Random regression models have been widely studied and evaluated for genetic evaluation at national level in many countries.RR models have the advantage of flexibility to account for the environmental and genetic components of the shape of the lactation curve.Currently, eleven countries are using TD records in their genetic evaluation systems for production in dairy cattle.Among them, eight countries use RR models (Interbull, 2006).
Genetic evaluations for production traits of Holstein cattle in Brazil are based on first lactation records adjusted to 305 days (Costa et al., 2006).Results from this and the previous study of Melo et al. (2005) suggest TD records may replace with advantages lactation data for genetic evaluations of milk yield.In addition RR models are better models than the repeatability model to fit test day milk yields of Holstein cattle in Brazil.Also it would be an opportunity to implement routine evaluations for persistency which may have greater economic value under subtropical conditions.Parameters estimated in RR test day models may be used to calculate genetic measures of persistency (Jakobsen et al., 2002;Cobuci et al., 2004).
Conclusions
The orthogonal Legendre Polynomial of order five to model animal genetic and permanent environmental effects best fitted the data.Residual variances were not homogeneous, but no significant differences were observed for genetic and permanent environmental variances when fitting different classes of ME across lactation.
Results from this study showed that using a five parameter LP to model both genetic and permanent environmental effects and assuming a homogeneous residual variance would be a parsimonious option to fit TD milk yields.
Further research is still needed to compare ranking of animals and expected genetic gain to decide whether replacing 305d records by TD records for breeding value estimation for production traits of Holstein cows in Brazil.
Figure 2 -
Figure 2 -Genetic variance estimates of test-day milk yields along lactation for LP models.
Figure 4 -
Figure 4 -Estimates of heritability of test day milk yield obtained by models LP3ME1, LP4ME1 e LP5ME1.
Figure 6 -
Figure 6 -Graphical illustration of additive genetic (above) and permanent environmental (bellow) correlation estimates for test day milk yields along lactation with model LP5ME1.
Table 3 -
Additive genetic, permanent environmental and phenotypic variances for daily milk yield at different days in milk (DIM) using LP models | 2018-12-23T11:04:59.066Z | 2008-04-01T00:00:00.000 | {
"year": 2008,
"sha1": "60c52424c81106539ac46b0920035989f75d5a71",
"oa_license": "CCBY",
"oa_url": "https://www.scielo.br/j/rbz/a/qvpCDsgr6TZS4w68VBchqCN/?format=pdf&lang=en",
"oa_status": "GOLD",
"pdf_src": "Anansi",
"pdf_hash": "60c52424c81106539ac46b0920035989f75d5a71",
"s2fieldsofstudy": [
"Agricultural And Food Sciences"
],
"extfieldsofstudy": [
"Biology"
]
} |
145026278 | pes2o/s2orc | v3-fos-license | Complex Sentences and their Punctuation in English Texts Composed by Latvian Students
Comma use in Latvian and English complex sentences differs significantly and this can cause major problems to Latvian users of English. In particular, the dependent clause that follows the independent clause is normally separated by a comma in Latvian complex sentences (Blinkena, 2009), whereas comma use/non-use in English complex sentences (Downing, Locke, 2006) depends on the semantic relation between the independent and the dependent clause. These comma use differences, as previous research reveals (Farneste, 2006b), can cause challenges in English complex sentence punctuation to students whose native language is Latvian. The goal of this study is to research Latvian students’ punctuation of English complex sentences extracted from a corpus of essays. The analysis covers the cases of nominal and relative dependent clauses that are tightly related to the independent clause they follow. Normally, as Downing and Locke (2006) note, such dependent clauses are not separated by a comma in English texts. The analysis shows that a strong semantic relation between the clauses has been relevantly recognised by students in numerous examples, which they have relevantly punctuated. However, the results also reveal unnecessary separations of relative and nominal clauses from the independent clause by a comma. These findings call for the focusing of students’ attention on the role of the semantic relation of clauses in English complex sentence punctuation by contrasting these sentences with the relevant complex sentences found in Latvian texts. In addition, further analysis of various sentence types and their punctuation in student-composed texts of different communicative purposes is required.
Introduction
Comma use in Latvian and English complex sentences differs significantly and this can cause major problems to Latvian users of English.The dependent clause that follows the independent clause is normally separated by a comma in Latvian complex sentences (Blinkena, 2009), whereas comma use/non-use in English complex sentences (Downing and Locke 2006) is related to the semantic interdependency between the independent and the dependent clause.These and other differences in comma use have been previously discussed by Farneste (2006aFarneste ( , 2006b)).She concludes that differences can cause English sentence punctuation challenges to students whose native language is Latvian.Farneste (2006a, 37) emphasises that first-year university students in Latvia, when asked if they discussed punctuation in English texts in English classes at their schools, predominantly point out that it was rarely done.The author also notes that there are very few tasks in English course books focusing on the communicative function of punctuation marks.Therefore, students often tend to rely on the punctuation rules that have been discussed in Latvian language classes.Meanwhile, the research on comma overuse in complex sentences in English texts written by Latvian students has not been taken up yet.
The goal of this study is to explore examples of complex sentences written by Latvian students in order to uncover how far their punctuation corresponds to the semantic relation between the independent and dependent clauses.
Seeking to achieve this goal, the following enabling objectives are set: (1) a comparative analysis of the theories on sentence classification according to clause relations and clause types; (2) extraction of examples of complex sentences from a corpus of essays composed by Latvian students (with the help of AntConc) that have been part-of-speech tagged with CLAWS7; (3) grouping of the extracted examples into sentences that contain nominal and relative dependent clauses; (4) analysis of the semantic relation of the dependent and independent clauses and their punctuation as proposed by the students.
Theoretical background
Sentences, 'the highest grammatical units' (Downing, Locke, 2006, 272), that contain more than one clause or verb phrase (Biber et al., 1999, 100, 120) as a central element are classified by linguists in terms of the relationship of the elements within the sentence.Researchers of the Latvian language (e.g.Auziņa et al., 2013, 831;Beitiņa, 2009, 173;Ceplītis et al., 1989, 110-111) label all such sentences as composite sentences (in the Latvian language: salikti teikumi) and use this as an umbrella term for all sentences that go beyond one clause.This term is not normally found in the grammar of the English language.However, it has been used by some linguists, for example, Lyons (1995, 157) in order to distinguish 'non-simple' sentences, which contain more than one clause, from simple sentences.It is also used by Davidson (2005, 30) in his punctuation manual 'How to Punctuate'.
Apart from compound and complex sentences, the so-called compound-complex sentences (in the Latvian language: jaukti salikti teikumi) are distinguished by the researchers of the Latvian language.They contain several clauses of equal as well as unequal status (e.g.see Auziņa et al., 2013, 831).In this connection Downing, Locke (2006, 273-174) explain that coordination and subordination of clauses in a connected discourse 'may be […] complex and variable', so they expand the notion of a complex sentence by claiming that 'a complex sentence can consist of any number of clauses of different types and in different combinations'.
Complex sentences or sentences with unequal relation of clauses are further described by linguists from two angles -the type of dependent clause and the semantic degree of subordination in its relation to the independent clause, both of which play an important role in punctuation solutions in English.According to traditional grammar, there are three main types of dependent clauses that are named after the relevant parts of speech: Noun, Adjective and Adverb clauses; however, there are terminological variations in the description of Noun and Adjective clauses.Noun clauses (Yule, 2006, 160) are referred to as nominal clauses by Biber et al. (1999, 193) or complement clauses by Trask (1993, 10) and Hurford (1994, 232), for example.The term 'adjective clause' or 'adjectival clause' seems to have been abandoned by linguists in favour of the term 'relative clause' (Biber et al., 1999;Downing, Locke, 2006;Yule, 2006, etc.).The term 'adverbial clauses' is generally accepted and has been widely used by linguists.
Although the previously mentioned dependent clauses are in a subordinate relation to the relevant independent clauses, the degree of subordination of dependent clauses Didaktinė lingvistika can vary in a complex sentence.Downing, Locke (2009, 275) emphasise that it is important to '[…] adopt the view that dependency is not an absolute property, but rather a question of degree'.They explain that the dependency degree shows the degree of integration of a dependent clause as 'perceived or imagined by the speaker or writer between events'.The tightest integration is referred to as embedding by linguists, which means that the dependent clause functions as a constituent of the independent clause.
On the basis of the aforesaid classification of complex sentences, some linguists point out that the English dependent clauses, which are embedded within complex sentences (e.g.Downing and Locke, 2009) and, therefore, do not require separation by a comma, are: nominal that-clauses (Example 1); nominal wh-interrogative clauses (Example 2) including whether and if, as well as relative (restrictive) that and wh-clauses (Examples 3 and 4).
(1) He saw that the bottles were empty.
(2) I'll ask where the nearest underground station is.
(3) The house that they live in is large.(4) Perhaps the people who were waiting are still there.
The perception of the degree of dependent clause subordination plays an important role in complex sentence punctuation in English.As it is explained in English punctuation manuals, a comma is used 'to make a slight break' (Davidson, 2005, 78), and for this function Lukeman (2007, 31) has metaphorically described a comma as 'the speed bump of the punctuation world' because 'with its power to pause, the comma controls the ebb and flow of a sentence'.This implies that the pause marked by a comma is relevant in the cases where the dependent clause is less tightly integrated.Comma variation caused by semantic dependency in English complex sentences obviously explains why English punctuation is considered a challenging issue of text composition by Latvian students of English (Farneste, 2006a(Farneste, , 2006b)).
The interrelation of an embedded subordinate clause separated by a comma reveals a noticeable, though neglected, aspect of comma use differences in Latvian and English complex sentences.Although some researchers discuss the semantic relation of clauses in Latvian complex sentences (e.g.Beitiņa, 2009), this phenomenon does not have the same impact on the punctuation in Latvian texts as it has in texts written in English.In Latvian complex sentences, as Blinkena (2009, 176) states, all dependent clauses, irrespective of their semantic dependency degree, are normally separated by a comma (see Example 5 and its translation into the English language -Example 6).
Thus, the theoretical analysis has revealed that comma use/non-use in English complex sentences, containing an independent clause followed by a dependent clause, is based on the type (e.g.nominal, relative) and semantic degree of subordination.The neglect of semantic dependency might cause punctuation confusion as, for example, in the following case: a nominal who-interrogative clause (Example 7), which is embedded in the main clause and, therefore, is not separated by a comma, might be confused with a non-embedded (i.e.non-restrictive/non-defi ning) relative clause introduced by who (see example 8) and, therefore, separated by a comma.The loose dependency degree of all the wh-non-restrictive/non-defi ning relative clauses should also be marked by the comma use.(7) They know who has left this paper on the table.(8) I'll give the CD to Ben, who likes music.
Method, procedure and results
The material for analysis was extracted from a Latvian student-composed text corpus and the analysed set contained 104 texts of approximately 21,319 words.The texts were written by students before their participation in a writing course.Their selection aimed at revealing if the students had used the above-discussed clause types and patterns, namely, the embedded clauses (nominal and relative) that are preceded by an independent clause, and what punctuation solutions they offered to serve as a departure point for the development of their writing skills.
In order to extract these clause types and patterns, corpus-based data extraction procedures were applied (McEnery et al., 2006;McEnery, Hardie, 2012).The clauses were extracted with the help of the AntConc concordancer.Seeking to ease data extraction, the texts were part-of-speech tagged by CLAWS7, which contains tags for the linguistic features functioning as subordinating conjunctions in complex sentences that are in the focus of this study (see Table 1).The tag CST of CLAWS7 was used as a query to extract nominal and relative that-clauses as conjunction concordance lines that were sorted into nominal and relative The students' texts, as expected, contained both nominal and relative that-clauses.The majority of these clauses (Table 2) were correctly presented by the students as embedded, as they had not been separated by a comma.Example 8 shows an embedded nominal clause, and Example 9 an embedded relative clause that, due to their strong semantic relation, are not separated by a comma.The students embedded their nominal clauses in such common structures as I think; I would like to tell; people think; I knew that; I must say; I can conclude' etc., as seen in Examples 8 and 10. (8) Some people think that travelling is a bad and useless way of spending money.(9) There are many routes that go through the city.
There were also occasional dependent nominal and relative that-clauses that, irrespective of their embedded information, were separated by a comma.The two sentences in Example 10 show comma use in nominal clauses, but example 11 shows comma separation of a restrictive relative clause.(10) I believe, that this way of transport is quite developed.// I think, that it is a beautiful city.(11) It has been circumstances, that are the most negative factor.
The overall percentage of nominal and relative clauses that were/were not separated by a comma was equal.It has to be noted that commas would normally be used in such sentences in the Latvian language, so the redundant comma use might have occurred due to the impact of their use in the Latvian language.The extraction of nominal interrogative clauses required the application of several tags.The tag RRQ extracted wh-general adverbs and the output included not only nominal wh-interrogative clauses, but also e.g.direct questions and adverbial clauses.Therefore, in order to extract nominal clauses that stand after an independent clause, the search was narrowed to VV* * RRQ and other tag modifications (Table 1).They were all placed in the advanced search option, as well as the output concordance lines were examined and sorted to check each instance for relevance.Presuming that the search options had enabled the extraction of all nominal interrogative clauses found in the texts, the results showed that students had used a small number of them.Although the majority of the clauses were not separated by a comma (as in Example 12), there were also cases where the students had separated these clauses by a comma, as in Example 13, where the author might have intended to insert a pause after 'wonder' for a dramatic effect.Obviously a dash would have been more relevant in that case.(12) I didn't know where they were.(13) One might wonder, what will happen.
It has to be noted that the number of these clauses extracted is too modest to serve for generalisations.
Although the PNQS tag extracted who in different position within a sentence, e.g. at the beginning of sentences, which were just a few cases in the students' texts, it allowed convenient extraction of relative who-clauses, which in their turn could be sorted into restrictive and non-restrictive clauses to check the relevance of comma use.Relative who-restrictive/embedded clauses that are normally used without a comma can cause punctuation challenges, as who is used to introduce both restrictive /embedded and non-restrictive/non-embedded clauses.Indeed, who relative restrictive clauses turned out to be challenging.Apart from numerous relevantly punctuated instances (Examples 14), there were also cases of the irrelevant insertion of commas (Examples 15).( 14) Of course there are people who mostly prefer walking.
There are a lot of people who are driving to work by car.
People who do not have their own car.(15) There is anybody, who would love.
I feel it more than people, who live in a city.( 16) Now factories who make cars think how to make their cars friendly to nature.
In addition, as can be seen in Example 16, there are clauses in which the use of who is doubtful, as who wrongly refers to factories, though it normally relates to humans.It is also noteworthy that cases of non-restrictive who relative clauses were not found in students' texts, which means that they did not display varied approaches to sentence structure to reach the communicative purpose of the text.
The PNQ0 tag was employed to check if the cases of nominal and relative whom clauses were found in the texts.It was determined that these clauses had not been used by the students. | 2019-01-02T14:51:55.029Z | 2016-12-07T00:00:00.000 | {
"year": 2016,
"sha1": "03f3f7efa8e8dd72d85f7fb65b03b44b87495775",
"oa_license": "CCBYSA",
"oa_url": "http://www.zmogusirzodis.leu.lt/index.php/zmogusirzodis/article/download/151/145",
"oa_status": "GOLD",
"pdf_src": "Anansi",
"pdf_hash": "03f3f7efa8e8dd72d85f7fb65b03b44b87495775",
"s2fieldsofstudy": [
"Linguistics"
],
"extfieldsofstudy": [
"Sociology"
]
} |
247587536 | pes2o/s2orc | v3-fos-license | Relationship between Professional Training of Dentists and Outpatient Clinical Production
Background The aim was to evaluate the association between the professional training of dentists and their outpatient production (OP) of clinical and collective/preventive procedures and the total number of procedures registered in a health information system. Methods It included all 19,947 primary dental care units participating in the Program for Improvement of Access and Quality of Primary Care (PMAQ-AB 2nd cycle) and the number of clinical procedures (CP), collective/preventive procedures (PP), and total procedures (TP) registered in the ambulatory information system between November 2013 and July 2014 for each participant oral health team. The outcome was being above the national median of procedures. The main variables related to training were the dentists specialising in family health, the level of training, and participation in permanent education. Effect estimates were calculated by multiple logistic regression. Results In the final model, controlled by contextual factor work process, family health specialists had higher chances (odds ratio (OR) = 1.13, 95% CI: 1.00; 1.27) of producing above the national median of CP than nonspecialists, OR = 1.06 (0.96; 1.18) for PP and OR = 1.17 (1.08; 1.27) for TP. Dentists taking permanent education had higher chances than those not taking it of producing above the national median for CP, PP, and TT, respectively, with OR = 1.40 (1.20; 1.62), OR = 1.24 (1.09; 1.40), and OR = 1.28 (1.18; 1.39). Conclusion Training in family health performs more procedures in primary care settings than those without training. However, this OP is influenced by variables related to the municipality and the work process, especially for PP. If the highest production observed is a consequence of training, then public health managers can not only encourage training policies such as permanent education policies to expand the use of services.
Introduction
Healthcare assessment is needed to evaluate if health policy has achieved its goals of transforming a situation towards a more desirable one [1]. The outpatient production of health teams, classified as process evaluation, is directly related to the healthcare system performance and represents the use of services (realised access) [2,3]. Outpatient production of clinical procedures constitutes an essential way of evaluating healthcare services, representing its interaction with the user [4]. The determinants of use health services and production of clinical procedures have been named as follows: (a) the health needs of users and their characteristics, (b) form of organisation of services, (c) health system policy, and (d) characteristics of health service providers. Health profes-sionals define the type and demand of resources consumed to solve the health problems of patients [5].
In Brazil, the outpatient production of oral health teams (OHT) is processed by municipal and state managers through the Outpatient Information System of the Unified Health System (in Portuguese, Sistema de Informação Ambulatorial do Sistema Único de Saúde, acronym SIA/ SUS) [6]. That production can be influenced by the infrastructure of healthcare facilities [7], by its oral healthcare network and care model [8][9][10], by the working process of the teams [11], through self-assessment [12], or by the training of dentists [13]. Araújo et al. highlight that professionals trained in the Brazilian Unified Health System (SUS) tend to know it better. Investments in training are of paramount importance for changes in the healthcare model, specifically in Primary Health Care (aka Family Health Strategy in Brazil, acronym FHS) [14].
The Brazilian National Health System has a legal responsibility in managing the training of healthcare personnel [15]. Despite that, no study has yet investigated the relationship between the training offered to dentists and their outpatient production. Recently, a study described the association between the variety of procedures according to different dentists' profile [16], but another study showed that the production of SUS is different according to geographical regions [17]. Due to local differences in the context of each catchment area of the OHTs, it would be expected that dentists with residence/specialisation in family health or public health may have a greater and more diversified number of procedures. Those professionals were trained to know the system [14] and also to include dental assistants as part of the team, different from the traditional biomedical model where dentist usually works alone. Dental assistants are known to have clinical competencies to increase productivity [18,19].
Therefore, this study was aimed at evaluating the association between the professional training of dentists and their outpatient production of clinical and collective/preventive procedures and the total number of procedures registered in a health information system.
Methods
This study included data from a census of all primary dental care units in Brazil during 2013 and 2014. Approximately 24,533 dental care units across the country and 19,946 (81.3% response rate) adhered to the 2 nd cycle of the Program for Improvement of Access and Quality of Primary Health Care (in Portugues called Programa de Melhoria do Acesso e Qualitade da Atenção Básica, acronym PMAQ-AB). Among the participant units, 17,259 OHT dentists comprised the current sample. All data is available at https:// aps.saude.gov.br/ape/pmaq/ciclo2/. Data from PMAQ-AB 2nd cycle was linked with data from the outpatient system (SIA-SUS) at primary dental care units' level from November 2013 to July 2014. All information is freely available at https:// datasus.saude.gov.br/.
2.1. Outcome Variables. The three outcome variables were the number of clinical and collective/preventive procedures and total procedures approved in SIA-SUS (see supplemental file for codes grouped into each category of procedures (available here)). Total procedures include clinical, preventive, and other procedures, such as home visits, nonplanned consultations (e.g., urgency), health surveillance, and nondental procedures (e.g., hypertension measurements taken by dentists). The statistical program R was used for data extraction and processing using the "microdatasus" and "tidyverse," respectively [20,21]. The "month_end" argument was extended to 3 months later due to its correspondence to the time interval in which the municipality can still make changes in the system [6]. The production was filtered by the dentist occupation codes, according to the Brazilian Classification of Occupations -CBO2002 [22]. The grouping was carried out by the "CNES" variable, the unique primary care unit identifier. In addition, the production of units that had more than one dentist (9.1%) was divided by the corresponding number of professionals. Original variables were counts of procedures at unit level; then, they were dichotomised in the median for further statistical analysis.
Main Predictor
Variables. The three main variables related to professional training were obtained from Module VI of the PMAQ-AB 2nd cycle. Two variables were created from the question "VI_3_2 -Which of these training process certifications do you possess?". The first corresponds to training with an emphasis on family health, being classified as follows: (a) "yes, in family health"-including professionals who answered "completed" to at least one of the questions referring to having specialisation, residence, master, or doctorate in family health/public health/collective health; (b) "yes"-for those who answered "concluded" for at least one of the variables of having a postgraduate degree different from family health; and (c) "no"-includes other professionals without a postgraduate degree. The second variable separated professionals who answered "completed" only for "Lato Sensu" or "Stricto Sensu," for both "Lato and Stricto Sensu" and "none." In the third variable, comprising the participation of permanent education, we considered the question "VI_6_1_11 -Does not participate in any form of permanent education".
Covariates.
In our study, we considered as potential confounding factors the following covariates: geographical region, city size, and percentage of residents in urban areas. All covariates were obtained from the 2010 census. The percentage of coverage of the FHS and OHT was obtained from the raw database of the E-Manager (E-Gestor). The variables related to the dentists working at OHT were obtained from the PMAQ-AB database: "VI_2_4 -How long have you worked in this primary care team?"; "VI_5_5 -Do you have an incentive, bonus, financial reward for performance?"; "VI_10_1_1 -Yes, from the DSC (Dental Specialties Centers)."; "VI_10_1_6 -No (Concerning Institutional Support)"; and "VI_7_5_2 -Which instrument/source was used? (Referring to Self-Assessment)". The inverse result of the variable "VI_10_1_6" was used to facilitate interpretation. The question "VI_7_5_2" was grouped with the answers "AMAQ" and "AMQ," other self-assessments, and self-assessment not included. Still, the question VI_2_4 allowed categorisation of professionals in "0-2 years," "3-6 years," and "more than 6 years" of work at the OHT.
In addition to those variables, the number of dental Xray, dental equipment, light curing unit, and dental ultrasound was collected from Module V of the PMAQ-AB, and a continuous variable was created with the sum of the equipment.
We theoretically considered as potential mediators of the relationship between the outcome and the main variables the question "VI_6_7 -How often does the unit host students, professors and/or researchers in teaching, research and/or community extension activities?". If the answer had weekly frequency, we classified it as "yes" and the rest as "no."
BioMed Research International
Moreover, variables related to the organisation of the dentists' schedule, "VI_13_1 -The team's clinical care schedule ensures" and "VI_13_6 -How are dental appointments scheduled at the Health Unit?", were grouped into "fixed days," if presented a specific day for scheduling. The hypothesis of these variables being mediators stems from the assumption that professionals with FH training would be more likely to develop teaching, research, and extension in partnerships with educational institutions, intending structural changes in their health unit towards a scheduling format and guaranteed access to SUS.
2.4. Statistical Analysis. Variables were described using mean and standard deviation for continuous variables and absolute and relative frequency for categorical variables. Bivariate analyses of the outcome with the other covariates were performed using the Kruskal-Wallis test with a significance level of 5%.
Effect estimates were calculated by multiple regression using the binomial model with a logit link function (logistic model), with the outcome being dichotomised by its median. The option to dichotomise was due to the asymmetric nature of the outcome data with several influential outliers and substantial standard deviations, as observed in a preliminary analysis.
Three models were tested in the regression analysis. The crude model includes only the main variables. Model 1 includes each main variable independently and the confounding variables described previously. Model 2 includes variables in model 1 (main predictors not adjusted by each other) and those that theoretically could mediate the effect of main variables on outpatient production. Finally, to test if the associations in each main variable differed by macroregion, we included interaction terms in regression models. Data were analysed using the R version 4.0.3 statistical software.
Results
It was observed that of the 19,947 participants in the external evaluation of the PMAQ AB 2nd cycle, 17,259 (86.4%) were OHT dentists that composed the analytical sample. All 17,259 units were units registered in the CNES database of healthcare facilities, and 81.4% of them had only one dentist, showing that in most observations, the unit's outpatient production refers to the dentist who answered to the interviewed professional at PMAQ-AB and is not influenced by the work of another professional.
According to Table 1, 57.1% of the interviewed dentists had additional training, mostly at graduation level. Specialisation in family health represents 29.5% of the sample, and those professionals produce, on average, a greater number of collective/preventive procedures and total production. Also, dentists who participate in permanent education have a higher average of procedures in all outcome variables. The mean of the three outcome variables was associated with all main variables.
Tables 2, 3, and 4 present the results of the adjusted models for each outcome variable. In the adjusted model 1, we observed that, after controlling for factors at the city level where the dentist works and the OTH work process, professionals with training in family health are the only ones who maintain a significant difference concerning 3 BioMed Research International the measurement of clinical procedures (p = 0:04) and the total number of procedures (p < 0:01). In addition, we emphasise that professionals who participate in permanent education produce, on average, a greater number of clini-cal and collective/preventive procedures and total procedures, both with p < 0:01.
The adjusted model 2 is controlled by variables in model 1 and potential mediators. Model 2 had a higher BIC than Table 2: Crude and adjusted odds ratio (OR) from logistic regression models (95% confidence interval) for the association of being above the median outpatient production of clinical procedures by the oral health teams participating in the second cycle of PMAQ-AB. Note: variables in the first column are not adjusted by one another. BIC: Bayesian information criteria. Model 1-adjusted by potential confounding factor: macroregion + population size + urban area (%) + FHS and OHT coverage (%) + team time + performance incentive + institutional support + DSC support + self-assessment. Model 2-adjusted by model 1 and mediation factors: receiving students + guaranteed scheduling + form of scheduling appointments.
4 BioMed Research International model 1, showing that it has a worse fit. The magnitude of main associations remained unchanged, meaning that those variables may contain substantial measurement error or they may not be mediators. The interaction results by macroregions (effect modification of "area of postgraduate course") were not statistically significant at a level of 5% and varied around OR = 1, sometimes in an uninterpretable way.
Discussion
The study shows that the outpatient production of specialists in public health or family health is larger than other specialists, but this result is partially influenced by the municipality context factors and work process. Regarding collective/preventive procedures, we found a smaller and statistically not significant difference in the adjusted model. However, for clinical procedures and total procedures, this speciality was the only one that presented a significant result. When explaining such a weak and nonsignificant association, one should consider that collective procedures include dental health education activities (see supplemental file) conducted by dental hygienists at a school environment [23]. Although more collective procedures have been associated with higher coverage of FHS [10], such increase is likely due to the presence of dental auxiliaries. Therefore, collective procedures may not be affected by training dentists who work mostly in clinical setting. Still, professionals who take part in permanent education perform more procedures. The present study has limitations and strengths. Outpatient production can be under-or overreported in informa-tion systems [8][9][10]. Nonetheless, no study has evaluated the reliability and validity of SIA-SUS, unlike other health information systems [24]. Barros and Chaves suggest that traditional dental procedures, such as the number of tooth extractions and restorations, show higher validity and reliability in their records [25]. A relevant limitation is that motivated professionals, who are predisposed to greater production, may also carry out training courses and permanent education, resulting in bias. As for strengths, we highlight the absence of selection bias, as the analyses used a census of health teams. Additionally, the information in the PMAQ-AB database was obtained in locus, minimising measurement errors.
The influence of participation in continuing education activities shows the relevance of this type of initiative, where the clinical topics in those education activities are present in the daily practice. Some studies [26,27] highlight that there is still conceptual confusion among health professionals and managers between the terms permanent education and continuing education, leading to a different interpretation of the results. In our study, the variable is a compilation of a group of questions that described only continuing education activities. Maciel et al. [26] reported the lack of studies that analyse permanent education in health linked to the work process of dental surgeons who work in the Family Health Strategy. We identified a couple of studies investigating the association between dentists' profile and a validated roster of 23 self-reported clinical procedures [16,28] included in the current study. They found that continuing professional training was associated with a larger variety of procedures but were not able to define the amount of each one. Our Table 4: Crude and adjusted odds ratio (OR) from logistic regression models (95% confidence interval) for the association of being above the median outpatient production of total number of procedures of the oral health teams participating in the second cycle of PMAQ-AB. Our data show that professionals with a Lato Sensu degree have greater production, thus expanding access to populations with large unmet needs. In Brazil, postgraduate courses have advanced the qualification of primary care professionals. By mid-June 2020, the Open University of SUS (Universidade Aberta do SUS, UNA-SUS) had already trained more than 90,000 professionals in specialisation courses, about 77,000 of them in family health [29], which means a big step forward in a possible change in the healthcare model. However, at the undergraduate level, the growing and unrestrained increase in new dental schools, mainly in the private sector, has contributed to increasing regional asymmetries in the supply of dentists and fostering an educational logic focused on the private market [30,31].
Conclusions
In conclusion, it was shown that dentists participating in the PMAQ-AB 2nd cycle with training in public health or family health performed more clinical procedures. However, other factors are essential in addition to professional training, especially in producing collective/preventive procedures. Future studies can confirm the current findings and investigate the processes that apparently lead professionals with broad-based education to have greater production. If the higher production observed is a consequence of training, then public health managers can encourage training policies and permanent education policies as a way to expand the use of services.
Data Availability
All information used in the current manuscript is freely available at Health Information Systems. PMAQ-AB microdata is available at the website of the Ministry of Health of Brazil: https://aps.saude.gov.br/ape/pmaq. | 2022-03-22T15:08:11.113Z | 2022-03-20T00:00:00.000 | {
"year": 2022,
"sha1": "56546cb4b5ea268b7c64adfd44fceae84473baae",
"oa_license": "CCBY",
"oa_url": "https://downloads.hindawi.com/journals/bmri/2022/5365363.pdf",
"oa_status": "GOLD",
"pdf_src": "Anansi",
"pdf_hash": "231c5b23381bafcbfdb605e91470f287223d1f9c",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
258917352 | pes2o/s2orc | v3-fos-license | allo-Predictors of breastfeeding duration on mothers who return to work: a systematic review
Introduction: the idea of continuing breastfeeding and work raises conflicting demands on the mother. Work correlates with early termination of breastfeeding and full-time work is associated with a shorter duration of breastfeeding than working part time. Objective: we aimed to systematically explore literature relating to factors related to breastfeeding duration in mothers returning to work. Materials and Methods: this research uses systematic review where systematic review is carried out following a combination of selected reporting items for systematic review and meta-analysis. Results: many selected articles were found related to factors affecting the duration of breastfeeding in breastfeeding mothers who returned to work. Discussions: delaying return to work until at least 15 weeks postpartum indicates a long duration of breastfeeding (at least 6 months). Part-time work is also positively related to the duration of feeding. Conclusions : employment and early return to work, support from social environment, perception of milk production, higher maternal occupational and education, household income, delaying return to work at least 15 weeks postpartum, working part time and longer maternity leave are predictors of breastfeeding duration on mother return to work.
Introduction
Breastfeeding is the foundation of child survival, nutrition and development as well as maternal health. The World Health Organization (WHO) recommends exclusive breastfeeding for the first 6 months of life, followed by continuing breastfeeding with appropriate complementary foods for up to 2 years or more. 1 According to the central statistics agency, the exclusive breastfeeding coverage in Indonesia in 2020 was 69.92%, while in central Java Province it was 76.30% (Central Statistics Agency, 2020). Nationally, based on Basic Health Research in 2018, it shows that the proportion of exclusive breastfeeding in infants aged 0-5 months is still low, namely 37.3%, partial breastfeeding 9.3% and predominant breastfeeding 3.3%. 2 The mother's work is cited as a major potential obstacle in exclusive breastfeeding and is a factor that can reduce the duration of breastfeeding. 3 The time the mother returns to work has an impact on the duration of breastfeeding. 4 In studies related to the duration of breastfeeding with the length of maternity leave, with mothers returning to work early for economic reasons more likely to give formula milk to their babies at the age of 6 months. 5 The mother's employment status has a significant negative relationship to the success of the mother in providing exclusive breastfeeding. These results suggest that working mothers increase the frequency of failure of exclusive breastfeeding. These results are no different from some studies on exclusive breastfeeding in different countries. Working mothers will face several obstacles in giving exclusive breastfeeding to their babies, including: time allo- cation, quality of togetherness with the baby, workload, stress, and the mother's confidence to give exclusive breastfeeding will be affected. Working mothers have low confidence in being able to give exclusive breastfeeding. 6 For many women overpaid work is a necessity and mothers face the challenges of work as well as breastfeeding. Work related characteristics affect breastfeeding continuation among working women such as how quickly women returning to work and whether they work full time or part time. [7][8][9] The idea of continuing breastfeeding and work raises conflicting demands on the mother. Work correlated with early termination of breastfeeding and full time work was associated with a shorter duration of breastfeeding than working part time. [10][11][12] Research looking at maternity leave and breastfeeding duration, shows that maternity leave is positively related to breastfeeding duration, especially for mothers with less flexible work, and longer leave is associated with prolonged breastfeeding 10 .
The socio-demographic characteristics of the mother are associated with the duration of exclusive feeding. The study showed that older, married, multiparous mothers, of high social class or non-smokers, would breastfeed longer. Occupational characteristics of the mother can also affect the duration of feeding. In the early 2000s, research in the UK and the Netherlands showed that mothers' jobs, in particular full-time jobs, were negatively related to breastfeeding. The characteristics of the mother's occupation, in particular the time the mother returns to work and the status of full/part-time work, relate to the duration of breastfeeding exclusively 8 .
The aim of this paper is to systematically explore literature relating to factors related to breastfeeding duration in mothers returning to work.
Materials and Methods
A systematic review was conducted following a mix of preferred reporting items for systematic review and meta-analysis (PRISMA). The databases ScienceDirect, SpringerLink, CINAHL, ProQuest, SAGE journals, and PubMed) were searched using the keywords breastfeeding duration, exclusivity, returning to work, statistical results and data, with no restrictions on the year of publication. The reference list of the identified papers was also examined with a focus on predictors affecting breastfeeding duration, breastfeeding practices and reasons for stopping breastfeeding after the mother returns to work back to work.
Most of the articles identified in the search specifically about the duration of breastfeeding in nursing mothers returning to work. A flowchart was developed in accordance with PRISMA guidelines for summarizing articles obtained in a literature search. PRIS-MA is used to illustrate the references that have been found, the number of exceptions and the criteria and the number of eventual inclusion in the full review. The PRISMA flow diagram of the search strategy is presented in Figure 1. All papers on the duration of breastfeeding in returning to work nursing mothers are considered for inclusion. In the results, the literature has been classified according to: the duration of breastfeeding, the practice of breastfeeding and the reasons why women after returning to work the duration of breastfeeding becomes short. In addition, the results and other implications of disclosure are also included. In each section, the identified types and qualities of the papers are described, and the papers are summarized and presented in Table 1.
Results
Based on the results of article searches, 6 selected articles were obtained related to factors that affect the duration of breastfeeding in breastfeeding mothers who return to work. The article is summarized in the form of columns whose names are researchers and journals, research titles, research methods, locations, participants and research results. Then the author conducts an analysis of the summarized article. A summary of the research on factors affecting breastfeeding duration in returning to work breastfeeding mothers is illustrated in Table 1.
Results from a systematic review showed that work was negatively related to the mother's ability to breastfeed over a long period of time. Those who do not work are more likely to breastfeed longer. In mothers who work part-time, self-employed mothers breastfeed longer than mothers who work as employees.
Discussion
Regarding leave during labor, the results of this study suggest that delaying returning to work can facilitate longer breastfeeding. The variable which has previously shown a significant association with the duration of breastfeeding i.e. maternal education is a strong predictor of continuing to breastfeed after returning to work. [18][19][20][21] Maternal education explains higher opportunities for initiation and duration of breastfeeding. 8,22,23 Mothers with high educated may have more controls over their work and schedules, which means they can relax for breastfeeding friendly work environment.
Duration of feeding is also affected by support. 24 From various literature shows that workplace support plays a role in the duration of breastfeeding after returning to work. 2,25,26 Organizational as well as managerial support increases the duration of exclusive breastfeeding by almost double. Co-worker support is critical in 27,28 There is a positive relationship between longer leave and duration of breastfeeding. 29 Mothers who aim to breastfeed longer differ in socio-economic and educational background as well as selfconfidence and aspirations of mothers. 30 Mothers may also have different jobs, with different working conditions and flexibility of work, and different patterns of returning to work and taking maternity leave. The mother's work is the main reason for the short duration of breastfeeding and even the cessation of breastfeeding. Studies from Ghana, Kenya, Brazil, Ecuador, and the Democratic Republic of the Congo concluded that the short duration of maternity leave correlates with the short duration of breastfeeding. [31][32][33][34][35][36] Duration of maternity leave (<3 months) is associated with a low duration of breastfeeding at 2 and 3 months compared to leave of 3 months or more. 37 The mother's control over time and space at work also contributes to the duration of feeding. Workplaces can provide more flexible work options. Flexibility of work schedule, division of work, location of work contributes to improved work-family balance so as to increase the duration of breastfeeding. [38][39][40][41]
Conclusions
Predictors of breastfeeding duration on mother who return to work are early return to work and employment, support from the social environment, the perception of milk production, higher maternal occupational and education, household income, delaying return to work at least 15 weeks postpartum, working part time and longer maternity leave. | 2023-05-27T15:18:21.200Z | 2023-05-25T00:00:00.000 | {
"year": 2023,
"sha1": "902bb0ddca4d171c4a30b4be4d1447f73614509d",
"oa_license": "CCBYNC",
"oa_url": "https://www.publichealthinafrica.org/jphia/article/download/2569/887",
"oa_status": "GOLD",
"pdf_src": "Anansi",
"pdf_hash": "4aacc552a7c7892b3875acc9a3d37f187ca38de1",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": []
} |
250692935 | pes2o/s2orc | v3-fos-license | Growth, structure and phase transitions of epitaxial nanowires of III-V semiconductors
We review and illustrate the impact of TEM on the study of nanowires of non-nitride III-V semiconductors, with particular emphasis on the understanding of the thermodynamics and kinetics of their formation assisted by nano-sized catalyst particles. Besides providing basic information about the morphology of the nanowires and their growth rate as a function of diameter, TEM offers insights into the peculiar crystalline structure that they adopt. We discuss the formation of the unusual wurtzite hexagonal crystalline phase and that of planar stacking defects in these nanowires and show that they are kinetically controlled. We also demonstrate the transformation of wurtzite into cubic sphalerite upon epitaxial burying of the nanowires. Nanowires are particularly interesting in that they allow the fabrication of precisely positioned quantum dots with well-defined geometries. In this respect, we discuss the formation of strained quantum-size inclusions in nanowires, their critical dimensions and the kinetic and thermodynamic factors governing the changes of the crystalline structure that sometimes occur around a hetero-interface.
Introduction
Free-standing wires with diameters ranging from hundreds down to a few nanometers are nowadays commonly fabricated from a large range of semiconductor materials [1][2][3][4][5][6]. These 1D nanostructures have remarkable physical properties and many potential applications, yet the control of their growth and final characteristics is still far from optimal.
In this paper, we illustrate the impact of transmission electron microscopy (TEM) on the study of nanowires (NWs) of III-V semiconductors with particular emphasis on the understanding of the thermodynamics and kinetics of their formation. To be specific, we restrict the scope of our study in terms of materials and growth methods, by considering III-V compounds containing no nitrogen and grown by Molecular Beam Epitaxy (MBE) assisted by nano-sized gold "catalyst" particles. However, some of our conclusions remain valid for other material systems, growth methods and catalysts. We growth of the NWs competes with that of the catalyst-free portion of the sample located between the NWs (the "non-activated" surface), which always occurs, albeit at a slower rate. This (111)B surface is typically fairly rough and grows by a combination of 2D nucleation and step flow whereas the NWs grow via the nucleation of new solid atomic layers at the top interface with the catalyst droplet. The droplet is a mixture of metal (here Au) with the wire elements. In most experiments, it is liquid during growth, as required for proper VLS. We checked this in situ for InAs NWs [8]. In some cases, the catalyst might however remain solid [3]. In MBE, the droplet is not a chemical catalyst which would facilitate the decomposition of precursor molecules, as in chemical vapour deposition methods, since the type-III elements are delivered as atomic beams and the type-V molecular beams are thermally decomposed. Instead, the droplet increases the growth rate at the top of the NW, probably by decreasing the nucleation barrier and by providing a dense feeding medium.
Nanowire
The basic geometrical features of the NWs are not only important for the properties of the devices into which they may subsequently be incorporated. They also provide invaluable insights into the growth processes. To our knowledge, in situ TEM has only been performed for NWs of type-IV elements [9][10], whereas III-V NWs have only been examined ex situ. We shall see in section 3.2 that the way in which the growth is terminated before the NW is removed from the epitaxy chamber may marginally alter its geometry and structure, but in general the NWs remain largely unaffected by the transfer.
The first two dimensions to be determined are obviously the length and diameter of the NW. In the simplest case, each NW grows with a uniform diameter determined by the size of the interface between the top facet of the NW and the droplet, which itself depends on the droplet volume and on its contact angle with the top facet. Hence, when the Au nanodrops are formed by annealing, because of their size distribution, a range of NW diameters is obtained (ordered arrays of droplets of uniform sizes may however be deposited ex situ, for instance through a mask patterned by electron lithography). The length/diameter relationship is best determined by examining a large number of such NWs by scanning electron microscopy. It provides key information about the processes that limit the growth velocity of the NWs. For instance, the inverse relationship between length and diameter that is often observed indicates that growth is limited by the diffusion of the atoms to the NWs from the nonactivated areas situated between the NWs [11].
Our NW volume measurements prove that the amount of material incorporated in the NW is larger than the amount impinging on the droplet. This confirms that part of the material reaches the top of the NW via diffusion from the substrate. In modern MBE machines the beams are inclined to the substrate normal (by about 30°), so that the NW material is also partly collected from the NW sidewalls. Initially overlooked in NW growth modelling, this fact is increasingly recognized as important [12].
In addition, individual NWs may have non-uniform diameters. The regular overall NW tapering observed at low growth temperature [6] indicates that the diffusing species tend to nucleate on the sidewalls of the NW before reaching the droplet. The localized tapering observed at higher growth temperature at the top of long NWs attests in turn that these species desorb if the NW length exceeds their diffusion length [4,12]. We sometimes observe double tapering at top and bottom; we attribute the former to a shadowing of the foot of the NW from the molecular beams by the neighbouring NWs.
Catalyst
In VLS growth, the catalyst droplet plays a critical part. Most often, the gold catalyst is first deposited as a film of roughly uniform thickness and then annealed to induce the formation of nanodroplets. Gold is usually deposited in a chamber separate from the epitaxy machine. On the contrary, we use an Au source located in the MBE chamber; this avoids any contamination of the coated substrate and significantly speeds up the fabrication cycle [4].
We have observed the catalyst particles not only after growth but also immediately after annealing, before exposing the sample to the growth fluxes. All chemical analyses performed in the TEM reveal a mixture of gold and type-III element. Type-V atoms are not detected, for two main reasons. First, their solubility in gold is very low (a few atomic percents at most). Second, since their equilibrium pressures are rather high, they probably escape from the droplet when the latter is cooled after annealing or after NW growth. Most often, the solid droplets are truncated spheres (spherical caps), another strong indication that they have been cooled from the liquid state. Even after mere annealing, before being exposed to the molecular beams, the particles are not pure gold. The TEM lattice images show that around each particle a few atomic layers have been removed from the top of the substrate and the corresponding type-III atoms have been incorporated in the droplet (Fig. 1). In the case of a GaAs substrate, we have identified by transmission electron diffraction (TED) that the particles are Au 7 Ga 2 crystals. This gives a lower limit to the Ga-content of the annealed droplets, since some of the dissolved Ga might have been lost during cooling. This content is consistent with the common view that VLS occurs at rather low temperature (430-620°C for GaAs in our MBE system) thanks to the formation of a eutectic between the catalyst and the substrate element(s). In the present case, gold liquefies by forming a eutectic with gallium, which may happen below 400°C for Ga concentrations of about 30 %.
We also analyzed the catalyst particles after growth of GaAs NWs [4]. Their composition depends much on the growth history and in particular on how growth was terminated. When the Ga and As fluxes are stopped simultaneously, we observe the Au 7 Ga 2 or AuGa crystalline phases. This is consistent with composition measurements made by other authors [13]. Together with our calculation of a typical temperature drop of only a few kelvins between substrate and NW apex during growth [14], this strengthens the picture of a liquid catalyst. It is only when growth is terminated by stopping the Ga flux while maintaining the As flux during cooling that we observe almost pure gold. In this case, most of the Ga atoms of the droplet have been used to continue forming NW layers with the As atoms of the beam (see section 3.2). and (111) respectively. Dashes mark the top of the substrate before its local dissolution in the droplets.
Prevalence of the wurtzite phase in nanowires of cubic III-V compounds
Under bulk form, all N-free III-V semiconductors crystallize in the cubic sphalerite structure, also known as zinc blende (ZB). A surprising feature of the NWs of these materials is that they very often adopt the hexagonal wurtzite (WZ) structure. This has been observed for most ZB III-V materials and growth techniques [1,3,4,15,16].
ZB and WZ are closely related structures. Both are stacks along a given direction (e.g. [111]B in ZB) of identical biatomic planes, or monolayers (MLs), which differ by their transverse positions. Three such positions (A, B and C) are allowed. ZB corresponds to a periodically repeated ABC sequence, whereas WZ consists of a periodic AB stack. ZB and WZ can be identified either by TED or by high resolution TEM (HREM), as shown in the inserts of Fig. 2. In HREM images, the 3-ML periodicity of ZB is clearly distinguished from the 2-ML periodicity of WZ ( Fig. 2(b)).
More complicated periodic stacks are also encountered in NWs, for instance polytypes [16]. Indeed, any aperiodic sequence is allowed, provided it does not include twice the same letter (A, B or C) consecutively. A stacking defect is a local disruption of the ZB or WZ periodicity (stacking fault) or a boundary between differently oriented grains of the same phase. Although often dominantly of WZ structure, the NWs usually contain stacking faults and sequences of ZB structure ( Fig. 3(a)). Since the coexistence of two phases impedes basic studies as well as applications, phase purity control has become one of the main challenges of III-V NW fabrication. The prevalence of WZ (a high pressure phase observed in no other nanostructure) in NWs of III-V cubic materials is surprising. Until recently, its formation had not been explained satisfactorily. The difference in cohesive energy between ZB and WZ bulk III-V materials is typically of a few tens of meV per III-V pair (24 meV for GaAs). Several ab initio calculations have shown that the total energy of NWs can indeed be lower in the WZ structure than in the ZB structure because of a lower specific energy of some specific sidewall facets [17][18][19] or of their vertical edges [20]. However, for the contributions of these 2D or 1D features to compensate the difference in bulk cohesive energies, the NWs must be very thin: the calculated critical radii under which WZ might be favoured over ZB are only of a few nm, whereas WZ is observed in NWs up to radii of several 100 nm.
Moreover, the standard theories of NW formation are based on nucleation [21][22][23] and the occurrence of planar faults normal to the NW axis (see e.g. Figs. 2(c) and 3(a)) suggests that, once a nucleus of critical size forms at the solid/liquid interface, it rapidly spreads out laterally over the whole interface [5,16] unless the wire is very wide [23]. Hence, the reason for the formation of WZ must be searched for in the formation of WZ nuclei rather than in the total energies of fully constituted NWs. This is done in sections 3.2 and 3.3, where we consider the specific case of GaAs.
TEM observation of the systematic formation of zinc blende sections in wurtzite nanowires
Although WZ is largely prevalent in GaAs NWs, TEM demonstrates that ZB forms systematically during two particular stages of the growth. The first one corresponds to the very beginning of growth. By growing GaAs NWs for only a few seconds, we could observe the foot of each wire before its burying by the growth of the non-activated surface. The samples were thinned to obtain crosssectional views, with the NWs still epitaxially attached to their substrate. The images identify unambiguously ZB as the sole phase formed in this initial stage ( Fig. 2(a)). When the NW height reaches about 30 nm, growth switches abruptly to WZ stacking ( Fig. 2(b)).
The second case where ZB forms systematically is when we terminate MBE growth by switching off the Ga flux while maintaining an As flux during cooling. Over a few tens of nm, the final section of the NW is of cubic structure ( Fig. 2(c)). Persson et al. had already reported the formation of a short terminal ZB section in the case of chemical beam epitaxy, when growth was terminated under As only [3]. In both cases, growth continues in the absence of a fresh supply of Gallium by consuming the Ga atoms present in the drop, whereas As can only be provided by the beam (see section 2.2). Since the volume and the base area of the drop then decrease, the final section of the NW is tapered (Fig. 2(c)).
Our TEM experiments reveal that ZB forms systematically in two transient growth regimes during which the supersaturation in the liquid droplet (with respect to the solid) is less than during steady NW growth. Before growth, the deposited undersaturated Au droplets dissolve the substrate locally to achieve equilibrium with it ( Fig. 1), and hence zero supersaturation. When vapour fluxes are turned on, the supersaturation increases until a permanent regime settles. The opposite happens during growth termination: then, the Ga concentration in the droplet, and hence supersaturation, decrease, since the atoms used to build the final NW MLs are not replaced. This shows that ZB systematically forms when the supersaturation is less than some critical value and suggests that, conversely, WZ formation requires a high supersaturation.
Nucleation at the triple phase line and wurtzite formation: a TEM-supported explanation
Following arguments developed by Ostwald at the end of the 19th century [24], the basic idea is that the phase that forms is not necessarily the stable bulk phase (ZB) but the one having the lowest nucleation barrier φ G ∆ , because the main dependence of the nucleation probability at temperature T is in . The formation of WZ may thus be searched for in its lower nucleation barrier as compared with ZB. Standard nucleation theory [25] gives the energy barrier for the formation of a 2D where a and b are shape-dependent coefficients ( π = a and π is the difference of chemical potential between liquid and solid per unit volume of solid, φ γ SN is the interface energy between the top facet of the NW and the nucleus, and φ γ l is the specific energy of the lateral interface between the nucleus and its environment. We argued elsewhere that, whatever the crystalline phase, nucleation at the triple phase line (TPL), the border of the top facet of the NW (where solid, liquid and vapour coexist), is strongly favoured with respect to nucleation anywhere else on the top facet [26]. The reason is simply that, in the former case, part of the pre-existing liquid-vapour interface is destroyed and replaced by a nucleus-vapour interface. Then, show that Si crystals tend to nucleate at the edge of Au-Si droplets deposited on a membrane [27][28].
To compare the free energies of formation of ZB and WZ nuclei at the TPL, one may consider that, because of the difference in cohesive energies, φ µ ∆ is lower for WZ than for ZB. However, a 2D nucleus is not intrinsically of ZB or WZ structure. In accordance with the description of ZB and WZ as different stacking modes of otherwise identical MLs, it is only the lateral positioning of the nucleus that decides if the sequence formed with the underlying MLs is locally of WZ or of ZB type [26]. Thus, it is preferable to use the same Eq. (2) shows that there are two conditions for the critical nucleus in WZ position to have a lower free energy of formation than the nucleus in ZB position. The first one is material-related: one must have Careful consideration of the atomic arrangements for the ZB and WZ nuclei positions indicates that this difference is not likely to arise from differences at the nucleus-liquid interface [26], although this may also play a part [29]. On the other hand, the nucleus-vapour interfaces that form during nucleation at the TPL are clearly different and the corresponding energy may be lower for the WZ nucleus position [26], as also suggested by calculations showing that the specific energies of some extended facets are lower for WZ than for ZB [17][18][19]. From Eq. (2), there is however a second condition for WZ nucleation to be favoured, namely that the supersaturation be high enough to overcome the larger NW-nucleus interface energy: To summarize, the frequently observed formation of the WZ phase in NW of cubic III-V materials is due to the preferential formation of nuclei in WZ lateral position at the triple phase line. This preferential formation in turns results from a lower energy of the edge of the nucleus when it is in WZ position as compared with ZB position. This condition recalls the arguments based on total energy calculations for fully formed NWs [18,20]. However, at variance with the latter purely thermodynamic explanation, ours is intrinsically kinetic, in that it is based on barrier-height-dependent nucleation probabilities. This kinetic character is also manifested by our second condition for WZ formation, namely the existence of a lower critical value of the chemical potential difference between liquid and solid. Even with differences WZ ZB l l γ γ − as low as a few tens of mJ . m -2 , we find that differences µ ∆ of the order of a few hundreds of meV per III-V pair are sufficient to promote WZ nucleation. This is indeed the order of the differences of chemical potential between vapour and solid that we estimate for our MBE growth conditions, although the actual µ ∆ between liquid and solid remains unknown. The nucleation of each new ML has a probabilistic character. The formation of a stacking defect in WZ ( Fig. 3(a)) may thus be interpreted as arising from the chance nucleation of a single ML in ZB position, occurring because µ ∆ is not sufficient to insure a near certain nucleation in WZ position. On the other hand, WZ formation requires nucleation of each ML in WZ position. This opens prospects of single-phase growth, by setting high supersaturation conditions for pure WZ or, alternatively, low supersaturation for pure ZB. However, our efforts in this direction have not been successful yet. In particular, supersaturation cannot be lowered under a limit below which the NWs cease to grow. Nevertheless, several TEM-based results from other groups strengthen our arguments. Following the latter, Shtrikman et al. managed to obtain pure ZB GaAs NWs in MBE by deliberately attempting to achieve a low supersaturation [30]. Caroff et al. noticed that, in InAs NWs grown by MOVPE, the proportion of the WZ phase increases at low growth temperature [31]. We propose that this can be due to the increase of both the chemical potential difference µ ∆ and the relative difference of nucleation probabilities at low T. Conversely, if the material and supersaturation conditions are such that ZB formation prevails, lowering the temperature may again increase the difference of nucleation probabilities, this time in favour of ZB, and hence reduce the density of stacking defects, as observed by Johansson et al. [5] and Joyce et al. [32].
Phase transformation upon epitaxial burying
Despite some successes, it remains very difficult to fabricate reproducibly single-phase III-V NWs in a single growth operation. We have however devised a method to transform highly faulted hexagonal WZ GaAs NWs into perfect cubic ZB ones [33].
Experiments
The method is based on deliberately altering the relative growth rates of the NWs and of the nonactivated substrate by changing the growth temperature G T . Whereas in our standard growth conditions, for C 620°≤ G T , the NWs grow faster than the non-activated substrate (section 1), the reverse holds at C 620°≥ G T , so that any previously grown NW will get buried by the substrate. 3(a)). We first wrap them in a nm-thick protective Al x Ga 1-x As shell to prevent their disappearance at elevated temperature. Because it results from epitaxial nucleation on the sidewalls of the NWs, the shell adopts the latter's WZ structure. By increasing the growth temperature to 640 °C, we then shift to a regime where the supporting non-activated cubic GaAs substrate grows faster than the wires, effectively burying them. The whole sequence is carried out in the same MBE chamber. The samples were examined by TEM at various stages of the burying process. We first notice that the burying layer naturally adopts the cubic ZB structure of the GaAs substrate from which it grows. Examination of partially buried NWs (Fig. 3(b)) shows that the part emerging from the burying layer has not changed: it is still WZ with stacking defects. However, below the top level of the burying layer, the material is entirely of cubic structure. This means that the buried portion of any NW has changed to ZB. Hence, the growing non-activated substrate transforms the buried portion of the NW from WZ to ZB. This phase transformation is very fast, since the phase boundary is right at the level of the top ML of the burying layer ( Fig. 3(b)).
Further growth leads to complete burying. Each NW can still be localized via the faint contrast of its AlGaAs shell (Fig. 3(c)). TED and HREM images (Fig. 3(d)) confirm that the whole sample is pure ZB. The burying is very efficient, in that it eliminates all the stacking faults ( Fig. 3(c,d)). We thus end up with perfect cubic NWs embedded in a cubic matrix. Thanks to the AlGaAs shell, each NW retains its one-dimensional character.
The mechanism of the phase WZ → ZB phase transformation as revealed by HREM
Our TEM experiments on partially buried NWs show that, when the burying material reaches the periphery of a NW, it transforms the hexagonal structure of the latter into its own cubic one ( Fig. 3(b)). To understand how TEM may also reveal the microscopic mechanism of the phase transformation, we must consider the burying process in more detail. First recall that, along ZB or [0001] WZ direction, any type-III atom sits directly above the underlying type-V atom. The A, B, or C character of each III-V ML is therefore determined by the lateral positioning of its type-V atoms [26]. Let us assume that the burying layer has already transformed the lower part of the NW to its ZB stacking ( Fig. 4(a), stage α) and consider how the next sequence of six yet unburied WZ MLs may transform into ZB.
When the advancing top ML of the burying GaAs reaches the NW sidewall ( Fig. 4(a), α), the As atoms in the burying ML and in the corresponding ML of the NW are not in lateral registry. This fault in the NW (with respect to ZB) can be eliminated by a rigid shift of the emerging part of the NW (As plane included), which brings two more MLs in ZB position ( Fig. 4(a), β). When the burying layer reaches the level of the next fault, two MLs above the previous one ( Fig. 4(a), γ), the process repeats ( Fig. 4(a), δ), and so on every other ML. Three shifts suffice to transform 6 MLs ( Fig. 4(a), ε,ϕ).
The lateral contact between the top burying ML and the out-of-registry NW ML produces a Shockley partial dislocation (albeit only partly surrounded by material). Since our NWs have vertical sidewalls parallel to the six { } (or the complementary set, depending on the ZB variant). As the dislocation glides along the (111)B plane into the NW, the As atoms shift into ZB position. This translation of the whole As layer fully eliminates dislocation and stacking fault [33].
A type II transformation may also be imagined, whereby only one or two MLs of the NW are translated at each step, all other MLs retaining their lateral positions (Fig. 4(b)). The same partial dislocation is initially generated (Fig. 4(b), α) but the first step of the process is now the translation of the sole NW ML that it borders (Fig. 4(b), β). This eliminates the fault in this ML but produces a characteristic twin-like ABCBA sequence [34]. The next two NW MLs cannot be translated before the third one, which occurs when the burying material reaches its level (Fig. 4(b), γ), again by translation of a single ML, producing a "double-twin" BCBAC sequence (Fig. 4(b), δ). Finally, the two already Fig. 4(a), ε,ϕ), albeit only if they are not yet entirely surrounded by the burying layer or if they can exchange atoms. The latter reservations make the type II process less likely than the type I. These two types of transformation of WZ into ZB can be distinguished by HRTEM via differences in stacking sequences at the interface between the untransformed and transformed parts of the NW. We first identify the topmost NW MLs that have already adopted the ZB stacking ( Fig. 4(c)). At any stage of a type I process ( Fig. 4(a)), the next NW MLs repeat the last two MLs of the ZB sequence (e.g. ABCABC/BCBC... at stage δ) whereas during a type II process ( Fig. 4(b)), the same sequence is found only at stage α or ϕ (ABCAB/ABAB...), other well-defined twin-like sequences occurring at stages β to ε. In the many HRTEM images examined, we found only the stacking sequence characteristic of the type I process (Figure 4(c)) and no evidence of the other sequences occurring in a type II process. This indicates that, at least for GaAs and in our burying conditions, type I is the dominant transformation mechanism, if not the only one.
Interpretation
The driving force for the type I phase transformation is the joint elimination of the partial dislocation that forms at the NW/burying material boundary and change of structure of the NW from WZ to ZB. The length of the partial dislocation segments formed between burying layer and NW and the area of the faulted interface between the transformed and untransformed parts of the NW determine the system energy. Fig. 4(d) describes schematically how these change during the transformation of each ML. As overgrowth proceeds, the NW sidewalls get surrounded by the top burying ML extending by step flow and the partial dislocation is generated at the periphery of the corresponding hexagonalshape NW ML. As soon as the dislocation forms along at least two sides of the hexagon, any of its parts can glide inward without any length increase and therefore, without any energy increase. The area of the NW ML swept by the dislocation transforms to ZB stacking with respect to the underlying ZB portion of the NW. We calculate that eliminating the fault corresponds to a decrease of energy of about 28 mJ·m -2 [34]. Hence, the total energy of the defects decreases continuously during each elementary step of the transformation (Fig. 4). Of course, when a NW ML is already in the correct ZB position, no dislocation is created and hence no driving force for translating the upper part of the NW. This happens in particular when there is a stacking defect or a ZB segment in the yet unburied WZ NW and explains why all such pre-existing defects ( Fig. 3(a)) get eliminated (Fig. 3(c)). According to this simple analysis, there is no energy barrier to the transformation, in agreement with the fact that the translation of the whole ML occurs very soon after the burying layer reaches it, which leads to abrupt horizontal interfaces between the transformed and untransformed parts of the NW (Figs. 3(b) and 4(c)). In addition, there is no critical NW radius for the transformation, which explains why the structure of any NW shifts to ZB. The NW geometry makes the WZ → ZB transformation much easier than in bulk samples where, even for the type I transform, there is a barrier corresponding to the nucleation of a transformed domain of finite critical size [35]. Here, the dislocation is formed when the burying layer reaches the NW, and nothing hinders the translation of the top part of the NW. Since three translation (Burgers) vectors adding to zero are allowed, the net tilt of each NW remains nearly null (Fig. 3(c)). The former analysis however neglects the Peierls barrier impeding the glide of the partial dislocation, which might kinetically limit the transformation. This might be responsible for the incomplete transformation that we tend to observe in InP NWs.
Thermodynamics and kinetics of heterostructure formation in NWs
Nanowires are particularly interesting in that they allow the fabrication of precisely positioned quantum dots with well-defined geometries. In this section, we consider the formation of heterostructures in NWs as regards strain relaxation and crystalline structure.
Critical dimensions for mismatched axial heterostructures in nanowires
A major advantage of NWs appears when one compares the growth of a given misfitting material on a bulk substrate and on an already grown NW base. In the latter case, the heterostructure is termed axial to distinguish it from radial (core-shell) heterostructures. For a given substrate/layer misfit, the critical thickness for plastic relaxation is always larger in the NW case because the sidewalls allow the strains efficiently to relax [36], all the more if the NW diameter is small. Moreover, for each material couple, there exists a critical NW radius below which the critical thickness becomes infinite, which allows the growth of arbitrarily thick layers. We derived simple formulas for the radius-dependent critical thickness and for the critical radius [36]. The former are consistent with our own TEM experiments (e.g. the absence of dislocation in the long insertions of Fig. 5) and with the limited amount of experimental data available in the literature.
Fault suppression and structural phase changes at an axial hetero-interface
In VLS growth, sharp axial heterostructures are invariably fabricated by switching the type-V element (or modifying the fluxes of several type-V elements) because the type-III element, present in high concentration in the droplet (see section 2.2), tends to be incorporated even after its supply from the vapour phase is interrupted (see section 3.2). Changing the fluxes changes the thermodynamic growth conditions and may have profound effects not only on the composition of the grown material but also on its structure. We indeed frequently observe structural changes at hetero-interfaces.
A first example is given by a double InP/InP 1-x As x /InP axial heterostructure ( Fig. 5(a)). A thick InP shell was later grown around the thin NW core. In the core, InP and InP 1-x As x (x ~ 0.35) both adopt the WZ structure. However, whereas InP is highly faulted, the InPAs insertion is free from any extended defect. Note that the stacking defects created during growth of the InP core propagate laterally in the WZ shell whereas the InP shell around the InPAs insertion is defect-free. The reason is that the shell material nucleates on the NW sidewalls, at variance with the burying material considered in section 4. In light of the discussion of section 3.3, the high density of faults in the WZ InP core indicates that the difference of chemical potential µ ∆ is high enough to favour WZ but not sufficient to insure that each ML will nucleate in WZ position. On the contrary, the absence of any fault even in long InPAs insertions (Fig. 5(b)) indicates that, in this case, the probability for nucleation in ZB position is negligible. Since the insertion is grown by adding an As flux to an unchanged P flux, it is tempting to attribute this effect to an increased µ ∆ . However, the deposited solid changes, and even though the liquid is still mainly Au-In, its composition might also change (as happens in chemical beam epitaxy [37]), so that the relation between the fluxes and µ ∆ is far from straightforward. Moreover, the contribution of the type-V elements to the chemical potential in the liquid is unknown. In addition, over and above the chemical potential condition, the balance between WZ and ZB nucleation also depends on the edge energy φ γ l of the nucleus, as discussed in section 3.3. When nucleation occurs at the TPL, the edge is partly in contact with the vapour and the surface energies (and reconstructions) of III-V materials are known to depend on the vapour fluxes [38]. The changes of chemical potential and edge energy might thus both play a part in the absence of fault in the alloy segment.
The formation of the heterostructure may have an even more striking effect: when a GaAs 1-x Sb x alloy is grown axially above a GaAs WZ NW base, it adopts the ZB structure [39]. In this case, we observed by TEM that, in the GaAs cap above the insertion, the transition to (faulted) WZ is not abrupt but involves a section of the 4H polytype. Since in the latter, half the MLs nucleate in ZB position and the other half in WZ position, this phase can be considered as intermediate between ZB and WZ [34]. It seems more likely that such a delayed ZB → WZ transition is due to a gradual return of the chemical potential in the droplet to its pre-insertion value after the Sb flux is stopped, rather than to a surface effect which probably follows the changes of fluxes more closely in time. Figure 5. (a) TEM bright field image of an InP 1-x As x axial insertion about 100 nm high in an InP NW. The NW core diameter during axial growth (about 12 nm) is indicated by dashed lines and the hetero-interfaces by horizontal black arrows. The whole core was subsequently wrapped in a thick InP shell (double-ended white arrows). The contrast along the vertical edges of the insertion is due to strain relaxation. (b) A long fault-free InP 1-x As x insertion in a sample of similar structure. | 2022-06-28T03:22:44.712Z | 2010-01-01T00:00:00.000 | {
"year": 2010,
"sha1": "31637de07cffd4c58587fbaf8d4caf003d22f6d4",
"oa_license": null,
"oa_url": "https://doi.org/10.1088/1742-6596/209/1/012002",
"oa_status": "GOLD",
"pdf_src": "IOP",
"pdf_hash": "31637de07cffd4c58587fbaf8d4caf003d22f6d4",
"s2fieldsofstudy": [
"Materials Science",
"Chemistry"
],
"extfieldsofstudy": [
"Physics"
]
} |
212527896 | pes2o/s2orc | v3-fos-license | Appraising emergency contraceptives knowledge and use among female youth corp members in Nigeria;
Emergency contraception (EC) works by utilizing hormones to inhibit or delay ovulation1–3 or by utilizing Copper-Intrauterine device/CU-IUD to prevent fertilization/implantation.2 In this regard, EC is the option left for a woman who got involved in an unprotected sex including forced sex, or failed in her regular birth control method and does not want to become pregnant.4,5 EC will forestall unintended pregnancies, which in the United State constitute about half of all pregnancies,1,6 will help prevent about 50% of abortion cases also in the United State1 and will reduce maternal deaths that might accrue in the developing world like Nigeria, from induced abortions most of which are carried out by untrained hands.5,7–9 Nevertheless, EC awareness and practices worldwide, are still dismal even among women who are sexually active.1,6,10,11 Poor knowledge and concomitant low use of EC have been blamed both on disparate understanding about the mechanism of action of EC use, giving room for wrong timing and inefficiency, just as poor sex education has, consequently decreases its use.2,8 Further, the fear of rise in cases of sexually transmitted diseases (STDs) with the popularization of EC was allayed by an earlier study4 which showed that knowledge and use of condom was correspondingly high among students who use EC. And EC pills used in the correct doses have been reported not to negatively affect fertility as a side effect.1,3
Introduction
Emergency contraception (EC) works by utilizing hormones to inhibit or delay ovulation [1][2][3] or by utilizing Copper-Intrauterine device/CU-IUD to prevent fertilization/implantation. 2 In this regard, EC is the option left for a woman who got involved in an unprotected sex including forced sex, or failed in her regular birth control method and does not want to become pregnant. 4,5 EC will forestall unintended pregnancies, which in the United State constitute about half of all pregnancies, 1,6 will help prevent about 50% of abortion cases also in the United State 1 and will reduce maternal deaths that might accrue in the developing world like Nigeria, from induced abortions most of which are carried out by untrained hands. 5,[7][8][9] Nevertheless, EC awareness and practices worldwide, are still dismal even among women who are sexually active. 1,6,10,11 Poor knowledge and concomitant low use of EC have been blamed both on disparate understanding about the mechanism of action of EC use, giving room for wrong timing and inefficiency, just as poor sex education has, consequently decreases its use. 2,8 Further, the fear of rise in cases of sexually transmitted diseases (STDs) with the popularization of EC was allayed by an earlier study 4 which showed that knowledge and use of condom was correspondingly high among students who use EC. And EC pills used in the correct doses have been reported not to negatively affect fertility as a side effect. 1,3 Although the knowledge and use of EC among youths has been described in various settings of Nigeria and elsewhere, this requires being updated, as a way of evaluating how the fight is improving against unintended pregnancies and it sequeles. And to our knowledge, no study of such has been dedicated to female Nigerian National Youth Service Corp (NYSC) members. The female youth corps member in Nigeria deserves a special evaluation of her knowledge and use of EC, if unintended pregnancies, back-street abortions and the consequences are to be reduced. This group of youths are usually fresh graduates of Universities/Polytechnics who are usually affected by a number of factors, considered to be associated with increased sexual activities; 1) Higher social interactions; this is associated with high sexual activities. 5 2) Averagely in her 20s; a stage characterised by increased sexual flares. 12 3) Observing a mandatory national assignment in an environment other than that she is indigenous to, thus parental influences are reduced which encourages more sexual behaviours. 13 4) Higher possibilities to associate and make new friends whose sexual tendencies may be higher and likely to influence hers. 14 5) On a grossly inadequate stipend of ₦650 (˂$2) per day to cover all her expenses including feeding, accommodation, travels which may push her further into risk of sex for benefits. 14 To this extent, we intend to evaluate the knowledge and utilization of EC in female youth corps members in Nigeria, which is a representative group of the various ethnic, religious and regions in the country. And we will also be examining how some socio-demographic factors could be associated with the knowledge and use of EC in this group.
Study background
We carried out our study for about 12 weeks among female National Youth Service Corps (NYSC) members observing their obligatory national assignment in Sokoto State, Nigeria. Nigeria is stratified geographically into ethnicities, with each being grossly at variance to the other in terms of religious beliefs, educational levels and some other social orientations.
Briefly about the NYSC scheme: After the Nigerian civil war ended in 1970, the NYSC scheme was established by decree 24 in 1973 and later repealed and replaced by decree 51, 1993 with the aim of reconciling and rebuilding relations within the Nigerian state. As part of its objectives to build common ties among the Nigerian youths and promote national unity and integration, youths (≤30 years) are assigned to a mandatory national assignment after their first or second degrees, for a period of 1 year, in a state other than their states of origin. By this, the NYSC members are exposed to ways of living of others Nigerians different from themselves. 15 Therefore, the NYSC members in any state are essentially a collection of Nigerians of various religious groups, ethnicities and regions.
Study design and population
It was a cross-sectional descriptive study targeted at serving female NYSC members using a self-administered, validated, structured questionnaires (Appendix 1).
Eligibility/inclusion criteria
Serving Female youth corps members, within the age bracket of 18-29yrs and who consented to participate in the study were recruited. Indeed, all participants in the study are Nigerians.
Exclusion criteria
Non-NYSC, non-consenting and Non-Nigerians and males were excluded from the study.
Sample size estimation
Kish and Leslie formula was used to arrive at the 177 sample size, given that 13.3% of the sexually active female students in University of Port-Harcourt, Nigeria had ever used Emergency Contraception. 16
Sampling technique
Stratified sampling method was employed Step 1 The respondents were categorised into zones. The Sokoto state NYSC scheme is divided into four (4) zones (Bodinga, Gwadabawa, Sokoto and Wurno zones) for ease of administration. And as at the time of our study, 949 female corps members were serving in the State (Appendix 2).
Step 2
A proportionate distribution of the respondents was made. The following were the distribution of the respondents, in accordance with each zone's contribution to the State 949 pool of female corps members; Step 3 Monthly clearance days were used for the recruitments (virtually all corps members turn in for clearance on these days). Every fifth successive and consenting female corps member on the clearance register was enrolled but after explaining to them the aim and objectives of the study and assuring them of confidentiality. 180 questionnaires were administered but 162 were completed and returned, bringing the response rate to 90%.
Evaluation
A scoring system was adapted for the 9 questions assessing respondents' knowledge of EC. Thus, ≥5 correctly answered questions are considered good, while ≤4 correctly answered are considered to have poor knowledge of EC.
Statistical analysis
Data were entered into and analysed using SPSS version 20.0. Age was expressed as the mean±S.D of the studied group. Charts were used to present frequencies. Tables 1-5 were used to present variables used in defining association between EC knowledge and some sociodemographic variables. Pearson Chi-Square test was used to validate the statistics at P-value of ≤0.05.
Ethical permission
This was obtained from the office of the state coordinator of the NYSC, Sokoto State (Appendix 2) and the informed consent of the recruited female youth corps members with/without that of their spouses were sought.
Socio-demographic stratification
162 of the 180 questionnaires served were completed and returned by the respondents. This gave a response rate of 90%. The mean age of the respondents was 25.4±2.5 years. The socio-demographic distributions of the respondents are as shown in Figures 1-6. It is obvious in fig 1a that the NYSC scheme at any state represents Nigeria in its distributions of its people by ethnicities. Similarly, it is obvious from Figure 2 that the female corps member has Christian's preponderance, in a Muslims dominated northern Sokoto state of Nigeria, which is an objective of the NYSC scheme. Figure 3 showed that most of the youth corps members are single which obviously is because of the category of Nigerians involved in the scheme. And Figure 4 showed how the respondents are spread across the 6 geopolitical zones of Nigeria. Figure 5 and Figure 6 depict the distribution of the respondents into whether they are graduates of a Polytechnic (HND) or a University or a degree awarding College of Education (1 st or 2 nd degrees).
Knowledge of emergency contraceptive
87 (53%) of the respondent know what EC is, 46 (28.4%) don't know, 22 (13.6%) were not sure what EC is while, 7 (4.3%) decline to answer whether or not they know EC. On the other hand, 37 (22.8%) of the respondents showed good knowledge of EC, 118 (72.8%) displayed poor knowledge of EC and 7 of them (4.3%) did not respond to the question. In the same vein, only 31 (35.6%) of the respondents that indicated they know EC were able to demonstrate good knowledge of EC (Figure 7). 29 of the respondents (17.9%) know hormonal/pills types EC, much fewer, 2 (1.2%) of them knows Copper-Intrauterine Device (Cu-IUD) and 131 (80.9%) of them declined answering the question. Similarly, on the question of whether respondents know any oral EC, only 26 (16%) answered in the affirmative, a huge number, 96 (59.3%) answered no and another handful of them, 40 (24.7%) did not answer the question. And about 53.8% (14/26) of the respondents that answered yes to oral EC were able to mention any examples ( Figure 8) and yet a lower portion, 50% (13/26) of those who answered yes to the question of oral EC knew how it is used.
Attitudes towards emergency contraceptives
While 44.4% (72) of the respondents believed it is morally alright to utilize EC when there is need, 32.1% (52) said it is not morally okay and 23.5% (38) of the respondents chose not to answer on that field. Only 23 (14.2%) of the respondents ever used an EC, a huge 127 (78.4%) of them never applied it and 12 (7.4%) of them did not answer the question. Furthermore, only 56.5% (13) of the 23 respondents that ever had reasons to apply EC were able to mention the type of EC they used. And 24 (14.9%) of respondents volunteered that they had ever been involved in an act that required EC, 112 (69.1%) of them answered no and 26 (16.0%) declined answering the question.
Practice of emergency contraceptives
Out of the 24 respondents above that ever got involved in an act that required EC only 9 (37.5%) of them access EC (Figure 9). It is obvious most of the respondents hardly access EC services, as most of them, 135 (83.3%) declined answering the fields. Only 10 (6.2%) access it through family planning clinic, 14 (8.6%) get through a private doctor and 3 (1.9%) others by other sources including chemists shops. Only 16 (9.9%) of our respondents believed EC was effective for them, 27 (16.7%) said it was not effective and 119 (73.5%) refused to answer that question. Most of the respondents, 115 (71.0%) have never been introduced to EC. But, 13 (8.0%) were introduced by a doctor, 16 (9.9%) by friends, 12 (7.4%) by sexual partner and 6 (3.7%) by other avenues including drug marketers. And just 12 (7.4%) of them ever introduced anybody to EC, while 101 (62.3%) never introduced anybody and the remaining 49 (30.2%) chose not to answer this question.
Figure 9
Respondents that ever got involved in an act that requires EC, those that access EC and those that ever got involved and accessed.
Correlations of EC knowledge with some sociodemographic variables
Out of the 162 respondents that participated in the study, 7 (4.3%) chose not to attempt the section fielded on the knowledge of EC and it is shown in the summary presented in Table 1. Our results showed that there is a positive relationship between tribe and EC knowledge (χ 2 =8.505, p=0.037), with the Yoruba and other tribes, aside from the Hausa and Igbo likely to show better knowledge of EC. There is no strong relationship between religion and EC knowledge (χ 2 =3.124, p=0.077). Similarly, there is no good correlation between the course a respondent studied and their knowledge of EC (χ 2 =1.174, p=0.556). But there was a strong positive correlation between EC knowledge and the geopolitical zone of the respondents (χ 2 =17.830, p=0.003). With the respondents from the North-central (N-C) and South-west (S-W) states likely to have better knowledge of EC than other geopolitical zones.
Discussions
The main findings of the present study are; in a representative group of female NYSC members of Nigeria, knowledge of EC is low (22.8%) even among those of them that signified they know it. That the concept of EC as seen in a significant number (32.1%) of this group of Nigerian youths as having moral issues, even when there is obvious need. Records of EC use are poor and the seeking attitude even when it is necessary was equally poor. It is obvious there is little or no encouragement from any person(s) as 115 (71.0%) of the respondents have never been introduced to EC to enhance its utilization. Although, a study 17 reported that there was no influence of ethnicity on the utilization of EC, we have been able to demonstrate in the index study that EC knowledge is related to tribe (χ 2 =8.505, p=0.037) and we showed that it was low among the respondents who are Hausa and Igbo by tribe, but the Yoruba and other tribes displayed high knowledge of EC. And this is supported by the demonstration for the first time in the present study that the respondents knowledge is influenced by their geopolitical zones of origin (χ 2 =17.830, p=0.003), that those respondents from the North-Central and South-Western states of Nigeria displayed higher knowledge of EC than their counterparts from the North-West, North-East, South-South, and South-Eastern states.
Previous studies have examined the knowledge and use of EC among students (male and female) of higher institutions of learning. 6,11,14,16,18 Most of the respondents in these study groups were still living off and under the guidance of their parents or some close relatives. None of the previous studies, form our search have investigated knowledge and use of EC among female graduates on mandatory national assignment away from their indigenous area, to usually a new environment without their parents close by and with a meagre stipend to cater for themselves. These, put together will increase their sexual behaviour, which may also upsurge the incidences of unwanted pregnancies and abortions that are usually handled mostly in developing countries like Nigeria by back-street abortionist. The consequences of which are usually not pleasant.
We found that about half of the respondents (53%) are aware of EC. This is similar to what was recorded among female undergraduates (58%), more than 10 years ago, in University of Benin South-South, Nigeria. 18 This is better than the 22.8% which in South Africa was reported on awareness of EC. 10 In the present study, a unified representative group showed lower EC awareness than what obtains in University of Benin. This may be explained by the relations we have earlier established to exist between EC knowledge and tribes and geopolitical zones. We have also recorded 22.8% of good knowledge of EC among our study group. This is low considering the importance of good EC knowledge in the group under consideration but it is better than the 18.5% reported among males and female students of institutions of higher learning in Osun state, South-West Nigeria. 5 Similar low EC knowledge was reported in South Africa 10 and same low knowledge EC was reported among women in California. 19 Oral EC are more popular, readily available and more convenient to use. 2 But in the present study, only 16.0% of the respondents know any oral EC and just about half of this number were able to mention any example of oral EC pills.
Less than half (44.4%) of the respondents in the present group believe it is morally oaky to use EC when there is indication. Other respondents either chose not to answer and yet some believed it is just not okay even if there is indication. Also, just about 56.5% of the respondents who know EC ever applied it. On the other hand, a study 10 in South Africa found a lower percentage (9.1%) of those who know EC have ever used it. This present study also found that as low as 37.5% of the respondents accessed EC even after being involved in an act that obviously required it.
Limitations of the study
The relatively small sample size and considering only female NYSC members, not the general population in the study, as well as the descriptive nature of the study makes inference making difficult in the current study.
Conclusion
It is obvious in the present study that the awareness, knowledge and utilization of EC among female Nigeria graduates is far below what is expected, even as the rate of EC use is expected to be higher, to avoid unplanned pregnancies. 20 The rate of non-responses to questions fielded was also high and makes one to think that people don't want to volunteer information related to sexuality even though our research team have as much as possible avoided sensitive questions in the questionnaire design. To this extent, more formal sex education should be encouraged to raise awareness and knowledge. 5,18 Information leaflets should also be distributed in advance to improve EC knowledge. 21 Although, a study 22 in Hong Kong showed most doctors don't support EC in advance. But some others 23,24 have shown abundantly that EC pills made available just as condoms will enhance the utilization. 17,24 By this, the incidence of unintended pregnancies will be brought down and therefore criminal abortions and the consequences will be minimised. | 2020-03-07T09:08:21.173Z | 2019-01-01T00:00:00.000 | {
"year": 2019,
"sha1": "3deae0900677555f8f5f5d0f52010c30027de41e",
"oa_license": "CCBYNC",
"oa_url": "https://medcraveonline.com/OGIJ/OGIJ-10-00463.pdf",
"oa_status": "GOLD",
"pdf_src": "MergedPDFExtraction",
"pdf_hash": "3deae0900677555f8f5f5d0f52010c30027de41e",
"s2fieldsofstudy": [
"Medicine",
"Sociology"
],
"extfieldsofstudy": []
} |
268287375 | pes2o/s2orc | v3-fos-license | circNOX4 activates an inflammatory fibroblast niche to promote tumor growth and metastasis in NSCLC via FAP/IL-6 axis
Background Cancer-associated fibroblasts (CAFs) orchestrate a supportive niche that fuels cancer metastatic development in non-small cell lung cancer (NSCLC). Due to the heterogeneity and plasticity of CAFs, manipulating the activated phenotype of fibroblasts is a promising strategy for cancer therapy. However, the underlying mechanisms of fibroblast activation and phenotype switching that drive metastasis remain elusive. Methods The clinical implications of fibroblast activation protein (FAP)-positive CAFs (FAP+CAFs) were evaluated based on tumor specimens from NSCLC patients and bioinformatic analysis of online databases. CAF-specific circular RNAs (circRNAs) were screened by circRNA microarrays of primary human CAFs and matched normal fibroblasts (NFs). Survival analyses were performed to assess the prognostic value of circNOX4 in NSCLC clinical samples. The biological effects of circNOX4 were investigated by gain- and loss-of-function experiments in vitro and in vivo. Fluorescence in situ hybridization, luciferase reporter assays, RNA immunoprecipitation, and miRNA rescue experiments were conducted to elucidate the underlying mechanisms of fibroblast activation. Cytokine antibody array, transwell coculture system, and enzyme-linked immunosorbent assay (ELISA) were performed to investigate the downstream effectors that promote cancer metastasis. Results FAP+CAFs were significantly enriched in metastatic cancer samples, and their higher abundance was correlated with the worse overall survival in NSCLC patients. A novel CAF-specific circRNA, circNOX4 (hsa_circ_0023988), evoked the phenotypic transition from NFs into CAFs and promoted the migration and invasion of NSCLC in vitro and in vivo. Clinically, circNOX4 correlated with the poor prognosis of advanced NSCLC patients. Mechanistically, circNOX4 upregulated FAP by sponging miR-329-5p, which led to fibroblast activation. Furthermore, the circNOX4/miR-329-5p/FAP axis activated an inflammatory fibroblast niche by preferentially inducing interleukin-6 (IL-6) and eventually promoting NSCLC progression. Disruption of the intercellular circNOX4/IL-6 axis significantly suppressed tumor growth and metastatic colonization in vivo. Conclusions Our study reveals a role of the circRNA-induced fibroblast niche in tumor metastasis and highlights that targeting the circNOX4/FAP/IL-6 axis is a promising strategy for the intervention of NSCLC metastasis. Supplementary Information The online version contains supplementary material available at 10.1186/s12943-024-01957-5.
Background
Metastasis is the leading cause of cancer-related mortality and consists of multistep processes driven by the cooperation of cancer cells with the tumor microenvironment (TME) [1].As predominant stromal components in the TME, cancer-associated fibroblasts (CAFs), also known as activated fibroblasts, actively orchestrate a supportive niche as "fertilized soil" and endow incipient cancer cells with traits needed to develop metastases [2].CAFs are also identified as a generalized biomarker of the fibrotic TME subtype that correlates with inferior therapeutic outcomes and prognosis across various cancers [3].With the development of single-cell sequencing approaches, tremendous insight has been yielded into the heterogeneity and plasticity of CAFs.The composition and functional states of fibroblasts differ extensively in the evolving TME, which makes targeting CAFs challenging in clinical settings [4][5][6].Therefore, reprogramming CAFs into a "normal" or quiescent state might be a promising approach that benefits early cancer treatment and inhibits metastasis.
Accumulating evidence has implied that aberrant fibroblast activation is an early event before cancer cell dissemination [7].When the interaction with cancer cells begins, local normal fibroblasts (NFs) are usually among the first cell types to be recruited and activated into CAFs, and ultimately establish a niche to facilitate metastatic cascades [8,9].Previous studies have shown that paracrine signaling molecules drive fibroblasts from homeostasis to activated states, including transforming growth factor β (TGF-β), platelet-derived growth factor (PDGF), fibroblast growth factor (FGF), and interleukin-1 [7].Moreover, direct physical contact with cancer cells via ligand receptor binding also leads to the transition of NFs into CAFs [10].More interestingly, CAFs maintain activated phenotypes even in the absence of cancer cells [11].However, the underlying mechanisms of fibroblast activation and phenotype switching that drive metastasis have not been extensively elucidated.
Fibroblast activation protein (FAP), a type II membrane serine protease, is widely acknowledged as one of the canonical CAF markers [12].FAP is overexpressed on activated fibroblasts and coincides with poor prognosis in cancers, which has made FAP-targeted imaging and therapy appealing [13].Indeed, over 28 different cancer entities have been accurately diagnosed in patients utilizing FAP ligands, including metastatic lesions [14].Cumulative studies have revealed that FAP-positive CAFs (FAP + CAFs) account for the main phenotypes of activated fibroblasts in the TME, and they execute crucial functions in promoting tumor growth, angiogenesis and metastasis, as well as the formation and maintenance of an immunosuppressive microenvironment [15][16][17].In addition, our previous work demonstrated that FAP orchestrated the immune microenvironment and served as a biomarker of immunotherapy resistance.Patients with high FAP expression had a significantly shorter progression-free survival in immunotherapy of non-small cell lung cancer (NSCLC) [18].However, the upstream epigenetic mechanisms by which FAP drives tumor progression are unclear and warrant further investigation.
In this study, we determined that FAP + CAFs are correlated with metastasis and poor prognosis in NSCLC patients.By investigating the critical mediator in fibroblast activation, we identified a novel circular RNA (cir-cRNA), circNOX4, that is significantly upregulated in CAFs and correlates with the worse overall survival of NSCLC patients.Importantly, circNOX4 fuels tumor growth and metastasis by activating the fibroblast niche via the miR-329-5p/FAP/IL-6 axis.Our results elucidate a novel mechanism of fibroblast activation and IL-6 signaling by CAFs to establish an inflammatory niche and highlight circNOX4 as a promising candidate for CAFtargeted therapy in NSCLC.
Patients and clinical samples
Tumor samples were collected from a cohort of 84 NSCLC patients at the Fourth Hospital of Hebei Medical University (Shijiazhuang, China) between August 2018 and March 2019 with the approval of the Medical Ethics Committee and informed consent from the corresponding participants (2018MEC005).All samples were histologically confirmed to have NSCLC.Patients' clinical information was collected and stored in a database, which was updated every 3 months by telephone followup.Overall survival (OS) was defined as the time interval from diagnosis to the date of cancer-related death or the last follow-up (January 2024).
Immunohistochemistry (IHC)
IHC staining was performed according to standard protocols as previously described [18].The staining was evaluated based on the intensity of FAP (ab207178, 1:250, Abcam, CA, USA) by two independent pathologists, and all scoring was completed blinded to the clinical information.The intensity of FAP was classified as 0, no staining; 1, weak intensity; 2, moderate intensity; and 3, high intensity.Scores of 0 and 1 were considered low expression, while scores of 2 and 3 were considered high expression.
Isolation, identification, and culture of fibroblasts from human NSCLC tissues
Five paired CAFs and NFs were isolated from human NSCLC tissues and normal tissues (more than 5 cm away from the tumor margin).Freshly isolated surgical specimens were acquired from patients with the approval of The Fourth Hospital of Hebei Medical University (2018MEC005).The patient's basic information is listed in Additional file: Table S1.For fibroblast isolation, tissue was initially stored and washed in PBS with penicillin/streptomycin (100 U/100 μg, BI, Israel).The tissue was then physically minced and digested enzymatically in a solution containing 2 mg/mL DNase (Gibco, Waltham, MA, USA), 5 mg/mL collagenase type I (Gibco, Waltham, MA, USA), 100 U/mL penicillin, and 100 μg/mL streptomycin at 37 °C for 2 h.After a series of filtration and centrifugation steps, cells were subsequently cultured in 15% fetal bovine serum (FBS, Gibco, Waltham, MA, USA), DMEM (Gibco, Waltham, MA, USA), 20 ng/mL EGF (PeproTech, Suzhou, China) and penicillin/streptomycin (100 U/100 μg) at 37 °C in humidified air with 5% CO 2 for subsequent expansion.Fibroblast populations were identified by morphology, western blotting and immunofluorescence staining (fibroblast markers α-SMA and FAP; epithelial cell marker EpCAM).Fibroblasts were subcultured at 80% confluence and used in experiments at passages 4-10.Antibody information is provided in Additional file: Table S2.
circRNA microarray
The circRNA microarray analysis (GEO accession number: GSE244065) was performed by the Sinotech Genomics Corporation, Shanghai, China.In brief, total RNA from three paired sets of CAFs and NFs was extracted by TRIzol reagent (Life Technologies, Carlsbad, CA, US) and purified using the RNeasy Mini Kit (Qiagen, GmBH, Germany).Then, RNA samples were utilized to generate biotinylated cRNA targets for the Sino Human ceRNA array V3.0.After hybridization, slides were screened by the Agilent Microarray Scanner (Agilent Technologies, Santa Clara, CA, US).Raw data were extracted and normalized, and circRNAs with a fold change of at least 1.5 and P values < 0.05 were selected for further analysis by R software.
Confirmation of the circular structure
Both complementary DNA (cDNA) and genomic DNA (gDNA) extracted from CAFs and NFs were amplified with circNOX4 primers (both convergent and divergent primers).Agarose gel electrophoresis was applied to analyze the RT-PCR products.The back-splice junction site of circNOX4 was verified by Sanger sequencing.Extracted RNAs were treated with RNase R (Geneseed, 3 U/μg, 37 °C, 20 min) to detect the stability of circRNA.
RNA isolation, cDNA synthesis, and quantitative RT-PCR (qRT-PCR) assay
In brief, total RNA was isolated using the E.Z.N.A.Total RNA Kit (OMEGA, Japan).cDNA was synthesized using HiScript III RT SuperMix (Vazyme, Nanjing, China) and amplified using a Light Cycler 480 II Real-Time PCR System (Roche, Basel, Switzerland) with ChamQ Universal SYBR qPCR Master Mix (Vazyme, Nanjing, China).GAPDH and U6 were used as internal controls via the classical ΔΔCt method.Primers were designed and synthesized by RiboBio (Guangzhou, China) and Tsingke (Wuhan, China).The primer sequences are listed in Additional file: Table S3.
RNA-fluorescence in situ hybridization (FISH)
FISH assays were performed on CAFs/NFs and NSCLC tissues as previously described [19].FISH probes were designed and synthesized by Servicebio (Wuhan, China).Dig-labeled probes specific for circNOX4 and biotinylated locked nucleic acid miR-329-5p probes were used during hybridization.The targeted sequences of the probes are provided in Additional file: Table S4.The FISH score for circNOX4 in NSCLC tissues was determined based on the integrated optical density values scanned by AIPATHWELL software (Servicebio Technology Co., Wuhan, China).The optimal cut-off value of 0.1 was determined by X-tile software.Based on this threshold value, patients were categorized into circNOX4 high-expression or low-expression subgroups for further analysis.
Collagen contraction assay
The fibroblasts were embedded in three-dimensional collagen matrices in 24-well plates with Rat Tail Collagen I (Corning, NY, USA) at a final concentration of 2 mg/ mL collagen and 5 × 10 5 cells per mL.A 500 μL aliquot of this mixture was dispensed into each well and allowed to solidify for 30 min at 37 °C.Gels were freed from the wells using a pipette tip after polymerization, and complete culture medium was added to the wells, followed by incubation for 12 h.ImageJ software was employed to measure the contractile capacity, which is defined by the following equation: [Area (well)-Area (gel)/Area (well)] × 100%.
Western blotting analysis
Standard western blotting was carried out by a wet transfer system (Bio-Rad Laboratories).Briefly, cells were lysed in lysis buffer (Solarbio, Beijing, China) supplemented with protease inhibitors.Total proteins were separated on SDS-PAGE gels and then transferred onto PVDF membranes (Millipore, Billerica, MA, USA).After blocking the membranes with 5% milk-TBST, they were incubated with appropriate dilutions of specific primary antibodies against FAP (ab207178, 1:1000, Abcam) and α-SMA (ab124964, 1:1000, Abcam).The blots were incubated with HRP-conjugated secondary antibodies and visualized using the ECL system (GE, Boston, MA, USA).
Cell viability assay
The cell viability of fibroblasts was measured by a cell counting kit-8 (CCK-8) assay (Solarbio, Beijing, China).A total of 2000 cells were seeded in 96-well plates per well, and 10 μl of CCK-8 reagent was added and incubated for 2 h at 37 °C.The absorbance was detected using a microplate reader (Tecan, Switzerland) at a wavelength of 450 nm.
Coculture of CAFs and cancer cells, conditioned medium (CM) preparation, migration and invasion assays
For coculture of CAFs and cancer cells, NSCLC cells were plated in the upper chamber of transwell apparatus (0.4 μm insert; Corning, Corning, NY, USA), and CAFs were cultured in the lower chamber.After incubation for 48 h, supernatants were collected for further measurements.For CM preparation, indicated fibroblasts were cultured to reach 80% confluence and then replaced with serum-free DMEM for 48 h.The supernatants were collected and centrifuged to remove cell pellets.The CM was subjected to cytological experiments or stored at -80 °C.
For the cell migration assay, NSCLC cells were cultured in six-well plates and scraped linearly to create an artificial wound.Then, NSCLC cells were incubated in CAF-CM or NF-CM for 48 h.For the cell invasion assay, the upper chambers of the Transwell plates (8 μm insert) were precoated with Matrigel (dilution, 1:8; BD Biosciences, Franklin Lakes, NJ, USA) at 37 °C for 30 min.NSCLC cells (5 × 10 4 /well) suspended in serum-free medium were seeded into the upper chambers.Indicated fibroblasts (3 × 10 4 /well) were seeded into the lower chambers.Subsequent procedures were performed according to previous studies [20].
Dual luciferase reporter assay
Wild-type (WT) or mutant (MUT) pmiRGlo-circNOX4 and pmiRGlo-FAP-3′UTR dual luciferase reporter vectors incorporating predicted binding sites for miR-329-5p and miR-624-5p were constructed by GenePharma (Shanghai, China).The sequences are provided in Additional file: Table S5 and Additional file: Table S6.Cells were seeded on 12-well plates at a density of 60% and then co-transfected with WT or MUT plasmids with miRNA mimics or a negative control (NC) using Lipofectamine 3000 reagent.After 48 h of incubation, the cells were collected and tested for luciferase activity with a Dual-Luciferase Assay System kit (Promega, Madison, WI, USA) according to the manufacturer's instructions.
RNA immunoprecipitation (RIP) assay
RIP was performed with an RNA immunoprecipitation kit (Geneseed, Guangzhou, China).CAFs were transfected with miR-NC or miR-329-5p for 48 h.Then, the cells were lysed with RIP lysis buffer and incubated with magnetic beads precoated with antibodies against Ago2 (ab186733, 1:50, Abcam) or IgG (ab172730, 1:50, Abcam).After the antibody was recovered by protein A/G beads and purification was completed, qRT-PCR was performed to detect the enrichment of circNOX4 and miR-329-5p in the precipitates.
Cytokine antibody array
A human cytokine array (AAH-CYT-G5; RayBiotech, Norcross, GA, USA) was used to measure the secretion levels of 80 cytokines in the CM of sh-circNOX4 CAFs and sh-NC CAFs according to the manufacturer's instructions.Quantitative array analysis was performed using the ImageQuant LAS4000 Scanner (GE, Boston, MA, USA).Cytokines were screened using the following integrated conditions: Fold change ≥ 1.2 and fluorescence intensity values > 300.
Mouse experiments
The animal experiments were conducted in accordance with national guidelines and approved by the Animal Research Committee of the Fourth Hospital of Hebei Medical University (SYXK2022-011).BALB/c nude mice (female, age 5 weeks) were purchased from Beijing HFK Bio-Technology Co., Ltd.
For the lung metastasis model, 2 × 10 5 A549 cells in 100 μL PBS were injected into the tail vein of BALB/c nude mice.The mice were randomly assigned to one of three groups: (1) untreated A549 cells, (2) A549 cells "primed" in vitro for 72 h in a transwell coculture model with CAFs transduced with empty vector (sh-NC), or (3) CAFs transduced with circNOX4 shRNA (sh-circNOX4).After 5 weeks, the mice were sacrificed and dissected, and the lungs were collected.Hematoxylin and eosin (HE) staining was used to observe the metastatic foci.
Ga-labeled FAPI-04 PET imaging of mice
The preparation of 68 Ga-Labeled FAPI-04 solutions was performed by previous method [22].Briefly, 68 Ga was generated by a 68 Ge-68 Ga generator and eluted with a solution of 0.1 M hydrochloride.The mixture of the 68 Ga solution (74 MBq), 1.0 M sodium acetate (95 μL) and 1.15 mM FAPI-04 (20 μL) were reacted at 95 °C for 10 min. 68Ga FAPI-04 PET imaging was performed up to 60 min after intravenous injection of 11.1 MBq of 68 Ga-FAPI-04 in tumor xenograft mice (n = 3) using a micro-PET/SPECT/CT machine (Novel Medical Equipment Ltd., China) under isoflurane anesthesia.PET images were reconstructed with the comprehensive image analysis software Pmod v4.201 (PMOD Technologies LLC, Switzerland) and were converted to SUV images.Quantification was performed using a region-of-interest (ROI) technique and expressed as SUVmax.
Statistical analyses
Statistical analyses were performed using SPSS 22.0 software (SPSS Inc., Chicago, IL, USA), GraphPad Prism 7 (GraphPad, San Diego, CA, USA) and R software V4.2.0 (RStudio, Murray Hill, NJ, USA).The data are presented as the mean ± SD of at least three biological replicates.Differences between the two groups were analyzed by t-tests.The associations between circNOX4 expression and clinicopathological characteristics were evaluated by the Chi-square test.The statistical significance of survival was estimated by the Kaplan-Meier method and Cox analysis.P values less than 0.05 were considered statistically significant.
FAP + CAFs correlate with tumor metastasis and poor survival in NSCLC patients
Given the supporting role of CAFs in tumor growth and metastasis [7], we first confirmed the contribution of FAP + CAFs in NSCLC.We queried FAP mRNA expression using the TCGA database and found that FAP was upregulated in lung adenocarcinoma (LUAD) and lung squamous cell carcinoma (LUSC) compared with normal lung tissues (Fig. 1A).The data also showed the copy number and mutation status of FAP gene were not correlated with FAP mRNA expression, indicating that the high expression level of FAP was potentially regulated by epigenetic mechanisms (Fig. S1A).Next, we analyzed the spatial cellular expression pattern of FAP across different datasets and cell types in the TME and confirmed that FAP was specifically enriched in fibroblast lineages (Fig. S1B).Moreover, we analyzed the GEO profile GSE127465, which revealed that fibroblasts were distributed into five clusters in NSCLC.We found that FAP was enriched in the C1 and C2 subpopulations, indicating the heterogeneous nature of CAFs in the TME (Fig. 1B, C).To uncover the biological function of FAP + CAFs in development trajectories, we constructed developmental pseudotime and discovered that FAP expression was relatively higher along the branch 1 cellular state, which was accompanied by increased protumor activities such as angiogenesis, inflammation, and metastasis (Fig. 1D-F).
These findings revealed that FAP exhibits cell lineageand developmental stage-dependent expression patterns while undergoing protumor activities.
To investigate the clinical relevance between fibroblast activation and the prognosis of NSCLC patients, we collected clinical samples from 84 NSCLC patients and performed IHC staining.The result showed that FAP was expressed in the stromal compartment in NSCLC tissues (Fig. 1G).Analysis of clinicopathological variables based on FAP expression level revealed that high expression of FAP was positively correlated with high stromal fraction and multiple distant metastases (Fig. 1H, I), indicating that FAP is closely related to the process of metastasis.Moreover, FAP expression was inversely correlated with overall survival in NSCLC patients (P = 0.023; Fig. 1J), in line with the clinical outcome data from the TCGA database (Fig. 1K).Taken together, FAP + CAFs are populated in metastatic tumors with inferior clinical outcomes in NSCLC patients.
Identification of circNOX4, a CAF-specific circRNA, that predicts poor prognosis for NSCLC patients
To gain insight into the contribution of circRNAs to aberrant fibroblast activation in the TME, we first isolated primary CAFs and NFs from human NSCLC tissues and adjacent normal tissues and identified the characteristics of fibroblasts (Fig. S2A-C).By performing circRNA microarray screening of CAFs and NFs, we identified 62 upregulated circRNAs in CAFs (cutoff value: |Fold change|> 1.5, P values < 0.05, Mean > 7, Fig. 2A and Fig. S3A).Then, we selected the top 5 upregulated circR-NAs as candidate circRNAs based on a competing endogenous RNA (ceRNA) network and detected 2 circRNAs (circ_0064142 and circ_0023988) that were enriched in CAFs through gel electrophoresis (Fig. S3A-C).However, Sanger sequencing of the reverse transcription product failed to confirm the circular splicing of circ_0064142, so we chose circ_0023988 (circNOX4) for further identification.Then, qRT-PCR analysis showed that circ-NOX4 was significantly upregulated in CAFs compared to NFs, human NSCLC cell lines (A549, PC9, H226 and H1581) and bronchial epithelial cells (BEAS-2B), indicating a potential role in fibroblast activation (Fig. 2B, C).Next, we identified the origin of circNOX4, which was generated from exons 6-11 of the NOX4 transcript (chr11:89,155,069-89185063) with a length of 476 nt.The back-splice junction site of circNOX4 was confirmed by Sanger sequencing (Fig. 2D).To further verify the circular property of circNOX4, the back-splice junction site of circNOX4 was amplified by RT-PCR using convergent and divergent primers, which only amplified a specific divergent band in cDNA but not in gDNA (Fig. 2E).Additionally, circNOX4 was RNase R-resistant, whereas the control linear NOX4 and GAPDH mRNAs were not (Fig. 2F).These results indicate that circNOX4 is a bona fide circRNA highly expressed in CAFs.
To determine whether circNOX4 was clinically relevant in NSCLC patients, we performed fluorescence in situ hybridization (FISH) in 84 NSCLC specimens.The results showed that circNOX4 was expressed in the intratumoral stroma but not in cancer cells (CK19-positive areas) (Fig. 2G).Additionally, a higher circNOX4 expression level was correlated with distant metastasis (Fig. 2G, Table 1).Next, we investigated the correlation between circNOX4 expression and patient survival.The overall survival of NSCLC patients with high circNOX4 expression was significantly shorter than that of patients with low circNOX4 expression (P < 0.0001; Fig. 2H).Univariate and multivariate Cox regression models revealed that high circNOX4 expression and distant metastasis were independent prognostic factors for NSCLC (Fig. 2I, J).Collectively, our findings suggest that a CAF-specific circRNA, circNOX4, correlates with metastatic progression and may be a prognostic marker in NSCLC patients.
circNOX4 contributes to tumor progression by mediating fibroblast activation
To delineate the functional role of circNOX4 in the intrinsic characteristics of fibroblasts, we knocked down circNOX4 in CAFs or overexpressed circNOX4 in NFs without altering the expression of the parental gene NOX4 (Fig. S4A-C).Then, we speculated that circ-NOX4 could regulate the phenotype of fibroblasts.The CCK-8 and collagen contraction assays exhibited that G Representative images of immunohistochemistry for high-or low-FAP expression in NSCLC tissues.(brown = FAP; blue = hematoxylin nuclear staining).Scale bar = 50 μm.H The proportion of patients with high/low stroma fraction in high/low FAP expression groups.I The proportion of patients with different metastasis status in high/low FAP expression groups.J Kaplan-Meier curves with patients stratified according to FAP expression (low vs. high) for OS.K Kaplan-Meier curves for LUAD and LUSC patients according to FAP expression levels in the TCGA database circNOX4 overexpression significantly enhanced the proliferation and contraction capabilities of fibroblasts, while circNOX4 knockdown exerted the opposite effects (Fig. 3A, B).Afterwards, we examined how circNOX4 affected the expression of CAF markers and regulators, including FAP, MMP2, MMP14, collagen type 1 alpha 1 (COL1A1) and PDGFRβ.Interestingly, qRT-PCR suggested a reversion of the above gene expression profiles.Notable among them was FAP, which displayed the greatest reduction following circNOX4 knockdown (Fig. 3C).In parallel, circNOX4 overexpression upregulated FAP expression while circNOX4 knockdown attenuated FAP expression, as shown by western blotting and qRT-PCR analysis (Fig. 3D and Fig. S4D).To extend our understanding of circNOX4 in CAF activation, we detected circNOX4 expression in NFs after stimuli of TGF-β1, a critical mediator of fibroblast activation and cancerstroma bidirectional communication within the TME (Fig. S5A).Consequently, the expression of circNOX4 and FAP increased in a dose-and time-dependent manner upon exposure to TGF-β1 (Fig. S5B, C).Collectively, these findings support a potent effect of circNOX4 on fibroblast activation.
As the activation of fibroblasts is linked to protumor properties, we explored the role of circNOX4 in supporting cancer cell migration and invasion.NSCLC cell lines were treated with fibroblast-derived conditioned medium (CAF-CM and NF-CM).As expected, the wound healing assays showed that medium from sh-circNOX4 CAFs drastically decreased the migration ability of A549 and PC9 cells.In stark contrast, overexpressing circNOX4 increased the migration ability of cancer cells (Fig. 3E, F).Meanwhile, the transwell coculture system demonstrated that sh-circNOX4 CAFs reduced the invasion of A549 and PC9 cells, while circNOX4-overexpressing NFs exerted the opposite effects (Fig. 3G, H).Overall, circ-NOX4 facilitates NSCLC migration and invasion by activating CAFs and maintaining a protumor phonotype.
circNOX4 functions as a miR-329-5p sponge to upregulate FAP in CAFs
To dissect the molecular mechanisms by which circ-NOX4 induces fibroblast activation, we detected its intracellular distribution.The results of FISH and subcellular fractionation assays showed that circNOX4 was mainly localized in the cytoplasm of CAFs and NFs (Fig. 4A, B).
It has been reported that cytoplasmic circRNAs function by sponging miRNAs or encoding peptides [23].We next observed that circNOX4 had a limited protein-coding potential with no predicted protein features using circR-NADb software (R Score < 1.6, Fig. S6A).Therefore, we propose that miRNA sponge activity could be a possible mechanism for its functional effect.
Next, to validate whether circNOX4 sponges miR-329-5p and miR-624-5p, we constructed luciferase reporters harboring wild-type and mutant forms of circNOX4 according to the predicted binding sites on miR-329-5p or miR-624-5p, respectively (Fig. 4G, H and Fig. S6D).Consequently, the miR-329-5p mimic could specifically target the wild-type form of circ-NOX4, resulting in markedly decreased luciferase activity, but not the mutant form (Fig. 4H and Fig. S6E).RIP assays further confirmed the interaction of miR-329-5p with circNOX4 in CAFs (Fig. 4I).Moreover, RNA-FISH revealed that circNOX4 and miR-329-5p co-localized in the cytoplasm of CAFs (Fig. 4J).Unfortunately, due to the low abundance of circNOX4, AGO2 was unable to accumulate circNOX4 in NFs (Fig. S6F).In addition, the Fig. 2 circNOX4 is upregulated in CAFs and correlates with poor prognosis of NSCLC patients.A Schematic illustration of the identification of circNOX4 upregulated in CAFs.B qRT-PCR analysis of circNOX4 expression in five paired CAFs and NFs.C qRT-PCR analysis of circNOX4 expression in CAFs, NFs, NSCLC cell lines (A549, PC9, H226, H1581), and human bronchial epithelial cells (BEAS-2B).D Diagram illustrating the genomic location and back-splicing mode of circNOX4 from the NOX4 host gene.The head-to-tail splice junction site was confirmed by Sanger sequencing using the divergent primer.E cDNA and gDNA of CAFs and NFs were amplified with convergent and divergent primers.GAPDH was the negative control.F PCR and qRT-PCR analysis of circNOX4, linear NOX4, and GAPDH in CAFs and NFs with or without RNase R treatment.G Representative images (left) and quantification (right) of circNOX4 by using RNA-FISH in NSCLC tissues with and without distant metastasis.CK19 indicated cancer cells.The nucleus (blue) was stained with DAPI.Scale bar = 40 μm, DM, distant metastasis.H Kaplan-Meier curves of OS for NSCLC patients with high (n = 48) and low (n = 36) circNOX4 expression.The cut-off value was calculated using X-tile.Statistical significance was determined by the log-rank test.I Forest plots of OS, as determined by the univariate Cox proportional hazards model.J Forest plots of OS, as determined by the multivariate Cox proportional hazards model, were constructed using variables significantly associated with OS by the univariate Cox proportional hazards model (P < 0.1).Data are expressed as mean ± SD and statistical significance was determined by two-tailed Student's t-tests and log-rank tests.*P < 0.05, **P < 0.01 and ***P < 0.001 (See figure on next page.)miR-624-5p mimic had no effect on the luciferase activity of the circNOX4 wild-type reporter (Fig. S6G).These results support that miR-329-5p directly binds to circ-NOX4 in CAFs, whereas miR-624-5p does not.Furthermore, similar experiments were conducted to confirm the binding of miR-329-5p with the 3′-untranslated region (3′UTR) of FAP.As shown in Fig. 4K and Fig. 4L, we observed a high interaction probability between the miR-329-5p RNA sequence and FAP protein by the RPISeq prediction algorithm (https:// www.pridb.gdcb.iasta te.edu/ RPISeq/), and the 3′UTR of FAP harbored specific sequences complementary to miR-329-5p.Dual-luciferase reporter assays showed decreased luciferase activity of the wild-type reporter of the FAP 3′UTR with the miR-329-5p mimic, while no significant differences were found with the mutant reporter (Fig. 4L).Additionally, the expression of FAP was significantly decreased with miR-329-5p mimic transfected into CAFs (Fig. 4M).The above evidence confirms that miR-329-5p binds to the 3' UTR of FAP in CAFs.
To verify whether circNOX4 regulated the expression of FAP through miR-329-5p, we performed rescue assays.Knockdown of circNOX4 significantly decreased the expression of FAP, while miR-329-5p inhibitor co-transfection neutralized this downregulation (Fig. 4N).Conversely, overexpression of circ-NOX4 significantly increased the expression of FAP, while miR-329-5p mimic co-transfection abrogated this alteration (Fig. 4O), indicating that the regulation of FAP by circNOX4 was dependent on miR-329-5p.In addition, we conducted collagen contraction assays to determine whether miR-329-5p was involved in circNOX4-induced fibroblast activation.The results showed that the inhibitory effect of circNOX4 knockdown was reversed by the miR-329-5p inhibitor (Fig. 4P).In summary, our results suggest that circ-NOX4 sponges miR-329-5p to upregulate FAP and induce fibroblast activation.
circNOX4 facilitates CAF-mediated tumorigenesis and metastasis of NSCLC in vivo
To evaluate how circNOX4 contributes to the TME and cancer progression in vivo, we constructed two different mouse models: the xenograft tumor model and the lung metastasis model.For xenograft tumor model establishment, A549 cells were subcutaneously co-transplanted with or without CAFs isolated from NSCLC tissue samples (Fig. 5A).We found that xenografts derived from cotransplantation with CAFs resulted in enhanced tumor volume and tumor weight compared with xenografts derived from A549 cells alone, suggesting that CAFs promoted NSCLC progression (Fig. 5B, and Fig. S7A).More strikingly, sh-circNOX4 CAFs significantly stunted tumorigenesis compared with A549 cells co-transplanted with shNC-CAFs (Fig. 5B, and Fig. S7A), indicating that knocking down circNOX4 efficiently disabled CAFs and impeded the development of invasive tumors.
Next, to intuitively visualize the expression of FAP in vivo, we utilized micro-positron emission tomography (PET)/CT imaging with a specific nuclide molecular probe ( 68 Ga-FAPI-04).Micro-PET/CT revealed tracer uptake in the tumor xenografts, as shown in the 3D MIP images (Fig. 5C).The quantitative results of FAPI-PET imaging suggested that SUV uptake in the xenografts was significantly reduced in the sh-circNOX4 group compared to the control group (Fig. 5D), indicating that knocking down circNOX4 could downregulate FAP expression and reverse fibroblast activation in vivo.Furthermore, we examined the effect of circNOX4 on the expression of FAP, the endothelial marker CD31, mesenchymal markers (N-cadherin and Vimentin), and metastasis-related genes (MMP2 and MMP9).All these factors are critically involved in angiogenesis and epithelial-mesenchymal transformation (EMT), which contribute to the process of metastasis [24,25].The IHC results showed that CAFs significantly increased the expression levels of FAP, CD31, N-cadherin, Vimentin, and MMPs.However, these effects were reversed by transfecting sh-circNOX4 into CAFs, implying that circNOX4 provides a pro-tumorigenic niche that facilitates metastasis in vivo (Fig. 5E and Fig. S7B, C).
To further substantiate the effect of circNOX4, we constructed lung metastasis mouse models.In brief, A549 cells were either injected through the tail vein or "primed" in a coculture system with CAFs for 48 h in vitro prior to injection (Fig. 5F).As shown in Fig. 5G, more lung metastatic foci were detected in mice that had been injected with A549 cells "primed" with fibroblasts, compared with mice injected with the control A549 cells.Notably, a lower number of lung metastatic foci was presented in the sh-circNOX4-CAF group than in the shNC-CAF group, as evidenced by reduced HE-stained tumors in the lungs (Fig. 5G, H).All these data thus suggest that circNOX4 potentiates tumor metastasis by inducing fibroblast activation in vivo.
The circNOX4/FAP axis establishes an inflammatory fibroblast niche by secreting IL-6
The activation of fibroblasts coincides with paracrine signaling changes in the TME.To examine whether circNOX4 promotes metastasis in a paracrine manner, we collected supernatants from sh-NC CAFs and sh-circNOX4 CAFs and adopted a cytokine antibody array.
The results showed that sh-circNOX4 CAFs secreted a decreasing panel of cytokines, including interleukin-6 (IL-6), chemokine C-C motif ligand 2 (CCL2), TGF-β2 and insulin-like growth factor binding protein 4 (IGFBP-4), which are reported as niche characteristic genes that facilitate cancer cell migration, invasion, and colonization in the metastatic process [26] (Fig. 6A and Fig. S8A).Additionally, Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis showed enrichment of inflammatory signaling pathways and cytokinecytokine receptor interaction pathways (Fig. 6B).Further validation by ELISA indicated that IL-6 was the most dramatically downregulated cytokine in sh-circ-NOX4 CAFs (Fig. 6C).Previous studies have revealed that IL-6, as one of the key pro-inflammatory cytokines in the TME, plays a pivotal role in initiating pro-metastatic inflammatory responses and immune cell recruitment.These results encouraged us to focus on IL-6 for further investigation.
To confirm whether IL-6 is mainly secreted by the CAFs, we initially detected the mRNA level of IL-6 in both CAFs and cancer cells.The results revealed significantly higher expression levels of IL-6 in CAFs compared to A549 and PC9 cells (Fig. S8B).Subsequently, we knocked down IL-6 in CAFs or cancer cells separately and established a transwell coculture system to measure IL-6 secretion level (Fig. 6D and Fig. S8C).ELISA analysis indicated that knockdown of IL-6 in CAFs drastically reduced its secretion within the coculture system; however, knockdown of IL-6 in A549 or PC9 only slightly lowered its secretion within the coculture system (Fig. 6E, F).We also found that IL-6 was primarily derived from the tumor stroma and fibroblast subsets in both the human and murine TMEs through bioinformatics analysis (Fig. S8D-F).Thus, these findings support that CAFs are the primary source for secreting IL-6.
Then, we determined the role of the FAP/IL-6 axis in CAF activation.Single-cell analysis from the TISCH database indicated that both FAP and IL-6 were enriched in subsets of fibroblasts (Fig. S8G).Moreover, IL-6 was positively correlated with FAP expression in the TCGA database of LUAD and LUSC cohorts (Fig. 6G).Gene set enrichment analysis (GSEA) showed substantial enrichment signatures of IL-6 signaling in NSCLC patients with high FAP expression (Fig. 6H).Importantly, the IL-6 secretion level in circNOX4-overexpressing NFs was significantly increased compared with the corresponding control, while a FAP inhibitor (linagliptin, 5 nm) hindered the CAF-secreted IL-6 level upon circNOX4 overexpression, suggesting that circNOX4 induced IL-6 secretion via FAP (Fig. 6I).
To examine whether the pro-tumoral function of circ-NOX4 on cancer cells was mediated by stimulating paracrine IL-6, rescue experiments involving circNOX4 and IL-6 were carried out.The wound healing assay and transwell assay together revealed that neutralizing IL-6 in a conditioned medium from circNOX4-overexpressing NFs greatly abrogated the migration and invasion capacity of NSCLC cells (Fig. 6J, K).In contrast, the exogenous addition of recombinant human IL-6 (rhIL-6) partially reversed the effects of circNOX4 knockdown on the ability of NSCLC cells to invade (Fig. 6K).Collectively, these results demonstrate that circNOX4 elicits a pro-metastatic inflammatory microenvironment by accelerating IL-6 secretion within the fibroblast niche and consequently boosts NSCLC cell migration and invasion.
Blocking IL-6 restrains tumor progression in vivo
To investigate the potential therapeutic role of the circ-NOX4/IL-6 axis, we established xenograft tumor models and performed experimental therapy with IL-6 antibodies (Fig. 7A).A549 cells were co-injected with CAFs into the axilla of nude mice.After fourteen days after implantation, the mice received an intratumoral injection of IL-6 neutralizing antibody (2 mg/kg) twice a week.Conceivably, overexpression of circNOX4 enhanced CAFmediated tumor growth compared to the control group.Of note, injection of the anti-IL-6 neutralizing antibody led to a reduction in tumor volume and weight (Fig. 7B-D).In further IHC analysis, the expression levels of IL-6 were increased in circNOX4-overexpressing xenograft tumors, accompanied by the increased expression of FAP, CD31, N-cadherin, Vimentin and the matrix metalloproteinases MMP2 and MMP9.However, the expression of the metastasis-related factors was partly reversed by treatment with anti-IL-6 antibody (Fig. 7E and Fig. S9A, B), indicating that IL-6 acted as a soluble mediator of remodeling the local inflammatory niche and metastatic microenvironment.In conclusion, we demonstrate that circNOX4 activates an inflammatory fibroblast niche to promote tumor progression.Our findings emphasize that the circNOX4/FAP/IL-6 axis could be exploited as a therapeutic target for CAF-based cancer therapy (Fig. 8).
Discussion
Cancer metastasis is fostered by activated stroma, where specific microenvironments called niches are shaped [1].Cumulative studies have highlighted the important role of CAFs in establishing these niches that facilitate tumor growth and metastasis by secreting cytokines, chemokines, growth factors, and exosomes [2].However, how quiescent fibroblasts are educated into CAFs has not been extensively elucidated.In this study, we investigated the role and the underlying the mechanisms of FAP + CAF activation and inflammatory niche formation.Based on clinical samples from NSCLC patients, we found that FAP + CAFs correlate with distant metastasis and poor prognosis in NSCLC patients.Mechanistically, we determined that circNOX4 functions as a miR-329-5p sponge to upregulate FAP expression and induces NF to CAF phenotypic conversion.The activated CAFs subsequently enhances the production of IL-6.Since IL-6 has broad and robust effects on maintaining an inflammatory milieu and promoting cancer development, the in situ fibroblast niche boosts NSCLC metastasis and progression.Our findings highlight that targeting the circNOX4/ miR-329-5p/FAP/IL-6 axis might be a novel strategy for cancer therapy in NSCLC.
In this study, we determined that FAP correlates with tumor metastasis and shortened survival of NSCLC patients.As a critical contributor to fibroblast activation, FAP is specifically expressed in the tumor stroma, which has prominent clinical implications [27,28].For example, FAP-specific drugs have the potential to efficaciously target lesions in cancers [29].However, disappointing results were shown in the treatment of depleting FAPpositive cells in clinical trials of metastatic pancreatic and colorectal cancers [30,31].Therefore, exploring the upstream modulators of fibroblast activation is of paramount importance.
Recently, the role of circRNAs in mediating the initiation and progression of cancers has been increasingly emphasized [32].Kristensen et al. demonstrated that circCDR1 was highly expressed in tumor stoma while its expression was absent in cancer cells, putting forward the notion that the intratumoral heterogeneity of circRNA expression should not be neglected [33].However, the roles and biological significance of circRNAs involved in CAFs remain largely unknown.We thus explored the potential involvement of circRNAs in primary fibroblasts from human NSCLC tissues.By conducting a circRNA array, we identified a novel CAF-specific circRNA, circ-NOX4.Clinically, the high expression of circNOX4 in the tumor stoma was correlated with distant metastasis and worse OS in NSCLC patients.Our study links cir-cRNA expression in CAFs with the prognosis of NSCLC patients, suggesting the therapeutic potential of targeting circNOX4 in the TME.Moreover, we found that circ-NOX4 was responsible for the process of fibroblast activation.Gain-and loss-of-function experiments showed that knockdown of circNOX4 "normalized" CAFs to a quiescent phenotype and downregulated the expression of FAP, while overexpression of circNOX4 in NFs was sufficient to induce an activated fibroblast phenotype with higher FAP expression.Recent studies reported that miR-30a directly targeted FAP and suppressed its expression [34].Additionally, the transcription factor TCF21 reprogrammed FAP-high CAFs to a state that lacked pro-tumorigenic properties [13].Our study provides novel mechanisms by which circNOX4 increases the expression of FAP, the key marker of fibroblast activation, thereby contributing to the tumor-promoting phenotypes of CAFs.
The functions of circRNAs depend on their specific subcellular locations [23].Aiming to determine the molecular mechanisms of circNOX4, we detected its subcellular location in fibroblasts and discovered that it was mostly distributed in the cytoplasm.Cumulative studies have demonstrated that cytoplasmic circRNAs always sponge miRNAs and organize ceRNA regulatory networks to regulate gene expression [23].By cross-referencing the prediction results of the databases, miR-329-5p was identified as one of the candidate target miRNAs of circNOX4 and was possibly involved in CAF activation.We further confirmed the circNOX4/miR-329-5p/FAP axis by conducting qRT-PCR, FISH, luciferase reporter assays, RIP, and rescue experiments.In previous studies, miR-329-5p served as either an oncogene or a tumor suppressor in different cancers [35][36][37].A circRNA-mediated increase in miR-329-5p expression levels enhanced tumor proliferation and inhibited apoptosis in acute myeloid leukemia [38].However, the expression pattern, biological function and clinical significance of miR-329-5p in the tumor stroma have not been reported.Our study provides the first evidence that the circNOX4/miR-329-5p sponge in the ceRNA network is crucial for regulating FAP expression in CAFs, indicating a promising avenue for endogenously antagonizing FAP by inhibiting circNOX4.
CAFs support cancer cells through various paracrine pathways, including the secretion of inflammatory cytokines [39].As a versatile cytokine, IL-6 has broad and robust effects on tumor-promoting inflammation [40], the EMT process [41], and metastasis.IL-6 has also been recognized as a key marker of the inflammatory fibroblast subtype recently [42].In this study, we identified CAFs, rather than cancer cells, are the predominant source for secreting IL-6 in the local niche.We also determined that IL-6 as the downstream mediator of the circ-NOX4/FAP axis in CAFs, which contributed to NSCLC metastasis.This was further confirmed by the antitumor effects seen in vivo when blocking IL-6 with a neutralizing antibody, as multiple metastatic-related factors were suppressed.Likewise, researchers found that IL-6 produced by CAFs promotes chemoresistance through the JAK-STAT3 signaling pathway and that targeting IL-6 is a strategy to improve the therapeutic efficacy of chemotherapy in gastric cancer [43].It should be noted, however, that cytokines released by CAFs could circulate to distant organs and establish a premetastatic niche.Thus, it is possible that IL-6 fuels cancer cell dissemination from primary sites and colonization to distant organs, which requires further investigation.
Conclusion
In summary, we unveiled a novel mechanism by which a circRNA-induced fibroblast niche contributes to the inflammatory microenvironment and metastatic progression in NSCLC.circNOX4 upregulates FAP, a crucial CAF marker, by sponging miR-329-5p.Moreover, the activated phenotype of CAFs promotes tumor growth and metastasis by enhancing IL-6 secretion and establishing an inflammatory fibroblast niche.Our findings shed new light on the therapeutic potential of targeting the circNOX4/miR-329-5p/FAP/IL-6 axis to reprogram CAFs into a quiescent state and improve clinical outcomes in NSCLC patients.
(
See figure on next page.)Fig. 1 FAP expression in the TME of NSCLC and its clinical significance.A The mRNA expression levels of FAP in tumor tissues and normal tissues in the LUAD and LUSC cohorts from the TCGA database.B Distinct fibroblast clusters in NSCLC _GSE127465 database as visualized by single-cell analysis.C The expression of FAP in different fibroblast clusters.D The cell development pseudotime constructed by CellTracer tools (http:// bio-bigda ta.hrbmu.edu.cn/ CellT racer).The cells were ordered in a pseudo-temporal manner using Monocle2.E FAP expression in cellular developmental pseudotime in different branches.F The dynamic variation in functional activities along developmental pseudotime of branch 1.
Fig. 3 Fig. 4
Fig. 3 circNOX4 induces fibroblast activation and promotes the migration and invasion of NSCLC in vitro.A The proliferation of circNOX4-knockdown CAFs and circNOX4-overexpressing NFs as detected by CCK-8 assays.B Representative photographs of collagen gel contraction by the indicated CAFs and NFs.C The expression of CAF-related genes in sh-NC and sh-circNOX4 CAFs as detected by qRT-PCR.D Effects of circNOX4 on the expression of FAP by western blotting analysis.E Schematic representation showing that fibroblastic conditioned medium was added to NSCLC cells in wound healing assays.F Wound healing assays in A549 and PC9 cells treated with conditioned medium from circNOX4-knockdown CAFs or circNOX4-overexpressing NFs.G Schematic representation showing the establishment of the transwell coculture system.H Transwell coculture assays of invasion of A549 and PC9 cells with circNOX4-knockdown CAFs or circNOX4-overexpressing NFs.Data are expressed as mean ± SD from three independent experiments.Statistical analyses used the Student's t-test.*P < 0.05, **P < 0.01, ***P < 0.001
Fig. 5 Fig. 6
Fig. 5 Knockdown of circNOX4 mitigates CAF-mediated tumor growth and metastasis of NSCLC in vivo.A Schematic illustration showing A549 cells were subcutaneously co-injected with or without CAFs stably transfected with sh-NC or sh-circNOX4 into nude mice (n = 6 per group).Images of nude mice and xenograft tumors from each group were shown in the right panel.B The volume of subcutaneous xenograft tumors (n = 6).C Representative images of micro-PET/CT imaging of the tumor-bearing nude mice after application of 68 Ga-FAPI-04.3D MIP Images were collected over 1 h after the molecular imaging probe administration.D Quantified results from Micro-PET/CT images of the tumor to muscle ratio in each group.E Representative images of IHC for FAP, CD31, N-cadherin, Vimentin, MMP2, and MMP9 in xenograft tumors.Scale bar = 100 μm.F Schematic illustration of lung metastatic tumor model construction in nude mice.G The potential for lung metastasis was determined via vein tail injection of the indicated A549 cells.HE staining was used to observe the morphological changes in lung tissues.H Histogram analysis of the metastatic nodule number per lung.*P < 0.05, **P < 0.01, ***P < 0.001
Fig. 7
Fig. 7 Suppression of IL-6 effectively restrains the tumor progression in vivo.A Schematic diagram illustrating that A549 cells were subcutaneously co-injected with NFs stably transfected with vector or circNOX4 into nude mice, and IL-6 antibody was locally injected twice a week.B Images of nude mice and xenograft tumors of each group (n = 8).C Tumor growth curves were shown in different treatment groups.D Tumor weights were measured in different treatment groups.E Representative images of IHC for FAP, CD31, N-cadherin, Vimentin, MMP2, and MMP9 in xenograft tumors.Scale bar = 100 μm.*P < 0.05, **P < 0.01, ***P < 0.001
Fig. 8
Fig. 8 Schematic illustration visualizing the molecular mechanism of circNOX4 in activated fibroblast niche formation and NSCLC metastasis.The proposed model indicates circNOX4 is back-spliced by exons 6-11 of the NOX4 gene.The circNOX4/miR-329-5p/FAP ceRNA network in the cytoplasm reprograms the CAF phenotype and establishes an inflammatory fibroblast niche via secreting IL-6, ultimately resulting in the progression and metastasis of NSCLC
Table 1
Clinicopathological parameters of NSCLC patients and correlation with circNOX4 expression (n = 84) | 2024-03-10T05:09:46.573Z | 2024-03-08T00:00:00.000 | {
"year": 2024,
"sha1": "448f81a17e3aa6ecb90c7d935d4530267ca6b55b",
"oa_license": "CCBY",
"oa_url": null,
"oa_status": null,
"pdf_src": "PubMedCentral",
"pdf_hash": "448f81a17e3aa6ecb90c7d935d4530267ca6b55b",
"s2fieldsofstudy": [
"Medicine",
"Biology"
],
"extfieldsofstudy": [
"Medicine"
]
} |
215410187 | pes2o/s2orc | v3-fos-license | COVID‐19, chronicle of an expected pandemic
What is COVID‐19? What are the causes, parameters, and effects of this disease? What are the short‐ and long‐term prospects? Philippe Sansonetti, Infectious disease specialist and Chief Editor of EMBO Molecular Medicine, explains why the fate of the epidemic is in our hands.
W hat is COVID-19? What are the causes, parameters, and effects of this disease? What are the shortand long-term prospects? Philippe Sansonetti, Infectious Disease Specialist and Chief Editor of EMBO Molecular Medicine, explains why the fate of the epidemic is in our hands.
It is difficult to believe in plagues when they fall on your head (Albert Camus, The plague (1947, Ed. Gallimard)).
The COVID-19 pandemic is a plague as described by Camus; it is urgent and vital for our society to realize this. It is not too late, but time is running out. It is the third coronavirus-associated epidemic globally in < 20 years after SARS in 2003 and MERS in 2012. During these two epidemics, we were first worried, then reassured after it was over, and we did not much afterwards in terms of preparedness, therapy, and vaccine. Today, in the absence of treatment and vaccine, the evolution of the COVID-19 epidemic is nonetheless still in our hands.
Charles Nicolle (1866Nicolle ( -1936, professor at the Collège de France and director of the Institut Pasteur in Tunis, wrote in Destin des maladies infectieuses (1933): So, there will be new diseases. It's a fatal fact. Another fact, also fatal, is that we will never be able to detect them from their origin. Knowledge of infectious diseases teaches men that they are brothers and united. We are brothers because the same danger threatens us, united because contagion comes to us most often from our fellow men. We also, from this point of view and whatever our feelings are towards them, stand together with animals, especially domestic animals.
He was already predicting the emerging pathogens of the late 20 th century and 21 st century.
What are coronaviruses?
Coronaviruses are a huge family of singlestranded positive RNA viruses. Alpha-coronaviruses are endemic and infect mammals including humans and cause, for example, mild respiratory and intestinal diseases in children. Beta-coronaviruses like SARS-CoV-2 (the official name of COVID-19 virus) on the other hand are well adapted to their reservoir, the bat, but not to humans, which explains why human infections are so damaging. Other members of this family, the gamma-and delta-coronaviruses infect birds and fish and, at least for the time being, have not been associated with human disease.
The identification of the virus
One thing that this epidemic has taught us is that we, scientists now, possess the technology and capabilities to quickly react: As soon as the virus emerged in the City of Wuhan, China, deep sequencing and bioinformatics allowed for the diagnostic of this novel virus in a few days. The viral genomic sequence was then quickly shared and used to develop a detection test based on realtime PCR for rapid diagnosis and for initiating epidemiological studies all over the planet. Now compare this with the years it took to identify HIV thirty years ago through conventional growing of the virus in in vitro culture. Molecular diagnosis has revolutionized this field, and despite the initial delays in communicating about this epidemic, Chinese doctors and biologists quickly reported the first evidence for SARS-CoV-2, and provided the first sequences, clearing the way for the global scientific community to further develop diagnostic tools and engage in a race to discover dedicated drugs and vaccines.
Crossing the species barrier and human responsibility
The name coronavirus comes from the Protein S (S for spike) and its crown-like shape. The S protein plays a central role for host specificity and pathogenicity as it binds to the ACE2 receptor on the target cell surface and mediates entry of the viral RNA. The phylogenetic tree shows that SARS-CoV-2 is very close to SARS 2003 and MERS 2012, and the diseases they cause are very similar. The same is true for their origin: Bats are the main reservoir of coronaviruses. The phylogeny is therefore relatively well known, and one can apply the lessons learned during previous outbreaks, even if no vaccine or therapy is available so far.
SARS-CoV-2 is a textbook case of an emergent pathogen after a species jump ("zoonosis"). We have observed these zoonoses for decades, particularly in tropical regions (Ebola is one such example; Marí Saéz et al, 2015). Generally, an emergent pathogen can give rise to two scenarios.
1
The virus is ill-equipped and has little capacity to mutate and therefore to adapt and stabilize in the new species. Human infection is abortive as the virus is not able to establish human-tohuman transmission. Still, under this scenario, the disease can be very severe as was the avian influenza H5N1 caused by direct bird-to-human transmission, reaching~60% mortality, but for which limited, if any, human-tohuman transmission was reported (Wang et al, 2008).
2
The virus is better adapted to its new host, thus better able to cross the species barrier, and its adaptation is facilitated by further mutations due to poor fidelity of the enzyme that replicates the positive strand of viral RNA. This is what happened with SARS-CoV-2, which jumped relatively easily from bats to humans via an intermediate mammal reservoir. The disease caused by SARS and MERS was more serious, and the mortality rates were higher (10% and 35%, respectively). SARS-CoV-2 however is often associated with milder pathologies. This is a typical trade-off: less virulent, but more transmissible versus more virulent but less transmissible. This balance is extremely important and defines the disease profile.
The natural reservoir is a certain bat species; it is impressive to see how far these animals are capable of carrying these emerging viruses, such as corona. This was the case with Ebola in Africa, and with the Nipah virus in Malaysia, which appeared in the late 1990s. As human behavior changes ecological conditions, these bats increasingly come into contact with animals which are susceptible for a zoonotic species jump and which further replicate the virus. This generates a risk zone around humans, since any contact with these intermediate animals can cause another species jump into humans. In the case of SARS 2003, it is assumed that the intermediate animal was the webbed civet, a feline particularly frequent in Asia; for MERS, it was the camel. For Ebola, it probably was great apes (Malvy et al, 2019) while direct bat-to-human transmission was also considered (Marí Saéz et al, 2015).
For SARS-CoV-2, some claim that the intermediate reservoir is the pangolin. Indeed, numerous studies have shown that the virus circulating in pangolins is very close to the one observed in humans, suggesting that the current pandemic results from human behavior and animal trafficking.
We are constantly threatened by these emerging diseases so-called anthropocene diseases, which are essentially or even exclusively linked to the omnipresent imprint that humans leave on the planet. What is true for the climate and the environment is also true for infectious diseases, in particular emerging ones, and the three are linked.
Pandemic spread of the virus
This is a story in three acts: (i) stochastic species jumps, (ii) possible transmission to and between human, and (iii) pandemic explosion. It is interesting to note that the distribution map of COVID-19 clusters and hotspots overlaps with the map of intercontinental air flights (which carried 4 billion passengers in 2019). We can clearly see the role transport plays in the transmission and global spread of these diseases, turning them into pandemics more readily than even before in human history. However, as we speak, more and more cases are being reported in the countries of the intertropical zones, reaching Southern America and Africa where containment will be difficult to achieve.
There are also important environmental aspects linked to the Anthropocene era, such as intensive agriculture and husbandry that profoundly modify ecosystems and cause unanticipated encounters between animal and human habitats, rising temperatures in equatorial regions that drive animal population migration to cooler regions and political instability that causes mass displacement of refugees. Also, different national healthcare systems and a lack of global information exchange and coordination can lead to a delay in diagnosis in some countries, but the comparison of thefts and foci of infection is striking. In how far these effects had an influence on the COVID-19 pandemic requires further study once the pandemic has passed its apex.
Parameters of the epidemic R0 (basic reproduction rate) is the average number of secondary infections produced when an infected individual is introduced into a population where all individuals are susceptible. If the R0 is less than 1, there is no epidemic situation; as soon as it is greater than 1, there is an epidemic. In the case of COVID-19, this number is between 2 and 3. This is therefore a typical epidemic situation. For the Spanish flu of 1918-1919, the R0 was 2.3; tuberculosis is 10 and therefore extremely contagious; and for measles, R0 ranges between 12 and 18.
The incubation period is 5-6 days, and all evidences converge to indicate that the patients are contagious from the onset, when they are yet asymptomatic, unlike SARS in 2003 where the contagion only appeared with the peak of viremia after several days after infection. The SARS-CoV-2 virus is highly contagious: People transmit while they are still asymptomatic, or just begin to experience mild symptoms.
The attack rate, the number of newly infected patients compared to the general naive population, is high, much higher than the seasonal flu. Severe cases are around 10-15% and require hospitalizations that last on average between 7 and 15 days, which is what threatens our health systems.
COVID-19 is therefore a disease with high epidemic potential that puts a major strain on the healthcare system. This is why governments decided to put in place different strategies to attenuate the progression of the disease.
The mortality rate is relatively low. When we will be able to get a full picture of this pandemic, we will most likely notice that it was between 1 and 2%. It seems higher during periods of exponential spread as is currently the case in the USA, France, Spain, or Italy. Not necessarily because the disease is more severe during this period, but likely because death count is indisputable, while it is difficult to assess the number of infections, which is always higher than observed. Undoubtedly, the majority of infected people develop a mild form of this disease, which makes accurate calculation of the mortality rate difficult at present. At the time of publication, mass serological testing is being rolled out in countries such as Germany, particularly in hotspot areas, which will give a much more accurate estimate of mortality rates. But the mortality rate increases due to the stress on the healthcare system and the availability of hospital beds. 1% mortality, 10% of severe cases, those are not large numbers statistically speaking, but compared to the number of infections, and taking into account the transmissibility and infectivity of the virus, we are reaching a situation that endangers our healthcare system. This is what legitimizes the flatten-the-curve policy. Yet, one good news: children younger than 10 years-old are mostly not ill even though they can get infected and transmit the virus.
Eco-pathology of beta-coronaviruses
The receptor for SARS-CoV-1 and SARS-CoV-2, angiotensin II converting enzyme (ACE2), is an enzyme attached to the surface of cells, including lungs, pneumocytes and endothelium, endocardium, kidney, liver, and intestine. It came as a surprise that the virus binds to an important enzyme for the regulation of blood pressure. This could explain the severity of the disease in the respiratory tract with instances of pneumonia and ultimately, acute respiratory distress syndrome (ARDS) observed in elderly subjects or individuals with chronic comorbidities (i.e., diabetes, hypertension, cardio-respiratory insufficiency, chronic immunosuppression). The virus targets the blood-lung barrier, affecting the oxygen exchange zone.
Acute respiratory distress syndrome has classic inflammatory signs, a so-called "cytokine storm" characterized by a significant increase in cytokines and proinflammatory chemokines. What is less typical for corona infections is the destruction of the alveolocapillary barrier, which seldom happens also in COVID-19, but when it does, requires fast intervention. ARDS can also occur in younger individuals during the healing phase, which, in this case though, may be linked to the immune response.
The adaptive immune response specific to the virus is still poorly understood. We know that this virus affects IFN-driven immunity, but we do not know why, or which effectors blunt this immune response. It is possible that the virus has acquired this strategy in bats that express constitutively low level of IFN. Further work is urgently needed to learn to control the disease efficiently and sustainably.
The future is in our hands
The future of the COVID-19 pandemic is in our hands. So far, we can only rely on prevention and symptomatic treatment of severe forms. Regarding prevention, health authorities have three main options. The first-which may seem cynical-considers that the more people become infected, the quicker the population will become immune until the epidemic peters out as the virus finds no more immunologically naive individuals. This is the principle of group/herd immunity. If 60% of the population were infected by SARS-CoV-2, the epidemic will stop. But this would come at the cost of a very brutal epidemic peak with a high number of severe cases and deaths. We saw it during the epidemic of the Asian flu in the United Kingdom in 1957. For a week to ten days, the health system imploded, because the health personnel were sick, the equipment insufficient, and the number of seriously ill patients exploded. Totally opposite is the "Chinese approach": to isolate cities and individuals, which undeniably appears effective in controlling the epidemic. The risk is that because only few individuals have been infected, the attack rate was reduced due to strict confinement, hence many individuals are still immunologically naive and vulnerable to a second viral wave and risk an epidemic rebound. The intermediate option, which has been taken in most of European countries, is to flatten the curve over time with different levels of isolation and social distancing, hoping that even if less than 60% of the population will be infected, above all, the healthcare system will hopefully be preserved. The approach is based on a concept of social distancing (two meters/6 feet apart) and individual hygiene. It seems that in addition to airborne contamination, contaminated hands are also a major vector, either by direct contact with an infected patient, or indirectly by contact with a contaminated surface on which the virus seems to be able to survive for many hours. So, no kiss, no hug, no handshake and absolute hand hygiene. Avoid touching face unless hands have been thoroughly washed with soap or decontaminated with a hydro-alcoholic gel. The importance of wearing masks for the general population is still debated in some countries that quickly ran short of these masks. In agreement with attitudes in Asian countries, wearing masks is now strongly advised to complement hygiene measures. Common sense starts to prevail.
Unfortunately, the measures that have been implemented so far have been clearly insufficient, as we have seen in Italy, Spain, France, UK and as we now see in the USA. Confinement and isolation at home have therefore been decided for an extended period at the cost of a huge economic risk, given the intrinsic dynamics of the epidemic. Failure of these measures in these countries will have to be compared with their success in Asian countries like South-Korea, Taiwan, Singapore, and to a certain extent in Germany. Large implementation of diagnostic combined with isolation of SARS-CoV-2 positive individuals may be part of the answer. No matter the strategy taken by health authorities, we need to convince ourselves that our fate is in our hands. We stepped into another world in only a few weeks. What we valued yesterday-our daily activities, our hobbies, our workmust be weighted in relation to the gravity of the situation.
Treatments
Finally, it is essential to find antiviral treatments for severely affected patients and to block transmissibility from individual to individual. A few options are as follows: (i) "repositioning" of drugs already tested for other viruses, such as HIV, or drugs that are already commercially available; (ii) better understanding of the pathophysiology of ARDS in order to develop a dedicated pharmacology using repositioned and novel molecules; and (iii) developing an effective vaccine. The risk of rebounds justifies the absolute necessity of a vaccine that is the only way to achieve durable elimination of COVID-19 and even eradication of SARS-CoV-2, including in low-income countries with fragile health systems. Design, development, validation, clinical studies, and registrations with regulatory agencies takes between 8 and 12 years on average for a standard vaccine. This is incompatible with the urgency of an emerging epidemic like COVID-19. Fortunately, novel paradigms of vaccine discovery and development are on their way to considerably shorten the delays. Nonetheless, we know that we would not have a vaccine available until at least next year. In the meantime, in spite of enormous difficulties, we need a global mobilization for more basic and applied research from academia and the pharmaceutical and vaccine industry. As much as we have been able to massively reduce the time to identify novel pathogens, we need to reduce the time to develop the tools of their control. Again, our future is in our hands.
Adapted from an article published in French in laviedesidees.fr, March 2020. | 2020-04-08T19:10:43.303Z | 2020-04-07T00:00:00.000 | {
"year": 2020,
"sha1": "934e504873afbb0a139e823670c69b47c9d4a877",
"oa_license": "CCBY",
"oa_url": "https://onlinelibrary.wiley.com/doi/pdfdirect/10.15252/emmm.202012463",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "e11c13d974ebb0001f6b7bde59c0cfd1e7556568",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
253044914 | pes2o/s2orc | v3-fos-license | Neuroprotective Effect of miR-483-5p Against Cardiac Arrest-Induced Mitochondrial Dysfunction Mediated Through the TNFSF8/AMPK/JNK Signaling Pathway
Substantial morbidity and mortality are associated with postcardiac arrest brain injury (PCABI). MicroRNAs(miRNAs) are essential regulators of neuronal metabolism processes and have been shown to contribute to alleviated neurological injury after cardiac arrest. In this study, we identified miRNAs related to the prognosis of patients with neurological dysfunction after cardiopulmonary resuscitation based on data obtained from the Gene Expression Omnibus (GEO) database. Then, we explored the effects of miR-483-5p on mitochondrial biogenesis, mitochondrial-dependent apoptosis, and oxidative stress levels after ischemia‒reperfusion injury in vitro and in vivo. MiR-483-5p was downregulated in PC12 cells and hippocampal samples compared with that in normal group cells and hippocampi. Overexpression of miR-483-5p increased the viability of PC12 cells after ischemia‒reperfusion injury and reduced the proportion of dead cells. A western blot analysis showed that miR-483-5p increased the protein expression of PCG-1, NRF1, and TFAM and reduced the protein expression of Bax and cleaved caspase 3, inhibiting the release of cytochrome c from mitochondria and alleviating oxidative stress injury by inhibiting the production of ROS and reducing MDA activity. We confirmed that miR-483-5p targeted TNFSF8 to regulate the AMPK/JNK pathway, thereby playing a neuroprotective role after cardiopulmonary resuscitation. Hence, this study provides further insights into strategies for inhibiting neurological impairment after cardiopulmonary resuscitation and suggests a potential therapeutic target for PCABI. Supplementary Information The online version contains supplementary material available at 10.1007/s10571-022-01296-3.
Introduction
The consequences of cardiac arrest (CA), including brain injury after resuscitation, are often fatal. It has been reported that only 3-7% of recovering re-establish the neurological state they exhibited before cardiac arrest (Lilja et al. 2015). There is a need to improve neurological outcomes and the quality of life for cardiac arrest survivors as more people survive cardiac arrest. Neuronal mitochondria play central roles in the pathophysiological processes of cardiopulmonary resuscitation (Lou et al. 2021). Ischemia-reperfusion injury is primarily induced by mitochondrial reactive oxygen species (ROS) generated after cardiac arrest. Inhibiting mitochondrial ROS levels reduced neuronal myocardial oxidative stress after the return of spontaneous circulation (ROSC) (Jeggo et al. 1985;Wang et al. 2021). Mitochondria can also control cell death via the induction of apoptosis Qiang Zhang and Haohong Zhan have contributed equally to this work. by releasing cytochrome c after injury (Nhu et al. 2021). Adaptive responses play a role in maintaining mitochondrial function when neurons are exposed to oxygen and glucose deprivation and reoxygenation (OGD/R). These strategies include mitochondrial fission and fusion, mitophagy, and mitochondrial biogenesis, which restore neurovascular homeostasis . Therefore, preventing mitochondrial injury after cardiopulmonary resuscitation may be an important strategy for reducing neurological damage after cardiopulmonary resuscitation.
MiRNAs are short nonprotein-coding RNA molecules that are evolutionarily conserved and ubiquitously expressed. Numerous studies have reviewed the role of miRNAs in regulating neuronal apoptosis, regeneration, the plasticity of neurons, and inflammation after cardiac arrest (Roitbak 2020). Due to the stability of miRNAs in the bloodstream and their function in regulating neurological impairment after ischemia-reperfusion injury, miRNAs have been the most promising newly discovered biomarkers and therapeutic targets of postcardiac arrest that may be used to alleviate neurological impairment (Stammet 2017). In this work, we found that the level of miR-483-5p was correlated with the prognosis of neurological dysfunction through Gene Expression Omnibus (GEO) database screening. However, it is unclear whether miR-483-5p affects the fate of nerve cells after cardiopulmonary resuscitation. The literature shows that the abnormal expression of miR-483-5p is associated with the quality control of mitochondria and with nervous system diseases (Nagaraj et al. 2021;Vazquez et al. 2013). Fan et al. found that miR-483-5p determined mitochondrial fission in tongue squamous cell carcinoma by targeting FIS1 (Fan et al. 2015). Additionally, miR-483-5p upregulated peroxisome proliferator-activated receptor gamma (PPARγ) in stromal cells , and this effect was closely related to mitochondria and neuroplasticity (Cheng et al. 2010).
TNFSF8 (also known as CD30L and CD153) is a member of the tumor necrosis factor receptor superfamily (TNFSF) (Shinoda et al. 2015) and is primarily expressed in immune system cells (Sun et al. 2013). Similar to TNF-a, TNFSF8 is a product of activated monocytes/macrophages and shows classical pleiotropic cytokine activities. Here, our study found that TNFSF8 was also expressed in neurons and it is the target gene of miR-483-5p. The prominent role of TNFSF in neurological disorder has been already demonstrated in different studies, targeting the miRNA-155/ TNFSF10 network restrains inflammatory response in the retina of Alzheimer's disease, TNFSF12 could promote mitochondrial fusion (Burgaletto et al. 2021;Shunkina Skuratovskaia et al. 2021). In summary, the causal mechanisms are unclear even though the potential role for miR-483-5p in nervous system function after ischemia-reperfusion injury has been found to be related to mitochondrial dynamics.
Mitochondrial biogenesis is part of mitochondrial quality control and is closely related to oxidative stress levels, mitochondrial-dependent apoptosis after ischemia-reperfusion (I/R) injury. The activation of mitochondrial biogenesis protects neurons from damage (He et al. 2020a, b). Many cell types, tissues, and disease states have been associated with aberrant AMPK/JNK translocation, including oxidative stress, apoptosis, autophagy, and cell death Zhang et al. 2020a, b). However, further study is needed to explore the role played by the AMPK/JNK pathway after cardiac arrest, as miR-483-5p may function in regulating mitochondrial quality control (Hayakawa et al. 2016). The present study was focused on the relationship between mitochondrial dysfunction after cardiac arrest and neuroprotective effects conferred by miR-483-5p and the role that the TNFSF8/AMPK/JNK signaling pathway may play in this process.
Microarray Data Source Identification of mDEGs
The expression profile data in this study were downloaded from the GEO public database (http:// www. ncbi. nlm. nih. gov/ geo/); GSE74198 and GSE34643 are miRNA datasets related to the prognosis of neurological function after cardiac arrest. The cerebral performance category (CPC) score has been widely used to evaluate the neurological function of patients with craniocerebral injury. In this study, a CPC score 1-2 was considered an indicator of a good neurological function, and a CPC score of 3-5 was considered an indicator of poor neurological function. The datasets were analyzed with the DEq2 package and limma R package in R language (4.0 vision). The miRNAs were differentially expressed in patients with good and poor outcomes in the two datasets. The screening conditions were set as a P value < 0.05 and an absolute value of LogFC > 0.7.
RNA-Seq Analysis and Functional Enrichment Analysis
Using Illumina HiSeqTM 2000, raw reads in fastq format were processed. Clean data were acquired by removing reads containing poly-N in the raw data, adapters, and low-quality sequences. The expression level of mRNA was calculated using RSEM (RNA-Seq by Expectation Maximization) (v1.3.1) and normalized to FPKM (fragments per kilobase per million reads) (Li and Dewey 2011). The sequencing results were stored with the GSE209531 database (https:// www. ncbi. nlm. nih. gov/ geo/ query/ acc. cgi? acc= GSE20 9531). Edge R software was used to analyze differential mRNA expression between the two groups (Robinsonet al. 1 3 2010). Differentially expressed mRNAs were identified on the basis of a | log2ratio |> = 1 and P value < 0.05. Enrichment analysis of Gene Ontology (GO) categories and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were performed using WebGestalt online software (http:// www. webge stalt. org). The WebGestalt database was used to annotate and visualize GO terms and KEGG pathways (Zhong et al. 2022). We identified independent predictors of a neurological outcome using binary logistic regression. The sva package with the ComBat function was used for removing batch effects in the two databases. Then, the four selected miRNAs were subjected to logistic regression analysis.
Animal Care and Treatment Groups
This research was approved and supervised by the Institutional Animal Care and Use Committee of Sun Yat-sen University (Approval number: SYSU-IACU-2022-001232). As progesterone is a neuroprotective steroid, it reduces proinflammatory cytokine expression and improves neurological outcomes after cardiac arrest (Espinosa-Garcia et al. 2020).
In order to eliminate the interference of progesterone, sixty male adult Wistar rats (350-400 g) were obtained from the Animal Experimental Center of Southern Medical University. Food and water were available ad libitum, and three animals were housed in each cage. The animals were randomly divided into five groups: sham group, CA+miR-NC group, CA+miR-Sponge NC group, CA+miR-483-5p group, and CA+Sponge miR-483-5p group. According to the manufacturer's instructions, an adeno-associated virus-9 (AAV-9, Hanheng, Shanghai, China) system was used to enhance or inhibit the expression of miR-483-5p. A nontargeting scrambled negatively matched shRNA was used as the control. After adequate anesthesia with pentobarbital (40 mg/ kg), the animals were placed on a stereotaxic apparatus. A total of 10 μL of the virus were injected intraventricularly at a rate of 0.3 μL/min 3 weeks before the establishment of CA models. Lateral ventricle injection localization was performed as followed: bregma: 1.0 mm lateral, 1.5 mm posterior, and 3.1 mm below the dural surface. Four weeks later, the rats were weighed and anesthetized with 4% sodium pentobarbital.
Cardiac Arrest Model
Our study involved a 10-min asphyxiation-induced cardiac arrest and cardiopulmonary resuscitation model (Geocadin et al. 2002). When the rats reached 350 g, they were anesthetized with 3% pentobarbital (40 mg/kg) prior to intubation. Intravascular catheters (PE50, Smiths Medical, Ashford, UK) were inserted into the right femoral artery and vein for dynamic blood pressure monitoring and drug delivery. After the connection of a cardiac electrical monitor, the rats were given an intravenous injection of vecuronium (MERCK, Cat. 50700-72-6, Darmstadt, Germany, 3 mg/kg). CA was defined as an average arterial pressure < 30 mmHg. After 10 min, the rats were given extracardiac compression (200 times/min) with ventilator-assisted ventilation (tidal volume: 5 mL, 80 times/min). Then, the rats were given an intravenous injection of adrenaline 0.02 mg/kg once every 3 min until the circulation in the rats returned to normal. Return of spontaneous circulation was defined as the rise of the mean arterial pressure (MAP) to at least 60 mmHg for more than 10 min. The ventilator was removed after the rats resumed spontaneous breathing. In the sham group, the rats were treated with pentobarbital (40 mg/kg) and intubated and then, the femoral vein and femoral artery were catheterized.
Live/Dead Cell Staining
We used a Live/Dead Cell Double Staining Kit (APExBIO, Cat. K2081, USA) to detect the effects of miR-483-5p on PC12 cells. PC12 cells were cultured in six-well plates for 24 h and then transfected with miRNA mimics or miRNA 1 3 inhibitors. After OGD/R induction, samples were washed twice with phosphate-buffered saline (PBS) and incubated for 20 min in the dark with a 5 μmol mixed staining solution consisting of Calcein AM-PI. They were then washed with PBS three times, and fluorescence microscopy was performed. Red fluorescence represents dead cells, and green fluorescence represents living cells. We counted the dead cells and determined their proportion among all cells.
Western Blotting and Isolation of Mitochondria
To obtain proteins, PC12 cells and hippocampal samples were collected and homogenized in RIPA lysis buffer (Thermo Fisher, Cat. 89901, IL, USA). The supernatant was collected after centrifugation at 14,000×g for 25 min at 4 °C. Protein concentration was quantified using a BCA assay kit (Thermo Fisher, Cat. 23225, IL, USA). Equal amounts of protein (40 μg) were electrophoresed on 8-12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gels and then transferred to 0.2-μm polyvinylidene difluoride (PVDF) membranes (Millipore, Cat. ISEQ00010, CA, USA), which were blocked with 5% nonfat milk and incubated overnight with primary antibodies at 4 °C. The primary antibodies were as follows: anti- (1:1000, Abcam, Cat. ab154856, USA), and anti-β-actin (1:1000, Abcam, Cat. ab8227, USA). The next day, after incubation with the HRP-conjugated secondary antibody for 1 h at room temperature, the membranes were detected with a chemiluminescence system (GEAI600, MA, USA). Immunoblots were visualized with a BeyoECL Star chemiluminescence reagent kit and quantified by densitometry using ImageJ software. VDAC1 and β-actin were used as the internal controls. Mitochondrial proteins were extracted from tissue using a mitochondria isolation kit (Beyotime, Cat. C3606, Shanghai, China) according to the manufacturer's instructions.
Transmission Electron Microscopy (TEM)
Following CA surgery, rats underwent transcardial perfusion with precooled PBS, followed by treatment with 50 mL of 4% paraformaldehyde. Hippocampal tissue was removed, cut into 1-mm 3 pieces, and maintained in 4% glutaraldehyde to observe the ultrastructure of the mitochondria. Four regions were randomly selected from three slices of each sample and viewed under ×100,000 magnification. The aspect ratio and proportion of individual mitochondrial ridge areas were measured as morphological parameters using ImageJ software (NIH, Bethesda, MD, USA). The aspect ratio was defined as the ratio of the major axes to the minor axes of the analyzed mitochondria. The proportion of an individual mitochondrial ridge area was defined as the ratio of the area of the mitochondrial ridge to the total mitochondrial area.
RNA Extraction, Reverse Transcription, and Quantitative RT-PCR
Following the manufacturer's instructions, total RNA from cells and fresh biopsied tissues was extracted with TRIzol reagent (Invitrogen, Cat. 15596108, CA, USA), and the RNA concentration was measured with a Nanodrop 2000 instrument. The RNA was transformed to complementary DNA (cDNA) with a PrimeScript RT Reagent Kit (Takara, Cat. RR047A, Tokyo, Japan). TB Green™ Premix Ex Taq™ II (Tli RNaseH Plus, Takara, Cat. RR820Q, Tokyo, Japan) was used with a CFX96 Real-Time PCR Detection System (Bio-Rad) to evaluate the relative mRNA levels.
Neurological Function Score and Water Maze Experiment
A blinded investigator assessed neurological function using established neurological deficit scores (NDSs). The NDSs of the surviving rats were evaluated 24 h after CA. We evaluated the degree of arousal, cranial nerve reflexes, motor assessment, sensory assessment, motor behavior, seizure, and simple behavioral responses and generated the overall NDS. A score of 80 represents normal brain function and 0 represents brain death. Low scores indicated poor performance. Additionally, Morris water maze tests were performed to assess spatial reference memory. The Morris water maze consisted of a black circular pool (160 cm in diameter and 55 cm high) filled with ink-stained water in which a transparent escape platform (diameter 10 cm) above the water surface was placed in one of the quadrants (the target quadrant). The swimming activity data were collected via a digital video camera placed above the pool, and the data were analyzed with a DigiBehave system (Jiliang Software Company, Shanghai, China). The Morris water maze test consisted of probe and visible platform tests. All rats were given five trials daily for seven consecutive days before CA to ensure that they could find the escape platform within 80 s. There was no significant difference in the time in which the rats in each group found the platform before CA. On Day 7 after CA, the probe test was performed to assess spatial memory retention within 80 s.
TUNEL Staining
An In Situ Cell Death Detection Kit, TMR red (Roche, Cat. 11684817910, Penzberg, Germany) was used according to the manufacturer's protocol for terminal deoxynucleotidyl transferase-mediated dUTP nick end-labeling (TUNEL) staining. PC12 cells were fixed with 4% paraformaldehyde for 30 min. The TUNEL solution was added and incubated at 37 °C for 1 h. After the solution was removed, and anti-fluorescence attenuating tablets containing DAPI were added to the cells. The rats were euthanized at 24 h after the operation. The paraffin-embedded hippocampal sections were heated at 60 °C, washed in xylene, and rehydrated with a graded series of ethanol and double-distilled water. Sections were placed in a plastic jar containing 200 mL of 0.1-M citrate buffer, and 300-W microwave irradiation was applied for 5 min. After rinsing twice with PBS, the sections were incubated in a dark, humidified chamber with the TUNEL reaction mixture for 1 h at 37 °C. Then, the sections were washed with PBS and incubated for 30 min. After washing, they were stained with 4′-6-diamidino-2-phenylindole (DAPI) for 15 min. The number of neurons in the stratum pyramidale (200 μm in length) within the CA1 region of the hippocampus was analyzed with a fluorescence microscope equipped with the Image-Pro Plus System (Media Cybernetics, USA). The number apoptotic neurons induced by ischemia-reperfusion was determined by comparing the number of TUNEL-positive cells with the number of DAPI-positive cells.
HE Staining and Nissl Staining
The morphological changes in the hippocampal CA1 area of the rats were determined via HE staining and Nissl staining. Rats were euthanized 24 h after cardiopulmonary resuscitation. Brain tissue was routinely embedded in paraffin and cut into 10-μm coronal sections. Sections of brain tissues were stained with HE or Nissl for routine histological examination, and the morphological changes were observed with a light microscope (IX81, Olympus, Tokyo, Japan).
Flow Cytometry Annexin V-FITC-PI Assay
Cultured PC12 cells were washed with PBS and then digested using EDTA-free trypsin after appropriate treatment. Cell suspensions were collected after centrifugation and suspended in 100 μL of 1× buffer. A FITC Annexin V Apoptosis Detection Kit (BD Pharmingen, Cat. C1062M, Shanghai, China) was used to detect the ratio of apoptotic cells to total cells. Propidium iodide and Annexin V-FITC (5 μL of each) were added to the solution and gently mixed. The final suspensions were then poured into flow cytometry tubes for flow cytometry examination.
Measurement of Mitochondrial Membrane Potential
A JC-1 mitochondrial membrane potential assay kit (Beyotime, Cat. C2003S, Shanghai, China) was used to measure the mitochondrial membrane potential (MMP) according to the manufacturer's protocol. JC-1 is a cell-permeable fluorescent dye that accumulates within mitochondria in a potentialdependent manner. Images were taken using an EVOS FL Imaging System. Alteration in the ionic equilibrium resulted in mitochondrial depolarization as indicated by a decrease in the red/green fluorescence intensity ratio. The ratio of fluorescence red intensity/green fluorescence intensity was analyzed using ImageJ software, and the value was calculated 1 3 on the basis of the control group. Changes in mitochondrial transmembrane potential in the hippocampus were analyzed for rats in each group using 50 mg of fresh hippocampal tissues cut into 1-mm 3 pieces in PBS and subsequently filtered through 200-mesh sieves. The filtrate containing the brain cells was collected and treated with the JC-1 working fluid. After incubation for 15-20 min at 37 °C in a 5% CO 2 incubator, the hippocampal cells were centrifuged (2000 RPM, 5 min), washed once, and suspended in PBS. We used a flow cytometer (CytoFLEX, Beckman Coulter, CA, US) (Ex = 488 nm; Em = 530 nm) to detect the MMP. Green fluorescence was detected through FITC channel FL-1, and red fluorescence was detected through PI channel FL-2. Apoptosis was indicated by an increase in the green fluorescence intensity/red fluorescence intensity ratio (Y. Chen et al. 2017).
ROS Determination by Flow Cytometry
Intracellular ROS levels in PC12 cells were detected with the oxygen-free radical sensitive probe DCFH-DA (Beyotime, Cat. S0033S, Shanghai, China). After OGD/R, the culture medium was discarded and basal medium containing 5-μM DCFH-DA was added to the plates and incubated for 30 min and then, the cells were digested with trypsin. The relative fluorescence intensity was measured by flow cytometry (excitation and emission at 488 nm and 530 nm, respectively). Rat brain homogenates were prepared according to a previously reported method (Jiang et al. 2010), diluted with ice-cold Trisbuffered saline (TBS) at a 1:10 ratio, and incubated for 45 min at room temperature with DCFH-DA (5 mM).
Oxidative Stress Assay
The supernatant of the PC12 cells and hippocampal homogenates were collected to assess the levels of oxidative stress. A lactate dehydrogenase (LDH) Cytotoxicity Detection Kit (Thermo Fisher Scientific, Cat. C20301, IL, USA) was used to determine intracellular LDH levels. In addition, the supernatant and homogenates were used to investigate both malondialdehyde (MDA) content and superoxide dismutase (SOD) activity using a SOD assay kit (Beyotime, Cat. S0101M, Shanghai, China) and Cell MDA assay kit (Beyotime, Cat. S0131S, Shanghai, China), respectively. Hippocampal ROS levels were measured with a Reactive Oxygen Species (ROS) Fluorometric Assay Kit (Beyotime, Cat. S0033S, Shanghai, China) on a microplate reader.
Statistical Analysis
SPSS 19.0 (SPSS, IL, USA) software was used for data analysis. Figures were drawn using GraphPad Prism 9 software.
All the data are expressed as the mean ± SD. A t test was performed to compare two groups, and a one-way analysis of variance (ANOVA) was performed for group of three or more. A P value < 0.05 was considered to be statistically significant.
High miR-483-5p Expression Was Associated with Neurological Prognosis After Cardiopulmonary Resuscitation and Alleviated Cell Injury After Ischemia-Reperfusion
Research has demonstrated that miRNAs play active roles in CA showing predicting capacity and potential therapeutic applications (Stammet 2017). To identify the miRNAs involved in regulating neurological prognosis, the GSE74198 and GSE343634 datasets containing miRNA expression profiles of the peripheral blood of patients after cardiac arrest were downloaded from the GEO databases, and data on differentially expressed miRNAs between patients with good and poor neurological prognosis were obtained. We found that four miRNAs (miR-574-5p, miR-671-5p, miR-320c, and miR-483-5p) were identified as the differentially expressed miRNAs that were upregulated in both datasets (Fig. 1A, B). The results of logistic regression analysis demonstrate that miR-483-5p exerted the strongest effect as an independent predictor of neurological prognosis (P = 0.002, hazard ratio = 2.381, Fig. 1C). Then, we measured the expression of miR-483-5p after OGD/R in PC12 cells. The level of miR-483-5p increased significantly at 2 h after ischemiareperfusion but decreased at 6-12 h after reperfusion. The expression level gradually returned to normal after OGD/R for 24 h (Fig. 1D). A cell viability assessment showed that overexpression of miR-483-5p (Fig. S1) reduced the extent of cell injury, suggesting that miR-483-5p mimics exert a protective effect against ischemia-reperfusion injury (Fig. 1E). In contrast, inhibition of miR-483-5p expression led to the opposite effect (Fig. 1E). These protective effects were also confirmed by live/dead cell staining (Fig. 1F).
miR-483-5p Attenuates Ischemia-Reperfusion-Induced Neuronal Apoptosis
As demonstrated by the CCK-8 assay and live/dead cell staining assay, miR-483-5p mimics significantly increased cell viability and decreased mortality in PC12 cells. To further characterize the function of miR-483-5p, we explored the effectiveness of miR-483-5p in regulating apoptotic cell death, which plays a decisive role in determining the prognosis of neurological function after cardiopulmonary resuscitation (Wang and Bennett 2012). The TUNEL assay revealed that overexpression of miRNA-483-5p attenuated 500 μm, n = 4. The data are expressed as the mean ± SD, **P < 0.01, ****P < 0.001 1 3 hypoxia-induced PC12 cell apoptosis; a miR-483-5p inhibitor reversed the effect of the miRNA on apoptosis (Fig. 2A). The apoptosis rate observed by TUNEL was consistent with that observed by flow cytometry assay (Fig. 2B). Cleaved caspase 3, the active form of Caspase 3, is a marker of apoptosis. Bax is a proapoptotic protein related to the mitochondrial apoptosis pathway. As shown by western blot assay, the protein levels of cleaved caspase 3 and Bax were increased after OGD/R, and the administration of miR-483-5p mimics significantly decreased their levels. After transfection with miR-483-5 mimics, Bcl-2 expression was increased after OGD/R, while Bax and cleaved caspase 3 expression were decreased (Fig. 2C). However, a miR-483-5 inhibitor led to the opposite effect. These results demonstrate that the mitochondrial apoptosis pathway was involved in OGD/Rinduced PC12 cell injury and that overexpression of miR-483-5p reduced the apoptosis rate triggered by OGD/R.
miR-483-5-p Alleviates PC2 Cell Ischemia-Reperfusion Injury by Promoting Mitochondrial Biogenesis and Decreasing Oxidative Stress
Mitochondrial biogenesis is closely associated with oxidative stress and mitochondrial damage and promoting mitochondrial biogenesis exerts a neuroprotective effect after cerebral ischemia-reperfusion injury (Cardanho-Ramos and Morais 2021). Proliferator-activated receptor-γ coactivator α (PGC-1α) is a transcriptional coactivator and a master regulator of mitochondrial biogenesis, including mitochondrial biogenesis in the nervous system (Besseiche et al. 2015). PGC-1α and its downstream signaling intermediates, such as nuclear respiratory factor 1 (NRF1) and mitochondrial transcription factor A (TFAM), are likely important to normal mitochondrial regulation after ischemia-reperfusion injury (Li et al. 2016). To determine the effect of miR-483-5p on mitochondrial biogenesis after OGD/R, we assessed the protein levels of PGC-1α, NRF1, and TFAM by western blotting. Additionally, we measured the levels of cytochrome c in the cytoplasm and mitochondria. The results revealed that mitochondrial biogenesis was inhibited with the levels of PGC-1α, NRF1, and TFAM decreased after OGD/R, and cytochrome c was released from the mitochondria. In the miR-483-5p mimic group, PGC-1α, NRF1 and TFAM protein levels were significantly increased, and a low level of cytoplasmic cytochrome c was detected (Fig. 3A). Consistently, the decline in mitochondrial membrane potential induced by OGD/R was reversed by the upregulation of miR-483-5p (Fig. 3B). We also detected ROS levels via flow cytometry and measured SOD activity, MDA levels, and LDH levels. Overexpression of miR-483-5p attenuated OGD/R-induced damage by inhibiting ROS generation in PC12 cells. It also inhibited the production of MDA and increased the activity of SOD (Fig. 3C, D).
MiR-483-5p Directly Targets TNFSF8 to Regulate Mitochondrial Biogenesis and Inhibit Apoptosis After Ischemia-Reperfusion
miRNAs generally bind to the 3′-UTR of their target mRNAs and suppress protein production by destabilizing the mRNA, abrogating their translation (Cannell et al. 2008). RNA sequencing was performed to investigate the potential mechanism of miR-483-5p effects on OGD/R-induced cellular injury. Eleven candidate genes (C1ql1, Rab3b, Dnajb7, KDM4E, Foxb2, Eif3h, Fam131a, Tnfsf8, Usp25, Caspase 6, and Atn1) were identified as the most likely target genes of miR-483-5p that contribute to resistance against ischemia-reperfusion injury. Their mRNA expression levels increased after OGD/R but decreased after transfection with miR-483-5p mimics in PC12 cells (Fig. 4A). GO and KEGG analyses indicated that these genes were enriched in "positive regulation of synaptic transmission, dopaminergic," "positive regulation of neurotransmitter uptake," and "neuron remodeling," "apoptosis," and "RNA transport" (Fig. 4B). Then, validation of reference genes for qPCR assays was performed, and the expression of four mRNAs (Usp25, Atn1, Tnfsf8, and Caspase 6) was found to be increased after OGD/R and was decreased in PC12 cells transfected with miR-483-5p mimics (P < 0.05). The opposite trend was observed in the miR-483-5p inhibitor group (Fig. 4C). As miR-483-5p possibly binds sites to the 3′-UTR of three genes, USP25, ATN1, and TNFSF8 (Fig. S2), but not Caspase 6, a potential indirect interaction between miR-483-5p and Caspase 6 was inferred. Western blot analysis indicated a consistent correlation between the protein and mRNA levels of ATN1, USP25, and TNFSF8 (Fig. 4D). Using short interfering RNAs (siRNAs), we silenced the protein expression of the ATN1, USP25, and TNFSF8 ( Figure S3). Next, we investigated the effects of protein expression silencing on apoptosis and mitochondrial biogenesis after OGD/R via western blot analysis. The inhibition of TNFSF8 significantly reduced the protein levels of cleaved caspase 3 and Bax while promoting the expression of mitochondrial biogenesis-related proteins, such as PGC-1α, NRF1, and TFAM (Fig. 4E). These findings suggest that miR-483-5p may reduce OGD/R damage by targeting TNFSF8. Finally, TNFSF8 was confirmed to be a direct target of miR-483-5p, as indicated via a dual-luciferase reporter assay (Fig. 4F).
miR-483-5p Promotes Mitochondrial Biogenesis and Inhibits Apoptosis by Regulating the AMPK/JNK Pathway after Ischemia-Reperfusion
Researchers found that oxidative stress-induced mitochondrial-dependent apoptosis is associated with AMPK/JNK signaling pathway activation (Lou et al. 2021). Furthermore, previous studies have shown that AMPK regulates mitochondrial biogenesis and that the JNK pathway is C Western blot analyses of apoptosis-related proteins. β-Actin was used as the loading control. The data are expressed as the mean ± SD, n = 5, *P < 0.05, and **P < 0.01, ***P < 0.001, and ****P < 0.0001 1 3 critical for regulating ROS-mediated apoptosis (Thirupathi and de Souza 2017). We measured the phosphorylation levels of JNK and AMPK by western blotting to further investigate their relationship with miR-483-5p. A significant decrease in AMPK phosphorylation and PGC-1α protein levels was observed in PC12 cells after OGD/R, Fig. 3 MiR-483-5-p alleviates ischemia-reperfusion mitochondrial injury in PC1 2 cells by promoting mitochondrial biogenesis and decreasing oxidative stress. A Western blot analyses of mitochondrial biogenesis-related proteins and the level of cytochrome c were measured both in mitochondria and the cytosol, n = 5. B Detection of mitochondrial transmembrane potential by JC-1. Red fluorescence: normal mitochondrial membrane potential, as indicated by the formation of JC-1 dimers; green fluorescence: depolarized membrane potential, as indicated by an increase the number of JC-1 monomers, Scale bar: 500 μm, n = 6. C Reactive oxygen species (ROS) were detected by flow cytometry and analyzed, n = 6. D SOD activity and MDA and LDH levels in PC12 cells after OGD/R. n = 3, technical replicates = 2, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 whereas JNK phosphorylation and Bax protein levels were markedly increased. Moreover, miR-483-5p mimics downregulated the expression of phosphorylated JNK and Bax and increased the protein levels of phosphorylated AMPK and PGC-1α (Fig. 5A). Subsequently, by the AMPK activity inhibitor doxorubicin was used, and it was found that after inhibiting the activity of AMPK, the phosphorylation level of JNK increased (Fig. 5B). As TNFSF8 is a target gene of miR-483-5p, we then explored the relationship between TNFSF8 and the AMPK/JNK pathway. By transfecting siRNA to inhibit the expression of TNFSF8 after OGD/R, the activity of AMPK increased and that of JNK decreased (Fig. 5C). In addition, we found that the protective effect of miR-483-5p were blocked when AMPK activity was inhibited and that protective effects of miR-483-5p were restored after treatment with d-epigalbacin (Fig. 5D). These results suggest that miR-483-5p plays a protective role that is mediated through the TNFSF8/ AMPK/JNK pathway. 1 3
MiR-483-5p Alleviates Brain Injury and Improves Neurological Function After Cardiac Arrest in Rats
We established a miR-483-5p overexpression group and miR-483-5p inhibition group in rats via intracranial injection of adeno-associated viral vectors. Four weeks later, qPCR was used to determine miR-483-5p expression in each group (Fig. 6A). Furthermore, we measured the mRNA expression of TNFSF8 and miR-483-5p in the rat hippocampus after cardiopulmonary resuscitation. MiR-483-5p expression increased 12 h after cardiopulmonary resuscitation but was decreased at 24 h. TNFSF8 mRNA levels increased significantly after cardiopulmonary resuscitation (Fig. 6B). Immunofluorescence examination of TNFSF8 protein expression in the rat hippocampus was performed, and the result was consistent with that obtained for the mRNA expression. Negligible protein expression of TNFSF8 was detected in the normal rat hippocampus. In contrast, clear expression of TNFSF8 was observed on some of the hippocampal neurons of rats after cardiac arrest (Fig. 6C). Upregulating Fig. 4 MiR-483-5p directly targets TNFSF8 to promote mitochondrial biogenesis and inhibits apoptosis after ischemia-reperfusion. A Heatmap of RNA sequencing in PC12 cells (11 genes were identified as candidate target genes). B GO enrichment analysis and KEGG enrichment analysis with the 11 identified genes. C Reference gene validation in PC12 cells via qPCR. Four genes that showed significant differences were identified. D Validation of potential target genes as determined by western blotting, n = 4. E Effect of potential target genes involved in apoptosis and mitochondrial biogenesis in PC12 cells as determined by western blotting, n = 5. F, G Binding site prediction and dual-luciferase reporter assays were performed to investigate the relationship between miR-483-5p and TNFSF8, n = 8. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 miR-483-5p inhibited the protein expression of TNFSF8, while inhibiting miR-483-5p increased the protein expression of TNFSF8 after cardiac arrest (Fig. 6C). Representative pictures showing H&E staining, Nissl staining, and TUNEL staining are displayed in Fig. 6D. HE staining revealed that hippocampal neurons in the CA1 region in the sham group were arranged in an orderly manner with normal structures, obvious nucleoli, and clear nuclei and abundant cytoplasm. Many swollen neurons with loosened structures, karyopyknosis, and vacuolar structures were observed in the miR-NC group after CA. The pathological changes in hippocampal neurons were ameliorated in the miR-483-5p group. After CA, the miR-483-5p inhibition group showed more obvious hippocampal damage than the miR-NC inhibition group. As shown by Nissl staining, the number of positive neurons in the CA1 hippocampus was significantly decreased after CA (P < 0.05) and was significantly increased after administration of miR-483-5p (P < 0.05). Neuronal apoptosis was detected by TUNEL assay, and the positive cells were stained red; miR-483-5p directly attenuated CA-induced hippocampal neuronal apoptosis in cardiac arrest. A Morris water maze test was performed and the neuropathy diabetes score (NDS) was determined to assess neurological function and memory 24 h after CA. Compared with the miR-NC group, the miR-483-5p group showed a higher NDS score (Fig. 6E). There was no difference in the baseline latency MiR-483-5p determines the activity of the TNFSF8/AMPK/JNK pathway. A MiR-483-5p enhanced the activity of AMPK while decreasing JNK pathway activity after OGD/R, as confirmed by Western blot assay, n = 5. B AMPK phosphorylation inhibits JNK phosphorylation after ischemia-reperfusion injury, n = 5. C Inhibition of TNFSF8 expression decreased AMPK activity while increasing JNK activity, n = 5. D Both AMPK activity inhibition and JNK activation inhibition counteracted the protective effect of miR-483-5p. Inhibition of JNK activity exerted a protective effect, n = 5. The data are expressed as the mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. Doxorubicin, an AMPK activator inhibitor; d-epigabacin, a JNK activator inhibitor time to target for the rats in each group before CA. The miR-483-5p group rats found the target platform faster, indicating good neural function. The miR-483-5p inhibition group spent a long time finding the platform (Fig. 6F).
miR-483-5p Improves Mitochondrial Biogenesis and Reduces Oxidative Stress-Induced Apoptosis in Rats After Cardiac Arrest
A flow cytometry-based assay with the cationic dye JC-1 was performed to measure mitochondrial membrane potential (MMP) in hippocampal tissue. The MMP was quantified by measuring the green-shifted monomers in the FL-1 channel and the redshifted JC-1 aggregates in the FL-2 channel. The results showed that the MMP of hippocampal tissue decreased, as indicated by enhanced green fluorescence after CA, while miR-483-5p inhibited the generation of green fluorescence (Fig. 7A). The mitochondria exhibited an intact outer membrane and dense cristae in the sham group; however, the mitochondria were clearly impaired, as indicated by fuzzy cristae and matrix swelling, after CA in the miR-NC and miR-NC inhibition groups. Overexpression of mR483-5p improved the mitochondrial morphology, that is, the mitochondrial cristae were distinct and little swelling in the matrix was observed compared with those in the model group. Moreover, the mitochondrial cristae and matrix in the miR-483-5p inhibition group were more severely damaged than those in the miR-NC inhibition group (Fig. 7B).
Performing western blot assays, we evaluated the effects of miR-483-5p on mitochondrial biogenesis and apoptosis after CA. The results showed that the protein expression levels of p-AMPK, PGC-1α, NRF1, Bal-2, and TFAM were significantly decreased in the CA group compared with those in the sham group (P < 0.05). In the miR-483-5p group after CA, the protein expression levels of p-AMPK, PGC-1α, NRF1, Bal-2, and TFAM were markedly increased compared with those in the miR-NC group (miR-483-5p inhibition reversed these effects, P < 0.05). Upregulated expression of miR-483-5p suppressed the expression of the proapoptotic proteins Bax and cleaved caspase 3 following cardiopulmonary resuscitation while increasing the expression of the antiapoptotic protein Bcl-2 (P < 0.05). The sponging miR-483-5p led to the opposite result: apoptosis (Fig. 7C). Additionally, we determined the oxidative stress level in the rat hippocampus for each rat group. Following cardiac arrest, the oxidative stress levels in the rat hippocampus increased significantly with increased LDH levels; however, MDA and SOD activity and LDH levels were reduced in the miR-483-5p group (Fig. 7D).
Discussion
Delayed brain mitochondrial dysfunction during ischemiareperfusion is the main reason for the severity of neurological injury (Kohlhauer et al. 2021). In the present study, our results revealed a close relationship between mitochondrial function and brain function after cardiac arrest. Indeed, mitochondrial homeostasis is necessary to maintain neuronal homeostasis as neurons have a high energy demand.
In addition, mitochondria are the main sources of reactive Fig. 6 MiR-483-5p alleviates brain injury after cardiac arrest in rats.
A Expression of adeno-associated virus carrying miR-NC, miR-483-5p, sponge miR-NC, and sponge miR 483-5p in rat hippocampus 4 weeks after transfection, as determined by qPCR, n = 2; technical replicates = 2. B Expression of miR-483-5p and tnfsf8 mRNAs in the rat hippocampus 24 h after cardiopulmonary resuscitation. C Immunofluorescence assay for TNFSF8 protein detection in the rat hippocampus, Scale bar: 15 μm, n = 3. D HE staining, Nissl staining, and TUNEL staining of hippocampal tissues in the CA1 region of rats from each group, Scale bar: 50 μm, n = 3. E, F An NDS score was obtained and a Morris water maze test was performed to assess spatial and related forms of learning and memory after cardiac arrest for 24 h, n = 3. The data are expressed as the mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 1 3 oxygen species. Our study showed that miR-483-5p targets TNFSF8 to protect against neurological impairment by promoting mitochondrial biogenesis and reducing oxidative stress damage (Fig. 8).
Postcardiac arrest brain injury (PCABI) caused by subsequent brain reperfusion following resuscitation is the leading cause of mortality and long-term disability. MicroRNAs have recently emerged as potential biomarkers of neurological prognosis, and they are also therapeutic targets in cardiac arrest diseases (Devaux et al. 2015). In this study, we focused on identifying the alteration of miRNAs and exploring their possible mechanisms after cardiac arrest. The logistic regression analysis revealed a strong association between miR-483-5p expression and neurological prognosis for CA patients (Fig. 1C). This finding was consistent with prior research showing that miR-483-5p was closely related to neurological disease. was the most abundant miRNA in blood plasma, specifically in prodromal and early-stage Alzheimer's disease (AD) patients, and upregulation of miR-483-5p decreased phosphorylation of neuronal microtubule-associated protein Tau, suggesting that it plays a neuroprotective role in AD pathology (Nagaraj et al. 2021). Expression of miR-483-5p promotes insulin-like growth factor-II (IGF-II) transcription, which is associated with neurogenesis and differentiation (Vazquez et al. 2013;Zhu et al. 2019).
To date, the role played by miR-483-5p in ischemic disease remains uncertain. Our study found that miR-483-5p exerts a protective effect by reducing ischemia-reperfusion injury in PC12 cells and improving neurological outcomes after CA in rats. Then, we investigated the potential mechanism of miR-483-5p. After ischemia-reperfusion, miR-483-5p promoted mitochondrial biogenesis by upregulating the protein expression of PGC-1α, inhibited the production of ROS, and reduced the mitochondrion-dependent apoptosis rate both in vivo and in vitro. Furthermore, miR-483-5p significantly increased the phosphorylation of AMPK while decreasing the phosphorylation of JNK. These data showed that miR-483-5p treatment exerted a strong neuroprotective effect after cardiac arrest.
Fig. 7
MiR-483-5p alleviates brain ischemia-reperfusion injury by inhibiting excessive mitochondrion-dependent apoptosis, promoting mitochondrial biogenesis, and decreasing oxidative stress in rats. A Changes in mitochondrial membrane potential was assessed by adding JC-1 to hippocampal neurons, n = 3; technical replicates = 2. B Electron microscopy revealed the morphology of mitochondria in the rat hippocampus after cardiac arrest (red arrow: mitochondrion with an intact outer membrane and dense cristae, severely impaired mito-chondria with fuzzy cristae and swelling matrix (Above scale bar: 2.00 μm; below scale bar: 500 nm), n = 3; technical replicates = 2. C The expression of proteins related to mitochondrial biogenesis and apoptosis in rat hippocampal tissues from each group was measured by Western blotting, n = 5. D SOD and MDA activity and ROS and LDH levels were measured in hippocampal tissues after cardiac arrest, n = 3; technical replicates = 3. The data are expressed as the mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 Multiple mitochondria-related mechanisms have been proposed to play essential roles in brain ischemia-reperfusion (I-R) injury. Mitochondrial dysfunction following ischemic stroke results in the loss of mitochondrial membrane potential, generation of ROS, and activation of intrinsic apoptosis (Li et al. 2018). In animal models, some therapeutic interventions targeting mitochondrial dysfunction have proven effective for brain injury following cardiopulmonary resuscitation. For example, palmitic acid therapy after CA inhibited neuroinflammation and mitochondrial dysfunction, improving neurological function (Wu et al. 2021). PD98059 protected cerebral cortex mitochondrial structure, increased survival rates and neurological deficit scores, and reduced the rate of mitochondrial permeability transition pore opening (Zheng et al. 2020). Therefore, mitochondria have become increasingly important targets for therapeutic interventions in cardiac arrest-induced brain injury.
Increasing evidence shows that the promotion of mitochondrial biogenesis protects neurons against mitochondrial insults and improves neurological outcomes after cardiac arrest in rats (Pan et al. 2014;Wang et al. 2016). PGC-1α has been reported to be a major regulator of mitochondrial biogenesis by inducing biogenesis via the activation of different transcription factors, including nuclear respiratory factor 1 (NRF 1) and mitochondrial transcriptional factor A (TFAM) (Piccinin et al. 2019). Increased PGC-1α expression was associated with a reduction in ROS production and a subsequent increase in mitochondrial biogenesis (Sharma et al. 2015). Evidence has also shown that increased mitochondrial ROS downregulated mitochondrial biogenesis (Chevtzoff et al. 2010), while upregulating the expression of PGC-1α, increasing the expression of components of the mitochondrial ROS defense system, such as glutathione peroxidase or SOD, in neural cells, thus protecting neural cells from oxidative stress-mediated death (St-Pierre et al. 2006). Our data showed that the PGC-1α, NRF 1, and TFAM protein expression levels were increased in the rat hippocampi overexpressing miR-483-5p after ROSC, and this outcome was accompanied by an excellent neurological prognosis. This finding indicates that miR-483-5p induced mitochondrial biogenesis and promoted recovery from I-R brain injury. Notably, miR-483-5p has been shown to regulate mitochondrial dynamics in other contexts. Fan et al. and Tian et al. showed that miR-483-5p attenuated mitochondrial fission and cisplatin sensitivity and inhibited apoptosis in tongue squamous cell carcinoma (Fan et al. 2015;Tian et al. 2019a, b). Moreover, miR-483-5p promoted the expression of PPAR-γ, which participates in the regulation of mitochondrial biogenesis ). Our results demonstrate that miR-483-5p affects brain injury after cardiac arrest by regulating mitochondrial biogenesis.
TNFSF8 functions through its interaction with CD30 and is reported to be a death ligand that induces the apoptosis and proinflammatory factor production (Ruggeri et al. 2002;Tur et al. 2009). Although TFNSF8 is rarely expressed in normal tissues and cells except when the immune system is activated (Abdalla et al. 2004), the expression of TNFSF8 has been widely detected in many tissues and cells that are under stress conditions. TNFSF8 was expressed in the cardiac myocytes of mice with myocarditis, and TNFSF8 was found to be moderately or highly abundant on ventricular myocytes after treatment with IFN-γ (Seko et al. 1999). Additionally, TNFSF8 has been observed in thyroid epithelial (follicular) cells, yolk sac carcinoma, and embryonal carcinoma (Pera et al. 1997;Ruggeri et al. 2002). Our study confirmed that TNFSF8 is expressed in hippocampal neurons. Although protein expression was relatively low in normal neurons, TNFSF8 protein expression was significantly increased and caused damage in hippocampal neurons after cardiac arrest. TNFSF8 was detected on PC12 cells. Furthermore, we demonstrate that TNFSF8 was the target gene of Fig. 8 Hypothetical mechanism through which miR-483-5p attenuates ischemia-reperfusion injury after cardiopulmonary resuscitation. miR-483-5p promotes mitochondrial biogenesis and reduces apoptosis via the TNFSF8/AMPK/JNK pathway miR-483-5p and that inhibiting the expression of TNFSF8 played a neuroprotective role after cardiac arrest. The inhibition of TNFSF8 expression alleviated mitochondrial injury and induced mitochondrial biogenesis.
AMP-activated protein kinase (AMPK) is a highly conserved serine/threonine protein kinase and a key regulator of energy homeostasis and cell survival under inflammatory and oxidative stress conditions (Sun et al. 2013). A study documented that AMPK inactivation or activation due to oxidative stress was involved in ischemic brain damage (He et al. 2020a, b). Activated AMPK can be measured to monitor mitochondrial function because it regulates the expression of peroxisome proliferator-activated receptor-γ coactivator (PGC)-1α, attenuating neuronal injury after ischemia-reperfusion (Tian et al. 2019a, b). Jun N-terminal kinase (JNK) activation induced ROS-modified mitochondrial fission and increased the mitochondrion-dependent apoptosis rate (Chen et al. 2021;Zhang et al. 2020a, b). Our results show that upregulating TNFSF8 inhibited the activation of AMPK while increasing the activity of JNK. JNK phosphorylation was suppressed in cells treated with the AMPK activation inhibitor doxorubicin. This finding indicated that JNK may be a downstream target of AMPK. The harmful effect of TNFSF8 was offset by treatment with the JNK activation inhibitor d-epigalbacin. This outcome suggested that TNFSF8 may function through the AMPK/ JNK pathway. Some studies have shown that the AMPK/ JNK pathway is related to the regulation of ROS generation and apoptosis. Lou et al. observed that inhibiting oxidative stress-induced mitochondrion-dependent apoptosis through the AMPK/JNK signaling pathway enhanced cardiac function (Lou et al. 2021). Chen et al. showed that metformin attenuated cadmium-induced neuronal apoptosis in vitro by blocking the AMPK/JNK signaling pathway . These results were consistent with those in our study. We found that when AMPK and JNK activity were both inhibited, the protein expression of PGC-1a, NRF1, and TFAM remained high, suggesting that the potential regulatory mechanisms between AMPK/JNK and the PGC-1α pathway are complex. To some extent, inhibition of JNK activity promoted PGC-1α protein expression. Zheng et al. reported that inhibition of JNK indirectly increased PGC-1α expression by upregulating PPAR-α (Zheng et al. 2017).
Our study was exploratory and had some limitations. The study showed that miR-483-5p suppressed the expression of Caspase 6 in PC12 cells. Caspase 6 is related to cognitive impairment and inflammation in the brain (Flores et al. 2022). However, Caspase 6 is not the direct target of miR-483-5p. Moreover, miR-483-5p reduced the expression of ATN1 and USP25. ATN1 mutation has been associated with neurodevelopment. Overexpression of USP25 resulted in microglial activation and induced synaptic and cognitive deficits (Zheng et al. 2021). Although ATN1 and USP25 carry potential binding sites for miR-483-5p, no further verification was performed. These results suggest that the neuroprotective mechanisms of miR-483-5p are multifaceted, and more research is needed.
Conclusion
In conclusion, the present study revealed that the miR-483-5p/TNFSF8/AMPK/JNK axis was critically involved in mitochondrial dysfunction in ischemia-reperfusion injury in vivo and in vitro. MiR-483-5p targets TNFSF8 to protect against neurological impairment by promoting mitochondrial biogenesis and reducing oxidative stress damage. These results indicate that miR-483-5p may be a new therapeutic target for improving neural function following cardiopulmonary resuscitation.
Author Contributions CH conceived of the presented idea. QZ and HZ carried out the experiment and wrote the manuscript. CL, CZ, HW, BL, DZ, and YL verified the analytical methods. SH, JC, and CW provided critical feedback and helped shape the research, analysis, and manuscript. CH and XL supervised the findings of this work and contributed to the final version of the manuscript. All authors discussed the results.
Data Availability
The data that support the findings of this study are available from the corresponding author, Xiao-Xing Liao, upon reasonable request.
Conflict of interest The authors declare no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. | 2022-10-22T06:16:32.578Z | 2022-10-20T00:00:00.000 | {
"year": 2022,
"sha1": "ad35458361d419e2630fecafa6f421d2ba543393",
"oa_license": "CCBY",
"oa_url": "https://link.springer.com/content/pdf/10.1007/s10571-022-01296-3.pdf",
"oa_status": "HYBRID",
"pdf_src": "Springer",
"pdf_hash": "ef8ccec4870ceacb47d41b41cae7f5bcd586f348",
"s2fieldsofstudy": [
"Medicine",
"Biology"
],
"extfieldsofstudy": [
"Medicine"
]
} |
233380848 | pes2o/s2orc | v3-fos-license | Determination of acceptance criteria for geometric accuracy of magnetic resonance imaging scanners used in radiotherapy planning
Highlights • Geometric accuracy of MRI-scanners in radiotherapy planning must be evaluated.• Phantom acquisitions with standard and clinical sequences were performed.• Geometric distortions were determined in several volumes of interest.• We recommend acceptance criteria for MRI-scanners in radiotherapy planning.• Explicit and simple acceptance criteria enable effective regulatory inspections.
Introduction
Magnetic resonance imaging (MRI) has become widespread in radiation therapy planning (RTP) in the 2010s. MRI has a superior soft-tissue contrast compared to computed tomography (CT), and the contouring of planning target volumes (PTV) and organs-at-risk (OAR) is more precise in magnetic resonance (MR) images in many anatomical regions, especially in head and pelvis [1][2][3].
Traditionally, MR-images have been utilized in RTP through coregistration with a planning-CT. The best registration accuracy is achieved, if the patient position is the same in the MR-images as in planning-CT. Recently, the number of RTP-dedicated MRI-scannerswith flat tabletops and large bores for flexible patient positioninghas increased at the radiotherapy clinics. Such scanners have also introduced the possibility of the MRI-only RTP, where the whole RTP from contouring to dose calculation is performed using MR-images [4,5].
MRI-only workflow in RTP has many benefits from improving accuracy of contouring to reducing the radiation dose to healthy tissues [2,6]. In addition, the systematic error of the MRI-CT-coregistration in the traditional RTP workflow is avoided. For these reasons MRI-only RTP is used in brachytherapy for gynecological cancers and is the suggested RTP method of American Society of Brachytherapy [7]. In external beam radiation therapy (EBRT), MRI-only RTP faces two major shortcomings: the first is the lack of attenuation information for dose distribution calculation, which has been solved by using synthetic-CTimages, and the second the limited geometric accuracy.
The geometric distortions in MRI arise from both system-related and patient-induced sources [8]. The system-related distortions include B 0field inhomogeneities and gradient nonlinearities. The subject-related distortions comprise of magnetic susceptibility and chemical shift artefacts. Additionally, the scanning sequence parameters affect the distortions. Thus, most modern scanners allow shimming and distortion correction algorithms for distortion reduction.
In MRI-only RTP, the distortions of PTV, bony structures and bodyoutline directly affect the dose distribution calculation. Additionally, the distortions can violate the deformable MRI-CT-coregistration. Hence, the MRI geometric accuracy is relevant in RTP regardless of the application. According to Weygand et al. [8] the distortions should be less than 2 mm for MRI-only RTP, and Bird et al. [9] have concluded that the geometric distortions no longer prevent the usage of MRI-only in RTP of pelvis region, provided the distortions are properly addressed.
The system-related geometric distortion in MRI can be measured with a phantom of known dimensions. Many phantoms have been suggested for the purpose, most based on signal-producing markers in a signal-suppressing medium or a grid structure in a signal-producing medium [10][11][12]. For example, Tortef et al. [11] and Walker et al. [12] produced large-field-of-view (large-FOV) phantoms, where vitamin capsules were inserted in signal-suppressing medium. The distortions are usually determined by detecting the markers or the grid intersections in the MR-image and comparing the positions to a geometricallyaccurate reference, for example a CT-image of the same phantom.
Multiple studies have also investigated the MRI geometric accuracy from RTP point-of-view as a part of MRI-only protocol commissioning or feasibility study [10,[13][14][15][16]. The dose uncertainty of the whole radiotherapy process should be less than 5% [17] as inaccurate dose in PTV or OAR or both can lead to an undesired clinical response. The required level of accuracy induces the demand for quality assurance (QA) through the patient care path from medical imaging to linear accelerators. International organizations (e.g. IAEA, AAPM) have established QA guidelines and acceptance criteria for the majority of the steps in the care path. For example, the American Association of Physicists in Medicine (AAPM) report no. 83 [18] sets requirements for CT-simulators and CT-simulation (e.g. table movement and laser accuracy).
According to the American College of Radiology (ACR) accreditation protocol at diagnostic MRI, the dimensions of the head-sized ACRphantom should not change more than 2% in a 2D plane [19]. AAPM report no. 100 [20] recommends that on MRI-scanners, which are used for treatment planning, geometric accuracy must be regarded. They also state, that the geometric distortions should be determined over a large-FOV, but do not specify the size of this large-FOV or the allowed distortion magnitude beyond the abovesaid ≤2%. Hence, the international QA guidance is lacking concerning the MRI-scanners in RTP. As the QA currently depends on the initiative of individual clinics, the intersite comparison and regulatory inspections of supervising authorities are difficult.
This study aimed to evaluate the performance level of MRI-scanners utilized in RTP. Multiple phantom measurements were performed to define the geometric accuracy of five different MRI-scanners at separate radiotherapy clinics. Acceptance criteria for the MRI-scanners in RTP are recommended.
Phantom
A GRADE-phantom (generic model TS1006, Spectronic Medical AB, Helsingborg, Sweden) was used for the measurements. The phantom was compatible with various MRI-scanners and contained around 1200 spherical signal-producing markers implanted into foam. The outer dimensions of the phantom were 400 mm × 490 mm × 535 mm (Width × Height × Length), and the markers defined a volume of 430 mm × 335 mm × 455 mm (W × H × L). The phantom had smaller markers in the center and a cross on the top.
Image acquisition
The phantom was imaged with five different MRI-scanners (hereafter denoted as A, B, C, D and E) from three major vendors. Three of the scanners had field strength of 1.5 T (A-C) and two 3 T (D and E); all the scanners had been installed between 2011 and 2018. Additionally, all the scanners had 70-cm-bores, flat tabletops, and outer set-up lasers (LAP-lasers). Details on the scanners are given in Supplementary Table S1.
On each scanner, the phantom was placed on the imaging table and the phantom's cross was visually aligned with the scanner's LAP-lasers. On all scanners, three sequences were acquired: a 2D fast-spin-echo (FSE), a 3D gradient-echo (GRE), and a clinical sequence. The 2D FSE and 3D GRE sequence parameters were chosen based on the GRADEphantom manual with small alternations between scanners. The clinical sequences were pelvic imaging sequences locally utilized at each clinic. In the 2D FSE and 3D GRE sequences, the phase-encoding direction was right-left and the receiving coil the scanner's intrinsic bodycoil. A maximal FOV (300-500 mm) without moving the imaging table was used and the vendor's 3D geometric-distortion correction was applied to each acquisition. The acquisition parameters for the used sequences are given in Supplementary Tables S2 and S3.
Repeated measurements were performed on Scanner A: the 2D FSE and 3D GRE sequences were acquired three times without moving the phantom (so-called single-setup) and seven times with new setups (repeated-setup) during a four-month period.
In addition, the phantom was imaged with CT (Siemens SOMATOM Definition AS, Siemens Healthcare GmbH, Erlangen, Germany) before and after the repeated-setup measurements. The CT-image was acquired using a FOV of 500 × 500 mm 2 with slice thickness of 1 mm, pixel size of 0.6 × 0.6 mm 2 , image matrix of 512x512, and a peak voltage of 120 kV. The CT-image was reconstructed with a high-definition extended-FOV of 512 × 512 mm 2 and isotropic voxel size (1 × 1 × 1 mm 3 ). The CT acquisitions were visually compared to affirm the phantom stability during the measurement period.
Data analysis
The MRI data was analyzed with the MriPlanner-software (v1.0.46, Spectronic Medical AB, Helsingborg, Sweden). The software used a geometrically accurate reference (i.e. our CT acquisition) and deformable registration to determine the distortions in the MR-images and produced text files that contained coordinates of real marker positions and deformed marker positions. Using the marker position information and a home-written MATLAB-code (R2019a, The MathWorks Inc., Natick, MA, US), the distortion magnitudes were determined in spherical volumes around the scanner isocenter with different diameters of the spherical volume (DSV), as well as in cylindrical volumes of interest (VOI) of different lengths along the z-axis (i.e. along the direction of the scanner bore; see Supplementary Fig. S1). Additionally, the one standard deviation (1SD) of the distortion magnitudes were determined marker by marker between the single-setups and the repeated-setups.
Results
An example of the measured geometric distortions as a spatial map with the corresponding phantom image is presented in Fig. 1.
For all acquired sequences, except the 2D FSE sequence of Scanner C, the mean and median distortion magnitude was <1 mm at DSV of 400 mm, and <2 mm at DSV of 500 mm (Fig. 2). The maximum distortion was <2 mm at DSV of 300 mm for all the scanners. On Scanner C, the mean and median value of the 3D GRE and the clinical sequence was <1 mm until DSV of 500 mm, but the 2D FSE sequence had the largest mean and median distortions in DSV of 400 mm and 500 mm. Overall largest maximum distortion was 49 mm of the 2D FSE sequence on Scanner A outside DSV of 500 mm.
In cylindrical VOIs of length ≤300 mm, the median distortion magnitude was <1 mm for all acquired sequences, except the 2D FSE sequences of Scanners A and C (Fig. 3). With the same sequences, the mean distortion was <2 mm in VOIs of length ≤300 mm. The maximum distortion magnitude was ≥2 mm for most of the sequences starting from the VOI of length 100 mm.
For the 2D FSE sequence, the marker-by-marker 1SD of single-setups had mean value of 0.3 mm and maximum value of 5.2 mm, and repeated-setups had mean value of 0.6 mm and maximum value of 9.0 mm (Fig. 4). For the 3D GRE sequence, the corresponding mean and maximum values were 0.1 mm and 0.3 mm for single-setups and 0.5 mm and 6.2 mm for repeated-setups, respectively.
Discussion
The aim of this study was to determine the present state of MRIscanners in several radiotherapy clinics, and to formulate acceptance criteria for such MRI-scanners. The measurements were repeatable and the measured distortions were found to be tolerable in clinically meaningful VOIs.
For all measured scanners and sequences, except a single sequence, the mean and median geometric distortion magnitude was <1 mm for DSV of 400 mm, <2 mm for DSV of 500 mm. In cylindrical VOIs along the z-axis, the maximum distortion magnitudes were >2 mm even in the smallest considered VOI with length of 100 mm. The characteristic values of distortion distributions were similar for all scanners and sequences at DSV of ≤300 mm. At the DSV of 400 mm the maximum distortion magnitudes are <2% of the VOI dimensions for all scanners and sequences, which satisfy the ACR and AAPM demands. With larger distances from the isocenter differences between scanners and sequences are apparent and the distortions can increase above the 2%threshold. The 3D GRE sequences are often more accurate further from the isocenter than 2D FSE sequences because vendor's 3D distortioncorrection algorithms function with their full capacity in 3D-volumes. The PTVs in RTP are rarely larger than 200 mm in diameter; hence, all the scanners are geometrically accurate enough for RTP with MRI-CT-coregistration.
The maximum distortion was >2 mm in all the VOIs along z-axis for a majority of the sequences, which could lead to a compromised geometric accuracy of a patients' body-outline. Kemppainen et al. [21] have reported that distortion magnitude increases rapidly along the z-axis off the isocenter. The same phenomenon is visible in our results for VOI lengths >200 mm with several sequences. Increasing distortion must be noted in MRI-only; only short longitudinal area nearby the isocenter can be planned accurately with MRI-only [22].
Kemppainen et al. [21] reported distortions less than 2 mm in the body-outline of pelvis area and less than 1 mm in PTVs (prostate, rectum, and gynecological) and OARs (rectum and bladder). Adjeiwaah et al. [15] measured maximum distortions of 2.17 mm in 200 mm radial distance from isocenter, and Gustafsson et al. [13] mean distortions <2 mm in 200-250 mm radial distance from isocenter in phantom studies. These studies were performed with clinical sequences for MRI-only commissioning, and the dose differences between distorted and planning-CTs were <1%, when a dose difference of <2% is generally accepted [23]. Compared to these results, our measured distortion magnitudes are similar especially on clinical sequences. The clinical sequences of Scanners A, C and E were MRI-only protocols and their geometric accuracy is sufficient for MRI-only RTP. The distortions were higher in the 2D FSE sequences, which emphasizes that the MRI-only protocols must always be locally commissioned.
According to our CT acquisitions, the phantom remained stable during our measurement period. The MRI scans were repeatable within radial distance <250 mm and distance <100 mm in z-direction, where the marker-by-marker mean 1SDs of single-setups and the repeatedsetups were less or equal compared to the measured distortions. Either due to the limited acquisition FOV or severe distortions, the analyses were limited to markers within <200 mm distance from isocenter in zdirection. Gustafsson et al. [13] have evaluated the intrinsic susceptibility of the GRADE-phantom, and deemed it insignificant (<0.5 mm in radian distances <250 mm measured with receiver bandwidth of 554 Hz/mm). Wyatt et al. [24] have evaluated the repeatability and set-up sensitivity of measurements with the GRADE-phantom and corresponding MriPlanner-sofware, and deemed the measurements repeatable but relatively sensitive to small (<1 mm or <1 • ) set-up errors. In addition, they reported that the analysis might detect markers incorrectly in the peripheral areas, if the distortion magnitude is close to the distance between the peripheral markers (ca. 30 mm). This could explain the high maximum 1SD of the marker-by-marker distortion in the single-setup and the repeated-setup measurements. Consequently, the maximum distortion magnitudes could be misidentified in the peripheral areas.
From the aspect of regulatory inspections performed by a supervising authority, the determination of geometric distortions at a couple of clinically meaningful DSVs would be the simplest solution for MRIscanner QA. However, differences between scanners could complicate standardization of such inspections; for example Scanners D and E in this study had remarkable smaller maximum FOVs than the others. Additionally, the MriPlanner-software does not directly produce distortion magnitudes at DSVs, but rather reports with the mean and maximum distortion magnitudes in different intervals of the radial distance from the scanner isocenter (e.g. 100-150 mm) and the text files mentioned in Section 3.2. Thus, to obtain results with DSVs, further analysis is necessary when this phantom and software are used. The inspection along z-axis also requires further analysis, and is less robust for different FOVs. Nonetheless, defining the distortions in the cylindrical volumes provides useful information as MRI-only RTP requires accurate representation of the patient body-outline. On the downside, our measurements with GRADE-phantom cannot be used to evaluate the sequenceand patient-specific susceptibility effects.
Required geometric accuracy of MRI depends on how the MR-images are used in RTP. In MRI-only, the mean distortion should be ≤1 mm at DSV of 200 mm and ≤2 mm at DSV of 400 mm (Table 1). Very large patients should not be simulated with MRI-only due to the possibility of deformed body-outline. In MRI-CT-coregistration, maximum distortion of 2 mm at DSV of 200 mm is acceptable, which is comparable to the ACR and AAPM requirements. The anatomic region where the PTV resides must be considered; MRI is best suited for targets lying in the center of the body, whereas peripheral targets (such as breasts) suffer from distortions that could hinder the treatment accuracy [16]. Furthermore, in stereotactical radiosurgery all error sources should be less than 1% [25]. This requires excellent geometric accuracy: 1 mm in less than 2 cm PTVs and 1.5 mm in larger PTVs [26]. As the systemrelated distortions are 3D by their nature, the phantom measurements should be performed over a 3D volume [27,28]. According to Ranta et al. [29] the QA measurements in 2D are sufficient once a standard level of scanner performance is established in 3D. Ranta et al. also concluded that the geometric accuracy could remain stable for years. Furthermore, the geometrical accuracy is not the only feature required from an MRI-scanner in RTP. The scanner should also have a radiologically acceptable image quality and the same patient positioning precision as treatment units [27,31]. Other properties of an MRI-scanner suitable for RTP include flat imaging tabletop, patient positioning laser system (LAP-lasers), a large bore of at least 70 cm, and large-FOV (preferably at least 50 cm). A comprehensive QA of image quality, geometrical accuracy, and mechanical properties of the scanner must be performed with regular intervals, at least once a year. The GRADEphantom together with ACR-phantom have been found feasible for these QA measurements [32] which remain the responsibility of a radiotherapy clinic. However, even the best system-related geometric accuracy and QA measurements do not replace critical revision of the In the future, a regulatory inspection of MRI-scanners could consist of phantom acquisitions with standardized QA sequences as well as clinical sequences. In MRI-CT-coregistration the geometric accuracy of diagnostic-level is sufficient. In MRI-only RTP mean distortion of ≤1 mm at DSV of 200 mm and ≤2 mm at DSV of 400 mm are accepted. The presented measurements with explicit and simple acceptance criteria enable effective regulatory inspections of the MRI-scanners and the comparison of the scanners for a supervising authority. Corresponding criteria will be necessary for MR-linear accelerators and MR-imageguided radiotherapy in the future.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Fig. 4.
One standard deviation (1SD) of the marker-by-marker distortion magnitude for single-and repeated-setups versus marker's radial distance (left) and distance along z-axis (right) from the scanner isocenter.
Table 1
Acceptance criteria recommendations of the geometric distortion for MRIscanners used in radiotherapy planning: external radiotherapy to conventional target volumes (MRI-CT-coregistration and MRI-only) versus diameter of spherical volumes (DSV); external stereotactical radiosurgery to small targets according to the size of the planning target volume (PTV); and internal radiotherapy (i.e. brachytherapy). | 2021-04-25T05:30:03.569Z | 2021-01-01T00:00:00.000 | {
"year": 2021,
"sha1": "537a515609f5ec9800acb8f716089dc9e3d5191b",
"oa_license": "CCBY",
"oa_url": "http://phiro.science/article/S2405631621000038/pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "537a515609f5ec9800acb8f716089dc9e3d5191b",
"s2fieldsofstudy": [
"Medicine",
"Physics",
"Engineering"
],
"extfieldsofstudy": [
"Computer Science",
"Medicine"
]
} |
56468674 | pes2o/s2orc | v3-fos-license | Boundedness of the number of nodal domains for eigenfunctions of generic Kaluza-Klein $3$-folds
This article concerns the number of nodal domains of eigenfunctions of the Laplacian on special Riemannian $3$-manifolds, namely nontrivial principal $S^1$ bundles $P \to X$ over Riemann surfaces equipped with certain $S^1$ invariant metrics, the Kaluza-Klein metrics. We prove for generic Kaluza-Klein metrics that any Laplacian eigenfunction has exactly two nodal domains unless it is invariant under the $S^1$ action. We also construct an explicit orthonormal eigenbasis on the flat $3$-torus $\mathbb{T}^3$ for which every non-constant eigenfunction belonging to the basis has two nodal domains.
Introduction
This article is concerned with the number of nodal domains of eigenfunctions of the Laplacian on certain 3-dimensional compact smooth Riemannian manifolds (P, G). The manifolds are S 1 = SO(2) bundles π : P → X over a Riemannian surface (X, g), and G is assumed to be a Kaluza-Klein metric adapted to π, i.e., G is invariant under the free S 1 action on P and there exists a splitting T P = H(P ) ⊕ V (P ) of T P so that dπ : H p (P ) → T π(p) X is isometric and so the fibers are geodesics. Thus, π : P → X is a special kind of Riemannian submersion with totally geodesic fibers in the sense of [3] (see Definition 1.1 and Definition 6.3 of the present article). The S 1 action commutes with the Laplacian ∆ G of the Kaluza-Klein metric G and one may separate variables to obtain an orthonormal basis of joint eigenfunctions φ m,j , Our focus is on the nodal sets of the real or imaginary parts of (1.2) φ m,j = u m,j + iv m,j and on particularly on the number of their nodal domains. Since ∆ G is a real operator, the real and imaginary parts (1.2) satisfied the modified eigenvalue system, Our main result (Theorem 1.5) is that when 0 is a regular value of φ m,j for all (m, j), then for m = 0, the nodal sets of u m,j , resp. v m,j , are connected and there exist exactly 2 corresponding nodal domains. The case m = 0 is special because φ 0,j is then real valued and is a pullback from the base X; in this case, the number of connected components of the nodal set (and the number of nodal domains) is the same as for the corresponding eigenfunction on X. Theorem 1.4 shows that it is a generic property of Kaluza-Klein metrics on S 1 bundles over Riemann surfaces that 0 is indeed a regular value of φ m,j for all (m, j). The precise statement requires a discussion of the geometric data underlying a Kaluza-Klein metric and how we allow it to vary when defining "genericity". An introductory discussion of the Kaluza-Klein metrics of this article is given in Section 1.1 and a more detailed discussion is given in Section 4 (see Theorem 4.1 and Lemma 4.9).
Adapted Kaluza-Klein metrics
We now define Kaluza-Klein metrics on a three-dimensional manifold P which is an S 1 bundle over a (usually) compact Riemannian surface X. In our main results, P is the unit co-circle bundle of an ample complex ANNALES DE L'INSTITUT FOURIER holomorphic line bundle L → X. Thus, P = S 3 or P = U (S 2 ) (the unit tangent bundle) when X = S 2 , and P = S * X (the unit conormal bundle) when the genus g 2. When g = 1, P can be the unit circle bundle of the theta line bundle over T 2 , or it may be the trivial circle bundle S * T 2 (see Section 9.1). Other examples where X has constant curvature are discussed in Section 9.
A Kaluza-Klein metric is determined by the following data: (i) A surface X equipped with a Riemannian metric g and a complex structure J, (ii) A nontrivial complex holomorphic line bundle L → X over a surface, (iii) A Hermitian metric h on L, (iv) A complex structure J L on L, (v) An h-compatible connection ∇ on L.
In this article, we fix J, L and J L and only vary the data (g, h, ∇). The unitary frame bundle for the Hermitian metric h is defined by The connection ∇ induces a connection 1-form on P h and a splitting T P h = H(P h ) ⊕ V (P h ) into horizontal and vertical spaces; see Section 2 for background.
Definition 1.1. -The Kaluza-Klein metric on P h is the U (1)-invariant metric G such that the horizontal space H p := ker dπ is isometric to T π(p) X, so that V = R ∂ ∂θ is orthogonal to H and is invariant under the natural S 1 action and so that the orbits of the S 1 action are unit speed vertical geodesics.
The data (g, h, ∇) determines a horizontal Laplacian ∆ H , a vertical Laplacian ∂ 2 ∂θ 2 , and their sum, the Kaluza-Klein Laplacian (1.1), As is well-known, sections s of powers L m of a complex line bundle L lift to equivariant functionsŝ : L * → C on the dual line bundle by the formulaŝ(z, λ) = λ(s(z)) where a point of L * is denoted by λ ∈ L * z . Under this identification, the horizontal Laplacian is equivalent to the Bochner Laplacians ∇ * m ∇ m on sections of L m . Thus, equivariant eigenfunctions of ∆ G of weight m on M are lifts of eigensections of ∇ * m ∇ m . See Section 3 and Lemma 6.6 for details. In proving genericity theorems it is easier to TOME 70 (2020), FASCICULE 3 work downstairs on X. But the nodal results pertain to the equivariant eigenfunctions on P h . Remark 1.2. -Not all S 1 invariant metrics on SX are adapted Kaluza-Klein metrics (see Section 6.1). Many of the techniques of this paper extend to general S 1 -invariant metrics on principal S 1 bundles over manifolds of all dimensions. For simplicity of exposition we restrict to dimension 3.
Nodal sets
We thus have two versions of the eigenfunctions of the Kaluza-Klein ∆ G , first as scalar complex valued equivariant eigenfunctions on P h and second as complex eigensections on X. In each version we have a nodal set, and we use the base nodal set on X to analyse the nodal set on P h .
We denote the eigensection corresponding to φ m,j as f m,j e m L in a local holomorphic frame. We mainly consider L = K X and then we write the section as f m,j (z)(dz) m . Let Then, f m,j (z)e −imθ = (a m,j (z) + ib m,j (z))(cos mθ − i sin mθ), so that with φ m,j = u m,j + iv m,j , (1.4) u m,j = a m,j cos mθ + b m,j sin mθ, v m,j = b m,j cos mθ − a m,j sin mθ.
See Section 6.2 for more details. We denote by Z fm,j the zero set of the eigensection f m,j e m L on X: Z fm,j = {z ∈ X : f m,j (z) = 0}.
It is easy to see that the zero set Z φm,j of φ m,j is the inverse image of Z fm,j under the natural projection π: Usually we study the nodal sets of the real and imaginary parts of the lift, not to be confused with the lifts of the real and imaginary parts of the local expression f m,j of the section (since the frame e m L must also be taken into account). In general, it is not obvious whether or not the zero set of f m,j is discrete in X.
The analysis is the same for real and imaginary parts and we generally work with the imaginary part, following the tradition for quadratic differentials.
Perhaps the most familiar setting for such Kaluza-Klein metrics and lifts of m-differentials to equivariant sections is that of hyperbolic surfaces of finite area. The reader familiar with Maass forms and operators may want to compare Kaluza-Klein notions with those of SL(2, R)-theory in Section 9.3.
Statement of results
Let (X, J, g) be a Riemannian surface with complex structure J. Let (L, h) be a Hermitian holomorphic line bundle over X, and let P h ⊂ L * → X be the principal S 1 bundle associated to h. Let ∇ h be an h compatible connection on L, and let G = G(g, h, ∇ h ) be the associated Kaluza-Klein metric on P h .
As mentioned above, weight 0 (invariant) eigenfunctions are special because they are real-valued (once they are multiplied by a suitable constant). We therefore separate the case m = 0 from the remainder of the discussion, and state the obvious (but interesting) Proposition 1.3. -For m = 0, invariant eigenfunctions (m = 0) of ∆ G are lifts π * ψ j of eigenfunctions ψ j of ∆ g on the base X, and the nodal set of π * ψ j is the inverse image under π of the nodal set of ψ j . The number of nodal domains of π * ψ j equals the number of nodal domains of ψ j .
Indeed, their nodal sets are inverse images of nodal sets on the base. Hence the number of nodal domains of "invariant" Kaluza-Klein eigenfunctions is the number for the corresponding eigenfunction on the base. Henceforth we always assume m = 0.
To prepare for our main result when m = 0, we first state a result on generic properties of equivariant Kaluza-Klein eigenfunctions. By "generic" properties of Kaluza-Klein metrics, we mean properties of residual sets in a suitable C k space of the data (g, h, ∇), often when only one component is varied and the others are fixed. The generic properties of concern in this article hold for many choices of Banach spaces of data, which could be suitable C k spaces or a Sobolev spaces H s . Our basic reference for genericity properties of eigenvalues/eigenfunctions in [20], and the reader TOME 70 (2020), FASCICULE 3 is referred there for background. Somewhat surprisingly, generic properties of eigensections of Bochner-Kodaira operators on complex line bundles do not seem to have been studied before.
The most general genericity results are stated in Theorem 4.1 in Section 4. In these results, we define "admissible data" such as (g, h, ∇), and prove the main genericity properties as different components of the data are varied. Since the general result requires some definitions from Sections 2-3, we only state the most elementary result here in the case where L = T * X is the canonical bundle, and where only the metric g on X is varied. But we also require that the Hermitian metric h on K is induced by g and that the connection ∇ h is the Chern (or equivalently, Riemannian) connection. Of course, K is not ample when g = 0, 1 but we are considering eigensections of Bochner-Kodaira Laplacians, not holomorphic sections, so ample-ness is not particularly relevant. Moreover, the proofs also work for K −1 , so that one could replace T * S 2 by the ample line bundle T S 2 or O(1) → CP 1 and obtain similar results.
Theorem 1.4. -Let (X, J) be a Riemann surface, and let L = K m . We consider Riemannian metrics g in the conformal class associated to J. We assume that h is the Hermitian metric induced by g and that ∇ is the Levi-Civita connection. Then, for generic metrics g in the class of J on X, (1) the spectrum of each Bochner Laplacian ∇ * g,h ∇ on C k (X, L m ) is simple (i.e. of multiplicity 1). Thus, the multiplicity of the eigenvalue λ = λ m,j of ∆ G is 1 if m = 0, and 2 if m = 0.
(3) all of the eigensections f m,j have isolated zeros and zero is a regular value. In particular, Z fj,m is a finite set of points; (4) If we lift sections to equivariant eigenfunctions φ, then φ and φ have zero as a regular value.
An important consequence of (1)-(2) of Theorem 1.4 is that, for generic data, all real Kaluza-Klein eigenfunctions are real/imaginary parts of equivariant eigenfunctions. Therefore, the results we prove for equivariant eigenfunctions hold for all possible eigenfunctions. Theorem 4.1 states similar results for three other types of variations of the data.
We can now state our main result. The first statement repeats (1)-(2) of Theorem 1.4 for the sake of clarity. (1) The eigenspace of ∆ G corresponding to λ = λ m,j = λ −m,j is spanned by φ m,j and φ −m,j = φ m,j . In particular, any real eigenfunction with the eigenvalue λ m,j is a constant multiple of T θ ( φ m,j ), where T θ is the S 1 action on P parameterized by θ.
(2) For m = 0, the nodal sets of φ m,j are connected.
(3) For m = 0, the number of nodal domains of φ m,j is 2.
Note from Weyl law that #{λ j,0 < Λ} ∼ Λ and that #{λ m,j < Λ} ∼ Λ 3/2 . Therefore as an immediate consequence of Theorem 1.5, we have the following: Corollary 1.6. -Let P → X be a non-trivial principal S 1 bundle with a generic Kaluza-Klein metric. For any given orthonormal eigenbasis, almost all (i.e., along a subsequence of density one) eigenfunctions have exactly two nodal domains.
The density one subsequence is of course the one with m = 0. Theorem 1.5 furnishes the first example of Riemannian manifolds of dimension > 2 for which the number of nodal domains and connected components of the nodal set have been counted precisely. The results for m = 0 may seem rather surprising, since in dimension 2 the only known sequences of eigenfunctions with a bounded number of nodal domains are those constructed in an ingenious way by H. Lewy on the standard S 2 [16] and those of Stern on a flat torus [4,18]. In those cases, the separation-ofvariables eigenfunctions have connected nodal sets but the complement of the nodal set has many components, i.e., nodal domains, saturating the Courant bound that the number of nodal domains of the jth eigenfunction (in order of increasing eigenvalue) is j. In the Kaluza-Klein case, all eigenfunctions for generic Kaluza-Klein metrics are separation-of-variables eigenfunctions and have connected nodal sets. But the connectivity is of a different kind than in dimension two and by a simple argument it induces connectivity of nodal domains. Remark 1.7. -When P → X is trivial, and P ∼ = S 1 ×X is endowed with the product metric, we have φ m,j = ψ j e imθ where ψ j is an eigenfunction of ∆ g on the base X. Hence φ m,j = ψ j cos mθ has many nodal domains, and the last statement in Theorem 1.5 fails. Hence, the "generic" set of metrics is not the full set of metrics. See Section 9.1 for flat tori for a product setting where the nodal results above do hold.
Outline of the proof
The nodal set of the lift real and imaginary parts of the lift φ m,j of f m,j e m L is very different over nodal versus non-nodal points of f m,j . Let (1.5) Σ := π −1 Z fm,j be the inverse image of the base nodal points. It is a union of fibers and is a finite union of fibers if and only if f m,j has a finite number of zeros. We refer to Σ as the "singular fibers" or singular set. We denote by X\Z fm,j the punctured Riemann surface in which the zero set of f m,j is deleted. A key statement in the nodal analysis is the following: is an m-fold covering space.
It follows that the topology of the nodal set is entirely determined by the combinatorics of gluing the sheets along the singular fibers. In fact, the gluing is rather simple and easily yields the following Theorem 1.9. -For all m = 0, the nodal set N φm,j is connected.
To count nodal domains, we need to make the assumption that there are just a finite number of zeros of f m,j and that at least one of them is regular.
When the zero set is transverse to the zero section, then the sum of the indices of the zeros is the first Chern class of K m , and in particular is non-empty when the genus of X is = 0, i.e., when X is not a torus.
For metrics satisfying Theorem 1.4 we prove Theorem 1.5 by using Proposition 1.8 together with some geometric observations on how the sheets fit together at the singular fibers. This is done using a Bers type local analysis of the eigensections (Section 5) and some geometric/combinatorial arguments in Section 8.
To put the nodal results into context, it is proved in varying degrees of generality in [8,9,11,12,13,14,17,22] that in dimension 2, the number of nodal domains of an orthonormal basis {u j } of Laplace eigenfunctions on certain surfaces with ergodic geodesic flow tends to infinity with the eigenvalue along almost the entire sequence of eigenvalues. By the first item of Theorem 1.5, the same is true for their lifts to the unit tangent bundle SX as invariant eigenfunctions of the Kaluza-Klein metric. But for higher weight eigenfunctions, the situation is virtually the opposite and the number of nodal domains is bounded.
ANNALES DE L'INSTITUT FOURIER
Remark 1.10. -Note that the geodesic flow on P with a Kaluza-Klein metric never is ergodic. To see this, observe that the hypersurface is invariant under the geodesic flow, and it divides SP into two subsets of positive measure:
Surfaces of constant curvature
Metrics g of constant curvature with their associated Hermitian metrics and connections on K → X are not generic. But they are of special interest, so we comment on what we are able to prove about them. Note that the standard metric on T 3 is Kaluza-Klein, as is the standard metric on S 3 or SO(3) = U (S 2 ). The standard metric on the unit tangent bundle PSL(2, R)/Γ over a hyperbolic surface is Lorentzian, but if one changes the sign of the vertical Laplacian it is also Kaluza-Klein.
Perhaps surprisingly, the results of Theorem 1.5 are valid for some orthonormal bases of eigenfunctions on flat 3-tori.
Theorem 1.11. -On the flat 3 torus T 3 , one can find an orthonormal eigenbasis for which all nonconstant eigenfunctions have two nodal domains.
Next we turn to hyperbolic metrics on a surface X of genus g 2. Then the total space P h = PSL(2, R)/Γ and the equivariant eigenfunctions of the Kaluza-Klein Laplacian ∆ are the same as joint eigenfunctions of the generator W of K and of the Casimir operator Ω. When the weight m is fixed, one may separate variables and obtain a Maass Laplacian D m on smooth sections of a complex line bundle π : K m → X, namely the bundle of m-differentials of type (dz) m , and are the usual weight m automorphic Maass eigendifferentials f m,j (z)(dz) m , Unfortunately, we are not able to verify that (any) Maass eigendifferentials have 0 as a regular value, i.e. the generic conditions needed for Theorem 1.5. Indeed, we do not know how to prove that (any) eigenfunctions φ 0,j of the TOME 70 (2020), FASCICULE 3 hyperbolic Laplacian have a discrete set of critical points, much less that 0 is a regular value of dφ 0,j . Regarding spherical harmonics on S 3 we will be brief, because we study random equivariant spherical harmonics in detail in a forthcoming article [15]. The round metric is a Kaluza-Klein metric, but of course not a generic one. In Section 9.2 we show that the joint eigenfunctions Y m1,m2 N of ∆ and of two commuting S 1 actions have very different nodal sets from the ones in Theorem 1.5. On the other hand, in [15] we show that "random" linear combinations of such Y m1,m2 N with m 1 fixed do satisfy the results of Theorem 1.5 and the nodal sets of their real, resp. imaginary parts, have just one nodal component. We also find their expected Euler characteristic. These results were motivated by the numerical discovery of Barnett et. al. that 3D random spherical harmonics on S 3 of fixed degree, the nodal set contains one "giant component" and many small components [2]. We are fixing the weight m 1 and therefore do not work with general random spherical harmonics. But in the N dimensional subspaces where m 1 is fixed with |m 1 | N, 2|N − m 1 , the nodal sets of the real and imaginary parts are connected and divide S 3 into just two components. For further discussion we refer to [15].
Geometric background
In this section we discuss the geometric data that goes into the construction of Kaluza-Klein metrics, which are defined in Definition 1.1. They are also the data needed to define Bochner Laplacians ∇ * ∇ and Kaluza-Klein Laplacians ∆ G . We plan to vary the data and study perturbation theory of eigenvalues and eigensections in Section 4.
Riemannian metrics on X and Hermitian metrics on L
Let (X, J, g) denote a Riemann surface with complex structure J and Riemannian metric g. We write g 11 = g( ∂ ∂z , ∂ ∂z ) and g 11 = g * (dz, dz), where g * is the dual metric. The complex structure gives a decomposition of T * X ⊗ C = T * (1,0) ⊕ T * (0,1) into (1, 0) resp. (0, 1) parts. We denote the area form of g by where the Kähler form ω is the (1, 1) form defined by g J (X, Y ) = ω(JX, Y ). Locally there exists a Kähler potential ψ defined up to a constant by 2.1.1. The space of isometry classes of metrics on surfaces It is well-known that the space of Riemannian metric tensors on a manifold X splits as a product Vol(X)×Met µ (X) of volume forms times metrics with a fixed volume form. Thus, one may separately consider metrics with a fixed volume form and conformal classes of metrics.
Choice of a complex structure J on X is equivalent to choice of a conformal class Conf(g 0 ) of metrics. The moduli space of conformal classes is the same as the (3g − 3)-dimensional moduli space M g of complex structures on X. In each conformal class, we may pick a background metric g 0 and represent other metrics in the form If we fix a complex structure J, then the Riemannian metrics in the corresponding conformal class are Kähler metrics, and may be parameterized by their Kähler potentials φ, where the area form ω φ of the Kähler metric is related to that of the reference metric by We let which may be identified with an open set in C ∞ (X). The Liouville field and Kähler potential are related by where ∆ 0 is the Laplacian of g 0 . The only difference in the two parameterizations of conformal metrics is that the area of metrics in K ω is fixed while it may vary in Conf(g 0 ). Thus Conf(g 0 ) ∼ = K ω × R.
In the case of Riemann surfaces, the area form is the symplectic form associated to a Kähler metric. Given a complex structure J, the Kähler metric g J can be recovered from its area form ω by the formula g J (X, Y ) = ω(X, JY ). Hence isometry classes of Kähler metrics with a fixed area form are parameterized by M g .
Complex line bundles L → X, connections and curvature
If we fix a complex structure J and holomorphic line bundle L → X, then a Hermitian metric h on L is determined by the length of a local holomorphic frame e L (i.e., a local holomorphic nonvanishing section) of L over an open set U ⊂ M ) by e −ψ = e L 2 h , where e L h = h(e L , e L ) 1/2 denotes the h-norm of e L .
Connections
In the real setting, a connection on a vector bundle E defines a covariant derivative In our complex setting, we assume L is a holomorphic Hermitian line bundle, i.e., we equip L with a complex structure J L , a connection ∇, and a Hermitian metric h. In a local frame e L it is defined by ∇e L = α ⊗ e L . We consider several types of compatibility conditions between this data: • An h-connection ∇ h is one compatible with h. In a unitary frame, the connection 1-form is iR valued and is denoted by iα. We denote the space of h-compatible connections by A h . • Or a J L -compatible connection. In a holomorphic frame e L the connection 1-form α is of type (1, 0). We denote the space of J Lcompatible connections by A C . • The unique Chern connection α J L ,h which is compatible with both J L , h.
This data induces: • The Hermitian metric h induces the principal bundle of h-unitary The connection 1-form in the frame e L is given by We denote the (1, 0) resp. (0, 1) parts of ∇ by ∇ 1,0 , resp. ∇ 0,1 .
Suppose that ∇ ∈ A C . Then if s = f e with e a local holomorphic frame,
ANNALES DE L'INSTITUT FOURIER
The holomorphic line bundle L also has a natural Cauchy-Riemann operator, . In a local holomorphic frame e, we write a smooth section s = f e and then It is well-defined since if e is another holomorphic frame and e = ge , then s = f ge and ∂ L s = ∂f ⊗ ge = ∂f ⊗ e.
The Chern connection ∇ associated to the Hermitian metric h is the unique metric connection ∇ : C ∞ (X, L) → C ∞ 1 (M, X ⊗ T * X) whose connection 1-form in a holomorphic frame e L has type (1, 0). The connection 1-form is given by ∇e L = α ⊗ e L with α = ∂ log |h|.
The metric g * is a Hermitian metric on K. Any Hermitian metric h on a line bundle L induces metrics h m = e −mφ on the tensor powers L m in the local frame e m L . The Hermitian metric and complex structure determine a Chern connection ∂ log h whose curvature 2-form Θ h is given locally by
Curvature form
Given a connection ∇ on L and a vector field V on X, the covariant derivative of a section s is defined by ∇ V s = ∇s, V . The curvature is W ] . If e L is a local frame and ∇e L = α ⊗ e L then Ω ∇ = dα.
Examples
• Let F * X be the unit co-frame bundle X, consisting of orthonormal frames of T * X. Then S m F * X is the bundle of real m-differentials, i.e., homogeneous polynomials of degree m in dx, dy or dz, dz. • When X is given a complex structure, we may decompose T * X ⊗ C = T * (1,0) ⊕ T * (0,1) into co-vectors f dz of type (1, 0) and gdz of type (0, 1). The holomorphic co-tangent bundle is usually denoted by K = K X = T * (1,0) and is called the canonical bundle. Its tensor powers K m are bundles of differentials of type f (dz) m with f (z,z) a smooth function.
• When the genus is 0, i.e., X = S 2 , K X is a negative line bundle and has no holomorphic sections. The associated circle bundle of frames is SO(3) ∼ = RP 3 = S 3 / ± 1. • When the genus is 1, then K X and T * X are trivial and F h ∼ = • When the genus is 1 we may twist L by a flat line bundle. This is not particularly relevant for this article except that we usually ignore this additional degree of freedom. There is an ample line bundle L → T 2 whose holomorphic sections are theta functions. The associated principal S 1 bundle is the reduced Heisenberg group, the quotient of the simply connected Heisenberg group by the integer lattice. • When the genus is 2 then X = H 2 /Γ where Γ ⊂ PSL(2, R). The associated S 1 bundle is SL 2 (R)/Γ. K X is ample and for m large there are many holomorphic sections of K m X . This is only significant in this article when we discuss splitting eigenspaces.
Canonical bundle and m-differentials
Let K → X denote the canonical bundle T * (1,0) X of (1, 0) forms f dz. Up to twisting by a flat line bundle, it is the unique ample line bundle on X. Hence there exists a Hermitian metric h on K with curvature form i∂∂ log h = ω. This should be distinguished from the curvature of g, which is given by dd c log g 11 = Kω h , where K is the scalar curvature.
In terms of the Hermitian metric h = e −φ0 on K, |dz| g = e −φ0 = g 11 . Also, We now regard g(X, Y ) = ω(X, JY ) as a Kähler metric. The co-metric * defines metric coefficients on T * X ⊗C by extending g * by complex linearity and induces the Hermitian metric, The associated (1, 1) form ω h is positive if the genus is 2 and if g is a metric of negative curvature K. It is negative if the genus is 0 and the metric g is of positive curvature. When the genus is 0 there do not exist Hermitian metrics with strictly positive or negative curvature forms.
ANNALES DE L'INSTITUT FOURIER
When dim X = 2 we write the area form as dA g = √ gdx or as ω g . The metric g induces metrics g on T X, T * X, T * (1,0) X = K X and on powers such as K m . On K m , the Hermitian metric induced by g is dz 2 g = e −φ = g 11 so φ = − log g 11 . The Chern connection on K is the same as the Riemannian connection.
Consider the complex line vector bundles T 1,0 X and (T X, J). They are isomorphic under the map X , the Chern connection D on the holomorphic tangent bundle T 1,0 is the Levi-Civita connection ∇.
Orthonormal frame bundles and m-differentials
If (X, g) is a Riemannian surface, and F is the SO(2)-bundle of orthonormal frames of T * X, then g determines a Riemannian connection on F . Similarly, if we fix a complex structure and define the principal S 1 bundle F h ⊂ T X associated to K X = T * (1,0) , then g determines a Hermitian metric on K X . We first discuss the real geometry and then the complex geometry.
The metric g = 2 i,j=1 g ij dx i ⊗ dx j on T X induces a co-metric g * on T * X, usually denoted by raised indices. It then induces metrics g ⊗n on powers S m T * X.
There always exists a basis of basic or horizontal 1-forms at a frame The Riemannian connection 1-form α is defined by the equations, where K is the scalar curvature. Dually, there exist vector fields ξ 1 , ξ 2 so that FASCICULE 3 Then define Then, The vector fields ξ 1 , ξ 2 are also of unit length since they are defined to be horizontal lifts of a unit frame.
Hilbert spaces of sections
Let (L, h) → X be a Hermitian holomorphic line bundle. We thus have a pair of metrics, h resp. g (with Kähler form ω φ ) on L resp. T X.
To each pair (h, g) of metrics we associate Hilbert space inner products where |s(z)| 2 h m is the pointwise Hermitian norm-squared of the section s in the metric h m . In a local holomorphic frame e L , we write In local coordinates z and the local frame e m L of L m , we may write s = f e m L and then In the special case where L = K X , we may use the frame dz in a local holomorphic coordinate z. In the local frames (dz) m of K m we may write sections as s = f (dz) m and then |s(z)| 2 h m = |f (z)| 2 e −mψ(z) and then, where dV g = ω φ is the area form of g.
Bochner Laplacians on line bundles
In this section, we give explicit local formulae for Bochner Laplacians where (L, h) → X is a Hermitian holomorphic line bundle, g is a metric on X, ∇ is a connection on L. In a local frame e L of L, with s = f e L , the inner product Hilb(g, h) on L 2 (X, L) takes the form, The inner product on L 2 (X, L ⊗ T * X) has the form, With no loss of generality, we fix J on X and assume that (g, It is a Hermitian metric on T 1,0 X and is compatible with J. (1) We also denote the Riemannian volume form by dV g = ω = dd c log g 11 . Remark 3.1. -Notational remark: We use G rather than g −1 or g * for the dual co-metric on 1-forms, because it is a convenient notation for later variations.
The Bochner Laplacian is the Laplacian on L 2 (X, L) determined by the quadratic form, Throughout we assume that g is J-compatible. In a local frame e L of L, , and the quadratic form is given by The adjoints are taken with respect to the volume form e −ψ dV g . We give local formulae for ∇ * h,g ∇ under several assumptions on ∇ and in correspondingly adapted frames (equivalently, choosing a gauge for ∇): (i) ∇ is h-compatible (see Section 3.1); in this case, we compute in a local unitary frame. Fixing h is equivalent to fixing the principal S 1 bundle P h → X, and varying the connection 1-forms α ∈ A h on P h (with fixed g). (ii) ∇ ∈ A C is compatible with a fixed complex structure J L on L (see Section 3.2); in this case we compute in a local holomorphic frame.
In the next section we fix J L and vary h (with fixed g). (iii) ∇ is compatible with both (h, J L ), hence is the Chern connection; see Section 3.3. The (1, 0)-part of the connection has the form α = ∂ log h and is parameterized by the Hermitian metric h on L; in the next section we consider its variation with h (with fixed g). (iv) When L = K X is the canonical line bundle, we let h be the Hermitian metric induced by g and let ∇ be the Levi-Civita connection. This is a special case of an h-compatible connection but is special because h is induced by g. Moreover, the Riemannian connection w.r.t. g is the Chern connection for the Hermitian metric g. In a holomorphic frame dz the connection form is ∂ log g 11 = ∂ log G(dz, dz). Also, dV g = g 11 dz ∧ dz. In the next section, we vary this connection by varying g on X.
There exist many formulae for Bochner Laplacians in the literature (see for instance [5]), but they often make assumptions on the compatibility of the connection with other data (the Hermitian metric or complex structure) and we need explicit dependence on the compatibility conditions so that we can perturb some of the data while holding others fixed. We therefore go through the calculations with explicit assumptions on the compatibility of ∇ with the data (g, h, ∇, J, J L ).
We also recall the general identities,
Calculation in a unitary frame
In this section we assume that ∇ is compatible with h. We recall from Section 2 that on a Hermitian line bundle (L, h), the set A h of connections on L which are compatible with the Hermitian metric is the affine space and Ω 1 (X) are the real 1-forms on X. The Hermitian metric determines the principal U (1) bundle P h of unitary frames of L and as before A 1 determines ANNALES DE L'INSTITUT FOURIER a connection 1-form α 1 on P h . On the base X, the connection 1-form is iR-valued in a unitary frame and we write it as iα with a real-valued α.
where ∆ g is the scalar Laplace operator.
Proof. -In a unitary frame, |e L | 2 h = e −ψ = 1 and this factor drops out. We leave it in until the last step for purposes of later comparison to other frames. Since ∇(f e L ) = df ⊗ e L + if αe L , and by (3.1), is the Hermitian norm-squared, so where in the last line we use that ψ = 0 in a unitary frame. Since α is real-valued, Thus, we get
Holomorphic line bundles: J L -compatible connections.
In this section we give a local formula for the Bochner Laplacian when L → X is a holomorphic line bundle and ∇ is compatible with the complex structure. Thus, complex structures J on X and J L on L are fixed. In a holomorphic frame, e L = e −ψ = 1 and ∇e L = α ⊗ e L , where α is of type (1, 0). We write ∇ = ∂ ∇ + ∂ ∇ for the decomposition of a connection into
The analogue of Proposition 3.2 is
Proof. -The proof is similar to that of Proposition 3.2, with two differences: (i) we use a holomorphic frame rather than a unitary frame and |e L | 2 h = e −ψ is not equal to 1; (ii) α is of type (1, 0) rather being iR-valued.
Note that , and integrating by parts the |df | 2 g term, and with | · | 2 denoting the Hermitian metric, we get
ANNALES DE L'INSTITUT FOURIER
We simplify the f G(df , α) term using that d Integrating the d * by parts gives
Canonical bundle: L = K and ∇ is the Riemannian connection
Let z be a local holomorphic coordinate and let dz be the associated section of K. Differentials of type (dz) m are sections of K m , the m-th power of the canonical bundle. The Riemannian metric on X induces a Hermitian metric h m on K m , namely |dz| h = |dz| g where g is the cometric. dz = dx + idy and at x + iy, |dz| = y.
The metric g on T X endows a Hermitian metric g * on K and the associated Riemannian connection ∇ g is the Chern connection with connection 1-form α = −∂ log g 11 in the frame dz. For simplicity of notation we write φ = − log g 11 . It induces connections and Hermitian metrics on K m with connection 1-forms mα. The associated Bochner Laplacian ∇ * m,g ∇ m,g on K m corresponds to the quadratic form
ANNALES DE L'INSTITUT FOURIER
Note that ω g = i 2 g 11 dz ∧ dz and the Laplacian on scalar functions is given Proposition 3.6. -Let ∇ m be the Chern connection for (K m , g * m ). Then, Proof. -This follows from Proposition 3.4. We give a direct proof. By the Bochner-Kodaira formula (3.2), it suffices to prove As above, we calculate the adjoint to be It follows that
Perturbation theory and genericity
In this section we prove generic properties of the eigenvalues and eigensections of Bochner Laplacians ∇ * g,h ∇ on complex holomorphic Hermitian line bundles (L, h) → X. Our ultimate goal is to deduce generic properties of Kaluza-Klein Laplacians on the principal U (1) frame bundles P h → X associated to h. First we discuss generic properties of Bochner Laplacians on the line bundles and then we draw conclusions for the Kaluza-Klein Laplacians. We prove that for generic data (g, h, ∇) (with fixed (J L , J)), eigenvalues of Bochner Laplacians ∇ * g,h ∇ are simple (multiplicity one) and all eigensections intersect the zero section transversally (i.e., have 0 as a regular value). This immediately implies that for the associated Kaluza-Klein Laplacians ∆ G on P h , all joint eigenfunctions of the U (1) action and ∆ G have simple joint spectrum and have 0 as a regular value. In Section 4.6, we discuss the multiplicity of the spectrum of ∆ G , hence proving a part of Theorem 1.5.
The main result of this section is: -For generic "admissible data" described below, and for every m, the spectrum of each Bochner Laplacians ∇ * g,h ∇ on C k (X, L m ) is simple and all of its eigensections have zero as a regular value. Moreover, if we lift sections to equivariant eigenfunctions φ, then φ and φ have zero as a regular value.
The generic admissible data is of the following kinds: (1) We fix h, g and vary the connection ∇ in A h . Fixing h is equivalent to fixing the principal U (1) bundle P h → X, and varying the connection 1-forms. (2) We fix (J, J L , g) and vary both h and ∇, assuming that ∇ ∈ A C is compatible with J L on L but not necessarily with h. (3) We fix (g, J L , J) and vary (h, ∇) assuming that ∇ is compatible with both (h, J L ), hence is the Chern connection of (L, J L , h). (4) We fix L = K m and also fix J and vary g in the conformal class associated to J. We assume that h is the Hermitian metric induced by g and that ∇ is the Levi-Civita connection.
The proofs in each of the cases are given in separate sections. Note that the functions relevant to this article are smooth sections of a complex line bundle L, and may locally be represented as complex valued functions u. We will prove that u : M → C has zero as a regular value, i.e., that du p = d u+id u is surjective. It follows that u, u are independent and nowhere vanishing on their zero sets, and that each has zero as a regular value
The Uhlenbeck framework
To study generic properties of the spectrum, we follow [20] and work with C r spaces of metrics and connections. We use the following notation:
ANNALES DE L'INSTITUT FOURIER
• We denote by G r (X) the Banach space of C r metrics on X. Since X is a surface and we usually fix the complex structure J, we only work with C r metrics in the associated conformal class Conf(J) and represent them in the usual Weyl gauge g = e ρ g 0 relative to a fixed background metric g 0 ∈ Conf(J). Thus, we may identify G r (X) ∼ = C r (X). We may also fix the area of the metrics with no loss of generality and then Conf(J) may be identified with the space K ω of Kähler metrics on X in a fixed cohomology class. This is simply a different choice of gauge in which we write the Kähler forms as ω φ = ω 0 + i∂∂φ and use the potentials φ rather than the Weyl gauge u to parameterize metrics. • We denote by H r (L) the Banach space of C r Hermitian metrics on L. Once we fix a local frame e L we may identify h ∈ H r (L) with the function ψ such that e L (z) 2 h = e −ψ(z) , and H r (L) is then equivalent to C r (X) except of course that the identification is frame dependent and the frame is only local (defined on the complement of a smooth closed curve in X, e.g.).
• We denote by A r (L) the space of connections with C r connection forms. As before, we also denote by A r h , resp. A r C , the h-compatible (resp. J L -compatible) C r connections.
• We denote by C r (X, L) the C r sections of L. We also denote by H s (X, L) the Sobolev space of sections with s derivatives in L 2 .
We define Here, the eigenvalue parameter λ in the domain is allowed to be complex even though at zeros of Φ L it is always real. This does not change the arguments in [20] but is needed so that λs spans the eigenspace when s is an eigensection. In [20] the eigenfunctions were real-valued, so this issue did not arise.
Recall that a linear map between Banach spaces is Fredholm if it has closed image and finite dimensional kernel and cokernel. The index of a Fredholm operator is the difference of the dimensions of its kernel and cokernel. A nonlinear map Φ : N → Y of Banach manifolds is Fredholm if its derivative DΦ n is Fredholm for every n ∈ N .
Our first goal, roughly speaking, is to prove that Φ is a Fredholm map of index 0, i.e., to prove surjectivity of the differentials D 2 Φ from tangent to L 2 (X, L). It is sufficient to pick the relevant types of frames and calculate the Bochner Laplacians in the frame as in Section 3.
Regarding the surjectivity, we need to prove density of the image and that the image is closed. Some care needs to be taken because sections of complex line bundles are "vector-valued", i.e., have two real components. As explained in [6], there are pitfalls to avoid when generalizing the arguments of [20] to the vector-valued case. But sections of line bundles are locally complex-valued functions and are essentially scalar functions, albeit with scalars in C.
Uhlenbeck's argument
We briefly review Uhlenbeck's proof that for generic metrics on compact C r Riemannian manifolds, all eigenvalues are simple and all eigenfunctions have 0 as a regular value.
Her framework is quite general and therefore uses the notation B for the relevant space of metrics or other geometric data, and L b for the Laplacian associated to b. The relevant functions are denoted by u and the space of such functions on a manifold M is denoted by C k (M ), even though they could be sections of a bundle over M . Then define Φ(u, λ, b) = (L b + λ)u, and put Then, We often write v =u, η =λ, D 2 (Φ)s = λ∆u, (∆ + λ)u + (∆ +λ)u = 0.
Further, let D 1 α denote the derivative of α along Q. Then,
ANNALES DE L'INSTITUT FOURIER
Also define J to be the image of D 2 Φ, Eigenfunctions and eigenvalues move continuously under perturbations of the operator. So it is easy to show that the set of metrics with for which the jth eigenvalue is simple is open. The difficulty is to prove that this set is dense.
To prove the first statement in Theorem 4.1 we need to verify the hypotheses of Theorem 4.2 and therefore need to prove Proposition 4.3, i.e., to determine the range of D 2 φ.
Base metric variations
In this section we fix (h, ∇, J, J L ) and vary only g = e ρ . Equivalently, we consider Kaluza-Klein metrics on a fixed U (1) bundleP h → M with a fixed connection α and vary the base metric g. Proof. -By Proposition 3.2, where ∆ g f is the scalar Laplace operator, where g = e ρ . Taking the variation δ with respect to ρ (and designating the variation with a dot), But each term is conformal to that of g with conformal factor e −ρ . Hence To prove that the image of D 2 Φ L is dense we argue by contradiction and suppose that there exists W ∈ L 2 (X, L) such that X δ g ∇ * ∇h(f e L , W (z)) h dV g = 0 for all ρ. But this implies that X ρ(f e L , W ) h dV g = 0 for all ρ, then W = 0. Write W = F e L so that the integral becomes, X ρfF e −ψ dV g = 0 for all ρ. This is only possible if fF e −ψ = 0. But f and e −ψ can only vanish on a set of measure zero, so F ≡ 0 almost everywhere. The image is closed because ∇ * ∇ is a Fredholm operator.
Varying the Hermitian connection
In this section we fix g, h and vary ∇ ∈ A h . In the application to Kaluza-Klein metrics, P h is fixed and the base and vertical metrics are fixed and only the splitting into horizontal and vertical is varied.
We recall from Section 2 that some of the variations are "trivial", i.e., are within a gauge equivalence class. Bochner Laplacians with gauge-equivalent connections are unitary equivalent by a gauge transformation, i.e., they have the same spectrum and their eigensections are related by a gauge transformations. Viewed in terms of line bundles over X, gauge equivalent connection forms are connection forms of a single connection in two different unitary frames, hence differ by a gauge transformation e iθ ∈ Map(X, S 1 ) taking e L → e iθ e L . The connection 1-form then changes by idθ ∈ Ω 1 (X, R). A unique representative of a gauge equivalence class is defined by the Coulomb gauge d * a = 0.
where ∆ g f is the scalar Laplace operator. Taking the variation with respect to α gives, for allα ∈ Ω 1 (X). We integrate d * g by parts to get, We may assume that the frame e L is unitary so that ψ = 0. If β ∈ Ω 1 (M, C) and X G(β, ν)dV g = 0 for all ν ∈ Ω 1 (M, R), then β = 0. Indeed, we may consider ν of the types ν = ν 1 dx, ν 2 dy separately to get orthogonality of the components β j with ν j . This reduces matters to the fact that if u, v are complex-valued and uvdV g = 0 for all v, then u ≡ 0. We conclude that On any open set U where f, F = 0 we may divide by ifF and write the solution as, This implies that on a dense open set and since α ∈ C ∞ , it is everywhere closed and hence the curvature of (L, h) is zero. This is impossible unless L is a topologically trivial line bundle, and the contradiction implies that F ≡ 0 except when dα = 0.
Proof of Theorem 4.1
As mentioned in the previous section, eigenfunctions move continuously under perturbations of the operator. So it is easy to show that the set of metrics with for which the jth eigenvalue is simple is open. The difficulty is to prove that this set is dense.
To prove the first statement in Theorem 4.1 we need to verify the hypotheses of Theorem 4.2 and therefore need to prove Proposition 4.3, i.e., to determine the range J of D 2 φ.
To complete the proof of Theorem 4.1 it suffices to prove: Proof. -Let G m,λ (z, w) be the kernel of the Green's function G m,λ : [ker(D m + λ)] ⊥ → [ker(D m + λ)] ⊥ for D m (g) + λ for a given background metric g. As above, one may use the Hermitian metric h on K or the associated Kähler metric g = ω J as the parameter space of metrics.
We need to show that for each x ∈ M , has 0 ∈ C as a regular value, i.e., that is surjective to C, where D 1 is the differential along Q with x ∈ M held fixed. Since x is fixed we may use a local coordinate z and frame (dz) m Equivalently, G m,λ (x, y) + u λ (x) = 0. This is not possible and the contradiction ends the proof.
Multiplicity of the spectrum of ∆ g
We begin by observing that λ m,j = λ −m,j and that φ −m,j = φ m,j . Then any real eigenfunction which is a linear combination of φ m,j and φ −m,j is e iθ0 φ m,j for some constant θ 0 . In local coordinates, φ m,j is φ(z)e imθ , and therefore we see that e iθ0 φ m,j = (T θ0 φ m,j ) = T θ0 (φ m,j ) .
For m 1 and m 2 such that |m 1 | = |m 2 |, we argue that λ m1,j1 = λ m2,j2 is satisfied for an open dense subset of metric G. This immediately implies the first, third, and the fourth statement of Theorem 1.5. Note that the eigenvalue moves continuously with respect to G. So it is sufficient to prove that Lemma 4.9. -Let P → X be a non-trivial principal S 1 bundle. Fix integers m 1 and m 2 such that |m 1 | = |m 2 |. Among all S 1 -invariant metric G on P , G satisfying λ m1,j1 = λ m2,j2 is dense.
Proof. -The deformation of the base of the Kaluza-Klein metric does not touch the vertical operator ∂ ∂θ and therefore the first order perturbation equations for infinitesimal deformations of the base metric g gives, Taking the inner product with φ m,j gives If there exist weights m 1 = m 2 for which we cannot split the eigenvalue λ m1,j1 = λ m2,j2 then for all infinitesimal base perturbations ρ we get Write φ m,j = f m,j (dz) m . Differentiation of the eigenvalue equation therefore gives the well-known formula Ḋ m1 f m1,j1 , f m1,j1 = Ḋ m2 g m2,j2 , g m2,j2 for every variation of g, where the inner product is that of g 0 . Recall from previous section that∆ H = ρ∆ H . Because −∆ H φ m,j = (λ m,j − m 2 )φ m,j we have for any ρ ∈ C ∞ (X), Thus, Integrating both sides against dV g and using that both eigenfunctions are L 2 normalized gives (λ m1,j1 − m 2 1 ) = (λ m2,j2 − m 2 2 ), i.e., |m 1 | = |m 2 |.
Local structure of eigensections at zeros
To study the nodal sets of real and imaginary parts of Kaluza-Klein Laplacians, we first study the zeros of the associated sections of the line bundles. For simplicity of exposition, we assume that L = K and describe the zero sets of eigen-m-differentials. Essentially the same discussion is valid for other line bundles.
We follow the notation and terminology in the theory of holomorphic quadratic differentials, even though our eigendifferentials are C ∞ , usually not holomorphic and of general weight m. Following a standard terminology for quadratic differentials, we call a point z such that f m,rj (z) = 0 a "regular point" and a point where f m,rj (z) = 0 a "critical point" or a "singular point".
After the first version of this article was written, we located some recent articles generalizing the geometric properties of quadratic differentials on Riemann surfaces to C ∞ higher order differentials [7,1] and to other line bundles. We now use the terminology and results of these articles but have retained some from our first version since it is important for us to lift to P h .
Trajectories of eigen-differentials.
The real and imaginary parts of the eigendifferentials ω m,j =f m,j (z)(dz) m are called binary differentials of degree m and the equation for the zero set of ω m,j is called a binary differential equation of degree m [7]. It is traditional to consider the nodal set f m,j (z)(dz) m = 0. If there exist exactly m solutions at a regular point where ω m,j (z) = 0 then ω m,j is called totally real in [7]. Our m-differentials are of a special type since they are real and imaginary parts of f m,j (z)(dz) m and therefore only have terms of the form (dz) m or (dz) m . The following is the key input into Proposition 1.8.
The kernel of f m,j (dz) m defines a smooth m-valued distribution on X with singularities where ω m,j = f m,j (dz) m = 0. The m line fields defines a web of m transverse singular foliations, whose leaves are called the trajectories: [19,Section 7] for holomorphic quadratic differentials. Illustrations of webs for higher order real differentials can be found in [7].
It lifts to a smooth curve (γ z0,θ0 (t),γ z0,θ0 (t)) in the nodal set upstairs. Since dπ is an isomorphism, the trajectories are special curves on the nodal set φ m,j = 0.
Non-degenerate singular points
The structure of the trajectories through a singular (zero) may be complicated in general if no conditions are placed on the degeneracy of the zeros. The purpose of Theorem 4.1 is to allow us to assume that the zeros are of first order, so that they are isolated and non-degenerate.
The structure of the trajectories of a totally real m-differential near an isolated singular point is discussed in [7]. As with vector fields, the key topological invariant of the singular point is its index.
Equivalently, in a small circle C around z 0 , choose a unit vector ] → X is a constant speed parametrization of C and let = L(C) be its length. Let X(t) be a smooth extension of X(0) along C(t). After a complete turn, X(2π) must be one of the 2m solutions of ω(X) = 0. After 2m turns X(2m ) = 0. Let θ(t) be a smooth determination of the angle between the tangent line to C and X(t). Then θ(2m ) and θ(0) differ by an integer multiple of 2π. The index of z 0 is defined by Thus, the index has the form s 2m with s ∈ Z. The following Lemma shows that singular points must exist when the genus of X is non-zero.
around a nodal point.
TOME 70 (2020), FASCICULE 3
Let p be a nodal point of f m,j . We Taylor expand the coefficients in Kähler normal coordinates for (J, g) in a disc z ∈ D(p, r) to get • g 11 = 1 + K(p)|z| 2 + · · · ; • ∂φ ∂z = ∂φ ∂z (p) + ω pz + · · · =z + · · · . Thus, the osculating constant coefficient operator is Let P k denote homogeneous polynomials of degree k in z,z. It is better to arrange the terms of the Taylor expansion of D m at p into terms Note that L −1 = 0 because dg 11 (p) = 0 and ∂φ(p) = 0, so neither the second or first derivative terms contribute at this order.
Also expand where f [k] ∈ P k is homogeneous of order k.
The following is the generalization of the Bers local expansion theorem to complex line bundles.
Lemma 5.8. -Let z 0 be a zero of f m,j (dz) m . The first non-zero homogeneous term f [n] of the Taylor expansion of an eigenfunction is a harmonic homogeneous polynomial. If the order of vanishing is n, f [n] (z) = a z n + ib z n . In particular, at a non-degenerate zero, the first homogeneous term is f m,j = a z + ib z.
is a homogeneous harmonic polynomial. In real dimension 2 the only possibilities are linear combinations of the real and imaginary parts of z k . By a well-known argument, the nodal set of the real and imaginary parts of f are topologically equivalent to those of the leading order homogeneous term.
This completes the proof of the Proposition 5.7.
Adapted Kaluza-Klein metrics
All of the Kaluza-Klein metrics are Riemannian metrics on principal S 1 bundles P h → X associated to C ∞ complex Hermitian line bundles L → X. Given P h we recover L as an associated line bundle. Let (X, g) be any Riemannian surface. We denote the genus of X by g. Let (L, h) → X be any complex line bundle with Hermitian metric h. Associated to L is the U (1) bundle P h of orthonormal frames. Let T = ∂ ∂θ generate the S 1 ∼ = U (1) action. We endow P h with a connection α, that is, an S 1 invariant 1-form on P h such that α(T ) = 1.
The connection defines a splitting into horizontal and vertical spaces. The vertical space is given by orbits of the S 1 action. The horizontal space is defined by H p = ker α and is isomorphic under dπ p to T z X where π(p) = z.
Definition 6.1. -The Kaluza-Klein metric on P h is the S 1 -invariant metric G such that the horizontal space H p := ker dα is isometric to T π(p) X, so that V = R ∂ ∂θ is orthogonal to H and is invariant under the natural S 1 action and so that the fiber is a unit speed geodesic.
A Kaluza-Klein metric on the principal S 1 bundle P h is thus determined by the pair (g, α) where g is a metric on X and where α is a connection 1form on F . In general, the metric and connection are chosen independently. In Section 2.6 we discuss the orthonormal frame bundle, where α is the Riemannian connection of g.
Given P h and any character χ m = e imθ of S 1 we obtain associated line bundles (resp, real rank 2 bundles) by For purposes of this paper it may be assumed that m 0.
We often assume that X is equipped with a complex structure J and that L is a holomorphic line bundle. Let D * h ⊂ L * be the unit co-disc bundle with respect to h and let P h = ∂D * h be its boundary, an S 1 bundle π : P h → X. Remark 6.2. -Not all S 1 invariant metrics on P h are adapted Kaluza-Klein metrics. It would be interesting to consider more general S 1 -invariant metrics on SX or on other manifolds (of all dimensions), as well as invariant metrics under more general compact Lie groups.
Geometry and analysis of Kaluza-Klein metrics
We use the term Kaluza-Klein metric or Kaluza-Klein metric in the sense of Definition 6.1 to denote metrics G on the unit tangent bundles π : P h = SX → X over surfaces for which the vertical and horizontal spaces are orthogonal and which is invariant under the free S 1 = SO(2) action (2) . They are special cases of Riemannian submersions with totally geodesic fibers isometric to R/2πZ. Definition 6.3. -If H is a compact Lie group, and if π : M → B is a principal H-bundle with fiber F , then one says that a connection θ is an H-connection for π if the H action preserves the horizontal spaces and preserves the connection 1-form. One says that a metric G is adapted to H if the fibers π −1 (b) are totally geodesic and isometric to F and such that the horizontal distribution of θ is the orthogonal complement to the vertical.
The following Lemma gives details on the equivalences of the various conditions and is implicitly contained in [3, Example 2.1] and is proved in [21,Theorem 3.5]. Hence we only sketch the proof. Lemma 6.4. -Suppose that S 1 acts freely on M and that G is an S 1 invariant metric for which all orbits are geodesics isometric to R/2πZ. Then G is a Kaluza-Klein metric and π : M → M/S 1 is a Riemannian submersion with totally geodesic fibers isometric to R/2πZ.
Proof. -Under the freeness assumption, we have an S 1 bundle π : M → X := M/S 1 . Let V = R ∂ ∂θ be the vertical space, i.e., the tangent space to the orbits. Let H x = V ⊥ x . The metric G determines a quotient Riemannian metric on X using the isomorphisms dπ x : H x → T π(x) X. By assumption, | ∂ ∂θ | G = 1 if we identify S 1 = R/(2πZ. The only non-trivial statement is that the orbits are geodesics. Let α(t) denote the orbit of a point x under the S 1 action. Let v(0) be a horizontal vector at α(0). Let σ(t, s) be the parallel translation of α(t) along a curve in X with initial tangent vector dπ(v(0)). Then ∂ ∂s σ(t, 0) is a horizontal vector field along α(t). The arclength of the curve t → σ(s, t) is constant in s. The first variation formula implies that α(t) is geodesic.
These adapted metrics are special cases of invariant metrics on an S 1manifold M . For instance, in the case of the non-free action of S 1 on the standard S 2 by rotations around the x 3 -axis, | ∂ ∂θ | g0 varies with the orbit, (2) A free action is one for which all isotropy groups are trivial.
ANNALES DE L'INSTITUT FOURIER
and almost no orbits are geodesics. It would be interesting to consider generalizations to all S 1 invariant metrics. Sections of K m , i.e., differentials of type (dz) m , lift to the dual line bundle K * = T (1,0) as equivariant scalar functions F : K * → C transforming by e imθ under the S 1 action of rotating a frame. Using the metric g we form the unit tangent bundle π : SX → X. In this section, we review the relevant formulae for lifted operators and review the fact that the Bochner Laplacians are Fourier components of the horizontal Laplacian on SX.
Lifts to P h
The natural inner product on L 2 (P h , dV G ) is given by Sections s of L m naturally lift to L * and F h bŷ It is straightforward to check that the lift of s ∈ C(X, L m ) satisfiesŝ(r θ x) = e imθŝ (x) and that Indeed, if x = r θ e L * (z) e L * (z) thenŝ(x) = e imθ e L (z) m h m . In the case of L = K X , the lift has the form, We define a orthonormal frame of T * X by ω 1 = e −φ dz := dz |dz| h as above, and let ∂ ∂z −1 ∂ ∂z = e φ ∂ ∂z be the dual frame. In local coordinates z,z on X and in this local frame we define local coordinates (z,z, θ) on SX corresponding to the point e iθ e φ ∂ ∂z . Then (dz) m lifts to the function, Consequently, the eigendifferential f m,j (dz) m lifts to φ m,j (z,z, θ) = f m,j (z)e imθ e mφ(z) .
In (1.2) we decomposed the lift into real and imaginary parts. We now relate them to the real and imaginary parts of f m,j .
If we take the inner product of u m,j and v m,j just along the fiber and use orthogonality of cos mθ, sin mθ and that 2π 0 (cos 2 mθ − sin 2 mθ)dθ = 0, and then integrate in dA(z) we get Lemma 6.5. -u m,j , v m,j = 0.
Eigenspace decompositions
The Kaluza-Klein Laplacian has the form is the horizontal Laplacian. The fact that the fiber Laplacian is ∂ 2 ∂θ 2 reflects the fact that S 1 orbits are geodesics isometric to R/2πZ.
The weight spaces are ∆ H -invariant, i.e., as an unbounded self-adjoint operator, Lemma 6.6. -TheBochnerLaplacian agrees with the horizontal Laplacian ∆ H . In the above local coordinates and frame,
Under the canonical identification
Note that except for the last identity, these statements are true for any isometric S 1 action, not just for adapted Kaluza-Klein metrics.
Equivariant decomposition
Since S 1 acts isometrically on (M, G) we may decompose into its weight spaces, where
ANNALES DE L'INSTITUT FOURIER
The weight spaces are ∆ H -invariant, i.e., as an unbounded self-adjoint operator, The lifting map gives a canonical identification
Connectivity of nodal sets of Kaluza-Klein eigenfunctions
Given the preparations in Section 5, it is now a simple matter to prove Theorem 1.9. The following is an immediate consequence of Lemma 5.1: -If 0 is a regular value, then N um,j ⊂ SX is a singular 2m-fold cover of X with blow-down singularities over points where f m,j (z)(dz) m = 0.
Indeed, the 2m zeros of ω m,j (v) = 0 in S z X give 2m points on the fiber π −1 (z) in P h . Since locally there exist 2m smooth determinations of the zeros, the nodal set is a covering map away from the singular points.
We have separated Lemma 7.1 from further geometric results on the map π : N um,j → X in the next section since it was stated separately from those results in Theorem 1.9.
Remark 7.2. -In the literature, π : N um,j → X is sometimes called a branched cover [1, Section 4], but as J. Y. Welschinger explained to us, the terminology is misleading since smooth branched covers are supposed to have z 1/m singularities over the branch points, just as for holomorphic branched covers, while the inverse image of a zero of f m,j is an S 1 orbit and the singularity is blown up. In some sense, π is locally like the projection of a vertical helicoid onto the horizontal plane.
Nodal domains of real and imaginary parts
We now give a sketch of the proof of Theorem 1.5. By Proposition 1.8 (and (1.5)), N φm,j \(N φm,j ∩ Σ) → X\Z fm,j is a 2m-sheeted cover. Moreover, P h \Σ → X is an S 1 bundle and i(P h \Σ)\N φm,j → X is a fiber bundle whose fibers consist of the punctured fibers π −1 (z)\N φm,j . The connected components of each punctured fiber consist of "arcs" along which φ m,j has a constant sign. We therefore express it as (P h \Σ)\N φm,j = P + P − where sign φ m,j = ± in P ± . Each π : P ± → X is a fiber bundle whose fiber consists of m arcs of the fibers of π : P h → X. Since the number of zeros in each regular fiber is 2m, the number of connected components of P ± is m. When we take the closure of these sets (i.e., add in the singular fibers, on which φ m.j = 0, the connected components of the closure are the nodal domains. It follows that there are 2m nodal domains. We now argue that the closure of P ± is connected, so that there exist exactly 2 nodal domains. We now use the local analysis in Section 5 of eigendifferentials of generic Bochner Laplacians around their zeros to determine how the sheets are connected at the singular fibers C j = π −1 (z j ), corresponding to singular points (i.e., zeros) of f m,j (dz) m i.e., we consider the maximal components P ±,j of in which φ m,j has a single sign. When we union the left side with m j=1 C j we glue together some of these domains along intervals of the singular fibers.
The gluing rule for the nodal domains is determined by the gluing rule for the nodal set, since the boundary of the each nodal domain is the nodal set. From the downstairs point of view, the gluing rule is the monodromy of the cover N ujm,j → X\Z(ω m,j ) If we fix a singular point z 0 , then we get a monodromy representation ρ : π 1 (X\Z(ω m,j )) → Aut(π −1 (z 0 )), determining how the sheets of the nodal set are changed as the point circles around z 0 .
By Proposition 5.7, the index of the singular points z 0 is ±1 m . In terms of the monodromy, this means precisely that each turn around a circle C enclosing z 0 lifts to an arc from one vector in the fiber to its nearest neighbor with the same sign of φ m,j (i.e., skipping the neighboring vector of the opposite sign).
It follows that both the + region and − region is connected in P h . Hence there are just two nodal domains.
Counting the number of nodal domains
We now give a more detailed presentation. Let D be an open disc. We first study connectivity of a certain graph that arise from a pair of partitions of D.
Let P and Q be partitions of D, i.e., P (resp. Q) is a collection of disjoint open-sets Ω P (1), . . . , Ω P (n P ) ⊆ D (resp. Ω Q (1), . . . , . Let c P : P → {0, 1} and c Q : P → {0, 1} be colorings of P and Q, and define the inversions of c P and c Q by c P = 1 − c P and c Q = 1 − c Q .
We now define a graph G m (P, Q, c P , c Q ) as follows: The vertex set is k=1 Ω Q (k) does not contain a closed curve. In particular, G m (P, Q, c P , c Q ) has at most 2m connected components. TOME 70 (2020), FASCICULE 3 Proof. -We first consider the case m = 1. To claim G 1 (P, Q, c P , c Q ) has only 2 connected components, it is sufficient to prove that if c P (Ω P (a 1 )) = c P (Ω P (a 2 )), then v 1,a1 and v 1,a2 are path-connected.
Because (P, Q) is a generic pair, one can find a chain of open-sets such that two adjacent open-sets have non-trivial intersection.
Observe that if Ω P (c) ∩ Ω Q (b) = ∅, then either is an edge, and likewise either is an edge. Therefore the above chain of open-sets corresponds to a path connecting v 1,a1 with either v 1,a2 or v 3,a2 . However, from the assumption c P (Ω P (a 1 )) = c P (Ω P (a 2 )), and from the construction of G 1 (P, Q, c P , c Q ), v 1,a1 cannot be connected to v 3,a2 , hence is connected to v 1,a2 . Now for the rest, note that G m is an m-covering of G 1 , and because v 1,1 and v 1,3 belongs to the different connected components of G 1 , any connected components of G m must contain at least one vertex of the fiber of v 1,1 or v 1,3 .
For a large class of colorings, we can deduce a much stronger result. There exist four open sets Ω P (a 1 ), Ω P (a 2 ), Ω Q (b 1 ), Ω Q (b 2 ) such that for i = 1, 2 and j = 1, 2, and that Then the graph G m (P, Q, c P , c Q ) has 2 connected components.
Proof. -Note that any connected component of G m must contain either one of v 4j,a1 or one of v 4j,a2 with j = 1, . . . , m, because G 1 has only two connected components.
The number of nodal domains of generic eigenfunctions
Let P be a principal S 1 bundle over a connected smooth compact Riemannian surface X with the covering map π : P → X. Let m be a fixed integer, and assume that φ ∈ C 1 (M ) satisfies the following conditions: the zero set of f (resp. f ) gives rise to a partition P = P U (resp. Q = Q U ) of U , and (3) (P U , Q U ) is a generic pair of partitions of U .
In this section, we prove the following theorem.
Theorem 8.5. -Fix any point x ∈ X such that φ(x, θ) = 0. Then any nodal domain of φ has a nonempty intersection with π −1 x. In particular, the number of nodal domains of φ is 2m. Assume further that φ has a regular zero. Then the number of nodal domains of φ is 2.
We begin with few observations in terms of fixed U and a local coordinate (x, θ) of π −1 U .
} for some integer k, and if πU 1 ∩ πU 2 = ∅, then U 1 and U 2 are contained in the same nodal domain of φ.
Proposition 8.7. -Any nodal domain of φ| π −1 U must intersect π −1 U ∩ {θ = kπ 2m } nontrivially for some integer k ∈ Z. Proof. -Assume for contradiction that Ω is a nodal domain of φ| π −1 U that is contained in From the equation we see that for each fixed x, φ(x, θ) either vanishes identically or has at most one sign change along the curve This implies that if x ∈ πΩ, then which contradicts the assumption that the zero set of f gives rise to a partition of U .
From these two propositions, we see that the nodal domains of φ| π −1 U can be understood from the nodal domains of the restrictions of φ| π −1 U to the 4m-hypersurfaces In particular, if we define c P U and c Q U in terms of the sign of f and f , then the number of connected components of G m (P U , Q U , c P U , c Q U ) is equal to the number of nodal domains of φ| π −1 U .
Proof of Theorem 8.5. -Let x ∈ X be a point where φ(x, θ) = 0, and let U be a sufficiently small neighborhood of x. We may assume without loss of generality that the vertices correspond to the nodal domains of the restrictions of φ| π −1 U to the hypersurfaces
ANNALES DE L'INSTITUT FOURIER
that intersect the fiber π −1 x. Then Lemma 8.2 implies that any nodal domain of φ| π −1 U must intersect π −1 x. Now assume that x is another point in U . Then we may restate this as "any nodal domain of φ| π −1 U that intersect π −1 x must intersect π −1 x", and equivalently, "any nodal domain of φ that intersect π −1 x must intersect π −1 x". Because we assumed that X is connected, by the freedom of choice of the pair of points x and x , any nodal domain of φ must intersect π −1 x. This proves the first part of the theorem.
For the latter part of the theorem, let p be a regular zero of φ, i.e., dφ : T p P → C is a surjection. Choose a sufficiently small neighborhood U ⊂ X of πp, and let f be the function that satisfies If d f and d f are linearly dependent, then a straightforward computation implies that dφ has rank 1, so d f and d f are linearly independent.
This implies that πp is a regular zero of both f and f . Also, linear independency implies that locally around πp, f = 0 and f = 0 define two curves intersecting transversally at πp. From this, we may find four open sets near p that are required for Lemma 8.3, and we infer that the number of nodal domains of φ| π −1 U is two. Now because any nodal domain of φ must intersect with π −1 x for some x ∈ U , any nodal domain of φ must contain one of the nodal domains of φ| π −1 U , from which we conclude that φ has only two nodal domains.
We are ready to prove our main theorem, Theorem 1.5.
Proof. -It is sufficient to verify the assumptions in Theorem 8.5 is satisfied. The first condition is trivial to verify. For the other conditions, note from the assumption that P → X is non-trivial, Z fm,j is non-empty, and Theorem 4.1 implies that it is discrete and consists only of regular zeros.
Remark 8.8. -If Z fm,j contains a closed curve that divides X into two connected components, then the number of nodal domain can be large. For instance, if f m,j vanishes on the boundary of small open disc U ⊂ X, and if it does not vanish on U , then φ m,j vanish identically on ∂ π −1 U , and therein, φ m,j has 2m-distinct nodal domains. In particular, Theorem 8.5 fails even if f m,j has a regular zero elsewhere.
Surfaces of constant curvature
In this section, we illustrate the geometry of Kaluza-Klein metrics and the Kaluza-Klein eigenvalue problem on unit tangent bundles of surfaces of constant curvature.
Flat tori
Let T 2 = R 2 /Z 2 . We use coordinates z = x 1 + ix 2 . Its unit tangent bundle is ST 2 = T 2 × S 1 . The connection is flat and ∆ H = ∆ is simply the Laplacian of T 2 . The Kaluza-Klein Laplacian is that ∆ G = ∆ + ∂ 2 ∂θ 2 on T 2 × S 1 . The Kaluza-Klein eigenfunctions are linear combinations of the product eigenfunctions, The multiplicity of the eigenvalue with fixed m is the number of ways of representing | k| 2 as a sum of two squares. They correspond to eigendifferentials f m, k (z)(dz) m = e i k, x (dz) m .
In the notation (1.2), The nodal sets of the imaginary part are given by, Z v m, k contains the set Note that φ m, k (x 1 , x 2 , θ) has no zeros on T 2 × S 1 and f m, k (z)(dz) m has no zeros as an m-differential on T 2 .
If we change the lattice to a general lattice L ⊂ R 2 , the eigenfunctions of T 2 change to e λ ( x) = e 2πi λ, x where λ ∈ Λ = L * , the dual lattice. For generic L, the eigenvalues have multiplicity 2 and the eigenspaces are spanned by the real and imaginary parts of e λ or equivalently by e λ and its complex conjugate e − λ . The same is true of the Kaluza-Klein eigenfunctions φ m, λ = e 2πi λ, x e imθ . Again, φ m, λ has no zeros. Using the bifurcation of nodal sets of eigenfunctions under generic paths of metrics of [20] We now give an explicit orthonormal eigenbasis of T 3 such that all of them have exactly two nodal domains, hence proving Theorem 1.11.
Assume without loss of generality that j 1 = 0. Then has two nodal domains by Theorem 8.5, because has a regular zero.
Case 2: exactly one m k is zero, and the other two are different. -From the same reasoning, each eigenfunction in the new basis in the following has two nodal domains: Case 3: exactly one m k is zero, and the other two are equal. -Again by the same reasoning, each of the following TOME 70 (2020), FASCICULE 3 has two nodal domains, and these are the basis of Case 4: exactly one m k is nonzero. -In this case, we consider orthogonal eigenfunctions which span Each of these has only two nodal domains from the following lemma. has only two nodal domains. for y such that 1 4 +2 sin 2 y = 1, there is x satisfying (9.1), and this completes the proof.
Kaluza-Klein metrics on S 3
Let (S 2 , g 0 ) be the 2-sphere with its standard metric of curvature 1. Then its unit tangent SS 2 = SO(3) = RP 3 = S 3 / ± 1 and the Kaluza-Klein metric is the standard metric of constant sectional curvature 1 on S 3 (divided by the antipodal group Z 2 ). The Kaluza-Klein Laplacian is therefore the standard Laplacian ∆ S 3 on Z 2 -invariant functions.
Since S 3 is a group, Alternatively, the eigenfunctions of S 3 are harmonic homogeneous polynomials on R 3 . Moreover, ∆| V N ⊗V N = N (N + 2) = (N + 1) 2 − 1. The eigenfunctions of RP 3 are those where N is even.
We need explicit separation of variables expressions for equivariant spherical harmonics, and therefore need to introduce coordinate systems. We use "axis -angle" Hopf coordinates (α, θ, φ) on S 3 ⊂ R 4 defined by sin α cos φ sin α sin φ cos α cos θ cos α sin θ .
As a result, these eigenfunctions do not satisfy the conditions of the generic Kaluza-Klein metrics to which our results apply, and their nodal sets are quite different.
As mentioned in the introduction, the numerical experiments of A. Barnett et al. [2] show that random spherical harmonics of degree N on S 3 also have different types of nodal sets than our generic eigenfunctions. Namely, the expected number of nodal domains has the asymptotics cN 3 for a certain c > 0. As proved in [15], the nodal sets of real/imaginary parts of random equivariant eigenfunctions with fixed m (a subspace isomorphic to V N ) have connected nodal sets. The difference is due to the fact that our random equivariant spherical harmonics are a thin subset of the random spherical harmonics of degree N on S 3 .
Hyperbolic surfaces H 2
Although it differs from our prior discussion in the compact case, let us consider a finite area hyperbolic real Riemann surface of constant negative curvature −1. Then X = S * g M = Γ\G where G = PSL(2, R). The total space X carries a Lorentz Cartan-Killing metric with indefinite Laplacian the Casimir operator Ω. It is well known that Ω = H 2 + V 2 − W 2 . We now change the sign of the third term to get the Kaluza-Klein Laplacian ∆ X = H 2 + V 2 + W 2 . The associated metric defines a Riemannian submersion π : X → M with fibers given by K-orbits. They are necessarily totally geodesic. It follows that the horizontal Laplacian H 2 + V 2 commutes with the vertical Laplacian W 2 . This is obvious because 0 = [Ω, The joint eigenfunctions of Ω, W are denoted by φ m,j . When m = 0 they are pullbacks of eigenfunctions of M = Γ\G/K.
In particular the number of nodal domains of φ j,0 on X is the same as the number of nodal domains of φ j on M . The former nodal sets are K-invariant and in the case of regular nodal components are 2-tori over circles.
The lift of weight m of an m-differential f (dz) m is given by Φ(x, y, θ) = y m/2 f (x + iy)e −imθ .
Automorphic forms on the full modular group
Now we consider the case Γ = PSL 2 (Z). Note that the quotient Γ\G is non-compact in this case. Nevertheless, it is known that −∆ G has infinitely many discrete spectrum, where corresponding L 2 integrable eigenfunctions can be chosen so that they are in one-to-one correspondence with Maass-Hecke cusp forms or holomorphic Hecke cusp forms. We refer the readers to [10] for detailed background. Assume that the zeros of φ m,ir are isolated. Then φ m,ir has only two nodal domains.
Proof. -The first statement of Condition 8.4 follows from the definition of Maass-Hecke cusp form, and the second statement follow from the fact that φ m,ir : X → C can not be scaled to a real-valued function, and that φ m,ir is analytic. The third statement follows from the assumption. Now, because the first Hecke eigenvalue is 1, the first Fourier coefficient of φ m,ir at the cusp does not vanish, meaning that i∞ is a regular zero of φ m,ir . We conclude the proof by applying Theorem 8.5.
Remark 9.6. -It is not hard to see that in the constant curvature case, the nodal set of φ 2,ir consists of the fibers over the critical point set C φir of φ ir . At this time, it does not seem to be known whether C φir is necessarily a discrete set of points in the case of hyperbolic surfaces. This cannot be proved by a purely local calculation, since the critical point set of rotationally invariant Dirichlet/Neumann eigenfunctions on a compact rotationally invariant submanifold C R of a hyperbolic cylinder H 2 / γ 0 consists of a union of S 1 orbits. Here, γ 0 is a hyperbolic element and γ 0 is the cyclic group it generates. Thus, negative curvature does not rule out codimension 1 critical point sets. One can put any negatively curved S 1 invariant metric on C R and obtain the same result, so it is not an effect of constant curvature. We conjecture that for compact hyperbolic surfaces without boundary, C φir is a finite set for every eigenfunction.
ANNALES DE L'INSTITUT FOURIER
When we have holomorphicity of φ, we may remove the assumption that the zeros of φ being isolated. For instance, we have: Theorem 9.7. -Let X = PSL 2 (Z)\H, and let φ m,0 be a Laplacian eigenfunction on PSL 2 (Z)\ PSL 2 (R) corresponding to a holomorphic Hecke cusp form F of weight m. Then φ m,0 has only two nodal domains.
Because we assumed that F is a Hecke cusp form, the first Hecke eigenvalue is 1. Therefore i∞ is a regular zero of φ m,0 , and now the theorem follows from Theorem 8.5.
Corollary 9.8. -There exist eigenfunctions on PSL 2 (Z)\ PSL 2 (R) that have only two nodal domains but with arbitrarily large eigenvalues.
It can be shown that the nodal set of Φ on the side is that of cos(24θ) = 0, and on the front is that of cos(12(ϕ + 2θ)) = 0, where we define ϕ = arccos(x). We compute the nodal set of the restriction to the top numerically using Mathematica. | 2018-06-12T18:52:33.000Z | 2018-06-12T00:00:00.000 | {
"year": 2018,
"sha1": "76d3ad862fd801419f2ee8154bd281acd00e848a",
"oa_license": null,
"oa_url": "https://aif.centre-mersenne.org/item/10.5802/aif.3329.pdf",
"oa_status": "GOLD",
"pdf_src": "Anansi",
"pdf_hash": "15307bf5288d28d5bf3a9dd5d57cdd25657c0c15",
"s2fieldsofstudy": [
"Mathematics"
],
"extfieldsofstudy": [
"Mathematics"
]
} |
119060072 | pes2o/s2orc | v3-fos-license | LensExtractor: A Convolutional Neural Network in Search of Strong Gravitational Lenses
In this work, we present our classification algorithm to identify strong gravitational lenses from wide-area surveys using machine learning convolutional neural network; LensExtractor. We train and test the algorithm using a wide variety of strong gravitational lens configurations from simulations of lensing events. Images are processed through multiple convolutional layers which extract feature maps necessary to assign a lens probability to each image. LensExtractor provides a ranking scheme for all sources which could be used to identify potential gravitational lens candidates significantly reducing the number of images that have to be visually inspected. We further apply our algorithm to the \textit{HST}/ACS i-band observations of the COSMOS field and present our sample of identified lensing candidates. The developed machine learning algorithm is much more computationally efficient than classical lens identification algorithms and is ideal for discovering such events across wide areas from current and future surveys such as LSST and WFIRST.
One of the main goals of observational cosmology is to constrain the main cosmological parameters that dictate the evolution of the Universe (Tegmark et al. 2004;Komatsu et al. 2009;Weinberg et al. 2013). Strong gravitational lensing has been utilized over the past few years to estimate and contain these cosmological parameters (Broadhurst et al. 2005;Suyu et al. 2013Suyu et al. , 2014Goobar 2016;Agnello et al. 2017;More et al. 2017). This is achieved through accurate lens modeling of such events and comparing the model predictions with observations (such as with observations of lensing induced time delays (Eigenbrod et al. 2005;Treu 2010;Suyu et al. 2014;Rodney et al. 2016;Treu & Marshall 2016). In a recent study, for example, Suyu et al. (2013) used combined WMAP, Keck and HST data on gravitational time delays in two lensed sources to constrain the Hubble constant within 4% in a ΛCDM cosmological framework.
One of the key aspects of gravitational lensing is its use as natural telescopes through boosting the observed signal and increasing the spatial resolution (Treu 2010). This is quite advantageous in searches for distant and/or faint objects at moderate observing costs and has been utilized extensively in various surveys in searches for such objects, the identification of which would not have been possible without it (Bolton et al. 2006;Heymans et al. 2012). Given that the number of identified lenses for different classes of galaxies rises sufficiently due to better lens finding algorithms, by modeling the lenses and using spectra stacking techniques, we can better understand the physical and chemical composition of farther and fainter galaxies which in turn would excel our understanding of galaxy evolution (Wilson et al. 2017;Timmons et al. 2016). In the past few years, deep diffraction limited observations have also taken advantage of gravitational lensing to extend the faint end of the luminosity function of galaxies by a few orders of magnitude (Atek et al. 2015) to produce the deepest images of the sky ever taken across multiple bands. Strong gravitational lensing events have been observed extensively in such surveys as galaxy-galaxy lensing in field surveys such as the Cosmic Assembly Near-infrared Deep Extragalactic Legacy Survey (CANDELS) (Grogin et al. 2011;Koekemoer et al. 2011) and the Cosmological Evolution Survey (COSMOS) (Scoville et al. 2007;Capak et al. 2007) or as cluster lensing from observations of nearby massive clusters (Postman et al. 2012;Treu et al. 2015;Lotz et al. 2017) with Hubble Space Telescope. These yield identification of the first generations of galaxies at z > 3 (and out to z ∼ 11; Oesch et al. 2015) to study galaxy formation and evolution at the epoch of re-ionization. This was, in fact, one of the main motivations behind Hubble cluster lensing studies such as CLASH and Frontier Fields (Postman et al. 2012;Lotz et al. 2017). These magnification provided by the strong lensing could potentially raise the observed flux by as much as a factor of hundred (depending on configuration). The power of gravitational lensing could also be used in the detection of low surface brightness emission from extended objects Figure 1. Schematic representation of an artificial neuron. The weighted sum of the neurons in the previous layer (green circles), plus the internal bias of the neuron, are mapped as the output of the neuron by an activation function. This model is captured by Equation 1. During the learning process, weights and biases of the neurons will be adjusted to achieve the desired network output.
such as far-infrared and radio emissions from dust and molecular gas at z ∼ 2 − 3. This is indeed one of the main techniques in observing these systems even in the era of powerful mm/radio telescopes such as ALMA.
Strongly lensed galaxies are normally targeted and identified from dedicated surveys (Bolton et al. 2006). Traditionally these lens identifications are either catalogbased, in which lensing events are identified by looking for objects in a lensing configuration, or pixel based, with the search starting from a set of pixels. These lensing searches are normally computationally challenging in that individual pixels are constantly compared with adjacent ones and they could be biased towards a given population and/or brighter objects. Recent far-infrared wide area observations (such as with Herschel ) significantly advanced these searches for lensed galaxies by adopting a simple efficient selection technique of lensed candidates through observations of excessive flux in the far infrared (as an indication of strong lensing events supported by number count distributions; Nayyeri et al. 2016;Wardlow et al. 2012). However such surveys are also biased towards populations of red dusty star-forming galaxies (missing any blue lenses) and are not always available across the full sky (the Herschel surveys that were targeted had ∼ 0.2 − 0.4 deg −2 lensing events, much lower than expected from optical surveys). Given that tests of cosmological models require simple unbiased selection function, it is important to have a complete unbiased catalog of lensing events.
We have entered the era of big data astronomy. Sky surveys such as the LSST, Euclid, and WFIRST will produce more imaging data than humans can ever analyze by eye. The challenges of designing such surveys are no longer merely instrumentational, but they also demand powerful data analysis and classification tools that can identify astronomical objects autonomously. Fortunately, computer vision has drastically improved in the last couple of years to make autonomous astronomy possible. The past couple of years has been the most exciting era in the field of machine learning (ML). Researchers from both the public and the private sectors have achieved landmarks in developing image recognition/classification techniques. One of the most exciting recent events in the ML community was the release of TensorFlow by Google, a parallel processing platform designed for development of fast deep learning algorithms (Abadi et al. 2016). Packages like Tensor-Flow, Caffe, and others have enabled researchers to develop very complex and fast classification algorithms. Among these deep learning programs, ConvNets have deservingly received a lot of attention in many fields of science and industry in the past few years (Krizhevsky et al. 2012). Complex ConvNets such as GoogleNet and AlexNet, which are publicly available, have achieved superhuman performance on the task of image classification. Google's TensorFlow has made it possible to easily develop parallelized deep learning algorithms which if integrated with Google's Tensor Processing Units (TPUs), could address the data mining challenges in the field of astronomy. The field of astronomy and observational astrophysics should take advantage of these new image classification algorithms. For instance, our team has been working on developing LensFlow, a ConvNet that can be used to search for strong gravitational lenses. Our work will be publicly available on Github after publication.
Non-machine learning computer algorithms have been previously used for finding gravitational lenses (Alard 2006;More et al. 2012). For instance, More et al. (2012) use two algorithms called RingFinder and ArcFinder. The former uses color information and the latter detects arc-like pixels. Initially, ArcFinder polishes the images by convolving a smoothing kernel. For each pixel, an estimator of elongation is calculated by taking the ratio between the sum of the flux of a few pixels along the horizontal line and the maximum value of a few nearby pixels along the vertical line which pass through the pixel in hand. This process is repeated for all pixels and those with smaller than an specified elongation threshold are set to zero to create a sharp arc map. An arc map that satisfies thresholds on the arc properties such as the size and surface brightness will be selected as an arc candidate for further visual inspection. Unlike deep learning, such classical algorithms are not easily parallelized and they can be computationally more expensive depending on the deepness of the ConvNets used. They also require threshold tuning which may cause insensitivity to smaller arcs. However, they do not require a massive training dataset. Even though more challenging, creating a large dataset could eliminate biases toward certain arclens morphologies. As we will discuss in Section 4, these algorithms do suffer from the same lens contaminants as ours. The hope is that machine has the capacity to improve its performance with better training datasets and improved architecture while remaining computationally efficient but such improvements are not trivial regarding classical algorithms.
Other researchers (Petrillo et al. 2017;Jacobs et al. 2017;Lanusse et al. 2017) also find deep learning a suitable solution for finding gravitational lenses. Lanusse et al. (2017) use residual ConvNets with 46 layers. Resid-ual ConvNets are modified ConvNet that do not suffer from layer saturation as ordinary ConvNets do. After adding more than 50 layers, the accuracy of ordinary ConvNets no longer improves and the training becomes more challenging. He et al. (2016) were able to overcome this issue by providing residual maps in between layers, which has been employed by Lanusse et al. (2017). They have simulated LSST mock observations in a single band and have trained and tested their network on these images. Jacobs et al. (2017) have trained their ConvNet using multiple color bands and have applied it to Canada-France-Hawaii Telescope Legacy Survey. Petrillo et al. (2017) have searched for lenses in Kilo Degree Survey by training their ConvNet on cataloged luminous red galaxies.
In our independently developed work, we focus on the morphology of the lenses and only rely on one color band, similar to Petrillo et al. (2017) and Lanusse et al. (2017). Our lens simulation method is very similar to Petrillo et al. (2017) where we both merge simulated arcs with real images of galaxies to preserve the complexity of the physical data. In contrast to others, we do not discriminate against different sources found in the COSMOS field. Artifacts, starts, and other sources have been included in our training dataset so LensFlow can be directly applied to fields without a need for a catalog with galaxy type information. The deepness of our ConvNet is comparable to Petrillo et al. (2017) and Jacobs et al. (2017) but it is shallower than Lanusse et al. (2017). As mentioned in Jacobs et al. (2017), the morphology of lenses are much simpler than the morphology of daily objects and human faces which extremely deep ConvNets are developed for. However, the cost to performance ratio of ConvNets with varying deepness has not been studied yet. The effectiveness of deeper ConvNets cannot be compared between ours (and Petrillo et al. 2017 ) and Lanusse et al. (2017) since they have not applied their algorithm to physical data. However, they have studied the change in the performance of their ConvNet by varying the Einstein radii and signal-to-noise ratio of their lenses.
This paper is organized as follow. In Section 2, we will explain the principal concepts underlying neural networks, supervised learning, and ConvNets. A supervised learning algorithm requires a large sample of labeled images, known as the training dataset. In Section 3.1, we will discuss the procedure we have taken to create our training and testing datasets. Before feeding the images to a ConvNet, they must be normalized and should be enhanced. The details of these methods are discussed in Section 3.2. Section 3.3 will lay out the architecture of LensFlow and Section 3.4 will illustrate LensFlow's performance on the testing dataset. We will conclude by sharing and analyzing the scan results of the COS-MOS field in Section 4. Throughout this paper, we assume a standard cosmology with H 0 = 70 kms −1 Mpc −1 , Ω m = 0.3 and Ω Λ = 0.7. Magnitudes are in the AB system where m AB = 23.9 − 2.5 × log(f ν /1µJy) (Oke & Gunn 1983).
2. DEEP LEARNING ALGORITHMS Artificial neural networks are inspired by biological neurons. Just like biological neurons, artificial neurons receive input signals and send out an output signal to The pixels in each box represent the weights of a convolving neuron which are connected to a 5 × 5 region input image. As these filters convolve over the entire input image, they generate 16 feature maps. Red pixels have a positive contribution and blue pixels have a negative contribution toward the activation of the convolving neuron. These filters are helpful for edge and texture recognition. . Two examples of convolutional layer feature maps. Image of a normalized physical lens has been shown in (1). After the normalizing and enhancing it with an anti-Gaussian function, the arc stands out (2). Feature map (3) shows an enhancement of the arc while feature map (4) highlights the edges along the antidiagonal orientation. Such feature maps are to be further compared and analyzed in the fully-connected layers to determine if the image is lensed or not.
other neurons (see Figure 1). The synaptic connections between neurons are known as weights and the output of a neuron is know as its activation. To reduce the computational time and simplify neural network models, neurons are placed in consecutive layers rather than having a connection with every other neuron. Neurons from one layer cannot talk to each other or to the neurons in arbitrary layers; they may only send their signal to the neurons in the succeeding layer. A neuron receives the weighted sum of the activation of all the neurons in the previous layer, adds an internal parameter known as the bias and maps this sum to a value computed by an activation function (e.g. sigmoid, hyperbolic tangent, rectilinear, softmax). This model can be stated mathematically by the following equation: Here, a l i is the activation of the neuron in hand (i.e. the i'th neuron in the l'th layer), f is the activation function of this neuron, a l−1 j is the activation of the neuron j in layer l − 1 (the previous layer), w l j→i is the synaptic weight connecting i'th neuron in layer l to the j'th neuron in layer l − 1, and b l i is the bias of the neuron to adjust its activation sensitivity. The first layer, i.e. the input layer, in a deep learning neural net acts as a sensory layer, analogous to the retina. As it gets analyzed, the information from the input layer travels through multiple layers until it reaches the final layer called the classification layer. Each class of images corresponds to a classifying neuron. In our case, we have a neuron corresponding to unlensed and another to lensed images. The neuron with the highest output determines which class an input image is placed in.
A neural net learns how to classify images by adjusting the weights between its neurons and the biases within them, having one goal in mind: minimizing the loss function C(x, y). The loss function, sometimes called the cost function, can take many forms but it has to captures the misfiring of the classification neurons, i.e. the deviation between the target class versus the predicted class. This is why such algorithms are known as supervised learning algorithms, in contrast to unsupervised techniques. A common choice for the loss function is the cross-entropy loss function with the following form (Nielsen 2016): a L j is the activation of neurons in the final (classifying) layer. x is the input data in the vector form and y represents the desired activations of the two classifying neurons. Of course, this function depends on the architecture of the neural net, weights, and biases, but they have not been expressed explicitly. As an example, if an image is a lens, its target output has to be (0.0, 1.0), meaning the activation of the unlensed neuron should be zero and the activation of the lensed neuron should be unity. During the training process of a neural net, images from a training dataset are presented to the network and the weights and the biases are adjusted to minimize the loss function for those images. The parameter space is massive and a change in one of the parameters of a neuron will affect the activation of a series of neurons in other layers. The first challenge is solved by minimizing algorithms such as the stochastic gradient descent (SGD) and the second one is solved via back-propagation. We won't go in the details of these two techniques, but it worths emphasizing on the stochasticity of SGD. Stochasticity refers to randomly selecting images from the training set and bundling them in one batch. The loss function for the batch is the average of the loss function for individual images. It is very important to use a batch rather than training the neural net with one image at a time. Loosely speaking, using a batch would drastically improve classification accuracy of the network since neurons learn the features in images rather than memorizing examples.
ConvNets are a class of neural networks with multiple convolutional layers. A convolutional layer consists of a set of convolving neurons (on the order of 10 neurons) which can be connected to a small rectangular region of an image. The set of weights of a convolving neuron is known as a filter and are subject to change as the Con-vNet learns. A filter scans an entire image by striding (convolving with specified steps) over the image and assembling its output into an image knows as a feature map. Feature maps contain information such as texture 5). In addition, gamma correction has been employed to adjust the contrast, resulting in an enhanced (top) image where the arcs stand out. Such enhanced and normalized images are the inputs of the ConvNet. and edges. See Figure 2 as an example of a set of filters in a LensFlow convolutional layer. Two extracted feature maps of a physical lens (not simulated) have been shown in Figure 3. 3. METHODOLOGY 3.1. Training and Testing Datasets Neural networks learn to classify data by learning from examples, referred to as the training dataset. A training dataset usually contains a few thousands of pre-classified images. Our training dataset contains 15, 000 unlensed sources and 15, 000 lensed sources. To make this dataset, we used SExtractor to obtain a catalog of sources from a few high-quality central tiles as well as a few low-quality edge tiles to include noisy images and artifacts. A cutout of 200×200 pixels was made around the identified sources and stored individually, labeled as unlensed. After training and finalizing the architecture of the ConvNet, using a source extraction software is not necessary since an entire tile can be divided into smaller tiles and scanned to identify locations with high lensing probability.
Creating lenses are more challenging and more involved. To create these lenses, we had two options; we could either artificially boost up the number of know lenses or simulate them. We tried both methods. For the first method, we used 17 out of 18 lenses discussed in However, the training dataset using the transformation method does not contain a wide range of lenses and lenses with different structures might be missed. For this reason, we have also created a simulated training dataset using LensTool (Kneib et al. 1996;Jullo et al. 2007;Jullo & Kneib 2009). LensTool receives input parameters of the lensing model and generates a lensed image of the background galaxy without a foreground lensing source. We will refer to these as arcs. Lens model parameters such as redshift of the background and foreground galaxies, their ellipticity, orientation, and their relative positions were randomly changed to create a wider range of arcs, from complete Einstein rings with different radii to arcs with different orientations and sizes. The code for generating these arcs which includes the range of model parameters will be included in our online distribution. After generating a wide range of arcs, they were merged with raw images. Both images, the raw sources, and the simulated arcs, were normalized and the pixels of the arc images were multiplied by a random number between 0.3 and 0.8 so a range of relative arc to foreground luminosity would be captured in the training set. For a limited number of generated lenses, other ranges were used to create extremely faint and extremely bright lenses. 85% of the lenses were generated by selecting elliptical sources while the remaining were sources randomly selected from tile 55. A bias like this is necessary since the foreground of the know COSMOS lenses are all elliptical. After examining the simulated lenses by eye and eliminating lenses with unnatural geometry, an eightfold transformation was performed to increase the number of lenses up to fifteen thousand. These transformations come from 8 elements of the symmetry group of the square, namely: 0 • , 90 • , 180 • and 270 • rotations, and horizontal, vertical, diagonal and anti-diagonal mirroring.
At the end, 500 lenses and 500 raw images were removed from the training set and placed in a separate set called the testing dataset to measure the performance of the ConvNet as the network gets trained on the training dataset.
Image Normalization and Enhancement
Before inputting the images to our neural net, we normalize and enhance them. For deep learning purposes, there are different methods of image normalization to choose from, but not normalizing is not a choice. This is due to the fact that raw images come in a wide range of values. However, a limited number of neurons are not capable of handling such variations. On top of that, the goal of a ConvNet is to learn geometric features rather than the statistical properties of the pixels, which could be unwillingly introduced while constructing the training dataset. To address these two issues, all input images must be normalized. Our method of normalization consists of two steps. In the first step, pixel values are shifted by their average so their new mean would be zero (Equation 3). In the second step, pixel values in an image are divided by their standard deviation (Equation 4).
p ij stands for the value of the pixel at row i and column j. i max and j max indicate the total number of rows and columns, respectively. In order to improve our classification accuracy, we made the arcs stand out by apply an anti-Gaussian function (Equation 5) on each image. This function will attenuate the central pixels which belong to the foreground galaxy (see Figure 4). The width of this function is about a quarter of the cutout width.
At this step, the images are ready for contrast adjustments which is done using the gamma correction. The pixels must be shifted up in order to eliminate zero or negative pixel values.
The normalization steps are applied again and at the end, the negative pixels are dropped.
This improves the image quality by removing lowintensity pixels which mostly include noise. Now images may be inputted to the neural net for training or for classification of new data. See Figure 4 for an example of image normalization and enhancement.
Architecture of LensFlow ConvNet
The architecture of the data determines the dimensionality of the ConvNet layers. We use 200 × 200 × 1 images where 1 indicate the number of color channels 2 . In our code, we have localized these parameters as input variables to ease the transition from one to multiple color bands. Classifying lenses with multiple color bands will be easier and more accurate since foreground and background sources have a color contrast. However, we have chosen to use one color channel so our algorithm can be sensitive to geometry rather than color contrast in order to expand its applicability to a wider range of bands as well as eliminating its need for multi-band images when unavailable. As it can be seen in Figure 5, after normalizing and enhancing these single-channeled images, LensFlow applies an average-pooling of kernel 4 × 4 and stride of 4. This means that the image is divided into 4×4 segments and the average of each segment will become a pixel of the down-sampled output image. These downsampled images are then fed to the first convolutional layer which consists of sixteen 5 × 5 filters. The hyperbolic tangent function is chosen as the activation function for the neurons in this layer as well as the upcoming convolutional layers. This layer outputs sixteen feature maps which should be interpreted as one image with 16 feature channels, i.e. a 50×50×16 image using our notation above. These feature maps are down-sampled using a max-pooling layer of size 2 × 2 and stride of 2 resulting in a 25 × 25 × 16 reduced feature map. This means Figure 5. Representation of data flow through the ConvNet layers. The data is down-sampled and fed to three convolutional-max-pooling layers. The data is then flattened into an array and is fed to 128 sigmoid neurons. These neurons are then connected to 2 classifying softmax neurons. These last two layers are responsible for learning the difference between unlensed and lensed images buy comparing and analyzing the feature maps.
the input of the max-pooling layer is divided into 2 × 2 tiles whose maximum values are selected to form downsampled feature maps. This 25 × 25 × 16 feature map is inputted to the second convolutional layer with twentyfive 5 × 5 filters where each filter will scan all channels by combining them linearly, hence outputting a 25 × 25 × 25 feature map to be down-sampled to 13×13×25. Another convolutional layer with thirty-six 4×4 filters paired with a max-pooling layer is followed, reducing the data size to a 7 × 7 × 36 feature map. This information is flattened into a 1-D array where it is fully connected to 128 sig-moid neurons. The output of these neurons is inputted to the classifying layer with 2 neurons with the softmax activation function: Here, each component of Z is the total weighted input plus the bias of each of the classifying neurons, z c = j a L−1 j w L j→c + b L c . c specifies whether we are talking about the unlensed neuron or the lensed neuron. This function enhances the training process and it will assign Notes. * : also identified in Faure et al. (2008). † : from ACS i-band.
a pseudo-probabilistic number to each class (since the probabilities across all classes add up to 1). The higher σ c=lensed for an image, the higher its probability to be a lens which can be used to rank the images based on their lensing probability.
Measuring Performance
To optimize our ConvNet, we have chosen a crossentropy function as our loss function which we minimize using Adam's Optimizer. During the training phase, 64 unlensed and 64 lensed images were placed in a batch of 128 images. This combination technique will prevent under-or over-representation of classes even if the train-ing size for different classes contain a different number of examples. This process was iterated one thousand times. An example of accuracy as a function of the number of iterations has been graphed in Figure 6 where the accuracy plateaus at 91.5%. To obtain this accuracy, any image with a lensing probability lower/higher than 0.5 was labeled lensed/unlensed. These predicted labels were compared against the true label of each image. Thus, accuracy is defined as the number of correct classifications over the number of images in the testing set. accuracy ≡ N (lensed|p lens > 0.5) N (lensed) + N (unlensed) (9) Figure 7. Normalized ranking of tracer lenses by LensFlow. Lens-Flow ranks images based on their lensing probabilities, the highest probability has a rank of 1, the second highest probability has a rank of two, etc. The normalized rank (i.e. absolute rank divided by the total number of unlensed and lensed samples) of top 10 (out of 15) identified COSMOS lenses that were added to 3488 COS-MOS sources has been plotted. The blue lines show the result of scanning images once and the orange lines represent scanning the 8 symmetry group transformations of the square applied to each image to artificially increase the number of data points. No significant improvement is detected. This plot indicate LensFlow can place the majority of the physical lenses in the top 8% minority. Figure 8. Examples of common lens contaminants. Due to their similar visual geometries with lenses, ring galaxies (1-2) and spiral galaxies (3) are often false positive classifications of LensFlow. Galaxies with satellite or tidal interactions (4 and 6) and artifacts from bright sources (5) also contaminate the lens candidates.
This result was achieved with a "ReLu" 3 neurons in the convolutional layers. After switching to hyperbolic tangent, our results improved to 96% meaning that for every 100 test images, 4 were misclassified. If the testing dataset contains simulated lenses, the reader should not translate these accuracies as a measure of physical lens classification performance, neither for our paper nor for others. Unlike physical lenses, simulated lenses may contain statistical artifacts and are limited in shape. Hence, the true performance of a classifier should be measured with its ability to locate physical lenses in a sufficiently large field.
3 A rectified linear unit is a neuron with the following common activation function:f (x) = max(0, x).
To optimize computational time while addressing this issue, we have mixed 15 known lenses from the COSMOS field (Faure et al. 2011) with 3.5 thousand images from the same field and have assigned a lensing probability to each image that was obtained from the output of the lens class neuron. These images were ranked based on their lensing probabilities. The relative ranking for these lenses, i.e. their rank divided by the total number of scanned images, has been plotted in Figure 7. We have also tried eightfold scanning where the probabilities of eight transformations discussed above (Section 3.1) have been summed to improve the accuracy. This technique does not show an improved ranking of the lenses. The majority of the lenses fall under top 8% where they can be further examined by eye. In the next section, we will discuss the images that contaminate the high-rank candidates. We will also discuss the remedies to further separate lenses from other sources.
RESULTS & DISCUSSION
A catalog of the strong gravitational lenses in the COS-MOS field has been generated previously (Faure et al. 2008) by looking at early type bright galaxies in the redshift range of 0.2 ≤ z ≤ 1.0 and visually inspecting and cataloging sixty high and low-quality lens candidates. In contrast, we have examined all sources in HST/ACS iband of the COSMOS field that are more extended than 200 pixels and are 1.5σ brighter than the background, i.e. 230,000 images. After scanning these images with LensFlow, we inspected the top outputs and selected 21 as good quality lens candidates which are presented in Figure 9. Their coordinates and other physical parameters are also listed in Table 1. The lens candidates previously identified by Faure et al. (2008) are marked in Table 1. Among sixty lens candidates presented in Faure et al. (2008), 25 were more extended than our image cutout size of 200 × 200 pixels which translates to larger than 3 × 3 (e.g. COSMOS0208+1422, COS-MOS0009+2455) while some others were considered as lens contaminates or low quality lenses rather than as good lens candidates during our visual selection process (e.g. COSMOS0055+3821). To resolve the former issue, a secondary scan can be performed by down-sampling the field to half its size to include larger lenses.
The main contaminants of our high lensing probability images fall into the main categories of, spiral galaxies with contamination arising from spiral arms and/or ring-like structures, tidally interacting galaxies and satellite/nearby galaxies in a lens-like configuration and image artifacts. Examples of these contaminants are shown in Figure 8. Given that these contaminants are similar to many of our training examples and that it is also time-consuming to separate them by human eye we created another training dataset which would eliminate such contaminants by only looking for perfect Einstein rings. After another scan of the field, we were able to identify the Einstein ring shown in the last two panels of Figure 9.
These indicate that perhaps, rather than increasing the number of layers in series, we must aggregate parallel ConvNets, each one trained to extract lenses from one class of contaminants only. Furthermore, reducing the down-sampling factor would, in turn, increase the resolution of the input images which might increase the gap Figure 9. Identified COSMSOS lens candidates by LensFlow. These candidates were visually selected from images with the highest lens probability assigned by LensFlow. Candidates (18) and (19) were identified by retraining LensFlow on perfect Einstein rings only. RA and Dec of these lens candidates are listed in Table 1. between spirals and lenses at a greater computational cost. And perhaps, it would be more efficient to have separated training datasets for visually distinct lenses, as tried above, rather than having one large training dataset with a variety of lenses. We will study these algorithms in our future publications.
SUMMARY
In this paper, we have emphasized the importance of gravitational lenses in the field of cosmology and have presented an introduction to neural networks including ConvNets. Furthermore, we have laid out the procedure for constructing simulated images for training and testing LensFlow. The importance and details of our normalization and enhancement methods have been discussed. Then, we have discussed the architecture of LensFlow, its accuracy on test data, and its performance on real data. The latter was done by scanning HST /ACS i-band images of the COSMOS field and listing the lens candidates that were visually separated from images with large assigned lens probability by LensFlow.
ACKNOWLEDGEMENT Financial support for this paper was provided by NSF grant AST-1213319 and GAANN P200A150121. Figure 1 and 5 were generated using www.draw.io. The backbone of our algorithm was initially inspired by Hvass Laboratories on Github (Pedersen 2016). Figure 2 was plotted using his code as well. We would like to thank Pierre Baldi for valuable feedback. We would also like to thank Noah Ghotbi and Aroosa Ansari for their assistance. | 2017-05-18T06:47:29.000Z | 2017-05-16T00:00:00.000 | {
"year": 2018,
"sha1": "69e2c1cf246c615f97584c3ed8cbb0092789395d",
"oa_license": null,
"oa_url": "http://arxiv.org/pdf/1705.05857",
"oa_status": "GREEN",
"pdf_src": "Arxiv",
"pdf_hash": "673394135624e049db521d02e872aa272c869fa4",
"s2fieldsofstudy": [
"Physics",
"Computer Science"
],
"extfieldsofstudy": [
"Physics"
]
} |
229438563 | pes2o/s2orc | v3-fos-license | Thermophysical properties of [EMIM][BF4] and [HMIM][PF6] imidazolium ionic liquids with MWCNTs
In this study, several ionanofluids (INFs) were prepared in order to study their efficiency as a cooling medium at 25 °C. The two-step technique is used to prepare ionanofluid (INF) by dispersing multi-walled carbon nanotubes (MWCNTs) in two concentrations 0.5 and 1 wt% in ionic liquid (IL). Two types of ionic liquids (ILs) were used: hydrophilic represented by 1-ethyl-3-methylimidazolium tetrafluoroborate [EMIM][BF4] and hydrophobic represented by 1-hexyl-3-methylimidazolium hexafluorophosphate [HMIM][PF6]. The thermophysical properties of the prepared INFs including thermal conductivity (TC), density and viscosity were measured experimentally. The TC measurement showed an enhancement of about 3% for INF and of 1% MWCNT in [EMIM][BF4] at a temperature of 298.15 K: the TC was 0.186 W/m.K, the kinematic viscosity was 100 centistokes (cSt), and the density was 1.283 g.cm−3. On the other hand, the TC of 1% MWCNT in [HMIM][PF6] INF enhanced by 5%. In this case, at a temperature of 298.15 K, the TC was 0.158 W/m, the kinematic viscosity was 1200 cSt, and the density was 1.294 g.cm−3. Furthermore, the stability of the prepared INFs was measured using the zeta potential method after 28 days of preparation. The results show very good dispersion of the nanoparticles in the ILs for all the prepared INFs. The zeta potential was -69.30 mV and - 45.34 mV for 0.5% and 1% MWCNT in [EMIM][BF4], respectively. On the other hand, zeta potential was -51.78 and -46.67 mV for 0.5% and 1% MWCNT in [HMIM][PF6], respectively. According to the obtained results, the preferable INFs to use as a cooling medium at 25 °C was the INF of 1 wt% MWCNT in [EMIM][BF4], since it provides better thermophysical properties than the other prepared INFs.
Introduction
Water was the first medium used as a heat transfer fluid (HTF) due to its low cost, availability, safety, good thermal stability at the low-temperature range, and its thermophysical properties. However, water has its limitations in high-temperature applications due to its high vapor pressure and associated corrosion problems. Thus, it has been replaced by other liquids such as oils and ethylene glycol. There is an urgent need to replace the conventional cooling liquid in the old heat exchanger units. Researchers have started to look for a new efficient, clean, durable, and environmentally friendly substitute with high flash point characteristics. Several researchers evaluated the use of nanofluid (NF), which is a mixture of different types of nanoparticles in the base fluid. Nanofluids (NFs) are tested and used in various applications, such as solar energy units [1][2], car radiators [3], and heat exchangers (HE) [4]. NFs can be prepared using two techniques: in the one-step technique, nanoparticles are formed inside the base fluid directly producing NF, while the two-step technique includes the previous preparation of nanoparticles followed by dispersing them in the base fluid [5]. A few researchers have made ZICMSE 2020 IOP Conf. Series: Materials Science and Engineering 987 (2020) 012001 IOP Publishing doi: 10.1088/1757-899X/987/1/012001 2 considerable efforts to prepare a surfactant-free NF of MWCNT in a different base fluid. For example, Rehman et al. used Jatropha seed oil as a base fluid using the two-step technique using an ultrasonic probe. This resulted in producing highly stable NF with TC enhancement of 6.76% [6]. NFs of carbon nanotube (CNT) in water or ethylene glycol represent good candidates for replacing conventional HTF. The disadvantage of these NFs are lower stability, higher vapour pressure, increased pumping power, which causes abrasion and attrition problems, and their high cost [7]. To overcome these limitations, NFs were developed to INF, which are a homogenous mixture of nanoparticles with ionic liquids (molten salts). New thermophysical properties were in the resulting ionanofluids, making them suitable to replace many liquids in different applications, especially in heat exchangers. A few researchers have tried to prepare hybrid mixture of IL and base fluid and to disperse nanoparticles in this mixture. For example, dispersing Al2O3 in a [C2mim][CH3SO3]:H2O mixture in a concentration range of 1-10 wt% causes a TC enhancement of about 3.9% -10.2% [8]. The current age of manufacturing tends to use economical low-cost and environmentally friendly methods in many factories. Thus, the use of INFs as a cooling medium in heat exchangers could revolutionise the field of heat transfer. Ionanofluid could be considered as a two-phase system, since it contains the liquid phase represented by the ILs and the solid phase represented by the nanoparticles. INFs combine the advantages of nanoparticles and ILs. Nieto de Castro and colleagues for the first time called this type of fluid 'ionanofluids' [9] [10], which have enhanced properties and negligible vapor pressure and can be designed for a specific application.
Carbon nanotubes
Carbon nanotubes (CNTs) are tiny microscopic tubes made of hexagonal graphene network rolled into cylindrical shells. This hexagonal shape is C6H6 molecule rings connected by single C-C and double C=C bonds. CNTs represent a suitable candidate for heat transfer enhancement due to their high stability level, great surface area, high aspect ratio, high thermal and electrical conductivity, good mechanical strength, and their low toxicity, making them environmentally friendly. The thermal conduction in carbon nanotubes is caused by phonons [11]. CNTs can be divided into two types: single-wall nanotubes (SWNTs) and MWNTs, depending on the number of shells forming the tubes. Furthermore, three different types of CNTs can be formed: armchair, zigzag, and chiral, depending on the graphene hexagonal network sheet rolling axis and the radius of the closing and the chiral vector, which discriminate CNTs into these shapes [12][13] [14].
Ionanofluid history
Many researchers have studied in depth various specifications of INFs, such as their preparation, stability evaluation, prediction of thermophysical properties, application in different fields, and their suitability as a heat transfer medium. Castro et al. presented several publications about INFs in the heat transfer field, reaching a moderate thermal conductivity enhancement of about 2-9% [10]. Furthermore, in their work, comparing traditional HTFs and ionic liquids showed that the addition of 1 wt% of MWCNT in ionic liquid gave the produced INF superior thermophysical properties compared to base ILs [15]. Wang et al. found that increasing MWCNT concentration by more than 1 wt% caused a reduction in the TC because of aggregation forming [16]. Murshed ). Our objective is to find an INF heat transfer medium that has suitable thermophysical properties (i.e. thermal conductivity, density, and viscosity) and that is highly stable over a long period of operation without using any surfactant.
Experimental work 2.1. Materials
MWCNT with CAS number 308068 and a purity of 95% was purchased from the USA from Cheap Tubes Company. Table 1 displays the MWCNT's specifications as received from the company. The weight of the nanoparticles and ionic liquids are measured using a highly sensitive electronic balance inside the laboratory hood to avoid air pollution. Following this, the dispersing process starts and includes two steps: 1-Using a magnetic stirrer for ten minutes at room temperature. 2-Sonicating the mixture using an ultrasonic probe homogeniser for 120 minutes.
Measurement of the thermophysical properties of INFs
Three thermophysical properties of INFs were measured, including the TC, density, and kinematic viscosity.
Thermal conductivity measurement.
The thermal conductivity measurement was carried out using a KD2 Pro apparatus. This device includes different sensors. The KS-1 sensor is a specialised sensor for measuring thermal conductivity/resistivity of liquid samples and insulating materials that have TC < 0.1 W/m·K. The KS-1 sensor must be inserted vertically into the liquid samples to prevent free convection. In addition, the KD2 Pro device must be calibrated using distilled water and glycerine before use.
Density measurement.
The density of the samples was measured using well-calibrated pycnometer with a size of 25 ml corrected to 24.766 ml volume capacity.
Viscosity measurement.
The kinematic viscosity of the samples was measured using capillary tube viscometers with different sizes (C, D, E, F, and G) immersed in a temperature-controlled water bath.
Zeta potential analysis
The Zeta Plus USA device was used to measure the stability of the prepared INFs. This device requires a very small amount of the sampling and gives accurate results within a short time and is suitable for dark colour samples.
The Raman spectra characterisation test of MWCNT nanoparticles was given by the company that produced it (Figure 1). The figure shows two bands: the disorder band (D-band) and the graphitic band (G-band). The ratio between the intensity of the D-band and the G-band, noted ID/IG, is related to the degree of disorder of the MWCNTs. An increase in ID/IG value corresponds to a higher proportion of sp3 carbon, which is generally attributed to the presence of more structural defects [23]. The G-band in the high-frequency region of the spectrum of 1610 cm -1 has an intensity of 5000 a.u. (arb. units), and the D-band of 1360 cm -1 has an intensity of 3500 a.u. Therefore, the ratio of ID/IG equals 0.7, which refers to low structure defects [24].
Transmission electron microscopy (TEM).
The TEM image shown in Figure 2 demonstrates the long tubular structure of MWCNTs tangled with each other. This type of structure is useful in increasing the efficiency of heat transfer.
X-ray diffraction (XRD).
The morphology of MWCNT was analysed using the XRD method. Figure 3 shows the XRD result, which displays a three-phase angle at the diffraction 2 theta; two maximum significant peaks at 25.7 and 44 °C refer to the carbon and graphitic structure of MWCNT. The third weak peak appears at 37.5 °C, which could be related to the types of catalyst used during the preparation of MWCNTs. Recent papers give the same result [25].
Atomic force microscopy (AFM).
The surface morphology of MWCNT nanoparticles can be seen using the AFM method. Figures 4 shows an AFM three-dimensional image for MWCNT nanoparticles. The sample size of MWCNTs was 21385*21385 nm. The AFM shows that the maximum apex of particles was 168.48 nm, the average roughness was 42.2 nm, the root mean square was 48.7 nm, the surface skewness was -0.0535, the surface kurtosis was 1.81, and the surface area ratio was 1.68.
Thermophysical properties of ILs and INFs 3.2.1. Thermal conductivity measurement of ILs and INFs.
The TC of the ILs and INFs of concentrations 0.5% and 1% MWCNT are measured in the temperature range 293-330 K and atmospheric pressure. Figure 5 shows
Kinematic viscosity measurement of ILs and INFs.
The kinematic viscosity of the two ILs and their INFs of the two concentrations of MWCNT nanoparticles were measured in the temperature range 20-90 °C and atmospheric pressure. In general, as shown in Figure 9, the kinematic viscosities of ILs and INFs decrease with temperature. The kinematic viscosity of the EMIMBF4 and its INFs at low concentration decreased almost linearly with increasing temperature. At tge higher concentration of 1% MWCNT in [EMIM][BF4], kinematic viscosity exhibits a sharp non-linear decline from 122.4 to 72.9 cSt for the temperature range 20-31 °C.
Zeta potential measurement
The stability of the produced INFs was detected using the zeta potential method. This test has been done for the prepared INFs, which includes two samples for EMIMBF4 with 0.5% and 1% MWCNT, and two samples for HMIMPF6 with 0.5% and 1% MWCNT on the 28th day of preparation. The obtained result of the zeta potential of EMIMBF4 with 0.5% MWCNT was -69.30 mV and with 1% MWCNT was -45.34 mV, as shown in Figures 11 and 12, respectively, which evidences the good stability of these INFs. Figures 13 and 14, respectively, which further supports their good stability.
Conclusions
The thermophysical properties of HTF are an essential part for heat transfer applications in different fields. In this research, TC enhancement occurred at about 3-5% for the prepared INFs. A slight increase in the INF density values caused by the addition of MWCNTs. On the other hand, INF viscosity values increased largely with increases in MWCNT concentration in the low-temperature range but decreased to low values at the high-temperature range. This is especially clear for the prepared INFs of [HMIM][PF6]. The experimental results also demonstrate the impact of different concentrations of nanomaterial upon physical properties. Furthermore, the zeta potential results show very good stability maintained for all the samplings over the tested period beyond the 28 days of preparation. Notably, no surfactant was used in the preparation of these INFs. In conclusion, the results and data presented in this work are of primary use to heat exchanger designers to provide clear guidance to select heat transfer mediums that are suitable to comply with the heat exchanger design specifications of various exchanger applications, such as electronic and/or solar panel equipment. | 2020-12-03T09:04:57.458Z | 2020-11-28T00:00:00.000 | {
"year": 2020,
"sha1": "ce5dbe154e578179c311ea7ec7379c5adbefda57",
"oa_license": null,
"oa_url": "https://doi.org/10.1088/1757-899x/987/1/012001",
"oa_status": "GOLD",
"pdf_src": "IOP",
"pdf_hash": "69cddf7e60920a8bf10b6f8e8c597a3fbd361e95",
"s2fieldsofstudy": [
"Materials Science",
"Chemistry"
],
"extfieldsofstudy": [
"Physics",
"Materials Science"
]
} |
5302881 | pes2o/s2orc | v3-fos-license | Drug Utilization and Inappropriate Prescribing in Centenarians
Objectives To use primary care electronic health records (EHRs) to evaluate prescriptions and inappropriate prescribing in men and women at age 100. Design Population‐based cohort study. Setting Primary care database in the United Kingdom, 1990 to 2013. Participants Individuals reaching the age of 100 between 1990 and 2013 (N = 11,084; n = 8,982 women, n = 2,102 men). Measurements Main drug classes prescribed and potentially inappropriate prescribing according to the 2012 American Geriatrics Society Beers Criteria. Results At the age of 100, 73% of individuals (79% of women, 54% of men) had received one or more prescription drugs, with a median of 7 (interquartile range 0–12) prescription items. The most frequently prescribed drug classes were cardiovascular (53%), central nervous system (CNS) (53%), and gastrointestinal (47%). Overall, 32% of participants (28% of men, 32% of women) who received drug prescriptions may have received one or more potentially inappropriate prescriptions, with temazepam and amitriptyline being the most frequent. CNS prescriptions were potentially inappropriate in 23% of individuals, and anticholinergic prescriptions were potentially inappropriate in 18% of individuals. Conclusion The majority of centenarians are prescribed one or more drug therapies, and the prescription may be inappropriate for up to one‐third of these individuals. Research using EHRs offers opportunities to understand prescribing trends and improve pharmacological care of the oldest adults.
Setting and Study Design
The study included a nationally representative sample of individuals in primary care from the CPRD who reached the age of 100 between 1990 and 2013, as described previously. 18 Practices can contribute data only when it is deemed of sufficient quality for research, according to a set of criteria that the CPRD group has developed. 19 This study, which received scientific and ethical approval from the Independent Scientific Advisory Committee for CPRD studies (ISAC Protocol 13_151), is based on fully anonymized data, and research ethics committee approval was not required.
Participants
A population-based cohort of centenarians was drawn from the CPRD between January 1, 1990, and September 30, 2013, as previously described. 18 Eligible participants were registered at CPRD general practices during the year in which they turned 100.
Measures
Drug utilization was first evaluated using the main chapters from the British National Formulary (BNF) as categories. The BNF is published every 6 months and provides healthcare professionals with up-to-date information about medicines. 20 Each participant's record was evaluated for drug prescriptions in the following drug categories: gastrointestinal; cardiovascular; respiratory; central nervous system (CNS); antimicrobial; endocrine; gynecological and urinary tract; neoplasms and immunosuppression; nutrition and blood; musculoskeletal; eye; ear, nose, and oropharynx; skin; immunological; and anesthetic. Centenarians were considered to have been prescribed a drug from a particular category if at least one relevant drug was prescribed during their 100th year. Drugs relevant to more than one category were counted in each. For example, prednisolone is included in the endocrine and gastrointestinal categories. Descriptive statistics were used to estimate the total number of prescriptions before reaching the age of 100, from the 95th year to the 100th year of life.
The frequency of PIP among centenarians was evaluated using the updated 2012 American Geriatrics Society (AGS) Beers Criteria. 21 An expert group developed the criteria in 1991, 22 with the latest revision by the AGS in 2012, 21 to provide explicit criteria for PIP in elderly adults. The criteria include three primary groups: a list of potentially inappropriate medication (PIMs) to avoid independent of disease or condition; a list of medications to be avoided in older persons with specific diseases or conditions; and a list of PIMs to be used with caution in older adults. 21 This tool has been widely used to evaluate PIP of drugs to elderly individuals in a variety of settings. [15][16][17][23][24][25] In the present study, we estimated the proportion of centenarians receiving one or more prescriptions from each overall Beers category, independent of disease or condition, including-anticholinergics, anti-thrombotics, anti-infectives, cardiovascular, CNS, endocrine, gastrointestinal, and analgesics. The proportion prescribed at least one specific PIM within each category was also estimated. The standard criteria were used, rather than the disease-specific criteria, because of the large population-based sample in the study. The three drugs that were most frequently prescribed inappropriately were evaluated, and descriptive statistics were used to determine the median number of PIMs in each category. Desiccated thyroid is not listed in the BNF, so it was excluded from the analysis. Glyburide is known as glibenclamide in the United Kingdom. The prescribing of nonsteroidal anti-inflammatory drugs (NSAIDs) in this study was considered inappropriate only if, as stated in the 2012 AGS Beers Criteria, they were not taken with a gastroprotective agent. All topical non-cyclooxygenase-selective NSAIDs were not considered as inappropriate and only oral formulations were included, as stated by the 2012 AGS Beers Criteria. Stata version 13.0 was used to conduct all analysis (Stata Corp., College Station, TX).
RESULTS
A cohort of 11,047 centenarians (8,982 women, 2,102 men), who reached the age of 100 between 1990 and 2013, was selected for analysis. Eighty-four percent were born between 1900 and 1913 and the remaining 16% between 1890 and 1899. The median annual number of prescriptions was 7 (interquartile range 0-12) during the 100th year. Table 1 shows the most frequently prescribed drugs at the age of 100. Drug utilization for each drug class was higher in women than in men, except for urinary tract drugs. The most frequently prescribed drugs overall were those affecting the CNS (53%), including hypnotics, anxiolytics, antidepressants, analgesics, drugs for nausea and vertigo, antiepileptics, and antidementia, as well as cardiovascular (53%) and gastrointestinal (47%) drugs. The distribution of prescriptions issued to men and women centenarians is shown in Figure 1. More than twice as many men as women did not receive any prescriptions in their 100th year, and the proportion of women was higher than men for each increasing category of multiple prescriptions.
The frequency of PIP according to the 2012 AGS Beers Criteria is presented in Table 2. Overall, 32% of centenarians were prescribed a PIM, and the three most frequently prescribed PIMs were temazepam, amitriptyline, and nitrofurantoin. The drug class with the highest proportion of PIP was CNS medications (23%). Approximately one-fifth of centenarians (19% women, 13% men) received anticholinergic drugs, with PIMs including chlorphenamine, hydroxyzine, and promethazine hydrochloride. Despite the high levels of prescribing for gastrointestinal, cardiovascular, and analgesic drugs, low levels of inappropriate prescribing (5%) were observed within these classes.
DISCUSSION
To the knowledge of the authors, this is the first population-based study describing medication use and PIP in centenarians. Only a minority of centenarians did not receive prescription medicines, with a higher proportion of men not receiving any prescriptions, consistent with their superior health status at age 100. 18 Almost 80% of women and 54% of men were prescribed at least one drug during their 100th year. Sex differences in overall prescribing between centenarians have not been reported previously. This disparity could be because of sex differences in health-seeking behavior, for example, not seeking a physician's advice, or nonadherence, as well as resulting from differences in health status. Up to one-third of centenarians were prescribed a PIM. The highest frequencies of PIP were attributed to the use of benzodiazepines (temazepam, diazepam), amitriptyline, and nitrofurantoin. The 2012 AGS Beers Criteria recommendation to avoid all of these drugs is "strong," and the reported quality of evidence is "high" for avoiding temazepam, diazepam, and amitriptyline and "moderate" for avoiding nitrofurantoin. 21 The 2012 AGS Beers Criteria recommend avoiding nitrofurantoin in individuals with creatinine clearance less than 60 mL/min because of concerns about lack of efficacy from inadequate drug concentrations in the urine. In view of the advanced age of the cohort, it is likely that most will have poor renal function and should therefore be using safer alternatives such as ciprofloxacin or trimethoprim. 26
Comparison with Existing Literature
Existing studies on drug utilization in elderly adults have often focused on younger cohorts of old people 7,9,10,12 or do not include centenarians. 9,13,14 There is scarce evidence about PIP in extreme old age. These studies focusing on younger elderly adults tend to rely on self-reported questionnaires and interviews, 7,15-17,24,27 resulting in a high risk of responder bias.
There have been no previous studies reporting PIP in a large group of centenarians. Reports of prevalence of PIP in individuals aged 65 and older are inconsistent, and few studies have used the updated 2012 criteria. One study 27 using the 2012 criteria reported a higher prevalence of PIP (44%) in Spanish individuals aged 65 and older than the present findings (32%) in centenarians. The most frequently prescribed PIMs were benzodiazepines, similar to the present data. Another study in New Zealand 24 also reported a higher PIM prevalence (42.7%) for communitydwelling individuals aged 75 and older than for those aged 100 and older in CPRD, whereas another 15 reported a prevalence of 17%. All three studies 15,24,27 used selfreported data. Several studies reported pain medications as the most commonly prescribed PIMs, 15,17,23,24 including two studies using the Screening Tool of Older People's potentially inappropriate Prescriptions/Screening Tool to Alert doctors to the Right Treatment (STOPP/START) criteria. 28,29 This is inconsistent with the present findings in centenarians, reporting CNS medications and anticholinergics as the most commonly prescribed PIMs. It is not stated in all these previous studies whether concurrent use of gastroprotective agents was considered alongside NSAIDs.
Strengths and Limitations
This study had the strengths of a large sample drawn from a representative population of U.K. general practices. In the United Kingdom, approximately 98% of individuals are registered with a family practice, ensuring that the present data are complete and nationally representative. Individuals aged 75 and older have an annual review of medicines, 30 and those with four or more medicines are reviewed every 6 months, although this review was not introduced until 2002. Using primary care EHRs allowed for the classification of prescribing according to drug category and of specific PIMs, but data were not available for several variables of interest, including whether an individual lived alone or whether they lived in an urban or rural location. EHRs circumvent the problem of recall bias, a limitation of many drug use studies relying on self-reported questionnaires or interviews to collect data on prescrip- tions. There is a possibility that some prescriptions will not be filled or consumed, and the data may not capture any over-the-counter or secondary care hospital prescriptions, so the present findings may underestimate drug usage of centenarians. Another limitation is that the sample was exposed to a nonuniform drug formulary because the study considers data over a 23-year span. Medical practice in primary care has evolved over time, and as a result, certain drugs used in 1990 may now be considered inappropriate. Several new drugs have also been introduced that could not be prescribed in 1990.
There are limitations of the 2012 AGS Beers Criteria, given their universal application without careful consideration of an individual's response to each drug. 15,25 There may also be some drugs used in the United Kingdom that were not captured. In 2003, the STOPP/START criteria were developed to identify potential errors in prescribing and prescribing omission in older people according to physiological system. 24,28 These criteria were created as a Europe-focused tool but are not as widely used as the 2012 AGS Beers Criteria to evaluate the epidemiology of PIP because of their specificity. Although a more-individualized tool may be favored in clinical practice, this approach is less feasible for epidemiological investigation of population-based samples. The 2012 AGS Beers Criteria were specifically designed to use pharmacy records with minimal additional clinical information so that they could be applied to chart reviews or computerized data sets. 21 According to the 2012 AGS Beers Criteria, the most commonly prescribed PIMs in the present cohort were benzodiazepines, tertiary tricyclic antidepressants, nitrofurantoin, and ibuprofen. The frequencies of PIP found in this study should be interpreted cautiously because each person's risk:benefit ratio for a drug will depend on his or her physiological and clinical status. Only an individual evaluation of each person will confirm the validity of these interpretations. Nevertheless, population-based studies provide useful epidemiological data on relative frequencies of PIP in large cohorts and help identify the inappropriate medications that are most frequently prescribed. This study also identifies specific drugs that may be given more attention in further research on the determinants of PIP in elderly adults.
Conclusion and Implications for Clinical Practice
Polypharmacy and multimorbidity in elderly adults present health professionals with a significant clinical responsibility. The limited empirically based evidence to guide drug prescription in very old adults means that physicians must base their prescribing decisions on clinical knowledge and prior experience with similar conditions, likely from younger cases. There is an urgent need for studies to explore the efficacy, safety, and harms associated with drug prescribing for different chronic conditions in very old adults. There is also the need to model the adverse clinical and economic consequences of inappropriate therapeutic decision-making in this group. This is the first study to use primary care EHR data to describe prescribing trends in UK centenarians, as well as the extent of PIP according to drug class. It provides proof of concept for using a large EHR database to evaluate appropriateness of prescribing and a basis for reevaluating indicators of appropriate prescribing as applied to this age group. This also offers valuable data for the modeling of future healthcare needs and costs of the oldest adults in the United Kingdom. | 2018-04-03T00:15:39.962Z | 2016-04-30T00:00:00.000 | {
"year": 2016,
"sha1": "2de1c9209a96e34945908b25dbc275e0a3db87ba",
"oa_license": "CCBY",
"oa_url": "https://doi.org/10.1111/jgs.14106",
"oa_status": "HYBRID",
"pdf_src": "PubMedCentral",
"pdf_hash": "2de1c9209a96e34945908b25dbc275e0a3db87ba",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
268361903 | pes2o/s2orc | v3-fos-license | Uniaxial‐Oriented Perovskite Films with Controllable Orientation
Abstract Perovskite films with large crystal size, preferred orientation, and facile fabrication process, combining advantages of single‐crystal and polycrystalline films, have gained considerable attention recently. However, there is little research on the facet properties of perovskite films. Here, (111)‐ and (001)‐oriented perovskite films with bandgaps ranging from 1.53 to 1.77 eV, and systematically investigated their orientation‐dependent properties are achieved. The (111)‐oriented films show electron‐dominated traps and the (001)‐oriented films show hole‐dominated traps, which are related to their atomic arrangement at the surface. Compared with the (001)‐oriented films, the (111)‐oriented films exhibit lower work function and superior water/oxygen robustness. For the wide‐bandgap films, the lattice of the (001)‐oriented film provides an unobstructed passage for ion migration. Comparably, the (111)‐oriented films exhibit suppressed ion migration and excellent phase stability. The optimized unencapsulated solar cells based on both (001) and (111) orientations show a similar high efficiency of ≈23%. The (111)‐oriented solar cell exhibits excellent stability, maintaining 95% of its initial efficiency after 1500 h maximum power point (MPP) tracking test, and 97% initial efficiency after 3000 h aging in ambient conditions. This work paves the way for the rational design, controllable synthesis, and targeted optimization of uniaxial‐oriented perovskite films for various electronic applications.
Introduction
Metal halide perovskites are expected to lead the revolution in photovoltaics with their unique optoelectronic properties of DOI: 10.1002/advs.202401184[3][4] Up to now, the power conversion efficiency (PCE) of metal-halide perovskite solar cells (PSCs) has surpassed 26%. [5,6]However, the further commercialization of PSCs is hindered by their short lifetime which is ≈1 order of magnitude lower than silicon solar cells. [7]10][11][12][13][14][15][16][17] Unfortunately, challenges for single-crystal perovskite photovoltaics remain in terms of material growth, uncontrollable thickness, and area, [18][19][20][21][22] unsatisfactory contact between single-crystal and charge transport layers [23,24], and high surface defect density. [25]As a result, the maximum PCEs of single-crystal-PSCs are still far below their polycrystalline counterparts. [12]It is worth noting that the polycrystalline films prepared by solution-process are still greatly attractive due to their facile, controllable, and low-cost fabrication process, [26][27][28] although the existence of grain boundaries and surfaces may lead to massive trap densities and reduce the PCE and stability of PSCs. [29,30][33] On one hand, compared with polycrystalline films, the higher phase purity and larger grains of perovskite films could effectively reduce the grain boundary and thus suppress defects and ion migration [34,35] ; on the other hand, the single orientation of perovskite provide an effective model for further investigating the working principles of PSCs that are related to the surface and grain boundaries. [36]For example, the (220) facet of MAPbI 3 single crystal exhibits higher electron and hole conductivities, and the (112) facet shows enhanced ionic conductance than that of (100) facet. [37,38][41] Similarly, the FAPbI 3 grains also show facet-dependent stability where the (111) facet is much more resistant to moisture than the (100) facet. [14]These results indicate that exquisite control of film orientation is an effective strategy to further optimize the photovoltaic performance of PSCs.
Although previous research has demonstrated the anisotropic properties of perovskites, such as photoelectric properties, carrier mobility, and defect distribution, and emphasized the impact of orientation on device performance, they mainly focus on singlecrystal perovskites or different grains in polycrystalline films.Few studies have focused on the orientation-dependent properties of perovskite films.Ma et al. [42] reported that carrier mobility and photocurrent of the (100) and (111) facets of FAPbI 3 -based perovskite film are much higher than those of the (110) facets.However, Zhang et al. [43] stated the opposite conclusion that the (110) facet had higher conductivity, enhanced charge carrier mobility, and better-matched energy level alignment compared to the (100) facet.Therefore, systematic investigations about the controlled synthesis and anisotropic film properties of perovskite films with different orientations are urgently needed to achieve PSCs with higher PCE and enhanced stability.
In this work, we synthesized Cs 0.03 (FA 0.90 MA 0.10 ) 0.97 Pb(I 1-x Br x ) 3 (x = 0.1-0.4)perovskite films with (111)-and (001)-preferred orientations by antisolvent engineering.All of the films were deposited with the same process, providing an objective platform to evaluate the intrinsic orientation-dependent film properties.For both normal bandgap (NBG) and wide bandgap (WBG) perovskite films, the (111)-oriented films exhibit lower hole trap density and higher electron trap density than the (001)-oriented films, which are related to their surface termination and defects types on different crystal facets.Moreover, the (111)-oriented films exhibit better stability than (001)-oriented ones due to the inorganic PbI 6 octahedra layer on their surface.For WBG perovskite, the commonly used (001)-oriented films show severe phase separation as the ion migration channels are parallel to the direction of the built-in electric field.In contrast, the (111)-oriented films exhibit excellent phase stability as it block ion migration channels along the direction of the electric field.Solar cells based on both (111) and (001)-oriented films show the champion PCEs of ≈23% for NBG PSCs and ≈19.8% for WBG PSCs.The (111)-oriented PSCs exhibited better stability as it maintained 97% initial PCEs after 3000 h aging in ambient conditions.This work provides guidance for controllable synthesis and targeted optimization of perovskite films with different orientations, as well as rational design for novel electronic applications, which would further promote the development of perovskite devices.
The GIWAXS was conducted to characterize the orientations in both in-plane and out-of-plane directions of NBG (x = 0.1) (Figure 1d-f) and WBG (x = 0.3) (Figure 1g-i) perovskite films.The diffraction rings at q z = 1.4 Å −1 are assigned to the (111) planes.The rings at q = 1.0 and 2.0 Å −1 are assigned the (001) and (002) planes, respectively.The azimuth angle () of 10°and 54.7°in Figure 1e,h indicate that the (001) and (002) planes are inclined at 10°and 54.7°with respect to the surface, which agrees with (111)-oriented films.Similarly, the of 0.0°and 45.0°fit well with the (001)-dominated perovskite films.These results are consistent with the XRD spectra.The lower signal of (111) plane compared with (001) plane is due to its weaker interaction with the grazing-incidence X-ray and the intensity loss of q z . [44]Compared with CB-treated film, the IPA and IPA + MACl/PEACl treated films show more focused diffraction spots, indicating the highquality perovskite films with well-aligned orientations of grains.This is a clear demonstration that antisolvent engineering is an effective method to control the crystal orientation of perovskite films.
Properties of (111) and (100)-Oriented Perovskite Film
To further investigate the impact of orientation on the properties of perovskite films, space charge limited current (SCLC) was conducted for NBG (x = 0.1, 1.6 eV) and WBG (x = 0.3, 1.7 eV) films.Some targeted precursor additives were added for the defect-passivation of films with different orientations.The details of the fabrication process are shown in the Experimental Section. Figure S3 (Supporting Information) shows that the precursor additives have little effect on the orientation of perovskite films.Considering the different atomic arrangements on different crystal facets, the defect distribution could differ with the film orientation.The trap density (n trap ) and carrier mobility (μ) were calculated from SCLC measurements for These results indicate that the perovskite films with preferred orientation have a lower density than random orientation (RO) films.Moreover, the hole trap density of (111)-perovskite films is lower than that of (001)-perovskite films, while the result of electron trap density is exactly the opposite.The carrier mobilities were derived from the SCLC results.The (111)-oriented films exhibit higher hole mobility but lower electron mobility than that of (001)-oriented films.The values of carrier trap density and mobility are summarized in Table S2 (Supporting Information).WBG perovskite films exhibit similar phenomena.We speculate that this is related to the different types of defects on (111) and (001) facets, due to their different atomic terminations.As shown in Figure 2h, Pb-dangling bonds dominate the (111) crystal plane which tends to capture free electrons, thus leading to a higher density of electronic defects.In contrast, the (001) facets have mainly MA vacancies which tend to capture holes and lead to a high density of hole defects.Surface electrical properties of (111)and (001)-oriented films measured by the Kelvin Probe Force Microscopy (KPFM) also confirm this speculation.As shown in Figure S6 (Supporting Information), the surface of (111)-oriented film exhibits lower work function than that of (001)-oriented film, since the exposed halogens on (111) facets provide sufficient excess electrons, and the exposed cations on the (001) facet acting as the hole doping agents.
Based on the difference in electron/hole trap density and atomic arrangement on (111) and (001) facets, it can be inferred that the (111)-oriented perovskite films prefer to interact with the electron-donating group (Lewis acids), whereas the (001)oriented perovskite films could be sensitive to electron withdrawing group (Lewis bases).Targeted surface passivation strategies should be designed for perovskite films with different orientations, which will be studied in further research.
Stability tests of NBG and WBG perovskite films with different orientations were tested.The perovskite films were stored in ambient air (16-30 °C, 10-30% relative humidity (RH)) for 30 days and then placed on 85 °C hot stage in air for 72 h for accelerated aging.The XRD spectra of fresh and aged perovskite films are shown in Figure 3a,b.The RO and (111)-oriented perovskite films exhibit less degradation than the (001) films.The appearance of a peak at ≈12.7°or ≈11.7°indicates the -phase perovskite degrades into PbI 2 or transforms into non-optical active -FAPbI 3 .Scanning electron microscope (SEM) images of RO, (111), and (001)-oriented perovskite are consistent with the XRD results (Figure S7, Supporting Information).The relative response spectral degradation ratios [45] (RSD: the ratio between the highest XRD peak of the aged film and the fresh film) with different orientations were calculated to quantify the stability of perovskite films.The RSDs of (001)-and (111)-oriented NBG perovskite films are 39.4% and 45.5%, respectively.The RSDs of (001)and (111)-oriented WBG perovskite films are 18.6% and 28.2%, respectively, proving that (111)-oriented films have a slight advantage in stability compared to (001)-oriented films.Time-Resolved Photoluminescence (TRPL) results before and after aging also confirm this conclusion, as the carrier lifetime of (111)-oriented films exhibited less deterioration after 2-3 h aging under 85 °C in air (Figure 3c-f; Table S3, Supporting Information).
Orientation-Dependent Ion Migration
Light-induced phase separation that significantly limits the PCE and stability of PSCs is a critical issue for I/Br mixed WBG perovskite films.[48][49] We studied the relationship between film orientation and phase separation in WBG films with different orientations.Figure 4a-c shows the in-situ PL results of WBG films with RO, (111), and (001) orientations under continuous laser illumination (405 nm, 20 W cm −2 ).All the samples show a new PL peak at ≈ 530 nm (corresponding to the PL peak of Cs y (FAMA) 1-y PbBr 3 ) [50] under 405 nm laser irradiation, indicating the phase separation of halide ions and A-site cations.Figure 4d shows the variation of peak intensity at 530 nm over time.The peak intensity of RO and (001)-oriented perovskite films increases continuously over time.In contrast, the (111)-oriented films exhibit much higher phase stability with a very low peak intensity at 530 nm even under prolonged laser irradiation up to 1000 s (Figure S8, Supporting Information).
To further verify the origin of phase separation, devices with structure ITO/perovskite/Au were prepared based on perovskite films with different orientations.A 20 V bias was added with the electric field from Au (front) to ITO (back).The steady-state PL of the front side and back side of the devices were measured (Figure 4e-g; Figure S9, Supporting Information).Before applying the bias, all films show the same PL peak position on both sides, indicating the homogeneity of perovskite films in the vertical direction.After applying 20 V bias for 5 min, a blue shift of PL peak at the front side and a red shift at the back side was observed for (001)-oriented and RO films, indicating phase separation caused by the electric field.In contrast, the (111) film shows almost no PL peak shift, exhibiting high resistance to ion migration.These results indicate that the photogenerated electric field is a major cause of ion migration and phase separation in the perovskite films.
The schematics in Figure 4h show the ion migration channels of (001) and (111) films under an external electric field.The (001)oriented film provides ion migration channels that are parallel to the electric field, which is most conducive to ion migration and phase separation.[53] In contrast, the tilted lattice of (111)-oriented film hinders ion migration along the direction of the electric field, which is more suitable for suppressing phase separation of WBG perovskites.One can conclude that the (111)oriented film is a good choice for traditional optoelectronic devices such as solar cells, light-emitting diode, and photodetectors.56] The current density-voltage (J-V) curves of WBG PSCs are shown in Figure 4i and Figure S10 (Supporting Information).The (111)-oriented PSCs show a PCE of 19.8%, a opencircuit voltage (V OC ) of 1.3 V, a short-circuit current (J SC ) of 21.1 mA cm −2 , and an fill factor (FF) of 72.6%.The (001)oriented PSCs show a PCE of 19.8% with a V OC of 1.3 V, a J SC of 21.1 mA cm −2 , and an FF of 71.9%.We further compared the stability of unencapsulated WBG PSCs by tracking the PCE evolution at maximum power point, as shown in Figure 4j.Three devices for each group were involved to ensure reliability (Figure S11, Supporting Information).To accelerate the phase separation, PSCs were operating under AM 1.5G and a superimposed ultraviolet (UV) illumination (365 nm, 50 mW cm −2 ).The (111)oriented PSCs retained ≈97% initial efficiency after 1000 s MPP tracking test, exhibiting the best working stability under intense illumination.However, the efficiencies of (001)-oriented and RO PSCs show severe degradation as the PCE quickly drops to 88% and 85% of the initial efficiency, respectively.The fast degradation could be attributed to the severe ion migration and phase separation of perovskite films under high-intensity light and heat, as well as new defects caused by the ion migration.
Device Performance
Solar cells based on (111)-and (001)-oriented perovskite films were fabricated and investigated to reveal the effect of orientation on device performance.We adopt the n-i-p structure (ITO/SnO 2 /perovskite/Spiro-OMeTAD/Au) for PSCs.The typical J-V characteristics and photovoltaic parameters of devices based on (111)-and (001)-NBG devices are shown in Figure 5a.The (111)-oriented perovskite shows a PCE 22.8%, with a V OC of 1.21 V, a J SC of 23.97 mA cm −2 , and an FF of 78.6% at the forward scan.The (001)-oriented perovskite device shows a similar PCE of 23.1% with a V OC of 1.21 V, a J SC of 23.99 mA cm −2 , and an FF of 79.9% at the forward scan.As shown in Figure 5b, the incident photon-to-current conversion efficiency (IPCE) values and integrated J SC of champion devices with (111) and (001) orientation are well matched with the J-V curves.Figure S12 (Supporting Information) shows the statistics of 50 devices based on (111) and (001) orientations.For long-term stability, the unencapsulated (111)-oriented perovskite device retained ≈97% of its original PCE after aging in the air for 3000 h, whereas the (001)-oriented PSCs only retained ≈76% initial PCE (Figure 5c).The stable power output (SPO) of the (111)-oriented device was 22.5% at a bias of 1.01 V, whereas the (001)-oriented device was 22.2% at a bias of 1.00 V (Figure 5d, red circle).After a 30 min test, the (001)-oriented PSC showed a slight light-induced degradation and the PCE dropped to 95% of its initial value, which is a common issue in PSCs with a normal device structure.In contrast, the (111)-oriented PSC exhibited almost no degradation and maintained ≈100% of its initial PCE with a stable photocurrent output (Figure 5d).Furthermore, the operating stability of unencapsulated devices was tested under a nitrogen atmosphere at room temperature.As shown in Figure 5e, the (111)oriented PSC remained at 95% of its initial efficiency after 1500 h MPP tracking test under one sun illumination, while the (001)oriented PSC slightly dropped to 90% of its initial efficiency.The excellent working stability of (111)-oriented PSCs could be mainly attributed to the superior environmental and phase stability of (111)-oriented perovskite films.For (001)-oriented PSCs, targeted water/oxygen protection and surface modification strategies should be designed to improve the film and device stability.
Conclusion
We achieved (111)-and (001)-oriented perovskite films with bandgap ranging from 1.53 to 1.77 eV by simple antisolvent engineering with a universal method.This provides an objective way to compare the effects of orientation on the intrinsic properties of perovskite films, eliminating the effects of precursor additives, which were commonly used in previous reports.We therefore performed comprehensive studies on the orientation-dependent film properties including defect distribution, ion migration, carrier mobility, and photo/electric stability for perovskite films based on RO-, (111)-, and (001)-orientations.We found that the intrinsic film properties were highly dependent on the film orientation.Therefore, targeted defect passivation and application scenarios for different film orientations were necessary and have been discussed to further improve the device's performance.The (111)-oriented films exhibit higher environmental stability due to their inorganic-dominated surface termination.Ion migration and phase separation were also studied to address these highly concerned issues in WBG perovskite films.The (111)-oriented film could effectively suppress the phase separation as the tilted lattice hinders ion migration stimulated by an electric field.In contrast, the (001)-oriented films exhibit pronounced ion migration, leading to severe phase separation in WBG films, which could be utilized in other electronic applications.As a result, the (111)-oriented PSCs exhibit higher stability than the (001) groups with a shelf lifetime of T 97 = 3000 h in an ambient environment.This work paves the way for rational design and controllable synthesis of uniaxial-oriented perovskite films.The systematic analysis of orientation-dependent properties provides guidance for the design of novel electronic applications of perovskites, which would further promote the development of perovskite photovoltaic and optoelectronic devices.
Experimental Section
Materials: Unless otherwise stated, all chemicals were purchased from Tokyo Chemical Industry Co., Ltd., and Xi'an Polymer Light Technology Crop., and all solvents were purchased from J&K Scientific Ltd.And used as received.The SnO 2 colloid precursor was purchased from Alfa Aesar (Tin (IV) oxide, 15% in H 2 O colloid dispersion).
Film Fabrication of Perovskite: 1.5 m Cs 0.03 (FA 0.90 MA 0.10 ) 0.97 Pb(I x Br 1-x ) 3 (x = 0.0-0.4)perovskite precursor was obtained by dissolving FAI, MAI, CsI, PbBr 2 , PbI 2 and 5% mmol RbCl in a mixed solvent of DMF:DMSO = 4:1, and adopted a one-step spin-coating process with a two-step spin-coating procedure (2000 rpm for 10 s followed by 6000 rpm for 20 s.).For (111)/(001)-perovskite, IPA/IPA+additive was dropped onto the spinning substrate during the second spin-coating step at the 12 s.The additive in IPA for NBG is 3 mg mL −1 MACl and for WBG is 1 mg mL −1 PEACl.For randomly oriented perovskite, CB was dropped onto the spinning substrate during the second spin-coating step at 12 s.The substrate was then immediately transferred on a hotplate and annealed, 120 °C annealed for 20 min for NBG perovskite film and 140 °C annealed for 20 min for WBG perovskite film.
Device Fabrication: For perovskite devices, the glass/ITO substrates were sequentially cleaned using distilled water, acetone, and isopropanol.The SnO 2 electron transporting layer was obtained by dissolving SnO 2 colloid dispersion in distilled water (SnO 2 : H 2 O = 1:6) and ultrasound for 4 min before spin-coated (4000 rpm, 30 s) on glass/ITO substrates in ambient air and annealed at 170 °C for 30 min. [57,58]Perovskite films were then deposited on the as-prepared substrate.After cooling down to room temperature, Spiro-OMeTAD (72.3 mg mL −1 in CB), tert-butylpyridine (29 μL mL −1 ), and bis(trifluoromethane)sulfonimide lithium salt (18.5 μL mL −1 , 520 mg mL −1 in acetonitrile) was subsequently deposited on top of the perovskite film by 4000 rpm for 30 s. Finally, 55 nm of gold electrodes were deposited on top of the devices by thermal evaporation in a vacuum of≈10 −5 Torr.For the SCLC test, the device structure of the electron-only device was glass/ITO/SnO 2 /perovskite/PCBM/Au and the structure of hole-only device was glass/ITO/PEDOT: PSS/perovskite/Spiro-MeOTAD/Au.All layers were fabricated using the same method as the solar cells except PE-DOT: PSS and PCBM.The PEDOT: PSS layer was spin-coated (4000 rpm 30 s) on the ITO/glass substrates in ambient air, and then annealed at 135 °C for 10 min.The PCBM layer was obtained by dissolving PC 61 BM in CB (10 mg mL −1 ) and spinning-coated at 2000 rpm for 30 s.
Device and Film Characterization: The current-voltage (J-V) characteristics of devices were tested via a Keithley 2635B source meter under simulated AM 1.5G (100 mW cm −2 ) solar irradiation in the air.A calibrated silicon solar cell from Newport Corp. was used to calibrate the light intensity.The J-V curves were tested from 1.3 to −0.1 and −0.1 to 1.3 V with a scanning speed of 0.02 V s −1 .The masked active area was 0.0805 cm 2 .The surface morphology of the perovskite films was characterized by a FEI Apreo C with an electron beam energy of 10 kV.TRPL spectra were measured on a platform built by ourself base on RedWave Labs APD using a 405 nm pulse laser and the light was illuminated from the perovskite film side.The KPFM was performed on perovskite samples in the air using AFM (Bruker Dimension Icon).For KPFM, the scanning was done in lift mode, during the first scanning the surface topography was directly obtained, and second scanning gave the surface potential information.The PL was executed by A laser scanning confocal microscope (Self-built platform) equipped with a 405 pulse laser and Mach-DSP scanner.XRD patterns were recorded on a Rigaku smartlab X-ray Diffractometer.The GIWAXS data were obtained at 1W1A Diffuse X-ray Scattering Station, Beijing Synchrotron Radiation Facility (BSRF-1W1A).EQEs were measured by a solar cell quantum efficiency measurement system (QE-R) supported by Enli Technology Co., Ltd.The absorption spectra were measured by a Hitachi UH4150 spectrophotometer.All the tests were taken in the ambient air at a temperature of 18-35 °C and relative humidity ≈ 10-30% except for SEM.
Statistical Analysis: All statistical analysis was performed with Origin-Pro 2021.The original data obtained from PL, TRPL, and XRD in Figure 1, Figure S2a,b (Supporting Information) were normalized.The other data were obtained by transferring the corresponding original data according to the calculation formula.Linear fitting was applied to SCLC.The biexponential decay function was applied to TRPL decays to infer the carrier extraction/recombination dynamics.GIWAXS (Jianyao Huang, GIWAXS Tools, Version [current version], https://gitee.com/swordshinehjy/giwaxsscript (accessed date).
Figure 2 .
Figure 2. a,b) Average SCLC results of NBG perovskite films treated with different antisolvents based on electron-only and hole-only devices.c) Histograms of trap densities of NBG perovskite films based on electron-only and hole-only devices.d,e) Average SCLC results of WBG perovskite films treated with different antisolvents based on electron-only and hole-only devices.f) Histograms of trap density of WBG perovskite films based on electron-only and hole-only devices.g) Diagram of the structure of SCLC devices.h) Diagram of types of defect and corresponding interface passivation agents for (111)-and (001)-oriented perovskite films.
Figure 3 .
Figure 3. Stability test of different oriented-perovskite films.XRD and RSD results of fresh and aged samples with RO, (111), and (001) orientation for a) NBG and b) WBG perovskite films.TRPL results of c,d) NBG and e,f) WBG perovskite films with (111) and (001) orientation before and after aging.
Figure 4 .
Figure 4.In situ PL of peak position and intensity of a) RO, b) (111)-oriented, and c) (001)-oriented WBG perovskite films with laser irradiation time.d) The variation of PL intensity at 530 nm with time of WBG perovskite with different orientations.e,f) PL spectra of the e) front and f) back side of the 001-oriented film before and after applying bias.g) Statistics of the PL peak positions before and after applying bias.h) Schematic of ion migration in perovskite films under an electric field.i) J-V curves and performance of WBG PSCs.j) Stability test of different orientations under the simultaneous irradiation of AM 1.5G and a superimposed UV light (365 nm, 50 mW cm −2 ).
Figure 5 .
Figure 5. Performance of NBG PSCs based on (111)-and (001)-oriented films.a) J-V curves and performance parameters, F and R represent the results of forward and backward scanning tests, respectively.b) IPCE values and its integrated J SC. c) Stability test of unencapsulated devices stored in the air.d) SPO results.e) MPP tracking results under one sun illumination. | 2024-03-13T06:17:56.004Z | 2024-03-11T00:00:00.000 | {
"year": 2024,
"sha1": "8b2b6eefd4eeeaff091ff6da5a5835bdd349b184",
"oa_license": "CCBY",
"oa_url": "https://onlinelibrary.wiley.com/doi/pdfdirect/10.1002/advs.202401184",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "d92526419ab9099bd019427e847c7954f8c81db9",
"s2fieldsofstudy": [
"Materials Science",
"Physics"
],
"extfieldsofstudy": [
"Medicine"
]
} |
14183171 | pes2o/s2orc | v3-fos-license | Brane World Models With Bulk Scalar Fields
We examine several different types of five dimensional stationary spacetimes with bulk scalar fields and parallel 3-branes. We study different methods for avoiding the appearance of spacetime singularities in the bulk for models with and without cosmological expansion. For non-expanding models, we demonstrate that in general the Randall-Sundrum warp factor is recovered in the asymptotic bulk region, although elsewhere the warping may be steeper than exponential. We show that nonsingular expanding models can be constructed as long as the gradient of the bulk scalar field vanishes at zeros of the warp factor, which are then analogous to the particle horizons found in expanding models with a pure AdS bulk. Since the branes in these models are stabilized by bulk scalar fields, we expect there to be no linearly unstable radion modes. As an application, we find a specific class of expanding, stationary solutions with no singularities in the bulk in which the four dimensional cosmological constant and mass hierarchy are naturally very small.
I. INTRODUCTION
The Randall-Sundrum (RS) model [1] for a warped 5-D geometry can account for the mass hierarchy, while providing a fine-tuning mechanism for canceling the 4-D cosmological constant without requiring a vanishing 5-D vacuum energy [1,2]. Variants of the model have been constructed in which the cosmological constant is exponentially small [3], and in which both the cosmological constant and the hierarchy problems may be solved simultaneously [4]. However, the RS model and similar models that solve the hierarchy problem are untenable: the radion-mediated interaction of matter on the visible brane dominates over the gravitational interaction [5,6]. This has the consequence that matter on the visible brane (with positive brane tension) gives a negative contribution to the square of the four dimensional Hubble parameter for homogeneous cosmological models [7,4]. Another manifestation of the problem is that the spacetimes are dynamically unstable: the 3-branes which they contain, whose positions must be arranged carefully to reproduce the mass hierarchy, move away from those special positions when slightly perturbed [8,9].
As suggested by Goldberger and Wise (GW) [5], the introduction of bulk scalar fields can solve these problems by giving a mass to the radion mode and stabilizing the positions of the branes. This is consistent with phenomenology if the radion mass is at or above the electroweak scale [10]. DeWolfe, Freedman, Gubser and Karch (DFGK) [11] showed how fully consistent stationary spacetimes could be constructed with bulk scalar fields, including the gravitational back-reaction of those fields; there have been further studies of spacetimes with bulk scalars along similar lines [12,13]. DFGK argued that the requirement that the induced metric on the brane be flat requires one fine-tuning of the parameters describing the solutions. Generalizing the class of solutions by allowing a non-zero effective 4-D cosmological constant Λ 4 relaxes this fine-tuning [11].
The purpose of this paper is to extend the work of DFGK [11] and Kakushadze [13], to discuss some general properties of 5-D spacetimes with a bulk scalar field, and to apply those properties to the solution of the cosmological constant and the hierarchy problems. The main points of this paper are as follows: • The principal problem associated with the introduction of a bulk scalar field is that generic stationary solutions contain timelike curvature singularities in the bulk at finite distances from the branes. This occurs both for static solutions (Λ 4 = 0) and for stationary solutions with cosmological expansion (Λ 4 = 0). Such singularities have already been encountered in work on self-tuning of the cosmological constant [14]. It is possible to avoid these singularities in two ways. One well-known way is to simply orbifold or otherwise compactify the fifth dimension in such a way that the singularity is never encountered. A second way, which is perhaps preferable, is to carefully choose the scalar field potential in such a way that the occurrence of singularities is prevented. In Sec. II below, we discuss a simple condition (first considered by Kakushadze [13]) for identifying such preferred potentials for static spacetimes, and present several examples. One interesting example is obtained by compactifying a 11-D spacetime down to 5-D with a single radion field which acts as a bulk scalar, as well as a 7-form field strength in 11-D descends to a non-dynamical 5-form field strength in 5-D that generates a potential for the scalar field (Sec. III C 1 below). By adjusting the 5-form field strength, this model can be rendered nonsingular.
• In the RS model, the metric's "warp factor" falls off exponentially as one moves away from the Planck brane, and this exponential fall-off underlies the RS solution of the hierarchy problem. In many models with scalar fields, the fall-off of the warp factor is faster than exponential, which facilitates solving the hierarchy problem (see the models in Secs. III A and III B below). This point was mentioned in passing for a specific model by DFGK.
• We show that allowing the four dimensional cosmological constant Λ 4 to be non-zero exacerbates the tendency to form spacetime singularities in the bulk. In particular, models with bulk scalar fields that are nonsingular when Λ 4 = 0 should be expected to become singular for nonzero Λ 4 , although the corresponding singularities are rather mild. The singularities are weak enough (see Eq. [40]) to be removed by simply requiring the bulk to become effectively AdS as the warp factor A(y) → 0. In Sec. IV below we derive a general method of constructing nonsingular models with cosmological expansion based on this idea.
• We can now construct models without bulk curvature singularities. With us living on the visible brane, it is possible to account for the extreme smallness of the electroweak scale and the cosmological constant simultaneously (Sec.V below). These models are analogous to those of Refs. [3,4], with the additional feature that the stability problems of Refs. [3,4] have been cured.
II. SETUP AND GENERAL CONSIDERATIONS
In this section we outline the general framework, review the procedure introduced by DFGK for constructing static and stationary solutions, and derive criterion under which bulk singularities do not occur.
A. Basic equations
We consider 5-D gravity plus a bulk scalar field with parallel 3-branes, for which the action is Here the coordinates are (x µ , y) for 0 ≤ µ ≤ 3, the bth brane is located at y = y b , g ab is the 5-D metric andg b µν is the induced metric on the bth brane. The brane tensions σ b and potential V are functions of the bulk scalar φ, and κ 2 is the 5-D gravitational coupling constant.
We seek solutions to the field equations of the form although we shall concentrate initially on the static case H = 0. For this ansatz, the Einstein and scalar field equations reduce to [11] where and primes denote derivatives with respect to y. Only two of these three equations are independent: the scalar wave equation follows from the other two via the Bianchi identities.
The jump conditions at the branes are where φ b = φ(y b ) is the value of the bulk scalar on the bth brane and The Ricci scalar is
B. Method of obtaining solutions
We now review the method of generating solutions introduced by DFGK. For any given solution of Eqs. (4), we can imagine inverting the relation φ = φ(y) to obtain y as a function of φ. Leaving aside questions of single-valuedness, we can imagine changing independent variables from y to φ. Since φ ′ (y) is also a function of y, we can likewise take φ ′ to be a function of φ. By analogy with the practice in supergravity theory, let us define a function W (φ) by In the H = 0 case, the first of Eqs. (4) then implies that, away from branes, so that The second of Eqs. (4) then becomes The procedure for obtaining solutions in the H = 0 case is (i) choose a potential V (φ) and superpotential W (φ) that are related by Eq. (13); (ii) integrate Eq. (10) to obtain φ as a function of y; (iii) combine this with Eq. (12) to obtain u as a function of y; and (iv) integrate Eq. (5) to obtain the warp factor A as a function of y.
C. Occurrence of singularities in the bulk
Combining the definition (10) of the superpotential W with the expression (9) for the Ricci scalar gives Eq. (14) is applicable to models with H = 0. When H = 0, substituting Eq. (13) into Eq. yields Eqs. (14) and (15) show why singularities are rampant in models with bulk scalar fields. For many choices of W (φ) and V (φ), it turns out that all of which are simple, and even well-motivated in some respects. Thus, unless φ is constrained to remain finite everywhere, a singularity will be encountered.
To see how to avoid singularities in general [13], consider first a spacetime with a single brane, at y = y b , and suppose that the value of the scalar field on that brane is φ b . From Eq. (10), we can solve for y as a function of φ: where Now if there exist finite values φ −∞ and φ +∞ of the bulk scalar field such that |I(φ b , φ ±∞ )| = ∞, then the spacetime can be infinite in the fifth dimension with φ → φ ±∞ as y → −∞ and y → ∞. In such cases, an infinite range in y is mapped onto a finite range in φ, thus potentially avoiding any singularities. From Eq. (17), divergences of I(φ b , φ) will occur only at zeros of W ′ (φ). Suppose now that W ′ (φ) has a zeros at both φ > φ b and φ < φ b , and that W ′ (φ) vanishes at those zeros at least linearly (but not like |φ − φ ±∞ | p with p < 1). If, in addition, neither W ′ (φ) nor V (φ) diverges in between these zeros, then the spacetime can extend infinitely in the fifth dimension away from the brane in either direction, without any singularities. A similar result can be obtained for spacetimes with more than one brane. In this case, the condition is that I(φ b , φ b ′ ) be finite for any two branes b, b ′ , and that for the two bounding branes, b ± , there exist φ ±∞ such that I(φ b ± , φ ±∞ ) = ±∞ with no intervening infinities in V (φ) and ∂W (φ)/∂φ.
In models where these conditions for avoiding singularities are met, V (φ) tends to a constant asymptotically, as a zero is approached. Thus, such models, for H = 0, tend toward the RS solution far from branes, provided that the scalar field potential is negative. However, there will also be regions in such solutions where the spacetime metric differs considerably from the RS model, and where the warp factor drops far faster than exponentially. We shall show how this comes about in a specific model below. The H = 0 solutions may also tend asymptotically toward their counterparts with uniform bulk cosmological constant [4]. However, the relatively harmless particle horizons characteristic of those uniform bulk cosmological constant models will generically be transformed in to curvature singularities, unless there are fortuitous cancellations, as can be seen from the A → 0 limit of Eq. (8).
Models where singularities would be inevitable in an uncompactified geometry can be truncated by compactification to a region that does not include any singularities. Such models need not approach the RS model, even asymptotically, and can have warp factors that vary much more rapidly than exponentially throughout the compactified fifth dimension.
III. MODELS WITHOUT COSMOLOGICAL EXPANSION
In this section we discuss several different choices of superpotential W (φ) and the properties of the corresponding solutions in the static case H = 0.
A. Even superpotential
The RS model corresponds to Choosing the + sign in W (φ) for y > 0 (and vice-versa), DFGK found φ(y) = φ(0)e −by , and therefore (via Eq. [4]) 3u/κ 2 = −a + bφ 2 (0)e −2by , which implies a transition between two regions of constant u, one at small by and the other at large by, where the model becomes equivalent to the RS model. This model has ∂W (φ)/∂φ = 0 at φ = 0, and is nonsingular as long as the maximum value of |φ| is finite. Thus, it avoids singularities by mapping an infinite range of y to a finite range in φ, as was discussed in Sec. II C. DFGK noted that they could also choose W (φ) = ±(a + bφ 2 ), with b > 0, for which Eqs. (18) and (19) can be rewritten in the form where the ∓ signs apply to Eqs. (18) and (19) respectively, and where Despite appearances, the differences between the two models (18) and (19) are not minor. For the + model, we have W (φ) = a + bφ 2 for y > 0, so that φ ′ = bφ, φ(y) = φ(0)e by and 3u/κ 2 = −a − bφ 2 (0)e 2by . This yields a warp factor A(y) which is an exponential of an exponential: This super-warped version of the GW model, obtained from an apparently insignificant change to the potential (20), becomes singular if the fifth dimension is not compactified, because R(φ) → −∞ as φ → ∞. Any compactification of the fifth dimension avoids the singularity by construction. The model can be altered slightly to become bulletproof against singularities by taking so that φ → φ ∞ and u → −(κ 2 /3)(a + bφ 2 ∞ /2) at large y, reducing asymptotically to the RS model. Note that the total change in the logarithm of the warp factor in the super-warped regime is limited to ∼ κ 2 φ 2 ∞ , which may be large.
B. Odd superpotential and Gaussian warp factor
The DFGK realizations of the GW stabilization mechanism for the RS model are based on choices of superpotential W (φ) that are even functions of φ. There are also simple models in which W (φ) is odd, first considered in Ref [13]; the simplest is with b > 0. (A constant term in W (φ) can be absorbed into the definition of φ.) For the model (24), we find that for y > 0. Thus, the warp factor in this model is Gaussian rather than exponential, and therefore falls off more rapidly than in the RS model. Note that A(y) has a maximum value, and all positive tension branes must be located away from that maximum. This model contains a singularity at y → ∞ since φ(y) diverges there. The singularity can be avoided by compactification, or can be removed altogether by setting W (φ) = ±2b(φ − φ 3 /3φ 2 ∞ ), in which case which tends to φ ∞ asymptotically. The warp factor for this model is given by and tends toward ln A(y) ≃ −4κ 2 bφ ∞ y/9 for by/φ ∞ ≫ 1, which is the same scaling as in the RS model. The maximum change in the logarithm of the warp factor in the Gaussianwarped regime is of order ∼ κ 2 φ 2 ∞ , similar to what we found for the model of Sec. III A above.
C. Exponential potential
The models considered so far become singular only in the asymptotic regime y → ∞. Thus, whether or not these singularities merit a potential-altering cure is largely a matter of taste, since compactifying to any finite size in the fifth dimension removes the singularity without any sort of fine-tuning. However, there are other models which are "spontaneously singular" in the sense that they develop singularities at finite y. An example is when the superpotential is an exponential, For this model, we have which diverges at y = (k 2 a) −1 e kφ(0) , implying a divergence of the Ricci scalar. The singularity is removed by the following slight modification of the superpotential, where q > k is a constant. The associated potential is For this model the relation between φ and y is given by which can be integrated numerically in general. The integral can be done analytically in special cases. For example, for q = 2k, q = 3k, and q = 5k we find the results respectively. For large values of y this model reduces to the RS model, since φ → φ ∞ from above and 3u/κ 2 → −2ae −kφ∞ (1 − k/q), a constant,
Specific realization obtained from dimensional reduction
A potential reminiscent of the potential (31) arises from compactification of an 11-D spacetime to 5-D; the result is Here it is assumed that the six compactified dimensions have a single associated radion field, ψ, the curvature associated with the compactified dimensions is c 6 > 0, Λ b is the descendant of an 11-D cosmological constant, and E is a 5-form field strength (descended from a 7-form field strength in 11-D). Although the potentials (31) and (34) are similar, there are no choices of parameter values for which the potentials match up term by term. This is because the exponents in three terms in Eq. (34) are in the ratios 1 : 1.5 : 3, whereas the exponents in Eq. (31) are in the ratio 1 : (k + q)/(2k) : q/k. However, we can find parameter choices for which the potentials match if we allow one of the two parameters c 6 and Λ b to vanish. If we drop the term proportional to Λ b and retain curvature in the six compactified dimensions, we find two possible identifications: and q = 3k, If we set c 6 = 0 but retain Λ b = 0, there are also two possibilities: and q = 5k, Whichever possibility we choose, we can regard the combination κ 2 a 2 e −2kφ∞ as a derived quantity, determined by either Λ b or c 6 , depending on which is nonzero. The 5-form field strength E, which involves a constant of integration, can then be adjusted to produce a nonsingular model. The 5-form fields therefore may play a central role in eliminating spacetime singularities from 5-D models derived by dimensional reduction from 11-D.
A. Bulk singularities at zeros of the warp factor
When the bulk geometry is anti-deSitter (AdS) space, metrics of the form (3) have nonsingular particle horizons at zeros of the warp factor A(y). From the expression (8) for the Ricci scalar, one might anticipate a singularity from the term ∝ 1/A, but that divergence is cancelled by divergences in the u 2 and u ′ terms for an AdS bulk.
However, when scalar fields are included, the surfaces where A(y) = 0 generically correspond to curvature singularities. This is the case even when we set φ ′ = 1 2 ∂W (φ)/∂φ and choose ∂W (φ)/∂φ so that φ remains finite over the entire range of y. The simplest way to see this is to consider the first of Eqs. (4) in a neighborhood of a point where A(y) = 0. Suppose that φ ′ is finite and nonzero in that neighborhood, so that as long as we restrict attention to a small enough region, we can take it to be a constant. If q 2 ≡ κ 2 (φ ′ ) 2 /3, then This equation would be exact for ∂W (φ)/∂φ = constant, which led us to the Gaussian model in Sec. III. Multiply by u = A ′ /A and integrate to find Here k 2 is a constant which may be positive or negative, although we shall take it to be positive without exception to recover the RS model when q = 0 = H. Substituting these results for u ′ and u 2 into Eq. (8), we find as A → 0. Thus, the Ricci scalar diverges at this point, which is a true singularity, unless q 2 = 0. The divergence is relatively mild (logarithmic). In fact, for H = 0 this is the same divergence as was found asymptotically in the Gaussian model of Sec. III.
B. Method of constructing singularity-free solutions
In the singularity-free models with H = 0, we saw that obtaining a mapping of the infinite y domain to a finite range of φ required that φ ′ → 0 asymptotically. For H = 0 models the singularity is absent, because A → 0 only asymptotically, and q 2 ln A → 0 asymptotically as well. For H = 0, nonsingular models can be constructed parametrically by the following procedure. Define a function g(A) by with g(0) = 0. Then Eq. (40) generalizes to Comparing this with the second of Eqs. (4) we obtain The procedure for obtaining a solution can now be summarized as follows: (i) pick a function g(A) and the integration constant k 2 ; (ii) solve Eq. (43) to obtain A as a function of y; (iii) solve Eq. (42) to obtain φ as a function of y; (iv) use Eq. (44) to compute the potential V (φ) corresponding to the solution just obtained. By construction, the Ricci scalar is finite at A → 0. Let us illustrate this procedure with the particular example with p > 0. (For small p, we expect this model to resemble the singular Gaussian model.) For the choice (46) we find Apparently, there is a maximum value of A(y) for such models; for small H 2 /k 2 we find A max ≃ (k/q) 2/p + H 2 /pk 2 . Note that in this particular model, freedom to rescale A(y) also implies that q may be rescaled arbitrarily; this need not be true for g(A) that are not scale-free. When H = 0, the solution to Eq. (47) is (apart from an overall ambiguity in the sign of φ ′ and hence φ) where we have located the maximum of A(y) at y = 0 and chosen φ = 0 at y = 0. As y ranges from zero to infinity, φ κ 2 p/3 ranges from 0 to π/2. Note that the maximum value of V (φ) is positive in this model, and occurs at the maximum of A(y); this is already apparent from Eqs. (43) and (44). In a particular realization of this model, with branes at specific values of φ, the maximum value of the warp factor may never be encountered. Finally, notice that for small values of p, the warp factor A(y) ∼ e −pk 2 y 2 for pky < ∼ 1, and, for any value of p, A(y) ∼ e −ky for pky > ∼ 1. These features are reminiscent of the modified nonsingular Gaussian model of Sec. III, and the potential (50) therefore represents an alternative way of regularizing the Gaussian model to avoid singularities. When H = 0, Eq. (48) shows that A(y) → 0 only when y → ∞. When H = 0, Eq. (47) shows that |A ′ /A| is larger than for H = 0, and we therefore expect the position of the zero of A(y) to move inward from y = ∞ to finite y. If y 0 is the zero, so that A(y 0 ) = 0, then A(y) may be found by inverting the equation and the scalar field may be found from Here φ(0) is the value at A = 0, and we have assumed φ ′ > 0. Eqs. (52) and (44) may be used to evaluate the potential V (φ), which will have a nontrivial dependence on H. For A < ∼ H 2/(1+p) , the scalar field plays an insignificant role, and we recover the vacuum solution , on the other hand, we can neglect H 2 , and we recover the solution (48) for A(y). Corrections to these approximate solutions may be obtained from Eq. (51). In general, the behavior of models with scalar fields is subtle for small values of H 2 , because nonsingular models require that the effects of the scalar field become insignificant as A(y) → 0, so that the H 2 terms become important in these regions despite the smallness of H 2 .
These nonsingular models appear to share common features with models developed previously without bulk scalar fields [4], particularly the occurrence of non-singular surfaces where A(y) = 0. Presumably, these surfaces are particle horizons similar to those which occur in the pure AdS case.
V. COSMOLOGICAL CONSTANT AND HIERARCHY PROBLEMS
To examine the implications of these models for the value of the cosmological constant, focus again on the model (46) and rewrite Eq. (43) as First, let us consider the S 1 /Z 2 orbifolded model, with a brane at each end, where the warp factors at these branes are A i , i = 1, 2, and the brane tensions are σ i . Note that it is possible for both brane tensions to be positive, if one of them is located at y > y 0 and the other at y < y 0 . We define the quantities Q 1 and Q 2 by Using the jump conditions (6) along with Eq. (53), where the constant term in V (φ) is adjustable (for example, via unimodular gravity), we find that In the regime A 2 ≪ A 1 , these formulae reduce to where we have assumed that Q 2 2 ∼ Q 2 1 . Note that Eq. (55) requires that Q 2 2 > Q 2 1 . The result (56) implies that the four dimensional cosmological constant Λ 4 ∝ H 2 /A 1 is suppressed by the ratio A 2 /A 1 of warp factors, which will be vanishingly small for suitable brane separations. Now suppose that our Universe lives on a test brane located somewhere between the branes 1 and 2. Denote by A U the warp factor on our brane, and let brane 1 be the Planck brane. Then, we find that the electroweak scale on our brane is given by m 2 EW = m 2 P A U /A 1 , where m P is the Planck scale. The expansion rate on our brane is H 2 U = H 2 /A U , and therefore Assuming that Q 2 i ∼ m 2 P , we see that H 2 U /m 2 EW can be exponentially small provided that A U ≫ √ A 1 A 2 . Roughly speaking, this will be the case if the test brane on which our Universe resides is closer to the Planck brane, brane 1, than it is to brane 2. Turning on a small positive brane tension for the visible brane does not change the above qualitative features. Even for models in which singularities are not absolutely prevented, they can still be avoided in the orbifold case, which requires a negative tension brane at one fixed point. If one only wants to solve the mass hierarchy problem, this negative tension brane is the visible brane [1]. If one also wants to solve the cosmological constant problem, the visible brane must be identified with a third brane between the orbifold boundaries, as in the nonsingular models discussed above.
This result can be generalized to other H = 0 spacetimes with no bulk curvature singularities, and remains valid for uncompactified, multi-brane models as well. It can also be generalized to cases where the branes are charged under some 4-form potentials [4]. In this case, the constant term (and/or other coefficients) in V (φ) depends on the 5-form field strengths, and the background values of these field strengths are determined by satisfying the jump conditions at the branes. With the warp factor normalized to unity at the Planck brane, Λ 4 is proportional to A p , the warp factor at the nearest particle horizon. This factor can easily be exponentially small. Since the visible brane must be closer to the Planck brane than the particle horizon, the warp factor A v at the visible brane can easily be exponentially small as well, providing a solution to the hierarchy problem [4], but A v >> A p , as observed in nature.
VI. DISCUSSION
In this paper, we have constructed several different models of 5-D brane worlds with bulk scalar fields. We gave a general procedure for constructing static (H = 0) models with no curvature singularities in the bulk. A specific model was constructed corresponding to compactifying from an 11-D theory with a 7-form field strength down to 5-D, where the 7-form field descends to a 5-form field strength. By adjusting the 5-form field strength, we were able to render the model nonsingular. In general, we found that nonsingular models are more warped than the RS model on relatively small distance scales, but tend towards simple exponential warping on large scales. Thus, the RS behavior is robust in the asymptotic regime, although stabilization of branes via the introduction of bulk scalar fields may require the branes to reside where the fields produce substantial extra warping of spacetime.
For expanding (H = 0) models, we found that singularities occur even for those scalar field potentials that do not give rise to singularities in the H = 0 case. We therefore were compelled to explore a different method for finding nonsingular models with H = 0. A subclass of these models incorporates a non-singular surface on which the warp factor A(y) vanishes, analogous to the particle horizons that occur in the pure AdS case [4]. One specific class of such models, which corresponds to a sinusoidally varying potential in the nonexpanding case, was treated in some detail. For this case we constructed an orbifolded model, with one bounding brane being the Planck brane, and the other bounding brane (on which the warp factor is exponentially smaller) having either positive or negative tension. In this model, we live on a test brane between the two boundaries branes, and the expansion rate on our brane is exponentially smaller than the local electroweak scale provided that our brane is located somewhat closer to the Planck brane than to the other brane. This feature does not appear to depend in an essential way on details of our particular model, and even can be shown to hold when singularities are avoided (by orbifolding) rather than prevented outright by choice of potential. Thus, there are models in which the cosmological constant and mass hierarchy problems can be explained simultaneously.
Two fine-tunings of the parameters of the RS orbifold model are required to get flat 4-D spacetime [1]. Allowing non-zero Λ 4 relaxes one of these fine-tunings, while the other can be eliminated by allowing the bulk cosmological constant to depend on 5-form field strengths, whose piecewise constant values are determined by boundary conditions at the branes. When the branes are relatively far apart, Λ 4 turns out to be exponentially small [4]. However, the models of Ref. [4] were dynamically unstable. Bulk scalar fields can be introduced to fix brane positions. When Λ 4 = 0, there are two sources of fine tuning in models with bulk scalar fields, one associated with expressing the potential V (φ) in terms of a superpotential, W (φ) [Eqs. (10) and (13)], and the other associated with adjusting W (φ) so that the solution becomes nonsingular. Given a potential V (φ), there is no guarantee that a W (φ) can be found that satisfies Eq. (13) without tuning the parameters [11]. For example, a general quartic potential of the form V (φ) = a 0 +a 2 φ 2 /2+a 4 φ 4 /4 contains three parameters, whereas Eq. (18) or (19) contains only two, implying that a relationship among a 0 , a 2 and a 4 is needed in order for V (φ) to arise from a superpotential W (φ). Moreover, Eq. (18) corresponds to a nonsingular model, whereas Eq. (19) is singular; the two independent parameters in a quartic potential have to be specifically adjusted to prevent singularities. The adjustment can be achieved by the introduction of adjustable 5-form field strengths, and we may speculate that, in an evolving bulk, the 5-form field strengths may relax naturally (e.g. via bubble nucleation) to values that prevent the occurrence of singularities. (A specific example of the role that 5-form field strengths might play is given in Eqs. (35)-(38) of Sec. III C 1.) In general, additional fine-tunings required to prevent singularities need not be invoked in an orbifold model, where compactification may simply avoid the presence of singularities that could otherwise arise in an uncompactified model.
When Λ 4 = 0, some fine tuning is still needed, apparently, for singularities to be absent, but we had to resort to a different method, not based on a superpotential, to construct nonsingular solutions in this case. The problem arises because singularities can only be prevented if φ ′ → 0 when A(y) → 0, a condition that does not follow readily from a description in terms of a superpotential, and therefore represents a different kind of tuning. This suggests that the bulk alone may be supersymmetric, making a description in terms of a superpotential W (φ) natural, but that the presence of branes breaks the supersymmetry completely, and leads to nonzero Λ 4 . As in the nonexpanding case, it is also possible to preserve models based on a superpotential without any special tuning of the parameters in W (φ) in a appropriately compactified or orbifolded model, but it is never possible to relate V (φ) directly to W (φ) when Λ 4 = 0.
We thank Csaba Csaki and Zurab Kakushadze for pointing out earlier works that are relevant to this paper. This research was supported in part by NSF grant PHY-9722189 (E.E. | 2014-10-01T00:00:00.000Z | 2001-10-08T00:00:00.000 | {
"year": 2001,
"sha1": "ba6ae848d6cc36393c081c89f539db6bc4e87213",
"oa_license": "CCBY",
"oa_url": "https://doi.org/10.1016/s0370-2693(01)01261-8",
"oa_status": "HYBRID",
"pdf_src": "Arxiv",
"pdf_hash": "98c91893dc3c4a3ca3dc8949514f538a0d2c8fee",
"s2fieldsofstudy": [
"Physics"
],
"extfieldsofstudy": [
"Physics"
]
} |
89724848 | pes2o/s2orc | v3-fos-license | In vitro anti-inflammatory activities of blacklip abalone (Haliotis rubra) in RAW 264.7 macrophages
ABSTRACT Abalone (Haliotis sp) has been used as a traditional functional food for many years by different cultures believing that consumption provides health benefits. We investigated the anti-inflammatory activity from blacklip abalone visceral waste. An extract was prepared from the viscera using digestion with the food grade proteases papain and bromelain followed by anion exchange chromatography (AEC). The anti-inflammatory potential of the extract and AEC fractions were investigated in comparison to a plant-derived positive control (quercetin) through the inhibition of lipopolysaccharide-stimulated nitric oxide production in RAW 264.7 monocytes. Significant anti-inflammatory activity was observed in response to the abalone extract and the unbound AEC material over a concentration (based on collagen) range of 12–96 µg/mL. Overall, results indicated that blacklip abalone extract has anti-inflammatory activity in vitro, warranting further investigation into the bioactive constituents that may be potential anti-inflammatory therapeutics.
Introduction
Inflammation is a protective mechanism in response to disease, physical trauma, noxious stimuli by chemical agents, heat, antigen-antibody reactions and microbial effects (Cheng et al., 2015). Inflammation is a complicated process engaging multifactorial networks of chemical-based signals helping the body to eliminate or limit the spread of injurious agents (Soni et al., 2014). Macrophages play a significant role in inflammation by producing nitric oxide (NO) which mediates many physiological functions, such as nonspecific host defences, anti-microbial defences and anti-tumor activities, as well as pathological processes including organ destruction in some inflammatory and autoimmune diseases (Turini & DuBois, 2002). NO is generated biochemically via oxidation of the terminal guanidine nitrogen of L-arginine by nitric oxide synthase. The inhibition of NO overproduction by blocking inducible nitric oxide synthase (iNOS) expression may be a useful strategy in the treatment of various inflammatory disorders (Qian et al., 2012). Overproduction of these factors leads to cell damage and inflammatory disease (Quang et al., 2014); therefore, inhibition of these inflammatory mediators is one of the important strategies in treating inflammatory diseases (Yoon et al., 2012).
Non-steroidal anti-inflammatory drugs (NSAIDs) are commonly used to treat inflammation and associated pain even though long-term usage of NSAIDs can lead to adverse gastrointestinal outcomes such as peptic ulcers (Takeuchi, 2012). Consequently, there is considerable interest in safer drugs (Prabhu, Nalini, Chidambaranathan, & Kisan, 2011) with alternative medicines from plant and marine sources gaining popularity as therapeutics due to their mild action and fewer side effects (Sun et al., 2016). The discovery of novel bioactive peptides from marine origins with specific cellular targets, may compliment the search for promising drug candidates. During the last decade, hundreds of marine-based bioactive molecules have been discovered with several undergoing clinical trials (Mayer, Rodríguez, Taglialatela-Scafati, & Fusetani, 2013;Suleria, Masci, & Globe, 2015). These bioactive molecules show anti-bacterial, anti-fungal, anti-cancer, anti-viral, as well as anti-inflammatory activities that can be used to promote the health status of humankind (Dang, Benkendorff, Green, & Speck, 2015;Sugiura, Nagayama, Kinoshita, Tanaka, & Matsushita, 2015), however, the biological activity of most of the marine compounds is still unclear and under investigation (Blunt et al., 2009).
Blacklip abalone (Haliotis rubra), a single-shelled marine mollusc, is harvested commercially in many international waters and widely cultured in eastern Asia. During the processing of commercial products, the viscera, considered inedible, are discarded and account for approximately 30% of abalone weight. Extracts have been prepared from abalone viscera with anti-inflammatory activity demonstrated in vitro and in vivo (Sun et al., 2010). Anti-inflammatory properties of water extracts and fermented hydrolysate of Haliotis discus hannai Ino have also been reported by Hasnat et al. (2015) using lipopolysaccharide (LPS)-stimulated RAW 264.7 macrophages. In this study, H. discus hannai Ino water extracts and fermented hydrolysate decreased NO production in LPS-stimulated RAW 264.7 macrophages in a dose-dependent manner at concentrations of 1 to 8 mg/mL protein with no detrimental effect on cell viability. Cheong et al. (2015) also reported the inhibitory effect of H. discus water extracts on LPS-induced NO production in zebrafish by using a fluorescent dye probe. Qian et al. (2012) also reported the anti-inflammatory effects of H. discus hannai intestinal digest in RAW 264.7 macrophage cells. According to their findings, the abalone intestinal digest suppressed LPS-induced production of NO via reduced iNOS expression in a dose-dependent manner.
Drug discovery for the treatment of inflammation often focuses on the decreased production of the pro-inflammatory mediator NO (Qian et al., 2012) as well as iNOS and the regulation of specific cytokines like tumor necrosis factor α (TNF-α) and interleukins 1β (IL-1β) and 6 (IL-6), and transcription factors such as nuclear factor κB (NF-κB) (Sun et al., 2010). In the current research, an extract was prepared from blacklip abalone viscera and fractionated prior to in vitro assessment of antiinflammatory activity. Cell viability and NO production was measured following treatment of LPS-induced RAW 264.7 macrophages with blacklip visceral extract and fractions.
Preparation of extracts
Extracts were prepared using 5.0% w/w food grade proteases papain and bromelain (Enzyme Solutions, Melbourne, VIC, Australia), combined in 1:1 ratio. For the purpose of determining the contribution of the two proteases to the subsequent assay measurements, enzyme-only control digests were prepared using the same concentrations and conditions but with no added abalone. Digests were incubated overnight (14-16 h) at 50°C, inactivated by heating at 95°C for 10 min, cooled on ice and centrifuged (Avanti® J-26XP1; Beckman Coulter, Brea, CA, USA) at 5940g for 10 min to remove undigested material (pellet). Supernatants were clarified using sequential filtration with 2 µm, 1 µm (Whatman™, GE Healthcare Life Science, Parramatta, NSW, Australia) and 0.45 µm (mixed cellulose ester, Merck Millipore, Bayswater VIC, Australia) filter paper and stored at -20°C. Due to likely interference with bioassays, salt ions were reduced in the abalone hydrolysate and pooled fractions using 3 kDa molecular weight cut-off (MWCO) spin columns (Centrifuge Filter Unit, Merck Millipore, Billerica, MA, USA). Samples (10 mL) were added into the spin column and centrifuged at 3270g for 30 min. This process was repeated using deionized water until salt was at or below 60 mM as determined by conductivity measurements (Metler Toledo-AG conductivity meter, VWR International, Dietikon, Switzerland) using a NaCl standard curve.
Estimation of protein content
Protein content was estimated in all samples and extracts using the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific, Waltham, MA, USA) with bovine serum albumin as a protein standard, according to manufacturer's instructions. All assays were performed in triplicate. Absorbance was measured at 562 nm using a Spectra-Max M3 System spectrophotometer (Molecular Devices, Sunnyvale, CA, USA) and concentrations determined using the associated instrument software (SoftMax-Pro 6.1).
Estimation of total phenolic contents
The total phenolic content (TPC) in the abalone extract was determined by using the Folin-Ciocalteu method described by Chang et al. (2002). Fifty microliters abalone extract or fractionated samples, sterile water (as the blank) or a standard solution of gallic acid (6.25, 12.5, 25, 50, 100 µg/mL in distilled water) was added to 50 µL 10% Follin-Cicocalteu's phenol reagent and 50 µL 1 M sodium carbonate solution in triplicate, in a 96-well plate (Nunclon Delta Surface, Thermo Fisher Scientific, Waltham, MA, USA). The plate was incubated for 60 min at room temperature in the dark. The absorbance was measured at 750 nm using a Spectra-Max M3 System spectrophotometer (Molecular Devices, Sunnyvale, CA, USA). TPC in the abalone extract and fractionated samples was expressed as mg gallic acid equivalents per gram abalone extract/ fractionated samples.
Estimation of collagen
Total collagen content (based on hydroxyproline detection) was estimated in the extracts using the QuickZyme Biosciences Total Collagen Assay Kit (BioScientific Pty Ltd) according to the manufacturer's instructions with modifications. Briefly, 125 µL standard (rat tail collagen) or samples were added to 125 µL 12 M HCl and hydrolyzed for 20 h at 95°C in screw capped tubes. Following the hydrolysis, 35 µL hydrolyzed samples and standards were transferred to a 96-well plate provided for the assay, mixed with 75 µL assay buffer and incubated at room temperature (∼21°C ) for 20 min. Seventy-five microliters of detection reagent was then added to each well and the resulting solution mixed and incubated at 60°C for 30 min. Absorbance was read at 570 nm and a linear standard curve constructed to interpolate the total collagen concentration. All assays were performed in duplicate.
2.4. Separation of blacklip abalone extracts using anion exchange chromatography via fast protein liquid chromatography Q Sepharose™ Big Beads (372.5 mL) was packed into an empty GE-XK 26/70 column (700 mm × 26 mm, GE Healthcare Life Science, Chicago, IL, USA). The packed column was attached to a fast protein liquid chromatography (FPLC) system (ÄKTA Lab-Scale Systems, GE Healthcare Life Sciences, Chicago, IL, USA). The column was equilibrated with deionized water (Buffer A) before a sample of approximately 800-900 mg GAG (sulphated polysaccharide basis) was loaded onto the column. Flow rate was set at 5 mL/min with a column pressure of 1 MPa. After isocratic washes of buffer A over 660 minutes (3300 mL), 15 and 50 mL fractions were collected on commencement of a 0-2 M NaCl linear gradient delivered over 330 min (1650 mL). Elution was monitored at 280 nm for protein content (Talaei et al., 2016), and 206 nm for glycosaminoglycan detection (Marks et al., 2001). Conductivity was also monitored throughout the run. The collected fractions were pooled on the basis of their interactions with dimethylmethylene blue (DMMB) dye. Overall, five anion exchange chromatography (AEC) pools were prepared from the linear NaCl gradient (numbered 1-5) along with the unbound material, initial column wash and final column wash. For further analysis, all pooled samples were desalted using 3 kDa MWCO spin columns and washed using deionized water to reduce salt concentration to 60 mM or less.
Cell culture studies
2.5.1. Culture of RAW 264.7 cells RAW 264.7 cells were cultured in growth media comprised of RPMI-1640 supplemented with 100 IU/mL penicillin, 100 µg/mL streptomycin and 10% FBS. Cells were grown in a humidified atmosphere with 5% CO 2 at 37°C. A mycoplasma test was performed routinely and there was no bacterial contamination.
2.5.2. NO assay NO production, measured as nitrite, was determined using the Griess Reagent. RAW 264.7 cells were seeded into 96-well plates at a density of 50,000 cells/well (in 100 μL) along with 20 µL 300 ng/mL LPS or 20 µL Quercetin (0.3-100 μM), with or without abalone extract or abalone AEC pools (in triplicate applications of 12,24,48,96,192 and 375 μg/mL on a collagen basis). The cells and various treatments were incubated at 37°C and 5% CO 2 for 48 h.
After 48 h, 50 µL cell media was removed from each well and transferred into a new 96well plate. Fifty microliters of various concentrations of sodium nitrite standard were also added in either duplicate or triplicate to the 96-well plate in preparation for the standard curve. Finally, 50 µL of Griess Reagent were added to each well for 15 min and absorbance measured at 540 nm. The nitrite concentration of each cell treatment was extrapolated from the sodium nitrite standard curve.
Cell viability
Cell viability was measured using the CellTiter 96® AQueous Non-Radioactive Cell Proliferation Assay for determining the number of viable cells. Following cell treatments in 96-well plates, 20 μL of MTS solution were added to 140 µL of media and incubated with the cells for 120 min at 37°C. Absorbance was measured using a Spectra-Max M3 System spectrophotometer at 492 nm.
Statistical analyses
All statistical analyses were conducted using a one-way analysis of variance with Dunnett's comparison tests or unpaired t-tests. These calculations were carried out using GraphPad Prism 5 Software for Windows (GraphPad Software, San Diego CA, www.graphpad.com). Significance was observed at p < .05.
Preparation of abalone extract and fractionation via AEC
Marine processing waste, often discarded due to limited know-how, market constraints and technological barriers, is a good source of bioactive molecules (Hayes & Tiwari, 2015). One of the biggest challenges and most important research areas in this field is the isolation and purification of these bioactive molecules for the development of LPS and 375,196,9,48,24 and 12 µg/mL abalone extract or AEC pools, was measured in triplicate (n = 3) using CellTiter 96 ® AQ ueous Non-Radioactive Cell Proliferation Assay. Cell viability is expressed as the mean percentage viable cells (±standard deviation) compared to LPS control. Statistical significance determined using a one-way analysis of variance with Dunnett's Multiple Comparison Test compared to LPS control with *p < 0.05. Figure 3. Screening of anti-inflammatory activity in RAW 264.7 cells by abalone extract and AEC pools. Anti-inflammatory activity in RAW 264.7 cells, following 48-h treatment with 300 ng/mL LPS and 375,196,9,48,24 and 12 µg/mL abalone extract or AEC pools, was measured in triplicate (n = 3) and expressed as mean percentage decrease in nitric oxide production (assayed as nitrite) ± standard deviation compared to LPS treatment control. 718 various therapeutic products (Orgueira et al., 2003). In the current research, an extract was prepared from blacklip abalone viscera using papain and bromelain enzymes and subjected to anion exchange chromatography. Total protein, collagen and phenolic contents were estimated in the extract and all AEC pools. In Figure 1, total protein, collagen and phenolic contents were higher in the extract and AEC-unbound material as compared to other AEC pools. Therefore, it appears as though the majority of peptides in the extract are cations, carry no charge or are weak anions.
Effect of abalone extract and AEC pools on inflammation and cell viability in vitro
When cells are activated with pathogenic substances, macrophages initiate and regulate inflammatory responses through a broad range of inflammatory mediators. LPS are among the most effective macrophage activators, and it has been known that LPS-stimulated macrophages produce excessive inflammatory mediators that have been in close relation with elevated anti-inflammatory response (Meng and Lowell, 1997). Therefore, the anti-inflammatory effect of the abalone extract and the AEC pools was investigated using LPS-stimulated RAW 264.7 cells. Inflammation was indicated by the cellular production of NO as measured using the Griess assay. A decrease in the production of NO by the positive control (Quercetin) or by cell treatment with abalone extracts or AEC pools, indicated anti-inflammatory activity. To ensure that the decrease in NO production was not associated with cell toxicity or cell death, RAW 264.7 cell viability was measured in response to all treatments using a screening MTS assay.
In Figure 2, cell viability results show that the extract and some AEC pools displayed some cytotoxic effects over the treatment concentration range 12-375 µg/mL (on a Figure 4. Cell viability in RAW 264.7 macrophages treated with LPS and abalone extract, unbound anion exchanged material or quercetin. RAW 264.7 cell viability, following 48-h treatment with 300 ng/mL LPS and 375,196,9,48,24 and 12 µg/mL abalone extract or AEC-unbound material, or 100, 30, 10, 3, 1 and 0.3 µM quercetin, was measured in three independent experiments in triplicate (n = 9) using CellTiter 96 ® AQ ueous Non-Radioactive Cell Proliferation Assay. Cell viability is expressed as the mean percentage viable cells (±standard deviation) compared to LPS control. Statistical significance determined using a one-way analysis of variance with Dunnett's Multiple Comparison Test compared to LPS control with *p < 0.05 decreased cell viability, and #p < 0.05 increased cell viability. Figure 5. Anti-inflammatory activity by abalone extract and AEC pools in RAW 264.7 macrophages. Anti-inflammatory activity in RAW 264.7 cells, following 48-h treatment with 300 ng/mL LPS and 375,196,9,48,24 and 12 µg/mL abalone extract or AEC-unbound material, was measured in three independent experiments in triplicate (n = 9) and expressed as mean percentage decrease in NO production (assayed as nitrite) ± standard deviation compared to LPS treatment control. Statistical significance determined for extract and unbound material using unpaired t-tests with 100 µM quercetin (*p < 0.05) and 30 µM quercetin (#p < 0.05).
collagen basis), when compared to LPS control cell viability. Cell viability decreased the most (over 90%) in response to 375 µg/mL AEC pool 5. The extract also decreased cell viability by approximately 35 and 30% in response to 375 and 195 µg/mL treatments; cell viability also decreased 10-20% in response to AEC pools 1 (12 and 24 µg/mL) and 2 (12-96 µg/mL). In Figure 3, the extract and AEC-unbound material suppressed LPSinduced production of NO in a dose-dependent manner. Anti-inflammatory activity without cell toxicity was observed in response to 12-96 µg/mL abalone extract and 96-375 µg/mL AEC-Unbound material.
To further investigate the anti-inflammatory effect of the extract and AEC-unbound material and related cytotoxicity, three additional in vitro experiments were performed in comparison to quercetin. Figure 4 shows the mostly cytotoxic effect of the extract (at 375, 96, 24 and 12 µg/mL) on cell viability whilst the 96,48 and 24 µg/mL) and quercetin (at 100, 30, 10, 3 and 1 µM) significantly increased or did not change the percentage of viable cells compared to the LPS control. The anti-inflammatory effects of these treatments are presented in Figure 5 and show that 100 and 30 µM quercetin and 375, 195 and 96 µg/mL abalone extract decreased nitric oxide production in a similar manner. However, in the absence of decreased cell viability, the AEC-unbound produced an anti-inflammatory response similar to 30 µM quercetin offering the most promising source of anti-inflammatory abalone molecules without cytotoxicity.
A previous study has also confirmed that abalone extracts have nitrite radical scavenging activities. In a study by Hasnat et al. (2015), an abalone extract showed dosedependent nitrite scavenging activities ranging from 25.14-36.28% at concentrations of 1-8 mg/mL w/v computing with α-tocopherol of 100 μg/mL. In a study by Rho et al. (2015), involving Fourier transform infrared spectroscopy (FTIR) analysis of abalone extract, glycoprotein and polysaccharide-rich fractions were consistent with anti-inflammatory and anti-oxidant bioactivities. Furthermore, abalone have been found to contain high contents of polyphenols and gamma-aminobutyric acid that is dependent upon how much seaweed abalone consume (Dang, Li, Speck, & Benkendorff, 2011;Stott, Takeuchi, Koike, Imada, 2003).
Conclusion
To the best of our knowledge, there is no published research addressing the antiinflammatory effect of blacklip abalone viscera. In this in vitro study, NO production was inhibited by molecules separated from blacklip abalone visceral extract, in particular the AEC-unbound material at a concentration range of 48-375 µg/mL (as measured on a collagen basis). Further strategies must be considered to better separate blacklip abalone visceral extract and improve the cellular responses to anti-inflammatory molecules present. Further research is also needed to investigate other anti-inflammatory markers, like cytokine secretion, that will help to elucidate the mechanism. experiments and analysis. Hafiz Ansar Rasul Suleria and Simone A. Osborne wrote the manuscript. Glenda Gobe supervised and provided advice to Hafiz Ansar Rasul Suleria, contributed reagents and edited the manuscript.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes on contributors
Dr Hafiz Ansar Rasul Suleria is currently working as an Honorary Fellow with the UQ Diamantina Institute, Translational Research Institute, the University of Queensland, Australia. He obtained his PhD in collaboration with the University of Queensland and CSIRO on an International Postgraduate Research Scholarship and Australian Postgraduate Award funded by the Australian Government.
Ms Rama Addepalli has over 25 years of experience in cell based and other biochemical assays for bioactive discovery. She has a BSc in Microbiology.
Dr Paul P. Masci is a research scientist with The University of Queensland, Faculty of Medicine, Translational Research Institute, Princess Alexandra Hospital, Brisbane, Australia. He has over 46 years of experience in Haemotology, Venomics, Biotechnology and Molecular Technology related to novel discovery. He has a BSc, MSc and a PhD in Haemotloogy, Venomics Biotechnology and Molecular Biotechnology.
Glenda Gobe is at the University of Queensland Diamantina Institute, Associate Professor Gobe heads the Kidney Disease Research Theme and is Director of Research Training. She is curator of the Chronic Kidney Disease -Queensland Biobank. She has a BSc in Biology and Chemistry; MSc in Physiology and Pharmacology; and PhD in Pathology. She is known internationally for expertise in molecular and cellular characteristics and determinants of apoptosis in disease.
Dr Simone A. Osborne is a research scientist with CSIRO and has over 12 years of experience in biotechnology and food science related to bioactive discovery and value adding. She has a BSc with first-class honours in Biotechnology and a PhD in Cellular and Molecular Biology. | 2019-04-02T13:06:47.950Z | 2017-04-10T00:00:00.000 | {
"year": 2017,
"sha1": "e39375d0f3488f2866fe8c654f1e122beee6e1e3",
"oa_license": "CCBYNC",
"oa_url": "https://www.tandfonline.com/doi/pdf/10.1080/09540105.2017.1310186?needAccess=true",
"oa_status": "GOLD",
"pdf_src": "TaylorAndFrancis",
"pdf_hash": "e1fa037d97de329aee36b45ce95182d32b9180d6",
"s2fieldsofstudy": [
"Biology"
],
"extfieldsofstudy": [
"Biology"
]
} |
9092809 | pes2o/s2orc | v3-fos-license | Accuracy of Self-reported Hypertension, Diabetes, and Hypercholesterolemia: Analysis of a Representative Sample of Korean Older Adults
Objectives This study will assess the accuracy of self-reported hypertension, diabetes, and hypercholesterolemia among Korean older adults. Methods Using data from the fourth Korean National Health Examination and Nutrition Survey (KNHANES IV, 2007–2009), we selected 7,270 individuals aged 50 years and older who participated in both a health examination and a health interview survey. Self-reported prevalence of hypertension (HTN), diabetes mellitus (DM), and hypercholesterolemia was compared with measured data (arterial systolic/diastolic blood pressure, fasting glucose, and total cholesterol). Results An agreement between self-reported and measured data was only moderate for hypercholesterolemia (κ, 0.48), even though it was high for HTN (κ, 0.72) and DM (κ, 0. 82). Sensitivity was low in hypercholesterolemia (46.7%), but high in HTN and DM (73% and 79.3%, respectively). Multiple analysis shows that predictors for sensitivity differed by disease. People with less education were more likely to exhibit lower sensitivity to HTN and hypercholesterolemia, and people living in rural areas were less sensitive to DM and hypercholesterolemia. Conclusion Caution is needed in interpreting the results of community studies using self-reported data on chronic diseases, especially hypercholesterolemia, among adults aged 50 years and older.
Introduction
The accuracy of self-reported cardiovascular disease (CVD) and its determinant factors are a pivotal issue for global public health in CVD prevention and management among older populations. CVD continues to be a leading cause of morbidity and mortality in most developed and developing countries [1e3]. In accompaniment with a rapidly growing elderly population concomitant with increasing life expectancy, South Korea (hereafter Korea) has seen a dramatically increasing trend in the risk of CVD. The prevalence of CVD risk factors such as of hypertension (HTN), diabetes mellitus (DM), and hypercholesterolemia among adults (30 years or older) has increased from 24.6%, 9.6%, and 10.7% in 2007, to 27.3%, 11.0%, and 14.9%, in 2013, respectively [4]. To prevent and manage the increasing health burden of CVD risk factors, the Korean government implemented the National Cardio-Cerebrovascular Disease Plan 2010e2015 [5], but patients with these chronic diseases are frequently excluded from the government's enrollment system, due to limited public knowledge and awareness of CVD risk factors [6].
Accurately assessing the prevalence and trend of these diseases is a prerequisite for societal disease management and medication compliance. While, in reality, self-reported health data through surveys are widely used due to cost-effectiveness, it is noted that complex variations between subjective and objective measures of CVD risk factors hamper us from a better understanding of the magnitude of CVD, its associated factors, and the effectiveness of government interventions for its prevention [7e10]. A handful of studies suggest that the prevalence of self-reported DM shows a relatively high level of agreement [11e14], but high cholesterol revealed a significant discrepancy [9e11,15], with a mixed result for high blood pressure. A Minnesota study using 2,037 participants aged 45 years or older suggests that agreement between reported and medical records was noticeable for both DM and HTN (k 0.71e0.80) [12]. In comparison to the American research, several European findings suggest similar agreement for DM (k 0.84e0.76), but a significantly lower agreement for HTN (k 0.63e0.51) and hypercholesterolemia (k 0.55e0.48) [9,11,13]. A recent European study of 12 countries estimated that nearly 70% of European adults were unaware of having high cholesterol levels [11]. These results imply that undiagnosed cases of chronic patients could significantly deteriorate the quality of CVD primary care and intervention, when relying only on self-reported health data. However, there is a substantial knowledge gap in the accuracy of self-reported chronic diseases among Asian older adults. In addition, it is still unclear why this discrepancy in agreement exists for CVD risk factors, while emerging research shows that some sociodemographic factors such as age, sex, and education can contribute to its accuracy [9,16e19].
The central objective of this epidemiological research is to investigate the accuracy of self-reported HTN, DM, and hypercholesterolemia among Korean older adults. The specific goals of the study are to assess: (1) whether Korean older adults have a higher accuracy of reporting CVD risk factors when compared with measured data; (2) whether the extent to which the observed variation between the two measures differs in HTN, DM, and hypercholesterolemia; and (3) whether the extent may be attributable to demographic factors, socioeconomic factors, and/or health behavioral factors. The study will also examine whether there are different determinant factors between DM, HTN, and hyperlipidemia. To achieve our aims, we will use the Fourth Korean National Health and Nutritional Examination Survey (KNHANES IV), 2007e2009, the representative national data [20].
Data and study population
This study is based on data from KNHANES, conducted from 2007 to 2009. This nationally representative cross-sectional survey included health interviews, health examinations, and nutritional surveys, intended to monitor the health and nutrition status of the Korean population. The details of this survey have been published in several publications [20]. Out of 24,871 individuals for whom both reported and measured data (health interviews and health examination) were available, 8,529 (34.3%) were over the age of 50 years. We obtained 7,270 observations for the final analysis, after eliminating all missing data.
CVD risk factors
CVD risk factors included HTN, DM, and hypercholesterolemia. Reported HTN, DM, and hypercholesterolemia cases were ascertained with the following questions. Has a doctor ever told you that you have: (1) high blood pressure or HTN? (2) hyperglycemia or DM? or (3) high blood cholesterol or hypercholesterolemia? Regarding the measured CVD risk factors, HTN was defined as a systolic blood pressure 140 mmHg, or diastolic blood pressure 90 mmHg, or currently using medication due to high blood pressure (Joint National Committee-6 definition) [21]. In the case of DM, individuals with fasting blood glucose levels of 126 mg/ mL or those under treatment are categorized as DM [22]. Hypercholesterolemia is considered as a fasting blood cholesterol level 240 mg/mL or using medication for the condition [23].
Predictors
Sociodemographic and behavioral characteristics were taken into account as predictors of the discrepancy between reported and measured data. Sociodemographic variables included sex, marital status, region (urban/ rural), education (elementary school/high school/college or higher), employment status (yes/no), and household income. The equalized household income levels were categorized into tertiles (high/middle/low), after household income was divided by the square root of the household size. This was to adjust for differences in disposable income by the numbers of people in the household [24].
We considered three behavioral factors: smoking (never/past/current); binge drinking (yes/no); and physical activity (yes/no). Binge drinking was defined as consuming alcohol more frequently than twice a week and more than seven glasses (50 mL; 5 glasses for women) on each occasion. Physical activity includes three types of exercise: (1) intensive activity for at least 20 minutes on 3 or more days a week; (2) moderate activity for at least 20 minutes on 5 days or more a week; and (3) walking for at least 30 minutes on 5 days or more a week. Self-rated health was divided into three groups: very good and good, average, and poor and very poor.
Statistical analysis
Descriptive statistics for proportions and mean were presented for all variables. Kappa statistics were employed to measure the accordance of reported and measured values. Sensitivity and specificity were calculated to examine the accuracy of reported HTN, DM, and hypercholesterolemia, compared with the gold standard measures of these CVD risk factors. We used multiple logistic regression models to assess factors associated with an individual's awareness of CVD risk factors. Awareness rates and sensitivity indicated the proportion of participants who reported having HTN, DM, and hypercholesterolemia among the cases of clinically measured CVD risk factors. SAS 9.3 was used for these statistical analyses. Table 1 shows the descriptive characteristics, the prevalence of CVD risk factors (HTN, DM, hypercholesterolemia), as well as the results of k, sensitivity, and specificity tests between self-reported and measured diseases among participants aged 50 years or older. The first two columns of Table 1 show general characteristics of the study population. The sample consisted of similar proportions of three age groups: 50e59 years (36.8%), 60e69 years (34.7%), and 70 years and older (28.6%). The proportion of women was higher than men (57.4% vs 42.6%). Nearly three in four respondents were currently married, while one in four was divorced, separated, or widowed. The majority of participants (63.1%) lived in urban areas, compared to 37% rural dwellers. While the highest percentage of respondents (56.7%) had an education level of elementary school or less, only 9.1% of them had earned a college degree or higher, and 17.1% had received a high school diploma. About half of the respondents were in paid employment. In terms of health behaviors, 16.6% were current smokers, 22.6% were past smokers. About 7.8% were binge drinkers, and 41.3% did not exercise. More than one third of participants reported their health was very good or good, and another third reported having poor or very poor health.
Results
The results in the right column show the prevalence and values of k sensitivity and specificity between measured and self-reported HTN, DM, and hypercholesterolemia. The prevalence of three measured CVD risk factors was higher than that of reported risk factors, especially in reported hypercholesterolemia. In general, both the reported and measured prevalence in HTN and DM seems to be much higher among those who were older, previously married, and had lower socioeconomic status (lower education, lower income, not employed). For hypercholesterolemia, the reported and measured prevalence shows a similar trend to those of HTN and DM, but the measured prevalence differs across sex (women: 21%, men: 10.9%), marital status (previously married: 20.0%, married: 15.6), district (urban: 18.4%, rural: 13.7%), and employment status (not employed: 19.4%, employed: 14%). It is noteworthy that these differences were much higher than those of HTN and DM.
Regarding the results of k and sensitivity tests, whereas HTN and DM showed substantial agreement (k: 0.72 and 0.82, sensitivity: 73.0 and 79.3, respectively), hypercholesterolemia showed moderate agreement (k: 0.48, sensitivity: 46.7). When considering demographic and socioeconomic factors overall, higher kappa and sensitivity between measured and reported HTN was found in the following three demographics: the middle age group (60e69 years), women, and the unemployed group. However, the differences were not found in DM, except not employed. By contrast, the agreement and sensitivity of hypercholesterolemia was higher among those with a higher level of education (k: 0.54 vs. 0.46, sensitivity: 57.4 vs. 41.6) and a higher income (k: 0.49 vs. 0.44, sensitivity: 52.1 vs. 39.8), compared with reference groups. Interestingly, the k and specificity in hypercholesterolemia were higher among those who were non-employed than employed, in all three CVD risk factors. For health behavioral factors and self-rated health, the prevalence of measured HTN and DM was highest among past smokers (51.9% and 18.6%, respectively), binge drinkers (52.5% and 16.2%, respectively), and the no physical activity group (50.3% and 16.3%, respectively). By contrast, compared to the counterparts, the kappa and sensitivity in HTN and DM were higher among non-binge drinkers and never smoked. For hypercholesterolemia, measured prevalence was higher among never smoked and binge drinkers. The prevalence, k, sensitivity in HTN, DM, and hypercholesterolemia were the highest among those who had poor and very poor self-rated health (k: 0.77, 0.85, and 0.52; sensitivity: 80.6, 87.5, and 52.6, respectively). It is noteworthy that specificity for all three diseases shows a high accuracy of > 95%, regardless of the individual's socioeconomic or health behavioral factors. Table 2 shows the results of multiple logistic regressions, examining the sociodemographic and behavioral factors associated with the sensitivity of CVD risk
110
H. Chun, et al Marital status did not influence sensitivity to all three risk factors. Compared with adults living in urban areas, rural dwellers were less likely to be aware of CVD risk factors, especially DM and hypercholesterolemia. According to one's education level, a socioeconomic gradient was found in awareness of HTN and hypercholesterolemia, but not in DM. Compared to those with college education or higher, ORs for sensitivity progressively decreased with a lower educational level. Participants who were not employed seemed to be more aware of their HTN and hypercholesterolemia than those who were employed. Regarding health behaviors, only smoking was related to sensitivity of hypercholesterolemia. Sensitivity in past and current smokers was lower than for those who have never smoked (OR and 95% CI for past smoker: 0.60: 0.38e0.93; current smoker: 0.41: 0.26e0.66). Regardless of CVD risk factors, poor and very poor self-rated health was significantly associated with a higher rate of sensitivity than good and very good self-rated health (OR and 95% CI for HTN: 2.07, 1.71e2.50; DM: 3.23, 2.21e4.72; and hypercholesterolemia: 2.09, 1.55e2.80).
Discussion
This study was conducted to determine the accuracy of self-reported HTN, DM, and hypercholesterolemia when compared with global gold standard of measured data: representative national data from KNHANES IV. Our study's findings show that sensitivity of selfreported data for HTN and DM among Korean older adults aged 50 years and older was relatively high (73.0% for HTN and 79.3% for DM), but that of hypercholesterolemia was low (46.7%). Specificity, by contrast, provided accurate results for all three diseases (> 95%). Reliability, based on k agreement, showed similar results: k Z 0.72, 0.82, and 0.48 for HTN, DM, and hypercholesterolemia, respectively. Reliability and validity tests show that self-reported data are relatively good for HTN and DM, but reported hypercholesterolemia suggests a significant reporting error. In terms of accuracy of reported DM, our results highly agreed with previous studies from other regions, such as Europe and the USA, which show high level of accuracy. Moreover, our findings were similar to those from European research of the agreement between selfreported and measured data for HTN and hypercholesterolemia. A Dutch study [13] showed the highest concordance for DM (k Z 0.84), a moderate level for HTN (k Z 0.63), and the lowest level for hypercholesterolemia (k Z 0.48). A study of the Spanish adult population [9] also suggested good agreement for selfreported DM (k Z 0.78), but moderate agreement for HTN (k Z 0.51), as well as poor agreement for hyperlipidemia (k Z 0.27). While respondents' awareness of the nature of their diseases can affect their accurate self-reporting, it is assumed that intervention programs for CVD risk factorsdpublic education campaigns, knowledge translation, and the infrastructure of health examination programs across countriesdmay influence these differences in awareness of CVD risk factors.
Given the significance of preventing or controlling CVD risk factors, important modifiable risk factors in specificity for these diseases have not been adequately assessed. Our multiple logistic regression analysis shows that demographic and socioeconomic factors, behavioral factors, as well as self-rated health status are more likely to be significantly related to specificity in at least one of the three chronic diseases, but the direction of association differs according to the disease. Regarding chronological age, the elderly group aged 60e69 years turned out to have a higher awareness of all three diseases in comparison to adults aged 50e59 years and those aged 70 years or older, which proved inconsistent with previous studies [8,18,25]. Further analysis shows that those aged 60e69 years appear to have more medical check-ups; thus it is plausible that this age group seems to be more aware of the possibility of developing a chronic disease. When it comes to sex differences in awareness of diseases, our study shows that women are more vigilant for HTN, but negligent for hypercholesterolemia, which is in considerably disagreement with previous findings from other countries [18,21,25]. One of our study's notable findings was that, although females reported a greater prevalence of hypercholesterolemia, they were less aware of their symptoms than men. For all three diseases, rural people were less sensitive than urbanites, but about DM and hypercholesterolemia most especially. A possible explanation is that, when compared with urban elderly, rural dwellers may suffer from a scarcity of information, knowledge, or accessibility to health care services. Our further analysis (data not shown) reveals that the rural elderly may have difficulties interpreting the results of medical check-ups, even though the Korean government provides, free of charge, medical check-ups every While numerous studies have highlighted a socioeconomic gradient in health, our results also observed that highly educated groups were more sensitive to HTN and hypercholesterolemia. The socioeconomic gap in our study can be attributable to broader inequalities in socioeconomic forces; in particular, evidence shows that being highly educated is closely related to an increase in health-related information and knowledge, which lead to a higher awareness of chronic diseases [2,16,19,25]. Income levels are also known to be a relevant factor associated with higher awareness and better management [26], but our study found this link only for hypercholesterolemia. In contrast to previous findings [2], economically inactive groups in our study were more likely to be aware of HTN and hypercholesterolemia. It can be proposed that, more than likely they are spending more time on health management in comparison to the employed who have difficulty in balancing work with family life, for Koreans have the longest workweek of all OECD countries.
In contrast to other studies [17,27], health-related behavioral factors such as drinking and physical activities were not substantially associated with sensitivity for CVD risk factors. Only smoking influenced an awareness of CVD risk factors on health-related behaviors. In particular, hypercholesterolemia's sensitivity between reported and measured data was significantly lower among past and current smokers than among nonsmokers. In addition to these factors, our findings show that individuals with average and poor self-rated health had higher rate of agreement and of sensitivity for all three CVD risk factors, when compared with those with good self-rated health (e.g., OR in poor selfrated health: 2.07 for HTN, 3.23 for DM, and 2.09 for hypercholesterolemia). This result implies that self-rated health may be a useful indicator of awareness of CVDrelated factors.
The limitations of this study are as follows. First, we cannot exclude the possibility of an over-or underdiagnosis of directly measured data. For example, in the case of white coat syndrome for HTN or DM, which is an anxiety-induced false positive reaction [25,28], the sensitivity level of self-reported data may be decreased. However, since our classification criteria for directly measured data did not cover all diagnosis criteria, except for fasting blood glucose levels, we also used the oral glucose tolerance test for DM diagnosis. Second, physician-related factors such as age or hospital setting were not controlled. The heterogeneity of disease criteria is limited to the results garnered from one country to another. Nevertheless, this study has strengths; it is based on a nationwide, largescale population, sampling scheme with a high response rate (74%), and the results of this study can be generalized to the Korean population as a whole. Moreover, this study's results provide diverse factors including sociodemographic and behavioral factors that influence awareness of CVD risk factors, which can use intervention programs of these diseases.
In conclusion, this study of Korean older adults reveals that reliability and validity of self-reported data for HTN and DM are generally acceptable, but for hypercholesterolemia is relatively poor. Caution is needed when interpreting the results of community studies using selfreported data on CVD risk factors, especially hypercholesterolemia, among adults aged 50 years and older. This result can be employed to find controllable risk factors for cardiovascular management that is specific to subgroups. In general, the elderly aged 60e69 years, who are highly educated, live in rural areas, are not engaged in the labor market, and have "poor" self-rated health are more likely to be aware of their three CVD risk factors. Compared to men, women are more likely to be vigilant of HTN, but less likely to be aware of hypercholesterolemia. Past and current smokers are particularly less sensitive for hypercholesterolemia. Our study's results imply the importance of health intervention programs and highlight the necessity for improving public knowledge and information. This requires further studies to elaborate more on the predictors of sensitivity and their strength of association. The results can also be used to find controllable risk factors specific to subgroups for cardiovascular management. | 2017-10-21T17:20:05.917Z | 2015-12-15T00:00:00.000 | {
"year": 2015,
"sha1": "89f077ad6a4cca86bf6c1b0c2485e796efbdaedc",
"oa_license": "CCBYNCND",
"oa_url": "https://doi.org/10.1016/j.phrp.2015.12.002",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "89f077ad6a4cca86bf6c1b0c2485e796efbdaedc",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
51898931 | pes2o/s2orc | v3-fos-license | Effectiveness of a Homeopathic Product in Recurrent Tonsillitis
We conducted a pragmatic, randomized, controlled clinical trial in Germany, Spain and Ukraine in patients aged 6-60 years with moderate recurrent tonsillitis [1]. The combined treatment of the homeopathic remedy SilAtro-5-90 (Atropinum sulfuricum D5, Hepar sulfuris D3, Kalium bichromicum D4, Silicea D2, Mercurius bijodatus D8) and symptomatic medication (test group) was compared to symptomatic medication alone (control group) over a period of 14 months. Thereby, SilAtro-5-90 was given during 3 treatment periods of 8 weeks each.
Introduction
Recent clinical practice guidelines recommend watchful waiting for patients moderately affected by recurrent acute throat infections (ATI). Alternative therapies with a beneficial safety profile such as homeopathic remedies may be an interesting option for patients suffering from a moderate recurrent tonsillitis during this waiting time.
The results showed that the hazard of getting an ATI was significantly lower in the test group than in the control group (hazard ratio=0.45; 95%-CI: 0.34-0.60; p<0.0001, intensity model; ITT). Furthermore, already during the first treatment period, patients in the test group were free of recurrent tonsillitis symptoms during significantly more days (MWU-test; p<0.0001; ITT) compared to those of the control group. Also the number of acute throat infections treated with antibiotics was significantly lower in the test group compared to the control group (37% vs. 58.2%; 95%-CI: 9.13-33.36; p=0.0008; Chi 2 test; ITT).
SilAtro-5-90 was well tolerated: from the 225 adverse events (AEs) in the test group, only 3 AEs (gastroenteritis, nausea and foul taste) were assessed as being related to SilAtro-5-90. The percentage of patients/investigators having rated the tolerability of SilAtro-5-90 as "very good" or "good" ranged from 98.4% at the end of the first treatment period to 100% at the end of the third treatment period.
These results expand previous research with SilAtro-5-90. The efficacy of SilAtro-5-90 had previously been demonstrated in acute tonsillitis based on several studies, whereof 2 randomized placebocontrolled double-blinded trials [2,3]. In these trials, children suffering from acute tonsillitis, who were treated with SilAtro-5-90 experienced a significant reduction of tonsillitis-associated symptoms by day 4 of the treatment compared to children in the placebo group.
Research of SilAtro-5-90 in recurrent tonsillitis is scarcer: an observational study, in which 605 patients with recurrent tonsillitis took SilAtro-5-90 during consecutive 8 weeks, showed that 78.8% of the patients did not have recurrent tonsillitis complaints anymore at the end of this 8 weeks treatment period. Also, the tolerability of SilAtro-5-90 was rated predominantly as "very good" or "good" by the investigators and the patients. Our above mentioned clinical trial now confirmed the effectiveness and tolerability results of this observational study in scientifically more rigorous controlled conditions. It additionally made the interesting and important observation that patients in the SilAtro-5-90 group used overall less antibiotics for the treatment of recurrent tonsillitis.
The mechanisms by which SilAtro-5-90 may reduce ATIs and recurrent tonsillitis-specific symptoms were not investigated in the current study. Our results, however, support the idea that SilAtro-5-90 may stimulate the body's own adaptive healing process, which is one of the general homeopathic principles [4]. Moreover, they contribute to the clinical evidence for the use of SilAtro-5-90 in patients with recurrent tonsillitis. Finally, our results suggest that patients in the SilAtro-5-90 group were less susceptible to bacterial throat infections than patients in the control group, which is in turn an indication for an increased self-healing power in patients having been treated homeopathically.
Conclusion
Whereas double-blind clinical trials had already previously demonstrated the effectiveness of SilAtro-5-90 in the treatment of acute tonsillitis, the results of our recent research described above, substantiate prior findings on the use of SilAtro-5-90 in recurrent tonsillitis. They support the notion that an integrative treatment approach where SilAtro-5-90 is given alongside mainstream symptomatic treatment may bring therapeutic benefit to patients suffering from recurrent tonsillitis. | 2019-01-25T14:12:07.783Z | 2018-01-01T00:00:00.000 | {
"year": 2018,
"sha1": "cb3ac848c1fac1b985b24a7ef9b9e95956f51134",
"oa_license": "CCBY",
"oa_url": "https://doi.org/10.4172/2161-119x.1000344",
"oa_status": "HYBRID",
"pdf_src": "MergedPDFExtraction",
"pdf_hash": "0c6dc1d28f6e33110691502b948033870b1aa3b6",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
265214482 | pes2o/s2orc | v3-fos-license | Pharmacological Treatment of Acute Unilateral Vestibulopathy: A Review
There have been few investigations on the epidemiology, etiology, and medical management of acute unilateral vestibulopathy (AUV). Short-term pharmaceutical resolutions include vestibular symptomatic suppressants, anti-emetics, and some cause-based therapies. Anticholinergics, phenothiazines, antihistamines, antidopaminergics, benzodiazepines, and calcium channel antagonists are examples of vestibular suppressants. Some of these medications may show their effects through multiple mechanisms. In contrast, N-acetyl-L-leucine, Ginkgo biloba, and betahistine improve central vestibular compensation. Currently, AUV pathophysiology is poorly understood. Diverse hypotheses have previously been identified which have brought about some causal treatments presently used. According to some publications, acute administration of anti-inflammatory medications may have a deleterious impact on both post-lesional functional recovery and endogenous adaptive plasticity processes. Thus, some authors do not recommend the use of corticosteroids in AUV. Antivirals are even more contentious in the context of AUV treatment. Although vascular theories have been presented, no verified investigations employing anti-clotting or vasodilator medications have been conducted. There are no standardized treatment protocols for AUV to date, and the pharmacological treatment of AUV is still questionable. This review addresses the most current developments and controversies in AUV medical treatment.
Introduction
Acute unilateral vestibulopathy (AUV) has replaced "vestibular neuritis" as the preferred term in the International Classification of Vestibular Disorders.AUV is distinguished by an acute onset of spinning sensation, which is usually accompanied by nausea, vomiting, and gait instability, with symptoms lasting at least 24 hours [1].It has an estimated frequency of 3.5 instances per 100,000 persons, generally affecting adults between the ages of 30 and 60 years, and both genders are equally affected [2].
Management of AUV is usually multi-stepped, involving symptomatic and causal treatment followed by central com-pensation enhancement, with antivertiginous drugs, corticosteroids, and physical therapy being frequently used [3].Despite this approach, there are currently no treatment protocols established for AUV.Therefore, the goal of this review was to synthesize current information on AUV care to aid physicians in selecting the best treatment choice for their patients.
Vestibular suppressants not only reduce subjective symptoms but also the nystagmus induced by AUV.However, it is generally accepted that all symptomatic therapies should be reduced in a short time, in order to avoid impairment of the central compensatory mechanisms [3].According to the actual knowledge, we here describe the most relevant vestibular suppressants drugs and causal therapies amenable to be used in AUV.It is important to note that the commercially available drugs widely vary depending on world regions and are dependent on national medicines policies.
A literature search of articles relevant to AUV pharmaco-logical treatment options was conducted using the PubMed and Cochrane databases, with "vestibular neuritis" and/or "acute unilateral vestibulopathy" as search terms.Articles published over the last 5 years were preferred and references from the main review papers were also used to detect other relevant articles.A narrative condensed review format was employed with the goal of providing a brief and concise tool for readers searching for an updated summary regarding the state-of-theart drugs used in the management of AUV.
Symptomatic Treatment
The use of vestibular suppressants may be an important step to alleviate the patient's symptoms during the first 2-3 days after AUV onset.Albeit the restoration of peripheral function may help in some cases, final recovery of static and dynamic symptoms is largely dependent on the compensation/ habituation/adaption of the vestibular, visual and somatosensory pathways [4][5][6].Hence, evidence suggests that the goals of pharmacotherapy in AUV should be carefully assessed before prescription as symptomatic therapy often decreases the vestibular tone imbalance that acts as the main central compensation driver.This fact likely has a negative impact on central compensation and long-term outcomes, further supporting the use of vestibular exercises to increase vestibular tone imbalance, particularly by head movements, increasing central compensation and adequate return to a full functional state [3,7,8].Table 1 summarizes the vestibular suppressive compounds denoted in this review.
Anticholinergics
Anticholinergic medicines are hypothesized to function in the vestibular system by inhibiting the activity of vestibular nuclei or cerebellar pathways [4,9,10], and are classified as solely anticholinergic or mixed anticholinergic+antihistamin ergic drugs [4,9,10].Scopolamine, an alkaloid, is an example of a pure anticholinergic medication that acts as a non-specific muscarinic receptor antagonist [11][12][13][14].Scopolamine blocks muscarinic receptors competitively, resulting in peripheral antimuscarinic effects as well as central sedative, antiemetic, and amnestic qualities.It is most typically used in vestibulopathies to reduce neuro-vegetative symptoms caused by parasympathetic activation by motion [12].Because pure anticholinergic medications tend to be less effective after symptom onset, they are mostly used as a prophylactic [15], with a transdermal application of 0.5 mg scopolamine 4 to 6 hours before the commencement of a travel being recommended in mo-tion sickness [16].Scopolamine can also be taken orally, although its usage is limited due to the short average life of scopolamine in the plasma and its side effects [10,12], the most common of which are impaired vision and dry mouth [10].Migraine episodes are another potential side effect of scopolamine use [17], and overdosage can cause mental symptoms such as hallucinations and impaired awareness [18][19][20].
Scopolamine administration in AUV is the subject of few investigations, and its late start of effect (90 minutes) and limited duration when given orally may restrict its use in AUV [21].Nonetheless, scopolamine has been shown to be effective in the treatment of vertigo in a caloric induced vertiginous state, with one study proving the use of a transdermal patch releasing 0.17 mg/day for 7 days to be effective in acute vertigo from multiple causes [22,23].Additional research addressing the usefulness of scopolamine patches in alleviating AUV symptoms is required [23].Current commercial transdermal formulations release around 1 mg of scopolamine over a 72hour period [24].
On the other hand, combined anticholinergic and antihistaminergic agents, although primarily designed for motion sickness avoidance, are more widely employed in AUV [25].Common examples of this kind of drug are meclizine, dimenhydrinate, cyclizine, and diphenhydramine [3,22,[26][27][28].The antihistaminic component may cause tiredness and long-term weight gain when used regularly [29].Meclizine and dimenhydrinate have been demonstrated to be equivalent to diazepam in the treatment of vertigo in AUV, but entail less sedation than benzodiazepines and should therefore be favored as a first therapy, particularly in the elderly [30][31][32].As a result, unlike scopolamine, which is virtually exclusively used to combat motion sickness, these medicines are also utilized in AUV [3,22,26,27].
Addiction may occur from anticholinergic use, and cautious withdrawal is suggested [31].Commonly used anticholinergic doses are displayed on Table 1.
Antihistamines
Despite a lack of robust evidence to support their efficacy, antihistamines have played a notorious role in vertigo management throughout the last decades [31], being the most often used medicines in the treatment of vestibular vertigo [33][34][35].The presence of all four types of histamine receptors (H1R, H2R, H3R, and H4R) in the inner ear indicates the relevance of histamine in proper ear function [33,34].H1 and H3 antagonists are the most commonly used antihistamines.Many antihistamines have simultaneous anticholinergic (e.g., meclinizine, cyclizine, dimenhydrinate, diphenhydramine, promethazine) or calcium blocker (e.g., cinnarizine, flunari- zine) effects, thus belonging in different drug classes [33].Hence, only pure antihistaminergic drugs will be considered in this section to facilitate systematization.Betahistine, an H1R agonist and H3R antagonist, is the most regularly used antihistamine in Europe; however, the drug is not licensed in the United States, where meclizine, cyclizine, and diphenhydramine prevail [33,36].Histamine has been shown in various in vitro and in vivo studies to play a role in the modulation of vestibular plasticity during the process of vestibular central compensation [33].Indeed, betahistine reduced the time required to reach compensation by 1 month in the setting of neurectomy for Meniére's disease [37].The authors of the the VIRTUOSO (Effectiveness of Betaserc ® [Betahistine Dihydrochloride] in Patients With Vestibular Vertigo in Routine Practice) study study suggest that using betahistine for 2 months could improve compensation in patients with vestibular vertigo [38].However, the variability in methodology and patient selection has prevented a definitive conclusion on the efficacy of betahistine in vestibular compensation in AUV patients [4].The major disadvantage of antihistamines is their sedative effect, which has a significant influence on optimal central compensation and everyday functioning, in addition to their direct modulation of central vestibular signaling [33].While betahistine may not be a viable alternative for the management of acute vertigo symptoms, its H1R agonism qualities and consequent augmentation of central compensation provide for a possible application in the recovery phase of vertigo through the decrease of chronic symptoms [33,37].
SENS-111 is an H4R antagonist undergoing clinical trials [39].In both in vitro and ex vivo experiments, selective H4R antagonists inhibited vestibular neuron activity, relieving vertigo symptoms in rats with AUV [39,40].Despite being associated with less sedation than other antihistamine classes and being suggested to be safe and tolerable [33,40], this medication failed to meet the primary endpoint in a phase 2 study [39].
Serotonin inhibitor
Ondansetron is a 5-HT3 receptor antagonist, interacting with these receptors in both the central and peripheral nervous systems [41].Ondansetron has traditionally been regarded as an anti-emetic, and its usage in peripheral vestibulopathies is intended to alleviate associated neuro-vegetative symptoms [42,43].However, it has also been demonstrated to improve vertigo [44,45].Indeed, early treatment with a combination of ondansetron, corticosteroids, and antivirals has been linked with a decrease in the vestibular deficit in acutephase AUV when compared to the standard H1R antagonists [45].Ondansetron is often provided orally since it is readily absorbed and produces a comparable anti-emetic effect as intravenous dosing, which is more suited when individuals present with excessive vomiting [46][47][48].Because ondansetron has no impact on dopamine receptors, it is not associated with extrapyramidal symptoms [49,50].
Relevant compounds for the treatment of vertigo include promethazine [42], prochlorperazine [51][52][53], and thiethylperazine [54], all of which have anticholinergic properties that result in sedative and anti-emetic effects by acting on the chemoreceptor trigger zone (CTZ) [4,51].Low-dose promethazine (6.25 mg) has been shown to have anti-emetic effects comparable to 4 mg ondansetron but superior vertigo suppression abilities [42,55].Phenothiazides are also linked to peripheral adrenergic receptor blockade and quinidine-like cardiac effects, along with lowering seizure thresholds [56].Sedation, coma, hypotension, extrapyramidal symptoms, and cardiac arrhythmias are the most prevalent clinical indicators of toxicity, in addition to anticholinergic side effects [56].
Antidopaminergics
Despite the fact that previously mentioned drugs such as meclizine and promethazine have anti-emetic activities, administration of dopaminergic antagonists may also be beneficial [57].Metoclopramide, domperidone, and droperidol are anti-emetic medications that can be used to alleviate neurovegetative symptoms in AUV [58][59][60][61], although they can cause extrapyramidal signs and QT prolongation [56].While metoclopramide and domperidone do not appear to be true vestibular suppressants, droperidol's nervous system activity was revealed through its use in anesthesia [59,61], with both droperidol and droperidol+fentanyl suppressing vestibular symptoms and being effective in the management of vestibular disease [59,61].
Benzodiazepines
Gamma-aminobutyric acid (GABA) is a primary neuroinhibitory transmitter for vestibular neurons [27].Benzodiazepines decrease the electrical activity of vestibular nuclei through their GABA agonist characteristics, leading to their therapeutic impact in vertiginous disorders [10].
Patients may benefit from the sedative and anxiolytic characteristics of benzodiazepines in addition to their inhibitory impact on vertigo [10], with diazepam, lorazepam, and clonazepam being the most often used benzodiazepines for vestibular suppression [10].Diazepam is associated with a significant reduction in the spontaneous electrical activity of neurons in the medial vestibular nuclei, modulating both preand post-synaptic actions of diverse groups on neurons [10,[62][63][64].Clonazepam on the other hand is especially useful in the treatment of migraine-related vertigo and postural vertigo [65,66].One should be aware that the long-term use of these drugs prolongs or even prevent central compensation of vestibular tone imbalance [27].
Calcium channel antagonists
Calcium is present in the endolymph [67], flowing into the cells of the crista ampullaris in response to its movement [68] and triggering an action potential that is transmitted centrally [10,51].Calcium channel blockers are postulated to inhibit this calcium flow, reducing the depolarization that could lead to vertigo [10,51].Flunarizine and cinnarizine are the two major calcium channel blockers used in clinical practice [51,69].Flunarizine is also a dopamine blocker and cinnarizine exhibits concomitant anti-histaminergic effects and blocks pressure sensitive potassium channels.The latter possibly provides a separate mechanism for the treatment of hydrops [69].
Flunarizine has been shown to be a powerful peripherally acting labyrinthine suppressant without the characteristic side effects of antihistamines and anticholinergics.Flunarizine has applications in the prevention of motion sickness, vertigo, and migraine, which constitutes an advantage in migraineurs [51,70,71].Cinnarizine is less potent than flunarizine [72].A fixed combination of cinnarizine 20 mg and dimenhydrinate 40 mg is available in Europe for patients with AUV.Studies confirm the benefit of this association on the management of AUV symptoms [72].
Neither drug is selective for a particular calcium channel subtype [51,70], thus exerting effects throughout the central nervous system.This can lead to potential toxicity, including weight gain, depression, sedation, or even parkinsonian symptoms, and while both drugs have been used in Europe, their use is rarer elsewhere [51].
Amino acids/metabolites
The exact role of N-acetyl-L-leucine in vertigo is unknown [73,74].Its primary mode of action is thought to be its influence on vestibular nucleus networks, with restoration of membrane potential of hyperpolarized/depolarized vestibular neurons.N-acetyl-L-leucine improves postural symptom compensation in a dose-dependent manner [75], possibly by activating the vestibulocerebellum and deactivating the pos-terolateral thalamus.This molecule is suggested to have a rapid antivertiginous effect when administered intravenously or orally in humans [73,75].N-acetyl-L-leucine is currently not used to treat vertigo in the United States.
Causal Treatment
Immune system, inflammation, and glucocorticoids Since Hiyoshi and Sekitani [83] discovered a possible association between the onset of AUV and influenza vaccination, immunological mechanisms have been formally implicated in the etiology of AUV [84].Bumm and Schlimok [85] later identified similarities between lymphocyte subpopulations and HLA-DR determinations in diseases of the inner ear and Bell's palsy.A recent study employing gene expression profiling and bioinformatics analysis discovered a difference in expressed immune system genes in AUV patients when compared to controls [86].According to the same study, neutrophil degranulation is one of the most prominent immunological pathways in AUV, leading not only to inflammation but also to a pro-thrombotic state in the vestibular end organ [86].These findings support Kassner, et al's [87] discovery of proinflammatory activation of peripheral mononuclear cells in AUV patients, and explain the widespread use of endovenous glucocorticoids, such as dexamethasone, methylprednisone, and prednisone (Table 2), in AUV in both Europe and the United States [88,89].Nonetheless, their usage is contentious, with no established standards for dosing or therapy length.Contrary to abrupt idiopathic hearing loss, intra-tympanic administration of corticoids in AUV has received less attention.Although the majority of authors endorse a brief course of high-dose glucocorticoids, a recent systematic review and meta-analysis determined that there is insufficient evidence to support their usage [90].These medications appear to offer short-term advantages in canal paresis but no long-term benefits in canal paresis or symptomatic recovery, with a small number of patients reporting adverse effects [90].Additionally, it was recently established from AUV rat models that acute anti-inflammatory modulators alter the post-lesional functional recovery and the endogenous adaptive plasticity pathways, which may eventually cause exacerbated and sustained vestibular and postural impairments instead [91].
Vascular mechanisms and blood related treatments
Vascular changes, as an epiphenomenon mediated by inflammation, have been linked to AUV by significant increases in plasma fibrinogen concentrations and decreased lipoprotein (a) levels [92].D-dimers, other marker of coronary artery disease, was also found to be higher in patients with AUV [93].However, whether vascular processes represent a causal cause of AUV remains unknown [93,94].Evidence suggests an epidemiologic link between cardiovascular risk factors and the prevalence of AUV, with comorbidities such as diabetes, dyslipidemia, hypertension, obesity, ischemic heart disease, and cigarette smoking being higher among AUV patients compared to the general population [95].
Inner ear infarctions cannot currently be seen using current imaging modalities [96].Nonetheless, Liqun, et al. [96] stated in a recent study that selective labyrinth embolism should be addressed in patients with AUV or auditory symptoms in the presence of concomitant acute infarctions, even if in nonanterior inferior cerebellar artery territories.To date, peripheral vestibular histology does not support vascular occlusion as an etiology of AUV, and there is no significant evidence supporting the use of anticoagulants or antiplatelet medications.
Ginkgo biloba extract (Table 2), a mild antiplatelet compound, has been shown to reduce vascular resistance and improve peripheral blood circulation.However, its use has been mainly studied as a central neuro-modulator in vestibular compensation [97,98].Preliminary animal model studies indicate this agent as efficient in vestibular compensation through modulation of cerebral vestibular networks [99][100][101].
Infection hypothesis and the use of antivirals
A viral etiology has long been suspected as being involved in AUV pathogenesis.The epidemiologic timing association between AUV onset and recent upper respiratory tract infec- tions, the detection of herpes virus in the vestibular ganglia, and the similarities found in postmortem histopathological findings in the vestibular nerve/vestibular sensory epithelium and other known viral disorders all lend support to this hypothesis [84,[102][103][104].Aside from the aforementioned data, there are very few trials utilizing antivirals in AUV, with Strupp, et al. [88] describing that valacyclovir (Table 2) did not enhance vestibular recovery in the setting of AUV.Hence, there are no available standardized guidelines for the use of antivirals.
Table 2 shows some examples of common causal treatments used in AUV, and Fig. 1 summarizes the proposed mechanisms and causal treatments in AUV.
Conclusion
AUV may have an undefined etiology, but its clinical presentation follows a common pattern for most patients, making it possible to design a comprehensive approach for treatment.Most times, impending vertigo lasting for days ensues and therefore, the first thing to do after diagnosis is to provide symptomatic relief to the patient, which is generally very symptomatic.Vestibular suppressants are almost always in the equation, in order to rapidly alleviate the patient's neuro-vegetative symptoms (predominantly nausea and vomiting) and intense rotatory vertigo.Pharmacological treatment of AUV is a mainstay in the management of this pathology, allowing for symptomatic relief of truly diseased patients.The clinician should be familiar with the vast and valuable arsenal of vestibular suppressant drugs, in order to be able to use them as wisely and quickly as possible, keeping in mind that central compensation can be compromised if the drugs are used inadvertently for long periods.
On the other hand, prognostic changing treatments to halt the "vestibular shutdown" progression in an early phase are still lacking.This probably reflects the same lack of consensus about AUV causal mechanisms, hindering the development of specific therapeutic targets, with corticosteroids being widely used, although not incontestably.This review ultimately aims to summarize the evidence and guide the clinician towards the best suitable solution in the treatment of AUV.
The topographic diagnosis in AUV is becoming a reality with the advent of video head impulse test [105].Topography of the involved segments may point to a vascular event in a specific arterial territory as the cause of the AUV, undermining the chance for an inflammatory cause.One example is the suggested selective ischemia at the area of the common cochlear artery when there is an abrupt functional loss of co- chlear epithelium, saccular macula and posterior canal ampulla [105].Furthermore, plasma biomarkers may play a role in separating causative pathways [93].
Thus, and parting from a focus on more efficient diagnostic exams, we believe that the future of the management of AUV may be a more tailored approach to its causal mechanism, as each AUV is unique, and a single case may be one of the multiple subtypes of AUV yet to be described.
Along with the advances in diagnosis, it is also vital to have full confidence when planning a course of treatment [106].Choosing an adequate treatment may become increasingly difficult as a lot of research is being conducted on AUV symptomatic treatment [39,75], causal treatment [93,94], and vestibular rehabilitation [107].Furthermore, and despite some debate on the subject [108], there is reasonable evidence that, whereas AUV affects the whole vestibular nerve, vestibular neuritis frequently spares a portion of the nerve [109,110].As a result, breakthroughs in physical and pharmacological therapy aimed at restoring residual vestibular function and speeding vestibular functioning are expected in AUV [111].Fig. 2 summarizes future perspectives in AUV diagnosis and treatment.
Summary recommendations
In general, there is insufficient evidence to support the effectiveness of anti-vertiginous drugs for AUV.However, betahistine may be effective for peripheral vertigo and may be considered (grade of recommendation: C) [112].A combination of cinnarizine and dimenhydrinate may be used for improving symptoms in the acute stage of vestibular neuritis (grade of recommendation: B) and can be used for a short period [112].Betahistine, ondansetron, cinnarizine, flunarizine, dimenhydrinate, and diazepam are among the most widely studied vestibular supressants and should be preferred.Corticosteroids resulted in a statistically significant improvement in vestibular neuritis [113], so that the authors recommend their routine use in AUV.Despite the multitude of therapeutic options, there is a lack for larger, higher quality studies regarding the pharmacological treatment of AUV.
majority of authors agree with a prompt course of high dosage of glucocorticoids for a short time period.Nevertheless, a recent systematic review and meta-analysis concludes there is insufficient evidence to support its use.
Fig. 2 .
Fig. 2. Acute unilateral vestibulopathy (AUV) treatment future perspectives: using video head impulse test towards topographic diagnosis, laboratorial data and imaging towards the causal diagnosis and scientific advancement towards targeted therapy and rehabilitation.
Table 2 .
Causal treatments for acute unilateral vestibulopathy | 2023-11-14T06:19:02.683Z | 2023-11-13T00:00:00.000 | {
"year": 2023,
"sha1": "11dbe3b7a27369222ad69f7fb5ed464a1d009907",
"oa_license": "CCBYNC",
"oa_url": "http://www.ejao.org/upload/pdf/jao-2023-00066.pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "79736c4ba69e5db79586c547b6b50473f50d5283",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": []
} |
266346712 | pes2o/s2orc | v3-fos-license | Pluronic F-127 Enhances the Antifungal Activity of Fluconazole against Resistant Candida Strains
Candida strains as the most frequent causes of infections, along with their increased drug resistance, pose significant clinical and financial challenges to the healthcare system. Some polymeric excipients were reported to interfere with the multidrug resistance mechanism. Bearing in mind that there are a limited number of marketed products with fluconazole (FLU) for the topical route of administration, Pluronic F-127 (PLX)/FLU formulations were investigated in this work. The aims of this study were to investigate (i) whether PLX-based formulations can increase the susceptibility of resistant Candida strains to FLU, (ii) whether there is a correlation between block polymer concentration and the antifungal efficacy of the FLU-loaded PLX formulations, and (iii) what the potential mode of action of PLX assisting FLU is. The yeast growth inhibition upon incubation with PLX formulations loaded with FLU was statistically significant. The highest efficacy of the azole agent was observed in the presence of 5.0 and 10.0% w/v of PLX. The upregulation of the CDR1/CDR2 genes was detected in the investigated Candida strains, indicating that the efflux of the drug from the fungal cell was the main mechanism of the resistance.
Fungal pathogens cause chronic conditions such as chronic mucocutaneous candidiasis, asthma, or chronic pulmonary aspergillosis, as well as life-threatening diseases such as fungaemia, pneumonia, or meningitis. 1 Benedict et al. estimated that in 2017, the treatment of fungal diseases cost more than $7.2 billion, including a total cost of $1.4 billion for hospitalizations due to Candida infections (n = 26,735). 2A constant increase of systemic and topical fungal infections has been observed over the past decade, with an increasing number of recurrent infections such as oral and vaginal candidiasis. 1,3,4hese can be caused by multidrug-resistant strains of fungi or non-Candida albicans Candida (NCAC) that are naturally resistant to commonly used treatments. 5As a consequence, total consumption of antifungal drugs has increased over the years worldwide (Figure 1A). 6,17n vulnerable groups of patients (e.g., treated with immunosuppressants), fungal infections are associated with high mortality.Therefore, early diagnosis and the selection of a safe and effective antifungal treatment determine the success of the therapy. 7As antifungal treatments are limited to only a few chemical classes, including azoles, echinocandins, polyenes, and flucytosine, increasing drug resistance is observed across clinically isolated fungal strains. 1 The resistance phenomenon is particularly important for azole derivatives, which are the most popular antifungals used for prophylaxis as well as empirical and directed therapy.Due to the limited options for antifungal therapies, the emergence of multidrug-resistance strains can have devastating effects on patient outcomes.Therefore, the search for new antifungal agents or new antifungal formulations that can help overcome multidrug resistance in fungi is of paramount importance for the future development of antimicrobial treatments.
Representatives of the Candida genus have developed various mechanisms of resistance to azoles (Figure 1B).One of them is an active efflux of the drug from the fungal cell by membrane transporters.Two main classes of efflux pumps have been identified in Candida strains, namely, ABC proteins and MFS pumps.Overexpression of genes encoding membrane proteins (e.g., CDR1, CDR2, CDR3, CDR4, CDR11, MDR1, and FLU1) results in enhanced azoles efflux. 8Another resistance mechanism involves alterations in an enzyme that is a target for azoles, namely, sterol 14α-demethylase encoded by the ERG11 gene.This enzyme is involved in the synthesis of ergosterol, which is an essential component of the structure of fungal cell membranes.A point mutation in the gene leads to the expression of a protein that is characterized by a binding site of lower affinity. 9Therefore, evaluating the mechanism of resistance for clinical Candida isolates might result in the design of more effective pharmaceutical formulations against fungal infections.
Fluconazole is a representative of azoles, registered by the European Medicine Agency as tablets and capsules (dose 50− 200 mg), the solution for infusion (2 mg/mL), syrup (5 mg/ mL), and suspension at concentrations of 10 and 40 mg/mL. 10he marketed products are dedicated to systemic use only, with 0.5% Elazor gel registered in Italy being the only exception authorized for topical application in the European Union. 10Fluconazole is characterized by high molecular flexibility; therefore, it is prone to form various polymorphs 11 that might affect its bioavailability in the formulation.For patients suffering from fungal keratitis, the solution for infusion can be compounded into eye drops as part of the course of personalized therapy. 12The topical route of administration could limit the systemic use of fluconazole and related side effects and the development of fluconazole-resistant strains.The possibilities include semisolid and liquid forms administered topically, such as gels, sprays, foams, or ointments that can be applied to the skin, mucous membranes, as well as into the eye.Such novel fluconazole formulations would primarily benefit patients vulnerable to recurrent fungal infections associated mainly with immunosuppressive or oncological treatments. 13,14t has been reported that some classes of polymers, in addition to their primary function as excipients, may exhibit additional properties affecting the final pharmacological effect of the applied formulation. 15Some polymers are characterized by intrinsic pharmacological activity, e.g., immunostimulating, antibacterial, antifungal, or antioxidant.Moreover, polymers themselves or conjugated with enzyme−inhibitor might inhibit certain enzymes, e.g., (chymo)trypsin, elastase, peptidase, and nuclease.They can interact with enzymes involved in drug metabolism (cytochrome P450 and CYP450), resulting in increased or decreased activity of some isoforms.They might also act as efflux pump inhibitors, resulting in a higher sensitivity of the cells to the treatment. 15With regards to Pluronics (PLX), inhibitory activity against ABC transporters such as Pgp (P-glycoprotein), MDR (multidrug resistanceassociated protein), and BCRP (breast cancer resistance protein) has been reported. 16The proposed mechanism of binding to the cell membrane depends on the properties of the block copolymer (Figure 1C). 16The structure of the copolymer, i.e., the hydrophilic/hydrophobic balance, determines the mechanism of its binding with the cell surface.Hydrophilic Pluronics absorb mainly at the membrane surface, whereas hydrophobic Pluronics are more likely to penetrate into the lipid bilayer.The insertion is explained by the interaction of propoxy blocks with fatty acid residues and ethoxy groups with polar groups in the lipid bilayer.Alterations in the lipid microenvironment may facilitate membrane transport of small molecules (e.g., drugs).
The aim of the study was to investigate whether the addition of PLX can increase the susceptibility of resistant Candida strains to FLU and whether there is a correlation between block polymer concentration and the antifungal efficacy of FLU when assisted by PLX.Fluconazole was selected as the model drug due to the increasing resistance of Candida strains to this drug and a lack of FLU topical formulations on the market.We have selected amphiphilic Pluronic F-127 that has a PPO/PEO ratio equal to ca. 0.3 (respective block lengths of PEO−PPO−PEO are 100−65−100) 19 to form micellar solutions and solutions gelling at various temperatures loaded with different FLU concentrations.
This work, for the first time, describes the enhanced activity of FLU in the presence of PLX against FLU-resistant Candida strains.To the best of our knowledge, this is the first extensive study covering a wide range of PLX concentrations, including the mechanistic considerations of the increased antimicrobial activity of the investigated formulations.
PLX and FLU Formulations.
We studied materials with increasing Pluronic F-127 concentrations, namely from slightly above the critical micellization concentration, CMC (0.08% w/v) 20 to 25.0% w/v.This series comprises a variety of formulations differing from each other in terms of their rheological properties.The detailed explanation regarding the concentration range of all the components is presented in Supporting Information, Section S1.They included both micellar solutions and solutions gelling at various temperatures (Table 1).Such a wide range of polymer concentrations enabled us to establish whether the polymer concentration might be a factor affecting the potential antifungal mode of action of the polymer.
Although the sol−gel transition temperature decreased with increasing polymer concentration, it was not affected by the addition of fluconazole (Table 2).The phase transition temperatures were below the temperature of the human body (36.6 °C) and human body surface (34.0 °C); therefore, the formulations with 20.0 and 25.0% w/v polymer content might be effectively administered topically or into body cavities when developed into the final pharmaceutical product.
At PLX concentrations of 10.0 and 15.0% w/v, the sol−gel transition was not observed (Supporting Information, Figure S1).PLX-20 compositions were liquid at room temperature.This might facilitate their application if formulated into the final drug form, such as eye drops or sprays with enhanced adhesive properties to the cornea, oral mucosa, or skin.The summary of the state of Pluronic F-127 formulations at particular polymer concentrations is presented in Supporting Information, Table S2.
Drug Content in Micellar
Solutions of Pluronic F-127.As Pluronic F-127 increases the solubility of the drugs, 20−22 we have studied whether micellar solutions up to 5.0% w/v loaded with fluconazole at its maximum solubility at a particular polymer concentration remained stable over prolonged storage at room temperature.Our results indicate that fluconazole was dissolved within a polymer solution throughout a whole range of investigated concentrations over 84 days (Supporting Information, Section S3).The drug content within the formulations remained within 100 ± 3.0%, depending on the sample.S6 and Table S5).
Antifungal Activity of
In contrast, susceptible strains of Candida did not show a statistically significant decrease in the absorbance value, indicating that PLX formulations are indeed interfering with the resistance mechanisms of Candida strains (Figure 3b and Table S6 in Supporting Information, FLU-MIC vs PLX-0.08_FLU-MIC,p = 0.320145, and FLU-MIC vs PLX-5_FLU-MIC, p = 0.060342, and Supporting Information, Figure S7 and Table S6 addition of either PLX-0.08_FLU-MIC or PLX-5_FLU-MIC micellar solutions of FLU (Table 3).In the four Candida strains susceptible to increased exposure to fluconazole (namely C. glabrata 1640, 634, 635, and 2453), a decrease in MIC values was observed after treatment with the PLX-0.08_FLU-MICformulation (Table 4).Block copolymers alone without the addition of fluconazole, namely, PLX-0.08 and PLX-5, did not cause ≥50% growth inhibition (except for C. albicans 3089 and C. tropicalis 3151 at 5.0% w/v polymer; thus, these two strains were excluded from the MIC evaluation and statistical analysis for PLX-5_FLU-MIC formulations).
Furthermore, addition of verapamil (series VER_FLU-MIC) induced a statistically significant decrease in the growth of resistant C. glabrata (Figure 3a and Table S4 in Supporting Information, FLU-MIC vs VER_FLU-MIC, p < 0.000001).This may indicate that the FLU resistance is a result of ABC and/or MDR efflux pump overexpression.A statistically significant decrease in absorbance values after incubation of yeasts with verapamil-fluconazole formulations was observed for the following strains (see Supporting Information, Figure S5 for details): C. glabrata 2586 (p = 0.005267), 2738 (p = 0.002483), 2853 (p = 0.000025), 1973 (p = 0.001151), and 1467 (p = 0.003504).The detailed statistical analysis involving all the studied Candida strains (C.krusei, C. albicans, C. glabrata, and C. tropicalis) showed similarly a statistically significant decrease in absorbance value (with respect to the FLU-MIC series, see Figure S6 and Table S5).In contrast, FLU-susceptible Candida strains did not show a decrease in absorbance when treated with VER-containing samples (Figure 3b, FLU-MIC vs VER_FLU-MIC, p = 0.111595 and Supporting Information, Figure S7 and Table S6).In 7 strains resistant to fluconazole, MIC decreased upon addition of verapamil (series VER_FLU-MIC, Table 3).The largest decrease of the MIC value upon addition of verapamil was observed for C. albicans 3057, C. tropicalis 3151, and C. glabrata 2853.In the case of strains susceptible to increased exposure to fluconazole, the addition of verapamil did not trigger a decrease in MIC values (Table 4).Control experiments with verapamil did not inhibit yeast growth, regardless of the strain (VER sample).
Among the investigated resistant strains, C. albicans 3057, C. glabrata 2586, C. glabrata 2738, and C. glabrata 2853 were characterized by a decrease of the MIC value upon addition of both Pluronic and verapamil, which might suggest that two of them share a similar mechanism of action.Hence, we suppose that the polymer might interfere to some extent with the efflux pump, allowing enhanced accumulation of the drug inside cells.
Effect of Pluronic F-127 Concentration on Antifungal Activity of FLU.
As some differences in the susceptibility of yeasts to fluconazole were observed between the series composed of 0.08 and 5.0% w/v Pluronic F-127, the antifungal effect of higher polymer concentrations was investigated further.Since the increased concentration of polymer results in increased viscosity of the Pluronic F-127 solution 23 that could hinder the use of the broth microdilution method, the cup plate method was used for the investigation of the antifungal activity of PLX formulations at 10.0, 15.0, 20.Solutions of Pluronic F-127 loaded with fluconazole exhibited antifungal activity when tested by the cup plate method.This suggests that the addition of block polymer, even at higher concentrations did not hinder the antifungal activity of fluconazole, similarly to the results obtained by the broth microdilution method.Fifteen resistant strains were selected for cup plate method studies (C.krusei ATCC 6258, C. albicans 3057, and C. glabrata 2586, 2738, 2853, 3010, 1467, 2124, 137, 140, 1941, 1973, 2342, 3154, and 773).The diameter of the zone of growth inhibition is summarized in Table 5.It is important to note that Pluronic F-127 without the addition of fluconazole, regardless of the polymer concentration (series PLX-10, PLX-15, PLX-20, and PLX-25), applied to the wells did not produce the inhibition zone.
For ten out of the 15 investigated strains, the inhibition zone increased The highest increase in the diameter of the zone of inhibition, as compared to the fluconazole solutions, was observed for PLX-10_FLU-4−15 formulations (Figure 4).In the series PLX-25_FLU-4−15 with 25.0% w/v Pluronic F-127 regardless of the strain, we observed a reduced inhibition zone to values lower than exhibited in the FLU-4−15 series (Supporting Information, Table S7).It is known that Pluronic F-127 at concentrations above 17% w/v undergoes a sol−gel transition at temperatures below 30 °C, resulting in the
ACS Infectious Diseases
formation of viscous gels that may limit the drug diffusion into the agar.As presented in Table 2, 25.0% w/v Pluronic F-127 was gelling already at ca. 21 °C, indicating that even before the incubation with fungi, its penetration into agar was limited.Therefore, 25.0% w/v Pluronic F-127 formulations were excluded from the statistical analysis.The phase transition temperature might also be the explanation for the decreased inhibition zones in PLX-20_FLU-15 in the above-mentioned strains.In 5 investigated strains (C.glabrata 1941, 1973, 2342, 3154, and 773), no inhibition zone around wells at the plates was observed both for formulations with neat fluconazole and fluconazole with PLX.For all C. glabrata strains, the inhibition zone was significantly larger when fluconazole was accompanied by block copolymers, regardless the copolymer concentration (Figure 5 and Table S8 in Supporting Information, FLU vs PLX-10_FLU, p < 0.000001, FLU vs PLX-15_FLU, p < 0.000001, and FLU vs PLX-20_FLU, p < 0.000001).The same phenomenon was observed when individual fluconazole concentrations loaded into polymer samples were considered (Supporting Information, Figure S8a and Table S8).The greatest changes were observed in the PLX-10_FLU series (Supporting Information, Figure S8b and Table S8).The statistical analysis of the combined data retrieved from experiments performed on all three different investigated resistant Candida strains showed a very similar pattern (C.krusei, C. albicans, and C. glabrata) and is presented in the Supporting Information (Figure S9 and Table S9).
Kinetics of the Antifungal Activity. As shown in
Figure S10 in Supporting Information, the progressive increase of absorbance in wells with treated yeasts was observed over the course of time.The plateau was apparent, especially in the investigated C. glabrata strains after ca.16−20 h of incubation.With increasing concentrations of fluconazole (in the range of 8−64 mg/L), a reduction of growth was observed; thus, growth curves of Candida showed dose-dependent inhibitory characteristics.When different concentrations of Pluronic F-127 were considered, 5% w/v caused faster inhibition of the yeast growth, as seen by the larger angle of the slope of the growth curves.This long-term kinetic study showed continuous inhibitory effects on fungal cells of the examined formulations, mostly starting at 5−6 h of incubation and lasting until 16−20 h of the experiment.It might be an indication of its prolonged action at the application site when developed into the final drug form.
2.6.ERG11, CDR1, and CDR2 Gene Expression Analysis.In order to determine the mechanism responsible for the enhanced resistance to fluconazole among the investigated Candida strains, the level of expression of the resistance-associated genes encoding drug transporters and sterol 14α-demethylase was established.The data on gene expression could provide an indication of the likely mode of action of Pluronic F-127.A detailed summary of the results from gene expression analysis in 15 C. glabrata strains resistant to fluconazole is given in the Supporting Information (Tables S10−S12).
Based on the real-time PCR results, the upregulation of the CDR1 and CDR2 genes is the major mechanism of resistance in the investigated strains (Table 6).Six out of 15 studied strains displayed a higher level of CDR1 expression in comparison to a reference isolate (2 −ΔΔCT value in the range 1.19−1.95).Moreover, the other six analyzed strains demonstrated a significant increase of CDR1 transcript (2 −ΔΔCT value between 2.01 and 8.85).Similarly, the CDR2 gene was overexpressed in nine strains as compared with the reference strain (2 −ΔΔCT value in the range 1.03−1.75),and in five strains, a significant overexpression was detected (2 −ΔΔCT value between 2.43 and 7.01).Each of the investigated strains displayed a higher expression level of either the CDR1 or CDR2 genes.Overall, the expression level of CDR1 genes was slightly higher compared to CDR2 genes.
In the case of the ERG11 gene, only three out of 15 strains expressed the ERG11 gene at a higher level than a control 2.7.Microscopic Imaging.Selected samples were analyzed using microscopy in phase contrast to detect any changes in growth and monitor yeast morphology (Supporting Information, Figures S11 and S12).The decreased amount of yeast cells without any changes in their morphology was noticed in wells characterized by decreased absorbance after incubation of fungi with fluconazole and polymer.
The dye Nile red that fluoresces in a hydrophobic environment was chosen to track the efflux in Candida cells treated with Pluronic F-127 and a known efflux inhibitor, verapamil.Nile red is a substrate for many types of transporters, both ABC and MFS (Cdr1, Cdr2, and Mdr1, respectively). 25As shown in the images (Figure 6 and Supporting Information, Figure S13), Candida cells stained with Nile red, in the presence of both polymer and verapamil, exhibited fluorescence.However, the amount of stained cells in nontreated samples 26 was larger in comparison to the treated cells, probably due to the inhibition of efflux pumps by both substances.It meant that the accumulation of the dye within cells was hindered by both verapamil and Pluronic F-127 affecting efflux pumps.
To evaluate whether polymers might interfere with the integrity of the fungal membranes as another possible way of acting, the cells incubated either with Pluronic F-127 or with amphotericin B were stained with propidium iodide.Propidium iodide is known to be a fluorescence dye that crosses damaged cell membranes and stains intracellular nucleic acids. 27Amphotericin B is a reference substance that, upon binding with cellular membranes, trigger the formation of pores, resulting in their death. 28Similarly to the case of Nile red staining, Candida cells, regardless of the substance used for incubation, exhibited fluorescence (Figure 6, and Supporting Information, Figure S14).The amount of stained cells treated with Pluronic F-127 was strain-dependent.The highest proportion was for the C. glabrata 2738 strain, exhibiting a significant overexpression of CDR1 and CDR2 genes, whereas the lowest was for C. glabrata 2853, characterized by a moderate overexpression of efflux pumps.As expected, Pluronic F-127 was less effective than a potent antifungal antibiotic when compared to amphotericin B-treated subculture.However, the results might indicate a linkage between the impaired function of the efflux pump and fungal membrane disruption when the mechanism of the action of the polymer considered.
DISCUSSION
Taking into account the developing antifungal resistance among Candida strains, the application of excipients that exhibit antifungal potential, i.e., antimicrobial polymers as functional pharmaceutical formulation additives, is perceived as a favorable solution in future antifungal strategies.Only a few reports are concerned with the study on the potential antifungal activity of neat excipients or their synergism with the antifungal substance.A direct comparison of these data poses difficulties as these macromolecules have different molecular weights and methods of preparation.In addition, different pathogenic strains are investigated.Some of the few examples reporting potential antifungal activity toward Candida species include hyaluronic acid and chitosan.However, the mechanisms underlying such an action remain unclear.Data on excipients with potential antifungal activity against Candida strains or their synergistic effects with an antifungal drug are summarized in Table 7.
In general, antifungal polymers are characterized by cationic and hydrophobic regions that are expected to interact with anionic components of the cell wall, as well as microbial phospholipids and membranes.In studies performed by Lo et al., six various chitosans with specific molecular weights were applied in combination with fluconazole, showing a great synergistic fungicidal effect against susceptible and drugresistant C. albicans and C. tropicalis both in liquid and agar media. 29Combination treatment exhibited a synergistic antifungal effect in both investigated drug-resistant strains (FIC index < 0.5). 29Grimling et al. have demonstrated that compositions comprising clotrimazole and high-molecularweight chitosan could be an effective solution in a topical antifungal formulation against non-Candida albicans Candida strains. 30The synergistic effect of clotrimazole and chitosan combinations was observed in tests carried out at pH 4 on C. glabrata strains. 30The inhibition of C. glabrata growth reached at least 90%, regardless of the drug/excipient weight ratio. 30tudies of the mechanism of action of both neat excipients and the developed formulations are rare.Shih et al. exceptionally attempted to evaluate the mechanism of the enhanced activity of chitosan, which is probably targeting the cell surface. 31The decrease in the expression of Ada2 genes was observed after exposure to 0.2% chitosan for 20 min and for 1 h in the wildtype strain C. albicans compared to polymer-untreated cells. 31hose genes are directly involved in the defining of the cell surface composition and its integrity by, e.g., regulating the MDR1 and CDR1 efflux pumps; 31 therefore, the authors concluded that chitosan might have altered the integrity of the cell surface of C. albicans.Since Pluronics were reported as one of the amphiphilic polymers that affect drug efflux transporters in cancer cells, resulting in sensitization and prevention of multidrug resistance, 16 we have decided to investigate the effect of Pluronic F-127 on the Candida cells, which also express drug efflux pumps responsible for their resistance to fluconazole.It has been reported in the literature that Pluronic P85 (PEO− PPO−PEO with block lengths of 25−40−25) 19 affected MDR cells already at concentrations below CMC (0.03 wt%). 32It was suggested that unimers are able to incorporate and translocate across the cellular membranes, where hydrophobic PPO chains of Pluronic are embedding into the membrane hydrophobic areas, causing so-called "membrane fluidization", resulting in alterations of the membrane structure and decreasing its microviscosity. 32It has been proposed that the formation of micelles at higher concentrations of block copolymer might result in hiding these hydrophobic PPO chains in the micellar core, resulting in their diminished availability to affect the cellular membranes. 32However, as described in our recent work, 20 molecular dynamics simulations along with experimental data showed an amphipathic microsegregated surface in Pluronic F-127 micelles with hydrophobic domains exposed to solvent instead of the typically accepted hydrophilic surface and hydrophobic core.This might be the reason for the fact that although the lowest studied concentration of Pluronic F-127 in the presented study was slightly above CMC, i.e., PLX-0.08_FLU-MIC, it preserved its efficacy in enhancing fluconazole activity, as described in the section "Antifungal 2.3".It indicated that the formation of PLX micelles did not affect neither the antifungal activity of fluconazole nor the investigated properties of polymers.Moreover, the drug−micelle interactions at 5.0% w/ v 35 did not affect the activity of the drug.
In the presented study, individual strains were vulnerable to even small amounts of Pluronic (slightly above CMC) loaded with fluconazole, manifesting itself in a decrease of MIC value in the broth microdilution method.A more pronounced effect of statistical significance was observed in the case of a higher Pluronic concentration (5.0% w/v) with incorporated fluconazole.In strains investigated by the cup plate method, the lowest investigated concentration of block polymer (10.0%w/v) loaded with fluconazole led to an increased inhibition zone of yeast growth.In strains in which the inhibition zone did not appear, we assume that the applied concentrations of both fluconazole and polymer were not sufficient to attenuate the enhanced fungal resistance of those strains.Overall, the highest investigated polymer concentration (15−25% w/v) did not cause the largest inhibition of fungi growth when all the results derived from the cup plate method were considered.This could be explained by the increased viscosity under experimental conditions.
As the addition of fluconazole to Pluronic F-127 led to the attenuated growth of the investigated yeasts when compared to neat fluconazole, the hypothesis for the reason for the observed enhanced fluconazole activity has been evaluated in several ways.The subsequent steps of the study are summarized in Figure 7. First, we have investigated whether fluconazole assisted by verapamil, a known example of the efflux pump inhibitor, would result in decreased yeast growth.A decrease in MIC value was observed in some strains after treatment with verapamil and fluconazole, meaning the inhibited yeast growth was resulting from pump inhibition.Similar outcomes for fluconazole used with Pluronic could indicate a similar mechanism for both components, i.e., Pluronic and verapamil.Next, we evaluated the main mechanism of resistance among the investigated strains.The gene expression data suggested that the expression of genes encoding efflux pumps is increased (CDR1 and CDR2).Thus, we have concluded that Pluronic might be responsible for targeting the efflux pumps and disturbing their function.This might lead to enhanced activity of fluconazole as it is not rinsed from the cell by nonfunctioning membrane transporters.If, on the contrary, the overexpression of ERG11 genes had been observed, it would have been an indication of the other mechanism of interference of Pluronic with fungal function.
In addition, we investigated the possible attenuated efflux and potential membrane permeability upon the addition of Pluronic by fluorescence microscopy.Two assays were involved, differing by the reference substance (verapamil and amphotericin B, accordingly), exhibiting various mechanisms of action (inhibition of efflux pumps and increased permeability of the fungal membranes, respectively).We observed the incorporation of Nile red, a substrate for efflux pumps, within the nontreated cells.The amount of stained cells upon incubation with Pluronic and verapamil was much lower in comparison to the treated cells, which could be explained by their disruptive effect on the membrane transporters.Propidium iodide, characterized by crossing damaged membranes, has accumulated within the cells when treated with Pluronic, however, to a lesser extent when compared to amphotericin B. In both approaches, the various amounts of stained cells (when cells were treated with Pluronic F-127 vs nontreated) indicated either impaired function of efflux pumps or attenuated integrity of the cell membrane.Therefore, the results might be interpreted as a mixed mode of action presented by the investigated polymer, resulting in enhanced activity when applied with fluconazole.
Taking into account the antifungal activity of FLU assisted by PLX presented in our study, we conclude that this block copolymer should be regarded as a valuable excipient when designing therapeutics against Candida infections.
Our results indicate that block copolymer itself does not cause ≥50% inhibition of yeast growth when absorbance of treated cells compared to nontreated control cells.Similarly, no inhibition zone is produced by the neat polymer.However, due to its surface-active properties, it may trigger the structural reorganization of the wall or membrane of the yeast or interrupt the activity of the efflux pumps, similarly to the mechanism reported for some types of membrane trans-porters. 16Therefore, a combination of PLX and FLU can result in synergistic antifungal activity that may arise from enhanced FLU penetration into the fungi or its retention in the cytosol.
CONCLUSIONS
Pluronic F-127-based formulations loaded with fluconazole increased the in vitro antifungal efficacy of FLU against resistant Candida strains.The absorbance measured in the broth microdilution method that corresponded to the yeast growth decreased significantly upon the addition of Pluronic F-127.These results agreed with the increased inhibition zone observed in the cup plate method in formulations containing this block copolymer.The highest efficacy of FLU was observed in the presence of 5.0 and 10.0% w/v of Pluronic F-127, as shown in the microbiological studies.Eight out of 15 investigated C. glabrata strains exhibited significant increases in the expression of genes encoding efflux pumps, and the subsequent seven tested C. glabrata strains displayed a higher expression level of CDR1 and CDR2 genes, indicating that Pluronic-based formulations affected the efflux mechanism.It was in line with the study of efflux pumps using verapamil that suggested that both Pluronic and verapamil might share a similar mechanism of action.The attenuated efflux and membrane integrity upon Pluronic treatment were visualized by fluorescence microscopy.Pluronic F-127 at concentrations of 20.0 and 25.0% w/v caused clear sol−gel transitions above 25 and 20 °C, respectively, in contrast to 10−15% w/v of polymer content characterized by no such phase transition.The presence of Pluronic F-127 enhanced the solubility of fluconazole, which was stable in the micellar solution upon storage for over 84 days at room temperature.
To the best of our knowledge, this is the first attempt to study the influence of Pluronic-based formulations on Candida growth.Not only the enhanced antifungal activity of fluconazole when assisted by block copolymer was observed, but also the potential mechanism of action of Pluronic F-127, comprising an impact on membrane transporters in fungal cells and membrane integrity, was presented in the study.Therefore, the results might facilitate the development of novel fluconazole formulations in light of the small amount of marketed fluconazole topical products.
MATERIALS AND METHODS
5.1.Chemicals.Pluronic F-127, RPMI-1640 medium with the addition of L-glutamine and without sodium bicarbonate, PBS (phosphate-buffered saline), propidium iodide, and amphotericin B were purchased from Sigma-Aldrich (USA).Nile red was obtained from Santa Cruz Biotechnology (USA).Fluconazole and verapamil hydrochloride were obtained from Pol-Aura (Poland).Sterile, purified water was purchased from Polpharma (Poland).3-morpholinopropanesulfonic acid (MOPS) was purchased from J&K Scientific (China).Mueller Hinton LAB-AGAR with glucose and methylene blue were both purchased from Biomaxima (Poland).Sabouraud dextrose agar with chloramphenicol was purchased from Liofilchem (Italy).Sabouraud dextrose liquid medium was purchased from Oxoid (Great Britain).Tryptic soy broth was bought from Biomaxima (Poland).Zymo Research�YeaStar RNA Kit was obtained from TK Biotech (Poland).The primers were ordered from Genomed (Poland).The Tran-Scriba Kit and the RT HS-PCR Mix Probe were supplied by A&A Biotechnology (Poland).
Candida strains.
Thirty Candida strains were involved in the study, including 3 from the American type culture collection (C.krusei ATCC 6258, C. albicans ATCC MYA-574, and C. albicans ATCC 64124) 37 and 27 clinical isolates from the collection of the Department of Pharmaceutical Microbiology and Parasitology at Wroclaw Medical University (C.albicans, C. glabrata, and C. tropicalis, see detailed list in Tables 3 and 4).C. albicans was chosen as the most frequent cause of severe fungal infections, while C. glabrata was selected as frequently developing multidrug resistance as a result of its haploid genome. 38,39Among the investigated strains, 21 were classified as resistant to fluconazole, whereas 8 strains were identified as susceptible to increased exposure based on EUCAST Antifungal Clinical Breakpoints. 40In the case of C. albicans and C. tropicalis, MIC values >4 mg/L indicated resistant strains. 40C. glabrata strains characterized by MIC values ≤16 mg/L were classified as susceptible to exposure, while those with MIC values >16 mg/L were classified as resistant. 40C. krusei ATCC 6258 was selected as the quality control strain 40 and is classified as naturally resistant to fluconazole.C. albicans ATCC MYA-574 strain overexpresses the ABC transporter genes CDR1 and CDR2 that encode ATP-dependent efflux pumps, while C. albicans ATCC 64124 has mutations in the ERG11 genes that affect the FLU binding to its target protein. 41−49 MIC values evaluated for the investigated strains were consistent with the previously published data. 50,51The strains were stored in tryptic soy broth with the addition of 15% glycerol and kept at −80 °C.Before experiments, strains were grown on Sabouraud dextrose agar at 35 °C for 24 h.Afterward, they were suspended in RPMI 1640 medium in double strength to a cell density of 0.5 McFarland.
Pluronic F-127 and Fluconazole
Formulations.Pluronic F-127/fluconazole formulations were obtained by a method called "direct dissolution", namely, dissolving the polymer in an aqueous solvent and adding the pharmaceutically active substance.The active substance is either in the form of a preprepared solution or added directly in solid form to the polymer solution. 52or the broth microdilution method, stock solutions of fluconazole and Pluronic F-127 were prepared in sterile purified water and then mixed in order to obtain a 2-fold concentrated solution (0.16% w/v and 10.0% w/v Pluronic F-127 loaded with 1−256 mg/L, 0.0028−0.84mM fluconazole) added into each well of a 96-well microtiter plate (series FLU-MIC, PLX-0.08,PLX-0.08_FLU-MIC,PLX-5, and PLX-5_FLU-MIC).
For the cup plate method and rheological studies, fluconazole and Pluronic F-127 solutions were prepared in sterile purified water, while Pluronic F-127 solutions loaded with fluconazole were prepared by the addition of fluconazole solution in sterile purified water at concentrations of 4.1, 8.For the drug content test, 10 mg/L fluconazole exceeding the maximum solubility within 5.0% w/v of Pluronic F-127 was added to the polymer solution in sterile purified water (at concentrations of 0.08, 0.1, 0.15, 0.5, 1.0, and 5.0% w/v), and the solution was stirred at room temperature for 24 h and then filtered using a syringe filter with a 0.20 μm pore size, resulting in the series PLX-0.08_FLU-S,PLX-0.1_FLU-S,PLX-0.15_FLU-S,PLX-0.5_FLU-S,PLX-1_FLU-S, and PLX-5_FLU-S.
Rheological Studies.
The rheological properties of Pluronic F-127 solutions (blank and loaded with fluconazole) were analyzed by the rotational rheometer Brookfield RVDV-III+ using cones CP40 and CP51 in a controlled shear rate mode. 23The temperature in the sample cup was controlled by a circulating water bath.For each experiment, 0.5 mL of the sample was used.An appropriate type of cone and rotational speed were selected for each of the series, taking into account their different viscosity ranges and the measurement limitations of the rheometer.The temperature of the phase transition and the temperature coefficient were evaluated based on viscosity vs temperature profiles.The spindle was rotated at 40 5.5.High-Performance Liquid Chromatography.The stability of fluconazole loaded into Pluronic F-127 solutions was analyzed according to the USP 32 monograph 53 using the 1260 Infinity (Agilent Technologies) HPLC system equipped with a 260 nm detector and the Zorbax column SB-C18 (4.6 × 150 mm) with 5 μm packing.The column temperature was 40 °C.A mixture of water and acetonitrile (80:20) was involved as a mobile phase in isocratic elution.The flow rate was 0.5 mL per minute.The injected volume of the sample was 20 μL.The retention time of fluconazole was 7 min.Samples were stored at room temperature for 84 days and filtered using a syringe filter with a 0.20 μm pore size diameter prior to each measurement.
5.6.MIC Determination.MIC determination was conducted according to the European Committee on Antimicrobial Susceptibility Testing (EUCAST) standards involving the broth microdilution method. 54RPMI 1640 medium with the addition of L-glutamine without sodium bicarbonate buffered to pH 7.0 was prepared in double strength to allow for 1:1 dilution throughout all experiments (final concentration 0.165 M MOPS and 2% glucose). 54In each experiment, control tests were involved, including RPMI 1640 medium without drug, polymer, and yeasts (sterility control) and suspension of tested yeasts in RPMI 1640 medium without drug and polymer (growth control).To examine whether the addition of polymer might inhibit yeast growth, control growth wells comprised suspensions of tested yeasts in RPMI 1640 medium with block polymer (series PLX-0.08 and PLX-5).50 μL of drug solution or drug loaded into Pluronic F-127 solution was added to one well of a 96-well microtiter plate, and then 50 μL of RPMI 1640 medium with yeasts was placed.After inoculation, plates were incubated for 24 h at 35 °C.Absorbance within wells was measured at 530 nm using a MultiScan Go Spectrophotometer (Thermo Fisher Scientific).The MIC values of fluconazole alone and of all combinations were read as the lowest concentration that caused growth inhibition by ≥50% in comparison to nontreated control cells.The final concentration of fluconazole in each well was in the range of 0.5−16 mg/L (0.0016−0.052 mM for susceptible Candida strains) and 4−128 mg/L (0.013−0.42 mM for resistant Candida strains) comprising the FLU-MIC series, whereas the final concentration of block polymer in the micellar series throughout all the wells was 0.08% w/v (series PLX-0.08_FLU-MIC) and 5.0% w/v (series PLX-5_FLU-MIC), respectively.The final concentration of yeasts in each well was approximately 1.5 × 10 5 cfu/mL.Every experiment was conducted at least three times.5.7.Investigation of Efflux Pumps with Verapamil.Verapamil is known as an inhibitor of ABC efflux pumps in the cell membranes of fungi.Therefore, the MIC determination assay was also performed in its presence to observe any potential changes in MIC values upon the addition of verapamil.Stock solutions of verapamil hydrochloride in sterile, purified water were prepared.It was added to the suspension of tested yeasts in RPMI 1640 medium with fluconazole, and the final concentration of 100 μM of verapamil in a well was obtained (series VER_FLU-MIC) according to the protocol described in section "MIC Determination".To verify whether neat verapamil might inhibit yeast growth, control growth wells comprised of suspensions of tested yeasts in RPMI 1640 medium with verapamil (series VER) were used.A decrease in the MIC value due to the presence of verapamil may indicate the resistance mechanism correlated to overexpression of ABC and/or MDR efflux pumps. 55 To investigate whether the addition of polymer might inhibit yeast growth, subsequent wells for control growth were filled with Pluronic F-127 solutions without the addition of fluconazole (PLX-10, PLX-15, PLX-20, and PLX-25).After inoculation, plates were incubated for 24 h at 35 °C.The diameter of the zone of growth inhibition was measured.Every experiment was conducted at least three times.
5.9.Studies on the Kinetics of Antifungal Activity.Studies on the kinetics of the antifungal activity were performed for the formulations with the best outcomes evaluated based on results from broth microdilution and cup plate methods (with a content of 5% w/v and 10% w/v of Pluronic F-127, series FLU-MIC, PLX-5_FLU-MIC, and PLX-10_FLU-MIC, respectively) according to the procedure described in the section "MIC Determination".The growth kinetics was investigated for the representative strains, C. krusei ATCC 6258, C. albicans 3057, C. glabrata 2586, C. glabrata 2738, and C. glabrata 2853.The absorbance mode was set for the reading every 20 min at 530 nm under constant shaking.It was programmed to take seventy-three individual measurements in total over a 24 h period.5.10.Microscopic Imaging in Phase Contrast.Particular wells with cultivated fungi in the presence and absence of fluconazole after 24 h at 35 °C on the 96-well microtiter plates were observed in phase contrast using a fluorescence microscope, an Olympus IX53.Final concentration of Pluronic F-127 in the studied wells was 0.08% and 5.0% w/v.Images were acquired with the Olympus cellSens imaging software.
5.11.Imaging by Fluorescence Microscopy.Cellular efflux of Candida species and the potential permeability of the fungal membranes were examined based on the method described by Iyer et al. with slight modifications. 25C. albicans 3057, C. glabrata 2586, C. glabrata 2738, and C. glabrata 2853 were selected as representatives for the study.Yeasts were cultured on Sabouraud agar plates at 35 °C for 24 h.Thereafter, they were suspended in a Sabouraud broth and incubated at 35 °C overnight with constant shaking.The density of all strains was diluted in Sabouraud broth to an OD 600 value of 0.02 in 1 mL (in order to obtain the final density of 0.01 that corresponds to the yeasts density in each well in broth microdilution method, see section "MIC Determination").Afterward, strains were cultured with agitation at 35 °C for ca.3−4 h until the exponential phase was reached.
In the experiment assessing the cellular efflux, three subcultures were set for each strain, i.e., stained, stained in the presence of a reference efflux inhibitor, verapamil, and stained in the presence of a studied substance (Pluronic F-127).Either the compound solution or sterile water (in the control subculture) was added to yeast suspension to allow for 1:1 dilution and incubated for either 20 min (verapamil) or 24 h (Pluronic F-127) at 35 °C under constant agitation.The final concentration of verapamil was 100 μM, and 10% w/v of Pluronic F-127.A stock solution of Nile red in DMSO at a concentration of 3.5 mM was added to each of the incubated samples to a final concentration of 7 μM and incubated for 20 min at 35 °C under constant agitation.Each subculture was transferred to an Eppendorf tube and centrifuged for 1 min at 14,000 rpm.The supernatant was removed, and the pellet was resuspended in 100 μL of PBS.
In the experiment assessing the potential permeability of the fungal membranes, three subcultures were set for each strain, i.e., stained, stained in the presence of a reference substance, amphotericin B, and stained in the presence of a studied substance (Pluronic F-127).Either the compound solution or sterile water (in the control subculture) was added to the yeast suspension to allow for 1:1 dilution and incubated for either 1 h (amphotericin B) or 24 h (Pluronic F-127) at 35 °C under constant agitation.The final concentration of amphotericin B was 2 μg/mL, and 10% w/v of Pluronic F-127.Each subculture was transferred to an Eppendorf tube and centrifuged for 1 min at 14,000 rpm.The supernatant was removed, and the pellet was resuspended in 100 μL of PBS.The cells were treated with propidium iodide prior to the imaging, and the final concentration amounted to 1 μM.
A few μL of Candida suspension in PBS was placed on a glass slide and covered with a coverslip.Visualization was performed using a 40× objective lens on the TRITC channel (excitation 544 nm, emission 570 nm) by a fluorescence microscope, an Olympus IX53.Images were acquired with Olympus cellSens imaging software.5.12.Gene Expression Analysis.Gene expression analysis was performed based on the method described by Szweda et al. with slight modifications. 58Yeasts were cultured on Sabouraud agar plates at 35 °C for 24 h.Two single colonies of each of the investigated strains were suspended in 4 mL of Sabouraud broth and incubated at 35 °C with constant shaking until an OD 660 value in the range of 1.5−1.9 was obtained.The density of all strains was diluted to a similar OD 660 of approximately 1.4, which corresponds to 3.8 × 10 7 cells.RNA was isolated using the YeaStar RNA Kit.In the first step, yeast cells were centrifuged at 500 g for 2 min.80 μL YR Digestion Buffer and 5 μL Zymolyase were added after the removal of the supernatant.
The suspension was incubated at 35 °C for 1 h, and then 160 μL of YR Lysis Buffer was added, followed by the addition of 245 μL of ethanol 96%.The mixture was transferred to the Zymo-Spin IIICG Column, centrifuged at 15,000g for 30 s, and afterward, flow-through was discarded.The same procedure was repeated twice with 200 μL of RNA wash buffer.In the next step, the column was transferred into a nuclease-free tube, and 60 μL of DNase/RNase-Free Water was added and centrifuged at 15,000g for 30 s.The concentration and purity of RNA were determined by NanoDrop2000C (Thermo Scientific).A reverse transcription reaction was performed with a TranScriba Kit.First, 10 μL of the mixture containing 140 ng of isolated RNA, 1 μL of oligo(dT) 18 , and DNase/RNase-Free Water was incubated at 65 °C for 5 min.Afterward, 4 μL of 5× reaction buffer, 2 μL of dNTP's mix, and 4 μL of TranScriba reverse transcriptase were added, and the mixture was incubated at 42 °C for 60 min.The reaction was terminated by temperature enhancement up to 70 °C for 5 min.The concentration and purity of cDNA were determined by NanoDrop2000C (Thermo Scientific).Analysis of the level expression of target genes ERG11, CDR1, CDR2, and a reference gene URA3 was performed by real-time PCR using the Toptical Gradient 96 Real-Time PCR system (Biometra).Primers and probes were chosen as previously described 58,59 (Supporting Information, Table S1), and solutions in buffer TE at concentrations of 10 μM were prepared.
The mixture for each real-time PCR reaction consisted of 10 μL of the 2× Real Time HS-PCR Master Mix Probe, 0.5 μM each primer solution, 0.25 μM probe solution, and 1 μL of cDNA filled with DNase/RNase-Free Water to the final volume of 20 μL.A 2-fold serial dilution of the pooled cDNA mixture of all investigated strains was used for standard curves, whereas 7 ng from cDNA of each of the studied strains was involved for gene expression analysis.Primer efficiency tests for each of the investigated genes were run in triplicate, while experiments for the evaluation of the level of expression of the studied genes were run in duplicate.The following conditions for amplification were set: initial denaturation at 95 °C for 5 min, 50 cycles of denaturation at 95 °C for 15 s, primer annealing at 59, 59, 62, and 55 °C (respectively for ERG11, CDR1, CDR2, and URA3) for 15 s, and elongation step at 72 °C for 15 s.The temperature transition rate was defined as 20 °C/s.5.13.Statistical Analysis.The results obtained in the experiments were variables on interval and quotient scales.The statistical methods used in the analysis were simple linear regression, nonlinear estimation, and tests to compare means.All results subjected to the statistical analyses performed were mean values (MV) calculated from 6 independent replicates.The precision of the obtained averages was evaluated by determining for each mean two descriptive statistics recommended by Pharmacopoeia XII, namely, standard deviation (SD) and relative standard deviation (RSD = SD/MV).A parametric MANOVA with a posthoc test for multiple comparisons (Fisher's Least Significant Difference test) was used to evaluate differences between MV means.All variables compared using the MANOVA test met the assumptions of normality of distribution and homogeneity of variance.The normality of the distributions of the compared variables was tested with three different statistical tests: the Kolmogorov− Smirnov test, the Lillefors test, and the W. Shapiro−Wilk test.Homogeneity of variance was assessed with the Brown− Forsyth and Levene's tests.Simple linear regression and nonlinear estimation were used to determine correlations between variables.In both cases, the loss function was minimized using the method of least squares.The statistical significance of the performed estimations was assessed by determining Pearson's r 2 correlation coefficients for linear models and R-determination coefficients for nonlinear ones, the statistical significance of which was evaluated by the t-test.
The overall relationships between all the statistically evaluated variables were analyzed and then visualized using multivariate analyses based on dimension reduction using the procedure of decomposing the matrix of results according to singular values: principal component analysis (PCA) and correspondence analysis.The constructed PCA models were estimated using the NIPALS iterative algorithm, with the convergence criterion set at 0.00001 and the maximum number of iterations equal to 100.The number of principal components was determined by determining the maximum predictive ability of Q 2 using the V-fold cross-check method, setting the maximum number of them at V max = 7.The obtained optimal PCA model was graphically visualized in the graph of the two principal components, characterized by the largest percentage contribution to the variance explained by the model (PC1 vs PC2) reduced to two components.The results of the PCA analysis shown in the PC1 and PC2 load graphs made it possible to preselect the variables having the most significant impact on the model built and to select the most significant relationships between them.The variables selected in this way were then subjected to further statistical evaluation.
In all the statistical analyses, the level of significance was adopted as α = 0.05.The statistical analyses were performed using STATISTICA PL 13.3 and Mathematica 10.0.
Livak's method was involved in the analysis of the level of gene expression. 24The method enabled us to establish differences between the expression levels of target genes in the investigated strains and a reference strain.In our study, C. glabrata 1004 from the collection of the Department of Pharmaceutical Microbiology and Parasitology at Wroclaw Medical University with the MIC value for fluconazole equal to 8 mg/L was designated as a reference strain.The level of gene expression of target genes was assessed by comparing the threshold cycle values (C t ) of the amplification of a target gene and a reference gene.If the 2 −ΔΔCT value amounts to 1, it means that the expression level of a target gene in the investigated and reference strains is the same.The 2 −ΔΔCT value in the range between 1 and 2 indicates a higher expression level of a target gene in the investigated strain in comparison to a reference strain, while the value above 2 stands for a significant enhancement of the expression level of the target gene with regard to a reference strain.
Figure 1 .
Figure 1.Consumption of antifungal drugs worldwide 17 (A), Candida mechanisms of resistance to azoles 18 (B), and mechanism of binding Pluronics to the cell membrane 16 (C).
a
For susceptible Candida strains.b For resistant Candida strains.
Figure 2 .
Figure 2. Effect of temperature on the viscosity of the investigated formulations: (a) PLX-20 and (b) PLX-25 with and without FLU addition; mean and standard errors are calculated from three experiments, shear rate was 0.192 s −1 .
.�C. krusei; C. a.�C.albicans; C. g.�C.glabrata; if no inhibition zone was observed, the diameter of the well is indicated in the table.
Figure 6 .
Figure 6. C. glabrata 2586 stained with Nile red and propidium iodide in the absence and presence of reference substances (verapamil and amphotericin B, accordingly) and Pluronic F-127.
Figure 7 .
Figure 7. Enhanced antifungal efficacy of Pluronic F-127-based formulations loaded with fluconazole is correlated with their potential mode of action affecting membrane integrity and the function of membrane transporters.
−57 5 . 8 .
Cup Plate Method.The antifungal activity of Pluronic F-127 at 10−25% w/v concentrations loaded with fluconazole was investigated by the cup plate method.The density of Candida cells inoculated on the surface of Mueller Hinton agar with the addition of glucose and methylene blue was 0.5 McFarland.Fourteen strains resistant to fluconazole and C. krusei ATCC 6258 were involved in the study.Wells with 7 mm diameter were cut with a sterile cork borer in the medium.20 μL of fluconazole solution (series FLU-4−15) or Pluronic F-127 solutions loaded with fluconazole (series PLX-10_FLU-4−15, PLX-15_FLU-4−15, PLX-20_FLU-4−15, PLX-25_FLU-4−15) was placed in each well on an agar plate.The final concentrations of fluconazole in each well were 25, 50, 75, and 90 μg.
Table 2 .
Temperature of the Sol−Gel Phase Transition of the Investigated Formulations Based on PLX-20 and PLX-25 a ). MIC values in the five studied resistant Candida strains (C.krusei ATCC 6258, C. albicans 3057, C. glabrata 2586, C. glabrata 2738, and C. glabrata 2853) also decreased upon the
Table 3 .
Antifungal Activity of Fluconazole, Fluconazole Loaded into Block Polymer Pluronic F-127 at 0.08 and 5.0% w/v, and Fluconazole in the Presence of Verapamil Hydrochloride in the Candida Strains Resistant to Fluconazole
Table 4 .
Antifungal Activity of Fluconazole, Fluconazole Loaded into Block Polymer Pluronic F-127 at 0.08 and 5.0% w/v, and Fluconazole in the Presence of Verapamil Hydrochloride in the Candida Strains Susceptible to the Increased Exposure to Fluconazole a A decrease in MIC values in comparison to fluconazole solutions.
Table 5 .
Antifungal Activity of Fluconazole and Fluconazole Loaded into Pluronic F-127 at 10−20% w/v Concentrations Evaluated by the Cup Plate Method in Resistance to Fluconazole Candida Strains (Mean Values Presented with Standard Deviation) a
Table 6 .
Summary of the Gene Expression Analysis among Resistant Strains Evaluated by Livak's Method 24 (Ranked by Increasing Level Expression of CDR1 and CDR2 Genes) in Regards to the Results Retrieved from the Broth Microdilution Method and the Cup Plate Method a
Table 7 .
Excipients with Potential Antifungal Activity against Candida Strains or with Synergistic Effects with Antifungal Drugs | 2023-12-19T06:17:39.755Z | 2023-12-18T00:00:00.000 | {
"year": 2023,
"sha1": "918a1369306e5234a14bff1da4321cd66322c7d0",
"oa_license": "CCBY",
"oa_url": null,
"oa_status": null,
"pdf_src": "PubMedCentral",
"pdf_hash": "9b3244b2e80f0abc3f73522255a0c8e77ae9fa53",
"s2fieldsofstudy": [
"Medicine",
"Materials Science"
],
"extfieldsofstudy": [
"Medicine"
]
} |
60440639 | pes2o/s2orc | v3-fos-license | A Vocoder-free WaveNet Voice Conversion with Non-Parallel Data
In a typical voice conversion system, vocoder is commonly used for speech-to-features analysis and features-to-speech synthesis. However, vocoder can be a source of speech quality degradation. This paper presents a vocoder-free voice conversion approach using WaveNet for non-parallel training data. Instead of dealing with the intermediate features, the proposed approach utilizes the WaveNet to map the Phonetic PosteriorGrams (PPGs) to the waveform samples directly. In this way, we avoid the estimation errors caused by vocoder and feature conversion. Additionally, as PPG is assumed to be speaker independent, the proposed method also reduces the feature mismatch problem in WaveNet vocoder based approaches. Experimental results conducted on the CMU-ARCTIC database show that the proposed approach significantly outperforms the baseline approaches in terms of speech quality.
INTRODUCTION
Voice conversion (VC) aims to modify the source speaker's voice to sound like that of the target speaker without changing the content information. The challenge is to transform the speaker identity while maintaining the speech quality. Various techniques have been proposed to convert spectral feature for speaker identity conversion. Among them, Gaussian mixture model (GMM) [1,2,3,4] is one of the most popular methods, where the spectral feature is transformed by a statistical parametric model. However, it is known the GMM method doesn't capture the spectral details thus suffers from over-smoothing problem [3,5]. To address these problems, frequency warping [5,6,7,8] and exemplar based methods [9,10] are also studied. More recently, with good regression performance, neural network methods are widely used in VC task, e.g. deep neural network (DNN) [11,12,13], long short-term memory (LSTM) [14] and generative adversarial networks (GAN) [15,16].
Despite the research progress, the quality of converted speech varies at run-time. One reason is that most of the This paper is submitted to INTERSPEECH 2019. existing techniques perform the speaker identity conversion and speech reconstruction on the intermediate features analyzed by parametric vocoders. Conventional parametric vocoders (STRAIGHT [17] and WORLD [18]) are designed based on certain assumptions, e.g. source filter model, time invariant linear filter. Additionally, to simplify mathematical formulation of the parametric model, some information, e.g. the phase information, are usually discarded. As a result, the artifacts are introduced in both speech-to-features analysis stage and features-to-speech synthesis stage. To address the features-to-speech synthesis issue a WaveNet vocoder [19,20,21,22] is proposed to directly estimate the time domain waveform samples conditioned on input features. Its effectiveness has been demonstrated in several voice conversion studies [23,24,25,26] to replace the traditional vocoders for high quality speech generation. However, these approaches continue to suffer from the feature mismatch problem between the training and generation of WaveNet vocoder. As a result, undesired noise-like signals are observed in the WaveNet generated speech as reported in [23,24,27].
In this paper, we introduce a vocoder-free voice conversion approach using WaveNet for non-parallel training data, where the traditional parametric vocoder is not required for either intermediate spectral feature extraction or speech reconstruction. Inspired by [28,29], the proposed method first encodeds a speech signal into speaker independent (SI) feature representations, e.g. Phonetic PosteriorGrams (PPG) [28,29]. Then, the WaveNet is trained to predict the corresponding time-domain speech signals with the SI features as the local conditioning. At run-time, the same SI features extracted by given speech are used to drive the WaveNet to generate the converted speech. Note that, the conversion model is trained between the SI features and the corresponding time-domain speech signals of the same speaker. Hence, the parallel data is not required for the proposed method.
This paper makes two main contributions, • Without using the parametric vocoder, the proposed approach prevents the feature extraction and speech reconstruction errors arising from the parametric vocoder; •
VOICE CONVERSION WITH WAVENET VOCODER
In this section, we discuss the advantages and limitations of the WaveNet vocoder based voice conversion techniques.
WaveNet Vocoder
WaveNet vocoder [20] is a conditional WaveNet [19]. It can reconstruct the time-domain audio signals conditioned on the acoustic features extracted from traditional vocoders, e.g. aperiodicity, f0 and spectral features. Given a waveform sequence x = [x 0 , x 1 , ..., x T ] and the additional local conditioning input h, WaveNet vocoder can model the conditional distribution p(x|h) as follows: In order to model the long-range temporal dependencies of audio samples, an architecture based on dilated causal convolutions and a gated activation unit is proposed. Deep residual learning framework is also utilized to speed up the convergence and train a deep model (e.g. 30 layers). For the i-th residual block, the gated activation function is expressed as: where * and • denote the convolution and element-wise product operator respectively. W and V are the trainable convolution filters, f and g denote the filter and gate, respectively. The WaveNet vocoder has been adopted in voice conversion tasks [23,24,25,26] to replace the traditional vocoders for high quality speech generation. One of the successful example is proposed in [23], where the WaveNet vocoder is integrated in a GMM based VC framework. Fig. 1 (a) and (b) show its conversion model and WaveNet vocoder training processes, while Fig. 1 (c) shows its conversion process. During training, two models are built. The GMM model is trained between the aligned source and target feature pairs for feature conversion. While, a WaveNet vocoder is trained with the acoustic features extracted from original target speech as the local conditioning input for speech generation. At run-time, the acoustic features extracted by the traditional vocoders are first converted by GMM VC model. The converted features are then used as the additional input of the WaveNet vocoder to generate the converted speech.
The Limitations
While the WaveNet vocoder based VC is able to generate high quality speech, unstable problems of converted speech generation are reported in recent studies [23,24,27]. This is because the converted features used in run-time generation are very different to the original target features used for training, which results in the noise-like signals or irregular impulses in some speech segments [27].
WAVENET APPROACH TO VOICE CONVERSION
Phonetic PosteriorGrams (PPG) based voice conversion [28,29] has been proposed to model the relationship between the PPG features to the corresponding acoustic features. PPG is a sequence of probability vectors estimated with an automatic speech recognition (ASR) system. As the ASR system is designed to generate the outputs invariant to the input speaker, the PPG feature is considered to be speaker independent. Hence, it can be easily applied for voice conversion with non-parallel data. In this paper, we investigate the effectiveness of using PPG as a local conditioning input of a WaveNet for vocoder free voice conversion. The proposed method does not rely on the intermediate features for speaker identity conversion. Moreover, as PPG feature is considered to be speaker independent, the proposed system reduces the feature mismatch between WaveNet training and run-time stages.
The proposed framework is presented in Fig. 2, which consists of two steps: (a) WaveNet conversion model training and (b) run-time conversion. The details will be described as follows. Fig. 2(a) shows the WaveNet conversion model training process. Given speech data of target speaker, we first extract PPGs L ∈ R D×N , where, D and N are the feature dimension and frame number respectively. In order to control the prosody of generated speech, f 0 and voiced/unvoiced flag (vuv) features are also extracted, denoted as F0 ∈ R 1×N and F vuv ∈ R 1×N , respectively. To facility the WaveNet training, the PPGs, f 0 and vuv are extended to match the temporal resolution of the time domain signals, denoted as L ∈ R D×T , F0 ∈ R 1×T and F vuv ∈ R 1×T . Then, the local conditioning input h in Eq.(2) can be expressed as h = [ L , F0 , F vuv ] . At run-time (see Fig. 2(b)), given a source speech, we first extract the PPG, f 0 and vuv features. A linear transformation is applied on the extracted f 0, expressed as: where µ x and σ x are the mean and variance of the input source speech sample's f 0 in logarithmic domain, respectively. µ y and σ y are the mean and variance of the target speaker's f 0 in logarithmic domain over all training samples. f 0 y is the converted f 0 of the target speaker. Then we adjust the temporal resolution of the PPG, converted f 0 and vuv features and feed them into the trained WaveNet conversion model to generate converted speech.
Database and feature extraction
The voice conversion experiments were conducted on the CMU-ARCTIC database [30]. Four speakers were selected consisting of two male speakers, bdl and rms, and two female speakers, slt and clb. Intra-gender and inter-gender conversions were conducted between following pairs: rms to bdl (M2M), clb to slt (F2F), clb to bdl (F2M) and rms to slt (M2F). 500 utterances were used for training, another 20 non-overlap utterances of each speaker were used for evaluation. WORLD vocoder [18] was used to extract the 513dimensional spectrum, 1-dimensional aperiodicity coefficients and F 0 with 5 ms frame step. Then 40-dimensional MCCs were calculated from the spectrum using Speech Signal Processing Toolkit (SPTK) 1 . The 42-dimensional phonetic posteriorgram (PPG) features were extracted by the PPG extractor trained on the Wall Street Journal corpus (WSJ) [31]. The detailed information can be found in [29]. All the audio files were resampled at 16 kHz. 1 https://sourceforge.net/projects/sp-tk/
Baselines and setup
The details of reference systems and the proposed vocoder free voice conversion methods were introduced as follows.
Reference Systems
• GMM-WORLD: We implemented the joint-density Gaussian mixture model with maximum likelihood parameter conversion [2] for feature conversion. The WORLD vocoder was used for speech generation. The source and target MCC features were aligned using dynamic time warping (DTW) [32]. Both static and its dynamic features were used in this implementation. The mixtures number of GMM is set to 128.
• GMM(GV)-WORLD: We use the same setting as GMM-WORLD, and the converted MCC features were enhaced by GV processing as proposed in [33].
• GMM-WaveNet: We use the same setting as GMM-WORLD with WaveNet vocoder for speech generation.
• GMM(GV)-WaveNet: We use the same setting as GMM(GV)-WORLD with WaveNet vocoder for speech generation.
The Proposed Vocoder-Free VC
• WaveNet-PPG: The proposed WaveNet based voice conversion system with non-parallel data. The 42-dimensional PPG was used as the local condition of the WaveNet.
• WaveNet-VC: The proposed WaveNet based voice conversion system with non-parallel data. The 42-dimensional PPG, voiced/unvoiced flag and converted f 0 were used as the local condition of the WaveNet. In total, the feature dimension was 44.
We trained the WaveNet vocoder and WaveNet conversion models for each target speaker. Both WaveNet vocoder and WaveNet conversion models shared the same network architecture. The WaveNet consisted of 3 dilated residual blocks. Each residual block contained of 10 dilated causal convolution layers. In each block, the dilation started from 1 and exponentially increased by a factor of 2. The hidden units of residual connection and gating layers was set to 512, while the skip connection channels was set to 256. The networks were trained using the Adam optimization method with a constant learning rate of 0.0001. The mini batch size was 15,000 samples and the training steps was set to 200,000. The waveform sample values were encoded by 16 bits µ-law.
Objective evaluation
We conducted objective evaluation to assess the effectiveness of WaveNet-VC approach. The Root Mean Square Error (RMSE) was employed as the objective measure the distortion between target and converted speech. Magnitude features were extracted every 5ms with a window of 25ms. For j th frame, the RMSE was calculated as: are the i th magnitude features of target and converted speech, respectively. F is the total number of the frequency bins. A lower RMSE indicates the smaller distortion. Table 1 shows the RMSE results for all the baseline methods. Firstly, we examine the effect of the f 0 and voiced/unvoiced flag as a additional condition for WaveNet voice conversion. It is observed that WaveNet-VC consistently outperforms the WaveNet-PPG in both intra-and inter-gender conversions.
Then, we further compare the performance of WaveNet-VC with other baseline methods. We observe that the WaveNet-VC performs similar to the WaveNet vocoder baselines, with averaged RMSEs over all the testing pairs of 55.88 dB, 54.83 and 56.53 dB respectively. The systems using WORLD vocoder outperform those with WaveNet vocoder. GMM-WORLD achieves the lowest RMSE of 46.89 dB.
Note that, the objective metric evaluates the spectral distortion that reflects the how close the generated voice is to the target speech. However, it is an indirect measurement. Typically, speech generated by traditional vocoders give a lower objective measure than that of WaveNet [20,22].
Subjective evaluation
AB preference tests and XAB tests were conducted to assess the speech quality and speaker similarity respectively. In AB preference tests, each paired samples A and B were randomly selected from the proposed method and one of the baseline methods, respectively. Each listener was asked to choose the sample with better quality. While, in XAB preference tests, X indicated the reference target sample, A and B were the converted samples randomly selected from the comparison methods. Noted that X, A and B have the same language content. The listeners were asked to listen to the samples, then decided A and B which is closer to the reference sample or no preference. For each test, 20 sample pairs were randomly selected from the 80 paired samples. 10 subjects participated in each tests. Only the WaveNet vocoder based VC baselines, GMM-WaveNet and GMM(GV)-WaveNet were included in the listening tests. Fig. 3. Results of quality preference tests with 95% confidence intervals for different methods.
The subjective results of quality preference tests are presented in Fig. 3. The results showed in Fig. 3 (a) suggests that the speech quality of WaveNet-VC significantly outperforms that of WaveNet-PPG. Similar results are also observed in Fig. 3 (b) and Fig. 3 (b), which suggest that the proposed WaveNet-VC significantly outperforms GMM-WaveNet and GMM(GV)-WaveNet in terms of speech quality. The subjective results of speaker identity are presented in Fig. 4. We observe in Fig. 4 (a) that WaveNet-VC consistently significantly outperforms WaveNet-PPG in terms of similarity. While, in the experiments of WaveNet-VC vs. GMM-WaveNet and WaveNet-VC vs. GMM(GV)-WaveNet (see Fig. 4 (b) and Fig. 4 (c)), the identification rates fall into each others confidence intervals. This indicates that they are not significantly different in terms of speaker identity.
CONCLUSIONS
This paper presents a vocoder-free voice conversion approach using WaveNet for non-parallel data. The proposed approach does not rely on the vocoder features for conversion, which reduces the feature mismatch problem in WaveNet vocoder based approaches. Experiment results show that the WaveNet-VC significantly outperforms the baseline methods in terms of quality, while maintain the speaker identity. | 2019-02-11T02:36:41.000Z | 2019-02-11T00:00:00.000 | {
"year": 2019,
"sha1": "c29614e2f3e14387e6be203b92d6d66860a948ba",
"oa_license": null,
"oa_url": null,
"oa_status": null,
"pdf_src": "Arxiv",
"pdf_hash": "c29614e2f3e14387e6be203b92d6d66860a948ba",
"s2fieldsofstudy": [
"Computer Science"
],
"extfieldsofstudy": [
"Engineering",
"Computer Science"
]
} |
250097990 | pes2o/s2orc | v3-fos-license | Synthesis and antibacterial activity of colloidal silver prepared by electrochemical method
Abstract Study was conducted on the electrochemical method of preparing colloidal silver at low DC voltage. A natural stabilizer was used as a carob bean gum from the carob bean. Using a UV-Vis spectrophotometer and a transmission electron microscope (TEM), colloidal silver (silver nanoparticles) was characterized. It is possible to produce colloidal silver of satisfactory quality by means of this electrochemical method. Silver nanoparticles have exhibited surface plasmon resonance as determined by UV-vis analysis. Based on the TEM analysis, the silver nanoparticles appear to be spherical except in sample E1, where they appear rod-shaped and micrometer-sized. Other three samples had particle sizes ranging from 23–52 nm. In a qualitative analysis of colloidal silver solutions, two types of antibacterial activity were observed. At all dilutions, some colloidal solutions showed better antibacterial activity against gram-positive bacteria (S. aureus, S. pyogenes, E. faecalis) while others displayed better antibacterial activity against gram-negative bacteria (S. enteridis, P. aeruginosa, E. coli).
Silver nanoparticles are particularly interesting because of their unique properties that can be used for antimicrobial applications, in biosensor materials, composite fibers, cryogenic superconducting materials, etc. Many studies have investigated the antibacterial efficacy of silver nanoparticles and the possibility of commercial application in the field of health care and medicine (Gurunathan, Choi, & Kim, 2018;Kasithevar, Periakaruppan, Muthupandian, & Mohan, 2017). Currently, the appearance of multidrug-resistant bacteria represents an increasing worldwide problem (Shrivastava, Shrivastava, & Ramasamy, 2018). Therefore, there is an urgent need to find an effective drug to treat infectious diseases (Kraker, Stewardson, & S. Harbarth, 2016). Given this fact, colloidal silver gained renewed interest (Barras, Aussel, & Ezraty, 2018;Tran et al., 2017). Colloidal silver can significantly reduce the duration and severity of many bacterial infections (Said et al., 2014). The main characteristic of colloidal silver is the high microbicidal capacity of silver ions in water at a very low concentration (Baranowski, 1995). Recent studies have found that interactions between silver ions and thiol groups play an important role in the inactivation of bacteria (Gupta & Silver, 1998). Colloidal silver is not yet understood exactly how it acts as a bactericide. Silver nanoparticles are believed to bind to the surface of membranes of cells, causing changes in properties such as permeability and respiration (Yliniemi et al., 2008).
Synthesis of silver nanoparticles can be carried out by various reduction methods such as chemical reduction, photoreduction, electrochemical reduction methods, or thermal decomposition. Factors affecting the synthesis of silver nanoparticles are the temperature of the solution, the concentration of the precursor, type of reducing agent, reaction time, and cleanliness of the glassware. Chemical reduction involves the reduction of a silver salt with a reducing agent in the presence of a stabilizer. The most commonly used methods for the preparation of silver nanoparticles in an aqueous solution are the reduction of silver nitrate with reducing agents such as borohydride (Barbir, Dabi c, & Mehe s, 2019), citrate (Gakiya-Teruya, Palomino-Marcelo, & Rodriguez-Reyes, 2018), or ascorbic acid (Chekin & Ghasemi, 2014). Stabilizers are substances used in the synthesis of nanoparticles to prevent their attachment by electrostatic or steric repulsion. In the case of silver nanoparticles, the most commonly used stabilizers are citrate, polyvinylpyrrolidone (PVP), and surfactin.
The electrochemical method is an alternative in the preparation of silver nanoparticles due to lower reaction temperatures, simple equipment, controllable products obtained, and short reaction times. Electrochemical methods have used metallic silver to prepare silver nanoparticles using PVP (Khaydarov, Khaydarov, Gapurova, Estrin, & Scheper, 2009). Among the main advantages of this method are the high purity of the end products, as well as their size and shape, which can be controlled by adjusting the current density without the need for expensive equipment and a vacuum (Rodr ıguez-S anchez, Blanco, & L opez-Quintela, 2000). Additionally, using the electrochemical method is environmentally friendly as it does not require the use of toxic reduction agents. Most of the products produced by this method are sold under the name colloidal silver. Colloidal silver is composed of silver ions and particles. Typical colloidal silver is 85% ionic silver and 15% silver particles. Silver ions obtained by electrolysis are often referred to as dissolved silver. Since most of the silver in colloidal silver is in the form of dissolved silver, it would be more technically correct to call them silver solutions (Haider & Mahdi, 2015).
In this work, the synthesis of colloidal silver was carried out by the low voltage electrochemical method DC. Carob bean gum synthesized from carob bean was used as a stabilizer. Antimicrobial activity of prepared colloidal silver solutions against six microorganisms: Staphylococcus aureus, Escherichia coli, Salmonella enteritidis, Enterococcus faecalis, Pseudomonas aeruginosa, and Streptococcus pyogenes has been investigated.
Preparation of carob bean gum
Carob beans, which come from the island of Solta, Croatia, are immersed in boiling water for 2 hours. After the seeds have softened, they are peeled and placed in an oven at 105 C. After drying, the carob seeds are crushed and brought to heat in high purity water at a ratio of 1:197. Heating is done at a temperature of 70 C for half an hour, with intense stirring at a speed of 600 rpm. The solution was filtered and the filtrate was treated with isopropanol in the ratio of 2:1 with stirring for about 5 minutes and then the solution was left for 12 hours to precipitate the locust bean gum. The slurry obtained from the solution was taken out and dried at room temperature, after which the carob bean gum was crushed.
Preparation of colloidal silver
In experimenting, two electrodes (anode and cathode) of high purity silver (99.99% Ag) in the form of a wire with a diameter of 3 mm were used. The electrodes were placed at a distance of 2 cm in a cell containing distilled water with a specific conductivity of 4.82 lS cm À1 and a carob bean gum as a stabilizer. Calculating the current and determining the electrode's surface area is essential prior to starting the electrolysis. The electrodes are connected with wires to a low voltage generator DC (power supply DC) and the required current is set. The magnetic stirrer is switched on and moderate stirring is set in the cell (300 rpm). Electrolysis was carried out for 100 minutes at room temperature and during the process silver concentration (TDS) was measured every 10 minutes using a conductometer Metler Toledo Seven Compact S230, Switzerland. The resulting colloidal silver solution was then stored in the dark until characterization of the silver nanoparticles.
Characterization of colloidal silver
The absorbance of the colloids formed was determined using UV-Vis spectrophotometer Analytik Jena SPECORDV R 200, Germany, in the wavelength range of 300 to 600 nm. For colloidal silver, the absorbance is in the range of 380 to 420 nm. The size and shape of silver nanoparticles were determined using a transmission electron microscope (TEM) Carl Zeiss EM10A, Germany.
Determination of antimicrobial activity of colloidal silver
The microorganisms used for the test were Staphylococcus aureus ATCC 25923, Escherichia coli ATCC 25922, Salmonella enteritidis ATCC 13076, Enterococcus faecalis ATCC 29212, Pseudomonas aeruginosa ATCC 10145, and Streptococcus pyogenes ATCC 13615. During this experiment, microorganisms were subcultured into Muller Hinton Broth (MHB) at 37 C for 18 h and their concentration was diluted from 10 8 to 10 3 cfu mL À1 in twofold MHB. For each sample of microorganism, 100 mL of colloidal silver solution (4 samples) and 100 mL of MHB were added to microplates (Lkhagvajav, Yaşa, Çelik, Koizhaiganova, & Sari, 2011;Sarkar, Jana, Samanta, & Mostafa, 2007). At 37 C, the microplates were then incubated for 24 hours. Each microplate was subsequently treated with 40 mL of TTC (2,3,5-Triphenyl tetrazolium chloride). A color change in TTC in a microplate from colorless to red-pink, an indication that it is positive (þ) (Eaton, Clesceri, Greenberg, & Franson, 2005). Table 1 shows the silver concentration values for four colloidal silver samples obtained by the electrochemical method under different reaction conditions. The designation E1 refers to a sample of colloidal silver obtained using a power supply DC as a current source. The designation E2 refers to a sample also obtained using a power supply DC with 5% by weight of carob gum as stabilizers. E3 denotes a sample of colloidal silver obtained using a low voltage generator DC, and E4 denotes a sample obtained using a generator with the addition of 5% by weight of carob gum.
Results and discussion
From Table 1, it can be seen that the silver concentration increases from the beginning to the end of electrolysis, which is expected for all four samples. After 100 minutes of electrolysis, the silver concentration ranged from 19 to 26 mg L À1 . A sudden jump in silver concentration occurs at 60 minutes for sample E1 and 40 minutes for sample E2. For the remaining two samples with the use of a generator, the increase in concentration was uniform.
UV-Vis spectroscopy is a very important and simple technique to confirm the formation of silver nanoparticles. The absorption spectrum of silver colloids was obtained in the range of 300-600 nm. The formation of silver nanoparticles during the electrochemical process is evident from the change in color of the solution from colorless to gray, which can be seen in Figure 1.
Metallic nanoparticles have free electrons, which lead to the formation of an absorption band of the surface plasmon resonance due to the mutual oscillations of the electrons in resonance with the light wave. The appearance of peaks in the region of 420 nm shows that silver nanoparticles possess surface plasmon resonance, as shown in Figure 2. From the plots, very low intensities and perturbations on the absorption curves in the region of wavelengths around 300 nm can be seen. This behavior can be attributed to the insufficient time of electrolysis and the formation of colloidal silver.
Previous studies indicate that silver colloids are characterized by UV-Vis spectrogram peaks in the wavelength range from 410 to 450 nm, and it was found that there is a relationship between the width of the absorption peak at 50% intensity and the particle size (Mulfinger et al., 2007). It is evident from the results that both the choice of source DC and the addition of a stabilizer do not affect the absorption maximum, as the values for all four samples are between 410-420 nm.
The shape and size of the obtained silver nanoparticles (Figures 3-5) were determined by TEM analysis. The scale on TEM images is 500 nm. Based on the analysis, it can be seen that all the particles except sample E1 are spherical (Figure 3). Moreover, the particles in sample E1 had a size in the micrometer range, so it can be said that no silver nanoparticles were formed. Silver nanoparticles were formed in the other three samples. The effect of stabilizer on the formation of silver nanoparticles can be observed in sample E2, where silver nanoparticles were formed, unlike sample E1. The average silver particle size in sample E2 was 23 nm, in E3 39 nm, and in sample E4 52 nm. Silver nanoparticles were unaffected by the addition of carob bean gum to sample E4. Figure 5 and Table 2 show the results of antimicrobial testing with colloidal silver solutions on different microorganisms.
The specific mechanism of antimicrobial activity of silver nanoparticles is still unclear and has not been fully explained. Nanoparticles containing silver can often release silver ions (Ag þ ), which can play a role in their antimicrobial activity. In order for silver to exhibit antimicrobial activities, positively charged Ag þ should essentially be ionized (Bapat et al., 2018;Klueh, Wagner, Kelly, Johnson, & Bryers, 2000). As Ag þ ions form complexes with nucleic acids, they actually interact specifically with nucleotides of these nucleic acids, unlike phosphate groups. Gram-positive and gram-negative bacteria have differences in membrane composition that may have a particular effect on the mechanism of action of prepared colloidal silver solutions ( Table 2). The membranes of gram-positive bacteria are composed primarily of anionic phosphatidylglycerol and cardiolipin lipids, whereas the membranes of gram-negative bacteria are composed primarily of zwitterionic lipid phosphatidylethanolamine (Epand & Epand, 2009;Pedron et al., 2017). Silver nanoparticles were found to be antibacterial against gram-positive and gramnegative bacteria by many scientists (Lara, Ayala-Nunez, Ixtepan Turrent, & Rodriguez Padilla, 2010;Shrivastava et al., 2007;Yogesha, Rabinal, & Ananthamurthy, 2012). Based on their findings, several factors affect antibacterial effectiveness. Generally, silver particles of a certain size are the most important. Silver particles measuring 9 nm in diameter are more effective at killing bacteria than silver particles measuring 64 nm in diameter in colloidal solutions, as reported by Lok et al. (2007) In their study, Sintubin and colleagues found that a different diameter of silver nanoparticle affects its antibacterial activity since the reduction in size results in a higher concentration of active silver (Sintubin et al., 2011).
In the Ostwald-Freundlich equation, silver nanoparticle shape and size play a role in producing Ag þ ions. As a result of their large surface areas, finer and spherical silver nanoparticles will promote Ag þ ion formation (Shanmuganathan et al., 2018). During our experiments, we found that the sample E1, in which the silver particles are not spherical and are microscopic, exhibited the least antimicrobial activity. By aggregating silver nanoparticles, less Ag þ ions are released, meaning this issue could be resolved by using stabilizers (carob gum), which modify the surface of silver nanoparticles. 10 8 10 7 10 6 10 5 10 4 10 3 10 8 10 7 10 6 10 5 10 4 10 3 Colloidal silver solutions S. aureus À À À À À À À À À À À À E1 À À À À À À À À À À À E2 À À À À À À À À À À À À E3 À À À À À À À À À À À À E4 Researchers have found that silver nanoparticles are more likely to kill gram-negative bacteria compared to gram-positive bacteria. Compared to gram-positive bacteria, gram-negative bacteria have thinner cell walls. Silver nanoparticles are less able to diffuse into the cellular environment because of the thick cell wall (Duval, Gouyau, & Lamouroux, 2019). In our study, we were unable to confirm this claim but obtained the opposite results. It follows that the method of obtaining silver nanoparticles is important for antimicrobial activity. The next step of the research will be to determine the Minimum Inhibitory Concentration (MIC) for all tested microorganisms.
Conclusions
It can be concluded from the presented results and discussion that electrochemical method was successful in synthesizing colloidal silver solutions. Silver nanoparticle surface plasmon resonance peaks at wavelengths 410-420 nm were detected by UV-Vis analysis. Silver nanoparticles were found to be spherical except in sample E1. The particles in samples E1 are rod-shaped and micrometer-sized. In the three other samples, the particle size ranged from 23 to 52 nm. Antimicrobial activities of colloidal silver solutions were analyzed quantitatively, revealing both positive and negative results. Some colloidal solutions had better antimicrobial activity on gram-positive (S. aureus, S. pyogenes, E. faecalis) and some on gram-negative (S. enteridis, P. aeruginosa, E. coli) bacteria at all dilutions.
Disclosure statement
No potential conflict of interest was reported by the authors. | 2022-06-29T15:11:07.806Z | 2022-06-26T00:00:00.000 | {
"year": 2022,
"sha1": "bec9fcf605f8d300b66fa6b332495ca307891e55",
"oa_license": "CCBY",
"oa_url": "https://www.tandfonline.com/doi/pdf/10.1080/25765299.2022.2093034?needAccess=true",
"oa_status": "GOLD",
"pdf_src": "TaylorAndFrancis",
"pdf_hash": "11ec639e766555351297655aa56f467a9abbf0fc",
"s2fieldsofstudy": [
"Materials Science"
],
"extfieldsofstudy": []
} |
258059705 | pes2o/s2orc | v3-fos-license | An extremal problem and inequalities for entire functions of exponential type
We study two variations of the classical one-delta problem for entire functions of exponential type, known also as the Carath\'eodory--Fej\'er--Tur\'an problem. The first variation imposes the additional requirement that the function is radially decreasing while the second one is a generalization which involves derivatives of the entire function. Various interesting inequalities, inspired by results due to Duffin and Schaeffer, Landau, and Hardy and Littlewood, are also established.
Introduction
In the present note we study some extremal problems concerning certain quantities over specific families of entire functions of exponential type. For ∆ > 0, we say that an entire function G : C → C has exponential type at most 2π∆ if, for all ε > 0, there exists a positive constant C ε such that |G(z)| ≤ C ε e (2π∆+ε)|z| , for all z ∈ C.
We adopt the usual convention that an entire function f : C → C is said to be real if its restriction to R is real-valued, as well as, that the function g * (z) is defined by g * (z) = g(z). For f, g ∈ L 1 (R) we denote by f * g their convolution, which is defined by (f * g)(x) = ∞ −∞ f (y)g(x − y) dy.
1.1. The one-delta problem. The classical one-delta problem is to determine the infimum where the family F consists of real entire functions f : C → C of exponential type at most 2π such that f ∈ L 1 (R), and f (x) ≥ 0 for all x ∈ R. This is a classical problem, and several of its variations are named after Carathéodory, Fejér and Turán. We refer to [9,11,14,16] for comprehensive information about its history and for some recent contributions. It is known that A = 1, and the unique extremal solution of the one-delta problem is the Fejér kernel, given by (1.1) space P W 2 such that f (z) = g(z)g * (z). Here, P W 2 is the subspace of L 2 (R) consisting of entire functions of exponential type at most π. Therefore, the one-delta problem can also be stated as finding (1.2) and clearly B = 1, too. Other L p −variations of this problem have also been studied in [4,6,13]. Note that (1.2) can be stated in yet another alternative way as follows: the inequality holds for every g ∈ P W 2 such that g(0) = 1, and (1.3) reduces to an equality if and only if g(z) = sin πz πz .
Our main goal is to study some natural variations of each of the above versions of the one-delta problem.
1.2.
Monotone-delta problem. The monotone-delta problem is to find where the family F 1 consists of real entire functions f : C → C of exponential type at most 2π, such that f ∈ L 1 (R), f (x) ≥ 0 for all x ∈ R, and f is radially decreasing, that is, f is increasing on (−∞, 0) and decreasing on (0, ∞). In the following theorem, we present some qualitative and quantitative information about this problem.
The following statements about the monotone-delta problem hold: (a) There exists an even function F ∈ F 1 with F (0) = 1 that extremizes (1.4).
(b) All the zeros of any even extremizer F lie in the set S = {z ∈ C : |Re z| > |Im z| > 0}.
We conjecture that the upper bound in part (c) is sharp in the sense that the first four significant digits of A 1 are those shown above. One of the reasons for this claim is that our proof of part (c) of Theorem 1 is constructive. We construct concrete examples for which the value 1.2771. . . is attained. To obtain A 1 , we first reformulate the monotone-delta problem (see Lemma 6 below) to the one of determining the infimum where the family F 2 consists of entire functions h : C → C of exponential type at most π such that xh ∈ L 2 (R) and |h(x)| = |h(−x)|. Then we find an explicit example of a function h 0 ∈ F 2 (see (2.9)). For this h 0 , we compute explicitly the quotient in (1.5), which turns out to be 1.2771 . . .. Despite that h 0 is not the extremal function for (1.5), our conjecture is that the value 1.2771 . . . is so close to the infimum A 1 , that they differ only in the decimal digits after the fourth one. In Section 3.2 we give a deeper discussion of numerical issues, and a sharper conjecture for the value of A 1 .
The monotone-delta problem has also been considered in R d , for d ≥ 2. In [5], using techniques from the theory of de Branges spaces, the authors found the exact solution of the monotone-delta problem when d is 2 even. Nonetheless, the authors state that the case when d is odd seems more subtle and remains open in general.
Despite that Lemma 6 below provides an integral representation of any function in F 1 , the first interesting explicit example of a function in this class we constructed was based on the classical method of Sonin, which was itself invented with the intention to obtain information about the monotonicity of the successive relative minima and maxima of certain oscillatory solutions of ordinary differential equations (see [18,Section 7.31]).
If g : C → C is a real entire function in P W 2 and satisfies a second-order differential equation of the form y ′′ + (B/x) y ′ + Cy = 0, with B, C > 0, Sonin's method suggests to construct the function It is clear that f ∈ F 1 . Moreover f (x) is a "lid" of g 2 (x) in the sense that f (x) ≥ g 2 (x) for every x ∈ R and f interpolates g 2 and possesses inflection points at its local maxima. Figure 1 shows Fejér's kernel K(x) and its lid f (x). 1.3. The one-delta problem with derivatives. The function in (1.6) appears in a classical inequality for entire functions. Duffin and Schaeffer [8, p. 239] proved that if a real entire function g : C → C of exponential type at most π is such that |g(x)| ≤ 1 for all x ∈ R, then (g(x)) 2 + (g ′ (x)) 2 π 2 ≤ 1, for all x ∈ R.
Inspired by this inequality, we prove that specific sums of the L 2 −norms of a function g ∈ P W 2 , normalized by g(0) = 1, and its consecutive derivatives, are bounded from below. Our result may be considered a variation of the one-delta problem where one wishes to minimize sums of L 2 −norms of an entire function and of its derivatives, and reads as follows: Note that when N = 0 and a 0 = 1, we recover the inequality (1.3), which once again shows that the latter is a natural result in the spirit of the one-delta problem. Moreover, choosing the polynomial P(x) = 1+aπ 2 x, we obtain the following corollary.
holds for every g ∈ P W 2 with g(0) = 1 and the unique extremal function for which (1.9) reduces to an equality is Observe that for a = 1/π 2 (1.9) reduces to the following lower bound for the integral of the function in Different choices of the polynomial P(x) allow us to obtain other interesting inequalities.
In particular, letting a → 1/π 2 in (1.10) we obtain which is exactly the L 2 −version of the classical Bernstein inequality that holds for every L p , p ≥ 1 (see [2,Theorem 11.3.3]).
Observe that the Bernstein inequality follows from Theorem 2 if we set P(t) = 1 + ε − t and let ε → 0.
Applying the same reasoning with P(t) = (1 + ε − t) N , we obtain: holds for every function of exponential type at most σ such that f ∈ L 2 (R). In particular, for N = 2, The latter is a curious result that resembles some classical ones, due to Landau and Hardy and Littlewood.
In 1913, Landau [15] proved that if f is a real function, f ∈ C 2 (R), and the inequalities f ∞ ≤ 1 and f ′′ ∞ ≤ 1 for the uniform norms of f and f ′′ on the real line hold, so does f ′ ∞ ≤ √ 2.
Hardy and Littlewood [12,Theorem 6] proved that, if y and y ′′ are in Moreover, the constant 4 is the best possible. The equality is attained if and only if y(x) = c Y (ax), where c and a are real constants and Theorem 7 in [12] states that, under the same requirements, the inequality holds with equality as before, but with a = 1.
Proof of Theorem 1
For f ∈ L 1 (R), we normalize the Fourier transform f of f as , we see that we may restrict our search for the infimum (1.4) to the even functions in F 1 . To prove the existence of an extremizer, we follow an argument in [4], which consists in showing that a certain weak limit is a viable candidate. Consider an extremizing function of exponential type at most 2π. Moreover, using the weak convergence, we conclude that, for every x ∈ R, Therefore, F is an even radially decreasing function such that F (x) ≥ 0 and F (0) = 1. Finally, using Fatou's lemma, we conclude that F ∈ L 1 (R).
Proof of b).
Let F be an even extremizer with F (0) = 1. Clearly, it has no real zeros. Indeed, since F is real, nonnegative and decreasing on the positive real axis, if it vanishes at x 0 > 0, it does for all x > x 0 which is impossible because F is entire and F (0) = 1. A similar argument shows that F cannot vanish at a negative x 0 . Therefore, all the zeros of F satisfy |Im z| > 0. Now, assume that F has a zero at z = ib, for b ∈ R. Since F is real-valued, it also has z = −ib as a zero. Consider the entire function Note that G(0) = 1 and G ∈ F 1 . Since we get a contradiction. Therefore, all the zeros of F satisfy |Re z| > 0. Now, assume that z = a + ib is a zero of F with |b| ≥ |a| > 0. Since F is real-valued and even, we have that z = a − ib, z = −a + ib, and z = −a − ib are also zeros. Note that all these zeros are different. Then, the entire function is in F 1 , and using that |b| ≥ |a|, it is easy to see that which gives a contradiction. We conclude that |b| < |a|.
Representation lemma.
The following lemma gives a representation for any even function in F 1 .
Lemma 6. If f ∈ F 1 , then it can be represented in R in the form where h : C → C is an entire function of exponential type at most π such that |h(x)| = |h(−x)| for all x ∈ R, and xh ∈ L 2 (R). Conversely, if f is a function of the form (2.1), then it has an analytic extension to C which is an even function in F 1 .
Proof. Let f ∈ F 1 be even. Then clearly lim x→±∞ x f (x) = 0. Integration by parts yields By the Plancherel-Pólya theorem, f ′ has exponential type 2π and so does −zf ′ (z). The monotonicity requirement implies −xf ′ (x) ≥ 0 for all x ∈ R, and then (2.2) yields −xf ′ ∈ L 1 (R). From the Krein decomposition theorem [1, p. 154], it follows that −zf ′ (z) = g(z)g * (z) for some g ∈ P W 2 . Moreover, since f attains its maximum at x = 0, then f ′ (0) = 0. Defining h(z) = g(z)/z, we rewrite the latter in the form where h is an entire function of exponential type at most π and xh ∈ L 2 (R). Since f ′ is odd, then |h(x)| = |h(−x)| for x ∈ R. Finally, integrating (2.3) appropriately, we arrive at (2.1). Conversely, assume the representation (2.1). Note that f has an analytic extension on C (also denoted by f ) of the form where [0, z] denotes the straight segment connecting 0 and z. Since h is an entire function of exponential type at most π, f is an entire function of exponential type at most 2π. From (2.1) it follows that lim which yields f ∈ L 1 (R).
Proof of c).
Since K(x) is the unique extremal solution for the one-delta problem, we have that 1 < A 1 .
On the other hand, as mentioned in the introduction, from Lemma 6, we can reformulate the monotone-delta problem as the one to determine where the family F 2 consists of those entire functions h : C → C of exponential type at most π such that xh ∈ L 2 (R) and |h(x)| = |h(−x)|.
We now transform this optimization problem over F 2 into another unrestricted, smooth optimization problem over R d+1 , so that we may construct functions h in a systematic way with standard numerical optimization methods. For this purpose, we make a couple of helpful observations. First, note that if h ∈ F 2 then h ∈ L 1 (R). In fact, by the Cauchy-Schwarz inequality, we have Therefore, h is continuous in R, and in particular h(±1/2) = 0. Denoting I = [−1/2, 1/2] we have that supp h ⊂ I. Therefore, by the Stone-Weierstrass theorem we may approximate h uniformly by a polynomial times χ I , where χ I denotes the characteristic function of the interval I.
With the previous observations in mind, we consider functions of the form is a polynomial of degree d. Note that the factor 1 4 − x 2 means that h (±1/2) = 0. Denoting a = (a 0 , a 1 , . . . , a d ) ∈ R d+1 , the infimum in (2.5), restricted to this class, becomes For all d ≤ 20 and 0 ≤ i ≤ d, it is easy to see by direct computation of f i that xf i ∈ L 2 (R), so that h = a · (f 0 , . . . , f d ) ∈ F 2 for all a ∈ R d+1 . The matrices N and D may be computed explicitly for a given d, and this is then a smooth optimization problem over R d+1 . Solving it numerically for d = 2, we find where P (x) = 242x 6 + 3001x 4 + 4176, see Figure 2. Additionally, we solve (2.7) for all d ≤ 20 and observe that, as the degree d increases, the sequence A 1,d decreases very slowly, showing only a tiny improvement from 1.27717... only in the fifth decimal digit. More precisely, we obtain A 1 ≤ 1.27713505... with those much more detailed calculations performed with large degree d of the polynomials g. In Section 3.2, we will show some tables with the results of these computations (see Table 1), and compare the results with those of another numerical approach. In Figure 3 and Figure 4, we plot the functions 4 h 0 and 600 91 h 0 , respectively, where, since h 0 (0) = 91 600 and h 0 (0) = 1 4 , we renormalized the plots accordingly.
3. Some functions in F 1 3.1. The lid function. In this subsection, we apply Sonin's method to construct a nice sequence of functions in F 1 . For any positive real numbers B and C, consider the differential equation Let y = g, g : R → R be a solution of the equation (3.1). The lid of g 2 is the function defined by Note that f (x) ≥ 0 for all x ∈ R, and This implies that f is radially decreasing. Moreover, if we suppose that the solution g has an analytic extension on C of exponential type at most π, and g ∈ L 2 (R), we conclude that f ∈ F 1 . Let us show some examples of lids. For α > 0, consider the Bessel function of the first kind of order α, which is defined by Let us remark some properties of the Bessel functions mentioned in [18, Section 1.71]. It is known (see [18,Equation 1.71.3]) that J α satisfies the differential equation Now, define the function g α (z) = J α (πz) (πz) α . A straightforward change of variables in (3.3) shows that g α satisfies the differential equation The function g α is an even entire function of exponential type π. Moreover, using the decay of J α (see [18, Equations 1.71.10 and 1.71.11] we see that g α ∈ L 2 (R). Therefore, inserting g α in (3.2) we actually construct the lid of g 2 α , with B = 2α + 1 and C = π 2 . In the particular case α = 1/2 we known that and therefore is the lid of K(x). Straightforward calculations show that the Fourier transform of f 1/2 is f 1/2 (ξ) = max{1 − |ξ|, 0} + 1 π 2 (g ′ 1/2 ) 2 (ξ). (3.4) Then the Fourier transform -convolution de Margan type law yields where we used the fact that g 1/2 (x) = χ I (x). This, together with (3.4), implies In particular, this example allows us to obtain the bound
3.2.
An L 2 −computational approach. Another natural approach for constructing functions in F 2 (and therefore in F 1 ), and computationally solving (2.5), starts by finding an orthonormal system for the space L 2 (R, x 2 dx). Note that F 2 is a Hilbert space with the inner product For odd integers k ≥ 1, we define the even functions and note that h k ∈ F 2 for all odd integers k ≥ 1. Gorbachev [10] previously considered this family of functions to obtain fine numerical estimates for other Fourier extremal problems, and the first and third authors [7] have also used this family for similar purposes in related extremal problems introduced by Carneiro, Milinovich, and Soundararajan [4]. Regarding this system, we can say the following: is a complete orthonormal system in the closed subspace {h ∈ F 2 : h is even.}.
Proof. Note that, if h ∈ F 2 is even, then xh ∈ L 2 (R) is odd. Furthermore, we have that (xh k )(t) = i(−1) k+1 2 √ 2 sin(πkt)χ I (t) =: s k (t). (3.5) To see this, since s k ∈ L 1 (R) ∩ L 2 (R), we may compute s k in a straightforward manner to verify that s k (x) = −xh k (x), and then we conclude (3.5) by Fourier inversion in L 2 (R). Now consider the operator defined by T h(t) := (xh)(t)e πit . By Plancherel's theorem and the Paley-Wiener theorem, T is a linear isometry, that is, f, g F2 = T f, T g L 2 (I) . Therefore, for odd positive integers k and j, we find that where δ kj = 1 if k = j, and 0 otherwise. Here, to compute the inner product over L 2 (I), we may apply the identity 2 sin(πkt) sin(πjt) = cos(π(k − j)t) − cos(π(k + j)t) and use that k ± j is an even integer. This shows that h k is orthonormal.
We now show that it is complete. Let h ∈ F 2 be even, such that h, h k F2 = 0 for all odd positive integers k. We must show that h ≡ 0. First, denote H(t) = (xh)(t), and note that, by Plancherel's theorem and (3.5), the condition h, h k F2 = 0 implies that I H(t) sin(π(2j − 1)t) dt = 0 (3.6) for all positive integers j. Actually, since sin(−x) = − sin x, (3.6) holds for all integers j. Now, since T is an isometry into L 2 (I), by the theory of Fourier series on L 2 (I), it is enough to show that T h, e j L 2 (I) = 0 for all integers j, where e j (t) = e 2πijt . In fact, for an integer j, we have T h, e j L 2 (I) = I H(t)e πit e −2πitj dt The first integral in the last line is 0 since H is odd, and the second integral is 0 by (3.6). Therefore, T h, e j L 2 (I) = 0 for all integers j, and then T h ≡ 0 and h ≡ 0, as desired.
Once we have a complete orthonormal system, we proceed to obtain numerical examples as follows. For a positive integer d, let F 2,d = span {h 2j−1 : 1 ≤ j ≤ d} ⊂ F 2 . Let Q ∈ R d×d be the matrix defined by Then, since h k are orthonormal, one can see that the reciprocal of the infimum in (2.5), when taken over the space F 2,d , satisfies where λ d is the largest eigenvalue (in absolute value) of Q, and the maximum is attained when for a ∈ R d an eigenvector of Q associated to λ d . We calculate the eigensystems numerically for d ≤ 1000.
We find that λ 1000 = 0.783002554..., giving a proof for the bound A 1 ≤ 1/λ 1000 = 1.277135042..., which is only slightly smaller than our example (2.8) in the proof of Theorem 1 -again coinciding in the first four decimal digits. Moreover, this also coincides with the first seven decimal digits given by the polynomials of degree d = 20 that we constructed with the approach in Section 2.4.
In Table 1, we compare the speed of convergence of the two numerical approaches we have presented.
The first approach is described in Section 2.4, with functions h defined as in (2.6), via polynomials g of some degree d. The second approach is the L 2 −approach described in the present section, with functions h defined by (3.7). In both cases, the parameter d is the number of degrees of freedom in the construction of the function h. In both cases, the bounds for A 1 appear to quickly converge to the first few decimal digits, yet we observe that in the polynomial approach, the bound for A 1 seems to converge much faster to more decimal digits with small values of d. Together, all of this gives evidence to the conjecture that the sharp value of A 1 , up to its first 8 significant digits, is Since g(0) = 1, then the extremal function is unique and it is is given by (1.8). | 2023-04-12T01:16:00.075Z | 2023-04-11T00:00:00.000 | {
"year": 2023,
"sha1": "3ae3fde79261a4e6b3970983144779c0d7ce3870",
"oa_license": null,
"oa_url": null,
"oa_status": null,
"pdf_src": "Arxiv",
"pdf_hash": "faa8d659c4913c98dfc0b507140d9c27a6e9b887",
"s2fieldsofstudy": [
"Mathematics"
],
"extfieldsofstudy": [
"Mathematics"
]
} |
235363018 | pes2o/s2orc | v3-fos-license | The African swine fever virus protease pS273R inhibits DNA sensing cGAS-STING pathway by targeting IKKε
ABSTRACT African swine fever virus (ASFV), a large and complex cytoplasmic double-stranded DNA virus, has developed multiple strategies to evade the antiviral innate immune responses. Cytosolic DNA arising from invading ASFV is mainly detected by the cyclic GMP-AMP synthase (cGAS) and then triggers a series of innate immune responses to prevent virus invasion. However, the immune escape mechanism of ASFV remains to be fully clarified. The pS273R of ASFV is a member of the SUMO-1-specific protease family and is crucial for valid virus replication. In this study, we identified pS273R as a suppressor of cGAS-STING pathway mediated type I interferon (IFN) production by ASFV genomic open reading frame screening. The pS273R was further confirmed as an inhibitor of IFN production as well as its downstream antiviral genes in cGAS-STING pathway. Mechanistically, pS273R greatly decreased the cGAS-STING signaling by targeting IKKε but not TBK1, and pS273R was found to disturb the interaction between IKKε and STING through its interaction with IKKε. Further, mutational analyses revealed that pS273R antagonized the cGAS-STING pathway by enzyme catalytic activity, which might affect the IKKε sumoylation state required for the interaction with STING. In summary, our results revealed for the first time that pS273R acts as an obvious negative regulator of cGAS-STING pathway by targeting IKKε via its enzymatic activity, which shows a new immune evasion mechanism of ASFV.
Introduction
African swine fever (ASF), a devastating disease for domestic pigs and wild boar, usually causes an acute hemorrhagic fatal disease with a mortality rate of up to 100% while being asymptomatic in the natural hosts [1,2]. Since African swine fever was initially described in Kenya in the 1920s, it has spread rapidly across many countries in sub-Saharan African, Caribbean, Eastern Europe [3]. In 2017, a large number of outbreaks of ASF occurred in Siberia near to the Russia-China border [4]. The first outbreak of ASF in China was reported on 3 August 2018 [5] where it caused unprecedented disaster for Chinese swine industry and food security due to currently unavailable vaccines in the world. Accordingly, animal slaughter and regional quarantine are the only method for the disease control. African swine fever virus (ASFV) is a large, complex, cytoplasmic double-stranded DNA virus. It is the sole member of the Asfarviridae family, but also the only DNA virus transmitted by arthropod ticks, soft ticks (Ornithodoros moubata) [6]. ASFV is an enveloped virus with icosahedral structure inside which is composed of four concentric layers from the central nucleoid, the core shell, the inner envelope to the outside the icosahedral capsid [7]. The average diameter of the ASFV is approximately 200 nm and genomes vary in length between 170 and 193 kb which depends on the isolates. ASFV contains between 151 and 167 open reading frames (ORFs), which encode structural proteins, viral DNA replication proteins and host defense escape protein etc [8].
The open reading frame (ORF) S273R encodes a 31-kD protein, and the protein pS273R belongs to the SUMO-1-specific protease family with the ability to catalyze the maturation of pp220 and pp62 polyprotein precursors into core shell matrix proteins [9]. The two significant polyproteins, pp220 and pp62, are cleaved by the pS273R protease to produce six primary structural components of the virus particle, among which p37, p34, p40, p150 derived from polyproteins pp220 and p15, p35 derived from polyproteins pp62 [10]. Therefore, the pS273R protease is of great significance for maturation and infectivity of the ASFV particle. The innate immune system is the body's first line of defense against pathogen infection. It utilizes a variety of pattern recognition receptors (PRRs) in cells to recognize and respond to pathogen associated molecular patterns (PAMPs) [11]. Among various PRRs, cyclic GMP-AMP synthase (cGAS) is a cytosolic double-stranded DNA sensor, which senses the presence of cytoplasmic DNA and catalyzes the synthesis of the second messenger cyclic GMP-AMP 2'3'-cGAMP) [12]. Then2'3'-cGAMP binds and activates the stimulator of interferon genes protein (STING), which upon stimulation, transfers from the endoplasmic reticulum (ER) to the trans-Golgi apparatus network (TGN), during which the kinase TBK1 is recruited and autophosphorylated [13]. TBK1 subsequently phosphorylates the transcription factor IRF3, and then IRF3 translocates into nucleus to activate the expression of antiviral type I interferons (IFNs) [14]. The DNA sensing cGAS-STING pathway is the relevant innate immunity for ASFV, however, the immune evasion of this pathway by ASFV has not been resolved. Here, we identified the viral protein responsible for the immune evasion and characterized its mechanism of action.
Plasmid construction and gene mutations
The genomic ORFs of ASFV China 2018/1 (GenBank submission No: MH766894) were codon optimized, synthesized and cloned into p3×FLAG-CMV-7.1 vector using Not I and Sal I sites. The 145 sequence confirmed ORF plasmids were used for screening of modulators of porcine cGAS-STING signaling pathway by ISRE promoter assay (Supplementary Figure S1). The S273R gene was amplified by PCR using Golden Star T6 Super PCR Mix polymerase from plasmid p3×FLAG-CMV-S273R and then was cloned into pCAGGS-HA vector using EcoR I and EcoR V sites. The critical enzyme active sites of pS273R were identified and point mutated using the mutation PCR primers designed by QuickChange Primer Design method (https://www.agi lent.com) (Supplementary Table S1). The mutation PCR was performed with KOD plus neo polymerase and pCAGGS-S273R-HA as the template. The PCR products were transformed into competent DMT E. coli after Dpn I digestion, and the resultant mutants were sequence confirmed. Truncated mutants of the pS273R, including pS273R (Δ1-20) and pS273R (Δ256-273) were obtained by PCR amplification and cloning into pCAGGS-HA using 2×MultiF Seamless Assembly Mix polymerase. Porcine cGAS, STING, IKKε, IKKβ, IRF3 and NF-κB p65 plasmids were constructed and characterized as we reported [15]; human TBK1, IKKε, IRF3-5D plasmids were preserved in our lab. The 3×FLAG-pCMV-sumo-1, 2, and 3 plasmids were all purchased from MiaoLingBio Company (Wuhan, China).
Promoter driven luciferase reporter gene assay
293T cells grown in 96-well plates (3-4 × 10 4 cells/well) were co-transfected by Lipofectamine 2000 with reporter plasmids, ISRE-luc, IFNβ-luc or ELAM (NF-κB)firefly luciferase (Fluc) (10 ng/well) plus Renilla luciferase (Rluc) reporter (0.2 ng/well), with or without the indicated porcine cGAS and STING plasmids as well as porcine S273R or vector control (10-20 ng/well). The total DNA amount for transfection was normalized with control vectors to 50 ng for each well. PAMs grown in 96-well plates were similarly transfected using the Lipofectamine 2000. At 24 h post transfection, cells were harvested and dual-luciferase assays were sequentially performed using the TransDetect Double-Luciferase Reporter Assay Kit. The fold changes were calculated relative to control samples after normalization of Fluc by Rluc.
Flow cytometry, virus TCID50 and plaque assay
PAMs in 6-well plate (8 × 10 5 cells/well) were transfected by Lipofectamine 2000 with porcine S273R expressing plasmids or vector control (1μg/well). About 24 h post-transfection, the transfected cells were infected or not with VSV-GFP (0.001 MOI) or HSV-GFP (0.01 MOI). After infection, the cells were harvested by trypsin digestion and washed 3× with PBS. The levels of GFP virus replication were analyzed by flow cytometry of the cell suspensions.
The supernatants of virus-infected PAMs were collected and tenfold serially diluted in DMEM medium with each dilution four to eight replications. The diluted supernatants were then used to infect Vero cells in 96-well plates for 2 h. The cells were then cultured in DMEM containing 2% FBS and grown at 37°C for 1 day (for VSV) or 2 day for (HSV-1). The cytopathic effects (CPEs) were counted and TCID50 was calculated with Reed-Muench method.
Vero cells seeded into 12-well plates and grown into monolayer were infected by the tenfold serially diluted cell supernatants from VSV infected cells for 2 h. Then, the infected cells were washed with PBS and overlaid by immobilizing medium of 1:1 mixture of warmed 2×DMEM with 4% FBS and a stock solution of heated 1.6% low melting agarose. Two days later, the immobilizing medium were discarded by tipping and cells were fixed and stained with crystal violet cell colony staining solution for 1 h at room temperature. After staining, cells were washed with tap water until the clear plaques appeared. The plaques were counted and photos were taken.
Co-immunoprecipitation and Western blot analysis
For Co-immunoprecipitation, 293T cells in 6-well plate (8 × 10 5 cells/well) were transfected for 48 h, and then cells were harvested and lysed in 500 μL RIPA buffer (50 mM Tris pH 7.2, 150 mM NaCl, 1% sodium deoxycholate, 1% Triton X-100) with protease inhibitors on ice for 30 min followed by centrifugation. The 50 μL cell lysate was saved as input controls and the remained were incubated with indicated antibodies overnight at 4 ℃, and then 30 μL 50% protein A/G bead solution was added for another 2 h. Later, the beads were thoroughly washed five times with RIPA and boiled in 40 μL 2×SDS sample buffer to obtain the elution which subjected to Western blotting. The elution samples together with input controls were boiled at 100 ℃for 5-10 min, separated by 6-10% SDS-polyacrylamide gels, and transferred to PVDF membranes. Membranes were then blocked using 5% nonfat dry milk Tris-buffered saline, with 0.1% Tween-20 (TBST) at room temperature (RT) for 2 h. Next the membranes were sequentially incubated with primary antibodies overnight at 4 ℃ and HRP-conjugated goat antimouse or anti-rabbit IgG for 1 h at RT. Protein signals was visualized and captured by Western blot imaging system (Tanon, Shanghai, China).
ASFV infection and S273R siRNA treatment
Primary PAMs were isolated from the lung lavage fluid of 4-week-old healthy piglets and maintained in RPMI-1640 medium containing 10% FBS. The ASFV genotype II positive samples from diseased pigs were inoculated into primary PAMs for virus culture and appearance of CPEs, and the fifth passaged ASFV was used for subsequent siRNA knockdown experiments. The primary PAMs (1 × 10 6 cells/well) were transfected with S273R siRNA and control siRNA (100 nM each, GenePharma) using Lipofectamine 2000, and 24 h later the transfected cells were infected with 0.01 MOI ASFV for another 72 h. The cells were harvested for RT-qPCR and Western blot analysis, respectively. All ASFV infection experiments were performed in the animal biosafety level 3 (ABSL-3) of Yangzhou University approved by the Ministry of Agriculture and Rural Affairs (-07140020201109-1). The animal experiment was in strict accordance with the Guidance for the Care and Use of Laboratory Animals of Yangzhou University (SYXK(JS)-2021-0026).
Statistical analysis
The data were the representative of three similar experiments and shown as the mean ± SD. The statistical significance was determined by analysis with the software GraphPad Prism 8.0, where p < 0.05 was considered statistically significant as determined by the paired two-tailed t-test analysis. In the figures, "*", "**" and "ns" denote p < 0.05, p < 0.01 and statistically not significant, respectively.
We first confirmed the inhibitory role of pS273R using different promotor assays in transfected 293T cells, and found that co-transfection of pS273R in 293T cells could significantly inhibit ISRE, IFNβ and NF-κB promoter activity induced by porcine cGAS-STING signaling (Figure 1(a)). Co-transfection of porcine cGAS and STING activated downstream IFNβ, ISG56, ISG60 and IL-8 gene transcriptions and the inductions of these gene transcriptions were significantly suppressed by the co-transfected pS273R (Figure 1(b)). Co-expression of porcine cGAS and STING led to the phosphorylation of TBK1 and IRF3, and downstream ISG56 production. The presence of pS273R restrained the phosphorylation of IRF3 and the production of ISG56, but retained the phosphorylation of TBK1 (Figure 1(c)).
In order to explore the effect of pS273R on endogenous porcine cGAS-STING signaling pathway, we used the PAMs stimulated by three different agonists, polydA:dT and 45bp dsDNA to activate cGAS, and2'3'-cGAMP to activate STING. In promoter assays, we found that pS273R significantly inhibited polydA:dT and2'3'-cGAMP activated ISRE and NF-κB promoter activity (Figure 2(a)). In RT-qPCR assay, the downstream gene transcriptions of porcine IFNβ, ISG56, IL-8 mRNA levels induced by polydA:dT (Figure 2(b)), 45bp dsDNA (Figure 2(c)) and 23'-cGAMP (Figure 2(d)) were all significantly inhibited by pS273R. Consistently, we also found that phosphorylation of IRF3 and expression of ISG56 were markedly decreased in PAMs in the presence of pS273R after polydA:dT stimulation (Figure 2(e)) and2'3'-cGAMP stimulation (Figure 2(f)). However, in both cases, the phosphorylation of TBK1 was not affected by pS273R ( Figure 2(e,f)). These results were perfectly in line with those from exogenous porcine cGAS-STING signaling pathway in transfected 293T cells ( Figure 1). In addition, we also checked the effect of pS273R on the IRF3 and NF-κB p65 nuclear translocations. The results showed that in the presence of pS273R,2'3'-cGAMP induced nuclear translocations of both IRF3 (Figure 3(a)) and p65 (Figure 3(b)) were inhibited. These results further consolidated the inhibitory function of pS273R. Taken together, our results clearly suggested that pS273R inhibits porcine DNA sensing cGAS-STING pathway.
ASFV pS273R disturbs the porcine cGAS-STING signaling pathway mediated antiviral function
The innate immune cGAS-STING signaling pathway is capable of sensing virus attack and induces an antiviral response [20]. To determine whether pS273R could disturb cGAS-STING mediated antiviral function, we selected two GFP viruses Herpes Simplex Virus-1 (HSV-1, a DNA virus) and Vesicular Stomatitis Virus (VSV, an RNA virus) to infect PAMs transfected with ASFV pS273R (Figures 4 and 5). The virus replications were examined by fluorescence microscopy, Western blotting, flow cytometry, RT-qPCR and TCID50 assay, respectively (Figures 4 and 5). We found that HSV-1 replicative GFP signals using MOI 0.01 and 0.1 were both enhanced in the presence of pS273R relative to controls under fluorescence microscopy (Figure 4(a)), by Western blotting (Figure 4(b)) and by flow cytometry analysis (Figure 4(c)). Using RT-qPCR, the HSV1 gB gene transcriptions were also upregulated with pS273R, in sharp contrast with the downregulated IFNβ and ISG56 genes (Figure 4(d)). Accordingly, the virus titers in the supernatants of PAMs infected with HSV1 of MOIs 0.01 and 0.1 were both increased relative to the controls in the TCID50 assay (Figure 4(e)).
The cGAS-STING pathway mediates a broad range of antiviral function including anti-DNA virus and anti-RNA virus activity, thus we also tested the role of pS273R during VSV replication ( Figure 5). Similarly, the VSV replicative GFP signals using MOIs 0.001 and 0.01 were both enhanced relative to the controls under fluorescence microscopy ( Figure 5(a)) and by flow cytometry ( Figure 5(b)). In RT-qPCR, the VSV glycoprotein gene transcriptions were upregulated with pS273R ( Figure 5(c)) in sharp contrast with the downregulated IFNβ and ISG56 gene transcriptions (Figure 2 (e,f)). Accordingly, the virus titers in the supernatants of infected PAMs with VSV of MOIs 0.001 and 0.01 were both increased with the S273R compared with the controls by the TCID50 assay ( Figure 5(d)) as well as by the plaque assay ( Figure 5(g)). Collectively, these results demonstrated that pS273R could enhanced virus replications by damaging cGAS-STING mediated antiviral functions.
ASFV pS273R negatively regulates the cGAS-STING signaling pathway by targeting IKKε
To investigate the mechanism of how pS273R suppresses the cGAS-STING signaling pathway, pS273R was first co-transfected with individual signaling molecules of cGAS-STING-IFN pathway including TBK1 (Figure 6(a)), IKKε (Figure 6(b)) or IRF3-5D (Figure 6(c)) into 293T cells and the downstream ISRE, IFNβ and NF-κB promoter activity were examined. The results showed that only IKKε but not TBK1 and IRF3-5D activated promoter activity were inhibited by pS273R (Figure 6(a-c)). The cGAS-STING signaling also activates downstream NF-κB and proinflammatory cytokines even though the strength is much weaker and the associated molecular details are not clear. Thus, the signaling activity of NF-κB relevant molecules IKKβ and p65 were also checked in the presence of pS273R by NF-κB promoter assay, and it turned out that no any inhibitory effect of pS273R was observed ( Figure 6(d,e)). Since only IKKε was affected by pS273R, the IKKε induced downstream gene transcriptions were examined by RT-qPCR, and the results showed that the IKKε induced downstream IFNβ, ISG56 and IL-8 gene levels were inhibited by pS273R in dose-dependent manners (Figure 6(f)), which is consistent with the promoter assays ( Figure 6(b)).
Next, the IKKε-induced signaling was examined by Western blotting (Figure 6(g)). Ectopic IKKε was able to activate the phosphorylations of TBK1 and IRF3 and the downstream ISG56 production as previously reported [21]. In the presence of pS273R, the IRF3 phosphorylation and ISG56 induction were inhibited, but once again TBK1 phosphorylation was not affected (Figure 6(g)). Finally, the IKKε induced antiviral activity was also investigated in the presence of pS273R (Figure 6(h-j)). IKKε exhibited obvious anti-HSV1 activity, however, in the presence of pS273R, the IKKε mediated anti-HSV1 activity was weakened as observed by fluorescence microscopy (Figure 6(h)), shown by RT-qPCR ( Figure 6(l)) and TCID50 assay (Figure 6(j)). Similarly, pS273R promoted the VSV replications by damaging IKKε mediated antiviral function (Sup Figure S2). Collectively, these results illustrated that pS273R inhibited cGAS-STING signaling pathway by targeting IKKε.
ASFV pS273R interacts with IKKε and disturbs the interaction between IKKε and STING
IKKε is a crucial IKK-related kinase that is significant for innate immune signaling, yet the interaction of STING and IKKε is still not clear. We first assessed the interaction between IKKε and STING by Co-IP, and the result showed that there was an obvious interaction between IKKε and STING (Figure 7(a)). Since pS273R targets IKKε and inhibits its activity, the interaction between pS273R and IKKε was also discernable in Co-IP assay (Figure 7(c)). However, there was no interaction between pS273R and STING in Co-IP assay (Figure 7(b)). Consistent results were shown for cellular colocalization between these proteins in PAMs by con- focal microscopy (Figure 7(e)). Obvious co-localization between IKKε and STING, co-localization between IKKε and pS273R, and no co-localization between pS273R and STING were observed (Figure 7(e)). The co-localization of STING and IKKε was so obvious that the formation of puncta was appreciated (Figure 7(e)). However, in the Co-IP assay and in the presence of pS273R, the interaction between IKKε and STING was disturbed in a pS273R dose-dependent manner (Figure 7(d)). These results indicated that ASFV pS273R blocks the interaction between STING and IKKε in a dose-dependent manner to evade host innate immunity.
The protease activity of pS273R is responsible for inhibition of porcine cGAS-STING signaling pathway
A recent study about the structural information of pS273R revealed that a catalytic triad of C232-H168-N187 and the regions spanning amino acids 1-20 and 256-273 of pS273R play essential roles in pS273R enzyme activity [22]. To explore the potential roles of these active sites of pS273R in the inhibitory function, we generated three-point mutants (H168R, N187A and C232S) and two truncated mutants (ΔN1-20 and , with 20 ng IKKβ plasmid and 10 ng pS273R (d), with 20ng p65 plasmid and 10 ng pS273R (e), plus 10 ng ISRE-luc, IFNβ-luc or NF-κB-luc and 0.2 ng pRL-TK plasmids, which were normalized to 50 ng/well by vector 3×FLAG-pCMV. After 24 h, the luciferase activities were detected using Double-Luciferase Reporter Assay. (f,g) pS273R plasmid (400 ng or 800 ng) were co-transfected with 400 ng IKKε into HEK293T cells. After 24 h, the cells were harvested and analyzed by RT-qPCR (f) and Western blotting (g). (h-j) the IKKε plasmid (500 ng) were co-transfected with 500 ng pS273R plasmid or empty vector into 293T cells for 24 h, then the transfected 293T cells were infected with 0.01 MOI or .1 MOI HSV-1-GFP for 36 h, the GFP signals were observed by fluorescence microscopy (H). the infected cells were harvested to measure the HSV-1 gene expression by RT-qPCR (i). the viral titer in the supernatant from HSV-1 infected 293T cells was measured by TCID50 assay (j). (a-c) HEK293T cells in 6-well plates (8 × 10 5 cells/well) were co-transfected with 1 μg each plasmid as indicated for 48 h and then the cells were harvested and subjected for Co-IP using the indicated antibodies and subsequent Western blot analysis. (d) Increasing amounts of pS273R plasmid (1 μg, 2 μg) were co-transfected with 1 μg STING-mCHERRY and 1 μg IKKε-MYC into 293T cells for 48 h and then cells were harvested and immunoprecipitated with anti-mCHERRY antibody and subjected to Western blot analysis. (E) PAM cells in 24-well plates (3 × 10 5 cells/well) were co-transfected with STING-GFP (0.5 μg) and IKKε-MYC (0.5 μg), with STING-GFP (0.5 μg) and pS273R-HA (0.5 μg), with IKKε-MYC (0.5 μg) and pS273R-HA (0.5 μg) for 24 h, and then cells were fixed and examined for cellular co-localization by con-focal microscopy. the co-localizations in multiple vision fields were analyzed using Image J, and the Pearson coefficient values from 10 positive cells were graphed and shown on the right. the value of Pearson coefficient reflects the level of co-localization, with 1 representing 100% co-localization. ΔN256-273) (Figure 8(a)). These enzyme inactive mutants were tested for inhibition of cGAS-STING signaling in various promoter assays. As shown in Figure 8(b), the inhibitory effects of the five mutants disappeared to varying degrees. Furthermore, the inhibitory effects of the five mutants on IKKε activated promoter activity also disappeared (Figure 8(c)). To figure out whether these five mutants could disturb STING interaction with IKKε, the Co-IP of STING and IKKε was performed in the presence of each pS273R mutants, and the results showed that the pS273R mutants lost the ability to disturb the interaction of STING and IKKε (Figure 8(d)), except the mutant H168R that is close to wild-type pS273R, also reflected by the remained inhibition of cGAS-STING mediated ISRE and IFNβ promoter activity (left and middle, Figure 8(b)). The cellular co-localizations between STING and IKKε in the presence of pS273R mutants were examined and the results showed that with the pS273R, the co-localized puncta of STING and IKKε disappeared, whereas in the presence of mutants except H168R, the significant co-localized puncta were maintained (Sup Figure S3). The mutant H168R exhibited kind of difference from other mutants although the reason is not known.
The pS273R mediated IKKε de-sumoylation likely contributes to inhibition of porcine cGAS-STING signaling pathway and replication of ASFV
It appeared that pS273R protease activity plays a critical role in the inhibition of cGAS-STING signaling as well as inhibition of IKKε activity. However, we did not observe the cleavage of IKKε and related signaling proteins, we wondered if pS273R, as a SUMO-1 protease, might affect the sumoylation modification of the target protein and its subsequent function. To test this hypothesis, we detected the IKKε status in the presence of sumo proteins and found that IKKε was sumoylated in the presence of sumo-3 (Figure 9(a)). Whereas pS273R was able to suppress the IKKε sumoylation (Figure 9(a)), as well as cell sumo-2/3 (Figure 9(b)). Secondly, we chose 2-D08, a unique inhibitor of SUMOylation [23,24], to evaluate the hypothesis. 2-D08 was confirmed to be inhibitory to the cell sumo-2/3 (Figure 9(c)). The results showed that 2-D08 could markedly inhibit the IFNβ, ISG56 and ng STING-mCHERRY, plus 10 ng ISRE-luc or IFNβ-luc or NF-κΒ-luc and 0.2 ng pRL-TK plasmid, along with 10 ng pS273R WT or pS273R mutants, which were normalized to 50 ng/well by vector pCAGGS. at 24 h post-transfection, luciferase activities were detected using Double-Luciferase Reporter Assay. (C) HEK293T cells were co-transfected with 20 ng IKKε, plus 10 ng ISRE-luc or IFNβ-luc or NF-κΒ-luc and 0.2 ng pRL-TK plasmid, along with 10 ng pS273R WT or pS273R mutants, which were normalized to 50 ng/well by vector pCAGGS. After 24 h, luciferase activities were measured. (D) Each pS273R mutants (1 μg) and WT were co-transfected with STING-mCHERRY (1 μg) and IKKε-MYC (1 μg) into 293T cells for 48 h, and then the cells were harvested and subjected for Co-IP using anti-mCHERRY antibody and subsequent Western blot analysis. IL-8 mRNA induction levels in dose manners upon either polydA:dT stimulation or2'3'-cGAMP stimulation (Sup Figure S4(a,b)). The inhibitor 2-D08 also inhibited cGAS-STING signaling induced IFNβ and ISG56 mRNA levels in transfected 293T cells (Sup Figure S4(c)). Analogously, 2-D08 was able to inhibit the interaction of IKKε and STING in Co-IP assay in a dose-dependent manner (Figure 9(d)). Additionally, K231 was reported as the only IKKε sumoylation location and the sumoylation of IKKε is the perquisite for IKKε function [25]. We made the IKKε mutant K231R and tested its activity and the interaction with STING. The results showed that K231R mutant exhibited substantial reduction of ISRE and NF-κB activities relative to wild-type IKKε (Sup Figure S5(a)). Further, K231R had impaired interaction with STING in Co-IP assay (Figure 9(e)) and did not interact with pS273R any more (Sup Figure S5(b)).
To test the role of pS273R in the IFN induction in the context of ASFV infection, we treated the primary porcine alveolar macrophages with S273R siRNA before ASFV infection. The results showed that, relative to control siRNA, the ASFV-induced IFNβ and ISG56 gene transcriptions were both upregulated with either of the two S273R siRNAs, accompanied by the decreased viral S273R and B646 L gene transcriptions and p30 protein expression (Figure 9(f,g)). Taken together, these data demonstrate that pS273R likely acts as an inhibitor of IKKε sumoylation to obstruct the interaction of IKKε and STING, suppress the cGAS-STING signaling and thus evade antiviral function.
Discussion
Many viruses are equipped to promote the replications and establish persistent infections in host. Therefore, viruses must evolved multiple mechanisms to antagonize the antiviral immune responses to achieve the immune evasion. Likewise, ASFV has developed multiple strategies to evade innate immune system, including modulation of type I IFN responses, inhibition of apoptosis, inhibition of autophagy and inhibition of inflammatory responses etc, with few of the corresponding viral proteins identified [26,27]. However, ASFV long double-stranded DNA genome codes for more than 150 proteins of which many are non-essential for viral replication in cells and their potential roles and the mechanism of action in evading the host's defenses are still unknown. Here, we revealed for the first time that pS273R plays as a negative regulator in the cGAS-STING pathway via genomic ORF screening and confirming experiments (Figures 1 and 3and Sup Figure S1). We also found that pS273R obviously promotes the replications of both DNA and RNA viruses (Figures 4 and 5), suggesting that pS273R facilitates their replications by damaging the cGAS-STING signaling which has been shown to have broad antiviral function [28]. Surprisingly, our results showed that pS273R inhibits the cGAS-STING signaling pathway by targeting and impairing the function of IKKε ( Figure 6). In the context of virus infections, pS273R has been observed to disturb IKKε mediated antiviral function ( Figure 6 and Sup Figure S2). The above results suggested that IKKε is likely to play an important role in the cGAS-STING pathway. Indeed, IKKε interacts with STING and the interaction can be disturbed by pS273R that interacts with IKKε but not with STING ( Figure 7). Therefore, our findings support that pS273R certainly acts on IKKε to impair the antiviral function and help virus evade host defense. IKKε has been previously reported to recruit and phosphorylate IRF3/7 in RIG-I-like receptors (RLR) and Toll-like receptors (TLR) signaling pathways [29,30], whereas it has been long believed that TBK1 acts as the only downstream kinase to mediate the cGAS-STING pathway to activate IFN production. However, one recent study showed that IKKε synergizes redundantly with TBK1 to mediated STING activated NF-κB signaling and inflammatory cytokine production, and on the other hand, IKKε also participates in STING activated IFN production even though to a much less degree [31]. The phosphorylation of porcine STING Ser365 (Ser366 in humans) is mediated by active TBK1, but the phosphorylation site of STING by IKKε in the absence of TBK1 is still not clear [31]. Therefore, our findings indicate that IKKε is actually significant for the activation of the STING-dependent DNA sensing pathway, though the mechanism was unclear, which is warranted for future investigation.
Many positive-strand RNA viruses and retroviruses encode polyproteins or polyprotein precursors that are cleaved by viral proteases to form vital components of the virus particle, which are critical for virus maturation and infectivity [32]. The ASFV pS273R protease, as a member of SUMO-1-specific protease family, could accurately cleave two polyproteins, pp220 and pp62, at Gly-Gly-Xaa sites to produce important proteins required for ASFV morphogenesis [10]. In this study, we found that the protease activity of pS273R is required for inhibition of the interaction of IKKε and STING and subsequent signaling ( Figure 8); therefore, it is possible that pS273R protease inhibits the interaction of STING and IKKε by cleaving the IKKε. However, neither IKKε nor STING expression was changed in the presence of pS273R. The ASFV pS273R protease consists of arm domain and core domain. While the arm domain is unique, the core domain shares similar structure fold and active triad to those of other SUMO-1 cysteine proteases, which contain Ubl (ubiquitin-like protein)-specific proteases (Ulp) in yeast and sentrin-specific proteases (SENP) in mammals [33,34]. These proteases mainly catalyze the deconjugation of SUMOylated proteins; and accordingly, it is well known that some viral proteases have the ability to deconjugate certain small protein modifications to interrupt host innate immune responses [7,35,36]. Therefore, it is also possible that pS273R protease could inhibit interaction of STING and IKKε by cleaving small protein modification such as sumoylation of IKKε. Indeed, the pS273R inhibits IKKε sumoylation and cell sumo-2/3 (Figure 9(a,b)). Sumoylation inhibitor 2-D08 suppresses the interaction of STING and IKKε, and the cGAS-STING signaling activity, certifying the function of pS273R protease in the cGAS-STING pathway (Figure 9(d) and Sup Figure S4). The mutation of sumoylation site of IKKε decreases its activity and interaction with STING (Figure 9(e) and Sup Figure S5).
The papain-like proteases from coronaviruses and arteriviruses are known to be anchored in double membrane vesicles, exerting deubiquitinating or deISGylating activity on the accessed cell proteins to suppress antiviral innate immunity [37,38]. Previous study showed that pS273R is incorporated in viral particles, localized at viral factory (VF) and weakly scattered throughout cytoplasm [9], which is different from pS273R cellular localization in transfected cells (Figures 3 and 7(e)) of this study. Nevertheless, at late stage of ASFV replication, the VF disrupts and reorganizes many cell organelles [39], and likely IKKε/STING complex is brought proximity to pS273R in VF or in the scattered viral particles in cytoplasm. Thus, our findings are able to match up to pS273R during ASFV infection and reflect the situations of papain-like proteases from coronavirus and arteriviruses. Our study presents clear and relevant evidence that pS273R protease acts as a sumoylation inhibitor and targets IKKε modification to interfere with the immune response. During revision of this paper, the work published by Zhao etc showed the pS273R exerts its protease activity to non-canonically cleave gasdermin D, disturb inflammasome-induced cell pyroptosis and promote ASFV replication [40]. Collectively, all these work explored different aspects of immune evasion by pS273R and broaden our understanding of immune evasion mechanism by ASFV.
In summary, upon infection of ASFV, porcine cytosolic cGAS senses the dsDNA of virus and catalyzes the synthesis of the second messenger cyclic GMP-AMP 2'3'-cGAMP), which binds and activates the adaptor STING. The oligomerized STING transfers from the endoplasmic reticulum (ER) to the trans-Golgi apparatus network (TGN); during the process, the TBK1 and IKKε as well are recruited, phosphorylated and activated. TBK1 and IKKε subsequently phosphorylate the IRF3/7 and activate IKKα/IKKβ, inducing expression of the interferons and inflammatory factors (Figure 9(h)). However, ASFV pS273R could antagonize the cGAS-STING signaling pathway by targeting the IKKε, thus achieving immune escape (Figure 9(h)).
Disclosure statement
No potential conflict of interest was reported by the author(s). | 2021-06-08T13:21:46.650Z | 2021-06-03T00:00:00.000 | {
"year": 2022,
"sha1": "b21806034f7ce4a18d12e67cc7c93e0bc7516bc6",
"oa_license": "CCBY",
"oa_url": "https://www.tandfonline.com/doi/pdf/10.1080/21505594.2022.2065962?needAccess=true",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "f56409eae06f91ac4a136082421cf9739aab4b1e",
"s2fieldsofstudy": [
"Biology"
],
"extfieldsofstudy": [
"Biology"
]
} |
218874551 | pes2o/s2orc | v3-fos-license | Who Are the People Willing to Pay for Physician Home Visits?
Background Since the recently announced Community Care Policy, there has been an opinion that Korea needs to establish an alternative medical model such as physician home visits. This study aimed to assess the need and willingness to pay (WTP) for physician home visits among the community-dwelling Korean older population and to determine the most important factors that influence older adults to decide to use a physician home visit service. Methods A total of 797 people aged 60 years or older who were randomly selected from a nationwide dataset using a multi-stage stratified sampling method answered a questionnaire on the need and WTP for physician home visits. Results A total of 39.3% of participants reported that they would like a physician home visit when they need help. Among older adults who needed physician home visits (n = 313), the WTP amount for physician home visits was 21,982 ± 17,546 KRW. Logit and Tobit regression analyses showed that the higher valuated need and WTP for physician home visits was associated with a lower level of physical/psychosocial functioning measured by EuroQol-five dimensions score (odds ratio [OR], 1.13; 95% confidence interval [CI], 1.01–1.27; P = 0.035) and a higher level of satisfaction when using community-based services such as public health centers (OR, 1.32; 95% CI, 1.02–1.72; P = 0.034), social welfare centers and Gyeong-ro-dang (OR, 1.61; 95% CI, 1.04–2.50; P = 0.033; β = 8.39; standard error, 3.63; P = 0.021). Conclusion This study provides evidence that the decision to pay for a physician home visit service is based upon the complex interactions among an individual's physical and psychosocial functioning, personal experiences of service utilization, and demographic factors. The value for physician home visits should be qualified based on the empirical data of WTP, which comes from a consumer-centered perspective.
INTRODUCTION
Physician home visits were widespread in the western history of medicine until the 19th century and then declined sharply with the formation of a hospital-oriented medical system in the mid-20th century. Over the last 30 years, Aging in Place has been emphasized due to population In recent years, there has been an opinion that Korea needs to establish an alternative medical model based on the community, including a primary geriatric physician system and homebased primary care for homebound older adults with increasing concerns about frailty. 13- 16 The Korean government recently announced the Community Care Policy in 2018, which includes: 1) strengthening community-based health care services, 2) ensuring health rights for vulnerable groups, 3) setting up a discharge pathway from hospital to returning home, and 4) building a community-based public-private partnership system. The role of primary care physicians in community care is to organize and coordinate various outpatient and home care services. An outpatient-oriented medical practice cannot successfully manage chronic diseases without home-based primary care, which enables continuous observation and monitoring of patients, especially for homebound older adults. According to Kim and Jang,15 at least 600,000 older adults (about 10% of the entire population aged 65 and over) were estimated to be homebound, not being able to go freely outside to use local outpatient clinics in Korea. These adults are the medically isolated people who are obviously excluded from the primary care system unless physician home visits are available in their hometown. Along this line, Lee 17 stimulated the vision and rhetoric to Korean primary physicians, "Now it's time to see the community -outside the clinic." However, compared to these social demands, there are only a handful of studies on physician home visits 17 and particularly little research on the valuation and willingness to pay (WTP) from the perspective of the older population.
The process of selecting medical services for users consists of several steps, such as problem recognition, information search, evaluation of alternatives, choice of medical institutions, and post-use behavior. 18 This process cannot necessarily be done rationally. Various criteria can be applied according to the subjective viewpoint of medical users. To enhance medical consumers' satisfaction, efforts to provide desired services based on empirical data from the perspective of consumers should be a basic and core strategy. 19 In addition, the cost of physician home visits, that is, who pays for them and how much they cost, is essential for achieving feasibility and quality of service. 20 WTP is a contingent valuation method that is commonly used to assess how patients value the personal utility of new healthcare technologies and services. WTP assesses the maximum amount of money an individual would pay for the health intervention and still consider him/herself better off. 21 In order to advance beyond the descriptive commentary on why patients need physician home visits, we aimed to conduct a nation-wide opinion survey from the perspective of consumers.
Our aim was to determine the need and WTP for physician home visits among the communitydwelling older population in Korea. Our second aim was to examine the hypothesis concerning the most important factors that influence older adults to make the choice to use a physician home visit service. Through the above analyses, our intention was to gain knowledge on which subgroup of older people needs physician home visits the most. Before outlining and testing the analytic model, we briefly discuss the hypothesis of the study.
What is the process through which people make a choice to pay for a physician home visit service? In the present study, we developed a theoretical model of the WTP for physician home visits based on a literature review of medical sociology and service marketing. 15,18-20 According to Fig. 1, the process is comprised of four stages: illness experience, help seeking, service utilization, and post-use behavior. An individual who gets sick will make a choice as to which type of service is more appropriate to relieve his or her symptoms through information gathering. This process might be influenced by a variety of factors, which include not only the current health problem (e.g., physical and psychosocial functioning) but also past experiences of healthcare service utilization (e.g., personal judgement, satisfaction). During the process, the value of physician home visits will be created personally. Finally, the individual will choose one of two types of healthcare services: 1) outpatient-oriented medical services or 2) community-based health and social services. Behavior after the selection will also affect the process of valuation. For example, the less the individual is satisfied with the use of outpatient-oriented medical services, the more the individual will prefer to use community-based health and social services.
Based on the theoretical process, the hypothesis of this study can be summarized as follows: H1. The lower the individual's physical or psychosocial functioning, the higher the individual will valuate physician home visits. H2.1. The less the individual is satisfied with the use of outpatient-oriented medical services, the higher the individual will valuate physician home visits. H2.2. The more the individual is satisfied with the use of community-based health and social services, the higher the individual will valuate physician home visits.
Study participants and survey
A total of 800 people aged 60 years or older were selected nationwide using a multi-stage stratified random sampling method in consideration of the distribution of gender, age and five wide residential areas including Seoul, Gyeonggi, Yeongnam, Honam, and Gangwon. Gender distribution were assigned to include men and women evenly. Age distribution were applied in accordance to the national age distribution by 10 years from 60 years old to 90 years old or over. As for the residential area, sampling was distributed evenly by five regions. The researchers trained interviewers on the questionnaires before the survey. Voluntary faceto-face direct interviews were conducted with older adults living in the community who were able to communicate without cognitive distortion and/or hearing loss and able to understand the purpose of this study. Participants with missing data were excluded; the resulting study population consisted of 797 participants.
Measurements
The outcome variable was the value for physician home visits, which was measured using two methods. First, the need for physician home visits was measured using the following question: "Do you want a doctor's home visiting service when you need help?" Second, WTP for physician home visits was measured using one open question measured in Korean won (KRW): "For a doctor's home visit, what is the appropriate amount per visit that you could pay?" Independent variables were categorized into two groups: 1) health-related factors and 2) service utilization factors. For health-related factors, variables of EuroQol-five dimensions (EQ-5D), number of falls, and subjective loneliness were measured to assess physical and psychosocial functioning, respectively. The EQ-5D is comprised of the following five dimensions: mobility; self-care; usual activities; pain/discomfort; and anxiety/depression. The results were summed to produce a global score ranging from 5 (best) to 15 (worst). The number of falls was also reported by participants to assess physical functioning based on their experience during the past year. We assessed psychosocial functioning as subjective loneliness (scored from 1 to 3). To assess service utilization, current experiences concerning outpatient-oriented medical services and community-based health and social services were measured. In this study, we supposed that currently available community-based health and social services in Korea included the following components: 1) National Preventive Home Visits (NPHV, provided by visiting nurses from a public health center); 2) Advanced Home Care Service (AHCS, provided by nurse practitioners from a general hospital); 3) Home Helper Visiting Service (HHVS, provided by home helpers financed by the Long Term Care Insurance); 4) Senior Health Promotion Program (SHPP, provided by a social welfare center or Gyeong-ro-dang); and 5) community service provided by a public health center. We regarded an outpatient-oriented medical service as equal to service from a local private clinic. All of the experiences of utilization were dichotomized, and satisfaction was scored on a Likert scale ranging from 1 to 5.
Information regarding demographic and socioeconomic status included gender, age, educational level, residential area, family size, marital status, social participation, household income, and occupational status (i.e., salary earners, self-employed or other jobs, out of work). Using information on household income and family size, the categorical variable of absolute poverty, indicating people living beneath the minimum cost of living, was formulated.
Statistical analysis
A statistical summary was first used to characterize study participants. The logit model for regression was used to analyze the association between the need for physician home visits (dichotomized variable) and each of the independent variables. Lastly, the Tobit model was used in order to identify the association between WTP (censored continuous variable) and the independent variables. In the present study, 61.5% of the study participants reported that they valuated physician home visits at zero KRW. This is the classic example of censoring, which Tobin 22 referred to in his study of household expenditures. It is reasonable to assume that dependent variables are censored to zero, because no one would easily suppose that a medical doctor would come to one's house if he or she could not pay a certain minimum amount (e.g., 10,000 KRW). In this case, he or she may respond with zero instead of reporting an exact value. The Tobit model has an advantage in that it can estimate the coefficient with less bias without loss of sample by assuming a censored normal distribution. All analyses were adjusted for gender, age, family size, residential area, educational level, marital status, social participation, household income (log transferred), poverty status, and occupational status. A P value less than 0.05 from two-sided tests was considered statistically significant. All analyses were conducted by Stata 15.
Ethics statement
The present study protocol was reviewed and approved by the Institutional Review Board (IRB) at Chung-Ang University (IRB No. 1041078-201711-HRSB-224-01). Informed consent was submitted by all subjects when they were enrolled.
RESULTS
The general characteristics of the study participants are described in Table 1. The mean age was 70.6 (standard deviation 7.5). Most participants were residing in an urban area (77.2%), and almost half of them did not participate in a labor activity (53.0%). About 25.6% of participants reported that their monthly household income was below the absolute poverty line. The mean and standard deviation of the EQ-5D score were 6.2 ± 1.6. Satisfaction related to using a local private clinic and public health center were recorded as 3.9 ± 0.5 and 3.7 ± 0.6, respectively. Table 2 shows the results of the valuation of physician home visits according to general characteristics. Three hundred and thirteen participants (39.3%) reported that they wanted a physician's home visit when they needed help. Among all study participants, the WTP amount for physician home visits was 8,683 ± 15,395 KRW. When we restricted the sample only for the subgroup of participants who reported that they needed a physician home visit service (n = 313), the mean and standard deviation of WTP was 21,982 ± 17,546 KRW. Table 2 also describes the group differences in value for physician home visits. Older adults who had an occupational status of out of labor or self-employed (P = 0.042) were more likely to need a home visit service than others. Those who were satisfied with a public health center (P = 0.046) were also more likely to need a home visit service compared with their counterparts. A younger age, below 75 years old (P = 0.030), higher educational level (P = 0.002), married status (P = 0.007), social participation (P = 0.018), and self-employed or out of labor as occupational status (P = 0.036) were also found to be significant factors associated with higher WTP for physician home visits. One thing to note is that only a weak correlation was observed between household income and WTP for physician home visits (Pearson's correlation coefficient, r = 0.053, n = 797), even though we analyzed the poor (r = 0.094, n = 204) and the non-poor (r = 0.021, n = 593) separately Fig. 1). This means that a physician home visit service is a kind of common necessity good that has an income elasticity of demand between 0 and 1. Health-related quality of life, EQ-5D score 6.2 ± 1.6 6.5 ± 1.6 5.9 ± 1.4 Self-rated health, score 2.7 ± 0.9 2. 8± 0.9 2.5 ± 0.8 Number of fall within 1 year, No.
0.5 ± 0.8 0.5 ± 0.8 0.3 ± 0.7 Subjective loneness, score 1.7 ± 0.6 1.7 ± 0.6 1.6 ± 0.6 Service utilization factors Outpatient-oriented medical services Satisfaction level for using local private clinic, score 3.9 ± 0.5 3.9 ± 0.5 3.9 ± 0.5 Community-based health and social services Satisfaction level for using public health center, score 3.7 ± 0.6 3.7 ± 0. 33 (9.4) Data are presented as mean ± standard deviation or number of case (%). KRW = Korean Won, EQ-5D = EuroQol-five dimensions, NPHV = national preventive home visits provided by visiting nurses from public health center, AHCS = advanced home care service provided by nurse practitioners from general hospital, HHVS = home helper visiting service provided by home helpers financed from the longterm care insurance, LTC = long-term care, SHPP = senior health promotion program provided by social welfare center or Gyeong-ro-dang. P values were calculated by χ 2 test, Fisher's exact test, and independent t-test. Total sum of scores are listed as follows: EQ-5D (15), self-rated health (5), subjective loneliness (3), satisfaction of local private clinic (5), satisfaction of public health center (5 3.63; P = 0.021). This finding means that persons who are currently using a community-based social service tend to pay 8,390 KRW more than those who are not using the service to purchase doctor home visits. Among the general characteristics, only higher educational status (i.e., above high school graduation) was significantly associated with an increased need (OR, 1.67; 95% CI, 1.05-2.67; P = 0.030) and WTP for physician home visits (β, 8.49; SE, 3.86; P = 0.028).
Tables 4 and 5 provide the results of the subgroup analyses, which show a moderate effect of gender and satisfaction level with using an outpatient-oriented medical service in Korea. Among female older adults (n = 445), health-related factors such as the EQ-5D score (OR, 1.21; 95% CI, 1.04-1.41; P = 0.015; β, 2.26; SE, 1.02; P = 0.028) and number of falls (OR, 1.41; 95% CI, 1.09-1.80; P = 0.008; β, 4.15; SE, 1.68; P = 0.014) were identified as significant factors associated with an increasing need and WTP for physician home visits. This means, for example, as the number of falls increased by one unit, the value of physician home visits generally increased 4,150 KRW among female older adults. When we restricted the study participants who were not satisfied with using a local private clinic (n = 115), the only factors that affected the need for physician home visits were identified as health-related factors, such as the number of falls (OR, 2.22; 95% CI, 1.08-4.57; P = 0.031) and subjective loneliness (OR, 3.19; 95% CI, 1.32-7.70; P = 0.010). Among the subgroup of non-satisfied, as the level of subjective loneliness increased by one unit (i.e., from not at all to sometimes, or from sometimes to always), WTP for physician home visits generally increased 17,520 KRW. Among this subgroup population, no other influencing factors reached significance at the 0.05 level. This indicates that valuation process for physician home visits among older adults is based upon the complex interaction between health-related factors, service utilization factors, and general characteristics of the population. 0.209 WTP = willingness to pay, KRW = Korean Won, OR = odds ratio, CI = confidence interval, β = beta coefficient, SE = standard error, EQ-5D = EuroQol-five dimensions, HQoL = health-related quality of life, NPHV = national preventive home visits service, HHVS = home helper visiting service, SHPP = senior health promotion program. a Local private clinic: satisfaction score of using local private clinic; b Public health center: satisfaction score of using public health center.
DISCUSSION
This study aimed to estimate the proportion of the Korean older population who needed physician home visits and suggest influencing factors associated with the value of physician 10/14 https://jkms.org https://doi.org/10.3346/jkms.2020.35.e158 Willingness to Pay for Physician Home Visits 0.323 WTP = willingness to pay, KRW = Korean Won, OR = odds ratio, CI = confidence interval, β = beta coefficient, SE = standard error, EQ-5D = EuroQol-five dimensions, HQoL = health-related quality of life; NPHV = national preventive home visits service, SHPP = senior health promotion program, HHVS = home helper visiting service. a Local private clinic: satisfaction score of using local private clinic; b Public health center: satisfaction score of using public health center; c Variable of 'use of HHVS' was omitted due to perfect collinearity. home visits. Overall, about 40% of older adults reported that they needed physician home visits. Those who were in need of physician home visits also reported that they were able to afford the service and willing to pay an average of 22,000 KRW per doctor's home visit. In this study, we also examined the most important factors that influenced older adults to decide to use a physician home visit service. Results showed that a higher valuated need for physician home visits was associated with an individual with a lower level of physical/psychosocial functioning (e.g., EQ-5D) and a higher level of satisfaction with using community-based health and social services (e.g., public health center, social welfare center or Gyeong-ro-dang).
To the best of our knowledge, no previous studies have published the results of a population survey concerning WTP for physician home visits in Korea. Currently, there is a heated debate about the fee for physician home visits under the Community Care Policy in Korea. In December 2018, for the first time in Korean history, legal foundations for physician home visits were established. According to the revised version of the National Health Insurance Act (41-5), a medical doctor can now conduct medical treatment in a patient's home (i.e., outside the clinic) if the patient has a condition (illness or injury) making it uncomfortable to move outside the house. These changes were announced in June 2019, and debates about how much the insurer (the National Health Insurance) should pay to physicians for home visits are being seriously discussed between the government and medical associations. With this background, this study emphasized that the fee for physician home visits should be valuated based on the empirical data of WTP, which comes from a consumer-centered perspective. According to the survey of the Korean older population in our research, the total fee for a physician home visit service could be set as high as 73,000-220,000 KRW, if we assume the copayment ratio (i.e., out-of-pocket payment share of the total fee for the service) is set as 10%-30%. This is a somewhat higher payment than the fee for a physician home visit set in the Primary Care Service for Disabilities (73,850 KRW, 10% copayment) and Home-based Hospice-Palliative Care Service (83,870 KRW, 20% copayment), which are currently being implemented.
This study is also significant in that it suggests a theoretical model of value creation for physician home visits. From the consumer-centered perspective, we assume that each individual develops a personal valuation system as cognitive processes that guide his or her choice decisions among various health-related service providers. In Korea, there are broadly two different settings of healthcare services: 1) outpatient-oriented medical services and 2) community-based health and social services. Value negotiation will occur as to which type of service is the most appropriate to relieve a patient's symptoms with consideration of the individual's physical and psychosocial functioning, personal experiences of service utilization, and basic demographic factors. This study provides evidence that choice decisions to pay for physician home visits are based upon complex interaction among these factors. The majority of participants (approximately 60%) did not report the necessity for physician home visit care. It is possible that older adults in need of physician home visits included home-bound elderly who are more likely to be vulnerable in terms of health status and economic condition. According to the previous study, 15 the prevalence of homebound older adults is estimated as 30.1%. In addition, the value-creating process of physician home visits can be influenced by past experiences of healthcare service utilization. Most of the elderly had little experience in physician home visiting in Korea, thus it would be difficult to respond positively to the need for visiting care.
On the basis of the empirical analyses, two subgroups of older adults were identified as most in need of physician home visits: 1) female older adults who have limited physical functioning 11/14 https://jkms.org https://doi.org/10.3346/jkms.2020.35.e158
Willingness to Pay for Physician Home Visits
and 2) older adults who feel loneliness and are not satisfied with using medical services in a local private clinic. These results have some implications. First, a gender difference was noticed in the valuation process of physician home visit services. Compared with men, older women in Korea were willing to pay for physician home visits when they encountered problems with physical functioning (e.g., recurrent falls). Perhaps this kind of difference may be a feature of Korean society with gender inequality in caring roles under the patriarchal tradition. 23 Second, psychosocial functioning, such as subjective loneliness, also affected the value-creating process of physician home visits. A previous study conducted in the Netherlands reported that preventive home visiting improved metal health rather than physical health. These results are consistent with previous studies that suggested that services in the home care setting should be provided with an interdisciplinary team. 5, 24 To be called home-based primary care, one of the alternatives to primary care in several countries, a physician should provide comprehensive and continuous healthcare services through interdisciplinary teams at a patient's home. 15,24,25 Third, community-based social services may contribute to expanding a new healthcare market for physician home visits. Interestingly, in this study, using healthcare services from a social welfare center or Gyeong-ro-dang was analyzed as the most significant factor that increased both the need and WTP for physician home visits. Maybe the learning effect of Aging in Place and Community Care Policy has worked. From the viewpoint of community-dwelling older adults, various types of community-based services may promote the value of physician home visits. This implies that community-based social services are not just a competitive substitute for outpatient-oriented medical services, but a cooperative complement that play a significant role in expanding the healthcare market into new areas. For example, the metropolitan government of Seoul initiated the 'Chat-Dong,' and 'Seoul-Care' recently. 15 Chat-Dong is the universal home visit program by nurses and social workers for older adults aged 65 and 70 years old. 'Seoul-Care' is the integrated chronic disease management program by multidisciplinary teams including physicians, nurses, social workers, nutritionists, and athlete trainers working in the public health center. Similar programs have been started inside Community Care Program to several regions in Korea from 2018. Connecting subjects and collaborating services with these community based social services can be both a benefit and an advantage for the physician home visit care.
This study has some methodological limitations. First, although the research design of this study helps identify general characteristics of people who are most in need of physician home visits, it is difficult to make causal inferences using a cross-sectional study. Second, not all confounding factors that affected the results were included in the analyses. For example, we were unable to measure variables concerning chronic disease or comorbidity. If we assume that the number of chronic diseases has a positive correlation with WTP for physician home visits, as is generally supposed, the results of our study would be overestimated by omitting that variable. Third, the results have limited generalizability due to the relatively small sample size. For example, the generalizability of the results to persons who are beneficiaries of the National Long-Term Care Service or persons with a homebound status is unknown. Although we used nationwide survey data selected by a multi-stage stratified random sampling method, further surveys with participants of a different age, functional status, and medical condition are needed to generalize the results of the study.
In conclusion, about 30% of community dwelling older adults reported that they need physician home visit services. And we can expect that consumer power might be stronger 12/14 https://jkms.org https://doi.org/10.3346/jkms.2020.35.e158 Willingness to Pay for Physician Home Visits than the Korean government has acknowledged because patients' WTP was somewhat higher than the fee for physician home visits set in the Primary Care Service for Disabilities. The decision to pay for physician home visits is based upon the complex interactions among an individual's physical and psychosocial functioning, personal experiences of service utilization, and demographic factors. The value for physician home visits should be qualified based on the WTP empirical data, which comes from a consumer-centered perspective. | 2020-04-30T09:09:31.692Z | 2020-03-25T00:00:00.000 | {
"year": 2020,
"sha1": "e64709ae6e19aeaabae9167bfbffb5795c229967",
"oa_license": "CCBYNC",
"oa_url": "https://doi.org/10.3346/jkms.2020.35.e158",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "b73d1c42e7ee55337bb85c216199f9e65cd77bf1",
"s2fieldsofstudy": [
"Medicine",
"Political Science"
],
"extfieldsofstudy": [
"Medicine"
]
} |
211146164 | pes2o/s2orc | v3-fos-license | Elitism in Mathematics and Inequality
The Fields Medal, often referred as the Nobel Prize of mathematics, is awarded to no more than four mathematician under the age of 40, every four years. In recent years, its conferral has come under scrutiny of math historians, for rewarding the existing elite rather than its original goal of elevating mathematicians from under-represented communities. Prior studies of elitism focus on citational practices and sub-fields; the structural forces that prevent equitable access remain unclear. Here we show the flow of elite mathematicians between countries and lingo-ethnic identity, using network analysis and natural language processing on 240,000 mathematicians and their advisor-advisee relationships. We found that the Fields Medal helped integrate Japan after WWII, through analysis of the elite circle formed around Fields Medalists. Arabic, African, and East Asian identities remain under-represented at the elite level. Through analysis of inflow and outflow, we rebuts the myth that minority communities create their own barriers to entry. Our results demonstrate concerted efforts by international academic committees, such as prize-giving, are a powerful force to give equal access. We anticipate our methodology of academic genealogical analysis can serve as a useful diagnostic for equality within academic fields.
Although mathematics is o en framed as objective and egalitarian, its access is not equally conferred. Recent a ention has been given to the Fields Medal, one of the most prestigious awards in math, and its elite community. When the award was rst conceived in 1930, it was in part designed to assuage international tensions [1]. e award was intentionally given to individuals that would otherwise not receive any recognition, rather than the best young mathematician.
Using social network analysis (SNA) and neural-based natural language processing (NLP), this paper analyses the ow of elite mathematicians between nations and lingo-ethnic categories. Analysis was performed on the Mathematics Genealogy Project, one of the most complete advisoradvisee databases maintained today with more than 240,000 mathematicians.
Results demonstrates the self-reinforcing behavior among the elite level in mathematics. is contrasts with prior conferral of the Fields Medal, which was a positive force in mending international relations, such as integrating Japan and Germany a er World War II [6]. We propose the Fields Medal can be used today to improve accessibility of mathematics to minority groups. e classi er for lingo-ethnic identity is textual, it would be more accurate to say we classify speci c languages that overlap signi cantly with ethnic or cultural identity. While the use of lingo-ethnic categorization as identity is shallow, our principal aim is to show, even at the most basic de nitions of ethnicity or culture through language, we nd evidence of inequality. is paper also o ers a methodological contribution. We show that combining network analysis, neural-based natural language processing (NLP), and well-maintained academic databases can serve as a powerful diagnosis for access and equity, and improve the practice of science.
Several studies on elitism within the production of mathematical knowledge have been conducted. Methods draw predominantly from the complex network perspective [7], leveraging network repositories such as citation and bibliometric networks. Gargiulo et al. studied the entire, connected giant component of the mathematical genealogy project, enriching the data using data mining techniques [5]. ey work focused on integrating math history with temporal network analysis, noting the elds evolution based on country, discipline, and the structure of scienti c families.
Prior investigated about the relationship between scienti c mentorship and winning the Fields Medal or Wolf Prize, but results were inconclusive. Rossi et al. studied the role of advisor-advisee relationships [4]. ey propose the genealogy index, adapted from the h-index which was initially developed by Hirsh [3]. Malmgren et al. studied the role of mentorship on protégé performance, focused on metrics of academic success like publication record [8]. Beyond scholarship, studies have also considered hiring practices [9] and departmental prestige [10]. e lack of metadata in these genealogies has limited the scope of investigation. is paper places elite community network ow as the focal point, contrasting the historical focus on the nation-state with the modern focus of identity.
HISTORICAL NETWORKS OF ELITE MIGRATION
We begin with a sketch of history. Figure 1a) captures the migration of elite mathematicians between ve key countries. e subgroup of elites was created by aggregating the shortest paths between Fields Medalists. is ensures that the full graph is connected, and conceptually, denotes a minimal graph that connects all the medalists together. Here, migration is determined by comparing where a mathematician earned their Ph.D. and where their students earned their Ph.D. It is reasonable to assume primary advisors have moved to the same country as their advisees.
Prior to WWII, Western European countries were the strong-holds of mathematical thought. Notably, France and Germany contained the highest proportion of elite mathematicians. Many Japanese mathematicians studied in Germany, before returning to Japan, as part of modernization during the Meiji restoration. Examples inclue Rikitaro Fujisawa, who studied at the Unviersity of Strasbourg with Elwin Christofeel, before returning [11]. He was instrumental to reforming mathematics education in Japan. Mobility pa erns of mathematicians among countries (traditionally strong en mass). 1a), the migration of elite mathematicians from 1800 to the present. 1b), the net flow of elite mathematicians between 7 key countries. 1c), the flow analysis of elite mathematicians between countries. Exporting means mathematicians flow out a country, importing means they flow in, and selfish means they are retained. Many countries only import at the elite level (etc in the bo om right). A full list is available in the appendix. e ow chart reveals mass ows of researchers due to historical events. By 1932, the Holocaust led to mass migration from Germany to the United States and other European countries, which accounts for the drop in green volume, including prominent scientist Albert Einstein. Similarly, we observe large amounts of out ow from Russia a er the cold war, greatly diminishing the presence of Russia mathematicians a er the 1990s, and the second Italian mass diaspora a er WWII. Beyond forced immigration, ow analysis also reveals the movement of reintegration. Japanese mathematicians immigrated to the United States following WWII, and continued throughout the 60s to the 90s. Twenty years later, Japanese mathematicians owed back toward Japan.
France is not shown in the Sankey ow chart (1a), but is historically one of the countries that produces the most elite mathematicians. e chord graph in 1b) shows the net ow of mathematicians over all time, with the color of the chord indicating net exports. e USA-GER chord is orange, which indicates a net out ow from USA to Germany. Only France exports more American mathematicians than it imports from the USA. In all other cases, the USA exports more to other countries. Figure 1c) shows the ow dynamics on a country level. In-ow is de ned as the number of incoming edges, out-ow as the number of outgoing edge, and self-ow the number of loops.
ese results are similar to Gargiulo et al. [5] with two striking di erences. First, the United States is a sel sh and importing country at the elite level, whereas in general it is sel sh and exporting. Secondly, there are many more importing countries compared to the general case, where most countries are exporting and sel sh. Notice, many of the countries that are exporting and sel sh are Western or part of the Soviet Union, where there were strong programs in mathematics. Other countries appear to import more at the elite level, because their "exports" are not as competitive as mathematicians exported from other countries. ese two points allow us to infer three things. First, elite mathematicians have more mobility, and in many cases can begin work in foreign countries. Second, the United States imports more compared to the general case, a racting more elite members. ird, countries considered traditional mathematics strong-holds can be observed in the lower le corner.
What this analysis tells us, beyond an exposé of diasporic history, is the elds medal served as a way to mediate tensions. In a similar way that the Olympics was held in Rome, Berlin, and Tokyo, the inclusion of internationally marginalized nations.
THE FLOW OF MARGINALIZED IDENTITIES
Upon analyzing the history of elite communities in mathematics, we turn to the present. As 1a) shows, today, there is signi cant ow between countries. lingo-ethnic categories of identity serve as a useful construct for understanding network ow. Figure 2a) shows the representation of identities, within three subgroups: all mathematicians (blue), mathematicians within the medalist subgroup, (green) and the medalists themselves (red). Fig. 2 compares elite representation of subgroups relative to their actual proportions. For instance, there is a higher proportion of French medalists (14%) compared to the general proportion (8%). In contrast, there is a signi cant number of East Asian mathematicians (14%) but very low representation in both the medalist family and medalists themselves (5% each). On the level of ow, Figure 2b) characterizes identities in terms of in-ow, out-ow, and self-ow. High in-ow means a higher likelihood of being mentored. High out-ow then corresponds to a greater likelihood to mentor others. High self-ow means higher likelihood of mentoring your own identity. e identity with the most self-ow is Japanese. However, when all mathematicians are considered, the Japanese are shown as green, that is to say opposite of sel sh. is indicates reinforcing behavior only occurs at elite levels.
However, once these groups are aggregated into larger groups-Greater European, Asian, African and Arabic-then di erences become evident. European names has high self-reinforcing behavior, whereas Asians names and African and Arabic names are much lower in the number of self-loops. is dispels a common myth that minority groups, due to homophily, tend to group together. is myth insinuate that barriers to entry are self-in icted. However, as we see from 2b), most minority groups are far away from the sel sh pole, with a healthy balance of in-ow and out-ow. Rather, increases in the quantity of self-loops occurs in the greater European subgroup.
OLD STRONGHOLDS, NEW POSSIBILITIES
It is understandable that, when considering all mathematicians, that there is a high levels of self-ow-studying in elite and o en foreign institutions is a privilege. However, the fact that high self-ow in identity at the elite level suggests institutions can do more to open access, given their greater access to resources. is has been the case for Japan.
Japan is unique among Asian countries and identities in that there are many Japanese Fields Medalists (3), with high representation in elite levels. Japan has been known for its rapid westernization during the Meiji restoration relative to other Asian counterparts. As early as 1872, their traditional form of math wasan was replaced by western science. Prussia, rather than the United Kingdom, was the primary source of westernization, and led directly to the establishment of the University of Tokyo [6]. A er WWII, mathematicians sought to re-establish international ties and formed the International Congress of Mathematicians and a new International Mathematics Union (IMU). Marshall Stone, a proponent of this movement, said it clearly: "in considering American adherence to a Union, it must be borne in mind that we want nothing to do with an arrangement which excludes Germans and Japanese as such." Indeed, we nd the ten founding members well-represented in the ternary diagrams, and not long a er founding, the Soviet Union joined. Revisiting Fig. 1a), we discover the density of elite mathematicians in Japan increases a er 1945.
What this says, is the Fields Medal can improve the status of marginalized populations. Mathematics historian Barany captures this aspiration, believing the elds medal should help "sculpt the future, rather than reward the past [1]." What we observe is the opposite, where the elite perpetuate the elite. Fig. 3 demonstrates this clearly, showing French Fields Medalist Laurent Schwartz and his lineage.
Within 5 generations a er Schwartz, 7 Fields Medalists emerge. In particular, Schwartz-Grothendieck-Deligne form a direct chain, as do Lions-Villani-Figalli. Note, Lions' father Jacque -Louis Lions was also a student of Schwartz. In other words, 13.3% of all Fields Medalists descended directly from Schwartz. Broadly, each of these all made contributions to some form of algebraic geometry or functional analysis. Fig. 4 further shows that all medalists can be traced to 9 connected components, with the largest one holding 44 out of 60 listed Fields Medalists. A mathematician that is French and a ends a Top 50 institution means they are 6.4 times more likely to gain membership into the elite circle. Here, the top 50 is de ned as the top institutions a ended by those in the elite group. Note, we de ned our Fields Medalist subgroup minimally, such that any other de nition of subgroup would yield a higher power ratio. On the other hand, being East Asian and a ending a Top 50 institution only a ords you 1.4 times the likelihood of gaining membership into this elite circle.
From this diagram, we infer that institution plays a large role in elite membership. However, notice an East Asian mathematician a top 50 school is 4.5 times less likely to be included than a French mathematician a ending a top 50 school. An Indian mathematician educated outside top 50 schools are 6 times less likely to be included than a French mathematician with the same education. Amongst non-elite institutions, being Japanese gives the best chance of inclusion, an a er-e ect of the e orts by the IMU.
CONCLUSION
In 2014, the late Iranian mathematician Maryam Mirzakhani won the Fields Medal. A talented star herself, her groundbreaking work on dynamics and geometry was encouraged by her Ph.D. advisor Curtis McMullen, also a Fields Medalist, at the elite institution Harvard University. is is by no means downplaying her achievements; rather, it serves to show the power recognition and elite communities have-all of which membership she rightly earned. Although the Fields Medal should serve to recognize under-represented researchers, the proper cultivation of talent through mentorship and institutional support should be the starting point.
In our evaluation of the present, there is a large under-representation of minority groups in not just Field Medalists, but also in the elite circle for mathematics. While institutional prestige a big factor, lingo-ethnic identity is also found to be highly relevant, the widest gap being 4.5 times the power ratio even at elite institutions. Given that elite institutions have more resources, they can take a bigger role in generating higher access for marginalized groups. Flow analysis also dispels the myth that under-representation arises from homophily-driven self-selection.
Although the French stronghold shows the old forces that govern mathematical knowledge remain strong, the presence of Japanese scholars also shows concerted e ort can be used as an integrating force. Concerted e orts by international academic commi ees, such as prize giving, are a powerful force to confer equal rights for knowledge production to traditionally marginalized groups. Beyond analysis, this network analytical methodology is a call for scienti c communities to use advisor-advisee databases to open knowledge production and scienti c access.
Graph Construction
e graph was constructed using the Mathematics Genealogy database. Nodes are mathematicians, and directed edges represent advisor-advisee relationships. e data set contained information (listed in order of completeness) on the academic, advisor-advisee links, school, PhD graduation year, country, and dissertation title and topic. e ID's of medalists were identi ed, then the shortest path was computed in a pairwise fashion. Analysis was conducted primarily using the Networkx package [12]. e subgroup of elites was created by taking the union of shortest paths between Fields Medalists. en, the full graph is connected, and denotes some form of minimal graph that connects all the medalists together. While it is possible to produce a minimal spanning tree, given the forest like structure of the genealogy, the shortest paths has more interpretive value.
Identity Classifier
Since lingo-ethnic identity is not included in the Mathematics Genealogy Project, a separate classi er is required. e identity categories were labeled using the ethnicolr package, which is a long-short term neural network (LSTM) trained on Wikipedia and the census [13]. Speci cally, the LSTM was based o the seminal work of Graves and Schmidhuber [14]. is package has found use in evaluating under-representation in other STEM elds such as biomedicine [15]. It achieves between 78% to 81% accuracy. Potential shortcomings of neural methods for categorization is the accuracy levels. However, for 13 individual categories (which would result in 7.7% accuracy if truly random), 81% is quite high. Additionally, since we are interested in comparison within individual demographics, any bias would be carried forward since the group of all mathematicians supersets the medalist subgroup and medalists themselves. e goal of using this classi er is not to a en de nitions of identity, but to use the best available tools for inference, in absence of concrete data.
Flow Analysis
Meso-graphs were constructed on a ributes of each mathematician. To turn a ributes into nodes, we constructed a mapping from mathematician to the meso-categories (lingo-ethnic and nationality of doctoral degree). Edges between meso-categories were simply the original directededges between mathematicians. Each edge is then weighed by the number of advisor-advisee relations between meso-categories.
Constructing Ternary Diagrams. We constructed the ternary diagrams through analysis of the meso-network. Every meso-network can be represented by a its adjacency matrix, which we denote M. e diagonal then accounts for self-loops, the rows excluding the diagonal elements the out going edges, and the columns excluding the diagonal element the incoming edges. Explicitly, for meso-category indexed by i, we have the following de nitions for in-ow (IF), out-ow (OF), and self-ow (SF).
We then normalize these values to represent each meso-category as a point in three dimensional space.
Note, all points lie on the plane described by x + + z = 1. We then transform this planar section onto the two dimensional plane using a translation and two rotations.
where R 1 rotates the plane up to the XY-plane, and R 2 aligns the simplex to the x-axis. | 2020-02-19T02:01:21.759Z | 2020-02-18T00:00:00.000 | {
"year": 2021,
"sha1": "092d9eaf00b0c1b60d788a168c34522e55530ed0",
"oa_license": "CCBY",
"oa_url": "https://www.nature.com/articles/s41599-020-00680-y.pdf",
"oa_status": "GOLD",
"pdf_src": "MergedPDFExtraction",
"pdf_hash": "9e724eb4ed8949a825d4fb87861630138845e1fb",
"s2fieldsofstudy": [
"Mathematics",
"Sociology"
],
"extfieldsofstudy": [
"Computer Science",
"Mathematics",
"Physics"
]
} |
251800301 | pes2o/s2orc | v3-fos-license | Analytical solution of the disordered Tavis-Cummings model and its Fano resonances
$\mathcal{N}$ emitters collectively coupled to a quantised cavity mode are described by the Tavis-Cummings model. We present complete analytical solution of the model in the presence of inhomogeneous couplings and energetic disorder. We derive the exact expressions for the bright and the dark sectors that decouple the disordered model and find that, in the thermodynamic limit, the energetic disorder transforms the bright sector to Fano's model that can be easily solved. We thoroughly explore the effects of energetic disorder assuming a Gaussian distribution of emitter transition energies. We compare the Fano resonances in optical absorption and inelastic electron scattering both in the weak and the strong coupling regimes. We study the evolution of the optical absorption with an increase in the disorder strength and find that it changes the lower and upper polaritons to their broadened resonances that finally transform to a single resonance at the bare cavity photon energy, thus taking the system from the strong to the weak coupling regime. Interestingly, we learn that the Rabi splitting can exist even in the weak coupling regime while the polaritonic peaks in the strong coupling regime can represent almost excitonic states at intermediate disorder strengths. We also calculate the photon Green's function to see the effect of cavity leakage and non-radiative emitter losses and find that the polariton linewidth exhibits a minimum as a function of detuning when the cavity leakage is comparable to the Fano broadening.
The minimal model to describe the cavity QED with multiple matter systems, or emitters as they are usually called, is the Tavis-Cummings (TC) model [22]. The model can be analytically solved in the absence of disorder using the bright (symmetric) and dark (nonsymmetric) states [23,24]. However, in the presence of a disorder in the transition energies of the emitters, these states are not decoupled so this transformation is rendered useless and we resort to brute force numerical solutions [25][26][27]. In some cases, the disorder can even invalidate the basic TC model and an extension is deemed necessary [28]. In natural systems, on the one hand, disorder both in the transition energies and couplings is inevitable. On the other, a tuneable disorder can be realised in artificial systems [29,30]. Thus, exploring the effects of disorder is important and solving the TC model analytically can be a significant step forward as it can provide us with an insight into the disordered cavity QED systems. Although the TC model describes a wide range of sys-tems, we illustrate our results here considering organic microcavities as an example. The transition energies of organic molecules usually have a distribution of width √ 2 × 0.1eV [31][32][33], which is a sizeable fraction of the typical Rabi splitting of 0.5eV [34] in these systems. Furthermore, the distribution of orientations and positions of the molecules in the cavity translates to a disorder in their couplings to the cavity mode.
Organic polaritons have interesting properties, e.g., condensation and lasing at room temperature [4,35,36]. Their effect on materials' properties, e.g., charge and energy transport [37][38][39][40], and chemical reaction rates [23,[41][42][43][44][45][46][47] is also studied recently. Besides the polaritons, we have a large number of dark states in such systems [23,48], that also play important role in various processes of interest, e.g., catalysis [27]. While the dark states were initially thought to be only a reservoir of incoherent excitations [49,50], they are now well appreciated for their coherence and delocalised nature [51][52][53]. Although, the dark states do not couple to the light, they can still be excited optically indirectly by exciting polaritons or higher energy emitter states. In addition, the electrical excitation predominantly creates the dark states due to their large density of states. The dark states can relax to the lower polariton state, which is important for creating large polariton population for condensation and lasing for instance, and their dynamics has been studied extensively [54][55][56][57][58][59][60][61][62]. The effect of disorder on the localisation of the dark state has also been numerically studied recently [26]. Due to their hugely important role in organic microcavities and other cavity QED systems, it is desired to know their exact form under realistic conditions, which can lead to a significant development of analytical methods.
Since real systems are often weakly coupled to their environment, the resulting homogeneous broadening of the cavity and molecular states is inherited by the polariton states, and can be treated using open quantum system approaches-master equations-or phenomenologically with the Green's functions [63][64][65] (or an equivalent quantum control method [66]). In fact, to obtain the optical absorption spectrum, the Green's functions can exactly treat the energetic disorder or the inhomogeneous broadening as well [30,67], as we will later show in this article. (Nevertheless, it has also been used to find an approximate solution by neglecting the effect of energetic disorder on the light-matter coupling strength [25].) However, only the photon Green's function can be computed this way, which alone cannot describe the eigenstates of the system, thus severely limiting the scope of the previous studies [25,30,66,67].
The exact numerical diagonalisation, another method that is simple and usually effective, has its own shortcoming. In a clean system with identical emitters, studying the collective effects in the thermodynamic limit is possible as the typical size that is considered large enough is N = 20, where N is the number of emitters. However, in an energetically disordered system, a realistically smooth distribution of states that converges the results requires N 10 6 , which is computationally intractable considering that all eigenstates of the Hamiltonian are to be computed. This is the reason that exact diagonalisation in such cases would even fail to capture the features in the optical spectrum that are given by the photon Green's function in the thermodynamic limit.
What is so important in the thermodynamic limit of the energetically disordered system that has not yet been understood? It is the emergence of the Fano's model [68], which can only be found if the Hamiltonian is first decoupled into its exact bright and dark subspaces. Fano's model describes the interaction of a discrete or localised state to a continuous band and exhibits Fano resonances [68][69][70][71][72]. It gives a characteristic asymmetric lineshape in excitation spectra that finds applications [73] in sensing and switching devices [74,75]. It is ubiquitous in physical systems [76][77][78][79][80][81][82][83] but, being a coherent phenomenon, finding it in a disordered system [84,85] is not common. The Fano's model has always been applied in the weak coupling regime but, as we will shortly see, it is also relevant in the strong coupling regime [72].
Here, we consider the Tavis-Cummings model with disorder both in emitter energies and their couplings to the cavity mode. We derive the exact expressions for the bright and the dark states that span the two decoupled subspaces of the model. We show that, in the thermodynamic limit, the bright states of the disordered system form a non-degenerate band whose interaction with the cavity mode gives the Fano's model that can be exactly solved. Figure 1 illustrates the formation of the bright band and the nature of the eigenstates of the system at varying levels of the disorder strength. In Fig. 1(a), when N emitters have the same transition energy, their states can be transformed to give a single bright state |S 0 that couples to the cavity and N − 1 dark states that are de-2 the cavity and molecular states is inherited by the polariton states, and can be treated using open quantum system approaches-master equations-or phenomenologically with the Green's functions [63][64][65] (or an equivalent quantum control method [66]). In fact, to obtain the optical absorption spectrum, the Green's functions can exactly treat the energetic disorder or the inhomogeneous broadening as well [30,67], as we will later show in this article. (Nevertheless, it has also been used to find an approximate solution by neglecting the e↵ect of energetic disorder on the light-matter coupling strength [25].) However, only the photon Green's function can be computed this way, which alone cannot describe the eigenstates of the system, thus severely limiting the scope of the previous studies [25,30,66,67].
The exact numerical diagonalisation, another method that is simple and usually e↵ective, has its own shortcoming. In a clean system with identical emitters, studying the collective e↵ects in the thermodynamic limit is possible as the typical size that is considered large enough is N = 20, where N is the number of emitters. However, in an energetically disordered system, a realistically smooth distribution of states that converges the results requires N 10 6 , which is computationally intractable considering that all eigenstates of the Hamiltonian are to be computed. This is the reason that exact diagonalisation in such cases would even fail to capture the features in the optical spectrum that are given by the photon Green's function in the thermodynamic limit.
What is so important in the thermodynamic limit of the energetically disordered system that has not yet been understood? It is the emergence of the Fano's model [68], which can only be found if the Hamiltonian is first decoupled into its exact bright and dark subspaces. Fano's model describes the interaction of a discrete or localised state to a continuous band and exhibits Fano resonances [68][69][70][71][72]. It gives a characteristic asymmetric lineshape in excitation spectra that finds applications [73] in sensing and switching devices [74,75]. It is ubiquitous in physical systems [76][77][78][79][80][81][82][83] but, being a coherent phenomenon, finding it in a disordered system [84,85] is not common. The Fano's model has always been applied in the weak coupling regime but, as we will shortly see, it is also relevant in the strong coupling regime [72].
Here, we consider the Tavis-Cummings model with disorder both in emitter energies and their couplings to the cavity mode. We derive the exact expressions for the bright and the dark states that span the two decoupled subspaces of the model. We show that, in the thermodynamic limit, the bright states of the disordered system form a non-degenerate band whose interaction with the cavity mode gives the Fano's model that can be exactly solved. Figure 1 illustrates the formation of the bright band and the nature of the eigenstates of the system at varying levels of the disorder strength. In Fig. 1(a), when N emitters have the same transition energy, their states can be transformed to give a single bright state |S 0 i that couples to the cavity and N 1 dark states that are de- coupled, hence the names. However, when the emitter energies form a distribution, a similar transformation at every energy produces bright and dark states at that energy, decoupling the dark sector completely and forming a non-degenerate band of bright states |S 0 (!)i with flat density of states (DOS). The eigenstates of the system in the single excitation space are formed by the coupling between |S 0 i or |S 0 (!)i and the cavity photon state |1 P i as shown in Fig. 1(b-e), where di↵erent scenarios that give polaritons [ Fig. 1 coupled, hence the names. However, when the emitter energies form a distribution, a similar transformation at every energy produces bright and dark states at that energy, decoupling the dark sector completely and forming a non-degenerate band of bright states |S 0 (ω) with flat density of states (DOS). The eigenstates of the system in the single excitation space are formed by the coupling between |S 0 or |S 0 (ω) and the cavity photon state |1 P as shown in Fig. 1(b-e), where different scenarios that give polaritons [ Fig. 1 For completeness, we also consider the effect of losses on the the optical response of the system and compute the exact photon Green's function that encodes it. We explore the effect of cavity and emitter losses on the optical absorption spectrum and find that, in organic microcavities, it should exhibit a minimum in the polariton linewidth around zero detuning. While preparing for the revised version of this manuscript, we found another work with some overlap [67]. The organisation of this article is as follows. The decoupling of the TC model into its bright and dark spaces is presented in sec. II, where sec. II A considers N identical emitters, sec. II B includes disorder in the couplings only, while sec. II C also takes account of energetic disorder. Section III contains the results related to the Fano's model. The emergence of the Fano's model and its solution in two possible scenarios [shown in Fig. 1(c,d)] are given in sec. III A. Fano resonances in two types of excitation spectra-optical absorption and inelastic electron scattering-are presented in sec. III B both in the weak and strong coupling regimes. Section IV explores the effects of energetic disorder, where sec. IV A considers the Rabi splitting and the nature of the eigenstates, sec. IV B discusses the differences from a homogeneous broadening, and sec. IV C analyse the Fano broadening of polaritons. Finally, the effects of losses are described in sec. V, where the photon Green's function is used to study the polariton lineshape and linewidth as well as the role of the distribution of emitters' energies.
II. MODEL AND ITS SOLUTION
Consider N two-level emitters coupled to a common cavity mode. The Hamiltonian of this system in the rotating wave approximation [86] is given bŷ where ω c is the energy of a cavity photon,â † ,â are its creation and annihilation operators, and σ ± i are raising and lowering operators for the ith emitter that has transition energy i and couples to the cavity mode with strength g i . Both g i and i are randomly distributed according to some probability functions.Ĥ commutes with the excitation numberN ex =â †â + N iσ + iσ − i , which allows its diagonalisation in an eigenspace ofN ex . We will focus on the single excitation subspace in this article.
In the following sections, we describe the decoupling of the emitters' Hilbert space into bright and dark sectors, where the bright sector couples to the cavity mode but the dark sector does not.
A. Identical Emitters
For identical emitters, i.e., when i = and g i = g for all i, the solution is already well known. In this case, a unitary transformation to the symmetric (bright) and non-symmetric (dark) superpositions of the emitter states block-diagonalises the TC model in the single excitation space. The bright state is created bŷ and produces the two polariton states due to its coupling to the cavity mode, which are given by, where δ = ω c − andΨ + U P/LP create the polaritonic states at energies ω U P/LP .
For the non-symmetric or dark states, being a degenerate manifold (at the bare exciton energy ), there is no unique representation and one can choose any complete set as basis states for the dark space. Usually, a set of delocalised discrete Fourier transform states is used, given byŜ In case of inhomogeneous couplings, (i.e., when a disorder in the couplings exists), these states no longer represent the dark sector. Here, we present another set that can be easily modified in such a case. Inspired from the maximally localised dark state, that is fairly known and is localised at site N , we can focus on how it turns out to be orthogonal to the bright stateŜ + 0 . We note that the contribution from the N th site is exactly cancelled by the sites 1 to N − 1. Designing another dark state on the same pattern to ensure its orthogonalisation toŜ + N −1 , we obtainŜ + N −2 that is maximally localised on emitter N − 1 but has zero component alongσ + N , i.e., on the site N . Repeating this process, we construct an orthonormal representation for the dark states where different dark states tend to localise on different molecules, albeit to a varying degree, given by [87] where j ∈ [1, N − 1].
B. Only off-diagonal disorder
It is well known that when a disorder in the couplings exists,Ŝ + 0 modifies tô where However, exact expressions for the dark states are not known. As we mentioned in the previous section, we modify the maximally localised states to obtain just that. These can be obtained by following the changes inŜ + 0 and imposing the orthogonalisation of the dark states to this bright state. That is, trŷ , where |GS is the ground state of the emitters. This gives, β j = G 2 j /g j+1 , and then normalisation GS|Ŝ − jŜ + j |GS = 1 gives α j = g j+1 /(G j G j+1 ). We thus obtain,
C. Fully disordered
Due to the presence of disorder in the diagonal terms in H, the above transformations no longer work. However, as we show below, the exact bright and dark subspaces still do exist and it is still possible to block diagonaliseĤ. Even though the energies { i } are randomly distributed, assuming their intrinsic linewidth γ to be negligible for the moment, we can count the number of states at any given energy and make bright and dark sectors for every energy level. Indexing these energy "levels" in ascending order, let's take K n to be the degeneracy of the nth level at energy ω n . Relabelling the emitters i → n, i n with i n ∈ [1, K n ] and dropping the subscript for brevity, i → ω n,i , g i → g n,i , and σ ± i → σ ± n,i . This allows us to make the bright and the dark states for every level, similar to Eqs. 9-10. For nth degenerate level, we havê where j ∈ [1, K n − 1] and G 2 n,l = l i=1 g 2 n,i . For the nondegenerate levels, only a bright stateŜ + n,0 =σ + n,1 exists.
Taking N as the number of the bright states or the number of levels,Ĥ can be written asĤ =Ĥ B +Ĥ D wherê whereĤ D is again already diagonal, and there are N D = N n=1 (K n − 1) = N − N dark states in total. Thus, to completely solve the disordered TC model, all that is left is the diagonalisation ofĤ B above.
H B can be diagonalised numerically. The size of the bright space could be negligibly small compared to the full space, i.e., N ≪ N . For instance, even at N = 10 23 , N could be taken as small as ∼ 10 2 − 10 3 for most realistic situations without compromising the thermodynamic limit. However, probably the best outcome of our approach is that, at large N , when {ω n } forms a continuous band, following Fano [68], analytical solution ofĤ B can also be found, as follows.
III. EMERGENCE OF FANO'S MODEL
In the thermodynamic limit, N → ∞, we can consider the continuous limit for the bright band. Taking where P c (g) and P (ω) are distribution functions for couplings and energies. Note that as long as we get the same value for g 2 , the specific form of P c (g) does not matter.
Taking ω 2 R ≡ n G 2 n,Kn and noting that P (ω)dω = 1, we can now write, ω 2 R = V (ω) 2 dω = N g 2 , which can be compared back to Eq. 15 above to write Thus, the continuous limit ofĤ B in Eq. 13 becomes, There are three ingredients in the Fano's model [68]: (i) a discrete state interacting with a (ii) continuum of states that form the eigenstates of the system to which (iii) transition from another discrete state [uncoupled to (i) and (ii)] is to be explored. Since N ex is a conserved quantiy,Ĥ B is decoupled in subspaces with different N ex . In the following we show thatĤ B,1ex -Ĥ B in N ex = 1 subspace -consists of (i) and (ii), whereas the transitions from the sole eigenstate ofĤ B,0ex (the ground state of the full system) to the eigenstates ofĤ B,1ex should give the Fano resonances.
In the single excitation subspace we can either have a cavity photon or a molecular excitation of the bright band. The corresponding states are |1 P ≡â † |vac and |S 0 (ω) ≡Ŝ + 0 (ω) |vac , where |vac = |0 P ⊗ |GS is the ground state of the full system that constitutes the subspace with zero excitation N ex = 0. So projectingĤ B to N ex = 1 subspace giveŝ We can see thatĤ B,1ex describes the interaction of a localised state |1 P with a continuum |S 0 (ω) . This model has been solved by Fano long ago [68] to explain the resonance bearing his name that occurs at small V (ω). Due to its relevance to the disordered TC model as shown above, we thoroughly explore this model in the strong coupling regime elsewhere [72] where the continuum bandwidth W is assumed to be the largest energy scale and all eigenstates lie within it. In real systems, e.g., organic microcavities, however, the distribution P (ω) is non-zero only in a finite energy window, which determines the bandwidth W of the bright states. So the following two scenarios are possible for the eigenstates of H B,1ex in general.
Small bandwidth: polaritons outside the continuum
At W 2ω R , depending on the cavity detuning from the centre of V (ω), we can have one or two polaritons as discrete states outside the bright band, along with a continuum of eigenstates at the original bright band energies, as shown in Fig. 1(c). These eigenstates can be easily calculated, see Appendix VIII A.
Assuming |Φ(ω) represents the discrete eigenstates of H B,1ex outside the bright band at energy ω =ω, we find that where the integration is over the bright band and the normalisation is left on purpose. The associated eigenvalues ω are given by the roots of where Note that there is no singularity in the above integral as, by definition, ω =ω in this case.
The roots of Eq. 20 can also lie within the bright continuum where F (ω) would be the principal value (p.v.) of the integral in Eq. 21. There can be up to three roots in the strong coupling case [72] that can be named after the states they belong to in the extreme cases of weak and strong coupling regimes by considering a localised distribution P (ω) [remember that V (ω) = ω R P (ω)]. If the width of P (ω) is taken to be σ, then at σ/ω R 1 we will only have a single rootω c at (or near) the cavity photon energy, while in the opposite case σ/ω R 1 two more rootsω LP andω U P exist at the lower and upper polariton energies.ω LP andω U P can be inside the continuum or outside it. When inside, the corresponding states |Φ(ω LP ) and |Φ(ω U P ) are still eigenstates of H B,1ex and produce polariton's Fano resonances. This will be explained in the following when we discuss the structure of the continuum eigenstates.
The eigenstates ofĤ B,1ex within the bright continuum have already been calculated by Fano [68], where he correctly handled the singularities involved. Assuming |Ψ(ω) is such an eigenstate at energy ω, we can write it as a superposition (Appendix VIII A), where The significance of the state |Φ(ω) can now be appreciated further by noting the sine and cosine terms in the coefficients of the two component states in this expression. For a physical process that can excite the system from |vac state to the eigenstate |Ψ(ω) , there are two transition paths available-via |Φ(ω) and |S 0 (ω) components-that can interfere if the corresponding amplitudes are finite for the two. This becomes quite interesting at the roots of Eq. 20ω inside the continuum, where the phase ∆(ω) jumps between ±π/2 and, since sine and cosine functions have opposite parity, the two paths interfere constructively on one side of the root and destructively on the other side, leading to an asymmetric lineshape there. This is the Fano resonance that will be discussed later in sec. III B. Since ∆ = ±π/2 atω, sin ∆ = ±1, cos ∆ = 0 there, and |Ψ(ω) = |Φ(ω) /πV (ω) showing that |Φ(ω) also represents the eigenstates inside the continuum at ω c ,ω LP ,ω U P . In the weak coupling case, when only a single rootω c exists, it will be called the photon's Fano resonance. In the strong coupling case, if the rootsω LP and/orω U P lie within the continuum, the resonances will be reminiscent of the discrete polariton states, and will hence be called polariton's Fano resonances.
Large bandwidth: polaritons inside the continuum
At W 2ω R , we do not have any discrete polariton eigenstates because all roots of Eq. 20 lie within the bright band, as shown in Fig. 1(d,e). In this case, all eigenstates are described by |Ψ(ω) in Eq. 22. This completes the solution of the disordered TC model in the thermodynamic limit for a generic distribution P (ω).
An interesting situation arises when a polaritonic root ω =ω LP/U P lies inside the bright band but the coupling V (ω) → 0 around its position. In such a case, ∆(ω) → 0 at ω =ω, but ∆(ω) → ±π/2 still holds. So, |Ψ(ω) → |S 0 (ω) at ω =ω, and |Ψ(ω) ∝ |Φ(ω) /V (ω) as always. This observation can be used to unify the above two cases of small and large bandwidth into the latter, by extending the bright band on either side and simultaneously reducing the coupling V (ω) → 0 in the extended region. Considering that V (ω) is sharply localised naturally as V (ω) 2 = P (ω) can be assumed to be a Gaussian distribution for energetic disorder in real systems, the above trick will be used henceforth and W will be assumed to be the largest energy scale.
Let's take where σ is the standard deviation that determines the width of the distribution and the mean energy ω 0 = ωP (ω)dω is taken as a reference. We can now evaluate F (ω) in Eq. 21 to obtain where erfi is the imaginary error functions. The contribution of the tail of the Gaussian (that may not be present in real systems) in the above integral is negligible (even around ω = ω on the tail -note that the p.v. of the integral in Eq. 25 measures the asymmetry of e −ω 2 /2σ 2 around ω in a small energy window as the denominator causes a suppression away from ω).
The response of the system to an optical, electronic, or a hybrid excitation can now be evaluated. To illustrate the characteristic nature of the eigenstates, let's first compare the inelastic electron scattering with the optical absorption. Later, we focus on the optical absorption as the main response function of the system.
Optical absorption vs. inelastic electron scattering
For a generic process that can excite the system from |vac to an eigenstate |Ψ(ω) , the excitation probability normalised with the bare excitation probability of the continuum state |S 0 (ω) is given by the Fano formula [68,72]. Let's assumeT to be the operator that describes the excitation of molecules in an inelastic electron scattering event. Obtaining this spectrum in experiments on microcavities is likely to be the easiest when a gas of molecules is confined between the cavity mirrors or flows through them. In typical organic microcavities with active molecules in a solid film of thickness ∼ 100nm sandwiched between two plane mirrors, the exposed region of the film along its thickness can be probed. We can consider the excitation of the dark molecular states (eigenstates ofĤ D in Eq. 14) as a background and focus on the bright band. The normalised probability for this process is given by [68,72], and the "asymmetry parameter" q is given by where we assumed the bare transition matrix element S 0 (ω)|T |vac is constant over the energy range of interest and used 1 P |T |vac = 0 as the electron scattering would not create cavity photons. For the optical absorption in our hybrid system, however, the bare transition amplitude to the bright continuum is zero, i.e., S 0 (ω)|â † |vac = 0, whereâ † is the transition operator for this process. We thus use the un- σ/ω R = 10 σ/ω R = 0.3 normalised probability given by, which is simply the photon spectral function. In the context of the weak coupling Fano resoance, the vanishing of S 0 (ω)|â † |vac also means that q = ∞, so the lineshape in the weak coupling regime is a Lorentzian whose width is given by the interaction strength [68,72]. The above result is exact as long as there is no cavity leakage or non-radiative emitter losses. We will include these using Green's function formalism later.
Fano resonances in the weak and strong coupling regimes
Both M(ω) and A(ω) give information about the eigenstates ofĤ B,1ex . However, their lineshapes are completely different due to the fact that they either excite the molecules or the cavity but not the both (as S 0 (ω)|â † |vac = 0 = 1 P |T |vac ). Figure 2 shows M(ω) and A(ω) at σ/ω R = 10, 0.3 and ω c = 0. At σ/ω R = 10 (blue dotted curves), the system is in the weak coupling regime [72] where we see the famous characteristic profiles aroundω c = 0 corresponding to q 0, ∞ -inverted Lorentzian dip in M(ω) and a Lorentzian peak in A(ω). In usual terms, the dip arises due to a destructive interference between two transition paths, one to the bare bright states |S 0 (ω) and the other to the modified (and broadened) discrete state |Φ(ω) (that now has components of the bright band as well). M(ω) drops to zero at the dip, indicating a complete destructive interference. For the same eigenstates, A(ω) behaves differently because it only excites the photon state.
At σ/ω R < 1, two new resonances appear atω LP andω U P where the phase angle ∆(ω) jumps discontinuously [72], which will now be seen in the two spectra. The black curves in Fig. 2 show M(ω) and A(ω) at and ω U P as a function of σ again at δ = 0 as in (a). The emitter weight W X in the state at the absorption peak is also shown, see the scale on the right. A finite splitting (ω R > 0) exists even in the weak coupling regime σ/ω R ≥ 1 [72]. Furthermore, even the strong coupling does not guarantee the hybrid excitations, W X 1 coexisting withω R ∼ ω R in the middle region.
σ/ω R = 0.3. We see sharp peaks around ω/ω R = ±1 in either spectrum, but the peak profiles of M(ω) and A(ω) are still very different. While A(ω) is not a superposition of two Lorentzians, as two (homogeneously) broadened polariton peaks would produce in case of identical emitters, it looks quite similar. However, M(ω) shows dips adjacent to these peaks where M(ω) vanishes, clearly establishing the difference between the eigenstates of our model from broadened polariton states, which will be discussed later in sec. IV B. While going from the weak to the strong coupling regime, a residual of original resonance still exists. At σ/ω R = 0.3, it is too small to be visible in A(ω) but shows its signatures in M(ω) that always vanishes at ω =ω c = 0.
A. Rabi splitting and polaritons
We now focus on the optical absorption and see how the Rabi splittingω R observed in the absorption spectrum depends on the energetic disorder σ, and wether a finite splitting necessarily means the formation of polaritons or the strong matter-light coupling. This is an important question as the optical spectrum can exhibit a similar two-peak structure even when the underlying quantum states are entirely different, as explained below. In the absence of energetic disorder,Ĥ B,1ex has only two polariton states given in sec. II A, which are observed in experimental optical spectra as two sharp peaks at ω U P , ω LP (Eq. 5,6) with a non-zero linewidth due to homogeneous broadening γ and cavity leakage κ (κ, γ ω R ). However, in the presence of energetic disorder, we have a continuum of states |Ψ(ω) that can also produce two peaks in the absorption just like the polaritons in the above case. In the following, we first present our results of optical spectrum which pertains to the states |Ψ(ω) . We will discuss the differences from the other case described above in sec. IV B. Figure 3(a) shows the evolution of the photon spectral weight A(ω) with σ for σ/ω R ∈ [0.2, 1.6] (shown red to yellow) for a resonant cavity, δ = ω c − ω 0 = 0. The roots of Eq. 20 are also shown as blue (ω U P ,ω LP ) and green (ω c = 0) dots to compare them with the peaks positions. At small σ, we have two sharp peaks for the upper and the lower polaritons which broaden and shift as σ increases, and finally merge together into a single broad central peak. At σ/ω R = 0.2, the peaks are ultra sharp and their locations agree toω U P ,ω LP (and the clean case ±ω R , Eqs. 5,6). This can be understood by ignoring the bright states away from ω 0 as they do not have significant effect on the polaritonic eigenstates formed by the coupling between |1 P and |S 0 (0) . However, at higher σ, still in the strong coupling case, the peaks are shifted slightly outwards fromω U P ,ω LP . This can also be explained as arising from the coupling of |1 P with the detuned bright states on either side of the ω = 0. Consider the bright states at ω 0. It will push both lower and upper polaritons upwards by an amount that would depend on their detuning from the polaritons, which is ω R − ω for the upper polariton but ω R + ω for the lower polariton. So the effect on the upper polariton will be stronger. Similarly, the bright states at ω 0 that pull down the polaritons will affect the lower polariton more strongly, leading to a net upward shift of the upper polariton and a net downward shift of the lower polariton. The bright states within the two polaritonic states will have this effect. At σ/ω R 1, the effective collective coupling of these states becomes weaker and comparible to the bright states outside this polaritons window, so all these states obtain significant photon component leading to excessively broad absorption peaks. At σ/ω R ≥ 1, ω U P ,ω LP do not exist and we are in the weak coupling regime with a single rootω c = 0, but, as we can see, A(ω) still keeps its two-peak structure at σ/ω R = 1, 1.2.
Thus, as σ increases, the interactions transform the polaritons into two polariton's Fano resonances, which are eventually replaced by a single photon Fano resonance at σ/ω R ≥ 1. For clarity, we call it a polariton's Fano resonance when the hybrid polaritonic character of the eigenstate is compromised due to Fano broadening. We can think of the polariton's Fano resonances as arising from the interaction of polaritons (formed by cavity mode and bright states that are resonant or more strongly coupled to it) and off-resonant or less strongly coupled bright states, as illustrated in Fig. 1(c-d) and also discussed in sec. III A 1. However, it is worth noting that this is not strictly the same as Fano resonances arising from a weak coupling between two discrete states and a continuum [68]. At large enough σ, there are not enough states or coupling strength to create the strong coupling resonances, so we obtain a single photon Fano resonance in this weak coupling regime.
It is also interesting to see that at ω =ω U P ,ω LP ,ω c , the expression for A(ω) in Eq. 30 can be simplified (as the first term in the denominator vanishes) to obtain ω R A(ω) = 2/π 3 σ/ω R exp(ω 2 /2σ 2 ), so that the size of the central peak is always proportional to σ. Since the spectral function is normalised, increasing σ thus decreases the linewidth of the central peak at σ/ω R > 1. This gives the correct picture in case of σ/ω R 1 where the coupling becomes too weak and the cavity photon state is only slightly perturbed by the interaction with the bright continuum.
In Fig. 3(b), the Rabi splittingω R or the position of the upper polariton peak in A(ω), and the actual location of the related resonanceω U P , are shown as a function of σ again at δ = 0. We see that, as σ increases,ω R slightly increases at small σ and then decreases sharply to zero at large σ where the two peaks in Fig. 3(a) merge together.ω U P also shows a similar behaviour but slightly quicker leading to an interesting situation where A(ω) has two peaks withω R > 0 even in the weak coupling regime, σ/ω R ≥ 1. In the strong coupling case, these peaks should exist because the first term in the denominator of Eq. 30 becomes non-monotonic [72]) (vanishes at multiple locations) but here it is sustained because its sum with the second term still bears a bimodal structure.
Thus we find important implications for the experiments on disordered systems: the Rabi splitting, or the anticrossing in the optical absorption or transmission spectrum alone cannot guarantee the formation of polaritons or the strong coupling. Besides, strictly speaking, these two terms should not be considered synonymous in case of disordered systems as there is a wide range of σ/ω R in the strong coupling regime where even the eigenstates containing the largest photon spectral weight are still far from truly hybrid polaritons. This will be further explained in the following section.
B. Inhomogeneous vs homogeneous broadening
Let's see how the effects of an inhomogeneous broadening or energetic disorder σ are different from that of a homogeneous broadening or the intrinsic linewidth γ of the emitter states. We can compare the spectrum in Fig. 3(a) to that of a system with no energetic disorder but a large homogeneous broadening, e.g., atoms in a microcavity as in Ref. [63] (see Fig. 2 there). While the spectra look similar, their physical interpretation depends on the nature of the quantum states of the system, which is different in the two cases. The homogeneous broadening of the emitter states γ broadens the polaritonic eigenstates or polaritons around their expected energy (±ω R at δ = 0) and keep their polaritonic character intact. In contrast, the inhomogeneous broadening σ (that creates the continuum of the bright emitter states in the first place) broadens the photon spectral function only without broadening the eigenstates, making the polariton states at the absorption peaks more excitonic or emitter-like as σ increases. That is, the polaritonic character of the eigenstates is distributed over larger number of eigenstates, making even the most polaritonic state nearly excitonic when the Fano broadening spans a large number of emitter-states in the bright continuum. This is best illustrated by considering a specific density of continuum states and numerically evaluating the emitter weight of the most polaritonic eigenstate. For example, taking one bright state every ∆ω = 0.003ω R , the emitter weight W X of the states at the absorption peaks is plotted in Fig. 3(b) [black curve] as a function of σ at δ = 0. We see that W X ∼ 1/2 (polaritons) at σ 0.3 where the Fano broadening is still smaller than the "width" ∆ω of a single eigenstate, but quickly approaches unity around σ 0.5 where the Fano broadening distributes the photon spectral weight over many eigenstates. Thus polaritons lose their meanings and the peaks in the optical absorption simply correspond to strongly absorbing emitter states. As the figure shows, this happens at σ/ω R as low as 0.5 withω R ω R , i.e., well below the threshold for the weak coupling σ/ω R = 1.
This raises another question that ought to be discussed here. What does ∆ω corresponds to in a real system? An intuitive answer to this question is that it plays the role of the intrinsic linewidth or the homogeneous broadening γ of the emitter states (i.e., ∆ω γ). Increasing the homogeneous broadening γ for the emitter states should extend the bright band on either side by γ/2 but reduce the number of such homogeneously broadened bright states in the band ( W/γ). This should enhance the polaritonic character of the states at the cost of their broadening. In the extreme case γ → W , the bright band will become a homogeneously broadened bright state that would produce two homogeneously broadened polaritons as observed in Ref. [63].
C. Fano broadening: shape and width of polaritonic peaks
Let's focus on the Fano broadening and the lineshape asymmetry in the strong coupling regime even when the corresponding weak coupling case exhibits a symmetric Lorentzian profile. We find that the Fano broadening appears even at fairly small σ. Figure 4(a) shows the scaled spectral weight ω R A(ω) at σ/ω R = 0.25 and δ/ω R ∈ [−1, +1]. We see that as a polariton peak gets near the bright emitter band, its linewidth increases and it transforms into a Fano resonance. This also accompa- nies a slight shift in the peak position compared to the clean case σ → 0, Eqs. 5,6, and the actual locations of the resonanceω U P ,ω LP , as discussed before for δ = 0. Considering the peak corresponding to the lower polariton, Fig. 4(b,c) show its linewidth Γ and a simple measure of the asymmetry of its lineshape, the ratio of the half linewidths of the low (Γ R ) and the high (Γ B ) energy side of the peak, as functions of σ for the same set of δ values as in Fig. 4(a). We see that Γ increases with σ much more strongly at large positive detuning (bluish curves) where the lower polariton is closer to the bright emitter band and would be more matter-like even at σ = 0. The lineshape asymmetry shown in Fig. 4(c), the deviation of Γ R /Γ B from 1, also shows the same trend.
We will now consider a small homogeneous broadening and treat it using the standard Green's function formalism. To treat both homogeneous and inhomogeneous broadening of the emitter states on equal footing, new methods need to be developed, which is clearly beyond the scope of the present work.
V. OPTICAL ABSORPTION IN THE PRESENCE OF LOSSES
So far, we have not considered the losses, a finite cavity leakage and non-radiative emitter decay that amount to homogeneous broadening of the bare cavity and emitter states. To see the effect of these losses on the spectral function A(ω) or optical absorption, we need the photon Green's function G(ω) as A(ω) = − G(ω), which will now be used instead of Eq. 30. Here stands for the imaginary part. Due to the absence of the coupling between the emitters or their bright states, G(ω) can where Σ is the self energy and κ, γ are cavity leakage and non-radiative emitter relaxation rates. Computing the self energy Σ(z) for the Gaussian P (ω) readily gives which can now be used with Eq. 31 to explore the optical absorption including the losses.
A. Lineshape and linewidth
Let's now include the effect of cavity and emitter losses on the total broadening of absorption peaks. In systems where κ γ, e.g., organic microcavities [25,88], we can ignore γ to understand how κ changes the picture discussed so far. Figure 5(a) shows Γ as a function of δ at σ/ω R = 0.25, γ = 0 and κ/ω R ∈ [0, 0.05]. We see that at large negative detuning, |δ| σ, where the polariton is away from the band and Fano broadening can be ignored, the effect of κ is the strongest due to larger photon fraction. As δ increases to zero and positive values, the effect of κ decreases but that of σ increases. This leads to an interesting situation where Γ develops a minimum at an optimum δ. The minimum gets deeper and more prominent if Γ at extreme δ values is large and of similar size. Noting that at δ/ω R = −1, Γ ∼ κ and at δ/ω R = +1, Γ ∼ Γ σ ≡ Γ(σ, δ/ω R = +1, κ = 0 = γ) [blue curve in Fig. 4(b)], the optimum κ to observe such a minimum is κ = Γ σ . This is shown in Fig. 5 We see that the smaller the σ, the larger the percentage change in Γ, and the larger the optimum δ. It should be observable in typical organic microcavities where κ ∼ 0.05eV [88], 2ω R ∼ 1eV [34], and σ ∼ √ 2 × 0.1eV [24,31,32] so σ/ω R ∼ 0.28, which gives the minimum at δ ∼ 0.
B. Dependence on the energy distribution
If the emitter transition energies are distributed more evenly, the distribution function is flatter in the middle and falls more sharply at the edges. We would expect it should result in a sharper transition between polaritons and their Fano resonances as the polariton energy starts overlapping the bright states band. Taking the extreme case of a flat distribution, P (ω) = 1/W, −W/2 ≤ ω ≤ W/2, 0 otherwise, we find that the transition actually becomes more gradual. This is because in this case all bright states would couple equally strongly to the cavity mode (as V (ω) = ω R / √ W = constant) and increasing W would push the polaritons away from the band edges at small to intermediate W . In this case, Eq. 32 gives Figure 6(a) shows the scaled spectral weight −ω R G(ω) for the flat distribution (Eq. 31 with Eq. 34) at δ = 0, κ/ω R = 0.05, γ/ω R = 10 −4 (κ, γ typical of organic microcavities, [25,88]) and W/2ω R = 0.01, 0.5 − 2.5 in increments of 0.5, marked with dotted vertical lines. We see that it has a feature resembling Fano lineshape with a dip but note the logscale, these structures differ from the weak coupling Fano profile indeed. At W < 2ω R , the peaks resemble isolated polaritonic states. Despite the change in lineshape at W ≥ 2ω R , they still remain quite hybrid for a while. W X for the lower polariton state at the peak position at δ = 0 = κ = γ is shown in Fig. 6(b). We see that W X increases from 1/2 to 1 over a wider σ range: W X 0.7 only at W/2ω R = 1 and it increases to 0.9 around W/2ω R = 1.5. Compare this to the gaussian energy distribution, where W X jumps to ∼ 0.95 around σ/ω R = 0.5 only. This sluggish increase in the emitter weight for the flat distribution can be understood as a result of poor correspondence of ω R itself to the Rabi splittingω R , because the bright states at the band edge couple equally strongly to the cavity mode and their effective detuning from it shifts the polariton energy more significantly. Another difference from the gaussian distribution case is that the photon Fano resonance appears before the polaritons' Fano resonances disappear. So the notion of the Rabi splittingω R also becomes relatively vague at W/2ω R 2 in this case. Interestingly, however, for polaritonic peaks, a significant linewidth narrowing occurs at W 2ω R , as can be seen in Fig. 6(a).
VI. SUMMARY AND CONCLUSIONS
We solve the disordered TC model by identifying and decoupling the exact bright and dark sectors of the emitters' Hilbert space. Explicit expressions for the bright and the dark states are presented, allowing the exact numerical diagonalisation for arbitrarily large systems. For the emitter states forming a continuum in the thermodynamic limit, we obtain Fano's model whose solution has already been around for a long time. The Fano resonances in the optical absorption and inelastic electron scattering spectra are presented both in the strong and the weak coupling regimes, where they correspond to polaritons or their Fano resonances and photon's Fano resonance, respectively. We find that the hybrid light-matter character of the excitations could be lost due to strong energetic disorder while the optical spectra still showing the Rabi splitting and anticrossing. We also explore the effect of cavity and emitter losses on the linewidth and lineshape by calculating the photon Green's function. We find that the polariton's linewidth should exhibit a minimum as a function of the detuning when the cavity losses are comparable to the maximum Fano broadening.
VII. OUTLOOK
We studied the model in the rotating wave approximation that ignores the counter rotating terms in the light-matter interaction. These terms introduce coupling within a given parity subspace (N ex even or odd) which are usually ignored below the "ultra-strong" coupling regime (ω R 0.3ω 0 ) due to a large energetic difference ( 2ω 0 ) between the coupled states. However, energetic disorder would reduce this energetic difference and, at large energetic disorder, the highest end of the bright band or the upper polariton in N ex = 1 subspace would come close enough to (or overlap with) the lowest end of the bright band or the lower polariton in N ex = 3 subspace such that their interaction cannot be ignored any longer. Similarly, the effects of an interaction that does not respect the excitation number parity (coupling to a bath, for instance) on the eigenstates of the system, would be enhanced even further as it would only require to reduce the gap ( ω 0 ) between the states of adjacent excitation spaces (e.g., N ex = 1 and N ex = 2). Investigation of such effects of the energetic disorder can be carried out in a future study. Considering a weak coupling between the bright and dark spaces due to dipole-dipole interaction and application of the two spaces in other systems, e.g., those described by the Anderson impurity model, and exploring the effects of Fano broadening on highly excited states such as polariton condensate and lasing would also be interesting future works.
ACKNOWLEDGMENTS
The author thanks Rukhshanda Naheed for fruitful discussions.
Eigenstates inside the bright band
Assuming |Ψ(ω) is an eigenstate ofĤ B,1ex at an energy within the bright continuum, we can again write it as a superposition, where the coefficients α, β can be written in terms of ∆(ω), F (ω) given in the main text as (see Ref. [68] for detailed derivation), α(ω) = sin ∆(ω)/πV (ω), Here, δ is the Dirac delta function. Substituting these back, we obtain, which along with |Φ(ω) in Eq. 19 at ω inside the bright band (where it does not represent an eigenstate in general), gives Eq. 22.
B. Calculation of G(ω)
The photon Green's function G(ω) can be obtained [30] from the resolvent R(ω) ofĤ B,1ex , as follows. We start from the discrete versionĤ B in Eq. 13. In the single excitation space, it can be written in the matrix form as, to take the inverse, where the subscripts with the matrix elements show the dimensions of each block. The matrix element of R(ω) corresponding to the cavity state is the photon Green's function G(ω), given by [89] G(ω) = (A − BD −1 C) −1 . Since D is diagonal, its inverse can be written directly, and we obtain where we have added the imaginary components κ, γ to ω c , ω n to include a finite cavity leakage and non-radiative emitter relaxation. Considering the continuous limit leads to Eq. 31. | 2022-08-26T01:16:06.329Z | 2022-08-25T00:00:00.000 | {
"year": 2022,
"sha1": "ebb6ced840fe8f4bae5eccbf7bac82a1754a8c7e",
"oa_license": null,
"oa_url": null,
"oa_status": null,
"pdf_src": "Arxiv",
"pdf_hash": "ebb6ced840fe8f4bae5eccbf7bac82a1754a8c7e",
"s2fieldsofstudy": [
"Physics"
],
"extfieldsofstudy": [
"Physics"
]
} |
257559816 | pes2o/s2orc | v3-fos-license | Dynamic CD8 + T cell responses to cancer immunotherapy in human regional lymph nodes are disrupted in metastatic lymph nodes
SUMMARY CD8 + T cell responses are critical for anti-tumor immunity. While extensively profiled in the tumor microenvironment, recent studies in mice identified responses in lymph nodes (LNs) as essential; however, the role of LNs in human cancer patients remains unknown. We examined CD8 + T cells in human head and neck squamous cell carcinomas, regional LNs, and blood using mass cytometry, single-cell genomics, and multiplexed ion beam imaging. We identified progenitor exhausted CD8 + T cells (Tpex) that were abundant in uninvolved LN and clonally related to terminally exhausted cells in the tumor. After anti-PD-L1 immunotherapy, Tpex in uninvolved LNs reduced in frequency but localized near dendritic cells and proliferating intermediate-exhausted CD8 + T cells (Tex-int), consistent with activation and differentiation. LN responses coincided with increased circulating Tex-int. In metastatic LNs, these response hallmarks were impaired, with immunosuppressive cellular niches. Our results identify important roles for LNs in anti-tumor immune responses in humans.
INTRODUCTION
Immune checkpoint blockade (ICB) immunotherapy targeting the PD-1/PD-L1 axis has revolutionized oncology, but the underlying mechanisms remain incompletely understood, especially in humans.CD8 + T cells are central effector cells that mediate the efficacy of ICB and have been extensively studied in the tumor microenvironment (TME).However, recent studies found that CD8 + T cells in the periphery, such as in secondary lymphoid organs (SLOs) including tumor-draining (td) lymph nodes (LNs), are integral for ICB response in mouse models.During the initiation of CD8 + T cell responses, naive CD8 + T cells are typically primed by dendritic cells (DCs) in the LN before trafficking through the blood to the tumor. 1 Blocking egress of lymphocytes from SLOs using the S1PR-blocking agent FTY720 abrogates ICB efficacy, 2,3 and targeted administration of anti-PD-L1 to the tdLN is sufficient to control tumor growth in mice. 46][7][8] Despite the potential importance of the tdLN in ICB-driven CD8 + T cell responses, a lack of LN sampling in clinical datasets leaves many unanswered questions about the relationship between immune responses in the LN and tumor in human cancer patients.
In settings of chronic antigen exposure, such as cancer and chronic infection, antigen-specific CD8 + T cells can become exhausted or dysfunctional, exhibiting elevated expression of inhibitory receptors such as PD-1. 9In mice, progenitor or precursor exhausted cells (Tpex) expressing the transcription factor TCF-1 undergo self-renewal and have high proliferative capacity. 10,113][14][15][16][17] In mice, Tpex are necessary for the maintenance of antigen-specific CD8 + T cell responses in settings of chronic antigen stimulation. 10,18In response to ICB in mice, Tpex are the primary exhausted subset that expands and differentiates into Tex-term, driving an increase of Tex-term in the tumor. 12,14ex-term themselves do not substantially proliferate in mice with ICB treatment, 12,14,19 though they may become more activated. 20f note, the tdLN is the primary site where Tpex are maintained and stimulated in mouse models of cancer, 21,22 suggesting that ICBresponsive CD8 + T cells may reside in the tdLN.
It remains unclear to what extent these findings translate to humans.4][25][26][27][28] A higher ratio of TCF-1 + CD8 + TILs to TCF-1 À CD8 + TILs in melanoma was associated with longer survival after anti-PD-1 treatment. 240][31] Furthermore, in addition to exhausted cells, potential tumor-reactive CD8 + T cells in human tumors and peripheral blood have been identified as Trm and Temra cells, respectively. 28Given that most patients do not currently exhibit durable responses to ICB, 32 identifying which CD8 + T cell subsets are targeted by ICB in patients, and where activation takes place anatomically, will be necessary to improve therapeutic efficacy.
In this study, we examined the relationship between CD8 + T cell responses across key anatomic sites, including TME, tdLN, and blood in human cancer patients.Patients with advanced head and neck squamous cell carcinoma (HNSCC) are routinely treated with surgery, including resection of regional LN, providing a unique opportunity to address these open questions.Through a combination of single-cell analyses by mass cytometry (CyTOF), single-cell RNA sequencing and TCR sequencing (sc-RNA+TCR-seq), sc-RNA and protein sequencing (CITE-Seq), and multiplexed ion beam imaging (MIBI), we examined CD8 + T cells across tissues from patients treated with surgery as standard of care (SOC) and patients treated with perioperative anti-PD-L1 ICB (atezolizumab) enrolled in a clinical trial at UCSF Medical Center (NCT03708224).We identified a subset of Tpex in uninvolved, regional lymph nodes (uiLNs) that were clonally related to Tex-term in the TME.Patients treated with ICB shortly before surgery exhibited a decrease in Tpex frequency in uiLNs with a concomitant increase in more differentiated Tex-int, which were localized in proximity to DCs.Changes in uiLNs correlated with an increase in proliferating Tex-int in peripheral blood.Moreover, regional lymph nodes with tumor metastasis (metLNs) exhibited an impairment in these responses to ICB associated with immunosuppressive cellular niches around Tpex and a reduction in the circulating CD8 + T cell response.These results highlight a central role for uiLNs in mediating responses to ICB, which may create new opportunities for next-generation immunotherapies focused on optimally harnessing these responses.
RESULTS
Tpex are increased in uninvolved lymph nodes of HNSCC patients Using mass cytometry, we profiled immune cells in the TME and LNs of 10 patients with locally metastatic HNSCC who had both tumor and matched uiLNs and/or metLNs available for analysis (Figure 1A; Table S1).All tissue specimens were collected immediately following tumor resection and neck dissection for each patient.To understand the immune composition present within each tissue, we manually gated major immune cell subsets (Figures S1A and S1B).As expected, CD4 + T cells and B cells were more frequent within uiLNs, while other immune cell subsets were more frequent in the tumor as a percentage of total immune cells (Figure 1B).CD8 + T cells were present at similar frequencies in both tissues (Figure 1B).
Tpex in LNs are clonally related to Tex within the TME If anti-tumor CD8 + T cell responses are coordinated across tissues, we would expect clonally related CD8 + T cells to be present in both LN and tumor.Therefore, we hypothesized that CD8 + Tpex cells in the LN are clonally related (i.e., share the same T cell receptor [TCR] sequence) to more exhausted CD8 + T cells within the TME.To assess CD8 + T cell clonality, paired tumor and LN tissues surgically resected as SOC treatment from 5 patients with HNSCC (Table S2) were subjected to sc-RNAseq, of which paired TCR-seq data were generated from 4 patients and paired single-cell protein expression data (CITE-seq) were generated from 3 patients.Unsupervised clustering identi-fied major cell populations in the TME and LN (Figure S2A), annotated based on the expression of canonical marker genes (Figure S2B).We identified CD8 + T cells based on a combination of RNA expression, protein expression (when available), and cluster annotation and sub-clustered these cells (Figures 2A, A B C D Tex-term (Figures 2C, 2D, and S2H). 28Indeed, our cluster annotations for Tpex and Tex mirror the clusters with the highest expression of these scores, respectively.In addition, cluster 7 exhibited higher expression of both scores, reminiscent of Texint, which was confirmed by analysis of differentially expressed genes (Table S3).
E F G
Based on TCR CDR3 sequences, we identified clonally related CD8 + T cells (termed ''shared clones'') in the LN and TME of each participant with available TCR-seq data (3-29 clones per participant; Figures 2E and S2I).Shared clones from cells localized in the LN were more likely to belong to the TCF7 + exhausted cluster (cluster 6), while their counterparts in the TME were more likely to belong to the Tex-int/Tex-term clusters (p = 0.0005, odds ratio = 5.71 by Fisher exact test) (Figure 2E).Moreover, cells belonging to shared clones in the LN exhibited significantly higher TCF7 + exhausted gene set scores (Figure 2F), while their counterparts in the TME exhibited significantly higher Tex-term gene set scores (Figure 2G).At the individual clone level, the difference between the average TCF7 + exhausted (Tpex) and average Tex-term scores was higher for cells located in the LN compared to those in the TME for the majority of clones (Figure 2H).Collectively, these results indicate that shared CD8 + T cell clones exist in the LN and TME across patients, albeit in different functional states.CD8 + T cells in the LN were more likely to be in a state resembling Tpex, while their counterparts in the TME were more likely to be in a Tex-term state.Because CD8 + T cells lose functional capacity as they differentiate from Tpex to Texterm, [12][13][14][15][16][17] these results underscore the importance of the LN in maintaining a reservoir of functional anti-tumor CD8 + T cell clones in humans, which could represent an important target for immunotherapies.
Tpex are primarily localized in the T cell zone in human uiLNs
We next sought to understand the Tpex architectural niche in the uiLN as well as what effect ICB might have on their organization.We recently treated 10 patients with human papilloma virus (HPV)-negative HNSCC with 1-2 cycles of the PD-L1 inhibitor atezolizumab prior to surgery through clinical trial NCT03708224 (Table S4).We obtained uiLN specimens from 9 of these patients and LN with tumor metastasis (metLN) from 4 patients.Additionally, we obtained uiLN resected as SOC treatment from HPV-negative HNSCC patients who were not treated with immunotherapy (n = 22).LN tissue cores were randomized across two tissue microarrays, of which sections were stained and analyzed by multiplexed ion beam imaging (MIBI) (Figure 3A, Table S5).Following image acquisition, pixels were clustered according to their intensity values for each measured parameter (Figure S3A), individual cells were segmented using the Mesmer algorithm, 36 and LN tissue regions were identified, including B cell follicles-/B cell-rich zones, T cell-rich zones, and metastatic tumor regions, when present (Figure 3B).Single cells were then clustered by FlowSOM based on their overall clustered pixel composition (Figure S3B), identifying major cell populations (Figures 3C, 3D, S3C, and S3D).
We first focused our analysis on CD8 + T cells within untreated uiLNs.CD8 + T cells were sub-clustered by FlowSOM and partitioned into 18 clusters based on their protein expression (Figures 3E, 3F, and S3E).Consistent with our prior results, a population of PD-1 + TCF-1 + Tpex (c18) was identified in the uiLN in addition to two clusters of Tex-int (c16, c5), which were distinguished by their differential expression levels of PD-1 and TCF-1.Leveraging this high-dimensional spatial information, we assessed where these cells were localized in uiLNs and with which other cell types they were adjacent.Tpex cells can express CCR7, 14 which promotes migration to T cell zones of LNs, 37 as well as CXCR5, 12 which in contrast drives migration toward B cell zones. 38Tpex were more likely to be localized in the T cell zones, though a fraction were also localized in B cell zones (Figure S3F).Within the T cell zone, Tpex (c18) and Tex-int (c16) were often found neighboring CD4 + T cells, Tregs, CD4 + CD8 + (double-positive) T cells, and DCs, which was readily observable in the MIBI images (Figures 3G-3I).
anti-PD-L1 ICB impacts the frequency of Tpex and Texint cells and their cellular neighborhoods in uiLNs
The identification of Tpex in uiLNs prompted the question of whether and how Tpex and their surrounding cell neighbors respond to ICB in humans.Therefore, we also evaluated uiLN tissues from patients who received anti-PD-L1 ICB prior to surgery.Compared with the SOC uiLN (Figures 3D and S3D), the frequency of major cell populations was not dramatically altered by therapy (Figure S4A).However, changes in the composition of CD8 + T cell subsets were clearly evident (Figures 4A, 4B, and S4B).We therefore hypothesized that anti-PD-L1 ICB may change the frequency and activation state of Tpex and Tex-int in uiLNs.Indeed, in the T cell zone, Tpex (c18) had the largest decrease in frequency among CD8 + T cell clusters in (A) Overview of cohort.uiLN samples were obtained from patients with HNSCC that received standard of care (SOC; n = 23) or 1 to 2 cycles of anti-PD-L1 treatment prior to surgery (n = 9).For anti-PD-L1-treated patients that had metastatic disease (n = 4), metLN samples were also obtained.Tissue cores were distributed across two tissue microarrays for analysis by multiplexed ion beam imaging (MIBI).(G and H) Percentage of (G) cluster 18 and (H) cluster 16 cells with a specific cell type as its neighbor in uiLN (T cell zone) from SOC-treated patients.(I) Representative image of an uiLN from a SOC-treated patient (patient 23) showing the spatial localization of CD4 T cells, Tregs, DCs, Tpex (cluster 18), and Texint (cluster 16).Cell identity overlaid onto the segmentation mask.Highlighted regions 1 and 2 are colored by the expression of CD8 + (red), PD-1 (green), TCF-1 (cyan), CD11c (purple), Ki-67 (yellow), CD4 (cyan), and FoxP3 (purple).See also Figure S3.(legend continued on next page) anti-PD-L1-treated patients compared to SOC-treated patients (Figures 4B and 4C).This change was specific to the T cell zone, as Tpex frequencies were not meaningfully altered in the B cell zone (Figure S4C).In contrast to Tpex, Tex-int (c16) had one of the largest increases among CD8 + T cell clusters in the T cell zone after treatment, though frequencies were somewhat variable across patients and LNs (Figures 4B and 4C).As a result of these effects, the ratio of Tpex to Tex-int significantly decreased in uiLNs from ICB-treated patients (Figure 4D), consistent with a shift toward more differentiated phenotypes (e.g., Tex-int) within the exhausted CD8 + T cell compartment after ICB.Other clusters with similar increases included a CD45RO À TIM3 + PD-1 À cluster (c17) and a CD45RO À HLA-DR + cluster (c12), the latter of which may represent recently activated CD8 + T cells (Figure 4B).While the proportion of Tpex that were proliferating (Ki-67 + ) did not change for most patients post-treatment, strikingly, 100% of Tpex were Ki-67 + in an uiLN resected from the only patient who experienced a major pathological response to therapy (patient 11) (Figure 4E).In contrast, Texint exhibited a trend toward increased proliferation in patients treated with anti-PD-L1 (Figure 4E).Next, we evaluated whether the spatial localization of Tpex and Tex-int cells was altered within uiLNs following ICB by assessing each cell's local neighborhood (STAR Methods). 39,40otably, both Tpex (c18) and Tex-int (c16) were more frequently localized in proximity to DCs (Figures 4F and 4G).Several clusters of CD8 + T cells had significantly higher numbers of DC neighbors after anti-PD-L1 treatment (Figure S4D), including Tpex and Tex-int, which exhibited the largest fold-change and the most significant p value, respectively (Figures 4H and S4D).These data are consistent with mouse cancer models in which DCs in the LN mediate responses to ICB. 41 However, the frequency of DCs was not different after anti-PD-L1 treatment (Figure S4E), nor did DC frequencies significantly correlate with Tpex or Tex-int frequencies (Figures S4F and S4G).Additionally, Tpex were more likely to be near Tex-int after treatment, consistent with ongoing activation and differentiation into a Tex-int functional state (Figure 4I).Interestingly, both Tpex and Tex-int had more neighboring PD-1 + and Ki-67 + cells around them after treatment (Figures 4J and 4K), suggestive of cellular niches containing activated and proliferating cells.Indeed, Tpex and Tex-int in proximity to DCs exhibited higher expression of PD-1 after treatment, suggesting that they were activated and/or differentiating, which was evident when visualizing MIBI images (Figures 4L and 4M).In summary, uiLNs resected after anti-PD-L1 ICB treatment exhibited a decrease in Tpex frequency with a concomitant increase in Tex-int that were localized close to Tpex, raising the possibility that treatment caused Tpex to transition to Tex-int.Furthermore, both Tpex and Tex-int were more likely to be located near DCs and exhibited elevated PD-1 expression compared to patients who had not received anti-PD-L1 immunotherapy, supporting a role for DCs in uiLNs in mediating anti-tumor CD8 + T cell responses.
Tex-int increase in the blood following ICB Given the significant CD8 + T cell responses within uiLNs of treated patients, we wanted to determine whether similar changes could be detected within the blood and tumor, as an expansion of activated CD8 + T cells in the blood has been previously associated with clinical response to ICB across several types of solid tumors, 6,7,[42][43][44][45] though the source of these cells has remained speculative.For all 10 patients treated with neoadjuvant anti-PD-L1 ICB, blood was collected at baseline just prior to their first infusion as well as on the day of surgery, and resected tumor tissue was available for analysis from 6 patients (Figures 5A and S5A).For 6 patients, blood was also collected at a follow-up visit approximately one month after their surgery.Fresh uiLN tissue was also available from one patient (11), collected during surgery (Figure S5A).As expected, mass cytometry analysis revealed clear differences in the frequency of major immune cell populations between matched tumor and blood from the same patients, regardless of when blood was collected, as well as in the uiLN sample (Figures S5B and S5C).
Unsupervised clustering of CD8 + T cells from all samples partitioned cells into 20 clusters, which reflected distinct differentiation states (Figures 5B and 5C).Cluster 6 expressed the highest levels of TCF-1 among the PD-1 + clusters, corresponding to Tpex, and was present in the uiLN sample; however, it was detected only at low levels in blood and tumor samples (Figures 5B, 5C, and S5E), implying that the decreased frequency of Tpex observed in treated uiLNs was not due to Tpex egress from LNs.In contrast, Tex-int clusters (c5, c9) were detectable in the blood (Figure 5D) and significantly increased after treatment (at the time of surgery) compared to baseline, along with a Temra cluster (c1) (Figures 5E and 5F).At the 1 month follow-up, similar changes were observed (Figures 5G and 5H).While most patients had some increase in Tex-int frequency at follow-up compared with baseline, two patients (3 and 11) had substantially larger increases (Figure 5H).Although the neoadjuvant clinical trial design precluded assessment of clinical response by radiology (RECIST criteria), patient 11 was the only patient with significant tumor treatment response microscopically (i.e., most of the tumor was necrotic) and is still alive nearly a year after surgery.Additionally, patient 3 has not had signs of tumor recurrence and is still alive nearly three years after surgery.
Tex-int cells express Ki-67 in the blood after anti-PD-L1 ICB To further evaluate the activation state of Tex-int (c5, c9) and Temra (c2) in the blood, we assessed the proportion of cells expressing the proliferation marker Ki-67.Consistent with their increased frequencies within blood after treatment, Tex-int (c5) and Temra (c2) exhibited significant increases in their proliferating fraction after ICB (Figure 5I), while the other Tex-int cluster (c9) did not (Figure S5F).Comparing the phenotypes of the Texint clusters, c5 expressed higher levels of the memory-associated co-stimulatory receptor CD27, while c9 expressed higher levels of the effector-associated transcription factor Tbet and TIM3, suggesting that c9 may be at a later differentiation state compared to c5 (Figure 5B).It is plausible that Ki-67 + Tex-int in circulation could originate in uiLNs or the tumor.Because we previously found increased proliferation in the uiLN, we also investigated whether there was evidence of proliferating Tex-int in the TME.Tex-int (c5) were significantly more proliferative in the blood than in the tumor of the same patients, where little proliferation was evident (Figure 5J).This was also true for the proliferative Temra population (c2) observed in the blood (Figure 5J).Consistent with a peripheral origin of Tex-int, these cells were also more frequent in blood compared to matched tumors after treatment (Figures S5G and S5H).Finally, we assessed whether these populations were correlated across uiLN, blood, and tumor from the same patients (Figures S5I and S5K).Indeed, both Tpex (c18) and Tex-int (c16) in the uiLN were positively correlated with both Tex-int clusters (c5, c9) in the blood (Figures 5K and S5J).While Tex-int (c16) in the uiLN were also positively correlated with Tex-int clusters in the tumor (c5, c9), Tpex (c18) in the uiLN exhibited negative correlations with these clusters in the tumor (Figures 5K and S5K).Therefore, increased frequencies of Tex-int in the tumor after treatment were associated with fewer Tpex within the treated uiLN and an increase in the frequency of Tex-int in the uiLN.These data are consistent with a model in which Tpex differentiate to Tex-int within the uiLN after treatment, which then transit through the blood to the TME.
Metastases within LNs alter immune composition
As with many solid tumors, HNSCC frequently metastasize to regional lymph nodes.Having established an important role for uiLNs in anti-tumor CD8 + T cell responses, we asked whether metastases in LNs (metLNs) would impact this response.Using mass cytometry, we profiled the immune cells present in paired metLNs and uiLNs from the same SOC patients with locally advanced disease, overlapping with the patients we had previously analyzed (n = 7).Paired metLNs and uiLNs were regional and ipsilateral to the same primary tumor; distant metastases were not identified at the time of surgery.Immune composition within metLNs was very similar to the tumor and distinct from paired uiLNs from the same individuals (Figures S6A-S6C).This pattern extended to CD8 + T cells, in which Tpex were almost absent in metLNs, while Tex-term were more abundant (Figures 6A and 6B).In contrast, no significant differences were observed between CD8 + T cell composition in metLNs and paired tumors (Figure S6D).Thus, metLNs exhibit a distinct immune environment as compared to paired uiLNs but recapitulate a similar immune microenvironment to the primary tumor from which they originated.
Having established that uiLNs are an important site for CD8 + T cell responses to ICB, we evaluated whether these responses remained intact or differed in metLNs.Of the 9 patients treated with anti-PD-L1 from whom uiLNs were collected, 4 patients also had at least one LN with evidence of metastasis, with metastatic cores (met-cores) from 3 patients available for MIBI analysis.Differences in major cell population frequencies mirrored the results from mass cytometry (Figures S6E and S6F).Within the non-naive CD8 + T cell compartment, Tpex (c18) and Tex-int (c16) were also among the clusters with the largest relative reductions in frequency within met-cores as compared to ui-cores after treatment (Figures S6G and S6H).Though these differences were not statistically significant, perhaps due to the smaller sample size, median Tpex (c18) abundance was reduced in met-cores compared to ui-cores from either immunotherapytreated or -naive patients, and the median frequency of Tex-int (c16) was even lower in met-cores from treated patients compared to ui-cores of patients who had not received immunotherapy (Figure S6I).
metLNs exhibit immunosuppressive niches surrounding Tpex and Tex-int after ICB We next evaluated the cellular neighbors of these CD8 + T cell clusters of interest in met-cores after treatment.Similar to uiLNs, Tpex and Tex-int were most likely to be localized next to CD4 + T cells, Tregs, and DCs (Figures S6J and S6K).However, in contrast to uiLNs, Tpex were less likely to be near Tex-int in met-cores after treatment (Figure 6C), and neither of these populations upregulated PD-1 when in proximity to DCs (Figure 6).These results are consistent with a failure of Tpex to become activated and differentiate into Tex-int in the metLNs.Therefore, we further evaluated whether their neighboring cells may exhibit more immunosuppressive properties.7][48][49] Moreover, Tregs neighboring both Tpex and Tex-int exhibited higher expression of the master regulator transcription factor, FoxP3, elevated levels of CD39 and TIM3 (Figures 6F and S6M) associated with enhanced suppressive function, [50][51][52][53][54] and increased Ki-67.In addition, neighboring CD4 + T cells expressed significantly lower levels of CD45RO and PD-1 along with higher levels of TCF-1, indicative of a more naive or quiescent phenotype (Figures 6G and S6N).These results collectively indicate that Tpex and Texint are localized in more immunosuppressive niches in metLNs compared to uiLNs after ICB (Figure 6H).
Given these differences, we hypothesized that perhaps patients with metLNs would exhibit muted CD8 + T cell responses in the blood after ICB.Indeed, patients with metLNs exhibited more modest increases in the frequencies of circulating Tex-int (c5, c9) as compared to patients with no evidence of LN metastases (Figure 6I).Moreover, patients with LN metastases also exhibited smaller changes in Ki-67 + Tex-int (c5, c9) and Temra (c2) cells in the blood after treatment (Figures 6J and S6O).Collectively, these data indicate that the changes in Tpex frequencies and cellular neighborhoods in metLNs are associated with blunted responses to ICB in the blood.
DISCUSSION
It has remained unclear whether regional LNs are important hubs of anti-tumor immunity and ICB response in patients with cancer.This information is critical to understanding the mechanistic basis of current immunotherapies in humans and for identifying new opportunities to improve responses by optimally harnessing the right cell subsets.
Given prior work showing the proliferative burst of Tpex observed in the tdLNs and tumor of mice after ICB treatment, 12,14,18 we originally hypothesized that the frequency of Tpex in the uiLNs of human patients might increase after ICB treatment.Surprisingly, the frequency of Tpex in the uiLNs decreased after immunotherapy in patients.Our data support a model in which ICB causes Tpex in human uiLNs to differentiate further into Tex-int, ultimately resulting in a decrease in the frequency of Tpex in the uiLNs at the post-treatment time points analyzed in this study.ICB has been shown to drive the differentiation of Tpex into Tex-int and/or Tex-term cells in mice, [12][13][14][15][16][17] and in humans, CD8 + T cells in the TME have been observed to adopt Tpex, Tex-int, or Tex-term fates. 25,27Indeed, we observed an increase in the frequency of PD-1 + TCF-1 À CD8 + T cells with low expression of inhibitory receptors, resembling Tex-int, with increased Ki-67 expression in the uiLNs after ICB.Based on the kinetics of Ki-67 expression as a function of the cell cycle, 55 these cells may be actively proliferating or have recently differentiated from dividing cells, such as Tpex.Consistent with this notion, Tpex and Tex-int were in closer proximity to one another in uiLNs after ICB.In blood samples, we also observed a corresponding increase in the frequency of Tex-int, but not Tpex, between baseline and post-treatment surgery timepoints.Tpex themselves were rare in blood, while Tex-int were not proliferative in paired tumors.Thus, our data suggest a model in human cancer patients in which ICB drives Tpex in the uiLNs to differentiate into Tex-int, which then transit through the blood and ultimately arrive at the tumor, where they become more terminally exhausted.
Because Tpex are critical to sustain endogenous and ICBmediated CD8 + T cell responses, determining the cellular niche that regulates Tpex in human patients will be critical for harnessing them more reproducibly across cancer patients.Our data suggest that Tpex and Tex-int are localized closely with DCs in the uiLN, and this interaction increases with ICB treatment.These observations parallel previous studies demonstrating that Tpex and Tex-int in the TME reside near antigen-presenting cells in human and mouse tumors. 25,56,57In mouse studies of chronic infection, DCs promote Tpex maintenance in the spleen. 58CD28 expression, the ligands of which are typically expressed on DCs, is necessary for the proliferative burst of CD8 + T cells after ICB. 59DC expression of PD-L1 also restricts CD8 + T cell responses in mouse tumor models. 41Thus, in human cancer patients, our data support the notion that ICB results in greater interactions between Tpex and DCs, promoting proliferation and differentiation of Tex-int cells.
Our data also add to a developing understanding of the impact of metastases in LN on the anti-tumor immune response.In a mouse model of melanoma, LN metastasis drove immunosuppression, including induction of Tregs and alterations in DC phenotypes, resulting in impaired CD8 + T cell responses. 60oreover, in human melanoma patients, metLNs exhibited higher expression of immunosuppressive genes and reduced lymphocyte activation associated with distant recurrence. 60,61ur data indicate that cellular niches surrounding Tpex and Tex-int become more immunosuppressive in metLNs from recently treated patients, including elevated expression of inhibitory proteins on DCs and Tregs as well as more naive and quiescent CD4 + T cells.Patients with metLNs also experienced weaker CD8 + T cell responses in the blood following treatment, supporting an important role for LNs in generating circulating CD8 + T cell responses that associate with clinical response.
In summary, our data highlight the important role of CD8 + T cell responses in human LNs at steady-state and after ICB immunotherapy while also revealing the disruption of these key processes by LN metastasis.These results lay a foundation for the future development of immunotherapies that optimally harness anti-tumor immunity in human LNs and inform immune-monitoring strategies for cancer patients treated with ICB.
Limitations of the study While our results indicate that key CD8 + T cell subsets in human
LNs have more DC neighbors after PD-L1 blockade, we have not yet determined the mechanisms responsible.It is possible that DC migration from the tumor into the LN is enhanced or that these cells change their relative localization with one another after treatment.Because we did not observe an increase in the total frequency of DCs after anti-PD-L1 treatment (Figure S4E), we favor the latter possibility.Additionally, the specific chemokines or adhesion molecules that regulate this remain to be determined.Several studies of Tpex in mouse models of cancer and chronic infection identified their expression of the gene Xcl1, [62][63][64] which encodes a DC chemoattractant that could play a role.The clusters annotated as Tex-int in the mass cytometry analysis and MIBI analysis had minor differences in their expression of some proteins, though the key proteins used to annotate these cell subsets were consistent.These differences may be attributable to distinct antibody clones and sample preparation methods between the assays; for instance, intracellular protein stores are accessible for staining by MIBI, while cell-surface antigens are stained for mass cytometry before cell permeabilization.Moreover, the molecular mechanisms that result in the observed reduction of Tpex and suppression of Tpex and Tex-int responses after ICB in metLNs remain to be determined.
STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following: Pathologic review of formalin fixed paraffin embedded tissues All formalin fixed paraffin embedded tissues used in this study were reviewed by a board certified oral and maxillofacial pathologist (K.B.J.).
METHOD DETAILS
Tumor, lymph node, and blood processing for mass cytometry Tumor and lymph node samples were finely minced and digested in Leibovitz's L-15 medium with 800 U/ml collagenase IV and 0.1 mg/ml DNase I with gentle agitation for 45 min at 37 C.After digestion, cells were filtered through a 70mm filter into PBS/5mM EDTA solution, spun down at 500g for 5 min at 4 C, the supernatant aspirated, and resuspended in fresh PBS/EDTA solution and kept on ice.Blood samples were mixed with Ammonium-Chloride-Potassium Lysis Buffer at room temperature for 3 to 5 min to lyse red blood cells, centrifuged at 300g for 5 min at 4 C, the supernatant aspirated, and resuspended in fresh PBS/EDTA solution and kept on ice.
Cells from both tissue and blood were then washed with PBS/EDTA and re-suspended 1:1 with PBS/EDTA and 50 mM cisplatin for 60 s at room temperature before quenching 1:1 with PBS/5mM EDTA/0.5% BSA to determine viability as previously described. 84ells were centrifuged at 500g for 5 min at 4 C and re-suspended in PBS/EDTA/BSA at a density between 1 3 10 ^6 and 10 3 10 ^6 cells per ml.Suspensions were fixed for 10 min at room temperature using 1.6% paraformaldehyde (PFA) and frozen at À80 C until ready to be run for CyTOF.
Tumor and lymph node processing for single-cell sequencing Tumor and lymph node samples were thoroughly minced with surgical scissors and transferred to GentleMACs C Tubes containing 800 U/ml Collagenase IV and 0.1 mg/ml DNase I in L-15/2% FCS per 0.3 g tissue.GentleMACs C Tubes were then installed onto the GentleMACs Octo Dissociator (Miltenyi Biotec) and incubated for 20 min (lymph node) or 35 min (tumor) according to the manufacturer's instructions.Samples were then quenched with 15 mL of sort buffer (PBS/2% FCS/2mM EDTA), filtered through 100mm filters and spun down.Red blood cell lysis was performed with 175 mM ammonium chloride, if needed.
Cells were then incubated with Human TruStain FcX Receptor Blocking Solution to prevent non-specific antibody binding before staining with Zombie Aqua Fixable Viability Dye and anti-human CD45 antibody in PBS/2%FCS/2mM EDTA/0.01%sodium azide and incubated for 25 min on ice in the dark.Live CD45 + and CD45-cells were sorted on a BD FACSAria Fusion.CD45 + and CD45-cells were pelleted and resuspended at 1x10 ^3 cells/ml in 0.04%BSA/PBS buffer before mixing in an 8:2 CD45+:CD45-ratio and loaded onto the Chromium Controller (10X Genomics) to generate 5 0 v1.1 gel beads-in-emulsions (GEM).For CITE-Seq samples, pooled 8:2 CD45+:CD45-cells were resuspended in Cell Staining Buffer (BioLegend) and stained with a pool of 137 TotalSeq-C antibodies (Table S5) according to the manufacturer's protocol before loading onto the Chromium Controller (10X Genomics) for GEM generation.The cDNA libraries were generated using all or a subset of Chromium Next GEM Single Cell 5 0 Library Kit for gene expression (GEX), Chromium Single Cell V(D)J Enrichment kit (10X Genomics) for T cell receptor (TCR), and Chromium Single Cell 5 0 Feature Barcode Library kit for antibody derived tag (ADT) according to the manufacturer's instructions.The libraries were subsequently sequenced on a Novaseq S4 sequencer (Illumina) to generate fastqs with the following mean reads per cell: 42,000 (GEX), 34,000 (TCR), and 5,700 (ADT).
Tissue microarray fabrication for multiplexed ion beam imaging
For all patient lymph node samples used for MIBI analysis, hematoxylin and eosin (H&E) stained tissue sections prepared by the UCSF Department of Pathology as part of normal patient care were retrieved and reviewed by K.B.J.In general, lymph nodes used in the study were selected from neck levels most likely to represent areas of lymphatic drainage from patient tumors.2mm diameter cores in both the cortex and paracortex regions of each lymph node were manually annotated on the H&E slides.In lymph nodes with metastatic disease, 2mm cores were also obtained from the tumor-lymph node interface.The slides were then used to select 2mm cores from their corresponding paraffin blocks using a 3D-Histech Tissue Microarray Master II robot computer assisted/imageguided production system.We created two, 5 3 12 tissue microarrays following the manufacturer's instructions.Cores of a single sample of normal human tonsil were also included in duplicate as controls on each microarray.
Antibody heavy metal conjugation for mass cytometry
The sources for all mass cytometry antibodies can be found in Table S5.Antibodies were conjugated to their associated metals with MaxPar X8 labeling reagent kits according to manufacturer instructions, diluted with Candor PBS Antibody Stabilization solution supplemented with 0.02% sodium azide, and filtered through an UltrafreeMC 0.1-mm centrifugation filter (Millipore) before storage at 4 C. Surface and intracellular master antibody cocktails were made and kept at À80 C.
Antibody heavy metal conjugation for multiplexed ion beam imaging
Antibodies were conjugated to metal-loaded MIBItags (Ionpath) according to manufacturer instructions, and stored at 4 C prior to use.
Mass-tag cellular barcoding for mass cytometry
Prior to antibody staining, mass tag cellular barcoding of prepared samples was performed by incubating cells with distinct combinations of isotopically-purified palladium ions chelated by isothiocyanobenzyl-EDTA as previously described. 85After counting, 1 3 10 ^6 cells were barcoded with distinct combinations of stable Pd isotopes for 15 min at room temperature on a shaker in Maxpar Barcode Perm Buffer.Cells were washed twice with cell staining media (PBS with 0.5% BSA and 0.02% NaN3), and pooled into a single 15 mL tube for subsequent staining and washing steps.
Mass cytometry staining
Barcoded cells were stained with Human TruStain FcX Receptor Blocking Solution at 20 mg/mL for 5 min at RT on a shaker.Surface antibody cocktail was then added with a 500ul final reaction volume for 30 min at RT on a shaker.Following staining, cells were washed twice with cell staining media.Before intracellular staining, cells were permeabilized for 10 min with methanol at 4 C. Methanol was then removed by washing the cells 2 times with cell staining media.The intracellular cocktail was then added to the cells for a 500uL final reaction volume for 1 h at RT on a shaker.Cells were washed twice in cell staining media to remove unbound antibodies and then stained with 1mL of 1:4,000,191/193Ir Cell-ID Intercalator Solution diluted in PBS with 4% PFA overnight.Before mass cytometry was run, cells were washed once with cell staining media, and twice with Cell Acquisition Solution.
Multiplexed ion beam imaging tissue section staining Tissue sections (5 mm thick) were cut from FFPE blocks of the tissue microarrays and mounted on gold-and tantalum-sputtered microscope slides (IonPath).Slides were baked at 70 C for 1 h followed by deparaffinization and rehydration with sequential washes in xylene (2x), 100% ethanol (2x), 95% ethanol (2x), 80% ethanol (1x), 70% ethanol (1x), and ddH2O (2x) with a Leica ST4020 Linear Stainer (Leica Biosystems) programmed for 30 s each.Tissues next underwent antigen retrieval by submerging slides in Target Retrieval Solution (pH 9) and incubating them at 97 C for 40 min and cooled down to 65 C in a Lab Vision PT Module (Thermo Fisher Scientific).Slides were further cooled to room temperature and washed in 1x phosphate-buffered saline (PBS) IHC Washer Buffer with Tween 20.Next, endogenous biotin and avidin proteins were successively blocked using an Avidin/Biotin Blocking System, for 10 min each.Tissues were then washed with wash buffer and blocked for 1 h at room temperature with 1x PBS IHC Wash Buffer with Tween 20 and 5% (v/v) normal donkey serum (Sigma-Aldrich).Two antibody panels were prepared.The first antibody cocktail was prepared in 1x PBS IHC Wash Buffer with Tween 20 with 5% (v/v) normal donkey serum and 0.5 mM EDTA and filtered through a 0.1-mm centrifugal filter (Millipore) prior to incubation with tissue overnight at 4 C in a humidity chamber.Following the overnight incubation, slides were washed twice for 5 min in wash buffer.The secondary antibody cocktail was prepared as described above and incubated with the tissues for 1 h at room temperature in a humidity chamber.Slides were dried under vacuum prior to imaging.
Mass cytometry data acquisition
Mass cytometry samples were diluted in Cell Acquisition Solution containing bead standards to approximately 1 3 10 ^6 cells/mL and then analyzed on a Helios mass cytometer (Fluidigm) equilibrated with Cell Acquisition Solution.A minimum of 10 3 10 ^6 cell events were collected for each barcoded set of samples at an event rate of 400-500 events/second.
Multiplexed ion beam imaging data acquisition
Imaging was performed using a MIBI-TOF instrument (IonPath) with a Hyperion ion source.Xe+ primary ions were used to sequentially sputter pixels for a given field of view.The following imaging parameters were used: acquisition setting: 80 kHz; field size: 800 3 800 mm, 2048 3 2048 pixels; dwell time: 1 ms; median gun current on tissue: 5.5 nA Xe+.
Data normalization and de-barcoding for mass cytometry
Bead standard data normalization and de-barcoding of the pooled samples into their respective conditions was performed using the R package from the PICI institute available at https://github.com/ParkerICI/premessa.
Mass cytometry batch normalization
Each group of barcoded samples was run with a control sample (replicates of a single normal human tonsil) to validate staining and for normalization between groups of barcoded samples.Bead standard data normalization and de-barcoding of the pooled samples into their respective conditions was performed using the R package Premessa from the PICI institute available at https://github.com/ParkerICI/premessa.All manually gated live, intact, single cells were downloaded as FCS files from CellEngine (CellCarta, Montreal, Canada).CytoNorm 65 (https://github.com/saeyslab/CytoNorm)was utilized to correct for batch effects.All markers were used for batch effect normalization.
Mass manual gating
Batch effect normalized FCS files were uploaded to CellEngine for manual gating (Figure S1A).
Mass cytometry clustering
Manually gated CD8 + T cells were downloaded as FCS files from CellEngine, and flowCore 66 was used to import FCS files into R.The FlowSOM clustering algorithm, 33 available through the CATALYST R/Bioconductor package, 67,68 was used to generate clusters based on CD8 + T cell specific markers (Granzyme B, CD38, CD127, CD45RA, TIM3, TIGIT, PD-L1, CD27, CD39, Tbet, CD103, FoxP3, CD69, CCR7, CD25, TCF-1, Pan-HLA-DR, PD-1, CD56, CD16, CD7, CD95).We ran FlowSOM with a 10 3 10 grid and utilized the built-in ConsensusClusterPlus 86 metaclustering step to obtain 20 CD8 + T cell clusters.Clusters were visualized using UMAPs via the umap package 87 and heatmaps via the ComplexHeatmap package, 88 both available through CATALYST.For paired differential abundance analyses, we used generalized linear mixed models implemented in the diffcyt R package. 69w level image processing for multiplexed ion beam imaging To remove background generated from gold, oxides and adducts, as well as compensate for channel cross talk.We used the Rosetta algorithm, which uses a flow-cytometry style compensation approach to remove spurious signals.All compensation parameters were first evaluated in a subset of 10 images to ensure contaminant signal was removed while preserving target signal.Next, finalized parameters were applied to the full image set.
Region masks for multiplexed ion beam imaging Region masks were generated to define histologic regions of each FOV including the T-cell-zone, B-cell-zone, tumor, endothelium, or other.The tumor mask was first generated by applying smoothing (Gaussian blur, radius 2 px) and a pixel thresholder to Keratin signal.To generate the B-cell-zone mask, tumor mask was subtracted from the FOV.In the remaining area, smoothing and pixel thresholder were applied to CD20 signal.T-cell zone and Endothelium masks were similarly generated using CD3 and CD31 signal, respectively and any remaining area of the FOV was classified as ''Other''.For each region, all holes were filled.
Single-cell segmentation for multiplexed ion beam imaging
To delineate the location of single cells in MIBI-TOF images, we performed cell segmentation using the pre-trained Mesmer convolutional neural network architecture. 36We used dsDNA as the nuclear marker and HLA Class 1, ABC as the membrane marker as input to the network.The output of Mesmer is the location of each cell in the image.Cells with area larger than 200 mm 2 were frequently due to out-of-focus regions of an image and therefore excluded from analysis.Additionally, FOVs with fewer than 3500 total cells were excluded from analysis.
Single-cell phenotyping for multiplexed ion beam imaging Single cell phenotyping was accomplished using a previously described method. 89Pre-processed MIBI-TOF images were first Gaussian blurred using a standard deviation of 2 for the Gaussian kernel.To account for both technical and biological confounders, pixels were normalized by their total expression, such that the total expression of each pixel was equal to 1.A 99.9% normalization was applied for each marker.Pixels were clustered into 100 clusters using FlowSOM 33 based on the expression of 24 phenotypic markers: Foxp3, CD11b, CD11c, CD138, CD14, CD16, CD163, CD20, CD21, CD3, CD31, CD4, CD45, CD56, CD68, CD8, E-Cadherin, HLADR, Keratin, MPO, PD-1, T-bet, Vimentin, BDCA3.The average expression of each of the 100 pixel clusters was found and the z-score for each marker across the 100 pixel clusters was computed.Using these z-scored expression values, the 100 pixel clusters were meta-clustered using consensus hierarchical clustering.These meta-clusters were manually inspected and adjusted by comparing with the MIBI-TOF images.Next, by applying the segmentation masks that delineate the boundaries of all cells in the images, we counted the number of each of the pixel clusters in each cell.This resulted in a pixel cluster by cell count table.These counts were then normalized by cell size.Using these frequency measurements as the feature vector, the cells were clustered using FlowSOM into 100 cell clusters.Similarly to the pixel clusters, the average expression of each of the 100 cell clusters was found and the z-score was computed.Using these z-scored values, the 100 cell clusters were meta-clustered using consensus hierarchical clustering.Each of the cell meta-clusters was then manually inspected and adjusted by comparison with the images, then annotated with its cell phenotype.Clustering in this manner resulted in better cluster definition than clustering using integrated expression. 89Cell populations were refined by inspecting biaxial plots of integrated marker expression for each single cell.
Neighborhood analysis for multiplexed ion beam imaging
Cell neighborhoods were produced as previously described. 40A cell neighbor matrix was generated, in which each row represents an index cell and columns indicate the number of each cell phenotype within a 50-pixel radius of the index cell.A similar approach was used to generate a cell neighbor matrix of cells positive for functional markers.Functional marker thresholds were determined by manually visualizing images.Each cell in the dataset was determined to be positive or negative using these manual thresholds, then the number of cells positive for each functional marker within a 50-pixel radius of each cell was determined.
Alignment of single-cell sequencing libraries
The raw fastq files were aligned using Cell Ranger v.3.0.2 and v.6.0.2 software with the default settings to the hg38 transcriptome and the TotalSeq-C Feature Reference provided by BioLegend for the GEX and ADT fastqs, respectively.The raw TCR fastq files for each participant from the lymph node and tumor were aligned jointly in order to call shared clonotypes across the tissues using Cell Ranger v.7.0.0 to the vdj GRCh38 v.7.0.0 reference.
Demultiplexing pooled single-cell samples
The GEX BAM alignment files from each of the pools were used as the input for dsc-pileup v.0.1beta to generate pileup files which were then used as input for freemuxlet v.0.1beta for doublet identification, inferred genotypes for individuals in each pool, and singlet assignment to those inferred genotypes.The percentage of shared single nucleotide polymorphism (SNP) assignments from the freemuxlet genotypes was used to match samples from the two individuals in the lymph node pool and the tumor pool to each other.The two samples in each pool were from a male and a female participant so the sex assignment (based on XIST expression and the percentage of total counts based on unique molecular identifiers (UMIs) mapping to Y chromosome genes; Figure S2J) was used to match the freemuxlet IDs to the clinical trial participant IDs.
Single-cell sequencing full dataset clustering
We filtered out 747 cells with less than 100 or more than 3,000 genes detected and filtered out 4,883 genes detected in less than 3 cells.We also filtered out 1,943 cells with more than 10% of total counts (UMIs) mapping to mitochondrial genes and 70 cells identified as red blood cells (based on HBB expression) or platelets (based on PF4 expression).For the multiplexed samples, we also filtered out genetic doublets (629 cells) identified by freemuxlet.The raw counts were normalized to 10,000 counts per cell and log (count + 1) transformed.We identified 1,098 highly variable genes which were scaled and used with the default settings in scanpy v.1.7.1 70 for PCA analysis.We used Harmony v.0.0.5 71 for batch correction in the PC space with each sample as a batch.The top 20 corrected PCs from Harmony were used for nearest neighbor detection followed by leiden 90 clustering and UMAP 91 projection.This analysis identified 22 clusters which we collapsed into 13 cell types based on marker gene expression.
Single-cell sequencing CD8 T cell clustering
For the ADT data, we filtered out isotype control antibodies (7 total in the 137 panel) and normalized the raw counts with CLR (count + 1) by cell.To identify CD8 T cells by protein, we used a Gaussian mixture model tool from sklearn v.0.24.1 to create positive and negative gates for CD3 protein, CD8 protein, and CD4 protein on the normalized expression for each marker.We assigned CD8 T cells by including 1) cells that expressed CD3 (CD3E or CD3D) and CD8 (CD8A or CD8B) by RNA and did not express CD4 by RNA or protein, 2) cells gated as CD8 T cells by protein [CD3 + CD8 + CD4-], and 3) cells in the clusters annotated ''CD8 T cell'' that did not express CD4 by RNA or protein.After removing 11 tumor cells (based on KRT14 expression) from the resulting cells, we subsetted the raw count matrix from the CD8 T cells (8,245 cells) and removed 9,251 genes that were not expressed in at least three cells.The counts from the remaining 13,305 genes were normalized to 10,000 counts per cell, log (count + 1) transformed, and scaled.We used the default settings in scanpy v.1.7.1 for PCA dimensionality reduction.We iteratively used Harmony v.0.0.5 for batch correction in the PC space with sex, cell cycle phase, and sample as the batch IDs.The top 20 corrected PCs were used for nearest neighbor detection followed by leiden clustering and UMAP projection.This identified 13 clusters which we collapsed into 9 CD8 cell sub-types based on marker gene expression.
Statistical analysis
For CyTOF and MIBI data, all statistical tests were performed in R. 93,94 The non-parametric Wilcoxon rank sum test was utilized to compare samples from anti-PD-L1 treated patients to standard of care treated patients and metLN to uiLN for MIBI data.The paired Wilcoxon rank sum test and generalized linear mixed models were used for paired CyTOF data.We used Spearman's correlation to measure the correlation between cluster abundances in paired data, which were plotted using the corrplot package. 73For multiple testing corrections, we applied Benjamini-Hochberg correction and statistical differences were declared significant at FDR < 0.1.For hypothesis testing, statistical differences were declared significant at p < 0.05.
The R packages dplyr 80 and reshape2 81 were used for data manipulation.Most plots were produced with ggplot2 82 and RColorBrewer. 83or single cell sequencing data, all statistical tests were performed in python v.3.8.8. 95We compared Tpex scores and Tex-term scores in shared clones in the tumor versus lymph node using a linear mixed effect model from statsmodels v.0.12.2 79 with tissue origin as a fixed effect and clonotype ID as a random effect.We used a Fisher's exact test from scipy v.1.6.1 74 to test for an association between tissue origin (lymph node or tumor) and cell sub-type assignment (cluster or Tex-term) for the cells in those cell sub-types that were shared clones.We used pandas v.1.2.3 75,76 and numpy v.3.3.4 77 were used for data manipulation in python.We used scanpy v.1.7.1 and matplotlib v.3.3.4 (Matplotlib: A 2D Graphics Environment) for creating plots in python.
Statistical parameters (e.g., sample size [n], etc.) are detailed in the main text and figure legends for each dataset.
ADDITIONAL RESOURCES
All samples used from subjects treated with atezolizumab were from clinical trial NCT03708224.Additional information about this clinical trial can be viewed at https://clinicaltrials.gov/ct2/show/NCT03708224. (legend continued on next page) (legend continued on next page)
Figure 1 .Figure 2 .
Figure 1.Tpex cells are increased in uninvolved lymph nodes of HNSCC patients (A) Overview of cohort.Paired tumor and lymph node samples (uninvolved [uiLN] and/or metastatic [metLN]) were obtained from patients with HNSCC (n = 10).Samples were analyzed by mass cytometry.(B) Paired differential abundance (DA) of main immune cell populations between uiLN and tumor (n = 9; paired Wilcoxon rank-sum test).The log2 fold changes are plotted against the negative log10 (nominal p values).Colors indicate if cell populations are significantly more abundant in uiLN (purple), tumor (blue), or not differentially abundant (False, gray) after Benjamini-Hochberg correction, FDR < 0.1.(C) Heatmap of markers used for CD8 + T cell clustering.Scaled median expression per marker is shown for cluster annotation.(D) UMAP of non-naive CD8 + T cell clusters and expression for a subset of markers.(E) Paired differential abundance analysis of non-naive CD8 + T cell subsets between uiLN and tumor (n = 9; generalized linear mixed models).See color scheme for (B).(F) Cluster 8 and 14 abundances (as percentage of non-naive CD8 + T cells) in paired samples from uiLN and tumor.p values obtained by generalized linear mixed models.(G) Spearman correlation between clusters in uiLN and clusters in tumors.Tpex clusters are highlighted in red boxes.See also Figure S1.
Figure 3 .
Figure 3. Localization of Tpex in human uiLN (B) Overview of the image analysis pipeline.(C) Heatmap of scaled median marker expression for cell lineage assignments.(D) Relative abundance of main immune cell types in uiLN (global, n = 23) from SOC-treated patients (n = 22).Sample ID represents patient ID followed by LN number.(E) Heatmap of markers used for CD8 + T cell clustering.Scaled median expression per marker is shown for cluster annotation.(F) Relative abundance of non-naive CD8 + T cell clusters in uiLN (T cell zone, n = 23) from SOC-treated patients (n = 22).
Figure 4 .
Figure 4. Anti-PD-L1 ICB impacts Tpex and Tex-int in uiLN (A) Relative abundance of non-naive CD8 + T cell clusters in uiLNs (T cell zone, n = 20) from anti-PD-L1-treated patients (n = 9).Sample ID represents patient ID followed by LN number.(B) Non-naive CD8 + T cell cluster ratios represented as log2 fold changes between uiLNs (T cell zone, n = 20) from anti-PD-L1-treated patients and uiLNs (T cell zone, n = 22) from SOC-treated patients.(C) Cluster 18 and 16 abundances (as percentage of non-naive CD8 + T cells) in the T cell zone of uiLNs from SOC-and anti-PD-L1-treated patients.
(
D) Cluster 18/cluster 16 ratio in uiLNs (T cell zone) from SOC-and anti-PD-L1-treated patients.(E) Percentage of cluster 18 and 16 proliferating cells in uiLNs (T cell zone) from SOC-and anti-PD-L1-treated patients.(F and G) Percentage of (F) cluster 18 and (G) cluster 16 cells with a specific cell type as its neighbor in uiLNs (T cell zone) from anti-PD-L1-treated patients.(H) Number of DC neighbors for cluster 18 and cluster 16 cells in uiLNs (T cell zone) from SOC-and anti-PD-L1-treated patients.(I) Number of cluster 16 neighbors for cluster 18 cells in uiLNs (T cell zone) from SOC-and anti-PD-L1-treated patients.(J and K) Number of (J) PD-1-positive and (K) Ki-67 + neighbors for cluster 18 and cluster 16 cells in uiLNs (T cell zone) from SOC-and anti-PD-L1-treated patients.(L) Expression of PD-1 on cluster 18 and cluster 16 cells with a DC neighbor in uiLNs (T cell zone) from SOC-and anti-PD-L1-treated patients.(M) Representative image of an uiLN from an anti-PD-L1-treated patient (patient 01) showing the spatial localization of CD4 T cells, Tregs, DCs, Tpex (cluster 18), and Tex-int (cluster 16).Cell identity overlaid onto the segmentation mask.Highlighted region is colored by the expression of CD8 + (red), PD-1 (green), TCF-1 (cyan), CD11c (purple), Ki-67 (yellow), CD4 (cyan), and FoxP3 (purple).(C-E and H-L) p values obtained by Wilcoxon rank-sum test.See also Figure S4.
Figure 5 .
Figure 5. Tex-int are present at higher levels in the blood following ICB (A) Overview of cohort.Blood samples were collected pre-treatment (baseline, 2-29 days before surgery), at time of surgery, and at follow-up (27-38 days postsurgery) from patients with HNSCC that received anti-PD-L1 treatment prior to surgery (n = 10).Additionally, tumor samples were collected at time of surgery.For a few patients, additional samples were collected (see complete overview in FigureS6A).Samples were analyzed by mass cytometry.(B) Heatmap of markers used for CD8 + T cell clustering.Scaled median expression per marker is shown for cluster annotation.(C) UMAP of non-naive CD8 + T cell clusters and expression for a subset of markers.(D) Frequencies of non-naive CD8 + T cell clusters.(E) Paired differential abundance analysis of non-naive CD8 + T cell subsets between blood at time of surgery and baseline (n = 10) (generalized linear mixed models).The log2 fold changes are plotted against the negative log10(nominal p values).Colors indicate if cell populations are significantly more abundant in blood at time of surgery (purple) or baseline (blue) or not differentially abundant (False, gray) after Benjamini-Hochberg correction, FDR < 0.1.(F) Cluster 5 and 9 abundances (as percentage of non-naive CD8 + T cells) in paired blood samples at time of surgery and baseline.p values obtained by generalized linear mixed models.(G)Paired differential abundance analysis of non-naive CD8 + T cell subsets between blood at time of follow-up and at baseline (n = 7) (generalized linear mixed models).See color scheme for Figure6C.(H) Cluster 2, 5, and 9 abundances (as percentage of non-naive CD8 + T cells) in paired blood samples at follow-up and baseline.p values obtained by generalized linear mixed models.(I and J) Percentage of cluster 2 and 5 proliferating cells in paired blood samples at time of surgery and baseline(I) and in blood and tumor at time of surgery(J).p values obtained by paired Wilcoxon rank-sum test.(K) Spearman correlation between cluster 16 and 18 in LN (MIBI data) and cluster 5 and 9 in blood (CyTOF data) at time of surgery.See also FigureS5.
Figure S2 . 2 (
Figure S2.Shared CD8 T clones have higher Tpex signature in the lymph node and higher exhaustion signature in the tumor, related to Figure 2 (A) UMAP of 35,551 cells from 5 paired tumor and draining lymph node samples (10 samples total).Cells are colored according to annotated cell type.(B) Dot plot of standard scale (subtract minimum and divide by maximum) column normalized expression for the indicated genes for each annotated cell type from all cells indicating percentage of cells with expression greater than zero (dot size) and mean expression for cells with nonzero expression (color).
Figure 4 (
Figure S4.anti-PD-L1 ICB impacts Tpex and Tex-int in uiLNs, related to Figure 4(A) Relative abundance of main immune cell types in uiLN (global, T cell zone, and B cell zone) from anti-PD-L1 treated patients.Sample ID represents patient ID followed by LN number.(B) Relative abundance of non-naive CD8 T cell clusters in uiLN (global and B cell zone) from anti-PD-L1 treated patients.Sample ID represents patient ID followed by LN number.(C) Percent Tpex (c18) of non-naive CD8 T cells in the B cell zone from untreated versus anti-PD-L1 treated patients.p-values obtained by Wilcoxon Rank-Sum Test.(D) Difference in the number of DC neighbors between untreated and anti-PD-L1 treated patients for each cluster of CD8 T cells (Wilcoxon Rank-Sum Test).Colors indicate if DC neighbors are significantly more abundant in anti-PD-L1 (purple) or SOC (blue) or not differentially abundant (False, gray) after Benjamini-Hochberg correction, FDR <0.05.(E) Percent DC of total immune cells from untreated versus anti-PD-L1 treated patients.(Fand G) Spearman correlation between % DC of total immune cells and (F) % Tpex (c18) or (G) % Tex-int (c16) of total non-naive CD8 T cells in samples from patients receiving SOC or PD-L1 blockade.
Figure S5 . 5 (
Figure S5.Transitional exhausted CD8 + T cells are present at higher levels in the blood following treatment with anti-PDL1 ICB, related to Figure5(A) Samples collected from each patient.Time (days) are relative to the first sample collected.(B) Relative abundance of main immune cell types.(C) Principal component analysis of gated main immune cell populations.
(
D) UMAP of all CD8 T cell clusters.(E) Frequencies of non-naı ¨ve CD8 T cell clusters.(F) Percentage of cluster 9 proliferating cells in paired samples from blood at time of surgery and baseline.p-value obtained by paired Wilcoxon Rank-Sum Test.(G) Paired differential abundance analysis of non-naı ¨ve CD8 T cell subsets between tumor and blood at time of surgery (n = 6) (generalized linear mixed models).See color scheme for Figure 6C.(H) Cluster 5 abundance (as percentage of non-naı ¨ve CD8 T cells) in paired samples from blood and tumor at time of surgery.P-values obtained by generalized linear mixed models.(I) Correlation between clusters in tumor and blood at time of surgery (Spearman correlation).(J and K) Correlation between clusters in LN (MIBI data) and clusters in blood (CyTOF data) (J) and clusters in LN (MIBI data) and clusters in tumor (CyTOF data) (K) at time of surgery (Spearman correlation).
TABLE
d RESOURCE AVAILABILITY B Lead contact B Materials availability B Data and code availability d EXPERIMENTAL MODEL AND SUBJECT DETAILS B Subject consent and biospecimen collection B Pathologic review of formalin fixed B Multiplexed ion beam imaging data acquisition B Data normalization and de-barcoding for mass cytometry d QUANTIFICATION AND STATISTICAL ANALYSIS B Mass cytometry batch normalization B Mass cytometry manual gating B Mass cytometry clustering | 2023-03-17T13:31:53.270Z | 2023-03-01T00:00:00.000 | {
"year": 2023,
"sha1": "8f0a2db44aa6f54c36288cd2a366de1033d22148",
"oa_license": "CCBY",
"oa_url": "http://www.cell.com/article/S0092867423001642/pdf",
"oa_status": "HYBRID",
"pdf_src": "ScienceParsePlus",
"pdf_hash": "fb856069c82448bd96a61c3ebab66900acf12f19",
"s2fieldsofstudy": [
"Biology",
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
49414578 | pes2o/s2orc | v3-fos-license | Complexity of Daily Physical Activity Is More Sensitive Than Conventional Metrics to Assess Functional Change in Younger Older Adults
The emerging mHealth applications, incorporating wearable sensors, enables continuous monitoring of physical activity (PA). This study aimed at analyzing the relevance of a multivariate complexity metric in assessment of functional change in younger older adults. Thirty individuals (60–70 years old) participated in a 4-week home-based exercise intervention. The Community Balance and Mobility Scale (CBMS) was used for clinical assessment of the participants’ functional balance and mobility performance pre- and post- intervention. Accelerometers worn on the low back were used to register PA of one week before and in the third week of the intervention. Changes in conventional univariate PA metrics (percentage of walking and sedentary time, step counts, mean cadence) and complexity were compared to the change as measured by the CBMS. Statistical analyses (21 participants) showed significant rank correlation between the change as measured by complexity and CBMS (ρ = 0.47, p = 0.03). Smoothing the activity output improved the correlation (ρ = 0.58, p = 0.01). In contrast, change in univariate PA metrics did not show correlations. These findings demonstrate the high potential of the complexity metric being useful and more sensitive than conventional PA metrics for assessing functional changes in younger older adults.
Introduction
The aging process is often accompanied by functional decline and increased risk of chronic diseases [1]. However, functional ability varies between older adults, depending on individual health condition and lifestyles. Physical inactivity is one of the known risk factors that can lead to morbidity and mortality [2]. The transition from work to retirement, which often occurs between 60 and 70 years of age, can involve a significant change in structured daily activities, with physical activity declining. A study on the Dutch population found that retirement introduces a reduction in physical activity from work-related transportation that is not compensated for by an increase in sports participation or an increase in non-sports leisure-time physical activity [3]. Thus, this population is of particular importance for addressing maintenance of their functional status.
In recent years, mobile health (mHealth) applications, incorporating wearable sensing technologies with modern mobile communication devices, are emerging. From the early adoption in younger populations for fitness tracking in particular, mHealth has been continuously developing and diversifying its applications for different populations [1,4]. The scalability of mHealth technologies enables data collection in diverse geographic locations over prolonged time periods [5]. This provides a new perspective in studying physical activities in real life and allows new analytical tools to be developed to analyze and present the data.
An earlier systematic review on body-worn, accelerometer-based physical activity monitoring revealed the challenge of achieving consensus on the reporting and the interpretation of the measurements provided by various mHealth applications [6]. Furthermore, the review pointed out that energy expenditure, walking time, and total activity are most frequently reported and are comparable variables across studies. In addition, several variables, such as walking time, number of steps, and cadence are the most widely adopted variables in research [5,7,8] to characterize walking pattern, the most common daily physical activity across all age groups. Descriptive statistics (e.g., mean, maximum values) are applied to the above walking parameters for analysis. Detrended fluctuation analysis proposed by Hausdorff et al. is used to quantify the stride-to-stride variability in supervised walking tasks [9]. However, physical activity involves multiple components and has more than one dimension. Different types of activities in daily life, as well as the quantity and the quality (performance) of each activity jointly determine a person's functional status. Thus, a variable that models a person's physical activity behaviour based on these aspects is warranted.
Complexity analysis as introduced by Paraschiv-Ionescu et al. [10] aims at combining both the quantity and quality dimensions of multiple, commonly performed daily activities into one metric to describe physical activity behaviour. This complexity metric has demonstrated discriminative power to distinguish groups of patients suffering from chronic pain [11,12]. Besides analyzing accelerometry-derived activities, the metric has been validated also for analyzing activity behaviour of older adults using an application based on wearable pressure insoles [13].
While several cross-sectional studies exist for clinical validation of complexity metrics, the sensitivity of those metrics for detecting changes in physical function over time has not yet been determined. Therefore, we aimed to examine the ability of the aforementioned complexity metric to detect change in a longitudinal intervention study conducted in younger older adults in comparison to conventionally applied univariate physical activity metrics. Within this context, we further explored the impact of smoothing sensor-based physical activity data on the complexity metric.
Study Protocol
The study is part of the larger PreventIT project [1], developing and testing an ICT-based mHealth system that enables early identification of risk for age-related functional decline, and engenders behavioural change in younger older adults (aged 60-70 years) in order to adopt a healthy, active lifestyle. Thirty participants, aged between 60 and 70 years, were recruited at three different sites, Trondheim (Norway), Amsterdam (the Netherlands), and Stuttgart (Germany), to participate in a 4-week pre-post pilot intervention study. All participants were instructed by an experienced physical therapist or exercise therapist to follow an adapted Lifestyle-integrated Functional Exercise (aLiFE) programme specifically developed for improving balance and strength and increasing physical activity in younger older adults [14]. aLiFE was taught during four weekly home visits, and the participants were asked to integrate the aLiFE activities into everyday routines. Pre and post intervention, the participants completed a balance and mobility assessment using the Community Balance and Mobility Scale (CBMS). The scale has been validated to capture high-level balance, gait, and mobility performances in healthy active younger older adults based on the quality of performing the tasks [15]. CBMS assessments were performed in the research hospital or university by trained assessors. In addition, daily physical activity (PA) of each participant was measured twice for one week with wearable sensors. The participants were instructed to wear an inertial sensor (DynaPort, MoveMonitor, McRoberts, The Hague, The Netherlands) at their lower back at the level of L5 using an elastic belt during the day and night. The sensor needed to be removed when showering or during any water activity and needed to be put back on afterwards. The sensor did not need to be recharged during the one week measurement. All sensors were collected at the end of the measurement and raw sensor data was downloaded for offline data analysis. The first measurement was prior to the start of the intervention period. The measurement was repeated during the third week of the pilot study to capture the change of daily activity patterns during the intervention.
Sensor Data Processing
The sensor consists of a 3D accelerometer with sampling frequency of 100 Hz. The recording start time of each sensor was registered on the device by manual insertion of a timestamp. A non-commercial activity classification software was used to extract quantitative as well as qualitative features of PA from raw sensor data. The software is an outcome of the FARSEEING EU project (FP7/2007-2013, grant agreement 288940). It has been applied in studies with dementia patients [16] and older people residing in independent-living retirement homes [17]. The software has been further developed based on two datasets of elderly subjects. The first one is the ADAPT dataset [18], where video recording was performed using ceiling-mounted cameras in lab settings and an action camera in free-living conditions. The second dataset is from the University of Auckland [17], where subjects performed both scripted and unscripted activities of daily living collected in a free-living environment. First, the algorithm estimates Metabolic Equivalents (METs); signals are filtered and processed as described in [19]. An interval is labelled as 'sedentary', if associated energy expenditure is below or equal to 1.5 MET [20]. Otherwise, the interval is labelled as 'active'. 'Sedentary' intervals with a mean angle between the vertical axis and the medio-lateral or the anterior-posterior direction of the trunk below 30 • are labelled as 'lying'. The 'active' intervals, where steps are detected are labelled as 'walking'.
Step detection is based on [21]. Each interval is then characterized by the category (label), the duration, and the activity counts (counts/minute) from which METs are estimated [19]. In addition, number of steps and the cadence (steps/minute) are extracted for each 'walking' bout. Data is labelled as 'non-wearing', if the sensor is detected lying flat with very low variance in acceleration signals for longer than half an hour.
The classified activities were sorted into natural days based on the registered timestamp. Days with less than 16 h of measurements (i.e., the first and the last day of the measurement) were excluded from further processing and analysis. Given the high resolution (1 s) of the PA output data, bouts of one activity may be interspersed with short episodes of other activities. For example, short breaks of a few seconds are often present during one walking bout due to environmental factors (e.g., walking episodes whiles shopping in a supermarket), which may lead to a string of several walking episodes rather than one continuous walking bout. Such short breaks during an activity introduce artificial changes in the dynamics of the PA time series, which are not relevant to one's Sensors 2018, 18, 2032 4 of 12 physical behaviour. Therefore, a smoothing technique was devised to filter such artefacts and to aggregate bouts in the original PA output belonging to the same PA category based on the following steps, as illustrated in Figure 1. First, we applied a moving forward sliding window of 30 s without overlap to smooth the PA time series [22]. The activity category of these 30 s was replaced by the activity with the highest density (counts) within the window. Second, the process was repeated for K folds. At each fold, the smoothing starts with a random shift between 1 and 30 s at the beginning of the time series. Third, the activity category of each second in the aggregated time series was determined by the majority vote of the K-fold smoothing. The sliding window length was chosen according to the 'barcode' design (explained in the later section Complexity analysis), where 30-s is the threshold of activity duration corresponding to indoor walking.
Sensors 2018, 18, x FOR PEER REVIEW 4 of 12 density (counts) within the window. Second, the process was repeated for K folds. At each fold, the smoothing starts with a random shift between 1 and 30 s at the beginning of the time series. Third, the activity category of each second in the aggregated time series was determined by the majority vote of the K-fold smoothing. The sliding window length was chosen according to the 'barcode' design (explained in the later section Complexity analysis), where 30-s is the threshold of activity duration corresponding to indoor walking.
Univariate Analysis
For each activity bout, its duration and activity counts per minute ('ActiCount') were estimated. In addition, the total number of steps and the average cadence (steps/min) were provided for each classified walking bout. PA was characterized by various univariate metrics including the percentage of time being sedentary, the percentage of time spent walking, the number of steps normalized to the measurement duration (in hours), and the mean cadence of walking bouts of each day. The univariate metrics were computed based on the original and the smoothed PA time series.
Complexity Analysis
Complexity was introduced in order to analyze the variability of biological and physiological time series data [23,24]. The technique was subsequently adapted and applied to analyze ambulatory activity patterns [25]. Paraschiv-Ionescu et al. proposed complexity analysis on a multivariate PA pattern ('barcode') derived from wearable sensor data [10]. The 'barcode' is constructed for the analyzed period based on the classified activity category, the duration and the intensity. The entropy rate of the resulting multi-state 'barcode' represents complexity. In the analysis of the pilot data, an adapted 'barcode' was used, where the 'ActiCount' was modeled according to a validation study presented in [19]. Entropy rate of the 'barcode' was computed in terms of Lempel-Ziv complexity based on the method described in [26]. (Additional materials are presented in Appendix A.1). Complexity of the 'barcode' generated from both the original and the smoothed time series of PA were computed to analyze the influence of activity classification on the calculated complexity.
Univariate Analysis
For each activity bout, its duration and activity counts per minute ('ActiCount') were estimated. In addition, the total number of steps and the average cadence (steps/min) were provided for each classified walking bout. PA was characterized by various univariate metrics including the percentage of time being sedentary, the percentage of time spent walking, the number of steps normalized to the measurement duration (in hours), and the mean cadence of walking bouts of each day. The univariate metrics were computed based on the original and the smoothed PA time series.
Complexity Analysis
Complexity was introduced in order to analyze the variability of biological and physiological time series data [23,24]. The technique was subsequently adapted and applied to analyze ambulatory activity patterns [25]. Paraschiv-Ionescu et al. proposed complexity analysis on a multivariate PA pattern ('barcode') derived from wearable sensor data [10]. The 'barcode' is constructed for the analyzed period based on the classified activity category, the duration and the intensity. The entropy rate of the resulting multi-state 'barcode' represents complexity. In the analysis of the pilot data, an adapted 'barcode' was used, where the 'ActiCount' was modeled according to a validation study presented in [19]. Entropy rate of the 'barcode' was computed in terms of Lempel-Ziv complexity based on the method described in [26]. (Additional materials are presented in Appendix A.1). Complexity of the 'barcode' generated from both the original and the smoothed time series of PA were computed to analyze the influence of activity classification on the calculated complexity.
Statistical Analysis
For each participant, the univariate metrics and the complexity metric of PA were analyzed for each day. The average value of each metric over the one-week measurement before (Week0) and during (Week3) the pilot study was calculated. According to study [27], participants having more than two days' sensor data during the one-week measurement, in both Week0 and Week3, were included for statistical analysis. The changes in PA metrics between Week0 and Week3 were computed. The change in CBMS score pre and post the pilot study was calculated. The primary analysis was to examine the correlations between the change in PA metrics and the change in CBMS score. Given the small sample size and ordinal data type (CBMS scores), spearman coefficient (ρ) was used to analyse the strength of correlation. In addition, a non-parametric effect size calculator, Cliff's Delta, was applied to measure the degree of overlap between the distribution of various variables extracted pre-and during/post interventions [28]. Cliff's Delta approaching 1 or −1 indicates absence of overlap, whereas 0 indicates overlap completely. Wilcoxon signed rank test was applied to examine the statistical significance of the change. Changes with a p value < 0.05 was considered statistically significant. The secondary analysis consisted of the impact of PA time series smoothing on the conventional metrics and the complexity metric. The analysis compared the aforementioned statistical outcomes before and after smoothing. Additional analyses on Week0 data compared the distributions of the length of sedentary and walking bouts before and after smoothing. The coefficients of variance (CV) of the daily complexity value of the original and the smoothed PA time series were compared.
Results
In total, 30 participants were included in the pilot study with 10 participants at each trial site. Due to technical problems with the sensor devices (no data could be retrieved), eight participants were excluded from further data processing and analysis (n = 5 in Week0 and n = 3 in Week3). One participant's CBMS score was not available at the baseline assessment and was excluded from further data analysis leaving 21 participants for statistical analyses. All participants in statistical analysis had minimum 3 days of data in Week0 and Week3. Table 1 summarizes the descriptive statistics of PA metrics and CBMS score for the included participants (n = 21, except the CV of complexity, where n = 25 from Week0 were analyzed) in the pilot study. The included participants were not significantly different from the excluded participants in CMBS scores at the pre-(Willcoxon ranksum test, p = 0.14) or the post-(p = 0.23) intervention assessment. To illustrate, Figure 2 shows one participant's barcode (based on smoothed PA time series) in Week0 (left) and Week3 (right). The barcode in Week3 shows richer colours filled in throughout several days of measurements, which was reflected by the higher mean complexity score of 0.120 compared to 0.111 in Week0. For the primary analysis, scatter plots in Figure 3 show the correlations. Spearman correlations between changes in PA metrics (based on original PA time series) and CBMS score (between the change in CBMS score and the changes measured by various conventional univariate metrics and the complexity metric) are based on original PA time series. Changes in univariate metrics had no significant association with the change as measured by the CBMS score. In contrast, complexity had a significant positive correlation (ρ = 0.47, p = 0.03) with the change in CBMS. Complexity was higher after intervention with an effect size of 0.18, which was comparable to the effect size as measured by For the primary analysis, scatter plots in Figure 3 show the correlations. Spearman correlations between changes in PA metrics (based on original PA time series) and CBMS score (between the change in CBMS score and the changes measured by various conventional univariate metrics and the complexity metric) are based on original PA time series. Changes in univariate metrics had no significant association with the change as measured by the CBMS score. In contrast, complexity had a significant positive correlation (ρ = 0.47, p = 0.03) with the change in CBMS. Complexity was higher after intervention with an effect size of 0.18, which was comparable to the effect size as measured by the CBMS score (0.20). Despite an increase in complexity post intervention, the change was not statistically significant. Changes in univariate metrics had smaller effect size and were not statistically significant (see Table 1). For the primary analysis, scatter plots in Figure 3 show the correlations. Spearman correlations between changes in PA metrics (based on original PA time series) and CBMS score (between the change in CBMS score and the changes measured by various conventional univariate metrics and the complexity metric) are based on original PA time series. Changes in univariate metrics had no significant association with the change as measured by the CBMS score. In contrast, complexity had a significant positive correlation (ρ = 0.47, p = 0.03) with the change in CBMS. Complexity was higher after intervention with an effect size of 0.18, which was comparable to the effect size as measured by the CBMS score (0.20). Despite an increase in complexity post intervention, the change was not statistically significant. Changes in univariate metrics had smaller effect size and were not statistically significant (see Table 1). For the secondary analysis, after smoothing the PA time series, multiple very short sedentary bouts were merged into one longer bout. Similarly, multiple walking bouts with short interruptions were concatenated to form a continuous walking bout (see Appendix A.2). Comparison of mean values of various univariate metrics presented in Table 1 indicated that smoothing had little impact on the percentage of sedentary time and the percentage of walking time, whereas the total number For the secondary analysis, after smoothing the PA time series, multiple very short sedentary bouts were merged into one longer bout. Similarly, multiple walking bouts with short interruptions were concatenated to form a continuous walking bout (see Appendix A.2). Comparison of mean values of various univariate metrics presented in Table 1 indicated that smoothing had little impact on the percentage of sedentary time and the percentage of walking time, whereas the total number of steps and mean cadence were reduced. The value of complexity for the smoothed PA time series was smaller, compared to the original complexity. The change in complexity for the smoothed PA time series resulted in a stronger association with the change in CBMS score (ρ = 0.58, p = 0.01 as shown in Figure 4. Association between complexity change and CBMS score change after smoothing PA time series.). Moreover, the mean CV of complexity of the participants in Week0 decreased from 0.11 to 0.07 after smoothing the PA time series (see Appendix A.2).
Discussion
Authors should discuss the results and how they can be interpreted from the perspective of previous studies and of the working hypotheses. The findings and their implications should be discussed in the broadest context possible. Future research directions may also be highlighted. The primary analysis of this study focused on the relevance of various conventional PA metrics and the complexity in the assessment of functional change after an exercise intervention in younger older adults. In addition, we analyzed the impact of smoothing PA time series data on the calculation of various PA metrics and the complexity metric. Despite a very short intervention, the change in complexity was significantly correlated with the change as measured by the CBMS score, whereas, the changes in conventional PA metrics did not show significant correlation with the change as measured by the clinical assessment. Moreover, smoothing the PA time series, to aggregate short activity bouts, improved the complexity metric in terms of stronger correlation with functional change as measured by a clinical assessment and higher measurement reproducibility as quantified by the CV of one-week measurements.
These results revealed that complexity is a useful and a more sensitive metric than conventionally applied univariate PA metrics in the assessment of functional change in younger older adults. Conventional PA metrics derived from wearable sensors, step counts, or cadence, might not be sensitive enough to capture the functional change after short interventions. Metrics characterizing one aspect of daily physical activity, such as the time spent walking or being sedentary, do not provide a comprehensive picture of the determinants of functional status. The complexity metric of physical behaviour, on the other hand, characterizes the quantity, the quality, and the dynamic changes between different activities and different performances while doing the same activity (such as a change in cadence) in the 'barcode'. Further, the entropy rate increases while the number of sub-
Discussion
Authors should discuss the results and how they can be interpreted from the perspective of previous studies and of the working hypotheses. The findings and their implications should be discussed in the broadest context possible. Future research directions may also be highlighted. The primary analysis of this study focused on the relevance of various conventional PA metrics and the complexity in the assessment of functional change after an exercise intervention in younger older adults. In addition, we analyzed the impact of smoothing PA time series data on the calculation of various PA metrics and the complexity metric. Despite a very short intervention, the change in complexity was significantly correlated with the change as measured by the CBMS score, whereas, the changes in conventional PA metrics did not show significant correlation with the change as measured by the clinical assessment. Moreover, smoothing the PA time series, to aggregate short activity bouts, improved the complexity metric in terms of stronger correlation with functional change as measured by a clinical assessment and higher measurement reproducibility as quantified by the CV of one-week measurements.
These results revealed that complexity is a useful and a more sensitive metric than conventionally applied univariate PA metrics in the assessment of functional change in younger older adults. Conventional PA metrics derived from wearable sensors, step counts, or cadence, might not be sensitive enough to capture the functional change after short interventions. Metrics characterizing one aspect of daily physical activity, such as the time spent walking or being sedentary, do not provide a comprehensive picture of the determinants of functional status. The complexity metric of physical behaviour, on the other hand, characterizes the quantity, the quality, and the dynamic changes between different activities and different performances while doing the same activity (such as a change in cadence) in the 'barcode'. Further, the entropy rate increases while the number of sub-patterns in the 'barcode' increases as illustrated in Figure 2. Since it captures more aspects of physical behaviour simultaneously, the complexity metric has the potential to capture the underlined important aspect in the aging process that the variety and dimension of activities decreases due to functional decline [29,30].
The 'barcode', as defined in Table A1 and illustrated in Figure 2, is constructed with generic activity features derived from the wearable sensors. This makes complexity a generic metric for PA data analysis in principle without constraints in specific sensor configuration or wearing position. The activity features required by the 'barcode', such as walking time and number of steps, are universally recognized by the state-of-the-art wearable sensors [10,12,25]. Selection of activity features to be included in barcode construction is a topic worth separate investigation. For example, efficacy and sensitivity to wearing position of a 'barcode' that states sensor-derived activity levels (sedentary, light, and moderate-to-vigorous) can be analyzed [31].
The complexity metric analyses the entropy rate of the 'barcode'. Changes in 'barcode' states depend on the richness of the activity performed but is also influenced by the resolution of the activity features. The higher the resolution, the more detailed the features in the activity performed can be described; however, the resolution becomes less resistant to noise in the activity data. As demonstrated in Figure 1, the original PA time series (the top bar) has second-by-second feature resolution, which shows frequent fast changes between sedentary and walking activities (for example, see between 1000 and 1250 s). In the original PA time series, almost all sedentary bouts were shorter than five minutes, and less than 5% of walking continued for more than one minute in all participants (see distribution of PA activity data before and after smoothing in Figure A1). It is plausible that very short bouts were artifacts of the activity features due to the noise in the accelerometry signal acquired at the waist. The complexity metric aims to capture the dynamic change of real activity patterns that are encoded in the 'barcode' rather than the signal noise. Thus, a pre-processing method to remove the noise, or smooth the activity features before barcoding for complexity analysis is necessary and important. We proposed a smoothing method in this study to remove the noise. As shown in the secondary analysis, the resulting complexity value was lower than the complexity of the original PA time series, indicating that there was less frequent change in the activities after smoothing. The smoothing method improved the reproducibility of the complexity metric, which implies that the smoothing removed irrelevant noise existing in the original PA time series. However, it preserved the clinically relevant activity patterns as shown in the significant correlation with the clinical outcome. We observed a decrease in step counts and mean cadence after smoothing, which was likely due to the procedure, where steps in very short walking bouts were removed, but no step was inserted in the concatenated walking bouts.
There are some limitations in the study presented in this manuscript. First, the sample size was relatively small. Due to missing data, only 21 participants were included in the statistical analyses. Even though, there was no significant difference in functional status as measured by the CBMS score between the included and the excluded subject, the strength of association and effect size of change measured in this pilot study will need to be confirmed in a larger cohort of comparable participants. Secondly, the assessment of daily activity with wearable sensors during the intervention (Week3) was collected one week prior to the post-intervention assessment of CBMS. For the short intervention pilot, this time discrepancy might have introduced bias. This bias likely has made our estimates of pre-or post-intervention change in PA metrics to be more conservative. Lastly, the window of 30 s chosen for the smoothing method is based on the lowest threshold for walk analysis in the 'barcode' design. Conceptually, this threshold corresponds to most indoor walking activities [7,10]. However, future research should conduct a systematic evaluation to confirm the optimized activity feature resolution for complexity analysis. In the context of PreventIT project, the on-going multi-national randomized controlled trial (RCT) study will provide a larger cohort data of 180 participants with comparable demographic and health profile as studied in this pilot. The RCT will assess the participant's CBMSs and monitor their PAs using a similar sensor configuration at baseline, 6-month, and 12-month follow-up. Correlation between the change in complexity and the change in CBMS will be validated after the longer intervention. Analyses on the association between complexity and CBMS at baseline will be conducted. In addition, relationship between complexity and each individual balance and mobility components assessed in CBMS will be analyzed.
Conclusions
This study demonstrated the clinical relevance of using a multivariate metric for physical behavioural complexity to capture change in functional status in a longitudinal study in younger older adults. The complexity metric showed higher sensitivity to functional change than conventionally applied univariate PA metrics such as sedentary time and step count. Complexity can be applied as a generic metric to analyze the daily life activity patterns derived from wearable sensors. A meaningful resolution of sensor-derived activity features is important for reliable complexity analysis. The complexity metric is a useful metric to be further developed for the outcome measure of the feasibility and the effectiveness of PreventIT interventions.
Appendix A.1 Adapted Barcode Design and Complexity Computation
A 'barcode' state was assigned to each second of the time series according to Table A1. The definition was modified for the type 'active' from the original design presented in [10] that applied the absolute value of acceleration. As acceleration is sensitive to the sensor wearing location, the acceleration based threshold was replaced by the 'ActiCount' in this study. The 'ActiCount' thresholds are determined based on a validation study presented in [19]. Complexity computation was according to Equation (A1), where 'nrPattern' is the total number of sub-patterns found in the 'barcode'. 'nBC' is the total number of 'barcode' states. In our analysis, nBC = 18. N is the total length of PA time series in seconds. Complexity = nrPattern * ( log 10 (nrPattern) log 10 (nBC) Figure A1 shows that, after smoothing, 90% of the sedentary bouts lasted up to 15 min, whereas only 10% of the walking bouts were longer than two and half minutes before a stop. Figure A2 compares the mean and standard deviation of daily complexity in one-week measurement before and after smoothing. Smoothing lowered the mean values and the CV of the complexity. Figure A1 shows that, after smoothing, 90% of the sedentary bouts lasted up to 15 min, whereas only 10% of the walking bouts were longer than two and half minutes before a stop. Figure A2 compares the mean and standard deviation of daily complexity in one-week measurement before and after smoothing. Smoothing lowered the mean values and the CV of the complexity. | 2018-07-04T00:07:14.088Z | 2018-06-25T00:00:00.000 | {
"year": 2018,
"sha1": "2f4a26bbb99042b5ec97e5e9cc4b0ea3763a9fa9",
"oa_license": "CCBY",
"oa_url": "https://www.mdpi.com/1424-8220/18/7/2032/pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "2f4a26bbb99042b5ec97e5e9cc4b0ea3763a9fa9",
"s2fieldsofstudy": [
"Psychology"
],
"extfieldsofstudy": [
"Medicine",
"Computer Science"
]
} |
251109527 | pes2o/s2orc | v3-fos-license | Daily work pressure and task performance: The moderating role of recovery and sleep
Whereas previous research has focused on the link between (mental and physical) workload and task performance, less is known about the intervening mechanisms influencing this relationship. In the present study, we test the moderating roles of daily recovery and total sleep time in the relationship between work pressure and daily task performance. Using performance and recovery theories, we hypothesized that (a) work pressure relates positively to daily task performance, and that both (b) daily recovery in the form of psychological detachment and relaxation, and (c) total sleep time independently enhance this relationship. Our hypotheses were tested in a 30-day diary study with 110 officer cadets on a cross-Atlantic voyage on a Naval sail ship. The results of multilevel modeling lend support to all three hypotheses. Taken together, our findings suggest that recovery and sleep duration between shifts play a key role in the relationship between daily work pressure and task performance. We discuss the implications of these findings for the stressor-detachment model.
Introduction
Work in high reliability organizations (HRO) can be very demanding. For example, naval flight operations in the military and computer operations in a nuclear power plant involve complex work activities in a high-risk environment, where individuals need to work in an almost error-free manner (Vogus et al., 2014). As an example from military operations, Bartone (2006) identified stressors like isolation, uncertainty, boredom, powerlessness, danger, and high workload as typical stressors soldiers face in their work. In such contexts, where severe unanticipated interaction of multiple failures is more likely to occur than in most other work environments-high-quality work performance is of paramount importance, particularly in periods of high peaks in demands and production (La Porte, 1996;Rochlin et al., 1998;Vogus et al., 2014). Hence, the possible detrimental effects of impaired task performance in these settings may be catastrophic, as witnessed for example in the Chernobyl nuclear reactor explosion in Ukraine in 1986 (LaPorte andConsolini, 1991).
High-quality work performance in these settings requires highly knowledgeable and technically skilled employees (Rochlin et al., 1998;Vogus et al., 2014). These employees must also be in optimal physical and mental states in order to obtain and maintain high levels of alertness, vigilance and situational awareness (Bakker, 2011), and thus able to fulfill their tasks effectively in dangerous situations, like during hurricanes or when confronted with an aggressive fire (Flin, 1996). Notably, several studies have identified recovery from work as an important mechanism behind employees' cognitive and physical functioning over time, by revitalizing depleted cognitive and physical resources (Sonnentag and Fritz, 2015). Indeed, Shimazu et al. (2012) have shown that lack of rest and detachment from work (i.e., lack of recovery) coincides with reduced work engagement, psychological distress, and physical complaints. In the same vein, several studies have shown that lack of sleep, as a highly prevalent challenge in operational contexts (Miller et al., 2011), impairs performance on a wide range of cognitive, emotional, and technical tasksand subsequently impairs task performance in professional settings by depleting personal resources (Harrison and Horne, 2000;Olsen et al., 2016). In contrast, recovery sleep after sleep deprivation shows restoring effects on these personal resources (Bonnet, 2011).
In their recent meta-analysis on employee recovery from work, Steed et al. (2021) point to the importance of both studying what people do to recover (e.g., recovery sleep) and their recovery experiences to capture a more comprehensive picture of the total recovery processes during both day and night. This may be especially important in HRO operational work-settings, where the quantitative workload and time pressure may fluctuate unexpectedly from very low to very high from one day to the next, and the consequences of impaired task performance can be devastating. Therefore, most HRO's apply rigorous selection processes based on trait research in order to secure employees generally match the work challenges. However, studies show that even employees who generally perform well, may have off-days, which may be detrimental in a HRO setting. Thus, in addition to knowledge about between-person differences in explaining work performance, we need to expand knowledge regarding within-person differences. Within-person variations in exposure to work pressure and recovery may explain why individual performance fluctuates from one work-shift to the next (Beal et al., 2005). To examine daily fluctuations in work pressure and recovery, as well as performance, we designed a quantitative diary study encompassing 30 days of measurement in two samples of naval cadets on two 75-day training voyages across the North Sea and the Atlantic during the storm season on a 100-years old sail ship.
With this study, we aim to contribute to the literature in at least three ways. First, we contribute to the recovery literature by investigating the possible moderating roles of recovery experiences and total sleep time in the link between work pressure and task performance-on a daily basis. Do experience-based recovery strategies like psychological detachment and relaxation, and amount of sleep in-between shifts, strengthen the positive relationship between daily work pressure and performance? Whereas previous research has established relationships of daily work pressure with wellbeing (e.g., Hetland et al., 2021), we focus on task performance in the current study. We use effortrecovery (Meijman and Mulder, 1998) and stressor-detachment (Sonnentag and Fritz, 2015) models to argue that effort expenditure at work is associated with acute load reactions (e.g., accelerated heart rate, fatigue) which demand compensatory effort in order to perform adequately at work. Specifically, we propose that recovery experiences and sleep can prevent an accumulation of strain and thus facilitate performance in response to daily work pressure. Second, we contribute to the work performance literature by providing important knowledge about what individual employees can do in order to optimize their task performance under high pressure. Third and finally, by conducting the study in an operational naval setting, our research also contributes to the field of operational psychology in HRO settings-adding knowledge about factors enhancing safety and operational performance. Thus, the present study may also provide novel insights relevant to the generalizability of evidence conducted in work settings where employees work under less constrained conditions like among knowledge workers and land-based industries.
Work pressure and task performance
Task performance can be defined as the effectiveness with which an employee performs activities that contribute to the organization's technical core, either directly by implementing a part of its technological process, or indirectly by providing it with needed materials or services (Borman and Motowidlo, 1993). For a naval officer, performing controlled navigation, leading combat or salvage operations, maneuvering a ship, and maintaining safety behavior are examples of such task behaviors, often corresponding to formal job descriptions. In the present study, we argue that naval cadets who are exposed to high levels of daily work pressure, will perform well-particularly if they have opportunities to recover between shifts.
According to the challenge-hindrance stressor model (Cavanaugh et al., 2000;LePine et al., 2005), there are two types of job demands, namely, hindrance and challenge job demands. Hindrance job demands refer to conditions and tasks that require effort and energy, but do not have much potential for learning and growth (LePine et al., 2005;Van den Broeck et al., 2010). Typical hindrance demands are role conflict, bureaucracy, and job insecurity, such job demands can actually thwart basic psychological needs (Coxen et al., 2021). In contrast, challenge job demands, including workload, task complexity, and deadlines present conditions and tasks that require effort and energy, but effectively dealing with them may result in growth, learning, and goal attainment. For example, short deadlines may require high levels of energy, but finishing tasks before the deadline can also promote mastery and competence.
Given that high work pressure on board of a sailing ship is often related to rapid handling of challenging and risky situations (e.g., a storm requiring hours of salvaging sails 40 m above deck at night), such work pressure may be perceived as an experience that motivates and stimulates performance, from day to day. Indeed, several previous diary studies have indicated that work pressure can act as a challenge job demand that is positively related to the willingness to invest considerable effort in work and perform well (e.g., Tadić et al., 2015;De Gieter et al., 2018). It should be noted that some diary studies did not find evidence for a direct positive relationship between work pressure and performance (e.g., Prem et al., 2018).
Work pressure may foster focus, involvement, and learning, and will challenge employees to use a variety of personal strengths and act proactively to reach their daily work goals. Approaching this issue from the opposite direction, Gardner (1986) argues that low levels of activation that may be due to a low work pressure, may cause apathy and low levels of performance. The idea here is that increases in work-related stimulation energize workers and their subsequent performance when the current stimulation level is low. At this backdrop, we also operationalize daily work pressure in the current study as work-pace demands and amount of work facing the officers ( Van Veldhoven, 2014). Our quantitative diary study with repeated assessments allows us to control for an individual's performance the previous day, and, consequently, examine to what extent daily work pressure is related to day-to-day changes in performance. We propose that daily work-pressure challenges naval cadets to invest all possible effort, which facilitates daily performance.
Hypothesis 1: Daily work pressure is positively related to daily task performance.
The impact of recovery on the work pressure-performance relationship The somewhat unclear association between work pressure and performance has led researchers to search for moderators of the relationship (de Jonge et al., 2012). Given the expected high expenditure of cognitive and physical resources when facing high work pressure in a HRO context like sailing a ship, we propose that the ability to regain personal resources, in terms of recovery, moderates this relationship. Recovery is a process during which individual functional systems that have been activated during a stressful experience return to their prestressor levels (Meijman and Mulder, 1998). This process can be viewed as the opposite of the strain process, encompassing both the ability to refrain from work demands (i.e., avoiding activities that call upon the same internal resources as those required at work), and gain new internal resources such as energy, selfefficacy, or positive mood, which may help to restore threatened resources (Sonnentag and Fritz, 2007). Sonnentag and Fritz (2007) have argued that employees may engage in several off-job time activities to recover from workrelated effort, but that recovery experiences are most important. The authors suggest four different recovery experiences, namely, psychological detachment, relaxation, mastery, and control. A review by Bennett et al. (2018) has shown that the first two strategies are most often investigated. Detachment from work refers to refraining from job-related activities and not thinking about work during non-work time. Relaxation implies low levels of mental or physical activation and little physical or intellectual effort. On a ship, mental distancing may take shape as planning of leisure activities at homecoming or daydreaming about family, and relaxation can be achieved during daily activities like training in the gym, watching movies, or reading books. However, such daily recovery experiences may be difficult to realize on a ship, where off-job time is spent on the workplace, physical space for leisure activities is sparse, and resting hours often disrupted by noise or even alarms involving all personnel. Therefore, it is likely that recovery will differ from day to day for individual officers. On the basis of previous studies, such lack in recovery is expected to impair task performance, particularly during high work pressure, due to sustained sympathetic activation (Meijman and Mulder, 1998). Impaired restoration of personal resources will mean that individuals have less energy and less cognitive capacity to carry out the work tasks (Binnewies et al., 2009;Hooff and Geurts, 2014).
Conversely, on days with low work pressure, high recovery may have a negative effect on performance, by adding more relaxation, and more resources into a state of already low activation, which may give rise to apathy and disengagement (Gardner, 1986). Furthermore, it can be expected that on days with low work pressure, a state of low recovery has less impact on performance, compared to a state of high recovery, because the body of resources needed to master work demands on such days are so low that performance will be maintained, despite lack of recovery. However, during days with high work pressure, such lack of recovery, and subsequently reduced personal resources, may lead to lower performance compared to high recovery states. Formally stated: Hypothesis 2a: Daily psychological detachment moderates the positive relationship between daily work pressure and daily task performance. This relationship is stronger for those reporting high (vs. low) psychological detachment between work shifts.
Frontiers in Psychology 03 frontiersin.org Hypothesis 2b: Daily relaxation moderates the positive relationship between daily work pressure and daily task performance. This relationship is stronger for those reporting high (vs. low) psychological detachment between work shifts.
The role of sleep
A central assumption in the present study is that recovery sleep and recovery experiences contribute to a person's state of being recovered in a complementary way. The negative effects of sleep loss on performance are well documented in research covering a wide range of cognitive tasks and sensory functions, including skills relevant to effective functioning in an operational setting, like decision-making and planning (Harrison and Horne, 2000) as well as mood and positive affect (Dinges et al., 1997). According to Hobfoll (1989) an individual facing such resource depletion, particularly over time, will avoid challenges and transfer assignments to others in order to avoid investing further resources, and subsequently protecting a sense of mastery and self-image. This claim is supported for example by Olsen et al.'s (2016) finding that lack of daily sleep strongly impaired leadership performance in a naval bridge-team. Conversely, studies have shown that only one night of recovery sleep after even several days of sleepdeprivation has a strong and almost immediate positive impact on cognitive, technical, and physical performance, underscoring the restorative effect of sleep as part of the recovery process (Bonnet, 2011).
In sum, we suggest that it is likely that lack of daily sleep, particularly in situations with high workload and challenges, by depleting cognitive, affective, and motoric personal resources, may impair daily job performance (and vice versa). Hence, the relationship between work pressure and task performance may be moderated by total sleep time and sleepiness. As for recovery, we suggest that this relationship will be sensitive toward levels of work pressure. Thus, high job demands including work pressure may profit from longer sleep duration because sleep replenishes depleted resources, needed to master work tasks. In contrast, when the work pressure is relatively low, task performance may be less dependent of such restoration of resources and may even increase apathy and reduce performance due to a larger gap between energy levels and task demands (Gardner, 1986). It is also noteworthy that sleep differs greatly from night to night for the individual officers, due to factors like noise, alarms, and weather conditions (Nordmo et al., 2020). Notably, even though recovery processes contribute to replenishment of resources, we argue that sleep represents a supplementary strategy adding to the effects of physical and mental recovery, as defined in the current study. While detachment from work and relaxation can be considered as cognitive and emotional recovery processes obtained during waking time, sleep has additional and unique recovery characteristics in terms of biological recuperative processes that has been described in terms of several cellular mechanisms (e.g., Bonnet, 2011). Thus, we propose: Hypothesis 3: Total sleep time moderates the positive relationship between daily work pressure and daily task performance. This relationship is stronger for those reporting high (vs. low) total sleep time between shifts.
Materials and methods
Participants and procedure A total of 115 Norwegian naval cadets from a Military University College participated in our study. As part of their leadership training, the cadets traveled across the North Sea and the Atlantic from northern Europe to North America by sail ship. The data for the present study were collected during two voyages that took place in the fall of 2010 and the fall of 2011. All cadets participating on the two voyages were invited, and all the invited cadets volunteered to take part in the study. As an incentive to participate the cadets were promised to receive an individual report based on the daily measurements to be used for their personal development when they returned from the voyage. The invited participants were informed about the objective of the study and gave their written consent to participate 2 or 3 days before leaving port. In the written consent, it was clearly stated that participation was totally voluntary and that the cadets were free to withdraw from the study at any point in time without providing any reason. During the voyage the cadets participate in several different work tasks like sailing maneuvers, safety drills, and maintenance of the ship. Both in 2010 and 2011, data were collected using two different questionnaires. First, prior to the voyages, participants responded to a general survey measuring individual differences including demographics and participants' general task performance. Secondly, the participants received a booklet with diary questionnaires for the first 30 days of their voyage including questions assessing work pressure, task performance, psychological detachment, relaxation, and total sleep time. We requested the cadets to fill out the questionnaire at 5 p.m. on each day. The initial sample of the first data collection consisted of 54 cadets, while the initial sample of the second data collection consisted of 61 cadets. In the initial sample of the second data collection, five of the participants did not respond to the general questionnaire and were therefore excluded from the final sample. The final sample, combining the two data collections, consisted of 95 male cadets (86.4%) and 15 female cadets (13.6%) making up a total of 110 participants. The mean age of the participants was 23.46 years (SD = 2.96), and their age ranged from 19 to 33. Of the 110 cadets participating in the study, 98 were from the naval branch, 11 from the armed forces, and 1 from the air force. Out of the possible 3300 measurement occasions we obtained 2381 responses, yielding a response rate of 72.15% across the 30 days.
Measures
All measures used in the present study were based on existing scales translated to Norwegian using translation and back-translation (Brislin, 1970).
Trait survey
In order to measure general task performance we used four items from the task performance subscale developed by Goodman and Svyantek (1999). The participants were asked to respond to the following statements about their usual performance when they are on duty. Example items are: "I achieve the objectives of my job, " and "I fulfill all the requirements of my job." The respondents were asked to respond on a five-point scale ranging from totally disagree (1) to totally agree (5). The scale showed reasonable reliability (Cronbach's α = 0.74).
Daily diary booklet
We used daily diaries to measure fluctuations in our study variables. All day-level questionnaires were adapted versions of existing scales. We adapted the time frame of the scales and the number of items so the questions could be answered on a daily basis (cf. Ohly et al., 2010). Moreover, some of the introductory text and items were also adjusted to fit the military context on board of the sail ship.
Day-level work pressure was measured with four items referring to quantitative and time pressuring aspects of work. Items were based on a scale developed by Van Veldhoven et al. (2002). The items are "Today, I have had to work extra hard in order to complete something, " "Today, I had to work fast, " "Today, I had too much work to do, " and "Today, I have been working under time constraints." Responses were given on a five-point frequency scale, ranging from not at all (1) to a very large degree (5). The average within-level reliability coefficient (Cronbach's alpha) was 0.91, and reliability coefficients were in the range from 0.82 to 0.95 across the 30 days. The Norwegian version of the scale have been used three previous studies functioning as theoretically expected Ågotnes et al., 2021;Hetland et al., 2021).
Day-level recovery was measured with six items from the Recovery Experience Questionnaire (Sonnentag and Fritz, 2007) where three of the items were from the psychological detachment subscale and three from the relaxation subscale. The items were following a headline stating "When I have not been on duty the last twenty-four hours. . ." Example items are "I have distanced myself from my work" (psychological detachment), and "I have done relaxing things" (relaxation).
We used the same answer format as for task performance. The internal consistencies for the two subscales were tested across all days. The average within-level reliability coefficient (Cronbach's alpha) for the psychological detachment subscale was 0.75, and reliability coefficients were in the range from 0.56 to 0.86 across the 30 days. For the relaxation subscale the average within-level reliability coefficient (Cronbach's alpha) was 0.89, and reliability coefficients were in the range from 0.77 to 0.94 across the 30 days. The Norwegian version of the scales has shown to be functioning as theoretically expected in a previous study .
Total sleep time was measured by one single item. The item was phrased in the following way: "How many hours have you been sleeping the last twenty-four hours?" followed by a line where to enter the number of hours.
Day-level task performance was assessed with four adjusted items from the task performance subscale developed by Goodman and Svyantek (1999). Example items are "Today, I have achieved the objectives of my job"; "Today, I have fulfilled all the requirements of my job." Responses were given on a five-point frequency scale, ranging from totally disagree (1) to totally agree (5). The average within-level reliability coefficient (Cronbach's alpha) was 0.89, and reliability coefficients were in the range from 0.75 to 0.97 across the 30 days. The Norwegian version of the scale has shown to be functioning as theoretically expected in three previous studies Sørlie et al., 2022).
Strategy of analysis
In order to capture the multilevel structure of the data, implying that the 30 daily measurements (level 1) of the study constructs were nested within individuals (level 2), we applied multilevel analyses by the use of MLwiN 3.05. In the analyses, the level 1 (day level) predictors were centered on the respective person mean, while the level 2 (person level) variable was centered on the grand mean. In the multilevel analyses, we tested four models in the prediction of task performance. First an unpredicted null model (model 1) was tested, followed by a predicted main effect model (model 2), and two interaction models (Model 3 and 4). In all of the predicted models, we included the respondents general task performance and the average score of all predictors on the person-level following the procedure suggested by Bolger and Laurenceau (2013, pp. 77-78). In main effect model, we added the day-level predictors (daily work pressure, psychological detachment, relaxation, and total sleep time). In the first of the two interaction models (Model 3), the three hypothesized interactions between work pressure and the three predictors were added to the main effect model, while in the second interaction model (Model 4) we also included previous-day task performance. By controlling for the uncentered levels of previous-day task performance Correlations below the diagonal are correlations on the between (person) level and correlations above the diagonal are correlations on the within (day) level. *p < 0.05, **p < 0.001.
in the second interaction model, the temporal stability in the construct is accounted for. Hence, the explained variance in daily task performance can be interpreted as a positive or negative change in task performance from the previous to current day. In contrast, the relationship in the model not controlling for previous-day task performance should be regarded as a simultaneous (cross-sectional) effect.
Pre-analysis
To rule out the potential danger of conceptual overlap between the daily measurement of work pressure and work performance, we conducted a multilevel confirmatory factor analysis using their respective observed indicators across the 30 days. Testing the expected correlated two-factor solution resulted in a good overall fit to the data (χ 2 = 219.85, DF = 38, CFI = 0.98, TLI = 0.98, and RMSEA = 0.04), as well as a good fit on the two specific levels (SRMR within = 0.02 and SRMR between = 0.03). On the within-level, factor loadings were in the range of 0.66-0.89, and on the between-level in the range of 0.85-0.98. The two constructs correlated positively on the within-level (r = 0.10, p < 0.001) and negatively on the betweenlevel (r = −0.28, p < 0.001). Hence, the multilevel factor analysis supported the validity of considering the two concepts as two separate constructs on both levels.
Descriptive statistics
Means, standard deviations, within person-and between person-level correlations for all study variables are presented in Table 1. A significant, but weak positive correlation was found between daily work pressure and task performance the same day at the within-level (r = 0.08, p < 0.001). Moreover, a significant negative association was found between psychological detachment and task performance (r = −0.07, p < 0.001), and between total sleep time and task performance (r = −0.04, p = 0.034). Daily relaxation was not related to task performance at the within-level (r = −0.03, p = 0.101). Table 2 presents the results from the multilevel analysis predicting daily task performance. Prior to testing the predicted models, an unpredicted model (null-model) should be tested to confirm that there is sufficient day-level variance in the current dependent variable. As shown in Table 2, the initial unpredicted model revealed significant variation in daily task performance on both the day-level (70.6%) and person-level (29.4%) allowing us to proceed with testing the hypothesized multilevel models.
Multilevel analysis
In hypothesis 1, we propose a positive relationship between work pressure and daily task performance. As can be seen in Table 2, in the main effect model a positive relationship was found between daily work pressure and daily task performance (B = 0.051, p < 0.001) after controlling for general task performance and the average scores for all predictors at the person-level. Hence, hypothesis 1 was supported. Moreover, the main effect model also revealed a significant negative relationship between daily psychological detachment and daily task performance (B = −0.042, p < 0.001), while no significant main effects were found for either relaxation, or sleep duration.
In hypothesis 2a and 2b, we predict that daily psychological detachment and relaxation positively moderate the relationship between work pressure and daily task performance. As shown in Table 2, the first interaction model investigating simultaneous effects (Model 3), revealed a significant interaction between work pressure and relaxation (B = 041, p < 0.001), while the hypothesized interaction effect between work pressure and psychological detachment was not significant (B = 0.034, p = 0.053). The work pressure × relaxation interaction effect is illustrated in Figure 1. The figure indicates higher task performance on days with higher work pressure among naval cadets who reported high relaxation between shifts, while task performance is unrelated to work pressure among those who reported low relaxation. To formally test the significance of the Daily task performance by work pressure, psychological detachment, relaxation, and total sleep time controlled for general task performance and average person-level scores for all predictors (N = 3300 occasions, N = 110 participants). **p < 0.01, *p < 0.05. slopes in the interaction, we conducted a simple slope test at the condition of ±1 SD for both the predictor and moderator. The results revealed a significant positive slope for those reporting high relaxation (Slope = 0.102, z = 4.649, p < 0.001), and a non-significant slope for those scoring low in relaxation (Slope = 0.018, z = −0.798, p = 0.425). In sum, the results from the interaction model investigating simultaneous effects supported hypothesis 2b, while hypothesis 2a was not supported in the simultaneous-effect model.
In the second interaction model (Model 4), we controlled for previous day's levels of task performance in addition to the main effects and interaction effects included in simultaneous effect model. Hence, a positive relationship represents an increase in task performance from the previous day to the current day, and a negative relationship would represent a respective decrease in task performance. As shown in Table 2, the work pressure × relaxation interaction effect was not significant (B = 0.013, p = 0.222) when predicting change in daily task performance, whereas now the work pressure × psychological detachment interaction became significant (B = 0.044, p = 0.023). In accordance with the hypothesized enhancing effect of psychological detachment in hypothesis 2a, Figure 2 indicates a clear positive relationship between work pressure and daily changes in task performance among cadets reporting high psychological detachment, while the corresponding relationship among those reporting low psychological detachment was almost flat. Noteworthy, cadets who reported high psychological detachment between shifts showed lower daily task performance compared to cadets scoring low psychological detachment when work pressure was low. Consistent with Figure 2, formal testing of the slopes revealed a significant positive slope for those reporting high psychological detachment (Slope = 0.091, z = 3.71, p < 0.001), and a non-significant slope for those reporting low psychological detachment (Slope = 0.015, z = 0.269, p = 0.529). Hence, the second interaction model predicting change in daily task performance supported hypothesis 2a, while hypothesis 2b was not supported in the model.
In hypothesis 3, we propose that sleep time during the previous night positively moderates the relationship between daily work pressure and daily task performance. As can be seen in Table 2, the work pressure × total sleep time interaction effect was significant in both interaction models (B = 0.035, p < 0.001 and B = 0.28, p = 0.010, in the simultaneous-effect model and change-effect model, respectively). The interaction effect from the change-effect model is illustrated in Figure 3. As can be seen, work pressure resulted in positive changes in task performance for individuals who reported high total sleep time, while there was no such effect for individuals who reported low total sleep time. Notably, cadets reporting high total sleep time showed lower daily task performance compared to cadets reporting low total sleep time when work pressure was low. In line with Significant interaction effect between daily work pressure and daily relaxation on daily task performance (Simultaneous-effect model).
predictions, formal testing of the slopes at the conditions of ±1 SD for both the predictor and moderator revealed a significant positive slope for those with many sleep hours in-between the shifts (Slope = 0.088, z = 4009, p < 0.001), while the slope for those with limited number of sleeping hours was nonsignificant (slope = 0.018, z = 0.918, p = 0.359). The pattern and slopes of the interaction effect revealed in the simultaneouseffect model was almost identical to the interaction revealed in the change-effect model. Hence, hypothesis 3 was supported in both interaction models.
Discussion
The present study tested the link between work pressure and day-to-day changes in task performance in an operational naval setting. The central aim was to investigate whether daily recovery experiences and sleep duration would qualify this relationship. The results provide evidence for a direct link between daily work pressure and daily task performance, supporting that work pressure generally acts as a challenge demand for all individuals. Moreover, as hypothesized, the findings revealed that recovery experiences and total sleep time in-between shifts positively moderated the work pressure-performance relationship. Nevertheless, some temporal differences were found in the functioning of relaxation and psychological detachment in the work pressureperformance relationship. Work pressure was positively related to simultaneous levels of task performance on the days that were preceded by relaxation, while work pressure was related to positive changes in task performance on the days preceded by psychological detachment. In what follows, we discuss the most important theoretical and practical contributions of the study.
Theoretical contributions
Individuals who work in HRO are often exposed to high job demands, but are expected to work in an error-free manner (Vogus et al., 2014). Particularly on the days that there are high peaks in demands-for example, when confronted with extreme weather conditions, high work pressure and intense training-high-quality work performance is crucial (La Porte, 1996;Rochlin et al., 1998;Vogus et al., 2014). In the present study, daily work pressure was positively related to job performance, but this relationship was relatively weak. Moreover, when controlling for previous-day performance, the work pressure-performance relationship disappeared. These findings indicate that work pressure was not a positive predictor of performance, which is inconsistent with the challengehindrance stressor framework (Cavanaugh et al., 2000;LePine et al., 2005). According to this framework, hindrance job demands such as role ambiguity and interpersonal conflicts present conditions that require effort and energy, but do not have growth potential and do not contribute to performance (LePine et al., 2005;Crawford et al., 2010). In contrast, challenge demands, including work pressure and job complexity present work tasks and conditions that require effort and energy, but efficient dealing with them results in growth, mastery, and goal attainment. It should be noted that Prem et al. (2018) in a previous diary study also failed to demonstrate the expected positive link between time pressure and performance on the within-level among knowledge workers in diverse occupations and industries. Hence, this suggests that the findings from the present study may provide useful insights that also generalize to more mundane contexts than the constrained HRO setting being the focus of the present study.
One possible explanation for the null finding in the present study may be that the work pressure was simultaneously Significant interaction effect between daily work pressure and daily psychological detachment on daily change in task performance (Change-effect model).
FIGURE 3
Significant interaction effect between daily work pressure and total sleep time on daily change in task performance (Change-effect model).
motivating and exhausting for the naval cadets-together resulting in a null effect on performance. Importantly, most previous studies on the hindrance-challenge stressor framework have used cross-sectional or longitudinal research designstesting differences in wellbeing (which would help to perform well) between individuals exposed to low versus high levels of work pressure. The present study shows that when the focus shifts to within-person effects-i.e., on variations in work pressure from day to day, work pressure may actually not contribute to explaining wellbeing and goal attainment (see also Breevaart and Bakker, 2018).
Another explanation for the null findings offered by the current study is that not all cadets recovered sufficiently inbetween the shifts to be able to respond to work pressure with excellent performance. According to both the effortrecovery model (Meijman and Mulder, 1998) and the stressordetachment model (Sonnentag and Fritz, 2015), recovery experiences can buffer the link between job demands and strain. The model proposes that psychological detachment from work offers opportunities to recover from work-related stress and build new physical and psychological resources (e.g., vitality). Such resources can then be used to perform well, and particularly on days with high work pressure. Hetland et al. (2021) expanded this model by proposing and empirically investigating that relaxation may similarly offer opportunities for recovery, so that employees have sufficient energy to optimize their performances on challenging days. Although their findings only supported the job demands × recovery interaction for psychological detachment, the current study found evidence for a boosting role of relaxation as well as sleep duration. Thus, cadets who relaxed and slept well inbetween shifts were better able to transform work pressure into adequate performance. These findings expand the stressordetachment model by showing that next to psychological detachment it is relaxation and sleep duration that presumably restore the energy reservoir so that employees can perform well on busy days.
Nevertheless, one interesting finding is that relaxation only boosts the work pressure-task performance relationship when looking at simultaneous effects within the same day, while the boosting role of psychological detachment only occurred when predicting changes in task performance from one day to another. A possible explanation to this is that detachment relates more explicitly to what happened yesterday (e.g., not thinking about work when off duty) than relaxation. The individual distances themselves psychologically from the events that happened at work, including all events that affected performance. This works particularly well when analyzing changes in job performance. On the days individuals realize an increase in their performance, they have to invest additional effort, which is more likely after shifts during which cadets have detached from previous-day's events that influenced previous-days' performance. When cadets detach from their work-related efforts, they can build new energy resources by engaging in leisure time activities such as studying, exercising, and socializing. These new resources can be used during the next day to deal with work pressure and perform well.
It should be noted that an interesting sub-finding in the present study is a negative relationship between psychological detachment and daily task performance. One possible explanation to this unexpected negative relationship could be that when individuals detach themselves psychologically from their work, they no longer invest energy in the work activity. In contrast, when still thinking about work during off-job time, people may solve problems that may facilitate performance in the short run. Indeed, Niks et al. (2017) found that employees who did not detach from work during leisure time were more creative.
A last noteworthy finding in the present study was that the items supposed to measure psychological detachment evoked inconsistent responses, resulting in an unreliable measure during four of the 30 days of study (in 13% of the cases). Although it is not unusual to see fluctuations in reliabilities across days in diary studies, these findings suggest that distancing from work was not perceived similarly across days. We can only speculate about the reasons for this, and this may be fruitful for future research in which scholars try to explain inconsistencies in their daily assessments. For example, it is conceivable that detachment from work is more difficult and psychologically different if one tries to recover on the work location as compared to home or another off-job location. Of course, the cadets who participated in the present study were each day requested to report the extent to which they could psychologically detach from work. Perhaps detachment was almost impossible on some of the days (e.g., days with stormy weather), making the concept erroneous during these days. The lesson for diary research is that the phenomenology of the concepts studied may fluctuate and be a function of the environment in which these concepts are studied.
Limitations and future research
The present study has several limitations that should be taken into account when interpreting the findings. First, we used self-reports, which may have led to common method variancewhich increases the risk of inflated correlations. However, the correlations in the present study were generally low, which speaks against inflated correlations. Still, future research may want to include other-ratings of job performance and a separation in time of the independent and dependent variables. Second, applying booklets usually gives researchers no or very limited control over when participants fill in the survey. It might be that some participants filled in several surveys at the same time. It might be that they responded to the surveys at bedtime rather than in the afternoon. However, in order to make sure that participants filled in the daily questionnaires at the correct time (each day at 5 p.m.), we asked squad leaders to motivate the cadets to fill out the booklets at the right time. Also, the researcher on board on the ship saw to it that the right procedure was followed. It should also be noted that at the time of the study, we did not have an alternative to booklets, because in the mid of the North Sea and the Atlantic Ocean cadets have no access to the internet.
Third, our sample of naval cadets was rather specific, and can of course not be taken as representative of the general working population. Still, the findings were generally in line with stressor-detachment and effort-recovery models, which have mostly been studied among white-collar, office workers. Moreover, our findings do have relevant implications for studying task performance in office workers. Our findings show that employee job performance changes as a function of fluctuations in work pressure (from very low to very high), and that recovery is crucial for a positive linkage between work pressure and performance. Previous studies among employees working in a range of occupations (e.g., Tadić et al., 2015;De Gieter et al., 2018) have indicated that such fluctuations can be found in most sectors, and that therefore psychological detachment, relaxation, and sleep most likely have a similar restorative function in other occupational groups. Nevertheless, as discussed above, the working conditions on board of the sail ship are rather unique in that the participants spend most of their time together during an extended period of time. Moreover, off-job time is spent on the same location as work, which most likely increases the risk of work-to-nonwork interference and decreases opportunities for distancing from work. Future research should therefore try to replicate the current findings in other occupational settings. Finally, the sample used in the present study was young (23 years), and most participants were male. Research has shown that with increasing age, individuals lose and gain personal resources (e.g., decreased fluid intelligence, increased crystallized intelligence, decreased physical health, increased wisdom and experience; Truxillo et al., 2015)-which may all affect how individuals deal with and respond to job demands. It would seem important and interesting to test age effects in the stressor-detachment model in future research.
Practical implications
The findings from the present study have several practical implications. First, in order to optimize employee job performance on days with high work pressure, organizations and their leaders should assure that employees have sufficient opportunities to recover and actually engage in recovery activities during their off-job time. This is especially important in a work setting where employees both work and recover within the same physical environment like on board of an oil-tanker, off-shore platform or during military operations on board of a ship. Hence, in such work settings organizations should assure that employees have appropriate conditions to disengage from work and relax. This could for example be done by providing access to a sufficient number of quiet rooms ideal for reading a book, using a tablet, and watching a movie (using head phones) or a gym and swimming pool where employees can exercise, relax, and socialize.
Second, the knowledge from the present study could also be used by HR departments and consultancy firms to develop training programs to improve and facilitate employees' recovery experiences and activities. We have shown that recovery experiences as well as sleep have the potential to facilitate employee functioning when confronted with high levels of daily work pressure. Trainers may use this knowledge to develop tools and exercises through which employees learn how, when, and where to detach from work and relax-by engaging in certain activities during leisure time (e.g., socializing, reading, exercise) and abstaining from other activities (e.g., using smartphones just before bedtime). Finally, the findings from the present study show that total sleep time has an additional and complementary boosting effect on the link between the effort invested by the employees and their performance. Hence, organizations and their leaders should also assure good facilities and conditions for employees' recovery sleep. They may distribute educative apps and video material developed to facilitate falling to sleep and maintaining consistent sleep through the whole sleeping period.
Conclusion
The present study shows that psychological detachment, relaxation, and sufficient sleep help dealing with daily job demands in the form of work pressure. Individuals who are able to detach from their work, relax during their leisure time and who get enough sleep seem able to change high work pressure into a challenge job demand that helps them to perform well. Since previous research has shown that work pressure is an important predictor of job stress that undermines performance (e.g., Taris, 2006), the current study shows that daily recovery is crucial for effective coping with daily job demands and efficacious daily job performance.
Data availability statement
The raw data supporting the conclusions of this article will be made available by the corresponding author on resonable request. | 2022-07-28T14:06:42.718Z | 2022-07-28T00:00:00.000 | {
"year": 2022,
"sha1": "cf18039afe2adc16ecff7c3ca03af8b33eff3dc0",
"oa_license": null,
"oa_url": null,
"oa_status": null,
"pdf_src": "Frontier",
"pdf_hash": "cf18039afe2adc16ecff7c3ca03af8b33eff3dc0",
"s2fieldsofstudy": [
"Psychology"
],
"extfieldsofstudy": [
"Medicine"
]
} |
237473255 | pes2o/s2orc | v3-fos-license | Severity of self-reported symptoms and psychological burden 6-months after hospital admission for COVID-19: a prospective cohort study
Objectives Few studies have reported clinical COVID-19 sequelae six months (M6) after hospital discharge, but none has studied symptom severity. Methods Prevalence and severity of 7 symptoms were estimated until M6 using the self-administered influenza severity scale in COVID-19 hospitalized patients enrolled in the French COVID cohort. Factors associated with severity were assessed by logistic regression. Anxiety, depression and health-related quality of life (HRQL) were also assessed. Results At M6, among the 324 patients (median age 61 years, 63% men, 19% admitted to intensive care during the acute phase), 187/324 (58%) reported at least one symptom, mostly fatigue (47%) and myalgia (23%). Symptom severity was scored, at most, mild in 125 (67%), moderate in 44 (23%) and severe in 18 (10%). Female gender was the sole factor associated with moderate/severe symptom reporting (OR = 1.98, 95%CI=1.13-3.47). Among the 225 patients with psychological assessment, 24 (11%) had anxiety, 18 (8%) depressive symptoms, and their physical HRQL was significantly poorer than the general population (p=0.0005). Conclusion Even if 58% of patients reported ≥1 symptom at M6, less than 7% rated any symptom as severe. Assessing symptoms severity could be helpful to identify patients requiring appropriate medical care. Women may require special attention.
Introduction
Since the first cases of SARS-CoV-2 infection in December 2019, clinical presentation of COVID-19 in its acute phase has been largely described ( Docherty et al., 2020 ;Richardson et al., 2020 ). However, long-term clinical sequelae of COVID-19 remain unclear. A few studies reported persistent symptoms 2 to 4 months post discharge ( Carfì et al., 2020 ;Garrigues et al., 2020 ;Xiong et al., 2021 ). More recently, Huang et al. described the 6-months consequences of COVID-19 and reported presence of fatigue or myalgia in 63% of patients in an inpatients single-center cohort in China ( Huang et al., 2021 ).
Here in a multicenter prospective cohort in France, we assessed self-reported symptoms 6 months after hospital admission for COVID-19, using the influenza severity scale and described the evolution following diagnosis. We also assessed anxiety, depression and health-related quality of life (HRQL) to measure the impact of these symptoms on patients' global health.
Study oversight
The French COVID cohort (NCT04262921) is a national prospective multi-center cohort study enrolling hospitalized patients with a RT-PCR virologically confirmed COVID-19 in 80 hospitals in France since January 24, 2020 ( Yazdanpanah, 2021 ). Briefly, patients were followed-up from hospital admission (D1) throughout hospitalization for COVID-19 and at discharge, 2 to 4 weeks after discharge, month 3 (M3) and 6 (M6). The assessment of symptoms, anxiety, depression and health-related quality of life using selfadministered questionnaires were proposed to the patients at each visit. All adult patients who fulfilled the self-administered symptoms questionnaire at months 3 (M3) and 6 (M6) by March 22 nd , 2021 were included in the present analysis.
Procedures
The self-administered symptoms questionnaire recorded the presence and the severity of the 7 following symptoms: fatigue, myalgia, headache, cough, nasal obstruction, sore throat and feverishness. All symptoms were rated by the patient using a fourpoint scale (0, none; 1, mild; 2, moderate; 3, severe) using a selfadministered questionnaire previously used in influenza infection ( Duval et al., 2010 ;Hayden et al., 1997 ). For each date of evaluation, a total score ranging from 0 to 21 was obtained by summing the points attributed for each symptom. Based on the definition of symptom alleviation used in influenza, reporting of at least one moderate or severe symptom was considered to reflect an abnormal state of health. Four additional COVID-19 symptoms (joint pain, dyspnea, anosmia and ageusia) which were not included in the influenza questionnaire were also collected by the practitioner (thereafter referred to as "practitioner reported symptoms") during M3 and M6 visits.
Anxiety and depression symptoms were assessed using the Hospital Anxiety and Depression scale (HADS), subdivided in the HADS-Anxiety (HADS-A) and the HADS-Depression (HADS-D) scales. Both scales contain 7 questions scored by the patients on a 4-point Likert scale (0-3) with higher scores indicating more severe anxiety/depression. The items scores were summed up separately for HADS-A and HADS-D, leading to two scores ranging between 0 and 21. Scores greater than or equal to 11 points indicated abnormal levels ( Zigmond and Snaith, 1983 ).
Health-related quality of life (HRQL) was assessed using the SF-12 Health Survey ( Gandek et al., 1998 ) including a Physical Component Summary (PCS) HRQL score and a Mental Component Summary (MCS) HRQL score. These scores range from 0 to 100, with a high value indicating good HRQL. A patient was defined as having an altered physical (or mental) HRQL if his PCS (or MCS) was lower than the 25 th percentile of the score distribution in the general French population of the same age group and gender ( Carrieri et al., 2003 ).
The self-administered questionnaires were collected using RED-Cap electronic data capture tools ( Harris et al., 2009 ) with a secured personal access given to each patient after hospital discharge.
Statistical analyses
Categorical variables were summarized as counts (percentage) and frequency distributions were compared with the Chi-square, the Fisher exact or the McNemar for paired samples tests as appropriate. Continuous variables were expressed as median [interquartile range (IQR)] unless otherwise specified and compared with the Mann-Whitney U test. Prevalence of symptoms is given with their 95% confidence interval, estimated by using the exact Clopper-Pearson method.
To assess the representativeness of the population of patients who fulfilled the questionnaire, demographic characteristics, comorbidities and clinical data at hospital admission were compared between patients who fulfilled or not the 7-symptoms questionnaire at M3 and M6 using logistic multivariate regression models. The latter were adjusted on age, sex and ethnic group, in order to assess for confounding variables.
Associations between having at least one moderate or severe symptom at M6 and baseline characteristics were assessed through univariate logistic regressions. All variables were put in the multivariate models. Variable selection was then performed by starting with a model that included all covariates and then excluding those that did not improve the overall fit as measured by the likelihood ratio test (LRT). A P-value cut-off point of .05 was used as a stopping rule for this backward manual selection. Two-way interactions between risk factors kept in the multivariate analysis (including "ICU during the acute phase") were tested. No imputation strategy was applied for missing data. Any case that had a missing value was discarded from the analysis. All tests were 2-sided and p-values < .05 were considered significant. All statistical analyses were performed using R version 4.0.2.
Ethics and regulatory issues
The study was conducted with the understanding and the consent of each participant or its surrogate. The French Ethics Committee (CPP-Ile-de-France VI, ID RCB: 2020-A00256-33) approved the study protocol.
As compared to the 2,736 patients who did not complete the 7-symptoms questionnaire at M3 and M6, the 324 patients were younger, less likely to have diabetes, chronic kidney disease at admission, or to have been hospitalized in ICU (Table S1).
Global burden of the disease at Month 6
The combinations of 7-symptoms score ≥ 3, anxiety, depression, and impaired physical and mental HRQL at M6 are presented in Figure S2; 116/225 (52%) patients presented at least one modality among total 7-symptoms score ≥3, HADS-Anxiety score ≥ 11, HADS-Depression score ≥ 11, impaired physical HRQL and impaired mental HRQL. Most frequent modality and combinations were impaired physical HRQL, total 7-symptoms score ≥3, and combination of impaired physical HRQL with total 7-symptoms score ≥3. Twenty-eight of the 135 (21%) patients who had professional activities before admission had not returned to work at M6. Barplot representing each symptom severity for the N = 324 patients who fully completed their 7-symptoms questionnaire at 3 and 6 months after diagnostic confirmation. The score for each symptom is given on a four-degree scale going from 0 to 3 (i.e. none, mild, moderate, severe). The corresponding percentages are given in each colored bar. Of note, patients presenting no symptom at all are not represented on this graph.
M6 evaluation as compared to previous evaluations
The proportions of patients who self-reported symptoms regardless of their severity at D1, discharge, 2 to 4 weeks after discharge, M3 and M6 are shown in Figure 2 a. Self-reporting of each symptom was not significantly different between M3 and M6. Similar results were obtained in a sensitivity analysis excluding patients who were admitted to ICU at any time during hospitalization ( Figure S3). Reporting of each symptom according to its severity at M3 and M6 are represented in Figure 1 .
Discussion
To our knowledge, this is the first study assessing the severity of persistent symptoms 6 months after hospital admission for COVID-19 and their evolution since admission. In our population, 56% of patients still reported at least one symptom at M6, but less than 7% rated any symptom as severe. Except for female gender, we did not identify any other factor linked to high burden of symptoms at M6 that could prompt specific follow-up management options, and thus improve patients' outcome.
Note: HADS is divided into an anxiety (HADS-A) and depression (HADS-D) subscale.
Each HADS item was scored on a 4-point Likert scale (0-3) with higher scores indicating more severe anxiety/depression. The items scores were summed up separately for HADS-A and HADS-D, leading to two scores ranging between 0 and 21. Scores between 11-21 points indicated abnormal levels. SF-12: an individual was defined as having an impaired physical (or mental) health-related quality of life if his Physical Component Summary (or Mental Component Summary) was lower than the 25 th percentile of the distribution of the score in the general French population of the same age group and gender.
The French COVID cohort was launched at the very beginning of COVID-19 pandemic when the first cases were identified in France and recruited patients in all types of French hospitals, throughout France including overseas territories. The database extraction date of March 22, 2021 made it possible to assess the health status of patients included between January 24, 2020 and September 22, 2020. The younger age of the population who responded to the Boxplots of the total score obtained by adding the scores obtained for each of the 7 self-reported symptoms. The score for each symptom is given on a four-degree scale going from 0 to 3, the total score is thus between 0 and 21. The red dots represent the mean values. questionnaire compared to those who did not, was probably explained by a greater propensity of younger people to complete a questionnaire online. The lower age of the respondents was logically associated with a lower proportion of comorbidity and of ICU admission.
The proportion of patients (approximately one out of two) who reported at least one symptom at M6 using the 7-symptom influenza questionnaire was high, with most common reported symptoms being fatigue and myalgia; this is consistent with previous studies on mid-term follow up of SARS-CoV-2 ( Carfì et al., 2020 ;Garrigues et al., 2020 ;Huang et al., 2021 ;Xiong et al., 2021 ) or long term follow-up of other SARS survivors ( Lam et al., 2009 ;Lee et al., 2007 ;Tansey et al., 2007 ), although the range of values was wide. This large range of reported symptoms could be due to differences in the characteristics of patients included in the cohort or filling the forms, such as the age, the proportions of women or of patients with comorbidities ( Huang et al., 2021 ).
Since we did not know patients' symptoms before they experienced COVID, we were not able to assess to what extent the M6 reported symptoms were related to the COVID episode or to a preexisting condition. For this reason, and taking into account the experience of self-questionnaires performed in influenza, we assessed the severity of the symptoms and defined symptom alleviation as seven symptoms were scored absent or only mild ( Duval et al., 2010 ;Hayden et al., 1997 ). This led to a fifth of the patients being considered to have a poor M6 clinical outcome. Beyond the proportion attributable to COVID that cannot be precisely estimated, on one hand, the association between a score higher than three and a poor physical quality of life, and on the other hand, the significantly higher proportion of subjects with an impaired physical quality of life compared to a control population argue for the COVID's responsibility for the symptoms reported by patients. This is all the more to be taken into account as the persistence of symptoms as well as the alteration in quality of life at M3 and M6 did not suggest a rapid improvement and suggests to extend the patient follow-up beyond M6. Considering the persistence of altered quality of life between M3 and M6, especially regarding mental health, supportive care should be provided as early as 3 months after hospitalization in patients who need it.
Female gender was found to be the sole risk factor of persistence of moderate or severe symptoms, consistent with previous studies in SARS-CoV-2 survivors ( Ghosn et al., 2021 ;Huang et al., 2021 ;Xiong et al., 2021 ), whereas women are prone to develop less severe acute COVID than men ( Richardson et al., 2020 ;Yazdanpanah and French COVID cohort investigators and study group, 2021;Zhou et al., 2020 ). The pathophysiology mechanisms underlying this finding remain to be explored. Neither ICU admission, nor reporting of moderate or severe symptoms at admission which could be considered as a proxy of the disease severity at admission, were associated with poor clinical outcome at M6. This study has several limitations. First, M3 and M6 selfadministered questionnaires were not completed by all survivors. On one hand, as the population was younger and therefore less prone to have been admitted to ICU during hospitalization, our study might underestimate the proportion of COVID-19 survivors with poor clinical outcome at M6 post-hospitalization ( Huang et al., 2021 ). On the other hand, one cannot rule out that the propensity to complete a questionnaire is higher among those with symptoms than those without. Second, M6 assessment took place between end of July, 2020 and March, 2021, a time period during which the health situation in France deteriorated with an alternation of curfews and lockdown measures, thus promoting an anxiety-provoking climate. Differences in the patient characteristics of those evaluated and those not, specific health care facilities bottlenecked at the time of follow-up, improvement of support of care strategies over the outbreak and evolution of virus character-istics did not allow us to extrapolate our results to all survivors of COVID-19. Third, the 7-symptoms questionnaire was previously used to assess symptom alleviation in the cute phase of influenza disease, but not long-term sequelae. However, when focusing on patients with a total score above or equal to 3 at M6, the same number of patients had a poor outcome and the determinants remained the same.
Identifying patients needing specific care after COVID-19 could be beneficial in terms of public health. The use of a simple webbased questionnaire with a scale of severity could be helpful to identify patients requiring appropriate medical care including psychological support within months following recovery, especially since close follow-up of all COVID-19 survivors after hospital discharge seems difficult to achieve in the context of successive epidemic waves and congestion in health-care facilities.
In conclusion, persistence of symptoms in more than half of the patients and impaired physical health-related quality of life at M6 promotes long-term follow-up beyond six months following recovery.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. | 2021-09-11T13:09:31.527Z | 2021-09-10T00:00:00.000 | {
"year": 2021,
"sha1": "9442ec803123b6326c3ee3b1e0a45aa1710fbeb8",
"oa_license": "CCBYNCND",
"oa_url": "http://www.ijidonline.com/article/S1201971221007244/pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "fe0047401a50623bb57b65a87a3e9580a506beff",
"s2fieldsofstudy": [
"Medicine",
"Psychology"
],
"extfieldsofstudy": [
"Medicine"
]
} |
104654112 | pes2o/s2orc | v3-fos-license | Antibacterial Effect of Aqueous and Alcoholic Ginger Extracts on Periodontal Pathogen Aggregatibacter Actinomycetem Comitans [An in Vitro Study] (Part 1)
Among natural food sources with antimicrobial activities, ginger rhizome has been used as widely grown food spices and medicinal crops for centuries. Furthermore, it possess antifungal and antioxidant properties due to the phenols – related constituents (gingerols) that constrain the growth of many Gram positive and Gram negative bacteria including some periodontal bacteria. The Actinomycetem comitans is a portion of the normal microbiota in numerous healthy individuals but is also a major etiological agent in some aggressive and chronic types of periodontitis. The present study was conducted to test the effect of aqueous and alcoholic ginger extracts on the growth of Aggregatibacter actinomycetem comitans in comparision to 0.2% chlorohexidine gluconate mouth wash and distilled water in vitro, determination of ginger extracts minimum inhibitory concentration and minimum bactericidal concentration and detection of active ingredients of ginger extracts by using the high-performance liquid chromatography as well as chemical elements.
Introduction:
The increased antibiotics application lead to development of antimicrobial resistance of the microorganisms that became a serious problem (1) .Consequently there is a wide need to find the substitute of chemotherapeutic medications for treatment of diseases especially those derived from plants that are easily obtainable and have less side effects (2) .Ginger (Zingiber Officinale) is a medicinal plant that has been used widely all over the world, for thousands of years and for treatment of a wide range of unrelated ailments including arthritis, cramps, sore throats, infectious diseases, pains, constipation, hypertension and fever (3) .Also alcoholic ginger extract exhibited antimicrobial effect agains ـــــــــــــــــــــــــــــــــــــــ (1) Dentist, Department Of Periodontics, College Of Dentistry, University Of Baghdad.(2) Assist.Prof., Department Of Periodontics, College Of Dentistry, University Of Baghdad.
Mutans streptococcus and Candida albicans (4)(5)(6)(7) , as well as against some Gram negative periodontal pathogenic bacteria (Porphyromonas gingivalis, Porphyromonas endodontalis and Prevotella intermedia) (8) .Periodontitis "is an inflammatory disease of toothsupporting tissues, characterized by loss of connective tissue attachment and alveolar bone".The sub gingival plaque bacteria is the primary etiological agent of periodontitis.It is mostly believed that periodontitis is a polymicrobial disease, and more relevant to complex interactions between specific pathogens than are individual species (9) .Three of these bacteria, Aggregatibacter actinomycetem comitans, Tannerella forsythia, and Porphyromonas gingivalis, were formally designated as etiological agents of periodontitis (10) .Numerous studies recommended that the consequence of periodontal treatment is superior if the -1 ) 2017 ( 5 ….Antibacterial Effect of Aqueous 2 detection of certain pathogens especially Aggrecatibacter actinomycetem comitans and Porphyromonas gingivalis can no longer present after therapy (11,12) .As ginger possesses antibacterial and antioxidant properties, and as there were no previous studies that evaluate the effect of ginger extracts (aqueous and alcoholic) against Aggrecatibacter actinomycetem comitans and compare the effect with 0.2% chlorohexidine and distilled water, this study was conducted.
Materials and method:
The present study involved two experiments in vitro and it was conducted at Microbiological laboratory for the postgraduate students in The Basic Science department/College of Dentistry/ Baghdad University.*Extraction of aqueous and alcoholic extracts of ginger Aqueous extract: 40 gm of ginger powder were placed in 160 ml of sterile distilled water (D.W.) and left at room temperature for 24 hrs.with continuous mixing by using magnetic stirrer.Then the mixture was filtered and dried using incubator at 40°C.The liquid evaporated, and the concentrated extract was left at the base of the baker (13) .Alcoholic extract: 40 gm of ginger powder were put in a glass jar then 500 ml of 99.9 ethanol alcohol was added and mixed well.The container was sealed with cotton and foil to prevent evaporation of alcohol and left at room temperature for 24 hours.Then, the contents were filtered after that concentrated by evaporating the solvent alcohol in a hot air oven at 40ºC for 24 hrs. (14).*The determination of ginger active ingredients by using the high-performance liquid chromatography (HPLC), was done in the Ministry of Science and Technology, Department of material research .Which, is a chromatographic technique applied to discrete the active ingredients of ginger essential phenols that depend on the retention time of each ingredient.Spectrophotometer was used for the quantitative determination of chemical elements of alcoholic and aqueous ginger extracts.*Preparation of a selective agar media for A.a : 40 gm of trypticase soy agar was dissolved in 1000 ml D.W., then heating till boiling, sterilization by autoclaving and leftward till cool down to about 40-45º C in a water bath, after that, 50 ml of sterile blood were added to 128 µg/ml bacitracin and 8 µg/ml malachite green and mixed well then the media was poured into sterile Petri dishes (15) .* Preparation of a selective liquid broth media for A. a: 30 gm of trypticase soy broth powder dissolved in 1000 ml of D.W., then heating till boiling, and sterilized using autoclave and left to cool down to about 40-45ºC in a water bath, next, aseptically adding 128 µg /ml of bacitracin and 8 µg /ml of malachite green mixed well then the media was poured into sterile glass tube,then left to cool down to about 25ºC and stored at the refrigerator till used (15) .*The plaque samples were obtained from patients attending the clinic at the department of Periodontics in the teaching hospital of Dentistry College / Baghdad University.The patients were informed about the goals of the study and patient's agreements and approvals were gained before samples collection.The sub gingival plaque samples were collected from 50 systematically healthy patients (males and females) with age range of (40-60) years old, suffering from chronic periodontitis, which required the existence of at least 4 sites with probing pocket depth ≥ 4mm and clinical attachment loss of (1-2) mm or more (16) .Samples were collected from one pocket for each patient, and from the deepest part of the periodontal pocket applying Gracey curette and the sample placed on a swab which is introduced into a transfer media immediately and transferred to the lab to be inoculated on the selective agar media under anaerobic conditions using anaerobic gas pack and anaerobic jar at 37°C for 48 hrs.*Isolation and identification of A.a colonies was made depending on the morphology of the colonies, Gram stain and several biochemical tests (Indole, oxidase, catalase, coagulase, urease and API tests), hemolytic ability and antibiotic sensitivity test.* First experiment: Sensitivity of A.a: Agar well diffusion method was done to test the sensitivity of A.a to different concentrations of ginger extracts (20%, 30%, 40%, 50%, 60%, 70%, 80%, 90% and 100%), 0.2% chlorohexidine gluconate (CHX) as a positive control and distilled water as a negative control.The A.a was spread on the selective media, several wells (4-6) of equal size and depth were prepared in each agar plate, each well was filled with 0.1 ml of the agent (ginger extracts) being tested and other wells filled with CHX and D.W. agents.Plates were incubated anaerobically for 72 hrs., a ruler was used for the measurement of inhibition zone.*Second experiment: Determination of minimum inhibitory concentration (MIC) and minimum bactericidal concentration (MBC) of alcoholic and aqueous ginger extracts against A.a: a.Serial dilution was performed: 9 ml of A.a broth is dispersed into test tube, and then 1ml of bacterial suspension was added to achieve 10 ml.Then from the first tube we took 1 ml and dispersed to the second test tube and complete with 9 ml of A.a broth to have the first dilution, then we repeat the procedure for 4 times in 4 sequential tubes to reach 5 dilutions .b.The MIC is the lowest concentration of an antimicrobial agent that will inhibit the visible growth of a microorganism (17) .Test tubes were labeled by the No. of the different concentrations of the ginger extracts (aqueous and alcoholic), after that 1 ml of bacterial suspension were added to each tube then 0.5 ml of the ginger extracts were added to its designated tube.Then the tubes anaerobically incubated for 48 hrs., after that, the tubes were examined to notice if any turbidity was present (turbidity indicates bacterial growth), the tubes that lack the turbidity were identified as the MIC.c.The MBC is the lowest concentration of an antimicrobial agent required to kill a particular bacterium (18) .A swab was taken from each tube that showed absence of turbidity and spread on the selective agar media plates and incubated anaerobically for 48 hrs., the plates that displayed no growth were identified as MBC.Statistical analysis was done using mean, standard deviation S.D., One-way Analysis of variance test (ANOVA) test, least significant difference (LSD) and Independent sample t-test.In the statistical evaluation, the following levels of significance were used, NS: P > 0.05, S: 0.05 ≥ P > 0.01 and HS: P ≤ 0.01.
Results:
*Isolation and identification of A.a : According to their morphological characteristics, A.a colonies were white to grayish in color, about 1 mm in diameter and they adhered well to the agar, microscopic examination revealed that A.a were gram negative rods, arranged in short or medium length chains.Biochemical tests revealed that A.a was oxidase and catalase positive but, it was indole, coagulase and urease negative,and according to API test, A.a was listed as (Haemophilus actinomycetem comitans), also it displayed positive hemolytic ability and it was resistant to both Kanamycin and Vancomycin antibiotics.*Sensitivity of A.a to different concentrations of ginger extracts,0.2%CHX and D.W. : The diameter of inhibition zones were found to increase as the concentrations of both ginger extracts increased.Hence, all aqueous ginger extract concentrations revealed mean values of inhibition zones less than the CHX, 100% concentration showed maximum mean value of inhibition zone was (12.77 mm).Alcoholic ginger extract concentrations 80%, 90% and 100% displayed mean values of inhibition zones higher than 0.2% CHX, whereas, D.W. showed no inhibition zone.The mean values of inhibition zones of alcoholic extract started at 20% concentration which was (8.59 mm) while, it became (15.38 mm) at 100% concentration.One way ANOVA test showed highly significant differences amongst different concentrations of each ginger extract, 0.2% CHX and D.W., Table (1).From Table (2), the concentrations of aqueous ginger extract revealed highly significant differences, but, the differences were significant between 50% with 80% and 90% with 100%.While non-significant differences between 20% with 30%, and 60% with 70% as well as, 50% with 40%, 60% and 70%, in addition, 80% with 90%, 70% and 60%.Alcoholic ginger extract demonstrated highly significant differences, except for concentrations 30% with 40%, 60% with 70% and 90% with both 80% and 100%,since there were nonsignificant differences, however there was significant difference between 80% with 100%.Table (3), demonstrated highly significance differences between CHX, D.W. and each concentration of aqueous and alcoholic ginger extracts, except, the concentrations 80% and 90% of alcoholic extract that revealed non-significant and significant differences respectively with CHX.Table (4), revealed that the mean values of inhibition zones of alcoholic extract in all concentrations higher than that of aqueous, with highly significant differences between mean values of A.a inhibition zones for the same concentration of both ginger extracts, but, there was significant difference at 40% concentration.The HPLC analysis, Table (5), of active ingredients displayed that alcoholic extract had higher contents of 10-gingerol and 6-shogoal than the aqueous ginger extract, while aqueous extract had higher contents of 6-gingerol and 8-gingerol as compared with alcoholic extract.While the detection of chemical elements of ginger extracts showed that Potassium present in highest concentration from the other elements, followed by Magnesium.
The least element concentration in both ginger extracts was Copper (Cu) element.However, alcoholic ginger extract had higher concentrations of all elements except, Copper, whereas Calcium (Ca) found only in the alcoholic extract, Table (6).The MIC of aqueous ginger extract was 80% concentration ,while of alcoholic ginger extract was 50%, CHX also showed bacteriostatic effect on A.a.The MBC of aqueous ginger extract was 100%, while of alcoholic ginger extract was 80%.
Discussion
The fresh ginger was described to contain fibers, protein, carbohydrates fat, lipids (including glycerides, phosphatidic acid, lecithins, and fatty acids), protease, minerals e.g.(iron, calcium, magnesium, potassium, and phosphorous), as well as it contains vitamins such as(thiamine, riboflavin, niacin and vitamin C) (19,20) .Ginger has strong antibacterial effect and to some extent antifungal activities (21) .Generally, numerous studies revealed that ginger antimicrobial effects primarily was due to the existence of oxygenated mono and sesquiterpenes phenolic constituents (shogaol, gingerol) (22) , which are lipid-soluble phenol compounds mostly isolated from the root of ginger (23,24) .Most of the phenols can change the cell permeability because they are protein denaturing agents, which may lead to swelling and rupture of the bacterial cells, also most of them are metal cheaters that bind to the active site of metabolic enzymes, decreasing the enzyme activities and therefore decelerating bacterial metabolism and reproduction (25) .*Results of A.a sensitivity to ginger extracts revealed that both alcoholic and aqueous extracts had the ability to inhibit the growth of A.a and the antimicrobial activity of both extracts increase when the ginger extracts concentration increased, so the diameters of inhibition zones was also increased, this may be due to the amount of the dissolved active ingredients of the extracts will be more abundant as the concentration increased.It is clear from this study that the concentrations of alcoholic extract had more antibacterial activity by displaying higher mean values of inhibition zones for A.a than the concentrations of aqueous extract, the comparisons between these two extracts showed almost highly significant differences between the same concentrations of the alcoholic and aqueous extracts that could be because of the amount of active component in the alcoholic ginger extract 10-gingerol (hence, its concentration was higher in alcoholic extract than the aqueous in this study) which is the primary active ginger phenol against periodontal anaerobic pathogenic bacteria, (8) .Indeed the polarity of the solvent (ethanol alcohol) which has great ability to dissolve ginger biologically active constituents (26) .*Detection of MIC and MBC of aqueous and alcoholic ginger extracts against A.a : The MIC required to inhibit A.a growth in broth media was 50% (0.5 g/ml) concentration of alcoholic ginger extract, while, MIC of aqueous ginger extract was 80% (0.8 g/ml) concentration and this represented the bacteriostatic activity of the extracts against A.a .The 0.2% Chlorohexidine gluconate applied in this experiment as a positive control displayed also bacteriostatic effect against A.a .The MBC of alcoholic ginger extract that kills A.a was 80%(0.8g/ml) concentration while, the MBC of aqueous ginger extract was 100% (1g/ml) concentration which represented the bactericidal effect of the extracts against A.a .Hence, gingerol and shogaol, the lipid-soluble phenols that are mainly isolated from the ginger root (27) , have various routes of action which means that these compounds not only attack cell walls and cell membranes i.e., affecting their permeability and release of intracellular constituents (e.g.ribose sodium glutamate), but they also interfere with (electron transport, membrane functions, nutrient uptake, protein and nucleic acid synthesis and enzyme activity).Thus, these compounds might have numerous invasive targets that lead to the inhibition of pathogenic bacteria (27) .There were no other studies to compare the results with.* Detection of the active constituents of ginger extracts and chemical elements: The HPLC analysis results of alcoholic and aqueous ginger extracts revealed that alcoholic extract had higher content of 6-shogoal and 10-gingerol that are the main active ginger phenol with powerful antibacterial effect against periodontal anaerobic bacteria, (8) .Since, several studies showed that the antimicrobial activity of ginger mainly caused by the presence of phenolic compounds (shogaol,gingerol) (24) ,while aqueous extract contained higher content of 6gingerol and 8-gingerol than alcoholic extract that are of less effect against periodontal pathogens (8) .A study done by Park et al.,2007 (8) ,which signified that the ethanol and n-hexane ginger extracts exhibited antibacterial effect against three of the anaerobic Gramnegative bacteria, Porphyromonas endodontalis, Porphyromonas gingivalis and Prevotella intermedia, that are responsible of periodontal diseases, they found that [10] -gingerol and [12] -gingerol effectively inhibited the growth pattern of these periodontal pathogens.*Detection of ginger extracts chemical elements: Similar elements were detected in both extracts but varied in their concentrations.Since alcoholic ginger extract exhibited higher concentrations of (K), (Mg), (Mn), (Fe) and (Zn) than the aqueous extract whereas, the other elements are almost equal and (Ca) found only in alcoholic extract, this could be due to the polarity of ethanol alcohol that has the capacity to dissolve ginger biological active component (28) .Since, (K), the third most copious mineral in human body, and it is a powerful element in general health improvement ; from other side, (Mg) revealed its significance in many biological and detoxification procedures, this reflect the powerful antioxidant effect of ginger.
Table ( 1
): The statistical analysis of A.a inhibition zones by different concentrations of aqueous and alcoholic ginger extracts, CHX and D.W.
Table ( 2
): Comparisons of mean values of A.a inhibition zones between each pair of different concentrations of aqueous and alcoholic ginger extracts by LSD test
Table ( 3
): Comparisons of mean values of A.a inhibition zones between each concentration of aqueous and alcoholic ginger extracts with CHX and D.W. by LSD test
Table ( 4
): Statistical analysis and comparisons between mean values of A.a inhibition zones for the same concentrations of aqueous and alcoholic ginger extracts
Table ( 5
): Concentrations of each constituent of aqueous and alcoholic ginger extracts Rodenburg JP, van Winkilhoff AJ, Winkel EG,Goene RJ,Abbas F, de Graff J. .Occurence of Bacteroids gingivalis, Table (6): Chemical elements concentrations in alcoholic and aqueous ginger extracts | 2019-04-10T13:11:51.797Z | 2017-04-16T00:00:00.000 | {
"year": 2023,
"sha1": "18831357e6252360775d37dd63c23a418987262e",
"oa_license": "CCBY",
"oa_url": "https://tjds.tu.edu.iq/index.php/tjds/article/download/189/121",
"oa_status": "HYBRID",
"pdf_src": "Anansi",
"pdf_hash": "fc34189e47f990ee9558784c908e889c3c925237",
"s2fieldsofstudy": [
"Medicine",
"Biology",
"Environmental Science"
],
"extfieldsofstudy": [
"Chemistry"
]
} |
664840 | pes2o/s2orc | v3-fos-license | Risk-taking behavior in juvenile myoclonic epilepsy
Objective Patients with juvenile myoclonic epilepsy (JME) often present with risk-taking behavior, suggestive of frontal lobe dysfunction. Recent studies confirm functional and microstructural changes within the frontal lobes in JME. This study aimed at characterizing decision-making behavior in JME and its neuronal correlates using functional magnetic resonance imaging (fMRI). Methods We investigated impulsivity in 21 JME patients and 11 controls using the Iowa Gambling Task (IGT), which measures decision making under ambiguity. Performance on the IGT was correlated with activation patterns during an fMRI working memory task. Results Both patients and controls learned throughout the task. Post hoc analysis revealed a greater proportion of patients with seizures than seizure-free patients having difficulties in advantageous decision making, but no difference in performance between seizure-free patients and controls. Functional imaging of working memory networks showed that overall poor IGT performance was associated with an increased activation in the dorsolateral prefrontal cortex (DLPFC) in JME patients. Impaired learning during the task and ongoing seizures were associated with bilateral medial prefrontal cortex (PFC) and presupplementary motor area, right superior frontal gyrus, and left DLPFC activation. Significance Our study provides evidence that patients with JME and ongoing seizures learn significantly less from previous experience. Interictal dysfunction within “normal” working memory networks, specifically, within the DLPFC and medial PFC structures, may affect their ability to learn.
Juvenile myoclonic epilepsy (JME) is the most common idiopathic epilepsy syndrome (Berg et al., 2010). Patients have been described as "immature" with impaired experience-related learning (Janz, 1985;De Araujo Filho & Yacubian, 2013), suggestive of frontal lobe dysfunction. This is corroborated by reports on impaired working memory (WM) and executive functions (Wandschneider et al., 2012) and functional and microstructural changes within the medial and dorsolateral prefrontal cortex (DLPFC) O'Muircheartaigh et al., 2011).
The Iowa Gambling Task (IGT) was developed for patients with lesions within the prefrontal cortex (PFC) presenting with impulsive decision making irrespective of possible future consequences (Bechara et al., 1994). Previous studies investigated the relationship of decision making with performance in other cognitive domains of the PFC, such as executive functions (EFs) and working memory (WM), arguing for a dissociation of decision-making abilities from other cognitive prefrontal lobe functions (Toplak et al., 2010). However, there is also evidence for an asymmetrical relationship of decision making and WM in healthy subjects (Suhr & Hammers, 2010) and in patients with frontal lobe lesions (Bechara et al., 1998;Manes et al., 2002), but also in nonlesional frontal lobe disorders (Duarte et al., 2012). Working memory performance may not depend on intact decision-making abilities, but WM dysfunction can mediate impaired decision making. Lesions or dysfunction of the anterior ventromedial prefrontal cortex (VMPFC) and orbitofrontal cortex seem to be associated with impaired decision making only, whereas pathology in the posterior (VMPFC) is associated with both WM and decision making impairment and the DLPFC with WM impairment only (Bechara et al., 1998;Brand et al., 2007).
These observations are corroborated by a recent functional MRI (fMRI) study investigating neuronal correlates of decision making during the IGT in healthy controls (Li et al., 2010).
Cortical activation patterns underlying decision making include the following: (1) the DLPFC for WM, in order to provide online knowledge during the decision-making process; (2) insula and posterior cingulate for emotional feedback; (3) medial orbitofrontal cortex to interlink the previous two processes; and (4) ventral striatum and supplementary motor area (SMA) to carry out decisions (Li et al., 2010). Impulsive decision making in the IGT has been reported recently in JME (Zamarian et al., 2013), and was specifically evident if seizures were poorly controlled.
We aimed to investigate the underlying mechanisms of decision-making behavior in JME, in particular to elucidate which part of the WM network, as defined by fMRI, is associated with IGT performance (Li et al., 2010).
Methods
This study was approved by the UCL Hospital Research Ethics Committee. Written informed consent was obtained from all study participants.
Twenty-one JME patients and 11 healthy controls (HC), matched for gender (JME: 12 female, HC: 6; Pearson chisquare p = 0.888), age (JME: median 33.5 [range: 22-64] years; HC: 28 [24-32]; Mann-Whitney U = 70.500, N 1 = 21, N 2 = 11, p = 0.074), and IQ (JME: median 109 [range: ; ; Mann-Whitney U = 51.000, N 1 = 20, N 2 = 7, p = 0.314) were assessed with the IGT. With regard to educational exposure, 10 patients were postgraduate, two obtained A-levels, six General Certificate of Secondary Education (GCSE), and in three patients, information about educational exposure could not be obtained. Performance was correlated with activations during an fMRI WM task (n-back). Imaging data sets from two controls and two patients were excluded because of lack of activations during fMRI. In one patient and two HCs, we could not record the dot back performance due to joystick failure.
Patients had a typical history of JME and had been diagnosed by experienced epileptologists according to the International League Against Epilepsy (ILAE) Classification (Berg et al., 2010). All clinical MRI scans were normal on qualitative analysis. Fourteen patients had been seizure-free for at least 12 months (see Table S1 for further clinical details). No patient experienced a seizure during the assessment.
Standardized frontal lobe tests
A battery of standardized tests assessing frontal lobe functions was administered to all participants. It included the following: 1 Letter and category fluency. The letters "F," "A," and "S" were presented for letter fluency, and the categories "animals," "fruits," and "vegetables" for categorical fluency. The number of words produced for each letter or category within one minute was totaled to provide an average score for letters and categories. 2 The Digit Span subtest of the Wechsler Adult Intelligence Scale (WAIS-III; Wechsler, 1981) for working memory; 3 The Trail Making Test (Reitan & Wolfson, 1985) for (a) Psychomotor speed: Total time to complete part A. (b) Mental flexibility: A score was derived by subtracting the time to complete for part A from the time to complete part B.
Iowa Gambling Task
In a computerized IGT version (Bechara, 2007), subjects chose from four decks of cards (A-D) with a total of 100 cards. All decks awarded virtual monetary gains and losses. The aim was to gain as much money as possible. Decks A and B were associated with higher gains and penalties, indicative of disadvantageous choices. Decks C and D were associated with lower gains and penalties (advantageous choices).
Before commencing the IGT, participants were instructed that their aim was to gain as much money as possible and to avoid losing money. However, they were not informed about the IGT's contingencies and had to learn them by feedback from their card choices, that is, their monetary gains and losses.
Scores of IGT performance 1 The total number of card selections from the disadvantageous decks (A and B) and the total number of selections from the advantageous decks (C and D) were counted. Subsequently, an overall net score was derived by subtracting the total number of disadvantageous from (1) was derived for five consecutive blocks of 20 cards. The IGT mainly measures decision making under ambiguity, as especially initial choices are made without explicit knowledge of their consequences and decisions are mainly guided by implicit information. However, it has been shown that healthy subjects optimize their performance during the task by learning the properties of each deck and subsequently change their strategy toward choosing from the advantageous decks only (Brand et al., 2007). By obtaining net scores for five consecutive blocks we aimed at characterizing changes in IGT performance and strategy over time (Brand et al., 2007;Li et al., 2010). 3 In addition, for each individual, changes in task performance indicative of experience-related learning throughout the task were analyzed. Similarly to Li et al. (2010), we subtracted the net score of the first block from the net score of the fifth block. A score above zero was indicative of a positive slope in the individual learning curve. A score of zero or below was indicative of a negative or unchanged slope in the learning curve. Accordingly, participants were divided into learners and nonlearners.
Statistical analysis
Data were analyzed using SPSS Statistics Version 17.0 (SPSS Inc., Chicago, IL, U.S.A.). Group differences were analyzed with statistical tests for nonparametric data. Chisquare tests were applied to categorical data.
Functional MRI paradigm
MRI data were acquired on a GE Excite HDx 3T scanner (GE Medical Systems, Milwaukee, WI, U.S.A) with a multichannel head coil as described previously . The WM paradigm consisted of dots presented randomly in four possible locations on a screen (Kumari et al., 2009). Participants monitor the locations of dots at a given delay of the original occurrence (0-, 1-, or 2-back) and respond by moving a joystick corresponding to the location of the current or previously presented dot. FMRI analysis was performed using SPM-5 (www.fil.ion.ucl.ac.uk/spm). For each subject, contrast images were created to explore the effect of "2-back minus 0-back" comparing the task with the highest WM load ("2-back") against the control task; hence by controlling for motor response and visual attention, only cortical activation due to the WM load was revealed.
Correlation analyses
To test for correlations between IGT performance and network activations during the fMRI task, a whole brain multiple regression analysis was performed and overall net scores of IGT performance were entered as a covariate.
We explored the effect of seizures on IGT performance at group level by creating a full factorial analysis of covariance with seizures (seizure-free/seizures) as a factor. Within the above analysis, a global conjunction analysis was performed to explore the effect of overall IGT performance between groups (HC, seizure-free patients, and patients with ongoing seizures).
We further explored the effect of seizures on learning throughout the IGT in a subgroup analysis of patients by creating a full factorial analysis with seizures (seizure-free/ seizures) and learning (learner/nonlearner) as factors.
The level for significance was p < 0.005 uncorrected.
Behavioral data
Performance on standardized frontal lobe test battery The results are detailed in Table 1. There were no significant group differences in performance on the frontal lobe test battery. Because learning to shift strategy toward more advantageous card choices may rely on executive functions (Brand et al., 2007), we conducted a subgroup analysis of frontal lobe measures in the patient groups of nonlearners versus learners, which showed no significant differences (see Table S2).
Analyzing changes in IGT performance throughout the task in each participant showed that the majority of patients and controls are learners (JME: 15 learner/6 nonlearner; HC: 8/3; Pearson chi-square p = 0.938). In a post hoc subgroup analysis, the majority of patients who were seizurefree improved throughout the task (12 learner/2 nonlearner), whereas only half of the patients still experiencing seizures changed their strategy toward more advantageous choices over time (3 learner/4 nonlearner; Pearson chi-square p = 0.040; Fig. 1).
In a further subgroup analysis, IGT performance expressed as the overall net score did not correlate with the performance on the fMRI WM task in either group: seizurefree patients r = 0.456, p = 0.117; patients with ongoing seizures r = À0.512, p = 0.299; healthy controls r = 0.481, p = 0.114.
In patients, poor IGT performance was associated with bilateral PFC activation. In controls, poor performance was associated with reduced deactivation in parts of the default mode network, that is, in the bilateral precuneus, cingulate gyrus, and right medial PFC (Fig. 2B). In a conjunction analysis of patients with ongoing seizures compared to both healthy controls and seizure-free patients, patients with seizures showed increased activity in the left DLPFC in association with poor IGT performance (Fig. 2C).
Using WM performance scores as an additional covariate of no interest in the fMRI regression analysis, similar neuronal association patterns were identified (Fig. 1, Supporting Information).
Within the group of patients still experiencing seizures, nonlearners showed greater activations in bilateral medial PFC and pre-SMA, as well as the left DLPFC and right superior frontal gyrus (SFG) when compared to learners (Fig. 3).
Discussion
We report network abnormalities associated with decision-making behaviors in patients with JME. We identified impaired IGT performance only in the group of JME patients still having seizures, which is similar to a previous study reporting continuing impairment in experiencerelated learning in patients with ongoing seizures compared to those without (Zamarian et al., 2013). Because both patient subgroups and controls performed equally well in the WM fMRI task, it is possible to relate cortical activation patterns with IGT performance. In JME, overall poor IGT performance was associated with bilateral DLPFC activation. Ongoing seizures and overall poor IGT-performance were associated with increased left DLPFC activation.
Overall IGT performance
Patients with ongoing seizures and seizure-free patients were comparable in their performance during the WM fMRI task. Patients and controls did not differ significantly in the behavioral measures of cognitive functions. This may be in keeping with reports that decision making is relatively independent from other frontal lobe functions, such as WM (Toplak et al., 2010). However, in a bigger cohort, Zamarian et al. (2013) could show that patients with JME performed significantly worse on overall IGT net scores. Similarly, when comparing seizure-free patients to those with ongoing seizures in our cohort, JME patients with ongoing seizures showed severe impairment of their IGT performance (Fig. 1).
Our fMRI findings suggest that crucial hubs within physiologic WM networks are implicated in impaired IGT task performance in patients. Studies in healthy subjects (Suhr & Hammers, 2010) and patients with nonlesional frontal lobe disorders (Duarte et al., 2012) describe an asymmetrical relationship between WM and IGT performance, in the sense that WM dysfunction may moderate decision-making impairment. The DLPFC is recruited during continuous updating of WM and has been identified as part of the neuronal circuitry underlying decision making both in fMRI studies in healthy individuals (Li et al., 2010) and lesional studies (Manes et al., 2002). In our cohort, left DLPFC activation was associated with impaired decision making only in patients with ongoing seizures.
Previous behavioral studies reported subtle working memory impairment in JME patients (Wandschneider et al., 2012). Causative brain pathology for frontal lobe dysfunction is not apparent on clinical MRI scans. However, there is evidence for microstructural abnormalities (O'Muircheartaigh et al., 2012), suggesting a thalamo-frontocortical network dysfunction and increased functional and A B C Figure 2. Association of working memory network activation with overall IGT performance. (A) Both groups, JME patients and controls, activate bilateral frontal and parietal working memory networks during the working memory task ("2-back minus 0-back"-contrast). (B) Overall net scores of IGT performance (C + D À [A + B]) were entered as a covariate. Negative correlation of IGTperformance with the "2-back minus 0-back" contrast across the whole JME group revealed bilateral prefrontal cortex activation. (C) Conjunction analysis of JME patients with seizures above seizure-free patients and controls, and negative correlation with overall IGT performance revealed hyperactivity in the left DLPFC. The contrast was masked by working memory network activations of healthy controls (p < 0.05 unc.) (CTR, controls; IGT, Iowa Gambling Task; JME, juvenile myoclonic epilepsy; sz, seizures; szfree, seizure-free). Epilepsia ILAE Epilepsia, 0(0):1-8, 2013 doi: 10.1111/epi.12413 structural connectivity between cognitive and motor networks , which correlate with disease activity. These network changes are present even in high functioning JME patients with no WM impairment on behavioral measures O'Muircheartaigh et al., 2011). Therefore, DLPFC hyperactivation may be the neuroanatomic correlate of a seizurerelated network dysfunction, implicating WM networks, and through this moderating impaired decision making. DLPFC hyperactivation has been seen in other patient groups with low WM capacity, such as patients with schizophrenia (Manoach, 2003), and may also reflect a necessary compensatory mechanism of cortical activation to achieve the WM performance required for the task. In controls, poor IGT performance was associated with decreased deactivation in the default mode (DM) during the active condition. DM areas are independent of task active areas, and suspension of baseline DM brain function allows reallocating neuronal resources to task-relevant brain areas. Inability to deactivate the DM during the active phase of tasks has been generally related to poor performance during the active task condition (Raichle et al., 2001) and may thus be the explanation for its relation to poor IGT performance in the control cohort. Similarly, in a cohort of patient with temporal lobe epilepsy and HC, a failure to deactivate parts of the DM network has been related to poor performance during the same WM fMRI task employed in our study . More specifically, for decision making, decreased deactivation in areas of the DM may imply that the process of learning probabilities of future events is impaired. This is corroborated by a recent study investigating decision making under uncertainty in healthy subjects (D'Acremont et al., 2013); the authors employed a gambling task to investigate how subjects learn probabilities of future events irrespective of their (immediate) value. The stage of learning during the task was associated with progressive deactivation of DM, including the angular gyrus, medial PFC, and posterior cingulate cortex.
Learning during the IGT
The effects of ongoing seizures and impaired learning throughout the task were associated with bilateral medial prefrontal cortex (PFC) and pre-SMA, right SFG, and left DLPFC activation. Brand et al. (2007) argued that the IGT comprises two phases: during the first trials, subjects learn to make choices from implicit information; during the second phase, subjects have acquired some explicit knowledge about the IGT contingencies, hence choices are made under risk, that is, with the knowledge of future consequences. The switch between the phases probably happens gradually, and the latter phase is more dependent on the integrity of the DLPFC and frontostriatal loops, as making successful choices under risk requires executive functions, such as set shifting (Brand et al., 2007). Similarly, in an event-related fMRI study, Lawrence et al. (2009) identified prefrontal subregions contributing to different stages of the IGT. In healthy individuals, activation in posterior frontal regions and the pre-SMA were positively correlated with general task performance. In terms of neuronal correlates of learning, the ventrolateral PFC and the pre-SMA were recruited early during the task, when subjects were acquiring knowledge about the contingencies. With more explicit awareness of task strategy and reduced uncertainty, a corresponding decrease in pre-SMA activation was observed in good performers. Accordingly, other studies identified the pre-SMA as an area involved in guidance of behavior in uncertainty situations and decision making under uncertainty (Ridderinkhof et al., 2004;Hartstra et al., 2010). In our cohort, pre-SMA activation is seen in patients with ongoing seizures who do not learn to shift their IGT strategy. This may imply that nonlearners do not gain explicit awareness about task strategies and continue to make choices under ambiguity; therefore, they do not show a decrease of pre-SMA activation over time.
Limitations
In terms of the possible impact of AED treatment on IGT performance, 8 patients were treated with monotherapy, 11 with combination therapy (10 patients with two AEDs, one patient with three), and one patient did not receive any AED. The majority of patients were on valproic acid (VPA; 13 patients), followed by levetiracetam (LEV; 9) and lamotrigine (LTG; 7). Generally, AEDs may affect cognitive abilities, and cognitive side effects vary among different agents and are more prominent with combination therapy Epilepsia ILAE (Hermann et al., 2010). Processing speed and attention are particularly vulnerable domains, however, these did not appear to be affected in our sample when comparing patients to healthy controls. LTG and LEV have been both associated with rather little negative effect on cognition (Hermann et al., 2010). For LTG, beneficial effects on frontal lobe functions have been described due to its mood-stabilizing properties (Pavuluri et al., 2010). One study in JME patients (Roebling et al., 2009) reported a negative effect of VPA treatment on a verbal memory test when compared to healthy controls. A previous fMRI study in our cohort, however, reported a dose-dependent normalizing effect of VPA on abnormal working memory network activation with motor-cortex coactivation . Because of the small subgroup sample sizes, specific AED effects on cognition could not be explored in more detail. As there was no difference in behavioral measures between groups, and the majority of patients was treated with AEDs, and the distribution of monotherapy versus polytherapy was equal in learners and nonlearners (see Table S1), we suggest that poor IGT performance is at least in part independent of AED treatment.
Conclusions
Our results add further evidence that ongoing seizure activity affects daily life decision making, and emphasize the importance to achieve full seizure control. Learning and IGT performance were impaired, especially in JME patients with ongoing seizures. Learning and nonlearning patients did not differ in behavioral frontal lobe measures, but did differ in IGT performance and neuronal network correlates within the DLPFC and thalamo-frontocortical loops. These structures are important for successful implementation of executive functions during decision making under risk (Brand et al., 2007), and the ventromedial PFC is crucial for overall successful task performance and learning to win (Lawrence et al., 2009). Our findings illustrate that neuropsychological deficits may be too subtle to detect on conventional behavioral measures, and underscore the importance of imaging studies and more challenging frontal lobe tasks, such as the IGT, which integrates decision-making behavior and executive functions, to detect differences in cognitive networks.
Supporting Information
Additional Supporting Information may be found in the online version of this article: Figure S1. Association of working memory networks activation with IGT performance. Table S1. Clinical details. Table S2. Neuropsychological test results in patients (learners vs. nonlearners). | 2016-04-27T00:50:03.749Z | 2013-10-18T00:00:00.000 | {
"year": 2013,
"sha1": "4be32361eca20e5f28774985737e5df57247385b",
"oa_license": "CCBY",
"oa_url": "https://onlinelibrary.wiley.com/doi/pdfdirect/10.1111/epi.12413",
"oa_status": "HYBRID",
"pdf_src": "PubMedCentral",
"pdf_hash": "71011326f9565e8d3d6ac22e9a5e52d031a824f2",
"s2fieldsofstudy": [
"Psychology"
],
"extfieldsofstudy": [
"Psychology",
"Medicine"
]
} |
9406775 | pes2o/s2orc | v3-fos-license | Mobility and falls in people with Huntington's disease
Objective: The aim of this study was to estimate the frequency of falls in people with Huntington's disease (HD) and make a preliminary assessment of tools appropriate for assessing the risk of falling. Design: Observational study. Setting: Hospital clinic. Subjects: 24 people with HD. Main measures: Balance was assessed using the Berg Balance Scale (BBS) and Timed “Up & Go” (TUG) test. Walking speed over 10 m was recorded. Long-term monitoring of walking activity was undertaken. Unified Huntington Disease Rating Scale (UHDRS) motor, Functional Assessment Scale (FAS), Independence Scale (IS) and Total Functional Capacity (TFC) scores were obtained as well as data about falls and stumbles. Differences between “recurrent fallers” (≥2 falls/year) and “non-fallers” (≤1 fall/year) for the range of outcome measures were investigated and probabilities calculated. Results: Mean (SD) age (years) of people with HD (n=24) tested was 56.6 (11.7) and BMI (kg/m2) 24.7 (5.5). Median (range) UHDRS motor scores were 48 (28-80). Ten (41.6%) patients reported ≤1 fall and 14 (58.3%) ≥2 falls in the previous 12 months. Recurrent fallers walked less (p<0.01) and slower than non-fallers. Their balance (BBS) (p<0.01) was worse and TUG scores were higher (p<0.01). People with HD had increased risk of falls if TUG scores were ≥14 s or BBS scores ≤40. Conclusion: A high proportion of HD patients have recurrent falls, and the BBS and TUG appear to be useful in falls risk assessment.
Huntington disease (HD) results in progressive loss of functional abilities. Despite reports highlighting falls-related injuries and associated balance problems, [1][2][3] there is no systematic research on falling and its risk factors in HD in the community. Formal assessment of falls is not part of the standard validated assessment protocols in widespread use.
The standard assessment tool for HD is the Unified Huntington Disease Rating Scale (UHDRS). 4 The UHDRS comprises semiquantitative clinical scales to assess motor function, cognitive function, behavioural abnormalities and functional capacity. UHDRS motor scores have been found to be sensitive to change over time 5 with higher scores indicating greater impairment for the motor assessment. The UHDRS functional scale comprises three components (functional assessment scale (FAS), the independence scale (IS), and the Total Functional Capacity (TFC)). Although UHDRS scores can be considered indicators of impairment and functional loss, their value in predicting likelihood of falls is unknown.
The Activities-Specific Balance Confidence Scale (ABC) 6 is a self-administered questionnaire used to assess fear of falling in older people and has discriminant ability to identify fallers from nonfallers and those who avoid activity due to fear of falling. 7 Additionally clinical measurements of mobility and balance, such as the Timed Up and Go Test (TUG) 8 and the Berg Balance Scale (BBS), 9 are known to relate to falls' risk in other conditions. 10 The TUG and the BBS have been reported to be useful in single cases of people with HD 1 but have not been formally validated for this use. Although anecdote suggests that people with HD are at risk of falls and do fall, falls' rate does not appear to have been investigated in depth. We aim here to assess fall prevalence and consider the above-mentioned potential measures of increased fall risk in the HD population.
METHODS Subjects
Patients with manifest HD attending the Cardiff HD clinic between December 2006 and November 2007 were offered participation in the study in accordance with local research ethics committee permissions (06/WSE03/70). All subjects provided informed consent prior to participation. Inclusion criteria were a genetically confirmed diagnosis of HD, age over 18, capacity for informed consent, no major concurrent psychiatric illness, presentation with motor signs and a score of 4 on the UHDRS motor diagnostic confidence rating.
Demographic data including age (years), and body mass index (BMI) (kg/m 2 ) and current medication were obtained. Scores on the motor section of the Unified Huntington Disease Rating Scale (UHDRS motor), the Total Functional Capacity (TFC), the Functional Assessment Scale (FAS) and the Independence Scale were recorded.
Balance and walking
Patients were assessed using the Berg Balance Scale (BBS) (maximum score 56 indicating good balance) and the ''Timed Up & Go'' (TUG) test. The modified Activities-Specific Balance Confidence (ABC-UK) Scale provided an estimate of balance confidence. Self-selected walking speed over 10 m was recorded in an indoor corridor. Stride length was also calculated. Long-term walking activity was recorded using using the Step Watch Step Activity Monitor (SAM) (Cymatech, Seattle, Washington) worn for seven consecutive 24 h periods. This device and protocol has been used previously in patients with chronic neurological disorders. 11 Indices extracted were mean daily 24 h step count and highest average step count sustained over any continuous 60 min period (sustained activity).
Falls and stumbles
Data were collected for falls and stumbles in the previous 12 months using a questionnaire. 12 The questionnaire was completed by the patient with assistance from their main carer. Participants were categorised as recurrent fallers if they reported >2 falls over the previous 12 months. If patients reported (1 fall, they were not considered to be recurrent fallers. 13
Data analysis
Descriptive statistics were used to quantify falls and scores on outcome measures. Independent t tests, Mann-Whitney U tests and x 2 tests were used where relevant to detect differences on outcome measures according to whether a person was a recurrent faller or non-faller. Logistic regression was used to plot the probability of falling for the each of the separate continuous falls risk outcome measures (ie, the BBS and the TUG). The alpha level was set at 0.05. The Statistical Package for the Social Sciences (SPSS) Version 12 (SPSS, Chicago) was used for all data analysis and Minitab Version 14 for graphing.
Sample size
A priori sample size calculations were conducted using WINPEPI software: 14 20 subjects were sufficient to detect a standardised difference in outcome measures of 1.25 between recurrent fallers and non-fallers (independent t test).
RESULTS
Twenty-four people with HD were recruited. Demographic data are presented in table 1. The most common medications prescribed for this group were antichoreic medication (41.6% of patients), benzodiazepines (16.6% of patients) and those with an antidepressant main action (50% of patients). Five (20.8%) people reported no falls in the previous 12 months, and five (20.8%) reported only one fall: thus 10 were classified ''non-fallers.'' Fourteen (58.3%) reported falling twice or more in the previous 12 months (classified as recurrent fallers). Of the recurrent fallers, 61.6% were also stumbling frequently, while in the non-fallers, 50% reported stumbling on a regular basis. Table 1 shows the outcome scores for the whole group and non-fallers and recurrent fallers subgroups.
There were no differences in age, BMI or gender between those who were classified as recurrent fallers and those classified as non-fallers. There were no clear differences in frequency of prescription of the anti-choreic medication between fallers (6/ 14) and non-fallers (4/10) or bensodiasepines in the fallers (3/14) compared with the non-fallers (1/10) although the numbers were too small to analyse these data statistically.
Logistic regression identified the continuous variables of TUG (s) and BBS score as significant predictors of falls. The TUG and BBS were not entered into a combined logistic regression model due to the high correlations between the measures. Predicted probabilities for falling plotted against the continuous variables identified a classification cut-off for increased risk of falls if TUG scores were >14 s or BBS scores (40. Figure 1 shows the plots of the probabilities obtained for falls for the TUG and BBS scores against the continuous TUG and BBS scores.
DISCUSSION
This study confirms that people with manifest HD do fall regularly. Only 20.8% of people did not report any falls in the previous 12 months. Fallers took fewer steps and walked more slowly than non-fallers, and their balance (BBS and TUG scores) and balance confidence (ABC-UK) were worse. Scores for nonfallers (n = 10) and recurrent fallers (n = 14) were different for the IS and FAS but not for the other UHDRS-related clinical scores (including UHDRS motor). This study was not specifically powered to detect differences in the UHDRS motor scores.
The TUG and BBS scores both predicted probability of falling and may therefore be considered for use in people with HD. The susceptibility of the TUG to cognitive impairment may however limit its applicability in HD; this test needs a clear understanding of the instructions as well as an interaction between patient, assessor and environmental setting. 15 The BBS, developed primarily to evaluate balance in elderly people and people with stroke, may be more robust in those with cognitive impairment.
Recurrent fallers were less active than non-fallers (step count and sustained step counts) and less independent. Both step count and sustained step counts were statistically significantly different between fallers and non-fallers. While a daily step count recording may mistakenly register chorea as steps, therefore inflating the step count of subjects with worse chorea, the sustained walking step counts should be more robust. It is unclear whether the reduction in physical activity is related more to perceived balance and confidence or is a direct result of the motor impairments seen in HD.
The importance of objective recordings is also an issue in the assessment of falls rate.
The extent to which carer bias may have influenced falls' reporting is unclear and emphasises the need for alternative methods of evaluation. Self-reported questionnaires are subject to bias, and failure to recall an injury is high using self report; calendar recording of falls events would be more accurate in recording falls and injuries. 16 17 While falls and stumbles do occur in healthy subjects, 12 it would be unusual for healthy people to report recurrent falls. We attempted to minimise report failure by careful involvement of carers in questionnaire completion, but one might anticipate potential under-reporting. Such underreporting limits the usefulness of the dichotomised outcome (1 or >2 falls except that the latter probably indicates a group falling more often. Reliance on patient/carer recall or recording of falls could be objectively augmented by techniques such as automated event recording 18 over prolonged periods (weeks). Falls, injury and loss of independent ambulation are often factors that precipitate admission to nursing homes, 3 and falls, risk of falls and basic falls management advice 19 should also be considered a priority in people with HD. Falling in HD is likely to be multifactorial in origin, 20 and factors such as home environment, medication and cognitive status also require further investigation. Reduced step count and physical activity as well as reduced scores on functional capacity and independence scales and reduced balance confidence may be indicative of a person at risk of falls. The BBS as well as the TUG also have potential as falls risk outcome measures in HD. | 2015-09-18T23:22:04.000Z | 2008-12-16T00:00:00.000 | {
"year": 2009,
"sha1": "4b5702e9c35bd936678076ff23ffda2cc17ba7d9",
"oa_license": "CCBYNC",
"oa_url": "https://jnnp.bmj.com/content/80/1/88.full.pdf",
"oa_status": "HYBRID",
"pdf_src": "BMJ",
"pdf_hash": "55072598c90999aee1c720801054dfee1b56a792",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Psychology",
"Medicine"
]
} |
248444969 | pes2o/s2orc | v3-fos-license | Un-cisgendering Womanism to Scrimmage with Misogyny: an Analysis of Cassandra (1999) by Violet Barungi
: Published in 1999, Cassandra is a remarquable novel through which Violet Barungi pinpoints various questions related to feminism and femalism. To delve into the issue of feminism and glass the idea on femocracy , the perspective of femism and kink has been developed and elaborated. Indeed, in a critical perspective, the paper has spotlighted the degree of the Ugandan women‟s failure in the struggle against patriarchy in their society. An analysis of probate pertinency has shown up the resolution of the female character to overthrow the social and cultural female status cushioned by so much domesticated vagrancies. However, confronted with pugnant realities, the female actant steps back to fall into the trap of sex and sexuality. A path to freedom is then cut short, a dream shattered on the erotic wall of venial pleasures. Emancipation is then wedged apart into the glooming splinters of masculinity and patriarchy. a factual a female aspiration” (Re through discursive activities. The system of manipulation and frontal opposition between men and Cassandra in the Lotus International Company bestows the narrative with a compositional field that amplifies itself in a network of paradoxical interactions. the reversals, tensions and surprises through the
INTRODUCTION
"One is not born, but rather becomes a woman" (Beauvoir, 1989: 17-18). These words from Simone de Beauvoir ring high and loud to tell much about female engagement in the struggle for emancipation. Indeed, the idea of liberating women from the double yoke of social feminicide and traditions is evoked and elaborated by many African intellectuals among whom Violet Barungi. Born in the District of Ibadan in the West of Uganda, Barungi is a woman who sharpens her struggle for women"s rights. She shoulders the vision to sweep through the female skirmish, dredging out the longstanding and sharp longing for women to walk across the borderline of social servitude and genderism. In this perspective, Barungi, in Cassandra, works out the issue of Ugandan women trapped between the hammer of traditions and the anvil of modernism and modernity. She dedicates herself to a self-writing through the perspective of dominomorphism defined in substance by C. Grignon and J.-Passeron as being the different social criteria that favour and put forward the culture of dominancy (1989:32). Indeed, Barungi problematizes the topical matter of women-being and woman-doing to highlight the critical emphasis intentionally voiced and enacted by a young and rebellious girl. The question on women"s struggle for emancipation, which has become "a major national and international significance in many African countries across Africa" (Mhabasonda 2002: 99), stirs to actions some women of Kampala who, in an intellectual-based awareness, vent their voices through the entrails of a female collective consciousness. Thus being, in this work, we target to raise the issue of probated pertinacity toward a patriarchal tradition. The whatnots of love as a hindering obstacle that rules away women from the horizon of emancipation is another point that will be put hand in glove with a medium stand on African Feminocentrism and Stewanism.
I. A Probated Pertinacity toward the Female Status Quo
Wafting from the Grec "Kassandre", the name Cassandra means the one who helps men. (Heems, 2019: 32). Devoted to gods and traditions, Kassandre was anointed with powers of prediction and would be consulted for any matter concerning the Kingdom of her origins. (Heems, 2019: 32). But, for having betrayed, she was punished by the god Apollo who deprived her of her powers and left her with no out of influency and with no power. Being a byword of the Grec personage, Cassandra, in the eponymous novel, follows a similar way of the cross. As a young girl, she loathes in a final way the all-powerful masculine complex of superiority Ugandan women suffer from. She builds cross-currents of opposing thoughts with the vice-grip of masculinity on women and principles herself for the sake of a grandeur of life that will profit her likes in her country.
Recruited in a Press House, Cassandra bears down on her stamina to engage a combat for a horizontal man-towoman relation. The conditions of women based on suffering and oppression are ably brought about by men"s insatiable desire for domination. In Cassandra"s eyes, the woman"s social position is, in Uganda in general and in Lotus International Company in particular, husky and lousy. Thus being, it is to be thought how to veto men"s power that must, according to her, stop going ad infinitum. She first targets women whom she believes, do not honour their femininity. Juliet, her foregoer in the company, is the first to be victim of Cassandra"s conceived dislike. She accuses her of not being enough decent and ethical to stand up to men. Being too mendacious and a bit fickle, Juliet swaps her bodily services for professional advantages. Cassandra"s hope to see her colleague in a more honourable state is dashed to ground and she is outraged and lets it know in the following: "Juliet belongs to the class of females whose ambition does not go beyond being the boss" pet, mistakenly equating a pat on the head with a badge of success" (…). She is cheap and vulgar and thinks everybody else"s like her" (Barungi, 1999: 11-12). Cassandra is formal and definitive; women should not be social ladders for men to atop themselves, profiting by the sexual facilities they are proposed as kickbacks to guaranty women"s promotion.
In the same logic of thoughts, Wakilo, in his capacity as the director of LI, selects Cassandra for a professional trip in Nairobi, in replacement of Juliet. The crux of such a decision is not to treat the new recruit with kid gloves. The boss drives at isolating the fresh heart-throb of the company to lull her into accepting a private intimacy. Marie, Cassandra"s colleague in LI, calls the former"s attention and opens her eyes through the following discussion: Marie: "ask yourself why you should be included in the team going for the seminar when there"re already two members from Editorial going. Doesn"t it strike you as odd especially when you go as a replacement for Juliet?" Cassandra: "hardly a replacement unless if you think I don"t qualify in my own right" Marie: "Oh I didn"t mean that and you know it, Cassandra. But merit has never been the criterion before and nobody"s going to believe that it is now. Take my advice and don"t wait until you"re forced to make a choice." Cassandra: "A choice between what? Marie: "Your job and your integrity." (Barungi, 1999: 15) Marie plainly informs Cassandra about the sexual harassment women are victim of in the LI Company. According to her arguments, the wherewithal offered to them for their recognition and progress in their places of works, is nothing but a self-dedication to unobtrusive give and take operations. Consequently, women stand as sexual toys in men"s hands. Women"s devoirs and rights are messed over and their personalities shrouded into bare and bleak existences. That most blow makes Cassandra learn the deepest lesson. Her reaction is that of a woman who bends her steps towards emancipation. She declares: That"s the trouble with women, always looking for a line of least resistance. It is any wonder men treat us the way they do if we"re not prepared to stand up for our own rights? How can we ever hope to achieve total liberation from all form of injustice unless we fight for them? (p. 16) These utterances are produced and woven into a descriptive and critical logic, turning Cassandra into a ligative force that must pull women towards emancipation. She carries on her shoulders, as a female freedom fighter, the duty of transforming discourses into actions to, decidedly, reduce the differential gaps between men and women. In this way, two axes emerge in her words: an axis of intensity and an axis of extensiveness. She extols her femininity, chides men"s gross malignancy and fumbles for a thawing of her society"s traditions. Unequivocally, the heroine changes her stripes. From a social prey, she becomes an iron lady who overturns the established order and imposes her vision around her social universe. She goes "against cultural, political and economic norms that grant social power and preference to men at the expense of women" (Dione 2017: 410). In addition to bearing a stiwanist cross to head along a combat route of sacrifices in the name of female freedom, Cassandra wears the dress of a female lawyer and attributes herself the duty to plead not guilty for the Ugandan woman. She aims at loosening the social stranglehold that buries the soul of women"s freedom in the depths of a tradition. She detaches herself from the masculinised understanding of her community and attacks the hegemonic will of men, for whom the only force that can tow women towards a fair and acceptable future is that grounded on masculinity.
Barungi reveals a factual truth about a female "emancipatory aspiration" (Re 2017: 112) through discursive activities. The system of manipulation and frontal opposition between men and Cassandra in the Lotus International Company bestows the narrative with a compositional field that amplifies itself in a network of paradoxical interactions. Hence the reversals, tensions and surprises that run through the narration.
Being aware of the impact of modernity on her daughters" generation, Cassandra"s mother manages to talk to her elder spawn through a diplomatic canvas. She tries to reason her into accepting the society in which she lives with all its imperfections and illogic parameters. As she belongs to the older generation, she remains bedded in traditions and lets it know to Cassandra: "certain decencies must be upheld even in your ultra modern society" (p.122). Mrs Mutano does not want hers daughters, Millie and Cassandra to be going out, gallivanting through a hive of flight activities. She insists on the necessity to build a correct and lawful conjugal life through the marital bounds. In her deep understanding of female fulfilment, the old lady, strongly, believes that a woman can only implement completion through "a husband, a home, children" (p.125). For Mrs Mutano, the ultimate goal of a woman cannot go beyond motherism, and mothering in a marital couple; hence the importance she assigns to marriage as a social institution. However, her position is deconstructed by her feminist daughter, Cassandra, according to whom times have changed and bygone époques have to be buried with their unmanners practices and beliefs. She states: Is marriage the only thing going for women, Mama? (…). Would it surprise you to learn that I don"t plan getting married? Marriage is not everything. You know. You have a couple of unmarried sisters yourself who look just fine to me, happier than some of the married women I know, as a matter of fact. (…). There are many married women who never stop wishing they were single again (…) haven"t you for instance, ever regretted giving up your job to become a full-time housewife? (…). You know, Maman, you live in the Dark ages. Today, women have more going for them than the subservient role designed for them by men. Marriage is no longer the only goal. (…). Single women are no longer looked upon by society with pity, at best, and as misfits, at worst. Quite a number of them are, in fact, happy, and living meaningful lives (pp.124-125).
By performing this discourse, Cassandra ceases to be an individual actant. She bears and makes the tradition bear the mark of an inalienable identity in front of the recalcitrant rebellious young girls on whom an obligation cannot be imposed to get them act in accordance with the goodwill of their fellow male citizens. Marriage, which is considered by traditions to be a social chain that links women to a Yes-Okay position, is reversed and portrayed as an unnecessary institution, a social nutshell very often devoiced and devoid of its content, its sense of happiness. Cassandra, through her utterances, indeed, dispraises and abases the "feudal heart of the society" (Beck, Ulrich, et al 1995: 25) to promote common-law couples for the benefit of women"s social margins. As such, she stands against Nego-feminism and the snail-sense feminism to put forward femalism and femalist ideas to, in so doing, give triumph to the emancipatory values which underlie the anti-patriarchal women"s struggle. Mrs Mutano"s daughter is therefore observed in a process of demystification and of total stripping of marriage in particular and traditions in general with their patriarchal finery.
The head-on confrontation between conformists and anti-conformists in Ugandan society is echoed in a daughter-mother discussion through which divergent opinions clash within the logic of the forbidden and the bravado. Barungi uses this dual face-off as a narrative device through which she textualises the fractured relationship between the overtones of orthodox duty and the obligation to rise up against the social status quo.
In Cassandra, Cassandra puts on surface a boulèsis (a strong will) to gain back her female poise. With a growing sense of all-occasion protest, she drives at liberating herself from any form of domination and exploitation. To keep the best for female bona fide, the young girl hangs on the balance to recoup the noûs praktikos and make modern times prevail over "dark ages" (p.124). She begrudges women like her mother of their aphasia and condemns their philosophy that consists in shirking their duty, their need to get out of their cage and face the daily oppression they are victim of. She puts a jaundiced view on patriarchy and, with a notch of arty commitment, breaks the domineering voice which targets "to persuade the powerless that their powerlessness is inevitable" (Ruthven 1984: 31).
In her permanent combat against the status quo, Cassandra, who has contracted a pregnancy out of marriage, decides to keep it against all odds. In spite of her society"s ferocity against single girls" gravidities, the young girl refuses to show undue concerns and avoid shirking her duty as a pregnant woman: "I intend to keep the baby. (…) I shan"t change my mind" (p.147). Her standpoint is opposed and challenged by her own sister, Mellinda. The latter who is taken aback by Cassandra"s stubbornness, looks down on her sister who trivialises what happens to herself. She proposes Cassandra to abort before being ridiculed and belittled by their male-dominated society. She warns her sister through these statements: You know, Cassandra, I sometimes despair of you, truly I do with your lofty views and unrealistic attitudes. But even you must know there are different rules for men and different rules for women. You may not like it but it doesn"t change the facts an iota. We live in a man-dominated society (p.150).
In this Ugandan society, the woman, in the eyes of traditions, remains a symbol of legitimate and legal submission. Her being and appearance remain the expression of a collectively accepted constraint and domination. On their shoulders, traditions place the burden of silence, assertion and conformism. Cassandra, who behaves with a perfect decorum, stays noncommittal in her views. Her answers are direct and laconic. To her sister"s criticism, she impudently, expostulates: "I"ve not lost sight of my goals in life, this is a minor setback (…). This is 1984 for heaven"s sake, not 1900" (p.148). A substantive argument about mentalities emerges on sight in a post-independence Ugandan society. Violet articulates a relationship of implication that unites the highly moral prohibition and the salvific duty. The author paints the narrative with a movement of logic where consecution and consequence govern the action in the storyline syntax of the R narration. The narrative space is thus enriched with catalytic functions and a parallelism is established between the conformist duty and the will to row against the current of customs. The plot then develops in a binary tonality, opposing the bearer of traditions and the progressive woman. Their contradictory points of views punctuate the sequences of the narrative on the axiologies of female resistance and feminine submission. The plot then gains in scenes and is divided, at this level of its message, into two fundamental sequences: the why-based sequence and the how-based occurrence. Indeed, Cassandra asks herself why the woman has to undergo the torments of ethno-patriarchy and her sister shows her how she has to stick to her tradition as a woman.
Violet Barungi bipolarises her message and brings out two forces that clash on the sensitive theme of the female body. If the intellectual, Cassandra expostulates on the necessity to spin, deconstruct and fidget the spine of the traditions, her sister mulls over how to sensitize and keep her abreast of the danger to disincline the traditional best-laid order. Hence the deliberate use of the warning by these words: "Would a little consideration for the people (…). I wash my hands of you (…). Like you want to show the world how tough you are, and how tenaciously you cling to your stupid principles" (pp.148-150). Mellinda dissociates herself from Cassandra"s combat. Good riddance to bad rubbish, Mellie seems to say to her sister. Her intolerance to the pregnant girl"s rebellious behaviour is sharp and pointy. However, straight in her boots, without fiddling or trembling before the powerful steamroller of tradition, Cassandra carries on her combat, sticking on her vision as an African feminist, and putting forward her full right over her full body. According to her "the logos (speech), the epitumia (reflection) and the ergon (action) are the top dogs items forfended by men to outstretch and avouch bluntly their lordship on women who are reduced to impotence and second hand objects. Traditions levy a model of private life which is to be scrutinized and controlled by parents, in-laws, friends and relatives. Life is to be lived in compliance with the does and don"ts of the society. But Cassandra refutes that standing point and cuts it clear in the following: "I never do anything that I don"t want to do" (p. 99).
Cassandra presents herself as a female revolutionary and thus embodies the expression of a sacrifice, a generosity of an individuality that confronts the very foundations of her society and consequently shows the way forward. She is painted as a heroine who embodies a virile cause and aims for glory in a liberating victory. Traditionalist men and women look down on and label her as a lost sheep whose efforts for emancipation will get her nowhere, if not to isolate her further from her own people and her own intrinsic identity.
In fact, in order to free women from the weight of silence imposed on them by the patriarchal system, Barungi raises a woman who, with a strong personality, stands up against the iniquity and blind haughtiness of a male power. In so doing, she inscribes a positional relationship where the action of the strongest is described as being transformational. She, as a committed writer, then reveals a social scheme that organises a cultural system where men are in junction with objects of positive values to make women state subjects, victims condemned to bend their backs before the domineering force of men. She gets Cassandra play a role which consists in questioning the social and cultural conception on men"s superiority over women. The male voice which rings loud and bears the tonality of chieftaincy is to be broken into conciliation and horizontality.
II. Love and Whatnots Irking Obstacles for Emancipation
Love! Woman! Between the actants is man, who bears the seal of quest and conquest. He sets his sights on the woman he wants to be attractive, loving and congenial. This is how the relationship between man and woman is, sometimes, determined and characterised. Love is then the feeling that unites male and female to give meaning to life, which is perpetuated through the channel of reproduction. In literature, love has remained an old theme, as old as the letter and the spirit of the word love itself which carries the message of affection that, according to André Maurois can be "a love-passion, a love-physical, a love-taste and a love-vanity" (Maurois, 1972: 47). In Cassandra, love is conjugated in the mode of passion and it eschews all forms of rationality.
Indeed, it is in the middle of the night that Cassandra and Ray mollify their sexual desire and thus sublimate their carnal conjunction into an illegal lovely union. The locative actant (Ray"s house) is enriched by a subject recipient and the narrative is axiologised. An euphoric atmosphere dominates the setting and favours the deployment of the interactions between the two lovers. While she is a virgin and unmarried, Cassandra makes a deliberate decision to give herself over to a married man. Her lover"s expectations take precedence over her principles that she sees tumbling down from the height of their rigidity. The first sexual act between Ray and Cassandra is free of any command and sounds like both a will and a duty to perform. The woman, who feels powerless in front of the man she loves, finds herself in a potentialized status: she is an object of desire and offers sexual delights to the man she dears without restraint or ulterior motives. Cassandra wants to be loved, neutralised and fixed on the dead centre of her existence as long as she is accepted in Raymond"s arms. In her discussion with George, she confesses: George: "you"re supposed to be a liberated woman" Cassandra: "I"m not that liberated, I"m afraid. Hard as it may be for you to believe, George, I too suffer from the modesty syndrome of sex." George: "Hallelujah, you"ve finally admitted to being human too!" (p. 27).
Sex is the weak point from which Cassandra suffers. Her commitment to the female cause is compromised by the exigencies of her sex. She cannot make a difference between satisfying her natural desire and her standpoint as a "liberal" woman. Her determination to resist male domination is combed by the harassing need of sexuality. On the one hand, she battles against patriarchy and man"s upper hand on women; on the other hand, she bows down and worships man"s virility and sexuality for the sake of her sexual well-being. In fact, Cassandra alchemizes her body into an offering that her boyfriend, Ray, willing, a field day, sacrifices under the altar of a male domination. She makes herself flesh in her desires and verbalizes her love in sexual acts to the goodwill of her man. Therefore, her female sex becomes a dumping ground for her feelings of inferiority. The young feminist, who used to martyr herself to move the mountains of male authority, has softened herself in front of love and given herself to be read in a logic of a very watchful and lustful submission. Her battles of yesterday become litanies with a shortened narrative and half-tinted commitments, and dreams of an innocent young girl.
Cassandra then discovers the world of sex and sexuality. And it is in this world of pleasures and temptations that the rebellious girl will lose the main meaning of her fight for freedom and the restoration of female dignity. The loss of her virginity marks the end of a match and the beginning of a life painted on the edge of chaotic and unbearable lightness grounded on a sexual life that goes straight to a total drift. In Cassandra and Raymond"s one-flesh union, passion prevails over morality and makes their love and love-makings bear the stamp of fornication and adultery. Actually, Cassandra"s boundless attachment to Raymond leads her to be more conciliatory and consensual. In the face of his love, she stops playing at elbows and becomes more interested in keeping her man than in emancipatory politics. Love wins out over reason; and her feminist struggle takes a socially compromising gasp. The sacredness of the feminist world collapses on the slab of the fighter"s germinal and fecund dream. A game is lost, and a "war" is compromised. The barometer of female resistance to the hegemonic traditions in the Ugandan society seems to be cooling under the mind-altering effect of sex and love. Cassandra turns her femininity into an erotic capital and wears the effigy of her sexual pleasures as a safety belt to protect her dream of keeping and preserving the man she cherishes with all her might. In a conversation with Ray, Cassandra avows her weakness and half renunciation to her principles and struggle: Ray: I know you take life seriously and determined to make it to the top of your career, I am right. Cassandra: That has been my dream so far, my only dream and the sole goal of my life. It was simple enough goal, possible too but now…Now something seems to have gone wrong with my calculations. I"m not getting the correct answers anymore. But as you said, there"re no given formulae to the riddle that is life; all one can do"s flounder in the dark (p.32).
The intellectual feminist"s dream is extinguished in a world of inequality and inequity. A paradox then arises between an ardent desire for freedom and a feminine power that seems to fall flat on the face of its impossible nature. This situation locks Cassandra"s life and that of her fellow women into a vicious circle, which takes on a rotating pattern that always brings the woman back to the zero point of her African feminist struggle.
Hegemonic feminism dreamt and roused by the young girl fades into a "situation in which women"s equality with men is denied" (Hall, 2001: 250). Cassandra falls apart with her bundle of sexism into the intersectional point of love. She loses her protesting voice and bows down in front of the strong assertion according to which "nothing exists that has not been made by man (…) the world is man"s word" (Leclerc, 1990: 74). Her submissiveness to men she loves is a turnaround that, read through an angle of stewanism and de Beauvoire"s conception of feminism, is an admission of failure and weakness of a female stand that can no longer question the belief that "with one voice, only one can speak. Man (Leclerc, 1990:75) the warning down of her female commitment is "supported by the existence of sexual desires like fetishism" (Giles, 1994: 339-357).
Morality and female dignity are sacrificed at the cost of love. In this way, Violet Barungi takes a critical look at her traditions, of which identity appears to be based on the perpetuation of the male-dominant system underpinned by a patriarchal spirit. Violet then focuses on the wrongness of a traditional system that has no other means than to create disparities that lead to social discrepancies and exclusions. The perpetual quest for happiness that Cassandra shows up contradicts her in her choices, and locks her in a permanent state of psychological and moral instabilities.
The unsubmissive girl comes up against the harsh realities of love and is disarmed by the unfathomable reasons of blind affection. Through this loosening and resignation in the face of the obstacle, Violet Barungi highlights the social barriers that obstruct the path to emancipation for Ugandan women in particular. Cassandra, indeed, bends over backwards and relegates her indignation to the second degree of her resistance. She expresses her choice in this way: "I should hate you [Raymond] to discover that underneath the veneer of dazzling wit was the usual humdrum female […]. Your friendship […] means more to me than anything else in the world" (pp. 33-35). Cassandra is disarmed by the unconditional nature of love and illustrates her powerlessness through a comparative statement which, in reality, shows resignation. In her view, there is no possible match between her love and her feminist combat. Out of many goals, only that about love is, afar, beyond any possible compromise. Thus being, she expresses a psychological state of abandonment, hence a pitiful position of a "model" female citizen who is being gangrened by the strength of masculinity that disempowers and disadvantages women. Her reaction then carries an assertive meaning which refers to a message of social and cultural establishment. She, undeniably, goes to the same direction than Annie Leclerc who wanders before stating in this way: "Man? What is Man? Man is what man brings into the world. We made children, they made man" (Leclerc, 1990: p.75).
Besides, Barungi follows and describes the adventures of the feminist Cassandra, bringing to the surface the assets of her actions and transformations. She turns her life into a narrative journey littered with conjunctions and disjunctions, punctuated by reflexive alterations through which her novel exposes the pulling and pushing factors of a struggle that is far from being won in advance.
Furthermore, the interactions in the marital space give rise to time effects that erode or strengthen feelings of love depending on whether one is married or is a mistress of a married man. While Cassandra finds joy in her illegal and extramarital relationship with Ray, Belinda suffers in her gut from the dimness of her marital life. This brings her to rebel against her husband"s expectations and makes their married existence a living hell. Belinda, thus, attacks the marital laws in her society, dragging her husband into the mud of humiliation and dishonour. Not only does she refuse to divorce, but she forces her man to distance himself from herself so that she can dispose of her own body. While Cassandra goes down in her resistance against male domination, Belinda flouts the laws of tradition that organize intimacy in the couple to control, by her own, her sex and her sexuality. A sort of a handing over seems to take place between Cassandra and Belinda. The latter, to Raymond"s authoritative assertion: "I"ll keep you for as long as it suits me" (p.215), she bitterly answers back: "I"ll not be tied to a crippled. Never!" (p. 215). The use of the adverb "never" in the sentence transforms the addressee into an allocator and homogeneous listener. This expressive and assertive utterance itself gives values to the woman"s refusal and rejection to abide by a man"s injunction. This places Belinda in a position of a "rebel" avenging herself against a former male-dominating power.
Belinda draws a curtain of illocutionary acts between herself and her husband in order to escape from men"s dominating potency. Thus, she uses a directive to tell her man, who is no longer a man,: "my husband, I admit, was once a ladies" man, but even you with your limitation, nurse, could do better than a crippled invalid with minimal chances of full recovery" (pp.216-217). Despite these assertions, Ray remains amorphous and apathetic. So Belinda refrains from blaming herself and develops her argument around an exclusive "comparison of superiority" through which she neutralizes any difference between him and "a crippled invalid" (p. 217).
Belinda, who hopes to take up the torch of femmism, retracts her decision to pursue her happiness with highranking military personalities. Indeed, her revolt against Raymond"s marital authority does not free her from male domination. Like Cassandra who abdicates and gives up the fight in the face of passionate love, Belinda trades in her independence and female authority for wealth and material happiness. She succumbs to a love-vanity and gives way to submission. The narrator informs: What about Belinda? Since the death of their son, Raymond and his wife had become more estranged than ever before although neither had made the move to legalise their separation. But now, at the time of husband"s accident, Belinda had been considering doing so. She just struck it lucky with one of the nouveau riche army officers and did not have time for her suffering husband (p. 214).
Through her renunciation of the struggle by and for women, Barungi shows the weakness of the feminist movement in Uganda in the 1980s. A struggle whose limits were matched only by the selfishness of those who engaged it. Both Belinda and Cassandra have succumbed to the submissive calls of love and have consequently lost sight of their principled attitude. If Mellie and her mother have given audience to the scrupulous respect of established norms, Cassandra and Belinda have moved from a firm commitment to a disengagement whose final objective is and remains the personal satisfaction of carnal and financial aims.
CONCLUSION
Violet Barungi, in Cassandra, has merged the phenomena of time to articulate a narrative identity. The movements of her main protagonist, in time and space, are expressed and illustrated as temporal values that signify a being and a becoming. Indeed, Barungi portrays Cassandra"s past in her capacity as a fighter, a protester, and a committed feminist woman, to detach her from the present, which she considers to be a sentimental salvation. She observes and is observed in a denominational movement from a principled, non-negotiable woman to a docile and submissive one at will. Her attempts to achieve a temporal syncretism between her past and her present collapse before the demands of a cornelian take-it-or-leave-it choice. Such a position sufficiently undermines her room for manoeuvre and subjects her to the man"s nirvana desires. Actually, Burungi, in Cassandra, has succeeded in highlighting the predicaments Ugandan women came across in their thorny and painful way toward emancipation.
Victim of a double prejudice related to patriarchy and femality, Cassandra incarnates the topical struggle of Ugandan women to unchain themselves from any form of relegation to a second class citizenry and second-gendered sex. Through her itineraries, points are handed out and female barometer to measure the mercurial value of the Ugandan women"s achievements strongly oriented towards another page of her woman-being and woman-becoming in the Ugandan society. Through Cassandra, Barungi has achieved a gripping on the female destiny rattled by the dynamism of her surroundings. She succeeded in transforming afar dreams into close realities. The narration follows a horizontal direction to contradict with the verticality in social, emotional and professional relations between men and women. A way is shaped out. A will is strengthened. A struggle is engaged. However, is a total female victory within reach? "That is the question!" (Shakespeare 1992: 53). | 2022-04-30T15:08:06.447Z | 2022-04-22T00:00:00.000 | {
"year": 2022,
"sha1": "4c9eb70ebbb62d0d7b480378da3becc50b496b02",
"oa_license": null,
"oa_url": "https://doi.org/10.36346/sarjall.2022.v04i02.003",
"oa_status": "GOLD",
"pdf_src": "MergedPDFExtraction",
"pdf_hash": "a0a1169e21d1626ee85a7fb519ef643368ef605c",
"s2fieldsofstudy": [
"Sociology"
],
"extfieldsofstudy": []
} |
17734847 | pes2o/s2orc | v3-fos-license | Alcohol use and health outcomes in the oldest old
Background As the United States population ages, an unprecedented proportion of the population will be aged 70 and older. Knowledge of alcohol use and its consequences in this age group is not well known. In light of the disparate findings pointing to negative outcomes with excessive drinking yet also benefits of moderate drinking, the true risk of alcohol use in late life needs more investigation. Methods This study examined the correlates and 2-year health outcomes related to alcohol use in 7,434 elders aged 70 years or older. Data was collected as part of the Assets and Health Dynamics of the Oldest Old (AHEAD), a nationwide health and economic study of elders. Data from Wave 1 and Wave 2 of AHEAD are presented. Frequency and quantity of drinking was assessed by self-report as was health status, lifetime alcohol or psychiatric problems, presence of chronic illness, functional impairment, and depressive symptoms. Cognitive status was assessed using a brief measure. Results Approximately 44% of the sample reported any alcohol use in the past three months, with the majority of drinking less than daily. Daily drinking was associated with being Caucasian, married, in relatively good health, and having good affective and cognitive status. Drinking was not associated with negative health outcomes two years later and was protective against stroke and functional impairment. Decline in drinking between Wave 1 and Wave 2 was strongly associated with poor health. Conclusion This study offers no evidence of negative health outcomes for drinking moderately and confirms the U-shaped curve often found in studies of alcohol and health. Nonetheless, cessation of drinking was associated with poor health suggesting the health benefits of moderate drinking may result from selection of a healthy group of people capable of sustained moderate drinking. Public health recommendations for moderate drinking must take this phenomenon into account.
Background
The age structure of the US population is making an unprecedented shift towards growing proportions aged 65 and older. The number of Americans aged 65 and older grew from 3.1 million in 1900 to 35.3 million in 2001 [1]. This corresponded to a growth from 4% of the population in 1990 to 12.4% in 2001. With the aging of the baby boomers, this is expected to rise to 70.3 million by 2030, about 20% of the total projected US population [2]. The highest rate of increase within those aged 65 and older is in the oldest-old, those aged 85 and older. This age group is expected to almost double between now and 2030, growing from 4.4 million to 8.9 million[2].
The aging of the US population calls for more research on the prevalence, risk factors and consequences of alcohol use in late-life. In contrast to the vast literature on alcohol use and abuse in younger age groups, relatively little is known about drinking patterns in the elderly, particularly the oldest-old. Of the two largest community based studies of psychiatric disorders, the Epidemiologic Catchment Area Study (ECA) and the National Comorbidity Survey, only the former included people aged 65 and older. The ECA estimates alcohol abuse in this age group to range from 1.9 to 4.6% for men and from 0.1% to 0.7% for women [3]. This study was conducted in the early 1980's. The National Household Survey on Drug Abuse was conducted more recently between 1991 and 1993 [4]. It examined patterns of alcohol use as well as dependence. It estimated that 54.9% aged 50 and older used alcohol, but alcohol dependence was 1.6%, comparable to that found in the ECA. This sparse literature demonstrates that there is significant risk for alcohol abuse in late-life, though both studies found the risk to be much lower when compared to younger age groups.
Drinking in late-life has become a more complicated issue in light of the growing literature on the health benefits of moderate drinking and the relatively lower risk for alcohol abuse. Whereas abstinence was once considered the healthiest option, more and more physicians and public health leaders are following an informal clinical policy of recommending moderate drinking, usually defined as one drink per day. Several studies in predominantly middle aged samples (mean age between 50 and 60 years) have found a benefit of moderate drinking, particularly for cardiovascular outcomes [5][6][7][8]. Both men and women who drink approximately 1 drink per day have a lower relative risk of coronary disease when compared to non-drinkers or heavy drinkers. This "U-Shaped" curve is well established though speculations on the reasons underlying the curve are controversial.
Those that propose a direct benefit of alcohol cite the effect of raising high-density lipoprotein (HDL) cholesterol levels in moderate users [9,10]. Furthermore, cardioprotective benefits may be related to alcohol's ability to decrease platelet aggregation [11] and increase the prostacyclin/thromboxane ratio [12,13], resulting in a reduced propensity to thrombosis. Others argue that the U-Shaped curve is an artifact of the demographic group most likely to engage in moderate drinking [12][13][14][15][16]. Therefore, it is not the alcohol that leads to health benefits, but rather social and personality factors that can sustain long periods of controlled drinking without leading to excessive use. Moreover, many abstainers are people who previously used alcohol, but due to the development of health problems, had to stop drinking [15,16].
In support of this, the British Regional Heart Study of middle aged British men found that 70% of non-drinkers are ex-drinkers; thereby contaminating comparisons of the long-term health risks and benefits of alcohol use [15,16]. Similarly, Krahn et al [14] found the demonstrated benefits of moderate drinking were no longer evident after controlling for baseline cognitive ability and educational attainment. This study had the benefit of measures of cognitive ability in the late teens presumably prior to any extensive alcohol use. However, subsequent studies have controlled for prior alcohol use or education level and the moderate benefit for cardiovascular outcomes remains [5,17].
It is unclear if these benefits of moderate drinking hold in late-life. A separate examination of elders only is needed because of the issue of competing risks, particularly in more vulnerable elderly. Taking into account competing risks means examining the full risk profile for a certain behavior rather than examining its impact on one outcome. Although cardiovascular health is enormously important in aging, the recommendation of moderate alcohol use must take into account alcohol's potential to cause other negative outcomes in late-life. The benefits of moderate drinking need to be weighed in light of the risk they pose to other key geriatric syndromes such as cognitive impairment, depression, falls, and hip fracture.
The negative effects of alcohol use relate mostly to excessive drinking. Alcohol abuse can lead to pancreatitis, cirrhosis, or alcohol-related cardiomyopathy [18]. It may also lead to impaired driving, falls, memory problems, depression, and sleep problems [19][20][21]. It is unclear whether moderate drinking in late-life increases risk for these outcomes. Looking at a range of health outcomes will demonstrate whether the benefit of moderate alcohol use increases other health risks. It will also test whether the benefit of alcohol use is specific to cardiovascular outcomes or instead confers an overall protective effect. If the later is true, there may be broader underlying correlates of moderate alcohol use that account for the positive health outcomes.
This study presents data collected as part of a nationwide community-based survey of elders aged 70 and older who participated in the Assets and Health Dynamics of the Oldest Old (AHEAD) [22]. It has three aims. First, it will present data on drinking patterns in the oldest old. It will then present correlates of drinking in late-life. Finally, it will present longitudinal data on the short-term consequences of drinking in late-life. We will determine the relation between drinking and a wide range of health outcomes to examine whether the benefits of alcohol use are specific to cardiovascular outcomes or whether they are associated with overall health. The goal of this study is to inform clinicians and public health policy makers of the benefits and risks of alcohol use in late-life.
Methods
AHEAD is a companion study to the Health and Retirement Survey and is intended to investigate the impact of health transitions on personal financial management, service and public program utilization, and intergenerational transfer of assets [22,23]. Wave 1 of AHEAD occurred in 1993-1994 and Wave 2 occurred in 1995-1996. The two sampling frames for the study were the 1992 screening of housing units enumerated for the Health and Retirement Survey and the Health Care Finance Administration's Master Enrollment file of Medicare enrollees who were living in a household. Primary respondents had to be 70 years or older and, if married, their partners participated regardless of their partner's age. Although the initial sampling frame excluded institutionalized elders, respondents who were institutionalized after Wave 1 remained in the study and were interviewed at Wave 2. All participants provided verbal informed consent and internal ethics review board approval was obtained.
The sample used in this study were all who participated in the first wave of AHEAD (N = 7,434) excluding spouses under the age of 70. Of these, 727 died between Wave1 and Wave 2. Of the living at Wave 2, 6,222 completed the second interview yielding a 93% follow-up rate. There is some variability in sample size for individual analyses due to missing data. Minor variability is due to non-response of participants and alters sample size by no more than ten observations. In addition, when proxy informants were used, the depression measure was not administered limiting the sample size to N = 6649 for analyses using this variable. Likewise, the cognitive measure was not administered when proxy interviews were done or the respondent was unable to complete the measure, limiting analyses including cognitive status to N = 6351. Demographic characteristics of the entire sample are provided in the second column of Table 2.
Alcohol use and health measures
All health assessments were based on self report, except for cognitive status, which was based on the Telephone Interview of Cognitive Status-Revised [24]. Mortality data was collected in interviews of nearest kin and from the National Death Index.
Alcohol use
In Waves I and II, all participants were asked about frequency and quantity of drinking in the three months prior to the interview. Two questions assessed quantity of drinking "Do you ever drink any alcoholic beverages such as beer, wine, or liquor?" which was followed by "In general, do you have less than one drink a day, one or two drinks a day, three or four drinks a day, or five or more drinks a day." Participants were also asked the first item from the CAGE [25]screen for alcohol abuse: "At any time in your life, have you ever felt you should cut down on drinking?".
Health variables
These too, were based on self-report. Participants were asked to rate their health as excellent, very good, good, fair or poor. Participants reported whether a medical doctor had diagnosed them with cancer, heart disease, diabetes mellitus, high blood pressure, lung disease, stroke, or arthritis. They also reported whether they were current, former, or never smokers.
Falls and hip fracture
Participants were asked "Have you fallen down in the past 12 months?" and "Have you ever fractured your hip?"
Functional impairment
Measure of participant's functioning on activities of daily living (ADL's) and instrumental activities of daily living (IADLs) were assessed. In the ADL assessment, participants reported whether they needed help walking, dressing, bathing, eating, getting into bed, and using the bathroom [26]. These items were selected based on the original instrument described by Katz et al. [27] and subsequent revisions by Kane and Kane [28] and Weiner et al. [29] The assessment of IADL's included meal preparation, grocery shopping, telephone use, taking medication, and managing money. Item selection was based on Fillenbaum's revision [30]of Lawton and Brody's [31]original measure of IADL's. Participants were coded categorically as having any versus no difficulty in ADL's and IADL's because a large majority of participants reported no difficulty in either.
Neuropsychiatric variables
The AHEAD cognitive measure was adapted from the Telephone Interview for Cognitive Status [24], which was modeled after the Mini-Mental State Examination [32]to be administered over the telephone. A total score (range 0 -35) was determined by summing the serial 7, immediate and delayed free-recall, and the mental state scale totals. Depressive symptoms were assessed using an abbreviated 8-item version of Radloff's Center for Epidemiologic Studies Depression Scale [33]. The abbreviated version demonstrated a comparable factor structure and internal consistency to the 20 item version [34]. In addition, all participants were asked "Have you ever seen a doctor for emotional, nervous, or psychiatric problems?"
Statistical analysis
All statistical analyses were conducted on SAS statistical software. Descriptive analyses were conducted using chisquare tests for categorical variables and a Wilcoxon ranksum test or the Kruskal-Wallis Test for continuous variables. The longitudinal analysis used logistic regression for categorical variables and a general linear model for continuous variables. Two sets of longitudinal analyses were conducted which examined the impact of Wave 1 drinking on health outcomes at Wave 2 occurring two years later. The first set controlled only for the Wave 1 value for the predicted Wave 2 outcome. The second set controlled for age, sex and education and the Wave 1 value for the predicted Wave 2 outcome. Because the results were largely similar, only the latter are presented, and the former are available upon request.
Similarly, generalized estimating equations were conducted on the initial bivariate analyses presented in Tables 3 and 4. These analyses controlled for correlated observations due to sampling of participants' spouses or partners if they were married. The results of the generalized estimating equations and the simpler Wilcoxon or Kruskal-Wallis test were practically identical (analyses available upon request) indicating that the correlated observations did not alter the relation between alcohol use and health outcomes, thus we opted to conduct the simpler models and present these results. To control for multiple comparisons and large sample size, a significance level was set at p < 0.001. All tests were two-tailed. Table 1 displays frequency and quantity of drinking in Wave 1 and Wave 2. In Wave 1, a little over 55% of the sample reported no drinking. Of the drinkers, the vast majority did not drink daily, while 8% drank one to two drinks per day. Two percent reported drinking three or more drinks per day. In Wave 2, occurring two years later, there was a general shift towards less drinking. In that wave, 63% reported no drinking while approximately 1.5% drank three or more drinks daily. Table 1 also presents the percentage within each drinking consumption group reporting ever in their lifetime needing to cut back on drinking. The proportion endorsing this item gradually increased with each level of alcohol consumption, from 12% in the non-drinkers to 50% in those drinking three or more drinks per day. The percentage endorsing ever having had psychiatric problems in their lifetime follows the U or J-shaped curve with the non- drinkers and the 3-or-more drinks/day having the highest rates (11.5% and 15.7% respectively) and the moderate drinkers having the lowest rates at 9.7%.
Results
The demographic correlates of drinking are provided in Table 2. Both gender and age showed a linear relationship, with greater drinking being associated with younger age and being male. Marriage was positively associated with moderate drinking. The relation to race and education was a bit more complicated. Moderate drinking when compared to no drinking was associated with being white and more years of education. A U-shaped pattern was seen for race, where moderate drinking was associated with being Caucasian, and either no-drinking or drinking three-or-more drinks per day was associated with being a minority.
Most of the negative health outcomes examined were higher in the non-drinkers and lower in the moderate drinkers (Table 3). Diabetes mellitus was also higher in the non-drinker group than the moderate drinkers. Cognitive function was higher for moderate drinkers and depression was lower. In contrast, the percentage currently smoking increased with increase in quantity of drinking. Table 4 presents the value of Wave 1 drinking in predicting negative health outcomes at Wave 2. Each row presents the predictive value of Wave 1 drinking for the Wave 2 outcomes while controlling for age, sex, education and the Wave 1 value for the specific outcome (e.g. Wave 1 stroke for the model predicting Wave 2 stroke). Even in the controlled models, drinking conferred a protective benefit for stroke, and ADL and IADL functioning. Protection from heart disease, diabetes, and hip fracture did not reach the a-priori threshold set for statistical significance. Drinking did not appear to be associated with falls. Similarly, Wave 1 drinking was not associated with poorer cognitive function at Wave 2, depressive symptoms, or total number of chronic illnesses.
Of note, 727 participants died between Wave 1 and Wave 2. Alcohol use was strongly associated with mortality (χ 2 (df = 3) = 38.5, p < 0.0001). Twelve percent of the nondrinkers died while the mortality rate for those drinking less than once a day, once to twice a day or three or more times a day were approximately the same ranging from 6.5% to 7.6%. Therefore, any health benefits presented in Table 5 could be underestimates, since mortality was higher in the non-drinkers.
The Wave 1 predictors of a decrease in drinking between Wave 1 and Wave 2 mirror the correlates of alcohol use in Tables 2 and 3 (Table 5). People who cut back on their drinking tend to be single, of minority status, and less educated. They are in worse health as indicated by the greater presence of heart disease, diabetes, and functional impairment. They also have significantly lower cognitive function and higher levels of depressive symptoms.
Discussion
In the oldest old, moderate drinking is associated with better health. However, it is also related to demographic factors that are strongly associated with health outcomes. The relative contribution of these two factors to the apparent health benefits of moderate alcohol use needs to be better understood.
In this study, a little over one-half of the sample reported no alcohol use in Wave 1. This grew to 63% of the sample in Wave 2. These percentages of non-drinkers are far higher than those presented for younger age groups [4,8].
Although this can reflect cohort differences, it is more likely to result from the cumulative impact of drinkers stopping use because of ill-health or general aging. Along these lines, with increasing age the likelihood of prescription medication use becomes greater, often involving medications that preclude or warn again concomitant alcohol use. The analysis of decline in drinking presented in Table 5 provides a snapshot of the process contributing to non-drinking and its association with poor health. Although it may be accelerated in late-life, it seems safe to assume that this process is occurring throughout the life course. Therefore, health comparisons between drinkers and non-drinkers without controlling for prior drinking status are highly confounded. Moreover, given the large socioeconomic factors contributing to drinking status, it is not clear if controlling for prior drinking status would be sufficient.
Nonetheless, in the controlled models presented in this study, alcohol use did not appear to be associated with any negative health outcomes including falls and hip fracture. Indeed, it was associated with some health benefits such as stroke and physical function. Drinking was also associated with lower mortality between Waves 1 and 2.
There appeared to be the potential for benefit for other health outcomes such as heart disease, but controlling for demographic factors mitigated this association considerably.
The Krahn et al study [14] noted earlier is unique in that it examined the relation between drinking and health while controlling for demographic and cognitive variables assessed long before this process of selection due to health on drinking has started. More studies like this are needed and some of the major longitudinal studies of aging, such as the Harvard Grant study [35]or the Seattle Longitudinal Study [36] should have the requisite data to conduct such an analysis.
This study suffers several limitations. All health variables are based on self-report, including that of alcohol use, a behavior that is often underreported. Moreover the questions forced respondents to group their alcohol use into apriori categories. Assessments of depressive symptoms and cognitive status used validated measures, yet are briefer versions than those used in more detailed studies. The longitudinal analyses are based on a two-year period which is relatively brief when compared to other studies demonstrating the long-term implications of moderate alcohol use. Drinking was associated with lower mortality, so it is likely the benefits of drinking at Wave 2 were underestimated because the most ill non-drinkers died and were not represented in the longitudinal analyses. Finally, the analyses presented did not correct for correlated observations, yet exploratory analyses described in the Methods section strongly suggested that the results would not be altered greatly by doing so.
More research is needed to elucidate the mechanism by which alcohol use may benefit health in late-life. In addi- tion, substance use varies widely by cohort. As the "babyboomers" age, cohort specific research should be done on this group because of their known greater use of both alcohol and drugs.
Conclusion
The health benefits of drinking observed in this study are not specific to cardiovascular illness, but are present for a range of medical and functional outcomes. Although the strong associations between moderate drinking and health outcomes supports the prophylactic use of alcohol, there also appear to be non-specific cultural factors that are tied to socioeconomic status. Given that socioeconomic status is one of the strongest correlates of health outcomes, these non-specific cultural factors need to be understood better before making unrestricted recommendations for moderate alcohol use. It should also be noted that moderate drinkers in this study were persons over the age of 70 who were able to maintain only moderate drinking, which may reflect a lower propensity to addictive behaviors or substance abuse disorders within this group. This is supported by the lower rates of smoking, depression and other psychiatric problems among the moderate drinkers. An informal clinical policy in favor of moderate drinking, then, may be appropriate only for a group with low risk for addictive disorders (e.g., no family of personal history of psychiatric disorders or substance abuse), but may not be appropriate for the general population. Due to missing data for proxy interviews, the n for analyses using the CES-D is 1737 and 903 for the two groups. For analyses using the TICS-R, the n = 1693 and 878 for the two groups. | 2018-05-08T18:25:39.774Z | 0001-01-01T00:00:00.000 | {
"year": 2006,
"sha1": "2db393f0329f4a95c3ba3ab1303acc7ce5cf68c3",
"oa_license": "CCBY",
"oa_url": "https://substanceabusepolicy.biomedcentral.com/track/pdf/10.1186/1747-597X-1-8",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "2db393f0329f4a95c3ba3ab1303acc7ce5cf68c3",
"s2fieldsofstudy": [
"Medicine",
"Psychology"
],
"extfieldsofstudy": []
} |
13562495 | pes2o/s2orc | v3-fos-license | Associations of Workplace Bullying and Harassment with Pain
The aim of this study was to investigate associations of workplace bullying and harassment with headache, stiffness of the neck or shoulders, lumbago, and pain of two or more joints. The subjects in this cross-sectional study were recruited from workers (n = 1,913) at 35 healthcare or welfare facilities in Japan. Because of non-participation or missing data, the number of subjects included in the analysis varied (response rate ≥ 77.1%). Workplace bullying and harassment were assessed using the Negative Acts Questionnaire. Depression was assessed using the Brief Job Stress Questionnaire. The frequency of pain experienced by workers in the previous month was evaluated using a four-point scale. Many of the associations of person-related bullying, work-related bullying, and sexual harassment with headache, stiffness of the neck or shoulders, lumbago, and pain of two or more joints were positive and significant (p < 0.05). Even after adjustment for depression, some of the associations remained significant (p < 0.05). For example, changes in the prevalence ratio for headache associated with a 1-point increase in the work-related bullying score were 1.05 (95% confidence interval (CI) 1.01 to 1.09) in men and 1.03 (95% CI 1.01 to 1.05) in women after adjustment for age, marital status, employment status, work shift, and depression.
Introduction
According to the International Labour Office (ILO), the International Council of Nurses (ICN), the World Health Organization (WHO), and the Public Services International (PSI), bullying (or mobbing) is "repeated and over time offensive behaviour through vindictive, cruel or malicious attempts to humiliate or undermine an individual or groups of employees," and harassment is "any conduct based on age, disability, HIV status, domestic circumstances, sex, sexual orientation, gender reassignment, race, colour, language, religion, political, trade union or other opinion or belief, national or social origin, association with a minority, property, birth or other status that is unreciprocated or unwanted and which affects the dignity of men and women at work" [1]. Many victims of these actions demonstrate depressive or somatic symptoms [1].
In a recent cross-sectional study, neck pain was associated with intimidation at work and with unwanted sexual attention in bivariate analyses in both genders [2]. In another recent study, new onset chronic neck pain during follow-up was associated with current workplace bullying, earlier bullying at the present workplace, and earlier bullying in another workplace in women, but not men [3]. However, to the best of our knowledge, the associations of workplace bullying and harassment with headache, stiffness of the neck or shoulders, lumbago, and pain of two or more joints have not yet been investigated. We hypothesized that workplace bullying and harassment are associated with headache, stiffness of the neck or shoulders, lumbago, and pain of two or more joints. Because a previous study reported gender difference [3], we analyzed the data in men and women separately.
Subjects
The subjects in this study were recruited from workers (n = 1,931) at 35 healthcare or welfare facilities in Japan. Questionnaires were mailed to the organizations and distributed to the workers. The purpose and procedure of the survey were explained to the participants in the documents. Written informed consent was obtained from all participants, who were not compensated for their participation. A total of 1,642 questionnaires were returned in sealed envelopes (response rate, 85.0%). Because of missing data, the number of subjects included in the analysis varied (Tables 1-4). This study was approved by the ethics committee of the Okayama Prefectural University.
Measures
From August to September, 2009, participants completed a self-administered questionnaire including background information such as age, gender, marital status, employment status, and work shift as well as measures of workplace bullying and harassment, depression, and pain.
Workplace bullying and harassment were assessed using the Negative Acts Questionnaire (NAQ) [4]. The NAQ is a self-administered questionnaire originally developed by Einarsen and Raknes that measures exposure to specific negative acts typical of bullying [4]. It contains items that refer to both direct and indirect behaviors, but it does not require respondents to label themselves as targets of bullying. Respondents indicated on a five-point scale (1 = never, 2 = now and then, 3 = monthly, 4 = weekly, 5 = daily) whether they have experienced the designated negative acts in the context of their job [4]. The NAQ was translated into Japanese using a back-translation method, and the internal consistency reliability and factor and construct validity are reportedly acceptable [1]. A cross-validation study revealed three subscales of the NAQ: person-related bullying (six items such as "Gossip or rumors about you"; score range, 6-30), work-related bullying (three items such as "Someone withholding necessary information so that your work gets complicated"; score range, [3][4][5][6][7][8][9][10][11][12][13][14][15], and sexual harassment (three items such as "Unwanted sexual advances"; score range, 3-15) [1].
The frequency of pain (headache, stiffness of the neck or shoulders, lumbago, and pain of two or more joints) experienced by workers in the previous month was evaluated using a four-point scale (1 = almost never, 2 = sometimes, 3 = often, 4 = very often). Participants who answered "2," "3," or "4" for a particular symptom were categorized as having that symptom; participants who answered "1" for a particular symptom were categorized as not having that symptom. Depression was evaluated using a self-reported Brief Job Stress Questionnaire (BJSQ) published in a research report on stress in the workplace and its impact on workers' health [5]. The BJSQ was developed in Japan with the support of the Japanese Ministry of Labour and has been widely used in Japan to evaluate work-related stressful situations in various clinical and occupational settings [6][7][8][9][10]. In the BJSQ, six items, such as "I feel sad," are used to measure workers' depression. The response option is based on the frequency in the previous month and is scored using a four-point scale (1 = almost never, 2 = sometimes, 3 = often, 4 = very often). We calculated a total score (range, 6-24) for the six items as a measure of depression. Cronbach's α for the depression scale has been reported as 0.86 [5] and was 0.89 in the present study.
Statistical Analyses
Variables of age and depression were compared using the unpaired t-test, variables of workplace bullying and harassment were compared using the Mann-Whitney U test, and categorical variables were compared using the chi-square test. SPSS version 20.0 were used for these tests. Based on the recommendation of Barros and Hirakata, prevalence ratios (PRs) and their 95% confidence intervals (CIs) were calculated by Cox regression with constant time at risk and robust variance using Stata/SE 10 [11]. All tests were two-tailed, and statistical significance was set at p < 0.05.
Results
Participant characteristics according to gender are shown in Table 1. On average, men were significantly younger than women. Scores for person-related bullying (mean score, 8.3 for men versus 7.8 for women) and sexual harassment (mean score, 3.38 for men versus 3.25 for women) were significantly higher in men than in women. Marital status, employment status, and work shift were significantly different between men and women. Significantly more women than men experienced each type of pain. Changes in the PR associated with a 1-point increase in the person-related bullying score are shown in Table 2. Headache was significantly positively associated with person-related bullying bivariately and after adjustment for demographics in both genders, and after adjustment for demographics and depression only in women. Stiffness of neck or shoulders was not significantly associated with person-related bullying. Lumbago was significantly positively associated with person-related bullying bivariately, after adjustment for demographics, and after adjustment for demographics and depression in women, but not in men. Pain of two or more joints was significantly positively associated with person-related bullying bivariately and after adjustment for demographics in both genders.
Changes in the PR associated with a 1-point increase in the work-related bullying score are shown in Table 3. Headache was significantly positively associated with work-related bullying bivariately, after adjustment for demographics, and after adjustment for demographics and depression in both genders. Stiffness of neck or shoulders was significantly positively associated with work-related bullying bivariately and after adjustment for demographics in both genders. Lumbago was significantly positively associated with work-related bullying bivariately, after adjustment for demographics, and after adjustment for demographics and depression in women, but not in men. Pain of two or more joints was significantly positively associated with work-related bullying bivariately and after adjustment for demographics in both genders.
Changes in the PR associated with a 1-point increase in the sexual harassment score are shown in Table 4. Headache and stiffness of neck or shoulders were significantly positively associated with sexual harassment bivariately and after adjustment for demographics only in women. Lumbago and pain of two or more joints were significantly positively associated with sexual harassment bivariately and after adjustment for demographics in both genders, and after adjustment for demographics and depression only in women.
Discussion
Many of the associations of person-related bullying, work-related bullying, and sexual harassment with pain were positive and significant. The point estimates of the PRs in men and women were similar, and their 95% CIs in men and women overlapped each other. We found significant associations more frequently in women than in men probably because of relatively large sample sizes in women. After adjustment for depression, the degrees of the associations of workplace bullying and harassment with pain decreased, but some of the associations remained significant.
Workplace bulling and harassment can cause depression [2,12,13]. Patients with depression are likely to report somatic symptoms [14]. Nakao and Yano reported an association between depression and somatic symptoms in Japanese workers [15]. Psychological distress has been reported to be a strong risk factor for neck pain in several studies [2]. Thus, depression might be an intermediate factor, rather than a confounding factor, in the association of workplace bullying and harassment with pain [16]. By adjustment for a possible intermediate depression, a direct effect of workplace bulling and harassment on pain, an effect of workplace bulling and harassment on pain that is not mediated by depression, could be calculated [16]. Depression can also be affected by pain [2]. If depression was affected by pain, we could not regard depression as a confounding factor either, interpretation of adjustment for depression should be made with caution, and adjustment for depression could underestimate the true associations of workplace bullying and harassment with pain [16].
Even after adjustment for depression, some of the associations of workplace bullying and harassment with pain remained significant. Mechanisms other than just depression appear to play a role in the relationship of workplace bullying and harassment with pain. For example, the relationship between corticosteroids and stress is well known. Recent scientific and clinical evidence has demonstrated the direct role that steroids play in the generation of chronic pain [17]. Stress reactions caused by bullying or harassment may affect health by direct biological effects, prolonged physiological activation and lack of restitution, or compromised lifestyle and health behaviors [18].
The strength of this study was that we used a relatively large sample of workers to obtain reliable results. However, we must also note several limitations. First, we need to be cautious in interpreting causality in our results because the study used a cross-sectional design. Second, because we used convenience sampling, the results may not be applicable to the entire workforce. Many of the participants were professional caregivers and working with night shift. However, because we included workers from 35 various facilities and obtained a response rate of at least 77.1%, some generalizability can be expected. For example, more women than men experienced each type of pain. This is consistent with previous studies on neck pain [2,13]. Third, the observed variables were self-reported. More objective measurements are needed in future studies.
The present study supports the association of workplace bullying and harassment with pain. If the causality from workplace bullying and harassment to pain was confirmed, to help prevent pain in workers, measures to prevent workplace bullying or harassment should be considered. Prevention may include the introduction of occupational guidelines against bullying and harassment, active monitoring of specific workplace bullying or harassment, and taking action to deal with bullying using criminal, civil, social, or occupational laws [19]. Other measures may also be necessary, such as improving workplace social support, which can reduce workplace bullying [1,20,21].
Conclusions
Many of the associations of person-related bullying, work-related bullying, and sexual harassment with headache, stiffness of the neck or shoulders, lumbago, and pain of two or more joints were positive and significant. Even after adjustment for depression, some of the associations remained significant. | 2016-03-22T00:56:01.885Z | 2013-09-25T00:00:00.000 | {
"year": 2013,
"sha1": "a8b1745fc296fb21207285ff277708a9ec5c6655",
"oa_license": "CCBY",
"oa_url": "http://www.mdpi.com/1660-4601/10/10/4560/pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "a8b1745fc296fb21207285ff277708a9ec5c6655",
"s2fieldsofstudy": [
"Psychology",
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
79788770 | pes2o/s2orc | v3-fos-license | Retroauricular Thyroidectomy Using a New Flexible , Single-Port Robotic Surgical System
최근 갑상선암의 발생률이 눈에 띄게 증가하면서 환자와 의료진의 관심이 많아지고 있고, 이러한 현상에 대한 연구도 많이 늘어나고 있다. 이렇게 증가되고 있는 갑상선암의 발생 률은 초음파를 통한 갑상선 검진의 대중화와 연관이 있다. 대부분의 갑상선암은 유두암(papillary thyroid cancer)으로 적절히 치료하면 이환률, 사망률과 재발률이 낮아 예후는 아 주 좋은 편이다. Emil Kocher가 1800년대 후반에 개방성 경부 절개 갑상 선 절제술을 고안하고 표준화한 후에 이 술식은 갑상선 절제 술의 주류로 인정받아 왔다. 이러한 개방성 갑상선 절제술 은 경험 있는 외과의에 의해서 행해지면 안전하고 효과적이라 고 평가받아 왔지만, 수술 후 생기는 경부의 상처에 대해 미 용적으로 불만을 가지는 환자들도 많다. 이러한 불만을 극복하기 위해 다양한 수술 접근법이 개발 되어 왔다. 최소 침습 비디오 보조 갑상선 절제술(minimally invasive video-assisted thyroidectomy)은 목에 2 cm의 횡 방향 절개를 하고 비디오 내시경을 보면서 수술하기 때문에 박리 범위와 조직 손상, 통증이 적어 회복 기간이 짧다. 이 접근법은 기존 방법에 비해서 절개 길이가 짧지만 켈로이드 체질이나 비후성 반흔의 소인이 있는 환자들에서는 여전히 눈에 띄는 상처가 생길 수 있다는 단점이 있다. 이 술식과는 다르게 경부에 상처를 전혀 안 내기 위해서 원위부에서 접근 (remote access)하는 내시경적 또는 로봇 접근법으로는 경액 와(transaxillary), 경유방(breast), 경액와-유방(axillo-breast), 후이개(retro-auricular), 경구강(transoral) 접근법 등이 있다. Retroauricular Thyroidectomy Using a New Flexible, Single-Port Robotic Surgical System
This study aimed to assess a new flexible, single-arm robotic surgical system to retroauricular thyroidectomy.Three fresh cadavers were used.Technical elements of the system and the whole surgical procedures were described in detail.This single-port flexible system could be used to successfully perform retroauricular thyroidectomy.The ideal angle to dock the patient-side cart was at a 90-degree angle to the operating table.When the cannula tip was placed 10-15 cm away from the skin incision, positioning and full movement of all four instruments without collisions were possible.Flexible three instruments and a stereoscope made the robotic dissection more efficient, safe and time-saving.We report the first preclinical evaluation of an innovative, flexible, single-arm robotic surgical system for retroauricular thyroidectomy.
Fig. 1 .
Fig. 1.External view of the flexible single-arm robotic surgical system.Four robotic manipulators (red star) take control of three En-doWrist Sp (Intuitive Surgical, Inc.) instruments and one flexible endoscope within the single instrument arm (red circle).Four instruments are delivered to surgical site through single port (cannula, red arrow).
Fig. 3 .
Fig. 3.The external view after docking of the system.The cannula tip was placed 10-15 cm away from the skin incision line.The flexible three instruments and endoscope moved freely without collisions during dissection at this distance.
Fig. 2 .
Fig. 2. The surgical console view during retroauricular thyroidectomy using the flexible single-port robotic system.The new graphical user interface in the black box in the lower mid part of stereoscopic display of the surgeon console shows the relationships of the four instruments in a real time.
Fig. 4 .
Fig. 4. The surgical view during retroauricular thyroidectomy using the flexible single-port robotic system.Through the right retroauricular incision, right thyroid lobectomy was performed.The right thyroid gland was retracted medially and anteriorly exposing the recurrent laryngeal nerve and Berry's ligament with a Pro-Grasp forceps.A monopolar cautery and a Maryland bipolar forceps were used to dissect the thyroid gland with care not to damage the recurrent laryngeal nerve.RLN: recurrent laryngeal nerve.
Fig. 5 .
Fig. 5.The surgical view during retroauricular thyroidectomy using the flexible single-port robotic system.Through the right retroauricular incision, left thyroid lobectomy was performed after right lobectomy had been finished.The left thyroid gland was retracted laterally and anteriorly exposing the recurrent laryngeal nerve with a ProGrasp forceps.During dissection of the thyroid gland with a monopolar cautery, a Maryland bipolar forceps were used to retract the cricoid and trachea posteriorly.RLN: recurrent laryngeal nerve. | 2019-03-17T13:10:34.584Z | 2017-12-21T00:00:00.000 | {
"year": 2017,
"sha1": "bada64fb608b25048d679f304c10d6da24b4dbb6",
"oa_license": "CCBYNC",
"oa_url": "http://kjorl.org/upload/pdf/kjorl-hns-2017-00115.pdf",
"oa_status": "GOLD",
"pdf_src": "Anansi",
"pdf_hash": "bada64fb608b25048d679f304c10d6da24b4dbb6",
"s2fieldsofstudy": [
"Physics"
],
"extfieldsofstudy": [
"Medicine"
]
} |
11407511 | pes2o/s2orc | v3-fos-license | Helper or cytolytic functions can be selectively induced in bifunctional T cell clones.
By using bifunctional T cell populations, we have shown in this report that elicitation of helper versus cytolytic function depends on the stimulatory signal at the membrane. Interestingly enough, the transduction of these signals is likely to be achieved via different metabolic pathways. Thus, helper function is associated with intracellular Ca2+ mobilization and PLC activation, while cytolysis can occur even in the absence of detectable levels of these second messengers. These results indicate that selective activation through the same membrane-transducing molecule may orientate T cell function through qualitatively or quantitatively different second messengers. This would be an important part of immune regulation.
T he two major T cell activities, i.e., helper and cytotoxic functions, are supported by two separate sets of lymphocytes in correlation with CD4 or CD8 expression, respectively. Exceptions to this rule have been reported in different experimental conditions providing evidence for CD4+ T cell clones displaying both helper and cytotoxic activities (1-2) . However, it is not clear whether both activities are simultaneously expressed or can be selectively induced by distinct activating signals. Using different combination pairs ofanti-CD2 mAbs for T cell activation, we demonstrate that it is possible to induce preferentially one of the two activities. Moreover, helper function is associated with calcium influx and phosphoinositide hydrolysis, whereas cytolysis can be observed in the absence of detectable amounts of these second messengers. This indicates that depending on the mode of activation through a same membrane transducing molecule, a single T cell can express either helper or cytotoxic function . This selective orientation might involve qualitative or quantitative differences in second messenger induction .
Materials and Methods
T Cell Clones. T cell clones were produced as previously described (2-4) . They are all helper and cytotoxic, specific for the influenza A/Texas virus, and restricted by HLA-DR molecule (2-4, and unpublished observations) . They were maintained in long-term culture by weekly restimulation with PHA and irradiated feeder cells in the presence of rUr2, as previously described (2-4). In all experiments described herein, T cells were used 7 d after their last restimulation .
T Cell Proliferation Assays. Cells were cultured in triplicate in 96-well round-bottomed microtiter plates, with various stimulators, as indicated, in a total volume of 200 td of culture medium for 3 d at 37°C, in 5% COZ atmosphere. During the last 6 h ofculture [3H]thymidine (0.8 P.Ci) (Amersham, les Ulis, France) was added.
T Cell Cytotoxicity Assays. Cytotoxicity was assessed by the standard "Cr-release assay described elsewhere (3). Briefly, effector cells (5 x 104 cells per well) were seeded into round-bottomed microtiter plates, in RPMI 1640 medium supplemented with 10% FCS. mAbs were added 20 min before addition of target cells . "Cr-labeled K562 target cells (5 x 103 cells per well) were added in a total volume of 200 P.1. After centrifugation at 100 g, cells were incubated for 4 h at 37°C in 5% COz atmosphere. Plates were then centrifuged again. 100 itl of supernatant was harvested from each well and the radioactivity was measured using a gamma counter.
Fc Rosettes and Inhibition of Binding by GT2, D66, and T21 Hybridoma Proteins. Rosette assays between IgG-coated SRBC and cells to be investigated have been previously described in detail (13) . Briefly, SRBC were coated with two different BALB/c mouse mAbs directed to SRBC (U182 .5, IgGl ; UN2, IgG2a) kindly provided by Dr. M .D. Scharff (Albert Einstein College of Medicine, New York, NY), and tested for rosette formation with K562 cells. Cells associated with at least five SRBC were scored as FcyR* . In another set of experiments, hybridoma proteins were purified from ascitic fluids by passing them over protein A-Sepharose columns (LKB-Pharmacia, Uppsala, Sweden) . ELISAs were performed as previously described (14) to control the purity of these preparations. Hybridoma proteins were then heat-aggregated by incubation at 60*C for 20 min . Pellets of 2 .5 x 105 K562 cells were incubated for 45 min at room temperature with 100 jal of heat-aggregated proteins diluted in PBS. IgGl-or IgG2a-coated SRBC were then added in 0.4 ml of PBS and rosettes were counted after 10 min centrifugation at 900 rpm at 4°C. Measurement of[Ca2*]i. T cells were washed and suspended at 5-7 x 106/ml in culture medium containing 3 IAM acetoxymethyl ester of indo 1, according to the method of Rabinovitch et al . (15) . Briefly, after incubation with indo 1 for 30 min at 37°C to effect loading, cells were washed, resuspended in the same medium at 106/ml, and analyzed on a cytofluorograph 50 HH cell sorter equipped with a 2150 computer (Ortho Pharmaceutical, Raritan, NJ). Analyses were conducted at 37°C at a flow rate of 500 cells. Cells were then transferred to spectrofluorometer cuvettes and after establishing the baseline [Cal*];, anti-D66+T111 mAbs (a) or anti-GT2+T11, mAbs in the presence of K562 cells (b) were added . PI Turnover. Effector cells were washed twice in a phosphatefree buffer containing 20 mM Hepes, pH 7 .2, 150 mM NaCl, 1 mM MgC12, 5 mM KCI, and 0 .1% Glc. The final pellet was resuspended in the same buffer and the cell suspension (10'/ml) was incubated with carrier free '2 P orthophosphoric acid (Amersham, 20 p.Ci/ml) for 1 .30 h at 37°C, until isotopic equilibrium . Cloned T cells as well as target cells were then washed twice ; 100-Al aliquots of the effector cell suspension were transferred into polypropylene tubes and after adding or not adding the K562 cell suspension (105 /tube), effector cells were stimulated with anti-CD2 mAbs for different time periods. Then, reaction was stopped as previously described (16). Radiolabeled phospholipids were located by autoradiography and their nature was determined by parallel migration of standard phospholipids visualized by exposure to iodine vapors . Spots concerning PI+PA were subsequently scraped from the plates and evaluated by liquid scintillation counting .
Results and Discussion
Two bifunctional human T cell clones, i .e., TA4 and M3ap20 specific for the influenza virus associated with HLA class II molecule and isolated from two different donors, were selected for the present study. As previously demonstrated, these monoclonal T cell populations exhibit a CD2''/3*/4+/8-phenotype and have the same specificity and restriction for both 214 Helper or Cytolytic Functions Induced in T Cell Clones helper and cytolysic activities (2-4) . Activation of T cell clones can be induced by mAbs to membrane proteins such as CD3 or CD2 (17)(18)(19) . In the case of CD2, combination pairs of mAbs are required, such as D66+Tl1, and GT2+T111, which recognize respectively D66 and T11, or GT2 and T11, epitopes on CD2 molecules (5-7) . Interestingly, stimulation of our T cell clones with the first combination, D66+T11,, was clearly more effective than GT2+Till to induce helper function. This observation led us to investigate whether expression of cytolysic function in these bifunctional clones would be reciprocally more efficiently induced by GT2+T11 1 combination . Consistent with this hypothesis, Fig . 1 shows that cytotoxic activity of clone TA4, as assessed by short-term cytolysis assay on the nonspecific K562 cell line, is preferentially triggered with anti-GT2+Tll, mAbs, whereas helper function, as assessed by proliferation assay, is preferentially triggered with anti-D66+Tl1, mAbs. It must be noticed that identical results were obtained with all T cell clones tested, i.e., TA4, M3ap20, and three more clones (data not shown) . This dichotomy was still observed in doseresponse experiments since, as illustrated with clone TA4 in Fig . 2, helper function was preferentially triggered with D66+T11, and K562 lysis preferentially with GT2+T11,, whatever the concentrations of purified mAbs used . Moreover, the opposite effects observed on the cytolysic function of our T cell clones with the two combination pairs of mAbs used cannot be related to a difference in the binding ability of GT2 (IgGl) and D66 (IgG2a) to the FcyRII present on K562 cells. Indeed, these cells express FcyRII that bind immune-complexed IgGl and IgG2a, as shown by their ability to form rosettes with both IgGl-and IgG2a-coated SRBC (Table 1). Furthermore, heat-aggregated GT2, D66, and Tl11 proteins were able to inhibit the rosette formation between K562 cells and IgG2a-or IgGl-coated SRBC (Table 1) . In these functional assays, helper and cytolytic activities were assayed following different experimental conditions . Indeed, proliferation was tested after 3 d of culture in the absence of K562 cells, while cytolysis was analyzed in the presence of the labeled targets, and after a short 4-h period of stimulation . Therefore, these assays did not directly investigate whether both activities could be simultaneously expressed in a single T cell upon a strictly identical mode of stimulation . Thus, helper function was tested in short-term assay (5 vs. 4 h for cytolysis) by analyzing IIr2, ID3, IIr4, and ID2R mRNA accumulation. Moreover, K562 cells, which are required for cytolysis, were added at the same E/T cell ratio . As illustrated in Fig. 3 with clone M3ap20, when both functional assays were performed under strictly identical conditions of stimulation, GT2+T11 preferentially induced cytolysis; conversely, D66+T111 selectively induced helper function. These 215 Eljaafari et al . Rosette assays between K562 cells and IgGl-or IgG2a-coated SRBC and inhibition of binding by hybridoma proteins were performed as described in Materials and Methods. 3 Heat-aggregated hybridoma proteins were added at a final concentration of 20 ug/ml. results strongly suggest that induction ofeither function depends upon different signals delivered through the same transducing CD2 molecule .
Concomitantly, we explore the second intracellular messengers induced by the two mAbs pairs. T cell helper function induction via CD2 molecule (20) or CD3/Ti antigen receptor complex (TCR) (21) has already been described to be associated with membrane phosphoinositide-derived second messengers : inositol 1, 4, 5 triphosphate (IP3) and diacylglycerol (DG). These two second messengers are the products ofa rapid, transient hydrolysis of membrane phosphatidylinositol 4, 5 biphosphate (PIP2) by phospholipase C (PLC) and act synergistically to activate protein kinase C (PKC), leading to protein phosphorylation and modulation of gene expression. IP3 is responsible for Cal+ mobilization from intracellular stores, whereas DG increases enzyme affinity for, Cat+ (22)(23)(24) . Phosphatidylinositol (PI) turnover induced by PIP2 hydrolysis, resulting in phosphatidic acid (PA), and PI accumulation has been described to be associated with extracellular Ca2 + influx (25)(26)(27) . In contrast, PKC activation, in the absence of free cytosolic [Ca2 +]i and PIP2 hydrolysis, has been reported to be absolutely required for cytotoxic triggering. Indeed, PMA can trigger T cell-mediated cytolysis with no increase in free cytosolic [Ca2+]; (28).
In our model of bifunctional T cells stimulated via CD2 molecule by anti-D66+T111 or anti-GT2+Till mAbs, cytosolic free Ca2+ increment was investigated by measuring fluorescence with the indo-1 indicator, and phospholipid hydrolysis was determined by analyzing PI+PA accumulation . As shown in Fig . 4 a, the helper function of M3ap20 clone triggered with anti-D66+Tll mAbs was accompanied by [Ca2 +]i increase. Interestingly, and as also reported by Alcover et al. (29) in other human T cell clones, a several minute latency in [Ca2 +]i rise after mAb addition was observed . This contrasts with the rapid increase observed in a Jurkat T cell line stimulated through CD2 molecule, as reported by Pan- 6) . Culture conditions are identical as those described in Fig. 1 . Then, total cellular RNA was extracted as detailed in Materials and Methods and analyzed by Northern Blotting. 10 Etg of RNA were transferred on nylon membranes and hybridized with 11,2, ID3, 11,4, IIJ2R riboprobes. (B) In the same experiment, the cloned T cells were tested for cytolytic activity under the same conditions of stimulation. Results are expressed as percentage of specific cytotoxicity calculated as described in Fig . 1. (Asterisk) .
In the absence of K562 cells, lysis could not be tested . taleo et al . (20). In both clones, helper function was associated with PIP2 hydrolysis as determined by PI +PA accumulation and as illustrated with clone TA4 in Fig. 5 . This is in good agreement with other reports, indicating that [Ca2 +]i rise and phospholipid hydrolysis are associated with acquisition of T cell helper function through CD2 stimulation (20,29) . In contrast, cytotoxic activity triggered with anti-GT2+T11 mAb, even in the presence of K562 cells, was not associated with detectable changes in [Ca2+], (Fig . 4 b) nor with or anti-D66+T111 in the presence or absence of K562 cells, radiolabeled phospholipids were extracted and located by autoradiography.
Results are presented as the cpm of (PA+PI) 32PO4 incorporation .
PI +PA accumulation (Fig. 5). It must be noticed that chronic treatment of our bifunctional T cell clones with phorbol ester abolished their ability to express helper and cytotoxic functions when subsequently stimulated by anti-CD2 mAbs (data not shown) . Since this treatment is known to inactivate PKC (30,31), one may assume that PKC activation is involved in the induction of both functions. Altogether, these data suggest that different second messenger metabolic pathways are triggered, depending on the combination pairs of anti-CD2 used for stimulation . This clearly correlates with subsequent expression of selective functions. Nevertheless, one cannot rule out that the GT2+T111 combination might be less effective at increasing steady-state levels of phosphoinositide turnover and Cat+ influx, both detected by relatively low sensitive methods. In that case, the same metabolic pathways would be involved but the threshold required for cytolytic function induction would be lower than that for helper function . Consistent with this hypothesis, we recently demonstrated that a secondary signal provided by Ilr2 in addition to GT2+Tllt can lead to helper function (16,32) and PA+PI accumulation (16). In addition, it must be noticed that in the present study, K562 cells in association with anti-GT2+Tllt provide a secondary signal leading to some proliferation (Fig. 1) associated with low I1r2, IL4, I1r2R mRNA accumulation (Fig . 3, lane 3) .
Finally, the induction of helper but not cytotoxic function after D66+T111 activation strengthens the assumption that GT2+T11 1 or D66+T111 mAbs trigger different functions through two distinct metabolic pathways. Indeed, ifthe low levels of GT2+T111-activated second messengers are sufficient to trigger cytotoxic activity but not helper function, the strong [Ca2 +]i and PA+PI increment observed after 1366+Tllt activation could not account for the absence of cytolytic activity. Thus, one hypothesis would be that the 1366+T11t combination would inhibit cytolytic activity. This is quite unlikely, since, as shown in Table 2 with the M3ap20 clone, D66+T11t combination, which does not by itselfinduce any cytolysis, is unable to abolish the CD3/TCR mediated cytolysis, even though a slight inhibition is occasionally observed .
Even if the exact biochemical mechanisms by which CD2 can differently orientate T cell functions are not fully understood, one may speculate that this phenomenon could be an important part in the human in vivo immune regulation, by directing the response towards cellular and/or humoral immunity. Although experiments are more difficult to carry out, further studies are now in progress to determine whether a similar dichotomy exist in TCR/CD3-mediated T cell activation .
We thank Michel Seman for helpful comments on the manuscript, Sylvie Benslama and Ms. J. Moncuit for excellent technical assistance, Isabelle Rivenez for typing the manuscript . | 2014-10-01T00:00:00.000Z | 1990-07-01T00:00:00.000 | {
"year": 1990,
"sha1": "cdcb303e5c9a18086575709a57e1e56778cc0457",
"oa_license": "CCBYNCSA",
"oa_url": "http://jem.rupress.org/content/172/1/213.full.pdf",
"oa_status": "BRONZE",
"pdf_src": "PubMedCentral",
"pdf_hash": "cdcb303e5c9a18086575709a57e1e56778cc0457",
"s2fieldsofstudy": [
"Biology"
],
"extfieldsofstudy": [
"Medicine",
"Biology"
]
} |
184485620 | pes2o/s2orc | v3-fos-license | Screening and enzymatic activity of high-efficiency gellan lyase producing bacteria Pseudoalteromonas hodoensis PE1
ABSTRACT Gellan is a widely used microbial polysaccharide and one of the more effective ways to expand its application value would be to investigate the mechanism of gellan lyase and to produce gellan oligosaccharide. In this study, efficient gellan degrading bacteria were screened. One of the strains with high efficient gellan degradation capacity was labeled PE1. Through physiological and biochemical analysis of 16S rDNA, the species was identified as Pseudoalteromonas hodoensis. The optimum conditions for enzymatic activity and how it was affected by metal ions were determined, and the results showed that the lyase activities were much higher than those of previously reported (about 20 times). The gellan degradation products were determined by thin-layer chromatography and the oligosaccharides were determined by high-efficiency liquid chromatography to analyze the action site of lyase. This study laid a solid foundation which elucidates the production and application of gellan oligosaccharides. Research highlights ● High efficiency gellan lyase producing bacteria ● Optimization of reaction conditions for gellan degradation ● Oligosaccharides were detected by TLC and HPLC to speculate the lyase action sites.
Introduction
Gellan gum is a linear acidic heteropolysaccharide produced by aerobic fermentation of Sphingomonas elodea. It is a microbial extracellular polysaccharide with unique physical and chemical properties as well as excellent application performance [1]. Gellan gum can be used as a new type of emulsifier, suspension agent, thickener, stabilizer, gel agent, sustained-release agent, film-forming material, and many more [2]. Gellan and its products can also be applied in the fields of food, medicine and chemical industry [3]. In order to investigate the mechanism of gellan lyase and to produce gellan oligosaccharide, would be one of the more effective ways to expand its application value [4]. Therefore, we chose to use efficient gellan lyase to improve its molecular structure and expand its application range.
Presently, two distinct reactions catalyzed by the enzyme are mainly used for the enzymatic depolymerization of polysaccharide chains: hydrolysis and lytic β-elimination(a common lysis mode of the acid polysaccharides) [5]. β-eliminative cleavage of a glycosidic bond occurs when the sugar is substituted with an acidic group next to the carbon forming a glycosidic bond which results in the breakage of the glycosidic bond adjacent to the carboxylate group, forming a new reducing end with the unsaturated sugar at the non-reducing end [6]. Because of this, the oligosaccharides of lysis products were endowed with new physiological activities. The chemical steps of this mechanism were proposed by Gacesa in 1987 [7]. According to previous studies, gellan is composed of linear tetrasaccharide repeating units of β-D-glucuronic acid, β-D-glucose,α-L-rhamnose, and β-D-glucose [8,9]. Based on the related acidic polysaccharides degradation mechanism, the gellan lyase breaks down the 1,4-glycosidic bonds associated with the β-D-glucuronic acid, forming unsaturated oligosaccharides. This resulted in reduced viscosity and molecular weight of gellan.
Currently, the industrialized degradation methods commonly used for gellan are irradiation and acid degradation, whereby the molecular weight of gellan is reduced by physical or chemical means, respectively [10]. However, these methods are in general limited by slow degradation rate, complexity and high cost of purification for the degradation products. Thereby using gellan lyase is the most ideal degradation method [3]. Furthermore, the reaction of gellan lyase can easily be monitored in real time, with the possibility of controlling the reaction and adjusting the relevant conditions during the reaction, so as to obtain the corresponding molecular weight of gellan gum for enzymatic β-elimination of oligosaccharides [11]. Gellan oligosaccharides also have antibacterial [12], antioxidant [13], and intestinal prebiotic [14] effects, which are important for the expansion of the application value of polysaccharides. However, the activity of related gellan lyase has not been strong. Therefore, this study screened for gellan lyase producing microorganisms which have a high efficient gellan degradation ability from abundant microbial resources in order to facilitate the degradation of gellan. Furthermore, we completed identification, comparative analysis and the evaluations of enzyme reaction conditions to perfect the related research of lyase.
Screening of gellan lyase producing bacteria
In order to isolate gellan-degrading bacteria, algae samples were collected from the South China Sea. The algae had obvious gelatinous substance, so we speculated that there might be bacteria capable of degrading polysaccharides. After the algae sample was collected manually, it was immediately stored at low temperature to complete strain screening. After the 1 gram of algae sample had been thoroughly ground in the mortar, 9 ml of sterile water was used to dissolve it fully. Then, 1 ml of dissolved liquid was mixed with 9 ml defined medium consisting of 0.1% (NH 4 ) 2 HPO 4 , 0.02% KCl, 0.01% MgSO 4 and 0.2% gellan, with gellan as the sole carbon and energy source. Samples were then cultivated at 37°C, at a pH of 7.0 for 24 h with shaking at 240 rpm. Afterward, 100 μl from flasks with registered growth were plated on Peptone-Yeast extract (PY) medium (0.2% peptone, 0.1% yeast extract) supplied with 0.5% gellan and incubated at 37°C for 5 days with gellan as the coagulant [15]. On the 5 th day, colonies with the visible formation of pits were isolated. Using a sterile inoculating tube, a small amount of the pits was streaked on fresh media. This was done at least three times to obtain pure single colonies for pure cultures which were further used for the study.
Physiological and biochemical identification of gellan-degrading isolates
The identification of the isolates was carried out by 16S rDNA sequence analysis. Firstly, DNA was extracted using a Chelex-100 genome kit. Amplification of 16S rDNA was then performed with general primers, 27F (5ʹ-AGAGTTTGAT CMTGCTCAG-3ʹ) and 1492R (5ʹ-TACGGYTA CCTTGTTACGACTT-3ʹ). PCR amplification was carried out with pre-denaturation at 94℃ for 3 min, 35 cycles(denaturation at 94°C,1 min; annealing at 52℃,1 min; extension at 72°C,1 min), and a final incubation at 72℃ for 5 min [16]. The purification, cloning, and sequencing of PCR products were completed at the Sangon Biotech (Shanghai) Co., Ltd. The sequencing products were listed on the Ezbiocloud website for homologous comparison [17] and the phylogenetic tree was constructed using MEGA7.0.26 software. Spore staining [18], capsule staining [19], flagellum staining [20] and gram staining [21] were carried out on high activity lyase strains and the specific morphologies of the strains were observed by scanning electron microscope [22]. Biochemical identification tubes were used to complete the tests of methyl red, Volpe test, arginine, ornithine, lysine, ONPG(O-Nitrophenyl-β-D-Galactopyranoside), sucrose, glucose, xylose, hydrogen sulfide, aesculin, urea, and nitrate to determine the physiological and biochemical properties of the bacterial strains [23].
Enzyme assays
Gellan lyase activity was assayed as described by Weissbach and Huritz [24]. Lyase activity was determined by increasing the absorbance continuously at 235nm using a spectrophotometer(Pye Unicam, SP8-500) based on the accumulation of unsaturated glucuronyl ends after cleaving of the glycosidic bond between glucosyl and glucuronyl residues in the molecular structure of gellan [25]. Strains were cultured in 2216 E enrichment medium and incubated at 30°C for 24 h, and then 10% of the inoculum was transferred onto PY fermentation medium for 36h. The fermented liquid was centrifuged at 8000rpm at 4°C for 15 min. The supernatant was then concentrated using a Millipore standard cassette system with an ultrafiltration membrane (MW cut-off 10 kDa) and precooled (4°C) 50 mM Tris-HCl buffer (pH 7.0) to obtain gellan lyase. Gellan lyase (1.5 ml of crude enzyme) was added to 15 ml of 0.05% gellan solution in 50 mM Tris-HCl buffer, pH 7.0 and incubated at 30°C. The absorbance changes of the gellan degradation were determined by Pye Unicam Spectrophotometer at 235 nm. The active unit was defined as the activity of gellan lyase to degrade gellan oligosaccharides to produce double bonds within one hour, thus increasing the absorbance value by 0.001 [15]. This assay was used only for the determination of enzyme activity. All experiments were carried out in triplicate and the average value was generated.
Optimization of reaction conditions for gellan degradation
To determine the optimal conditions for gellan degradation, samples were exposed to varying temperatures and pH and various compounds (see Table 1 for details). For the pH, temperature and various compounds, the detection were in the range of 6-8, 25-40°C and 1mM of various compounds separately in 0.05% gellan solution in 50 mM Tris-HCl buffer respectively. The lyase activity was detected at 235nm by spectrophotometry every 30 min. Using the lyase activity detection method described above, the optimal conditions for gellan degradation were accordingly determined.
Thin layer chromatography analysis
The degrading activity and degradation products of gellan were detected and identified by thin layer chromatography (TLC) [26]. Gellan and its constituent monosaccharides were used as reference materials to compare the positions of degradation products, which were used to verify the corresponding lyase activity. 10 μl of the degradation products mixture was spotted onto a silica gel 60 F 254 plates and developed with a solvent system consisting of n-butanol/acetic acid/water (5:4:1,v/v/v) in a TLC developing tank [27]. The ascending development of the TLC plate was carried out at room temperature for 2h, and then the plate was transferred to the fume hood for natural air drying. The plate was soaked in 1 l of a methanol solution containing 30%(v/v) sulfuric acid and then dried in a 110°C oven for 20 min to visualize the spots [28]. All experiments were carried out in triplicate.
High-performance liquid phase analysis
The lyase products were determined by HPLC to analyze the action site of lyase generally, 100 ml of the crude degradation products mixture passed through the activated carbon adsorption column. After three column volumes were washed with deionized water, 25% ethanol eluent was collected for three column volumes. Then, 25% ethanol eluent through rotary evaporation (110rpm, 55°C to evaporate the alcohol. After that, the eluent passed through the anion exchange column, collected the 50mM NaCl eluent for three column volumes. Demineralization was accomplished by dimensional exclusion chromatography (Dextran gel G-15, Scope of separation <1500Da). Glucan standards and purified products were analyzed by high-performance liquid phase (Younglin Instrument ACME-9000, Seoul, South Korea) equipped with a Sugar KS-802 column (Shodex, Tokyo, Japan) and a RI-detector at 45°C. The column heater was set at 75°C, and deionized water was used as the mobile phase at a flow rate of 1ml/min.
Isolation of gellan lyase producing bacteria
Samples were directly inoculated onto PY medium with gellan as the coagulant and the sole carbon source, and gellan lyase producing bacteria had pits forming as showed in Figure 1. Microbial growth was not easily determined because of the turbidity of the gellan samples. Using a sterile inoculating tube, a small amount of the pits was streaked on fresh media. This was done at least three times to obtain pure single colonies. The subcultures were considered pure after microscopic observation of one type of bacterium per culture. Four strains of gellan-degrading bacteria were purified from the pits of the plates, and one of them was determined to have high efficient gellan degradation capacity by preliminary lyase activity detection and was labeled PE1(NCBI registration number: MK788321).
Identification of gellan-degrading microorganisms
For the identification of the gellan-degrading isolates, 16S rDNA sequencing was carried out. The PCR product amplified by conventional 16S rDNA primers was sequenced and compared with the BLAST program. Four strains of gellan lyase producing bacteria were classified ( Table 2). The isolated PE1 showed high identity to Pseudoalteromonas hodoensis (98.95%).The phylogenetic tree was constructed to confirm the relationship with other Pseudoalteromonas species (Figure 2).
Morphology and physiology
Pseudoalteromonas hodoensis was obtained by experimental screening which produced gellan lyase with a high degradation ability. It formed obvious pits on PY medium, forming opaque, pale yellow round colonies with smooth edges, and moist surfaces with colony size of 2-3.5mm (Figure 3(a)). The bacteria were also without nuclei, spore formation or capsules, and with strictly anaerobic respiratory metabolism, with no fermentation. They were rod-shaped, with an average length of 2.3-3.1μm and diameter of 0.4-0.9μm (Figure 3(b)), and were highly efficient at gellan lyase production to reduce gellan molecular weight and viscosity. They were able to grow in a pH range of 5.5-8.5 at temperatures (20-50°C). The optimal conditions for their maximal growth were pH 7.0 and 30°C. The physiological and biochemical analyses results showed that the strain had no spores, flagella or capsules, and was gram-negative ( Figure 3C). The arginine, ornithine, lysine, citrate and urea tests were positive, while the nitrate reduction, ONPG, sucrose, glucose, hydrogen sulfide, methyl red, MR-VP and indigo matrix tests were negative (Table 3). The enzyme was preincubated for 30 min at 30°C in Tris-HCl buffer, pH 7.0, containing different ions and inhibitors at final concentration 1 mM and the activity was then determined at 30°C, pH 7.0. The difference value was calculated by the difference between absorbance at the initial time 0 min and reaction time at 60 min. At the same time, relative activity was determined by comparing the difference between different ions or compounds and the blank control.
Lyase activity detection
Gellan lyase (1.5 ml of crude enzyme) was added to 15 ml of 0.05% gellan solution in 50 mM Tris-HCl buffer with a pH of 7.0. After incubation at 30°C, the absorbance changes were monitored and recorded every 30 min in a Pye Unicam Spectrophotometer at 235 nm. An activity curve of lyase was then generated from the resulting data. The corresponding analytical curve showed that the light absorption value of the lyase strain at OD 235 increased by 0.824U/ml/h within 24h, which was much stronger than the measured lyase activity of the previously reported Geobacillus stearothermophilus (0.03 U/ml/h) [15].
Optimization of the lyase conditions
To determine the optimal reaction conditions for gellan degradation, the effects of temperature (25-45°C), pH (6.5-7.5), and various compounds were examined. The gellan-degrading products were determined based on the accumulation of unsaturated glucuronyl ends. The absorbance at 235 nm was measured at different temperature ranges every 30 min (Figure 4), the lyase was active in the range of 25-45°C, with an optimum at 30°C. Similarly, the OD 235 was measured at the different pH ranges every 30 min ( Figure 5). From the results, it could be deduced that the lyase was active in the range of 6.5-7.5, with an optimum at 7.0. The effect of the various compounds on gellan lyase activity was also established ( Table 1). In the presence of a number of metal ions in a low concentration (1 mM) a stimulating effect on the lyase activity was observed for almost all tested ions. Metal ions in high concentration (1 M) inhibited lyase activity. The increased activity in the presence of K + , NH 4 +, and Na + at such a concentration was no more than 10%. Mn 2+ had the most promoting effect on PE1 lyase, by up to 138.39%. The lyase was sensitive to all the inhibitors, with the highest degree of inhibition towards urea.
Thin-layer chromatography and high-performance liquid chromatography
Under the optimum culturing conditions determined above, lyase degraded gellan efficiently and the culture supernatant showed single spots on the TLC plates, which was thought to be an oligosaccharide product. On a TLC plate, spots of gellan-degraded products were distinguished from the components of gellan polymer ( Figure 6). Gellan is formed by linear tetrasaccharide repeating units, hence its rise in the TLC plate was extremely slow, but rhamnose and glucose, being monosaccharides, had very fast ascending development. On the TLC plate, the spots of gellan-degrading products were lower than those of rhamnose and glucose, but higher than gellan. This indicated that the molecular structure of gellan was degraded by the action of the lyase. When the reaction products of lyase were analyzed by HPLC. Results displayed that the purified oligosaccharides samples had a single absorption peak at 235nm. Compared with the glucan standard substance in the same condition, we confirmed almost all the degradation products were the minimum enzymatic degradation products(tetrasaccharide units containing double bond) (Figure 7) [29]. Therefore, we speculate that the lyase through β-elimination broke down the 1,4-glycosidic bonds associated with the β-D-glucuronic acid, formed unsaturated oligosaccharides. It was consistent with the degradation mechanism of acidic polysaccharides.
Discussion
Although gellan is of great interest for many commercial applications, the expensive downstream processing impairs the economic viability of gellan production, and consequently, it is the most expensive among the various food gums [30]. Additionally, its highly viscous properties have limited its utility, particularly in the food industry [31]. Therefore, we chose to use efficient gellan lyase to improve its molecular structure and expand its application range. Emphatically, the newly isolated strain of Pseudoalteromonas hodoensis is the first reported gellan lyase producing bacteria in marine Pseudoalteromonas. It has very high efficiency in degrading gellan gum compared with the bacteria producing gellan lyase obtained in previous studies, such as in the work by Derekova et al (2006) on Geobacillus stearothermophilus(crude lyase 0.03U/ml/h vs 0.824U/ml/h) [15]. At the same time, the optimum action conditions of lyase were 30°C and pH 7.0, so environmental impacts are mild. And the oligosaccharide molecular weight produced by gellan lyase was single, this indicates that the lyase can degrade gellan specifically and the purification cost is very low. Furthermore, through bio-enzymatic degradation, its reaction is controllable with low energy consumption and high lyase yield [32]. By comparison, it has great advantages over the existing industrial degradation methods. Therefore, it has very important applications in fermentation processes as well as in biomass energy utilization [33].
The application of degradation products in new fields will provide a new direction for the market application of polysaccharides [34]. Such as Salina et al. have found that gellan oligosaccharides can promote the growth and antibacterial activity of Eucommia ulmoides [35]. Polysaccharide lyases are believed to provide tremendous commercial benefits, but their production at industrial levels still remains a challenge. At present, the lack of suitable high activity lyase is the biggest obstacle for its application. The ideal lyase should be characterized by high activity, mild operating conditions, and a single product. Therefore, we believe that lyase from Pseudoalteromonas hodoensis met these . HPLC analysis of the gellan degradation products. Gellan was converted to a tetrasaccharide (4,5-ene-glucuronyl-glucosylrhamnosyl-glucose) by the gellan lyase. criteria and is suitable for a wide range of industrial applications.
Through preliminary detection, we can conclude that the lyase obtained from Pseudoalteromonas hodoensis has strong ability to degrade gellan. In addition, the preliminary analysis of the active site of lyase has also been completed. As part of our future works, we will be purifying and collecting lyase to detect its yield, and the whole genome of the strain will be sequenced to analyze the relevant genes associated with lyase production. This is a preliminary exploration on our part to obtain lyase in large quantities for subsequent oligosaccharide preparation and industrial application. A promising way might be cloning the genes, and this is also part of our follow-up research. We would also like to continue developing the application fields of gellan oligosaccharides by detecting the antibacterial, antioxidant and intestinal probiotics activities of the oligosaccharides. The application of polysaccharide derivatives has been greatly enhanced by the broadening research on polysaccharide lyases and their products. | 2019-06-12T14:56:11.486Z | 2019-01-01T00:00:00.000 | {
"year": 2019,
"sha1": "9496fffdc5999ff3151a24e5b0731b8656627ee2",
"oa_license": "CCBY",
"oa_url": "https://www.tandfonline.com/doi/pdf/10.1080/21655979.2019.1628882?needAccess=true",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "ff3edb9af4d12a9d1129b66dcd3fe59f2a2b9944",
"s2fieldsofstudy": [
"Biology",
"Environmental Science",
"Chemistry"
],
"extfieldsofstudy": [
"Chemistry",
"Medicine"
]
} |
125834649 | pes2o/s2orc | v3-fos-license | 1/2 spin-isospin fermions close to the unitary limit
The equal mass three-fermion system having 1/2 spin-isospin symmetry is study around the unitary limit. The two body system has two different scattering lengths, a0 and a1, corresponding to the spin singlet state S = 0 and the spin triplet state S = 1, respectively. The unitary limit is defined when the two quantities a0, a1 → ∞. The three-nucleon system is located very close to this limit, the singlet and triplet n − p scattering lengths are large with respect to the range of the nuclear interaction. The ratio of the two is about a0/a1 ≈ −4.31. This value defines a plane in which a 0 and a 1 can be varied. Using a nucleon-nucleon spin dependent potential with variable strength it is possible to study the behavior of the three-nucleon binding energy along that plane. This analysis can be considered an extension of the Efimov plot for three bosons to the case of three 1/2-spin-isospin fermions.
Introduction
Weakly bound states have always attracted some attention because the particles have a large probability to be outside the interaction range and the system has universal characteristics. The limiting case is the unitary limit in which the two-body scattering length diverges, 1/a = 0. In this case the two-body system is resonant and can be described reasonable well in the zero-range approximation. In a series of papers [1,2] V. Efimov has shown that a system of three identical bosons interacting through a two-body short-range potential has a geometrical series of energy levels at the unitary limit accumulating at zero energy. The ratio between the energies of two consecutive states is constant and, remarkably it results independent of the particular form of the interaction. This behavior has been denoted Efimov effect and its universal characteristic has given to this effect a particular relevance. In fact in the last two decades an enormous amount of work has been dedicated to study this effect in different fields as molecular, atomic, nuclear and particle physics.
The spectrum of the three-boson system close to the unitary limit, in the zero-range limit, is described by the Efimov radial law where Δ(ξ) is an universal function. A parametrization of this function in the range −π < ξ < −π/4 can be found in Ref. [3]. The quantity s 0 ≈ 1.00624 is a universal number and κ * 2 1234567890 ''"" defines the energy 2 κ 2 * /m of the level n = n * at the unitary limit. Knowing the value of κ * the complete spectrum is determined as a function of a. The above equations have been derived in the zero-range limit in which the two-body energy is E 2 = 2 /ma 2 . The three-body sector shows a discrete scale invariance (DSI), the two-body low energy observables can be written in terms of a, making this quantity a control parameter (see for instance Ref. [3] and references therein). When the two-body potential has a finite range, the Efimov radial law can be modified as it was discussed in Refs. [4,5]. In the following the case in which the particles are 1/2-spin-isospin fermions [6] is discussed.
Effective description of 1/2-spin-isospin fermions close to the unitary limit
The zero-range energy spectrum given by Eq.(1) is unbounded from below as has been shown by Thomas [7]. Finite-range potentials cure this pathology limiting the energy spectrum of the three-body system from below being the lowest state the ground state of the system. For values of a much bigger than the range of the interaction the three-body energy spectrum has many excited state with this number diverging as a → ∞. Eq.(1) is still valid for the highest levels, however the ground and first excited levels show significant finite-range effects. In order to take into account these corrections the Efimov radial law is modified as follow [4,5], Two modifications have been introduced, the first one takes into account the fact that in the case of finite range interactions the two-body scattering length a and the energy length, a B , defined by E 2 = 2 /ma 2 B , with E 2 the two-body binding energy if a > 0, or the two-body virtual-state energy in the opposite case, a < 0, are slightly different. The second modification consists in the introduction of a finite-range parameter, the shift Γ 3 n , depending on the energy level. Recently it has been shown that the value of the shift is almost the same for very different potentials as the atomic helium-helium LM2M2 interaction of Aziz [8] and a two-parameter gaussian potential in which the strength and range have been fixed to reproduce the values of E 2 and a given by that interaction. Other examples are a combination of yukawians, as the MT-III interaction [9], describing the s-wave triplet n − p state or a nonlocal gaussian potential. Interestingly the value of the shift for the ground state in the three cases results to be almost the same, Γ 3 0 ≈ 0.8 [10,11]. Following those findings we would like to discuss the case of fermions having 1/2 isospin symmetry in which the interaction is different in the singlet and triple state as is the case of the two-nucleon system. The experimental data for the zero energy p − n system are a 0 = −23.740 ± 0.020 fm and a 1 = 5.424 ± 0.003 fm and the respective effective range values are r 0 ef f = 2.77 ± 0.05 fm and r 1 ef f = 1.753 ± 0.008 fm (as reported in Ref. [12]). To describe these data it is possible to use a gaussian potential with different strength in the singlet (S = 0) and triplet (S = 1) states. To simplify the description, the same range (in the following we consider r 0 = r 1 = 1.65 fm) is considered. The values V 0 = −37.90 MeV and V 1 = −60.575 MeV approximate well the experimental data and, for S = 1, the deuteron binding energy is well described too. Varying the strengths V S it is possible to explore the three-nucleon binding energy as the values of a S move toward the unitary limit. This study is done fixing the ratio a 0 /a 1 = −4.31 to its experimental value, accordingly both strengths are related. It should be noticed that there are other possibilities to make this study. For example, the ratio a 0 /a 1 = 1 corresponds to explore the three-boson system. In shown determined by the values of a 0 and a 1 . As mentioned, when the two scattering lengths are equal, a 0 /a 1 = 1, the corresponding plane is the boson plane for three equal particles very well studied (see for instance Ref. [4]). In the figure this is indicated by the (red) solid line. The (blue) solid line corresponds to a plane in which the nuclear physical point is included, this plane is called the nuclear plane. The continuation to the fourth quadrant given by the (blue) dashed line does not give new information since in this case the singlet and triplet potentials are exchanged and the system has the same binding energy. The two axes correspond to planes in which the singlet or triplet scattering lengths are on the unitary limit. The energy values E n 3 could be different in the different planes, however when the two scattering lengths are at the unitary limit the energy value is unique and it is that of the boson spectrum.
Three-body spectrum of 1/2-spin-isospin fermions close to the unitary limit
The spin-dependent potential defined in Eq.(3) can be used to explore the spectrum of the threenucleon system close to the unitary limit. With the strengths and range defined in the previous section the low energy two-body data are reasonable well described. Extending the analysis to the three-body case a binding energy of E 3 = −10.2 MeV is obtained to be compared to the experimental value of the triton E( 3 H) = −8.48 MeV. The observed over binding is common characteristic when attractive gaussian potentials are used. In order to have a quantitative agreement with the experimental value a three-body force has to be included (see for instance Ref. [13]). Here we are interested to study the variation of the three-body binding energy as a function of a 1 with the ratio a 0 /a 1 fixed to the nuclear value, thus we do not include such a three-body force.
The results of the present study are given in terms of the binding momentum, defined from the relation ( 2 /m)(K n 3 ) 2 = E n 3 , with n = 0, 1 for the ground and first excited state respectively. In Fig.2 first excited state. For comparison the binding momentum values in the case of three bosons are given too by the (black) solid and dashed lines. The (red) solid line is the two-body biding momentum. The binding momentum K n 3 is shown in units of κ * , the binding momentum of the ground state at the unitary limit and both, the fermion and boson values, are the same and correspond to a binding energy of ( 2 /m)κ 2 * = 3.6 MeV. The excited state in this limit has a binding momentum equal to K * /22.9 showing a slight finite-range correction (the zero-range theory predicts the value κ * /22.7). In the figure the (negative) values of the triplet scattering lengths, a 0 − and a 1 − , at which the three-boson ground and excited states disappear into the three-nucleon continuum are shown. Their ratio of about 17.6 is well below the prediction of 22.7 given by the zero-range theory. In the case of the three-fermion system it is interesting to notice that the first excited state disappears into the n − d continuum at the value a 1 * ≈ 20 fm, well before matching the nuclear physical point corresponding to a 1 = 5.4 fm). Conversely, in the case of the three-boson system, the excited state remains bound and always below E 2 . This fact has been studied before in the case of a system of three helium atoms. The present analysis shows that in the case of three nucleons the difference in the single and triple potential strengths is such that the excited state disappear into the n − d continuum before reaching the physical point. Evidence of the presence of this state embedded in the n − d continuum has been given in the literature [14,15] and the curvature of the n − d effective range function close to the threshold is one of this [16,17]. Finally, at the physical value, the two-and three-body energies are explicitly displayed. As it is well known, in order to reproduce the triton binding energy a three-body force has to be included.
The spectrum of the three-fermion system can be described by Eq.(2) with a value of the shift depending on the ratio a 0 /a 1 . For the ground state level the numerical results can be described with Γ 0 3 ≈ −0.2, a negative value. If we compare this value to the value of about 0.8 obtained in the case of three bosons we can conclude that there is a particular value of the ratio at which the shift Γ 0 3 ≈ 0. In this case Eq.(2) will be equivalent to the zero-range equation. However this 5 1234567890 ''"" would valid for the ground state level, the shift depends on the level and, as we have shown, the ratio with the excited state at the unitary limit is 22.9 (not 22.7) for a gaussian potential independently of the shift value.
Conclusions and Outlook
The present study analyzes the particularities of the Efimov plot in the case of 1/2-spin-isospin fermions. In order to make this study a particular way of constructing the Efimov plot has been selected. The ratio of the singlet and triplet scattering lengths has been kept fixed, accordingly the binding momentum K n 3 and a 1 define a plane similar to the boson case, in particular the values at the unitary limit coincide. The present work has to be consider a first step in the study of the Efimov plot for three nucleons. A deeper analysis including some results in the four-nucleon case and considering a three-body force are given in Ref. [13]. Extensions of the present study consist in the investigation of universal characteristic of the virtual state of the triton and the breakup reaction close to the deuteron threshold [18,19]. | 2019-04-22T13:10:54.929Z | 2018-03-01T00:00:00.000 | {
"year": 2018,
"sha1": "98a7315282543148803d685d2ee5181ba32a477a",
"oa_license": null,
"oa_url": "https://doi.org/10.1088/1742-6596/981/1/012021",
"oa_status": "GOLD",
"pdf_src": "IOP",
"pdf_hash": "1fefa444b8a15e7af45cd1c401284262b6b935ae",
"s2fieldsofstudy": [
"Physics"
],
"extfieldsofstudy": [
"Physics"
]
} |
209097887 | pes2o/s2orc | v3-fos-license | Exact synchronization in partial deterministic automata
An automaton is a synchronizing if it has an input word that transfers it from any state to a particular state. There are two versions of synchronization in partial deterministic automata: careful synchronization and exact synchronization. In this paper we focus on the exact version; we survey the complexity of testing exact synchronization and describe a SAT solver based algorithm for calculating the length of the shortest exact synchronizing word.
Introduction
A partial finite automaton (PFA) ࣛ is a triple 〈ܳ, Σ, ߜ〉, where ܳ and Σ are finite sets called the state set and the input alphabet, respectively, and ߜ: ܳ × Σ → ܳ is a partial function. The elements of ܳ and Σ are called states and letters, respectively, and ߜ is referred to as the transition function.
(The set ߜ(ܵ, )ݒ in the right-hand side is defined by the inductive assumption since the word ݒ is shorter than ).ݓ Observe that the set ߜ(ܵ, )ݓ may happen to be empty, in which case we say that ݓ is undefined at ܵ. In particular, if ߜ(ܵ, )ݓ is empty and ܵ consists of a single state ,ݍ we say that ݓ is undefined at ;ݍ otherwise ݓ is said to be defined at .ݍ When we deal with a fixed PFA ࣛ, we write .ݍ ݓ for ,ݍ(ߜ )ݓ and ܵ. ݓ for ߜ(ܵ, .)ݓAccordingly, we may simplify the notation for ࣛ, writing ࣛ = 〈ܳ, Σ〉 rather than ࣛ = 〈ܳ, Σ, ߜ〉. An automaton 〈ܳ, Σ〉 is said to be complete if .ݍ| ܽ| = 1 for all ,ݍ( ܽ) ∈ ܳ × Σ. We use the acronym CFA for the expression "complete finite automaton".
A CFA ࣛ = 〈ܳ, Σ〉 is called synchronizing if it possesses a word ݓ ∈ Σ * whose action leaves the automaton in one particular state no matter at which state in ܳ it is applied: .ݍ ݓ = ݍ ᇱ . ݓ for all ,ݍ ′ݍ ∈ ܳ. Any ݓ with this property is said to be a synchronizing word for the automaton. We refer the reader to the [2] of the forthcoming "Handbook of Automata Theory" for a discussion of the rich theory of complete synchronizing automata as well as their diverse connections and applications.
Any word w satisfying )1ܥ( − )3ܥ( is called a carefully synchronizing word for ࣛ. Thus, when a carefully synchronizing word is applied at any state in ܳ, no undefined transition occurs during the course of application.
If a word w satisfies the condition ,)3ܥ( it is called an exactly synchronizing word for ࣛ. Thus, ݓ can be undefined on some states in ܳ but there must be a state at which w is defined and .ݍ ݓ = ݍ ᇱ . ݓ whenever ݓ is defined at any ,ݍ ′ݍ ∈ ܳ. Clearly, a carefully synchronizing word is exactly synchronizing but the converse needs not be true. A PFA is said to be exactly synchronizing if it possesses an exactly synchronizing word. The class of exactly synchronizing PFAs is much larger than that of carefully synchronizing PFAs: for instance, if one adds to a synchronizing CFA a new state at which no letter is defined, one gets an exactly synchronizing PFA which is not carefully synchronizing since the condition all )1ܥ( fails for each letter.
Both careful synchronization and exact synchronization have interesting connections and numerous applications. In particular, exact synchronization is relevant in biologically inspired computing where exactly synchronizing words appear under the name "constants" in the study of so-called splicing systems, see [8]. Another cause of interest in exact synchronization is provided by so-called ߳-machines, important models in the theory of stationary information sources, see [9,10].
In [11], the present author and Volkov used SAT-solvers for finding shortest carefully synchronizing words in carefully synchronizing PFAs. Here we apply the same approach to exact synchronization. In section 2, we overview known results related to the problems of determining whether or not a given PFA is exactly synchronizing and of finding shortest exactly synchronizing words in exactly synchronizing PFAs. In Section 3 we show how instances of the latter problem can be efficiently encoded into instances of the Boolean satisfiability problem (SAT). Section 4 presents some of our experimental results.
Testing Exact Synchronization
A PFA ࣛ = 〈ܳ, Σ〉 is said to be strongly connected if for every pair ,ݍ( ݍ ᇱ ) ∈ ܳ × ܳ, there exists a word ݓ ∈ Σ * such that ݍ ᇱ = .ݍ .ݓ It turns out that for strongly connected PFAs, exact synchronization behaves similarly to synchronization of CFAs. In particular, checking whether a given strongly connected PFA is exactly synchronizing can be done in polynomial time, and for strongly connected exactly synchronizing PFAs with ݊ states, there exits a cubic in ݊ upper bound on the minimum length of exactly synchronizing words. Both these facts are straight forward consequences of the following observation by Travers Unfortunately, sufficiency in Proposition 1 does not extend to general PFAs. In the absence of strong connectivity, properties of exact synchronization became rather similar to those of careful synchronization. The following facts were established by Berlinkov [12, Corollary 1]: Proposition 2. Testing a given PFA with 2 input letters for exact synchronization is PSPACE-complete. There is a series of ݊-state PFAs with 2 input letters with the minimum length of exactly synchronizing words of magnitude 2 ஐ() .
Due to the complexity of the problems related to exact synchronization, one has to invoke some tools that have proved to be efficient for dealing with computationally hard problems. In this paper we use a satisfiability solver as such a tool.
Encoding
For completeness, recall the formulation of the Boolean satisfiability problem (SAT). An instance of SAT is a pair (ܸ, ,)ܥ where ܸ is a set of Boolean variables and ܥ is a collection of clauses over ܸ . (A clause over ܸ is a disjunction of literals and a literal is either a variable in ܸ or the negation of a variable in ܸ.) Any truth assignment on ܸ, i.e., any map ߮: ܸ → {0,1}, extends to a map ܥ → {0,1} (still denoted by ߮) via the usual rules of propositional calculus: ܥ has a satisfying assignment (i.e., a truth assignment on ܸ that satisfies )ܥ and NO otherwise.
Nowadays, many efficient SAT solvers, i.e., specialized programs designed to solve instances of SAT, have become available. Modern SAT solvers handle instances of SAT with hundreds of thousands of variables and millions of clauses within a few minutes. This remarkable progress has given rise to the following approach to computationally hard problems: one encodes instances of such problems into instances of SAT that are then fed to a SAT-solver. Here we apply this approach to the following problem: ESW (the existence of an exactly synchronizing word of a given length): Input: a PFA ࣛ and a positive integer ℓ (given in unary); Output: YES if ࣛ has an exactly synchronizing word of length ℓ; NO otherwise. We have to assume that the integer ℓ is given in unary because with ℓ given in binary, an efficient reduction from ESW to SAT is hardly possible in view of Proposition 2. Now, given an arbitrary instance (ࣛ, ℓ) of ESW, we construct an instance (ܸ, )ܥ of SAT such that the answer to (ࣛ, ℓ) is YES if and only if so is the answer to (ܸ, .)ܥ Here we use the same set of variables as in [11] but the set of clauses is essentially different. One may think that since the definition of an exactly synchronizing word is obtained from the definition of a carefully synchronizing word by omitting the conditions )1ܥ( and ,)2ܥ( one can get an encoding for ESW by just omitting groups of clauses that control these conditions in the encoding used in [11]. However, it is easy to exhibit counterexamples to show that this naive approach fails.
The set ܸ of variables in our encoding of (ࣛ, ℓ) consists of ݉ℓ letter variables ݔ ,௧ with 1 ≤ ݅ ≤ ݉, 1 ≤ ݐ ≤ ℓ, and ݊(ℓ + 1) state variables ݕ ,௧ with 1 ≤ ݆ ≤ ݊, 0 ≤ ݐ ≤ ℓ. The letter variables encode the letters of a hypothetical exactly synchronizing word ݓ of length ℓ: namely, we want the value of the variable ݔ ,௧ to be 1 if and only if the -ݐth letter of ݓ is ܽ . The intended meaning of the state variables is as follows: we want the value of the variable ݕ ,௧ to be 1 whenever the state ݍ belongs to the image of ܳ under the action of the prefix of ݓ of length ,ݐ in which situation we say that ݍ is active after ݐ steps. We see that the total number of variables in ܸ is ݉ℓ + ݊(ℓ + 1) = (݉ + ݊)ℓ + ݊.
The set ܥ of clauses in the encoding of (ࣛ, ℓ) consists of four groups. The group ܫ of initial clauses contains ݊ one-literal clauses ݕ , , 1 ≤ ݆ ≤ ݊, and expresses the fact that all states are active after 0 steps. For each ݐ = 1, … , ℓ, the group L of letter clauses includes the clauses Clearly, the clauses (1) express the fact that the -ݐth position of our hypothetical exactly synchronizing word ݓ is occupied by exactly one letter in Σ. Altogether, ܮ contains ℓ ቀ (ିଵ) ଶ + 1ቁ clauses.
The next group of clauses encodes the transitions of ࣛ. For a state ݍ ∈ ܳ, let ܲ ݍ( ) stand for the set of all preimages of ݍ under the action of the letter ܽ , that is, ܲ ൫ݍ ൯ ≔ { ∈ ܳ | . ܽ = ݍ }. Consider for every ݐ = 1, … , ℓ and every ݆ = 1, … , ݊, the following formula: ( Observe that the equivalence (2) just expresses, in the language of propositional logic, the fact that the state ݍ is active after ݐ steps if and only if some preimage of ݍ under the action of the -ݐth letter of ݓ is active after ݐ − 1 steps. A direct conversion of (2) into a conjunctive normal form leads to quite a bulky system of clauses. Instead, we use the following lemma. . Thus, if (2) is satisfied by The implication (5) is equivalent to the clause ݕ¬ ,௧ ∨ ⋁ ݕ ,௧ିଵ ೖ ∈ బ ( ೕ ) , and we see that ߮ satisfies the clause (3) with ݅ = ݅ . The implication (6) is equivalent to the system of clauses ݕ ,௧ ∨ ݕ¬ ,௧ିଵ where ݇ runs over the indices of states in ܲ బ ݍ( ). Hence, if ߮ satisfies (6), we see that ߮ satisfies all clauses (4) with ݅ = ݅ . Besides that, ߮ satisfies all clauses (3) and (4) with ݅ ≠ ݅ because each of these clauses includes ݔ¬ ,௧ as a literal and ߮൫ݔ ,௧ ൯ = 0 for all ݅ ≠ ݅ .
Thus, we have shown that if ߮ satisfies the equivalence (2), then ߮ also satisfies all clauses (3) and (4). All our arguments are reversible, and therefore, the converse claim holds as well.
We collect clauses of the form (3) and (4) for all ݐ = 1, … , ℓ and ݆ = 1, … , ݊ in the group ܶ of transition clauses. There are ℓ݉݊ clauses of the form (3); as for clauses of the form (4), its number for each triple (݅, ݆, )ݐ depends on the cardinality of the set ܲ ݍ( ), which clearly does not exceed ݊. Hence the number of clauses of the form (4) does not exceed ℓ݉݊ ଶ whence ܶ contains at most ℓ݉݊(݊ + 1) clauses in total.
It may be worth explaining how the clauses of the form (3) and (4) are understood in the case when one of the sets ܲ ݍ( ) happens to be empty. In (3) (3) reduces to ݕ¬ ,௧ ∨ ݔ¬ ,௧ . The latter clause simply means that if the -ݐth letter of ݓ is ܽ and the state ݍ has no preimage under ܽ , then ݍ cannot be active after ݐ steps. As for (4), the clauses of this sort disappear for all ݅ such that ܲ ݍ( ) is empty.
The final group ܵ of synchronization clauses describes the situation at the end of the synchronization process when the action of the word ݓ is completed. It consists of the following clauses: Clearly, the clauses in (7) express the fact that that exactly one state remains active after ℓ steps, which corresponds to the requirement |ܳ. |ݓ = 1. The group ܵ contains ( Moreover, the exactly synchronizing words of length ℓ for ࣛ are in a one-to-one correspondence with the satisfying assignments of (ܸ, .)ܥ Proof. We keep the notation and terminology introduced in the course of the construction of our encoding (ࣛ, ℓ) → (ܸ, .)ܥ In particular, ࣛ = 〈ܳ, Σ〉 with ܳ = ݍ{ ଵ , … , ݍ } and Σ = {ܽ ଵ , … , ܽ }. The fact that the instance (ܸ, )ܥ can be constructed in polynomial of ݊, ݉, and ℓ time follows from the estimates of |ܸ| and |C| established above. Now suppose that ࣛ has an exactly synchronizing word of length ℓ and fix such a word .ݓ We define a truth assignment ߮: ܸ → {0,1} as follows: for the letter variables ݔ ,௧ with 1 ≤ ݅ ≤ ݉, 1 ≤ ݐ ≤ ℓ, if the ݐ ୲୦ letter of w is a ୧ , 0 otherwise.
We see that any exactly synchronizing word of length ℓ for ࣛ determines a satisfying assignment for (ܸ, )ܥ and vice verse. Moreover, from the above proof it is clear that if we start with an exactly synchronizing word ݓ of length ℓ, construct from ݓ a satisfying assignment ߮ for (ܸ, ,)ܥ and then build an exactly synchronizing word from ߮, we get back the word .ݓ Thus, the correspondence between the exactly synchronizing words of length ℓ for ࣛ and the satisfying assignments of (ܸ, )ܥ is indeed one-to-one.
Experimental results
In this section we present some of our experimental results. The main aim of these experiments is the estimation of the expected length of the shortest exactly synchronizing word for PFAs with two input letters, ݊ states, and precisely one undefined transition. In [11] a similar series of experiments have been performed for careful synchronization, and it appears to be natural to look at similarities and differences between the two versions of synchronization.
As in [11], the number ݊ in our series of experiments did not exceed 100. We have generated uniformly at random a sample of 5000 automata for each ݊ ≤ 20, 2000 automata for 25 ≤ ݊ ≤ 60, and 700 automata for 65 ≤ ݊ ≤ 100. In order to find an exactly synchronizing word of minimum length for generated PFAs, we have considered ESW instances (ࣛ, ℓ) with each ࣛ in the sample and applied the encoding constructed in Section 3 to solve ESW instances with the help of a SAT solver. Then the minimum length been calculated via binary search on ℓ; we refer the reader to [13,14] where the procedure of binary search on ℓ is presented in detail. We have used MiniSat 2.2.0 [15,16]; the algorithm outlined above has been implemented in C++ and the code has been compiled with GCC 4.9.2. In our experiments we have used a personal computer with an Intel(R) Core(TM) i5-2520M processor with 2.5 GHz CPU and 4GB of RAM.
For each ݊, we have calculated the average length ℓ(݊) of shortest exactly synchronizing words for generated PFAs. Finally, the least squares method has been used to find a function that best reflects how ℓ(݊) depends on ݊, and it turned out that our results are reasonably well approximated by the following expression: ℓ(݊) ≈ 0.12 + 0.44݊ − 0.004݊ ଶ + 0.000018݊ ଷ (9) The relation between this approximation and our experimental data is shown in figure 1. In figure 2, we see that the relative standard deviation of ℓ(݊) gradually decreases as the number of states grows.
Finally, we compare our experimental results with those of [11] where the average length of shortest carefully synchronizing words for the same class of PFAs (PFAs with two input letters and precisely one undefined transition) has been evaluated. As expected, the average length of shortest exactly synchronizing words turns out to be smaller than the average length of shortest carefully synchronizing words−recall that every carefully synchronizing word is exactly synchronizing but the converse needs not be true. However, both values seem to follow the same pattern at least for the special class of PFAs used in our experiment. These observations are illustrated in figure 3.
Conclusion
We have approached the problem of computing an exactly synchronizing word of minimum length for a given PFA via the SAT solver method. For this, we have developed a new encoding that essentially differs from ones used in the earlier papers by the author [13] and by the author and Volkov [11,14]. We have implemented our algorithms and performed a number of experiments that confirm that our approach works sufficiently well even with very basic SAT solver and very modest computational resources. Now we are designing new experiments. For instance, it appears to be interesting to compare the minimum lengths of exactly synchronizing words for exactly synchronizing PFAs with a fixed number of states and letters but with different number of transitions. | 2019-11-07T15:08:51.623Z | 2019-10-01T00:00:00.000 | {
"year": 2019,
"sha1": "79271ea293106baa2ee985f327f783d0d7cc3f35",
"oa_license": null,
"oa_url": "https://doi.org/10.1088/1742-6596/1352/1/012047",
"oa_status": "GOLD",
"pdf_src": "IOP",
"pdf_hash": "2df3e6fab6dfe93532c933a875e19591c47fca27",
"s2fieldsofstudy": [
"Computer Science",
"Physics"
],
"extfieldsofstudy": [
"Physics",
"Computer Science"
]
} |
246979419 | pes2o/s2orc | v3-fos-license | Grapevine Gene Systems for Resistance to Gray Mold Botrytis cinerea and Powdery Mildew Erysiphe necator
Grapevine is one of the world’s most economically important fruit crops. It is known that Vitis vinifera is a host for a large number of pathogenic agents, which significantly reduce the yield and berry quality. This forces the agronomists to use a huge amount of fungicides. Over the last few decades, alternative methods for solving this problem have been developed and continue to be developed. Such new technologies as marker-assisted selection, bioengineering of the rhizosphere, genetic engineering (transgenesis, cisgenesis and intragenesis) allow the production of pathogen-resistant cultivars. However, they are linked to a number of problems. One of the most promising methods is the creation of modified non-transgenic cultivars via CRISPR/Cas9-targeted mutagenesis. Therefore, researchers are actively looking for target genes associated with pathogen resistance and susceptibility. This review elucidates the main mechanisms of plant—pathogen interactions, the immune systems developed by plants, as well as the identified genes for resistance and susceptibility to the biotrophic pathogen Erysiphe necator and the necrotrophic pathogen Botrytis cinerea.
Introduction
The grape family (Vitaceae) includes about 16 genera and 900 species [1][2][3][4]. The genus Vitis L. includes over 70 woody climber species, spread mostly in the temperate regions of the northern hemisphere. These species are classified into 3 groups of origin: European/West Asian; East Asian; American. The European/West Asian group includes only one species-Vitis vinifera L., which includes almost all varieties cultivated for high quality fruits. However, these cultivars are characterized as being unstable to abiotic and biotic factors. The East Asian and American groups include all the rest species of grapes. They are mostly of no practical value due to the low quality of the fruit and still remain poorly studied [1,3,5]. Two subgenera of Vitis have been commonly recognized: (1) subg. Vitis (2n = 38) comprises the vast majority of species and is distributed in northern South America, Central and North America, Asia, and Europe. (2) subg. Muscadinia (2n = 40) includes only 2-3 species and its distribution includes only the southeastern United States, northeastern Mexico, Belize, Guatemala, and the Caribbean [1][2][3]5,6]. The recognition of the two distinct subgenera, defined on the basis of morphological, karyological and cytological differences, is still debated [3,5]. The grape production is widespread throughout the world. Grapevine is one of the economically most important crops yielding berries and is mainly used for wine production [7,8]. According to data from the Food and Agriculture Organization (FAO, 2019), its cultivation area covers over 7.2 million hectares and the total grapes in the Russian Federation have been increasing annually over the past 10 years (Figure 1).
One of the main global problems of viticulture is the decrease in quantity and quality of harvest due to biotic stresses, mainly caused by bacteria, fungi and oomycetes, which causes severe diseases of grapes, including gray mold (caused by Botrytis cinerea Pers: Fr.), downy mildew (caused by Plasmopara viticola de Bary), powdery mildew (caused by Erysiphe necator (Schw.) Burr), also formerly known as Uncinula necator (Schwein.) Burr.) and grape black rot (Guignardia bidwellii (Ellis)). The majority of European grapevine cultivars of V. vinifera are susceptible to pathogen attack [10][11][12].
Gray mold causes significant economic losses in viticulture all around the world, reaching 20-50% of yield losses in grapes, due to rotting of ripe bunches in the post-harvest period. The high relative humidity and moderate temperatures during the grapevine vegetative cycle favour the development of this fungal pathogen [13]. The causative agent of the disease is Botrytis cinerea, a necrotrophic fungus with a short biotrophic phase. B. cinerea infects more than 1400 different plant species [14]. In the vineyard, this pathogen is a part of the microflora and usually lives in the soil on dead plant parts [15]. The disease infects fruits during ripening, causing necrotic areas with extensive fungal growth, which gives the characteristic appearance of gray rot. As a consequence, the grapes become unsuitable for wine making. Infection of berries is usually initiated by airborne conidia from overwintered sources [16,17]. During contact with the plant, B. cinerea causes cell death by producing phytotoxins and cell wall degrading enzymes and controls the host metabolism to facilitate colonization [13,18,19].
Powdery mildew (PM) is one of the most serious fungal diseases of grapes (V. vinifera) caused by the ascomycete E. necator, an obligate biotrophic pathogen that spreads through the air by conidial sporulation. It was introduced from America to Europe in the middle of the nineteenth century. Powdery mildew easily infects all green tissues (leading to chlorosis and premature aging and shedding of leaves, and forming white or gray powdery bloom on green stems), inflorescences and berries. Fruit infection leads to uneven ripening, wrinkling or cracking of the berries, resulting in fruit rotting, reduced yields and deterioration of the wine quality [20,21]. Fruit quality deteriorates greatly, acidity rises, and anthocyanins and sugar levels decrease. Even a 5% of infection can lead to unpleasant odors of the wine [22]. Host defense, which conditions ontogenic resistance, is known to operate early in the infection process, in the absence of major anatomical barriers. Until recently, grape berries were thought to remain susceptible to powdery mildew (Erisyphe necator) until late in their development. However, the development of ontogenic resistance is actually quite rapid in berries, and fruits become nearly immune to after fruit One of the main global problems of viticulture is the decrease in quantity and quality of harvest due to biotic stresses, mainly caused by bacteria, fungi and oomycetes, which causes severe diseases of grapes, including gray mold (caused by Botrytis cinerea Pers: Fr.), downy mildew (caused by Plasmopara viticola de Bary), powdery mildew (caused by Erysiphe necator (Schw.) Burr), also formerly known as Uncinula necator (Schwein.) Burr.) and grape black rot (Guignardia bidwellii (Ellis)). The majority of European grapevine cultivars of V. vinifera are susceptible to pathogen attack [10][11][12].
Gray mold causes significant economic losses in viticulture all around the world, reaching 20-50% of yield losses in grapes, due to rotting of ripe bunches in the post-harvest period. The high relative humidity and moderate temperatures during the grapevine vegetative cycle favour the development of this fungal pathogen [13]. The causative agent of the disease is Botrytis cinerea, a necrotrophic fungus with a short biotrophic phase. B. cinerea infects more than 1400 different plant species [14]. In the vineyard, this pathogen is a part of the microflora and usually lives in the soil on dead plant parts [15]. The disease infects fruits during ripening, causing necrotic areas with extensive fungal growth, which gives the characteristic appearance of gray rot. As a consequence, the grapes become unsuitable for wine making. Infection of berries is usually initiated by airborne conidia from overwintered sources [16,17]. During contact with the plant, B. cinerea causes cell death by producing phytotoxins and cell wall degrading enzymes and controls the host metabolism to facilitate colonization [13,18,19].
Powdery mildew (PM) is one of the most serious fungal diseases of grapes (V. vinifera) caused by the ascomycete E. necator, an obligate biotrophic pathogen that spreads through the air by conidial sporulation. It was introduced from America to Europe in the middle of the nineteenth century. Powdery mildew easily infects all green tissues (leading to chlorosis and premature aging and shedding of leaves, and forming white or gray powdery bloom on green stems), inflorescences and berries. Fruit infection leads to uneven ripening, wrinkling or cracking of the berries, resulting in fruit rotting, reduced yields and deterioration of the wine quality [20,21]. Fruit quality deteriorates greatly, acidity rises, and anthocyanins and sugar levels decrease. Even a 5% of infection can lead to unpleasant odors of the wine [22]. Host defense, which conditions ontogenic resistance, is known to operate early in the infection process, in the absence of major anatomical barriers. Until recently, grape berries were thought to remain susceptible to powdery mildew (Erisyphe necator) until late in their development. However, the development of ontogenic resistance is actually quite rapid in berries, and fruits become nearly immune to after fruit set [23]. High relative humidity is conducive to production of conidia. Atmospheric moisture in the 40% to 100% relative humidity range is sufficient for germination of conidia and infection. Free moisture, especially rainfall, is detrimental to survival of conidia [24][25][26][27][28][29]. Pesticides and fungicides, inorganic substances, above all sulfur, are widely used against powdery mildew and gray mold. However, this has a number of disadvantages, such as their negative impact on the environment, the high cost of chemicals and of their use. Worldwide, on average 35% of all pesticides are used for viticulture, which accounts for only 0.005% of the world's arable land [30][31][32]. In addition, populations of pathogens rapidly develop resistance to fungicides [33]. All this makes fungicide application program development a challenging task for winegrowers. Thus, there is a growing interest of modern viticulture in searching for new alternative environmentally friendly methods of protection against pathogens [11,[33][34][35]. Since the 19th century, breeding programs based on generative hybridization have been developed [36,37]. High pathogen resistance is observed among grapevine species from East Asia (V. pseudoreticulata, V. romanetii, V. amurensis and V. piasezkii Maximowicz) and North America (Vitis riparia Michx., V. berlandieri Planch (syn. V. cinerea var. helleri (Bailey) M.O. Moore), V. rupestris Scheele, V. aestivalis Michx (Muscadinia rotundifolia (syn. Vitis rotundifolia Michx.), V. labrusca L.). Due to its ability to be easily crossed with V. vinifera, these wild species are actively and successfully used in breeding as a rootstock, providing resistance to abiotic and biotic stresses [5,12,[37][38][39][40][41][42][43][44][45][46][47][48][49][50][51][52]. However, grape breeding is a slow process, for example, only after six successive crosses between Vitis vinifera and Muscadinia rotundifolia the complete resistance to powdery mildew was developed, which was associated with the transfer of the Run1 resistance gene to V. vinifera [53].
One of the alternative strategies to reduce the use of pesticides is related to the bioengineering of the rhizosphere, which represents the development of methods to stimulate the spread of introduced or local populations of beneficial microbes, the creation of synthetic microbial communities to stimulate plant growth, disease and stress resistance [54]. The plant ecosystem functioning and for sustainable farming the quality of the rhizosphere microbiome is of great importance [55,56]. It is known that the use of beneficial bacteria inhabiting the rhizo and/or endosphere of plants as biocontrol agents provides protection of their host plants against pathogens either by direct interaction with the pathogen or by promoting the development of plants induced resistance (IR) [57][58][59]. Inoculation with biocontrol agents could lead to changes in the whole plant root transcriptome due to a complex plant response [60]. After inoculation of grapes with various potential biocontrol agents (P. oligandrum, Trichoderma spp., Streptomyces sp., Pseudomonas protegens), the development of some pathogens was suppressed and plant growth was stimulated [8,[60][61][62][63][64][65][66]. In addition, arbuscular mycorrhizas (AM) fungal inoculation enhanced the growth of grapevine, together with rhizobacterium, increasing plant height and total dry weight [67][68][69]. Putatively effective and good candidates are known for bacterial biocontrol of E. necator, B. pumilus B-30,087 [70], Bacillus strains ATCC 55,608 and 55,609, which produce antifungal substances including zwittermicinA, playing a vital role in the interaction [71], B. subtilis [8,72]. The following biocontrol beneficial strains demonstrated a protective activity against B. cinerea on dual culture plates, on in vitro plantlets and in field conditions: Pantoea agglomerans, Acinetobacter Iwoffii, Burkholderia phytofirmans, Micromonospora spp., Streptomyces spp., Cupriavidus sp., P. fluorescens, P. putida, Pseudomonas aeruginosa, Bacillus circulans, Bacillus subtilis [8,57,[73][74][75][76][77][78][79][80][81][82][83][84][85]. Pathogen resistance was correlated to a different extent with phytoalexins and oxidative burst production [77,78,85]. In addition, it was concluded that different strains could be more appropriate for treatment of specific organs [75]. The bacterial strain oxB (which is closely related to Cupriavidus campinensis) was detected, which actively decomposes oxalic acid, an important virulence factor of B. cinerea, thus limiting the symptoms of gray mold on the leaves and greatly reducing the symptoms of the disease caused by B. cinerea in inflorescences in laboratory conditions [84]. However, there are several problems linked to this strategy-the success of bioinoculants or biocontrol agents utilization in the field depends on the target culture, product availability and application options, environmental conditions, and costs [86]. Expansion of the knowledge of plant-microbial interactions and its application may be useful in future plant breeding and farming programs [87].
Enormous efforts to develop resistant varieties have prompted researchers to search for new breeding technologies based on the knowledge of genes and genomes. One approach is a combination of classical selection with marker-assisted selection (MAS) for introducing certain loci into new varieties [88]. Different countries have developed various new breeding programs aimed at obtaining grape varieties combining disease resistance with high quality berries [4,[89][90][91][92]. The use of molecular markers allows to speed up and to make the selection process cheaper, and also to ensure the creation of varieties carrying new combinations of genes, which are almost impossible by traditional breeding methods [93].
The second approach is associated with the interference into the genome with the use of genetic engineering. It includes several methods that can purposefully alter the genome of plants in order to obtain resistant grape varieties with the desired grape properties for growers and consumers [94][95][96][97][98][99][100]. One of the methods is the genetic modification of grapes to provide resistance to fungal diseases. For example, transgenic grape plants were created, carrying the rice chitinase gene, which increased resistance to powdery mildew and anthracnose (Elisinoe ampelina (de Bary) Shear). However, transgenesis raises concern among winemakers due to the introduction of alien genes into elite grape varieties and their potential impact on wine characteristics, and genetically modified organisms (GMOs) are prohibited from being grown in many countries [101].
As alternatives to transgenesis, two methods of crop transformation and plant breeding have been developed-cisgenesis and intragenesis. Both concepts imply the modification of plants only with genetic material obtained from the same or closely related species [102,103]. Cisgenesis is a closer method to traditional breeding, since plants are subjected to genetic modification by one or more genes that retain their natural genetic composition with all its regulatory elements. However, unlike traditional breeding, cisgenic crops contain exclusively the gene or genes of interest without any undesirable genetic element. Intragenesis allows the use of hybrid genes that can incorporate genetic elements from different genes and loci, thus promoting the creation of new genetic combinations, new expression models and new plant properties [104]. The application of cisgenesis and intragenesis is limited to several species, mainly due to the lack of knowledge about the required regulatory sequences. Cisgenic grape lines are currently being developed. Chardonnay grape was transformed with the use of the construct, containing PR protein VVTL-1 (Vitis vinifera thaumatin-like protein), which inhibits spore germination and hyphal growth. The resulting transformants were characterized by a decreased severity of powdery mildew and black rot [105,106].
CRISPR/Cas9 technology is one of the most promising genome-editing methods, which can be used for targeted gene modification, suppression/activation of gene expression, subtle modification of gene function, and epigenome editing. CRISPR/Cas9-targeted mutagenesis opens the possibility to obtain non-transgenic modified plants carrying specifically determined mutations that are stably inherited over generations. CRISPR/Cas9targeted mutagenesis. The legislation of some countries does not prohibit the cultivation of such modified non-transgenic organisms; therefore, CRISPR/Cas9 editing is considered a promising tool for improving the resistance of grape varieties. Researchers conduct search for new target genes of grapes for being modified with CRISPR/Cas9 technology that will increase resistance to pathogens [93,99,100].
To date, the genomes of cultivated grapes V. vinifera [107,108] and wild V. vinifera sylvestris [109], as well as the B. cinerea, have been sequenced [110,111]. Modern genetic studies of grapes have also allowed to identify gene systems of resistance/susceptibility to the most common pathogens-gray mold and powdery mildew. In the frameworks of this review, we have systematized the available to date scientific data about these gene systems.
Gene Systems for Resistance and Susceptibility to Botrytis cinerea 2.1. Mechanisms
Infection of grapevine inflorescences with B. cinerea occurs at the stage of flowering with air conidia, which settles in the surface cells of the host. At the stage of immature green berries, the pathogen spends a long time in the host tissues without symptoms [15][16][17][112][113][114][115]. The resistance of immature berries to B. cinerea is the result of several associated mechanical and chemical processes, but they have not yet been fully studied [13]. The transition of the pathogen from the resting stage to the phase of active infection occurs during the berry ripening, which is facilitated by physiological and biochemical changes that are activated in berries during ripening, as well as by signals associated with ripening [13,116,117]. Although B. cinerea has been well studied in various plant species, there is limited information regarding the mechanisms of resistance and susceptibility of the Vitis ssp. genotypes. to B. cinerea. causing gray mold. Transcriptome and metabolomic analyzes of grapes and B. cinerea at different stages of plant-pathogen interaction allowed identification of 3610 genes with differential expression [116][117][118][119]. The genes VIT_01s0011g05420 and VIT_11s0016g02070 were the only genes shared at three different stages of infection and they were always activated by B. cinerea. The function of the VIT_01s0011g05420 gene is not yet known [116]. The VIT_11s0016g02070 gene presumably belongs to the basic helix-loop-helix family; it is a transcription factor (TF) involved in hormonal interactions and host immunity [120].
For B. cinerea, the expression of 9927 genes was identified, both unique as well as common for different stages of infection. 222 common genes were identified for all stages of infection, most of them were predicted/hypothetical proteins, ribosomal proteins, and housekeeping genes [116]. It is known that the fungus B. cinerea actively promotes plant susceptibility using many virulence factors [121,122]. At the early stages B. cinerea utilises mRNA and effector proteins to suppress the early host cells death and immune responses, which allows the fungus to anchor inside the host and to accumulate its biomass before the necrotrophic phase. In response to plant infection, a whole cascade of interrelated defense mechanisms is activated, which come into action at certain stages of the pathological process [123,124]. The plant recognizes peptides and proteins such as extracellular pathogen-associated molecular patterns PAMP, or intracellular pathogens effectors delivered to host cells, which trigger signal transduction and activate rapid defense responses that involve massive reprogramming of transcription within the plant. This leads to activation of gene expression (for example VvGLP3), accumulation of reactive oxygen species (ROS, oxidative burst), activation of genes encoding antimicrobial proteins, and of genes of the polyphenol biosynthesis pathway for the production of phytoalexins and precursors for strengthening the cell wall (biosynthesis of monolignols (VvPAL, VvCOMT, VvCCoAMT), stilbenoids (VvSTSs) [119].
The plant cuticle is the first barrier against the pathogen conidia germination. It consists mainly of two types of lipids, cutin and cuticular waxes, which may be the main candidates for signaling early interactions between plants and pathogens. It is known that the absence of any of these components in mutants led to changes in genes expression at the early stages of infection and germination of B. cinerea decreased consequently [125,126]. Oleanolic acid (OA) and n-fatty alcohols (C22, C24 and C26) are the main components of the cuticular waxes of mature grapes [127]. OA is a naturally occurring triterpenoid that is associated with membrane fluidity and potentially affects signaling by many ligands and cofactors. The presence of OA can play a regulatory or a stabilizing role in the germination of B. cinerea. The second component, n-fatty alcohols, play an important role as inducers of germination at the early stages; however, they participate in the acceleration of germination only in a certain combination-the complete mixture of alcohols (C22, C24 and C26) was designated as "full-fatty-alcohols" (FFA) [126]. The involvement of pectin methylesterases (PMEs) and PME inhibitor families in maintaining cell wall integrity has also been detected [116].
Initiation of Infection
One of the earliest cellular responses to flower infection with B. cinerea is the production of ROS, and the genes encoding oxidative stress enzymes (GST, ascorbate oxidase, 2OG-FeII oxygenase, cytochrome P450 monooxygenases activation [12]). In the investigation of Wan et al. (2015) the grape leaves resistance to B. cinerea was analyzed in the genotypes of Chinese wild grape species and of cultivated V. vinifera. The low rates of infection and good resistance to fungi were characteristic to the Chinese wild species: V. amurensis; V. yenshanensis; V. qinlingensis P.C. He (Qinling grape from Qinling Mountain region)clone "Pingli-5", V. adstricta. Many Chinese wild species demonstrated simultaneous resistance to several fungi: for example, "Pingli-5" was resistant to gray mold, anthracnose, powdery mildew and downy mildew [12,41]. Oxidative stress disrupts the redox balance in the affected tissues, thereby contributing to the progression of the disease. Antioxidant enzymes such as Peroxidase (POD), catalase (CAT) and superoxide dismutase (SOD) protect plant cells from oxidative stress and maintain redox balance by removing ROS generated during pathogens attack. Pathogen-resistant "Pingli-5" was characterized by CAT and POD increased activities throughout the experiment, but almost no change in SOD activity was observed, with the exception of its increase at 4 hpi (hours post inoculation), which corresponded to the minimal induction of ROS [12,128]. Similarly, in the genotype of V. vinifera "Ju mei gui", a small amount of ROS was found after inoculation, which indicates the antioxidant enzymes role in maintenance of redox balance and cells protection from ROS destruction. The leaves of "Ju mei gui"were characterised by reduced germination of fungi and spread of infection. The elevated levels of POD activity and no significant changes in SOD activity were detected in "Ju mei gui" as well [129].
At the beginning of the initiation of infection, in 24 hpi, the genes encoding membrane receptor-like kinases (RLK), such as CLV1, WAK1, BAK1, associated with the immune system response to necrotrophic pathogens, are activated. WAK1-damage-related receptor recognizes oligogalacturonides derived from the plant cell wall due to cell wall degradation [117,119]. There is a rapid and strong induction of the expression of genes encoding about 100 transcriptional regulators (TFs) in infected flowers, which promotes the activation of specific defense pathways. The genes encoding WRKY, NAC, MYB, and ethylene-responsive element-binding proteins were among the most well-studied and most important identified TF genes. TF WRKY33 is involved in the defense reaction of the host plant, and its expression level was detected to increase strongly already after 12 hpi, but it decreased after 48 hpi [117,119,130]. TF Myb14 regulates the biosynthesis of stilbenes [131]. At the early stage of infection, grapevine genes encoding various classes of pathogenesis related proteins (PR), like osmotin, thaumatin, chitinases, Beta 1-3 glucanase, Bet v I allergen, PR1 and PR10, regulated by TF, are activated. As a result, oligosaccharides are formed, which in turn enhance the synthesis of phytoalexins, low molecular weight substances of a secondary origin, which are synthesized at the sites of pathogen penetration and suppress its development. Phytoalexins maintain species immunity to non-specialised pathogens, and varietal resistance to specialised pathogens [117,119]. Even during the green berry stage, when the pathogen was dormant, the genes encoding PR proteins were still activated. The involvement of these genes in response to B. cinerea was also observed in ripe berries, but their action was ineffective against bunch rot development [116]. In response to the infection of grapes with B. cinerea, the phenylpropanoid defense pathway is activated. At the stage of infection initiation, an increase in the concentration of compounds such as resveratrol, viniferin, miyabenol, isogopeaphenol, catechin and proanthocyanidins was observed. These substances mediate plant protection, suggesting that they contributed to the inhibition of the pathogen before maturation [117][118][119][132][133][134].
Genes involved in the synthesis and signaling of phytohormones in response to infection are differentially expressed. Their production starts relatively fast in plant tissues after its interaction with fungal elicitors [135,136]. Salicylic acid (SA) is associated with local resistance to biotrophs and with the onset of systemic acquired resistance (SAR). Biotrophic pathogens generally induce the SA-mediated defense response, which activates various downstream physiological immune responses such as programmed cell death and ROS accumulation [137]. Jasmonic acid (JA) and ethylene (ET) control the resistance to necrotrophs and induce systemic resistance. Semi-biotrophic pathogens induce both SAand JA-mediated signaling responses [135,136,138]. It has been suggested that SA-and JA-mediated signaling pathways are antagonistic, and if plants defend against a particular pathogen via the SA-dependent pathway first, then JA-signaling is inhibited [136,139]. The phytohormones ET, JA and SA have been found to play an important role in the interaction between the flower and B. cinerea. The involvement of gibberellic acid (GA) and abscisic acid (ABA) has also been identified [119]. High content of JA in resistant cultivar "Ju mei gui", blocked B. cinerea infection [129].
The Green Berry Stage, 4 Wpi (Weeks Post Inoculation)
At the green berry stage, the amount of expressed B. cinerea genes is relatively small. The pathogen is able to maintain its main metabolic activity during dormancy, high expression of fungal ribosomal genes, elongation factors, ATP synthesis, ATP-dependent molecular functions related genes, as well as 34 CAZyme genes, is registered, which implies the preservation of the pathogen ability to obtain energy from the host plant [117,119]. The expression level of the CWDE genes and genes encoding phytotoxins is either strongly reduced or not detected-the virulence factors of the fungus are disabled, probably due to the dormant state [116][117][118]140]. There is a molecular relationship between B. cinerea and immature berries during the dormant period-the virulence genes of the pathogen are inactive, and an increased immune response is observed in plants [117]. Preformed and induced defense mechanisms, including skin features of immature berries such as polyphenols in cell walls of berry skin and the thickness of the epidermal cell layer complex, have been proposed as part of ontogenic resistance to B. cinerea [119,133].
At the green berry stage, the VvPAL and VvSTS genes are induced in grape plants, which are associated with the immune pathway of phenylpropanoid biosynthesis, leading to the production of stilbenes (resveratrol and its derivatives) and lignin, which inhibit the growth of B. cinerea. The activation of the expression of 2 grape genes encoding putative enzymes involved in lignin biosynthesis, VvC3H (p-coumarate 3-hydroxylase) and VvCCR (cinnamoyl CoA reductase), was characterized, and the increase in the amount of the corresponding metabolites (affeic acid, ferulic acid, and chlorogenic acid) was confirmed at the green berry stage. At the ripe berry stage, their expression decreased. The expression of VvEXT (extensin-like protein) gene encoding proline-rich extensin-like protein was also induced at the green berry stage. Papillas (callose-enriched encasement of haustorium) were formed under the appressoria in the green berry, during early infection, but they were not detected in the mature berry [141]. The activation of the VvRbohD gene was identified, which encodes the putative superoxide-generating NADPH-oxidase, that leads to the formation of O2-, and its transformation into H 2 O 2 depends on the activation of the VvGLP3 gene (germin-like proteins). GLPs were among the most inducible genes during the green berry stage and the early ripening stage [142,143]. Since the expression of the SA marker gene VvPR1 was increased, it was suggested that the SA pathway may also be involved in basal resistance at this stage [118]. Similar to the stage of infection initiation in the flower, in the green berry there was a detected activation of the expression of the genes encoding membrane-localized RLKs: Clavata1 receptor kinase (CLV1), Brassinosteroid insensitive 1-associated kinase 1 (BAK1), and Wall-associated kinase 1 (WAK1) [117,119]. In response to infection, genes of various PR protein families, including PR10, were strongly induced due to dormant B. cinerea. At this stage B. cinerea induces the expression of key TFs that play an important role in the interactions between plants and microbes. Lignification at the site of entry is one of the main defense mechanisms used by plants to arrest the development of B. cinerea [118,119]. It was observed that there was an up-regulation of SA, ET signaling marker genes, and genes related to auxin metabolism, that contributed to the increase in the protective ability of the hard green berry [117,118].
The Ripe Berry Stage, 12 Wpi
During the process of ripening, grape berries undergo several modifications that reduce their natural resistance to the pathogen. The mechanical resistance in the cell wall decreases, contributing to the appearance of microcracks. Moreover, the sugar concentration increases and the concentrations of organic acids and of some compounds associated with biotic resistance decrease. The hormonal balance and pH also change [140]. VvJAZ1, a marker gene for JA, and transcripts of various enzymes of the JA biosynthesis pathway (phospholipase, lipase, allene oxide synthase, jasmonate O-methyltransferase) are the most activated at the ripe berry stage [118]. It is likely that the signals associated with ripening, as well as physical and chemical changes at this stage, play an important role in triggering the transition of a necrotrophic pathogen from prolonged dormancy to an active infection phase [13,116]. The sensitivity of ripe berries is significantly correlated with phenolic compounds in the cell walls of berry skin and is negatively correlated with the total content of tannins in the skin and with water activity (Aw) on the surface of the berry [144].
Gene Systems for Resistance and Susceptibility to Botrytis cinerea
Currently, a number of genes have been identified in grapevines that may confer resistance or tolerance to B. cinerea, including VvPLDs, VvPR10.1, VvPR10.3, VvNPR1.1, VaSTS19, WRKY57, VlWRKY3, VqWRKY52, VqSTS21, VqAP13, ERF, VvXYLP, etc. The introduction of genes VaSTS19 and VvNPR1.1 into A. thaliana enhanced its resistance to B. cinerea. On the contrary, overexpression of the VqTLP29, VlWRKY3, VqWRKY52, VqSTS21, VqAP13 genes in A. thaliana increased the susceptibility to B. cinerea [139,[145][146][147][148][149]. The phospholipase D gene family (PLD) has been found to play an important role in the regulation of cellular processes in plants, including ABA signaling, programmed cell death, growth regulation and stress responses. Structural analysis showed that the PLD gene family can be divided into 6 subgroups: α, β/γ, δ, ε, ζ and ϕ, which had descended from 4 original ancestors as a result of a series of gene duplications [150]. PLD α1 participates in various plant processes, such as ROS production and accumulation of jasmonic acid at the wound site, and promotes ABA signaling inhibiting the negative regulator ABI1. PLD α1 and its lipid product PA are intermediates between important cellular regulators in plant cells. PLD α1 interacts directly with Ga proteins [151]. PLDδ positively regulates plant resistance to stress, including oxidative stress. Its role lies in signaling of damage caused by ROS. It is activated by oleic acid and strongly binds to the plasma membrane and the cytoskeleton of microtubules. PLDδ is activated by hydrogen peroxide (H 2 O 2 ) and the resulting PA reduces the programmed cell death caused by H 2 O 2 . PLD δ and PA participate in the H 2 O 2induced activation of MAP kinase cascades [152]. Thus, PLDδ mediates plant responses to ROS [151].
In grapes, the VvPLD genes were also characterized and classified into 6 types (VvPLDαs, VvPLDβs, VvPLDδs, VvPLDε, VvPLDρ, VvPLDζ) and 3 groups (C2-PLD, PXPH-PLD and SP-PLD). The function of VvPLD genes upon infection V. vinifera cv. Cabernet Sauvignon with B. cinerea was studied. It was revealed that the genes VvPLDβ1, VvPLDβ2, VvPLDδ2, VvPLDρ and VvPLDζ were activated, and expression of the genes VvPLDα and VvPLDδ was suppressed. Identified genes can serve as candidates for the role of resistance genes [153].
The WRKY TF family has been detected to play a role in biotic stress responses. The VqWRKY52 gene found in the Chinese wild Vitis quinquangularis encodes the WRKY III gene family member and plays an important role in the SA-dependent signal transduction pathway, inducing cell death during hypersensitive response (HR). In transgenic A. thaliana lines, its overexpression increased the sensitivity to B. cinerea due to increased HR [146]. When the VvWRKY52 gene was switched off in grapes by CRISPR/Cas9-targeted mutagenesis, a phenotype with increased resistance to B. cinerea was observed in transgenic lines [154]. In Arabidopsis gene WRKY33, a functional homologue of VvWRKY33, plays a key role in plant defense, regulating redox homeostasis, SA signaling, ET-JA-mediated cross-communication and phytoalexin biosynthesis, providing resistance to B. cinerea [130]. WRKY57 belongs to Group IIc of WRKY TF family [155]. Loss of function of WRKY57 en-hanced resistance against B. cinerea, and this resistance was associated with the JA signaling pathway, especially COI1. JAZ1 and JAZ5, direct targets of WRKY33, were identified as direct target genes of WRKY57. WRKY57, the same as WRKY33, binds to the promoters of the JAZ1 and JAZ5 genes via a W-box sequence. WRKY57 and WRKY33 competitively regulate JAZ1 and JAZ5. WRKY57 compromises B. cinerea resistance via competing with WRKY33 to transcriptionally regulate JAZ1 and JAZ5. WRKY33 acts as a transcriptional repressor of JAZ1 and JAZ5 [156]. Transgenic VlWRKY3 lines of A. thaliana showed increased sensitivity to B. cinerea, probably due to the interaction of the SA and methyl jasmonate (MeJA) signaling pathways, or the influence of other genes. It has previously been reported that WRKY3 responds to salt and drought stress as well as MeJA and ET in V. labrusca × V. vinifera cv. 'Kyoho' [147,157].
Among the Pr genes, the highest activation of expression was observed for genes VvPR10.1 and VvPR10.3. This suggests that they are the main candidates for the role of genes for resistance to B. cinerea. In grapes, VvPR10.1 is associated with resistance to P. viticola and is regulated by VvWRKY33 [119]. In grapes (V. vinifera), the regulation of 35 ERF genes (Ethylene responsive factor) in response to B. cinerea was analyzed, and their important functions in plant protection, in addition to their known role in development, were identified. In the study, the genes VvERF071, VvERF072, VvERF066 and VvERF099 of the IX ERF subfamily showed the highest levels of expression at key time points of infection. After leaves infection with B. cinerea of two cultivars (the susceptibile "Red Globe" and the resistant "Shuangyou") the destruction and strengthening of the cell wall was significantly induced respectively. The analysis of cis-elements of the promoter showed that most of the ERF genes contain elements of the CGTCA-motif and TGACG-motif, which are known to be involved in the response to MeJA, especially VvERF071 and VvERF072, suggesting that they may be involved in the response to the pathogen infection. The identified TGAelement and AuxRR-core are involved in the response to auxin. ABRE, TCA-element and ERE-elements are involved in the reaction for ABA, SA and ET, respectively. VvERF099 has a gibberellin responsive motif P-box [158]. It was also detected that expression levels of VvERF016 from group VII increased after infection by B. cinerea. In A. thaliana, the VvERF016 gene homologue (AtERF72, also named AtEBP) interacts with ACBP4 (acyl-CoA binding protein 4) and probably regulates ethylene-related signaling and defense mechanisms in plants [158,159].
Thaumatin-like protein (TLP) is a large family in plants, and individual members play different roles in various responses to biotic and abiotic stresses. TLP family is involved in pathogen resistance. The expression of the TLP gene is widely influenced by E. necator and B. cinerea. VqTLP29 overexpressing transgenic grape lines of diseases resistant V. quinquangularis increased resistance to powdery mildew, but increased susceptibility to gray mold, by regulating signal transduction through the SA and JA/ET pathways. Treatment with MeJA at concentrations of 10 to 100 µmol/L can effectively induce resistance to B. cinerea. Induced disease resistance was closely associated with increased production of H 2 O 2 , enhanced expression of the VvNPR1.1 gene, and accumulation of stilbene phytoalexins such as transresveratrol and its oligomer (trans-) ε-viniferine [160].
Recently identified sugar transporters SWEETs have been characterized as bi-directional, low-affinity sugar carriers, probably operating by an uniport mechanism [161]. On the one hand, pathogens can alter the expression level of genes encoding various sugar carrier proteins, including the SWEET genes using sugars as a source of carbon and energy [162]. On the other hand, it has been found that high sugar levels in plant tissues can also increase plant resistance enhancing the oxidative burst in the early stages of infection, thereby increasing cell wall lignification, stimulating flavonoid synthesis and inducing certain PR proteins [163]. VvSWEETs transporters play an important role in sugar mobilization during grape berry development, and their expression is transcriptionally reprogrammed in response to B. cinerea infection. In the grapevine, the SWEET family is composed by 17 members, and only VvSWEET4 and VvSWEET10 were functionally characterized [164,165]. VvSWEET7 and VvSWEET15 are the key SWEET genes at the stages of grapes development and maturation, localized in the plasma membrane. VvSWEET7 transports both glucose and sucrose. Expression of VvSWEET7 in berries peaks at the green berry stage and of VvSWEET15 at maturity. They are characterized by increased expression in response to infection with B. cinerea at the green berry and ripe berry stages, respectively. In addition, B. cinerea infection suppresses the expression of VvSWEET17a and activates the expression of VvSWEET2a at the green berry stage, suppresses the expression of VvSWEET10 and VvSWEET17d at the veraison stage and the expression of VvSWEET11 at the ripe berry stage. Since the grape resistance is high at the green berry stage, it is likely that overexpression of VvSWEET2a and VvSWEET7 may be a defense mechanism [19].
Xylogen-like proteins (XYLPs), a glycoprotein family with high molecular weights, are widespread in plants [166]. XYLPs are involved in various plant growth, development and stress responses, such as cell xylem differentiation, programmed cell death, abiotic tolerance and hormone signal transduction, such as SA, JA, and ABA. Six XYLP genes were identified in V. vinifera (named VvXYLP01-VvXYLP06 according to their chromosome location). It was detected that upon infection with B. cinerea the level of VvXYLP02 expression significantly increases in the resistant genotype, but decreases or slightly increases in the sensitive genotype. Overexpression of VvXYLP02 in A. thaliana significantly increased resistance to B. cinerea. In addition, the JA treatment significantly increased VvXYLP02 expression in V. vinifera [149]. Chitosan can play a key role in plants resistant to B. cinerea infection. Chitosan induces protective genes and protects berries quality from infection through JA signaling. Suppression of expression of negative regulators in fruit disease-resistance expression-VvHDAC19 (histone deacetylase 19) and VvTPR3 (Topless-related protein 3)-with the application of chitosan led to an increase in grape resistance to B. cinerea. TPR3 and HDAC19 interact with each other and their activity is suppressed by JA and chitosan. The synthesis of phenolic compounds was activated and the strengthening of the cell wall was observed, oxidative stress was modulated and the production of JA in ripe fruits was induced, the growth of B. cinerea was suppressed [167].
The VvAOS gene, which plays an important role in the biosynthesis of JA, and the VvCOI1 gene, which encodes the receptor of jasmonate [168] and regulates the level of JA signaling, both affect the process of berries ripening through the regulation of fruit pigmentation and cell wall metabolism, and also affect the process of berries infection by B. cinerea.
The accumulation of JA occurs relatively quickly in plant tissues and cells after injury or exposure to fungal elicitors [169,170]. Suppression of the VvAOS gene leads to a decrease in the JA content. The samples with overexpression of the VvCOI1 gene (VvCOI1-OE) were characterized by a high level of JA and a delay in the B. cinerea infection progression. Even though the overexpression of VvCOI1 in strawberries did not affect directly the JA content, it induced plant defense genes such as PPO, SOD, POD, PAL, BG and chitinase, which protected fruits from infection by gray mold [171]. Cysteine-rich receptor-like kinases (CRKs) play an important role in disease resistance of plants, including A. thaliana, Oryza sativa L. (rice), Hordeum vulgare L. (barley), Triticum aestivum L. (wheat) and Medicago truncatula Gaertn. [172][173][174][175][176][177][178]. Four CRKs have been identified in V. amurensis (VaCRK1-VaCRK4). Overexpression of VaCRK2 increases resistance to B. cinerea in A. thaliana. A grape cultivar resistant to gray mold had a significant increase in the VaCRK2 gene expression after inoculation with B. cinerea and treatment with methyl jasmonate. VaCRK2 expression induced genes associated with the JA signaling pathway, genes associated with pathogenesis (PR), and accumulation of ROS [179].
Gene Systems for Resistance and Susceptibility to Erysiphe necator
To resist pathogens, plants have developed an innate immune system that exists in two main forms: (1) the primary species-specific immune response known as PTI (PAMPtriggered immunity) is activated when the host's extracellular plasma membrane receptors (PRR-pattern-recognition receptors) interact with pathogen-specific molecules. For example, callose synthase GSL5/PMR4 is responsible for the biosynthesis of most of the callose. Disabling this gene leads to a dysregulation of SA-mediated defense reactions and a loss of susceptibility to powdery mildew pathogen. On the other hand, overexpression of GSL5/PMR4 leads to the increase of callose accumulation in the papilla and complete resistance to the penetration of E. necator [180]. This cellular rearrangement is observed in mlo (MILDEW RESISTANCE LOCUS O) mutants resistant to pathogen penetration [180,181].
(2) The secondary, or race-cultivar-specific immune response, known as ETI (effectertriggered immunity), is initiated by pathogen virulence factors, which are recognized directly or indirectly by the intracellular receptors of the host plant, usually linked with resistance genes (R-genes). Then, subsequent signaling pathways are activated, that ultimately leads to numerous protective reactions, localized hypersensitivity reactions at the site of infection, and programmed cell death [181,182]. It has been suggested that wild Chinese and wild American grapes interact with E. necator via different cellular and molecular defense mechanisms [46]. Wild North American grape species (Muscadinia rotundifolia (syn. Vitis rotundifolia), V. rupestris, V. riparia, and V. Aestivalis; V. labrusca, V. Berlandieri) [183], due to long co-evolution with E necator, have acquired ETI resistance to powdery mildew. As a consequence, this resistance can be quickly overcome by new strains of E. necator. Thus, the resistance of grapes obtained by the introduction of dominant R-genes into elite cultivars through classical breeding is unlikely to be durable, and a broad spectrum genetic resistance is required for breeding [46,[184][185][186][187]. Wild Chinese species (V. pseudoreticulata, V. romanetii and V. piasezkii) exhibit persistent immunity to E. necator after its penetration into cells via programmed cell death mediated by NBS-LRR, as well as the formation of papilla. Thus, wild Chinese grapes represent an important genetic source for enhancing the sustainability of cultivated grapes [46,[188][189][190][191].
R-Genes, NBS-LRR
The most characterized R-genes belong to the NBS-LRR (leucine-rich nucleotide-binding site repeat) gene superfamily, one of the largest gene families in plant genomes. The majority of plant disease resistance genes encode NBS-LRR proteins and contain conserved nucleotide binding motifs (NB) and leucine rich repeat (LRR) motifs. NBS-LRR proteins are divided into two subfamilies based on their N-termini: TIR-NBS-LRR (TNL) and non-TNL [192].
NBS-LRR genes have been extensively studied in various plant species such as A. thaliana, O. sativa, Populus trichocarpa (poplar), Zea mays (maiz), etc. [193]. These genes were also identified in grapes by various research groups [108,192,193]. It was revealed that they are widely represented in the grape genome, and localized in many chromosomes. Among all detected NBS-LRR genes, 63 are associated with response to powdery mildew infection. Some of them were found to localize at loci REN1 (chromosome 13), REN3 (chromosome 15), REN5 (chromosome 14), REN6 (chromosome 9), REN9 (chromosome 15) and RUN1 (chromosome 12) [193]. Based on the studies of the NBS-LRR genes in different plants, a correlation between the number of NBS-LRR genes and the size of plant genome was established (with the increase in size of plant genome the number of NBS-LRR genes increases) [194][195][196][197][198]. The E. necator susceptible NBS-LRR proteins were characterized as being about 1000 amino acids in size [193]. Various types and amounts of regulatory elements associated with stress-induced gene expression were identified in the sequence of promoter regions of NBS-LRR proteins: hormone responsive regulatory elements: Salicylic acid responsive element (TCA element), Ethylene responsive element (ERE), Methyl jasmonate responsive element (CGTCA and TGACG-motifs) and Abscisic acid responsive element (ABRE); defense and stress responsive element (TC-rich repeats); wound and pathogen responsive element (WUN-motif and W-box); stress responsive element (GT1 Box), cis-acting regulatory elements (CARE). TFs bind to these regulatory elements and activate the defense mechanisms of V. vinifera in response to stress conditions caused by powdery mildew, for example NAC TFs bind to GT1 box and WRKY binds to W-box [193]. It was observed that the levels and timing of expression of the NBS-LRR genes differ greatly in susceptible and resistant V. vinifera lines in response to E. necator. Some of the NBS-LRR genes are induced at later stages of powdery mildew infection, and its expression is higher in resistant lines.
Several NBS-LRR genes were detected to be expressed in resistant cultivars only, as well as several NBS-LRR genes of resistant cultivars that were involved in the early response to powdery mildew infection [193].
S-Genes, MLO
An alternative approach is based on obtaining recessive mutations of susceptibility genes (S-genes). Expression of S-genes is induced by pathogens, and encoded proteins suppress plant immunity system. Inactivation of S-genes required for the successful penetration of a particular pathogen does not cause a strong activation of plant defense. Their knock out leads to recessively inherited non-race-specific resistance, and is one of the main models of PTI non-host resistance [188,199]. S-gene-mediated resistance is usually more durable as the pathogen must overcome its dependence on S-factor of the host. The disadvantage of such resistance is a possible formation of pleiotropic phenotypes, for example, the formation of papilla in the absence of any pathogen and the premature onset of leaf senescence [200][201][202]. Some genes of the Mildew Locus O (MLO) family are a typical example of powdery mildew S-genes. For dicotyledons, these are all S-genes from group V of the MLO family [203][204][205]. In grapes, 17 representatives of the MLO gene family were identified, and suppression of a certain group of these genes led to resistance to E. necator [203][204][205][206][207]. For the first time resistance to powdery mildew mediated by inactivation of the MLO gene was revealed in barley in 1992 and for a long time has been considered as a unique form of resistance. Disruption of the MLO gene in barley provides persistent resistance to powdery mildew, which has been confirmed by the fact that during many decades no new E. necator strains capable to overcome this mlo-mediated resistance have been found in the fields [199]. Later, it was revealed that MLO genes are conserved within the plant kingdom, and their non-functional forms reduce the susceptibility to powdery mildew in some species, such as barley [208], Arabidopsis [209], pea [202], tomato [210], wheat [211] and pepper [212]. In Arabidopsis, 3 genes (AtMLO2, AtMLO6 and AtMLO12) were identified, which are induced in response to powdery mildew infection, and their joint inactivation led to complete resistance to powdery mildew [209].
In the study carried out by Feechan and colleagues (2008) and Winterhagen and colleagues (2008), three VvMLO genes were identified in grapes, in accordance with the Feechan nomenclature named: VvMLO3 (VvMLO11 (W) (W-Winterhagen's nomenclature)), VvMLO4 (VvMLO13 (W)) and VvMLO17, which showed similarity in homology of translation products sequences and transcriptional response to infection with other MLO genes of Arabidopsis and tomato, associated with susceptibility to powdery mildew. In response to E. necator rapid induction of genes VvMLO3, VvMLO4 and VvMLO17 and accumulation of large amount of their transcripts was detected [203,204,213]. MLOs are calmodulin-binding proteins. Pharmacological studies suggest that the influx of Ca2+ ions is important for MLO functioning [213,214]. Therefore, Ca2+ may be a candidate signal for the induction of the E. necator responsive genes VvMLO3, VvMLO4, VvMLO9 and VvMLO17, since plant cells generate a transient Ca2+ signal in response to pathogen attack [203,204,215,216]. Further research has revealed grape MLO genes in clade V, which are associated with susceptibility to powdery mildew: VvMLO6 (W), VvMLO7 (W) and VvMLO11 (W). It is likely that VvMLO6/VvMLO7 (W) can be a functional homologue of AtMLO2.
In the experiment, the incomplete knock out of these MLO genes via RNAi significantly reduced the susceptibility of V. vinifera to powdery mildew. VvMLO7 (W) is the main candidate gene for susceptibility to E. necator; it was inactive in resistant cultivars, and a decrease in its expression in transgenic mlo lines correlated with the level of susceptibility to powdery mildew. Inactivation of VvMLO6 (W) and VvMLO11 (W) was not directly related to susceptibility, the expression of these two genes was significantly reduced in both resistant and susceptible lines; however, both genes additionally contributed to increasing resistance [205,206]. Pleiotropic phenotypes were not found in mlo-lines under greenhouse conditions [205].
PM-Responsive REN and RUN Loci
In the genotypes of various grape species quantitative trait loci (QTLs) associated with powdery mildew resistance were identified: Run (Resistance to Uncinula necator) and Ren (Resistance to Erysiphe necator) [45]. These QTLs were mapped to chromosomes 9, 12, 13, 14, 15, 18, and 19, where various resistance gene analogs (RGAs), encoding resistance proteins of the TIR-NBS-LRR and CC-NBS-LRR types, are also localized. Loci Run, Ren provide qualitative resistance [193]. The main QTLs of the North American Muscat grape V. rotundifolia, associated with powdery mildew resistance, were located on chromosome 12-Run1 [217], and on chromosome 18-Run2.1 (V. rotundifolia "Magnolia") and Run2.2 (V. rotundifolia "Trayshed") [218]. The Run1 locus was found to comprise a family of seven putative Toll/ interleukin-1 receptor (TIR)-NB-LRR-type R genes [186]. Run1 confers ETI hypersensitivity phenotype, was introgressed from V. rotundifolia to V. vinifera. The locus Ren1 was identified on chromosome 13 in the cultivar «Kishmish vatkana» of V. vinifera [219]. Ren1 was located in the region of the genome containing NB-LRR sequences, which corresponds to its resistance phenotype (ETI). Hence, both Ren1 and Run1 may be non-durable considering their probable type of functioning. V. rupestris B38 showed an enhanced penetration resistance to a powdery mildew isolate with the ability to overcome the Run1 gene, making it an interesting resistance source to prolong the durability of this gene. V. rupestris B38 showed an enhanced penetration resistance to E. necator isolate with the ability to overcome the Run1 gene, making it an interesting resistance source to prolong the durability of this gene [44]. M.A. Dalbó and colleagues (2001) investigated a new locus of resistance on chromosome 14, later named Ren2, in a population of grape cultivars crossing: "Horizon" and resistant "Illinois 547-1" [45,220]. Run2, Ren1 and Ren2 have a lower incidence of programmed cell death after infection than Run1, which leads to further formation of secondary hyphae [221]. In the case of Ren1, the fungus can receive sufficient nutrition to maintain its life cycle; however, the level of sporulation is about 10 times lower than that observed in susceptible genotypes [219].
Locus Ren3 was identified on chromosome 15 in the cultivars of V. vinifera «Regent» and «Villard Blanc» [45,222]. Later, two regions were identified in this locus-more distinct boundaries of Ren3 and the new locus Ren9 [223]. Moreover, a new qualitative resistance locus, named Ren11, was found on chromosome 15. It was found to be effective in nearly all vineyard environments on leaves, rachis, berries, and stems. Locus on chromosome 8 and 9 had additive effects with Ren11 on the stems [224].
Locus named Ren4 was identified on chromosome 18 in the wild Chinese species V. romanetii [218,225]. Chromosome 18 contains large clusters of the TIR-NBS-LRR category of genes and surpasses all other chromosomes for the number of resistance genes [218]. Locus Ren8 was also identified on chromosome 18 during and after a dry growing season characterized by high temperatures. This locus is likely to mainly affect the leaf resistance [226].
The Ren5 locus was identified on chromosome 14 in V. rotundifolia, which affects both the development of mycelium and the intensity of sporulation. Ren5 manifests its effect after the formation of the first appressorium, delaying and then interrupting the development of mycelium. In the region of Ren5 locus, the presence of 7 NBS-LRR genes, as well as the VvEDR1 defense reaction regulator, was revealed. Thus, Ren5 is likely to belong to the class of NBS-LRR genes responsible for disease resistance [227].
The dominant powdery mildew resistance loci, designated as Ren6 and Ren7, were found in the wild Chinese V. piasezkii grapevine on chromosomes 9 and 19, respectively. Ren6 is associated with complete resistance, whereas Ren7 is responsible for partial resistant with reduced colony size. Genotypes containing Ren7 or Ren6 exhibited HR to powdery mildew infection. In the case of Ren7, HR was much more slower, it occurred in epidermal cells after the penetration of appressoria and the beginning of the secondary hyphae development, whereas in the Ren6 genotypes HR was more severe and was caused by appressoria of germinated spores [190].
The locus Ren10 was detected on chromosome 2, where the Myb-like genes controlling berries color are located [95,107,108,228,229]. Grapevine plants with the Ren10 haplotype showed less colonization by E. necator and decreased sporulation. Ren10 was probably introgressed from a North American species [95].
The major QTL associated with powdery mildew susceptibility in V. vinifera cultivar "Chardonnay" was located on chromosome 9 and named Sen1 (Susceptibility to Erysiphe necator 1) [230].
The majority of loci (Run1, Run2, Ren4, Ren5, Ren6 and Ren7) confer resistance to powdery mildew after the pathogens penetration in cells by inducing programmed cell death (ETI) [188,190,217,227]. The Ren6 locus contributes to a strong restriction of hyphal growth [187,190]. In the case of Ren4, resistance is associated with papilla formation [187,188].
Other Gene Systems for Resistance and Susceptibility to Erysiphe necator
The VvDOF3 TF was identified in V. vinifera, the overexpression of which enhanced resistance to powdery mildew through the SA signaling pathway. DOF proteins are plant-specific TFs that play an important role in plant growth, development, and defense responses. The highest level of VvDOF3 expression was observed in leaves and roots. Rapid induction of VvDOF3 expression occurred under the attack of E. necator and under the influence of SA and JA phytohormones. Transgenic A. thaliana overexpressing VvDOF3 was characterised by increased expression of the SA-sensitive PR1 gene and high concentration of SA [231]. Overexpression of VqTLP29 in V. quinquangularis caused activation of the SA signaling pathway, and could promote JA biosynthesis after infection with powdery mildew. Studies have shown that the level of PR5 expression increases after infection with powdery mildew, with an increase in the expression of PR1 and NPR1. Overexpression of PR5 can increase resistance to biotic and abiotic factors by activating many protective genes in the SA or JA/ET signaling pathway [160].
The WRKY gene family is one of the largest TF families in plants, playing an important role in diverse plant developmental and physiological processes, including plant defense signaling pathways. In grapes 59 identified WRKY genes were classified into 3 groups (I-III) [157]. Upon infection of grapes with E. necator, TFs VvWRKY19, VvWRKY48 and VvWRKY52 are activated [205]. The genes VvWRKY48 and VqWRKY52 belong to Group III, and gene VvWRKY19 belongs to Group II. VqWRKY52 was strongly induced by SA, but not by JA. In transgenic A. thaliana lines its expression was strongly induced 24 h after inoculation with powdery mildew, which contributed to an increase in resistance to E. necator, but at the same time it increased sensitivity to B. cinerea [146,157].
During investigations carried out on grapes, defense-related genes were identified in basal resistance, belonging to the innate immune system of plants, which does not depend on resistance genes: EDS1 (ENHANCED DISEASE SUSCEPTIBILITY 1), EDL2 (EDS1-LIKE 2), EDL5 (EDS1-LIKE 5), and PAD4 (PHYTOALEXIN DEFICIENT 4). The EDS1-PAD4 complex takes part in the SA-mediated defense pathway, and was detected to be more complicated in grapes than in Arabidopsis. EDS1 is a key regulator of basal resistance in various plants, including grapes. Pathogen effectors target EDS1. EDS1 expression is induced by SA and grapes resistance increases with increasing SA levels. Grapevines gene VvEDS1 is a functional ortholog of Arabidopsis gene AtEDS1, VvEDS1-LIKE (EDL)-paralog of AtEDS1. VvEDS1 is expressed higher in veins than in the rest of the leaf infected by E. necator [232]. In Arabidopsis, EDS1 interacts with PAD4 and SAG101, which in turn compete for EDS1 to form nuclear heterodimers. In the grapevine there was reported up to five SAG101-like genes. VvEDS1, VvEDL2 and VvPAD4 may form a complex to regulate the defense system. VvEDL2 was proposed to be a non-functional version of VvEDS1 and either competes with VvEDS1 when interacting with VvPAD4 or regulates the amount of functional VvEDS1 dimers. These protein interactions provide resistance to E. necator [232].
It is well known that an increase in the concentration of stilbene phytoalexins has an important role in global defense mechanisms and is observed in pathogen-resistant grape cultivars, when the growth of pathogen appressorium is suppressed [233]. In the Chinese wild grape species V. pseudoreticulata, resistant to powdery mildew, the VpSTS29/STS2 encoding stilbene synthase was identified. Stilbene synthase catalyzes the synthesis of various stilbenoids, including an antimicrobial low molecular weight natural phytoalexin-resveratrol and piceid, derivative glucosides of resveratrol. Overexpression of VpSTS29/STS2 in powdery mildew susceptible V. vinifera cultivars increases the amount of STS, which induces reprogramming of gene expression. The SA signaling pathway is activated, which induces transcriptional reprogramming of the downstream WRKY-MYB complex, which in turn regulates the synthesis of stilbenes and PR to enhance plant defense at the loci of infection. Expression of STS then promotes the production of the more toxic pterostilbene and ε-viniferin, which enhance basic immunity by inducing programmed cell death. Expression of the grape STS gene improves resistance to biotrophic and semi-biotrophic pathogens, but not necrotrophic organisms. The VqSTS5 gene from V. quinquangularis contributed to the resistance to powdery mildew infection [234]. Among the studied 31 VqSTS genes from V. quinquangularis, it has been revealed that VqSTS21 was up-regulated in response to E. necator. VqSTS21 overexpressing lines of Arabidopsis exhibited enhanced resistance to E. necator, via up-regulation of genes involved in SA-mediated signaling pathway, but displayed increased susceptibility to necrotrophic B. cinerea, with suppressed JA-mediated signaling pathway [139]. Another STS gene VaSTS19, isolated from a Chinese wild grape, V. amurensis. cv. "Tonghua-3", enhanced resistance to powdery mildew and gray mold in Arabidopsis transgenic lines, and contributed to the accumulation of more callose amount in comparison to nontransgenic control plants. Analysis of the expression of several diseaserelated genes suggested that VaSTS19 expression enhanced defense responses though SA and/or JA signaling pathways [235].
The Early-Response to Dehydration six-like (ERD6l) is one of the largest families of sugar transporters in plants. For grapevines, 18 genes of the ERD6l family were identified, one of which (VvERD6l13) was characterized in detail. When berries were infected either by necrotrophic pathogen B. cinerea or by the biotrophic E. necator, the VvERD6l13 gene was activated. VvERD6l13 localizes in the plasma membrane and mediates H+-dependent sucrose transport, thereby suggesting that VvERD6l13 is a sugar transporter involved in the sugar mobilization process and is induced in response to biotic stress. VvERD6l13 expression was detected at all developmental stages, with the highest level at the ripe berry stage. Two sucrose-sensitive elements were detected in the promoter region of this gene, as well as regulatory cis-acting elements, associated with both biotic and abiotic stress and responding to various hormones. It can likely facilitate the extraction of apoplasmic sucrose, thereby limiting the availability of nutrients for the pathogen and reducing the progression of infection [18].
The VvCSN5 gene (subunit 5 of the COP9 signaling complex), negatively associated with grapevine powdery mildew resistance, has been identified. Inactivation of VvCSN5 led to higher levels of expression in leaves of several marker genes linked to plant protection processes (VvPR1, VvPR3, VvPAD4 and VvRBOHD). As a result, an increased resistance to E. necator was noted to be manifested by the deposition of callose on the cell walls at the sites of pathogen penetration attempts and the death of infected epidermal cells. Gene VvCSN5 may be positively correlated with the JA-signaling pathway [191].
Aspartic proteases (AP), one of the superfamilies of proteolytic enzymes, are the extracellular enzymes that may play a role in the degradation of pathogenesis-related proteins in leaves. APs are involved in the regulation of many biological processes, such as the recognition of pathogens, pests and the induction of protective reactions [236,237]. In grapes, 50 AP genes have been identified. It was revealed that the AP13 gene is susceptible to infection caused by E. necator, as well as to salicylic acid exposure. As a consequence of infection caused by B. cinerea and exposure to MeJA, the VqAP13 transcript levels in V. quinquangularis cv. 'Shang-24 decreased, but increased when exposed to ET and SA. Transgenic A. thaliana lines overexpressing VqAP13 showed increased resistance to E. necator and accumulated more callose than wild-type plants, while its resistance to B.
cinerea inoculation was reduced. The results suggest that functioning of VqAP13 promotes the SA-dependent signaling pathway but suppresses the JA signaling pathway [148].
Genetic Editing Capabilities
CRISPR/Cas9-targeted mutagenesis technology allows editing of plant genomes, potentially disabling one or more genes in order to study their function, and is a valuable tool for creating disease-resistant varieties [99,100]. This two-component system is based on the Cas9 nuclease, which is able to cut a double-stranded DNA molecule in a selected target in the presence of a guide RNA (gRNA). To date, there have been several reports on the use of CRISPR-Cas9 for genome editing in grapevine [154,207,[238][239][240][241]. The application of CRISPR/Cas9-targeted mutagenesis allowed to reveal that VvMLO11 (W) belongs to S-genes promoting infection of grapes with powdery mildew, and the targeted mutation of VvMLO11 (W) leads to enhanced resistance to powdery mildew in edited lines, which was associated with host cell death, cell wall apposition (CWA) and H 2 O 2 accumulation. A homozygous Vvmlo3 mutant seedling with knocked out VvMLO3 (VvMLO11 (W)) died from massive leaf necrosis in tissue culture medium. Significantly increased resistance to powdery mildew was observed in several heterozygous VvMLO11 (W) lines in comparison with WT plants, which suggested that VvMLO11 (W) is one of the functional homologues of AtMLO2 in grapevines. Simultaneous directed mutagenesis of all three MLO genes (VvMLO6 (W), VvMLO7 (W) and VvMLO11 (W)) may be necessary for the development of complete resistance of grapevines to powdery mildew in the future [207]. The VvWRKY52 gene knock out in grapes by means of CRISPR/Cas9-targeted mutagenesis led to a transgenic line phenotype with increased resistance to B. cinerea [154].
Conclusions
Over the past decade, significant progress has been made in the study of the grapes system of genes for resistance and susceptibility to Erysiphe necator and Botrytis cinerea. The most important direction of the future viticulture is based on infections suppression with the least amount of chemicals, development and introduction of new pathogen-resistant varieties, preservation and enhancement of nutrient content, obtaining higher yields, and restoration of crop properties unintentionally lost during domestication. Currently, breeding and interspecific hybridization have been replaced by new technologies of cisgenesis / intragenesis and editing of genes (or genome) as more relevant new biotechnologies. This has improved the breeding efficiency and has dramatically increased the range of plant properties that can be obtained.
Author Contributions: Writing-original draft preparation and editing, J.F.; writing-review and editing, Y.U., N.T., E.K. and R.I.; conceptualization and supervision, R.I. and E.K. All authors have read and agreed to the published version of the manuscript.
Funding:
The work is supported by the Sirius University of Science and Technology project: GNZH-RD-2008. | 2022-02-20T16:20:06.757Z | 2022-02-17T00:00:00.000 | {
"year": 2022,
"sha1": "641ba3cce6ba7f3e138397504ca8d1c95b6a256d",
"oa_license": "CCBY",
"oa_url": "https://www.mdpi.com/2073-4395/12/2/499/pdf",
"oa_status": "GOLD",
"pdf_src": "Adhoc",
"pdf_hash": "eccf6ebdc2eca249a9a198c8b36fdeaae300667e",
"s2fieldsofstudy": [
"Agricultural And Food Sciences",
"Biology"
],
"extfieldsofstudy": []
} |
270866375 | pes2o/s2orc | v3-fos-license | The History of Classification Systems for Periprosthetic Femoral Fractures: A Literature Review
Periprosthetic femoral fractures (PPFFs) following total hip arthroplasty (THA) present a significant clinical challenge due to their increasing incidence with an aging population and evolving surgical practices. Historically, classifications were primarily based on anatomical fracture location, the stability of the implant, and bone quality surrounding the implant. We critically analyzed 25 classification systems, highlighting the emergence and adaptations of key systems such as the Vancouver classification system (VCS) and the Unified classification system (UCS), which are lauded for their simplicity and effectiveness yet require further refinement. VCS, developed in 1995, categorizes fractures based on the site, implant stability, and bone quality, and remains widely used due to its robust applicability across different clinical settings. Introduced in 2014, UCS expands the VCS to encompass all periprosthetic fractures with additional fracture types, aiming for a universal application. Despite their widespread adoption, these systems exhibit shortcomings, including the incomplete inclusion of all PPFF types and the imprecise assessment of implant stability and surrounding bone loss. These gaps can result in misclassification and suboptimal treatment outcomes. This paper suggests the necessity for ongoing improvements in classification systems to include emerging fracture types and refined diagnostic criteria, ensuring that they remain relevant to contemporary orthopedic practices and continue to facilitate the precise tailoring of treatment to patient‐specific circumstances. This comprehensive historical review serves as a foundation for future innovations in classification systems, ultimately aiming to standardize PPFF treatment and improve patient prognosis.
Introduction
P eriprosthetic fracture has become a frequent complica- tion of total hip and knee arthroplasties with rates up to 15% in major registries. 1 The most common fracture localization of the periprosthetic fracture is the femur, and periprosthetic femoral fracture (PPFF) after total hip arthroplasty (THA) has become a severe clinical disease which has been found to be much more common after uncemented hip replacement. 2,3][19][20][21][22][23][24] More young patients with end-stage hip disease are willing to undergo THA, and tend to have a higher range of motion after surgery, which also increases the risk of PPFF. 25,26The Mayo Clinic Joint Registry identified the overall incidence of the PPFF following THA was 4.1% from 1969 to 1999. 27In 2016, the data further updated by the Mayo Clinic Joint Registry from 1969 to 2011 showed that the incidence of intra-operative PPFFs was 1.7% and 20-year probability of post-operative PPFFs was 3.5% after primary THA, while the incidence of intra-operative PPFFs was 12% and 20-year probability of post-operative PPFFs was 11% after revision THA. 19,28According to the cases analysis reported from the Swedish National Hip Arthroplasty Register of periprosthetic fracture after THA, the accumulated incidence of PPFFs was 0.4% for the primary THA and 2.1% for the revision of THA. 29 According to the Australian Prosthesis Registry, the incidence of periprosthetic fractures associated with THA was 10%. 30A multicenter retrospective cohort study in the UK revealed that patients with PPFFs were predominantly elderly, frail, female, and more likely to experience the unified classification system B type fractures (the fracture around the stem or just below it), with females being a significant predictor for PPFF type. 313][34] Since PPFFs frequently occur in the elderly with severe bone loss, surgical complications are common, making treatment challenging and prognosis poor. 35For instance, the rate of PPFF-related infection, a typical surgical complication, stands at 7.3%.This commonly results in multiple re-operations, possible non-union, a decreased clinical function, and long-term antibiotic treatment. 363][54][55] The peri-implant periprosthetic survival analysis, a multicenter study from Spain, reported a 6.2% inhospital mortality rate for PPFFs. 56][59] In comparison with primary THA, 60 it is a significant clinical challenge for orthopedic surgeons to take aseptic loosening, bone loss and osteoporosis as well as general state of patients into consideration when treating PPFF. 11,29,40- 42,61-63In addition, each periprosthetic fracture poses a unique challenge to the orthopedic surgeons because of the variables including the relationship of the fracture to the implant, the specifics of the implant including wear, and the functional demands of the patient also should be taken into account. 64To obtain an ideal prognosis, it is very important to establish reasonable treatment principles and providing patients with the individualized treatment. 5Take history as a mirror, we can know what is going on.Understanding the history of classification systems for PPFF will allow us to better understand and grasp the treatment of PPFF.This study provides a literature review of reported classification systems for PPFF to provide reference for the classification and standardized treatment of PPFF in our clinical work and provide evidence for further improvement of PPFF classification system.
Methods
A comprehensive literature search was conducted using PubMed and the Cochrane Library, employing the keywords "periprosthetic femoral fracture," "periprosthetic fracture," and "peri-implant fracture."The search encompassed literature available until March 10, 2024.Additionally, the review included a search of published reports from the American Association of Orthopedic Surgeons (AAOS) annual meeting proceedings spanning the years 1990 to 2023.
The inclusion criteria specified studies discussing the classification systems or new types of PPFF.The exclusion criteria comprised: (i) duplicate articles; (ii) articles without full-text access; (iii) irrelevant titles, abstracts or full texts; and (iv) editorials, letters, and commentaries.
During the literature selection process, we conducted a quality assessment to ensure the reliability and relevance of the review findings, adhering to established inclusion and exclusion criteria.Evaluation metrics included study design, methodological rigor, the appropriateness and accuracy of outcome measures, and reproducibility.Furthermore, references of all relevant articles were manually cross-referenced to identify potential additional studies.Ultimately, 25 articles were included (Figure 1).
The Basis Factors for the Classification Systems for PPFF
B ased on the literature search and quality assessment, we identified 25 classification systems or new types for PPFF (Table 1) (Figures 2 and 3).6][67][68][69][70][71][72][73][74][75][76][77][78] The classification systems vary, with some focusing on the anatomical location of the fracture, while others emphasize the stability of the implant and the surrounding bone stock.We describe the history and evolution of classification systems for PPFF in the following sections: (i) the classification systems for PPFF at the early stage; (ii) the milestone of classification systems for PPFF-the Vancouver classification system (VCS); (iii) the modification for VCS-the unified classification system (UCS); (iv) disadvantages of the VCS and UCS; (v) other classification systems supplementary to VCS and UCS; and (vi) the modified version of the UCS for PPFF.
The Classification Systems for PPFF at the Early Stage B ased on the anatomical location of the fracture, Parrish and Jones classified PPFF into four types: fractures in the trochanteric area, fractures in the proximal portion of the shaft, fractures in the mid-shaft, and fractures in the distal part of the shaft of the femur. 65Whittaker et al. classified PPFF into three types according to the relationship of the stem of the prosthesis with the fracture line and the degree of displacement: Type I, at the intertrochanteric level about the middle portion of the stem; Type II is an oblique or spiral fracture about the stem tip; and Type III, fracture is severely displaced with the distal fragment dislodged from the stem of the prosthesis. 66Based on the location relative to the femoral component of the total hip prosthesis, Johansson et al. classified PPFF into three types: Type I, fractures were those in which the fracture occurred proximal to the tip of the prosthesis; Type II, fractures were those in which the fracture line extended from the proximal portion of the femoral shaft to beyond the distal tip of the prosthesis; and Type III, fractures were those in which the fracture was entirely distal to the tip of the prosthesis. 67According to the anatomical location of the fracture, Bethea et al. extended the classification to three types: Type A, those that fractured at the tip of the femoral component, either transversely or with a distal spiral, were designated; Type B, spiral fractures around the femoral component were designated; and Type C, fractures proximal to the tip of the stem with comminution around the stem were designated. 68Based on the anatomical location of the fracture, Cooke and Newman classified PPFF into four types and proposed treatment regimens: 69 Type 1, comminuted fractures around the stem of the prosthesis.Treatment regimens: early revisions; Type 2, oblique or spiral fractures around the shaft of the prosthesis.Treatment regimens: conservative treatment or revision operation; Type 3, fractures are transverse at the level of the tip of the prosthesis and are unstable.Treatment regimens: primary internal fixation; and Type 4, fractures are entirely distal to the prosthesis.Treatment regimens: internal fixation.Jensen et al. emphasized the stability of the post-injury prosthesis as a predictor of prosthetic revision and classified PPFF into three types: Type 1, located around the proximal two thirds of the prosthetic femoral stem; Type 2, extending proximally and distally from the area of the femoral stem tip; and Type 3, extending distally from the area of the femoral stem tip. 70Roffman and Mendes classified PPFF according to the stability of the prosthesis after THA. 71Based on the fracture site and morphology, PPFFs were classified into five types by Schwartz et al.: Type A, involving the greater trochanter; Type B, longitudinal fracture of proximal femur; Type C, longitudinal fracture of distal femur; Type D, complete metaphyseal fracture; and Type E, complete fracture in distal femur prosthesis. 79AAOS put forward a classification system according to the anatomical location of the fracture: Type I, proximal to the intertrochanteric line; Type II, vertical splitting fracture, not exceeding the inferior margin of the lesser trochanter; Type III, fracture extends below the inferior margin of the lesser trochanter, but not exceeding the lower 1/3 junction of the prosthesis; Type IV, fracture spanning the prosthesis tip; Type IVa, spiral fracture; Type IVb, transverse fracture; Type V, severely comminuted Type-III and Type-IV fracture, extends below the stem tip; and Type VI, fracture distal to the stem tip. 72According to the anatomical location of the fracture, Mont and Maar divided PPFF into six types: Type 1, intertrochanteric; Type 2, proximal femur; Type 3, spanning the prosthesis tip; Type 4, distal to the prosthesis tip; Type 5, comminuted, blowout; and Type 6, supracondylar. 73According to the anatomical location of the fracture, Beals and Tower classified PPFF into four types: Type I, fracture in the trochanteric region; Type II, proximal metaphyseal/disphyseal fractures that do not involve the stem tip; Type IIIA, proximal diaphyseal fractures at the stem tip with less than 25% disruption of the prosthetic interface; Type IIIB, proximal diaphyseal fractures at the stem tip with greater than 25% disruption of the prosthetic interface; Type IIIC, supracondylar fracture at the tip of a long femoral stem; and Type IV, supracondylar fracture distant to the stem tip. 74hese classifications did not take into account all three factors of the site of fracture, the stability of implant and surrounding bone stock at the same time, so their guiding significance for the treatment of PPFF was limited.
The Milestone of Classification Systems for PPFF-The Vancouver Classification System (VCS) D uncan and Masri proposed the Vancouver classification system (VCS) in 1995, which was established according to the site of fractures, the stability of the femoral implant and the surrounding bone quality of the proximal femur. 75CS is the most widely accepted classification system nowadays. 69,80It classified PPFF into three types: A, B and C, of which Type A includes two subtypes and Type B includes three subtypes (Table 2, Figure 4).Type A is the fracture in trochanteric region.Type A is further subdivided into Type A G involving the greater trochanter and Type A L involving the lesser trochanter.Type B is the fracture around the stem or just below it.Type B is further subdivided into B1 (Well-fixed stem), B2 (loose stem with good proximal bone stock), and B3 (loose stem with poor-quality bone stock).Type C is the fracture occurring well below the tip of the stem.
According to VCS, the treatment regimens for each type as follows (Table 2): 8,26,46,59,[81][82][83][84][85][86][87] Type A, conservative treatment is recommended if the fracture displacement is small, and internal fixation is recommended if the fracture displacement is large and associated with severe pain and dysfunction; Type B1, open reduction and internal fixation (ORIF) or THA revision with a long stem; Type B2, THA revision with ORIF; Type B3, THA revision with ORIF and management of bone defects such as bone graft; Type C, ORIF.
The clinical effectiveness of the VCS for PPFF has been extensively validated through several studies.Studies have consistently demonstrated the VCS's robust interobserver and intraobserver reliability and its validity across different clinical settings and populations.Naqvi et al. reported substantial agreement among consultants (κ = 0.69) and trainees (κ = 0.61) in using the VCS, with even higher intraobserver reliability ranging from 0.74 to 0.90. 40The system also showed strong validity, aligning well with intraoperative findings with an 81% agreement rate and a κ value of 0.68.The study noted a potential issue with the underdiagnosis of implant instability when relying solely on preoperative radiographs, suggesting the necessity for intraoperative assessments to ensure accurate treatment planning.Brady et al. further corroborated these findings by assessing the VCS's reliability and validity, showing that the VCS helps in accurately reflecting the surgical and clinical realities and can be applied consistently across different levels of clinical expertise. 42Moreover, Rayan et al. extended the validation of the VCS into a European context, confirming its applicability and reliability across different geographic and clinical environments. 45This cross-continental validation supports the universal applicability of the VCS, ensuring that it holds under various healthcare systems with different levels of resources and expertise.However, Lee et al. identified limitations within the context of cementless femoral stems, especially in differentiating stable (B1) from unstable (B2 and B3) fractures. 88The interobserver reliability showed moderate agreement (κ = 0.45), indicating a potential need for refinement of the VCS when used with newer implant technologies.
The VCS provides a structured framework for classifying PPFF, guiding clinical decisions based on the site of fractures, implant stability and bone quality.This system not only facilitates clear communication among clinicians but also helps in planning surgical interventions and predicting outcomes.It promotes standardized care across different clinical settings and by various medical professionals.while the VCS is invaluable for managing PPFF, its moderate κ values in certain contexts highlight the necessity for ongoing evaluation and adjustment to ensure its continued relevance in all surgical scenarios.
The Modification for VCS-The Unified Classification System (UCS) I n 2014, Duncan and Haddad proposed UCS for all-site periprosthetic fractures. 5The UCS aims to reach the following goals: 89 (i) by adding Type D, Type E and Type F, the VCS was expanded and updated; (ii) so regardless of the which bone is broken and the joint involved, the groupings and treatment principles can cover all periprosthetic fractures; and (iii) after the popularization, the classification can eliminate language barriers and we may communicate with simplicity and clarity.For the PPFF, the UCS classification added relatively rare and new types of the fracture (Type D, Type E, Type F) and expanded the VCS (Table 3, Figure 5).Type D is the fracture of the femoral shaft between the hip and knee replacements.Type E is the fracture involving both the acetabulum and femur after hip replacement.Type F is the fracture involving only the acetabulum after hip replacement.
The treatment principles of Type A, B and C in UCS were the same as those in VCS.1][92] Furthermore, Thomas et al. reported that monoblock tapered stems can be an acceptable modality in the treatment of Type B2 fractures, while Capone et al. indicated that for Type B2 and B3 fractures, revision with modular stems is preferable to monoblock stems. 91,93A multicenter study has indicated that for Type B fractures around cemented polished taper-slip femoral components, ORIF is associated with lower reoperation rates, shorter waiting times for surgery, reduced transfusion needs, and lesser demands for critical care compared to revision surgery. 94Conversely, research by Scalici et al. found that patients with Type B and C fractures undergoing revision surgery had better functional outcomes than those treated with ORIF. 92Therefore, there is no definitive operative technique for all UCS B fractures at present.Meanwhile, several studies also applied the principles of treatments for new types as follows (Table 3): 5,89,95-99 Type D, logical individual treatments (revision of one, both or neither joint replacement and ORIF); Type E, logical individual treatments according to the "block-out" analysis in relation to each component of the joint replacement; Type F, logical individual treatments (non-operative approach with protected weight-bearing, delayed and relatively straightforward conversion to THA, initial displacement).
The clinical effectiveness of the UCS for PPFF has been scrutinized across various studies.The UCS, aimed at providing a comprehensive framework for all periprosthetic fractures, demonstrates high reliability and validity across different clinical settings and fracture types.Vioreanu et al. validated its high interobserver and intraobserver reliability, suggesting that it could be consistently applied by different clinicians across international settings. 100Huang et al. confirmed the reliability of the UCS, observing substantial agreement in clinical assessments. 101Specifically, for 23 Type B cases, there was 79.71% agreement with intraoperative 2.
findings and a mean κ value of 0.694 for validity, indicating substantial agreement.De Meo et al. reported the UCS showed a slightly higher validity (κ = 0.64) compared to the VCS (κ = 0.56), suggesting its enhanced ability to match intraoperative findings. 102Nevertheless, Jain et al. pointed out the UCS has moderate reliability but only fair validity when applied to PPFF around cemented polished taper-slip (PTS) stems. 103The findings indicate that UCS may not adequately define PTS stem loosening, indicating the need for potential refinements in its use.Moreover, the lack of perfect agreement in the κ values across these studies further confirms the necessity for ongoing improvements to the UCS.Furthermore, Schopper et al. concluded that despite the UCS's advancements, the VCS remains more frequently used in the literature and clinical practice, likely due to its established efficacy and simpler application. 104In summary, the UCS possesses the capability to improve surgical planning and outcome prediction, while also necessitating ongoing adjustments and validations to ensure its clinical effectiveness.
Limitations of the VCS and UCS
P revious studies have shown that there were two following limitations during the clinical practice of the VCS and UCS: (i) not all types of PPFF are included; and (ii) there are individual variations in the evaluation of the stability of femoral prosthesis and the judgment of bone loss in proximal femur.
First, some fracture types cannot be grouped into either of the above two classifications, so doctors and researchers often encounter confusion in the choice of treatment options, or even choose the wrong treatment.For example, Mallory et al., 105 Van Houwelingen and Duncan, 77 and Capello et al. 76 reported that the pseudo A LT periprosthetic fractures including a segment of the proximal medial femoral cortex, which is actually the new B 2 fracture.
FIGURE 5
The unified classification system for periprosthetic femoral fracture.For definitions of the fracture types, see Table 3.
Egrise et al. reported the clinical and radiological results and identified risk factors of this type of fractures. 106Fan et al. reported the variant A GT periprosthetic fracture (a fracture of the greater trochanter with lateral cortical extension) and femoral stem destabilization, which is also actually the new B 2 fracture. 107These fracture patterns have not been previously described in the VCS and the UCS.Without a correct type differentiation, the treatment will surely become a failure.In addition, PPFF accompanied by stem fracture after hip arthroplasty were not classifiable under the original VCS or the UCS.
Second, Orthopedic surgeons have different opinions about the stability of femoral prosthesis in Type B fracture and the presence of bone loss in the proximal femur, which even may bring about inaccurate treatments. 108,109It has been pointed out definitely in the verification articles of the VCS and UCS. 40,42,45,100,101Grammatopoulos et al. retrospective reviewed the radiographs of some patients who were considered to have a Vancouver B1 fracture, found some subtle features, such as subsidence of the stem into the centralizer, that were characteristic of a B2 fracture pattern. 110If we cannot distinguish this fracture pattern from Vancouver B1, it may lead to treatment failure.
Other Classification Systems Supplementary to VCS and UCS N inan et al. presented the Coventry classification system, which grouped the periprosthetic fractures into "happy hips" and "unhappy hips" based on stability of the stem.According to the therapeutic principle of VCS, fracture fixation alone is required in the "happy hips," and revision of the prosthesis is needed in the "unhappy hips." 111 Capello et al. presented their modified version of VCS. 76Type TG: same as type AG of VCS.Type TL: same as type AL of VCS.Type A1: fracture of medial cortex that includes the residual neck, calcar, and lesser trochanter and is displaced medially, with a well-fixed stem.Type A2: fracture of medial cortex that includes the residual neck, calcar, and lesser trochanter and is displaced medially, with a loose stem.Type B1: same as type B1 of VCS; Type B2: same as type B2 of VCS.Type B3: same as type B3 of VCS.Type C: same as type C of VCS.This study reported the new A category which represented the pseudo A LT periprosthetic fractures supplementing the VCS.The new A category was described to encompass the "clamshell" pattern supplementing the B type of VCS with A1 nominating this fracture with a stable stem and A2 nominating this fracture with an unstable stem.The "clamshell" fracture originates at the medial base of the greater trochanter and ends at the distal medial cortex of the lesser trochanter, and the lateral cortex is intact.Similarly, Mallory et al., 105 Van Houwelingen and Duncan 77 also reported the periprosthetic fracture of the lesser trochanter including a segment of the proximal medial femoral cortex which is not the traditional Type A L but really the Type B2 fracture.
Huang et al. proposed PPFF accompanied by stem fracture supplementing VCS and UCS, and provided corresponding treatments. 78They were divided into two types (Type A and Type B), and Type B has two subtypes.Type A: PPFF accompanied by stem fracture after hip arthroplasty in which the proximal portion of the fractured femoral prosthesis is stable.Treatment regimens: remove the distal portion of the fractured femoral prosthesis, followed by open reduction and internal fixation (ORIF).If the patient is in poor physical condition and a future revision is not possible, ORIF alone is performed without removal of the distal portion.Type B: PPFF accompanied by stem fracture after hip arthroplasty in which the proximal portion of the fractured femoral prosthesis is unstable.Type B is divided into two subtypes.Type B1: the proximal portion of the fractured femoral prosthesis is loose and the surrounding bone quality is good.Treatment regimens: revision of THA is performed with a longer stem.Type B2: the proximal portion of the fractured femoral prosthesis is loose and the bone bed is of poor quality.Treatment regimens: revision of THA with a particular stem.The study also identified three cases of periprosthetic femoral fractures accompanied by stem fracture after hip arthroplasty through retrospective analysis, which were not classifiable under the VCS and UCS.This validated the clinical relevance and effectiveness of the new fracture pattern classification.Similarly, Malhotra et al. also presented the unusual fractures through the femoral shaft and the femoral prosthesis and discussed the treatment of removal of the broken implants and subsequent revision replacement surgery. 112or remedying the defects of the VCS and UCS in evaluating the stability of femoral prosthesis and overcome PPFF failures by objective evaluation, Baba et al. presented new classification for PPFF. 43Type 1: cementless stem.Type 1A: the main fractured region involves the porous-coated area which is the region firmly bonding the stem to bone and is necessary to stabilize the stem, so the stem is likely to be unstable when this region is fractured.Type 1B: the main fractured region is outside the porous-coated area.Since the stem is firmly bonded to bone in the porous-coated region, the stem is stable when the fractured region is distal to it.Type 2: cemented stem.Type 2A: The main fractured region involves the stem.Fixation of bone and cement may be broken, and the stem is likely to be unstable.Type 2B: the main fractured region is distal to the stem.The stem is stable because fixation of bone and cement are not broken, and fracture occurs distal to it.The reliability and validity of the Baba classification was improved when plain radiograms, CT imaging and implant information were given. 113When supplemented with additional diagnostic information, this classification system exhibited superior interobserver reliability (κ = 0.94) compared to the VCS, and demonstrated substantial intraobserver reliability among experts (κ = 0.81).Additionally, the Baba classification achieved a 95% concordance with intraoperative findings, underscoring its clinical validity and utility in accurately guiding surgical interventions.They also presented treatment regimens for each type as follows: Types 1A and 2A, intra-operative stem stability test is necessary, the combination of stem revision and ORIF when the stem is unstable and ORIF alone when the stem loosening is not confirmed; Types 1B and 2B, revision with a long stem or ORIF if it is the implant-tip fracture of a cement stem and ORIF alone if not.Stoffel et al. introduced an algorithmic approach to identify loose stems around proximal femoral periprosthetic fractures using patient history, stem design, and plain radiographs, which can help to detect stem loosening and improve curative effect. 114rammatopoulos et al. described the 'spiral' fracture pattern 110 and Phillips et al. described the comminuted 'burst' pattern 115 with cemented stems in B2 type of VCS.The "spiral" fracture is often in association with a separate wedge fragment and significant comminution.The comminuted "burst" pattern often splits along the cement mantle, similar to an "ax."Based on the spiral, comminuted "burst," and the previously described clamshell pattern 76 in B2 type of VCS, Karam et al. reported the newly observed "reverse clamshell" in B2 type. 116This kind of fractures has an intact medial cortex, and originates from the medial calcar and passes through the lateral cortical outlet.All these fracture patterns belong to B2 fractures which indicate the implant instability, and the treatment regimen still refers to the treatment principle of B2 in VCS.
Gonz alez-Martín et al. divided the proximal femur into three zones (medial, lateral and distal) according to the Gruen system 117 and propose a sub-classification of B2 type in VCS to help: B2.1 (1 fractured zone) and B2.2 (≥2 fractured zones). 118They suggested B2.1 type treated via ORIF had a lower risk of complication than B2.2 type.
These classification systems supplementary to VCS and UCS aim to address specific scenarios and shortcomings in the VCS and UCS.Each system introduces unique categories and treatment strategies based on fracture characteristics, implant stability, bone quality or implant design, enhancing the diagnostic and therapeutic precision for PPFF.These supplementary systems collectively enrich the original frameworks, offering a more nuanced approach to managing the diverse and complex scenarios encountered in PPFF treatment.
The Modified Version of the UCS for PPFF I n 2018, based on previous study in 2015, 78 Huang et al. further modified the application of UCS in the femur side through literature search and case collection and provided corresponding treatment methods 119 (Table 4, Figure 6).The modifications of UCS are as follows: (1) add two new B2 subtypes: B2PALT/B2PAGT (the pseudo ALT/AGT: fracture in trochanter region including a segment of the proximal medial/lateral femoral cortex); (2) add a new FS category to encompass stem fracture alone or accompanied by PPFF, with FSO designating stem fracture alone, FS1 designating this fracture with the proximal portion of the fractured femoral prosthesis being stable, FS2 designating this fracture with the proximal portion of the fractured femoral prosthesis being loose and the surrounding bone quality being good, and FS3 designating this fracture with the proximal portion of the fractured femoral prosthesis being loose and the bone bed being of poor quality; and (3) delete Type F which does not apply to the femur.
The modified version of the UCS also applied the treatment regimens for the new types as follows (Table 4): Type B2PALT and B2PAGT, revision THA with a longer stem, along with ORIF of the fracture; Type FS (FSO, FS1, FS2, FS3), removal of the stem, and revision THA with logical individual treatment.In addition, it mentioned that if fracture is at the tip of the stem in B1 type, the cortical allograft struts were recommended for use.And a new study further confirmed that the longer stem revision and internal fixation (LSRIF) which obtained the excellent clinical outcome is the main treatment for Type B2PALT fractures, while simple ORIF is an option for elderly patients in poor condition. 120his new modification compensated for the deficiency of some fracture types in classic UCS and VCS and refined the application on the femur side.Fan et al. evaluated the clinical effectiveness of this modified UCS for PPFF. 121he study revealed almost perfect agreement among consultants (κ = 0.882) and substantial agreement among trainees (κ = 0.776) for interobserver reliability.Intraobserver reliability showed substantial to almost perfect agreement, with κ values ranging from 0.701 to 0.972.Validity assessments involving 299 type B cases demonstrated 89.854% agreement with intraoperative findings (κ = 0.849), indicating almost perfect agreement.These findings affirm the modified UCS's high reliability and validity in classifying PPFF, enhancing its utility in precisely evaluating implant stability and informing treatment decisions for PPFF.They also suggested that the trainee surgeons mastered the original VCS and subsequent UCS, while consultants mastered the modified UCS. 102s mentioned above, Huang et al. suggested the cortical allograft struts for fracture at the tip of the stem in B1 type on the basis of the stable implant. 119Fractures at the tip of the stem have also been reported in other studies.Yasen et al. recommended to use the cortical allograft struts with ORIF when the implant is stable, and a long stem revision with ORIF when the implant is loose for the treatment of fractures at the tip of the stem. 59Baba et al. indicated that the therapeutic policy for implant-tip fracture was controversial, revision with a long stem or ORIF may be chosen by different surgeons. 43oo et al. adopted a long-stem femoral prosthesis augmented with autologous bone grafts and a new cable-plate construct for the fracture at the tip of the stem, and fracture healing was achieved. 97Hence, the fracture at the tip of the stem is a special pattern of PPFF which is difficult to treat because of the lack of rotational stability and may not be the common B1 type.Current classification systems do not mention the requirement for special classification of the fracture at the tip of the stem.Whether it requires special classification perhaps remains to be further analyzed in future studies.
The ultimate goal of the classification systems for PPFF is to guide treatment and improve prognosis.In order to achieve this goal, the classification system must reflect the conditions of injury comprehensively, be reliable and effective, and be easy to master.The VCS and UCS for PPFF has been shown to be reliable, valid, and easy to master, but there is still room for further improvement.
Discussion
67][68][69][70][71][72][73][74]79 Developed over decades, these systems provide a structured approach to categorizing fractures based on their anatomical location or implant stability, aiding in clinical decision-making.They have several advantages.First, these systems innovatively categorized PPFF according to specific anatomical details, improving diagnostic precision by clearly identifying fracture locations-critical for planning targeted interventions.Additionally, by delineating fractures into distinct categories based on location or implant stability, these early systems enabled the development of standardized treatment protocols-from conservative management to revision surgery-tailored to each fracture type.This standardization has streamlined treatment processes and established care benchmarks applicable across various healthcare settings, thus enhancing consistency and optimizing outcomes.Furthermore, the structured categorization fostered by these systems has bolstered clinical research and communication within the orthopedic community by standardizing study inclusion criteria, facilitating outcome comparisons across different patient groups, and improving the efficiency of findings dissemination.These systems also paved the way for future advances in classification systems by highlighting the early models' limitations, allowing new systems to incorporate additional crucial factors like bone quality.
The classification systems at the early stage, which focused solely on anatomical location or implant stability, provided foundational frameworks for understanding and managing PPFF but had notable disadvantages.][69][70][72][73][74]79 First, inadequate evaluation of implant stability can lead to unsuitable treatment choices, as the system fails to effectively navigate between conservative management, fracture fixation, or revision surgery needs.Inadequate initial surgical repairs often fail to provide long-term stability and functionality, significantly increasing the likelihood of requiring secondary revision surgery.Furthermore, implant instability can lead to continuous micro-movements at the fracture site, which not only delays healing but may also cause further damage to the boneprosthesis interface.This can reduce the functional lifespan of the implant and potentially lead to a cascade of mechanical failures.In addition, Bone defects add complexity to the surgical repair of PPFF, often rendering standard fixation techniques like plates, screws, or intramedullary rods insufficient.As a result, revision surgeries and bone grafting frequently become necessary components of the treatment plan.To manage the instability caused by bone defects, enhanced fixation methods using longer or more robust prosthetic stems may be required to provide stability across the fracture site.Ignoring the surrounding bone quality can result in poor outcomes such as loosening of the implant, nonunion or malunion of the fracture, increased pain, and limited mobility.Alternatively, classification systems that are solely focused on implant stability often fall short by FIGURE 6 The modified version of the unified classification system for periprosthetic femoral fracture.For definitions of the fracture types, see Table 4.
overlooking both the precise fracture morphology and location, which are crucial for selecting the appropriate surgical strategy and ensuring biomechanical integrity, as well as ignoring the surrounding bone quality, leading to potentially detrimental effects. 71he VCS and UCS are pivotal in the classification of PPFF.The advantages of these two classification systems lie in the simultaneous consideration of the site of fractures, the stability of the femoral implant and the surrounding bone quality, thus making up for the shortcomings of the classification systems at the early stage. 69,75,80The VCS facilitates precise intervention strategies by distinguishing between fracture types A, B, and C, each subdivided further to reflect different clinical scenarios.The UCS expands on the VCS by including additional fracture types (D, E, F), thereby covering all periprosthetic fractures and aiming to simplify global communication within the orthopedic community. 5,89][102] Despite their extensive applications, both VCS and UCS have limitations that may affect their clinical utility.The VCS and UCS are sometimes criticized for not encompassing all PPFF types, which can lead to potential misclassifications and inappropriate treatment choices.6][107] Additionally, the clinical effectiveness of these systems in differentiating between stable and unstable fractures, particularly in the context of the application of cementless femoral stems or cemented PTS stems, has shown only moderate agreement, suggesting a need for refinement. 88,1039][110] Continuous validation and adjustment, informed by clinical outcomes and emerging research, will enhance their reliability, validity and applicability in all surgical scenarios.
The introduction and adaptation of various classification systems supplementary to the VCS and UCS have significantly enhanced the management of PPFF.These systems provide tailored approaches to diverse fracture scenarios, showcasing their advantages.The Coventry classification system introduced by Ninan et al. categorizes fractures based on stem stability, simplifying treatment decisions between "happy hips" that require fracture fixation alone and "unhappy hips" that need prosthesis revision. 111Capello et al. introduced a modified version of the VCS to address pseudo ALT fractures, offering a nuanced view that enhances diagnostic accuracy and treatment precision.This inclusion of specific fracture types such as the "clamshell" pattern further refines surgical planning. 76The classifications proposed by Huang et al. provide a framework for dealing with stem fractures, a scenario inadequately covered by previous systems, thus enhancing treatment specificity and potentially improving outcomes. 78Moreover, the inclusion of unusual fractures through the femoral shaft and the femoral prosthesis by Malhotra et al. highlighted the need for removal of broken implants and revision surgery, emphasizing the classification's adaptability to diverse clinical scenarios. 112The Baba classification system addresses the limitations of previous systems by focusing on implant design and the physical integrity of stems, which is critical for setting appropriate therapeutic strategies. 43This system's ability to differentiate between the stability of cementless and cemented stems based on the location of the fracture relative to the porouscoated area enhances treatment specificity.The classification by Gonz alez-Martín et al. introduces a sub-classification for B2 fractures, facilitating more targeted treatment options based on fracture zone risks, which could reduce complications associated with over or under-treatment. 118While these adaptations expand the applicability of the VCS and UCS, they have several disadvantages, such as complicating the clinical decision-making process.Each new sub-classification introduces additional criteria that clinicians must consider, which can vary significantly in different clinical settings and require specific diagnostic tools not universally available.For instance, the system by Baba et al. enhances the detail by differentiating fractures based on the location of the break relative to the implant's porous-coated area, which is crucial for determining the stability of cementless stems. 43However, this requires precise imaging and can lead to variability in treatment approaches. 113Huang et al. proposed classifications for PPFF with stem fractures, which were not previously categorized, highlighting the need for specific treatments. 78While these additions aim to cover the broad spectrum of PPFF scenarios, they also require precise diagnostic capabilities and may increase the cognitive load on clinicians, potentially complicating decision-making processes.Additionally, these classification systems may rely on advanced imaging for precise application, which may not be applicable to all clinical settings, especially those with limited resources.These systems collectively expand the granularity with which PPFF can be assessed and treated, providing frameworks that potentially improve patient outcomes by allowing for more customized treatment plans.However, their effectiveness depends on the clinical setting and available resources, necessitating ongoing evaluation and adaptation to ensure they remain relevant and beneficial across all orthopedic practices.
The modified UCS introduced by Huang et al. 119 presents several distinct advantages.By integrating new subtypes, such as B2PALT/B2PAGT for specific trochanteric fractures, and the FS category for stem fractures, this system offers a more granular and precise classification.This level of specificity aids in accurately assessing the stability of the implant and the quality of the surrounding bone, which is crucial for determining appropriate treatment strategies.Such detail-oriented categorization helps to standardize treatment approaches, potentially reducing variability in clinical outcomes and improving prognostic accuracy.Additionally, the expansion of fracture types beyond the scope of the original VCS and UCS addresses previous limitations, allowing for a more comprehensive evaluation of complex fracture scenarios.However, the modified UCS also presents certain disadvantages.The complexity introduced by the additional fracture subtypes and categories may pose a steep learning curve, particularly for less experienced clinicians, potentially leading to inconsistencies in fracture classification and treatment.This complexity could impede the widespread adoption of the system, as thorough training and familiarity with the classification are required to utilize it effectively.Moreover, while the system enhances the granularity of classification, it may rely on detailed clinical data and imaging not always available in all healthcare settings, limiting its practical applicability in less resourced environments.These factors suggest that while the modified UCS offers substantial improvements in the classification and management of PPFF, its application may be constrained by the specificities of clinical practice settings and the expertise of the medical personnel.
Conclusion
T he history of classification systems for PPFF following THA reveals a nuanced evolution, shaped by the growing complexities of patient demographics, surgical advancements and a deepening understanding of patient outcomes over several decades.Most classification systems at the early stage primarily focused on the anatomical location of fractures.These foundational frameworks established the basis for further innovations by identifying crucial areas requiring enhancement, notably the integration of evaluations concerning the surrounding bone quality and the implant stability.
The introduction of the VCS marked a significant milestone by offering a more comprehensive framework that considered not just the location of the fracture but also the implant stability and the bone quality.Its widespread adoption underscored the system's utility in facilitating communication among clinicians and standardizing treatment approaches.Furthermore, the UCS was developed to create a more inclusive framework by expanding the categories of the VCS, aiming to standardize the classification of all periprosthetic fractures regardless of their location.This was a step towards universal applicability, intending to simplify global communications.However, the VCS and UCS occasionally fail to encompass all fracture types, and there are individual variations in the evaluation of the stability of femoral prosthesis and the judgment of bone loss, which may lead to misclassification and suboptimal treatment strategies.
The evolution of classification systems for PPFF has paralleled advances in orthopedic surgery and a deeper understanding of biomechanics and patient-specific factors.
The integration of emerging technologies and interdisciplinary approaches promises to refine these systems further, enhancing both diagnostic precision and treatment efficacy.Looking forward, several improvements are expected to further improve these systems: (i) the integration of advanced imaging techniques, such as high-resolution MRI and 3D computed tomography, can provide more detailed insights into fracture morphology and the biomechanical environment around the implant.These technologies could help refine fracture classification by revealing subtle features that current systems may overlook, such as micro-cracks in the bone or early signs of implant loosening; (ii) incorporating biomechanical modeling and simulation into classification systems could significantly improve the predictive accuracy of treatment outcomes.By simulating different scenarios of implant and bone interaction, doctors could predict potential complications and tailor interventions more precisely to individual patient anatomy; (iii) AI and machine learning offer robust tools for analyzing large datasets to predict outcomes, personalize treatment plans and identify previously unrecognized patterns in fracture presentation.By applying these technologies, future classification systems could dynamically incorporate real-time data from clinical outcomes to update and refine prognostic models continuously.;(iv) enhancing collaboration between engineers, biologists, computer scientists, and clinicians could foster the development of integrative classification systems that reflect both clinical realities and theoretical advancements.This would not only enhance the functionality of these systems but also ensure they are adaptable to the rapid pace of technological change; and (v) there is a pressing need to ensure that new classification systems are validated across diverse populations and healthcare settings.This would involve multicenter studies to test the applicability and reliability of revised classifications, ensuring they are robust and universally applicable.
In summary, while the VCS and UCS have set a strong foundation, the dynamic nature of medical science necessitates their continuous evolution.The integration of advanced technologies and multidisciplinary approaches will likely play a critical role in the next generation of classification systems, ensuring they are capable of addressing the complex and varied needs of the global patient population.These advancements will further enhance our ability to deliver personalized, precise, and effective treatments to patients with PPFF.
FIGURE 1 A
FIGURE 1 A brief flow chart of selection process.
FIGURE 2
FIGURE 2 Timeline of the overall development of classification systems for periprosthetic femoral fracture.
FIGURE 4
FIGURE 4 The Vancouver classification system for periprosthetic femoral fracture.For definitions of the fracture types, see Table2.
TABLE 1
Classification systems and their bases.
TABLE 2 The
Vancouver classification system for PPFF.
B3 Loose stem with poor-quality bone stock ORIF with THA revision and management of bone defects Type C Fracture occurring well below the tip of the stem ORIF Abbreviations: ORIF, open reduction and internal fixation; PPFF, periprosthetic femoral fractures; THA, total hip arthroplasty.
TABLE 3
The unified classification system for PPFF.
TABLE 4
Modified version of the unified classification system. | 2024-07-02T06:17:30.469Z | 2024-06-30T00:00:00.000 | {
"year": 2024,
"sha1": "4455495b513422197e732ca4b97c9abf14d9700e",
"oa_license": "CCBYNCND",
"oa_url": "https://doi.org/10.1111/os.14149",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "6cb541d19768dca23d1d3c7ccc395f4621e28c6d",
"s2fieldsofstudy": [
"Medicine",
"History"
],
"extfieldsofstudy": [
"Medicine"
]
} |
213670004 | pes2o/s2orc | v3-fos-license | Introductory Chapter: Swarm Intelligence - Recent Advances, New Perspectives, and Applications
Swarm intelligence has emerged as one of the most studied artificial intelligence branches during the last decade, constituting today the most high-growing stream on bioinspired computation community [1]. A clear trend can be deduced by analyzing some of the most renowned scientific databases available, showing that the interest aroused by this branch has been in crescendo at a notable pace in the last years [2]. Undoubtedly, the main influences behind the conception of this stream are the extraordinarily famous particle swarm optimization (PSO, [3]) and ant colony optimization (ACO, [4]) algorithms. These meta-heuristic lighted the fuse of the success of this knowledge area, being the origin and principal inspiration of their subsequent research. Such remarkable success has led to the proposal of a myriad of novel methods, each one based on a different inspirational source such as the behavioral patterns of animals, social and political behaviors, or physical processes. The constant proposal of new methods showcases the capability and adaptability of this sort of solvers to reach a near-optimal performance over a wide range of high-demanding academic and real-world problems, being this fact one of the main advantages of swarm intelligence-based meta-heuristics.
Introduction
Swarm intelligence has emerged as one of the most studied artificial intelligence branches during the last decade, constituting today the most high-growing stream on bioinspired computation community [1]. A clear trend can be deduced by analyzing some of the most renowned scientific databases available, showing that the interest aroused by this branch has been in crescendo at a notable pace in the last years [2]. Undoubtedly, the main influences behind the conception of this stream are the extraordinarily famous particle swarm optimization (PSO, [3]) and ant colony optimization (ACO, [4]) algorithms. These meta-heuristic lighted the fuse of the success of this knowledge area, being the origin and principal inspiration of their subsequent research. Such remarkable success has led to the proposal of a myriad of novel methods, each one based on a different inspirational source such as the behavioral patterns of animals, social and political behaviors, or physical processes. The constant proposal of new methods showcases the capability and adaptability of this sort of solvers to reach a near-optimal performance over a wide range of high-demanding academic and real-world problems, being this fact one of the main advantages of swarm intelligence-based meta-heuristics.
Brief history of swarm intelligence
The consolidation of swarm intelligence paradigm came after years of hard and successful scientific work and as a result of the proposal of several groundbreaking and incremental studies, as well as the establishment of some cornerstone concepts in the community.
In this regard, two decisive milestones can be highlighted in swarm intelligence history. First of these breakthrough landmarks can be contextualized on horseback between the 1960s and 1970s. Back then, influential researchers such as Schwefel, Fogel, and Rechenberg revealed their first theoretical and practical works related to evolving strategies (ES) and evolutionary programming (EP) [5][6][7]. An additional innovative notion came to the fore some years later from John H. Holland's hand. This concept is the genetic algorithm (GA, [8]), which was born in 1975 sowing the seed of the knowledge field today known as bioinspired computation. All the three outlined streams (i.e., ES, EP, and GA) coexisted in a separated fashion until the 1990s, when they all erected as linchpin elements of the unified concept evolutionary computation.
The second milestone that definitely contributed to the birth of what currently is conceived as swarm intelligence is the conception of two highly influential and powerful methods. These concrete algorithms are the ACO, envisaged by Marco Dorigo in 1992 [9], and the PSO [10], proposed by Russell Eberhart and James Kennedy in 1995. Being more specific, the PSO was the method that definitely lit the fuse of the overwhelming success of swarm intelligence, being the main inspiration of a plethora of upcoming influential solvers. Therefore, since the proposal of PSO, algorithms inheriting its core concepts gained a great popularity in the related research society, lasting this acclaim until the present day [11][12][13]. For the modeling and design of these novel approaches, many inspirational sources have been considered, commonly categorized by (able to collect these sources in three recurring groups): • Patterns found in nature: we can spotlight two different branches that tie (fall) together within this category. The first one is related to biological processes, such as the natural flow of the water (water cycle algorithm, [14]), chemotactic movement of bacteria (bacterial foraging optimization algorithm, [15]), pollination process of flowers (flower pollination algorithm, [16]), or geographical distribution of biological organisms (biogeography-based optimization, [17]). The second inspirational stimulus is the behavioral patterns of animals. This specific trend is quite outstanding in recent years, yielding a design based on creatures such as bats (bat algorithm, [18]), cuckoos (cuckoo search, [19]), bees (artificial bee colony, [20]), or fireflies (firefly algorithm, [21]).
• Political and social behaviors: several human conducts or political philosophies have also inspired the proposal of successful techniques. Regarding the former, we can find promising adaptations of political concepts such as anarchy (anarchic society optimization, [22]) or imperialism (imperialist competitive algorithm, [23]). With respect to the latter, social attitudes have been also served as inspiration for several methods such as the one coined as society and civilization [24], which emulates the mutual interactions of human and insect societies, or the hierarchical social meta-heuristic [25], which mimics the hierarchical social behavior observed in a great diversity of human organizations and structures.
• Physical processes: physical phenomena have also stimulated the design of new swarm intelligence algorithmic schemes, covering a broad spectrum of processes such as gravitational dynamics and kinematics (gravitational search algorithm, [26]), optic systems (ray optimization, [27]), or the electromagnetic theory (electromagnetism-like optimization, [28]). A recent survey published by Salcedo-Sanz [29] revolves around in this specific sort of methods.
In addition to the above-defined categories, many other fresh branches spring under a wide range of inspirations such as business tools (brainstorming optimization, [30]) or objects (grenade explosion method, [31]).
It is also worth mentioning that besides these monolithic approaches aforementioned, there is an additional trend which prevails at the core of the research activity: hybridization of algorithms. Since the dawn of evolutionary computation, many efforts have been devoted to the combination of diverse solvers and functionalities aiming at enhancing some capabilities or overcoming the disadvantages of well-established meta-heuristic schemes. Obviously, memetic algorithms (MAs), conceived by Moscato and Norman in the 1980s in [32,33], beat this competition. Despite MAs were initially defined as hybridization of GAs and local search mechanisms, MAs rapidly evolved to a broader meaning. Related to SI, today is straightforward to find hybridization of SI meta-heuristic schemes with separated local improvement and individual learning mechanisms in the literature. Some examples of this research trend can be found in [34][35][36][37][38].
Finally, up to now, SI methods have been applied to a wide variety of interesting topics along the years. Being impossible to gather in this introductory chapter all the applications already addressed by SI paradigms, we refer the reader to some remarkable and highly valuable survey works specially devoted to outline the application of SI algorithms in specific domains. In [39] a survey dedicated to geophysical data inversion was published. In [11] the latest findings of portfolio optimization are studied. An additional interesting work can be found in [12] focused on summarizing the intensive work done related to the feature selection problem. Intelligent transportation systems are the crossroads of the works gathered in [40], while in [41] authors conducted a comprehensive review of SI metaheuristics for dynamic optimization problems. We acknowledge that the literature focused on all these aspects is immense, which leads us to refer the interested readers to the following significant and in-depth surveys [42][43][44].
Motivation behind the book edition
With reference to the scientific production, SI represents the most high-growing stream in today's related community, with more than 15,000 works published since the beginning of the twenty-first century. Analyzing the renowned Scopus® database, a clear upward trend can be deduced. Specifically, scientific production related to SI grows at a remarkable rate from nearly 400 papers in 2007 to more than 2000 in 2018. In fact, the interest in SI has been in crescendo at such a pace that the number of published scientific material regarding this field is greater than other classical streams such as evolutionary computation every year since 2012.
Thus, and taking advantage of the interest that this topic arises in the community, the edited book that this chapter is introducing gravitates on the prominent theories and recent developments of swarm intelligence methods and their application in all the fields covered by engineering. This material unleashes a great opportunity for researchers, lecturers, and practitioners interested in swarm intelligence, optimization problems, and artificial intelligence as a whole. | 2019-12-05T09:05:18.458Z | 2019-12-04T00:00:00.000 | {
"year": 2019,
"sha1": "265792a63ee9e1e47e1dce148484421b2009e7bb",
"oa_license": "CCBY",
"oa_url": "https://www.intechopen.com/citation-pdf-url/69981",
"oa_status": "HYBRID",
"pdf_src": "ScienceParsePlus",
"pdf_hash": "002b783ddb6107fae75cbff7bee261f831dd6f58",
"s2fieldsofstudy": [
"Computer Science"
],
"extfieldsofstudy": [
"Engineering"
]
} |
232145241 | pes2o/s2orc | v3-fos-license | On the Certificate Revocation Problem in the Maritime Sector
. Maritime shipping is currently undergoing rapid digitalization, but with increasing exposure to cyber threats, there is a need to improve the security of the ship communication technology used during operations across international waters, as well as close to local shores and in ports. To this aid, there are ongoing standardization efforts for an international maritime Public Key Infrastructure, but the inherent properties of limited connectivity and bandwidth make certificate revocation a problematic affair compared to traditional Internet systems. The main contribution of this paper is an analysis of certificate revocation techniques based on how they fulfil fundamental maritime requirements and simulated usage over time. Our results identify CRLs (with Delta CRLs) and CRLite as the two most promising candidates. Finally, we outline the pros and cons with these two different solutions.
Introduction
Maritime shipping is currently undergoing rapid digitalization. The introduction of new communication technologies onboard ships, such as the upcoming VHF Data Exchange System (VDES) [21], enables a wide variety of new digital services, such as digital ship reporting, electronic port clearance, search and rescue communications, vessel traffic services and broadcast of maritime safety information. These services will all require information security, and a prevalent solution to establish and deploy a Public Key Infrastructure (PKI) for distributing digital certificates and securing the integrity and confidentiality of the information exchange. Several different research groups have in parallel worked to define, implement and test the characteristics of a PKI for the maritime domain [7,12,13,23], and the concept has now been acknowledged by IMO 1 and brought into the ongoing standardization by IALA 2 [5]. However, there is a significant challenge related to certificate revocation yet to be solved. This is a crucial part of any PKI, and unlike typical Internet applications that can be constantly online, ships tend to be offline for long periods of time or sailing in the open sea where connectivity and bandwidth can be both poor and expensive. The consequences could be delayed awareness of revocations and less trust in the PKI itself. Even though some previous work has evaluated and compared different revocation mechanisms [15,27,28,30], there are no previous studies that address the specific maritime challenges.
The main contribution of this paper is an analysis of certificate revocation techniques based on requirements fulfilment and simulated use in a maritime setting over time. The paper is organised as follows. Section 2 provides the background to our work, including a description of the envisioned maritime PKI as well as an overview of existing solutions for certificate revocation. Section 3 presents the fundamental requirements and Sect. 4 gives an analysis of and simulation benchmarks for the solution candidates. Section 5 discusses these results and Sect. 6 concludes the paper.
The Maritime PKI
A normal PKI depends on a hierarchy of trust, e.g. as depicted in Fig. 1. There are three layers in this model: 1. A Trusted International Root Certificate Authority (CA) that issues certificates to its subordinates. The root CA is envisioned to be operated by IMO, since they are a trusted entity by the majority of maritime stakeholders around the world. 2. A number of (intermediate) Issuing National CAs, that would typically be the Flag State administrations associated with each country. 3 3. End entities, which are the ships, ports and coastal services that need to communicate securely.
In addition, an entity called "Revocation issuer" responsible for issuing information about invalid certificates, will be needed. This role could also be handled by the root CA.
Existing Revocation Solutions
Revocation of certificates in a PKI ecosystem can happen for a number of reasons, such as change of ship name, change of association between the end entity and the issuing CA, or compromise of the corresponding private key [10]. Affected entities should be informed as fast as possible after a revocation, and we have described existing revocation mechanisms that we have considered for the maritime sector. Certificate Revocation List (CRL). (RFC 5280 [10]) are issued at regular time intervals. When a user wants to check the validity of a certificate, he needs to have a local copy of the CRL installed. This is usually be achieved by pulling the CRL from the CA's CRL distribution endpoint. While this method works well for PKIs with relatively few end entities, the solution does not scale well, since a full and complete CRL will list all (unexpired) certificates that have been revoked. To counter this problem, delta-CRLs can be used, which only include certificates whose revocation status has changed since last update. A drawback with CRLs (and delta-CRLs) is that the validity check is done "offline" and there is a risk that end entities accept certificates that have been revoked. The use of CRLs to revoke certificates in the Maritime PKI has been proposed by Froystad et al. [13] and the Maritime Cloud Development Forum [12]. Figure 2 presents the principle: a CRL issuer collects CRLs from all entities entitled to issue such lists, and creates a joint CRL that is distributed to all the end entities in the PKI. However, neither [13] nor [12] specify the use of delta-CRLs or discuss the risk of using obsolete CRLs. CRLs from multiple sources are collected and distributed to end entities through a CRL issuer [13].
Online Certificate Status Protocol (OCSP). (RFC 6960 [26]), was designed to provide more timely revocation information compared to CRLs. The protocol is "online"; to verify a certificate's validity, the user sends an OCSP request to the CA, asking for the status of one or more given certificate(s). The CA will respond with the certificate(s)'s status (good/revoked/unknown), along with a response validity interval. However, the solution comes with some drawbacks: -If the CA is unavailable, the validity of certificates cannot be confirmed.
-Increased latency since the OCSP request needs to be confirmed before the certificate can be used. If the CA is unreachable, it can take several seconds before the request times out, 4 which will be a no-go for time critical applications. -While the OCSP response is signed, error messages are not. This may open up to interception attacks [18]. -The identity of the user is revealed to the CA each time the user sends an OCSP request, possibly creating privacy issues.
To counter OCSP bandwidth issues, RFC 5019 [11] defines a lightweight profile that minimises the communication bandwidth and client-side processing.
OCSP Stapling/OCSP Must Staple.
To solve some of OCSP's problems, RFC 6961 [24] and RFC 7633 [25] (commonly referred to as "OCSP Stapling" and "OCSP Must Staple", respectively) define a method where a server makes a request to the OCSP service and get a signed message with its certificate status that it can then "staple" to its certificate during the Transport Layer Security (TLS) handshake. This method is more efficient than the original OCSP because the server can cache OCSP messages, thus reducing the latency. This also partially solves the privacy issue with OCSP, as the user does not need to reveal its identity each time. Still, OCSP Stapling suffers from being an "online" solution, which is not suitable for all maritime applications.
CRLSet [3]/OneCRL. [1] are currently used by Google and Firefox to revoke CA certificates stored in their web browsers. Revocation lists are built internally by the respective companies and pushed to the clients daily. The lists only include a subset of the most critical revoked certificates. The main benefit is that bandwidth is minimised. In contrast to OCSP, these solutions do not reveal communication patterns, as all end entities receive the same lists. The common downside of CRLSet and OneCRL is that end entities must be online on a daily basis.
CRLite. allows to push all TLS certificate revocations to all browsers. Initially described in a research paper by Larisch et al. [19], it has since 2019 been tested by Mozilla Firefox as its new certificates revocation method [8]. CRLite relies on cascading bloom filters to efficiently push out the revocation information, which is "a simple space-efficient randomized data structure for representing set to support membership queries" [22]. The answer to a membership query is probabilistic; if the answer is negative, we know the element is not in the set, but if the answer is positive then chances are that this is a false positive.
To remove false positives, CRLite takes advantage of Certificate Transparency (CT), which aims to fix several flaws in the SSL certificate system by "providing an open framework for monitoring and auditing SSL certificates in nearly real time" [2]. In practice, certificate logs are built by collecting all certificates issued by trusted CA [2,20]. In the context of CRLite, a first bloom filter is b.uilt using the revoked certificates data and then, using the CT log, a new bloom filter (smaller than the first one) containing all false positives of the first bloom filter is built. This operation can be repeated until there is no more false positive, thus the name of cascading bloom filters. Larisch et al. [19] emphasize the following advantages of CRLite over CRLs and OCSP: -Small Size: About 10 MB is needed to represent the status of all Web certificates, and 560 kB in average for daily updates. While OCSP also covers all certificates, it is online. CRL on the other hand is not efficient to handle so many certificates. -Update Frequency: An OCSP response is valid for 4 days in average for the Web, and a CRL has usually a lifetime of 7 days. CRLite is believed to be more up-to-date because it involves daily update. -Failure Mode: CRLite covers all certificates and allows end entities to operate in a hard fail mode. -Privacy: The privacy issues caused by OCSP are avoided as CRLite caches the information locally. -Deployment and Auditing: CRLite is easily deployed and can be audited.
-Speed: Telemetric data shows that CRLite is faster than OCSP in 99% of the cases [16].
There is also some criticism towards the Bloom filter cascades. Holzhauser [14] shows that CRLite will suffer a scalability problem when the number of certificates increase. She also shows faults in the CRLite design that can lead to higher than expected number of false positives.
Short-Lived Certificates. Which are valid from a few hours up to 2-3 days, represent an alternative strategy. Firefox for instance, does not check such shortlived certificate for revocation if they are valid for less than 10 days [8]. This will efficiently remove the need to revoke certificates, given an acceptable risk that illegitimate keys will probably not be used before the certificates expire. The obvious downside is the need for frequent certificate renewal and distribution, which does not go hand in hand with maritime operations.
Fundamental Requirements for Revocation
We perform an initial filtering of suitable revocation mechanisms by identifying the fundamental requirements imposed by the maritime sector. These are presented below. While some ships call at port on a regular basis, others might be out at sea for several weeks or even months [13]. They usually rely on different technologies to get connectivity depending on their position: VDES, SatCom, GSM/LTE or even Wi-Fi when at shore. Table 1 developed by Frøystad et al. [13] gives an overview of their properties. Some will in many cases be too expensive or not available at all while at sea. It should therefore be possible to use a cache of revocation information, e.g. when vessels encounter each other on open sea.
Requirement 1 -Offline support:
The revocation mechanism should be able to operate when the vessel has no internet connectivity.
Chrome and Firefox currently push incomplete CRLs with only high value certificates in them, but this is not acceptable in the maritime context. The solution needs to be complete, meaning that revocations must be shared between intermediate CAs and eventually known to all end entities. Of course, with a solution that operates offline from time to time, there will be some delay before a revocation information is available to the end entities. The update frequency of the revocation information is left for discussion.
Requirement 2 -Completeness:
The revocation mechanism should inform about all revoked vessels in a timely manner.
While the revocation mechanism should not be dependent on the communication link, its bandwidth usage should be as low as possible to ensure acceptance affordability in the wider maritime community. In practice, this means within the capacities given by Table 1.
Requirement 3 -Bandwidth:
The bandwidth usage of the revocation mechanism should be within the capacity of the available communication link.
Requirements-Based Selection of the Revocation Mechanism
We now use the requirements identified in Sect. 3 to do an initial filtering of the candidate solutions.
R1 -Offline support:
Amongst the previously presented revocation mechanisms, only CRL, DeltaCRL, CRLSet/OneCRL and CRLite are truly offline mechanisms, meaning that the ships can stay off-line for a long period of time and the mechanism will still work, even though it will be on outdated data. On the contrary, OCSP Stapling, OCSP Must Staple and Short-Lived Certificate all rely on periodic connection to update their staple or certificate in order to work. They are therefore not considered viable solutions for the maritime sector. R2 -Completeness: All revocation mechanisms but CRLSet and OneCRL are complete or can be. CRLSet and OneCRL by definition only include high value revocation information to allow a quick reaction to critical events like of a CA compromise.
R3 -Bandwidth:
This parameter is difficult to evaluate. A known problem with CRL is their growing size when the number of certificates in the system growths. However, the deltaCRL is less dependent on this parameter, and more on the system revocation ratio. CRLite was conceived with lowbandwidth usage in mind, but it has only been applied to the web by Mozilla, and the web has very different parameters than the maritime sector. Table 2 shows which mechanisms fit our requirements the best. As can be seen, CRL/DeltaCRL and CRLite meet al.l three requirements, but there is an uncertainty on their respective bandwidth usage, which we analyse further below.
Bandwidth Analysis for CRL/DeltaCRL and CRLite
In order to estimate bandwidth requirements for the different revocation solutions, we need to estimate some parameters for the PKI. This includes the expected number of certificates in the system and the expected revocation frequency.
The number of merchant ships in the world is varying, but as of January 2019, there were 96,295 registered ships in the total fleet world wide, whereof 51,684 were commercial ships of 1000 gt and above [6]. In addition there will be shore users communicating with the ships (ports, VTS, applications human users etc.), but this number is lower than 1000. To simplify, we approximate that the total number of end entities that will be enrolled in the maritime PKI will be around 100 000. Based on the simulation, we observed that the result remains the same with more entities in the PKI, which covers the case of several certificates per ship.
Revocation of the digital certificates from the end entities in a PKI can happen for a number of reasons, but in the maritime domain, change of flag is expected to be the main driver for revocation. We foresee that the frequency of other reasons are negligible in comparison. All commercial ships must be registered with a country, which is known as its Flag State. Ships normally change their flags in connection with sale and purchase transactions. However, ship owners may also do this to avoid the stricter marine regulations imposed by their own countries. In practice, many ships are therefore registered under a flag that does not match the nationality of the vessel owner ("flag of convenience"). The Flag State with the largest number of registered ships is Panama (6465 ship as of January 2019), followed by China (4039 ships), Liberia (3456) and the Marshall Island (3454) [6]. Even though the total number of Flag States is fairly large (117 as of January 2019), we do not foresee that all of these will operate their own Intermediate CA. However, a ship will still need to obtain a new digital certificate when it changes its flag. A study from 2008 [9] provides an indication of the frequency of flag changes. The study uses data collected from 35,261 port state control inspections on 7,547 vessels, carried out between 2002 and 2008. The data shows that 25.3% of all the inspected ships have had at least one change of flag during this time period. Further, 9.5% of all the ships have had at least one change in flag since their previous inspection, where the average number of inspections per ship in this time period was 4.05. Unfortunately, the paper does not include information on how many times the ships have re-flagged, when they have changed their flags "more than once," but we can use an approximation of the yearly revocation ratio of around 5%.
We also need to know how long the certificates will be valid. This will impact our analysis, because when revoked certificates expire, they will not be included in the transmitted revocation information anymore. The validity period of end entity certificates in a maritime PKI was studied by several independent research groups [12,13] who proposed to set it to 3 years. We thus chose to fix this parameters to 3 years for our simulation.
Theoretical Approach 5
Size of the CRL: In order to evaluate the size of the CRL, we will consider the scenario in which the CRL contains the most certificates. Given that the certificates have a 3-year validity period, they are removed from the CRL once they are not valid anymore. Thus the maximum number of certificates in the CRL is 15 000 (3 × 100000 × 0.05). After doing some tests, we calculated that in average the size (in bytes) of a CRL is given by S(n) = 277 + 50 × n, where n is the number of certificates to be included in the CRL (with a reason code), 277 the size of an empty CRL in bytes and 50 the average size added by the addition of a certificate to the CRL (empirically determined). So, for the maritime sector we end up with a maximum CRL size of 750 kB.
Size of the DeltaCRL. We need to have an idea of the number of certificates that will be included in it, both newly revoked certificates and those that need to be removed from the CRL. If we make the assumption that the CRL and DeltaCRL are issued on a weekly basis, then there is around 100 newly revoked certificates for each DeltaCRL. In addition, we will assume that there is about the same number of certificates removed from the CRL. Following the same formula as above, the size of the DeltaCRL should be around 10 kB.
Size of the CRLite Filter:
We need to find an estimation of the size of the filter that needs to be downloaded by the end entities in order to check for the revocation status of a certificate. In their original paper on CRLite [19], the authors present a way to set the parameters of the different filters to have the smallest possible size of the overall bloom filter cascade. We followed that methodology and chose the filters' parameters to minimise the overall size of the bloom filter cascade. For those condition, the overall size of the filter is given by: S bf c = 4.92 * |R|, where R is the set of revoked certificates. In our case, the result yields 73 800 bits, or 9.2 kB.
Size of the Delta CRLite Filter:
There is no easy way to theoretically estimate the size of the delta filter for CRLite. This needs to be determined in an empirical manner.
As it can be observed from the estimated sizes of the different "payload" for each mechanism, these sizes are smaller than what is normally found and used on the Internet. However, there is still a 75-factor between the size of the CRL and the size of the optimised CRLite filter. The DeltaCRL is about the size of the optimised CRLite filter.
Empirical Approach
In order to get a better idea of the sizes for the different revocation mechanisms, we developed a PKI simulator, consisting of a Root CA, Intermediate CAs, End entities and a CRL issuer. The parameters of the simulator are its duration in time, the revocation ratio and frequency, along with the PKI (number of intermediate CA and end entities). The simulator can also determine the growth of the PKI, as we can assume that not all ships will be part of it from day one, and thus get a better idea of what will happen when a real PKI is be deployed. The following steps are taken for each iteration: 1. Renew certificates that are about to expire. 2. Revoke random certificates based on the revocation ratio parameter. Our implementation of the CRLite bloom filter cascade follows Mozilla's implementation available on Github [17] and the implementation of the delta filter follows what is described in the CRLite paper [19].
For the parameters, we used 100 000 end entities and a revocation ratio of 5% as above. The revocation frequency was set to 7 days, which is the common revocation frequency for a CRL. We also estimated that the PKI will start with 1000 entities and then grow to 100 000 over a 5 year period. However, this was only to get an idea of the system evolution when integrating new components. What we really care about is the system in its "steady" state, which is why ran the simulation over a period of 20 years. Figures 3 and 4 Table 3 present the results of the simulation.
CRLite Vs CRL:
The results presented in Fig. 3 show that the size of the CRL is much bigger than the size of any other mechanism, with an average size of 356 kB for the simulation. Even if this is much smaller than CRLs from the Internet world, this is still too big to be downloaded over a low-speed network. As presented in Table 3, even if the size of the DeltaCRL remains small (with an average of 8.6 kB), the delta filter is even smaller with an average size of 2 kB. It is also interesting to note that the size of the delta filter is almost constant once the system has reached its equilibrium (no more ship being added), and is not much influenced by big changes in the end entities certificates. Indeed, the certificates having a 3-years validity, large amount of already revoked certificates expire every 3-years, leading to substantial changes in CRL (and thus deltaCRL as well). The "waves" pattern that can be observed is the direct consequence of the initial certificates' expiration. The size of the filter is in average 39.5 kB, but like the CRL, it is based as a reference to get the delta filter, and is not sent over low-speed communication channels. The size of the optimised filter, calculated every day for comparison, is close to the size of the DeltaCRL, with an average size of 10.1 kB. Based on the analysis of the simulation, it seems like CRLite is indeed well suited for low-bandwidth usage, not only for the Internet world, but also for the maritime sector, with a payload being five times smaller than the DeltaCRL.
CRL and DeltaCRL
Using solely CRL as the revocation mechanism for the maritime PKI is not possible as shown by the simulation: the size of the PKI grows fast and does not meet the low-bandwidth usage requirements. It must be coupled with the use of DeltaCRL.
The main advantage of using CRL/DeltaCRL is that it is a well-known and standardised revocation mechanism. It is also already implemented in commercial PKI solution and is thus more easily acceptable. However, ships can stay at sea for long periods of time, and might have to download several DeltaCRL (or the full CRL) to catch up. Moreover, it is known that CRL/DeltaCRL does not scale well for the Internet. This is true as well for the maritime sector where the constraints are even more strict regarding the internet access and the bandwidth usage. Finally, the CRLs must be collected from all CAs to create a joint CRL by the CRL issuer. This CRL is then transmitted to the end entities. That means that different states might have to trust not only the Root CA but also the CRL issuer.
CRLite for the Maritime Sector
The second solution is to adapt CRLite to the maritime sector. To the best of our knowledge, this has not been proposed before. A web browser and a ship are very different in nature, and while the main concept can remain the same, the update frequency along with the push/pull model might have to be adapted to fit the need of the maritime sector.
The first advantage of CRLite over CRL is the smaller size of the payload that needs to be delivered to the end-entities. Based on our simulation, the size of the CRLite payload is five times smaller than the equivalent payload for the DeltaCRL. As explained in the background section, CRLite relies on having both the revocation information and the valid certificates information in order to create the filter. To achieve that, CRLite relies on Certificate Transparency, which, even if it is out of the scope for this paper, harden the security of the PKI as a whole. Finally, the authors of the original CRLite paper proposed a way to create the filter in a distributive manner and not involving only the issuer. This is an interesting property for the maritime sector where different states have to collaborate but do not necessarily trust each other.
On the other hand, CRLite is a recent technology (at least compared to CRL), and is neither standardised nor field-tested, which can be an issue to be accepted by the maritime organisations. There is also a lack of formal security analysis and research done. A bachelor thesis from ETH Zurich analysed CRLite and more specifically the usage of Bloom Filter Cascade and concluded that the mechanism presents some weaknesses [14]. In particular, Bloom Filter Cascades do not adapt and scale well with the market growth. This argument does not hold for the maritime sector however, as the amount of certificates is much lower than in the Internet world, where CRLite has been proven to work with around 36M certificates (which is more certificates than the maritime PKI will ever have). Finally, another negative point compared to CRL is that the end entities will get no information on the revoked certificate. When checking for the status of a certificate with CRLite, the answer is either "valid" or "revoked." Depending on how applications plan to handle revocation information, this can be a problem.
Common Topics to both Solutions
No matter which solution is chosen, the frequency of the updates needs to be determined. CRL and DeltaCRL are usually issued on a weekly basis. For CRLite it is on a daily basis. In our simulation however, we used a weekly basis for both methods as a mean for comparison. Choosing the update frequency comes down to answering the question: "How long do we want the ships to accept revoked certificates." The answer to this question might vary between different ship owners and flag states.
Related to this, the importance of the revocation information may vary depending on the reason why a certificate is revoked. For instance, a certificate being revoked because the ship has changed its flag state is not a security threat by itself, but a certificate revoked because CA compromise is. Different priority could thus be given to different revocation information. The notion of "scopes" is described in RFC 5280 [10], and CRLs (and DeltaCRLs) can be issued with a scope. For instance, it is possible to have a CRL for certificate revoked with the reason code "keys compromised" and another one with all the remaining reasons. It is also possible to implement different frequencies for the different scopes, thus allowing reducing the bandwidth costs as well. Splitting the revocation information in scopes is also feasible in CRLite, but as there are very few cases of key compromises compared to other reasons, creating a filter for those might not be justifiable.
How applications handle the revocation information is also another important issue. Currently, in the Internet world, browsers tend to apply a "soft-fail" techniques, meaning that if it can't verify the certificate validity, it will consider it valid, creating a feeling of false security for the user. In the maritime world, it will be important to think about the failure scenarios, how the information is communicated to the user and what are the process to respond to those failures.
Looking Elsewhere
Something that has been out of our research scope is to analyse solutions that are still on a conceptual level. For instance, a blockchain-based certificate transparency and revocation mechanism for the web has been proposed by Wang et al. [29] The idea is to remove the trust from the CA, and to transfer it to the end-entities which are in this case the browsers. Servers can then publish their certificates to a public blockchain, and a browser will accept the certificate received during the SSL/TLS negotiations if and only if it matches the ones in the public blockchain and if it is not revoked. It is very much likely that this and similar solutions will require a degree of connectivity that could be difficult to obtain in a maritime setting.
Conclusion
In this paper we have identified two potential candidate solutions for revocation of certificates in the maritime sector: 1) CRLs (with DeltaCRLs) and 2) CRLite (with Delta filters). Both these solutions can operate when vessels are offline and they both inform about all revoked vessels in a timely manner. However, our results from simulating the behaviour of these two different solutions over time show that will be significant changes in terms of required bandwidth. While the size of the CRLs itself will have an average size of 365 kB, the size of the DeltaCRLs is expected to be relative small (8.6 kB). Still, CRLite is a much better solution in this respect, with 39.5 kB filter size and 2.74 kB Delta filter size. However, as explained in the discussion, there are pros and cons with both solutions and the final choice will be a trade-off between selecting a more wellknown and mature technology (CRL), or going for a potentially more efficient, but less tested, solution (CRLite). | 2021-03-09T14:16:10.782Z | 2020-01-01T00:00:00.000 | {
"year": 2021,
"sha1": "22bc53e4a6934630132d8f7766bbf0cecaf132d1",
"oa_license": "CCBY",
"oa_url": "https://link.springer.com/content/pdf/10.1007/978-3-030-70852-8_9.pdf",
"oa_status": "HYBRID",
"pdf_src": "MergedPDFExtraction",
"pdf_hash": "59ceb87d21806cc88920ac04e83170b3cd92d294",
"s2fieldsofstudy": [
"Computer Science"
],
"extfieldsofstudy": [
"Computer Science"
]
} |
247260181 | pes2o/s2orc | v3-fos-license | Fingerprint As A Means of Identification In The Crime Of Murder ( Study At Polresta Pematang Siantar )
Article history: Received Jan 01, 2020; Revised Feb 03, 2020; Accepted Mar 13, 2020; Online Apr 30, 2020. The crime of murder is currently one of the most complicated and very serious problems faced by individuals, society, law enforcement, and the government. The impact of this act of killing was felt very strongly by the community, namely the disruption of security and order in society. Therefore, this crime of murder needs to be handled or tackled seriously. It is in this part of the investigation that the Police can take the first step, namely managing the crime scene (TKP) by identifying any evidence found such as objects that are strongly suspected of being related to the incident, including the victim, objects around the victim or in the vicinity of the victim. around the crime scene. And for the purposes of investigation, the Police are authorized or entitled to take several actions that have been regulated or determined by the Law on the Indonesian National Police No. 2 of 2002 in Article 15 paragraph (1), including taking fingerprints and other identities and taking pictures of someone. This is the reason behind the author's interest in writing a thesis with problems including: How is the process of formulating fingerprints (dactyloscopy) in identifying a crime of murder; What is the role of fingerprints as a means of identification in a murder crime and what are the obstacles faced in the process of identifying a murder crime by using fingerprints as a means of identification. The research used in this thesis is a normative juridical research method which in research the author generally uses library materials or secondary data as basic data materials in research activities. Through this thesis, it is hoped that it can provide input for law enforcement officers, especially the POLRI in uncovering the crime of murder through fingerprints as a means of identification. Regarding the realization of the recognition and protection of one's right to life, this is a very serious matter because, as stated earlier, this right is the most essential right for every human being. However, recently, more and more problems have been faced with regard to the realization of this right to life, among which the most common is the deprivation of the right to life. Deprivation of the right to life that we usually or often encounter is by coercion and violence which is generally known in our society as an act of murder or a crime of murder. However, recently, more and more problems have been faced with regard to the realization of this right to life, among which the most common is the deprivation of the right to life. Deprivation of the right to life that we usually or often encounter is by coercion and violence which is generally known in our society as an act of murder or a crime of murder. However, recently, more and more problems have been faced with regard to the realization of this right to life, among which the most common is the deprivation of the right to life. Deprivation of the right to life that we usually or often encounter is by coercion and violence which is generally known in our society as an act of murder or a crime of murder.
Introduction
Human rights in principle are a set of rights that are inherent in the nature and existence of humans as creatures of God Almighty and are His gift that must be respected, upheld by the State, law, government and everyone for the sake of honor and protection of human dignity.
The crime of murder is currently one of the most complicated and very serious problems faced by individuals, society, law enforcement, and the government, in which the forms or types of methods used in committing the crime of murder are very diverse. As we can see from the mass media, examples of the forms or methods used in committing this type of crime are as follows; mutilation, a criminal act of terrorism which is an act that threatens the safety of the lives of many people and even deprives the right of life of those who are victims, or kills preceded by violence (persecution).
The impact of this act of murder is very strongly felt by the community, namely the disruption of security and order within the community itself, so that this is one of the important reasons human rights and obligations as citizens cannot run in balance as they should. Therefore, this crime of murder needs to be handled or tackled seriously. Which in handling or tackling it is closely related to the role of all levels of society together with the government and law enforcement authorities in this case especially the Indonesian National Police as their role as investigators and investigators as determined by law.
The police themselves to uncover the occurrence of a crime of murder is not an easy thing, the initial procedure carried out by the police is an investigation to find and find an incident that is suspected of being a murder crime, to then determine whether or not further action can be taken, namely an investigation. It is in this part of the investigation that the Police can take the first step, namely managing the crime scene (TKP) by identifying any evidence found such as objects that are strongly suspected of being related to the incident, including the victim, objects around the victim or in the vicinity of the victim. around the crime scene. And for the purposes of investigation, the Police are authorized or entitled to take several actions that have been regulated or determined by the Law on the Indonesian National Police No. 2 of 2002 in Article 15 paragraph (1), including taking fingerprints and other identities and taking pictures of someone.
Taking fingerprints by the police or investigators is considered as one of the efforts or ways that should not be missed to be carried out, where fingerprinting can be carried out on the victim's body, objects attached to the victim's body, objects around the victim, or other objects. -Other objects in the vicinity of the incident which are suspected to be related to the occurrence of the crime of murder.
So in other words, taking fingerprints by the investigator is one of the most important actions based on the reason that the fingerprints of every human being are different from one human to another. Then after obtaining the fingerprint, the investigator can use it as a means of identification in a murder crime. By properly and accurately processing these fingerprints as a means of identifying the crime of murder, it is hoped that it can provide a bright spot in revealing who is the perpetrator of the murder, so that material truth can be obtained regarding what happened related to the murder act.
In addition to the police, the government itself in its efforts or forms of responsibility for handling the crime of murder has also made special regulations and certain agencies, such as the Prosecutor's Office and Medicine (psychological or forensic) which later if a murder crime case occurs So in other words, taking fingerprints by the investigators is one of the most important actions based on the reason that the fingerprints of every human being are different from one human to another. other. Then after obtaining the fingerprint, the investigator can use it as a means of identification in a murder crime. By properly and accurately processing the fingerprints as a means of identification and then determining whether or not further action can be taken, namely investigation. It is in this part of the investigation that the Police can take the first step, namely managing the crime scene (TKP) by identifying any evidence found such as objects that are strongly suspected of being related to the incident, including the victim, objects around the victim or in the vicinity of the victim. around the crime scene. And for the purposes of investigation, the Police are authorized or entitled to take several actions that have been regulated or determined by the Law on the Indonesian National Police No. 2 of 2002 in Article 15 paragraph (1), including taking fingerprints and other identities and taking pictures of someone.
Taking fingerprints by the police or investigators is considered as one of the efforts or ways that should not be missed to be carried out, where fingerprinting can be carried out on the victim's body, objects attached to the victim's body, objects around the victim, or other objects. -Other objects in the vicinity of the incident which are suspected to be related to the occurrence of the crime of murder. So in other words, taking fingerprints by the investigator is one of the most important actions based on the reason that the fingerprints of every human being are different from one human to another. Then after obtaining the fingerprint, the investigator can use it as a means of identification in a murder crime. By properly and accurately processing these fingerprints as a means of identification .
Method
The research used in this thesis is a normative juridical research method in which the author's research generally uses library materials or secondary data as basic data in research activities. Method of collecting data In writing this thesis used data collection methods as follows: Library Research, namely, Primary legal materials such as; the Criminal Code (KUHP), the Indonesian Police Act; secondary legal materials such as; official government documents in the form of court decisions. Tertiary legal materials or supporting legal materials which include materials that provide instructions and explanations of primary and secondary law as well as primary, secondary, tertiary materials outside the field of law; Field Research (Field Research) That is by conducting research directly into the field. In this case the author directly conducts research at the Identification Center.
BARESKRIM Pematang Siantar Police and conducted an interview with Brigadier M. Nasib (a member of the Identification Center of BARESKRIM Pematang Siantar Police). Data analysis In this paper, the data analysis used is qualitative, in which the secondary data and primary data obtained are then analyzed qualitatively to be able to answer the problems in this thesis.
The Role Of Fingerprints In The Process Of The Identification Of The Crime Of Murder And Obstacles That Faced In Utilizing
a. Facilities or Elements Affecting the Role of Fingerprints in the Identification Process. In the use of the fingerprint system as a means of identification, of course there are elements that affect the role of the fingerprint. Among them are the quality of human resources; supporting/auxiliary equipment; and technology systems. 1) Human Resources Given how important the role of fingerprints in an effort to reveal a crime and catch the perpetrators. So knowledge about fingerprints for every Polri investigator from subordinates to superiors is actually a must that must be possessed, because the evidence is overwhelming that the largest percentage of criminals or criminals can be caught because of fingerprint proof. Fingerprints alongside other human traits are powerful tools for searching and finding criminals. So be careful with every fingerprint found at the crime scene (TKP). Furthermore, it is known that the job of taking fingerprints is not an easy job, but a job that requires perseverance, foresight and patience. So to maximize the function or role of fingerprints in the identification process, it must be carried out professionally by officers who are certainly experts in the field of dactyloscopy (fingerprints). Because if the retrieval or development of fingerprints is carried out unprofessionally, the fingerprint data in the computer database of the Criminal Investigation Unit of the National Police will be useless or of no benefit at all. And these investigators who specialize in the field of dactyloscopy should be evenly distributed throughout the police resorts. 57 2) Support/auxiliary equipment The availability of equipment and materials needed for officers who take and develop fingerprints found at crime scenes is also very supportive in carrying out the process of identifying a crime. These tools can assist officers in taking latent fingerprints that are found or found in different places or circumstances which may cause difficulties or obstacles in processing (development). Regarding the equipment used as a support or assistant in fingerprint processing, among others, as described in the description in the previous chapter.
3) Technology system
However, no matter how sophisticated the equipment used, the usefulness and success of the role of fingerprints in the identification process also requires a sophisticated technological system and keeps up with the times. So that the use of technological systems that are in line with the times and advances in electronic and information technology, in particular the field of fingerprints (dactyloscopy) plays a very important role in the system of increasing national resilience and security from being undermined by criminals who can endanger economic, political, and even cultural resilience. nation. One of these technology systems is; computer software AFR (Automatic Fingerprint Recognition System) and CAAFIS (Computer Aided Automatic Fingerprint Identification System) which is a place to store or database of fingerprint data obtained. 58 Again, that in order to use or utilize a sophisticated technological system, the main thing that must be considered is the availability of operator personnel who have knowledge of the system and are skilled at using it.
b. The Role of Fingerprints as a Means of Identification of the Crime of Murder. As stated earlier that at the crime scene (TKP) it is certain that there are fingerprints or other traces, as well as at the location / scene of the murder case. This is based on the idea that to commit a crime such as a murder crime, the perpetrator takes several actions that are likely to come into contact with objects around the crime scene (TKP), so fingerprints are likely to be left on these objects. except, the perpetrator uses hand pads (paper, cloth or gloves).
And based on the results of the study revealed the secret of fingerprints, namely, there have never been two people who have the same and permanent fingerprints. Starting from the results of this study, people use fingerprints to reveal a crime and try to find the culprit.
Knowledge of fingerprints for every Polri investigator from subordinates to superiors, is actually a must that must be possessed, because the evidence is overwhelming that the largest percentage of criminals or criminals can be caught because of fingerprint proof. Fingerprints alongside other human traits are powerful tools to search for and locate criminals. So be careful with any fingerprints found at the crime scene (TKP), because these fingerprints or fingerprints can be collected and then matched with the fingerprints of criminals who And to systematize and streamline the use of fingerprints in the Police itself, a task force has been formed that specifically handles fingerprint issues and fingerprint examinations at the National Police Headquarters (Police Headquarters) laboratory are handled by fingerprint experts. In general, the benefits of this fingerprint function are: 1) As information material in the investigation process, which is the AK-23 fingerprint blank that can be used to take/record 10 fingerprints, and contains complete individual data which includes characteristics, general, special/signals, photos and signatures . This is used to complete information in an effort to trace the history of the perpetrator of the crime/murder crime. 2) As evidence, because fingerprints are one of the material evidences, which never change and are not the same for everyone, so this fingerprint is very effective, efficient and accurate. More broadly, the fingerprint system has a dual role in the investigation of criminal cases, especially cases of murder, namely: a) As an effort to trace the criminal history of the arrested murder suspects and as documentation of the criminals who were sentenced by the court. b) As an effort to track down the perpetrators of crime, especially the crime of murder whose identity is not / not yet known, but accidentally and without realizing it, leaves fingerprints (latent fingerprints) at the crime scene (TKP). In order for this fingerprint system to play a role / function in the process of investigating or identifying a crime, especially a crime of murder, it must be supported by several things as described in detail in the previous description, which are briefly as follows: c) Adequate and even distribution of specialist investigators/experts in the field of fingerprints (dactyloscopy) throughout the Police Resort (POLRES). d) There are trained fingerprint takers. e) Availability of equipment and materials needed for fingerprint takers and development officers. f) There are methods of retrieval, processing and analysis that have been mastered by the officers concerned. g) Mastery of modern technology systems, centered on information systems.
The use of the fingerprint system in identifying a crime or crime such as murder is still rare in Indonesia. However, even so in a murder where the perpetrator may be wearing a mask or the perpetrator of the murder has never been identified as the perpetrator of the crime, then with the help of this fingerprint system the perpetrators can finally be caught and brought to justice and the fingerprints that have been developed or processed can function as evidence in the evidentiary process.
Constraints in the Identification Process Using Fingerprints as a Means of Identification.
Based on the research results obtained from the Pematang Siantar Police, namely with Mr. Brigadier M. Nasib at the BARESKRIM Identification Center of the Pematang Siantar Police, it is generally known that the obstacles faced in the utilization or use of the fingerprint system in the identification process, which include: a. Human resources, meaning that the number of personnel or self-identification officers who master or are experts in the field of fingerprint identification is still inadequate.
b. The incompleteness of supporting/auxiliary equipment in fingerprint identification as well as facilities and infrastructure which are also not evenly distributed among each Police Resort. The use of technology systems that are centered on information systems is still very limited, due to the lack of mastery of technology systems by the identification officers themselves. Thus causing the application of fingerprint processing quickly and accurately is also limited.
Sample case.
The use of fingerprints as a means of identification in the crime of murder is not often found in Indonesia, but there are several, for example: The criminal case of murder with the defendant named Harri Hutabarat has been charged with the crime of murdering the victim named Ng Teng So alias Samsudin with a summary of the incident/case as follows: On Monday, July 21, 2008 at around 06.30 WIB, the complainant named Ng Teng Yu alias Berkat as the owner of the Andalas shop Jl. Merdeka intersection Jl. Bandung No. 104 Ex. Dwikora district. West Siantar Pematang Siantar city. When the reporter, Ng Teng Yu alias Berkat was about to open the door of the shop, he saw that the side of the shop was filled with crowds of people, and when the reporter approached the crowd, he found a lying man who he knew named Ng Teng So alias Samsudin who was none other than his real brother. the complainant himself, and the condition of the person was covered in blood and covered with a sarong, which was suspected to have been killed by an unknown person. Furthermore, the reporter reported the incident to the West Siantar Police, Pematang Siantar City, because the victim Ng Teng So alias Samsudin was allegedly murdered. And on Thursday, July 24, 2008 when officers from the West Siantar Police on patrol saw a man lying on the road in an unconscious state which was later found to be due to an electric shock, then the officers took the man to Djasamen Saragih Hospital, Pematang Siantar City for received treatment, not long after the man regained consciousness and received treatment, the officers interrogated him until it was discovered that his name was Harri Hutabarat, then the officer continued by asking the man what caused him to be electrocuted, and according to his confession because he intends to steal the power cable behind the Horas Market Police Post, West Siantar District, Pematang Siantar City, and alias Samsudin a fruit seller at the intersection of Jl. Merdeka -Jl. Bandung, who is suspected of having been killed, the man initially said he did not know anything about the incident, and later said he did not feel involved in the incident. Then the officers continue to investigate and identify until clues and evidence are collected to be used in an effort to find the suspect who committed the murder, to then be used as a defendant in court and the evidence is sufficient to be faced in the evidentiary process in court.
Then the evidence is processed by officers, one of which is fingerprint identification, which at this stage several things are done as follows: How to carry out fingerprints; researching/developing fingerprints; filing; latent fingerprint investigation; latent fingerprint formulation; fingerprint investigation of corpses/victims; fingerprint investigation of the suspect/perpetrator; make fingerprint photos; expose fingerprints as evidence in the judicial process.
Fingerprint identification is carried out on objects found at the crime scene (TKP) and around the murder scene, which include penknife, dagger, blue mancis and other items, by processing and developing latent fingerprints obtained from these objects go through the stages or processes as described the previous description. And in addition to dactyloscopy identification, signal identification and photographic identification were also carried out.
And based on the results of the investigation and identification as well as conducting interrogations that have been carried out by the Polsek officers against several witnesses, then the investigators have concluded that the suspect in the murder is none other than the suspect against Harri Hutabarat. Previously, interrogation and taking fingerprints as well as signaling data from the person concerned had been carried out. So it was found that the pattern/main shape of the fingerprint painting found at the crime scene with that of Hari Hutabarat's brother was found, which means in other words the investigator or identification officer found the identity between the comparison fingerprint and the latent fingerprint. Likewise, by matching Harri Hutabarat's signaling data obtained through identification of evidence with statements from witnesses. So the officers raised Harri Hutabarat's status from a witness to a suspect perpetrator and then a defendant in the judicial process. Fingerprints and other data that have been developed or processed can be used as evidence in the evidentiary process.
So that at the time of the trial, in the evidentiary process several items of evidence as mentioned above, and also the results of processing and developing latent fingerprints were obtained. Which through the results of the examination/comparison or fingerprint identification, the panel of judges consists of: AMSiringoringo, SH, MH (Chairman Judge); Dahlia Panjaitan, SH (Member Judge); J. Simarmata, SH (Member Judge), who is authorized to try this murder case, includes him as evidence which will become one of several considerations of the panel of judges in making or formulating the contents of its decision.
Conclusion
.In the case of a murder crime, the precise and accurate formulation of fingerprints plays a very important role in the investigation and proof process, with the following stages: searching for fingerprints at the crime scene or around the crime scene; then confirm the location of the fingerprint; taking fingerprints (eg fingerprints from people who were right at the crime scene or around the crime scene); and lastly examination of the comparison (formulation) of fingerprints in the identification process. This last stage is the most important stage because it is carried out to determine whether the two fingerprints have the same painting shape and are identical fingerprints, meaning by seeing whether the flow of papillary lines between the two fingerprints is the same.
To determine two identical or identical fingerprints, there are several factors that must be assessed orseen, among others, are through: a) The basic form of fingerprint painting between latent fingerprints and comparison fingerprints, namely: the two fingerprints being compared must be the same, although their identity cannot be determined, if other factors are not/not fulfilled; b) Characteristics of the fingerprint papillary lines (detailed galtons) the type and shape of the detailed galtons between the two fingerprints (latent fingerprints and comparison fingerprints) must be of the same shape, position, and direction; c) The number of equation points (galton details of the same type, shape, direction and position) then: if the equation points are 11-12 or more, it means that the identity is definite; if the point of the equation is 8-10, it means that the identity still has to be strengthened with things like; the clarity of fingerprints, the presence of cores or deltas of the principal forms of fingerprint paintings that are rarely found.
In general, the benefits obtained from this fingerprint function are: As information material in the investigation process, which is the AK-23 fingerprint form that can be used to take/record 10 fingerprints, and contains complete individual data which includes characteristics features, general, special / signal, photos and signatures. This is used to complete information in an effort to trace the history of the perpetrators of crimes/criminal acts of homicide; As evidence, because fingerprints are one of the material evidences, which never change and are not the same for everyone, so this fingerprint is very effective, efficient and accurate. More broadly, the fingerprint system has a dual role in the investigation of criminal cases, especially cases of murder, namely: (1) As an effort to trace the criminal history of the arrested murder suspects and as documentation of the criminals sentenced by the court; (2) As an effort to track down the perpetrators of crimes, especially the crime of murder whose identity is not / not yet known, but accidentally and without realizing it, leaves fingerprints (latent fingerprints) at the crime scene (TKP). In general, it is known that the obstacles faced in the utilization or use of the fingerprint system in the identification process, among others: Human resources, meaning that the number of personnel or self-identification officers who master or are experts in the field of fingerprint identification is still inadequate. The incompleteness of supporting/auxiliary equipment in fingerprint identification as well as facilities and infrastructure which are also not evenly distributed among each Police Resort. The use of technology systems that are centered on information systems is stillvery limited, due to the lack of mastery of the technology system by the identification officer himself. Thus causing the application of fingerprint processing quickly and accurately is also limited. Difficulties -difficulties also encountered by officers or fingerprint experts in identifying, namely: 1) Two fingerprints will be difficult or may not be identified if they are of different types; 2) Fingerprint marks found at the crime scene (TKP) often show imperfect or blurry shapes because they are scratched, stained, or overlapped with other fingerprints. Although this does not make identification impossible. Comparing fingerprints that have been recorded and those obtained at the crime scene is not a special science, but depends on the expertise and experience of the expert. If it meets the characteristics between the two fingerprints (latent and recorded fingerprints), then identification can be done. Differences in the conclusions of experts occur if the characteristics of fingerprints are obtained to a minimum. So sometimes other experts argue that the data is insufficient for identification; Another difficulty is that fingerprints should have been recorded in the Police file beforehand. However, the file system in the Police is still not good, and has not been systematically arranged, and there is still a lot of incomplete fingerprint data, meaning that it is not recorded up to ten fingers. then identification can be done. Differences in the conclusions of experts occur if the characteristics of fingerprints are obtained to a minimum. So sometimes other experts argue that the data is insufficient for identification; 3) Another difficulty is that fingerprints should have been recorded in the Police file beforehand. However, the file system in the Police is still not good, and has not been systematically arranged, and there is still a lot of incomplete fingerprint data, meaning that it is not recorded up to ten fingers. then identification can be done. Differences in the conclusions of experts occur if the characteristics of fingerprints are obtained to a minimum. So sometimes other experts argue that the data is insufficient for identification; 3) Another difficulty is that fingerprints should have been recorded in the Police file beforehand. However, the file system in the Police is still not good, and has not been systematically arranged, and there is still a lot of incomplete fingerprint data, meaning that it is not recorded up to ten fingers. fingerprints should have been recorded in the Police file beforehand. However, the file system in the Police is still not good, and has not been systematically arranged, and there is still a lot of incomplete fingerprint data, meaning that it is not recorded up to ten fingers. fingerprints should have been recorded in the Police file beforehand. However, the file system in the Police is still not good, and has not been systematically arranged, and there is still a lot of incomplete fingerprint data, meaning that it is not recorded up to ten fingers. | 2022-03-08T16:31:36.228Z | 2020-04-30T00:00:00.000 | {
"year": 2020,
"sha1": "8cb0d23f6286b8808c9166d639f3c4f3cf971c46",
"oa_license": "CCBYNC",
"oa_url": "https://iocscience.org/ejournal/index.php/JLS/article/download/1620/1167",
"oa_status": "HYBRID",
"pdf_src": "Anansi",
"pdf_hash": "347e6826bc564e869c6ef8f6cbe02be3acc54bfd",
"s2fieldsofstudy": [
"Law"
],
"extfieldsofstudy": []
} |
263310901 | pes2o/s2orc | v3-fos-license | Extreme cases: Math education within the U.S. prison system
This paper summarizes the author's insights and experiences of teaching mathematics to incarcerated students within the U.S. prison system.
Pulling into a visitor parking space for the first time, I watch a stray dog hesitantly making her way toward the barbed wire fences. The surroundings are austere and unforgiving. Lockhart Correctional Facility is about an hour's drive from the city of Austin and miles from the nearest town center, situated, intentionally, in the middle of nowhere. I wonder briefly how the dog ended up here and feel a pang of sadness that I have nothing to offer her. It will not be the last time I struggle with these feelings in this parking lot; they only intensify when I think about my intelligent and resourceful students being left here to fend for themselves.
I pocket my ID and car keys, grab the box of printouts I made for class, and head toward the barbed wire gate. With sunlight still streaming through the lobby's glass doors behind me, the first checkpoint feels similar to airport security. I empty my pockets, place everything on a table to be searched, remove my belt and shoes, and walk slowly through the metal detector. A female correctional officer (CO) outlines my body with her wand before patting me down, paying special attention to the bottoms of my feet, while another CO searches my class materials. He warns me that highlighters and papers with "too much ink" are not permitted, but thankfully all my materials are allowed through today. I gather them up and head toward the next security checkpoint, where any glimpse of daylight or sense of familiarity is gone.
The main guard station remotely buzzes me through a door after surveilling me on camera, and I hand over my ID in exchange for a badge after sign-ing in with them. I am buzzed through one last door and suddenly step out into prison, narrowly avoiding the lines of people moving seamlessly in and out of the cafeteria. I hear my pulse in my ears until I notice that they are all smiling and welcoming me. Someone exclaims, "God, I miss wearing jeans!" We all laugh. Another woman approaches me to show me to my classroom, and we walk down a hallway under fluorescent lighting and a large hemispherical mirror that shows everything around the corner. "We are so grateful and excited you are here!" she tells me, and I feel the same.
When we arrive at the classroom, she apologetically explains that I will have to wait while the COs release my students to come to class. Although class technically begins at 6 p.m., we usually don't start until 6:30, but I enjoy the wait. That half hour is the most important part of teaching for me; without a phone or any electronics to distract me, I wind up studying all the art and handouts displayed on the walls. This classroom is also used by vocational and re-entry programs, and the exercises that the students complete are somber reminders of the real stakes here. "My goal is to get sober for my children." "I want to learn how to forgive myself and earn forgiveness from others." "When I get out, I will start a new career with my certifications to support my family." These messages reiterate to me that I am primarily here to build their confidence in themselves and their problem-solving abilities. This confidence gives them a sense of freedom even in incarceration, as a formerly incarcerated friend will later share with me.
The twelve students in my class trickle in, and we introduce ourselves. Even though I learned about average state prison populations during the Texas Prison Education Initiative's (TPEI) orientation, I am still surprised at the large variance in age, from about early 20s to mid 60s. I feel a flutter of anxiety about the pace I've decided on for my lectures.
The students need to pass my course as a prerequisite for any credit-bearing TPEI math course through UT Austin, so I want to focus on building their abstract problem-solving abilities. In particular, I have decided to spend a lot of time developing their intuition in subject areas they had likely seen before, like integer addition; I am worried the students will be bored with my decision.
"So, just out of curiosity, what do y'all think about math in general?" I ask, quickly adding, "No judgment here, I do it for a living and still feel love-hate about it." We laugh, but it's hard to keep smiling encouragingly when I hear about their previous experiences in math classes. The students are diverse in age, race, ethnicity, and sexual orientation, but they unfortunately had one major thing in common: all of them proclaimed to be "bad at math," and many had dropped or failed out of school specifically because of math. 1 I tell them my plan for the course: we are going to start over with math and look at it from a new perspective. Anything that they need to know for the course, I will teach them, and it would actually be preferable to wipe away any memory of subjects where they had unsuccessful learning outcomes. Even things they have seen before, like integer addition and multiplication, I want us to consider with fresh eyes. "Our course objective, and the real beauty of mathematics," I tell them, "is using specific instances of a problem to understand how it works in full generality." And that is exactly what we did.
As mathematicians, we often look to "extreme cases" of a problem to gain intuition for a general solution. Mathematics education within the prison system is perhaps the most extreme case, and I believe it sheds light on how we can teach more effectively in a university classroom.
Satisfying Constraints
In addition to the usual concerns about teaching a math course, teaching math in prison comes with a unique set of challenges. With respect to the aforementioned issue of bringing class materials inside, there are certain canonical classroom resources that are prohibited or impossible to obtain. There is no computer, internet access, or even a desk for lecture materials, just a small white board at the front of the room that, just like in university classrooms, is unreliably equipped with dry erase markers. There are no office hours, recitations, or means of communicating with students outside of class time. If students miss class for a variety of legitimate reasons, like having to work late at their jobs in prison or not being allowed out of their dormitories, they have no way to access material or get in touch with me until the next class.
During the summer, for example, I taught an elective course called "The Art of Mathematics," in which students investigated several math topics, like algorithms and infinite set cardinalities, that served as inspiration for their own art work. TPEI provided sketchbooks, folders, and colored pencils, but I was not permitted to leave the colored pencils with the students when class was over. The other class materials required certain stickers to denote that TPEI had given them to these students, in hopes they would not accidentally be confiscated.
I also brought in art books and textbooks that I had at home to pass around the room, which I was told to collect at the end of each class. Though I did not, I was very tempted to ignore this rule on two occasions. Once when an older student was poring through The Math Book by Clifford A. Pickover and asked me excitedly if she could check it out like a library book. Then again when a younger student asked me if she could borrow Baby Rudin because she was curious about analysis proofs; she'd also asked me earlier that evening which prerequisite textbooks or classes she'd need in order to teach herself analysis some day. Later in the course, I learned the latter student was actually studying astronomy in college before her arrest; unfortunately, she'd been ousted from the major due to her grades in the required math classes.
The culmination of the math art elective course was to be a gallery night, a classroom exhibition of their work on the last day. Together, we would reflect on the various topics we'd covered, discussing the mathematical ideas and artistic choices that spoke to us in each other's work. I'd invited the TPEI program coordinators, Max and Chloe, who were also excited to see what the students had created.
When we arrived at the classroom, we waited for an unusually long time before a couple of students came in and mentioned they'd had a difficult time getting to our room. Around 6:45, when Max went back to the central guard station to ask if all TPEI students had been called for class, we were told that a fight had broken out in one of the dorms, and it was now on lockdown. As a result, despite no involvement in the fight, about half of my students were unable to come and present the art pieces they'd worked on all summer, and by the time the ones who could make it finally arrived, we only had about 45 minutes left. Everyone involved was disappointed, but those of us in attendance tried to make the best out of the remainder of the evening. The exhibit was breathtaking, and the influences and creativity in every piece were truly awe-inspiring. Max and Chloe were blown away and agreed that the students' exhibit looked and felt professionally curated.
Since we couldn't bring cameras inside to take pictures, my proposal to the students was, with their permission, to borrow their sketchbooks overnight, take photos of their work, then return their sketchbooks to them the next evening at another TPEI instructor's class. That way, students wouldn't have to tear out pages from their sketchbooks and would only be without them for less than a day. However, I was honest about my worry that there could be some unforeseen issue getting the sketchbooks back to them, which was always a risk, and I told them I understood completely if they wanted to hold on to their work just in case. I was moved almost to tears when every student there that night actually gifted me some of their art work, tearing pages out of their sketchbooks for me to take home and keep. Those pieces are now framed and displayed in my school office, and Figure 1 contains photos of their art work. In (a), we see a sketch, created with a straight-edge, colored pencils and deoderant for shading, inspired by M.C. Escher's work with H.S.M. Coxeter on the "limit of infinite smallness" [Wie10]. Figure 1(b) shows a selfportrait of the aspiring astronomy student contemplating math topics from the course. In (c), the student illustrates the concept of yin and yang using precise mathematical symmetry. In Figure 1(d), the sketch on the left is a play on the Poincaré disk, using circular objects of different radii that the student collected from around the prison. On the right in Figure 1(d), she depicts her interpretation of Escher's limit of infinite smallness. In (e), we see an interpretation of Fibonacci numbers arising in nature, and in (f), a self-portrait of the student in front of a tiled background, with a new (imaginary) tattoo showcasing her love of math on her right elbow. Finally, the photo in (g) showcases a 3-D quilling sculpture of an elephant to represent me, named "Lil Kate," made completely out of small paper strips, dyed with food coloring, and adhered together with a mixture of coffee creamer and water, and in (h), a proportionallyaccurate illustration of the Fibonacci spiral.
Developing Intuition
One of the biggest challenges of teaching in prison is being diligent in my language choices. I actively try to avoid using idioms or explanations that could trigger mental blocks for students, especially about math subjects in which they'd experienced difficulty or unsuccessful learning outcomes. Since I am hoping to build up their intuition on a new mathematical foundation, it's important not to repeat the same explanations that caused them confusion the first time. Moreover, because of the already difficult environment we are in, I don't want to say something that would cause students any more distress.
A small but pervasive example of being more conscious of my language in the classroom was swapping out the term "homework" for "assignment" in my college algebra course. But I also wanted to make sure my writing achieved the same objectives as my classroom language and tone. So every week, I wrote lecture notes that sounded exactly like I would speak during class, while still including the usual textbook definitions, examples, formulas, etc.-and while trying to avoid anything that would elicit a mental block toward the subject. This style of lecture notes was especially important in situations where students missed class. Since there are no dedicated times like office hours, appointments, or recitations for them to get help, I tried to make my notes reminiscent of our classroom: thorough, but not dry, while still containing the explanations they need to understand the material.
Out loud and in text, before introducing any new topic in the course, I wanted to provide motivation for why we were talking about it, but really address why they should care about it. As a stereotypical example, I did not have to ask how my students felt about fractions; they all groaned loudly when I said, "Today, we'll be talking about fractions." I had anticipated this, mostly because of certain family members and friends who have had an identical response to dealing with fractions. I jokingly threw up my hands and acquiesced over the groans: "Okay, fine, we'll talk about one of my favorite things in math first instead . . . prime numbers!" We then detoured through prime numbers, prime factorizations, division trees, greatest common divisors, and least common multiples before we finally circled back to fractions. Suddenly, the two biggest obstacles to them mastering fractions previously (adding and simplifying) were reduced to problems they had just tackled in a very different setting. Once we'd gotten around the initial hurdle to approaching a subject they'd already convinced themselves they'd "never understand," my students no longer experienced the same mental block toward it. Instead, I witnessed them appreciate the power of abstraction, to the point that one of them exclaimed, "Holy crap, I can actually help my kids with their homework now!" after adding three fractions with different denominators using their least common multiple.
By the time we got to solving linear equations and word problems, the students had a completely different outlook on the material. After hearing it so much on the first day, I had banned the phrase "bad at math" from our classroom, but even if I hadn't, that was not how the students felt anymore. We had spent multiple class periods developing intuition for the properties of addition, subtraction, multiplication, and division of real numbers, from an abstract perspective but also with the concrete example of debt to understand computations with negative numbers. What I'd worried would seem boring to the students was actually what initially piqued their curiosity: we took a specific instance of a problem, then abstracted it away with algebraic tools to study the problem in full generality. Once they'd seen a given topic from these two different angles, they also had plenty of "low-stakes" opportunities to practice. I frequently reminded them that it's normal to make mistakes when trying something new. So by the time I said to them, "Today, we are going to solve some word problems," the students were no longer groaning, or feeling anxious about using variables to represent unknowns; their confidence in their abstract problem-solving abilities had been strengthened by the time we spent in the low-stakes material, building a solid foundation from which to work. Upon solving one of the linear systems arising from a word problem, one of my students proudly announced, "I feel like a mathematician!" "You are," I replied.
When I taught the art elective course, I ascribed to a similar ideology. Each mathematical topic (algorithms, proportion, infinity, and abstraction) was paired with complementary art work from different time periods and regions of the world to illustrate the idea. It was surprising how much of the material, which I'd created with the intention of introducing them to advanced undergraduate math they wouldn't have seen before, illuminated other mathematical concepts for them that I hadn't even anticipated. When we talked about proportions, for instance, we discussed how the ancient Greeks thought the ideal relationship between the width w and height h of a building is given by the "Golden proportion" After we talked about its relationship to the Fibonacci sequence, I mentioned offhandedly that nowadays we can solve explicitly for the Golden ratio φ using the quadratic formula; a student immediately raised her hand and asked to see how that would work. After I showed them the trick of set-ting h = 1, they were amazed to see an "elementary" formula they'd rotely memorized in school being used to answer such a seemingly unrelated question. Later in the same class, when we were discussing the Archimedean method of approximating π, I wrote the more familiar equation C = πd as π = C/d to emphasize the ratio. To my surprise and dismay, the vast majority of them had never seen or considered π as a ratio, and it was emotional to watch them eagerly discovering that amazing property of circles every mathematician discovers at some point in their career. I began to think deeply about how we are motivating these concepts for students when they first encounter them. My students' comments in class cast many mathematical concepts in a new light for me as well. For example, in talking about algorithmic constructions with girih tiles [LS07], we discussed how it would be possible with rudimentary tools to fabricate these tiles with precise angular measurements and arrange them into astonishingly intricate designs. We talked about the idea of decomposing the polygonal tiles into triangles, and how an understanding of triangles unlocks a lot of possibilities for the tiles' fabrication and design. One of the students mused casually, "So that's why they teach an entire class about triangles. They're like the atoms of shapes." I've since begun borrowing that phrase to motivate the subject of trigonometry.
During another class period, we talked about the cardinality of the natural numbers. I chuckled to myself as I listened to a familiar debate among the students. "Why would 0 be included? You don't count anything with zero fingers." Fortunately, they were satisfied that it didn't much matter with respect to cardinality after we wrote out the bijection between the natural numbers and the integers. However, that awareness came back to bite me in the form of a deeply profound and unexpected question that I received while explaining Cantor's diagonalization argument. I had just demonstrated his proof by contradiction: if we try to enumerate all of the elements s 1 , s 2 , s 3 , . . . of the set T of infinite sequences of 0's and 1's, it is always possible to construct an element s ∈ T that differs from s k in the kth position, so that element of T actually wasn't enumerated in our list.
"But why can't you just call that element s 0 ? Then wouldn't you be able to count them all with counting numbers since it doesn't matter if we include 0?" The question was so subtle and clever that it caught me off guard. "You're thinking like a mathematician," I replied, before we spent the next few minutes verifying that Cantor had indeed gotten it right, though probably not on his first try.
Solving Problems
Once I was talking about my experience teaching math in prison with a friend of mine who was formerly incarcerated. He'd served a ten-year sentence over the entirety of his 20s, during which time he had the opportunity to take several math courses for high school degree equivalency. He's now earning his bachelors degree while working full-time as an operations manager at a water treatment facility. While math courses equipped him with necessary skills for his new career, he credits those classes with something even more important. As a creative writer, my friend had always felt more passionate about writing classes in school. He admitted to me that he'd never liked that there was "only one right answer" in his math classes. But once he was incarcerated, solving math problems became a mentally stimulating and comforting activity. With so much time to think and reflect, he realized how many problems in life lack a clear-cut solution, and he began to appreciate the existence and uniqueness of the solutions to the problems in his math assignments.
I told my students about this conversation with my friend, and I asked them if they felt similarly about learning math while incarcerated. Several of them participate in entrepreneurial and vocational training programs, and those students echoed the importance of math for their new career paths. In fact, one student made parole during our spring algebra class, and she still wanted to finish the course remotely post-incarceration to help her earn her EMT certifications. (This situation is one of many in which the TPEI administrative team of volunteers is essential to the program's success.) Several other students said that my friend's point about taking comfort in having a right answer really resonated with them, even though they'd had terrible previous experiences with math. One of them was the student who'd had to drop her astronomy major because of her math classes, who'd asked about teaching herself Baby Rudin. Another student had been concurrently earning her high school diploma during our class ("only 50 years late!" as she'd often say), and she explained to me how learning math again had helped her find balance in her life. "Everything in math has an opposite," she said, "I am a 'yin and yang' type of person, and I guess that's why I love math. You can always add back anything you subtract, or multiply anything you divide." Her friend chimed in, "Plus, even just knowing there are infinitely many possible choices to start from gives you options for all of the gray areas in life." Yet another student, who was nervous about her upcoming parole hearing the following week, reiterated that being able to solve a problem and find a correct answer out of infinitely many possible choices was going to be vital to her success post-incarceration. Like so many of her classmates, she'd also been interested in STEM growing up; she wanted to be an astronaut when she entered high school, before dropping out because of math. "I'm nervous about finding a job and a permanent place to live, but I have people to help me for now," she said. "I probably can't be an astronaut anymore because of my felony charge, but I still want to enroll in college and earn my B.S. once I'm back on my feet." In spite of all the problems and uncertainties that she faced post-incarceration, she'd already begun identifying solutions. On the last day of the art elective class, after the exhibition, I announced that I would be teaching a brand new course in the fall that had never been offered by TPEI. It is a credit-bearing math course through UT Austin that is a prerequisite for UT's calculus sequence. My student-the aspiring astronomer who'd been enthralled by Baby Rudincame up to talk to me after the other students had said their thank-you's and goodbye's. "I really want to take the course, but I've failed this subject before and I'm worried I will again," she said nervously. "I haven't done math in a long time." Thinking back over all of the insights she'd shared during the course, I smiled and reassured her that I knew she could do it and that I would be there to help her. "Besides," I replied, "you've been thinking like a mathematician this whole time." | 2023-10-02T06:42:07.695Z | 2023-08-03T00:00:00.000 | {
"year": 2023,
"sha1": "08a4d598edbbd7c5eb5806820e65e62292cec823",
"oa_license": null,
"oa_url": null,
"oa_status": null,
"pdf_src": "Arxiv",
"pdf_hash": "08a4d598edbbd7c5eb5806820e65e62292cec823",
"s2fieldsofstudy": [
"Education"
],
"extfieldsofstudy": [
"Mathematics"
]
} |
233605417 | pes2o/s2orc | v3-fos-license | Microphysical process of precipitating hydrometeors from warm-front mid-level stratiform clouds revealed by ground-based lidar observations
Mid-level stratiform precipitations during the passage of warm front were detailedly observed on two occasions (light and 15 moderate rain) by a 355-nm polarization lidar and water-vapor Raman lidar, both equipped with waterproof transparent roof windows. The hours-long precipitation streaks shown in the lidar signal ( X ) and volume depolarization ratio ( δ v ) reveal some ubiquitous features of the microphysical process of precipitating hydrometeors. We find that for the light rain case, precipitation that reaches the surface begins as mixed-phase hydrometeors fall out of a shallow liquid cloud layer at altitudes above the 0 ° C isotherm level, and the depolarization ratio magnitude of falling hydrometeors increases from the liquid-20 water values ( δ v < 0.10) to the ice/snow values ( δ v > 0.25) during the first 100 − 200 m of their descent. Subsequently, the falling hydrometeors yield a dense layer with an ice/snow bright band occurring above and a liquid-water bright band occurring below (separated by a lidar dark band) as a result of crossing the 0 ° C level. The ice/snow bright band might be a manifestation of local hydrometeor accumulation. Most falling raindrops shrink or vanish in the liquid-water bright band due to evaporation, whereas a few large raindrops fall out of the layer. We also find that a prominent δ v peak (0.10 − 0.40) always 25 occurs at an
Abstract. Mid-level stratiform precipitations during the passage of warm fronts were detailedly observed on two occasions (light and moderate rain) by a 355 nm polarization lidar and water vapor Raman lidar, both equipped with waterproof transparent roof windows. The hours-long precipitation streaks shown in the lidar signal (X) and volume depolarization ratio (δ v ) reveal some ubiquitous features of the microphysical process of precipitating hydrometeors. We find that for the light-rain case precipitation that reaches the surface begins as ice-phase-dominant hydrometeors that fall out of a shallow liquid cloud layer at altitudes above the 0 • C isotherm level, and the depolarization ratio magnitude of falling hydrometeors increases from the liquid-water values (δ v < 0.09) to the ice/snow values (δ v > 0.20) during the first 100-200 m of their descent. Subsequently, the falling hydrometeors yield a dense layer with an ice/snow bright band occurring above and a liquid-water bright band occurring below (separated by a lidar dark band) as a result of crossing the 0 • C level. The ice/snow bright band might be a manifestation of local hydrometeor accumulation. Most falling raindrops shrink or vanish in the liquid-water bright band due to evaporation, whereas a few large raindrops fall out of the layer. We also find that a prominent δ v peak (0.10-0.40) always occurs at an altitude of approximately 0.6 km when precipitation reaches the surface, reflecting the collisioncoalescence growth of falling large raindrops and their subsequent spontaneous breakup. The microphysical process (at ice-bright-band altitudes and below) of moderate rain resembles that of the light-rain case, but more large-sized hydrometeors are involved.
Introduction
An observation-based understanding of the microphysical processes of precipitation is essential for weather/climate modeling and predictions. Such processes are difficult to observe since they involve a variety of hydrometeor sizes, shapes and phases at different altitudes, all of which are affected by cloud dynamics (Aggarwal et al., 2016;Pfitzenmaier et al., 2018). In situ aircraft observations deliver data on the sizes and numbers of hydrometeors only for small sampling volumes at single altitudes at any given time during preplanned case studies (Barrett et al., 2019). Although lidar and radar can measure the time-resolved vertical profiles of bulk backscattering quantities, retrievals of the microphysical properties of hydrometeors require numerous assumptions (e.g., the hydrometeor shape and size distributions). Furthermore, in most cases, ground-based lidar cannot penetrate high enough to sample complete precipitating hydrometeor layers (only profiling the lower part of a layer) due to signal attenuation. Thus, information about their source clouds is usually not available (Sassen et al., 2005;Di Girolamo et al., 2012;Mega et al., 2012). There is also a lack of systematically observed lidar data on precipitation processes, because most lidar systems are not protected from precipitation. Cloud/precipitation radars are insensitive to small raindrops and droplets in cloud layers. Therefore, the microphysical processes of precipitation formation are not well understood thus far.
Satellite observations have revealed that cold clouds are the major source of liquid precipitation over land (Mülmen-städt et al., 2015). Heterogeneous ice formation pertinent to cold clouds is believed to lead to the generation of rain (Field and Heymsfield, 2015;Bühl et al., 2016;Pfitzenmaier et al., 2018). The ice formation process has been studied extensively by observing liquid-layer-topped ice virgae, because ground-based lidar and radar can reliably sample the entire height ranges of ice virgae and their apparent source cloud bases de Boer et al., 2011;Bühl et al., 2016Bühl et al., , 2019. In stratiform cloud layers at temperatures above −20 • C, precipitating bulk ice particles (ice virga) occurred after bulk liquid phases had formed overhead de Boer et al., 2011). This suggests that the heterogeneous nucleation of ice proceeds via the freezing of supercooled droplets de Boer et al., 2011). Our polarization lidar observations have revealed the detailed vertical structures of falling virgae and their supercooled liquid source cloud layers, indicating that the depolarization ratio values of falling hydrometeors increase rapidly with decreasing altitude on the top of the virgae (Cheng and Yi, 2020).
To study the microphysical processes that occur at altitudes ranging from the apparent source cloud base down to the near-surface during surface precipitation, a 355 nm polarization lidar and a water vapor Raman lidar at the Wuhan University atmospheric observation site were equipped with waterproof transparent roof windows. According to an artificial water-splashing experiment, water accumulation on the lidar roof windows yielded nearly height-independent lidar signal (X, range-corrected signal) attenuation, whereas neither the X vertical structure nor the profile of the volume depolarization ratio δ v (the magnitude and vertical structure) were altered. In addition, water accumulation on the roof windows hardly impacted the lidar-observed subcloud water vapor mixing ratio (q v ) profiles. This allows us to systematically observe precipitation processes (light and moderate rains). Based on our lidar observations obtained on two warm-front occasions, a complete microphysical process is revealed for precipitating hydrometeors pertinent to warm-front-related mid-level stratiform precipitation (the ice-nucleating processes are not covered). This paper first depicts the relevant instrumentation and methodology. Section 3 presents two light and moderate warm-front precipitation cases observed at our lidar site. The summary and conclusions are given in Sect. 4.
Lidar
Precipitating hydrometeor observations were obtained with two newly developed lidars equipped with waterproof transparent roof windows at the Wuhan University atmospheric observatory (30.5 • N, 114.4 • E; 73 m above sea level). The roof windows were designed to project out from the sur-roundings, avoiding a heavy water accumulation on the window glass during rainfall. The two lidars can simultaneously deliver the sequential profiles of the range-corrected signal X, volume depolarization ratio δ v and water vapor mixing ratio q v . All the observation sessions started with clear-sky conditions and ended when heavy precipitation occurred. This allowed us to capture the evolving layer structures of light and moderate precipitation events as well as their precursor clouds present over our mid-latitude site.
Polarization lidar
The polarization lidar has a configuration similar to our 532 nm system (Kong and Yi, 2015), but the transmitter employs a frequency-tripled Nd:YAG laser. It produces emissions of ∼ 150 mJ per pulse at 355 nm with a repetition rate of 30 Hz. A Brewster polarizer is added to improve the polarization purity of the transmitting laser (up to ∼ 10 000 : 1). After beam expansion, the beam with a divergence of 0.15 mrad is transmitted vertically into the atmosphere (to the zenith). The backscattered light is collected by a 20 cm Cassegrain telescope. The field of view (FOV) of the receiver is ∼ 1 mrad. After collimation, the elastically backscattered light passes an interference filter (with a 0.3 nm bandwidth centered at 355 nm) and is then incident on a polarization beam-splitter prism (PBS). To decrease the cross talk between the two orthogonal polarization channels, two additional polarizers are placed on the two output sides of the PBS. The light exiting from the two polarizers is focused onto two photomultiplier tubes (PMTs). The signals from the two PMTs are gathered by a PC-controlled twochannel transient digitizer (TR40-160, manufactured by Licel).
The raw lidar data are stored in both analog and photon counting modes with a range resolution of 3.75 m and a temporal resolution of 1 min. Based on a method originally proposed by Newsom et al. (2009) that was further developed by Zhang et al. (2014), the stored analog and photon-count data are glued to form a reasonable photon-count profile with a large dynamic range. For the cases in this study, the altitude range of signal gluing was ∼ 1.2-3.3 km. The range and temporal resolution of the processed photon-count profiles are 30 m and 1 min, respectively. The starting altitude of the lidar measurements is ∼ 0.3 km, determined based on the overlap of the laser and the field of view of the telescope. The altitude values referenced in this article are all relative to sea level.
The range-corrected lidar signal X is utilized to represent the backscattering intensity (returned laser power) of cloud particles and gravitationally falling hydrometeors (Ansmann et al., 2008). The volume depolarization ratio δ v , defined by the ratio of the perpendicular-to parallel-polarized backscatter coefficients, can be obtained from the two-channel lidar signals along with the relative gain of the parallel and perpendicular channels. The relative gain is determined in advance using a conventional method (Freudenthaler et al., 2009). The magnitude of the δ v value allows us to identify whether the dominant backscattering is attributed to ice crystals or water droplets in a given backscatter volume (Shupe, 2007). In general, liquid water droplets suspended in the atmosphere are nearly spherical and produce a very low depolarization ratio (close to zero) for single scattering at exactly 180 • , while ice crystals, which are usually nonspherical, generate a quite large depolarization ratio in the 180 • backscattering direction. For some mid-level stratiform precipitations, gravitationally falling hydrometeors form initially at altitudes above the 0 • C isotherm level. They fall often as icephase-dominant hydrometeors at subzero temperatures during their early descent. After the falling hydrometeors pass through the 0 • C isotherm level, the snowflake-to-raindrop (ice-to-raindrop) transition can yield a shallow layer of relatively smaller lidar echoes (a local X minimum) that is called "lidar dark band" (Sassen and Chen, 1995;Di Girolamo et al., 2012). The lidar dark band can be used to differentiate between the altitudinal regions with ice-containing particles above the dark band and pure liquid raindrops below the dark band.
It should be mentioned that the particle depolarization ratio δ p is conceptually a more suitable quantity in discriminating spherical and nonspherical particles (hydrometeors) in virga/cloud than the volume depolarization ratio δ v . But, the volume depolarization ratio δ v represents the more basic lidar measurement. In order to validly utilize the δ v magnitude in discriminating spherical and nonspherical depolarizations, we have examined the relationship between δ p and δ v . The δ p magnitude is a well-defined function of δ v , lidar backscattering ratio R and molecular depolarization ratio δ m (Cairo et al., 1999). The molecular depolarization ratio δ m has a value of ∼ 0.004 in terms of our lidar receiver bandwidth (0.3 nm) (Behrendt and Nakamura, 2002). Information about the R value range is available from a combined consideration of the earlier lidar measurements and our current observations on precipitation-related cloud/virga. The typical values of R for enhanced aerosol load are around 2 and for optically thin clouds up to around 10 (Lampert et al., 2010). The R values are ∼5-8 on the upper part of typical shallow (∼ 400 m thick) evaporating ice virgae (see Fig. 4 in Cheng and Yi, 2020). In this study, the R value should certainly be larger than 7 on the precipitation-related virga layer. Based on the analysis for the δ p expression (Cairo et al., 1999) for clouds and virgae, the particle depolarization ratio δ p has a quasi-linear dependence on the volume depolarization ratio δ v and a very weak dependence on lidar backscatter ratio R (when R ≥ 5). This favorable feature of the functional dependences allows us to utilize δ v in discriminating whether the dominant lidar backscattering is attributed to spherical or nonspherical particles in a given backscatter volume. If R min is the minimum of the R value range for the clouds/virgae of interest (e.g., R min = 7 for the precipitation-related virgae in this study), the discrimination criterion of spherical particles expressed by δ v (z) (equivalent to δ p < 0.1) takes the form (see Appendix A for mathematical derivation) The discrimination criterion of nonspherical particles expressed by δ v (z) (equivalent to δ p > 0.2) is given approximately by As noticed from the right-hand sides of Inequalities (1) and (2), the absolute differences between the discrimination threshold values expressed by δ p (0.1 and 0.2) and by δ v are small for clouds/virgae with R min > 7. The unambiguous cloud-phase discriminations based on the volume depolarization ratio δ v in earlier literature (Wang and Sassen, 2001;Intrieri et al., 2002;Shupe, 2007;Ansmann et al., 2009;Lampert et al., 2010) have confirmed the functional relationship between δ p and δ v mentioned above. This allows us to employ δ v with very little threshold-value change in discriminating whether the dominant lidar backscattering is attributed to spherical or nonspherical particles in a given backscatter volume. Specifically, at altitudes above the dark band, the δ v -based discrimination criteria are δ v < 0.09 for spherical water drops/droplets and δ v > 0.17 for ice crystals (based on the above discrimination criteria when R min = 7), while an enhanced depolarization ratio (δ v > 0.1) at altitudes below the dark band indicates the presence of large raindrops. We examined the multiple-scattering-induced depolarization ratio enhancements for an opaque cloud layer composed of dense spherical water droplets by putting a motorized iris on our polarization lidar system. It is indicated that for a receiver FOV of 1 mrad, the enhanced depolarization ratio δ v values due to multiple scattering increased from ∼ 0.03 at the X peak altitude to a maximum value of ∼ 0.27 at the weaksignal cutoff altitude with increasing penetration of laser light into the opaque water-droplet cloud layer. Note that for the same receiver FOV (∼ 1 mrad), the multiple-scatteringinduced depolarization ratio δ v values were all less than 0.04 within the laser light penetration range in a slightly dense water-droplet cloud layer (Hu et al., 2006). Combining the earlier multiple-FOV polarization lidar measurements (Hu et al., 2006) and our similar observations yields a suggestion that for the 1 mrad receiver FOV the multiple-scatteringinduced depolarization ratio values larger than 0.10 should result from an opaque water-droplet cloud layer (see Figs. 2 and 4 in Yi et al., 2021). In other words, for the 1 mrad receiver FOV, the vertical structure of hydrometeors and aerosols present above a dense water-droplet cloud layer with δ v values larger than 0.1 is undetectable by ground-based lidars. An artificial water-splashing experiment was performed on the lidar roof windows to examine the effects of water accumulation. A comparison of the lidar profiles with and without water accumulation on the lidar roof windows is given in Enhanced lidar signal (X) and depolarization (δ v ) values at altitudes around 4.0 km resulted from an optically thick (opaque) water-droplet cloud layer, because there existed a high X value and near-zero δ v value (∼ 0.008) on the cloud base (∼ 3.9 km) (Wang and Sassen, 2001), and also there initially existed a monotonic rapid increase in both the values of X and δ v with increasing penetration of laser light into the layer. The cloud-related structures shown in both the X and δ v profiles were consistent before and after water splashing (particularly, cloud base altitudes). This comparison clearly shows that water accumulation on the lidar roof windows yielded nearly height-independent lidar signal (X) attenuation, and neither the cloud-related X vertical structure nor the profile of the volume depolarization ratio δ v (the magnitude and vertical structure) were altered. This result is physically reasonable.
Water vapor Raman lidar
The configuration of the water vapor Raman lidar used in this study is similar to our 45 cm aperture Raman system (Wu and Yi, 2017), but the current Raman lidar shares the same transmitter with our 355 nm polarization lidar depicted above. It detects inelastic Raman backscatter from water vapor at 407 nm and nitrogen molecules at 387 nm as well as detecting elastic backscattered light by using a 20 cm receiver telescope. The water vapor mixing ratio q v , which is defined as the mass ratio between water vapor and dry air in a given volume, can be obtained from the Raman signals representing water vapor and nitrogen molecules (Whiteman et al., 1992). The Raman lidar system was calibrated by corresponding local radiosonde measurements. A comparison analysis showed that the lidar-derived water vapor mixing ratio profiles agree well with the coincident radiosonde data (the relative deviation is less than 10 % when the water vapor field is horizontally homogeneous on a scale of ∼ 20 km). During the daytime, the water vapor Raman signal is quite noisy at high altitudes due to strong sky background light, so the water vapor mixing ratio profiles are available only at altitudes below ∼ 2 km. A similar artificial water-splashing experiment to that described above was performed on the water vapor Raman lidar roof window. Water accumulation on the roof window hardly had an impact on the obtained subcloud q v profiles.
Radiosonde
The radiosondes were launched at 08:00 LT (00:00 UTC) and 20:00 LT (12:00 UTC) every day from the Wuhan weather station (∼ 23.4 km away from our lidar site). Profiles of the pressure, temperature, relative humidity, wind speed, and direction from the near surface up to a height of 20-30 km were measured. The obtained radiosonde profiles were used to quantitatively determine the meteorological conditions pertinent to the precipitation events and their precursor clouds. The temperature measurement error was less than 1 • C, and the uncertainty in the relative humidity measurement is less than 5 % when the temperature was higher than 10 • C (Nash et al., 2011).
All-sky camera and rain gauge
The cloud photographs are recorded every 2 min by a groundbased all-sky camera located at our lidar site. A tippingbucket rain gauge is used to measure the precipitation rate at the surface. It has a sampling interval of 1 min. For each 0.1 mm of precipitation, the bucket tips and empties, yielding an output signal. The cloud layer was mostly characterized by a mixed phase and had subcloud ice virgae during the later descent ( Fig. 2a and b). After the subcloud virgae reached an altitude (∼ 3.0 km) that was lower than the 0 • C level (∼ 3.6 km), as measured by a conventional radiosonde at approximately 20:00 LT on 27 December at the Wuhan weather station (∼ 23 km away from our lidar site), falling raindrops (precipitation streaks in the X and δ v contour plots) that reached the ground were frequently observed beneath the 3 km altitude until 05:38 LT on 28 December 2017 when the lidar operation terminated. Long survival time of falling ice crystals at altitudes below the 0 • C level might be ascribed to cooling of the surrounding air during their evaporation and melting. The associated water vapor mixing ratio, q v , increased steadily with the descent of the cloud layers (Fig. 2c).
In particular, a high-concentration moisture layer appeared in the subcloud region during the rainfall event. This moisture layer resulted mainly from the evaporation of snow/ice particles and raindrops. Corresponding photographs of the sky taken by a ground-based camera at our lidar site are given at the top of Fig. 2. The light rain lasted for ∼ 8 h and yielded an accumulated rainfall amount of 2.6 mm (rain gauge data obtained at our lidar site). Interestingly, a humid aerosol layer also moved downward from ∼ 4.2 km at ∼ 20:00 LT on 25 December to ∼ 2.3 km at 20:00 LT on 27 December 2017, which appeared to be associated with the warm front. Figure 3 presents the radiosonde profiles that are pertinent to the warm-front cloud at different stages and during precipitation, together with the 1 h mean lidar profiles obtained during the radiosonde launches. The temporally varying cloud properties (e.g., falling cloud base, increasing cloud thickness and variable cloud types) between 20:00 LT on 26 December and 20:00 LT on 27 December 2017 coincided with the classical picture of preceding upglide clouds of an advancing warmfront system. Accordingly, a downgoing moist layer was observed strengthening and broadening with time during this period ( Fig. 3b and c). At the cloud base (except cirrus), the relative humidity over ice had values close to the relative humidity threshold of 84 % that is conventionally used to determine the cloud base heights (Wang and Rossow, 1995;Zhang et al., 2018). Furthermore, the radiosonde data showed that the southwesterly wind mostly prevailed at the cloud altitudes (Fig. 3d, e and f), and the air pressure at altitudes of ∼ 0-5 km dropped continuously by ∼3-5 hPa in the period (not shown here), which did belong to the typical warm-front features.
Associated meteorological conditions
The radiosonde released at 08:00 LT on 28 December 2017 provided measurements of the meteorological conditions when precipitation reached the surface, although the lidar measurements had already terminated (at 05:38 LT) ∼ 2 h earlier. As seen from Fig. 3b (red), the relative humidity reached a maximum of 98 % with respect to water in an altitude range of ∼ 3-4 km, immediately above the tops of the liquid precipitation streaks (at ∼ 3 km; see Fig. 2a and b). Water vapor at altitudes of ∼ 3-9 km was advected from the southwest, as seen in the wind component profiles (Fig. 3f, red). The high water vapor mixing ratios observed at alti-tudes below ∼ 3 km came from the evaporation of falling raindrops.
Microphysical process of precipitating hydrometeors for the light warm-front rain
The X and δ v precipitation streaks were visible in the period between 23:51 LT on 27 December and 05:36 LT on 28 December 2017 ( Fig. 2a and b). The streaks extended from the starting height (∼ 0.3 km) of the lidar measurements to an altitude of ∼ 2.88 km when surface precipitation occurred. A lidar dark band (X minimum) appeared persistently on the top of the precipitation streaks at an ∼ 2.88 km altitude except when the dark band was concealed by a drifting small-scale cloud (at 2.2-2.6 km altitudes during 04:18-05:36 LT on 28 December). This is consistent with earlier lidar observations of stratiform precipitation (Sassen and Chen, 1995;Demoz et al., 2000;Roy and Bissonnette, 2001;Di Girolamo et al., 2012). An inapparent local depolarization (δ v ) minimum was also persistently present at an altitude of ∼ 2.76 km, lying just ∼ 100 m below the dark-band minimum (Fig. 2b). The local δ v minimum represented the completion of the melting process of most falling ice/snow particles. Note that the δ v value decreased as a whole from the ice/snow (including partially melted large particles) values (0.10-0.34) at altitudes above the lidar dark band to the small liquid drop level (≤ 0.04, far less than the δ v -based discrimination threshold value of spherical particles when the lidar backscatter ratio R ≥ 5) at an altitude ∼ 100 m below the dark-band minimum. The lidar dark band definitely differentiates the altitude regions of precipitating ice-containing hydrometeors occurring above and liquid raindrops occurring below. Although the rainfall-induced water accumulation on the roof window of the lidar varied with time, the precipitation streaks and dark band were steadily reasonably displayed in the X and δ v time-height plots ( Fig. 2a and b). This is consistent with the result of our water-splashing experiment.
To further clarify the microphysical process of precipitating hydrometeors, two sets of representative lidar profiles (X, δ v and q v ) for the period that precipitation reached the surface (in Fig. 2) are plotted in Figs. 4 and 5. Figure 4 gives three 1 min X and δ v profiles from 01:12 to 01:14 LT on 28 December 2017 and a 1 h averaged q v profile centered at 01:13 LT on the same day. The lidar dark band appeared at a 2.88 km altitude at approximately 01:13 LT, while the local δ v minimum (< 0.04, far less than the δ v -based discrimination threshold value of spherical particles when R ≥ 5) was located at a 2.76 km altitude. These altitudes represent a typical lidar signature of the snowflake-to-raindrop transition for a variety of stratiform precipitation events. An ice-containing bright band (ice bright band hereafter) with δ v values ranging from ∼ 0.13 to ∼ 0.39 was visible at altitudes ∼ 3.0-3.45 km, just above the lidar dark band (Fig. 4); these altitudes correspond to the "relative lidar bright band" in the literature water vapor mixing ratio q v measured by a water vapor Raman lidar on 26-28 December 2017, which exhibited the passage of a warm front and the resulting hours-long light rain. A sliding average of 60 min was applied to the Raman lidar data. The precipitation streaks surrounded by magenta lines are zoomed in to show their details. Shown on the top of the figure are the corresponding photographs of the sky taken by a ground-based camera at our lidar site, with the third photograph exhibiting the sky illuminated by a 532 nm laser beam during the onset of rainfall. (Sassen and Chen, 1995;Di Girolamo et al., 2012). The ice bright band peaked on its bottom (∼ 3.0 km). It showed a variable vertical structure and intensity (in both X and δ v ) on the timescale of minutes, representing the presence of small-scale fluctuations in the precipitating ice crystals and snowflakes. A liquid-water bright band appeared as a layer of relatively large particle backscatter values, located at ∼ 1.50-2.76 km altitudes, just below the lidar dark band (Fig. 4). It is called "weak lidar bright band" in the literature (Sassen and Chen, 1995;Di Girolamo et al., 2012). The δ v values in the water bright band were ∼ 0.03-0.06, indicating that the enhanced lidar backscattering therein was caused mainly by high-concentration quasi-spherical raindrops with diameters ≤ 1 mm. The water bright band actually represents a major precipitation-related lidar backscattering layer in the liquidphase stage of the light precipitation event. The water bright band appeared to have a larger vertical extent (∼ 1.26 km) than that of the lidar ice bright band.
The lidar q v profile (Fig. 4c) shows an enhanced water vapor mixing ratio at altitudes from ∼ 1.7-3.4 km, indicating the subcloud evaporation of precipitating hydrometeors. In particular, q v was maximized (5.95 g kg −1 ) around the water bright band center (at ∼ 2.34 km), suggesting that this altitude was a primary subcloud evaporation region for this light warm-front precipitation event. Furthermore, the q v values in the water bright band increased as precipitation continued (Fig. 2c). Combining the vertical structures of X, δ v and q v in the water bright band (Figs. 2 and 4) yields the suggestion that most falling small-sized raindrops shrunk or vanished in the water bright band due to evaporation, whereas a small portion of large-sized raindrops survived via collisioncoalescence processes and fell out of the water bright band. At altitudes below the water bright band, the precipitationrelated lidar backscattering (X) apparently weakened (Fig. 4a, in which the enhanced X values at altitudes from 0.3-0.7 km resulted from the boundary layer aerosols), indicating low-density raindrops there, whereas δ v first increased with decreasing height and then decreased after reaching a maximum (0.13-0.16) at an altitude of approximately 0.6 km (Fig. 4b). Here the magnitude and altitude variation of the lidar depolarization ratio δ v values allow us to identify where large-sized raindrops form and break up. Falling small-sized raindrops (equivalent diameter ≤ 1.0 mm) are quasi-spherical (Pruppacher and Klett, 1997) and yield small δ v values (generally less than 0.1), whereas falling large-sized raindrops (equivalent diameter > 2.8 mm) become nonspherical (with flat or hollow bottom in falling direction) (Pruppacher and Klett, 1997) and lead to large δ v values (larger than 0.1). In fact, prominent δ v peaks (∼ 0.1-0.4) at altitudes of approximately 0.6 km are always observed in the δ v profiles related to reaching-surface precipitation in the present lightrain case (Fig. 2). The δ v maxima at an altitude of ∼ 0.6 km are much larger than the typical values (< ∼ 0.07) observed by our 355 nm polarization lidar at approximately the same altitude during rainless days. Here we can exclude a possibility that the δ v maxima (∼ 0.1-0.4) at ∼ 0.6 km altitude resulted from multiple scattering by dense droplets around this altitude. As mentioned above, for the 1 mrad receiver FOV, a dense water-droplet cloud layer with the multiplescattering-induced depolarization ratio δ v values larger than 0.1 is optically opaque. In contrast to this situation, in our case, when the prominent δ v peak (∼ 0.1-0.4) around 0.6 km altitude occurred, the vertical structure of the precipitation streaks at altitudes far above 0.6 km (e.g., ice bright band, lidar dark band and lidar water bright band) was unambiguously detected by our polarization lidar, indicating that the enhanced depolarization ratios around 0.6 km altitude cannot be caused by multiple scattering from dense spherical water droplets therein. Furthermore, since most falling raindrops evaporated and vanished in the liquid-water bright band as indicated by the enhanced water vapor mixing ratio therein and rapidly decreasing lidar signal on the bottom of the water bright band, small droplets at altitudes below the water bright band were hardly dense enough to generate a strong multiple scattering with δ v ≥ 0.1. Therefore, our observational results suggest that sparse large raindrops that fall out of the water bright band with higher fall velocities further grow in size by collecting smaller raindrops along their fall paths. They grow to sizes at which spontaneous breakup occurs at an altitude of approximately 0.6 km. In brief, our lidar observations reveal for the first time (to our knowledge) the collisioncoalescence growth and subsequent spontaneous breakup of falling raindrops that actually take place in the natural atmosphere. They represent the posterior microphysical processes necessary for the reaching-surface precipitation production. Interestingly, the size maximization of falling raindrops as shown by the strongest nonspherical shapes (maximum depolarization ratio values) always appeared at an altitude of ∼ 0.6 km for a variety of mid-level stratiform precipitations (in light of our observations). Obviously, the explanation to this ubiquitous feature needs further observational and modeling efforts. As seen in Fig. 2b, the boundary layer aerosols had little impact on the δ v precipitation streaks. In addition, at altitudes below 1.5 km, the q v values decreased with increasing altitude, reflecting a normal altitude distribution of the boundary layer water vapor.
Based on the radiosonde temperature data obtained at approximately 20:00 LT on 27 December 2017 (Fig. 3a, orange), the 0 • C isotherm level was at an altitude of ∼ 3.6 km, and a warm-front-related inversion layer appeared just below the 0 • C level with a local temperature maximum (2.2 • C) at 3.33 km and a local minimum (1.0 • C) at 2.84 km. The lidar dark band (at 2.88 km, with a temperature of ∼ 1.0 • C) was located ∼ 720 m below the 0 • C level. In comparison with the results reported in the literature (Sassen and Chen, 1995;Demoz et al., 2000;Sassen et al., 2005;Di Girolamo et al., 2012), the observed ∼ 720 m distance of the dark-band minimum to the 0 • C level and the low dark-band temperature (∼ 1.0 • C) are somewhat peculiar for light-precipitation cases. In the current case, the melting process might be delayed by the temperature structure (with a small lapse rate) of the inversion layer. However, it should be mentioned here that the radiosonde launching site was ∼ 23.4 km away from our lidar site. Figure 5 presents three 1 min lidar X and δ v profiles displaying the time span from 02:30 to 02:32 LT on 28 December 2017 and a 1 h averaged lidar q v profile centered at 02:31 LT on the same day, depicting the microphysical process of precipitating hydrometeors for slightly strong precipitation that reached the surface during the light rain event. Although the water bright band and aerosol backscatter layer below the dark band became evidently weak compared to those seen in Fig. 4a (due to precipitation attenuation), the altitude of the dark-band minimum (2.85 km) was very close to that (2.88 km) obtained from Fig. 4a. The magnitude (∼ 0.03) and occurrence altitude (2.76 km) of the local δ v minimum were consistent with the corresponding values (less than 0.04 and 2.76 km, respectively) observed in Fig. 4b. Furthermore, the depolarization maxima (∼ 0.17) associated with reaching-surface precipitation still appeared at an altitude of approximately 0.6 km, which was also similar to that seen in Fig. 4. The observational facts confirm the result gathered from our water-splashing experiment in which the thin liquid-water layer on the roof windows of the lidars caused nearly altitude-independent attenuation on the X profiles and had no effect on the δ v profiles. The profile characteristics shown in Fig. 5 are mostly similar to those mentioned above for Fig. 4, but some newly emerging features need to be illustrated. Figure 5 exhibits an ice bright band stronger than the concurrent water bright band. This result is different from our observations obtained at ∼ 01:12 LT (Fig. 4) but is consistent with earlier lidar observations (Sassen and Chen, 1995;Di Girolamo et al., 2012). The ice bright band observed at approximately 02:30 LT had the X maxima at its bottom (at an altitude of ∼ 3.0 km) and a small vertical extent (∼ 0.21 km, from 3.00 to ∼ 3.21 km due to precipitation attenuation). The X maxima corresponded to the local minima of the depolarization ratio ( Fig. 5a and b). Interestingly, this inverse relationship between the backscatter and depolarization values on the bottom of the ice bright band is nearly ubiquitous in the precipitation lidar profiles obtained in the present case. Since the depolarization δ v showed moderate minima (∼ 0.08-0.10) at an altitude of ∼ 3.0 km (Fig. 5b), the ice bright-band maxima observed at approximately 02:30 LT might reflect backscattering from high-concentration partially melted large particles therein. On the band's altitudinal extension (from 3.06-3.21 km), the markedly enhanced depolarization values (∼ 0.17-0.34) indicate the presence of ice crystals and large snowflakes (Sassen and Chen, 1995;Di Girolamo et al., 2012). The water vapor mixing ratio q v showed slight enhancements at altitudes of ∼ 1.5-3.0 km at approximately 02:30 LT compared with that measured at approximately 01:12 LT.
As seen from the X and δ v precipitation streaks at altitudes below ∼ 1.5 km ( Fig. 2a and b), precipitation that reached the surface was intermittent. During periods without reachingsurface precipitation, our lidars were able to sample both a complete virga (from the rain to the snow regions) and a shallow mixed-phase cloud layer immediately above the virga under weak optical attenuation conditions. Such an example is shown in Fig. 6. The lidar profiles above the dark band clearly exhibit the typical structure characteristics of a liquidtopped mixed-phase cloud (a shallow liquid cloud layer and ice virga below) (see Fig. 6 in Wang and Sassen, 2001). The mixed-phase cloud top layer (at altitudes of ∼ 4.6 km) was of high X values and very low δ v values (∼ 0.01), while lower part of the cloud was characterized by significantly lower X values and higher δ v values (with a maximum up to ∼ 0.33). Furthermore, the cloud top layer had a maximum water vapor mixing ratio q v and a temperature of ∼ −8.5 • C (based on radiosonde data at ∼ 20:00 LT on 27 December). Combining with the schematic representation of commonly observed mixed-phase cloud layers (see Fig. 1 in Bühl et al., 2016), the current observations suggest that the cloud top layer should mainly be composed of liquid droplets (that were not dense enough to yield detectable multiple scattering), and the lower part of the cloud was mainly precipitating ice crystals (falling ice virga). The liquid-topped mixedphase cloud (a liquid cloud layer and ice virga below) (Bühl et al., 2016) might be fundamental monomers that constitute mid-level precipitating stratiform clouds. According to the expressions on the right-hand sides of Inequalities (1) and (2), the δ v -based discrimination threshold values were, respectively, 0.09 for spherical particles and 0.17 for nonspherical particles when the lidar backscatter ratio R had a value of 7 (the minimum of the R value range) on the upper part of the precipitation-related virga (Lampert et al., 2010;Cheng and Yi, 2020). Thus, the δ v magnitude of the falling virga increased from the liquid-water values of ∼ 0.02-0.07 (< 0.09) at an altitude of 4.38 km to the ice/snow values of ∼ 0.21-0.33 (> 0.17) at an altitude of 4.02 km. The falling ice crystals yielded a very weak ice bright band at an altitude of ∼ 3.0 km and then melted into liquid drops at an altitude of ∼ 2.76 km (the local δ v minimum). During their further descent, the liquid drops fully vanished due to evaporation, leaving a lidar-detectable rain virga (water bright band) without reaching-surface precipitation. In contrast to the situation during precipitation that reached the surface, no clear-cut δ v enhancement occurred at an altitude of approximately 0.6 km when there were only virgae suspended in air. Similar results were discerned for other lidar profiles shown in Fig. 2, in which a complete mixed-phase cloud layer could be detected.
During the light warm-front rain event, since the reachingsurface precipitations and virgae occurred alternately on a small timescale from a few minutes to tens of minutes and since their precipitation streaks had nearly the same darkband structures ( Fig. 2a and b), both reaching-surface precipitation and virgae would come from the same source cloud (because a warm-front cloud system is generally widespread and slowly varying). Reaching-surface precipitation (drizzle) arose when the precipitation rate was high below the shallow water-droplet-dominated cloud layer (apparent source cloud), while virgae without reaching-surface precipitation took place when the subcloud precipitation rate was slightly low. Therefore, the current lidar observations reveal the microphysical process of precipitating hydrometeors related to light warm-front rain. Both reaching-surface rainfall and virgae suspended in air began as ice-phase-dominant hydrometeors fell out of a liquid apparent source cloud layer at altitudes above the 0 • C isotherm level. The depolarization ratio magnitude of falling hydrometeors increased from the liquidwater values (δ v < 0.09) to the ice/snow values (δ v > 0.20) during the first 100-200 m of their descent. Subsequently, the falling hydrometeors yielded a dense layer with an ice/snow bright band occurring above and a liquid-water bright band occurring below (separated by a lidar dark band) as a result of crossing the 0 • C level. In the ice/snow bright band, large particles would form via the cold rain processes (riming and aggregation), because the broad size distributions of the pristine hydrometeors falling out of the apparent source cloud base could lead to local accretion. The production efficiencies of large particles would depend on the magnitude of the rain rate below the apparent source cloud base and size distributions of the pristine falling hydrometeors. The local depolarization minimum (δ v ≤ 0.04, far less than the δ v -based discrimination threshold value of spherical particles when R ≥ 5) was persistently observed immediately beneath (∼ 100 m below) the lidar dark-band minimum (X minimum). This displayed that the completion of the melting process of most falling ice particles took place at altitudes (hundreds of meters) below the 0 • C isotherm level. The liquid-water bright band (with a geometrical thickness of ∼ 1 km) just below the lidar dark band was characterized by enhanced X values and small δ v values. There existed a high-concentration moisture (large q v values) in this bright band. These features indicate that the liquid-water bright band resulted from gravitationally falling, dense evaporating liquid drops. In terms of the lidar-measured profiles during reaching-surface precipitation, at altitudes below the water bright band, the precipitation-related lidar backscattering apparently weakened, while δ v first increased with decreasing altitude and then decreased after reaching a prominent maximum at an altitude of ∼ 0.6 km. The lidar profiles for the virgae showed narrower and weaker water bright bands than those observed when precipitation reached the surface. Moreover, during virga occurrence, there was no perceptible depolarization enhancement at an altitude of ∼ 0.6 km. By combining the abovementioned lidar observations, a picture on the microphysical processes of falling hydrometeors in liquid-phase stage emerged. After going through the dark band, most falling raindrops shrank or vanished in the water bright band due to evaporation, whereas a few large raindrops survived and fell out of the water bright band when the rain rate below the apparent source cloud base was high enough. The large raindrops might come from both the complete melting of large falling ice/snow particles and collision-coalescence formation in the dense water bright band. Sparse, large raindrops with high fall velocities further grew in size by collecting smaller raindrops along their fall paths. At an altitude of ∼ 0.6 km, the large raindrops grew to the sizes at which spontaneous breakup could occur, yielding reaching-surface precipitation. When the rain rate below the apparent source cloud base was low, nearly none of the large raindrops fell out of the water bright band. Consequently, there were only virgae suspended in air (without reachingsurface precipitation). Figure 7 shows an example of moderate warm-front precipitation that occurred on 4 March 2019. Both the descending precursor clouds and the X and δ v precipitation streaks are generally similar to those seen in the first example (Fig. 2). The precursor clouds are cirrus (photo I in Fig. 7), altostratus (photo II) and altocumulus (photo III). The reachingsurface precipitation started just after the subcloud ice virgae reached an altitude (∼ 2.7 km) slightly lower than the 0 • C level (∼ 3.0 km). The δ v precipitation streaks show the upper portion (ice bright band) containing ice/snow particles (mostly δ v > 0.3) and the lower portion (water bright band and below) being composed of liquid drops (δ v ≤∼ 0.12 except for the δ v maxima that occurred due to raindrop-size growth at an altitude of ∼ 0.6 km). The δ v values in both the ice bright band and water bright band (Fig. 7b) were generally larger than their counterparts in the light-rain example (Fig. 2b), indicating that more large ice/snow particles and raindrops were involved in the moderate precipitation than in the light-precipitation event. Partially melted, large falling particles sometimes concealed the lidar dark band produced by the melting effect of most relatively small-sized particles in precipitating hydrometeors, making the band somewhat fuzzy (Fig. 7a). Accordingly, the altitude of the local δ v minimum (on the lidar dark band) became somewhat unsteady (Fig. 7b). The δ v maxima at an altitude of approximately 0.6 km (Fig. 7b) were apparently larger than those shown in Fig. 2b, indicating that more breakupsize raindrops formed via collision-coalescence processes therein than in the light-rain case. Specifically, the δ v maxima at an altitude of ∼ 0.6 km were as high as ∼ 0.27-0.35 at ∼ 23:38 LT, which corresponded well to the large rainfall rate of 3.2 mm h −1 measured from our rain gauge on the ground. The (apparent) source cloud for this moderate rain event was invisible by the lidars due to strong optical attenuation. Therefore, the following analysis was limited to the ice bright band and below. A strong southerly wind prevailed at altitudes of 0-12 km in light of the radiosonde data obtained at 20:00 LT on 4 March 2019. A high-concentration moisture layer appeared in the subcloud region at altitudes from ∼ 0.5 to ∼ 3.0 km during the rainfall event, indicating the subcloud evaporation of precipitating hydrometeors. The moderate rainfall lasted for ∼ 14 h, yielding an accumulated rainfall amount of 23.9 mm on the ground.
Associated meteorological conditions
The conventional radiosonde profiles associated with the moderate warm-front precipitation and its precursor clouds and the 1 h mean lidar profiles obtained during the radiosonde launches are plotted in Fig. 8. At ∼ 08:00 LT on 4 March 2019, the sky was nearly cloudless (Fig. 8d, blue), and high relative humidity occurred only at altitudes below 1.2 km (Fig. 8b, blue), while the northwesterly wind prevailed at altitudes from 1.7-10.5 km. This indicated that the warm front had not yet reached our lidar site. At ∼ 20:00 LT on 4 March, a moist layer occurred at altitudes ranging from ∼ 4.8 to 8.0 km with increased relative humidity over water of 80 %-95 % (Fig. 8b, green). An evaporating ice virga was observed at altitudes from ∼ 3.6-4.6 km ( Fig. 8d and e, green), just below the moisture layer peak. The apparent source cloud of the virga was invisible by lidars due to strong optical attenuation. A potential occurrence region for the apparent source cloud ranged in altitude from 4.8-6.0 km, where the relative humidity was larger than 90 % (Fig. 8b, green). The southerly wind prevailed at altitudes from 0-12 km (Fig. 8f, green), indicating that the moisture layer and altocumulus (photo III in Fig. 7) were precursors of the warm-front precipitation event. The radiosonde profiles obtained at 08:00 LT on 5 March 2019 showed the meteorological conditions during the moderate warm-front precipitation event after the lidar measurements had already terminated (at 00:51 LT on 5 March). As shown in Fig. 8b (orange), the relative humidity over water had values of 97 %-98 % at altitudes from 0-5.65 km, corresponding to a precipitation rate of ∼ 1.8 mm h −1 (rain gauge record) at approximately 08:00 LT on 5 March. The air pressure from the radiosonde data at Wuhan showed a persistent decrease (by ∼ 2-4 hPa at altitudes of ∼ 0-5 km) during the observational period from the precursor clouds to precipitation (between 08:00 LT on 4 March to 08:00 LT on 5 March) that reflected the warmfront passage.
Microphysical process of precipitating
hydrometeors for the moderate warm-front rain identical dark-band locations (the X minima is located at ∼ 2.04 km, and the local δ v minima is located at ∼ 1.96 km). The dark-band minima appeared ∼ 960 m below the 0 • C level at a radiosonde temperature of ∼ 6.0 • C. Such a long survival time of falling ice crystals at altitudes below the 0 • C level was due to cooling of the surrounding air during their evaporation and melting. At 22:22 LT, a weak X peak occurred at the dark-band altitudes with δ v values ranging from ∼ 0.21-0.29, indicating that partially melted large particles passed through the dark band. As seen from Fig. 9, the X and δ v precipitation streaks had complicated vertical structures at altitudes below the dark band and showed strong variations on the timescale of minutes. In particular, enhanced depolarization (0.07-0.12) occurred within the water bright band. These profile details confirm that large-sized particles sometimes fell out of the ice bright band during the moder- The radiosonde profiles quantitatively present the meteorological conditions pertinent to the warm-front clouds and precipitation.
ate warm-front precipitation event, concealing the lidar dark band produced by the melting effect of most relatively smallsized particles in precipitating hydrometeors. This effect appears to explain why the lidar dark band became fuzzy for the moderate precipitation event (Fig. 7). Note that the δ v maxima (∼ 0.2), which occurred at an altitude of approximately 0.6 km, were slightly larger in the moderate warm-front rainfall than those observed in the light-rainfall warm-front example. This suggests a larger concentration of raindrops of spontaneous breakup sizes around this altitude. The water vapor mixing ratio q v had values ranging from 3.4-4.4 g kg −1 at altitudes from 1.0-3.0 km (from the bottom of the water bright band to the 0 • C level). Figure 10 gives three 1 min lidar X and δ v profiles representing the period from 23:37 to 23:39 LT on 4 March 2019 and a 1 h averaged lidar q v profile centered at 23:38 LT on the same day; these profiles exhibit the vertical structures of the X and δ v precipitation streaks as well as the water vapor mixing ratio observed when the surface precipitation rate was highest (3.2 mm h −1 ) during the moderate warm-front precipitation event (yielding thick liquid-water accumulation on the roof windows of the lidars; see photo V in Fig. 7). The ice bright band, dark band and water bright band were roughly discernible in the three 1 min X profiles despite the considerable fluctuations that occurred on the timescale of minutes. Large ice/snow particles occurred on the ice bright band (at an altitude of approximately 2.5 km), because the δ v values were larger than 0.3 therein. The dark band located ∼ 700 m below the 0 • C level (3.0 km) had δ v values ranging from 0.13-0.19 and a temperature of 4.3 • C, reflecting that there were partially melted large particles present in the dark band. In the height range of the water bright band, the depolarization ratio increased from ∼ 0.04-0.06 at an altitude of approximately 2.09 km to ∼ 0.12-0.15 at an alti- tude of 0.9 km, indicating that more large raindrops formed via collision-coalescence processes therein than in the lightrainfall warm-front example (Figs. 4 and 5). The δ v maxima observed at an altitude of ∼ 0.6 km were as high as ∼ 0.27-0.35 corresponding well to the high measured rainfall rate of 3.2 mm h −1 (rain gauge record on the ground). As mentioned above, for the 1 mrad receiver FOV, if such large δ v values (∼ 0.27-0.35) came from the multiple scattering by a dense water-droplet cloud layer around 0.6 km altitude, the cloud layer would be optically opaque. It would conceal the vertical structure of the precipitation streaks at altitudes above 0.6 km. In contrast to this situation, as seen from Fig. 10, Figure 10. Three 1 min lidar X and δ v profiles covering the period from 23:37 to 23:39 LT on 4 March 2019 and a 1 h averaged lidar q v profile centered at 23:38 LT on the same day, exhibiting the vertical structure of the X and δ v precipitation streaks as well as the water vapor mixing ratio when the surface precipitation rate was highest (3.2 mm h −1 ) during the studied moderate warm-front precipitation event (yielding thick liquid-water accumulation on the roof windows of the lidars; see photo V in Fig. 7). the vertical structure of the precipitation streaks at altitudes above 0.6 km was clearly discerned by our ground-based polarization lidar, indicating that the enhanced depolarization ratios around 0.6 km altitude cannot be caused by multiple scattering from dense spherical water droplets therein. Furthermore, since most falling raindrops evaporated and vanished in the liquid-water bright band as indicated by the enhanced water vapor mixing ratio therein and rapidly decreasing lidar signal on the bottom of the water bright band, small droplets at altitudes below the water bright band were hardly dense enough to generate a strong multiple scattering with δ v ≥ 0.1. Therefore, it is suggested that the prominent δ v peak at an altitude of approximately 0.6 km reflected the collision-coalescence growth of falling large raindrops and their subsequent spontaneous breakup. The q v values at altitudes from ∼ 0.7-3.0 km ranged from 5.3-7.3 g kg −1 (Fig. 10c), indicating overall moisture enhancement compared to those values measured at the onset of the moderate warm-front precipitation event (Fig. 9c).
Summary and conclusions
Observations of precipitation and associated precursor clouds were made with two co-located lidars (a 355 nm polarization lidar and water vapor Raman lidar) equipped with waterproof transparent roof windows at the Wuhan University atmospheric observatory (30.5 • N, 114.4 • E; 73 m above sea level). The lidar observations obtained during reachingsurface precipitation events indicate that the rainfall-induced liquid-water accumulation on the roof windows of the li-dars yielded a nearly height-independent lidar signal (rangecorrected signal X) attenuation, whereas neither the X vertical structure nor the magnitude or vertical structure of the volume depolarization ratio (δ v ) were altered. Furthermore, the liquid-water accumulation on the roof windows of the lidars also had nearly no effect on the obtained subcloud profiles of the water vapor mixing ratio measured by the Raman lidar. These observations are consistent with the results of our artificial water-splashing experiment on the roof windows.
Warm-front precipitation events and their precursor cloud evolution were reported in this paper based on two case studies corresponding to light and moderate rainfall occurring at the Earth's surface. The lidar-observed precursor clouds showed a systematic descent for each case. The descending clouds changed gradually from cirrus and altocumulus to altostratus before rainfall occurred, with gradually increasing moisture, and the southwesterly wind prevailed over most altitude ranges of the cloud layers. These features indicate that, in each case, a warm front was approaching our lidar site. The precursor clouds had underlying ice virgae in their later descent phases. When the subcloud virgae reached an altitude slightly below the 0 • C level, rainfall at the surface began. The hours-long precipitation streaks shown in the lidar signal (X) and volume depolarization ratio (δ v ) profiles reveal some ubiquitous features of the microphysical processes of precipitating hydrometeors.
For the light warm-front rain event, since the reachingsurface precipitations and virgae occurred alternately over a short timescale from a few minutes to tens of minutes and since their respective precipitation streaks had nearly the same dark-band structures, both reaching-surface precipitations and virgae originate from the same source cloud (because a warm-front cloud system is generally widespread and slowly varying). Through an analysis combining the lidar profiles of reaching-surface precipitations and virgae, we find that the reaching-surface precipitation began as icephase-dominant hydrometeors fell out of a liquid apparentsource-cloud layer at altitudes above the 0 • C isotherm level. The depolarization ratio magnitude of falling hydrometeors increased from the liquid-water values (δ v < 0.09) to the ice/snow values (δ v > 0.20) during the first 100-200 m of their descent. Subsequently, the falling hydrometeors yielded a dense layer with an ice/snow bright band occurring above and a liquid-water bright band occurring below (separated by a lidar dark band) as a result of crossing the 0 • C level. In the ice/snow bright band, larger particles formed by riming and/or aggregation, because the broad size distributions of the pristine hydrometeors falling out of their apparentsource-cloud base could lead to local accretion. The completion of the melting process of most falling ice particles appeared at altitudes (hundreds of meters) below the 0 • C isotherm level, as indicated by the local depolarization minimum located immediately beneath (∼ 100 m) the observed lidar dark-band minimum. After going through the dark band, most falling raindrops shrunk or vanished in the water bright band due to evaporation, whereas a few large raindrops survived and fell out of the water bright band when the rainfall rate below the liquid apparent-source-cloud base was high enough. Large raindrops might originate from both the complete melting of falling large ice/snow particles and collision-coalescence formation in the dense water bright band. We also find that a prominent depolarization δ v peak (0.10-0.40) always occurred at an altitude of approximately 0.6 km when precipitation reached the surface, reflecting the collision-coalescence growth of large falling raindrops (sparse large raindrops with high fall velocities further grew in size by collecting smaller raindrops along their fall paths) and subsequent spontaneous breakup. The δ v peak observed at an altitude of ∼ 0.6 km provides an indicator in advance (∼ 2 min) of precipitation that reached the surface.
For the moderate warm-front rain event, although the apparent source cloud was undetected, owing to strong attenuation, the lidar-detectable microphysical process (at the altitudes of the ice bright band and below) was similar to that observed in the light-rain case. However, the δ v values in both the ice bright band and water bright band were generally larger than their counterparts in the light-rainfall case, indicating that more large ice/snow particles and raindrops were involved in moderate precipitation. Furthermore, the X and δ v precipitation streaks had complicated vertical structures at altitudes around and below the dark band and showed strong variations on the timescale of minutes. These profile details suggest that large particles sometimes fell out of the ice bright band during moderate precipitation, concealing the lidar dark band produced by the melting effect of most relatively small particles in precipitating hydrometeors. Thus, the lidar dark band became fuzzy. The δ v maxima observed at an altitude of approximately 0.6 km were also larger than those observed in the light warm-front rain case. This suggests larger concentrations of raindrops with spontaneous breakup sizes around this altitude.
has a value range of [0.2− 0.24 R min +0.2 , 0.2). Since δ v,threshold (R) is a slowly varying function of R as seen from Eq. (A7) (particularly when R min ≥ 5), the discrimination criterion of nonspherical particles expressed by δ v (z) can be written approximately as When R min = 7, the discrimination criterion of nonspherical particles is given by δ v (z) > 0.167, which is equivalent to δ p (z) > 0.2 approximately. In conclusion, the particle depolarization ratio δ p has a quasi-linear dependence on the volume depolarization ratio δ v and a very weak dependence on lidar backscatter ratio R (when R ≥ 5). This favorable functional dependence allows us to utilize δ v in discriminating whether the dominant lidar backscattering is attributed to spherical or nonspherical particles in a given backscatter volume. If R min is the minimum of the R value range for interested clouds/virgae (e.g., R min = 7 for the precipitation-related virgae), the discrimination criterion of spherical particles expressed by δ v (z) is given by Eq. (A5), while the discrimination criterion of nonspherical particles expressed by δ v (z) is given approximately by Eq. (A8).
Data availability. Lidar data used to generate the results in this work are available from the corresponding author with permission (E-mail: yf@whu.edu.cn).
Author contributions. YY performed the lidar measurements, made the data analysis and wrote the initial article. FY conceived the project, led the study and finalized the article. FL, YZ and CY built the lidar systems for precipitation observations. YH participated in scientific discussions and suggested analysis. All authors discussed the results and commented on the article.
Competing interests. The contact author has declared that neither they nor their co-authors have any competing interests.
Disclaimer. Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. | 2021-05-04T22:05:17.649Z | 2021-04-07T00:00:00.000 | {
"year": 2021,
"sha1": "e24ca457cd0921a084c20d918cfe6c44e6ca076a",
"oa_license": "CCBY",
"oa_url": "https://acp.copernicus.org/articles/21/17649/2021/acp-21-17649-2021.pdf",
"oa_status": "GOLD",
"pdf_src": "MergedPDFExtraction",
"pdf_hash": "f440c9ecb1d99edd3ca6c29211d022f09d2debe4",
"s2fieldsofstudy": [
"Environmental Science",
"Physics"
],
"extfieldsofstudy": [
"Geology"
]
} |
212540107 | pes2o/s2orc | v3-fos-license | RISK PERCEPTION IN E-COMMERCE: A HOLISTIC REVIEW OF EMERGING ONLINE SHOPPING IN INDIA
This paper examines the perceived risks associated with online shopping. Perceived risk defined as a consumer’s belief about the potential losses or other negative outcomes from transacting on the internet. Fifty-eight percent of internet users describe online shopping as a frustrating, confusing, and overwhelming activity. Even though majority of the online shoppers have perceive risk, religion, income and occupation are not found to be significant factors in risk perception.
Introduction
Modern world is rightly described as 'Internet world' because daily life is all about finger tips in this electronic era.Electronic commerce is a powerful concept and process as it has fundamentally changed the current human life.Electronic commerce is one of the main areas of Information, Technology and Communication (ICT).This style of trading has received a major boost in recent years due to the enormous benefits associated with it.It is claimed that electronic commerce has surpassed many of the limitations of traditional business.Online shopping is a form of electronic commerce which allows consumers to directly buy goods or services from a seller over the internet using a web browser.It is also known as E-tail, from electronic retail or E-shopping.
The volume of online shopping has been rapidly growing all over the world.The percent of online users who have made an internet purchase was 85% in 2010 in the US, and the number of US digital shoppers is expected to grow from 137 million in 2010 to 175 million in 2017.Clearly, the statistics imply the crucial role of e-commerce in world trade and there have been significant benefits both for sellers and buyers.Online shopping improves market efficiency and enhances welfare because of the ability of shoppers to make better quality decisions while buying online.This is because there appears to be a strong link between product varieties of online shopping and social welfare.
Online buying is growing exponentially throughout the world.UK is biggest online shopping nation in the world followed by Denmark, Norway and Korea.In Nigeria and other African countries the number of users is still far below the world average which is around 30%.Now-adays due to the availability of faster internet networks e-commerce activities are expanding at a faster rate in Nigeria, South Africa and Kenya.In South Africa, 51% of individuals with internet access shop online whereas, in Kenya, only 18-24% makes online purchases.In Nigeria approximately 28% of the population has internet access according to ITU figure.According to a study the swiftness of online shopping in Pakistan is sluggish compared to India and other countries.The total value of e-commerce activities in India has surpassed Rs.5.7 billion during 2004-05 and Rs.23 billion by the year 2006-07.According to Rajan Anandan, VP & Managing Director of Google India, with approximately 8 million Indians shopping online in 2012, online shopping industry in India is growing rapidly and will continue to see exponential growth.According to a survey conducted by IMRB International and IAMAI, there was an estimate of 137 million internet users in the country as of June 2012 of which 99 million were from urban parts of the country, the remaining 38 million were from rural India.So, online shopping is flourishing not only in metros and urban areas but also in rural areas.
Although there is an increase in the overall volume of e-commerce, there is still a need to enhance the total value created by online shopping in order to provide further benefits in terms of raising welfare.But this is a challenging process because the most significant obstacle to reaching this goal is the high level of consumers' perception of the risk associated with the shopping on the internet.Perceived risk is an important barrier for online consumers because it threatens the operation of the e-commerce, making security issues a fundamental concern.Risk is a much more salient concern in e-commerce than in many other social and economic interactions involving uncertainty due to the nature of the financial transactions and multiple levels of uncertainties.However, a vast body of the literature investigates increase in the volume of e-commerce by reducing risk issues.Hence, both scholars and online retailers are increasingly concerned about reducing risk perceptions in online shopping.
Statement of the Problem
Emergence of online shopping has considerable impact on modern life style.The development of e-commerce has increased the popularity of online shopping worldwide especially with the advent of Technology Acceptance Model (TAM).It was developed by Davis (1989), adopted from the Theory of Reasoned Action, which is designed to explain human behaviours in general.In accordance with this theory, the TAM proposes that the use of computers is determined by behavioural intention, which measures individual's intention to perform a specific behaviour.The TAM proposes two specific beliefs; perceived ease of use and perceived usefulness, which determine individual's intention to use a technology.
Business to Consumer (B2C) is an internet and electronic commerce (e-commerce) model that denotes a financial transaction or online sale between a business and consumer.B2C is also known as business-to-consumers.India is one of the major countries having a wide range of internet related activities.It was estimated that increasing internet penetration and growing preference for shopping online will drive the e-commerce market in India to USD 15 billion by 2016 with a whopping 100 million people going online to shop, today.
Despite of all these merits, online shopping has risks too.Perceived risk defined as a consumer's belief about the potential losses or other negative outcomes from transacting on the internet.Fifty-eight percent of internet users describe online shopping as a frustrating, confusing, and overwhelming activity.Past research has found that a major inhibitor of online shopping is the uncertainty or perceived risk associated with online purchasing.These risks are great barriers to the online purchase intention of customers.It includes social, financial, physical, performance, time, and psychological risks.
Most of the studies on e-commerce have concentrated mainly on risk aspect in various developed countries such as Saudi Arabia, Malaysia, and USA.All these studies show that customers dealing with online shopping face several risks.But the interesting question is why people still prefer online shopping to traditional shopping even if there are perceived risks.However, there is a need for systematic investigation of risk perception in online shopping since the literature on people's preference over traditional shopping in the presence of perceived risk has not received sufficient attention.So this particular study is an attempt to identify the various types of risk involved in e-commerce and why people still depends on online purchasing in the presence of perceived risk in a developing country like India.
Literature Review
There are very limited studies, which are directly relevant to the present study.They have been taken from journals, articles, PhD thesis and unpublished research work, concentrated mainly in developed countries like Saudi Arabia, USA, Malaysia, New Zealand etc. Still many recent researches have been found in developing countries like India and Jordan as well.
Dr. PanicosGeorgiades (2000), observed in his study that there were no differences between males and females in all three occupational groupings with regard to security and convenience of e-commerce.In particular respondents' attitudes toward security were found to be in agreement in that "they did not feel confident with the provision of information concerning their personal and financial details and that technology backing the internet is reliable".This suggests that companies can standardize their communications strategies aiming at alleviating the fears of internet users with regards to security concerns.
Sami Alsmadi (2002), study pointed out that most Jordanian consumers are likely to have enough knowledge and skills in using the computer and dealing with the internet, and have reasonable access to internet services, with a positive impression about the current presentation and promotion of companies' web sites on the internet.The study concluded that the issue of security of online transactions seems to be a major factor that restricts the willingness to make a better use of online shopping and also no significant differences in consumer attitudes due to demographic variables, with the exception of income.
Internet and Online Association of India (2005), made a survey of e-commerce security in 2005.
The study was conducted through online and gathered information regarding the solicited information on the user's profile, internet usage, their perception of the security associated with transacting online, their areas of concern and factors that would increase their faith of online transactions.It was found that 45 percent of window shoppers at e-commerce sites represent an audience that shopping sites make informed decisions, a huge opportunity not exploited by marketers.
Syed Shah Alam and ZaharahBakar (2008), have investigated in their study the relationships between young consumers' perceptions of the factors that influence their intention to buy through online.The analytical results are generally consistent with consumers' perceptions of the customer service, reliability and trust of online purchasing.Trust has received the most consistent support as factors that influence online buying.
Amar Cheema and Purushottam Papatla (2009), made an attempt to study the relative importance of online information versus offline information for internet purchase.The study found that the relative importance of online information is higher for utilitarian products and the relative importance of online information decreases with increasing consumer internet experience and consumers' trust.
Objectives
The major objectives of this study are: 1) To identify the nature and magnitude of risk perception in online shopping.
2) To understand people's preference on online shopping over traditional shopping in the presence of perceived risk.3) To examine the scope and future of online shopping in a world of casino economy.
Data and Methodology
To identify the types of risk perceived by online shoppers when considering an online purchase, data has been collected from both primary and secondary sources.Primary data for the study has been collected from 100 samples by using random sampling method.The study is purely based on sample surveys conducted among online shopping consumers in Kozhikode locality and sample consumers are those who buy commodities through online for more than one year.The sample survey was conducted by using a questionnaire.
The secondary data has been collected from online shopping surveys and reviews collected from various articles published in several journals, and E-journals and websites of Times of India, outlook, Deccan herald, CIOL and from various other references.For analysing the data and arrive at the logical conclusions, mathematical tools like percentage, ratios etc. and statistical tools like mean, standard deviation etc. have been used.To understand whether there is any difference in the perception about various types of risks, test of proportion (Z), Kruskal -Wallis (non-parametric test) H test and χ2 test of association has been used.
The Kruskal -Wallis Test
To assess whether there is any significant difference in risk perception of the consumers dealing with online shopping, the non-parametric alternative to one way ANOVA namely Kruskal -Wallis Test has been employed.
The Kruskal -Wallis hypothesis for differences in population is: H o : All k populations have the same distribution.H 1 : Not all k populations have the same distribution.
The Kruskal -Wallis test statistic is: Where 'n' is the total sample size from k populations n1, n2...nk and R j sum of the ranks from sample j (j=1, 2... k).We reject H o if sample H exceeds critical H, at chosen level of significance.
Chi-square Test (χ 2 )
To assess the association between socio-economic characteristics (income, sex, religion and occupation) of sample consumers and risk perception in online shopping, we use the Chi-square test for independence.
The hypothesis for independence is: The two classification variables are independent of each other.H 1 : The two classification variables are not independent.
The Chi-square test statistic for independence is: 2
Test of Proportion
To assess whether majority of the respondents have high risk perception, Z-test for proportions has been employed.The hypothesis for proportions is: The test statistics is:
Results and Discussion
This section outlines the major findings of sample survey.It examines the various socioeconomic characteristics of the sample consumers and also assesses the various types of risk perceptions involved in e-commerce and why people still depends on online purchasing in the presence of perceived risk.The study is based purely on sample survey conducted among online shopping consumers and the sample consumers are those who buy commodities through online for more than one year.The sample survey was conducted by using a questionnaire.
The Kruskal -Wallis test of consumers' preferences on various online websites and its corresponding ranks are given below: As the table suggests that there is difference in consumers rank and their preference over various online websites.Majority of the consumers prefer Flipkart and Amazon online websites.(p=0.000)It has been observed that the easy of accessibility is the major reason behind the consumers' preference over various online shopping websites (20%).It is followed by the availability of more variety commodities (16%), trustworthiness (14%), and timely delivery of the commodities (10%).The other reasons such as offers (06%), discount (06%), replacement guarantee (06%), customer friendly (06%), time saving (06%), attractive (04%), and high quality (04%) are least important in the consumer's preference on various online shopping websites.Books are the most purchased commodities in online shopping (36%) followed by footwear (18%), gadgets (16%) and clothing (16%) items respectively.The other commodities like fashion accessories (10%), automobiles (02%), etc. are purchased less quantity.Most of the people choose cash on delivery method for purchases.
Consumers Risk Perception in Online Shopping
We examine below the extent of risk perception by online shoppers and the various socioeconomic factors determining it.Table 2 clearly suggests that majority of the consumers perceive the existence of risk associated with online shopping.Test of Proportion suggests that majority of the consumers perceive high risk associated with online shopping (Z=6.51,P=0.000).
We also examine if there is any association between socio-economic characteristics of consumers such as income, sex, occupation and religion and their risk perception.To assess the association between income and risk perception, χ2 test of independence has been performed.Test results are given in Table 3. Where, SR-Security Risk, TR-Time Risk, FR-Financial Risk, PR-Product Risk, QR-Quality Risk D-Damage.
As the table suggests there is no association between risk perception and income (χ 2 =22.40,P=0.0978).
To assess the association between sex and risk perception, χ2 test of independence has been performed.Test results are given in Table 4.As the table suggests there is no association between risk perception and sex.(χ 2 =5.47, p =0.3614).
To assess the association between occupation and risk perception, χ2 test of independence has been performed.Test results are given in Table 5.As the table suggests there is no association between risk perception and occupation.(χ 2 =2.43, p =0.7873, p > 0.05).
To assess the association between religion and risk perception, χ2 test of independence has been performed.Test results are given in Table 6.It is noted that there are no differences in the reasons behind the online shopping even in the presence of perceived risk.
Conclusion
This paper investigates the risk perception in online shopping.Online shopping has a bright future in India for a variety of reasons.Easy accessibility, variety of choices, rate comparisons, the availability of feedback on the products etc. are some of the important factors that boosts up the scope of online shopping.However, the risks associated with online shopping are an important constraint which should be adequately addressed to boost the morale of online shoppers.Majority of the online shoppers opined that there are various types of risks such as product risk, time risk, financial risk etc. associated with online shopping.However, risk perception of the online shoppers is not significantly related to their income, sex, religion and education.Thus measures should be taken by online companies to mitigate various types of risks associated with online shopping.
/www.granthaalayah.com©International Journal of Research -GRANTHAALAYAH[237] Survey (Figure in brackets is p-value, * indicates significance at 5% level.)Where,A-Trust in exchange process B-To get wide variety choices C-Growing tendency of banking and network facilities D-Time saving E-Economic incentive F-Easy access G-Not yet cheated so far.
Table 1 :
Preference of Online Websites and Kruskal -Wallis (H) test
Table 2 :
Consumers Risk Perception in Online Shopping
Table 3 :
Income and Risk Perception
Table 4 :
Sex and Risk Perception
Table 5 :
Occupation and Risk Perception
Table 6 :
Religion and Risk Perception
Table 7 :
Reason behind Online Shopping in the Presence of Perceived Risk [Shafeeque et. | 2020-03-07T13:12:00.712Z | 2017-06-30T00:00:00.000 | {
"year": 2017,
"sha1": "374e24e22452e3bea9a5594c52b4a94aca2b172e",
"oa_license": "CCBY",
"oa_url": "https://www.granthaalayahpublication.org/journals/index.php/granthaalayah/article/download/IJRG17_A06_369/1904",
"oa_status": "HYBRID",
"pdf_src": "ScienceParsePlus",
"pdf_hash": "0edd4bbd53b11111a5395ae6f83441720023dac7",
"s2fieldsofstudy": [
"Business"
],
"extfieldsofstudy": [
"Psychology"
]
} |
4578657 | pes2o/s2orc | v3-fos-license | Preoperative intra-aortic balloon pump improves the clinical outcomes of off-pump coronary artery bypass grafting in left ventricular dysfunction patients
Severe left ventricular (LV) dysfunction patients undergoing off-pump coronary artery bypass grafting (OPCAB) are often associated with a higher mortality. The efficacy and safety of the preoperative prophylactic intra-aortic balloon pump (IABP) insertion is not well established. 416 consecutive patients with severe LV dysfunction (ejection fraction ≤35%) undergoing isolated OPCAB were enrolled in a retrospective observational study. 191 patients was enrolled in the IABP group; the remaining 225 patients was in control group. A total of 129 pairs of patients were propensity-score matched. No significant differences in demographic and preoperative risk factors were found between the two groups. The postoperative 30-day mortality occurred more frequently in the control group compared with the IABP group (8.5% vs. 1.6%, p = 0.02). There was a significant reduction of low cardiac output syndrome in the IABP group compared with the control group (14% vs. 6.2%, p = 0.04). Prolonged mechanical ventilation (≥48 h) occurred more frequently in the control group (34.9% vs. 20.9%, p = 0.02). IABP also decreased the postoperative length of stay. Preoperative IABP was associated with a lower 30-day mortality, suggesting that it is effective in patients with severe LV dysfunction undergoing OPCAB.
Scientific RepoRts | 6:27645 | DOI: 10.1038/srep27645 dysfunction that underwent OPCAB and performed a propensity-matched analysis to determine whether these patients benefited from preoperative prophylactic IABP insertion.
Material and Methods
Patient population. The present study was a historical, single-center, observational cohort study conducted at Beijing Anzhen Hospital, Capital Medical University. A total of 18,719 consecutive adult patients underwent isolated OPCAB in this hospital from January 2009 to December 2014. The left ventricular ejection fraction (LVEF) of all patients was calculated from an echocardiography assessment performed before the OPCAB operation. A total of 446 (2.4%) patients were identified as having severe LV dysfunction (LVEF ≤ 35%). Among those patients, 30 cases that presented with hemodynamic instability or severe femoral artery stenosis were excluded (Fig. 1). Finally, 416 patients were categorized according to preoperative prophylactic IABP insertion into either the IABP group or the control group. Data were retrospectively extracted from an institutional registry of OPCAB patients and ICU clinical database. This study received approval from the Beijing Anzhen Hospital Research Ethics Committee and all operations were performed in accordance with relevant guidelines and regulations. Because this was a retrospective observational study, the institutional review board of the Beijing Anzhen Hospital Ethics Committee waived informed consent. Data collection was subjected to the supervision of the Beijing Municipal Administration of Hospitals.
Surgical technique. The detailed information of the OPCAB procedure has been previously described 7 .
Six experienced cardiac surgeons (≥ 200 cases/year) performed all of the procedures. A cardiopulmonary bypass (CPB) circuit was placed on stand-by during the procedures. Conversion to CPB was considered if there was any evidence of hemodynamic instability concerns, such as ventricular arrhythmia, hypotension (systolic pressure ≤ 80 mmHg), and cardiac arrest during OPCAB procedures. IABP support. Patients in the IABP group received IABP support prior to the induction of anesthesia, followed by continuous IABP during the entire procedure, and postoperatively if needed. Each patient provided written consent for the IABP insertion. The control patients did not receive IABP support preoperatively. IABP treatment was initiated in the control group when cardiac index (CI) could not be maintained at a level ≥ 2.0 L/ min/m 2 , regardless of a high inotrope status.
Two types of IABP system were used at our center, namely, the Arrow System and the Datascope CS300 System. IABP balloon was selected according to the height of the patients. The IABP balloon was connected to a Datascope or Arrow pump. IABP insertion through the best femoral artery was possible in all cases according to the results of an ultrasound examination of bilateral lower extremities, and the corrected placement was assessed by a chest X-ray. The IABP variables were settled at a 1:1 balloon inflation synchronized with the electrocardiogram or aortic blood flow. Heparin was systematically used for anticoagulation. IABP support was terminated with hemodynamic stability after the operation.
Primary and secondary outcomes and definitions. The primary endpoint was postoperative 30-day mortality (death occurring within 30 days after surgery). The secondary endpoints were major postoperative complications, such as low cardiac output syndrome (LCOS), myocardial infarction, bleeding requiring reoperation, tracheotomy, renal failure requiring dialysis, stroke, and ICU stay, and postoperative length of stay (LOS) 21 . The IABP-related complications were also recorded and analyzed for safety purposes.
Statistical analysis. Continuous variables are shown as the mean and standard deviations or the median and interquartile ranges, and categorical variables are shown as frequencies and percentages. Continuous variables were compared using Student's t-test or Mann-Whitney U-test. Categorical variables were compared using the Chi-square test or a Fisher's exact test. Limb ischemia requiring surgical intervention 1 (0.8%) 0 (0%) 1.00 Severe bleeding 0 (0%) 0 (0%) 1.00 Balloon failure or leak 0 (0%) 0 (0%) 1.00 Propensity score matching was undertaken by estimating the probability of receiving preoperative IABP support (that is, the propensity score) using a logistic regression model including the potentially confounding covariates shown in Table 1. Each IABP group patient was matched with a non-IABP-supported patient based on the propensity score by the method of nearest neighbor matching within a caliper (caliper = 0.25 × SD[logitP] 22 ) using the psmatch2 command in Stata 23 . After matching, continuous variables that followed a normal distribution were compared using the paired sample t-test; otherwise, the Wilcoxon's matched-pairs signed-rank test was used. The McNemar's test was used to evaluate the comparability of categorical factors after matching. The in-hospital survival rates after surgery of matched samples were shown as Kaplan-Meier survival cures and the effect of IABP vs. non-IABP was presented as a hazard ratio (HR) with associated 95% confidence intervals (CIs) from the Cox regression model.
Statistical analyses were conducted using Stata software version 11 (Stata Corp, College Station, Texas, USA). Two-sided testing was used with a p-value significance level of less than 0.05.
Patient baseline characteristics.
A total of 416 patients were included in this study (Fig. 1), with 191 patients in the IABP group and 225 patients in the control group. There were significant differences documented in the demographics and comorbidities of the patients (Table 1). After the propensity-score matching, 129 matched pairs were obtained (Fig. 1). No differences in demographics or preoperative risk factors were found between the two groups (Table 1). Figure 2 shows a Love plot for the absolute differences in the baseline covariates before and after matching; a jitter plot for propensity-score distribution is also presented.
OPCAB procedural characteristics are detailed in Table 2. Importantly, the mean number of anastomoses were comparable between the two groups (p = 0.22). However, there were few patients converted to on-pump CABG in the IABP group (0 vs. 6 [4.6%], p = 0.04).
Postoperative 30-day mortality. Clinical outcomes are presented in Table 2. The 30-day mortality was 1.6% in IABP group compared to 8.5% in the control group (p = 0.02). The Kaplan-Meier survival curves of the two groups before and after matching are shown in Fig. 3. The preoperative prophylactic IABP insertion was an independent predictor of survival after adjusting for the propensity score using the Cox regression model (HR 0.17, CI 0.04-0.79, p = 0.02).
Postoperative complications. The postoperative complications are summarized in Table 2. There was a significant reduction in postoperative LCOS in the IABP group (14 vs. 6.2%, p = 0.039). There were no significant differences in the required transfusion of red blood cells (2 vs. 2 units, p = 0.07) and fresh frozen plasma (600 vs. 600 ml, p = 0.37). The postoperative myocardial infarction, reoperation for bleeding, tracheotomy, hemodialysis, and neurologic events were comparable between the two groups ( Table 2). There were no statistically significant differences in the duration of mechanical ventilation (47.1 ± 49.1 vs. 37.9 ± 40.9 h, p = 0.16). However, the required prolonged mechanical ventilation occurred more frequently in the control group (34.9 vs. 20.9%, p = 0.02). ICU stay was comparable between the two groups. The mean postoperative LOS was shorter in the IABP group (interquartile range) of 9 (7-12) vs. 10 (7-16) days (p = 0.03) ( Table 2).
No significant interactions were observed between preoperative IABP insertion and any of the 12 sub-groups with respect to the in-hospital mortality, as shown in the hazard-ratio plots in Fig. 4.
IABP-related complications.
Post-cardiotomy IABP was used in 21 patients in the control group, including 3 who required weaning from CPB, and 18 who occurred LCOS. The mean duration of IABP support was shorter in the IABP group (135.9 ± 80.9 vs. 76.4 ± 33.4 h, p < 0.001). There were no cases of IABP-related mortality. No severe bleeding at the IABP insertion site, or balloon failure in any of the patients. Lower limb ischemia requiring surgical intervention was observed in 1 patient (0.8%) in the IABP group (Table 2).
Discussion
This large regional study of propensity-matched patients indicated that preoperative prophylactic IABP insertion was associated with reduced postoperative 30-day mortality in severe LV dysfunction patients who underwent OPCAB. Furthermore, prophylactic IABP insertion resulted in significant reduction of postoperative LCOS incidence and shorter postoperative hospital stay.
From the perspective of pathophysiology, the positive effect of IABP insertion is believed to increase coronary blood flow while simultaneously decreasing myocardial oxygen demand. Consequently, preoperative prophylactic IABP assistance provides better hemodynamic stability in crucial times of higher oxygen demand when the heart is displaced in OPCAB procedures 24,25 .
Although there are certain advantages in theory, the results have been controversially debated in clinical practice. Some studies have shown a positive effect of preoperative prophylactic IABP insertion in improving the outcomes of high-risk patients 13,14 . The strongest evidence supporting preoperative IABP insertion for high-risk patients undergoing CABG comes from published meta-analysis studies 12,14,17 . However, many contemporary studies have challenged the effectiveness of preoperative IABP in high-risk patients undergoing CABG 15,16 . Worse outcomes were shown in a recent propensity-score matching study 16 , in which the preoperative IABP insertion in patients undergoing CABG after acute myocardial infarction was associated with increased in-hospital morbidity, greater transfusion requirements, and longer postoperative ICU stay.
The results of previous studies have been controversial for several possible reasons. First, there is no standard definition of a high-risk patient. Various conditions, including severe LV dysfunction, left main disease, diffuse coronary disease, and reoperation, have been suggested for preoperative prophylactic IABP insertion 14 . Second, the criteria for prophylactic IABP insertion have not been well defined 26 . A distinction is lacking between therapeutic use for patients with preoperative cardiogenic shock or hemodynamic instability and prophylactic use for patients with preoperative hemodynamic stability. In many previous studies including patients with hemodynamic instability, the indication has been more likely for therapeutic, rather than prophylactic insertion. Third, the results were also possibly affected by the severity of the patients who were selected to receive preoperative IABP support. Finally, most of the procedures were conducted in CABG patients. IABP insertion before surgery was suspended during CPB. The benefit from IABP support was relatively low.
Our study investigated for the clinical effects of preoperative prophylactic IABP insertion in patients with severe LV dysfunction that underwent selective OPCAB. The IABP group of patients received preoperative prophylactic IABP support to increase the safety of OPCAB procedures. The patients who received preoperative IABP support for hemodynamic instability, cardiogenic shock, and emergency operations were excluded. Comparatively, IABP was still working during OPCAB procedures in our study. These patients were more likely to benefit from preoperative IABP insertion. Therefore, preoperative prophylactic IABP insertion was associated with a lower rate of conversion to on-pump CABG, which has been associated with increased in-hospital mortality 9 .
The preoperative prophylactic IABP insertion in our study was not associated with an increased rate of IABP-relationship complications (that is, limb ischemia requiring surgical intervention, severe bleeding at the IABP insertion site, and embolism). The incidence rate of IABP-related complications was low, similar with a previous study 27 . Therefore, IABP insertion is safe in those high-risk patients.
Limitations.
Our study had several limitations. First, this study was subject to the limitations inherent in any retrospective, observational study from a single center. The nonrandomized design might have affected our results, owing to unmeasured confounds, procedural bias, or detection bias. Despite the benefits of propensity matching, it is possible that there are additional confounds that were not accounted for in our adjustment algorithm. The whole study depends on the accuracy of propensity score matching and many pitfalls may be hidden. Second, it is generally believed that the experience of the surgeon can influence the results of OPCAB. Six experienced cardiac surgeons performed the OPCAB procedures in this study. However, our hospital is an international center for cardiovascular clinical and research. All of the surgeons followed the same standard OPCAB procedure of our hospital. Furthermore, this study was also limited to patients undergoing isolated OPCAB. Patients requiring concomitant cardiac surgical procedures, and/or those with mechanical complications of acute myocardial infarction, such as acute mitral regurgitation or myocardial rupture, were excluded from this study. Therefore, the results of this study cannot be extended to these extreme high-risk patient populations. Finally, the present study was conducted in the setting of a high-volume tertiary cardiovascular center in a developing country; therefore, the results might not be generalizable to other centers in different situations. | 2018-04-03T03:25:04.927Z | 2016-06-09T00:00:00.000 | {
"year": 2016,
"sha1": "014fdd252a84ed8a15e019d93897e711181db22f",
"oa_license": "CCBY",
"oa_url": "https://www.nature.com/articles/srep27645.pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "014fdd252a84ed8a15e019d93897e711181db22f",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
258330575 | pes2o/s2orc | v3-fos-license | Differences in Trapezius Muscle H-Reflex between Asymptomatic Subjects and Symptomatic Shoulder Pain Subjects
In chronic shoulder pain, adaptations in the nervous system such as in motoneuron excitability, could contribute to impairments in scapular muscles, perpetuation and recurrence of pain and reduced improvements during rehabilitation. The present cross-sectional study aims to compare trapezius neural excitability between symptomatic and asymptomatic subjects. In 12 participants with chronic shoulder pain (symptomatic group) and 12 without shoulder pain (asymptomatic group), the H reflex was evoked in all trapezius muscle parts, through C3/4 nerve stimulation, and the M-wave through accessory nerve stimulation. The current intensity to evoke the maximum H reflex, the latency and the maximum peak-to-peak amplitude of both the H reflex and M-wave, as well as the ratio between these two variables, were calculated. The percentage of responses was considered. Overall, M-waves were elicited in most participants, while the H reflex was elicited only in 58–75% or in 42–58% of the asymptomatic and symptomatic participants, respectively. A comparison between groups revealed that the symptomatic group presented a smaller maximum H reflex as a percentage of M-wave from upper trapezius and longer maximal H reflex latency from the lower trapezius (p < 0.05). Subjects with chronic shoulder pain present changes in trapezius H reflex parameters, highlighting the need to consider trapezius neuromuscular control in these individuals’ rehabilitation.
Introduction
The nervous system (NS) acts through the relationship between excitation and inhibition [1], which could be altered in conditions such as pain [2]. In acute musculoskeletal pain conditions, the presence in the muscle of substances such as bradykinin, substance P or serotonin [3] leads to reduced nociceptor thresholds [4] as a protective response against injury [4,5]. However, with time, the repeated discharge of the peripheral nociceptors [6] and an impaired balance between descending inhibitory mechanisms and the pain facilitator pathways can lead to changes in the central modulation of sensory input from the muscle [2,3]. Thus, increased central excitability [5] of dorsal horn neurons [3] as well as enhanced response to inputs (hypersensitivity), known as central sensitization [2], may be experienced by some subjects with chronic pain, such as shoulder pain [5]. At the same time, an adaptation of the motor strategy to protect the painful area [7,8] and maintain the motor output [8], could lead to changes at the muscular level [3][4][5]7,8]. Both painful and non-affected muscles seem to change their mechanical behavior and activity [8,9]. These adaptations to pain are heterogeneous between subjects [10], vary with tasks and muscles [7,8] and their protective action could lead to consequences such as, for example, increased load to non-painful structures and decreased movement and variability [7].
Shoulder pain conditions stand out for their high prevalence in the general population, their prolonged recovery, and associated recurrences [11,12]. Some muscular changes [7,13] have been related to CNS adaptation mechanisms [14]. The adaptations in muscular mechanical behavior [8], together with possible changes in motoneuron or cortical excitability [7] and sensitization [2][3][4][5], might explain the maintained shoulder pain condition [2] and the high prevalence of pain recurrences in shoulder non-degenerative conditions [15] and could be the reason why rehabilitation is not always effective [10] or quick [2]. Beyond the alterations in the rotator cuff and other scapular musculature, previous studies have reported changes in trapezius muscle activity in subjects with maintained or recurrent shoulder pain. These changes include inhibition and/or changed timing of activation [16] of the lower (LT) and middle (MT) portions of trapezius, together with inconsistent findings regarding the upper portion (UT) [17,18]. Thus, for UT, some studies reported excessive activation [19][20][21][22][23] while others reported decreased UT activation [17,18,24]. These opposite findings regarding UT could possibly be related to two conditions. Some authors believe that increased UT activity could be related to the pain and the trigger points usually found over it [25], or to a compensatory strategy to elevate the arm during shoulder pain through increased clavicular elevation and, consequently, scapular elevation [21]. By contrast, other authors have associated impaired UT activity with the scapular dyskinesis presentation, namely scapular depression and upward rotation [18]. Nevertheless, it is important to mention that although this tonic scapular stabilizer [24] is considered by some authors to be responsible for scapular upward rotation, together with serratus anterior [18,26], the main role of UT is in clavicular motions (rotation, retraction and elevation). Its influence on the scapula occurs as a consequence, contributing in this way to scapula elevation and only for approximately 25% of upward rotation [27,28]. In addition to the mentioned adaptations, a previous study [29] also reported changes in a long-latency trapezius muscle reflex when comparing UT and LT of subjects with shoulder instability with healthy subjects. More specifically, in the case of shoulder instability, LT showed delayed or absent reflexes evoked from the afferents of the forearm and hand, which could reflect an impaired LT contribution to scapular stability during the use of the arm and hand [29].
Considering these facts, and remembering that muscular activity is exposed to the descending and reflex controls [30] that contribute to the scapular stability and movement [29], and that could be altered with shoulder pain, studies regarding muscular modulation in other shoulder pain conditions may add helpful knowledge [31] to improve the rehabilitation process [29]. The H reflex neural circuit occurs via a mainly monosynaptic connection in the spinal cord, whereby electrical stimulation of the group Ia afferents leads to an excitatory volley onto the motoneurons, evoking motoneuron depolarization/excitation and a consequent activation of the muscle fibers [32,33]. Thus, it can evaluate the modulation of monosynaptic reflex activity in the spinal cord [31,34,35] by assessing motoneuron excitability [31,34,35], independently of the muscle mechanoreceptors [33]. H reflex assessment could then be a useful tool to identify adaptations in the function of spinal structures following pain conditions or also therapeutic interventions [33]. Previous studies involving painful conditions in other body regions have found alterations in H reflexes. Painful areas in the gastrocnemius were associated with a decreased threshold and increased amplitude of H reflexes [36] but reductions in H reflex amplitude [9,37], as well as increases in threshold [37,38] and latency [37], occurred in the vastus medialis in patellofemoral pain [9] and soleus [37,38] in chronic low back pain and lumbosacral radiculopathy [37,38].
Unusually, the trapezius innervation is split between the C3/4 nerve (sensory information) and the accessory nerve (motor component) [13,31,39,40]. Thus, the amplitude of the maximum H reflex (evoked through stimulation of the C3/4 nerve) is little influenced by the M-wave [13], as the absence of motor axon direct activation [39] avoids collision between action potentials of the evoked reflex responses and the antidromic impulses evoked from the motor axon stimulation [13,31]. Despite the mentioned facts and the changes related to shoulder pain, the H reflex of the trapezius muscle [13,30,31,39,41] has not been studied in people with shoulder pain.
The present study, by investigating a neurophysiologic variable possibly related to shoulder pain adaptations, aimed to assess the possible differences in trapezius neural excitability between symptomatic and asymptomatic subjects. The reduction in H reflexes in other painful conditions [9,[36][37][38] supports the hypothesis of a decreased H reflex response, with increased threshold and latency in symptomatic subjects. The results of the present study could be helpful for the rehabilitation field by improving knowledge regarding motor control changes in pain conditions, but could also be useful for technological development regarding H reflex assessment.
Subjects
Twenty-four volunteers (ten male and fourteen female) took part in this study. Participants, recruited from a School of Health via personal contact or email, were included in the symptomatic or asymptomatic groups, according to the eligibility criteria.
While the asymptomatic group included healthy subjects that had had no shoulder pain events in the last 2 years, the symptomatic group included subjects that presented continuous or intermittent (2 or more episodes in the last 3 months) chronic shoulder pain. Chronic shoulder pain fitted the criteria of: (a) pain in the upper arm, specifically in shoulder, deltoid and/or scapular areas; (b) non-specific or associated with a diagnosis (except if mentioned in the exclusion criteria); and (c) lasting more than 3 months. Subjects with a history of shoulder fracture, dislocation, tears, infection or neoplasm; shoulder surgery; cervical and/or thoracic pathologies or pain associated with active movements of these regions; neurological disease; and/or body mass index outside the range 18.5-30 kg/m 2 [42][43][44] were excluded.
Subjects of the two groups were matched regarding gender, age, and dominant upper limb. Before participation, all subjects read and signed the informed consent form, and the study was approved by the local ethical committee (CE0071A).
Instrumentation
To characterize the participants, two self-reported questionnaires were applied to the symptomatic group and the scapular dyskinesis classification test was applied to all the subjects. The numeric rating scale (NRS), which, according to a previous study [45] has an intra-rater intraclass correlation coefficient (ICC) of 0.84, was used to quantify the pain intensity from 0 ("no pain") to 10 ("unbearable pain") [46]. The Portuguese version [47,48] of Shoulder Pain and Disability Index (SPADI), with an internal consistency reliability of α = 0.75 (for pain) and 0.84 (for functional activity), was used to characterize the shoulder function from 0 ("no pain/no difficulty") to 100 ("worst pain imaginable/so difficult required help") [49]. Scapular dyskinesis test: scapular position and motion were evaluated, in standing position, at rest and during shoulder abduction and lowering movements [50]. Based on visual observation, a trained physiotherapist classified each participant's scapular presentation with a type of scapular dyskinesis (intratester reliability of k = 0.49-0.59 [51]) as: (1) type I-when the scapula's inferior angle was posteriorly displaced/prominent [51][52][53][54]; (2) type II-when the scapula's medial border was prominent [51,52,54]; (3) type III-when there was excessive elevation of the scapula's superior border and/or excessive/insufficient scapular upward rotation [51,52,54]; and (4) type IV-when there was symmetry between the scapula of the symptomatic side and the contralateral one [53,54].
To record surface electromyographic (EMG) activity and register the H reflex, the Biopac Systems Inc.-MP 160 Workstation™ (Biopac System Inc., Goleta, CA, USA) was used with a sampling frequency of 2000 Hz, input impedance of 100 MΩ, a common mode rejection ratio (CMRR) of 95 dB and a gain of 1000 and an analog-to-digital converter of 12 bits. Data were collected on UT, MT and LT using steel surface electrodes (TSD150, Biopac Systems Inc., Goleta, CA, USA), bipolar configuration, with an 11.4 mm contact area and an inter-electrode distance of 20 mm. To evoke the H reflex and M-wave through nerve stimulation, the constant current stimulator STMISOLA (Biopac System Inc., Santa Barbara, CA, USA) was used. The locations to place the stimulation electrodes were found through a motor point pen (Compex Iberica, Barcelona, Spain), which was later replaced by self-adhesive Ag/AgCl surface electrodes (Lessa, Barcelona, Spain). Acqknowledge ® software (version 3.9, from Biopac Systems Inc., Goleta, CA, USA) was used to create and apply custom-written scripts for stimulator control, EMG signal acquisition, filtering and analysis. An auxiliary tool developed in Python (Python Software Foundation, Beaverton, OR, USA) was used to identify and observe the H reflex and M-wave, as well as to extract the parameters of interest.
Positioning of Electrodes for H Reflex and M-Wave Recordings through Electromyography
To assess the H reflex and M-wave responses of the trapezius muscle, steel surface electrodes were positioned to record EMG from three parts of the trapezius of the painful (or more painful) shoulder. After the skin had been shaved, abraded, and cleaned, ethanol electrodes were placed: (a) UT: over a 2 cm laterally to the midpoint of the line between the spinous process of C7 and the posterior tip of the acromion [55,56]; (b) MT: midway on a horizontal line between the root of the spine of the scapula and the T3 spinous process [19,55]; and (c) LT: obliquely, at 2/3 of the distance along the line from the root of the spine of the scapula to the T8 spinous process [56] (Figure 1). The quality of the raw signals was checked [55,57] before the data were recorded.
Electrical Stimulation of C3/4 and Accessory Nerves
To allow the stimulation of nerves responsible for the trapezius sensory (C3/4 nerve) and motor (accessory nerve) innervation, the process started with the search for the optimal cathode locations. For this, adhesive electrodes for the anodes were placed just below the middle point of the clavicle [13,30,31,35,39,41,58], or over the mastoid process [13,31,39,58], for the C3/4 nerve or the accessory nerve, respectively. Then, single electrical stimuli (pulse width of 1 ms and intensity of 10 mA) were applied while moving a hand-held motor point pen around the appropriate areas of the neck: (a) around the anterior surface of UT above the clavicle, for C3/4 nerve stimulation [13,30,35,39,41,58]; and (b) behind the sternomastoid muscle and between the level of the jaw and the upper border of the trapezius, for the accessory nerve [13,39,58] (Figure 1). At the locations where the largest responses (H reflex (when C3/4 was stimulated), or M-wave (when accessory nerve was stimulated)) were found, the motor point pen was replaced by self-adhesive electrodes.
During the search for cathode location and also during the process to evoke both H reflex and M-wave, the participants were asked to maintain a sitting position [13,30,31,35,39,41,58,59] with 2/3 of the thigh supported, the knee and hip at 90 • of flexion, the feet on the floor [13,31,39,58] and with the head, pelvis and trunk in a neutral position. Both hands were supported on a height-adjustable table, with the shoulder at 90 • abduction in the scapular plane, with the forearm in neutral position and the elbow extended. As trapezius H reflex is difficult to elicit at rest [13,31], throughout the H reflex stimulation the same position was adopted but with no support provided to maintain the shoulder position. Therefore, participants maintained the shoulder position through isometric contraction [30,39].
Electrical Stimulation of C3/4 and Accessory Nerves
To allow the stimulation of nerves responsible for the trapezius sensory (C3/4 nerve) and motor (accessory nerve) innervation, the process started with the search for the optimal cathode locations. For this, adhesive electrodes for the anodes were placed just below the middle point of the clavicle [13,30,31,35,39,41,58], or over the mastoid process [13,31,39,58], for the C3/4 nerve or the accessory nerve, respectively. Then, single electrical stimuli (pulse width of 1 ms and intensity of 10 mA) were applied while moving a handheld motor point pen around the appropriate areas of the neck: (a) around the anterior surface of UT above the clavicle, for C3/4 nerve stimulation [13,30,35,39,41,58]; and (b) behind the sternomastoid muscle and between the level of the jaw and the upper border of the trapezius, for the accessory nerve [13,39,58] (Figure 1). At the locations where the Then, after placing the stimulation and recording electrodes, percutaneous electrical stimulation was delivered separately to the accessory nerve and the C3/4 nerve ( Figure 1) by a constant current stimulator controlled with custom-written scripts. The M-wave recruitment curve was evoked by accessory nerve stimulation of 1 ms rectangular pulses [13,35,39,58], 10 s apart, with gradually increased (0.5 mA steps) intensity [13,31,39]. Once the M-wave seemed to be quite stable [34], only one electrical pulse was delivered at each intensity. Recordings were made with the subject relaxed [13,31,39,58] and until there was no further increase in M-wave amplitude in each of the three-trapezius portions in 3 consecutive stimulations, despite an increase in stimulus intensity [13,31,39]. The maximal M-wave (Mmax) was used to normalize the amplitude of the H reflex [13].
C3/4 stimulation was delivered to collect H reflex recruitment curves. Given reflex variability [59], sets of 10 electrical monophasic positive rectangular pulses [30,31,39,41] of 1 ms [13,35,39,58] were delivered with a 10 s interval [60], for each intensity. The first set started with the lowest intensity capable of evoking an H reflex, which was identified by a sequence of single impulses of increasing intensity and at 5 s apart. Then, the intensity was gradually increased by 0.2 mA steps until no further increase in H reflex amplitude, of any trapezius portion, was detected, despite increasing stimulus intensity [13,31,39].
Maximal Voluntary Isometric Contractions (MVICs)
At the end of the protocol, the maximal voluntary isometric contractions (MVICs) of each trapezius portion were assessed to normalize the EMG activity level that was produced during the isometric submaximal contraction needed to evoke the H-reflex [UT: 23.0 ± 14.1% and 24.3 ± 14.3% of MVIC; MT: 8.9 ± 12.1% and 11.6 ± 10.7% of MVIC; LT: 13.9 ± 8.4% and 19.0 ± 10.1% of MVIC, for symptomatic and asymptomatic group, respectively]. Each MVIC test was based on the recommendations of Cools et al. [55] and Ekstrom, Soderberg and Donatelli [61]: (a) for UT: shoulder abduction at 90 • with the neck in same-side inclination, opposite-side rotation and extension against manual resistance applied at the head and above the elbow, with the subject seated [61]; (b) for MT: shoulder horizontal abduction and external rotation against manual resistance applied above the elbow, with the subject prone [61]; and (c) for LT: shoulder abduction (diagonally at 135 • ) against a manual resistance, applied with the subject prone [55,61]. Three trials of 5 s MVIC were performed against manual resistance, with 1 min of rest [39] between repetitions to minimize fatigue [62,63]. The 3 central seconds of each contraction were considered, and the mean EMG was calculated.
Data Processing
The EMG signals were digitally filtered with a second-order band-pass Butterworth filter with cut-off frequencies ranging between 20 and 1000 Hz. Since for the H reflex 10 stimuli were delivered at each intensity, the 10 responses were overlaid as represented in Figure 2, and averaged for UT, MT, and LT. The latencies of H reflex and M-wave were measured from the stimulus artifact to the onset of the potential. The peak-to-peak amplitudes of H reflex and M-wave were calculated through the mathematical difference between the maximum and minimum values of the compound muscle action potential, for each trapezius portion (Figures 3 and 4). While the maximum value of H reflex peak-topeak amplitude (Hmax) presents information about the maximum number of motor units reflexively activated [13,38], the maximum peak-to-peak amplitude of M wave (Mmax) represents the maximum motor response [38]. UT, MT, and LT Hmax/Mmax amplitude ratios were calculated for each subject.
Sensors 2023, 23, x FOR PEER REVIEW 7 of 18 measured from the stimulus artifact to the onset of the potential. The peak-to-peak amplitudes of H reflex and M-wave were calculated through the mathematical difference between the maximum and minimum values of the compound muscle action potential, for each trapezius portion (Figures 3 and 4). While the maximum value of H reflex peakto-peak amplitude (Hmax) presents information about the maximum number of motor units reflexively activated [13,38], the maximum peak-to-peak amplitude of M wave (Mmax) represents the maximum motor response [38]. UT, MT, and LT Hmax/Mmax amplitude ratios were calculated for each subject.
Statistical Analysis
The sample size was calculated with G*Power software (Kiel University, Germany) using an effect size (d = 1.218) determined from the difference in latency of a reflex in the lower trapezius found previously between healthy subjects and subjects presenting non-
Results
Each group contained five male and seven female participants with the demographic characteristics described in Table 1. Groups were well matched at baseline, except for scapular positioning, where the asymptomatic group had only 33% of the subjects presenting some type of scapular dyskinesis while the symptomatic one presented it for 100% of the subjects. Symptomatic subjects also presented a moderate level [46] of shoulder pain (4.42 ± 1.38) and a lower level [49] of functional shoulder disability (17.81 ± 9.02).
Statistical Analysis
The sample size was calculated with G*Power software (Kiel University, Germany) using an effect size (d = 1.218) determined from the difference in latency of a reflex in the lower trapezius found previously between healthy subjects and subjects presenting non-traumatic shoulder instability [29]. It was found that a sample of 10 individuals per group was sufficient to detect differences with a power of 0.8 and an alpha of 0.05. To account for a possible inability to measure the H reflex [13], 12 subjects were included in each group.
Statistical analysis was carried out using the Statistical Package for Social Science (SPSS) version 27 (IBM, Inc., Chicago, IL, USA).
The normality of the data distribution was tested through the Shapiro-Wilk test and histogram analysis. When this assumption was not verified, if possible, a logarithmic or an inverse tangent transformation [64] was applied. Sample homogeneity was tested through the t-student test, Mann-Whitney, or chi-square/Monte Carlo. In all tests, p < 0.05 was For the comparison of the study variables (the current needed to evoke Hmax (current at Hmax), Hmax, Mmax, the percentage of the ratio between Hmax/Mmax (% Hmax/Mmax) and Mmax and Hmax latencies) between groups, the unpaired-sample t-test or the Mann-Whitney test were used. The values of effect size (Cohen's d) are also presented, except for the variables where a non-parametric test was used. Values higher than 0.8 were considered to represent a large effect size, those of approximately 0.5 a moderate effect size, and those less than 0.2 a small effect size [65].
Results
Each group contained five male and seven female participants with the demographic characteristics described in Table 1. Groups were well matched at baseline, except for scapular positioning, where the asymptomatic group had only 33% of the subjects presenting some type of scapular dyskinesis while the symptomatic one presented it for 100% of the subjects. Symptomatic subjects also presented a moderate level [46] of shoulder pain (4.42 ± 1.38) and a lower level [49] of functional shoulder disability (17.81 ± 9.02). Footnote-Missing data occurred for H reflexes (indicated by response frequency) when it was not possible to identify the H reflex after application of the adhesive electrodes or when, instead of an H reflex, an M-wave response (identified through its latency) occurred during the C3/4 nerve stimulation, possibly due to the stimulus spreading or the anatomical variation of the C3/4 nerve [58]. Significant results were signed with an *.
Comparison of Trapezius H Reflex and M-Wave between Symptomatic and Asymptomatic Subjects
Statistically significant differences between groups (p = 0.020 and 0.012) were observed for the UT Hmax/Mmax and LT Hmax latency, respectively (Table 2 and Figures 5 and 6). The symptomatic group presented a decreased UT Hmax/Mmax (6.20 ± 5.05%), of less than half of the value for the asymptomatic group (18.58 ± 11.61%), and an increased LT Hmax latency (14.10 ± 3.15 ms), compared with the asymptomatic one (10.71 ± 1.04 ms). In Figure 5, note the low dispersion of UT Hmax/Mmax values in the symptomatic group, while in Figure 6, note the opposite for the LT Hmax latency. No statistically significant differences (p > 0.05) were found between the symptomatic and asymptomatic subjects in % Hmax/Mmax from MT and LT, in Hmax latency from UT and MT, nor in current at Hmax, Hmax, Mmax, and Mmax latency from all trapezius portions (Table 2 and Figures 5 and 6). Moderate effect sizes were observed for the current needed to evoke UT Hmax, the UT Hmax amplitude and UT and MT Hmax latency in the comparisons between groups (Table 2).
Discussion
Chronic pain has been associated with mechanical and neurophysiological adaptations [66][67][68][69][70][71]. Adaptations in the nervous system [69] have been shown to occur at a supraspinal level [7,29] but also at a spinal level, since muscle pain, mediated by muscle afferents, could alter the muscle spindle discharge and its effect on the motoneuron pool [39]. Therefore, in conditions of pain [33], the H reflex in superficial muscles that have muscle spindles [60] has already been used to give information regarding spinal neurophysiological changes in excitability [5,9,72], pain modulation [31], and integration of peripheral information [9]. The association between the trapezius muscle and musculoskeletal conditions, such as shoulder pain [31], motivated the present study, which compared trapezius spinal excitability between asymptomatic subjects and symptomatic subjects with chronic shoulder pain [58].
Symptomatic subjects reported a moderate shoulder pain intensity (~4.5 out of 10) but a relatively low impact on shoulder function (~18 out of 100), despite all showing scapular dyskinesis. They had significantly lower values of UT Hmax/Mmax than in the asymptomatic group, indicating decreased excitability of the motoneuron pool [35]. Such reduced excitability may suggest the existence of changes in muscle activity during chronic shoulder pain [7], such as a possible decrease in muscular activation [9] from scapula feedback mechanisms [29]. This seems to agree with the findings of previous studies [9,73]. Previous studies have hypothesized decreased activity of the agonist muscles' motoneurons in muscle pain conditions [73] and that the size of the reflex response is adapted according to the ongoing muscular activity [74]. However, in the present study, the UT ongoing EMG during the recording of the H reflex was 23% and 24% of the maximum, whereas Hmax/Mmax was 6.20% and 18.58% for the symptomatic
Discussion
Chronic pain has been associated with mechanical and neurophysiological adaptations [66][67][68][69][70][71]. Adaptations in the nervous system [69] have been shown to occur at a supraspinal level [7,29] but also at a spinal level, since muscle pain, mediated by muscle afferents, could alter the muscle spindle discharge and its effect on the motoneuron pool [39]. Therefore, in conditions of pain [33], the H reflex in superficial muscles that have muscle spindles [60] has already been used to give information regarding spinal neurophysiological changes in excitability [5,9,72], pain modulation [31], and integration of peripheral information [9]. The association between the trapezius muscle and musculoskeletal conditions, such as shoulder pain [31], motivated the present study, which compared trapezius spinal excitability between asymptomatic subjects and symptomatic subjects with chronic shoulder pain [58].
Symptomatic subjects reported a moderate shoulder pain intensity (~4.5 out of 10) but a relatively low impact on shoulder function (~18 out of 100), despite all showing scapular dyskinesis. They had significantly lower values of UT Hmax/Mmax than in the asymptomatic group, indicating decreased excitability of the motoneuron pool [35]. Such reduced excitability may suggest the existence of changes in muscle activity during chronic shoulder pain [7], such as a possible decrease in muscular activation [9] from scapula feedback mechanisms [29]. This seems to agree with the findings of previous studies [9,73]. Previous studies have hypothesized decreased activity of the agonist muscles' motoneurons in muscle pain conditions [73] and that the size of the reflex response is adapted according to the ongoing muscular activity [74]. However, in the present study, the UT ongoing EMG during the recording of the H reflex was 23% and 24% of the maximum, whereas Hmax/Mmax was 6.20% and 18.58% for the symptomatic and asymptomatic groups, respectively. Thus, the impairment in the UT reflex response was not the result of differences in ongoing muscle activity, but rather suggests alterations in the reflex loop. Oliveira Silva et al. [9] have also reported similar results of a decrease in percent Hmax/Mmax from the vastus medialis in subjects with patellofemoral pain, when compared with asymptomatic subjects. Such results confirm the presence of neurophysiological changes in both pain conditions, beyond the structural and biomechanics adaptations [9]. Impairment of the transmission between the Ia afferents and the motoneurons [9,38] could occur at several levels: (a) at the site of stimulation where the afferent volley might be reduced [29,38], (b) presynaptically, where presynaptic inhibition or homosynaptic post-activation depression can alter the neurotransmitter released given the same Ia afferent volley [29,38], and (c) at the motoneurons, through changes in their excitability that makes the Ia input less effective at producing output [7,29,38,75].
In the current study, the symptomatic group also presented increased LT Hmax latency, suggesting a delayed recruitment of the motoneurons through the C3/4 nerve stimulation. One previous study that used stimulation of the ulnar nerve to compare long-latency reflex responses from UT and LT of subjects with shoulder instability and healthy subjects, also reported similar differences between groups for LT latency [29]. Such findings may represent a lower efficiency of the feedback mechanisms acting in scapular stability [29].
Given changes in H reflex parameters in both UT and LT, it may be important to consider whether this could indicate a generalized decrease in trapezius feedback mechanisms [29] and, possibly, a reduction in muscle activation [9]. Although there were no other significant differences between groups, moderate effect sizes were also observed for lengthening of the UT and MT Hmax latencies, (d = −0.918 and −0.505, respectively), and Hmax/Mmax values for MT and LT tended to be lower for the symptomatic group. Together, these non-significant results point in the direction of a generalized effect. Moreover, the lack of results could be due to the reduced sample sizes. The number of participants in whom it was possible to evoke the H reflex was not the same for each group and each part of the trapezius. Thus, it may have limited the identification of significant results. However, it should be noted that, regardless of wider effects, the isolated statistically significant deficits for UT Hmax/Mmax and LT Hmax latency could represent an impact on the function of the shoulder complex, given the possible imbalance among trapezius portions [29]. Reflex motoneuron activation seems to have an impact on scapula stability and, consequently, on upper-limb functionality. Thus, such an imbalance regarding the muscular control of the portions of trapezius could reflect an impairment of the feedback mechanisms that act during scapula motion, as well as differences in the coupling of forces responsible for scapular movement and stability. In consequence, impairments in scapula-humeral rhythm could be found [29]. In the present study, the results regarding scapular positioning assessed at rest and during motion seem to reinforce the hypothesis that impairments at UT Hmax/Mmax and LT Hmax latency could be associated with changes in scapular stability and motion, as all the participants of the symptomatic group had altered scapular positioning, compared to only 34% of the participants of the asymptomatic group.
In the current study, a moderate effect size (d = 0.586) was also observed for the current intensity needed to evoke the UT Hmax. Lower values were observed in the symptomatic group, perhaps because the symptomatic group did not have the ability to increase the H reflex to values equal to the asymptomatic one, noting the moderate effect size for the UT Hmax (d = 0.957). This may be in line with the results of a previous study [38] that evaluated the H reflex in patients with low back pain compared with healthy subjects. In that case, in the group with pain, the H reflex size grew more slowly as the stimulus intensity increased, and the H reflex threshold was higher [38].
In short, impairments in the scapula's reflex control could compromise its stabilization and, consequently, the function of the shoulder and the whole upper limb [29]. Thus, rehabilitation should also be planned considering this neurophysiological parameter, with the expectation that it could contribute to longer-lasting results. A previous study suggested the use of proprioceptive exercises and sensory stimulation to enhance the H reflex through more inputs [9]. Moreover, in cases of different adaptations between trapezius portions, exercises focusing on scapula motor control and the balance of muscular activity [55,[76][77][78] could also be important for subjects suffering from chronic shoulder pain associated with changes in H reflex variables such as Hmax/Mmax. Feedback from wearable sensors could even allow the retraining of scapula control in daily activities [79].
Limitations
The present study had some limitations. The first one was the long duration of the protocol, given the necessity of searching for a very specific location for the nerve stimulation, together with the recording of 10 stimuli for each intensity to evoke the H reflex (justified given the high variability of the reflex [60]). Such a factor could be associated with higher demands and fatigue, which could affect the scapular muscles' motor control [80] and the EMG signals [81], and could also limit the applicability and prevent quick access to the evaluation results. However, previous studies have reported inconsistent effects with decreased, increased or unchanged ratios of Hmax/Mmax after fatigue [82]. Therefore, we believe that fatigue likely did not have a very strong influence on the results obtained in the present study. Based on these limitations, the development of tools to facilitate the performance of this measurement in clinical practice is required. Another limitation was the inability to identify an H reflex for the three-trapezius portions of all subjects when the motor point pen was changed for adhesive electrodes, which consequently led to a lower final-sample size. This occurred, even though the H reflex was evoked during an isometric contraction to ensure that the motoneurons were close to their threshold and more easily activated by the Ia afferent volley [58]. However, it was expected, since it also happened in previous studies [13,30,39,58]. Finally, the present study only assessed the reflex response, and the findings do not allow insight into other nervous system structures or other scapular variables to better characterize the origin and/or proportions of the adaptations. Future studies assessing the H reflex, together with variables related to voluntary control or the scapular kinematic and muscular activity levels, could be important.
Conclusions
Subjects with chronic shoulder pain present decreased UT Hmax as a percentage of Mmax, as well as a longer LT H reflex latency. Despite the moderate effect of size and the same tendencies being observed for the other trapezius portions, a significant finding for only one portion may be important, since it could indicate an imbalance between the activity of these scapular stabilizers. Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.
Data Availability Statement: Not applicable.
Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results. | 2023-04-26T15:19:28.580Z | 2023-04-23T00:00:00.000 | {
"year": 2023,
"sha1": "6c7f11b5d90ae8ea98919eeb1650fc984d2c661d",
"oa_license": "CCBY",
"oa_url": "https://www.mdpi.com/1424-8220/23/9/4217/pdf?version=1682247303",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "4fcf9983980cb0618dfb933cbbee4ae76aff63fd",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Computer Science",
"Medicine"
]
} |
21154388 | pes2o/s2orc | v3-fos-license | Spontaneous Transomental Hernia Combined with Incarcerated Inguinal Hernia: A Case Report
An 85-year-old woman was admitted to our hospital because of appetite loss and vomiting. She had no previous history of abdominal surgery, trauma, or intra-abdominal inflammation. Following the diagnosis of a small bowel obstruction due to an incarcerated left inguinal hernia, manual hernia repositioning was attempted, and proceeded smoothly. However, 42 hour after the procedure, she suddenly developed abdominal pain without recurrence of the incarcerated inguinal hernia. Abdominal computed tomography revealed signs of small bowel strangulation; therefore, emergency surgery was performed 48 hour after admission. On laparotomy, a defect was found in the greater omentum, and a part of the ileum was incarcerated by the defect. The operative diagnosis was spontaneous transomental hernia. Incision of the greater omentum was performed, and no ischemic changes were observed in the ileum; therefore, bowel resection was not performed, and the left inguinal hernia orifice was repaired. The postoperative course was uneventful. The transomental hernia in our case might have been induced because of the incarcerated inguinal hernia. For patients plagued with persistent small bowel obstruction after successful manual repositioning of an incarcerated inguinal hernia, concurrent internal hernia should be considered as a possible cause of the persistent small bowel obstruction.
Introduction
Transomental hernia (TOH) is a rare type of internal hernia, in which a loop of the small bowel herniates through an abnormal defect in the omentum, resulting in strangulation of the herniated bowel, with significant morbidity and mortality [1][2][3][4][5][6]. The incidence of TOH is rare, accounting for only 1-4% of total internal hernia cases [1], and most of the omental defects are acquired because of a previous intraabdominal event, such as abdominal surgery, trauma, or intraabdominal inflammation [2][3][4][5][6]. Here, we report a rare case of spontaneous TOH that presented 42 hour after successful manual repositioning of an incarcerated inguinal hernia. We also reviewed the relevant literature and discussed a possible mechanism of the presentation of TOH in our case.
Case Report
An 85-year-old woman with hypertension was admitted to our hospital because of appetite loss and vomiting that started approximately 24 hour prior to admission. She had no previous history of abdominal surgery, trauma, or intra-abdominal inflammation. On admission, her general condition was good, and vital signs were all within normal limits. On physical examination, her abdomen was mildly distended, but there was no tenderness, and a compressible hard mass, measuring approximately 50 mm in diameter, was palpated in the left groin. Plain abdominal computed tomography (CT) revealed dilated and fluid-filled small bowel loops in the abdominal cavity and a fixed small bowel loop in the left groin ( Figure 1). The patient was diagnosed with a small bowel obstruction (SBO) due to an incarcerated left inguinal hernia. Since there was no local redness or sign of peritoneal irritation, manual hernia repositioning
Journal of Gastrointestinal & Digestive System
was attempted, and proceeded smoothly. We hospitalized the patient after the procedure to observe her progress and considered elective definitive surgery for the inguinal hernia within several days. The following day, she remained stable, but nausea and abdominal distension persisted, and 42 hour after the manual hernia repositioning, she suddenly developed abdominal pain. Body temperature was 36.9°C; heart rate, 102 beats/min; respiratory rate, 20 breaths/min; and blood pressure, 118/83 mmHg. On physical examination, her abdomen was severely distended with tenderness; however, there was no recurrence of the incarcerated inguinal hernia. Laboratory results revealed leukocytopenia (4,300/mm 3 ) and elevated levels of blood urea nitrogen (48.7 mg/dL), creatinine (0.84 mg/dL), and C-reactive protein (19.8 mg/dL). Plain abdominal CT revealed a cluster of dilated small bowel loops, a small amount of ascites, conversion of the mesenteric vessels, and a beak sign in the right lower abdomen ( Figure 2). The patient was strongly suspected of developing a strangulated SBO, and emergency surgery was performed 48 hour after admission. On laparotomy, an omental defect without a hernia sac, measuring 40 mm in diameter, was found near the right edge of the greater omentum, and 50 cm of the ileum was incarcerated by the defect (Figure 3). The operative diagnosis was spontaneous TOH. To reduce the incarceration of the ileum, incision of the greater omentum from its right edge to the defect was performed, and no ischemic changes were observed in the ileum. Therefore, bowel resection was not performed, and the left inguinal hernia orifice was repaired with mesh graft. The postoperative course was uneventful. Following rehabilitation for muscular atrophy of the lower legs, she was discharged 27 days later. At 1-year follow-up, she was doing well, with no sign of recurrence of the hernias. Retrospective reading of the initial CT on admission was performed. In the initial CT, conversion of the mesenteric vessels and a beak sign, which were revealed in the follow-up CT on the 2 nd hospital day, were not observed at the same slice levels as in the follow-up CT; however, mesenteric edema was retrospectively identified at the same slice levels.
Discussion
Internal hernia is a characteristic condition of SBO, in which a loop of the small bowel herniates through a congenital or acquired peritoneal or mesenteric aperture within the abdominal cavity, resulting in strangulation of the herniated bowel [1,[7][8][9][10][11]. Internal hernia is an unusual cause of SBO, accounting for only 0.6-5.8% of total SBO cases [7]. The types of internal hernia and frequency of each type are reported as follows: paraduodenal (53%), pericecal (13%), foramen of Winslow (8%), transmesenteric (8%), intersigmoid (6%), and transomental (1-4%) [1,8]. Thus, TOH is a rare entity among various causes of SBO. In children with TOH, the omental defect is usually a congenital anomaly caused by embryonic developmental failure [2,3]. In contrast, most of the omental defects in adult TOH cases are acquired because of a previous intra-abdominal event [2][3][4][5][6]. In a few adult TOH cases, however, the defect might be created because of senile atrophy and/or weakness of the omentum without a predisposing history, as in our patient, and these cases are referred to as "spontaneous" TOH [2][3][4][5].
TOH is known to be more prone to develop strangulation of the herniated bowel than other types of internal hernia [2,4,7,9,10]. Therefore, for the treatment of TOH, emergency surgery must be attempted as soon as possible. However, the preoperative diagnosis of TOH is difficult because clinical manifestations are nonspecific, as [3][4][5][6]11]. As a result, precise diagnosis in most TOH cases, as in our case, is made during emergency surgery for strangulated SBO [3][4][5][6]. Recently, several studies have reported the usefulness of CT, especially multi-detector row CT, for the preoperative diagnosis of TOH [2,9,[12][13][14]. In addition to CT signs of small bowel strangulation, (1) a cluster of dilated small bowel loops in the right paracolic gutter (2) medial and posterior displacement of the ascending colon and cecum by dilated loops and (3) absence of omental fat between the dilated loops and anterior abdominal wall, are suggested as characteristic CT findings of TOH [12,13].
In our case, conversion of the mesenteric vessels and a beak sign, which were revealed in the follow-up CT, were not observed in the initial CT; however, mesenteric edema, which may represent early ischemic change in the small bowel, i.e., an early CT sign of small bowel strangulation, was identified in the initial CT. Accordingly, we speculate a possible mechanism of the presentation of TOH in our case as follows. Initially, an abnormal defect was created in the greater omentum because of senile atrophy, which remained asymptomatic for some time. Under this condition, incarceration of an inguinal hernia incidentally occurred, leading to SBO. This SBO induced an increase in intra-abdominal pressure, which allowed the ileum to herniate through the formerly asymptomatic omental defect. Although the incarcerated inguinal hernia was successfully reduced with manual repositioning and the associated SBO was released, the transomental herniation of the ileum gradually progressed after the procedure, resulting in strangulated SBO due to TOH. Therefore, the TOH in our case might have been induced because of the incarcerated inguinal hernia through an interesting mechanism. To the best of our knowledge, there is no other case report of TOH combined with inguinal hernia in the literature.
SBO due to incarcerated inguinal hernia is a common gastrointestinal disease. Approximately 70% of incarcerated inguinal hernia cases are successfully reduced with manual repositioning [15], which may also facilitate the release of the associated SBO. If the associated SBO cannot be released and persists after successful manual hernia repositioning, the physician should consider recurrence of the incarcerated inguinal hernia or vascular compromise of the reduced small bowel [16], and further examination followed by emergency surgery must be performed. In our case, when the persistent SBO was recognized, we initially suspected stenosis of the reduced small bowel due to vascular compromise. However, the follow-up CT revealed signs of small bowel strangulation, as opposed to a stenotic small bowel segment. Further, on laparotomy, TOH proved to be the cause of the SBO. Therefore, we suggest that for patients plagued with persistent SBO after successful manual repositioning of an incarcerated inguinal hernia, concurrent internal hernia including TOH should be considered as a possible cause of the persistent SBO.
Conclusion
We experienced a rare case of spontaneous TOH combined with incarcerated inguinal hernia. The TOH in our case might have been induced because of the incarcerated inguinal hernia through an interesting mechanism. For patients plagued with persistent SBO after successful manual repositioning of an incarcerated inguinal hernia, concurrent internal hernia including TOH, although rare, should be considered by physicians. | 2019-03-17T13:08:09.888Z | 2017-01-01T00:00:00.000 | {
"year": 2017,
"sha1": "ba5ea7c6c66a4cf541a956406a8d53b6383f05b3",
"oa_license": "CCBY",
"oa_url": "https://doi.org/10.4172/2161-069x.1000532",
"oa_status": "HYBRID",
"pdf_src": "MergedPDFExtraction",
"pdf_hash": "e40f0be6012c4cd299639131fa95d0eec19b5ba0",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
237938189 | pes2o/s2orc | v3-fos-license | Postharvest Treatment of Hydrogen Sulfide Delays the Softening of Chilean Strawberry Fruit by Downregulating the Expression of Key Genes Involved in Pectin Catabolism
Hydrogen sulfide (H2S) plays several physiological roles in plants. Despite the evidence, the role of H2S on cell wall disassembly and its implications on fleshy fruit firmness remains unknown. In this work, the effect of H2S treatment on the shelf-life, cell wall polymers and cell wall modifying-related gene expression of Chilean strawberry (Fragaria chiloensis) fruit was tested during postharvest storage. The treatment with H2S prolonged the shelf-life of fruit by an effect of optimal dose. Fruit treated with 0.2 mM H2S maintained significantly higher fruit firmness than non-treated fruit, reducing its decay and tripling its shelf-life. Additionally, H2S treatment delays pectin degradation throughout the storage period and significantly downregulated the expression of genes encoding for pectinases, such as polygalacturonase, pectate lyase, and expansin. This evidence suggests that H2S as a gasotransmitter prolongs the post-harvest shelf-life of the fruit and prevents its fast softening rate by a downregulation of the expression of key pectinase genes, which leads to a decreased pectin degradation.
Introduction
The Chilean strawberry [Fragaria chiloensis (L.) Mill.] is a native species from South America and the maternal progenitor of the commercial strawberry Fragaria × ananassa Duch. [1]. The Chilean strawberry is a non-climacteric fruit that possesses remarkable organoleptic properties, such as good taste, aroma, nutritional value and an exotic white fruit appearance, having a great potential to become a new exotic berry fruit for the worldwide market [2][3][4]. Besides, F. chiloensis fruit is emerging as a new model for studying several ripening-associated processes in strawberries [5,6], such as anthocyanin biosynthesis and plant cell wall disassembly [7,8]. Studies on fruit softening are important as this influences the postharvest life of highly perishable fruit [9].
The ripening-associated softening of fleshy fruit has been largely described as a direct consequence of enzyme-mediated cell wall disassembly [10]. Events such as depolymerization and solubilization of hemicelluloses and pectins within the cell wall often occur in many fleshy fruits, including strawberry [7]. Genes encoding for proteins involved in cell wall disassembly, such as polygalacturonase (PG), pectate lyase (PL), pectin methylesterase (PE), endo-β-1,4-glucanase (EG), β-galactosidase and expansins (EXPs), increase their expression during strawberry ripening [7,[11][12][13][14]. Previous reports have shown that Chilean Fruit exposed to 0.2-1.2 mM NaHS solutions displayed a longer shelf-life than untreated fruit, nevertheless, the maximum effect was reached with 0.2 mM NaHS ( Figure 1A). In subsequent experiments, 0.2 mM NaHS was employed, extending the shelf-life period (14 days) ( Figure 2). NaHS-treated fruit reached their shelf-life limit after 6 d of storage; while untreated fruit reached this stage after 2 d, thus the shelf-life period of NaHS-treated fruit was nearly tripled (Figure 2A). Interestingly as shown in Figure 2B, after 6 d of storage, untreated fruit evidenced severe signs of fungal infection, while NaHS-treated strawberries were not affected.
Changes in the external color of strawberries during storage, both in control and NaHS-treated fruit, were evaluated throughout the shelf-life period ( Figure 2C). Untreated fruit showed a significant decrease in lightness (L*), while fruit exposed to 0.2 mM NaHS maintained this color parameter during 4 d of shelf-life ( Figure 2C). In addition, control and NaHS-treated fruit displayed a similar decreasing pattern of other color parameters (a* and b* values) during shelf-life ( Figure 2D,E). Color measurements were not reliable at longer shelf-life periods due to the presence of fungal development on the surface of control strawberries.
Effects of H 2 S Treatment on Fruit Color during the Shelf-Life of Chilean Strawberry Fruit
Changes in the external color of strawberries during storage, both in control and NaHS-treated fruit, were evaluated throughout the shelf-life period ( Figure 2C). Untreated fruit showed a significant decrease in lightness (L*), while fruit exposed to 0.2 mM NaHS maintained this color parameter during 4 d of shelf-life ( Figure 2C). In addition, control and NaHS-treated fruit displayed a similar decreasing pattern of other color parameters (a* and b* values) during shelf-life ( Figure 2D,E). Color measurements were not reliable at longer shelf-life periods due to the presence of fungal development on the surface of control strawberries.
Effects of H 2 S Treatment on Fruit Softening and Respiration Rate of Chilean Strawberry Fruit during the Shelf-Life Period
Changes in fruit firmness were followed during the storage period in control and NaHS-treated fruit. Untreated Chilean strawberry fruit evidenced a rapid decrease in firmness throughout the shelf-life period; nevertheless, fruit treated with 0.2 mM NaHS remained firmer than untreated fruit ( Figure 3A). Major differences in firmness between treatments were recorded after 4 d at 20 • C, with values of 1.61 N for NaHS-treated fruit and 0.76 N for untreated fruit. Firmness values of untreated fruit were not reliable after 6 d of storage.
remained firmer than untreated fruit ( Figure 3A). Major differences in firmness between treatments were recorded after 4 d at 20 °C, with values of 1.61 N for NaHS-treated fruit and 0.76 N for untreated fruit. Firmness values of untreated fruit were not reliable after 6 d of storage.
As H2S is an acid molecule, titratable acidity (TA) was determined to investigate whether exposure to NaHS treatment influences changes in acidity ( Figure 3B). TA values increased both in NaHS-treated and control fruit, although with a different pattern. In untreated fruit, TA increases immediately after harvest, while in NaHS-treated fruit, there is a delay as no changes in acidity take place during the first two days. At 6 d of storage, the same acidity level was recorded in both fruit conditions. Changes in respiration rate were also determined during the shelf-life period ( Figure 3C). The production of CO2 of NaHS-treated fruit was maintained at a lower rate during shelf-life, and in contrast, it increased constantly in untreated fruit until 4 d. The respiration rate of untreated fruit was not measured at 6 d of the storage period due to the presence of fungus on the fruit surface, which also releases CO2, interfering with the determination of the fruit tissue.
Changes in soluble solids content (SSC) were also followed in control and NaHStreated fruit during shelf-life ( Figure 3D). SSC decreased in both fruit groups during shelflife with no differences. As H 2 S is an acid molecule, titratable acidity (TA) was determined to investigate whether exposure to NaHS treatment influences changes in acidity ( Figure 3B). TA values increased both in NaHS-treated and control fruit, although with a different pattern. In untreated fruit, TA increases immediately after harvest, while in NaHS-treated fruit, there is a delay as no changes in acidity take place during the first two days. At 6 d of storage, the same acidity level was recorded in both fruit conditions. Changes in respiration rate were also determined during the shelf-life period ( Figure 3C). The production of CO 2 of NaHS-treated fruit was maintained at a lower rate during shelflife, and in contrast, it increased constantly in untreated fruit until 4 d. The respiration rate of untreated fruit was not measured at 6 d of the storage period due to the presence of fungus on the fruit surface, which also releases CO 2 , interfering with the determination of the fruit tissue.
Changes in soluble solids content (SSC) were also followed in control and NaHStreated fruit during shelf-life ( Figure 3D). SSC decreased in both fruit groups during shelf-life with no differences.
Effects of H 2 S Treatment on Cell Wall Polymer Solubilization during the Shelf-Life of Chilean Strawberry Fruit
Cell wall fractionation was performed from AIR samples prepared from F. chiloensis fruit samples subjected to H 2 S treatment. Total cell wall yield was subjected to a sequential fractionation procedure to separate several pectin fractions. The fractionation of pectins in WSF, CSF, and NSF fractions corresponds to loosely-, ionically-, and covalently-bound pectins, respectively, while KSF is mainly associated with the hemicellulose fraction. During the postharvest period, significantly higher solubilization of pectins was observed in the control fruit than in H 2 S-treated fruit (Table S1, Figure 4). The solubilization of pectins advised as the increment in WSF is accompanied by a decrease in CSF and NSF fractions in non-treated fruit ( Figure 4B-D). In H 2 S-treated fruit, pectin solubilization was delayed as WSF did not increase and CSF and NSF did not decrease as in the control fruit. All this evidence suggests that H 2 S treatment delays pectin degradation in Chilean strawberry fruit during the shelf-life period.
Effects of H2S Treatment on Cell Wall Polymer Solubilization during the Shelf-Life of Chilean Strawberry Fruit
Cell wall fractionation was performed from AIR samples prepared from F. chiloensis fruit samples subjected to H2S treatment. Total cell wall yield was subjected to a sequential fractionation procedure to separate several pectin fractions. The fractionation of pectins in WSF, CSF, and NSF fractions corresponds to loosely-, ionically-, and covalently-bound pectins, respectively, while KSF is mainly associated with the hemicellulose fraction. During the postharvest period, significantly higher solubilization of pectins was observed in the control fruit than in H2S-treated fruit (Table S1, Figure 4). The solubilization of pectins advised as the increment in WSF is accompanied by a decrease in CSF and NSF fractions in non-treated fruit ( Figure 4B-D). In H2S-treated fruit, pectin solubilization was delayed as WSF did not increase and CSF and NSF did not decrease as in the control fruit. All this evidence suggests that H2S treatment delays pectin degradation in Chilean strawberry fruit during the shelf-life period. Regarding the solubilization of hemicelluloses, an increase in its content was observed in cell wall material obtained from control fruit during the shelf-life period; however, in H 2 S-treated fruit, this increase is delayed in time (Table S1). The content of hemicelluloses of control fruit after 2 d of shelf-life was reached after 6 d of storage by H 2 S-treated fruit.
Effects of H 2 S Treatment on the Expression of Genes Involved in Pectin Degradation during Strawberry Shelf-Life
To gain molecular insights to explain fruit firmness changes and pectin solubilization in H 2 S-treated fruit, the expression of genes involved in cell wall metabolism was analyzed during shelf-life ( Figure 5). The expression of genes encoding enzymes involved in pectin solubilization such as polygalacturonase (FcPG1) and pectate lyase (FcPL1), displayed a drastic and fast reduction in response to H 2 S treatment ( Figure 5A,B). By contrast, untreated fruit displayed a slow reduction in the expression level of both genes. The expression of these genes reached the lowest levels after 6 d of shelf-life in both treatments. Remarkably, a gene encoding an isoform of expansin (FcEXP2) showed a similar expression profile to that of FcPG1 and FcPL1 ( Figure 5C). The expression of a gene encoding for the enzyme xyloglucan endotransglycosylase/hydrolase 1 (FcXTH1), involved in molecular modifications of hemicellulose did not change its transcriptional level during the first two shelf-life days either in control or H 2 S-treated fruit ( Figure 5D). From the fourth day, a strong reduction in FcXTH1 transcript levels was observed in H 2 S treated fruit, while in contrast, in non-treated fruit, there is an increment in their levels. Furthermore, the expression of a gene that encodes for endo-β-1,4-glucanase 1 (FcEG1), with cellulase activity, recorded a similar pattern to that of FcXTH1 both in H 2 S-treated and controls groups ( Figure 5E); however, FcEG1 transcripts were not detected in untreated fruit at 6 d of storage.
Principal Components and Correlation Analyses on Studied Parameters in H 2 S-Treated Strawberries
To understand the impact of H 2 S treatment in strawberry fruit concerning the analyzed parameters and to identify correlations between them, principal component analysis (PCA) and correlation analysis were performed. The responses are mainly explained by the first two main components (PCA 1: 45.22%; PCA 2: 22.18%) revealing that the most significant variables for the principal component 1 (PC1) were those related to CSF, NSF, and the fruit color (value of L* parameter), which showed negative correlations with treatment time, respiratory rate, and decay index. The principal component (PC2) showed a higher relationship factor for the FcPL1, FcEXP2, FcPG1 genes, and the fruit firmness, showing a negative correlation for the FcXTH1 gene and TA. It is worth noting the strong relationship between the ionically-bound pectin fraction (CSF) and the FcPL1 gene. H 2 S treatments influenced the fruit responses mainly at 4 d of treatment ( Figure 6B). Control fruit showed, at this time, a higher contrasting relationship for the H 2 S-treated fruit, which experienced major changes in the analyzed variables. FcEG1). The data were analyzed by Tukey test (p < 0.05) per time. Each value represents mean ± SD (n = 4). Asterisks indicate significant differences between treatments. For details, see Section 4.
Principal Components and Correlation Analyses on Studied Parameters in H2S-Treated Strawberries
To understand the impact of H2S treatment in strawberry fruit concerning the analyzed parameters and to identify correlations between them, principal component analysis (PCA) and correlation analysis were performed. The responses are mainly explained by the first two main components (PCA 1: 45.22%; PCA 2: 22.18%) revealing that the most significant variables for the principal component 1 (PC1) were those related
Discussion
Until now, most of the research about the role of H2S on the maintenance of fruit quality during postharvest storage has mainly been focused on the antioxidant system and its effect on the ethylene pathway. In this sense, the reduction in the accumulation of ROS and the increase in ascorbic acid, flavonoids, phenolics, and the enzymatic activities of ascorbate peroxidase (APX), catalase (CAT), peroxidase (POD), and superoxide dismutase (SOD) has been reported in fruits such as apple, pear, kiwifruit, grape, strawberry, and mulberry [36,37,[39][40][41][42]. Regarding interference on ethylene pathway by H2S treatment, it has been reported in tomato, a downregulation of fruit ripening-related genes such as those encoding for ethylene-responsive transcription factors ERF003 and H 2 S-treated strawberry fruit showed a strong positive correlation amongst firmness, the CSF and covalently-bound pectin (NSF) fractions, and genes involved in pectin solubilization (FcPG1, FcPL1, FcEXP2), exhibiting a higher significance than the control fruit ( Figure 6C,D). Furthermore, exclusively in the treated fruit, the decay index was negatively correlated to these studied variables. On the other hand, the decay index, which indicates the fruit decay, evidenced a negative correlation with the FcXHT1 gene in the treated fruit, which showed a positive correlation in the control fruit. Additionally, this gene showed a negative correlation for the loosely-bound pectin fraction (WSF) in treated fruit. However, in the control group, the FcXHT1 gene was positively correlated with the WSF fraction, which is also positively correlated with the respiratory rate ( Figure 6C,D).
Discussion
Until now, most of the research about the role of H 2 S on the maintenance of fruit quality during postharvest storage has mainly been focused on the antioxidant system and its effect on the ethylene pathway. In this sense, the reduction in the accumulation of ROS and the increase in ascorbic acid, flavonoids, phenolics, and the enzymatic activities of ascorbate peroxidase (APX), catalase (CAT), peroxidase (POD), and superoxide dismutase (SOD) has been reported in fruits such as apple, pear, kiwifruit, grape, strawberry, and mulberry [36,37,[39][40][41][42]. Regarding interference on ethylene pathway by H 2 S treatment, it has been reported in tomato, a downregulation of fruit ripening-related genes such as those encoding for ethylene-responsive transcription factors ERF003 and DOF22 [35]. In apple and kiwifruit, H 2 S was linked with the suppression of the expression of genes involved in ethylene biosynthesis and signal transduction [39] and the inhibition of ethylene production [41], thereby supporting the counteractive role of H 2 S in the ethylene pathway in climacteric fruits. Certainly, less is known about the specific role of H 2 S on cell wall degradation and related gene expression during postharvest of fleshy fruits. In the present work, we showed evidence about the effect of H 2 S on the pectin and hemicellulose catabolism and the expression of genes encoding for the main strawberry cell wall-modifying enzymes.
A quick decay, softening, and loss of fruit peel lightness are common processes during the senescence of Fragaria chiloensis fruit [43,44]. In this work, it was found that the shelflife of Chilean strawberry fruit was extended by H 2 S treatment by an effect of optimal dose, reaching a most favorable effect with a dose of 0.2 mM NaHS donor, which was evidenced by the largest decrease in decay index and extending the shelf-life limit of treated fruit, around thrice compared to control fruit. The extended strawberry shelf-life was accompanied by a maintained firmness and color lightness, which indicates a significant delay in the senescence process of strawberry fruit tissue. These findings were consistent with previous studies [37] that have reported that H 2 S prolongs the postharvest shelf-life of strawberry (F. × ananassa 'Bao Jiao') by an antioxidative role in fruit. Interestingly, Hu et al. [37] recorded that the optimal dose of H 2 S donor NaHS was 0.8 mM, which is four times higher than the 0.2 mM required for the optimum effect in F. chiloensis fruit (Figure 1). This result might be explained by differences in the perception and signaling pathway of H 2 S of these strawberry species. It is worth noting that F. chiloensis is the maternal parental of F. × ananassa and the above-described differences might be derived from the genetic cross during the mating process between the parental strawberry species, i.e., F. chiloensis × F. virginiana [1]. Alternatively, the lower dosage of H 2 S required for decay control in F. chiloensis could also be explained by its rusticity, as it is still a native undomesticated species that preserves its natural defensive strategies [4].
The physiological events during the ripening and senescence of strawberry fruit mainly involve softening associated cell wall modification [7]. The results reported in this work have shown that H 2 S treatment delayed softening on fumigated F. chiloensis fruit, which remained significantly firmer than untreated fruit throughout the monitored days ( Figure 3A). Similar results in firmness preservation triggered by H 2 S applications have been reported in strawberry 'Bao Jiao' and 'Fengxiang' cultivars [37,38], banana Brazil' [45,46] and kiwifruit 'Jinkui' [41,47]. Besides, in the present research, the respiration rate and titratable acidity were also delayed by the H 2 S exposure on Chilean strawberry fruit, confirming the role of this molecule as an inhibitor of respiration rate as has been described in other non-climacteric fruits such as F. × ananassa [37,38] and mulberry (Morus indica 'Dianmian-1') [42]. It has been previously reported that the high softening rate of Chilean strawberry fruit contributes to its fast postharvest decay [12]. Furthermore, the respiration rate is an important factor in determining the postharvest deterioration of strawberry fruit [37]. Therefore, a delayed softening and respiration rate, due to the H 2 S treatment, influences an extended postharvest shelf-life by a decrease in the senescence and deterioration process.
Regarding changes in cell wall-associated polymers, the effect of H 2 S treatment on cell wall polymer solubilization during the shelf-life of Chilean strawberry fruit was investigated. Pectin solubilization of H 2 S-treated fruit was delayed compared to control fruit as loosely-bound pectin fraction (WSF) did not increase as much as in control fruit up to the fourth day, and consequently, ionically (CSF)-and covalently (NSF)-bound pectin fractions did not decrease as in control fruit ( Figure 4A-D). All this evidence suggests that H 2 S fumigation delays pectin degradation in Chilean strawberry fruit during its postharvest shelf-life period, which is related to the evidenced delay in softening rate in H 2 S-treated strawberry fruit. Interestingly, differences between control and H 2 S-treated fruit were also observed in hemicellulose-related fraction (KSF) in all days of treatment (Table S1), supporting the idea that H 2 S could affect hemicellulose metabolism during postharvest storage. NaHS treatment has been recently observed altering the contents of cellulose and hemicellulose in alfalfa [48]. As far as we know, these results are the first report of the effect of H 2 S treatment on the cell wall polymer contents in fleshy fruit. Therefore, the effect of H 2 S on cell wall polymers especially hemicellulose and cellulose in fleshy fruit needs further characterization.
Plant cell wall disassembly during softening is a direct result of specific enzymatic activities, where pectin degradation plays a major role [10]. Previous reports have evidenced that decreased pectin depolymerization, by an antisense knockdown of a key pectinase gene (pectate lyase), leads to a reduction in the softening rate of strawberry fruit [13]. Thus, to gain molecular insights about how H 2 S delays the cell wall polymer solubilization, the relative expression of key genes involved in pectin and hemicelluloses degradation during strawberry postharvest shelf-life was studied. The results evidenced that H 2 S downregulates the expression of genes involved in pectin and hemicellulose metabolism. Genes encoding key enzymes involved in pectin solubilization such as polygalacturonase (FcPG1), pectate lyase (FcPL1), and an isoform of expansin (FcEXP2), a non-enzymatic protein associated with Chilean strawberry softening [49], exhibited a drastic and fast reduction in response to H 2 S treatment ( Figure 5A-C). It has been reported that the softening rate of the Chilean strawberry fruit reflects the expression of polygalacturonase and pectate lyase genes [12]. Additionally, the expression of expansin genes in strawberry varieties is highly related to contrasting fruit firmness [11,14]. Furthermore, a higher relationship factor for the FcPL1, FcEXP2, FcPG1 genes and the fruit firmness was observed ( Figure 6A). Additionally, H 2 S-treated strawberry fruit showed a strong positive correlation between firmness, the ionically-and covalently-bound pectin fractions (CSF and NSF), and genes involved in pectin solubilization (FcPG1, FcPL1, FcEXP2), exhibiting a higher significance than in control fruit ( Figure 6C,D). Therefore, the effects displayed by H 2 S treatment on decreasing the softening rate, pectin solubilization and the expression of genes involved in pectin depolymerization are highly associated with this previous evidence. Moreover, it has been reported that PG and PE activities are decreased in H 2 S-treated strawberry fruits, prolonging their shelf-life [37,38]. Furthermore, the expression of a gene encoding for an isoform of xyloglucan endotransglycosylase/hydrolase (FcXTH1), involved in molecular modifications of hemicellulose, and a gene that encodes endo-β-1,4-glucanase (FcEG1), with cellulase activity, did not change its transcriptional level until the fourth shelf-life day of H 2 S-treatment, exhibiting a significant increase and a strong reduction, respectively ( Figure 5D,E). H 2 S treatment could probably affect the synthesis of the hemicellulosic polymers by an unknown mechanism that needs further characterization.
Plant Material and Treatments
Ripe F. chiloensis fruit were harvested from a commercial orchard at Purén, the Araucanía Region, Chile (latitude 38 • 04' S; longitude 73 • 14' W). The collected strawberries were immediately transported to the laboratory. Fruit of similar size and without external damage and microbial infection symptoms were selected. A total of 540 fruits were used for the following experiments.
Hydrogen sulfide (H 2 S) fumigation was performed as described by Hu et al. [37] with some modifications in sealed chambers (5 L) using sodium hydrosulfide (NaHS) as an H 2 S donor. Each chamber contained six perforated clamshells, which contained six strawberries, respectively. Initially, 200 mL of NaHS aqueous solutions at 0.2, 0.4, 0.8 and 1.2 mM were placed inside independent chambers to fumigate the fruit at 20 • C for up to 5 d. It is worth noting that the H 2 S-donor (NaHS) and control (water) solutions (200 mL) and the chamber atmosphere were renewed each 24 h, by opening the lid of treatment and control chambers. For the control treatment, water was used instead of NaHS solution. Longer treatments (up to 14 d) were performed at 0.2 mM H 2 S.
Evaluation of Decay Index
Decay index was determined as described by Ayala-Zavala et al. [50] and Hu et al. [37] with modifications. Fifty strawberry fruit per treatment were selected for the assessment of the decay index. Each fruit was classified into four ranks according to decay area: 0, no decay; 1, decay surface less than 10 %; 2, decay surface between 10% and 30%; 3, decay surface between 30% and 50%; 4, decay surface more than 50%. The decay examination was recorded every day using the whole set of fruit per condition. The decay index was calculated by the following equation: decay index = [∑(rank × fruit quantity per rank)/number of fruit x higher rank (4)] × 100%. The experiment was repeated in two different harvest seasons.
Analysis of Fruit Chromaticity
The external color of individual strawberry fruit was analyzed with a colorimeter (model CR-200, Minolta), which measure L*, a*, and b* values, where L* indicates lightness, a* indicates chromaticity on a green (−) to red (+) axis, and b* indicates chromaticity on a blue (−) to yellow (+) axis. Two measurements on each equatorial side were performed per fruit. Values reported corresponding to the mean ± SE of three fruit per experimental condition. The initial fruit L*, a*, and b* values were a mean of 61.0, 6.1, and 17.0, respectively.
Analysis of Fruit Respiration Rate
CO 2 production rate (expressed in µmol CO 2 kg −1 s −1 ) was determined in three independent experiments by a CO 2 and O 2 analyzer (BRIDGE Analyzers, Inc., Bedford Heights, OH, USA). CO 2 percentage was measured after 1 h of incubation of three strawberry fruit in hermetic flasks. The initial fruit respiration rate had a mean of 1.0 µmol CO 2 kg −1 s −1 .
Fruit Firmness Measurement
Firmness (N) was measured using the FirmTech II (BioWorks, Wamego, KS, USA) provided with a flat tip of 2 cm. Samples of six strawberry fruit were evaluated per each experimental condition. Two measurements on each equatorial side were performed per fruit with a penetration depth of 1 mm. The values reported corresponds to the mean ± SE of six fruit per condition. After these analyses, the peduncle and calyx of each fruit were removed, and the fruit was cut into pieces, frozen under liquid nitrogen and stored at -80 • C for further determinations. The fruit from each experimental condition was mixed to provide a bulk of fruit samples. The initial fruit firmness had a mean of 1.7 N.
Determination of Soluble Solids, pH and Titratable Acidity
Two grams of frozen fruit tissue were homogenized in water with a disperser T25 digital Ultra-turrax®(IKA, Staufen, Germany) and adjusted to 25 mL final volume. The mixture was filtrated through miracloth, and the juice was analyzed for soluble solids content (SSC), pH, and titratable acidity (TA). SSC (expressed as %) was measured at 20 • C using a hand-held temperature compensated refractometer (Atago Co., Tokyo, Japan). TA (expressed as m Eq 0.1 kg −1 of fresh weight (FW)) was determined by titration using a pH meter and automatic titrator PH-Burette 24 (Crison, Barcelona, Spain) of an aliquot of 5 mL of strawberry juice with 20 mM NaOH until reaching pH 8.2. The pH of the juice was recorded per each replicate. Three independent fruit extractions were prepared from each biological sample, and values correspond to mean ± SE.
Cell Wall Extraction and Cell Wall Fractionation
Cell wall material was extracted according to Figueroa et al. [44] with some modifications. Five grams of ground frozen fruit tissue were homogenized in 40 mL of 95 % ethanol and boiled for 45 min. The insoluble material was filtered through miracloth and sequentially washed with 15 mL of boiling ethanol, 15 mL of chloroform/methanol (1:1, v/v) and 15 mL of acetone. The residue (Alcohol Insoluble Residue, AIR) was dried overnight at 37 • C and weighed. The results of three replicates per treatment were expressed as mg AIR per g −1 FW.
The fractionation of cell wall material was performed using a sequential chemical treatment of AIR as previously described [44]. The water-soluble, the 50 mM trans-1,2diaminocyclohexane-N,N,N',N'-tetraacetic acid (CDTA)-soluble, the 50 mM Na 2 CO 3soluble, and the 4 M KOH-soluble fractions were obtained and named as WSF, CSF, NSF, and KSF fractions, respectively. Two independent extractions were obtained from each experimental replicate. The concentration of uronic acid (UA) and neutral sugars (NS) in the different cell wall fractions were determined colorimetrically as previously described [51,52]. The results were calculated using standard curves prepared for galacturonic acid and glucose for UA and NS, respectively. Measurements were performed in triplicate, and the results were expressed as mg of galacturonic acid (UA) or glucose (NS) per g of AIR.
Analysis of Gene Expression
Total RNA was isolated from 2 g of frozen fruit samples using a modified CTAB method [53]. Three independent RNA extractions were prepared from each biological sample. One microgram of total RNA was treated with DNAse (Turbo DNA-free kit, Ambion®, Life Technologies) to eliminate genomic DNA. The RNA quantity and purity were estimated at 260/280 nm by NanoDrop™ 1000 Spectrophotometer (Thermo Scientific, Waltham, MA, USA). The RNA integrity was visualized in electrophoresis in 1.5 % (w/v) agarose gel stained with GelRedTM (Biotium, Hayward, CA, USA). The cDNA was synthesized with a first-strand cDNA synthesis kit (Thermo Scientific, Waltham, MA, USA). Quantitative reverse transcription PCR (RT-qPCR) was performed in a Stratagene Mx3000P (Agilent Technologies, Santa Clara, CA, USA). Each reaction consisted of 20 µL containing 2 µL of (1:10 diluted) cDNA, 1 µL of primer mix 10 µM, 10 µL of Sybr Green PCR Master Mix 2X (Stratagene ®, Agilent Technologies), and nuclease-free water to reach the final reaction volume. The cycling conditions were 1 cycle of denaturation at 95 • C for 5 min, followed by 40 cycles (two-segment) of amplification (95 • C/15 s, 60 • C/45 s) and a final melting cycle (95 • C/1 min, 55 • C/30 s and 95 • C/30 s). Each RT-qPCR reaction was performed in three technical replicates and the mean was used for further analysis. A control without template was included in each RT-qPCR run. Fluorescence was measured at the end of each extension step. The specific primer sequences for the genes encoding polygalacturonase 1 (FcPG1), pectate lyase (FcPL), endo-β-1,4-glucanase 1 (FcEG1), expansin 2 (FcEXP2), xyloglucan endotransglycosylase/hydrolase 1 (FcXTH1), and glyceraldehyde 3-phosphate dehydrogenase (FcGAPDH) from F. chiloensis were obtained from a previous report [44]. The relative expression levels were normalized by 2 −∆∆CT method [54] using GAPDH as the reference gene. The results were expressed in arbitrary units assigning the value of one unit to time zero. The data were analyzed by Tukey test (p < 0.05) per time. Each value represents mean ± SD (n = 4).
Statistical Analysis
The results were compared by two-way analysis of variance (ANOVA) (time, treatment) and Duncan's multiple range test at the 5 % level of significance. Shapiro-Wilk and Levene's tests were used to verify normality and homoscedasticity, respectively. Principal component analysis and a Pearson correlation analysis were performed to evaluate the differences between the treatments (control and treated with H 2 S) using R 1.3.1093. The data were presented with their means and standard errors.
Conclusions
In this work, we demonstrated a novel role of H 2 S as a gasotransmitter prolonging the postharvest shelf-life of the Chilean strawberry fruit by preventing its decay and its fast softening rate through a decreased pectin degradation, which reflects the downregulation effect on the expression of key pectinase-related genes. These pieces of evidence provide useful information about the biological function of H 2 S acting as a plant gasotransmitter and transforms the Chilean strawberry fruit as an emergent model for studying the quick softening rate and decay in non-climacteric fruit. In addition, this biological system could provide, in the future, valuable information about the role of H 2 S as a cell wall modifier in plants, adding new insights into the regulation of postharvest physiology of fruit and vegetables during storage. Furthermore, the use of this elicitor emerges as a potent tool for the exogenous application of horticultural products for storage and shelf-life preservation, albeit the approval for use of H 2 S gas on fresh foods is still pending at the global level. | 2021-09-28T05:24:15.631Z | 2021-09-01T00:00:00.000 | {
"year": 2021,
"sha1": "cd6029d51c55222eafd99ed09b723f8bafb7c1d0",
"oa_license": "CCBY",
"oa_url": "https://www.mdpi.com/1422-0067/22/18/10008/pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "cd6029d51c55222eafd99ed09b723f8bafb7c1d0",
"s2fieldsofstudy": [
"Environmental Science",
"Agricultural and Food Sciences"
],
"extfieldsofstudy": [
"Medicine"
]
} |
119362788 | pes2o/s2orc | v3-fos-license | Testing Supersymmetry with Lepton Flavor Violating tau and mu decays
In this work the following lepton flavor violating $\tau$ and $\mu$ decays are studied: $\tau^- \to \mu^- \mu^- \mu^+$, $\tau^- \to e^- e^- e^+$, $\mu^- \to e^- e^- e^+$, $\tau^- \to \mu^- \gamma$, $\tau^- \to e^- \gamma$ and $\mu^- \to e^- \gamma$. We work in a supersymmetric scenario consisting of the minimal supersymmetric standard model particle content, extended by the addition of three heavy right handed Majorana neutrinos and their supersymmetric partners, and where the generation of neutrino masses is done via the seesaw mechanism. Within this context, a significant lepton flavor mixing is generated in the slepton sector due to the Yukawa neutrino couplings, which is transmited from the high to the low energies via the renormalization group equations. This slepton mixing then generates via loops of supersymmetric particles significant contributions to the rates of $l_j \to 3 l_i$ and the correlated $l_j \to l_i \gamma$ decays. We analize here in full detail these rates in terms of the relevant input parameters, which are the usual minimal supergravity parameters and the seesaw parameters. For the $l_j \to 3 l_i$ decays, a full one-loop analytical computation of all the contributing supersymmetric loops is presented. This completes and corrects previous computations in the literature. In the numerical analysis compatibility with the most recent experimental upper bounds on all these $\tau$ and $\mu$ decays, with the neutrino data, and with the present lower bounds on the supersymmetric particle masses are required. Two typical scenarios with degenerate and hierarchical heavy neutrinos are considered. We will show here that the minimal supergravity and seesaw parameters do get important restrictions from these $\tau$ and $\mu$ decays in the hierarchical neutrino case.
INTRODUCTION
The present strong evidence for lepton flavor changing neutrino oscillations in neutrino data [1] implies the existence of non-zero masses for the light neutrinos, and provides the first experimental clue for physics beyond the Standard Model (SM). These oscillations also give an important information on the neutrino mixing angles of the Maki-Nakagawa-Sakata matrix (U M N S ) [2]. The experimentally suggested smallness of the three neutrino masses can be explained in a very simple and elegant way by the seesaw mechanism of neutrino mass generation [3]. This mechanism is usually implemented by the introduction of three heavy right-handed (RH) Majorana neutrinos whose masses, m M i , can be much higher than the SM particle masses. The smallness of the light neutrino masses, m ν i , appears naturally The most elegant solution to this hierarchy problem is provided by the introduction of the symmetry relating fermions and bosons, called supersymmmetry (SUSY). When the seesaw mechanism for the neutrino mass generation is implemented in a SUSY context, the SUSY scalar partners of the neutrinos, i.e. the sneutrinos, do also contribute to the radiative corrections of the Higgs boson masses and cancel the dangerous contributions from the Majorana neutrinos, solving in this way the hierarchy problem of the simplest non-SUSY version of the seesaw mechanism.
The best evidence of supersymmetry would be obviously the discovery of the SUSY particles in the present or next generation colliders. However, there are alternative ways to test supersymmetry which are indirect and complementary to the direct SUSY particle searches. These refer to the potential measurement of the SUSY particle contributions, via radiative corrections, to rare processes which are being explored at present and whose rates are predicted to be highly suppressed in the SM. Among these processes, the Lepton Flavor Violating (LFV) τ and µ decays are probably the most interesting ones for various reasons. On one hand, they get vanishing rates in the SM with massless neutrinos and highly suppressed rates in the SM with massive netrinos. The smallness of these rates in the non-SUSY version of the seesaw mechanism for neutrino mass generation is due to their suppression by inverse powers of the heavy scale m M . On the other hand, although these decays have not been seen so far in the present experiments, there are very restrictive upper bounds on their possible rates which imply important restrictions on the new physics beyond the SM. These restrictions apply even more severely to the case of softly broken SUSY theories with massive neutrinos and the seesaw mechanism, since these give rise to higher rates [4], being suppressed by inverse powers of the SUSY breaking scale, m SU SY ≤ 1 TeV, instead of inverse powers of m M .
We will be devoted here in particular to the LFV τ and µ decays of type l j → l i γ and l j → 3l i where the present experimental upper bounds are the most restrictive ones [5,6,7,8,9], specifically, where the heavy Majorana neutrinos are still active, down to the electroweak scale. We will assume here a Minimal Supergravity scenario (mSUGRA) with universal soft parameters at M X and the breaking of the electroweak symmetry being generated radiatively. This scenario is also refered to in the literature as the constrained MSSM (CMSSM).
The above LFV processes have previously been studied in the SUSY-seesaw context by several authors [10,11,12,13,14], under some specific assumptions for both seesaw parameters, m D (or Y ν , since they are related by m D = Y ν v sin β, with tan β = v 2 /v 1 being the ratio between the two MSSM Higgs vacuum spectation values) and m M , and for the mSUGRA parameters, M 0 , M 1/2 , A 0 , sign(µ) and tan β.
Our present study of these decay channels updates, completes and corrects the previous anlayses in several respects. First we include, by the first time to our knowledge, the full set of SUSY one-loop contributions to the l j → 3l i decays, namely, the photon, the Z boson, and the Higgs bosons penguin diagrams, and the box diagrams. The most complete computation so far of these l j → 3l i decays was done in [10] where the contributions from the photon and Z boson penguin diagrams and from the box diagrams were included, but they focused on the particular choice of degenerate heavy Majorana neutrinos and they presented numerical results just for µ → 3e decays. We extend this previous study by including in addition the Higgs penguin diagrams mediated by the three neutral MSSM bosons, H 0 , h 0 and A 0 , and correct their results for the Z penguin contributions. We also extend their study in that we present results for the three decays, µ → 3e, τ → 3µ and τ → 3e and consider both possible scenarios, degenerate and hierarchical heavy neutrinos.
The contributions from the Higgs penguin diagrams, in the SUSY-seesaw model, were firstly analized in [13]. They worked in the large tan β limit and used the mass insertion approximation to account for the induced effect from the intergenerational slepton mixing in the SUSY contributing loops. There it was concluded that these Higgs-mediated contributions can be very relevant in the large tan β region, because the radiatively induced LFV Higgs-τ -µ couplings grow as tan 2 β (and, in consequence, the BR(τ → 3µ) as tan 6 β), and also because the SUSY one-loop contributions do not decouple in these couplings. These large tan β enhacement and SUSY non-decoupling behaviour were also found in the LFV Higgs boson decays, H 0 , h 0 , A 0 → l ilj [15,16]. A more exhaustive study of the τ → 3µ and other Higgs-mediated LFV τ decays, including an estimate of the Higgs contributions, were done in [17]. However, these previous numerical estimates of the Higgs contributions to LFV τ and µ decays were done in the context of a generic MSSM (see also [18]), where the Higgs boson mass, or equivalently m 0 A , is an input parameter and can take small values of the order of 100 GeV which produces larger rates. A more recent study on the LFV Higgs decays has been done in [19] in the SUSY-GUT SU(5) context. We instead work here in the mSUGRA context where all the MSSM particle masses are quantities derived from the mSUGRA parameters. We will see here that this and the requirement of compatibility with the present experimental lower bounds for all the SUSY particle masses [20] do indeed constraint the contribution from the Higgs penguins.
In the present work we also include the predictions for the l j → l i γ channels which, for the context we work with, are interestingly correlated with the l j → 3l i rates. This correlation has been studied previously in the generic MSSM context in [17] and in a similar mSUGRA context in [14], but in this later the dominant photon penguin approximation was used. We will update this comparative analysis of the l j → l i γ and l j → 3l i rates, in the mSUGRA context, including the full contributions, and considering the very recent upper bounds for τ → µγ [7] and τ → eγ [8]. In addition, we also require the input seesaw parameters to be compatible with the present neutrino data. For this comparison with the neutrino data we use the parametrization first introduced in [12] for the study of the µ → eγ decay.
Our final goal will be to use the SUSY contributions to all the above LFV τ and µ decays as an efficient way to test the mSUGRA and seesaw parameters. With this goal in mind we will analize here the size of the branching ratios in terms of the mSUGRA and seesaw parameters and will explore in detail the restrictions imposed from the present experimental bounds. We will find that for some plausible choices of the seesaw parameters, being compatible with neutrino data, there are indeed large excluded regions in the mSUGRA parameter space.
The present work is organized as follows. In section II we will review the basic aspects of the MSSM extended with three RH neutrinos, their SUSY partners and the seesaw mechanism for neutrino mass generation. The lepton flavor mixing in the slepton sector and in the mSUGRA context will be explained in section III. There we also include the exact diagonalization of the sfermion mass matrices, both in the slepton and in the sneutrino sectors. The analytical results of the LFV l j → 3l i decays will be presented in section IV. The numerical results for all the LFV τ and µ decays will be presented in section V. Finally, section VI will be devoted to the conclusions.
THE MSSM EXTENDED WITH RH NEUTRINOS AND SNEUTRINOS
In this section we briefly review the additional basic ingredients that are needed to extend the MSSM in order to include three right handed neutrinos, their corresponding SUSY partners, i.e. the sneutrinos, and the generation of neutrino masses by the seesaw mechanism. We follow closely the notation of refs. [12,16] to describe the SUSY-seesaw scenario and the connection with neutrino data. For the other sectors of the MSSM we assume here the standard conventions as defined, for instance, in [21,22].
We start with the Yukawa-sector of the MSSM-seesaw that contains the three left handed (LH) SM neutrinos ν o L,i and three extra right handed (RH) massive neutrinos ν o R,i , whose Yukawa interactions provide, after spontaneous electroweak symmetry breaking, together with the right handed neutrino masses, the following mass Lagrangian containing the Dirac and Majorana mass terms, where, Here m D is the 3 × 3 Dirac mass matrix that is related to the 3 × 3 Yukawa coupling matrix Y ν and the MSSM Higgs vacum expectation value, The other MSSM Higgs doublet gives masses to the charged leptons by m l = Y l < H 1 >, where Y l are the Yukawa couplings of the charged leptons and < H 1 >= v 1 = v cos β. The remaining 3 × 3 mass matrix involved in the seesaw mechanism, m M , is real, non singular and symmetric, and provides the masses for the three
RH neutrinos
The mass matrix M ν is a 6 × 6 complex symmetric matrix that can be diagonalized by a 6 × 6 unitary matrix U ν in the following way: This gives 3 light Majorana neutrino mass eigenstates ν i , with masses m ν i (i=1,2,3), and three heavy ones N i , with masses m N i (i=1,2,3), which are related to the weak eigenstates via, The seesaw mechanism for neutrino mass generation assumes a large separation between the two mass scales involved in m D and m M matrices. More specifically, we shall assume here that all matrix elements of m D are much smaller than those of m M , m D << m M , and the predictions of the seesaw model are then given in power series of a matrix defined as, In particular, the previous diagonalization of the mass matrix M ν can be solved in power series of ξ. For simplicity, we choose to work here and in the rest of this paper, in a flavor basis where the RH Majorana mass matrix, m M , and the charged lepton mass matrix, m l , are flavor diagonal. This means that all flavor mixing of the LH sector is included in the mixing matrix U M N S . By working to the lowest orders of these power series expansions one finds, in the flavor basis, the following neutrino 3 × 3 matrices, Here, m N is already diagonal, but m ν is not yet diagonal. The rotation from this flavor basis to the mass eigenstate basis is finally given by the MNS unitary matrix, U M N S . Thus, and the diagonalization of M ν in eqs. (2) and (3) can be performed by the following unitary 6 × 6 matrix: As for the U M N S matrix, we use the standard parametrization given by, where c ij ≡ cos θ ij and s ij ≡ sin θ ij .
Regarding the sneutrino sector, and because of SUSY, the introduction of three RH neutrinos, ν R , leads to the addition of the three corresponding SUSY partners,ν R . Thus, there are two complex scalar fieldsν L andν R per generation, as in the charged slepton case where there arel L andl R . The difference is that in the sneutrino sector, the seesaw matrix ξ is involved, as in the neutrino sector, and gives rise to a natural suppression of the RH sneutrino components in the relevant mass eigenstates. This fact makes the diagonalization procedure simpler in the sneutrino sector than in the charged slepton one.
In order to understand properly this feature of the MSSM-seesaw model, we will illustrate in the following the simplest case of one generation, where this suppression already manifests.
For this, we follow closely [23]. The generalization of this decoupling behaviour of theν R components to the three generations case is straightforward and we omit to show it here, for brevity.
One starts by adding the new terms in the MSSM Lagrangian that involve the ν R and/or ν R . In particular, the usual MSSM soft SUSY breaking potential must be modified to include new mass and coupling terms for the right handed sneutrinos, which for the one generation case are the following, where mM , A ν and B M are the new soft breaking parameters. These are in addition to the usual soft parameters of the slepton sector, mL, mẼ and A l . The sneutrino mass terms of the MSSM-seesaw model can then be written in the one generation case as, with, Notice that, in the sneutrino sector, there are several mass scales involved, the soft SUSYbreaking parameters, mL, mM , B M and A ν , the Dirac mass m D , the µ-mass parameter, the Z boson mass m Z and the Majorana neutrino mass m M . Our basic assumption in all this work is that m M is much heavier than the other mass scales involved (except M X ), m M >> m D , m Z , µ, mL, mM , A ν , B M . The size of B M has been discussed in the literature [23] and seems more controversial. For simplicity, we shall assume here that this is also smaller than m M . In this large m M limit, the diagonalization of the previous sneutrino squared mass matrix is simpler and leads to four mass eigenstates, two of which are light, ξ l 1 , ξ l 2 and two heavy, ξ h 1 , ξ h 2 . In the leading orders of the series expansion in powers of ξ the mass eigenstates and their corresponding mass eigenvalues are given by (We correct in the definitions of M 2 ± and ξ l 2 some typos with wrong signs of ref. [16]), Here we can see that the heavy states ξ h 1,2 will couple very weakly to the rest of particles of the MSSM via theirν L component, which is highly suppresed by the small factor ξ and, therefore, it is a good approximation to ignore them and keep just the light states ξ l 1,2 , which are made mainly ofν L and its complex conjugateν * L . One says then that the heavy sneutrinos decouple from the low energy physics.
The generalization of the previous argument to the three generations case leads to the conclusion that, in the seesaw limit, ξ ≪ 1, the physical sneutrino eigenstates,ν β (β = 1, 2, 3) are made mainly of theν L, l states with l = e, µ, τ respectively, and their corresponding complex conjugates. The process from the weak eigenstates to the mass eigenstates is simplified to the diagonalization of a 3 × 3 sneutrino mass matrix. This is to be compared with the more complex case of charged sleptons where the corresponding process requires the diagonalization of a 6×6 slepton mass matrix. This will be presented in the next section, where the most general case with lepton flavor mixing is considered.
To end this section, we shortly comment on the parameterization that we use to make contact with the neutrino data. It was first introduced in [12] to study the µ → eγ decay and used later by many other authors. The advantage of this parameterization is that instead of using as input parameters the seesaw mass matrices m D and m M it uses the three physical light neutrino masses, m ν i , the three physical heavy neutrino masses, m N i , the U M N S matrix, and a general complex 3 × 3 orthogonal matrix R. With our signs and matrix conventions, the relation between the seesaw mass matrices and these other more physical quantities is given by, where R T R = 1 and, as we have said, m W up to m M . Therefore, we will also include these running effects in the numerical results for all the branching ratios presented in this work.
Regarding the matrix R, we will consider the following parameterization: where c i ≡ cos θ i , s i ≡ sin θ i and θ 1 , θ 2 and θ 3 are arbitrary complex angles. This parameterization was proposed in ref. [12] for the study of µ → eγ decays. It has also been considered in ref. [24,25] with specific values for the θ i angles to study the implications for baryogenesis in the case of hierarchical neutrinos. And it has also been considered by [16] to study the LFV Higgs boson decays into l ilj .
Finally, for the numerical estimates in this work, we will consider the following two plausible scenarios, at the low energies, being compatible with data: • Scenario A: with quasi-degenerate light and degenerate heavy neutrinos, • Scenario B: with hierarchical light and hierarchical heavy neutrinos, In the two above scenarios, we will fix the input low energy data to the following values, ∆m 2 sol = 0.008 eV, ∆m 2 atm = 0.05 eV, θ 12 = θ sol = 30 o , θ 23 = θ atm = 45 o , θ 13 = 0 o and δ = α = β = 0 (See for instance, ref. [26]). Some results will also be presented for the alternative choice of small but non-vanishing θ 13 .
III. GENERATION OF FLAVOR MIXING IN THE SLEPTON SECTOR
Once the three ν R and the threeν R are added to the MSSM particle content, lepton flavor mixing is generated in the slepton sector. This can be seen as the result of a misalignment between the rotations leading to the mass eigenstate basis of sleptons with respect to the one of leptons, which is generically present in the SUSY-seesaw models. This misalignment is radiatively generated from the Yukawa couplings of the Majorana neutrinos and can be sizable, in both, the charged slepton and sneutrino sectors. Usually, it is implemented via the Renormalization Group Equations (RGEs), which we take within the context of mSUGRA extended with three right-handed neutrinos and their SUSY partners. In consequence, we assume here universal soft-SUSY-breaking parameters at the large energies M X >> m M , which must now include the corresponding parameters of the neutrino and sneutrino sectors, namely, Here, M 0 and A 0 are the usual universal soft SUSY breaking parameters in mSUGRA, We present next the slepton mass matrices, relevant to low energies, that include the lepton mixing generated from the neutrino Yukawa couplings by the RGEs. For the charged slepton case and referred to the (ẽ L ,ẽ R ,μ L ,μ R ,τ L ,τ R ) basis, the squared mass matrix can be written as follows, where, The soft SUSY breaking mass matrices and trilinear coupling matrices above, mL ,ij , mẼ ,ij and A ij l , with i, j = e , µ , τ , refer to their corresponding values at the electroweak scale which we get with the SPheno program.
After numerical diagonalization of the M 2 l matrix one gets the physical slepton masses and the six mass eigenstates (l 1 , .....,l 6 )≡l which are related to the previous weak eigenstates For the sneutrino sector, the 3×3 squared mass matrix, referred to theν ′ = (ν e, L ,ν µ, L ,ν τ, L ) basis can be written as follows, where m 2 L,ij are the same as in the previous charged slepton squared mass matrix. After diagonalization of the M 2 ν matrix one gets the relevant physical sneutrino masses and eigenstates,ν β (β = 1, 2, 3) which are related to the previous statesν ′ α by the corresponding 3 × 3 rotation matrix,ν ′ = R (ν)ν , and is such that, Finally, in order to illustrate later the size of the misalignment effects in the slepton sector we define the following dimesionless parameters, is an average slepton squared mass. These parameters have also been considered by other authors in a more model independent approach and with the purpose of getting bounds from experimental data. Some of these bounds can be found in [28,29,30].
For all the numerical results presented in this paper, we will set values for all the following input parameters and physical quantities: • mSUGRA parameters: M 0 , M 1/2 , A 0 , sign(µ) and tan β.
• physical quantities: In this section we present the analytical results for the LFV τ and µ decays into three leptons with the same flavor, within the mSUGRA-seesaw context that we have presented in the previous sections. We perform a complete one-loop computation of the τ and µ decay widths for all the three possible channels, τ − → µ − µ − µ + , τ → e − e − e + and µ → e − e − e + , and include all the contributing SUSY loops. We present each contribution separately, γ-penguin, Z-penguin, Higgs-penguin and boxes. The contributions from the Higgs-penguin diagrams are, to our knowledge, computed exactly by the first time here. We have also reviewed the analytical results in [10] and correct their results for the Z-penguin contributions. Notice that we make the computation in the physical mass eigenstate basis. That is, we consider the one-loop contributions from charged sleptons,l X (X = 1, .., 6), sneutrinosν X (X = 1, 2, 3), . In all this section we follow closely the notation and way of presentation of [10]. The interactions in the physical mass eigenstate basis that are needed for this computation are collected in the form of Feynman rules in Appendix A.
First, we define the amplitudes for the l − decays as the sum of the various contributions, In the following we present the results for these contributions in terms of some convenient form factors.
A. The γ-penguin contributions
The diagrams where a photon is exchanged are called γ-penguin diagrams and are shown in fig. 1. The result for the γ-penguin amplitude contributing to the l − j → l − i l − i l + i decays is usually written as, where q is the photon momentum and e is the electric charge. The photon-penguin amplitude has two contributions in the MSSM-seesaw from the chargino and neutralino sectors respectively, as can be seen in the structure of the form factors, The neutralino contributions are given by On the other hand, the chargino contributions are Notice that in both neutralino and chargino contributions a summation over the indices A and X is understood. Notice also that we have not neglected any of the fermion masses. If we neglect these masses in the previous formulas we get the same result as in [10]. The expressions for the N and C couplings are given in the Appendix A.
B. The Z-penguin contributions
The diagrams where a Z boson is exchanged are called the Z-penguin diagrams and are shown in fig. 2. The amplitude in this case is where, as before, L(R) . The expressions for these form factors are the following: Notice that all the loop functions are evaluated at zero external momenta which is a very good approximation in these decays. That is, The expressions for the couplings are collected in Appendix A and the loop functions [31] are given in the Appendix B. Notice that our result for the Z-penguin contributions differs significantly from the result in [10]. In fact, these authors did not consider all the diagrams in these Z-penguin contributions, which we think is not justified.
C. The box contributions
The box-type diagrams are shown in fig. 3. We have computed these diagrams and found a result in agreement with [10]. The amplitude for these box contributions can be written The different neutralino contributions are, where The chargino contributions read, where
D. The Higgs-penguin contributions
The diagrams where a Higgs boson is exchanged are called the Higgs-penguin diagrams.
These are shown in fig. 4 and have been computed here by the first time. These are usually not considered in the literature. In particular, in the most complete study so far of [10] these Higgs-penguin diagrams were not included. However, they are expected to be relevant at large tan β [13]. We will therefore include them here. Specifically, we include the contributions from the three neutral MSSM Higgs bosons, h 0 , H 0 and A 0 and consider all SUSY loops.
In this case, the amplitude can be written as, where H p (p = 1, 2, 3) = h 0 , H 0 , A 0 and The values of the couplings are given again in Appendix A and the loop functions in Appendix B. Correspondingly, the result for the chargino contribution is given by, where H Again the values of the couplings and the loop functions are given in Appendices A and B respectively.
The decay width for l − j → l − i l − i l + i can be written in terms of the form factors given in the previous sections as [10]: where Notice that we have corrected the result in ref. [10] for the term that goes with A L 2 + A R 2 .
V. NUMERICAL RESULTS FOR THE LFV BRANCHING RATIOS
We present in this section the numerical results for all the branching ratios of LFV τ and µ decays in the context of the mSUGRA-seesaw scenario that has been introduced in the previous sections. We focus on the following LFV decays, τ − → µ − µ − µ + , τ − → e − e − e + and µ − → e − e − e + , and the radiative decays τ − → µ − γ, τ − → e − γ and µ − → e − γ. The reason to consider these radiative decays together with the decays into three leptons is that there are insteresting correlations among them that provide additional information in testing SUSY.
Specifically, we show in this section the correlations between the ratios of τ − → µ − µ − µ + and τ − → µ − γ; between τ − → e − e − e + and τ − → e − γ; and between µ − → e − e − e + and For the numerical estimates of the radiative decays we use the formula of [10], which is given in terms of the A L,R 2 as, but we use our expressions for the form factors in eqs.(32), (33), (35) and (36) that include the lepton mass contributions. We explore here in full detail the size of the SUSY contributions to all these LFV l j → 3l i and l j → l i γ decays as a function of all the mSUGRA parameters, M 0 , M 1/2 , A 0 , tan β and sign(µ) and the seesaw parameters m N i , i = 1, 2, 3 and R or, equivalently, θ 1 , θ 2 and θ 3 . In all this numerical analysis we require compatibility with the neutrino data and with the present upper experimental bounds for all these branching ratios [5,6,7,8,9], as given explicitely in the introduction. We also demand the complete set of SUSY particle masses, which we derive with the SPheno program, to be above the present experimental lower bounds [20]. The numerical values of the total τ and µ widths (lifetimes) are taken from [20]. We show first the results for the scenario A with quasidegenerate light and degenerate heavy neutrinos and next the most interesting scenario B with hierarchical light and hierarchical heavy neutrinos.
A. Degenerate case
We show in figs. 5 through 8 the numerical results of the branching ratios for the LFV τ and µ decays in scenario A with degenerate heavy neutrinos of mass m N . We show our predictions for the three channels, τ − → µ − µ − µ + , τ − → e − e − e + and µ − → e − e − e + , and similarly, for the comparison with the leptonic radiative decays, l j → l i γ, we also show in the plots the correlated decay, τ − → µ − γ, τ − → e − γ and µ − → e − γ, respectively.
The results of the branching ratios for the τ − → µ − µ − µ + and τ − → µ − γ decays as a function of tan β are illustrated in fig. 5. In these plots we set m N = 10 14 GeV and assume the matrix R to be real. Notice that in the degenerate case with real R these LFV ratios do not depend on the particular choice for R. This can be easily understood because the dependence on R drops in the relevant factor, (Y * ν Y T ν ) ij , appearing in the dominant δ ij LL slepton mixing, and due to the property R T R = 1. From this figure we also see that the predicted rates for both channels are well below their respective experimental upper bounds for all tan β values, eventhough the total rates grow fast with tan β. We also see clearly the mentioned correlation between the τ − → µ − µ − µ + and τ − → µ − γ rates.
In fact, this correlation is an inmediate consequence of the dominance of the γ-penguin contributions which clearly governs the size of the τ − → µ − µ − µ + rates. This dominance is illustrated in fig. 5.(a). where the various contributions are shown separately. In fact, the contributions from the γ-penguin diagrams are almost undistinguishable from the total rates for all tan β values. For low tan β values the next dominant contribution is from the Z-penguin diagrams, but this is still more than one order of magnitude smaller than the γ-penguin contribution. The contributions from the box diagrams are even smaller. We tan β also learn that the Z and boxes contributions do not depend significantly on tan β, while the photon contribution goes approximately as (tan β) 2 at large tan β. In this large tan β region it is interesting to note that the total Higgs contribution becomes larger than the Z contribution and the boxes, due to the fact that it grows approximately as (tan β) 6 which leads to the approximate values of 1 440 , 1 94 and 1 162 for (l j l i ) = (τ µ), (τ e) and (µe), respectively. As will be seen later it also works extremely well in the other channels. These nearly constant values of the ratios of branching ratios will be showing along this work.
Obviously, if these ratios could be measured they could provide interesting information.
In fig. 5(c) we have included our predictions for |δ 23 LL |, |δ 23 LR | and, |δ 23 RR |, as defined in eqs. (24), (25) and (26) respectively, as a function of tan β. These are the flavor changing parameters that are the relevant ones for the τ decays having µ in the final state. It is also interesting to compare them with the predictions in the leading logarithmic approximation where the generated mixing in the off-diagonal terms (i = j, i, j = 1, 2, 3), through the running from M X down to m M , is given by and, in consequence, it predicts the hierachy, |δ 23 LL | > |δ 23 LR | > |δ 23 RR |. As expected from the leading-log approximation, we see in fig. 5(c) that |δ 23 LL | is much larger than |δ 23 LR | and |δ 23 RR |. However, we get |δ 23 RR | larger than |δ 23 LR | and it can be indeed two orders of magnitude larger than |δ 23 LR | at large tan β. It is clear that, at least for our choice here of A 0 = 0, the leading-log approximation does not fully work. We also learn from this figure that the size of the mixing is always small in the degenerate case, being the largest |δ 23 LL | about 3 × 10 −3 . We next comment on the relevance of the choice for the m N values. In fig. 6 we have illustrated the τ → µ − µ − µ + and τ → µγ branching ratios as a function of m N for degenerate heavy neutrinos and tan β = 50. The explored range in m N is from 10 8 GeV up to 10 14 GeV which is favorable for baryogenesis. Both rates have the same behaviour with m N which corresponds approximately to BR(τ → µ − µ − µ + ), BR(τ → µγ) ∝ |m N log(m N )| 2 .
As before, these two predicted branching ratios are well bellow their experimental upper bounds, even at the largest m N value of 10 14 GeV. In the last plot of fig. 6 tan β get. However, as already said, the more interesting region of low M 0 and/or M 1/2 values, being close to 100 GeV, is not allowed by the present experimental lower bounds on the MSSM particle masses.
In summary, in the case of degenerate heavy neutrinos, we get LFV τ and µ decay rates which are still below their present experimental upper bounds, for all the explored values of the seesaw and mSUGRA parameters, which have been required to provide a full MSSM spectrum with masses being compatible with the present experimental bounds.
B. Hierarchical case
We next present the results for hierarchical neutrinos, scenario B, which are much more promissing. In this case the choice for R is very relevant. The results for the general complex R case and for the particular mass hierarchy (m N 1 , m N 2 , m N 3 ) = (10 8 , 2 × 10 8 , 10 14 ) GeV, are shown in figs. 9 through 14. This particular choice for the heavy neutrino masses seems to generate a proper rate for baryogenesis via leptogenesis in the hierarchical case [24]. We will later explore other choices as well.
From these figures we first confirm that the LFV τ and µ decay rates are much larger in the hierarchical case than in the degenerate one. This is true even for the case of real R, which corresponds in our plots to the predictions at arg(θ 1 ) = arg(θ 2 ) = arg(θ 3 ) = 0.
Furthermore, we get severe restrictions on the maximum allowed decay rates coming from the experimental upper bounds.
The predictions for BR(τ − → µ − µ − µ + ) and BR(τ → µγ) as a function of |θ 2 | are depicted in fig 9. Here θ 1 and θ 3 are set to zero, and arg(θ 2 ) = π/4. From now on the arguments of θ 1 , θ 2 and θ 3 are written in radians. The other parameters are set to tan β = 50, M 0 = 400 GeV, M 1/2 = 300 GeV, A 0 = 0 and sign(µ) > 0. In fig. 9(a) we show separately the various contributions to BR(τ − → µ − µ − µ + ). The dominant one is again the photonpenguin contribution (which is undistinguisible from the total in this figure) and the others are several orders of magnitude smaller. We also see that the relative size of the subdominant contributions have changed respect to the previously studied degenerate case. Now the Higgs contribution is larger than the boxes one and this is larger than the Z one. This is so because the largest tan β = 50 value has been set. All the rates for τ − → µ − µ − µ + in this plot are within the allowed range by the experimental bound, which is placed just at the upper line of the rectangle. In contrast, one can see in fig .9(b) that, for the chosen mSUGRA and seesaw parameters, the predicted BR(τ → µγ) are clearly above the experimental bound.
The dependence of |δ 23 LL,LR,RR | with |θ 2 | is shown in fig. 9(c). We see that |δ 23 LL | can reach very large values, up to 0.4, for |θ 2 | = 3 and arg(θ 2 ) = π/4. We have checked that this particular choice of θ 2 = 3e iπ/4 gives rise to large neutrino Yukawa matrix elements |Y 33 ν | and |Y 23 ν | of the order of 1, which are the responsible for this large mixing in the slepton sector.
It is also interesting to compare the MSSM spectrum for this hierarchical case with the In fig. 10 we show the predictions of BR(l − j → l − i l − i l + i ) and BR(l j → l i γ) as functions of |θ 2 |, for all the channels and for the different values of arg(θ 2 ) = 0, π/10, π/8, π/6, π/4.
In this fig. 10 we see that all the rates obtained are below their experimental upper bounds, except for the processes τ → µγ and µ → eγ, where the predicted rates for complex θ 2 with large |θ 2 | are clearly above the allowed region. The most restrictive channel in this case is τ → µγ where compatibility with data occurs just for real θ 2 and for complex θ 2 but with |θ 2 | values near the region of the narrow dip. We also see that the rates for BR(µ → 3e) enter in conflict with experiment at the upper corner of large |θ 2 | and large arg(θ 2 ) = π/4. Even more interesting are the predictions for BR(l − j → l − i l − i l + i ) and BR(l j → l i γ) as functions of |θ 1 |, due to the large values of the relevant entries of the Y ν coupling matrix, which are illustrated in fig. 11. Concretely, |Y 13 ν | can be as large as ∼ 0.2 for |θ 1 | ∼ 2.5 and arg (θ 1 ) = π/4, and |Y 23 ν | and |Y 33 ν | are in the range 0.1 − 1 for all studied complex θ 1 values. The results for BR(l − j → l − i l − i l + i ) and BR(l j → l i γ) as functions of |θ 1 |, for different values of arg (θ 1 ), are illustrated in fig. 12. Here θ 2 and θ 3 are set to zero. The same set of mSUGRA parameters and heavy neutrino masses as in fig. 10 are taken for comparison. We see clearly that the restrictions are more severe in this case than in the previous one. In fact, all the rates cross the horizontal lines of the experimental bounds except for BR(τ − → µ − µ − µ + ) With the purpose of exploring other choices of the mSUGRA parameters, we have also generated results for the specific value A 0 = −100 and found very close predictions to the A 0 = 0 case, the lines in the plots being nearly undistinguisable respect to this case. We have also run the alternative case of sign(µ) < 0, and found again very close predictions to the sign(µ) > 0 case, with the lines in the plots being undistinguisable from this case. GeV and (10 8 , 2 × 10 8 , 10 12 ) GeV. We conclude, that the relevant mass is the heaviest one, m N 3 , and the scaling with this mass is approximately as the scaling with the common mass m N in the degenerate case. Because of this, the rates for the two first sets are nearly undistinguisable, and the rates for the third set are about four orders of magnitude below.
Last but not least, we consider the very interesting case where the angle θ 13 of the U M N S is non vanishing. It is known that the present neutrino data still allows for small values of this angle, θ 13 < 10 o . The dependence of BR(µ − → e − e − e + ) and BR(µ → eγ) with this θ 13 is shown in fig. 16 where we explore values in the 0 < θ 13 < 10 o range. We choose these two channels because they are the most sensitive ones to this angle. For this study we assume the most conservative choice of R = 1, and set the other parameters to the following values: In summary, we obtain in the hierachical case much larger rates than in the degenerate one, and one must pay attention to these values, because the rates in several channels do get in conflict with the experimental bounds. More specifically, the choice of a complex R matrix with large modules and/or large arguments of θ 1 and/or θ 2 and a light SUSY spectrum is very constrained by data. We also confirm that the experimental upper bounds of the processes l j → l i γ are more restrictive than the l − j → l − i l − i l + i ones but all together will allow to extract large excluded regions of the mSUGRA and seesaw parameter space.
A more precise conclusion on the excluded regions of this parameter space deserves a more devoted study.
VI. CONCLUSIONS
We have shown in this paper that the LFV τ and µ decays do provide a very efficient tool to look for indirect SUSY signals. Whereas the predicted rates for these processes are negligible within the SM, the SUSY scenario considered here provides in contrast significant rates which are at the present experimental reach for some of the studied channels. This scenario consists of the well known mSUGRA extended with three right handed neutrinos and their SUSY partners, and with the needed neutrino masses being generated via the seesaw mechanism. The reason for these significant rates is because of the important lepton flavor mixing that is generated in the slepton sector due to large Yukawa neutrino couplings, which is transmited via the RGE running from the large energies down the electroweak scale.
With the motivation in mind of testing SUSY we have studied exhaustively in this work the particular decays τ → 3µ, τ → 3e and µ → 3e, and the correlated radiative decays τ → µγ, τ → eγ and µ → eγ. All of these channels have quite challenging experimental bounds and they are expected to improve in the future . We have explored the dependence of the branching ratios for these LFV processes with the various parameters involved, namely, the mSUGRA and seesaw parameters. We have computed and analyzed in full detail all the contributions from the SUSY loops to the l − j → l − i l − i l + i decays. Our analytical results for these decays correct and complete previous results in the literature. In particular we have presented the results for the separate contributions from the γ-penguin, the Z-penguin, the Higgs-penguin and the box diagrams and shown explicitely the γ-penguin dominance. In the numerical estimates we have presented results for both the l − j → l − i l − i l + i and the correlated radiative decays l j → l i γ.
For the degenerate heavy neutrinos case, we have got rates for all the studied LFV τ and µ decays that are below the present experimental upper bounds. The largest rates we get, within the explored range of the seesaw and mSUGRA parameter space, are for the τ decays. First of all, we get much larger branching ratios than in the previous case and secondly they are in many cases above the present experimental bounds.
We have analyzed in detail the behaviour of the branching ratios with the mSUGRA and seesaw parameters also in the hierarchical case. The largest ratios found are again for τ → µγ and τ − → µ − µ − µ + decays. All the LFV τ and µ decay rates are mainly sensitive to tan β, the heaviest neutrino mass m N 3 , which we have set to m N 3 = 10 14 GeV, and the complex angles in the R matrix θ 1 and θ 2 , which have been taken in the range 3 < tan β < 50, 0 < |θ i | < 3 and 0 < arg(θ i ) < π/4. For the values of these parameters at the upper limit of this studied interval we have found that some of the predicted branching ratios are clearly above the corresponding experimental upper bounds. The most restrictive channels being µ → eγ, µ → 3e and τ → µγ. Therefore, we get in this region important restrictions on the posible values of the mSUGRA and seesaw parameters. In particular, for θ 2 = 2.8e iπ/4 , we get that the whole studied range of 100 GeV < M 0 , M 1/2 < 800 GeV with tan β = 50 is totally excluded by τ → µγ. Values of M 0 and M 1/2 in the low region below 250 GeV are also excluded by τ − → µ − µ − µ + data. The case of θ 1 is even more restrictive, because the predictions for µ → eγ, µ → 3e and τ → µγ totally exclude a light SUSY scenario, for practicaly all θ 1 values.
Perhaps, the most striking result is that even for the most conservative choice of R = 1, that is θ 1 = θ 2 = θ 3 = 0, there are also important restrictions at low M 0 , M 1/2 and large tan β values. In particular, for tan β = 50, values lower or equal than M 0 = 250 GeV and M 1/2 = 150 GeV are totally excluded by τ → µγ, µ → eγ and µ − → e − e − e + data.
For this conservative choice of R = 1 we have also found the surprising result that both µ → eγ and µ − → e − e − e + place important restrictions on the allowed values for the U M N S angle θ 13 . For values lower or equal than M 0 = 250 GeV and M 1/2 = 150 GeV and for tan β = 50 and m N 3 = 10 14 GeV, we get that values of θ 13 larger than 1 degree are not allowed by these LFV data.
In conclusion, it is clear from these results that the LFV τ and µ decays studied here do restrict significantly the mSUGRA and seesaw parameter space. A more refined analysis of the restrictions on this multidimensional parameter space, deserves a further study.
Photon interactions
The Feynman rules for the photon interactions that are used in this work are given by,
Neutralino interactions
The Feynman rules for neutralinos that take part in the one-loop diagrams computed here are the following: where (1,3,5)X + tan θ W N * A1 R (l) (2,4,6)X (A1) C is the charge conjugation matrix and P L,R = 1∓γ 5 2 . Here R (l) and N are the rotation matrices in the charge slepton and neutralino sectors, respectively. The definition of N can be found in [21,22]. where Q (ν) We have used here and everywhere the short notation s W = sin θ W and c W = cos θ W .
Higgs boson interactions
The Feynman rules for the three neutral Higgs bosons read as, and H p (p = 1, 2, 3) = h 0 , H 0 , A 0 . We have also used here the standard notation for the MSSM soft-gaugino-mass parameters M 1,2 and the µ parameter. | 2019-04-14T02:38:46.076Z | 2005-10-30T00:00:00.000 | {
"year": 2005,
"sha1": "d9463692a7d5b5814066b00741dc8ea614252701",
"oa_license": null,
"oa_url": "http://arxiv.org/pdf/hep-ph/0510405",
"oa_status": "GREEN",
"pdf_src": "Arxiv",
"pdf_hash": "863052c93b582618efcb9aac76d1da308b66dbec",
"s2fieldsofstudy": [
"Physics"
],
"extfieldsofstudy": [
"Physics"
]
} |
4072094 | pes2o/s2orc | v3-fos-license | Developmentally defined forebrain circuits regulate appetitive and aversive olfactory learning
Postnatal and adult neurogenesis are region- and modality-specific, but the significance of developmentally distinct neuronal populations remains unclear. We demonstrate that chemogenetic inactivation of a subset of forebrain and olfactory neurons generated at birth disrupts responses to an aversive odor. In contrast, novel appetitive odor learning is sensitive to inactivation of adult born neurons, unveiling that developmentally defined sets of neurons may differentially participate in hedonic aspects of sensory learning.
B r i e f c o m m u n i c at i o n s Postnatal neurogenesis occurs in distinct regions of the mammalian central nervous system (CNS) including the hippocampus and the olfactory bulbs (OB). In rodents, neurons generated in the subependymal zone and rostral migratory stream move to the OB to integrate into existing circuits as interneurons 1 . Recent findings using toxininduced cell ablation 2 , blockade of subependymal zone neurogenesis with mitotic inhibitors 3 or irradiation 4 have implicated adult-born OB neurons in a range of olfactory functions that includes odor sensing, odor discrimination, olfactory memory and social recognition. Olfactory learning has also been linked to enhanced survival of adult-born OB neurons 5 , and optogenetic activation of these neurons facilitates olfactory memory and performance 6 . However, there are challenges associated with nonautonomous effects of toxin-based cell ablation on surrounding cells and tissues, as well as incomplete and potential off-target effects of antimitotic drugs on production of new OB neurons. Moreover, use of optogenetics for selective modulation of OB neurons is difficult because of the spatially restricted penetration of light into neuronal tissue in vivo 7 , which imposes the requirement for multiple fiber optic implants in the OB to control the activity of dispersed neurons. These technical obstacles have limited the ability to directly investigate the behavioral and physiological roles of continuous neuronal integration in the olfactory bulbs 8 .
The chemogenetic approach using designer receptors exclusively activated by designer drugs (DREADDs) 9 is far less invasive and an ideal approach for acute control of activity in newborn OB neurons. DREADDs are mutagenized muscarinic acetylcholine receptors that fail to respond to acetylcholine but can be activated by systemic administration of the physiologically inert small molecule clozapine N-oxide (CNO), a metabolite of clozapine. CNO-mediated activation of distinct DREADDs results in depolarization or hyperpolarization of targeted neurons. To selectively express the inhibitory version of DREADDs (Gi) in postnatally born OB neurons, we injected an adeno-associated virus serotype 2 (AAV2) containing Cre-responsive FLEX elements 10 flanking the inverted Gi sequence into the anterior cerebral ventricles of postnatal day 42 (P42) Nestin-cre male mice ( Fig. 1a; referred to as 'P42-Gi' hereafter). To avoid expression in stem cells and immature neuroblasts, the vector contained the 'FLEXed' Gi sequence downstream of a human synapsin promoter (pSYN; Fig. 1a) 11 . Transduced P42-born OB neurons were tracked 4 weeks after injection (Supplementary Fig. 1a) using coexpression of an AAV2-pSYN::EGFP reporter with either hemagglutinin (HA)-tagged or mCherry-fused Gi (Supplementary Fig. 1b,c). AAV2-mediated expression was largely confined to mature neurons in the OB while absent in subependymal zone and rostral migratory stream cells (Supplementary Fig. 1d, Postnatal and adult neurogenesis are region-and modalityspecific, but the significance of developmentally distinct neuronal populations remains unclear. We demonstrate that chemogenetic inactivation of a subset of forebrain and olfactory neurons generated at birth disrupts responses to an aversive odor. In contrast, novel appetitive odor learning is sensitive to inactivation of adult-born neurons, revealing that developmentally defined sets of neurons may differentially participate in hedonic aspects of sensory learning. The distribution and morphology of EGFP + neurons in the OB corresponded to cells in the granule cell layer; some had extensive apical dendrites and others resembled Golgi or star-shaped Blanes cells (Supplementary Fig. 1f) 12 . To test whether EGFP + neurons were active in the OB, we administered vehicle or CNO injections and perfused the mice 90 min later, followed by staining for c-Fos, a reporter for neuronal activity (Supplementary Fig. 1a). The OBs of CNOadministered P42-Gi mice contained significantly fewer EGFP + c-Fos + double-labeled neurons compared to controls (Fig. 1b,c). Thus, selective expression of DREADD-Gi can be targeted to and inhibit a subset of adult-born neurons after CNO administration.
Since newborn OB neurons are responsive to novel odors and are stimulated to form new synapses 13 , we tested P42-Gi responses to novel odors. Four weeks after AAV2-Gi administration, CNO-and vehicle-injected mice explored clean test cages for 45 min while video recorded (Fig. 2a). Basal behaviors were prevalent in both groups during habituation (Supplementary Fig. 2). Both sets of P42-Gi mice were then tested on ability to find novel appetitive odors buried under the bedding in the test cage (10-min trials; treat A, peanut butter cookie; treat B, mint chocolate cookie). Vehicle-injected mice had a significantly higher number of successful trials than their CNO-administered counterparts over 7 d of repeated trials (Fig. 2b). Moreover, the performance of the vehicle group but not the CNOinjected mice improved during the 7 d (Fig. 2c). CNO-administered Nestin-cre mice without AAV2-Gi injections performed similarly to vehicle-injected P42-Gi mice (Supplementary Fig. 3a-c), ruling out unanticipated effects of CNO. To further test the significance of novelty in responses to appetitive odors, CNO-withheld P42-Gi mice were pre-exposed to either treat in seven daily trials ( Supplementary Fig. 3d). This was followed by a second 7-d test trial during which half of the mice received vehicle and the other half CNO injections. All mice in both groups successfully found the learned treats during the entire period (Supplementary Fig. 3e). Thus, a small subset of P42-born neurons in the OB are critical for successfully retrieving novel, but not learned, appetitive odors.
We next asked whether activity of P42-born neurons also plays a role in the expression of fear responses to an aversive odor novel to our mice. To mimic a predator odor we used 2,5-dihydro-2,4, 5-trimethylthiazoline (TMT), an extract from the red fox anal gland, which is an established mediator of rodent freezing responses requiring intact olfactory bulbs 14 . P42-Gi mice were exposed to TMT-treated filter paper placed on one side of a test cage and video recorded for B r i e f c o m m u n i c at i o n s 10 min (Fig. 2a). To our surprise, both vehicle-and CNO-injected mice responded to TMT with robust freezing (Fig. 2d). Neither the responses during the 7 d of repeated exposures (Fig. 2e) nor time spent immobile during the 7 d (Fig. 2f) were significantly different from vehicle controls. As in the appetitive tests, administration of CNO in Nestin-cre mice without intraventricular injections of AAV2-Gi failed to disrupt responses to TMT (Supplementary Fig. 4).
Since heterogeneous populations of neurons begin to integrate into the OB circuitry during perinatal development [15][16][17][18] , we next asked whether neurons generated soon after birth include a population whose activity is critical for freezing responses to TMT later during early adulthood. P0 Nestin-cre pups were intraventricularly injected with AAV2-Gi (P0-Gi) and allowed to survive to P70 as were P42-Gi mice. Since subventricular stem and progenitor cells were transduced in newborn mice, labeled cells were more abundant than in P42-Gi mice and were distributed in various forebrain regions outside the OB (Fig. 3a,b). However, the number of counted neurons and their percentages as a fraction of the sampled population transduced in the forebrain was far greater in the OB and olfactory related regions (for example, olfactory nuclei) than in non-olfactory areas (Fig. 3a- Supplementary Fig. 5). For example, only a miniscule fraction of neurons were found in the amygdala, which is known to regulate freezing behaviors (Fig. 3c). Attempted injections into the olfactory ventricles to focus AAV2 transduction to OB neurons continued to result in spread to various forebrain regions (Supplementary Fig. 6). We presume the spread is due to the open communication between the olfactory ventricles and the lateral ventricles at P0. Despite the larger number of forebrain cells being transduced, CNO-and vehicleadministered P0-Gi mice spent similar proportions of time on exploratory and other basal behaviors when exposed to a novel environment (Supplementary Fig. 7).
Comparison of vehicle-and CNO-injected P0-Gi mice during individual trials of TMT exposure (Fig. 3d) revealed that a higher proportion of the CNO group failed to freeze (Fig. 3e) and responses did not change over the 7 test days (Fig. 3f,g). CNO-induced disruption of freezing was independent of fear-associated acoustic startle response (Supplementary Fig. 8), thus ruling out the potential impact of transduced neurons outside the olfactory circuitries in freezing. Furthermore, performance of P0-Gi mice in the appetitive odor test revealed no significant difference from controls (Fig. 3h,i). These findings suggest that the freezing response to TMT is most likely associated with olfactory circuitries established perinatally. These forebrain circuitries exhibit little to no involvement in detection and learning of novel appetitive odors and/or related exploratory behaviors.
Since dorsal OB glomeruli are known to participate in eliciting avoidance behaviors in mice in response to aversive odors 14,19 , we next assessed whether AAV2-targeted cells in the P0-Gi OBs were preferentially differentiating in dorsal or ventral glomeruli. Mapping the distribution of EGFP + neurons in P0-Gi OBs revealed no preference, with substantial distribution in granule, mitral and glomerular layers of both the ventral and dorsal OB, as well as in the accessory olfactory region (Supplementary Fig. 9). This was in contrast to the largely granule cell layer preference of P42-Gi transduced neurons, where appetitive odor detection could be disrupted by CNO administration. Accordingly, the observed influence of P0-born neurons on detection of the aversive TMT odor is correlated with their differentiation in the periglomerular and mitral cell layers in the OB, in addition to several other olfactory-related nuclei and regions in the forebrain.
Lastly, to examine the proportions of virus-infected cells among newly generated neurons in the OB, we administered bromodeoxyuridine (BrdU) during the day of AAV2 injections to target progenitors synthesizing DNA at the time of surgery ( Supplementary Fig. 10). Only a fraction of BrdU + cells labeled at P0 or P42 were EGFP + , and likewise only a fraction of EGFP + cells were BrdU + by P70. This finding suggests that despite inhibition of only a fraction of developmentally defined cells in the forebrain, activity in their untransduced cohorts is insufficient for mediating detection of new odors. We postulate that developmentally distinct neurons become components of networks wherein small alterations to ensemble activity result in disruption of behavioral outputs. There is precedence for this concept in a recent finding that ensemble activity of central interneurons can be modulated even when only a few neurons in the network are experimentally activated 20 . The precise mechanism of how inactivation of only a fraction of newly integrated cells inhibits a behavior such as finding a novel appetitive odor remains to be determined.
Thus, our study has unexpectedly exposed two sets of developmentally early-and late-derived neuronal ensembles that independently regulate detection and/or associative learning of novel aversive versus novel appetitive odors. Most of these neurons differentiate within olfactory-associated regions, although we cannot neglect the potential role of neurons targeted by our approach in other forebrain regions. Moreover, since a larger population of neurons were targeted at birth as compared to the lower densities targeted in young adult mice, we cannot rule out the possibility that their mere density allows disruption of aversive odor detection. This finding raises the possibility that, unlike attractive odorants, aversive and potentially danger-indicating stimuli may have an innate and developmentally early component in olfactory and related forebrain circuitries. It is tempting to speculate that a hard-wired or innate circuitry established immediately after birth is critical for detection of aversive odorants. This may afford the animal evolutionary or physiological advantage early in life, when it must instinctively protect its immediate survival. In contrast, learning new appetitive odors and maintenance of satiety later in life may be necessary for long-term survival. The precise specification, identity and connectome of neuronal circuits regulating hedonic odor detection and learning remain to be elucidated. Whether or not this theme applies to other sensory and complex behavioral modalities remains to be tested.
Methods
Methods, including statements of data availability and any associated accession codes and references, are available in the online version of the paper.
oNLINe Methods
Animals. Mice were housed, maintained, and used in experiments under the regulations and approval of Institutional Animal Care and Use Committee at North Carolina State University. Male Nestin-cre mice (Tg(Nes-cre)1KLN/J; Jackson Lab, #003771) at either day of birth (P0) or postnatal day 42 (P42, 6 weeks of age) were used in this study. Mice were maintained on a 12 h light:12 h dark cycle with access to water and food ad libitum. None of the animals were used in prior experimental procedures unrelated to this study and mice were obtained from independent litters as much as possible.
Viral microinjections.
High-titer viral stocks of adeno-associated virus 2 (AAV2) carrying the constructs pAAV-pSYN-DIO-EGFP (6 × 10 12 particles per mL) together with pAAV-pSYN-DIO-HA-hM4D(Gi)-IRES-mCitrine (2 × 10 12 particles per mL) or rAAV2-pSYN-DIO-hM4D-mCherry (4 × 10 12 particles per mL) were purchased from the Vector Core at University of North Carolina, Chapel Hill. P42 naive Nestin-cre males were anesthetized using freshly prepared Avertin (375 µg/kg body weight). AAV2 viral stocks (5 µl) were bilaterally injected into the lateral ventricles using a stereotaxic apparatus with coordinates relative to bregma of 0.4 mm lateral, 2.5 mm deep. Following injections, mice (referred to as P42-Gi) were allowed to recover under a heat lamp, then placed back in their home cage and allowed to survive to P70, at which time behavioral experiments were performed. P0 Nestin-cre pups were anesthetized by hypothermia and AAV2 viral stocks were injected (2 µl) into the lateral ventricle ( relative to bregma, 0.2 mm lateral, 1.5 mm deep) or olfactory ventricles (0.8 mm anterior, 0.4 mm lateral, 1.5 mm deep). Following injections, pups were recovered under a heat lamp and returned to the nursing dam. P0-Gi mice were weaned normally and subjected to behavioral tasks at P70. Some P0-Gi and P42-Gi mice were intraperitoneally administered BrdU (Sigma; 100 mg/kg body weight; dissolved in saline) once at the end of the surgery for AAV2 injections. Mice were perfused following final testing as described later. novel appetitive odor detection task. On each day of behavioral testing, mice were intraperitoneally injected with clozapine-N-oxide (CNO; 1 mg/kg body weight dissolved in sterile H 2 O) twice, with 4 h between the injections. Control mice were P0-Gi or P42-Gi mice that only received vehicle injections of sterile H 2 O at same volume as would be given with CNO. Additionally, Nestin-cre mice without intracerebral AAV2-Gi injections were administered CNO to control for possible effects on behavioral responses. Mice were moved from their home cage to a testing cage immediately after the second CNO or vehicle injections. This cage was similar to the home cage with fresh bedding, food pellets and water, and was equipped with video cameras outside the cage. Mice were allowed to habituate to the testing cage for 45 min, after which they were briefly moved to their home cage while the testing cage was prepared for appetitive novel odor detection task. As a source of novel appetitive odors either peanut butter cookie (treat A; 1 g per slice) or mint chocolate cookie (treat B; 1 g per slice) was buried approximately 2 cm below the surface of bedding in the testing cage. All mice used for testing had never been exposed to either treat before testing. Mice were placed in the testing cage and video recorded for 10 min, and time to finding the treats was quantified post hoc. When mice correctly located and recovered the buried treat, it was recorded as a successful trial, with the time to success measured in the video. Each mouse performed the test once daily for 7 consecutive days. Data were plotted and appropriate statistical analyses were performed as indicated.
To test the effects of pre-exposure to treats on novelty task performance, P42-Gi mice were subjected to the task without i.p. injections for 7 consecutive days 3 weeks after intraventricular injections of AAV2-Gi (beginning at P63). Following pre-exposure, mice were divided into two groups (n = 6 each) with one group administered CNO and the other vehicle. Odor detection was video recorded and analyzed in the testing cage as described above. novel aversive odor detection task. For testing of responses to an aversive odor, the same behavioral procedure described above was applied. Fear freezing response was used to measure aversion to an established aversive odor presented by 2 µl of 2,5-dihydro-2,4,5-trimethylthiazoline (TMT; 1 mg/µl) on a filter paper placed on one side of the testing cage and visible to mice on the bedding. Mice were placed in the testing cage and responses were video recorded, and post hoc analyses were performed as described above. Controls were the same as those described for the appetitive odor tests. Freezing behavior was defined as any immobility excluding respiration during the 10-min test period. Latency to freezing was also measured to assess rate of response to TMT.
Acoustic startle test. Mice were acclimated to an SR-LAB Startle Response System (San Diego Instruments) with enclosure centered over a motion sensor for 15 min before testing without any restraint. The system is coupled to a computer to record the responses of mice. Acclimation was followed by trials without sound stimulus to control for effects of background noise. This was followed by a habituation period during which 10 maximum 'pulse noise' stimuli (120 dB, ~40 ms per pulse) were presented. Mice were then exposed to 84, 88, and 92 dBs of prepulse noises in random order and random intervals. This was followed by presentation of a single startle pulse noise followed by a final train of the same prepulse noises. Responses of mice were recorded by the apparatus in millivolts and data were transferred to Microsoft Excel for analysis. Responses were averaged over 7 d per mouse and presented as a dot plot ( Supplementary Fig. 8). Indices of prepulse inhibition were calculated as percentage of pulse test over average responses during the initial trains of pulse noise. tissue processing and immunohistochemistry. Mice were intracardially perfused with 4% paraformaldehyde (PFA; in 0.1 M PBS; pH 7.4) 90 min after the last testing session, and brains were collected and postfixed overnight in 4% PFA at 4 °C. Brains were sectioned on a vibratome at 50 µm in the sagittal plane and blocked in 0.1 M PBS containing 10% serum and 1% Triton X-100 (Sigma) for 1 h at room temperature. For staining of BrdU, sections were washed in 2 N HCl for 10 min followed by three 5-min washes in 0.1 M sodium borate (pH 8.5), followed by blocking. Sections were incubated with primary antibodies in 0.1 M PBS containing 1% serum and 0.3% Triton X-100 overnight at 4 °C, followed by washing in PBS and incubation with appropriate secondary antibodies for 1-2 h at room temperature for visualization the next day. Antibodies and stains used: goat anti-cFos (Santa Cruz, sc-52-G; 1:1,000), chicken anti-GFP (Abcam, ab13970; 1:2,000), rabbit anti-RFP (Abcam, ab62341; 1:1,000), rabbit anti-HA (Cell Signaling, 3724S; 1:500), mouse anti-NeuN (Millipore, MAB377; 1:1,000), mouse anti-BrdU (Becton Dickinson, 347580; 1:100), guinea pig anti-Dcx (Millipore, AB2253; 1:1,000), and DAPI (nuclear stain; Sigma, D9542; 1:2,000). Labeled sections were imaged on Fluoview 1000 (Olympus) or C1 (Nikon) confocal microscopes and double labeling was confirmed by z-stack analyses in IMARIS software. cell counting and analyses. Percentages of double-labeled cells were calculated from 60× (Olympus oil lens) confocal image stacks, 0.5 µm intervals in the z plane and 1,024 × 1,024 resolution in the xy plane (0.24 µm per pixel). Images were obtained from sagittal sections of P42-Gi or P0-Gi brains that were perfused 90 min after CNO or vehicle injections. Counting frames (210 × 210 µm) were captured from random areas in the olfactory bulbs and were analyzed in Image J using the Cell Counter plugin (ImageJ, US National Institutes of Health). Colocalization was confirmed in z-stacks of confocal images. For analysis of c-Fos colocalization with EGFP, all cells in multiple frames were counted until a total of 300 EGFP + cells was reached in each animal. For analysis of overlap of AAV2 transduced cells (EGFP + ) with HA-or mCherry-tagged Gi (Supplementary Fig. 1) and with BrdU (Supplementary Fig. 10), all cells in multiple frames were counted until a total of 100 EGFP + cells was reached in each animal. Data were transferred to Microsoft Excel for statistical analyses.
Number of EGFP + cells in various nuclear and cortical forebrain structures in P0-Gi and P42-Gi mice were counted in tile-scanned confocal images of sagittal sections from three mice in each group (18 sections containing EGFP + cells in areas and nuclei identified in pilot studies, 20× objective, 1,024 × 1,024 resolution, 0.62 µm per pixel). DAPI counterstain was used to delineate architectonic boundaries using the Allen Brain Atlas P56 mouse sagittal reference panels (http://atlas.brain-map.org/atlas?atlas=2).
Statistics. Data collection and analyses were not performed blind to the conditions of the experiments. No data points were removed from statistical analyses. Sample sizes for behavioral experiments were not determined statistically but are in accordance with common practice 21,22 . All statistical analyses were conducted in Microsoft Excel and GraphPad Prism. Data are presented in dot plots with lines indicating mean ± s.e.m., and interquartile ranges are available when appropriate in the Supplementary methods checklist. Temporal dot plots show individual values for each mouse over 7 d of testing. Trend lines were obtained using GraphPad Prism to generate linear regressions and associated 95% confidence intervals of proportions of failed trials or average time spent immobile over time for each group.
For cellular data, two-tailed unpaired t-test was employed for comparison of percentages of cell numbers between groups (normality was assumed but not formally tested). Cellular data were collected from randomly selected regions and sections. Behavioral data could not be tested for normality owing to small sample size; therefore, nonparametric statistics were employed. Assignment of mice to vehicle or CNO groups was arbitrary but not formally randomized. A two-tailed Mann Whitney U test was used when comparing time spent in behaviors or acoustic startle response between groups. Fisher's exact test was used when comparing binary trial outcomes (success or failure; freeze or no freeze) between groups. Hypothesis testing of linear regression slopes was used to determine whether the rate of change in binary outcomes and time spent immobile was different between groups. Significance was defined as P < 0.05. data availability. Data that support the findings of this study are available from the corresponding author upon reasonable request. | 2017-11-08T01:05:28.926Z | 2016-11-05T00:00:00.000 | {
"year": 2016,
"sha1": "f03cbf1dde4be812bbef702f43dfab4c4cf023f6",
"oa_license": "implied-oa",
"oa_url": "https://europepmc.org/articles/pmc5191939?pdf=render",
"oa_status": "GREEN",
"pdf_src": "SpringerNature",
"pdf_hash": "d71a6e1ae79b5617614042b2247a16929164d059",
"s2fieldsofstudy": [
"Biology",
"Psychology"
],
"extfieldsofstudy": [
"Psychology",
"Medicine"
]
} |
252032929 | pes2o/s2orc | v3-fos-license | Independent Association of Thyroid Dysfunction and Inflammation Predicts Adverse Events in Patients with Heart Failure via Promoting Cell Death
Thyroid dysfunction and inflammation are individually implicated in the increased risk of heart failure. Given the regulatory role of thyroid hormones on immune cells, this study aimed to investigate their joint association in heart failure. Patients with pre-existing heart failure were enrolled when hospitalized between July 2019 and September 2021. Thyroid function and inflammatory markers were measured at the enrollment. The composite of all-cause mortality or rehospitalization for heart failure were studied in the following year. Among 451 participants (mean age 66.1 years, 69.4% male), 141 incident primary endpoints were observed during a median follow-up of 289 days. TT3 and FT3 levels were negatively correlated with BNP levels (r: −0.40, p < 0.001; r: −0.40, p < 0.001, respectively) and NT-proBNP levels (r: −0.39, p < 0.001; r: −0.39, p < 0.001). Multivariate COX regression analysis revealed that FT3 (adjusted HR: 0.677, 95% CI: 0.551–0.832) and NLR (adjusted HR: 1.073, 95% CI: 1.036–1.111) were associated with adverse event, and similar results for TT3 (adjusted HR: 0.320, 95% CI: 0.181–0.565) and NLR (adjusted HR: 1.072, 95% CI: 1.035–1.110). Restricted cubic splines analysis indicated a linear relationship between T3 level and adverse events. Mechanistically, primary cardiomyocytes showed strong resistance to TNF-α induced apoptosis under optimal T3 concentrations, as evidenced by TUNEL staining, flow cytometry analysis, and LDH release assay as well as increased expression of Bcl-2. Thyroid dysfunction and inflammation are independently associated with cardiovascular risk in heart failure patients, which may concurrently contribute to the ongoing cardiomyocyte loss in the disease progression.
Introduction
Thyroid hormones (THs) play an indispensable role in development as well as organ homeostasis. They have been implicated to regulate the cardiovascular system, including cardiac rhythm, contractility, hypertrophy, and vascular resistance [1,2]. Thyroid dysfunction, even a minor alteration of circulating THs, may induce or exacerbate cardiovascular disorders towards heart failure (HF) [3]. Indeed, either hypothyroid or hyperthyroid states were identified to be related to 58% and 85% increases in cardiac death relative risk compared with euthyroid status [4]. Moreover, both overt and subclinical thyroid dysfunction can adversely affect cardiac function [3]. Most importantly, thyroid dysfunction is modifiable, which brings a potential therapeutic option for patients with heart failure. However, the significant gap concerning the exact molecular basis underlying the interaction between the thyroid and cardiovascular systems becomes a major hurdle to optimally manage patients with both thyroid and cardiac abnormalities.
In the past decades, the immune system has been regarded as an important target of THs [5]. The systematic inflammatory status is affected by TH levels. Hypothyroidism 2 of 14 tends to suppress the activation of inflammatory response, while hyperthyroidism commonly enhances the activity of neutrophils and lymphocytes [6]. Interestingly, an ongoing inflammatory response is considered as a major regulator in the pathogenesis of heart failure [7]. Since the 1990s, elevated circulating levels of tumor necrosis factor α (TNF-α) have been noticed in HF patients [8]. Accumulating evidence has highlighted an important role of the inflammatory response, featured by accumulated circulating pro-inflammatory cytokines and immune cells in the heart, and in acute or chronic heart failure resulting from distinct etiologies [9][10][11][12]. Targeting inflammation could be a potential therapeutic utility, as revealed by the subanalysis of the CANTOS trial [13]. However, whether the prevalence or prognostic significance of thyroid dysfunction in HF is related to inflammation has not been investigated.
Interestingly, patients with chronic heart failure may present with overt or subclinical hypothyroidism [14]. The increase of proinflammatory cytokines such as TNF-α have been noticed to be associated with low-T3 syndrome [15], which may result from the adverse effects of inflammation on the conversion of thyroxine (T4) to triiodothyronine (T3). However, hypothyroidism seems to undermine the pro-inflammatory role of neutrophils, macrophages, and lymphocytes, which is supposed to be beneficial to heart failure. On the contrary, previous studies showed that administration of low levels of T3 can promote cardiac function to some extent [16,17]. Little is known about the interconnection among thyroid dysfunction and inflammation in HF. To address this important gap, we aimed to explore the joint association of thyroid dysfunction and inflammation with cardiovascular risk in heart failure patients. We hypothesized that these two factors may simultaneously promote ongoing cardiomyocyte loss, leading to adverse events in heart failure.
Study Population
From July 2019 to September 2021, 509 patients hospitalized with HF in the Second Affiliated Hospital, Zhejiang University School of Medicine, with complete clinical data and which provided written informed consent were enrolled. HFs were defined based on the 2022 heart failure guidelines [18]. Patients without echocardiogram, lack of thyroid function test, or loss of visit were excluded. Patients with hyperthyroidism or hypothyroidism were also excluded. Generally, Probable HF hospitalizations (n = 509) were eligible for inclusion. Fifty-eight patients met the exclusion criteria, and 12 patients were lost to follow-up. Finally, 451 validated HF patients were included ( Figure 1). The study was conducted on the grounds of the Declaration of Helsinki and approved by the Ethics Committee of the Second Affiliated Hospital, Zhejiang University School of Medicine.
commonly enhances the activity of neutrophils and lymphocytes [6]. Interestin ongoing inflammatory response is considered as a major regulator in the pathogen heart failure [7]. Since the 1990s, elevated circulating levels of tumor necrosis f (TNF-α) have been noticed in HF patients [8]. Accumulating evidence has highligh important role of the inflammatory response, featured by accumulated circ pro-inflammatory cytokines and immune cells in the heart, and in acute or chron failure resulting from distinct etiologies [9][10][11][12]. Targeting inflammation could b tential therapeutic utility, as revealed by the subanalysis of the CANTOS tri However, whether the prevalence or prognostic significance of thyroid dysfunc HF is related to inflammation has not been investigated.
Interestingly, patients with chronic heart failure may present with overt or sub hypothyroidism [14]. The increase of proinflammatory cytokines such as TNF-α ha noticed to be associated with low-T3 syndrome [15], which may result from the a effects of inflammation on the conversion of thyroxine (T4) to triiodothyronin However, hypothyroidism seems to undermine the pro-inflammatory role of neut macrophages, and lymphocytes, which is supposed to be beneficial to heart failure. contrary, previous studies showed that administration of low levels of T3 can p cardiac function to some extent [16,17]. Little is known about the interconnection thyroid dysfunction and inflammation in HF. To address this important gap, we ai explore the joint association of thyroid dysfunction and inflammation with cardiov risk in heart failure patients. We hypothesized that these two factors may simultan promote ongoing cardiomyocyte loss, leading to adverse events in heart failure.
Study Population
From July 2019 to September 2021, 509 patients hospitalized with HF in the Affiliated Hospital, Zhejiang University School of Medicine, with complete clinic and which provided written informed consent were enrolled. HFs were defined ba the 2022 heart failure guidelines [18]. Patients without echocardiogram, lack of function test, or loss of visit were excluded. Patients with hyperthyroidism or hy roidism were also excluded. Generally, Probable HF hospitalizations (n = 509) wer ble for inclusion. Fifty-eight patients met the exclusion criteria, and 12 patients were follow-up. Finally, 451 validated HF patients were included ( Figure 1). The stu conducted on the grounds of the Declaration of Helsinki and approved by the Committee of the Second Affiliated Hospital, Zhejiang University School of Medicin
Thyroid Hormone Sampling and Subgroups of Thyroid Status Definition
We used electrochemiluminescence Immunoassay (ECLIA) for the analysis of THs immediately after sampling. The reference ranges were thyrotropin (TSH), 0.35 to 4.94 mIU/L, free thyroxine (FT4), 9.01 to 19.05 pmol/L. Further categorical analysis assigned patients to the following groups based on FT4, free triiodothyronine (FT3), and TSH levels through our institution's cutoffs. Euthyroidism refers to TSH of 0.35 to 4.49 mIU/L; subclinical hypothyroidism (SCH) refers to TSH of 4.94 to 19.9 mIU/L with normal FT4; subclinical hyperthyroidism refers to TSH levels under the reference range with normal FT4 levels; additionally, low-T3 syndrome refers to TSH of 0.35 to 4.94 mIU/L with decreased FT3 [19].
Data Collection and Clinical Outcome
For patients with multiple admission records, the latest medical record was adopted as the baseline record. Clinical status, comorbidities, medication, intervention, laboratory results, and echocardiogram results were obtained from this clinical database. Echocardiogram results were left ventricular ejection fraction (LVEF), left ventricular end-diastolic volume (LVEDV), interventricular septum thickness in diastolic phase (IVSd), left atrium (LA) dimension (defined as the largest diameter of left atrium in parasternal long axis view), left ventricular internal diameter in diastolic phase (LVIDd) and left ventricular posterior wall thickness in diastolic phase (LVPWd). Laboratory results include brain natriuretic peptide (BNP), N-terminal pro-B type natriuretic peptide (NT-proBNP), cardiac troponin T (cTnT), kinase isoenzyme-MB (CK-MB), CRP, full blood counts (like neutrophils, lymphocytes, monocytes), creatine (Cr), hemoglobin (Hb), alanine aminotransferase (ALT), glycated hemoglobin (HbA1c), serum glucose (Glu), and triglycerides (TG). The quality of medical therapy in heart failure was measured by Heart Failure Collaboratory (HFC) score. The quality of medical therapy was considered as sub-optimal, acceptable, and optimal treatment when the score was <3, 3-4, and ≥5, respectively [20]. The primary endpoint was the composite of all-cause mortality or rehospitalization for heart failure.
Cell Culture
Neonatal rats (1 to 3 day old) were supplied by the Zhejiang province animal center, China. A neonatal heart dissociation kit (No 130-098-373, Miltenyi Biotec, Bergisch Gladbach, Germany) was utilized for the primary neonatal rat ventricular cardiomyocytes prepared. Cells were cultured in a complete medium with 10% FBS, 1% penicillin-streptomycin, and Brdu (1:200), and then incubated in a 37 • C humidified incubator with 5% CO 2 . T3 was provided by Sigma (No T074, St. Louis, MO, USA) and cells were treated with 1, 10, 30, 60, and 120 ng/mL T3 as well as 10 ng/mL TNF-α for 24 h to induce an inflammatory response based on previous studies [21,22].
Cell Apoptosis Assay and Apoptosis Detection by Western Blot
Apoptotic cardiomyocytes were detected by TUNEL labeling (Beyotime, Institute of Biotechnology, Haimen, China). Nuclei were stained by DAPI, and TUNEL-positive cells were quantified by the high content-screening method. The percentage of apoptotic cells was considered as the apoptotic index, and each group was randomly chosen by 16 fields of each well. Meanwhile, AnnexinV-FITC/PI Apoptosis Detection Kit (MultiSciences Biotechnology) was used for detecting cardiomyocytes apoptosis based on flow cytometry. In the LDH release assay, cells were planted in a 96-well plate, and treated with TNFα along with various concentrations of T3 (0 ng/mL, 10 ng/mL, 30 ng/mL, 60 ng/mL and 120 ng/mL) for 24 h. Then cells were managed by the LDH Cytotoxicity Assay kit (Beyotime, Institute of Biotechnology, Haimen, China) and determined under the absorbance at 490 nm. SDS-PAGE gels and PVDF membranes were used for the protein separation and transferring respectively (Millipore corporation, Shanghai, China). Five percent milk was prepared for the membranes blocked for 60 min before taken to the antibodies against Bcl-2 (1:500 dilution) and Bax (1:1000 dilution) at 4 • C temperature overnight. The membranes were then washed by 1 xPBST 3 times and taken with anti-Rabbit secondary antibodies for 60 min. β-actin served as a loading control.
Statistical Analysis
Continuous variables in both thyroid dysfunction and euthyroid groups were tested for the normal distribution. The results were presented as number (%), mean ± standard deviation (if data fitted normal distribution), or median ± quartile (if data did not fit normal distribution). Categorical variables were compared among groups using chi-squared test or Fisher's exact test. Continuous variables were compared using unpaired Student's t-test or Kruskal-Wallis's test as appropriate. Univariate and multivariate COX regression analysis was constructed to investigate the association of thyroid function and inflammation values with the primary endpoint. Cubic splines were taken to evaluate the linearity between T3 and the incidence of study endpoints. Quantitative data are presented as the mean ± standard deviation of three independent experiments. p values <0.05 were considered statistically significant. The R package and python were used for all statistical analyses with two-tailed p values of 0.05.
Study Subjects and Baseline Characteristics
Four hundred and fifty-one participants (69.40% males) with a median age of 66 years old were finally included in this study. Table 1 shows the baseline characteristics of the study population. Among the 451 patients included, 109 patients (24.17%) were classified as having thyroid dysfunction. In the thyroid dysfunction group, cardiac function parameters including NT-proBNP, BNP, and cTnT levels were nearly two-fold higher. FT4, total thyroxine (TT4), FT3, and total triiodothyronine (TT3) levels were lower in the thyroid dysfunction group, whereas TSH, as well as TPOAb levels, were higher in those patients.
Correlation of Thyroid Function with Cardiac Parameters and Inflammation Values
As baseline, characters exhibited a significant difference of HF biomarkers, such as BNP and NT-proBNP based on thyroid status; we then sought to detect the association between THs and cardiac function. A significant trend to a negative association was also present in the TT3 levels with BNP (r: −0.40, p < 0.001) and NT-proBNP (r: −0.39, p < 0.001), respectively. Next, we assessed the correlation between inflammation values and thyroid function. There were negative correlations between NLR as well as neutrophils with FT3 (r: −0.22, p < 0.001; r: −0.11, p = 0.022, respectively) and TT3 values (r: −0.23, p < 0.001; r: −0.12, p = 0.014, respectively) ( Figure 2).
The Protecting Role of T3 in TNF-α Induced Neonatal Rat Ventricular Cardiomyocyte Apoptosis
The TUNEL method was utilized to explore the functional role of T3 on cardiomyocytes apoptosis. The cardiac apoptosis index was significantly increased compared with the control group when exposed to TNF-α. When the cardiomyocytes were further exposed to T3 with various concentrations, the apoptotic index first increased at 10 ng/mL and decreased dramatically at 30 ng/mL compared with TNF-α alone. The result indicated that 30 ng/mL of T3 alleviated cardiomyocyte apoptosis obviously ( Figure 5A,B). Consistently, the flow cytometry analysis with Annexin V-FITC/PI double staining also suggested that a proper concentration of T3 supplementation may decrease cardiomyocyte apoptosis compared with TNF-α alone, whereas too low or too high T3 may not exert beneficial effects on cardiomyocytes under TNF-α stimulation ( Figure 5C). LDH release assay was used to detect cell damage among each group. As shown in the Figure 5D, cells treated with TNF-α and 30ng/mL T3 significantly reduced LDH release, which consists with the TUNEL results. The expression of Bcl-2 and Bax were then determined by TNF-α or with various concentrations of T3 (10, 30, 60, and 120 ng/mL) for 24 h. We found an increased Bcl-2 protein expression when pretreated with T3, especially at 10 ng/mL, compared with the TNF-α only. These results indicated that low concentrations of T3 could prevent TNF-α-induced cardiomyocyte apoptosis ( Figure 5E).
Discussion
It has been previously studied that hospitalized HF patients can suffer transient thyroid dysfunctions such as subclinical thyroid dysfunction or low T3 syndrome [23]. SCH is considered to have an increased TSH level above the upper reference limit with a normal level of serum T4 and T3 [24]. Meanwhile, low-T3 syndrome is characterized by decreased serum T3 levels with serum TSH levels in the normal range, and changes of T3 concentration can reflect the severity of the illness [25]. The presence of low-T3 syndrome or SCH has been reported following pathologic response to acute illness such as pneumonia or myocardial infarction [26,27]. Additionally, a higher risk of developing thyroid dysfunction was detected in patients with more advanced stages of HF [28]. In the present, our study investigated thyroid function changes in the pre-existing HF subjects and suggested that lower levels of TSH as well as TT3 were found associated with more severe symptoms of HF at baseline. Further analysis indicated that inflammation plays a vital role in the HF-mediated thyroid dysfunction. A possible explanation for the inverse relationship between TSH level and inflammatory response can be attributed to the . Data are presented as the mean ± SEM. * p < 0.05 versus control group, # p < 0.05 versus TNF-α group evaluated by one-way analysis. ** p < 0.01, *** p < 0.005, # p < 0.05, ### p < 0.005. (C). Flow cytometry analysis of apoptosis by Annexin V-FITC/PI double staining and statistical analysis of apoptosis rate. Data are presented as the mean ± SEM. * p < 0.05, *** p < 0.005. (D). LDH release assay with cardiomyocytes treated with TNF-α and in presence or absence of T3 (0 ng/mL, 10 ng/mL, 30 ng/mL, 60 ng/mL and 120 ng/mL) for 24 h. Data are presented as the mean ± SEM. * p < 0.05, *** p < 0.005. (E). Western blot assay was used for detection cell apoptotic protein Bcl-2 and Bax expression and statistical analysis for Bcl-2/Bax ratio. Data are presented as the mean ± SEM. ** p < 0.01.
Discussion
It has been previously studied that hospitalized HF patients can suffer transient thyroid dysfunctions such as subclinical thyroid dysfunction or low T3 syndrome [23]. SCH is considered to have an increased TSH level above the upper reference limit with a normal level of serum T4 and T3 [24]. Meanwhile, low-T3 syndrome is characterized by decreased serum T3 levels with serum TSH levels in the normal range, and changes of T3 concentration can reflect the severity of the illness [25]. The presence of low-T3 syndrome or SCH has been reported following pathologic response to acute illness such as pneumonia or myocardial infarction [26,27]. Additionally, a higher risk of developing thyroid dysfunction was detected in patients with more advanced stages of HF [28]. In the present, our study investigated thyroid function changes in the pre-existing HF subjects and suggested that lower levels of TSH as well as TT3 were found associated with more severe symptoms of HF at baseline. Further analysis indicated that inflammation plays a vital role in the HF-mediated thyroid dysfunction. A possible explanation for the inverse relationship between TSH level and inflammatory response can be attributed to the higher degree of cardiac decompensation mediated endogenous stress in HF patients. Increased endogenous stress accounts for higher levels of cortisol, which further reduces TSH concentrations [29]. As a positive loop, decreased TSH concentrations resulted in higher adrenergic activity and heart rates which could, in turn, aggravate cardiac function [30]. On the other hand, HF status could lead to the transient increase in pro-inflammatory factors, like IL-1 and TFN-α, accounting for temporary changes in peripheral as well as central THs by the role of TSH on the thyroid, and, subsequently, resulting in decreased T 3 and T 4 levels in circulation [31].
Besides, in the setting of HF status, neutrophils accumulation acts as the initial inflammatory response followed by the infiltration of mononuclear phagocytes, including monocytes, macrophages, and dendritic cells. Neutrophils are responsible for the maintenance of inappropriate immune responses and can destroy tissue through the enzyme myeloperoxidase (MPO) and free radicals in the process of inflammation and immunity [32]. After one week, increased lymphocytes involved in the responses of the immune system, including T cells, B cells and macrophages present with anti-inflammatory effects (express IL-10) appear around this time point for inflammation resolution and LV remodeling [33]. The link between immune cells and THs metabolism was first noticed in the 1970s when radioactively labeled THs was found to appeal to the sites of bacterial infection [34]. Recently, based on the strong immunostaining of the TH-inactivating deiodinase DIO3 detected in infiltrating leukocytes in bacterial infection mice models, neutrophils are confirmed to metabolize THs [35]. Furthermore, in the cerebral spinal fluid of bacterial meningitis patients, there is a marked change of THs level characterized by increased T4 and rT3 concentrations, which is consistent with elevated DIO3 activity detected in infiltrating neutrophils [36]. Besides, lymphocytes like T cells and B cells could induce autoimmune thyroiditis and decreased plasma THs level by producing antibodies against thyroid antigens [37]. According to the above study, the correlation between THs with immune cells can be explained by either the deiodination role or the direct immunoinflammatory injury [38].
Meanwhile, the subtle changes in thyroid function may be more remarkable in preexisting HF patients [19]. A previous study by Tomohiro et al. based on 274 subjects demonstrated that SCH or an increased TSH level could independently predict hospitalization for worsening HF [15]. Besides, in a prospective cohort study with 1365 pre-existing HF patients, patients with isolated low T3 levels or SCH with TSH ≥ 7 mIU/L indicate poorer prognosis as well [19]. Both neutrophils and lymphocytes have been determined to be strongly and independently associated with advanced HF patients' hospitalization, survival, and survival free from heart transplant [39]. However, in patients with decompensated AHF, Uthamalingam et al. showed that NLR rather than absolute neutrophil or lymphocyte counts independently was related to the in-hospital mortality and post-discharge clinical outcomes independent of the LV function [40]. Additionally, as complete blood count involves automatic leukocyte subsets distribution analysis and is widely utilized, NLR values can be a great tool in the disease survival or prognosis assessment, probably more efficient than CRP [32]. NLR has already been used to determine the disease severity in thyroiditis, CVD, malignancies, and other inflammatory-related diseases [32,[41][42][43]. Our univariate COX analysis suggested that NLR (HR: 1.090, 95% CI: 1.060-1.130, p < 0.001) was significantly associated with an elevated risk of composite cardiovascular events in included subjects, and the multivariate analysis showed the same results. However, the follow-up period of the present study is still not long enough, which may cause difficulties in predicting the influence of thyroid hormones and inflammatory markers on the prognosis of HF patients in long run. Further studies with longer follow-up durations are needed.
Inflammatory cytokine TNF-α plays a vital role in cell apoptosis stimulation and extracellular matrix degradation. In our study, TUNEL-positive cells and the release of LDH were increased in cardiomyocytes stimuli with TNF-α, which was decreased by the low concentrations of T3 (30 ng/mL). Apoptosis is characterized by cell shrinkage, DNA fragmentation, chromatin condensation, and apoptotic bodies [44]. Ferroptosis, an irondependent regulated cell death [45], can cause ROS and lipid peroxidation production [46]. Unlike typical apoptosis and autophagy, ferroptosis presented with reduced mitochondrial volume in morphology, increased density of the mitochondrial membrane, and intracellular iron content [47]. Recently, endoplasmic reticulum stress (ERS) has aroused much attention in the ferroptosis progression, which is a kind of warner under the endoplasmic reticulum dysfunction circumstance by activating the transcription factor 4-C/EBP homologous protein (ATF4-CHOP) pathway [48]. During the process of ferroptosis in mitochondria, iron ions and NADPH oxidase interaction can cause ROS production [49], which is the major stimulator for ERS [50]. Liang et al. suggested that in insulin-resistant β-cells, the expression of ERS apoptotic-related proteins, such as CHOP as well as caspase-12, can be upregulated by high levels of T3, accounting for cell apoptosis through ERS [51]. According to this, a lower concentration of T3 in protecting cell apoptosis maybe attributed to inactive ERS.
Additionally, we demonstrated that TNF-α could down-regulate Bcl-2 expression in neonatal rat ventricular cardiomyocytes. Cell fate is always influenced by the apoptosisrelated signals, for example, Bcl-2 family members [52]. In response to deleterious events, the Bax can be activated and exhibit an amino acid homology and form heterodimers with Bcl-2 in the mitochondrial outer membrane to present with a pro-apoptotic function [53].
Conclusions
Our study suggested that HF status-mediated inflammatory responses could lead to thyroid dysfunction, which further aggravates HF progression. Additionally, low T3 concentrations could alleviate cardiomyocyte apoptosis. Besides, NLR can be a strong and independent poor predictive marker in thyroid dysfunction patients with pre-existing HF.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/jcdd9090290/s1, Table S1: Interaction analysis of NLR, TT3 and FT3. Informed Consent Statement: Written informed consent has been obtained from the patients to publish this paper.
Data Availability Statement:
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Acknowledgments:
The authors wish to express their gratitude for the help and support provided by the staff at the Second Affiliated Hospital, Zhejiang University School of Medicine.
Conflicts of Interest:
The authors declare no conflict of interest. | 2022-09-03T15:11:08.496Z | 2022-08-31T00:00:00.000 | {
"year": 2022,
"sha1": "208a6715d5e8d3cf4e016945e9ec4ee919ffd07d",
"oa_license": "CCBY",
"oa_url": "https://www.mdpi.com/2308-3425/9/9/290/pdf?version=1661939105",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "0ffe3f1cdd5d3446ffb193a13499bd7ed70b70d7",
"s2fieldsofstudy": [
"Medicine",
"Environmental Science"
],
"extfieldsofstudy": [
"Medicine"
]
} |
261547258 | pes2o/s2orc | v3-fos-license | Recent progress in high-loading single-atom catalysts and their applications
Single-atom catalysts (SACs) attract significant attention owing to their high catalytic activity, high metal atom utilization efficiency, and well-defined and configurable active sites. However, achieving single-atom dispersion of active metals at high metal loadings remains challenging, limiting the performance of SACs in many practical applications. Herein, we provide a comprehensive review of recent methods developed for synthesizing high-loading SACs, critically exploring their advantages, limitations, and wider applicability. Additionally, we showcase the benefits of high-loading SACs in the oxygen reduction reaction (ORR), water electrolysis, photocatalytic hydrogen production and CO oxidation. Although great recent progress has been made in the synthesis of high loading SACs, simple and universal routes that allowed the pre-programmed preparation of single metal and multi-metal SACs with specific metal coordination need to be discovered.
Introduction
2][3][4][5] As early as 2011, Zhang and his collaborators 6 University.Subsequently he worked as a postdoctoral researcher in the Westlake University, the University of Auckland (UoA, New Zealand) and the Technical Institute of Physics and Chemistry, Chinese Academy of Sciences.8][9] SACs have many advantages over traditional metal nanoparticle-based catalysts.Firstly, their active sites, which typically consist of a metal atom coordinated by 2-5 heteroatoms (N, O, S, B or P) on a support, can be pre-programmed, allowing researchers to explore reaction mechanisms at the molecular level.Secondly, SACs have very high metal atom utilization efficiency (∼100% in some cases) and ultra-high reactivity, surpassing the performance of metal nanoparticle-based catalysts on both counts. 102][13][14][15] For the above reasons, SACs have become a main focus of many research groups, leading to rapid advancements in the field in the past few years. 16For example, Xin and coworkers 17 recently used dissolution and carbonization methods to synthesize/characterize monometallic SACs for 37 different metal elements, establishing the largest SAC library reported to date.This work is significant as it allows benchmarking of SACs for different applications, guiding rational improvements in catalyst design.
In SACs, single atoms are very active and have high surface free energies. 18Accordingly, to stabilize the atomically dispersed metals against aggregation, the local coordination of the metal atoms is important.Most studies to date relating to SACs utilize N-doped carbon supports, with the isolated metal atoms typically having MN 4 , MN 3 or MN 2 first coordination shells.Such SACs are typically prepared by pyrolysis routes using metal salts and nitrogen/ carbon-rich precursors, with the metal contents generally kept low in order to avoid metal agglomeration into nanoparticles.This limits the SAC loading in the derived M 1 -N-C catalysts to typically a low weight percent.0][21] For example, Sa et al. reported a "silica protective layer-assisted" strategy that effectively prevented the aggregation of Fe atoms during hightemperature pyrolysis synthesis of Fe 1 -N-C catalysts.However, due to the low loading of Fe as the active centre (1.9 wt%), the overall performance of the catalyst was not greatly improved. 22Researchers are now targeting highloading SACs with metal loadings >5 wt%, aiming to significantly boost the overall activity of the catalyst through increasing the number of available active sites. 23Table 1 summarizes recent research towards high-loading SACs for specific applications, with most studies realizing a metal loading greater than 5 wt%.
A further advantage of preparing high-loading SACs is synergistic effects between adjacent metal atoms, which can have a dramatic effect on catalytic activity and selectivity. 24,25lectronic and spin interactions between nearby atomically dispersed metal atoms can modulate and enhance SAC performance, stimulating new research into dual metal single atoms catalysts.An overarching goal of current SAC research is the discovery of novel synthetic strategies for the preparation of high-loading SACs with high activity, selectivity, and stability (in particular suppressing metal agglomeration). 26,27][30] In this review, we discuss and summarize the main synthesis methods used to prepare high-loading SACs, such as pyrolysis methods, wet chemistry methods, electrochemical methods, atomic layer deposition (ALD) method, and so forth.Then we explore how the adjustment of some specific parameters in these classic synthesis methods enables the synthesis of high-loading SACs. 48,49The review then explores some important applications of highloading SACs, including the oxygen reduction reaction (ORR), water electrolysis, photocatalytic hydrogen generation, CO oxidation, and other catalytic reactions.Finally, the challenges and future prospects of high-loading SACs in industrial catalytic applications are examined.
Synthetic methods towards highloading SACs
The synthesis of SACs requires care to avoid metal agglomeration as much as possible. 50,51The dispersion or agglomeration of metal atoms involves competition between the relative strengths of the metal-support interactions and metal-metal bonds. 41The nature of the support and its affinity for particular metals are therefore vital to realizing SACs with high-loadings.A strong metal-support interaction is vital to achieving a high metal atom loading and for enhancing catalytic activity. 52,53The main purpose of this section is to provide an overview of the different synthesis methods commonly used to prepare SACs and then to explore whether these methods can be modified for the synthesis of high-loading SACs.
Industrial Chemistry & Materials
Mini review
Pyrolysis methods
Pyrolysis methods are widely used for SAC synthesis (especially M 1 -N-C catalysts), with precursor heating under an inert atmosphere creating N-C catalysts with sites for immobilizing metal atoms.However, during the pyrolysis process, metal atoms are highly susceptible to agglomerate to form nanoparticles, with post-synthetic treatments in acid often used to selectively dissolve the metal nanoparticles to leave only SACs. 43How to effectively design precursors to achieve physical or chemical confinement of metal atoms and avoid laborious post-synthetic treatments is an important challenge for the preparation of high-loading SACs. 32,42,54,55he use of metal-organic frameworks (MOFs) as precursors allows the synthesis of SACs with spatially separated metal sites, tunable structures, and flexible coordination geometries.For example, Qiao et al. revealed the formation mechanism of high-density single atoms with electronic metal-metal interactions using in situ spectroscopy characterization studies and theoretical calculations. 24Using ZIF-67 as a starting material, they discovered that highdensity Pt ions could be embedded into the ZIF-67 skeleton through ion exchange with Co to form Pt-N coordination sites.This process was in line with the hard and soft acidbase theory (HSAB), which states that ion exchange can occur when the acid-base interaction between the introduced metal and ligand is stronger than the host metal-ligand interaction.During the subsequent pyrolysis process, the Pt-N bond in the first shell breaks at lower temperatures and evolves into a Pt-O-Pt coordination, followed by Co-N cleavage to form Co-O, which promotes the formation of high-density Pt SACs on a Co 3 O 4 support (Pt loadings up to 6 wt% or 1.1 at%) with Pt-Pt interactions between adjacent Pt atoms.The SAC loading achieved was higher than most reported Pt-based ORR catalysts.In addition, Sun et al. prepared a layered g-C 3 N 4 carrier (NGK) from kaolin by thermal oxidation calcination. 42Subsequently, Fe ions were assembled into the "sixfold cavities" of NGK through high-temperature treatment, allowing Fe atoms to combine with N atoms and form a stable atomic structure, as shown in Fig. 1a.Through introducing kaolinite, g-C 3 N 4 and nitrogen vacancies, the anchoring efficiency of Fe single atoms was enhanced, thus avoiding Fe agglomeration.Thanks to the carrier effect of kaolinite, the load of Fe single atoms was successfully increased to 14.2 wt% (∼3.4 at%).The obtained Fe singleatom composite (Fe SA-NGK) possessed high catalytic activity for the ORR.
To prepare edge-defect-rich high-loading Co SACs, Xu et al. developed a boric acid-assisted pyrolysis strategy (Fig. 1c). 56Using chitosan, cobalt nitrate, and boric acid as precursors, a Co-N-B-C catalyst rich in exposed sheet edges and with high Co loadings was synthesized by a one-step pyrolysis method.During pyrolysis, B 2 O 3 was formed by the dehydration of boric acid.Since the temperature in the pyrolysis was much higher than the melting point of B 2 O 3 , B 2 O 3 remained in a liquid state.Chitosan carbonization thus occurred within the molten B 2 O 3 medium, yielding a N-doped carbon skeleton with abundant pores and rich in edge defects after washing to remove B 2 O 3 .The use of molten B 2 O 3 was very effective in preventing Co single atoms from agglomerating.The reported Co monoatomic loading was 4.2 wt% (∼0.9 at%), though it is expected that much higher Co SA loadings could be achieved by optimization of this novel method.To further improve the loading of Co SAs, Wang et al. mixed dicyandiamide, formaldehyde, and Co(NO 3 ) 2 •6H 2 -O in distilled water, stirred them vigorously, then heated the mixture at 100 °C for more than 12 h to completely evaporate the solvent. 57In this process, formaldehyde and dicyandiamide were polymerized, whilst Co ions were fully coordinated by the obtained dicyandiamide-formaldehyde resin.The purple solid was then heated in a tubular furnace to 600 °C under Ar and held at that temperature for 2 h, then cooled to room temperature and heated to 400 °C in a H 2 /Ar (5%) atmosphere for 2 h.The Co SAC prepared by this method had a remarkable Co loading of 23.58 wt% (∼5.9 at%) and very uniform Co dispersion (Fig. 1b), representing a notable breakthrough in the synthesis of high-loading SACs.
Nowadays, pyrolysis methods remain very popular for synthesizing SACs due to their simplicity.The above examples demonstrate that by judicious selection of precursors and pyrolysis parameters, it is possible to reduce metal agglomeration and greatly increase the overall loading of atomically dispersed metals.Pyrolysis is likely to remain one of the most important methods for the synthesis of highloading SACs in the future.
Wet chemical methods
Among traditional methods used for SAC synthesis, wet chemical methods are indispensable (especially in the field of catalysis) due to their simple operation and the possibility of mass production.Wet methods include coprecipitation methods, impregnation methods, and ion exchange methods, amongst others.Generally, these methods consist of the following steps: i) The metal precursor is attached to the surface of a carrier through impregnation, adsorption, ion exchange, coprecipitation, or other methods; ii) The modified carrier is dried or heat treated (sometimes this step can be omitted); iii) Reduction or activation.Generally, the reaction conditions used in wet chemical methods are mild and more easily controlled compared to those of pyrolysis methods.However, many parameters in the preparation process influence the performance of the obtained SACs.The parameters need to be strictly controlled, including the addition speed of the precursor, stirring speed, reaction temperature, reaction time, reduction/activation conditions, etc.In addition, some metal atoms loaded onto the carrier may aggregate, making it difficult for them to be transformed into single atoms.High-loading SACs are generally difficult to prepare using wet methods.
In 2020, Wu et al. prepared a kind of metal (M) centered catalyst (M = Cu, Pt, Pd, etc.) with sulfur (S) and nitrogen (N) coordination by cation exchange (Fig. 1d). 58The synthesis method involved converting the surface part of CdS nanorods into subnano/atomic layer Cu 2 S via cation exchange.Then, a 3-aminophenol/formaldehyde (3-AF) layer was deposited on the CdS/Cu 2 S surface and the resulting sample was heated at 900 °C under an N 2 atmosphere.The 3-AF layer was converted into a nitrogen-doped carbon (NC) shell in situ.Then Cd 2+ was gradually reduced by the NC and evaporated due to the low boiling point of Cd.The volatile S species selectively vulcanized the NC.The Cu species diffused into the edge-enriched S and N defects to form a monatomic copper catalyst.Lang et al. added a mixture of chloroplatinic acid and ferric nitrate into a sodium carbonate solution dropwise.Via this co-precipitation method, a Pt 1 /FeO x catalyst with a Pt loading up to 1.8 wt% was finally prepared. 59As shown in Fig. 1e, Pt was uniformly dispersed on the FeO x support in the form of single atoms.Even with heating at 800 °C, the dispersed Pt atoms were resistant to agglomeration.
The wet chemical method remains common for SAC preparation, with the SACs prepared by this method being characterized by good metal dispersion and low susceptibility to forming metal clusters.However, the loading of SACs prepared by this method is difficult to increase.Studies concerning the synthesis of high-loading SACs by the wet chemical method are still relatively rare.
Atomic layer deposition (ALD)
1][62] It can be used for the preparation of uniform thin films, single atoms of regular dispersion, subnano clusters, and nanoparticles (NP) in porous materials.Compared with other deposition techniques such as chemical vapor deposition (CVD), physical vapor deposition (PVD), or electrochemical deposition, ALD possesses significant advantages, which are summarized as follows: 63 i) the amount of deposited metal can be accurately controlled; ii) metal atoms are evenly dispersed; iii) the method exhibits surface chemical selectivity; iv) the method can be scaled and applied industrially.
In an early study, Sun et al. used (methylcyclopentadienyl)-trimethyl platinum as the precursor to deposit platinum on a graphene carrier via ALD under nitrogen purging (Fig. 2a). 64During the oxidative decomposition of MeCpPtMe 3 , some precursor ligands reacted with surface adsorbed oxygen to form CO 2 , H 2 O, and hydrocarbon fragments.This provided a self-limiting growth energy sufficient to construct a platinum-containing monolayer during ALD.More oxygen adsorbed on the Pt surface to form a new adsorbed oxygen layer, thus completing an ALD cycle.By this approach, the amount of Pt deposition could be accurately controlled by adjusting the number of ALD cycles.The average size of Pt particles on graphene was 0.5, 1-2, and 2-4 nm for 50, 100, and 150 ALD cycles, respectively.Pt loading on graphene was 1.52 wt% (∼0.09 at%), 2.67 wt% (∼0.17 at%) and 10.5 wt% (∼0.72 at%), respectively.In these samples, Pt co-existed in the form of single atoms, small Pt clusters, and Pt NPs.Gu et al. adopted atomic layer deposition (ALD) to deposit monatomic copper and nickel on a carrier (Fig. 2b). 65The developed experimental method involved first making an atomic copper "gripper" on g-C 3 N 4 by Cu ALD (also known as Cu 1 /g-C 3 N 4 ), where the saturation Cu loading was about 11.2 wt%.Subsequently, Ni y Cu 1 /g-C 3 N 4 (y is the atomic ratio of Ni to Cu) was synthesized by depositing Ni atoms on subsaturated Cu 1 /g-C 3 N 4 by NiO x ALD.It was shown that the Ni loading decreased with increasing Cu loading, demonstrating that open N py sites together with adjacent Cu atoms provide anchoring sites for guest atoms (i.e.Ni) at subsaturation Cu coverage, similar to chelating ligands in organometallic chemistry.Using this approach, a compact atomic copper "gripper" (8.1 wt%/∼0.17at%) on g-C 3 N 4 was used to stabilize a relatively high-loading of monatomic nickel catalyst (3.1 wt%/∼0.65 at%).Thanks to the atomic layer deposition method and the metal-carrier interaction, SACs with high metal loadings were obtained.
To sum up, ALD is an emerging technology for the synthesis of SACs, clusters, and nanoparticle catalysts.It is expected to play an increasingly important role in the preparation of high-loading SACs for different applications in the future.
Electrochemical methods
Compared with thermochemical methods, electrochemical methods represent a milder, more environment-friendly and energy-efficient route towards high-loading SACs.Generally, electrochemical syntheses of SACs are carried out at near ambient temperature and pressure.Also, the synthesis parameters can be fine-tuned by changing the input electrochemical technique (e.g., linear sweep or cyclic voltammetry, CV) or by continuously adjusting the applied current, voltage, sweep rate, and reaction time to obtain the desired product.By this electrodeposition approach, Pt atoms could be immobilized on carbon nanotubes or graphite.In addition, Pt SACs on CoP-based nanotube arrays (supported on nickel foam) were successfully prepared by controlled potential cycling. 67Zhang et al. used a standard three-electrode system to grow Co(OH) 2 nanosheets on a glassy carbon working electrode. 68By adding a low concentration of IrCl 4 to 1 M KOH electrolyte, a C-Ir 1 / Co(OH) 2 catalyst was successfully obtained over ten scanning cycles.The monatomic Ir was evenly dispersed on the Co(OH) 2 substrate, with the mass loading of Ir around 2.0 wt%, verified by inductively coupled plasma atomic emission spectrometry (ICP-AES).
Electrochemical methods represent a relatively new SAC synthesis method, being both simple and easily controllable.However, similar to wet chemical methods, achieving high single atom loadings remains challenging, motivating the search for improved synthetic methodologies.
Other methods
In addition to the methods discussed above, there are many other methods now being used to prepare SACs, including electrospinning, ion exchange, click chemistry, mass-selected soft-landing methods, and so on. 48uan's team reported an interlayer nano spatial restriction strategy in 2019 for the first time (Fig. 2c). 33A simple electrostatic ion exchange method was used to embed potassium ions into graphitic carbon nitride (g-C 3 N 4 -K).Then, the prepared g-C 3 N 4 -K was mixed with Pt(NH 3 ) 4 Cl 2 solution to load Pt onto the surface of the material.Here, the negatively charged g-C 3 N 4 -K induced the spontaneous adsorption of Pt ions via electrostatic attraction.Through ion exchange with K + , Pt ions were inserted into the interlayer nanospace and confined by the adjacent layers.Finally, the obtained yellow powder was heated in an Ar atmosphere to enhance the interaction between Pt atoms and the support.The loading of Pt could be controlled by adjusting the amount of Pt(NH 3 ) 4 Cl 2 solution used in the synthesis, with the maximum loading of Pt single atoms reaching 10.4 wt% (∼0.71 at%).
The click chemistry-limited strategy devised by Zhao et al. differs from conventional chamber-limited and lattice-limited strategies (Fig. 2d) in that not only does it ignore strict molecular size effects or symmetry requirements, it also facilitates the controlled and directed synthesis of catalysts. 69he authors successfully prepared Co-N-C SACs by clicking cobalt porphyrin units onto conducting substrates via amination reactions.Under the guidance of click chemistry, metal-containing complexes are attached to modified substrates to ensure the effective confinement of metal atoms in terms of precursor design for the construction of single-atom sites.Considering the high reaction specificity between carboxyl and amino groups, the authors chose the amidation reaction as the click reaction to graft transition metalcontaining complexes onto conducting substrates.The use of cobalt(II) meso-tetra(4-carboxyphenyl)porphyrin and amino functionalized carbon nanotubes (CNT amino) allowed the efficient synthesis of Co SACs.Xia et al. used a modified molecular fusion route to synthesize well-crystallized GQDs-NH 2 dispersions. 32They added different volumes of IrCl 3 stock solution (5 mg mL −1 ) to 30 mL of the GQDs-NH 2 dispersion (1 mg mL −1 ), then freeze-dried the resulting dispersion.Pyrolysis of the freeze-dried product created an Ir SAC with Ir loading up to 41.6 wt% (3.84 at%).Based on that, the same method was used to prepare Ni SACs with a high atomic Ni loading amount reaching 15.4 wt% (3.6 at%). 32
Synthesis strategies and applications for different methods
There are several strategies to achieve high-loading SACs.First, defect engineering, though constructing more diverse and dense defect sites on the surface of the carrier to capture single atoms, thereby synthesizing high-loading single atom catalysts.Second, the space limitation strategy.It rationally uses diverse porous materials (ZIF, MOFs, COFs, etc.) to encapsulate and anchor mononuclear metal precursors, and achieve high loading and their uniform spatial distribution.Finally, for regulating coordination sites and coordination groups on the surface of the support to capture mononuclear metal precursors, a variety of coordination sites with different coordination properties are used to regulate the coordination environment of single atoms.Different ligands can provide different electron densities and coordination forces to regulate the catalytic performance of the active site.For example, the addition of a complex can make it form coordinate covalent bonds with metal ions, and wrap the metal ions.The metal ion is fixed by providing a site for the metal ion to form a coordinate covalent bond with, so as to prevent it from further reacting with other ions or molecules, thus stabilizing the metal ion and preventing it from agglomerating or aggregating in the reaction process.
After analyzing the synthesis strategies for the preparation of SACs, different synthesis methods can successfully prepare high-loading SACs through reasonable application of these strategies, and the industrialization prospects of these methods are discussed.First, the pyrolysis method can achieve excellent atom dispersion and has shown promise in industrial applications.However, the high temperature and energy requirements can make it costlier compared to wet chemistry methods.Additionally, the scalability of hightemperature pyrolysis needs further exploration.Wet chemistry methods, such as sol-gel and impregnation techniques, have been widely used.They are relatively simple and cost-effective, making them attractive for large-scale production.However, achieving high dispersion and stability of single atoms can be challenging.The electrochemical method for preparing SACs has strong industrialization prospects due to its advantages in terms of scalability, costeffectiveness, and market benefits.It can be easily scaled up using conventional electrochemical cells and equipment.This allows for potential large-scale production of SACs, meeting the demands of various industrial applications.Also, the electrochemical method uses relatively low-cost and abundant metal precursors.These factors contribute to the potential cost advantage of electrochemically synthesized high-loading SACs.However, ongoing research and development efforts to address challenges and optimize synthesis parameters will be crucial in realizing the full potential of electrochemically synthesized high-loading SACs in various industrial applications.Atomic layer deposition (ALD) offers great potential for atomically precise catalysts, but the complexity of synthesis and the high cost of precursor materials and equipment may limit their industrial scalability.
Considering these, wet chemistry methods currently appear to have great industrialization prospects for SACs.They are relatively cost-effective and can be easily scaled up for mass production.However, the loading of SACs prepared by this method is difficult to increase.Studies concerning the synthesis of high-loading SACs by wet chemical methods are still relatively rare.In the actual research and application, the pyrolysis method to produce high loads of single atoms still occupies a dominant position, but it is hoped that a mild and controllable wet chemical method can be developed in the future.
Oxygen reduction reaction (ORR)
The ORR reaction is a key reaction in a number of nextgeneration electrochemical devices including fuel cells and metal-air batteries. 70,71In a proton exchange membrane fuel cell (PEMFC), the design of the cathode "oxygen reduction" electrocatalyst is more challenging than the anode "hydrogen oxidation" electrocatalyst. 72This is due to the following reasons: i) oxygen reduction is a four-electron transfer process that requires fast reaction kinetics, and its exchange current density is only 1/100th that of the anodic hydrogen oxidation reaction (a two-electron transfer process).As such, the ORR is five orders of magnitude slower than the anodic hydrogen oxidation reaction.As such, the ORR is the rate limiting step in the overall PEMFC electrocatalytic reaction; ii) the oxygen reduction process is relatively complex, involving many proton and electron transfer steps and multiple intermediates, resulting in a lower energy conversion efficiency and increased oxygen reduction overpotential.3][74] In a wide variety of electrocatalysts, high-loading SACs can enhance the catalytic activity and selectivity of the ORR compared to conventional catalysts.The unique atomic dispersion of metal atoms in SACs can improve the efficiency of oxygen reduction, leading to improved fuel cell performance and energy conversion.
Pt-nanoparticle based electrocatalysts have excellent catalytic properties and efficiencies as ORR catalysts. 4By comparison, single-atom Pt catalysts are typically ineffective in catalyzing the breakage of O-O bonds during ORR catalysis, and thus they cannot catalyze the four-electron reduction of O 2 efficiently.At low metal loadings, Pt SACs typically favor twoelectron products (H 2 O 2 ) rather than four-electron products (H 2 O). 75Ambarish Kulkarni et al. mentioned that the essence of the four-electron pathway is to allow O-O bond dissociation in the adsorbed *OOH. 75Lei Zhang concluded that when the distance between Pt atoms is greater than a few angstroms (Å), neighboring Pt single-atom sites exhibit no significant interaction. 76In this case, each Pt single-atom site remains isolated and unsupported in ORR catalysis.When the distance between two Pt single atoms is approximately 3 Å, both oxygen atoms in *OOH can be adsorbed by these two Pt atoms, effectively stretching the O-O bond and activating its cleavage, and exhibiting superior 4-electron ORR performance.Therefore, when designing atomically dispersed Pt catalysts for 4-electron ORR electrocatalysis, achieving a high metal density is crucial. 72,75Li et al. applied a photochemical solid-phase reduction method to prepare high-loading Pt SACs (3.8 wt%, ∼0.24 at%). 77The catalyst exhibited excellent HER and ORR electrocatalytic properties in acid and alkaline electrolytes.The catalyst delivered a high half-wave potential of up to 0.89 V, along with high specific mass activity and specific activity, low Tafel slope, low hydrogen peroxide yield and high stability, thus superior all-round performance compared with a commercial 20 wt% Pt/C catalyst containing platinum nanoparticles (Fig. 3a-f).Although a high-loading Pt SAC has excellent 4-electron ORR performance, the limited reserves of Pt and its high price are obstacles to the large-scale uptake of PEMFCs.Further, supported Pt nanoparticles usually undergo agglomeration or dissolution during electrocatalytic processes, which can be highly detrimental to performance.To address these problems, researchers are now pursuing non-precious metal SACs for the ORR and other electrochemical reduction reactions that traditionally have used Pt nanoparticle-based catalysts.For example, by codoping carbon materials with nitrogen and transition metals (TM), non-precious metal SACs with excellent electrocatalytic ORR activity can be created, serving as cathodic catalysts for proton exchange membrane fuel cells.In addition, their ORR activity can be improved by increasing the loading of TM SAC active sites.The catalytic activity of SACs depends on both the intrinsic catalytic activity and density of the active sites.In general, the intrinsic activity of TM single atom sites for a particular TM is relatively fixed, thus increasing the active site density is considered the most effective strategy for improving the catalytic activity in acidic electrolytes.Kucernak et al. prepared monoatomic Fe catalysts supported on N-doped carbon (Fe-NC) with a Fe loading up to 7 wt% (∼1.59 at%) using a ZIF-8 precursor. 44The presence of Fe single atoms was verified by 58 Fe low-temperature Mössbauer spectroscopy and Fe K-edge X-ray absorption spectroscopy, both of which demonstrated a Fe-N 4 coordination.The catalyst exhibited excellent catalytic activity for the ORR in fuel cells.Xia et al. prepared Ni-N-C catalysts using a selfassembly method with graphene quantum dots, achieving Ni loadings up to 15.4 wt% (3.6 at%). 32The Ni atoms were atomically dispersed on the carrier with no agglomeration being observed.Excellent ORR activity was reported.These transition metals such as Fe, Co, Mo and Ni have been prepared to form high-loading SACs that demonstrated excellent catalytic performance for the ORR (Fig. 3g-i).Xin et al. prepare a high-loading Mo SAC. 39A precursor powder containing ammonium molybdate ((NH 4 ) 6 Mo 7 O 24 ) and glucose was placed in a crucible and heated at 650 °C for 4 h under an Ar atmosphere.The obtained Mo SAC had a metal loading of 9.54 wt% and very good electrocatalytic properties.
In conclusion, high-loading SACs are very promising alternatives to traditional Pt nanoparticle catalysts for the ORR in PEMFCs and other applications.Much work continues, aimed at increasing the site density of single atoms on carbon supports, essential for realizing fast reaction kinetics.
Photo/electro-catalytic water splitting
Hydrogen is a promising energy carrier in the shift away from polluting fossil fuel energy.Water electrolysis using renewably generated electricity is arguably the cleanest way to manufacture hydrogen fuel at scale, with the only byproduct being oxygen. 70,78The reaction rates of the cathode hydrogen evolution reaction (HER) and anodic oxygen evolution reaction (OER) both affect the overall electrolytic water reaction efficiency.Nowadays, the main challenge of water electrolysis is to reduce the overpotentials of the HER and OER. 79Researchers are now exploring various types of catalysts for the HER and OER in order to lower the overpotentials of these reactions.Platinum and certain precious metal oxides (RuO 2 and IrO 2 ) represent the benchmark HER and OER catalysts.SACs are now being pursued as lower cost alternatives to these precious metal nanoparticle-based catalysts for both the HER and OER. 80nd high-loading SACs have shown promise in improving the catalytic efficiency and stability of water splitting reactions.They can provide active and durable catalytic sites, enhancing hydrogen production and oxygen evolution kinetics.
For example, Chen et al. used nickel foam as the substrate and synthesized atomic ruthenium (Ru)-loaded nickel hydroxide ultrathin nanoribbons (R-NiRu) with a high atomic Ru loading amount reaching ∼7.7 wt%, exhibiting a low overpotential of 16 mV for the HER at 10 mA cm −2 and a Tafel slope of 40 mV dec −1 in aqueous 1.0 m KOH solution. 81n addition, Cheng et al. constructed catalysts with highdensity Pt (1.9 at%) and Ir (2.6 at%) single atoms anchored on Co(OH) 2 by a facile one-step approach.Remarkably, Pt 1 / Co(OH) 2 and Ir 1 /Co(OH) 2 only required 4 and 178 mV at 10 mA cm −2 for the HER and OER, respectively. 35In addition, the assembled Pt 1 /Co(OH) 2 //Ir 1 /Co(OH) 2 system showed a mass activity of 4.9 A mg noble metal −1 at 2.0 V in an alkaline water electrolyzer, which is 316.1 times higher than that of Pt/C//IrO 2 .With a view towards direct solar energy conversion in hydrogen fuels, researchers are also pursuing photocatalytic hydrogen production.Zeng et al. constructed Pt single-atom photocatalysts with ultra-high Pt loading (8.7 wt%/∼0.58at%) using the interlayer sub-nanospace of layered carbon nitrides to confine Pt atoms. 33Both theoretical calculations and experimental results showed that the interlayer interactions can significantly change the electronic structure of Pt atoms, causing the charge density of confined Pt atoms to change and facilitate the adsorption of protons, significantly reducing the energy barrier of photocatalytic hydrogen generation.The prepared Pt single-atom photocatalyst delivered a high photocatalytic H 2 production rate of 22 650 μmol g −1 h −1 under visible light and an apparent quantum yield (AQY) of 22.5% at 420 nm, which was much higher than most g-C 3 N 4 and polymeric-based catalysts (Fig. 4a-d).In addition, it has been shown that surfactants can play an important role in the synthesis of high-loading SACs.Zhou et al. prepared a Pt SAC with a metal loading of 12.0 wt% (∼0.83 at%) with the help of surfactants. 34In their work, polyvinylpyrrolidone (PVP) molecules were selectively adsorbed on the surface of MOF sheets to form 2D MOF nanosheets.The prepared PtSA-MNS (ultrathin 2D MOF nanosheets coordinated with Pt single atoms) exhibited excellent photocatalytic activity for hydrogen generation under visible light and good stability (Fig. 4e and f).
Industrial Chemistry & Materials Mini review
These examples demonstrate that high-loading SACs can be beneficial for hydrogen generation via water electrolysis and photocatalysis, with high-loading platinum SACs showing excellent activity for both of these processes.
CO oxidation reaction
The CO oxidation reaction is an important reaction, 82 having relevance to direct methanol fuel cells, neutralization of vehicle exhaust emissions, and removal of trace amounts of CO from hydrogen-rice gas.SACs provide an ideal platform for studying the reaction mechanism of CO oxidation at a molecular level, whilst also offering the ability to greatly accelerate the CO oxidation reaction.4][85] Based on that, highloading SACs can exhibit higher catalytic activity and selectivity than traditional catalysts in CO oxidation.The atomic-level dispersion of metal atoms in SACs provides efficient exposure of active sites, leading to enhanced CO oxidation performance and better tolerance to poisoning.
Pt-based catalysts have been widely studied for CO oxidation.By loading Pt single atoms on the surface of metal oxides, each Pt atom can be used as an active site for the reaction, thereby offering great advantages over Pt nanoparticle catalysts in terms of Pt atom utilization efficiency.In fuel cells, Pt nanoparticle catalysts are vulnerable to CO poisoning, 86 making the CO oxidation reaction very important.Pt SACs are not easily poisoned by CO, offering a great advantage over their nanoparticle counterparts.In traditional nanoparticle catalysts, CO molecules can bind to multiple metal atoms, inhibiting their ability to facilitate catalytic reactions.However, in Pt SACs, each metal atom is isolated and surrounded by ligands or cocatalysts, preventing CO molecules from adsorbing onto multiple metal sites.This unique atomic-level dispersion of Pt atoms reduces the probability of CO poisoning.Furthermore, the ligands or co-catalysts surrounding the Pt atoms in SACs can help stabilize the catalyst and modify its electronic structure, optimizing its activity and selectivity.This allows for enhanced catalytic performance and increased resistance to CO poisoning.Recently, Kim and his colleagues used CeO x -TiO 2 (two-component oxides) to support a compact Pt single (or double) atom catalyst, obtaining a catalyst with excellent CO oxidation performance by adjusting the interface between the two oxide phases. 87The CeO x -TiO 2 interface was created by adding 1 wt% Ce on TiO 2 nanoparticles, which stabilizes Pt single-atoms by strong electronic interactions, physically disintegrates Pt into Pt-SAs and activates the interface-mediated MvK mechanism of CO oxidation, while preserving its catalytic activity against COpoisoning.Also, the author found that the loading of Pt SACs with addition of 1 wt% of Ce to the TiO 2 support is higher than that of the TiO 2 support, resulting in superior ORR activity (Fig. 5a-c).
In order to overcome the scarcity and high cost of platinumbased catalysts, researchers are now pursuing highperformance and high-loading SACs based on earth abundant metals for CO oxidation.Zhu et al. systematically studied the geometric structure, electronic structure, and stability of various monatomic metals supported by oxygen functionalized Ti 2 C (Ti 2 CO 2 ) M 1 /Ti 2 CO 2 (M = Fe, Co, Ni, Cu-Ru, Rh, Pd, Ag-O s , Ir, Pt, Au) using density functional theory. 88It was found that an Fe 1 /Ti 2 CO 2 SAC showed excellent catalytic performance for CO oxidation at low temperatures.The adsorption of O 2 and CO on the Fe 1 atom of Fe 1 /Ti 2 CO 2 was extremely favorable (Fig. 5d-g).This work guides the experimental synthesis of lowcost SACs for CO oxidation.
Other catalytic reactions
In addition to the above applications, high-loading SACs have also been applied in many other catalytic reactions, such as oxidative conversion of methane, organic synthesis, CO 2 reduction and other applications. 8,89,90The advantages of high-loading SACs, such as improved atom dispersion, catalytic activity, and selectivity, make them valuable in a wide range of catalytic processes.
According to the International Energy Agency (IEA), the global emission of CO 2 from the burning of fossil fuels was estimated at 33.8 billion tons in 2022, 300 million tons more than the previous year.In order to reduce global CO 2 emissions, recycling of CO 2 is essential.High-loading SACs can play a role in CO 2 RR by improving the activity, selectivity, and stability of the catalyst.SACs can offer precise control over the reaction intermediates, allowing for the production of specific products, such as hydrocarbons or methanols.Converting CO 2 into methanol has high economic value.Fig. 6a shows the mechanism of photocatalytic CO 2 reduction to methanol.The reaction to produce methanol from CO 2 is a six-electron reduction process.Therefore, the photocatalyst activity is closely related to the localization of photogenerated electrons on the catalyst surface.Ma et al. loaded an ultrahigh density of Co-N 2 C single-atom active sites on a g-C 3 N 4 photocatalyst, achieving a very high loading of 24.6 wt% (∼6.2 at%). 91Photocatalytic tests showed a methanol generation rate of 941 μmol g −1 over 4 h, which is 13.4 and 2.2 times higher than pure phase g-C 3 N 4 (17.7 μmol g −1 ) and CoO x /g-C 3 N 4 -0.2 (423.9 μmol g −1 ), respectively.In addition, the photocatalytic activity of the Co/g-C 3 N 4 -0.2SAC did not decrease significantly after 12 cycles (∼48 h).Besides being able to convert CO 2 into methanol, reducing it to commercial carbon monoxide is also a good choice.Zhao et al. used a general cascade anchoring strategy for the mass production of a series of M-NC SACs with a metal loading up to 12.1 wt%. 38Among them, the Ni-NC SAC exhibited the potential for CO 2 reduction to CO and showed an 89% Faraday efficiency at −0.85 V with 30 mA cm −2 current density for CO.
In conclusion, high-loading SACs are now being widely used in photocatalysis, electrocatalysis and thermal catalysis. 92,93It's important to note that while high-loading SACs can offer advantages in these applications, there may still be challenges to overcome, such as stability under harsh reaction conditions or the requirement for specific support materials.Further research and development efforts are ongoing to optimize the design, synthesis, and utilization of high-loading SACs in different catalytic reactions.We hope that high-loading SACs will find end use in the chemical industry and energy sector.
Summary and outlook
This article summarizes recent advances in the synthesis and application of high-loading SACs, highlighting ways in which the electronic structure of the catalysts, active site density, and ultimately catalytic activity can be improved by maximizing the metal atom loading. 94Thereafter, we explored applications of high-loading SACs in different catalytic reactions, including the ORR, water electrolysis, photocatalytic hydrogen production, CO oxidation reaction, and other catalytic reactions.6][97] The controlled preparation of SACs with high loading still presents significant challenges. 98,99Particular issues that need to be resolved include: i.The best way to describe the SAC loading has been disputed for a long time.Most of the scholars prefer to report metal loadings in SACs as a weight percentage (wt%).However, atom percentage (at%) actually depicts the number of single atomic active sites in SACs much more accurately (since at a fixed at% of metal, the wt% value will increase with the atomic number of the metal atoms which can lead to confusion).1][102][103] Alternatively, mass activity could be used to compare metal SACs in terms of their catalytic properties, which is particularly useful when comparing SACs and metal nanoparticle catalysts.
ii.Although many strategies have been discovered to prevent the agglomeration of single atoms during the preparation of high-loading SACs, aggregation of atoms is inevitable above certain threshold loadings.It is hoped in the future that even more effective strategies can be found to prevent agglomeration from happening.
iii.Although many methods have been found to synthesize high-loading SACs, often these methods make it difficult to prepare catalysts in large quantities.Universal methods for the large-scale and controlled preparation of SACs with high single atom site densities are urgently needed.Such universal methods are key to the industrial scale preparation of SACs with high loadings.
iv.The majority of SACs reported to date with high loadings are of the type M 1 -N-C (metal single atoms on N-doped carbons).For SACs based on oxide supports, metal loadings are typically much lower than M 1 -N-C catalysts, limiting the range of application of SACs in catalysis and other applications.More work needs to be directed towards SACs based on other supports (metal oxides, metal nitrides, metal carbides, metal sulfides, etc.).
v. When analyzing the industrialization prospect of different methods for preparing SACs, several factors need to be considered, including cost, scalability, performance, and market demand.Considering that, wet chemistry methods currently appear to have great industrialization prospect because they are relatively cost-effective and can be easily scaled up for mass production.However, the loading of SACs prepared by this method is restricted.Studies concerning this are still relatively rare.In the actual research and application, the pyrolysis method to produce high loads of single atoms still occupies a dominant position, but it is hoped that a mild and controllable wet chemical method can be developed in the future.
put forward the concept of SACs for the first time.Since that time, SACs have increasingly become the Jiahui Luo Jiahui Luo received her B.E. degree from Central South University of Metallurgical Engineering in 2021.She is pursuing her M.E.degree in Ganjiang Innovation Academy, Chinese Academy of Sciences under the supervision of Prof. Lishan Peng and Qingjun Chen.Her current research interest focuses on electrocatalysts for the oxygen reduction reaction Lishan Peng Lishan Peng is currently an associate professor in the Ganjiang Innovation Academy, Chinese Academy of Sciences.He obtained his Ph.D. degree in chemical Chen received his Ph.D from East China University of Science and Technology in 2012.After that, he worked as an assistant professor in Shanghai Advanced Research Institute, Chinese Academy of Sciences.From 2013 to 2021, he continued his research as a postdoc/ researcher/professor in the University of Toyama (Japan), Norwegian University of Science and Technology (NTNU, Norway), and Institute of Process Engineering, Chinese Academy of Sciences, successively.Since 2021, he has been a professor at the Ganjiang Innovation Academy, Chinese Academy of Sciences.His research interests are hydrogen energy and fuel cells.
Fig. 1
Fig. 1 (a) Schematic diagram for preparation of high-loading Fe SACs by pyrolysis.Reprinted with permission from ref. 42.Copyright 2022, Wiley-VCH GmbH; (b) STEM and EDX electron microscopy images of Co SACs with 23.58 wt% load obtained by pyrolysis.Reprinted with permission from ref. 57.Copyright 2023, Springer Nature; (c) Schematic diagram of a novel boric acid assisted pyrolysis strategy for producing SACs with abundant edge defects and high loading.Reprinted with permission from ref. 56.Copyright 2023, Wiley-VCH GmbH; (d) Schematic diagram of SACs prepared by cation exchange for sulfur-nitrogen dicoordination.Reprinted with permission from ref. 58.Copyright 2020, American Chemical Society; (e) Electron microscopy of the Pt 1 /FeO x catalyst prepared by the coprecipitation method.Reprinted with permission from ref. 59.Copyright 2019, Springer Nature.
Fig. 2
Fig. 2 (a) Schematic diagram of deposition of Pt by atomic layer deposition.Reprinted with permission from ref. 64.Copyright 2013, Springer Nature; (b) High-resolution HAADF-STEM image of Ni 1 Cu 2 /g-C 3 N 4 , where isolated atoms, triangular trimers and linear trimers are represented by yellow circles, red triangles and green rectangles, respectively.Reprinted with permission from ref. 65.Copyright 2021, Springer Nature; (c) Schematic diagram of Pt SAC and restricted Pt single-atom coordination structures prepared by an interlayer nanospace confinement strategy.Reprinted with permission from ref. 33.Copyright 2020, Elsevier; (d) Schematic diagram of three different limiting strategies (chamber-limited strategy, lattice-limited strategy and click-limited strategy).Reprinted with permission from ref. 69.Copyright 2022, American Association for the Advancement of Science.
Fig. 3
Fig. 3 (a-f) The ORR performance of single-atom platinum catalysts.Reprinted with permission from ref. 77.Copyright 2021, Elsevier; (g) The ORR activity of a single-atom Fe catalyst.Reprinted with permission from ref. 44.Copyright 2022, Springer Nature; (h and i) Steady-state ORR polarization curves and corresponding Tafel of Fe-NC SAC, Pt/C and control samples.Reprinted with permission from ref. 38.Copyright 2019, Springer Nature.
Fig. 4
Fig. 4 (a) Visible light (λ > 420 nm) photocatalytic HER activity of g-C 3 N 4 , Pt NPs/G-C 3 N 4 -K and SA Pt/g-C 3 N 4 under different Pt loads; (b) Photos of reaction using 10 mg SA Pt/g-C 3 N 4 -8.7;(c) Stability test and (d) SA Pt/g-C 3 N 4 -8.7 of the apparent quantum test.Reprinted with permission from ref. 33.Copyright 2020, Elsevier; (e and f) Characterization of photocatalytic activity, (e) photocatalytic production rates of H 2 of Pt SA MNS, Pt NP MNS and MNS and (f) cyclic stability of Pt SA MNS and Pt NP MNS.Reprinted with permission from ref. 34.Copyright 2019, Wiley-VCH GmbH.
Fig. 5
Fig. 5 (a) DFT calculation of the binding tendency of Pt SA to fully oxidized or reduced Ce ions at a CeO x -TiO 2 interface; (b) Mars-van Krevelen-type CO oxidation pathway catalyzed by Pt 2 clusters fixed at the CeO x -TiO 2 interface; (c) Results of CO-TPR analysis for 0.25 PT and 0.25 PCT catalysts.Reprinted with permission from ref. 87.Copyright 2022, The Royal Society of Chemistry; (d-g) DFT calculation of the binding energy PDOS diagram and PEDD diagram of metal (M) monoatom in M 1 /TiC 2 O 2 .Reprinted with permission from ref. 88.Copyright 2020, Elsevier.
Table 1
Summary of recently reported high-loading single atom catalysts (SACs) Tavakkoli et al. used bulk Pt foils and single-walled carbon nanotubes (SWCNTS) coated on glassy carbon electrodes as counter electrodes and working electrodes, respectively, to electrochemically deposit Pt atoms onto SWCNTS through a continuous voltage cycle. | 2023-09-06T15:09:50.740Z | 2023-01-01T00:00:00.000 | {
"year": 2023,
"sha1": "4c50d844b799b36f0fc088268c6af7a757507a0d",
"oa_license": "CCBYNC",
"oa_url": "https://pubs.rsc.org/en/content/articlepdf/2023/im/d3im00062a",
"oa_status": "GOLD",
"pdf_src": "Anansi",
"pdf_hash": "8af9c5919dbf9d2cb70f1c1c75333ba2e055d7a3",
"s2fieldsofstudy": [
"Chemistry",
"Materials Science"
],
"extfieldsofstudy": []
} |
121882804 | pes2o/s2orc | v3-fos-license | Over spinning a black hole?
A possible process to destroy a black hole consists on throwing point particles with sufficiently large angular momentum into the black hole. In the case of Kerr black holes, it was shown by Wald that particles with dangerously large angular momentum are simply not captured by the hole, and thus the event horizon is not destroyed. Here we reconsider this gedanken experiment for black holes in higher dimensions. We show that this particular way of destroying a black hole does not succeed and that Cosmic Censorship is preserved.
Introduction
Black holes can be formed through the collapse of matter, through sufficiently high-energy collisions of particles or quantum fluctuations in the early universe. Basically any process capable of confining a large portion of matter in a small enough space. Once formed, black holes are hard to kill. Quantum processes aside, no known classical mechanism can destroy a black hole. One of such processes was considered by Wald [1] many years ago and revisited recently [2,3,4,5,6]. It consists in throwing a point particle at a (four-dimensional) Kerr black hole with just the right angular momentum to spin the black hole up in such a way that eventually the horizon is disrupted. Indeed, the angular momentum of Kerr black holes is bounded by J ≤ M 2 , thus if it were possible for the black hole to capture particles of high enough angular momenta, then one might exceed this bound, possibly creating a naked singularity. Wald showed this cannot happen, as the potentially dangerous particles (i.e., those with large enough angular momentum) are never captured by the black hole [1].
The purpose of this short letter is to extend Wald's analysis to other spacetimes, in particular the Myers-Perry family of rotating black holes in higher dimensions [7]. Such an analysis is interesting because it allows one to test Cosmic Censorship in a very simple, yet realistic scenario.
2. Effective potential for radial motion of a test particle We are interested in analysing the motion of a test particle in a D-dimensional rotating black hole. More precisely, we will be considering an asymptotically flat stationary black hole described by the Myers-Perry solution [7]. Unlike Kerr black hole, where there is only one possible rotation axis, and therefore there is only one angular momentum parameter, in higher dimensions there are several choices of rotation axis and there is a multitude of angular momentum parameters, each referring to a particular rotation plane [7]. Here, we will focus on two possible situations: (i) a black hole with all the angular momenta equal or (ii) a singly spinning black hole. In addition, we focus exclusively on the intuitively most dangerous process: particles falling in along the equator. Therefore, we use the effective "2+1" dimensional metric along the equatorial plane. It is as well easy to define the conserved energy, E, and the angular momentum L (per unit testparticle mass m 0 in the case of time-like geodesics 1 ) associated to the time-like and rotational Killing vectors of the background geometry. The equatorial motion in the geometry just described can be reduced to the radial equatioṅ r 2 = V r [8,9]. For a single rotational plane where ∆ D = r 2 + a 2 − M r 5−D . Similar equations can be written when all angular momenta are equal. For instance, for even D = 2(d + 1) we find, while for odd D = 2d + 1 we obtain On the previous equations δ 1 = 1, 0 for timelike and null geodesics, respectively, and a dot stands for derivative respect to the proper time of the particle. The parameters a and M are related to the angular momentum and mass of the black hole (cf. Ref. [9]) The radial motion is completely governed by the potential V r . If there are turning points outside the event horizon, then a particle coming from infinity can not reach the event horizon. Thus, the analysis we want to make is to study the maximum value of L for which there are either no turning points, or all of them lie inside the event horizon.
3. Spinning-up a black hole by throwing point particles Now, we address the main issue of this paper: can we spin-up a black hole with mass M 0 and angular momentum J 0 in general D spacetime dimensions? For that, we throw in a particle of mass m 0 with angular momentum δ J = m 0 L and energy δ M = m 0 E, such that δM ≪ M 0 and δJ ≪ J 0 . Upon absorption of this particle, the dimensionless spin of the black hole changes to j = j 0 + δj , where the subscript stands for initial parameters of the black hole and 1 For massless particles, the quantities E and L may be regarded as the energy and angular momentum.
Single rotation parameter
As in four dimensions, in D = 5 we can also spin the black hole to the extremal limit and not further than that [10,9]. What about general D? We have numerically searched for the critical angular momentum, and computed δj in Eq. (6). The results, which are summarised in Fig. 1, are clear: neutral black holes in four and five spacetime dimensions with a single rotation cannot be spun-up past extremality. For larger D, there is no extremal limit, and the black holes can be spun-up to an arbitrarily high rotation.
All angular momenta equal
As we are dealing with a Myers-Perry black hole with equal angular momenta, by throwing only one point particle, we would end up with a Myers-Perry black hole with different angular momenta where the condition of all angular momenta being equal does not apply and we could conclude erroneously a violation of the cosmic censorship. How can we avoid this situation? Instead of throwing in one test particle, we consider d particles following similar geodesics along the d orthogonal rotation planes. The final black hole will also have all angular momenta equal. In this situation the expression (6) should read and the results are presented in Fig. 2. In full analogy with the singly spinning case in D = 5 in which the spin is bounded, we cannot exceed the extremal limit by throwing in test particles.
Conclusions
We have shown that a D-dimensional Myers-Perry black hole is immune to the throwing of point particles: in the geodesic approximation employed here, particles which are captured by the black hole have an angular momentum which is sufficiently low so as to be harmless; in fact sufficiently low that they are never able to spin-up the geometry past the extremal value. | 2019-04-19T13:11:23.617Z | 2011-09-22T00:00:00.000 | {
"year": 2011,
"sha1": "e745eb1ac8e425dd62a95ce1d88c8642307f7c96",
"oa_license": null,
"oa_url": "https://doi.org/10.1088/1742-6596/314/1/012064",
"oa_status": "GOLD",
"pdf_src": "IOP",
"pdf_hash": "0421aa175ca2309b89b33f02fbe8c68ccf23d409",
"s2fieldsofstudy": [
"Physics"
],
"extfieldsofstudy": [
"Physics"
]
} |
195186867 | pes2o/s2orc | v3-fos-license | Introduction of Procalcitonin Testing and Antibiotic Utilization for Acute Exacerbations of Chronic Obstructive Pulmonary Disease
Background: The majority of acute exacerbations of chronic obstructive pulmonary disease (AECOPD) are triggered by nonbacterial causes, yet most patients receive antibiotics. Treatment guided by procalcitonin (PCT), a sensitive biomarker of bacterial infection, safely decreases antibiotic use in many controlled trials. We evaluated PCT implementation for inpatients with AECOPD at a large academic hospital. Methods: All patients admitted for AECOPD during the first 6 months of PCT-guided therapy were eligible for inclusion in this retrospective cohort study. Patients with PCT performed were compared with those without PCT. The primary outcome was antibiotic days of therapy (DOT). Secondary outcomes included 30-day readmission and mortality. Results: Of the 238 AECOPD admissions, 73 (31%) had PCT performed. Procalcitonin-tested patients were more likely to meet systemic inflammatory response syndrome (SIRS) criteria, require intensive care unit (ICU)-level care, and have a longer length of stay (LOS) compared with those without PCT. Even after adjustment for these factors, PCT-tested patients received more inpatient DOT and there was no difference in total DOT. However, a low PCT value (<0.25 ng/mL) was associated with a 25.5% (P ⩽ .001) decrease in intravenous (IV) antibiotic DOT. Guideline-recommended follow-up testing was rare (12%). Procalcitonin measurement had no effect on 30-day readmission or mortality. Conclusions: In this real-world analysis of inpatients with AECOPD, PCT-guided therapy was poorly adopted by providers and was not associated with a decrease in total antibiotic DOT. However, a low PCT level was associated with a 25.5% decrease in IV antibiotic DOT, suggesting increased comfort stepping down from IV to PO therapy.
Introduction
Chronic obstructive pulmonary disease (COPD) affects 10% to 20% of Americans >40 years and is the third leading cause of death in the United States. 1,2 Acute exacerbations of COPD (AECOPD), defined as acute worsening in baseline symptoms beyond normal daily variations warranting a change in therapy, are associated with significant morbidity and mortality. 3 When the exacerbation is severe enough to warrant admission, 85%-87% of patients receive antibiotics. 4,5 Using antibiotics in severe AECOPD can decrease treatment failure and is consistent with the Global Initiative for Chronic Obstructive Lung Disease (GOLD) guidelines. 3,6 However, the majority of AECOPD may not be impacted by antibiotics; 22% are noninfectious and only 55% of infectious etiologies are bacterial. 7 Overuse of antibacterial therapy is not without consequence. Selective pressure is a primary mechanism in creating resistant organisms. 8 Specific to respiratory illness, broad spectrum antibiotics have been linked to methicillin-resistant Staphylococcus aureus (MRSA), methicillin-resistant Streptococcus pneumoniae (MRSP), and fluoroquinoloneresistant Streptococcus pneumoniae. [9][10][11] Furthermore, the intestinal dysbiosis from antibacterial therapy can nearly triple a 2 Infectious Diseases: Research and Treatment patient's risk of all-cause diarrhea and predispose them to Clostridium difficile colitis, a nosocomial pathogen responsible for nearly 30 000 deaths per year. 6,12,13 Considering the unintended consequences of antimicrobial therapy and the often nonbacterial cause of AECOPD, a specific biomarker for bacterial respiratory infections would be valuable. Procalcitonin (PCT)-a precursor peptide stimulated from bacteria-specific pro-inflammatory mediators (IL1-β, tumor necrosis factor [TNF], and interleukin [IL]-6) 14 -is such a biomarker. 15 Since 2004, there have been multiple randomized trials and a Cochrane review showing that PCTguided therapy safely reduces antibiotic exposure in acute respiratory illness (ARI). [16][17][18] Specific to AECOPD, a decrease in antibiotic exposure was shown in a randomized trial (ProCOLD) 19 and recent meta-analysis. 20 From an economic standpoint, it has been estimated that widespread use of PCTguided therapy in the United States in case of ARI could save US$1.6 billion annually. 21 Although PCT-guided antibiotic therapy shows promise for safely reducing antimicrobial use in AECOPD, its utility is extrapolated from predominantly randomized, controlled settings outside of North America. As hospitals increasingly introduce PCT to help antibiotic decision making, it is important to understand whether the benefits are achievable in realworld settings.
Patients and study design
We conducted a retrospective analysis using electronic records at a large academic hospital in Ann Arbor, Michigan. Data were collected over the first 6 months of PCT availability (October 1, 2013-March 31, 2014). Institutional rollout of PCT testing included dissemination of institutional PCT guidelines ( Figure 1) to department chairs, making the guidelines available online, and education by infectious diseases (ID) faculty at general medicine, surgery, and emergency department (ED) weekly conferences. Inpatients ⩾18 years of age with an International Classification of Diseases, Ninth Revision 22 (ICD-9) principal admission diagnosis of AECOPD were eligible and records were manually reviewed to confirm the diagnosis. Patients transferred from outside facilities were excluded. Requirement for patient consent was waived and the study was approved by the University of Michigan Health System Institutional Review Board (HUM00085581).
Measurements
In addition to standard demographic data, antibiotic days of therapy (DOT) and type (intravenous [IV] or oral) were recorded. Patients on chronic antibiotics (N = 6) at admission Ulrich et al 3 were included, but days on chronic antibiotics were excluded. Days of therapy included antibiotics initiated in the emergency department or while inpatient. If a patient was discharged with outpatient antibiotics, the anticipated course based on prescriptions was added to inpatient DOT to give total DOT. Systemic inflammatory response syndrome (SIRS) criteria were defined in concordance with the American College of Chest Physicians/Society of Critical Care Medicine (ACCP/ SCCM) 1992 conference. 23 Patients were classified into GOLD classes using forced expiratory volume in 1 second (FEV1) on prior pulmonary function testing (PFT) when available. 3 Acute exacerbations of COPD within the last year and home oxygen use was determined with detailed record review. Procalcitonin measurement was at the discretion of the admitting and inpatient physicians. Procalcitonin was performed from 0.5 mL of serum using the VIDAS BRAHMS PCT assay (bioMérieux, Inc., Durham, NC, USA), a US Food and Drug Administration (FDA)-approved automated immunofluorescent assay. Procalcitonin was considered low if ⩽0.25 ng/mL. When PCT was performed, time to re-testing was compared with institutional protocol (Figure 1).
Outcomes
Our primary outcome is antibiotic DOT. Secondary outcomes include protocol re-testing compliance, length of stay (LOS), 30-day readmission, and mortality.
Statistics
All analyses were conducted in R, version 3.2.2 (R Foundation for Statistical Computing, Vienna, Austria). Variables were examined for out-of-range values and histograms were used to visualize continuous variables. Descriptive statistics included proportions for categorical variables and measures of central tendency/spread for continuous variables. These initial analyses assisted in reconstructing variables for analysis, including logarithmic transformations for DOT and LOS due to the nonnormal distribution of data points. P values <.05 were considered statistically significant for all analyses. To analyze relationships between predictor variables and the primary/secondary outcomes, simple linear or logistic regression was employed. Multiple linear or logistic regression models were employed for adjusted analysis and variables were included in the final models based on significance on unadjusted analysis. Backwards elimination was then employed to result in a final candidate model. Interactions were modeled and included if significant.
PCT testing and outcomes
A total of 238 patients were admitted with AECOPD during the study period: 73 (31%) had PCT testing and 165 (69%) did not (Table 1). There were no significant age or gender differences between the 2 groups. Prior PFTs were available in 160 (67%) patients with no significant differences in FEV1 or GOLD classification. On unadjusted analysis, patients with PCT testing during the admission were more likely to be admitted to the intensive care unit (ICU; odds ratio [OR]: 2.21, 95% confidence interval [CI]: 1.02-4.81, P = .042), meet SIRS criteria (OR: 1.5, 95% CI: 1.09-2.06, P = .012), and have a baseline home oxygen requirement (OR: 2.24, 95% CI: 1.28-3.92, P = .005). In total, 171 (72%) patients received antibiotics while inpatient. Patients who underwent PCT testing were more likely to receive antibiotic therapy (OR: 3.85, 95% CI: 1.79-8.3, P = .001) and receive IV antibiotics (OR: 2.7, 95% CI: 1.51-4.8, P = .001), and these associations held after adjustment for the presence of SIRS and LOS. Procalcitonin testing had no effect on discharge antibiotic prescription rates (P = .493). On average, patients with PCT testing were admitted longer (5.4 ± 3.8 vs 3.1 ± 3.3; P < .001) and had more inpatient DOT (2.8 ± 1.8 vs 2.0 ± 1.4; P < .001) than those without PCT. Despite the differences in inpatient antibiotic DOT, PCT measurement had no effect on total DOT (P = .140; Figure 2). Only 9 (12%) patients with PCT testing had guideline-recommended re-testing. There was no difference in 30-day mortality (P = .213) or readmission (P = .514) after adjusting for LOS.
PCT level and outcomes
Patients with PCT testing were further divided into groups with low (⩽0.25 ng/mL) and high (>0.25 ng/mL) results; 57 (78%) of PCT results were low. All 16 patients with high PCT results received antibiotics, and 48 of 57 (84.2%) patients of the low PCT results received antibiotics. Despite high antibiotic imitation regardless of PCT level, a low PCT result was associated with a 27.2% lower (P < .001) IV antibiotic DOT (Table 2; Figure 3). This difference persisted after adjustment for SIRS criteria and LOS (Table 3). However, PCT level had no effect on inpatient DOT (P = .098) or total antibiotic DOT (P = .160). Furthermore, PCT level had no effect on use of antibiotics at discharge (P = .906), 30-day mortality (P = .747), or readmission rates (P = .237).
Discussion
During the introduction of PCT-guided therapy at our large academic hospital, patients admitted with AECOPD had PCT sent infrequently (31%), and those tested had higher acuity (SIRS, ICU) requiring longer admissions. Even after adjusting for these patient factors, PCT-tested patients received more inpatient DOT and there was no difference in overall DOT when compared with patients without PCT testing. Importantly, there was a 25.5% reduction in IV DOT after adjusting for LOS and SIRS. There was no difference in 30-day readmission or mortality.
Our results are contrary to many prior randomized controlled trials and prospective studies showing PCT-guided 4 Infectious Diseases: Research and Treatment therapy decreases antibiotic utilization in respiratory illness. [15][16][17]19,24,25 The discordant results are likely due to differences in guideline compliance; patients in prior studies were either randomized 16,17,19 or required to have PCT testing on admission with a qualifying diagnosis. 15,24,25 This bypasses the reality of provider-driven decision making. When given a choice, our results show that providers sent PCT infrequently, usually on patients with more severe disease. A possible explanation for this may be a greater awareness of PCT's role in sepsis and ICU settings [26][27][28][29][30][31] or mistrust in the reported test sensitivity. It could be the medical provider's tendency to send more tests on sicker patients. 32 Irrespective of the cause, our results suggest that providers sent PCT to confirm an already high clinical suspicion for bacterial infection rather than using it as a tool to delineate bacterial infection from nonbacterial causes. Our findings are concordant with the recent ProACT study, an open-label randomized controlled trial of PCT-guided therapy across 14 US hospitals of 1664 patients with lower respiratory tract infection (LRTI), 32% of which had AECOPD. 33 This trial found no decrease in antibiotic DOT between the PCT measurement and usual care. Similar to our study, clinicians' adherence to PCT guidelines was poor (49.2% for COPD diagnoses) and a significant portion of patients with low PCT still received antibiotics. Our findings, together with ProACT, highlight the challenges with implementation of a PCTguided model and suggest that achieving a PCT-driven decrease in antibiotic DOT requires high adoption rates for initial measurement and subsequent guideline-based therapy.
Regional differences are likely playing a significant role toward our results. In AECOPD, the sentinel study showing a benefit of PCT-guided therapy (ProCOLD) was conducted in Switzerland, 19 and a meta-analysis has shown a similar benefit in 8 trials conducted outside of North America. 20 Our study examined an academic institution in the United States, with a population and provider culture more similar to studies that failed to show PCT benefit for AECOPD: the ProACT trial 33 and a recent database review utilizing patients across 500 U.S. hospitals. 34 The reason for regional differences in PCT benefit likely ties back into guideline compliance. In the largest study examining PCT use in LRTI outside of trial settings, compliance was 76% in Switzerland and 74% in France, but only 34% in the United States. 25 The provider compliance in our studywith 31% of patients having an initial PCT-is concordant to these findings. Procalcitonin compliance may be higher in European countries due to greater familiarity with PCTguided therapy and a difference in the health care culture surrounding guidelines; for example, compared with the United States, European guidelines are more likely to have implementation tools to help providers. 35 Despite poor guideline compliance and minimal effect on total antibiotic DOT, we found a 25.5% decrease in IV antibiotic DOT if PCT result was low. This persisted after adjustment for differences in patient acuity. Although a low result is highly suggestive of nonbacterial infection and the algorithm discourages antibiotics, providers were more comfortable "stepping-down" therapy from IV to oral. The utility of this is debatable, but any decrease in broad spectrum antibiotics is welcome from a stewardship perspective and is associated with less antimicrobial resistance. 36 Furthermore, with no difference in readmission or mortality between groups, testing is unlikely to cause harm.
A major strength of our study is our large cohort in a realworld setting, where consecutively admitted patients with AECOPD were all eligible for inclusion. Our cohort is larger than the randomized trial AECOPD arms 16,17,19 and similar in size to the largest retrospective studies on PCT and antibiotics in AECOPD. 37,38 To reduce confounding from facilities making antibiotic decisions without the option of sending PCT, we excluded patients transferred from outside facilities. To fully characterize the cohort, we included important COPD baseline disease factors such as PFT, home oxygen use, and recent AECOPD, which allowed determination of GOLD classification in the majority of patients.
Our study has several limitations. First, we used a retrospective cohort of AECOPD patients based off ICD-9-coded admission diagnoses. To ameliorate misclassification, we confirmed each patient's diagnosis through extensive chart review, but it is possible that some admissions for AECOPD were not In an effort to increase generalizability, we included both ICU and non-ICU patients. However, a recent trial has shown that PCT-guided antibiotic management in ICU patients admitted with AECOPD does not affect antibiotic exposure and may worsen mortality outcomes. 39 In the future, PCT testing should be limited to non-ICU inpatients with AECOPD. Much like the ProACT trial, our study was also limited by the implementation of PCT testing, algorithm availability, and provider education without active surveillance or prospective audit and feedback. Although active stewardship may increase compliance with PCT-based guidelines, we believe our design more realistically approximates testing rollout in a US health care environment where it can be difficult to obtain adequate funding for stewardship measures. 40 An additional limitation was our limited timeframe including the inpatient stay, outpatient discharge antibiotics, and 30-day outcomes; there may be benefits to PCT guidance extending beyond this time that were not realized. A final possible limitation was our inclusion of discharge azithromycin suppression of AECOPD, which could affect DOT, but may not be unnecessary as it can be effective over months to years. 41 However, the majority of prescriptions were for short-course (5-10 days) azithromycin therapy which has been shown to offer no benefit in AECOPD. 42 Current GOLD initiative and Veterans Affairs and the Department of Defense (VA/DoD) guidelines do not recommend testing PCT in AECOPD. 3,43 Our study suggests that there are significant difficulties with implementing PCTguided therapy in AECOPD, and we found no association with decreased antibiotic DOT. Nevertheless, a change in national guidelines with a subsequent increase in PCT compliance may be what is needed to have medical centers approximate the PCT benefit seen in prior AECOPD studies. Furthermore, our results suggest that current guidelines may be missing an opportunity to decrease IV antibiotic DOT when PCT returns low. Future analysis of secular trends is warranted as increased familiarity of PCT will increase use of this biomarker toward PCTguided therapy; with this, the benefits observed in clinical trials may emerge in real-world settings of AECOPD.
Conclusions
In patients admitted with AECOPD, introduction of PCTguided therapy was poorly adopted and was not associated with the antibiotic reducing effects seen in previous controlled trials. Our study highlights the difficulty with PCT implementation, as providers tested primarily higher acuity AECOPD patients requiring a longer LOS. Even after adjusting for these patient factors, PCT testing did not associate with a decrease in overall antibiotic DOT. However, patients with a low PCT received 25.5% less IV antibiotic DOT, suggesting comfort with stepping-down therapy and a possible benefit from an antimicrobial standpoint. Abbreviations: DOT, days of therapy; IV, intravenous; LOS, length of stay in days; PCT, procalcitonin; SIRS, systemic inflammatory response syndrome. | 2019-06-21T23:05:39.867Z | 2019-01-01T00:00:00.000 | {
"year": 2019,
"sha1": "715ee242e1c8c389a81e0c7364cd78959f8a1095",
"oa_license": "CCBYNC",
"oa_url": "https://journals.sagepub.com/doi/pdf/10.1177/1178633719852626",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "715ee242e1c8c389a81e0c7364cd78959f8a1095",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
214359870 | pes2o/s2orc | v3-fos-license | Classification of provinces in Indonesia based on 2014 presidential election participation levels using fuzzy c-means
The presidential and vice presidential elections held every five years is a benchmark for democracy in Indonesia. However, the expected level of participation cannot be achieved. Election socialization as an effort to increase the level of participation in each province is different because the level of participation in each province is also varied. Therefore, there is a need of regional grouping based on the level of electoral participation in order to facilitate the electoral socialization process. This study uses fuzzy c-means with two attributes, namely the number of voters and the number of voting rights users. All provinces in Indonesia are classified into two clusters. If the provincial participation rate is below the national participation standard (69.58%) then the province is considered to have a low participation rate and vice versa. Therefore, the first cluster constitutes provinces with high electoral participation rate and the second cluster constitutes provinces with low electoral participation rate. The results of the classification are 9 provinces in first cluster and 24 provinces in the second cluster.
Introduction
Indonesia has many provinces and in terms of sovereignty, it "is in the hands of the people and is implemented according to the Constitution" [1]. This statement in the 1945 Constitution confirms that Indonesia is a democratic country, from the people by the people and to the people, including in the election of the president and vice president, in which citizens elect directly in the election. Article 6A of the 1945 Constitution states that the President and Vice President shall be elected as a pair by the people. General elections shall be conducted in a direct, general, free, secret, honest and fair every five years [1], so elections are usually called democratic parties because not every year happens.
Elections that are routinely carried out certainly produce data that can be processed mathematically. For example, it is predicting election results and predicting presidential candidates to win or lose. Research conducted by Jiao using fuzzy adaptive network shows that fuzzy can predict the results of the presidential election ideally [2]. Whereas Singh forms a model to calculate the percentage of win or lose from a candidate or party in the election using fuzzy [3]. In addition to the election results, the level of election participation also received attention. Voter participation is one indicator of the results of a general election with a strong degree of legitimacy. Participation rate is often a reference for election organizers whether the elections they hold have a strong appeal to citizens to be involved. Not surprisingly, General Election Commission (KPU) as the election organizer monitors the participation rate which will always be attempted to increase [4]. KPU commonly targets national participation rates. In 2014 KPU targeted 75% for election participation rates of the President and Vice President. However, voter participation in the 2014 presidential election missed the KPU's goal. Moreover, voter participation also declined compared to 2009 elections and April 2014 legislative elections. Voter participation in the 2014 presidential election based on data reported by KPU was only 69.58%. Generally the causes of decreasing participation differ in each region so that the treatment to increase the participation level of each region also varies. Therefore, grouping regions based on the level of electoral participation is deemed necessary and it is expected that the election socialization in the future could be carried out more easily. In this paper the Indonesian provinces will be divided into two clusters, with two attributes or variables, namely number of voters and number of voting right users. Provinces with a participation rate below the national participation standard of 69.58% are considered to have a low participation rate. Meanwhile, the provinces with a participation rate above the national participation level are considered to have a high level of participation. Fuzzy C-Means (FCM) is greater than SOM neural networks, K means and traditional hierarchical clustering algorithms in clustering. FCM shows stability even though there are outliers and overlapping data [5]. Clustering using FCM was used to classify areas in West Java based on economic potential. The study produced nine clusters based on the most potential economic in each region [6]. Similar to the study, this study will cluster provinces using FCM. This study uses FCM, which is commonly used for grouping data. Based on the background above, this grouping needs to be carried out because the level of voter attendance as a form of election participation becomes very important. It relates to the level of legitimacy and the future of democracy. The low level of voter attendance in the election can pose a risk for a country's democracy [7].
Methodology
The data processing use rules in clustering with FCM. The data were collected from the results of the 2014 presidential and vice presidential elections on the General Election Commission web page (kpu.go.id). The variables used in this study are variables that represent the level of regional participation. These variables are: V1: number of voters, V2: number of voting right users. After determining the variable, data were processed using FCM.
Fuzzy C Means
Fuzzy Clustering is one technique for determining optimal clusters in a vector space based on the Euclidian normal form for distances between vectors. There are several data clustering algorithms; one of which is FCM. FCM is a data clustering technique where the existence of each data point of a cluster is determined by the degree of membership. This technique was first introduced by Jim Bezdek in 1981. The first basic concept of FCM is to determine the center of the clusters, which marks the average location for each cluster. In the initial conditions, the cluster center is still not accurate. Each data point has a degree of membership for each cluster. The second basic concept of FCM comprises fixing the cluster center and the degree of membership of each data point repeatedly. It can be seen that the center of the cluster will move towards the right location. This looping is based on the minimization of the objective function which describes the distance from the data points to the cluster center weighted by the degree of membership of the data points [8]. Membership degrees between zero and one are used in fuzzy clustering instead of crisp assignments of the data to clusters. The resulting data partition improves data understanding and reveals its internal structure. Partition clustering algorithms divide a data set into clusters or classes, where similar data objects are assigned to the same cluster, whereas dissimilar data objects should belong to different clusters [9]. The FCM algorithm is as follows [8]: (2) 6) Calculate the partition matrix changes with i=1,2,…,n; k=1,2,…,c.
Results and Discussion
Since its post-reform election history, Indonesia has held elections four times, including the last 2014 Election. If compared with the 1999 election when the president is still elected by the MPR, voter participation in the legislative elections in 2014 is very high. Voter participation in the 1999 election reached 97.7%. This means that only 2.3% of voters did not vote. A number of voters registered in the 1999 election were 117,815,953 people. However, after the 1999 elections, the tendency of voter participation rates has declined [2]. The election participation rate is calculated by the following formula:
× 100%
Nr : number of voters Vo : number of voters who give voting rights Based on this formula, the level of community participation in the 2014 presidential and vice presidential election was 69.58%. The number of voters is the sum of the Final Voters List (DPT), Additional Voters List (DPTb), Voters registered in the Special Voters List (DPK) and Additional Special Voters List (DPKTb). Table 1 demonstrates results of the total votes from each province in the 2014 presidential and vice presidential elections. The data were collected from www.kpu.go.id. The next step for clustering is determining number of cluster, degree, maximum iteration, the smallest error expected, first objective function, and first iteration. So, there are two clusters, thus the degree used is two and the results of the cluster center calculation, first objective function, and matrix partition changes are not too large. The smallest expected error is 10 −5 and the desired maximum iteration is 5. Afterwards, the center of cluster and objective function was observed using U Matrix random. In the first iteration, the condition cannot stop due to the following calculation. Since it meets the requirements of | 2 − 1 | < the iteration stops. The clustering results table is presented according to the new U matrix. Cluster determination is based on the largest value of the U matrix. For example, The Province of Aceh has membership values on matrix U 0.14 and 0.86. The value of 0.14 is the value of Aceh's membership in the first cluster and 0.86 is membership value in the second cluster. As the membership value in second cluster is bigger, so The Province of Aceh goes into second cluster. From the calculation results according to the algorithm, the first cluster (provinces of high electoral participation rate) consists of the provinces of North Sumatra, South Sumatra, Lampung, DKI Jakarta, West Java, Central Java, East Java, Banten and South Sulawesi. The second cluster (provinces of low electoral participation rate) consists of 24 other provinces, namely Aceh Province, West Sumatra, Riau, Jambi, Bengkulu, Bangka Belitung, Riau Islands, DIY, Bali, NTB, NTT, West Kalimantan, Southeast Kalimantan, South Kalimantan, East Kalimantan, North Sulawesi, Southeast Sulawesi, Central Sulawesi, Gorontalo, West Sulawesi, Maluku, North Maluku, Papua and West Papua.
Conclusion
Clustering with the Fuzzy C-Means method with two iterations has fulfilled the stop condition according to the algorithm, namely (| 2 − 1 | < < 10 −5 )). Provinces are classified into 2 clusters. The first cluster consists of the provinces with high electoral participation rate and the second cluster consists of the provinces with low electoral participation rate. This clustering method results in 9 | 2019-12-05T09:15:58.233Z | 2019-10-01T00:00:00.000 | {
"year": 2019,
"sha1": "867108f175f1c7aa10beb5c2955207dc3becd645",
"oa_license": null,
"oa_url": "https://doi.org/10.1088/1742-6596/1320/1/012022",
"oa_status": "GOLD",
"pdf_src": "IOP",
"pdf_hash": "8a69a056e779622b9fd35e80e5c84d389f673a90",
"s2fieldsofstudy": [
"Political Science"
],
"extfieldsofstudy": [
"Physics",
"Computer Science"
]
} |
260717926 | pes2o/s2orc | v3-fos-license | Potential additional effects of iron chelators on antimicrobial- impregnated central venous catheters
COPYRIGHT © 2023 Itoh, Tsutani, Mitsuke and Iwasaki. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Potential additional e ects of iron chelators on antimicrobialimpregnated central venous catheters
Introduction
Central venous catheters (CVCs) play a clinically important role in various treatments, including venous pressure monitoring, the infusion of drugs, such as anti-cancer chemotherapy and antibiotics, parenteral nutrition, and blood transfusion (Saugel et al., 2017). However, compared with peripheral venous catheters, patients with indwelling CVCs are at increased risk of thrombosis, embolism, and infection (Ball and Singh, 2023). Catheter-related infections (CRIs) are thought to be associated with the formation of bacterial and fungal colonies on the catheter (Hanna et al., 2004). Microorganisms adhere to device surfaces, produce an extracellular polymeric matrix, and form biofilms. The biofilm-forming microorganisms then become a source of medical device-associated infection (Donlan, 2002). CRI is defined as a positive catheter tip culture and systemic inflammatory findings, while catheter-related bloodstream infection (CRBSI) is defined as CRI-criteria plus detection of the same organisms as in the catheter tip culture in peripheral blood cultures (Garner et al., 1988;Rockholt et al., 2023). CRBSIs have a reported mortality rate of 12%−40% depending on a variety of factors, such as comorbidities, the type of CVC used, and the type of pathogen, and are associated with longer hospital stays and increased medical costs (O'Grady et al., 2011;Böll et al., 2021;Ball and Singh, 2023).
Several preventive measures exist for CRIs, such as thorough hand hygiene, sterilization of the CVC site, and avoidance of catheter insertion at the femoral site. It is widely accepted that CVCs, which are foreign substances that can be a source of infection, should be removed as soon as possible. However, in clinical practice, situations arise in which long-term indwelling is required. Therefore, in such special situations, the Centers for Disease Control and Prevention (CDC) recommends the use of antimicrobial agents or antiseptic-impregnated catheters with comprehensive and adequate infection control measures, as shown Supplementary Figure S1 (O'Grady et al., 2011). Antimicrobial-impregnated catheters are a promising option for cases that require long-term CVC use (Böll et al., 2021). The purpose of this article was to .
summarize previous attempts to use antimicrobial/antiseptic impregnated catheters and provide a new perspective.
Antimicrobial/antiseptic-impregnated catheters
The control of CRI remains a major challenge to overcome; hence, a multifaceted attempt should be conducted. Certainly, antimicrobial-impregnated catheters have achieved some success to date, but no established clinical efficacy has been achieved with antiseptic substances (e.g., silver), which have been studied for many years (Viola et al., 2017). One such catheter is the chlorhexidine/silver sulfadiazine catheter, but most studies have found that its effect on reducing CRBSIs is not significant (Viola et al., 2017). Minocycline/rifampicin and miconazole/rifampicin catheters have also been shown to reduce the incidence of CRBSIs in several studies, where the incidence of antibiotic resistance, which remains a concern, was low (Viola et al., 2017;Reitzel et al., 2020;Böll et al., 2021). In clinical trials of patients with cancer, minocycline/rifampicin-impregnated catheters have also been shown to reduce the incidence of CRBSIs when used for longer than 2 months (Hanna et al., 2004). The combination of chlorhexidine with minocycline and rifampicin on catheters has been reported to have enhanced antimicrobial activity (Raad et al., 2012). Chlorhexidine acts on cell membranes, minocycline on protein synthesis, and rifampin on RNA synthesis. Thus, the mechanism of action of the combination of these three drugs may be additive. However, the confirmed enhancement of activity is only the result of in vitro studies, and the possibility of clinical application needs to be confirmed in clinical trials (Viola et al., 2017). As discussed above, a combined approach with different mechanisms of action may enhance antimicrobial activity and broaden the spectrum. Therefore, new approaches that are not antimicrobial agents or antiseptic substances may produce additional benefits.
Hypothesis
From a different perspective, attempts have been made to inhibit iron, which is essential for the growth and proliferation of pathogenic microorganisms, as a strategy to increase antimicrobial activity (Schwarz et al., 2019;Scott et al., 2020). Specifically, the strategy is to use iron chelators in combination with antibacterial and antifungal agents. We hypothesize that the impregnation of catheters with antimicrobial agents and iron chelators would increase their antimicrobial and anti-biofilm activity and inhibit microbial colony and biofilm formation, resulting in reduced CRI incidence and longer catheter use, as shown in Supplementary Figure S2. The surface and lumen of the catheter are impregnated with antimicrobial and ironchelating agents, which are assumed to inhibit bacterial and fungal colony and biofilm formation by exerting antimicrobial and iron-chelating activities on the bacteria and fungi that attach to the surface.
Support for the hypothesis Antimicrobial activity of iron chelators
Because of its redox potential, iron is involved in many biochemical reactions. Iron is an essential element for all living organisms, including bacterial pathogens (Dev and Babitt, 2017). It also plays an important role as a cofactor in various intracellular pathways, such as DNA synthesis and repair, electron transfer, oxygen transport, and immune response (Scott et al., 2020). Iron levels in the human body are strictly regulated to protect against the toxic effects of free iron, which catalyzes the generation of reactive oxygen species (ROS), and the utilization of iron by invading microorganisms (Dixon and Stockwell, 2014). Through nutritional immunity, humans inhibit microbial growth by starving microorganisms of the metal ions they need (Hood and Skaar, 2012;Palmer and Skaar, 2016). Bacteria and fungi bind iron using high-affinity chelating compounds, called siderophores, as a strategy to acquire iron. Numerous types of siderophores have been reported. Siderophores are categorized into three main structural families-carboxylates, catecholates, and hydroxamates-with the catecholate siderophore, enterobactin, showing the highest affinity for iron, even higher than that for the host iron-binding protein, transferrin (Ellermann and Arthur, 2017). Iron chelators may be useful in the treatment of infectious diseases by limiting the availability of iron, thereby inhibiting the production of ROS, and by inhibiting microbial growth due to nutrient restriction (Lehmann et al., 2021;Paterson et al., 2022).
To date, the combinations of vancomycin and deferasirox against methicillin-resistant Staphylococcus aureus (Luo et al., 2014), doxycycline, and the iron chelator CP762 against Pseudomonas aeruginosa (Faure et al., 2021), and the triple combination of doxycycline, deferasirox, and thiostrepton (an antibiotic against gram-positive bacteria) against P. aeruginosa and Acinetobacter baumannii (Chan et al., 2020) have shown promising effects both in vitro and in animal models. For fungi, animal model experiments of combination therapy with liposomal amphotericin B (L-AMB), micafungin, and deferasirox against Aspergillus and Mucor (Ibrahim et al., 2011), animal experiments of combination therapy with L-AMB and deferasirox against Aspergillus fumigatus (Ibrahim et al., 2010), and in vitro experiments of combination therapy with AMB and deferoxamine against Cryptococcus spp. (Chayakulkeeree et al., 2020) have indicated that these combinations are useful. Despite promising experimental data for the combination of deferasirox and L-AMB for mucormycosis (Ibrahim et al., 2007(Ibrahim et al., , 2011Schwarz et al., 2019), unexpectedly, a higher mortality rate at 90 days was seen with the combination than with L-AMB alone in a randomized controlled trial (Spellberg et al., 2012). Patients were enrolled at multiple centers with heterogeneous patient populations, resulting in an imbalance in underlying diseases and risk factors, with more patients in the deferasirox group having hematologic malignancies, neutropenia, and pulmonary infections. This imbalance in the patient's background may have influenced the patient's prognosis (Spellberg et al., 2012). Therefore, clinical trials are inconclusive regarding the efficacy of iron chelation assisted therapy.
Nevertheless, iron-chelation therapy still has potential for clinical applications. Different types of iron-chelating agents .
/fmicb. . have been shown to have different antimicrobial activities. Of those tested, a water-soluble agent containing iron-selective copolymers (denying iron to bacterial infections; DIBI) was the most effective (Iron Binding Polymers | Fe Pharma, 2021; Lehmann et al., 2021). Recent studies have reported the promising finding that a combination of DIBI and antimicrobials may overcome drug-resistant bacteria and fungi. DIBI in combination with ciprofloxacin has been shown to be successful in treating ciprofloxacin-resistant A. baumannii-infected mice (Parquet et al., 2019). Moreover, the combination of DIBI and fluconazole was found to inhibit the growth of fluconazole-resistant Candida albicans in an in vitro study (Savage et al., 2018). These findings suggest that the appropriate selection of a type of iron chelator and its combination with an antimicrobial agent may be useful for dealing with a broad spectrum of infectious pathogens, and may overcome antibiotic resistance mechanisms.
Anti-biofilm activity of iron chelators
One notable problem related to CVC implantation is biofilm formation. Here, we further explain the relationship between iron chelators and biofilm formation. It has already been mentioned that iron is essential for bacterial and fungal growth and proliferation, but it is also required for biofilm formation (Nazik et al., 2015;Coraça-Huber et al., 2018;Firoz et al., 2021). Microorganisms produce biofilms that protect them against antimicrobials, making them drug resistant (Høiby et al., 2010;Singh et al., 2010;Brauner et al., 2016;Thi et al., 2020). Hence, methods are needed to inhibit the formation of biofilms and attack the microorganisms that have formed within them. Iron-chelation therapy has been tested in P. aeruginosa to explore adjunctive therapy for chronic P. aeruginosa infection. Deferiprone, a synthetic iron chelator, has been shown to inhibit biofilm formation (Houshmandyar et al., 2021). Furthermore, the combination of iron chelators and antimicrobial agents has shown enhanced biofilm inhibitory activity; examples include combination therapy with the iron (VI) chelator N, N'bis (2-hydroxybenzyl) ethylenediamine-N, N'-diacetic acid and colistin in an in vitro experiment (Mettrick et al., 2020), and with the iron chelators deferoxamine or deferasirox and tobramycin (Moreau-Marquis et al., 2009). Lactoferrin, a component of human secretions, also has iron-chelating activity (Singh, 2004), and ALX-109, an investigational agent containing lactoferrin and hypothiocyanite (a bactericidal agent), has shown activity against P. aeruginosa biofilms in combination with tobramycin and aztreonam (Moreau-Marquis et al., 2015). Deferasirox has shown biofilm inhibitory effects on the periodontal bacterium Prevotella intermedia (Moon et al., 2013), and lactoferrin treatment has significantly reduced proinflammatory cytokines production by gingival fibroblasts infected with P. intermedia. Moreover, an observational clinical trial of patients with periodontitis showed that treatment with lactoferrin reduced proinflammatory cytokines such as interleukin 6, edema, bleeding, pocket depth, and gingival and plaque indices in the crevicular fluid and improved .
/fmicb. . Fourteen compounds and one iron analog reported to have iron chelating activity were selected and tested for synergism with thiostrepton. The triple combination of thiostrepton, deferasirox, and doxycycline was the most effective against P. aeruginosa and A. baumannii isolates.
Ibrahim et al. (2010)
In vitro and in vivo study
Deferasirox L-AMB
A. fumigatus Deferasirox demonstrated in vitro fungicidal activity against A. fumigatus and prolonged survival in mice with invasive pulmonary aspergillosis. Deferasirox plus L-AMB synergistically improved survival and reduced pulmonary fungal burden compared to either agent alone.
R. oryzae; A. fumigatus
Triple therapy with L-AMB, micafungin, and deferasirox improved survival and reduced tissue fungal burden in mice with mucormycosis, but to a lesser extent in aspergillosis.
Parquet et al. (2019)
In vitro and in vivo study
DIBI Ciprofloxacin
A. baumannii DIBI inhibited clinical A. baumanii isolates at MICs below those of typical antibiotics; low-dose nasal administration of DIBI after intranasal challenge with ciprofloxacin-resistant A. baumanii LAC-4 significantly reduced the bacterial burden in mice and DIBI also inhibited the spread of infection to the spleen. Given the ciprofloxacin resistance of LAC-4, treatment of infected mice with ciprofloxacin alone was ineffective, but treatment with DIBI greatly enhanced the therapeutic effect. Savage et al. (2018) In vitro and in vivo study DIBI Fluconazole C. albicans DIBI inhibited the growth of C. albicans in vitro at relatively low concentrations, and this inhibition was reversed by the addition of iron; the combination of DIBI and various azoles showed stronger growth inhibition than azoles alone, as shown in fluconazole-resistant C. albicans. In an experimental model of C. albicans vaginitis, DIBI in combination with fluconazole significantly improved clearance of infection in mice inoculated with fluconazole-sensitive strains.
(Continued)
Frontiers in Microbiology frontiersin.org . /fmicb. . Houshmandyar et al. (2021) In vitro study Deferiprone P. aeruginosa Deferiprone showed moderate activity against P. aeruginosa cells grown on plankton and was effective in inhibiting biofilm formation. Mettrick et al. (2020) In vitro study HBED Colistin P. aeruginosa HBED, a synthetic hexadentate iron chelator, inhibited the growth and biofilm formation of all clinical strains of P. aeruginosa under aerobic and anaerobic conditions; HBED in combination with colistin significantly enhanced the microcolony killing effect of colistin, resulting in almost complete biofilm removal.
Moreau-Marquis et al. (2009)
In vitro study Deferoxamine; Deferasirox Tobramycin P. aeruginosa Tobramycin in combination with deferoxamine or deferasirox reduced the P. aeruginosa biomass of established biofilms by approximately 90% and reduced viable bacteria by 7 log units; neither tobramycin, deferoxamine, nor deferasirox alone had such a significant effect. The combination of tobramycin and iron chelators also prevented biofilm formation on human respiratory cells.
Moreau-Marquis et al. (2015)
In vitro study ALX-109 [combination of an investigational drug containing lactoferrin and hypothiocyanite (a bactericidal agent)] Tobramycin; Aztreonam P. aeruginosa ALX-109 inhibited P. aeruginosa biofilm formation but did not affect established biofilms; ALX-109 enhanced the ability of tobramycin and aztreonam to inhibit P. aeruginosa biofilm formation and reduce established biofilms. Moon et al. (2013) In vitro study Deferasirox P. intermedia Deferasirox exhibited potent antibacterial activity against P. intermedia, partially inhibited bacterial growth, and significantly prolonged bacterial doubling time. Deferasirox also significantly reduced the biofilm-forming activity and biofilm formation of P. intermedia. In the ATP bioluminescence assay, Deferasirox significantly reduced biofilm bioactivity. Lin et al. (2012) In vitro study PGG S. aureus PGG may be a candidate for the development of anti-biofilm products for clinical use due to its anti-biofilm activity and low cytotoxicity. Impairment of S. aureus biofilm formation by PGG was shown to be due to iron chelation. Iron supplementation complemented the effect of PGG and restored biofilm formation. Richter et al. (2017) In vitro and in vivo study Deferiprone Gentamicin; Ciprofloxacin
S. aureus
The combination of deferiprone-GaPP significantly reduced bacterial load and increased survival of S. aureus small colony variant-infected C. elegans in an artificial wound model. When deferiprone-GaPP was combined with gentamicin or ciprofloxacin in a colony biofilm model, deferiprone-GaPP was able to enhance the activity of gentamicin or ciprofloxacin.
Coagulasenegative staphylococci
Deferiprone alone had only a moderate inhibitory effect on biofilm growth, but the combination of deferiprone and the respective antibiotics significantly reduced bacterial counts by 2-3 logs compared to the effect of the antibiotic alone. In addition, the combination of deferiprone and antibiotics showed severe to complete destruction of the biofilm, which was not seen when antibiotics were administered alone. Nazik et al. (2015) In vitro study Deferoxamine; Deferiprone
A. fumigatus
Deferoxamine concentrations below 1,250 µM had no effect, while 2,500 µM increased biofilm formation or preform biofilms of A. fumigatus. 156 to 2,500 µM deferiprone inhibited biofilm formation in a dose-response manner. In preform biofilms, deferiprone above 625 to 1,250 µM showed an inhibitory effect compared to controls. The results of deferoxamine-induced enhancement of biofilm formation were attributed to A. fumigatus acquiring many iron chelator complexes through a siderophore-like mechanism.
A. baumannii, Acinetobacter baumannii; A. fumigatus, Aspergillus fumigatus; DIBI, denying iron to bacterial infections; GaPP, gallium protoporphyrin; HBED, N, N'-bis ( clinical adhesion levels (Berlutti et al., 2011). Several studies were conducted on Staphylococcus spp. 1,2,3,4,6-Penta-O-galloylβ-D-glucopyranose, a plant-derived ingredient with iron-chelating properties commonly used in Chinese medicine, has shown biofilm inhibitory effects on S. aureus (Lin et al., 2012). The combination of deferiprone and the heme analog gallium protoporphyrin in combination with gentamicin or ciprofloxacin increased the biofilm inhibitory effects in an S. aureus colony biofilm model (Richter et al., 2017). Additionally, for coagulase-negative staphylococci, deferiprone showed enhanced antibacterial activity in combination with clindamycin, gentamicin, or vancomycin by disrupting biofilms (Coraça-Huber et al., 2018). Regarding fungi, deferiprone inhibited biofilm formation by A. fumigatus, whereas deferoxamine had the opposite effect, and promoted biofilm formation. The result of the promotion of biofilm formation by deferoxamine may be due to A. fumigatus acquiring more iron chelator complexes through a siderophore-like mechanism (Nazik et al., 2015). In addition, deferoxamine is used by zygomycetes and A. fumigatus as a siderophore to promote bacterial growth (Boelaert et al., 1993). Thus, depending on the combination of the type of microorganism and iron chelator, the microorganism has a tolerance mechanism and takes advantage of iron starvation. However, iron chelators have a different point of action than do antibacterial and antifungal agents, and thus, may carry less risk of inducing drug resistance in microorganisms. This is a major advantage of combining drugs with different mechanisms of action.
Mechanisms of antimicrobial and anti-biofilm activity of iron chelators and antimicrobial-impregnated catheters
The iron chelator CP762 in combination with tetracyclines against P. aeruginosa showed synergistic effects. Among the tetracyclines, doxycycline showed particularly high synergism (Faure et al., 2021). CP762 is a hexadentate hydroxypyridinone iron chelator that has high affinity and selectivity for iron and does not utilize many of the bacterial iron siderophore receptors, making it unlikely to donate iron to pathogenic microbes (Piyamongkol et al., 2005;Zhou et al., 2011). Tetracycline binds to the 30S bacterial ribosome via a magnesium bridge (White and Cantor, 1971;Pioletti et al., 2001), but iron can inhibit this mechanism by binding to the magnesium binding site (Faure et al., 2021). Sequestration of iron by iron chelators may minimize iron binding to tetracycline, which would facilitate its complexation with low-affinity ions such as magnesium, which is necessary for binding to the bacterial ribosome (Faure et al., 2021).
Conclusion
Most iron chelators examined in the past have been effective in inhibiting biofilm formation by bacteria and fungi, and their antimicrobial activity has been shown to be enhanced when used in combination with antimicrobial agents (Figure 1, Table 1). Therefore, we believe that the inclusion of iron chelators in antimicrobial-or antifungal-impregnated catheters could help overcome microbial resistance mechanisms and create CVCs with a lower risk of CRIs. Further research on iron chelators and antimicrobial-impregnated catheters, and confirmation of their effects on reducing the incidence of CRIs in clinical trials, are warranted. | 2023-08-09T15:02:38.994Z | 2023-08-07T00:00:00.000 | {
"year": 2023,
"sha1": "3772966b29bee46d430cd4cc90a46420ec0a3a11",
"oa_license": "CCBY",
"oa_url": "https://www.frontiersin.org/articles/10.3389/fmicb.2023.1210747/pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "cbd969057ca8afae12e49a5f3028df7105c6076d",
"s2fieldsofstudy": [
"Biology"
],
"extfieldsofstudy": []
} |
251119481 | pes2o/s2orc | v3-fos-license | Immunogenicity and safety of a SARS-CoV-2 inactivated vaccine (CoronaVac) co-administered with an inactivated quadrivalent influenza vaccine: A randomized, open-label, controlled study in healthy adults aged 18 to 59 years in China
Background Studies are needed for evidence of inactivated COVID-19 vaccine co-administered with influenza vaccine. Methods A randomized, open-label, controlled study was conducted in Zhejiang Province, China. Eligible healthy adults aged 18–59 years underwent randomization at a ratio of 1:1:2 to receive inactivated quadrivalent influenza vaccine (IIV4) either concomitantly with the first (C1 subgroup) or the second (C2 subgroup) dose of CoronaVac, or 14 days after the first dose of CoronaVac (S group). The primary purpose of the study was to prove the non-inferiority in seroconversion rate of antibody against SARS-CoV-2. Results Overall, 480 participants were enrolled, with 120, 120, and 240 randomly assigned to the C1, C2, and S groups, respectively. As lower bound of the two-sided 95% confidence interval (CI) of the difference for the seroconversion rate of antibodies against SARS-CoV-2 was over −10%, the immune response for CoronaVac in the C group (93.1% [89.0, 96.0]) was non-inferior to that in the S group (95.2% [91.5, 97.6]) in the per-protocol set. A lower GMT of antibody against SARS-CoV-2 was observed in the C group as compared to the S group (27.5 vs. 38.1, P = 0.0001). Decrease of immune response to CoronaVac was mainly observed in participants received IIV4 concomitantly with their second dose of CoronaVac (C2 subgroup), with a seroconversion rate of 89.7% (95CI: 82.6%-94.5%) and a GMT of 23.3. The occurrences of vaccine related adverse reactions were no more than 20% and comparable among different groups. Most of the adverse reactions were mild and moderate. Conclusion Co-administration of inactivated COVID-19 vaccine and seasonal influenza vaccine, especially the administration regimen that the seasonal influenza vaccine co-administered with the first dose of the inactivated COVID-19 vaccine, would be feasible.
Immunogenicity and safety of a SARS-CoV-2 inactivated vaccine (CoronaVac) co-administered with an inactivated quadrivalent influenza vaccine: A randomized, open-label, controlled study in healthy adults aged 18 to 59 years in China
during the COVID-19 epidemic could result in poor clinical outcomes among patients and an increased disease burden on the society [8][9][10].
Annual vaccination is the most effective strategy to prevent seasonal influenza [11][12]. To date, it is believed that the COVID-19 vaccine is a powerful approach to prevent COVID-19 diseases, especially severe illness and death [13][14][15]. High immunization coverage of two vaccines could facilitate effective control of the spread of corresponding viruses [16]. However, the Technical guideline for the use of COVID-19 vaccines (1st Edition) issued by the National Health Commission of the People's Republic of China requires a minimum interval of 14 days between the COVID-19 vaccine and any other vaccines. Additional clinical visits are inconvenient for many people and thus could have a negative impact on vaccine uptake. Concomitant administration may help to ease the inconvenience, minimize distress caused by multiple injections, and reduce vaccine stocking and administration costs associated with separate vaccinations [17]. However, evidence about the safety and immunogenicity of concomitant administration of inactivated COVID-19 vaccine and influenza vaccine are limited up to date.
Therefore, we conducted a phase 4 clinical trial to prove that the immunogenicity of CoronaVac was non-inferior between healthy adults aged 18-59 years who received the CoronaVac and IIV4 concomitantly and those who received two vaccines separately with 14 days apart.
Study design and participants
We conducted a randomized, open-label, noninferiority, phase 4 clinical trial in Kaihua County, Zhejiang, China, from March to May 2021 (clinicaltrials.gov number, NCT04801888). The study protocol and informed consent form were approved by the ethics committee of the Zhejiang Provincial Center for Disease Control and Prevention (ZJCDC). This study was performed by investigators from the ZJCDC following the Declaration of Helsinki of the World Medical Association Good Clinical Practice and the Chinese regulatory requirements. Before enrolment, written informed consent was obtained from each participant.
Healthy adults aged 18-59 years who provided written informed consent were eligible for enrolment in this study. The exclusion criteria included: history of any direct or indirect contact with suspected or confirmed COVID-19 cases within the previous 14 days, history of infection with SARS-CoV-2, any history of a severe allergy to any vaccination, any specified comorbidities that may influence the immune response of vaccination or lead to severe adverse events, and any prior administration of the influenza vaccines, investigational products, blood products, or immunosuppressive therapy, within a defined period (Supplementary appendix 1). The eligibility of participants was assessed through medical history inquiry and physical examination by the investigators.
Vaccines
CoronaVac is an inactivated vaccine against COVID-19 developed by Sinovac Life Sciences Co., Ltd. The production process of CoronaVac was consistent with vaccines used in the previous series of clinical trials [18][19][20] Both vaccines were prepared in Good Manufacturing Practiceaccredited facilities. Both vaccines were stored at 2-8°C and administered intramuscularly to each participant, with different vaccines injected into different arms. Two doses of CoronaVac (lot number: 20200412) were given 28 days apart and IIV4 (lot number: 202007007) was given in single dose.
Randomization
Each eligible participants was assigned a random number according to the sequence of enrolment and then randomized into two groups (C group: concomitant administration group, S group: separate administration group) at a ratio of 1:1. Meanwhile, participants in the C group were concomitantly randomly assigned into two subgroups (C1 and C2) at a ratio of 1:1. The randomization was carried out according to a randomization list (block size = 8) prepared by an independent statistician using SAS 9.4 (SAS Institute Inc., Cary, NC, USA). The random allocation was sealed using a numbered ''scratch card". The opaque material covered on the grouping information were not allowed to be scratched off unless the randomization has completed.
Procedures
After obtaining the first blood samples from all the participants at baseline (Day 0), two doses of CoronaVac and one dose of IIV4 were administered: for participants in the C group, IIV4 was administered concomitantly with the 1st dose (C1 subgroup, Day 0) or 2nd dose (C2 subgroup, Day 28) of the CoronaVac; for participants in the S group, IIV4 was administered between the two doses of CoronaVac, 14 days apart from each dose. Subsequent blood samples were collected as scheduled: on Day 28 (before the 2nd vaccination) and Day 56 after the first dose from participants in the C groups; on Day 14 (before IIV4 administration), Day 42 and Day 56 after the first dose from participants in the S group. Blood samples collected on Day 14 and Day 42 in the S group were for the detection of antibodies against influenza viruses only.
The participants were observed for immediate adverse events (AEs) for 30 min after each dose. Fever or any other AEs occurring within 28 days after each dose were recorded on diary cards (solicited AEs were collected during Days 0-7), which were checked by investigators to assure completeness and accuracy. The AEs were graded according to a scale issued by the National Medical Product Administration [21].
Immunogenicity assessment
The Immunogenicity of CoronaVac and IIV4 were evaluated by comparing the seropositivity rate and seroconversion rate of neutralizing antibodies, and the geometric mean titer (GMT) and geometric mean increase (GMI) of neutralizing antibody to live SARS-CoV-2 (virus strain SARS-CoV-2/human/CHN/CN1/2020, GenBank number MT407649.1) or to four seasonal influenza strains (A/Gu angdong-Maonan/SWL1536/2019, A/Hongkong/2671/2019, B/ Washington/02/2019, B/Phuket/3073/2013) at baseline and 28 days after the last dose of each vaccine. The primary endpoint for immunogenicity was the seropositive rate of neutralizing antibodies against SARS-CoV-2 28 days after the second dose of Coro-naVac. The neutralizing antibody titers against SARS-CoV-2 and influenza virus were measured by the quantified micro cytopathogenic effect assay (CPE) [22] and the hemagglutinin inhibition (HI) assays [23], respectively.
For antibody responses to SARS-CoV-2, seropositivity was defined as the neutralizing antibody titer to be ! 1:8, seroconversion was defined as that the neutralizing antibody titer is < 1:8 before immunization and ! 1:8 after immunization, or the neutralizing antibody titer is ! 1:8 before immunization and increases by 4 times and above after immunization. For antibody responses to influenza strains, seropositivity was defined as the HI antibody titer against certain antigen to be ! 1:40, seroconversion was defined as that the HI antibody titer is＜1:10 before vaccination and ! 1:40 after vaccination, or the HI antibody titer is ! 1:10 before vaccination and increased by four times after vaccination.
Statistical analysis
The assessment of immunogenicity was performed in corresponding per-protocol population, i.e., in those who were eligible, had received full doses of corresponding vaccines, and donated all blood samples required with corresponding detective neutralizing antibody values. Participants who received at least one dose of any vaccine were included in the safety analysis set (SS). The incidence rates of adverse reactions within 7 days, adverse reactions within 56 days, and SAE within 56 days were calculated as endpoints for the safety assessment.
The sample size was calculated using NCSS-PASS software (version 11.0) according to the seroconversion rate of SARS-CoV-2 neutralizing antibody at 28 days after the 2nd dose under a noninferiority design between the C and S groups. Non-inferiority would be considered as achieved if lower bound of the two-sided 95% CI of the difference for the seroconversion rate of antibodies against corresponding virus was > -10%. Assuming the seroconversion rate of SARS-CoV-2 neutralizing antibody in the S group was 90%, we estimated that a sample size of 205 in each group would achieve 90% power to detect a 10% difference between groups with an a of 0.05. Considering the potential loss to follow-up of 15%, the final sample size was 480 participants, with 240 in each group. The chisquare test or Fisher's exact test was used to analyze categorical data. The Wilcoxon rank sum test or log-transformed Group t test was used to compare the log-transformed antibody titer. Statistical analyses were performed by independent statisticians using SAS 9.4, and the hypothesis tests were two-sided with an a value of 0.05.
Participants
In total, 484 volunteers were recruited from March 23 to March 30, 2021, of whom 480 (99.7%) were enrolled and underwent randomization, with 120, 120, or 240 assigned into the C1 subgroup, C2 subgroup or S group, respectively. Overall, 459 participants were included in the PPS including 116 in the C1, 116 in the C2 subgroup, and 227 in the S group for the CoronaVac immunogenicity analysis. A total of 462 participants were included in the PPS including 118 in the C1, 116 in the C2, and 228 in the S group for the IIV4 immunogenicity analysis. Information on recruitment, enrolment, randomization, withdrawal and study completion were illustrated in Fig. 1. Participants in three groups were comparable in terms of mean age, sex, ethnicity, height and weight (Table 1).
Immunogenicity
At baseline, all participants were seronegative for SARS-CoV-2 ( Table 2, Fig. 2). The baseline seropositivity rates and GMT of antibodies against four types of influenza strain (A/H1N1, A/H3N2, B/ Yamagata, B/Victoria) showed no difference among the three groups. Pre-existing antibodies against A/H3N2 and B/Yamagata strain were observed with the pre-vaccination seropositivity rates to be over 60% and approximately 30%, respectively ( Table 2, Table 3).
Twenty-eight days after the 2nd dose of CoronaVac, 216 out of 232 (93.1%) participants in the C group and 216 out of 227 (95.2%) in the S group developed neutralizing antibodies against SARS-CoV-2 with no significantly different seroconversion rate (P = 0.351). Participants in the S group had higher GMT (38.1 vs. 27.5, P < 0.001) and GMI (19.0 vs. 13.8, P < 0.001) of antibodies against SARS-CoV-2 as compared to the C group.
In the analysis performed between subgroups within the C group, only 29 out of 116 (25.0%) participants and 30 out of 116 (25.9%) in the C1 and C2 subgroup, respectively, developed neutralizing antibodies against SARS-CoV-2 28 days after the 1st does of CoronaVac, with no significantly different seroconversion rate (P = 0.880). Twenty-eight days after the 2nd dose of CoronaVac, participants in the C1 subgroup showed significantly higher seroconversion rate (96.6% vs. 89.7%, P = 0.038), GMT (32.6 vs. 23.3, P = 0.015) and GMI (16.3 vs. 11.6, P = 0.015) of antibodies against SARS-CoV-2 as compared to the C2 subgroup. Additional analyses were performed to explore whether pre-vaccination antibodies against influenza virus would affect the immune response against SARS-CoV-2. Participants without pre-existing antibodies against influenza viruses A/H3N2 or B/Yamagata showed similar immune responses against SARS-CoV-2 among different groups (Supplementary appendixes 2-3).
There was no difference, in terms of seropositivity rate, seroconversion rate, GMT and GMI of antibodies against four types of influenza strains that contained in the IIV4 vaccine between participants in C1/C2 subgroup and S group, except that: 1) as compared to S group, participants in the C group had a higher seropositivity rate of antibody against A/H1N1 (93.6% vs. 87.3%, P = 0.021) with more participants getting higher antibody titer; 2) participants in the C1 group showed a higher seroconversion rate (95.8% vs. 88.8%, P = 0.046) of antibodies against A/H3N2 than that in the C2 subgroup.
Comparison of antibody responses of C1 subgroup with S group showed similar result with the comparison result between C group and S group that participants in the S group had higher GMT (38.1 vs. 32.6, P = 0.01) and GMI (19.0 vs. 16.3, P = 0.01) of antibodies against SARS-CoV-2 as compared to the C1 subgroup, while the participants in the C1 subgroup had a higher seropositivity rate of antibody against A/H1N1 (94.9% vs. 87.3%, P = 0.025) as compared to the S group (Supplementary appendix 4).
According to predefined non-inferiority margin, non-inferiority of the C group versus the S group in antibody responses against 4 types of viruses was achieved, not including antibody against the B/Yamagata. However, non-inferiority of the C group versus the S group in antibody responses against B/Yamagata was achieved if analysis was performed based on participants without prevaccination antibodies against corresponding virus strain based on an exploratory analysis.
Safety
Overall, 91 participants reported 130 vaccine-related adverse reactions during the 0-56 Day period after the first dose, with a rate of 20.0% in the S group, 16.7% in the C1 subgroup, and 19.2% in the C2 subgroup, respectively. The occurrence rates of both solicited and vaccination-related unsolicited AEs were similar among three groups (Table 4). Relative higher occurrence rate of pain at the injection site within 0-7 days after co-administration of IIV4 and CoronaVac in the C2 subgroup and after administration of IIV4 in the S group were observed (Supplementary appendixes [5][6][7][8]. The most frequently reported adverse reaction was a pain at the injection site, with an occurrence rate over 10% in all three groups. Most of the AEs were mild and moderate. Only one grade 3 (severe) fever with a maximum axillary temperature of 39.3°C in the C1 subgroup was observed and verified not due to SARS-CoV-2. This case was resolved within one week. No vaccine-related serious adverse events were observed in this study.
Discussion
This is the first report on the immunogenicity and safety of inactivated COVID-19 vaccine concomitantly administered with IIV4 among adults aged 18-59 years. Our study indicated that both concomitant or separate administration of CoronaVac and IIV4 could induce sufficient immunogenicity and had satisfactory safety profiles that concomitant administration strategy would be feasible. The outcome provided the scientific evidence for improving the immunization strategy of both inactivated COVID-19 vaccine and IIV4.
In the previous phase 2 clinical trial of CoronaVac conducted in the same age group, the seroconversion rate of antibody against SARS-CoV-2 was 97.4% (114 out of 117) with a GMT of 44.1 (95 %CI: 37.2-52.2) in participants receiving 3 lg CoronaVac following a 0-28 Day schedule [18]. The same vaccine formulation and vaccination schedule were used in this study. In our study, the immune response against SARS-CoV-2 in participants receiving COVID-19 and influenza vaccines separately showed similar results with previous phase 2 trial, while a moderate decrease of the immune response to CoronaVac was observed in the participants *Non-inferiority was achieved, as the lower bound of the two-sided 95% CI was > À 10% for seropositive rate and seroconversion rate. N, n: number of participants; SARS-CoV-2: severe acute respiratory syndrome coronavirus 2; C group: concomitant administration group, IIV4 was administered concomitantly with the 1st or 2nd dose of the SARS-CoV-2 inactivated vaccine on Day 0 or Day 28, respectively; S group: separate administration group, IIV4 was administered separately on Day 14; BI: before injection; GMT: geometric mean titre; GMI: geometric mean increase of titre.
For antibody responses to SARS-CoV-2, seropositivity criteria: The neutralizing antibody titer is ! 1:8; seroconversion criteria: the neutralizing antibody titer is < 1:8 before immunization and ! 1:8 after immunization; or the neutralizing antibody titer is ! 1:8 before immunization and increases by 4 times and above after immunization. For antibody responses to influenza strains, seropositivity criteria: The protective level of HI antibody titer against certain antigen is ! 1:40; seroconversion criteria: with 1:10 as the minimum dilution of HI antibody against certain antigen, the HI antibody titer is ＜1:10 before vaccination and ! 1:40 after vaccination; the HI antibody titer is ! 1:10 before vaccination and increased by four times after vaccination.
receiving vaccines under a concomitant administration regimen (IIV4 with the first or second CoronaVac). However, a similar immune response against SARS-CoV-2 was detected 28 days after the first vaccination in participants receiving either their first Cor-onaVac dose and IIV4 concomitantly or their first CoronaVac dose alone. The unexpected decrease in the immune response against SARS-CoV-2 seemed to be mainly observed in participants receiving IIV4 concomitantly with their second dose of CoronaVac, suggesting that this regimen may induce a slight interference with the immune response to CoronaVac. Up to date, three reports of co-administration of seasonal influenza vaccine or high-dose seasonal influenza vaccine with a protein subunit COVID-19 vaccine with Matrix-M adjuvant (NVX-CoV2373) [24][25], or an adenoviral vector (ChAdOx1) or mRNA COVID-19 vaccine (BNT162b2) [26] have been published. In the study of NVX-CoV2373 [24], decrease of the immune response against NVX-CoV2373 was observed following a concomitant regimen of seasonal influenza vaccine with the first dose of NVX-CoV2373 which may suggest an immune interference. However, there was no comparison group included to study the concomitant administration of the seasonal influenza vaccine with the second dose of NVX-CoV2373, as well as no detection of the immune response to the COVID-19 vaccine before the second dose [24]. Therefore, it is not clarified whether the immune interference observed in the study of NVX-CoV2373 would be consistent with our findings if the comparison group was included and the detection was performed. However, for the study of NVX-CoV2373 concomitantly administered with a high-dose quadrivalent influenza vaccine in adults aged ! 65 years [25], no decrease or increase of immune response against each influenza strain (A/H1N1, A/ H3N2, B/Yamagata and B/Victoria) or SARS-CoV-2 were observed, which is different from the previous NVX-CoV2373 study and our findings, but similar with the study of ChAdOx1 and BNT162b2 [26], in which no interference of the immune response to COVID-19 vaccines were observed in participants receiving seasonal influenza vaccine with the second dose of ChAdOx1 or BNT162b2. Further studies are needed to explain this finding and assess whether such limited differences in immunogenicity are clinically relevant. Moreover, the low seropositivity of antibody against SARS-CoV-2 twenty-eight days after the first vaccination with CoronaVac provided additional evidence in line with results acquired in the previous phase 2 clinical trial of CoronaVac that at least 2 doses are needed for the primary immunization of inactivated COVID-19 vaccines.
Considering the immune response of antibodies against influenza strains, antibody GMT against influenza A/H3N2 was observed to be relatively high in both study groups (703.3 to 724.8) while the value was 193.7 in a previous phase 3 clinical trial [27]. As the baseline value of seropositivity rate of antibody against influenza A/H3N2 and the GMT were respectively 66.7% and 40.0, this could indicate that participants in our study already had relatively high pre-existing antibodies as compared to those in a previous phase 3 clinical trial. Furthermore, our reanalysis of noninferiority of antibody responses against influenza B/Yamagata based on data from participants with or without pre-existing antibodies were different, suggesting that corresponding pre-existing antibodies may affect post-vaccination immune responses and further study are needed to address this aspect.
Concomitant administration of CoronaVac with IIV4 and separate administration of the corresponding two vaccines 14 days apart in this study showed satisfactory and good safety profiles. The occurrence of vaccine-related adverse reactions through all treatment groups was no more than 20%, within which 98.9% were mild or moderate. These findings were similar to previous clinical trials of both inactivated COVID-19 vaccine and seasonal influenza vaccine [18,27], and also to other inactivated COVID-19 vaccines Table 3 Antibody responses to SARS-CoV-2, 4 types of influenza strain before first vaccination and 28 days after last vaccination (per protocol sets) by subgroups. For antibody responses to SARS-CoV-2, seropositivity criteria: The neutralizing antibody titer is ! 1:8; seroconversion criteria: the neutralizing antibody titer is < 1:8 before immunization and ! 1:8 after immunization; or the neutralizing antibody titer is ! 1:8 before immunization and increases by 4 times and above after immunization. For antibody responses to influenza strains, seropositivity criteria: The protective level of HI antibody titer against certain antigen is ! 1:40; seroconversion criteria: with 1:10 as the minimum dilution of HI antibody against certain antigen, the HI antibody titer is ＜1:10 before vaccination and ! 1:40 after vaccination; the HI antibody titer is ! 1:10 before vaccination and increased by four times after vaccination. studies [28,29], indicating that co-administration of these two vaccines did not increase reactivity in any concerning manner.
There are some limitations of this study. First, no parallel control groups with participants receiving CoronaVac or IIV4 only were included during the time of this study, which may have affected the explanation of the absolute immune response of the concomitant administration of the corresponding two vaccines. However, the phase 2 clinical trial of CoronaVac and phase 3 clinical trial of IIV4 were performed within 1-3 years before this study in a similar population (18-59 aged adults). The results of this study showed similar immunogenicity and safety profiles with two previous clinical trials, which may indirectly indicate the reliability of this study. Second, as the primary endpoint of this study focused on the immune response against SARS-CoV-2, the sample size of each group may not have been sufficient to power the conclusion about influenza viruses, and further studies are needed. Third, this study was conducted on 18-59 aged people that the result of this study couldn't be used to support the immunization strategy of concomitant injection of CoronaVac and IIV4 in older people aged 60 years and older, and further studies are needed. Finally, further immunological studies are warranted to precisely explain the differences of the immune response to CoronaVac when IIV4 is co-administered with the first or with the second dose of CoronaVac, which has not been previously reported in the literature to our knowledge.
Conclusions
Co-administration of inactivated COVID-19 vaccine and seasonal influenza vaccine, especially the administration regimen that the seasonal influenza vaccine co-administered with the first dose of the inactivated COVID-19 vaccine, would be feasible.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. | 2022-08-04T05:06:48.942Z | 2022-07-01T00:00:00.000 | {
"year": 2022,
"sha1": "5980205319748c9291d4392d3f5dda1ac516f8a8",
"oa_license": null,
"oa_url": "https://doi.org/10.1016/j.vaccine.2022.07.021",
"oa_status": "BRONZE",
"pdf_src": "PubMedCentral",
"pdf_hash": "5980205319748c9291d4392d3f5dda1ac516f8a8",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": []
} |
233854184 | pes2o/s2orc | v3-fos-license | Investigation and Protection of Ancient and Famous Trees Resources of Daxiong Mountain
In the paper, the authors presents current situation of ancient and famous trees resources in Daxiong Mountain; the characteristics of ancient and famous trees resources is analyzed: (1) There were a large number of trees, and most of them appear in communities. (2) There are many old and tall trees with beautiful tree shapes and different postures. (3) There are many rare and precious old trees. Third, investigated the Problems existing in protection and management of ancient and famous trees in Daxiong Mountain and the reasons. Final, protection countermeasures of ancient and famous trees in Daxiong Mountain is put forward.
Taking the ancient trees growing in groups of more than 3 trees as an ancient tree group, there are 124 ancient tree groups in Daxiong mountain, forming communities' area of 475.6 hm² Lizhong word area has the largest communities area of 180.1 hm², and its number of ancient and famous trees is 6,025, which is related to the large number of secondary forests. The second is the Daxiong work area, with the communities' area of 104.2hm² and 3090 ancient and famous trees (Table 3).
There were a large number of trees ,and most of them appear in communities.
The ancient and famous trees of Daxiong Mountain are scattered all over the mountains and fields. At the foot of the mountain, tall old trees are obvious. And they can be seen everywhere in the mountain. In each work area of the Daxiong mountain, Crataegus hupehensis, with DBH more than 20cm and height over 6m can be seen everywhere. Some largest Crataegus hupehensis, whose DBH were more than 50cm, can be said to be unique in Hunan province. On A hillside of mountaintop, there are patches of Cornus controversa, with DBH above 35cm, branches low and branches above 20cm in diameter, which are relatively rare. However, in the montane elfin forest and secondary forests, some small trees are hundreds of years old.
There are many old and tall trees with beautiful tree shapes and different postures.
At the foot of the mountain, the Pine king, with DBH 115cm, height 45m, age more than 320 years, is tall and straight, natural and unrestrained. There are two large ginkgo trees from Xiongshan Temple and Xiquan Temple with DBH more than 300cm. The base of the two tree is densely sprouted , unique even stalactitic adventitious roots formed. The roots are really rare. In the old-growth forest of Lizhong, the Cercidiphyllum japonicum's DBH is 173 cm and its height is 14m. It is more than 600 years old and has a very beautiful shape (Table 4). (Table 4).
Problems existing in protection and management of ancient and famous trees in Daxiong Mountain and the reasons.
The statistical analysis showed that among the 423 scattered ancient and famous trees in Daxiong Mountain [7], 113 were in average growth and 21 were in poor growth condition or even dying. The situation was not optimistic.Damage to ancient trees can sometimes be seen. Such as a maple of Shiliping was eroded by termites; By the Xiongshan Temple, one of Carpinus viminea and one of Castanea henryii were in the dying state, and Another Carpinus viminea had badly broken branches.
There is a dying of Lauraceae obtusiloba by the tourist trail on the mountaintop. In Baxianyan scenic spot the leaves of a maple were eaten up by insects. In Jinping work area, one of Castanopsis eyrei and one of Schima argentea were seriously hollow, and one of Castanopsis eyrei, had broken tip. We investigated the reasons for these things. One was the physiological function decline of the ancient trees. Ancient trees were basically older and overmature trees, which had weakened their absorption and regeneration ability and were difficult to meet the needs of growth. Second, the pests and EMCEME 2020 IOP Conf. Series: Earth and Environmental Science 692 (2021) 032009 IOP Publishing doi:10.1088/1755-1315/692/3/032009 6 diseases affected them. Most of the ancient trees had not been effectively protected and managed, and some were weak and vulnerable to pests and diseases. Third, natural disasters affect them. Some ancient trees have not been managed for a long time and suffer from natural disasters, such as lightning strikes and other natural disasters. The phenomenon of broken branches and broken trunks often occurs. Fourth, environmental factors affect them .
The artificial over-stepping and the laying of cement or other hard impermeable floor tiles around ancient trees made the soil around ancient trees rigid, air permeability and water permeability decrease, which affected the water absorption and gas exchange of the trees, therefore the nutrients needed by trees could not be replenished in time. Fifth, human damage. With the gradual acceleration of tourism development in Daxiongshan, tree damage, root system destruction, artificial logging and other phenomena had occurred from time to time in the construction process. In villages and courtyards, people randomly piled up wastes and waste materials near ancient trees, dumping sewage and waste water, which changed the physical and chemical properties of the surrounding soil and further damaged the ancient trees. Sixth, insufficient investment in protection funds. At present, there is basically no fund for the protection of ancient and famous trees at all levels, and the people responsible for the protection of ancient and famous trees are generally the district heads and branch secretaries of the work areas. The owners of ancient trees next to houses are the protectors, and there is almost no fund for the custodians. Therefore, many ancient and famous trees are basically left to die by themselves and have not been effectively protected. Seventh, protection and management measures are backward, and effective modern management means have not been established, resulting in inadequate management.
Protection countermeasures of ancient and famous trees in Daxiong Mountain
First, we should broaden financing channels and increase capital input. We should establish a diversified investment mechanism and encourage all sectors of the society to donate money for the adoption of ancient and famous trees. Second, we should strengthen the management of the ancient and famous trees. We should make clear who is responsible for the care of ancient and famous trees. For old trees with long age and weak growth potential, it is necessary to timely take rescue treatment and protection measures, and carry out related research to restore the tree potential [8][9][10][11][12].
Third, establish the digital information platform of ancient and famous trees. The information database of species, quantity, distribution and growth status of ancient and famous trees should be established comprehensively and systematically.Fourth, make full use of various media to publicize the knowledge of ancient tree protection. Fifth, strengthen publicity and law enforcement efforts to effectively protect ancient and famous trees. We should crack down on the man-made destruction of ancient and famous trees. | 2021-05-07T00:04:18.103Z | 2021-03-01T00:00:00.000 | {
"year": 2021,
"sha1": "859d011a24db004ed3104d6fc6e2835bab2335ae",
"oa_license": null,
"oa_url": "https://doi.org/10.1088/1755-1315/692/3/032009",
"oa_status": "GOLD",
"pdf_src": "IOP",
"pdf_hash": "115f5506e6bed7872cdb15503ac74ea02f8e693d",
"s2fieldsofstudy": [
"Environmental Science"
],
"extfieldsofstudy": [
"Physics"
]
} |
59325260 | pes2o/s2orc | v3-fos-license | Neutrinophilic two Higgs doublet model with dark matter under an alternative $U(1)_{B-L}$ gauge symmetry
We propose Dirac type active neutrino with rank two mass matrix and a Majorana fermion dark matter candidate with an alternative local $U(1)_{B-L}$ extension of neutrinophilic two Higgs doublet model. Our dark matter candidate can be stabilized due to charge assignment under the gauge symmetry without imposing extra discrete $Z_2$ symmetry and the relic density is obtained from $Z'$ boson exchanging process. Taking into account collider constraints on $Z'$ boson mass and coupling, we estimate the relic density.
I. INTRODUCTION
An origin of tiny neutrino masses is one of the unsolved issues in the standard model (SM) where neutrino mass type can be either Dirac or Majorana type. Majorana type neutrino mass can be realized in many scenarios such as type-I seesaw mechanism in which heavy SM singlet right-handed neutrinos are introduced. On the other hand, Dirac type neutrino mass can also be obtained as charged leptons by introducing right-handed neutrinos without Majorana mass term. In such a scenario, neutrinophilic two Higgs doublet model is suggested to avoid large hierarchy in the Yukawa couplings where only one Higgs doublet with small vacuum expectation value (VEV) has Yukawa interaction among lepton doublets and righthanded neutrinos giving neutrino masses [1][2][3]. This kind of Higgs doublet model can be constructed by imposing symmetry such as global U(1) symmetry [1,3]. It is also interesting to consider realization of neutrinophilic two Higgs doublet model based on an exotic U (1) gauge symmetry such as U(1) B−L . Then we consider alternative U(1) B−L charge assignment for right-handed neutrinos [4][5][6][7] since original U(1) B−L charge assignment is not suitable due to universal B − L charge for leptons including right-handed neutrinos.
In this paper, we construct a neutrinophilic two Higgs doublet model based on an alternative U(1) B−L gauge symmetry which introduces three right-handed neutrinos ν R 1 , ν R 2 and ν R 3 with B −L charge −4, −4 and 5 to cancel gauge anomalies. We also assign B −L charge −3 to one of the Higgs doublets, and the B − L charged Higgs doublet only has Yukawa couplings among SM lepton doublets and right handed neutrinos ν R 1,2 . Thus a Dirac type neutrino mass matrix is obtained after electroweak symmetry breaking and the smallness of neutrino mass can be explained by the small VEV of the B − L charged Higgs doublet by choosing parameters in the scalar potential appropriately. Furthermore ν R 3 is stabilized by an accidental Z 2 symmetry due to the charge assignment and can be a good dark matter (DM) candidate. Here we emphasize that three fermion contents inducing active neutrino mass and providing DM are required by anomaly cancellation condition, and our charge assignments for the fermions naturally guarantee the stability of DM without inducing further symmetry such as discrete Z 2 symmetry. Then we discuss phenomenology of the model such as Z ′ boson at collider experiments and DM physics. The constraints on Z ′ mass and U(1) B−L gauge coupling can be obtained by the data of LEP experiment and the current LHC experiments. Taking into account the constraints, DM relic density is estimated by Higgs doublet H that is the same as SM. In addition, we introduce isospin singlet scalars 1 At non-renormalizable level, we would have dimension 7 operator such asν c Ri X R ϕ * 10 (ϕ 3 ) 3 . We consider such a term is highly suppressed by sufficiently large cut-off scale and does not affect stability of DM and phenomenology. {ϕ 10 , ϕ 3 } with nonzero B − L charges denoted by subscripts where ϕ 10 is required to give X R mass and ϕ 3 is necessary to avoid massless Goldstone boson from Higgs doublet Φ. All the field contents and their assignments are summarized in table I. Under these framework, one finds the following renormalizable Lagrangian: where we assume the parameters in the potential are real,H ≡ (iσ 2 )H * with σ 2 being the second Pauli matrix, a runs over 1 to 3, and i runs over 1 to 2.
A. Scalar sector
The scalar fields are parameterized as where Q = {3, 10} indicates B−L charges for singlet scalar fields, the lightest mass eigenstate after diagonalizing the matrix in basis of (w ± , φ ± ), which is massless, is absorbed by the SM W ± boson, and two of the mass eigenstate in CP odd boson sector (φ I , z, z ′ Q ) are also absorbed by the SM Z boson and B − L gauge boson Z ′ , as Nambu-Goldstone(NG) bosons. The VEVs are obtained by applying the conditions ∂V /∂v Q = 0, ∂V /∂v H = 0 and where we assume the relation v 2 The small v φ can be realized taking trilinear coupling µ to be sufficiently small. Note that z ′ 10 is dominant component of NG boson which is absorbed by Z ′ boson since we consider VEV of ϕ 10 is much larger than the others. After ϕ 10 developing VEV at high energy scale, our scalar sector has the same structure as discussed in ref. [3]. Then CP-odd component of ϕ 3 ; z ′ 3 , becomes physical massless Goldstone boson(GB) due to a global symmetry in the scalar potential. However the existence of this physical Goldstone boson does not cause serious problem in particle physics or cosmology since it does not couple to SM particles directly except for Higgs boson whose couplings are well controlled by the parameters in the potential, and decouples from thermal bath in early Universe. Here we discuss the condition that GB decouples from thermal bass at sufficiently early Universe following discussion in ref. [8]. Since scalar and gauge bosons which couples to GB are heavy their interactions with GB decouple at sufficiently high temperature.
We thus focus on interaction between GB and the SM fermions. The ratio between collision and expansion rates is roughly given by where we take Boltzmann constant as 1, m pl is Planck mass, and m f denotes an SM fermion mass; the process GB GB ↔ f f is induced by the Higgs-ϕ 3 mixing for m f < T . The decoupling occurs when R(T ) ∼ 1 and we obtain decoupling temperature T d ∼ 0.42/λ 2/5 GeV temperature and cosmology is not affected by GB; note that SM fermions with m f < m τ decouples earlier due to smaller couplings.
In our scenario, one finds that GeV since v φ is expected to be tiny in order to generate the active neutrino masses. Thus charged component w ± in H is approximated to be NG boson which is absorbed by W ± boson while the φ ± from Φ is physical charged Higgs boson. Similarly the CP-odd boson z is absorbed by the neutral SM gauge boson Z while φ I is physical CP-odd Higgs. The masses of physical charged Higgs and CP-odd Higgs are approximately given by [3] The mass matrix for the CP-even scalars has 4 × 4 structure in a basis of (h, φ R , ϕ 3R , ϕ 10R ).
In our analysis, we omit details of the matrix assuming SM Higgs is the lightest component among four physical CP-even scalar bosons. In addition, we assume mixing among SM Higgs and other CP-even scalars are small to avoid experimental constraints for simplicity. More details of the scalar sector can be found in refs. [1,3], and we focus on Z ′ and DM physics in the following analysis below.
B. Fermion sector
The masses for charged-leptons are induced via y ℓ after symmetry breaking, and active neutrino masses are also done via y ν term where neutrinos are supposed to be Dirac type fermions. Their masses are symbolized by m ℓa ≡ v H y ℓa / √ 2 and m ν ai ≡ v φ y ν ai / √ 2. Since the charged-lepton mass matrix is diagonal, the neutrino mixing matrix V ab is arisen from diagonalizing the neutrino mass matrix squared as where V is measured PMNS matrix by the neutrino oscillation data [9]. Notice here that one of the active neutrinos is massless due to the rank two matrix. Thus one can parametrize the neutrino mass matrix in terms of observables and arbitrary parameters in the following form: where O a ′ i is generally an arbitrary three by two matrix with complex values, satisfying GeV 2 . Note here that correlations between them seem to occur due to the manner of our parametrization. Since (m ν ) ai = v φ y ν ai / √ 2, the order of Yukawa coupling is 10 −12∼11 /v φ ; when v φ ∼ O(KeV) the order of coupling is around 10 −6 which is similar to electron Yukawa coupling.
The third right-handed neutrino obtains Majorana mass term from the term with f X after ϕ 10 developing the VEV: ϕ 10 = v 10 / √ 2. The Majorana mass is simply given by (II.12)
III. PHENOMENOLOGY
In this section, we consider phenomenology of our model focusing on Z ′ boson and dark matter candidate.
A. Z ′ boson
Here we consider constraints from collider experiments for Z ′ boson mass m Z ′ and U(1) B−L gauge coupling g BL . The gauge interactions of Z ′ and fermions are given by where Q f BL is the charge of U(1) B−L symmetry, and f SM denotes all the electrically charged fermions in SM. Here Z ′ mass is given by m Z ′ ≃ 10g BL v 10 as we assume v 10 ≫ {v 3 , v φ }.
Firstly, we discuss constraint from the LEP experiment. Since Z ′ couples to SM leptons, we obtain the following effective interactions considering Z ′ is sufficiently heavy; where ℓ ′ = e, µ and τ . In this case, we obtain constraints from the analysis for the process e + e − → ℓ ′+ ℓ ′− with the data of measurement at LEP [11]: In our following analysis, we take into account the constraint.
We next discuss the Z ′ production at LHC. In hadron collider experiments, Z ′ boson can be produced via the process qq → Z ′ where q indicates SM quarks. Here we estimate the production cross section using CalcHEP [12] implementing the relevant interactions with the CTEQ6 parton distribution functions (PDFs) [13]. Then Z ′ decays into B − L charged particles where we only consider fermions assuming masses of scalar bosons are greater than m Z ′ /2. The decay width is given by where f denotes any fermion in the model, N f c is color factor, and we used m f /m Z ′ ≪ 1 for fermions except for top quark and X. The branching ratio for a mode Z ′ → ff is given by The branching ratio for charged lepton modes are less than ∼ 0.05 since Z ′ dominantly decays into ν R 1,2 due to the charge assignment. In Fig. 2, we show the product of cross section and branching ratio, σ(pp → Z ′ )BR(Z ′ → ℓ + ℓ − ) where ℓ = e and µ, at √ s = 13 TeV as a function of m Z ′ with g BL = {0.1, 0.05, 0.01} which is compared with the limit from the LHC data [14]. We thus find that m Z ′ 2.8 TeV is required for g BL = 0.1 while m Z ′ ≃ 500 GeV is still allowed for smaller gauge coupling g BL = 0.01. Note that the constraint on Z ′ mass is weaker than that in original U(1) B−L case [15] since our Z ′ dominantly decays into light right-handed neutrinos ν i which has larger charge than the other SM fermions.
More parameter region will be tested by the data of the future LHC experiments with larger integrated luminosity.
B. Dark matter
In this subsection we discuss a dark matter candidate; X R . Firstly, we assume that any contributions from the Higgs mediating interactions are negligibly small and DM annihilation processes are dominated by the gauge interaction with Z ′ ; we thus can easily avoid the constraints from direct detection searches such as LUX [16], XENON1T [17], and PandaX-II [18,19]. Here we discuss the condition for Higgs portal interaction from direct detection constraints. The nucleon-DM interaction is induced by Higgs portal interaction via mixing between the SM Higgs and ϕ 10 . The relevant effective interaction is given by [3] L eff = q f X m q sin θ cos θ 2 √ 2vm where θ is mixing angle for Higgs-ϕ 10 mixing, m q is a mass of quark q and we assumed heavier scalar boson is much heavier than m h . Then X-N spin independent scattering cross section is obtained as where m N is the nucleon mass, µ N X = m N m X /(m N + m X ) is the reduced mass of nucleon and DM and f N is effective coupling constant for Higgs-Nucleon interaction. For simplicity, we estimate the cross section applying f n ≃ 0.287 for neutron. Finally we estimate the cross section as Therefore direct detection constraints can be satisfied with small mixing angle such as sin θ ≪ 0.1 even if coupling f X is O(1). We next consider direct detection via Z ′ exchange.
The relevant effective interactions between DM and the SM quarks is given by where DM has only axial vector current due to Majorana property and we defined Λ Z ′ ≡ 6m Z ′ /(5g BL ). The operator in Eq. (III.9) induces spin dependent operator s ⊥ X · q and s X · ( s N × q) [20][21][22]; q is transferred momentum, s X(N ) is spin operator of DM(nucleon) and ⊥ indicate direction perpendicular to q direction. In Ref. [22], the lower limit of Λ Z ′ is given as ∼ 1 TeV which is obtained by data from PandaX, LUX and XENON1T including spindependent direct detection results [18]. This constraint is much weaker than the constraint from collider search of Z ′ as shown in Fig. 2. Thus constraint from direct detection is not stringent in our model.
Relic density:
We have annihilation modes via gauge interaction as XX → Z ′ → ff to explain the relic density of DM: Ωh 2 ≈ 0.12 [23], and their relevant Lagrangian in basis of mass eigenstate is given in Eq. (III.1). Then the relic density of DM is estimated by [24] Ωh 2 ≈ 1.07 × 10 9 , (III.10) where g * (x f ≈ 25) is the degrees of freedom for relativistic particles at temperature x 2 ) is given by [25] J where we implicitly impose the kinematical constraint above, C f = 1/2 for neutrino pairs including two light right-handed neutrino ν i in the second line of Eq. (III.1) otherwise C f = 1, and the width of Z ′ is given by Eq. (III.4).
In fig. 3, we show the relic density in terms of M X , where we fix parameters m Z ′ = {500, 1000} GeV and g BL = 0.01 which are allowed by the collider experiments. We find that the correct relic density can be obtained near the Z ′ pole since we need resonant enhancement due to small gauge coupling.
IV. CONCLUSION
In this paper, we have discussed neutrinophilic two Higgs doublet model with alternative for the right-handed neutrinos. Then two right-handed neutrinos ν R 1,2 have Yukawa coupling with SM left-handed neutrinos and we obtain 2 × 3 Dirac neutrino mass matrix predicting one massless neutrino. In addition, X R ≡ ν R 3 can be a good dark matter candidate since it is stabilized by an accidental Z 2 symmetry in our model due to the charge assignment of Then we have discussed phenomenology of the model such as Z ′ boson production at collider experiments and dark matter physics. Our Z ′ can be produced at LHC and can decay into SM leptons. We thus have estimated the cross section and branching ratios for Z ′ → ℓ + ℓ − and compared resulting values of the product of the cross section and branching ratio with the current LHC constraint. We find that Z ′ should be heavier than ∼ 2.8 TeV for gauge coupling g BL = 0.1 while m Z ′ ≃ 500 GeV is still allowed for smaller gauge coupling g BL = 0.01. We have found that the constraint on Z ′ mass is weaker than that in original U(1) B−L case, since our Z ′ dominantly decays into light right-handed neutrinos ν i that has larger charge than the other SM fermions. We have also estimated relic density of dark matter which is determined by the thermally averaged cross section of the processes XX → Z ′ → f f where f is any fermions in the model. We have shown that the observed relic density can be obtained with Z ′ mass and g BL allowed by the collider constraints. Our model can be further tested in future by both collider and dark matter experiments. | 2018-02-25T14:51:20.000Z | 2017-08-29T00:00:00.000 | {
"year": 2018,
"sha1": "557f12fff7a74d01d9a82095bba2bb10db8e4dea",
"oa_license": "CCBY",
"oa_url": "https://link.springer.com/content/pdf/10.1140/epjc/s10052-018-5667-6.pdf",
"oa_status": "GOLD",
"pdf_src": "Arxiv",
"pdf_hash": "557f12fff7a74d01d9a82095bba2bb10db8e4dea",
"s2fieldsofstudy": [
"Physics"
],
"extfieldsofstudy": [
"Physics"
]
} |
62789858 | pes2o/s2orc | v3-fos-license | Gravitational crystal inside the black hole
Crystals, as quantum objects typically much larger than their lattice spacing, are a counterexample to a frequent prejudice that quantum effects should not be pronounced at macroscopic distances. We propose that the Einstein theory of gravity only describes a fluid phase and that a phase transition of crystallization can occur under extreme conditions such as those inside the black hole. Such a crystal phase with lattice spacing of the order of the Planck length offers a natural mechanism for pronounced quantum-gravity effects at distances much larger than the Planck length. A resolution of the black-hole information paradox is proposed, according to which all information is stored in a crystal-phase remnant with size and mass much above the Planck scale.
Introduction and main ideas
The black-hole information paradox (see e.g. [1,2,3,4,5,6,7,8,9,10] for reviews) is one of the greatest challenges for a successful marriage between theory of gravity and quantum theory. Many proposals for the resolution of the paradox, such as massive black-hole remnant [1], fuzzball [11], energetic curtain [12], firewall [13], or Planck star [14], require a large deviation from classical gravity at macroscopic distances. On the other hand, classical gravity is generally expected to be a very good approximation at macroscopic distances, which raises justified skepticism about all such proposals of large deviations from classical gravity at macroscopic distances.
Nevertheless, it is not true that quantum effects cannot be significant at macroscopic distances. At low temperatures, some materials show macroscopic features such as superfluidity or superconductivity [15], which cannot be explained by classical physics. Furthermore, there are even examples which do not require low temperature. For instance, white dwarf stars avoid gravitational collapse owing to the electron degeneracy pressure [16], which is a consequence of the quantum Fermi-Dirac statistics. Similarly, neutron stars avoid gravitational collapse owing to the quantum degeneracy pressure of neutrons [16].
Finally, even if these examples look somewhat exotic, there is a totally non-exotic example; probably the best understood and most ubiquitous macroscopic quantum object is a crystal [17]. Crystals can be found everywhere in our everyday macroscopic lives. And yet, they are quantum objects which can be thought of as macroscopic "molecules". Indeed, the typical lattice spacing between the neighboring atoms in a crystal is a few Bohr radii, where the Bohr radius (in Gaussian CGS units) is a quantum distance, determined by the Planck constanth, electron mass m, and electron charge e. The lattice spacing in the crystal determines many of its macroscopic properties. In particular, the fact that spacing is of the order of a 0 implies that entropy density s in a crystal is of the order The analog of (1) in quantum gravity is the Planck length l Pl , given by in the unitsh = c = 1, where m Pl is the Planck mass. It is often assumed in the literature that effects of quantum gravity should not be seen at distances much larger than l Pl . Yet, the analogy between (1) and (3) suggests that such an assumption may not be justified. Just as the Bohr radius has a macroscopic manifestation in matter crystals, the Planck length could have similar macroscopic manifestation in a gravitational crystal. In particular, the analogy with (2) suggests that the entropy density of such a gravitational crystal could be of the order of the Planckian entropy density Can a gravitational crystal be described by the Einstein-Hilbert action of gravity? The classical and semi-classical gravitational phenomena have many similarities with phenomena in condensed-matter physics [18,19,20], but most of these condensedmatter phenomena are properties of fluid phases, while similarities between gravity and the properties of crystals are not so pronounced. Guided by general principles of condensed matter physics [17], we suggest that this is because the classical Einstein theory of gravity is really an effective macroscopic description of a fluid. In other words, we propose that the classical Einstein equation only describes the fluid phase of some unknown fundamental degrees of freedom. If so, then the crystal phase cannot be described by quantization of the Einstein-Hilbert action, implying that the Einstein-Hilbert action is not fundamental. Roughly, this is analogous to the fact that ice cannot be described by quantization of a non-fundamental macroscopic fluid equation, such as the Euler equation or the Navier-Stokes equation. The transition from the fluid phase (described by general relativity) to the crystal phase (described by as yet unknown theory) occurs via a phase transition (for a related idea see also [21]). Unfortunately, in the absence of detailed knowledge about the fundamental theory, the details of such a phase transition cannot be described explicitly from first principles.
A simple model
Even though we cannot derive the properties of such a gravitational crystal from first principles, some essential qualitative properties can be described by a simple model. For that purpose, we assume that crystallization happens under extreme conditions inside the black hole, in the core formed around the center at r = 0. Denoting by r core the radius of the crystallized core, general relativity is valid only in the fluid phase at r > r core . The physics in the core for r ≤ r core is not described by the Einstein-Hilbert action, and a priori the entropy of the core does not need to scale with area. Therefore, to model the core, we take the simplest possible assumption, i.e. we assume that the effective metric in the core is the flat Minkowski metric and that its entropy scales with volume in accordance with (4). (The flat metric in the crystal phase seems natural from a condensed-matter point of view because, in condensedmatter physics, the fundamental degrees of freedom do not describe metric. The existence of a curved metric as an effective large-distance description is characteristic for the fluid phase [19,20], but not for the crystal phase.) Hence the volume of the core is V core = 4πr 3 core /3 and its entropy is where α ∼ 1 and we work in Planckian units l Pl = m Pl = 1. (For a related model of a core inside the black hole see also [22].) On the other hand, the fluid phase outside the core is described by general relativity, so the fluid-phase entropy is the standard Bekenstein-Hawking entropy where A = 4πR 2 is the black-hole area, R = 2M is the black-hole radius, and M is the black-hole mass. The total black-hole entropy S bh is the sum of the two contributions Now let us assume that initially there is no core, so that the initial black-hole entropy is the initial Bekenstein-Hawking entropy where M 0 is the initial black-hole mass. Owing to the Hawking radiation [23], the mass M decreases and the black hole absorbs entropy of the ingoing Hawking quanta. This entropy cannot be absorbed by the fluid phase because the fluid-phase entropy A/4 decreases due to the decrease of M. Consequently, a crystal phase must be created in the center of the black hole in order to absorb the incoming entropy. The incoming entropy is the entropy of entanglement with external Hawking radiation, so the incoming entropy is equal to the radiation entropy S radiation . Hence the total black-hole entropy must be equal to The radiation entropy has been calculated by Page [24,25,26]. His result can be written as where η ≈ 1.5. Equating (7) with (9) and using (5), (6), (8) and (10), we obtain which describes how r core increases as the black-hole mass M decreases due to Hawking radiation. As the black-hole radius R = 2M shrinks, the crystal radius r core grows. This black-hole shrinking and crystal growth is a continuous process which lasts as long as Hawking radiation is created at the horizon at R = 2M. This happens as long as general relativity is valid at r ≥ R. However, at some point r core becomes equal to R, at which point general relativity ceases to be valid at the horizon. At this point there is no reason to expect any further creation of Hawking radiation, so the process of crystal growth stops at that point. At this critical point we have r core = R = 2M, so (11) gives which is a cubic equation for M.
The exact solution of this cubic equation can be expressed in an analytic form [27], but such an expression is rather cumbersome and not very illuminating. It is much more illuminating to find an approximate analytic solution of (12). Since α and η are of the order of unity, while M 0 ≫ 1, it is not difficult to see that the solution of (12) satisfies M ≪ M 0 . Hence (12) can be approximated by the equation the analytic solution of which has a very simple form with To better understand the physical content of (14), it is useful to restore the physical units in which m Pl = 1. In such units, (14) can be written as Since M 0 ≫ m Pl , this implies a hierarchy The mass M given by (16) is the mass of the black-hole remnant, the mass and size of which are much above the Planck scale. As a possible resolution of the black-hole information paradox, a massive remnant with mass of the order of (16) has also been suggested by Giddings in [1]. The problem with his suggestion, in his own words, was the acausal behavior of the core behind the horizon. Namely, if one assumes that general relativity is valid in the core, then causal evolution of the core is not compatible with an increasing core radius r core . But in our approach this is not a problem, because the core is in the crystal phase, while general relativity is valid only in the fluid phase outside the core.
Discussion
In the literature one can find many other proposals for the black-hole remnant (see [28] for a review), most of which have mass on the Planck scale or some other fixed scale which (unlike (16)) does not depend on M 0 . To resolve the black-hole information paradox for an arbitrarily large M 0 , such a remnant should be able to absorb an arbitrarily large amount of information. But object with a bounded mass and unbounded phase space is expected to have an unbounded probability of production in various quantum processes [1,2,3,4,5,6], contradicting the fact that such productions are not observed. There are also various counterarguments [28] according to which the overproduction of such objects should not necessarily be expected. But whatever one thinks of such counterarguments, we emphasize that in our scenario the mass of the remnant (16) increases with M 0 , so our remnant does not lead to such a problem at all.
In our scenario the entropy of the remnant is proportional to the volume rather than the area. At first sight, this may seem to be incompatible with AdS/CFT duality [29,30]. Nevertheless, there are two possibilities for making it compatible with AdS/CFT. The first possibility is that AdS/CFT duality is a property of an approximative theory (for example, it is possible that string theory is only an approximative theory of quantum gravity), while the gravitational crystal is described by some more fundamental, as yet unknown theory. The second, more interesting possibility, is that AdS/CFT itself allows entropy which scales with volume, as proposed by the non-holographic version of AdS/CFT [31].
Conclusion
Inspired by insights from condensed matter physics, in this paper we have considered a possibility that general relativity is only an effective large-distance description of a fluid phase of some unknown fundamental degrees of freedom. From that point of view, it seems natural to assume that the same fundamental degrees of freedom can also exist in the crystal phase, and that entropy of the crystal phase can scale with volume rather than area. Under these assumptions, we have proposed a natural resolution of the black-hole information paradox, according to which all the information needed for unitarity of Hawking radiation is absorbed by the crystal core inside the black hole. Even if there is no such a core in the initial black-hole state, the creation of a core (via a phase transition of crystallization) is forced by incoming information of ingoing Hawking quanta. This provides a simple mechanism for crystal growth, which eventually leads to the result that the final state of black-hole evaporation is a remnant in the crystal phase, with mass and size much above the Planck scale. Even though such a large remnant requires effects of quantum gravity at macroscopic distances, the idea that these effects are a consequence of crystallization makes such effects rather natural; the violation of classical general relativity in a macroscopic gravitational crystal should not be more "unexpected" than the violation of the classical Navier-Stokes fluid equation in a macroscopic piece of ice. Hence the gravitational crystal inside the black hole, as a candidate for a massive remnant which stores information and resolves the black-hole information paradox, seems to be an attractive idea worthwhile of further investigation. | 2015-09-08T10:04:13.000Z | 2015-05-15T00:00:00.000 | {
"year": 2015,
"sha1": "f798887c04f5f010c4a323eded441ceff148cf56",
"oa_license": null,
"oa_url": "http://arxiv.org/pdf/1505.04088",
"oa_status": "GREEN",
"pdf_src": "Arxiv",
"pdf_hash": "f798887c04f5f010c4a323eded441ceff148cf56",
"s2fieldsofstudy": [
"Physics"
],
"extfieldsofstudy": [
"Physics"
]
} |
231990332 | pes2o/s2orc | v3-fos-license | Palliative radiotherapy after oesophageal cancer stenting (ROCS): a multicentre, open-label, phase 3 randomised controlled trial
Summary Background Patients with advanced oesophageal cancer have a median survival of 3–6 months, and most require intervention for dysphagia. Self-expanding metal stent (SEMS) insertion is the most typical form of palliation in these patients, but dysphagia deterioration and re-intervention are common. This study examined the efficacy of adjuvant external beam radiotherapy (EBRT) compared with usual care alone in preventing dysphagia deterioration and reducing service use after SEMS insertion. Methods This was a multicentre, open-label, phase 3 randomised controlled trial based at cancer centres and acute care hospitals in England, Scotland, and Wales. Patients (aged ≥16 years) with incurable oesophageal carcinoma receiving stent insertion for primary management of dysphagia were randomly assigned (1:1) to receive usual care alone or EBRT (20 Gy in five fractions or 30 Gy in ten fractions) plus usual care after stent insertion. Usual care was implemented according to need as identified by the local multidisciplinary team (MDT). Randomisation was via the method of minimisation stratified by treating centre, stage at diagnosis (I–III vs IV), histology (squamous or non-squamous), and MDT intent to give chemotherapy (yes vs no). The primary outcome was difference in proportions of participants with dysphagia deterioration (>11 point decrease on patient-reported European Organisation for Research and Treatment of Cancer quality of life questionnaire-oesophagogastric module [QLQ-OG25], or a dysphagia-related event consistent with such a deterioration) or death by 12 weeks in a modified intention-to-treat (ITT) population, which excluded patients who did not have a stent inserted and those without a baseline QLQ-OG25 assessment. Secondary outcomes included survival, quality of life (QoL), morbidities (including time to first bleeding event or hospital admission for bleeding event and first dysphagia-related stent complications or re-intervention), and cost-effectiveness. Safety analysis was undertaken in the modified ITT population. The study is registered with the International Standard Randomised Controlled Trial registry, ISRCTN12376468, and ClinicalTrials.gov, NCT01915693, and is completed. Findings 220 patients were randomly assigned between Dec 16, 2013, and Aug 24, 2018, from 23 UK centres. The modified ITT population (n=199) comprised 102 patients in the usual care group and 97 patients in the EBRT group. Radiotherapy did not reduce dysphagia deterioration, which was reported in 36 (49%) of 74 patients receiving usual care versus 34 (45%) of 75 receiving EBRT (adjusted odds ratio 0·82 [95% CI 0·40–1·68], p=0·59) in those with complete data for the primary endpoint. No significant difference was observed in overall survival: median overall survival was 19·7 weeks (95% CI 14·4–27·7) with usual care and 18·9 weeks (14·7–25·6) with EBRT (adjusted hazard ratio 1·06 [95% CI 0·78–1·45], p=0·70; n=199). Median time to first bleeding event or hospital admission for a bleeding event was 49·0 weeks (95% CI 33·3–not reached) with usual care versus 65·9 weeks (52·7–not reached) with EBRT (adjusted subhazard ratio 0·52 [95% CI 0·28–0·97], p=0·038; n=199). No time versus treatment interaction was observed for prespecified QoL outcomes. We found no evidence of differences between trial group in time to first stent complication or re-intervention event. The most common (grade 3–4) adverse event was fatigue, reported in 19 (19%) of 102 patients receiving usual care alone and 22 (23%) of 97 receiving EBRT. On cost-utility analysis, EBRT was more expensive and less efficacious than usual care. Interpretation Patients with advanced oesophageal cancer having SEMS insertion for the primary management of their dysphagia did not gain additional benefit from concurrent palliative radiotherapy and it should not be routinely offered. For a minority of patients clinically considered to be at high risk of tumour bleeding, concurrent palliative radiotherapy might reduce bleeding risk and the need for associated interventions. Funding National Institute for Health Research Health Technology Assessment Programme.
Introduction
In the UK, data from 2015-18 showed more than 9000 new cases of oesophageal cancer and approximately 8000 deaths due to oesophageal cancer each year.Worldwide in 2012, more than 450 000 new cases were diagnosed, with greater than 80% of new cases and deaths occurring in low-income and middle-income nations. 1 Most patients present with incurable disease.For patients with advanced disease, median survival is 3-6 months, 2 with the majority requiring intervention for dysphagia. 3,4lthough the ideal intervention for dysphagia palliation has not been defined, options include chemical and thermal ablation, self-expanding metal stents (SEMS), and radiotherapy and chemotherapy alone or in combination.6][7] Median time to recurrent dysphagia in stent comparator 11,12 and brachy therapy studies 8 is 11-12 weeks and it has profound effects on independence, social function, and QoL.Hospitalisation and re-intervention account for most stent-related costs, 5 imposing a substantial burden on both NHS resources and a vulner able population with a median overall survival of 3-6 months.In line with Cochrane review research recommendations, 6,7 com bination of SEMS with other treatments might reduce costs and patient burden by reducing adverse events and re-interventions at a time when patients are approaching the last weeks of life.
The Radiotherapy after Oesophageal Cancer Stenting (ROCS) study was developed in response to a UK National Institute for Health Research (NIHR) call for research proposals into aspects of palliation, and aimed to address uncertainties in the evidence base for interventions
Research in context
Evidence before this study Before study commencement we searched Medline and the Cochrane Database of Systematic Reviews for prospective trials of dysphagia palliation interventions published from Jan 1, 1995, to Jan 1, 2011, using the terms "dysphagia", "stent", "oesophageal cancer OR carcinoma", "radiotherapy" and "brachytherapy".We included randomised controlled trials (RCTs), both masked and unmasked, reported in English.Studies of interest were those with patients with incurable primary squamous cell carcinoma or adenocarcinoma of the oesophagus or oesophagogastric junction.We identified a UK National Institute for Health Research Health Technology Assessment and a Cochrane systematic review confirming the efficacy of self-expanding metal stents (SEMS) in relieving dysphagia.These studies showed that efficacy of SEMS alone was limited by early problems with pain, decline in general aspects of quality of life (QoL), and later complications such as haemorrhage, tumour overgrowth, and recurrent dysphagia within 12 weeks.They called for evidence of interventions in combination with a stent to improve outcomes, and evidence on QoL and cost-effectiveness.These findings were confirmed by a further systematic review during the present study period.We identified two studies of brachytherapy versus stenting that showed longer dysphagia-free survival and more stable health-related QoL with brachytherapy.The literature search confirmed low access to brachytherapy for this patient group across the UK.A search of ClinicalTrials.govdid not reveal any current prospective studies of the more widely available palliative external beam radiotherapy (EBRT) in combination with stenting for improving dysphagia outcomes.
Added value of this study
We believe that the present study is the first sufficiently powered RCT to test the efficacy of palliative EBRT in combination with stenting for improving dysphagia outcomes in advanced oesophageal cancer, and the first RCT to show an effect of EBRT on tumour bleeding risk.It also provides prospective data on the effect of radiotherapy on re-interventions and service use.The addition of radiotherapy after oesophageal stent insertion does not reduce dysphagia deterioration and adds significantly to the cost of treatment.Reduction of upper gastrointestinal bleeding was observed in patients randomly assigned to receive radiotherapy after stent insertion, which warrants further investigation.This trial provides detailed findings on the poor outcomes in this patient group, which are rarely the focus of multicentre prospective research.
Implications of all the available evidence
Patients with advanced oesophageal cancer and dysphagia have limited treatment options and outcomes are often poor.Re-interventions are common following stent insertion.This study shows that when stent insertion is required to palliate dysphagia, most patients will not additionally benefit from-and should not be routinely offered-locoregional radiotherapy.Radiotherapy might be reserved for a minority of patients at high risk of tumour bleeding, but this strategy should be balanced against the burdens of treatment.Our study highlights the considerable unmet needs of this patient group.We hope our findings will challenge upper gastrointestinal services to establish better evidence in this field on other combination therapies and how multidisciplinary support can help patients and families negotiate symptom burden and improve QoL.
Study design and participants
The ROCS study was designed as a multicentre, parallelarm, open-label, phase 3 randomised controlled trial with an internal pilot phase examining recruitment.It was done in cancer centres and acute care hospitals across Scotland, England, and Wales.A description of the original trial protocol has previously been published. 13A summary of changes to the original ROCS protocol are provided in the appendix (p 2).The final trial protocol is available online.Ethics approval for the study was given by the Wales Research Ethics Committee 2 in October, 2012 (reference 12/WA/0230).
Participants were patients with incurable oesophageal carcinoma (histologically confirmed excluding small cell carcinoma; or clinical or radiological evidence of invasive tumour and at least high-grade dysplasia of a non-small cell type on histology), already referred for a stent as primary palliation of dysphagia by members of the local upper gastrointestinal multidisciplinary team (MDT), age 16 years or older, having an expected survival of at least 12 weeks, deemed clinically able by the MDT to tolerate radiotherapy, having completed a baseline (post-consent) European Organisation for Research and Treatment of Cancer (EORTC) quality of life questionnaire-oesophagogastric module (QLQ-OG25), 14 unsuitable for radical treatment (oesophagectomy or radical chemoradiotherapy) because of patient choice or medical reasons, and with the ability to provide written informed consent.We excluded patients planned to receive endoscopic treatment of the tumour, other than dilatation, in the peri-stent period (except for required emergency interventions), those with a tumour length of greater than 12 cm (or tumour growth within 2 cm of the upper oesophageal sphincter), those with presence of a tracheo-oesophageal fistula, or pacemaker in the proposed radiotherapy field, those who had previous radiotherapy to the area of the proposed radiotherapy field, and those who were pregnant.Patients in whom brachytherapy or EBRT was already planned after stent insertion were not included as we were concerned that further addition of trial radiotherapy might increase the risk of toxicity.The number of such patients was small (brachytherapy accounts for <2% of dysphagia inter ventions in the UK and EBRT is rarely used for immediate dysphagia relief in the UK in the participants of interest 10 ).Patients were approached and randomly assigned either before or after stent insertion to allow pragmatic accommodation within the clinical pathway.
Randomisation and masking
Patients were randomly assigned 1:1 to receive EBRT (plus usual care) or usual care alone after stenting by the method of minimisation with a random element (80:20) via a central telephone randomisation system developed by, and based at, the Centre for Trials Research at Cardiff University (Cardiff, UK).Minimisation was stratified to ensure balanced treatment allocation by a number of potential confounding factors: treating centre, stage at diagnosis (I-III vs IV), histology (squamous or nonsquamous), and MDT intent to give chemotherapy (yes or no).Participants were enrolled and assigned their trial group by the local principal investigator or research practitioner.The research practitioner was responsible for subsequent follow-up data collection.The study was necessarily open label and neither the patients nor the treating clinicians were blinded to treatment allocation.However, classi fication of some events was blinded, detailed herein.
Procedures
SEMS insertion was done in the EBRT group and the usual care group as per standard procedures at each centre.Stent type and length were determined by the treating clinician.When possible, the stent length was chosen to ensure that at least 2 cm of normal oesophagus was covered by the stent above and below the tumour.
Usual care was implemented in both groups according to local MDT practice to include, as needed, post-stent dietetic advice, referral for palliative and supportive care interventions (eg, blood transfusion and supportive oncology), and community-based health-care and socialcare follow-up.
In the EBRT group, the study protocol mandated that radiotherapy begin within 4 weeks of stent insertion and preferably 2 weeks.Treatment dose was prespecified at each centre, preferably 20 Gy in five fractions per day over 1 week or, at the treating clinician's discretion, 30 Gy in ten fractions per day over 2 weeks.Treatment was administered according to each centre's normal radiotherapy procedures without corrections for inhomogeneity in dose calculation.In the event of severe radiotherapy side-effects or treatment machine unavailability, gaps in treatment of up to 7 calendar days were allowed.If the patient missed more than 7 consecutive calendar days during radiotherapy treatment, then they were withdrawn from the trial and further treatment given at the clinician's discretion.Radiotherapy quality assurance was monitored by the NIHR Radiotherapy Trial Quality Assurance Group.
Follow-up at home was planned 1 week after stent insertion (before any radiotherapy; forming the baseline measurement for the primary outcome), and every 4 weeks thereafter for up to 1 year (finishing when the last patient See Online for appendix For the final protocol see https://www.cardiff.ac.uk/ centre-for-trials-research/ research/studies-and-trials/view/ rocs enrolled had been followed up for 12 weeks).At each timepoint, participants completed the EORTC QLQ-OG25, EORTC QLQ core 30 (QLQ-C30; version 3.0), 15,16 and EuroQol 5D questionnaire 3 level (EQ-5D-3L). 17,18Data were also collected on WHO performance status, stent complications, toxicities (as per the National Cancer Institute Common Terminology Criteria for Adverse Events [CTCAE] version 4.03), other treatments, and health-care and social-care resource use.If a home visit was not possible, or if patients preferred, data were collected by phone.Additionally, phone calls every 4 weeks were introduced midway between home visits to maximise capture of the primary outcome data only.Serious adverse events were collected in real time via a designated contact service from time of informed consent until 60 days after stent insertion.
During the pilot phase, an embedded qualitative study based on longitudinal interviews explored the feasibility of trial recruitment in a subset of patients during their first 8 weeks after stent insertion, relating to patient experience of the trial and recruitment process, the effect of trial interventions on daily life, and the experiences of living with advanced oesophageal cancer.30 longitudinal interviews in 15 patients (nine in the EBRT group and six in the usual care group) were done in their first 8 weeks of trial involvement.Interviews were analysed with the Braun and Clarke framework for thematic analysis. 19Full results of this qualitative study will be reported separately.
Outcomes
The primary outcome was deterioration in a dysphagia event within 12 weeks after stent insertion, defined as: two consecutive deteriorations of more than 11 points from baseline 20 in patient-reported dysphagia score on the EORTC QLQ-OG25, with the first being taken as the event timepoint (consecutive deteriorations were specified because patients undergoing radiotherapy might tem porarily show worsening of dysphagia secondary to radiation-induced oesophagitis); one deterioration and no more data possible (patient withdrew completely or died before next visit); one deterioration and patient missing on the next visit, with patient withdrawal or death within 4 weeks of the missed visit; additional dysphagia-related primary events consistent with the relevant change in dysphagia score (additional stent insertion, hospital admission for dysphagia, overgrowth or undergrowth of the stent, grade ≥3 dysphagia [CTCAE v4.03], or additional radiotherapy to the oesophagus or stent region; assessed and confirmed by the chief investigators [DA and AB] as tumour related and reviewed by an independent gastro enterologist, all masked to treatment group, as a dysphagia-related event); or death from any cause.In patients showing deterioration at one assessment but with missing data at a subsequent assessment, deterioration was timed at the previous assessment.
Secondary outcomes were overall survival, QoL (including WHO performance status), morbidity (upper gastrointestinal-related bleeding event or hospital admission for a bleeding event, first dysphagiarelated stent complication, or re-intervention), dysphagia deterioration-free survival (DDFS), post-stent chemotherapy or additional radiotherapy, patient experience (to be reported in detail elsewhere), and cost effectiveness.Overall survival was calculated from the date of stent insertion to the date of death from any cause.QoL was measured with the EORTC QLQ-C30, EORTC QLQ-OG25, and EQ-5D-3L.The prespecified main patient-reported outcome items were the global health score from the QLQ-C30 and four scales from the EORTC QLQ-OG25: odynophagia, pain or discomfort, eating restrictions, and eating in front of others.Upper gastrointestinal bleeding events were confirmed by the chief investigators who were masked to the study group and reviewed by a masked independent gastro enter ologist.These events could include blood transfusion, haematemesis, other descriptions of upper gastro intestinal haemorrhage or bleeds, or interventions related to bleeding (such as argon plasma coagulation or additional radiotherapy).If there was no clinical evidence that anaemia was due to a bleed then it was not considered.Data on treatment with antiplatelet drugs, anticoagulants, and nonsteroidal anti-inflammatory drugs other than aspirin were collected at each clinical assessment visit.Stent complications were defined as restenting, repeat endoscopy, overgrowth or undergrowth of the stent, stent blockage, stent fracture, stent slippage, and stent-related pain (grade ≥2 on the CTCAE v4.03).Dysphagia-related stent events (blindly assessed) only influenced the primary outcome if an event had not already been identified in the patient-reported OG25 assessments.Re-interventions (dyphagia-related or not related) were defined as additional stent insertion, stent removal, endoscopic intervention (including laser therapy and alcohol injection), and other palliative radiotherapy (including brachytherapy and additional EBRT for dysphagia).For monitoring of safety, toxicity was measured throughout follow-up with the CTCAE v4.03.
Statistical analysis
Originally the sample size calculation was based on a time-to-event analysis for the primary endpoint requiring 496 participants.However, during the recruitment phase of the trial, in view of lower than expected eligible patient numbers and substantial missing data after 12 weeks reflecting patient deterioration, the independent data monitoring committee (IDMC) recommended a revised sample size calculation, based on comparison of proportions with an event by week 12 rather than a timeto-event analysis.No early analysis was done that might have influenced this recommen dation.To detect a reduction in the proportion of patients with deterioration from 40% to 20% required 164 patients (82 patients per group; 80% power at a two-sided α level of 5%), with a total of 220 to be recruited to allow for 25% loss to followup.This difference in proportions was larger than that for the original sample size sought but was in line with the difference sought in other studies of stent or non-stent interventions for malignant dysphagia. 8,21,22The changes were approved by the independent trial steering committee and ratified by the funder following further independent review.
All statistical analyses followed a predefined statistical analysis plan agreed with the IDMC.Our modified intention-to-treat (ITT) population was defined as all patients who had a stent inserted (otherwise no benefit from radiotherapy was expected) and returned a baseline EORTC QLQ-OG25 (an eligibility criteria).The perprotocol (PP) population was defined as the subgroup of the modified ITT population that was alive and had not withdrawn from trial treatment at 4 weeks after stent insertion, and, in the EBRT arm, had received at least one fraction of radiotherapy to compare those who could have received radiotherapy in the usual care arm with those who did in the EBRT arm.
Analysis of the primary binary endpoint of deterioration in dysphagia symptoms by 12 weeks was primarily done in the modified ITT population with complete case data.Complete cases were defined as having complete data for the dysphagia subscale of the QLQ-OG25 questionnaire at baseline, week 4, week 8, and week 12, or having died with complete data before week 12.In the absence of a documented dysphagiarelated event, missing dysphagia scores between two non-event dysphagia scores were assumed to be no event.Multivariable logistic regression was used to adjust for randomisation stratification factors and obtain odds ratios (ORs) and 95% CIs for any treatment effect in the primary analysis and all sensitivity analyses.We did three sensitivity analyses: using the same complete case population but treating death by 12 weeks without earlier deterioration as no deterioration; imputing missing data using a best-case scenario that assumed no deterioration in a missing QLQ-OG25 form immediately before an QLQ-OG25 form that showed deterioration (or a dysphagia-related primary event), or that assumed no deterioration in a missing QLQ-OG25 form immediately before death; and imputing missing data using a worst-case scenario that assumed deterioration in a missing QLQ-OG25 form immediately before an QLQ-OG25 form that showed deterioration (or a dysphagia-related primary event), or that assumed deterioration in a missing QLQ-OG25 form immediately before death.As further sensitivity analyses, all analyses were repeated in the PP population.
As a secondary endpoint per IDMC guidance, DDFS was calculated in the ITT population from the date of stent insertion to the date of deterioration in dysphagia (as per the primary outcome definition).We analysed overall survival and DDFS using Kaplan-Meier plots and Cox regression (with the usual care group as the reference for the treatment effect measured by hazard ratios [HRs] and 95% CIs), with patients without events being censored at the time of last contact and adjusted for randomisation stratification factors with treating centre included as a shared frailty.We tested the model fit and assumptions using Cox-Snell residuals and Schoenfeld's global test.
QoL data and WHO performance status scores, prespecified in the statistical analysis plan, were analysed by the same method: the distributions of the variables were tested for normality with the Shapiro-Wilk test, kernel density, normal probability, and normal quantile plots, and either mean scores (or median scores if the was evidence of non-normality) Time to first morbidity event was compared between trial groups by competing risks regression (used to calculate subhazard ratios and 95% CIs), with death as a competing risk, adjusted for randomisation stratification factors, and with cumulative incidence functions plotted by trial group and median time to event calculated with the stci command in STATA.Treatment-emergent grade 3-4 toxicity was reported in the modified ITT population.Risk ratios were calculated in a post-hoc analysis to compare rates of toxicities and post-stent chemotherapy or additional radiotherapy between treatment arms.
We assessed cost-effectiveness from a UK NHS and Personal Social Services perspective using a combined decision tree and Markov model (appendix p 33), comparing the costs of usual care alone to EBRT over a time horizon of 12 weeks after stent insertion (extended to 12 months in a sensitivity analysis).We considered the intervention implementation costs of EBRT and the cost of subsequent health and social care resource use (primary, secondary, hospice, and social care including any cancer treatment and medications received).Health and social care resource use was captured with a Client Service Receipt Inventory.As the analysis was less than 12 months, discounting was not applied.The model calculated the incremental cost per quality-adjusted lifeyear (QALY) gained based on dysphagia data, health-care resource use, and health utilities derived from the EQ-5D-3L responses collected at the follow-up points every 4 weeks and the UK EQ-5D value set. 23Costs were applied by calculating number of patients in each stage of the model and multiplying by mean cost for that stage.In the base-case analysis, patient-level data was used to populate the model with the first 12 weeks of data.The health-care costs in primary, secondary and social care for both EBRT and control groups post-randomisation were summated and mean absolute cost difference per patient (including 95% CIs and p values) were calculated with SPSS (version 26).Independent sample t-tests were used for comparisons with a 5% significance level.For the 12-month time horizon, the model structure remained the same.However, costs, utilities, and transition probabilities between the health states of stable or worsening dysphagia were updated to include the data from 13-52 weeks.
Where data were missing, mean patient-level interpolation was used.Sensitivity analyses were undertaken to test the robustness of the cost utility analysis considering the uncertainty in input parameters such as costs and outcomes and in different scenarios.In a deterministic, univariate sensitivity analysis, we changed intervention and health-care costs and outcomes individually within plausible ranges (using 10%, 20%, and 30% of the mean value).Scenario analyses were used to test different assumptions and recalculate the incremental cost per QALY gained (eg, based on different populations: complete cases and all available cases).The time horizon was also extended to 12 months to explore longer term effects of the intervention.In a probabilistic sensitivity analysis, we used non-parametric bootstrapping to address joint parameter uncertainty and assess the effect on the incremental cost during 1000 simulations, undertaken with random sampling from distributions of costs and outcomes (with replacement).
Detailed statistical methods are presented in the protocol. 13In all analyses, a p value of less than 0•05 was considered to indicate significance.All statistical analyses were done with STATA 16.This study is registered as an International Standard Randomised Controlled Trial, ISRCTN12376468, and with ClinicalTrials.gov,NCT01915693.
Role of the funding source
The funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report.
Results
1252 patients were screened, 546 were found to be eligible, and 220 were enrolled and randomly assigned at 23 hospital sites (appendix p 21) between Dec 16, 2013, and Aug 24, 2018.Of these patients, 112 were allocated to receive usual care alone and 108 to receive adjuvant EBRT (plus usual care) after stent insertion (figure 1).The modified ITT population comprised 102 patients in the usual care group and 97 in the EBRT group.Unless otherwise stated, all results described herein are for the modified ITT population.Baseline characteristics are shown in table 1 and details of stents used are given in the appendix (p 3).The randomisation stratification factors of tumour histology, stage at diagnosis, and intended chemotherapy after stent insertion (and treating centre [data not shown]) were well balanced between trial groups as were other baseline characteristics.
Compliance with radiotherapy is shown in the appendix (p 4) for patients assigned to EBRT.Of the 97 patients in the EBRT group, 15 (15%) died or were withdrawn from the trial before radiotherapy treatment was given (figure 1).One (1%) participant chose a reduction in the planned radiotherapy from 30 Gy in ten fractions to 15 Gy in five fractions.All remaining participants received the protocol radiotherapy except one (1%) participant who received 8 Gy in one dose as this was the local practice (classified by the trial management group [AB, DA, JB, GG, AN, LN, MS, JF, ST, AM, JS, TC, and DF] as a minor deviation as an appropriate palliative dose; and patient retained in the PP population).Of the modified ITT population, 74 patients in the usual care group versus 75 patients in the EBRT group had complete data at week 12 for the primary endpoint (table 2).The baseline characteristics of patients with and without complete data were well balanced (appendix p 6).Our primary analysis of complete case data showed no significant difference in the proportion of patients with a dysphagia deterioration event up to 12 For EORTC QLQ-C30 and EORTC QLQ-OG25 responses, the proportion of missing data at each timepoint (including withdrawals, deaths, and loss to followup) was balanced across trial groups and increased from 0-1% at the first post-stent assessment to 23-25% at week 4, 36-41% at week 8, 49-59% at week 12, and 59-65% at week 16 (appendix pp 10-11).The appendix (pp 12-17) shows the results of the linear mixed models for the EORTC QLQ-C30 and QLQ-OG25 questionnaire scales and WHO performance status.The prespecified main subscales or items of interest were global health (QLQ-C30; appendix p 22), odynophagia (appendix p 22), pain or discomfort (appendix p 23), eating restrictions (appendix p 23), and eating in front of others (appendix p 24; all QLQ-OG25).We found no evidence of a time versus treatment interaction in any of these items (p values >0•05; appendix pp 12-13).
Evidence of a time versus treatment interaction was observed for dysphagia (p=0•013; appendix p 12).The median dysphagia scores were higher (ie, worse swallowing problems and more severe dysphagia) at week 4 in the EBRT group than in the usual care group, but the same by week 8 (appendix p 24).This short-term deterioration was expected, hence the requirement for two successive deteriorations in the definition of the primary endpoint.We also found evidence of a time versus treatment interaction for pain measured by the EORTC QLQ-C30 (p=0•005; appendix p 14). Median pain score was higher at weeks 8, 12, and 16 in the EBRT group (appendix p 25).Furthermore, we observed a time versus treatment interaction for constipation (p=0•009; appendix p 14).The mean score was higher at weeks 8, 12, and 16 in the EBRT group, mirroring the pain scales and possibly related to greater use of analgesia.We also observed significant time-treatment interactions for trouble swallowing saliva, anxiety, body image, and cognitive functioning.We found no evidence of time-treatment treatment interactions for any other QoL scales or WHO performance status (appendix pp 12-17, 25-26).Median time to first upper gastrointestinal-related bleeding or hospital admission for a bleeding event was longer with EBRT than with usual care (65•9 weeks [52•7-not reached] vs 49•0 weeks [95% CI 33•3-not reached]; adjusted subhazard ratio 0•52 [95% CI 0•28-0•97], p=0•038; appendix p 26).Table 3 shows the analysis of upper gastrointestinal-related bleeding events by trial group, for the total study period and up to week 16.Blood transfusion was the most common event, followed by haematemesis and upper gastrointestinal haemorrhage or bleed.
We found no evidence of differences between trial groups in time to first dysphagia-related stent complication or re-intervention event (appendix p 27), including additional stent insertion (appendix p 28), repeat endo scopy (appendix p 29), overgrowth or undergrowth of the stent (appendix p 30), or stent-related pain (appendix p 31).
Table 4 shows the proportions of patients with treatment-emergent grade 3-4 toxicities by trial group between week 0 and week 16.The most common grade 3-4 adverse event was fatigue, reported in 19 (19%) of 102 patients receiving usual care alone and 22 (23%) of 97 receiving EBRT.RRs were calculated (not shown) for each toxicity in a post-hoc analysis, and all of the 95% CIs included 1, but we noted that the point estimates of the RRs were greater than 2 (indicating increased adverse events in the EBRT group) for vomiting and abdominal pain.However, the absolute numbers of patients with abdominal pain were low in both groups (table 3).Three deaths (two in the usual care group, one due to a fall [eventually categorised as hospital-acquired pneumonia after admission for the fall] and one due to myocardial infarction; and one in the EBRT group due to multifocal ischaemic stroke) were reported via the real-time serious adverse event system (appendix p 9), none of which were considered to be related to treatment.
In our qualitative study, participants in the EBRT group described whether potential benefits of radiotherapy were worthwhile against additional burdens of hospital attendance and of pain and fatigue.Participants from both study groups revealed ongoing challenges with eating despite stent placement.They suggested that the technical intervention of stenting does not address physical and social eating concerns and symptoms.Information around diet, symptom control, and general medical management throughout the course of the disease was often scarce.Full results of this quantitative analysis will be reported elsewhere.
The costs of health-care and social-care service use in the 12 weeks after stent insertion for the usual care and EBRT groups are reported in table 5 and the appendix (p 32).No significant differences were found in care settings, medication costs, or total care costs (excluding EBRT cost) per patient between groups.Mean EBRT intervention cost was £1297•34 (SD 296•38) per patient, which explained the significant difference in total mean cost (including EBRT cost), of £1831•77 (95% CI 387•43-3276•11, p=0•013) favouring usual care.The basecase cost-utility analysis (12-week time horizon) showed that EBRT was more costly and had lower efficacy than stent only, and thus was not a cost-effective treatment at a willingness-to-pay threshold of £20 000 per QALY gained (appendix p 19).This conclusion did not change when the time horizon was extended to 12 months and when parameters for cost and outcomes were varied in sensitivity analyses (appendix p 20).
Discussion
Our results show that palliative radiotherapy does not reduce the proportion of patients with recurrent dysphagia at 12 weeks, nor does it improve overall survival or reduce service use.The addition of radiotherapy to stent insertion in advanced oesophageal cancer is therefore not routinely recommended.Patients in the radiotherapy group did have significantly fewer bleeding events, an effect which persisted and increased with time.From a clinical perspective, these findings suggest that radio therapy could be considered for patients deemed at increased risk of bleeding rather than for all patients, to minimise treatment burden.
ROCS was designed to address gaps in the evidence on improving dysphagia outcomes and NHS and personal burdens in patients with advanced oesophageal cancer, [5][6][7] with use of a widely available intervention (EBRT) combined with stenting.The choice to assess palliative stent therapy was intended to reflect real-world UK practice, where SEMS is the predominant option for rapid dysphagia relief in advanced oesophageal cancer, accounting for more than 90% of endoscopic and radiological interventions. 10ROCS' target population therefore excluded patients with non-severe dysphagia being considered for inter ventions other than a stent, and those too unwell to have a stent.Although intra luminal brachytherapy might be considered an appropriate alternative to stenting particularly in patients with longer term survival prospects, [7][8][9] few services in the UK have access to brachytherapy, and it accounts for less than 2% of dysphagia interventions in the NHS. 10 Similarly, incorporation of endoluminal radiotherapy with a stent as a single modality has been shown to lower dysphagia scores with time compared with a stent alone, 24 but again, the equipment and expertise is not widely available and cost is likely to be substantially higher. 25 6 with the 30 Gy in ten fractions dose available if prespecified by the treating clinician.The timing of the primary endpoint at 12 weeks reflects the mean stent patency reported by Homs and colleagues 8 and other studies, 11,12 and the median overall survival of around 19 weeks in both groups confirms that our participant population accurately reflects the wider clinical population. 2 The finding that almost twice as many patients in the control group received their preplanned chemotherapy is noteworthy.This might reflect treatment burden in the radiotherapy group discouraging planned chemotherapy uptake, or could reflect participant or clinician assessment that an active treatment had already been given in the form of radiotherapy.Exploration of patient experiences in our qualitative study addressed previous gaps in understanding patient and family experiences of advanced oesophageal cancer 5 and QoL trade-offs. 27The results reflect the challenges of their lived experiences of eating restrictions, concerns over nutrition and diet, and a trial and error approach to combating these.These resonate with other findings 28 and emphasise the need for more structured, proactive multidisciplinary approaches in patients receiving palliative stent therapy. 27lthough to our knowledge, this study is the first of its kind and has been completed in a patient group with a poor prognosis, it does have some limitations.Originally the sample size calculation was based on a time-to-event analysis for the primary endpoint requiring 496 participants.Due to recruitment and data capture challenges associated with the poor prognosis of the study population, this approach was revised during the study, on advice of the IDMC, to a sample size calculation based on comparison of patient proportions with an event by week 12.Although the revised primary outcome might have affected the ability of the study to detect a true effect for EBRT, the consistency of the results across sensitivity analyses is robust, including the secondary analysis of DDFS.Whether death was treated as an event or not did not alter the primary outcome.In pragmatically allowing two radiotherapy schedules in the EBRT group, we did not aim to seek a difference between doses and the small number of patients receiving 30 Gy precludes any such analysis (64 [78%] of 97 received 20 Gy in five fractions).We also acknowledge the large number of QoL secondary endpoints assessed in this study and associated issues that arise with multiple comparisons, hence we urge caution in the overinterpretation of significant findings found amongst them.Finally, this study was inevitably open-label and some outcome measures could be prone to assessment bias, particularly the adverse events known to be sideeffects of radiation.We attempted to mitigate this bias with the use of a comprehensive panel of secondary outcome measures and use of masked assessors to review bleeding events.The baseline QoL assessments (week 1 post-stent) were done after randomisation (and therefore might have been suscep tible to bias), but we noted that most scores were balanced between groups at that timepoint.
In summary, the ROCS study confirms that patients with advanced oesophageal cancer requiring a stent to improve dysphagia will not benefit further from the addition of concurrent palliative radiotherapy.In addition, the study provides detailed data on the poor outcomes in these patients, which are rarely the focus of multicentre prospective research.For patients with a long-term prognosis and considered to have a markedly increased risk of tumour bleeding, concurrent palliative radiotherapy might reduce bleeding risk and the need for associated interventions.Future research should focus on alternative, readily accessible interventions that might be effectively combined with stenting or compared as a monotherapy, and effective, multidisciplinary, supportive interventions that address the multidimensional concerns around eating and nutritional intake.Contributors DA, AB, LN, and JB conceived the idea for this study.DA, AB, AN, TC, JS, JB, JF, and GG were grant applicants (AB and DA were co-chief investigators).CH, CP, AB, and DA wrote the first draft of the manuscript.All authors were involved in acquisition of data and critical revision of the manuscript for important intellectual content.LN provided senior management of the study, overseeing the trial manager (MS) at the CTR.CH provided scientific leadership of the study at the Centre for Trials Research at Cardiff University.CP did the statistical analysis with oversight by CH.AN was responsible for the qualitative research.BS was responsible for the health economic study design and analysis.MJ designed and created the decision analytical model, under supervision of DF.JB was responsible for the quality of life study design.MS was responsible for trial management and contributed at all stages of study reporting.ST and JF were responsible for the patient and public perspective throughout study development and implementation and contributed to study reporting.AM and JS were responsible for radiotherapy quality assurance.CP and CH accessed and verified the data.All authors had full access to all the data in the study and had final responsibility for the decision to submit for publication.
Declaration of interests
DA reports grants from the UK National Institute for Health Research (NIHR), during the conduct of the study; grants from Roche and Boehringer Ingelheim, outside the submitted work; and has given advice to Roche on the development of multidisciplinary team software but has received no financial recompense for this.AB reports grants from the NIHR and Marie Curie, during the conduct of the study.DF reports being an active member of the European Organisation for Research and Treatment of Cancer Quality of Life Group.GG reports grants from Janssen-Cilag, Novartis, Astex, Roche, Heartflow, Bristol Myers Squibb, and BioNtech, grants and personal fees from AstraZeneca, and personal fees from Celldex, outside the submitted work.JS reports personal fees and non-financial support from Janssen Oncology, non-financial support from Bayer, and personal fees from Astellas, Novartis, and AstraZeneca, outside the submitted work.ST reports personal fees from the NIHR Evaluation, Trials and Studies Coordinating Centre Prioritisation Panel of the NIHR Health Technology Assessment Programme, outside the submitted work.All other authors declare no competing interests.
Data sharing
The Centre for Trials Research at Cardiff University is a signatory of the AllTrials initiative and aims to make research data available when possible, subject to regulatory approvals, any terms and conditions from external providers, patient confidentiality, and all laws concerning the protection of personal information.De-identified participant data and study documents (including the study protocol, statistical analysis plan, participant information sheet, and informed consent form) are generally freely available with publication until Nov 30, 2028, but recipients are expected to acknowledge the original creators in any public use of the data or in publishing research results that are based wholly or in part on the data; anyone requesting access to data will be asked to agree to the terms of the CC BY 4.0 license.We might ask the requestor to cover reasonable cost for preparing and providing the data (for example physical storage and postage, when dataset size makes providing data by electronic means impractical).Please send requests for access to data related to the current study to the open data team (opendata@cardiff.ac.uk) for assessment, providing sufficient detail to uniquely identify the dataset sought and appropriate contact details for the requestor.
Figure 1 :
Figure 1: Inclusion of patients receiving a stent for primary palliation of dysphagia ITT=intention-to-treat. QLQ-OG25=quality of life questionnaire-oesophagogastric module.PP=per-protocol.*Research staff unavailable.†The number of weeks before radiotherapy should have been started.‡Two patients withdrawn due to patient choice; one patient uncontactable; and one patient relocated.§Seven patients withdrawn due to patient choice; and one at clinician request.
Figure 2 :
Figure 2: Dysphagia deterioration-free survival (A) and overall survival (B) by trial group up to 1 year DDFS=dysphagia deterioration-free survival.EBRT=external beam radiotherapy.
complete case data (death without earlier deterioration as non-event)
Data are n (%) unless otherwise specified.NA=not applicable.*Adjusted for randomisation stratification factors.†Missingat least one assessment at week 4, week 8, or week 12 and unable to impute a non-deterioration between two adjacent non-deteriorations.
Table 2 : Patient status and primary endpoint analyses at 12 weeks after stent insertion in the modified ITT population
The last date of follow-up was March 6, 2019.Total deaths from all causes are summarised in the appendix (p 9).
By contrast, palliative EBRT is widely available across the UK, and at Costs per patient are summarised as mean (SD).*Visits at home by community nurses or care assistants (eg, to help with personal care).†This result was no longer statistically significant after Bonferroni-Holm correction for multiple comparisons of total care costs.
Table 5 : Cost of care resources used per patient in the 12 weeks after randomisation in the modified ITT population lower
cost.The prespecified dose in this study, 20 Gy in five fractions, reflected the most widely used palliative dose used across the UK at the time of study design, | 2021-02-23T06:16:22.848Z | 2021-02-18T00:00:00.000 | {
"year": 2021,
"sha1": "48dccf38f3420c9a1b21bd3ae88e78384ddeed2d",
"oa_license": "CCBY",
"oa_url": "https://doi.org/10.1016/s2468-1253(21)00004-2",
"oa_status": "HYBRID",
"pdf_src": "ScienceParsePlus",
"pdf_hash": "521e6fa0b108d64ef17299faa1324927850a170b",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
244852824 | pes2o/s2orc | v3-fos-license | Effect on Communication Ability of Dental Hygienist in Problem-Solving Ability: Focusing on the Modulatory Effect of Cooperative Self-Efficacy
Purpose: This study investigated the effect on the communication ability of dental hygienists on problem-solving ability, focusing on the modulatory effect of cooperative self-efficacy. Materials and Methods: The study was conducted from November 4 to December 20, 2019, and it involved 213 dental hygienists working in the Busan, Ulsan, and Gyeongnam regions in Korea. The collected data were analyzed using SPSS 25.0 program. The general characteristics were analyzed using frequency analysis, the average score of each variable was analyzed using means and standard deviation, and the relevance of each variable was analyzed using correlation analysis. Hierarchical regression analysis was used to investigate the modulatory effect of cooperative self-efficacy. Results: Communication ability (p<.01), cooperative self-efficacy (p<.01), and problem-solving ability (p<.01) of dental hygienists showed significant correlations.As a result of cooperative self-efficacy analysis, in case of positive self-efficacy, the effect of communication ability on problem-solving ability was significant (p<.01). Cooperative self-efficacy was significant, as it partially mediated the relationship between communication ability and problem-solving ability (p<.001). Conclusion: This study provides basic data for improvingproblem-solving ability and effective communication with patients and subjects by improving the cooperative self-efficacy of dental hygienists.
Introduction
The Korean society has a demand for talented individuals who can respond flexibly and act quickly in solving problems with others using the expertise and information of members of the organization [1]. In a rapidly changing medical environment, the level of education of patients is increasing, and information acquisition through various media has become easier, and the levels of knowledge and awareness of medical care is high [2,3]. Dental hygienists in charge of clinical practice management and education are required to be cooperative and capable of solving various problems through effective communication within the organization. Communication ability refers to the ability to share messages about one's thoughts and feelings in relationships with others and to effectively accept the messages of others [4,5]. In the recent dental environment, dental hygienists should improve their communication ability to actively interact with patients during consultation and treatment [6]. Dental hygienists should resolve potential problems through verbal and nonverbal communication when various situations arise in dental clinical practice [7]. Therefore, communication ability is closely related to problem-solving ability [9]. Problem-solving ability refers to the ability to solve problems quickly, intellectually, and creatively through the recognition of the difference between the current state and the goal to be reached. Dental hygienists should have problem-solving abilities related to patient management and treatment to quickly solve problems related to patient safety [8]. Cooperative self-efficacy is a belief, confidence, or expectation in person's ability to successfully achieve goals. Persons with high self-efficacy have excellent self-control when performing certain actions, and they have good judgment, which facilitates good decisions makings [9]. A dental hygienist with high self-efficacy will demonstrate confidence and better attitudes and behaviors during the management of patients, which will facilitate good quality of care [10]. Cho and Kim et al. reported that self-efficacy and problem-solving ability have a positive effect on the performance of given tasks [9,11]. Therefore, enhancing the cooperative self-efficacy of dental hygienists is important, given the demands of dental care and services in the rapidly changing medical environment. Although some studies have analyzed communication ability and problemsolving ability of nurses and dental hygienists [9,11], only a few of them focused on the effect of communication ability of dental hygienists on problem-solving ability, focusing on the effect of cooperative self-efficacy. Therefore, this study investigates the relationships between communication ability, problem-solving ability, and cooperative self-efficacy for dental hygienists, and the effect of communication ability on problem-solving ability, focusing on the modulatory effect of cooperative self-efficacy. In addition, this study will provide basic data for appropriate methods and strategies for accurately assessing the level of self-efficacy and increasing cooperative self-efficacy, communication ability, and problem-solving ability. This will improve the quality of care and the satisfaction of patients and subjects in dental practices.
Research Design and Subjects
The subjects of this study were dental hygienists working in dental hospitals and clinics in the Busan, Ulsan and Gyeongnam regions in Korea, and a convenient sampling method was used. Questionnaires were given to dental hygienists who agreed to participate in the study and signed the consent form. The questionnaires were delivered in person at the dental hospitals and clinics, and they were filled between November 4 and December 2, 2019. Based on Cohen's power analysis, the required minimum sample size under the conditions of a significance level of 5% (two sides), power of 80%, and an effect size of 0.5, was calculated using G*power 3.1.3. The minimum sample size required was 200, and 240 questionnaires were distributed to compensate for dropout. Two hundred and twenty-five dental hygienists responded to the questionnaires. Finally, 213 questionnaires were analyzed, excluding 12 copies with biased or missing responses.
Research Method
A structured questionnaire was used to collect data for this study. We used a tool developed by Lee at al. [12], to evaluate the questionnaire responses and assess communication ability was composed of 49 questions. The responses to the questions were scored using a Likert 5-point scale, and a high score indicated high communication ability. The modified and supplemented questionnaire of Alavi [13], which consisted of 19 questions, was used to evaluate cooperative self-efficacy. Each question was framed to allow the response to be scored on a Likert 5-point scale, with higher scores indicating higher cooperative self-efficacy. Negative content in each category was analyzed by reverse coding for the consistency of the questionnaire.
Data Analysis
The SPSS statistical program (ver. 25.0, IMB, Armonk, NY, USA) was used to analyze the data of this study. The general characteristics of the study subjects were presented using frequency and percentage, whereas the scores of communication ability, problem-solving ability, and cooperative self-efficacy were presented as mean± standard deviation. The relationshipsbetween the factorswere analyzed for significance using correlation analysis, and the modulatory effect of cooperative self-efficacy was analyzed for significance using hierarchical regression analysis, including the interaction of the independent and the control variables. Mediation effect analysis and Sobel test were additionally conducted. Cronbach's α of the test tool of this study was 0.829 for communication ability, 0.936 for problem-solving ability, and 0.940 for cooperative self-efficacy. The reliability coefficient was 0.80 or higher, indicating a high degree of internal consistency of the questionnaire tool.
Results
The general characteristics of the study subjects are shown in Table 1. Most of the participants were women (95.3%). The age distribution was as follows: 25-29 years old were 53.1%; 30-39 years old, 38.5%, 40 years old, 8.4%. Regarding the academic background of the participants: 62.9% graduated from technical college, 31.0% graduated from university, 6.1% were graduates and postgraduates. Regarding the total clinical experience, 32.9% had less than 3 years, 36.6% had less than 3-7 years and 30.5% had 7 years or more of experience.
The scores for communication ability, cooperative self-efficacy, and problem-solving ability are shown in Table 2. The scores were 3.12 points for communication ability, 3.28 point for cooperative self-efficacy, and 3.40 points for problem-solving ability.
Correlation analysis results of dental hygienist communication ability, cooperative self-efficacy, and problem-solving ability are shown in Table 3. Communication ability showed a statistically significant positive (+) correlation with cooperative self-efficacy (r= 0.342) and problem-solving ability (r= 0.527). Cooperative self-efficacy showed a statistically significant positive (+) correlation with problem-solving ability (r= 0.624).
The moderating effect of cooperative self-efficacy on the effect of communication ability on problem-solving ability is shown in Table 4. Amoderated regression analysis was used to evaluate the moderating effect of cooperative self-efficacy on communication ability and problem-solving ability, and cooperative self-efficacy was found to be significant. On analyzing the main effects of communication ability, the magnitude of the effect of communication ability on problem-solving ability was 0.255. The analysis of the moderating effect of cooperative self-efficacy revealed that the magnitude of the interaction between communication ability and cooperative self-efficacy was .912. This means that the influence of communication ability on problem-solving ability increases with increasing cooperative self-efficacy.
The results of the hierarchical regression analysis of the mediating effect of cooperative self-efficacy on the effect of communication ability on problem-solving ability are shown in Table 5. During the first stage, communication ability significantly affected problem-solving ability (β= 0.527, p< 0.001). During the second OPEN ACCESS https://scidoc.org/IJDOS.php stage, communication ability was significantly associated with the mediating variable, cooperative self-efficacy (β= 0.120, p< 0.001.). During the third stage, where the independent and mediating variables were analyzed together, cooperative self-efficacy was significant as a mediating variable (β= 0.503,p< 0.001.), and the β coefficient of the independent variable was also significant (β= 0.355, p< 0.001). This means that cooperative self-efficacy partially mediates the relationship between communication ability and problem-solving ability. The analysis of the mediating effect showed that the mediating effect of cooperative self-efficacy was significant.
Discussion
This study investigated the degrees of communication ability, cooperative self-efficacy, and problem-solving ability of dental Classification M ± SD (n=213) Communication ability 3.12 ± 0.29 Cooperative self-efficacy 3.28 ± 0.53 Problem-solving ability 3.40 ± 0.40 Table 3. Correlation between communication ability, cooperative self-efficacy, and problems solving ability.
Communication ability
Cooperative
self-efficacy
Problem-solving ability Communication ability -Cooperative self-efficacy .342** -Problem-solving ability .527** .624** -**p< 0.01 The data were analyzed by person correlation coefficient. hygienists, and the effect of communication ability on problemsolving ability focusing on the modulatory effect of cooperative self-efficacy. The mean scores of the dental hygienists were 3.40 points for problem-solving ability, 3.28 points for cooperative self-efficacy, and 3.12 points for communication ability( Table 2).
The communication ability score of dental hygienists in the study by Lee et al. [12] was 3.63 points, 3.44 points in the study of nurses from Jeong and Shin [14], and 3.34 points in the study by Lim and Kim [15]. The communication ability score of Korean dental hygienists in this study, was lower than that of the previous report. Communication ability is essential for the interactive role of dental hygienists in solving and practicing real problems. The average score of the creative problem-solving ability of the dental hygienists was 3.40 points, whereas the score of cooperative self-efficacy was 3.27 points in the study by Kim and Sim [6], which was similar to the results of this study. Problem-solving ability affects the speed and efficiency of decision making when solving a given problem, whereas cooperative self-efficacy affects job satisfaction and organizational commitment [6]. Therefore, communication ability, cooperative self-efficacy, and problemsolving ability are necessary for dental hygienists, and continuous improvement and enhancement of these capabilities through education and seminars are required. Based on the analysis, cooperative self-efficacy had the highest statistically significant positive correlation with problem-solving ability (p< 0.01)( Table 4),and communication ability showed the highest statistically significant positive (+) correlation with problem-solving ability (p< 0.01), followed by cooperative self-efficacy (p< 0.01)( Table 4).Park and Ko [16] reported that communication ability was significantly valuable during problem-solving, and their results were similar to the results of this study. In addition, the study of Yun et al. [17] reported that the communication ability of nurses has a great influence on patient safety competencies.
This is thought to be because communication ability enhances problem-solving ability in the medical environment. To improves the job performance of dental hygienists, it is necessary to improve the problem-solving ability in the evidence-based practice of dental hygiene [4]. It is believed that various problems can be solved in the field of dental clinical practice, and effective communication is key to increasing cooperative self-efficacy. Communication ability affects the moderating effect of cooperative self-efficacy and problem-solving ability, focusing on the mediating effect.In other words, high cooperative self-efficacy has a great positive effect on problem-solving ability, whereas communication ability has a positive effect on problems solving ability. Based on the high cooperative self-efficacy of dental hygienists, it is necessary to share information through participatory education and seminars to positively improve communication and problemsolving abilityby enhancing cooperative self-efficacy using various programs in the workplace. Furthermore, effective communication will facilitate a better quality of dental care to patients and subjects and induce positive effects on medical satisfaction. This study was involved dental hygienists working in dental hospitals and clinics in the Busan, Ulsan and Gyeongnam regions, and it is difficult to expand and interpret or generalize the results of dental hygienists in Korea and other countries. This study provides important basic data for appropriate measures and strategies for increasing cooperative self-efficacy, communication ability, and problem-solving ability.
Conclusion
This study confirmed the effect of communication ability on problem-solving ability, focusing on cooperative self-efficacy, based on the results of the analysis of the relationships between communication ability, cooperative self-efficacy, and problemsolving ability of dental hygienists. The communication ability of dental hygienists has a positive effect on problem-solving ability and cooperative self-efficacy. Therefore, communication ability can be an appropriate target for improving problem-solving ability. Furthermore, effective communication of dental hygienists will induce a positive effecton problem-solving during the consultation and treatment of patients and subjects. | 2021-12-04T16:05:45.268Z | 2021-11-11T00:00:00.000 | {
"year": 2021,
"sha1": "4f94e4fdc83d164e3dd774eeb2bcaa0dc058084b",
"oa_license": null,
"oa_url": "https://doi.org/10.19070/2377-8075-210001002",
"oa_status": "GOLD",
"pdf_src": "Adhoc",
"pdf_hash": "e92d0e1e620fb71259cd582197e970d1a0343659",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": []
} |
157371793 | pes2o/s2orc | v3-fos-license | Achieving Research University: Indonesian Case
Today many universities have the vision to become a research university, including in Indonesia. It is based on the desire to play a role in advancing science for the benefit of humanity as well as to enhance the university reputation at the international level. However, in the case of Indonesia, it can only be done by several universities, given the large number of universities with very different capabilities. In addition, another problem is human resources, infrastructure, and research funding. Various targets indicator used to determine its success include the number of publications, patents and industrial products. There is an urgent need to improve all factors that can accelerate the increase in research in Indonesia universities, and has been started by the policy of the current government.
Introduction
Rapid development of higher education is currently happening around the world [1] [2]. Changes in the form of a university education in the past become a research university is not inevitable. The public demands of university is not only producing graduates but also expected to provide new alternative findings of the research. The fast development of lifestyle demands a new technology capable of meeting those needs [3]. University and industry partnership is very important and play important role in this matter [4] [5].
These conditions make the universities transform itself into a research university that is able to produce the technology needed by the community. Furthermore, changes continued after a research university and then apply knowledge and become an entrepreneur university [6]. This development is increasingly affecting many universities in the world, including in Indonesia.
This brief paper discusses the state of research universities in Indonesia. Given that there are still many obstacles faced in order to become a research university, it will also discuss what policies are needed to support the development of research universities in Indonesia.
Development of university in Indonesia
Indonesia has several types of higher education institutions, those are university, institute, polytechnic, college, academy, and community college. University is higher education institutions that produce graduates in various fields up to the doctoral level. Institute is also institution of higher education which only on specific field up to doctoral level, while college also has specialized field, but only up to bachelor level. Polytechnic is a higher educational institution that produces diplomas in various fields up to Diploma level 4, while the academy only to the Diploma level 3 on specific field. For a community college that has recently been developed in Indonesia, only up to Diploma level 1 or 2. All universities in Indonesia initially were intended to produce graduates with degree or diploma. Its main target is to produce people who are experts in a particular field as much as possible with a good quality according to the needs of society and the market. Based on the development of higher education in the world, several major universities and institute in Indonesia began to transform themselves into research universities [7]. However, considering the real situation of universities in Indonesia, only a few are able to run the process as well as a research university.
The number of universities in Indonesia is currently more than 4,000 institutions. This number is very large, but has not been able to accommodate all existing high school graduates annually. Most universities in Indonesia are private, only about 300 are state university. The ability and capacity of these universities is very different both in financial, management, infrastructure, the number of students and professors, as well as other things are concerned.
Constraints research university
To become a research university, it takes the ability and adequate institutional capacity. Research universities require various aspects that are not easily met by universities in Indonesia. Being a successful research institutions not only need financial capabilities, but also support human resources and infrastructure facilities [8]. On top of that, there is a need to have good management in higher education management for its sustainability [9], with university strong and visionary leadership [10].
Initially, Indonesian government giving financial support to public universities only, because the existing rules do not allow the state to help private institutions. Therefore, private universities should seek their own funding from both student fees and from other businesses. Currently, the government has begun providing assistance to private universities, with methods permitted by the rules and regulations in Indonesia. For example, the government has been providing a wide range of research funding schemes may be accessed by anyone, from both public and private universities. However, given that government funds are limited, the support given to college do not meet all of the needs. Therefore, to become a good research university, generally financial constraints being faced by these universities.
In terms of infrastructure, the general conditions of all the universities are very different. Even among public universities was also not the same. Several universities in the top 10 in Indonesia, is generally adequate and able they are expected to develop into a research university. However, laboratory and other supporting facilities at the university also need to be improved, as well as other universities. On the other hand, some major private university established by large companies, generally have adequate infrastructure. Some universities also have a laboratory facility, but still needs improvement. Among all universities in Indonesia today the availability of human resources with high research ability has not been evenly distributed. The number of researchers who holds a doctor or professor is still small. On the other hand, research universities require human resources with high research capabilities. They must also be able to follow the development of science through articles published in good scientific journals in order to develop cutting-edge research topic
Necessary policies
To support the increase of nation competitiveness, the universities have to work hard in order to carry out cutting edge research to deliver the products required by the community and the market. Research and education is the main foundation of the progress of a nation [11]. It is necessary for the central and local government policy, in addition to college policy to encourage the performance of research at each university.
In regard to financial policy, consistency is required of the central government in allocating research funds to universities. Allocation of funds should be given to universities on a competitive basis to produce good research results based on the criteria of a good proposal. The central government should have specific targets of those funded universities. Currently the central government, through the Ministry of Research Technology and Higher Education has set a target of 7 + 1 field of research, namely the areas of food, renewable energy, health, transportation, information technology and communication, defence and security, materials, and maritime. It takes courage for the central government to focus on the few universities that has been able to become a research university grow faster with the allocation of special funds with specific targets.
Government policies are needed to be able to encourage some universities selected through the allocation of special funds for the development of the infrastructure necessary for the development of research. In general, some universities have had the basic infrastructure that has can be used for research, but for certain things required special infrastructure.
In order to improve the ability of the existing human resources, it is also required special government policies to encourage this. Among others is by providing scholarships to students who have the ability to develop research and education levels. Related to this, the government has provided many scholarships, for example through the program LPDP. To enhance the experience of Indonesian academics to conduct research, government policy is needed to provide an adequate and flexible funding to develop research collaboration with other researchers from developed countries. Accelerate the recruitment of human resources who have the capacity and capability of high research also must be implemented immediately.
Conclusion
Improving the ability of universities in Indonesia in the research development, the central government policy is needed in particular to improve the competitiveness of the nation. The policy is primarily in financial assistance for the implementation of research, development of infrastructure and improvement of human resource capability. | 2019-05-19T13:04:58.652Z | 2017-02-01T00:00:00.000 | {
"year": 2017,
"sha1": "e2d59607a15e43aff551a4dfa8128af3aa441576",
"oa_license": null,
"oa_url": "https://doi.org/10.1088/1755-1315/55/1/012072",
"oa_status": "GOLD",
"pdf_src": "IOP",
"pdf_hash": "21719179b71566a73390260a3a078cabd1b09fe1",
"s2fieldsofstudy": [
"Education"
],
"extfieldsofstudy": [
"Business"
]
} |
55039662 | pes2o/s2orc | v3-fos-license | A Structural Approach on Students ’ Satisfaction Level with University Cafeteria
This study was carried out to identify the relationship between the food quality, price fairness, staff performance, and ambience of the university cafeteria with students’ satisfaction. The survey method was employed in testing the proposed hypotheses via a structured self-administered questionnaire. This survey was conducted in Universiti Sultan Zainal Abidin (UniSZA) and a total of 93 undergraduates were selected for questioning via convenience sampling method. The results were generated by using the Structural equation modeling (SEM) technique via AMOS 21.0 computer program with maximum likelihood estimation. Based on the SEM technique, food quality and price fairness are the two most important dimensions that influence the students satisfaction on café service quality. Next, the students give less priority to staff performance and ambience. The results were differing from the previous study. The university cafeteria should take serious measurement in improving the food quality and price for long term sustainability.
Introduction
The service quality and customer satisfaction are very much related to any kind of service business.These are the two most important determinants that will provide the competitive advantages in long run.The food service sector is highly fragmented with large number of small medium players such as restaurants, hotels and resorts and institutional food service (Stanton et al., 2000); whereas, the increase number of students in public and private universities in Malaysia (Dollah et al., 2012) forced university management to improve the food service at the mentioned higher learning institutions.In Malaysia, these higher learning institutions provide cafeteria which offers variety of menus and comfortable surroundings where the students experience a sense of home and where they can engage in leisurely conversation and interactive activities with their peers (Norhati & Nurhafisah, 2013).
Most of the food service sectors, eager to improve the service quality to satisfy their existing customers and devote additional resources in capturing new ones (Noel-Levitz, 2010).The university cafeterias also need to improve and upgrade their services to maintain and chasing after new customers.Therefore, this study attempts to identify the relationship between the food quality, price fairness, staff performance and ambience of the university cafeteria especially among Universiti Sultan Zainal Abidin (UniSZA) students.
Service Quality and Customer Satisfaction
Hundreds of service quality scholars have been discussing on customer satisfactions for decades.The customer satisfaction is the heart of business marketing (Andaleeb & Conway, 2006).Customer satisfaction is influenced not only by service quality perceptions but also by personal and situational factors and price (Aldridge & Rowley, 1998;Patterson & Johnson, 1993;Robinson, 1999;Rowley, 1997;Zeithaml, Bitner, & Gremier, 2008).The customer judgment of product or service itself will provide a pleasurable level of consumption related to fulfillment (Oliver, 1997).The judgment of students in university cafeteria is important for the success of a cafeteria in any higher learning institution.The students' satisfaction in institutional food service sectors depends on food quality, food variety and price fairness (Xi & Shuai, 2009).Whereas, Chang, Norazah, and Tam (2014) mentioned that the university students' satisfaction on university cafeteria depends on food quality, price fairness, staff performance and ambience.
Food Quality
Food quality is very much related to customer satisfaction in measuring students' satisfaction on cafeteria service level.Food quality is the quality characteristics of food that is acceptable to customer (McWilliams, 2000).The appearance of foods encompasses several basic sensory attributes such as colour, opacity, gloss, visual texture and perceived flavour (Imran, 1999).Therefore, the degree of satisfaction with university cafeteria depends mostly on the quality of meals, diversity of food, food hygiene and environment (W.G. Kim & H. B. Kim, 2004).Hence, the following hypothesis is posited.
H1: Food quality has a positive influence on the level of student satisfaction with the university cafeteria.
Staff Performance
The staff performances at each food outlets are very important in increasing the level of customer satisfaction.The employees especially in service industry play vital role in the success of food service outlets.Numerous scholars explored various factors triggering a customer willingness to spread a positive electronic word-of-mouth and came to the conclusion that service quality is in fact the strongest predictor, while affecting post purchase behaviour through two altruistic mediators: expressing positive feelings and also an urge to help a restaurant company as a thank you for a great dining-out experience.The interaction between the cafeteria staff and students, such as friendly gestures (e.g.smiles and greeting and high levels of responsiveness, cleanliness and quick service) is important as it influences student satisfaction with the service quality (Barlett & Han, 2007).The following hypothesis is hence developed: H2: Staff performance has a positive influence on the level of student satisfaction with the university cafeteria.
Price Fairness
Several studies have been carried out by many scholars on price fairness or price and value.Price fairness means the judgment of whether an outcome and the process to reach an outcome are reasonable or acceptable (Bolton & Shankar, 2003).The price to be paid for a service determines the level of quality to be demanded (Soriano, 2003).He also stressed that the price (value) of the meal and service are equally important when compared to other service dimensions.The recent studies by Ng (2005) and Xi and Shuai (2009) did consider price and value in assessing students' service quality in dining hall services.Martin-Consuegra, D., Molina, A. and Esteban, A. (2007) found that perceived price fairness positively influences customer satisfaction.The subsequent hypothesis is thus proposed: H3: Price fairness has a positive influence on the level of student satisfaction with the university cafeteria.Bitner (1992) coined a new term servicescape with physical constituents of service environment.The servicescape consists of optimum temperature, noise, furnishings and layout combine together to influence the customer satisfaction and repeat patronage level.In addition, the design of the cafeteria environment influences the consumer's food choices and eating behaviors which call the personal food environments to promote wellness, combat obesity and complement interventions at higher levels (Raman & Chinniah, 2011;Wansink, Painter, & Ittersum, 2001).Furthermore, the physical setting influences customers' perceptions of service quality (Hensley & Sulek, 2007;Norhati & Hafisah, 2013).Hence, it is posited that: H4: Ambience has a positive influence on the level of student satisfaction with the university cafeteria.
Ambience
The suggested research framework is demonstrated in Figure 1.
Methodology
A structured self-administered questionnaire survey was successfully conducted among 93 undergraduates from Universiti Sultan Zainal Abidin (UniSZA), Malaysia via convenience sampling method.Initially, 100 questionnaires were distributed in May 2013 of which 7 were returned incomplete.The questionnaire was prepared in English and then translated into Malay by the author and reviewed by two bilingual linguists.Section A of the questionnaire contained demographic questions relating to gender, age and race, while Section B required respondents to indicate levels of agreement on factors such as food quality (9 items), staff performance (5 items), price fairness (3 items), ambience (6 items) and students' satisfaction (5 items).These variables were adapted from Kim and Kim (2004), Martin-Consuegra et al. (2007), Story, Kaphingst, Robinson-O'Brien, & Glanz (2008); Raman and Chinniah (2011) and were measured using a 5-point Likert scale from 1 (strongly disagree) to 5 (strongly agree).Data was analyzed via Structural Equation Modeling (SEM) technique using AMOS 21.0 computer program with maximum likelihood estimation as it has the ability to ensure the consistency of the model with the data and estimate effects among constructs instantaneously.
Structural Equation Modelling
A two-step SEM approach was employed guided by maximum likelihood method of estimation: a measurement model and a structural model.
Measurement Model
Confirmatory factor analysis was performed to test the validity of each construct in the model, including item loading, construct reliability, and average variance extracted (AVE).Results are presented in Table 2.Each of the standardized loadings items is greater than 0.50 on their expected factor after removal of items that are not meeting the recommended value.Thus the construct validity is acceptable.The results also infer that the CR scores for all constructs exceeded the acceptable level of 0.70, indicating a relatively high level of constructs reliability.All AVE values are greater than the cut-off value 0.50.All in all, the current data have a good convergent validity.The shared variances between factors were compared with the squared root of AVE for each construct to examine discriminant validity.The results in Table 3 shows the shared variances of the construct with other constructs were lower than the squared root of AVE of the individual factors, confirming discriminant validity.Hence, each construct was statistically different from the others.
Structural Model
The structural equation model was evaluated by examining fit indices and variance-explained estimates.The measure of the fit of the tested model was done through examining several goodness-of-fit indices (Table 4).The results indicated that the χ 2 of the model was 204.839 with 138 degrees of freedom (χ 2 /df = 1.484) and a root mean square error of approximation (RMSEA) of 0.063.The fit indices value for CFI, GFI, NFI, CFI, and IFI were above 0.90 and RMSEA below 0.08, indicating a satisfactory fit.The results in Figure 2 demonstrate that all independent variables accounted for 89% of the total variance in students' satisfaction (R 2 =0.89).As a consequence, the results are a sign of adequate model fit between the proposed research model and the empirical data.The standardized parameter estimated for the structural model regarding the relationship of the independent variables (food quality, price fairness, staff and ambience) and dependent variable (students' satisfaction with university cafeteria) is summarized in Table 5. Figure 2 shows the standardized theoretical paths linking these variables.Hypothesis 1 posited that food quality significantly influences students' satisfaction with university cafeteria.This hypothesis was supported with a β 1 =0.535, p<0.05.Similar significant findings appears for H3, i.e. price fairness (β 2 =0.370; p=0.006).Next, staff and ambiance both had insignificant influence on students satisfaction with university cafeteria (β 3 =-0.121,p=0.288; β 4 =0.262,p=0.183), suggesting H2 and H4 are not supported.
Discussion
From the initial hypotheses of this study, the above result indicates that two hypotheses are accepted (H1 and H3) and the other two are rejected (H2 and H4).With regard to hypothesis 1, the result defined that students' satisfaction with university cafeteria is highly affected by food quality, implying H1 is significant.This finding is positively associated with those of earlier researchers (Hwang et al., 2003;W. G. Kim & H. B. Kim, 2004;Qin & Prybutok, 2009;Raman & Chinniah, 2011).Food quality aspects such as careful handling, cleanliness while serving to customers, quality offered and menu variation are considered important by university students dining at the cafeteria.
Contrary, with regard to hypothesis 2, the result explains that staff performance has no influence on the level of student satisfaction with the university cafeteria.The unsatisfactory services offered by the university cafeteria operators, staff unfriendliness, unresponsiveness (lack of smiles and greetings) slow service and unreasonable product prices contribute to this finding.The results are incompatible with preceding research (e.g.Barlett & Han, 2007;Herrmann et al., 2007).
Furthermore, with regard to hypothesis 3, the result describes that students' satisfaction with university cafeteria is highly affected by price fairness, implying H3 is accepted.This is resemblance with prior findings (Herrmann et al., 2007;Martin-Consuegra et al., 2007;Oliver & Swan, 1989), implying price fairness becomes more important to students as they acquire more information from the menus to make price comparisons and judgments whether the payment is higher or lower in relation of their expectations of the services rendered.
With the regard to the final hypothesis 4, the result explains that ambience has no influence on the level of student satisfaction with the university cafeteria.This implies that the spatial arrangement of seating, quality of interior design, and suitability of background music do not influence students' satisfaction level with the service quality of the university cafeteria operators besides the food packaging, plate size and design, lighting and dining companions.This finding is similar with that earlier research (Hensley & Sulek, 2007;Namkung & Jang, 2009;Norhati & Hafisah, 2013;Story et al., 2008).
Conclusion
The study revealed that students' satisfaction with the university cafeteria is influenced more by food quality and price fairness than staff and ambience.It is therefore important for university cafeteria operators to keep on improving the quality of food serve to the customers to maximize their satisfaction level.Furthermore, they should also offer an attractive menu at a reasonable price on the food variations and serve their customer in a proper ambience that can excite their interest in dining at the cafeteria.
This research has important implication for research and practice in the planning and designing new marketing strategies when launching new university cafeteria, depending on their target market.One of the major implications of this research is that food quality and price fairness are the important factors that the cafeteria providers should considers regardless of their target market; whether students, academicians or the public who visit and dine at the cafeteria.Remarkably, staff performance and ambience factors do not affect satisfaction of the students' with the university cafeteria.Thus, the university cafeteria should invest in these issues through staff training and development, using fresh foods in the menu choices, providing an attractive and cozy ambience and choosing furniture suitably designed at university cafeteria.
Although this study has provided useful information which may help the university cafeteria providers to improve their service quality and students' satisfaction, there are a few limitations to this research.Therefore, it is recommended that future studies expand the number of respondents and include more respondents from other age groups such as university academicians and administrative staff in order to provide more representative results and improve sample for generalizability.Since the study was carried out in Malaysia the results may not be fully generalizable for other countries, as beliefs and perceptions may differ e.g. between developed versus developing countries, and Islamic versus non-Islamic countries.This study can be further expanded using demographics, market environment, and the ideology and culture of the students as mediating and moderating variables.
Figure
Figure 1.Research model
Figure 2 .
Figure 2. The result of the research model Table1details the descriptive statistics of the demographic profiles of respondents.Female respondents represented 79.6 per cent of the sample, while male respondents represented the remaining 20.4 per cent, with 17.2 per cent aged 20, 20.4 per cent aged 21, 22.6 per cent aged 22, 15.1 per cent aged 23, 12.9 per cent aged 24, 6.5 per cent aged 25, 4.3 per cent aged 26 and 1.1 per cent aged 27.In terms of race, the majority were Malay (86.0 per cent), followed by Chinese (8.6 per cent), Indian (2.2 per cent), others (2.2 per cent) and Kadazandusun (1.1 per cent.The science and non-science streams represented about 74.2 per cent and 25.8 per cent, respectively.
Table 5 .
Relationships with students' satisfaction on university cafeteria | 2018-12-07T02:40:10.911Z | 2014-08-22T00:00:00.000 | {
"year": 2014,
"sha1": "f0288041e69fe6e0e4824bc5bbcb53d1def5833a",
"oa_license": "CCBY",
"oa_url": "https://ccsenet.org/journal/index.php/ass/article/download/39703/21986",
"oa_status": "GOLD",
"pdf_src": "Anansi",
"pdf_hash": "f0288041e69fe6e0e4824bc5bbcb53d1def5833a",
"s2fieldsofstudy": [
"Business"
],
"extfieldsofstudy": [
"Psychology"
]
} |
84726000 | pes2o/s2orc | v3-fos-license | Evaluation of some infiltration models and hydraulic parameters
The evaluation of infiltration characteristics and some parameters of infiltration models such as sorptivity and final steady infiltration rate in soils are important in agriculture. The aim of this study was to evaluate some of the most common models used to estimate final soil infiltration rate. The equality of final infiltration rate with saturated hydraulic conductivity (Ks) was also tested. Moreover, values of the estimated sorptivity from the Philip’s model were compared to estimates by selected pedotransfer functions (PTFs). The infiltration experiments used the doublering method on soils with two different land uses in the Taleghan watershed of Tehran province, Iran, from September to October, 2007. The infiltration models of Kostiakov-Lewis, Philip two-term and Horton were f itted to observed infiltration data. Some parameters of the models and the coefficient of determination goodness of f it were estimated using MATLAB software. The results showed that, based on comparing measured and model-estimated infiltration rate using root mean squared error (RMSE), Horton’s model gave the best prediction of f inal infiltration rate in the experimental area. Laboratory measured Ks values gave signif icant differences and higher values than estimated final infiltration rates from the selected models. The estimated f inal infiltration rate was not equal to laboratory measured Ks values in the study area. Moreover, the estimated sorptivity factor by Philip’s model was signif icantly different to those estimated by selected PTFs. It is suggested that the applicability of PTFs is limited to specif ic, similar conditions. Additional key words: final infiltration rate, saturated hydraulic conductivity, sorptivity.
Introduction
Infiltration is the term applied to the process of water entry into soil (Hillel, 1980). Evaluation of soil infiltration characteristics and determination of the final steady infiltration rate are required for increased irrigation water use efficiency, the design of irrigation systems, and decreased water and soil losses which are important factors in agriculture. The final steady infiltration rate of a soil is the minimum rate that water enters into the soil (Mbagwu, 1993). Since measuring the final infiltration rate is time consuming, several models have been proposed to determine this parameter. Empirical models such as those of Kostiakov (1932) and Horton (1940), and physical models such as that of Philip (1957) are the most commonly models to estimate final infiltration rate. Singh (1992) stated that the various models can estimate different values of the final soil infiltration rate which seems incorrect as the parameter is soil-dependent. The ability of the models to estimate infiltration rate has been examined by many researchers. Gifford (1976) observed that among the Horton, Kostiakov and Philip's models, the Horton model gave the best fit of infiltration data in mostly semi-arid range lands in Australia, but only under specific conditions. Roohian et al. (2005) suggested that Horton's model gives an acceptable estimate of final infiltration rate under given soil texture conditions. However, complicated factors that influence the final infiltration rate are a major reason for the different applicability of the models. Machiwal et al. (2006) observed infiltration was well described by the Philip's model in wasteland in Kharagpur, India. Navar and Synnott (2000) evaluated the infiltration rates of soils under four land uses in north-eastern Mexico. Among the infiltration models of Horton, Philip, a modified Kostiakov and Green-Ampt, the modified Kostiakov model gave the best fit. Hajabbasi (2006) evaluated the Kostiakov, Horton, and Philip's infiltration models under different tillage and rotations in a clay-loam in North-west Iran and reported that the Horton's model gave the best prediction of infiltration rate in that region.
Soil hydraulic parameters, such as saturated hydraulic conductivity (K s ) and sorptivity (S) are important factors that govern the soil infiltration rate. Several workers have evaluated the Philip's model parameters, transmissivity (A) and sorptivity factors in their experiments (Sutikto et al., 1993;Mohammadi and Refahi, 2006). The A factor in the Philip's model, may approximate to K s for long periods (Swartzendruber and Youngs, 1974;Rawls, 1992). The fit of measured infiltration data to the Philip's model leads to A= K s , as time approaches infinity (Philip, 1969). In Horton's model, the infiltration capacity approximates to a constant minimum (non-zero) rate of i c , as time approaches infinity (Horton, 1940;Hillel, 1998) that can be equal to K s . Mishra et al. (2003) showed that A varied from 50 to 75% of K s values. Mbagwu (1995) also showed that values of the final infiltration rate and K s were not equal over long periods.
Sorptivity is function of soil suction which is important for knowing soil hydraulic properties. This parameter is defined as a physical quantity that shows the capacity of a porous medium for capillary uptake and release of water into soil (Philip, 1957). This parameter can be estimated by defined PTFs as the relationship between soil hydraulic and other more available measured properties (Bouma, 1989) which can be used to estimate hydraulic parameters such as S factor.
The aim of this work was to evaluate the most common models used to estimate the final infiltration rate of soils, and compare values of the estimated final infiltration rate with laboratory measured K s on undisturbed soil columns, and finally, to compare values of the S factor estimated by the Philip's model and PTFs.
Characteristics of the research area
The study was conducted on part of the Taleghan watershed in Iran. Eight points were selected at two sites (36°08' 10.8" N, 50°40' 40.6" E, at 2,236 m asl and 36°08' 58.6" N, 50°43' 12.6" E, at 1,453 m asl, respectively) so the sites had the least difference in soil properties in the area studied. These sites were in rangelands with similar soil surface conditions and homogeneous soils. Soil depth and plant cover at points 1, 2, 3, and 4 were slightly greater than at points 5, 6, 7, and 8. The history of these rangelands showed that they were cropped for several years in the last century at points 3, 4, 7, and 8 using animal tillage. According to the Taleghan watershed study report (1993) the soils are calcareous and are classified as Typic Xerorthents. The depth of the Ap and C horizons were mostly 15 and 30 cm respectively and the boundary between horizons was smooth. The soil structure was massive in the soils studied. Soil texture in this area varied from clay-loam to silty clay loam. Minimum and maximum daily temperatures were approximately -5.6 and 17°C, respectively. Mean annual rainfall varied from 464 to 796 mm and the land had a slope of about 15%.
Soil sampling and analysis
Soil properties were determined by sampling at 0-15 and 15-30 cm soil depths. At each site, 48 undisturbed soil sample cores of 5 cm height and diameter and 98.12 cm 3 volume were taken (four replicates). They were used to measure the saturated hydraulic conductivity (K s ) using the constant head method (Klute and Dirksen, 1986), total porosity (by measuring saturated water content), and initial moisture content (gravimetrically). Darcy's equation was used to calculate the K s value. Additionally, 24 composite soil samples were collected for measurement of soil chemical properties. Organic matter and particle size distribution were measured by the modif ied Walkley-Black wet oxidation method (Allison and Moodie, 1965) and the Bouyoucos hydrometer procedure (Gee and Bauder, 1986). Dry bulk density was determined by the core method (Davies et al., 1973). Infiltration was measured using a doublering infiltrometer (Bouwer et al., 1999), with 25 cm inner diameter and 60 cm outer diameter rings (four replicates at each site) installed in the soil. Grass was carefully cut at the soil surface with shears in the inner and outer rings. A constant water head of approximately 15 cm was kept in both rings and amount of the water drop and the depth of added water were recorded for two hours. The EC of the water used was 0.49 dS m -1 , and sodium (Na) content was 20 mg L -1 . The data were fitted to three chosen infiltration models and the bestfit model selected. The soil Na (in a soil saturation extract) and calcium carbonate content (CaCO 3 ) were measured by flame emission and a volumetric method (Allison and Moodie, 1965), respectively. Soil pH and electrical conductivity (EC) were determined in soil saturation extract using the procedures of Page et al. (1982).
Estimation of infiltration model parameters
In this study, a few commonly used infiltration models, such as Kostiakov-Lewis, Philip two-term, and Horton's model were fitted to the infiltration data. The modelestimated final infiltration rate and sorptivity factor were determined using MATLAB software. The values of the measured final field infiltration rate with the measured K s in the laboratory and estimated S factor, by the Philip's model and PTFs were compared using a t-test (SAS/STAT, 1985).
Kostiakov-Lewis model
The modified Kostiakov model for long time periods is: where a and b are the equation ' s parameters (a > 0 and 0 < b < 1); i(t) and i c are the infiltration rate and presumed final infiltration rate in cm min -1 at time t in min, respectively.
Horton's model
The Horton ' s infiltration model (Horton, 1940) is: where i 0 is the initial infiltration rate in cm min -1 ; t is time; and k is the infiltration decay factor.
Philip two-parameter model
The Philip two-term model (Philip, 1957) is: where A is a transmissivity factor in cm min -1 as a function of soil properties and water content; S is sorptivity which is a function of soil matric potential (cm min -0.5 ).
Sf =
[5] where λ and h b are the Brooks-Corey pore-size distribution index and bubbling pressure head (cm). These parameters are defined based on clay and sand %, and porosity (cm 3 cm -3 ).
Selection of best-fit infiltration model
The results of the soil analyses are shown in Table 1. As expected, the soil analysis results showed that soil texture varied from clay-loam to silty clay loam, the soil pH from 7.7 to 8.0 and the EC from 0.23 to 0.58 dS m -1 at soil depths of 0-15 and 15-30 cm. Thus, there was no salinity or sodicity in soils of the area studied. Table 1 shows there were no notable differences among the soil properties studied. In this study, the goodness of fit of the selected models and their ability to estimate the final soil infiltration rate was evaluated using root mean squared error (RMSE). The R values were high (0.99) and equal for all points. Values of RMSE and i c showed that estimated infiltration rates by the Horton's model, were closer to the measured ones (
Final infiltration rate and K s
The results showed that the saturated hydraulic conductivity, measured in the laboratory, was positively related to the observed final infiltration rate (R = 0.612) (Table 2, Fig. 1). The results showed that measured values of the saturated hydraulic conductivity were significantly higher than the observed final infiltration rate in the field at all 8 points, (P > F = 0.0299, P = 0.05) ( Table 2). Measured K s was approximately 2-3 times the observed final infiltration rate at most sample points.
Estimated sorptivity factor
The sorptivity factor, in the Philip's model, changed from 0.315 to 2.984 among the different sample points (Table 3, Fig. 2). Statistical analysis at p < 0.01 (SAS/STAT, 1985) showed signif icant differences between the sorptivity factor from the Philip's model and the one estimated by PTFs. From Figure 2 it can be seen that sorptivity values estimated by PTFs had high deviations from the Philip's sorptivity factor. Further, the correlation between parameters of the Philip's model (S and A) was poor and the relationship between them was not significant (R = 0.127) (Fig. 3).
Discussion
In this study, the Horton's model gave the best fit of the results for all sample points. The different values of RMSE, for each model, at the different points is probably due to changed conditions at the points such as soil particle size distribution (Table 1), because of a high dependency of infiltration rate on soil texture (Rawls, 1992;Mohammadi and Refahi, 2006). The suitability of the Kostiakov and Philip's models for estimation of the infiltration rate can be site-specific (Mbagwu, 1993). In addition, complicated conditions and regional soil variation can signif icantly affect measured and estimated values. Thus, spatial variations in short distances and probable preferential flow at some points such as P4 can affect the data set obtained and consequently the model parameters. The result of this study agrees with the earlier result of Hajabbasi (2006), who found that the Horton's model is applicable in a clay loam soil of northwest Iran. Thus, the applicability of infiltration models should be tested under different conditions, because the various models can suppose different final infiltration rate values for a soil, which is not correct, while the final infiltration rate is generally a soil-dependent parameter (Singh, 1992;Mishra et al., 2003). In this study, the low predictive ability of the Kostiakov-Lewis model for estimating the final infiltration rate could be because it takes longer to obtain a steady infiltration rate because of a low initial soil moisture content (Navar and Synnott, 2000). It also seems that a longer time duration, more than 2 hours, may be required for application of the Philip's model (Philip, 1969;Mbagwu, 1993). Also, the Philip's two-term model sometimes can not accurately describe the measured field data (Gosh, 1983).
Significant difference between the Philip's sorptivity factor in P4 compared to other sample points may be due to the impossibility of accurately determining the value of the final infiltration rate using the Philip's model. Overall, spatial soil variability is important in infiltration modelling. Significant differences between the observed f inal inf iltration rate and K s can be attributed to the effect of trapped air under field conditions (Radcliffe and Rasmussen, 2000) or disturbed core samples producing preferential flow in the laboratory (Maheshwari, 1996;Mohammadi and Refahi, 2006). Thus, complicated conditions in the field, differences between laboratory and field conditions, and spatial soil variability can be good reasons for the observed results (i.e. Taleghan watershed). Differences between K s and the final infiltration rate (in Table 2) were similar to those of Mohammadi and Refahi (2006).
In this study, the variation in the sorptivity parameter showed no clear pattern in the study region. This agrees with the results of Machiwal et al. (2006). Soils in the field are more heterogeneous than laboratory soil cores and this is a possible reason for the differences. Thus, PTFs should be developed for hydrologic parameters of models on a larger scale than soil core. Also, PTFs derived from one database, in a particular region and condition, is not applicable under all conditions (Kern, 1995;Wösten et al., 2001). Thus, PTFs should be applied to similar conditions to the soils investigated. Moreover, soil pore size distribution is an important factor in Evaluation of some infiltration models and hydraulic parameters 215 spatial variation of Philip's sorptivity parameter (Sharma et al., 1980;Machiwal et al., 2006), whereas total porosity was considered in selected PTFs in this study. Other possible reasons for the difference could be due to variation in soil effective porosity, pore size distribution (Sharma et al., 1980), matric properties, and probable preferential flow. These can be possible reasons for differences between estimated values of sorptivity factor by Philip's equation and PTFs. However, the relationship between the sorptivity factor and effective porosity needs to be evaluated in more detail. Wide variation in model parameters can also be due to non-uniform initial soil moisture. The poor correlation between parameters of Philip's model (S and A) observed in this study, is confirmed by the findings of Talsma (1969).
In conclusion, in this work infiltration models were evaluated. The results showed that Horton's model can be used for estimating final soil infiltration rate. Only, one or some of the infiltration models are better and appropriate for a specific study site. This investigation showed that Horton's model was better than the Kostiakov-Lewis and Philip's model at most sample points in the Taleghan watershed in Iran. Thus, infiltration models should be tested for their ability to estimate the final infiltration rate of each location and it should be documented at each site. It was found that the observed final infiltration rate has no equal concept to the measured saturated hydraulic conductivity for all conditions and sites. Moreover, the determined sorptivity factor is a variable parameter under field conditions and estimated values of this factor by PTFs were significantly different compared with the values obtained for Philip's sorptivity factor. Thus, the use of various models and PTFs is restricted and this problem needs to be considered in these sorts of studies. It is suggested that in the future there is a need to evaluate the parameters of infiltration models and the relationship among them, accurately and for the calibration of PTFs and infiltration models under different conditions. Soil spatial variability has an appreciable effect on infiltration properties and hydraulic parameters that need to be evaluated in more detail. | 2019-03-21T13:05:08.810Z | 2010-03-01T00:00:00.000 | {
"year": 2010,
"sha1": "9bc9058ed8a06b9107994cf0ae6199b82ab4b8b8",
"oa_license": "CCBY",
"oa_url": "http://revistas.inia.es/index.php/sjar/article/download/1160/1157",
"oa_status": "GOLD",
"pdf_src": "MergedPDFExtraction",
"pdf_hash": "9bc9058ed8a06b9107994cf0ae6199b82ab4b8b8",
"s2fieldsofstudy": [
"Agricultural and Food Sciences",
"Environmental Science"
],
"extfieldsofstudy": [
"Mathematics"
]
} |
236157675 | pes2o/s2orc | v3-fos-license | It Requires More Than Intelligence to Solve Consequential World Problems
What are consequential world problems? As “grand societal challenges”, one might define them as problems that affect a large number of people, perhaps even the entire planet, including problems such as climate change, distributive justice, world peace, world nutrition, clean air and clean water, access to education, and many more. The “Sustainable Development Goals”, compiled by the United Nations, represent a collection of such global problems. From my point of view, these problems can be seen as complex. Such complex problems are characterized by the complexity, connectivity, dynamics, intransparency, and polytely of their underlying systems. These attributes require special competencies for dealing with the uncertainties of the given domains, e.g., critical thinking. My position is that it is not IQ, but complex problem-solving competencies for dealing with complex and dynamic situations, that is important for handling consequential global problems. These problems require system competencies, i.e., competencies that go beyond analytical intelligence, and comprise systems understanding as well as systems control. Complex problem solving is more than analytic intelligence.
Introduction
The editors of this Special Issue have asked me to respond to the statement: "How intelligence can be a solution to consequential world problems". What exactly are consequential world problems? Why should intelligence (intelligent behavior) be relevant to their solutions? What other skills would help in solving global problems? The answer I give here to these questions is based on my personal opinion, rather than on an evidence-based review of the literature.
Who defines "consequential world problems"? It is not easy to scale and prioritize the challenges to mankind, because there are many different perspectives. Lomborg (2007) tried to define "the world's biggest problems" in terms of costs and benefits. However, for me (as a German, remembering that the Nazis used the term "life unworthy of life" [in German: "lebensunwertes Leben"] to evaluate different qualities of life), this approach suffers from the issue of giving a monetary value to human life. I cannot follow the idea that you can compensate a human life for money, in line with the belief that "all lives are equal".
How, then, to define global problems? As "grand societal challenges", one might define them as problems that affect a large number of people, perhaps even the entire planet, including problems such as climate change, distributive justice, world peace, world nutrition, clean air and clean water, access to education, and many more.
The Challenges
The world around us presents many challenges to mankind-not only human-driven climate change but also the demand for food and water for a global population that may soon reach 8 billion people. Clean air and clean water are not widely available in large parts of the world, driving migration from failed states to more promising ones.
The United Nations have compiled 17 "Sustainable Development Goals" that represent a collection of such global problems. These goals are concerned with the survival of planet Earth and its inhabitants. They are the blueprint to achieve a better and more sustainable future for all. They address the global challenges we face, including poverty, inequality, climate change, environmental degradation, peace, and justice.
What is psychology's answer to these grand societal challenges? The research on problem solving is my preferred perspective on these challenges. Problem solving is a processual view on intelligence (see, e.g., Sternberg 1982). This research arena has produced various different approaches for dealing with nonroutine situations.
Proposed Solution: The "General Problem Solver"
The "General Problem Solver" (Newell et al. 1959) was, at its time (in the 1950s and 1960s), a promise to solve all problems with an algorithm. However, that promise did not hold-as it turned out, it was more a kind of "special" problem solver that could deal with well-defined problems but not with ill-defined ones. Real-world (ill-defined, "wicked") problems normally show a series of attributes that simple (well-defined) problems lack:
•
There is no clearly defined goal; • There is more than just one ("the one and only") solution; • Information needed for the solution has to be actively searched for; • There are dynamics in the problem situation (one cannot wait forever for a solution).
Today's version of this is called "artificial intelligence". This again promises to solve problems of many different types, but again, there is some disillusion concerning AI's abilities when considering "adversarial perturbations" (e.g., Metzen et al. 2017) and missing explanations from deep-learning "black-box" algorithms (e.g., Rudin 2019).
Dealing with Uncertainty: Solving Complex Problems
As a consequence of the limitations of problem-solving research focused on puzzles, towers, and other "toy problems", the use of computer-simulated microworlds was proposed (e.g., Brehmer and Dörner 1993). The use of microworlds within a controlled laboratory environment allowed the introduction of attributes that characterize uncertain situations and complex problems (see Dörner and Funke 2017;Funke 2019). According to Funke (2019, p. 167), a complex problem shows the following features: (1) complexity, in the sense that many variables are involved; (2) connectivity, reflecting the fact that relationships exist between variables; (3) intransparency, referring to missing or inaccessible information important for the problem-solving process; (4) dynamics, in the sense of the possibility of changes of a given situation over time; and (5) polytely (from the Greek word 'polytelos', meaning many goals), referring to there being many goals and objectives that are possible and could be pursued.
How are complex problems different to complicated problems? For complicated problems, solutions exist, and experts are needed (and found); for complex problems, there are no experts available, because these problems are novel and have no known solution (see Kurtz and Snowden 2003).
Contrary to expectations, the successful handling of complex problems does not correlate with intelligence (as measured by classical tests of intelligence), according to early studies (for a review, see Kluwe et al. 1991;Wenke et al. 2005). However, Wittmann and Süß (1999) pointed out that these negative results might be due to a Brunswikian asymmetric relationship between the dependent variables (single-act criteria on the side of complex problems and multiple-act criteria on the side of intelligence).
Research with "minimally complex problems" (as they were developed within PISA 2012; see Csapó and Funke 2017; Funke and Greiff 2017) showed that scores from these tasks explained about 10% of the variance in grade point average beyond test intelligence (Wüstenberg et al. 2012). However, concerning the validity of the "minimally complex systems", questions and doubts remain (see, e.g., Funke 2014;. To explore and control minimally complex problems, strategies such as VOTAT ("vary one thing at a time") may be helpful (Schoppek and Fischer 2017), but in real life, they cannot be recommended as a useful strategy. For example, you normally cannot vary the mood of your boss (or other potentially influential variables) to test the possible hypothesis that, while in a sad mood, she/he might give you more difficult tasks than when in a happy mood. In complex situations, simple variations of single factors are neither possible (because there are too many variables) nor conclusive (you cannot control the influence of other explanatory causes).
Systems competencies seem to be more important than intelligence (Funke et al. 2018). In the 21st century, skills such "critical thinking" (e.g., finding and evaluating relevant information; Halpern and Dunn 2021), understanding the complexities and dynamics of large systems, and dealing with uncertainty better represent the requirements of life than finding the correct answer to a number-series task (as tends to be requested in conventional tests of intelligence).
What are the cognitive proficiencies of systems thinking? Systems thinking involves a collection of cognitive and noncognitive features. It is a holistic approach, switching between a bird's eye view and a detailed view, remaining calm and persistent in the pursuit of goals, and showing empathy for the acting persons, instead of focusing on simple causeeffect chains, with a preference for systemic thinking about loops with positive or negative feedbacks.
Based on the proposed model, I illustrate more in detail how system competencies could be applied in practice. For this purpose, I considered an important contemporary real-world problem, namely climate change. Analytical intelligence (based on rational choice assumptions) might argue that, with respect to the anticipated rising sea level and myself living far above sea level, there is no threat to me as a person. However, with empathy (as part of a broader understanding of intelligence), one could imagine the threat to people living on an island in the Pacific Ocean exposed to a rising sea level.
Ethical Values, Character Formation, and Wisdom
Our current concept of intelligence is missing an ethical dimension. During the education process of individuals, we do not solely teach facts and knowledge; we also teach values. The education process, in general, is a process of character formation. One of the long-term results of character formation can be seen in the development of wisdom. Tuchman (1984, p. 21) defined wisdom as: "the exercise of judgment acting on experience, common sense and available information." Is wisdom the result of successful character formation?
In her recent review, Glück (2019 , Table 16.1, p. 310) presented twelve definitions of wisdom. Only one of them mentions "values" explicitly, namely the "balance theory of wisdom" from Sternberg (1998). According to that theory, wise people know that different people can have different values. This idea of "value relativism" in wise people is also one of the five criteria for wisdom within the Berlin wisdom paradigm (see, e.g., Baltes and Staudinger 2000). However, to know that there are different perspectives on dilemmata does not imply that one has clear moral values-it is a kind of metaknowledge, free of any special content. Values allow for "sinn" (German, "meaning") in life.
Similarly, Fischer (2015) argued for a context-free view of wisdom and saw it as "independent of one's values and context." On the other hand, Fischer collected 12 propositions that were commonly known to wise men from four different cultures (Socrates, Jesus, Confucius, and Buddha). These four wise individuals show parallels concerning certain wise content (e.g., Proposition 10: "Good people (and children) make good company"). Once again, there is no comment about the acquisition of these pieces of wisdom. Reading such "wise" propositions does not make us a wise person instantaneously. To become a wise person is a process that normally demands time and life experience. Wisdom should be one of the competencies of good leaders.
The University of Cambridge Institute for Sustainability Leadership (CISL), as reported by Visser et al. (2016), examined leadership theories and leadership development framed within the United Nations Sustainable Development Goals (SDGs), which were launched in September 2016. The CISL summarized the elements of a 'good' global leader into a model based on earlier research by Visser and Courtice (2011). This approach argued that a leader operating with a global perspective, and in a complex context, should have seven characteristics: capacity to be a systems thinker, proficiency in navigating complexity, open-minded, long-term thinker, interdisciplinary, inclusive, and globally conscious.
Conclusions
Complexity continues to increase, and creative solutions to complex problems are urgently needed (e.g., Mainzer 2009;Puccio 2017). To address societal needs, intelligence must be enriched by an ethical dimension. Sternberg (2019Sternberg ( , 2021a uses the term "adaptive" as a qualification for a new perspective of intelligent behavior. From my point of view, intelligence in all its forms is adaptive. The new perspective is on society and on the survival of mankind and therefore takes value into account. Here, "menschenbild" (our view on man) matters; the metaphors of man have shifted from "homo oeconomicus" (skeptically: Sen 1977) over "homo ignorans" (Hertwig and Engel 2016) to "homo curans" (man that cares; see, e.g., Tronto 2017). The deployment of "transformational intelligence" (Sternberg 2021b) also comes into play. To quote Fyodor Dostoyevsky (1821-1881): "It takes something more than intelligence to act intelligently" (from his book Crime and Punishment). This "something more" could be values and wisdom. "Values" adds to the noncognitive variables and "wisdom" to the cognitive variables. This could make an important difference. | 2021-07-22T06:18:01.057Z | 2021-07-19T00:00:00.000 | {
"year": 2021,
"sha1": "5d248d516b84f03c325de84ef4ed303650061282",
"oa_license": "CCBY",
"oa_url": "https://www.mdpi.com/2079-3200/9/3/38/pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "06a48b136902db03abb3f0dadd25e966d653e053",
"s2fieldsofstudy": [
"Psychology"
],
"extfieldsofstudy": [
"Medicine"
]
} |
251929747 | pes2o/s2orc | v3-fos-license | O4-8 Measuring capabilities for physical activity-related health outcomes: A Systematic review
Abstract Background Health promotion projects commonly measure health outcomes and behavior to provide proof of effectiveness. An alternative concept that focuses on the real opportunities a person can choose from to influence their health is Amartya Sen's capability approach. Numerous tools have been developed to measure capability change in general, but it remains unclear which ones can be applied specifically to physical activity (PA).We therefore conducted a systematic review to identify appropriate tools to measure capabilities for physical activity and health and provide information on their quality. Methods The review included a total of 6,850 articles published between 2000 and June 2019 that were identified via searches on PubMed, EbscoHost, and ProQuest. Screenings of titles/abstracts and full texts were conducted independently by two researchers using Endnote X9 and Microsoft Excel. Identified tools are currently being analyzed regarding their indicators, evaluation methods, quality, and the extent to which they address capabilities for physical activity. Results The screening resulted in a total of 49 articles included in the analysis. Preliminary results show a diverse use of methods for measuring capabilities for healthy lifestyles. Preliminary results show that three categories of instruments can be identified: (a) Five studies employed secondary data analysis of specific datasets to extrapolate capabilities for healthy living; (b) five articles dealt with measuring capabilities using qualitative approaches (interviews, video recordings); (c) 39 articles reported on a total of 10 different questionnaires to measure capabilities. We identified only one instrument (employing both a questionnaire and qualitative measures) that explicitly measured capabilities for PA, albeit only for a specific target group. Conclusions The identified articles show that capabilities for healthy lifestyles are mostly measured by questionnaire. Available tools are mostly target group- and setting-specific. Currently, there is a dearth of tools that explicitly cover capabilities for PA, especially across settings or target groups. Therefore, more research is needed to work towards the development of universally applicable tools to measure PA capability.
Background
Using Amartya Sen's capability approach (CA) to conceptualize physical activity (PA) promotion projects has been suggested as a promising alternative to conventional theories, as it focuses on the real opportunities people have to engage in PA rather than on PA behavior alone. The Capital4Health research consortium used the CA to conceptualize and implement PA projects in four different settings across the life-course. The aim of the study was to evaluate the implementation of the CA in these projects and to develop recommendations for the use of the concept in future PA promotion projects. Methods Based on an overarching analytical framework, we investigated the utilization of the CA in the individual projects using document analysis, workshops, and group interviews with project teams. Results were used to develop a set of draft principles and recommendations for effectively employing the CA in future projects and as a bridging framework for larger consortia. A participatory process combining elements of action research and Delphi surveys is currently conducted with all members of the Capital4Health consortium to arrive at final set of agreed-upon principles.
Results
Preliminary results show that the use of the CA varied substantially between projects and settings, but that a number of common conclusions can be drawn. A future framework for using capabilities for PA promotion may focus on three areas: (a) Project conceptualization should address both the target group and relevant multipliers. It should also consider both individual PA competences and structural factors, (b) Evaluation should cover both capabilities for PA as well as actual changes in PA levels to assess health impact on multiple levels, (c) The CA may be very useful to research consortia for developing shared goals and evaluation frameworks. This, however, requires a shared knowledge base and agreement about central theoretical concepts.
Conclusions
The CA constitutes a potentially useful theoretical basis for both individual PA promotion projects and multi-setting research consortia. However, the application of the concept is Abstract citation ID: ckac094.032 O4-8 Measuring capabilities for physical activityrelated health outcomes: A Systematic review Maike Till 1 , Susanne Ferschl 1 , Karim Abu-Omar 1 , Peter Gelius 1 1 Department of Sportscience and Sport, Friedrich-Alexander-Universitä t Erlangen-Nü rnberg, Erlangen, Germany Corresponding author: Maike.till@fau.de
Background
Health promotion projects commonly measure health outcomes and behavior to provide proof of effectiveness. An alternative concept that focuses on the real opportunities a person can choose from to influence their health is Amartya Sen's capability approach. Numerous tools have been developed to measure capability change in general, but it remains unclear which ones can be applied specifically to physical activity (PA).We therefore conducted a systematic review to identify appropriate tools to measure capabilities for physical activity and health and provide information on their quality.
Methods
The review included a total of 6,850 articles published between 2000 and June 2019 that were identified via searches on PubMed, EbscoHost, and ProQuest. Screenings of titles/ abstracts and full texts were conducted independently by two researchers using Endnote X9 and Microsoft Excel. Identified tools are currently being analyzed regarding their indicators, evaluation methods, quality, and the extent to which they address capabilities for physical activity.
Results
The screening resulted in a total of 49 articles included in the analysis. Preliminary results show a diverse use of methods for measuring capabilities for healthy lifestyles. Preliminary results show that three categories of instruments can be identified: (a) Five studies employed secondary data analysis of specific datasets to extrapolate capabilities for healthy living; (b) five articles dealt with measuring capabilities using qualitative approaches (interviews, video recordings); (c) 39 articles reported on a total of 10 different questionnaires to measure capabilities. We identified only one instrument (employing both a questionnaire and qualitative measures) that explicitly measured capabilities for PA, albeit only for a specific target group.
Conclusions
The identified articles show that capabilities for healthy lifestyles are mostly measured by questionnaire. Available tools are mostly target group-and setting-specific. Currently, there is a dearth of tools that explicitly cover capabilities for PA, especially across settings or target groups. Therefore, more research is needed to work towards the development of universally applicable tools to measure PA capability. Keywords: Capability Approach, measurement tools, systematic Review | 2022-08-31T05:06:10.817Z | 2022-08-27T00:00:00.000 | {
"year": 2022,
"sha1": "ea7342a3ee2047c2e706afdae90356d0e283afe6",
"oa_license": "CCBY",
"oa_url": null,
"oa_status": null,
"pdf_src": "PubMedCentral",
"pdf_hash": "ea7342a3ee2047c2e706afdae90356d0e283afe6",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": []
} |
270000607 | pes2o/s2orc | v3-fos-license | Multimodal Learning for Mapping the Genotype-Phenotype Dynamics
How complex phenotypes emerge from intricate gene expression patterns is a fundamental question in biology. Quantitative characterization of this relationship, however, is challenging due to the vast combinatorial possibilities and dynamic interplay between genotype and phenotype landscapes. Integrating high-content genotyping approaches such as single-cell RNA sequencing and advanced learning methods such as language models offers an opportunity for dissecting this complex relationship. Here, we present a computational integrated genetics framework designed to analyze and interpret the high-dimensional landscape of genotypes and their associated phenotypes simultaneously. We applied this approach to develop a multimodal foundation model to explore the genotype-phenotype relationship manifold for human transcriptomics at the cellular level. Analyzing this joint manifold showed a refined resolution of cellular heterogeneity, enhanced precision in phenotype annotating, and uncovered potential cross-tissue biomarkers that are undetectable through conventional gene expression analysis alone. Moreover, our results revealed that the gene networks are characterized by scale-free patterns and show context-dependent gene-gene interactions, both of which result in significant variations in the topology of the gene network, particularly evident during aging. Finally, utilizing contextualized embeddings, we investigated gene polyfunctionality which illustrates the multifaceted roles that genes play in different biological processes, and demonstrated that for VWF gene in endothelial cells. Overall, this study advances our understanding of the dynamic interplay between gene expression and phenotypic manifestation and demonstrates the potential of integrated genetics in uncovering new dimensions of cellular function and complexity.
Introduction
Understanding the intertwined relationship of genotype and phenotype has long captivated biologists.
Wilhelm Johannsen was the first person who coined the terms and attempted to quantify them.In his book "Elemente der exakten Erblichkeitslehre" (1909) [johannsen1909elemente], he attempted to provide an "exakten" description to biological phenomena by relating genotype to phenotype, both conceptually and mathematically.A century later, we still remain significantly distant from achieving a precise quantification of this relationship.In recent years, however, new sequencing techniques, such as single-cell RNA-seq (scRNA-seq), have propelled our understanding forward.These technologies reveal the complex dynamics of gene expression at a cellular resolution, highlighting the vast landscape of genotypes influenced by myriad factors [norman2019exploring].However, these techniques fail to provide a comprehensive mapping of how genotypic combinations give rise to emerged phenotypes.
Current methods in mapping the genotype-phenotype relationship such as forward or reverse genetics, while theoretically capable of dissecting this relationship, fall short in practice due to the sheer scale and complexity involved [replogle2022mapping].In human cells, a combination of thousands of genes gives rise to an incredibly diverse phenome landscape [freimer2003human].Furthermore, the advent of scRNA-seq, while groundbreaking, introduces its own set of challenges.The technique's ability to uncover thousands of gene expression changes across cells offers a detailed view of the transcriptomic landscape but also complicates the task of drawing meaningful biological conclusions from these high-dimensional datasets [kiselev2019challenges, stegle2015computational].
This complexity is exacerbated by the vast amounts of data generated in the field, necessitating advanced computational techniques for analysis.
Recent developments in machine learning, particularly the adaptation of self-supervised transformer architectures from the field of natural language processing (NLP), have shown promise in analyzing complex biological datasets [theodoris2023transfer, cui2024scgpt].In these approaches, each gene is a word and each cell state can be modeled as a sentence equivalent to natural language.
Current models, however, often lack the interpretability needed to fully understand the sources of variability and fail to capture the dynamic interplay between genotypes and related phenotypes.They primarily focus on resolving cellular heterogeneity without addressing the multifaceted nature of gene expression patterns, which are not only cell-type specific but also highly responsive to organism-level phenotypic characteristics such as age and sex.
Here, we propose the application of self-supervised language models to map the genotypephenotype landscape simultaneously and introduce the concept of integrated genetics as opposed to forward and revere genetics.In this approach, the phenotypes such as sex, age, anatomical tissue and cell types are integrated and learned with genotypic information measured by scRNA-seq data.This enhances our understanding of the biological context of gene expression and ultimately the genotype-phenotype relationship.In particular, we apply this approach to Tabula Sapiens cell atlas [the2022tabula], creating an integrated foundation model, called PolyGene, comprised of nearly 500,000 human cells across 24 organs from various age ranges.Our results show that PolyGene successfully learns genotype-phenotype annotations and also offers new insights into contextualized gene network dynamics.Ultimately, our work lays the groundwork for a phenotype-informed deep learning framework that can be fine-tuned for a broad range of downstream applications, potentially revolutionizing our understanding of the genotype-phenotype relationship and accelerating the discovery of therapeutic targets.
Model Overview
PolyGene is a multimodal foundational model to extend the application of attention-based deep learning methods in scRNA-seq data analysis.Our model leverages the unique characteristic of language models, which is their ability to integrate multimodal information by incorporating phenotypic metadata.Multimodal machine learning offers several unique advantages [jayagopal2022multimodal], such as enhanced learning for a more comprehensive understanding of the data [kline2022multimodal], improved robustness in dealing with noisy or incomplete data [cheng2023robustness], and cross-modal understanding between different data modalities.These advantages enable the model to perform tasks that require bridging the gap between various types of data [lin2023multimodality].Therefore, we hypothesize that our method can lead to a more accurate biological contextualization, where a better understanding of variation sources and heterogeneity is necessary.
In our approach, for each cell, the gene expression values and associated phenotypes are embedded into a vector representation and passed into the network (4).We have incorporated different data modalities of the information of single-cell transcriptomes as well as corresponding phenotypes where each gene expression was measured and encoded jointly as the inputs to the model (figure1).With this structure, PolyGene can learn the relationship between phenotypes (sex, age, tissue, and cell type) and contextualized phenotype-genotype relationship.Furthermore, PolyGene develops a more meaningful biological understanding of the expression levels in the context of its preceding (figure 2a).Dimension reduction on contextualized embeddings significantly improves the characterization of cell types, the tissue of origin, and even the age and sex of the donors with notable distinctions (figure 2c and figure S3).Further, the classification of these contextualized embeddings has unveiled the presence of two distinct subclusters within several tissues, including the thymus, trachea, subcutaneous adipose tissue, and the parotid gland (figure 2b).This finding demonstrates the model's sensitivity to subtle intratissue variations, potentially indicative of distinct functional states or developmental stages.Moreover, our analysis highlights a pronounced heterogeneity within the blood and lung tissues, which suggests a higher degree of cellular diversity in these tissues.This Next, we explored the potential of contextualized embeddings in analyzing tissue similarities.By leveraging these high-dimensional embeddings, we conducted a comprehensive examination of the relationships between different tissue types using Pearson correlation coefficients.The results demonstrate a landscape of tissue similarity with encoded biological relationships (figure 2c).An example is the high similarity score observed between the cardiac ventricle and cardiac atrium tissues which aligned with their close functional and anatomical relationships.The tissue similarity results also reveal an unexpectedly high similarity between the tongue, the retinal neural layer, and the kidney.
This finding, although initially unexpected, aligns with previous research indicating the tongue's diagnostic relevance for chronic kidney disease [chen2021tongue, bemelman2014failing].This intricate link highlights the potential of cross-tissue biomarkers in systemic disease diagnosis.The ability of multimodal integrated genetics to independently discover such relationships indicates its utility in revealing novel biological insights beyond traditional analysis methods.
Beyond individual pairwise comparisons, our analysis has revealed a structured hierarchy within the tissue similarity landscape.Specifically, muscle tissues emerge as a distinct cluster, demonstrating inherent similarities among themselves.Yet, this cluster is not isolated; it forms part of a larger conglomerate that includes the aorta, vasculature, and uterus (figure 2b).This broader aggregation suggests a deeper, developmentally rooted connection between these tissues, hinting at shared origins or functional pathways.We observed a categorization of all examined tissues into three major subgroups.Except for some abnormalities, these subgroups align well with the primary germ layers from which they originate i.e. mesoderm, endoderm, and ectoderm.These results demonstrate the model's potential in understanding tissue differentiation and organogenesis based on embedding vectors.
Embeddings encode a dynamic genotype-phenotype relationship
Integrated genetics methodology creates contextualized embedding outputs which can serve as multidimensional representations of genotypic and phenotypic data, encapsulating the intricate associations and variations within biological contexts.The high-dimensional characteristics of the embeddings allow novel analysis of specific phenotypic traits or states associated with certain genotypic profiles.To understand what genes universally have influenced the phenotypes, we calculated the cosine similarity between phenotype embeddings and gene embeddings (Extended data 4).Our results showed that H4C3 (H4 clustered histone 3) is the most prominent gene among all the phenotypes.For tissue phenotype, urokinase plasminogen activator receptor (PLAUR) and transcription factor XBP1 are among the most similar genes observed in all tissues (Extended data 4a).XBP1 is also observed in cell type, age and with lesser impact in sex phenotype.For cell types, however, lysosome C (LYZ) gene has the highest universal similarity (Extended data 4b).Follistatin (FST) gene is one of the most similar genes to the age phenotype embedding which shows the role of tissue activin and cellular proliferation in aging (Extended data 4c).
Further analysis identified differentially embedded genes (DEGs) among phenotypes.Our results showed that for tissue phenotype, IGHV3-7, CCL17 and IGKV6D-21 are the DEGs (Extended data 4e).For cell type phenotype, CCL3 (C-C Motif Chemokine Ligand 3, CA1 (Carbonic anhydrase 1) and CCL4L2 (C-C motif chemokine 4-like) are the top different genes between all cell types (Extended data 4f).The differentiating genes in age phenotype are RGS1 (regulator of G-protein signaling 1), CA1 and CCL3 (Extended data 4g), and for sex phenotype CA1, CCL17 (CC chemokine ligand 17) and EREG (Epiregulin) are shown to be embedded significantly different between male and female (Extended data 4h).Overall, the DEGs show that immune-related genes play the most important role in differentiating the embeddings of various phenotypes.In particular, CA1 and CCL17 molecules play a larger role in differentiating contextualized phenotypes.
We hypothesized that embeddings represent an information-rich representation of each phenotype and genotype.Focusing on tissue phenotype, the embedding analysis distinctly separated tissue subtypes (3a).Specifically, for different muscle cells, genes like DNAJB1, GEM, and GPRC5A were identified as similar across different muscle tissues (3b), while HMGN2P40, CCNB1, and CTD and GNLY genes were notably different (3c).This methodology extended to other tissues, with vasculature and cardiac tissues examined for the impact of genotypes on emergent phenotypes (3).For instance, for cardiac tissue, the maternally expressed gene 3 (MEG3) gene was identified as the top DEG that distinguishes ventricle embeddings from atrium embeddings.Previously, this gene has been implicated in ventricular dysfunctions such as myocardial fibrosis and compensatory cardiac hypertrophy, conditions that predominantly affect ventricular rather than atrial tissue.
[piccoli2017inhibition, zhang2019stat3, mi2023inhibition].Notably, kidney and tongue tissues were found to share significant similarities, attributed to genes such as transcription factor ELF3 (which regulates epithelial cell differentiation), CRYAB (molecular chaperone in skeletal muscle), LGALS4, PDK4 (highly expressed in skeletal muscle), and CD59 (expressed in smooth muscle cells), highlighting the molecular variations among difference muscle cells.Overall, our results demonstrate how high-dimensional embedding analysis can help with understanding the genetic diversity contributing to phenotypic heterogeneity.On the contrary, the analysis of KRT8 (Keratin 8) gene embeddings presents a different scenario.
Contextualized genotype encodings across phenotypes
Here, the small and large intestines exhibit a high degree of similarity when clustered based on KRT8 embeddings, aligning more closely with traditional understandings of tissue similarity (4.This outcome points out the heterogeneous nature of gene function across tissues, with KRT8 demonstrating a more uniform pattern of influence within the intestinal tract.We have explored this polyfunctional characteristic of genes in the next section in more detail.
To demonstrate that the genotypes are dynamic and context-dependent, we examined the distance between gene embeddings in various tissues and cell types.This examination aimed to uncover how the same genes manifest different degrees of similarity or divergence within and across biological contexts.For instance, for aorta, spleen, and subcutaneous adipose, certain genes-NICOA7, XBP1, BNIP3L, and CTSD-exhibited distinct patterns of similarity.These genes, despite being expressed across the examined tissues, showed closer embedding proximity between the aorta and subcutaneous adipose than between the aorta and spleen.This suggests a more nuanced interaction of these genes within the cardiovascular and fat storage systems compared to their roles in the immune function represented by the spleen.Conversely, the embedding distances for TIMP1, CD55, and HSPA5 across these tissues were notably less variable, indicating a more uniform functional role or expression pattern across diverse biological systems.
Expanding our analysis to cell types further revealed the dynamic nature of gene embeddings.
The FBXO7 gene, known for its multifunctional role across various physiological processes, showed significantly closer embeddings between memory B cells and naive B cells than between memory B cells and kidney epithelial cells.This observation highlights the similarity in FBXO7's function or interaction network within B cells, in contrast to its role in epithelial cells.On the other hand, the SMS gene, which encodes the spermine synthase enzyme, maintained a consistent embedding similarity across B cells and epithelial cells, suggesting a universal role that transcends cell type specificity.
Gene Network Structure is Context-dependent
A significant contribution of our work is the illumination of dynamic gene networks that operate across different phenotypic contexts.The integration of phenotypes allowed for identifying embedded gene expression patterns and context-dependent interactions that can be also investigated in the structure of the gene interaction networks.By including cell metadata in our model, we can encode the conditions in which gene expression occurs and sources of variation into the gene embeddings.In this section, we characterize the relationship between these gene embeddings in different contexts.
First, we will analyze the overall structure of gene networks across different phenotypes.
Metabolic, protein, and gene interaction networks have been reported to exhibit scale-free behavior which makes them robust and efficient in processing information [khanin2006scale, arita2005scale].However, the context-dependent architecture of the network from RNA expression data has not been reported before.Here we used similarity between gene embeddings to construct the gene network (where nodes are genes edges are the similarity score higher than a specific threshold).
Our results show that the networks follow a scale-free pattern, where the distribution of connections to nodes follows a power law (Extended data 5).This indicates that most nodes have few connections, while a small portion of important nodes have a huge number of connections (hubs).The degree distribution results across various cell types demonstrate the scale-free nature of all networks, albeit with the phenotype-dependent variation that manifests itself in the slope of power law (Extended and old endothelial cells (bottom) which is more pronounced in the specialized region of the network (depicted by orange color).Genes that are highly different between the two networks are scaled for visibility.c, similarity analysis between gene embeddings reveals differentially embedded genes between young and old endothelial cells (left) and genes with a higher degree of similarity that do not change during vascular aging (right) data 5).Additionally, the construction of gene interaction dendrograms facilitates a deeper understanding of the networks' hierarchical organization.These dendrograms reveal clusters of genes with similar embeddings, each representing a distinct functional unit, such as signaling pathways or cellular response mechanisms (Extended data 5).
Specifically, we investigated the role of aging in the structure of genetic network in endothelial cells (ECs).Our results show that aging alters the slope of the power-law distribution within the genetic network, with a pronounced effect on low-degree nodes-those with fewer connections (5a).These findings reveal a significant restructuring of the network.Specifically, we observed that specialized nodes, which typically have fewer connections, begin to form more linkages, thereby diluting their specialized roles within the network (5b) To pinpoint the genes most susceptible to these age-related structural changes, we delved into differentially embedded genes within aged ECs.
Our analysis highlighted several genes such as KCNH8 (potassium voltage-gated channel subfamily, Kv12.1),DNAJA4 (DnaJ Heat Shock Protein Family (Hsp40) Member A4), and EGLN3 (Egl-9 Family Hypoxia-Inducible Factor 3) as exhibiting the most significant alterations in their network embeddings.These genes are proposed as candidates for further investigation into their roles in vascular aging, potentially offering new insights into the molecular underpinnings of age-related vascular changes.
Polyfunctional gene embeddings represent genetic heterogeneity
The extraction of high-dimensional gene embeddings from our model offers a unique perspective into the diverse contexts and functions where genes are active.Through these embeddings, we can gain an insight into the functional heterogeneity of genes, which we call polyfunctional genes.This is defined as the gene's ability to form multiple, functionally distinct clusters even within the same cell type.Contextualized embeddings reveal the complex regulatory landscapes that genes navigate, the distinct networks that they construct, and potentially the multifaceted roles they play in various biological processes.
To demonstrate the concept of the polyfunctional gene, we have analyzed VWF (von Willebrand Factor) gene within endothelial cells.We identified two functionally distinct clusters that are not phenotype dependent (Extended data 6): the first aligns with VWF's well-documented role in blood coagulation and critical function in hemostasis (cluster 1 in 6a).In contrast, the second cluster reveals a lesser-known aspect of VWF's functionality, related to its involvement in the management of reactive oxygen species and the cellular response to glucocorticoid stress (cluster 0 in 6a)).The mechanism of this polyfunctional manifestation could be RNA splicing, which has been observed previously to play an important role in mediating the expression of VWF in endothelial cells [yuan2013role, hawke2016characterization].
Similarly, our examination of the CD55 gene, known for its regulatory role in the complement system, across T cells presents a similar pattern of functional diversity.The embeddings reveal six distinct clusters (6b), each representing different aspects of CD55's function or suggesting variations in the gene's expression sources within the same cell type.These clusters range from their canonical role in protecting cells from complement-mediated lysis to potentially novel functions that warrant more investigations.This diversity within the gene embeddings in a specific cell type highlights the existence of complex, multifunctional roles that a single gene can embody, which challenges our conventional understanding of gene function.However, polyfunctionality is not a universal characteristic of all genes.For instance, we showed that for transcription factor KLF5, no distinct functional clusters were observed in endothelial cells (Extended data 7).In general, genes that encode surface proteins show higher diversity in their embeddings compared to genes that translate to intracellular proteins.
These findings illustrate the power of contextual gene embeddings in uncovering the broad spectrum of functional heterogeneity inherent in genes.By dissecting these polyfunctional clusters, our method enriches the gene function lexicon and opens new avenues for exploring the dynamic interplay between genes and their expression contexts.This research on the complex roles of genes enhances our understanding of biology and sets the stage for future studies on the molecular mechanisms that influence cellular behavior and organism physiology.One of the main applications of polyfunctionality could be in improving the precision of targeting the special function of a gene in drug discovery and cell therapy.
Discussion
A central goal of genetics is to understand the complex relationship between gene expression profiles and emergent phenotypes.However, studying this relationship is challenging because of the vast array of possibilities and the bidirectional nature of the relationship, which necessitates a holistic approach.Here, we introduce the concept of integrated genetics to navigate the genotype-phenotype landscape simultaneously and provide a multimodal foundational model to accomplish it for human transcriptomics.By modeling these landscapes jointly, we can gain insights into the coordinated mechanisms underlying genetic expression and phenotypic manifestation.
We demonstrated that the resultant embeddings from our approach contain meaningful infor- Additionally, we explored the polyfunctionality of genes through contextualized embeddings and demonstrated the diverse roles a single gene can play in various biological processes (demonstrated for VWF and CD55 genes in fig 6).The tissue-specific polyfunctional characteristics of the genes provide greater insight into the side effects observed when targeting genes clinically; thus, our analysis could aid in designing more accurate and safe therapeutics.
Our study aims to decode the intricate genotype-phenotype relationship by integrating contentrich transcriptomics data and advanced machine learning.By incorporating various factors such as cell type, tissue of origin, and donor demographics into our analysis, we have substantially improved the precision of pattern recognition.This integration allows for the use of information-rich, highdimensional embeddings that offer new analytical capabilities and enhance our understanding of
Methods
In this section, we will provide an overview of the PolyGene architecture and key features.Previous works [yang2022scbert, theodoris2023transfer, cui2024scgpt] on gene expression data from scRNA-seq data have been limited to statistical analysis of a large number of genes, for example, clustering and cell type annotation.This approach mirrors the pre-2010 natural language processing (NLP) research, which primarily focused on clustering, classification, topic modeling, and other statistical analyses.However, the advent of word embedding catalyzed a shift towards more granular text analyses, such as Named Entity Recognition (NER), sentiment analysis, and semantic interpretation [mikolov2013distributed].We advocate for a parallel methodological evolution in genomics, leveraging the exponential growth in genetic data availability across various species.Concurrently, the integration of advanced machine learning frameworks, notably the Transformer architecture's success in NLP and related domains, has paved the way for analogous applications in genetics.
It's been known for a long time that there is a connection between genes in cells in terms of a network also known as Gene Regulatory Network (GRN).Despite this, research in this domain has predominantly been theoretical, with rudimentary models like Stuart A. Kauffman's Random Boolean Networks (RBN) focusing on the overarching characteristics of these networks under varying conditions, such as network stability contingent on nodes and connectivity [kauffman2003random, shmulevich2004activities]. In essence, the methodology to elucidate the complex interrelationships among genes that result in intriguing phenotypic manifestations-phenomena that single gene analyses cannot account for-remained elusive.
The limitations of existing models employing Transformer [vaswani2017attention] architectures in genomics arise from their reliance on a constrained subset of genes, organized by their expression levels from highest to lowest.This ordering premise is problematic for several reasons.Firstly, it presumes, without clear justification, that a descending expression-based sequence represents a gene's natural or functionally relevant order.More critically, this method's validity is compromised by its sensitivity to expression value disparities; minor fluctuations or noise can alter the gene sequence, thereby impacting the analysis.This issue is not encountered in text analysis, where the inherent sequential nature of language enables Transformer models to extract meaningful patterns by evaluating the distance between tokens.For instance, in the textual context, the change of word order in phrases like "it is clear" to "is it clear" significantly shifts the meaning.This natural order is not inherently present in gene expression data.
A potential solution for the positional embedding problem in genetics is its complete elimination, yet this introduces new challenges in loss function computation.The challenge stems from the ambiguity in comparing input tokens with output when the input is permutation invariant.One approach to address this is by generating numerous permutations of the input to refine loss calculation or by adopting methodologies like Set-Transformer.However, these solutions are inefficient for practical implementation due to their computational demands.In our study, we adopt an innovative strategy, employing a consistent positional encoding for each gene, complemented by padding for unexpressed genes within a specific cell.This method effectively resolves the issues associated with loss function computation while circumventing the constraints imposed by gene order.
Input representation
To model our problem we start with a representation of the genes in a single cell C that includes all of them as a vector.For the sake of consistency and without loss of generality, we select a specific and fixed order for these genes: (1) Each gene value is the expression of the gene in that cell that is normalized to a range.To make the values of the expression comparable, we have to normalize the value by dividing it by the dispersion rate.To choose a threshold for distinguishing expressed from non-expressed genes, we conduct a histogram analysis of gene expression values across all cells.This analysis reveals a significant concentration of expression values below 0.1, followed by an increase, forming a distinct bump in the histogram.Such a distribution pattern indicates that a threshold of 0.1 is an effective demarcation for identifying genes as expressed or non-expressed.
After determining the threshold, the rest of the expression range is divided into B equal-sized bins and each gene is represented as the number of the corresponding bin (higher expression corresponds to higher bins).More specifically each gene value belongs to a bin: (2) Given the extensive number of genes, the resultant input vector length, if every gene is included, would be impractically large for efficient computation.To address this, our initial step involves sorting the genes based on their "variability" or the degree to which their expression levels differ across various cells.We then assign a rank to each gene according to this variability criterion.This ranking process enables us to prioritize genes based on their variability, thereby streamlining the input size by focusing on those genes that are likely to provide the most informative insights into cellular functions and states.We can model each cell by T highly variable genes (from all N genes): (3) To pass the genes into the network we need to first embed them in a vector representation.A gene embedding is constructed by mapping its value which is the bin number into an initially random embedding of size h.This way, the cell is represented as a matrix X as follows: (4) In which X i is of size h × 1 and is defined as:
Model architecture
To calculate the contextualized embedding of each gene, a denoiser also known as BERT (Bidirectional Encoder Representations from Transformers) [devlin2018bert] has been employed.The model is dependent on both directions of the input (unlike autoregressive models that are only dependent on the previous words/tokens).In this study, a cell, denoted as X and characterized by a sequence of genes, undergoes a process to generate its corrupted counterpart, X.This corruption involves the random masking of the proportion r of its genes.This procedure can be equivalently conceptualized as independently and randomly masking each gene within the sequence.X = Bernoulli(r) r ∈ [0, 1] (6) The task of MLM (Masked Language Modeling) is to reconstruct the expression values of masked genes based on the known genes.The corrupted version is then fed into the Transformer model.
The Transformer models is decomposed of L layers of Transformer blocks H θ : The Transformer block takes an input of length T , and outputs the same size in the output: In which H θ is the composition of the following functions: LayerNorm and FeedForward are the usual layers used in modern neural networks.MultiHead function is the multi-head attention module that is defined as follows MultiHead(X) = Concat(z 1 , ..., z k )W 0 (10) where Concat(.) is the concatenation function, and the diag(.) is a diagonal matrix with the input vector as the diagonal.Also, 1 L is the all-ones vector of length L and exp(.) is applied element-wise.
All the matrices W Q i , W K i , W V i and W 0 are trainable parameters.
It should be noted that W p is the position embedding that in the first layer of the Transformer maps the discrete inputs into initial random embeddings, in order to distinguish each gene we rely on its position.Note that, unlike text, there is no natural order in gene sequences.So here the position embedding makes it easy to distinguish between each gene.W e is the gene embedding for each expression.The size of position embedding is B. The input to the first transformer block is: In which G is the binned matrix.The last layer of the Transformer is passed through a feedforward layer and results in the "logits" that will be passed through a softmax function to get the probabilities over the expression values.
The MLM task is to minimize the cross entropy between the probability distribution from the model and the hard labels (one hot encoding) from the dataset: Because one hot encoding has been used the equation 15 will reduce to: In which I M (t) is the indicator function for the masked tokens (M ).Here, the independence assumption between the prediction of each gene has been made.
This methodology is pivotal as it enables the extraction of contextualized embedding for each gene, considering the interactions and relations with other genes within the same cell.By capturing the nuanced interplay among genes, this approach allows us to construct a more refined and dynamic model of the gene interaction network.Additionally, a significant advantage of this method over preceding approaches is that the computation of the loss function is not contingent on the precision of gene ranking.
When the number of bins is larger than 2 (B > 2), we are dealing with Ordinal regression.The original implementation of the BERT doesn't have this problem because there is no natural order in the tokens but expressions are binned into an ordered list.For this case, we have to change the loss function to take into account this subtlety.Our approach to solving this problem is using the soft labeling technique.In this technique instead of focusing all the probability on the label (one hot encoding above) we take r percent of it and distribute it equally to other bins, so each other bin has r B−1 probability.Even though this doesn't account for the distance (the bin number 9 is closer to 8 than bin number 1) it still helps the network to converge faster.
To address this challenge, we employ a soft labeling technique.Traditionally, a one-hot encoding scheme would allocate a probability of one to the true label and zero to all others.Soft labeling, however, moderates this approach by redistributing a portion of the probability across all bins.Specifically, we allocate a fraction r of the probability to the true label, while the remaining probability is evenly dispersed among the other bins, with each receiving a probability of r B−1 .This method doesn't explicitly account for the ordinal proximity between bins (e.g., bin 9 is closer to bin 8 than to bin 1); nevertheless, in practice it has the same effect because it facilitates faster network convergence by providing a gradient across adjacent categories, thereby enhancing the learning process regarding the ordinal relationship between bins.
Multi-Task Learning: Integrating Genotype and Phenotype
One of the biggest challenges faced by traditional approaches of modeling genomics data is the problem of relating the Genotype and Phenotype.These relationships are not often very straightforward, linear, or intuitive to understand.In this paper, we use Multi-task learning to understand this problem better.Multi-task learning is an approach in machine learning that focuses on improving generalization in transfer learning using different signals in the training data.Some of the seemingly unrelated signals in the data can be used as inductive biases to learn a shared representation that helps wildly different tasks.
Each cell has other phenotypic features such as sex, tissue type, cell type, developmental stage, and diseases.The formula 1 is expanded and the input to the model can be represented as follows: in which p i 's are the phenotypes and C represents a cell as the input to the model.The model is designed to perform a multi-task learning approach, where the objective is to not only predict the gene expressions but also the phenotypic values at the same time.The masking probability for phenotypes and genotypes are different: phenotypes are masked with higher probability because the number of phenotypes is lower than the genotypes and this helps the model to be more robust.In our experiments, we used %50 for phenotypes masking probability versus %15 for genotypes.
This approach has been very successful in other domains of machine learning such as computer vision and NLP [raffel2020exploring], but their power has not been used extensively in the domain of genomics data.There are several advantages to this technique, for example, it helps to generalize better in phenotypic domains where there is data scarcity.Also, it gives much more flexibility in exploring a joint model that incorporates both phenotypic and genotypic features where we can control and observe the effects of each phenotype on gene expressions and vice versa.
Pre-training
In this study, we implemented an unsupervised learning strategy by employing a masking algorithm.
The input data is randomly masked and a model is trained to make predictions based on the remaining inputs.We randomly mask the non-zero gene expression and then reconstruct the original inputs by model predictions using the remaining genes.This approach essentially enables the model to learn predictive patterns and relationships within the data, fostering a deeper understanding of the gene interactions and dependencies by attempting to reconstruct the original input based on the partial information available.author=Mi, Shan and Huang, Feng and Jiao, Mingli and Qian, Zhuang and Han, Mingming and Miao, Zheng and Zhan, Heqin, journal=Redox Report, volume=28, number=1, pages=2224607, year=2023, publisher=Taylor & Francis
Fig. 1
Fig. 1 Overview of the model.Gene expression data from single-cell RNA-seq and associated phenotypes are embedded as input to a language model.The model learns the genotype-phenotype relationship in an integrated manner, where genes and phenotype are masked simultaneously in the input layer.The outputs of the model are highdimensional embeddings which are information-rich
Fig. 2
Fig.2Resolving heterogeneity in phenotypes using contextualized embeddings.a, UMAP dimension reduction visualization of human cell atlas, colored by tissue type, using gene expression data (left), using output embeddings on the same sample cluster tissue more distinctly (right).b, Pearson correlation between tissues' embeddings demonstrates a hierarchical grouping of different tissues, similar tissues have a higher correlation.c, linear dimension reduction on age embeddings separates different age groups clearly
Fig. 3
Fig.3Contextualized embeddings from integrated genetics help to reveal intricate genotypephenotype relationship.a, UMAP visualization of tissue phenotype embeddings for various tissue types that are considered similar such as muscle, vasculature, cardiac, and tissues are considered different such as kidney, tongue, and retinal tissue.b, similarity analysis between phenotype embeddings of tissue and genotype embeddings present in those tissues can identify genes that have similar roles in these tissues.c, similarity analysis between various tissues can help with identifying differentially embedded genes that have been different in all observed contexts Fig. 4 Dynamic phenotype relationships.Measuring similarity for gene embeddings in different contexts reveals a dynamic relationship between phenotypes.a, analysis of IL2 embeddings demonstrates a pattern of similarity between tissues that do not specifically correlate with the anatomical location of the tissues.b, KRT8 embeddings show a different pattern from IL2 which is more similar to the anatomical organization of the tissues.c, evaluating Euclidean distance between gene embeddings for different tissues (left) and cell types (right) demonstrates that gene embeddings are context dependant of Genes in Old Endothelial Cells
Fig. 5
Fig.5Aging restructures the gene network in endothelial cells.a, degree distribution analysis on endothelial cells shows a scale-free pattern with different characteristics of the networks for endothelial cells of donors under 40 years old (left), and cells from donors more than 60 years old (right).b, Clustering of gene embeddings demonstrates a hierarchical structure in the network highlighting the changes in subnetworks between young endothelial cells (top) and old endothelial cells (bottom) which is more pronounced in the specialized region of the network (depicted by orange color).Genes that are highly different between the two networks are scaled for visibility.c, similarity analysis between gene embeddings reveals differentially embedded genes between young and old endothelial cells (left) and genes with a higher degree of similarity that do not change during vascular aging (right)
Fig. 6
Fig. 6 Contextual gene embeddings capture polyfunctional characteristics in VWF in endothelial cells and CD55 in T cells.a, Clustering of VWF gene's embeddings in endothelial cells reveals two distinct clusters (middle).In cluster 0, VWF interacts with genes that are involved mostly in reactive oxygen species metabolic process (lef), and cluster 1 shows the more canonical function of VWF in blood coagulation.b, Contextualized embeddings for CD55 gene in T cells show 6 distinct clusters (left), the networks that clusters form for clusters 0, 1, and 5 are depicted (right) mation about the genotype-phenotype relationship.Specifically, analyzing embeddings can resolve cellular heterogeneity and significantly refine the accuracy of phenotype annotations, facilitating advanced clustering and similarity analysis demonstrated in fig 1.At the tissue level, our analysis shows potential cross-tissue biomarkers and subtle intra-tissue variations that can be used for systemic disease diagnosis (illustrated in fig 2 and fig 3).Our study highlights the phenotype-dependent nature of gene networks (fig 4), revealing scale-free patterns of the gene network (shown in fig Extended data 5).Specifically, we demonstrated the phenotype-dependant variations in the gene network structure for endothelial cells during aging and proposed potential genes involved (fig 5).
genetic dynamics.The insights derived from our multimodal foundational model open up possibilities for precision medicine to exploit individual genetic profiles, enabling the development of customized therapeutic strategies.Such strategies could be utilized to improve efficacy and minimize adverse effects by predicting potential cross-tissue side effects.Furthermore, this approach holds the promise of integrating more clinical data modalities both related to observed phenotypes and measured genotypes.This could significantly alter therapeutic development, leading to more effective, safer, and finely targeted interventions that are optimized for individual genetic and phenotypic profiles, thereby dramatically improving patient outcomes.
Fig. Extended data 1 Overview of Tabula Sapiens dataset.The human atlas includes nearly 500,000 cells from 24 different tissues listed.More than 400 cell types from 15 donors with different demographics are included in a multimodal dataset.For more details about the Tabula dataset refer to [the2022tabula]
Fig. Extended data 3
Fig. Extended data 3 Visualization of embeddings compared to expressions The embeddings obtained from the integrated foundation model show a higher resolution of cellular heterogeneity.a Principle Component Analysis (PCA) on gene expressions (left) and embeddings (right) colored by sex phenotype.b PCA on gene expressions (left) and embeddings (right) colored by age phenotype.a UMAP visualization of gene expressions (left) and embeddings (right) colored by cell type phenotype.
Fig. Extended data 4 Similarity analysis on phenotype embeddings a-d Top similar genes to tissue, cell, age and sex phenotypes.e-h Top different genes to tissue, cell, age and sex phenotypes | 2024-05-26T05:20:13.715Z | 2024-05-16T00:00:00.000 | {
"year": 2024,
"sha1": "ab4827f60ad81afcd6b7c4af4bb58fc23e2809c1",
"oa_license": "CCBY",
"oa_url": "https://www.researchsquare.com/article/rs-4355413/latest.pdf",
"oa_status": "GREEN",
"pdf_src": "PubMedCentral",
"pdf_hash": "ab4827f60ad81afcd6b7c4af4bb58fc23e2809c1",
"s2fieldsofstudy": [
"Computer Science",
"Biology"
],
"extfieldsofstudy": [
"Medicine"
]
} |
271417006 | pes2o/s2orc | v3-fos-license | A comparative spectral assessment approach of SEBAL and SEBS for actual evaporation estimation in Ardabil Province
Evapotranspiration is a crucial process in the Earth's water and climate cycle, responsible for transforming water from liquid to water vapor. This transformation plays a vital role in the global water cycle and has a significant impact on the climate, weather patterns
Introduction
In recent years, with the development of remote sensing technologies, many mapping techniques and analyzes have been developed [1].Remote sensing has long been an important and effective tool for monitoring land cover, with its ability to quickly provide broad, precise, unbiased and easily accessible information regarding the spatial variability of the land surface.
Problems such as global warming and climate change have increased their negative impact in the world in recent years [2].On the one hand, the use of water resources has been limited, and on the other hand, water consumption in the agricultural sector has increased.while water consumption in the agricultural sector has increased.On the other hand, changes in land use and land cover are also influential factors [3].Therefore, efficient use of available water resources, especially in agricultural uses, which constitute a significant portion of a country's water consumption, is essential.The agricultural sector consumes the largest volume of freshwater in the world, making an efficient method for managing freshwater resources crucial in each country.In this regard, the first step is to calculate the water requirements of plants.Evapotranspiration is a good indicator for assessing irrigation efficiency and overall water consumption of plants.Land surface evapotranspiration (ET) is of prime interest for environmental applications, such as optimizing irrigation water use, irrigation system performance, crop water deficit, drought mitigation strategies, and accurate initialization of climate prediction models especially in arid and semiarid catchments where water shortage is a critical problem [4].
Evaporation is a process through which water from the Earth's surface and water bodies returns to the atmosphere.This process involves the transfer of energy, where water molecules acquire 600 calories of heat, reach the state of vaporization, and are consequently released into the air.The amount of evaporation is influenced by various factors, including solar radiation, air dryness, water temperature, water concentration and color, wind speed, surface type, absolute humidity of the air, and atmospheric pressure.
Evapotranspiration potential, or simply potential evapotranspiration, refers to the amount of evaporation and transpiration that would occur under conditions where sufficient moisture is available throughout the entire period, or in other words, the water level that would be evaporated if moisture resources were present.On the other hand, actual evapotranspiration is the amount that occurs under the natural conditions of a specific region, and its value increases with more available water.However, the actual evapotranspiration will never exceed the potential evapotranspiration.
Asadi and Valizadeh Kamran [5] conducted a study comparing the algorithms SEBAL, METRIC, and ALARM for estimating actual evapotranspiration of wheat crops in the Parsabad-Moghan region in northwest Iran, which is one of the main agricultural areas in the country.The research utilized the Surface Energy Balance Algorithm for Land (SEBAL), Mapping Evapotranspiration at High Resolution with Internal Calibration (METRIC), and Analytical Land-Atmosphere Radiative Transfer Model (ALARM) as the research tools.Twelve satellite images from Landsat 7 and 8 were used, covering the crop development period from 2016 to 2019, and the results were compared with lysimeter data.
Yang et al. [6] conducted a study to estimate evapotranspiration (ET) by combining Bayesian Model Averaging (BMA) with machine learning algorithms.The objective of this study was to reduce errors and uncertainties among multiple ET models to improve daily ET estimation.The results indicated that the BMA method outperformed the eight individual models.Four significant models obtained through the BMA method were ranked based on Random Forest, SVM, SEBS, and SEBAL.The combination of BMA with machine learning can significantly improve the accuracy of daily ET estimation, reduce uncertainties among models, and leverage the distinct advantages of empirical and physical-based models to obtain more reliable ET estimates.
Wei et al. [7] conducted a study on rice growth stage identification and evapotranspiration (ET) estimation in paddy fields using an improved SEBAL model with the consideration of the practical application of the surface resistance equation.Comparison between the estimation results and covariance data showed that SEBALR can provide more accurate ET estimates compared to the original surface energy balance algorithm for land, with a root mean square error (RMSE) of 1.02 mm•d-1, mean relative error (MRE) of 22.97%, and Pearson correlation (R 2 ) of 0.790.Ma et al. [8] conducted a study on estimating regional actual evapotranspiration (ET) using an improved SEBAL model.In this study, a Surface Energy Balance model based on SEBAL and improved sensible heat flux computations (Y-SEBAL) was proposed and used for simulating actual ET at a large scale.The results showed that the Y-SEBAL model's simulation performance was highly consistent with covariance measurement data, with an R value of 0.82, agreement index of 0.90, and root mean square error of 0.81 mm/d.The performance validation indices were better than those of the SEBAL, MOD16, and SSEBop models.The Y-SEBAL model demonstrated the highest sensitivity to wind speed, reaching 0.714.
Therefore, our goal of this research is to estimate plant water needs and optimal management of water resources.For this purpose, SEBAL and SEBS spectra were compared to estimate actual evaporation and transpiration in the study area and the efficiency of these two algorithms was compared.In this regard, we continued to introduce these two algorithms, using each of these two algorithms, we calculated the rate of evaporation and transpiration and compared their results.
Material, methods and case studies
In this study, in order to estimate real evaporation and transpiration based on SEBAL and SEBS methods, we will use satellite images and meteorological data.The satellite image used is the spectral data of OLI and TIRS sensors of Landsat 8 satellite.Meteorological data was also prepared from Iran Meteorological Organization.
Case studies
The study area is located in the northwest of Iran, specifically in Ardabil province, between the cities of Parsabad and Bilasuvar (Moghan Plain), (Figure 1).The area has an average elevation of 100 meters above sea level.The dominant crop in the study area at the time of image acquisition and actual evapotranspiration estimation is wheat.
Material and dataset
The main data required for implementing the SEBAL (Surface Energy Balance Algorithm for Land) model are satellite images and weather data.Analysis of the climate data series collected can provide insights about the changes in climatic conditions in the region [9].
Satellite data
The used images must be cloud-free.In this research, the OLI sensor images from Landsat 8 satellite were utilized.The Digital Image used corresponds to the date of 04/04/2021, and the local time is approximately 11:00 AM.To calculate evapotranspiration, bands 1 to 7, and also band 10 (thermal band) were utilized.
Climate data
The climate data used in the model and for calculating the reference evapotranspiration (ETr) are presented in Table 1.In this research, data from two synoptic stations, Parsabad and Pileh Savar, were utilized, and the final values were obtained by averaging the corresponding measurements from these two stations.ETr represents the evapotranspiration of well-irrigated crops, which is used to calculate the sensible heat in the cold pixel area and ETrF.ETrF is similar to the crop coefficient (Kc), representing the ratio of instantaneous ET (ETinst) calculated for each pixel to the ETr calculated from climate data for the image time.ETrF is used for extrapolating ET from the image time to the 24-hour period (ET24) or longer.Its value ranges between zero and one ET24: Generally, daily ET values are more commonly used than instantaneous ET values.SEBAL calculates ET24 assuming that ETrF is constant over the 24-hour period (i.e., relatively constant throughout 24 hours).ET24 can be calculated (Equation 1) as follows [10]: ET24=ETrF×ETr-24 (1) ET24: ET24 is the total accumulated evapotranspiration over a 24-hour period for the satellite passing day.It is calculated by summing up the hourly ETr values during the satellite passing day.
The SEBAL method
According to the definition, the total evaporation and transpiration from all surfaces of vegetation are referred to as evapotranspiration (ET).Regardless of the partial amount of water used in metabolic activities of plants, evapotranspiration can be considered as equivalent to the water consumed by the plant.Evapotranspiration is a process resulting from the turbulent transfer of energy.The complete energy balance equation can be expressed as Equation (2), where (Rn) represents the net incoming radiation to the surface, (G) is the soil heat flux, (H) is the sensible heat flux, and ETλ denotes the latent heat flux.In this equation, the term (Rn) corresponds to the net radiation received by the surface, while (G) and (H) represent the soil heat flux and sensible heat flux, respectively.Most plants use less than one percent of the received solar radiation during the day for photosynthesis.The heat storage in plants during the day is negligible, and thus, both photosynthesis and heat storage in plants can be disregarded in the energy balance equation.Satellite data provide continuous and temporally diverse information about spectral reflectance and surface radiation.To calculate surface heat fluxes (sensible and latent heat), a multi-stage energy balance algorithm based on physical principles (SEBAL) has been designed.SEBAL uses surface temperature, surface reflectance, and the normalized difference vegetation index (NDVI) as inputs to estimate surface heat fluxes for various land surface covers.By relying on satellite data and a limited amount of meteorological data, the SEBAL model is capable of estimating evapotranspiration.This model calculates the net incoming radiation (Rn), soil heat flux (G), and sensible heat flux (H) (Equation 2), and then subtracts the soil heat flux and sensible heat from the net incoming radiation to obtain the remaining energy, which is equivalent to the energy used for evaporation and transpiration (ET), representing the energy spent on converting water from liquid to vapor (Equation 2 and 3).
Calculation of net radiation
Net radiation (Rn) at the surface is obtained by considering all incoming and outgoing radiation fluxes [10].The value of net radiation should fall within the range of 100 to 700 W/m².
To generate the net radiation raster layer, the input parameters of surface albedo, outgoing longwave radiation, and incoming shortwave and longwave radiation were calculated as 0.895, 314.714, 6584.314, and 714.314 respectively.Raster layers of vegetation index, emissivity, and surface temperature were used as inputs to calculate the outgoing longwave radiation at the longwave band.Finally, the net incoming radiation (Rn) was computed.
Calculation of surface albedo (α):
Surface albedo is defined as the ratio of electromagnetic energy reflected from the soil and vegetation surface to the incoming energy at that surface.
To calculate surface albedo, satellite images were first loaded using the metadata file in ENVI 5.6 software.The Radiometric Calibration command was then used to calculate radiance and spectral reflectance for each band.Subsequently, αtoa (atmospherically-corrected albedo) was computed using the Equation 4 and 5: αtoa: Atmospherically-corrected albedo ρi: Spectral reflectance for each band ωi: Weighting coefficient for each band λi: Wavelength of the band λmin and λmax: Minimum and maximum wavelengths, respectively The weighting coefficients (ω) for each band of Landsat 8 are shown in Table 2. Surface albedo is the ratio of the reflected solar radiation from the soil and vegetation surface to the incoming solar radiation [11].Albedo is influenced by the surface characteristics such as vegetation, soil, and other cover types.The calculation of albedo is performed by correcting Equation ( 4) for atmospheric transparency effects (Equation 6).
Γsw: Atmospheric transparency.path_radianceα: Path radiance-induced albedo, with a value between 0.25 and 0.40.In the SEBAL model, a value of 0.30 is recommended [12], and in this study, a value of 0.30 was selected (Equation 7).
Z: Elevation above sea level in meters.This elevation should represent the elevation of the study area, and it is often recommended to use the elevation of the nearest weather station.In this study, an elevation of 100 meters was considered.
Calculation of incoming short wave radiation (Rs↓)
The shortwave incoming radiation (Rs↓) is the actual solar radiation that reaches the Earth's surface under clear atmospheric conditions.It can be calculated using Equation ( 8): Gsc: Solar constant (1367 W/m²) Cosθ: Cosine of the solar zenith angle dr: Inverse square of the relative distance between the Earth and the Sun Γsw: Solar radiation atmospheric transmission factor The calculated values for dr and θ for the studied region are 0.96 and 23.772°, respectively.
Long wave output radiation (↑) RL
The outgoing longwave radiation (RL↑) is a type of thermal radiation with a wavelength longer than 8 micrometers that is emitted from the Earth's surface into the atmosphere.Its magnitude varies based on the spatial and temporal location and ranges between 200 to 700 watts per square meter.Equation ( 9) can be used to calculate RL↑: ε˚: Surface emissivity, which is the ratio of emitted thermal radiation to that of a perfect blackbody (0 ≤ ε˚ ≤ 1) σ: Stefan-Boltzmann constant (5.67×10^-8W/m²/K^4) Ts: Surface temperature in Kelvin Please note that the value of RL↑ is influenced by the surface temperature and emissivity of the area at the time of imaging.
Long wave input radiation (RL ↓)
The input longwave radiation (RL↓) is the thermal radiation from the atmosphere towards the Earth's surface, measured in watts per square meter (W/m²).It can be calculated using equation (11) and its value varies depending on the spatial and temporal location of the imaging, typically ranging between 200 to 500 W/m².Equation (10) describes the calculation of RL↓: RL↓ = εa × σ × Ta^4 (10) where: εa: Atmospheric emissivity, which is the ratio of emitted thermal radiation from the atmosphere to that of a perfect blackbody (0 ≤ εa ≤ 1) σ: Stefan-Boltzmann constant (5.67×10^-8W/m²/K^4) Ta: Air temperature near the surface in Kelvin Please note that the value of RL↓ depends on the atmospheric emissivity and the air temperature near the Earth's surface at the time of imaging.
Vegetation indicators
The SEBAL (Surface Energy Balance Algorithm for Land) model accepts vegetation indices as input data to calculate surface emissivity, surface temperature, and outgoing longwave radiation in the energy balance equation.Therefore, two important vegetation indices used in this model are presented below:
Normalized difference vegetation index (NDVI)
NDVI is a vegetation index that is sensitive to vegetation cover but cannot eliminate the effects of background soil.Its values range between 1+ to 1-where positive values indicate healthy vegetation.NDVI is calculated using the following formula, by placing near-infrared (NIR) and red bands in the Equation 11 and executing it (Figure 2): NDVI = (NIR -R) / (NIR + R) (11)
Leaf area index (LAI):
LAI is a vegetation index that represents the ratio of the area covered by the vegetation canopy to the ground area beneath it.It is commonly calculated using the relationship between LAI and NDVI (Equation 12): LAI = 4.04 * Ln(NDVI) + 7.04 (12) In this study, after performing radiometric and atmospheric corrections on the input imagery, the vegetation indices NDVI and LAI were generated (Figure 3).
Surface temperature
In order to generate the surface temperature layer, thermal band number 10 of Landsat 8 was utilized.After performing radiometric and atmospheric corrections on band 10, radiance and brightness temperature were calculated using Equations 13 and 14, respectively.These calculations were performed to derive the surface temperature layer in Kelvin using the available thermal band data from Landsat 8 after radiometric and atmospheric corrections (Figure 4).
Surface emissivity
Surface emissivity is the ratio of the thermal radiation emitted by the Earth's surface to that emitted by a blackbody at a specific temperature.In the SEBAL model, two emissivity are defined: the narrowband emissivity (εNB) that represents the behavior of surface emission in the narrow thermal band (with a small bandwidth), and the broad emissivity (ε₀) that represents the behavior of surface emission in the broad thermal band (ranging from 6 to 14 micrometers).The values of these emissivities are calculated using empirical Equation 17 and 18.
For water with α < 0.47 and NDVI < 0, and for snow with α > 0.47 and NDVI < 0, the values of both emissivities are set to 0.985 for ε₀ and 0.99 for εNB.
After calculating the required parameters, the net surface radiation (Rn) is computed as the final output in the SEBAL model (Figure 5).
Soil heat flux (G)
Soil heat flux is the amount of heat stored in the soil and vegetation cover on the Earth's surface due to molecular conduction processes.In the SEBAL model, the ratio of G/Rn is calculated using an empirical equation (Equation 19) presented by Allen (2000).In Equation 19, the temperature is in degrees Celsius (Figure 6).For clear and deep water and for snow, the ratio is set to 0.5.The values of this ratio for other land cover types are provided in Table 3.
Sensible heat flux (H)
Sensible heat flux represents the amount of heat loss from the surface to the air through processes of convection and molecular conduction due to temperature differences.Sensible heat flux is calculated using Equation (20) for heat transfer (Figure 7).
In Equation (20), ρ is the air density (kg/m³), Cp is the specific heat of air (1004 J/kg/K), dT is the temperature difference (T1-T2) between two heights (Z1 and Z2), and rah is the aerodynamic resistance for heat transfer (s/m).
Initially, using data from synoptic weather stations, the wind speed at a distance of 200 meters above the ground (U 200) is calculated.Then, the friction velocity (U*) is estimated for each separate pixel.Finally, the aerodynamic resistance parameter (rah) is calculated for each pixel.After determining the aerodynamic resistance and air density, the coldest and warmest pixels are selected, and based on those, Hcold and Hhot are calculated.Subsequently, dtcold and dthot are determined, leading to the final calculation of dttotal.
After obtaining the initial H, the aerodynamic resistance is corrected, and the sensible heat flux is recalculated.This cycle continues until the average values of aerodynamic resistances converge.In this study, the convergence of rah values were achieved after repeating this cycle five times (Figure 8).
Cold and warm pixels
Sabal uses two reference pixels to determine the boundary conditions in the energy balance equation, which are called cold and warm pixels.Cold pixels are selected from fully irrigated fields and are free from moisture stress, appearing green and vibrant within the study area.In these pixels, the surface temperature and near-surface air temperature are assumed to be equal.Warm pixels are chosen from drylands with no vegetation cover [13] To select these two pixels, a complete understanding of the study area, familiarity with the spectral behavior of phenomena, and proficiency in interpreting images are necessary.The accuracy of calculating evapotranspiration in Sabal relies on the precise selection of these two reference pixels."
Calculation of Hcold and Hhot
Hcold and Hhot are calculated using Equation 25 and 26 respectively.
λ: Latent heat of evaporation ETr: Reference evapotranspiration for the cold pixel In Equation ( 26), the reference evapotranspiration for the reservoir pixel is calculated using the FAO Penman-Monteith method and estimated to be 1743.0.
The value of Hhot for the selected warm pixel is estimated to be 262.499.It is important to note that the value of Hhot should not be less than the value of Hcold.
Calculation of "dtcold" and "dthot"
In the context of the original Equations 27 and 28, "dtcold" and "dthot" are likely terms used in a specific scientific or engineering domain to represent temperature differences or changes.Without further context or specific information about the equations, it's challenging to provide a more precise translation.If you can provide more details or the full equations, I would be glad to assist further.
Modification of aerodynamic resistance
In order to correct the aerodynamic resistance, the length of the Monin-Obukhuv length (L) needs to be calculated (Equation 29).If the value of L is negative, it indicates atmospheric instability, and if L is zero or positive, it signifies atmospheric stability.In order to converge the aerodynamic resistance, the correction cycle was repeated 5 times.which is shown in Figure 9.The instantaneous evapotranspiration (ETinst) can be calculated as (Equation 31): ETinst: Instantaneous evapotranspiration (mm/hr) λ: Latent heat of evaporation of water or the amount of heat needed to evaporate one kilogram of water (J/kg) 3600: Conversion factor from seconds to hours The value of λET can be obtained from Equation (32).To calculate the daily evapotranspiration (ET24), which has more practical significance compared to instantaneous evapotranspiration, the following steps are taken: ETr-24 is obtained by summing up the hourly values of ETr during the day of interest, which can be derived from satellite data.ETr-24 is estimated using CROPWAT software with synoptic data at 5.4 mm per day for the study area.
After calculating Rn, G, and H, the latent heat flux (ETλ) is determined, and then the instantaneous actual evapotranspiration (ETinst) is estimated (Figure 10).Finally, the daily actual evapotranspiration is calculated in millimeters per day (Figure 11).
The SEBS method
SEBS is based on the Crop Water Stress Index (CWSI; [14]), idea in which the surface meteorological scaling of CWSI is replaced with planetary boundary layer (PBL) scaling.It uses the contrast between wet and dry areas appearing within a remotely sensed scene to derive ET from the relative evaporative fraction.
The basis of this method is to use the energy balance equation and calculate the latent heat flux as the residual of this equation for each pixel.This approach follows similar theoretical principles as the SEBAL algorithm.
The required input data include layers generated from satellite images and data obtained from weather stations.The output of the SEBS algorithm, unlike the SEBAL algorithm, provides daily actual evapotranspiration.
Evapotranspiration
The surface energy balance is commonly written as (Equation 34): where Rn is the net radiation flux, G0 is the soil surface heat flux, H is the sensible heat flux, and λE is the latent heat flux.The unit of energy balance terms is watts per square meter.
To estimate the evaporative fraction, SEBS makes use of energy balance at limiting cases at dry limit and wet limit, such that the relative evaporation (ratio of the actual evaporation to the evaporation at wet limit) can be derived as (Equation 35): where the H wet is sensible heat flux at the wet limit and H dry sensible heat flux at the dry limit.The estimations of H wet and H dry were detailed by Su [15].The evaporative fraction (ratio of latent heat flux to available energy) is estimated by (Equation 36 and 37): where λE wet is the latent heat flux at the wet limit (i.e., the evaporation is only limited by the available energy under the given surface and atmospheric conditions).The latent heat flux (λE) can then be calculated by (Equation 38): Finally, the daily actual ET can be written (Equation 39): where ρw is the density of water (1, 000 kgm -3 ) and Rn is the average daily net radiation in this equation.Moreover, the soil heat flux G0 for 24 h is normally assumed negligible (G average).
Daily actual evapotranspiration
The results of the daily evapotranspiration calculated by the SEBS algorithm are presented in Figure 12.
Results
The results obtained from the SEBAL and SEBS algorithms indicate that the SEBAL algorithm exhibits a broader range of actual evapotranspiration values (0.74 to 5.8 millimeters) compared to the SEBS algorithm (1.25 to 8.85 millimeters), demonstrating its greater capability in distinguishing areas with different evapotranspiration rates.On the other hand, the implementation of the SEBAL algorithm is more complex and time-consuming compared to SEBS.The results showed that both algorithms have relatively high capabilities in calculating instantaneous evapotranspiration using spectral data.Estimating plant water consumption on a pixel-by-pixel (spatial) basis is a unique advantage of spectral methods, as other empirical methods provide a single value estimation for all farms and different varieties of a crop.Generally, satellite data has the potential to estimate evapotranspiration for different plant species.Additionally, due to the pixel-based nature of satellite data, it allows for estimating surface properties such as temperature, emissivity, and actual evapotranspiration within a specific region instead of point-based estimation (at a stationary location).This capability is perhaps one of the most important characteristics of satellite data, as it enables the investigation and analysis of spatially distributed environmental characteristics.Therefore, it is recommended that when investigating spatial and temporal changes in environmental variables, the use of raster data, or satellite data, is highly advantageous as it can significantly aid in such investigations with minimal time and cost.The valuable results obtained from such analyses can be crucial for resource management.Additionally, considering the scarcity and inadequate distribution of weather stations and the subsequent unavailability of synoptic data in the country, employing methods based on digital data is highly suitable.
Discussion
The study aimed to estimate actual evapotranspiration using the SEBAL and SEBS algorithms and spectral data from the OLI and TIRS sensors of the Landsat 8 satellite in the Mughan plain of Ardabil province.The results of the analysis revealed valuable insights into the performance of these algorithms and their applicability in the specific study area.
Firstly, the comparison between the SEBAL and SEBS algorithms demonstrated distinct differences in their estimated values of actual evapotranspiration.The SEBAL algorithm showed a wider range of values, indicating its ability to discern variations in evapotranspiration rates across different areas in the Mughan plain.On the other hand, the SEBS algorithm provided a more limited range of values, suggesting a somewhat less nuanced representation of the spatial distribution of evapotranspiration.Secondly, while the SEBAL algorithm exhibited superior capabilities in distinguishing areas with different evapotranspiration rates, it also presented challenges in terms of computational complexity and time consumption.This issue needs to be considered when implementing SEBAL for large-scale or timesensitive applications.
Furthermore, the utilization of spectral data from the OLI and TIRS sensors of Landsat 8 enabled a pixel-based approach to estimate evapotranspiration.This pixellevel estimation offers a significant advantage over pointbased methods since it allows for a more comprehensive analysis of surface properties, such as temperature, emissivity, and actual evapotranspiration, within specific regions.This spatially distributed information provides valuable insights into the environmental characteristics of the study area.
Overall, the results indicated that both SEBAL and SEBS algorithms have relatively high capabilities in estimating instantaneous evapotranspiration using spectral data.This finding highlights the potential of satellite data for accurately estimating evapotranspiration for various plant species.However, it is essential to consider the trade-off between the finer spatial resolution and computational complexity when choosing the most suitable algorithm for a particular study.
The results of this research, in comparison with the findings of previous studies such as the calculation of actual evapotranspiration using the SEBAL algorithm by Asadi and Valizadeh Kamran [5], Wei et al [7], and Ma et al [8], as well as the estimation of actual evapotranspiration using the SEBS algorithm by Yang and colleagues [6] and Matinfar & Soorghali [4], are consistent.
Conclusion and suggestions
In conclusion, the study successfully demonstrated the application of SEBAL and SEBS algorithms in estimating actual evapotranspiration in the Mughan plain.The findings contribute valuable information to water resource management, agricultural planning, and environmental studies in the region.Additionally, the use of spectral data from satellite sensors opens up possibilities for further investigations into the spatial distribution of environmental variables, enhancing our understanding of the local climate and water balance.the application of the SEBAL and SEBS algorithms, along with spectral data from the OLI and TIRS sensors of the Landsat 8 satellite, proved to be a successful approach in estimating actual evapotranspiration in the Mughan plain of Ardabil province.The study provided valuable insights into the spatial distribution of evapotranspiration rates, shedding light on the water balance dynamics and environmental characteristics of the region.
The findings of this research lead to several key conclusions and offer valuable suggestions for future studies: Algorithm Performance: Both SEBAL and SEBS algorithms demonstrated the capability to estimate actual evapotranspiration using satellite data.However, further comparative analysis and validation against ground-based measurements are recommended to identify the strengths and weaknesses of each algorithm for specific study areas and environmental conditions.Spatial and Temporal Variability: The observed wide range of evapotranspiration values highlights the spatial variability in water use and transpiration rates across the Mughan plain.It is essential to consider this variability in water resource management and agricultural planning to optimize irrigation practices and ensure sustainable water use.
Data Integration and Validation: Integrating data from multiple satellite sensors and ground-based measurements can enhance the accuracy of evapotranspiration estimates.Validation of remote sensing-derived results through field measurements is critical to ensuring reliable and precise evapotranspiration calculations.Long-Term Monitoring: Continuous monitoring of evapotranspiration over time can provide valuable insights into climate trends and changes in water availability.Establishing long-term monitoring networks and using historical satellite data can contribute to a better understanding of the region's hydrological dynamics.
Application in Water Resource Management: The estimated evapotranspiration data can be instrumental in water resource management and decision-making processes.Utilizing this information can aid in the sustainable use of water resources and improve irrigation practices in the Mughan plain and other similar regions.Climate Change Implications: Considering the potential impact of climate change on evapotranspiration patterns is essential for anticipating future water availability and planning for adaptation measures.Future studies should explore the correlation between evapotranspiration and climate change indicators.
Capacity Building: Promoting capacity building initiatives among researchers and practitioners in remote sensing and hydrological modeling can enhance the application of these techniques in water resource management and environmental studies.Data Accessibility: Ensuring open access to remote sensing data and making relevant datasets publicly available can foster collaboration and enable more researchers to contribute to advancing the understanding of evapotranspiration dynamics.
In conclusion, this study lays the groundwork for further investigations into evapotranspiration dynamics in the Mughan plain and serves as a valuable resource for water resource management and agricultural planning.The implementation of the suggested improvements can lead to more accurate and comprehensive assessments of evapotranspiration in the region and contribute to sustainable water management strategies.
Table 1 .
Climate data used in the research.
Table 3 .
Values of G/Rn for some land cover types. | 2024-07-25T15:18:24.719Z | 2023-12-09T00:00:00.000 | {
"year": 2023,
"sha1": "6515c25342065f21fb95f5d6486f8a9585a4f2d3",
"oa_license": null,
"oa_url": "https://doi.org/10.26833/ijeg.1344679",
"oa_status": "GOLD",
"pdf_src": "ScienceParsePlus",
"pdf_hash": "5960015dc523008dee1a0cf13de82288b645d821",
"s2fieldsofstudy": [
"Environmental Science"
],
"extfieldsofstudy": []
} |
17405457 | pes2o/s2orc | v3-fos-license | CpG-Oligodeoxynucleotide Treatment Protects against Ionizing Radiation-Induced Intestine Injury
Background the bone marrow and the intestine are the major sites of ionizing radiation (IR)-induced injury. Our previous study demonstrated that CpG-oligodeoxynucleotide (ODN) treatment mitigated IR-induced bone marrow injury, but its effect on the intestine is not known. In this study, we sought to determine if CpG-ODN have protective effect on IR-induced intestine injury, and if so, to determine the mechanism of its effect. Methods and Findings Mice were treated with CpG-ODN after IR. The body weight and survival were daily monitored for 30 days consecutively after exposure. The number of surviving intestinal crypt was assessed by the microcolony survival assay. The number and the distribution of proliferating cell in crypt were evaluated by TUNEL assay and BrdU assay. The expression of Bcl-2, Bax and caspase-3 in crypt were analyzed by Immunohistochemistry assay. The findings showed that the treatment for irradiated mice with CpG-ODN diminished body weight loss, improved 30 days survival, enhanced intestinal crypts survival and maintained proliferating cell population and regeneration in crypt. The reason might involve that CpG-ODN up-regulated the expression of Bcl-2 protein and down-regulated the expression of Bax protein and caspase-3 protein. Conclusion CpG-ODN was effective in protection of IR-induced intestine injury by enhancing intestinal crypts survival and maintaining proliferating cell population and regeneration in crypt. The mechanism might be that CpG-ODN inhibits proliferating cell apoptosis through regulating the expression of apoptosis-related protein, such as Bax, Bcl-2 and caspase-3.
Introduction
The small intestinal epithelium is continuously and rapidly replaced by cells proliferation within the crypts and the subsequent migration of their progeny to the villous epithelium. Intestinal epithelial cells are ultimately derived from these proliferating cell [1]. Although the multiple chemical, physical, infectious and inflammatory agents could cause intestine injury, ionizing radiation (IR) is the most extensive factor for intestine injury, which affects the rapidly proliferating cell population. In small intestine, there is a significant amount of proliferating cell located within crypt, including intestinal stem cells and proliferative crypt cells. Intestinal stem cell is thought to occupy a niche located at approximately position 4 above crypt base. The distribution of proliferative crypt cell present in each crypt remains a subject of speculation (cell position from 5 to 16, approximately). The stem cell and its immediate descendants, proliferative crypt cell, are arranged in a functional hierarchy, where hierarchical position is related to topologic position in the crypt [2]. When mice were exposed to high dose of IR, a large number of proliferating cell located within the intestinal crypt are killed, except fewer survival proliferating cells [3]. These surviving proliferating cells play a central role in the regeneration of the intestinal crypt. They proliferate, form regenerative crypts and eventually repopulate the entire epithelium. With higher doses of IR, the number of surviving proliferating cells may be insufficient to regenerate the crypt, and consequently the number of crypt decrease [1]. Loss of these crypts after IR prevents normal re-epithelialization of the intestinal epithelium, which leads to different degrees of villous fusion and attenuation, as well as intestinal epithelial cells hypertrophy. These changes result in the acute radiation-induced gastrointestinal syndrome (RIGS) presenting malabsorption, electrolyte imbalance, diarrhea, weight loss and death. Therefore, the development of agents for prevention or treatment of IR-induced intestine injury has primarily focused on the protection of intestinal crypt, the mechanism responsible for the maintenance of proliferating cell population and regeneration.
Many studies indicated that bacterial products could affect the intestinal epithelial cellular response to IR injury. Toll-like receptors (TLRs) were found to play an essential role. These receptors recognize pathogen-associated molecular patterns (PAMPs) associated with pathogens [4], mediate interaction between bacterial production and intestinal epithelium, and influence the intestinal epithelial cellular response to IR injury [5]. Different reports have showed that lipopolysaccharide (LPS), a TLR4 ligand, is a radioprotective agent for mice intestine tissue through a prostaglandin E2 (PGE2)-mediated mechanism [1]. Probiotic, a TLR2 ligand, also has similar effect and mechanism [5]. Flagellin or CBLB502, a TLR5 ligand, may reduce the intestine IR injury by NF-kB activation [6,7]. However, due to toxicity and other issues, these ligands are limited to use as a clinical agent.
CpG-oligodeoxynucleotide (ODN), unmethylated short single stranded synthetic DNA molecules, is a TLR9 ligand, which does not present know adverse effects and has been clinically evaluated and used as vaccine adjuvants and antitumor immunotherapy agent [8][9][10]. Our previous study found that CpG-ODN mitigated IR-induced bone marrow injury. In experimental models of IR injury, CpG-ODN given to mice resulted in the inhibition of white blood cell (WBC) apoptosis and enhancement of the bone marrow cell recovery [11]. Recently, G.Pedersen et al and Jongdae L et al reported that CpG-ODN by TLR9 in intestinal epithelium contributed to intestinal homoeostasis [12,13]. Preliminary studies of CpG-ODN indicate that it prevents IR injury in the intestine [14]. However, no information is available on the treatment effect of CpG-ODN on the intestine IR injury, nor any mechanism was described. In the present, our purpose was to determine if CpG-ODN treats IR-induced intestine injury, and if so, to determine the mechanism of its effect.
Reagent
CpG-ODN was synthesized at Shanghai Sangon Biological engineering Technology Services Co Ltd (Sangon, China). CpG-ODN sequences used in this study were 59-TCG TCG TTT TCG GCG CGC CG-39 (regular letters represent phosphorothioate, bold and italic letters represent phosphodiester) [15]. The compound was diluted with phosphate buffer saline (PBS) to a final concentration of 250 mg/ml and stored at 4uC.
Treatment
Mice were treated with either 0.2 ml CpG-ODN (250 mg/ml) or 0.2 ml PBS via intraperitoneal (i.p) injection. CpG-ODN was given 30 min, 24 h and 48 h after IR, unless stated otherwise. Mice were divided into four groups as follows: a normal group (the non-irradiated mice with PBS treatment), a CpG-ODN group (the non-irradiated mice with CpG-ODN treatment), a IR control group (the irradiated mice with PBS treatment) and a IR plus CpG-ODN group (the irradiated mice with CpG-ODN treatment).
Irradiation
Irradiation was done with the 60 Co c-radiation at SMMU facility. Each mouse was placed inside a specially designed and well-ventilated acrylic container with dimensions of 8.0 6 3.5 6 3.5 cm. The mice were then subjected to 8 Gy total body irradiation under standard laboratory conditions. The dose rate was 0.8 Gy/min.
Survival and Body Weight
Mice were treated with CpG-ODN after IR. To determine the survival and body weight, daily monitoring was performed for 30 days. Survival was expressed as percent survival. Average body weight was calculated considering the initial body weight of the animal as 100%.
Histological Assessment
Mice were euthanized by cervical dislocation at 0.5, 1, 2, 3, 5, 7 days after IR. The appropriate segments of the small intestine were rapidly excised, rinsed with PBS, and pinned on a corkboard. Portions of the specimen were fixed for 48 h in 10% neutral buffered formalin. Slides 5 mm-thick were prepared and stained with hematoxylin and eosin (H&E).
Measuring and Scoring
The microcolony survival assay is a reproducible method of assessing injury to the intestinal crypt and has been widely used since it was first described in 1970 by Withers and Elkind [16]. The number of surviving crypts per circumference was counted in 3 sections per mice, using a Nikon ECLIPSE microscope (a 4006 magnification was applied). Only transversely sectioned crypts of 10 or more epithelial cell (excluding Paneth cells) were counted. Crypts consisting of less than 10 non-Paneth cells were considered to have no surviving crypt, and not to contribute to the repopulation of the epithelium. The width of all longitudinally crypts per circumference was also measured under microscope with the aid of a micrometer graticule (5 mm/100 points, Shanghai optical instrument factory, Shanghai, China) placed in an eyepiece with a 10 6 magnification. The width of each crypt was measured at the widest point with a 40 6 magnification. Because of the variation due to different treatments, the mice crypt diameters were corrected according to the factor Dc/Dt, where D represents the crypt transverse diameter, c is normal and t is treated. The number of surviving crypts per circumference for each group was obtained by multiplying the counted crypt by the appropriate correction factor [17].
TUNEL Staining
Apoptotic cells were determined by terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick end labeling (TUNEL) staining. Mice were euthanized 5 hours after IR. Small intestine was rapidly excised, rinsed with PBS, pinned on a corkboard, and fixed in 10% neutral buffered formalin. Portions of the specimen were paraffin embedded and 5 mm sections were cut. Sections were deparaffinized, permeabilized with proteinase K (20 mg/ml, Sigma) at 23uC and rinsed 4 additional times with distilled water followed by incubation with a solution made up of TdT (1 ml/200ml of mix solution), bovine serum albumin (1 mg/ ml), and biotin-16-dUTP (1 nmol/50 ml of mix solution) in TdT buffer. Sections were then incubated in saline citrate for 15 min, rinsed and incubated with 2% bovine serum albumin at 23uC. Endogenous peroxidase activity was blocked by incubating the tissue sections with 3% H 2 O 2 . The labeled DNA fragments were detected with peroxidase-conjugated antibody elicited against biotin. Slides were developed with DBA (diaminobenzidine tetrahydrochloride), and counted after stained with hematoxylin.
BrdU Labeling
Proliferative cells were determined by BrdU incorporation. S phase cells were labeled in vivo by administering BrdU (i.p, 2.5 mg/mice, Sigma, USA) 2 hours before euthanasia. Mice were euthanized 3 days after IR. Small intestine was rapidly excised, rinsed with PBS, pinned on a corkboard, and fixed in 10% neutral buffered formalin, and embedded in paraffin. Sections (5 mm) of paraffin embedded small intestine were deparaffinized and incubated with 3% H 2 O 2 in methanol for 15 min. Sections were incubated with 2N HCl for 1 h, washed in PBS, and then incubated in 0.1% trypsin for 15 min at 37uC. Sections were stained for BrdU incorporation using a BrdU staining kit (Cat.No.KT-077, Kamiya Bioedical Company, California, USA) according to the manufacturer instructions.
Positional Analysis
The intestinal crypts per circumference were scored on a cell positional basis for labeled cells as previously described in detail for the small intestine by Pritchard D.M for the small intestine [19]. Cells were numbered from the base of the crypt, designated cell position 1, and scored as labeled cells, up to the crypt-villus junction. Data are presented as plots of labeled cells index percentage against cell position along the crypt and as mean labeled cells index percentage for each mice group.
Statistical Analysis of Data
For the survival data, the difference in 30 days survival between was analyzed by Kaplan-Meier plots. For body weight, crypt survival, positional labeled cell and the number of labeled cell, the difference between IR plus CpG-ODN group and IR control group was analyzed by the two-sided, non-paired Student's t-test. Data from all the experiments are expressed as mean 6 standard error of mean (SEM). The software analysis program SPSS 13.0 (Release 12.0 K; SPSS Inc., Chicago, USA) was used. All differences were considered significant if p was less than 0.05.
Results
CpG-ODN Diminishes Weight Loss and Improves Survival after IR Mice exposed to 8 Gy produce a characteristic RIGS comprising weight loss and death [3]. This process is divided in four phases: prodromal, latent, critical and recovery [20]. In the prodromal, mice consistently lose body weight. In the latent, they begin to regain weight and reach a plateau. In the critical, mice begin to loss weight again. In the recovery, body weight slowly returns to normality. Irradiated group treated with CpG-ODN showed an increase in the body weight, when compared with irradiated control group in all phases, especially in the prodromal (from 4 to 6 days after IR) and recovery phases (from 16 to 30 days after IR). Furthermore, treating with CpG-ODN after IR injury, body weight came back to normal earlier than that of irradiated control group ( fig. 1A). Death of exposed animal is a typical endpoint in the study of IR injury and described as the most attributable consequence to the combined effects of RIGS and bone marrow failure. Although irradiated mice in the groups either treated with or without CpG-ODN began to die after 9 days of IR, the 30 days survival of the CpG-ODN treated group increased 46% as compared with the control one ( fig. 1B). These results suggest that CpG-ODN decreased mice weight loss and improved the survival after IR injury.
CpG-ODN Enhances Intestinal Crypts Survival after IR
The microcolony assay is a reproducible method of assessing IR injury to the crypt. Mice exposed to 8 Gy IR crypt cell was death.
In an attempt to compensate for the reduced input of cells from crypt to the epithelium, some crypts with surviving proliferative cell increase their size beyond normal value ( fig. 2A). Because of variations in the crypt diameter of mice subjected to different treatment affect the crypt count, a correction for small differences in the crypt size of each group was described. The number of surviving crypts per circumference evaluated considering the correction factors applied ( Table 1). The results showed that IR caused a significant and continual reduction in the number of surviving crypt 3 days after IR. However, the irradiated group treated with CpG-ODN, the number of surviving crypts was significantly larger than that of irradiated control group 3 days after IR. Especially on day 7, CpG-ODN treatment increased the number of surviving crypts from 60% to 92% (fig. 2B). These suggest that CpG-ODN enhanced intestinal crypts survival after IR injury.
CpG-ODN Maintains Proliferating Cell Population and Regeneration after IR
Proliferating cells, including intestinal stem cells located at position 4 above crypt base and proliferative crypt cells located in position 5,16, play an important role in forming the crypt [21]. In order to determine whether CpG-ODN could maintain these proliferating cells population and regeneration after IR, the distribution and the number of their apoptosis and proliferative ability were evaluated by TUNEL and BrdU assays respectively. Firstly, we found that CpG-ODN inhibited proliferating cell apoptosis after 8 Gy of IR ( fig. 3A). The distribution results showed that in the irradiated control group, the apoptotic cell index was .50% from position 3,13. CpG-ODN treatment resulted in a significant index reduction. The apoptotic cell index was .50% only from positions 5,6 ( fig. 3B). Furthermore, CpG-ODN inhibited cell apoptosis at position 4 from 77.8% to 48.7% ( fig. 3C). Position 4 is the position of intestinal stem cell that plays a key role in the regeneration of the crypts after IR exposure [18]. The result showed a 2.2-fold decrease in the number of apoptotic cell in irradiated group with CpG-ODN treatment compared with irradiated control group (fig. 3D). Secondly, CpG-ODN promoted proliferating cell regeneration (fig. 3E). The proliferating cell distribution showed that in irradiated control group, the proliferating cell index was higher than 50% at any position of irradiated control mice. CpG-ODN treatment resulted in a significant promotion as the proliferating cell index was .50% in position 4,9 ( fig. 3F). CpG-ODN promoted cell proliferation at position 4 from 31.0% to 50.8% (fig. 3G). The result showed a 1.9-fold increase in the number of proliferating cell in the irradiated CpG-ODN treated group compared with the irradiated control group (fig. 3H). These results suggest that CpG-ODN maintained the population and regeneration of proliferating cell located within the crypt after IR injury.
CpG-ODN Regulates the Expression of Bax and Bcl-2 Expression after IR
To describe the underlying mechanism of CpG-ODN, we investigated the expression of Bax and Bcl-2 proteins as the prominent members among Bcl-2 family. Mice were exposed to 8 Gy. Bax expression in crypt showed a position-dependent fashion ( fig. 4A). The positive Bax cell index in exposed untreated mice was .5% at positions 2,7. The index reached a maximum of 16.1% at position 5.
CpG-ODN Regulates the Expression of Caspase-3 after IR
In order to elucidate the role of death cascades, the expression of caspase-3 protein was measured. Caspase-3 interacts with caspase-8 and caspase-9, playing a central role in the executionphase of cell apoptosis. Mice were exposed to 8 Gy of IR. Results showed that IR also induced caspase-3 expression in a positiondependent fashion (fig. 5A). The distribution of caspase-3 positive cell showed that for the irradiated control group, caspase-3 positive cell index was .5% at positions 3,7, the index reached a maximum of 16.0% at position 5 and fall off rapidly to below 1%. For exposed mice treated with CpG-ODN, caspase-3 positive cell index was .5% only at position 5. The index reached a maximum of 5.5% at position 5 and fall off gradually to below 1% ( fig. 5B). CpG-ODN reduced the number of caspase-3 positive cell at position 4 from 9.0% to 3.0% (fig. 5C). The results showed a 2.8fold decrease in the number of caspase-3 positive cell for the irradiated group with CpG-ODN treatment compared with irradiated control group (fig. 5D). These results suggest that CpG-ODN down-regulated caspase-3 expression at positiondependent fashion after IR injury.
Discussion
The replacement process of proliferating cells in the crypt maintains normal homeostasis of intestine, these proliferating cells replicate and differentiate in order to continuously and rapidly replace the terminally differentiated epithelial cell. IR-induced intestine injury from direct cytocidal effects on these proliferating cell and subsequent loss of the mucosal barrier, results in RIGS [3].
Under the present protocol, the protective effect of CpG-ODN on IR-induced intestine injury is manifested indirectly by significantly reduction of body weight loss ( fig. 1A), improving mice survival ( fig. 1B). The direct evidence showed the increase in the number of surviving crypt for the irradiated group with CpG-ODN treatment was about 32% compared to irradiated control group on day 7 after IR exposure ( fig. 2B). After radiotherapy at the abdominal or pelvic region, a reduction of the intestinal crypt number is an important factor associated with the side effects that limit the dose delivered to patients. The RIGS syndrome is caused by the loss of a large amount of crypts. The 32% increase of survival crypts is expected to reduce the side effects observed after radiotherapy. It is generally known that the fate of the crypt is determined by the replacement of proliferating cells located within the crypt, including intestinal stem cells and proliferative crypt cells. If all cells die, the crypt is 'sterilized' and disappears. However, if one or more proliferating cells survive, they rapidly proliferate and regenerate the crypt with subsequent reconstitutions of the epithelium. After high dose of IR, most crypt cells undergo apoptosis, or stop proliferating temporarily or permanently, but a few proliferating cells still survive, playing a central role in the regenerating of crypts after IR injury. In particular, there are some stem cells which increase production of proliferative crypt cell, undergo rapid clonal expansion, follow by differential into the mature cells and eventually repopulate the mucosa [3,22]. TUNEL results show that IR induce many crypt cell apoptosis, and the apoptotic response is highest (index .50%) for proliferating cells located at positions 3,14, including stem cell located at cell position 4. However, treatment of CpG-ODN result in a 2.2-fold decrease in the number of apoptotic cell and the apoptotic response is highest for proliferation cells located at positions 5,6, exclusive of stem cell ( fig. 3B and fig.3 C). BrdU results show that a 1.9-fold increase in the number of proliferating cell for the irradiated group with CpG-ODN treatment compared with the irradiated control group (fig. 3F). From these results we conclude that CpG-OND maintains proliferating cell, especially stem cell, located within the intestinal crypt population and the regeneration after IR exposure.
Although the effect of CpG-ODN is not fully understood, one possible mechanism is inhibition of apoptosis. Apoptosis has been known as the main mode of crypt cell death after IR injury, which is modulated through the activation of a particular set of molecular proteins. Bcl-2 family molecules are major molecules of the caspase cascade on cell apoptosis after IR injury [18]. Bax is a proapoptotic cytoplasmic protein that translocate to the mitochondria in response to apoptotic stimuli. Bcl-2 is present in the outer the mitochondrial membrane and is thought to act as an antiapoptotic protein at least partly by inhibiting the translocation of Bax as well as the release of cytochrome c from the mitochondria. [23,24]. The study results show that there were less apoptosis for the irradiated group with CpG-ODN treatment than for the irradiated control group. Positional analysis indicated that the difference was largely confined to proliferating cell in intestinal crypt at positions 3,14, approximately, where Bax expression was down-regulated and Bcl-2 expression was up-regulated. These conclusions suggest that CpG-ODN inhibiting IR-induced proliferating cells apoptosis is associated with the expression of Bax and Bcl-2. In an attempt to better characterize regulation effect of CpG-ODN on apoptotic activity in proliferating cell, the expression of caspase-3 was assessed. Caspase-3 is now considered a key biochemical hallmark of apoptosis. Its regulatory pathways converge on a common cell apoptosis effector called the caspase family and the activation of caspase [24]. The distribution and the number of caspase-3 positive cell were generally consistent with that of Bax positive cell. This result demonstrate that CpG-ODN inhibiting proliferating cell apoptosis may be involved to the regulation of cytochrome c from mitochondria, which interacts with Apaf-1 and subsequently activates caspase-3 via proteolytic processing. Although some key step in the cascade is not determined in the present study [25].
Overall, this study found that CpG-ODN treatment protects against IR-induced intestine injury. The treatment effect is mediated through enhancing intestinal crypt survival and maintaining proliferating cell located within the intestinal crypt population and regeneration. The mechanism might be that CpG-ODN inhibited proliferating cell apoptosis through regulating the expression of apoptosis-related protein, such as Bax, Bcl-2 and caspase-3. The signaling pathway may be involved to the regulation of cytochrome c from mitochondria, which subsequently activates caspase-3. A better understanding of the mechanisms related to the treatment effect of CpG-ODN on IR injury will require further studies. | 2016-05-04T20:20:58.661Z | 2013-06-21T00:00:00.000 | {
"year": 2013,
"sha1": "b79647d465d749b2554222d4b3d40e1f9ba0c7f5",
"oa_license": "CCBY",
"oa_url": "https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0066586&type=printable",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "b79647d465d749b2554222d4b3d40e1f9ba0c7f5",
"s2fieldsofstudy": [
"Environmental Science",
"Medicine"
],
"extfieldsofstudy": [
"Biology",
"Medicine"
]
} |
55543426 | pes2o/s2orc | v3-fos-license | The Ability of Water Plants to Reduce the Level of Mercury Pollution in Water Quality in Irrigation
This research was conducted on July – October 2013 about a mercury analysis which has been performed in Environmental Engineering Laboratory of Engineering Faculty, Andalas University. The level of mercury that is permitted by Government Regulation Republic Indonesia No. 82 of 2001 at the fourth grade for water are at 0.005 mg/l. In that analysis, mercury contents with 0.020169 mg/l at irrigated areas in Batang Hari River. This research aims to find out the ability of water lilies (Salvinia molesta), wood lettuce (Pistia stratiotes), and water hyacinth (Eichhornia crassipes) to decrease the content of water level. This research used experimental methods and the initial content of heavy metals mercury (Hg) by using 0.02 mg/L, 0.06 mg/L, and 0.1 mg/L. The results at decreasing concentrations of heavy metals mercury will be compared with the quality standard of heavy metal mercury at the fourth grade of water. The result showed that water lilies (Salvinia molesta), wood lettuce (Pistia stratiotes), and water hyacinth (Eichhornia crassipes) were able to fix the water quality for irrigation which contaminated heavy metal (Hg). Then, mercury concentration reached a quality standard for irrigation at early concentration 0.02 mg/L during the 15 days and at early concentration 0.1 mg/L during 35 days. From the analysis, it was found that Water hyacinth (Eichhornia crassipes) is the best plant to decrease the concentration of heavy metals mercury.
Introduction
One of natural resources that have an important role in supporting the growth of the plant is water. Water serves as a carrier element of nutrients, assimilation, parts of the plant body and transpiration for the plants. Observing from the water functions, the availability of water must be considered both of quantity and quality. The increasing rate of population growth and human needs for water has caused many negative impacts on the availability and quality of water in the form of pollution and environmental damage. In recent years, the rate of growth of industry and mining in developing countries shows the graph of continuing to rise. Water pollution by heavy metals like Mercury, Lead, Cadmium, Cobalt, Zinc, Arsenic, Iron, Copper and other compounds, originally spread in small concentrations but in its process, it will be accumulated or concentrated so that at certain concentrations it can cause the negative impacts to environment. The common heavy metals like Cd, Pb, Co, Zn and Cr etc. are phytotoxic at both low concentration as well as very high concentration are detected in waste water. If these metals are presented in sediments then these reach the food chain through plants and aquatic animals.
River that has been polluted by industrial waste, mining and house waste contains some elements of heavy metals. Moreover, if it is used to irrigate crops, it will certainly cause the plants that contain heavy metals. From the observation in various rivers in West Sumatra, it can be found that the gold mining content with mercury. One of the causes of environmental pollution by mercury is gold processing tailings disposal which are processed in the amalgamation. From the previous studies, in the district of Dharmasraya, Batang Hari River, it was found that the pollution of Heavy Metal Mercury (Hg) gives the impact to this river due to the process of gold mining, meanwhile, in that area, the river water is a source of irrigation for agricultural land [1].
Heavy Metals Mercury (Hg) is the only metal that is liquid at room temperature. Mercury, both metals and methyl mercury (CH 3 Hg + ), typically enters the human body through digestion. It also can be originated from fish, scallops, shrimp, and the contaminated waters. Mercury in metallic form is not dangerous, because it is only 15% that can be absorbed by human body. But, once it is exposed to nature, at under certain conditions, it could react with methane that has been derived from the decomposition of organic compounds to form toxic methyl mercury which contains toxic. In the form of methyl mercury, most of it will accumulate in the brain. Because of the absorption in a great level, in a short time, it causes variety of disorders. Starting from the destruction of the balance of the body, it cannot concentrate with deaf and various other disorders.
Mercury can dissolve and seep into the ground and there is also the entrance to the plant metabolism and accumulates in all tissues of plants that affects the outcome of these plants. It contains mercury in a quite large part, reducing the amount of chlorophyll of green plants, reducing plant growth, impairing root growth and having a function as well damage leaves and lowers production. Remedial that is measured (remediation) on contaminated land is needed as an effort to reduce the impact of heavy metals on the environment, crops and water creatures. There are many techniques that can be conducted to attempt recovery (remediation) of the contaminated land; one of them is Phytoremediation that utilizes water plants as a cleaning agent. Some of the plants be able to work as phytoremediation agent such as: water liliess, wood lettuce and water hyacinth. The kinds of these plants are aquatic plants that are found in river, beach, pond, and lake in West Sumatera. These plants has an ability called "hyperaccumulator", relatively endured to all of the contaminator and be able to accumulate at enough tissues. Therefore, in this research, these are the plants that reduce the level of mercury pollution in water quality for irrigation.
Materials and Methods
The tools that are used in the implementation of this study are; plastic buckets, 140 ml sample bottles, MVU (Mercury Vaporation Unit). The materials are ; water, Hg (NO 3 ) 2 1000 mg/l and water liliess, wood lettuce and water hyacinth, as well as tools and other materials that support during the study. This study aims to test the ability of water plants (phytoremediation to reduce the pollution of heavy metal mercury (Hg) until it reaches the limit of class-4 of water quality or water quality for irrigation. The treatment involved with three water plants as an agents of phytoremediation such as Water lilies (Salvinia molesta), Wood lettuce (Pistia stratiotes), and Water hyacinth (Eichhornia crassipes). The selection of the three water plants is easily on findings on various waters in West Sumatra. This study used laboratory-scale experiments by using three water plants; water liliess (Salvinia molesta), wood lettuce (Pistia stratiotes), and water hyacinth (Eichhornia crassipes). Levels of mercury (Hg) that will be tested to be lowered by Phytoremediation treatment with three plants are 3 levels of concentration and be repeated 3 times with a 20 cm of depth of solution of 20 cm were 0.02 mg/L, 0.06 mg/L, and 0,1 mg/L.
Plants are selected based on the same weight and covered the entire surface of the container research. The further testing of water samples would be used as a medium to grow three kinds of plants in the laboratory. It is done to measure whether the water contains Hg. If the sample contains Hg, then it will perform an additional heavy metal Hg concentrations until it reaches to three levels, namely 0.02 mg/L, 0.06 mg/L, and 0,1 mg/L. The next step to insert the solution of heavy metals into a plastic box measuring in 80 × 54 × 50 cm and accordance with the desired of water level at 20 cm. The observations are done every five (5) days until Hg concentrations in the water have reached the quality standard for irrigation, which is 0.005 mg/L. When the water sampling will be conducted, the high level of water in the bucket cultivated as the initial state which is 20 cm.
To investigate the decrease of heavy metal concentration, analysis was conducted from water samples in the laboratory by using Inductively Coupled Plasma Mass Spectrometry (ICP-MS). The results of the analysis will be compared to the quality standards for irrigation water. Observations on the study at each treatment is conducted every 5 days to determine the reduction tendency of concentration Hg. The observation will be completed if the reduction in the concentration of heavy metals Mercury has been in the raw water for irrigation. Data which is obtained as a result of decreasing in water pollution by the water lilies (Salvinia molesta), Wood lettuce (Pistia stratiotes) and Water hyacinth (Eichhornia crassipes) would be compared to a quality standards for irrigation. Phytoremediation index calculation (IFR) is done based on data from the treatment. The collected data are calculated to find the declined rate of mercury concentration during the activity. The declined rate of mercury concentrations known as phytoremediation index (IFR) was obtained with formula:
Results and Discussion
The sample tests were conducted at the Laboratory of Environmental Engineering, Andalas University. The average concentrations (levels) of heavy metal Mercury (Hg) in the solution (research vessel) with an initial concentration of 0.02 mg/L during the observation can be seen in Table 1. Table 1: Levels of Hg in solution with initial concentration of solution 0.02 mg/L. Table 1 shows the average concentration of Hg exists in research containers, with an initial concentration Hg 0.02 mg/L. For all the treatments, the quality standard will be achieved after sampling comes on the 15th day. In the treatment with water lilies and wood lettuce, Hg concentrations remains in the solution was right at the limit of water quality standards for irrigation which is 0.005 mg/L on the 15th day. In the treatment with Water hyacinth plants, Hg concentrations remains in the research container is only 0,001 mg/L on the 15th day or it is far under the limit of water quality for irrigation.
The process of absorption and accumulation of heavy metals by plants gets through three processes, namely metal absorption by the roots, metal translocation from roots to other parts of the plant, and the localization of the metal in certain parts of the plant [2]. The reduction trend in heavy metal mercury (Hg) during the observation of the initial concentration of 0.02 mg/L can be seen in Figure 1. Based on Figure 1, it can be seen that the length of time which is required by Water lilies, Wood lettuce and Water hyacinth to reduce the concentration of mercury (Hg) up to the quality standard limits for irrigation with an initial concentration is at 0.02 mg/L. In the containers of Water liliess and Wood lettuce, it takes 15 days to reach the quality of water standard of irrigation and the concentration of mercury (Hg) for irrigation (0.005 mg/L). In the container of Water hyacinth, it takes a shorter than 15 days. Based on the interpolation method, Water hyacinth plants only take 11.67 days to reach below the water quality standard for irrigation. The reduction level of Mercury concentrations then is known as Phytoremediation index (IFR) obtained from the reduction of the initial concentration to the last concentration and it is divided by the initial concentration multiplied by one hundred percent. Phytoremediation index calculation (IFR) is based on data from the treatment. The collected data is calculated to reduce the level of Mercury (Hg) concentration during the activity. The percentage of reduction level of Mercury (Hg) concentration with initial concentration 0.02 mg / L can be seen in Figure 2. The average concentrations of heavy metals Mercury (Hg) in the solution (research container) with an initial concentration of 0.06 mg/L during the observation can be seen in Table 2.
Days
Water lilies (mg/L) Table 2, it can be seen that the average concentrations of Hg exist in research containers with an initial concentration of Hg is 0.06 mg/L. For all treatments, the quality standard of the samplings achieved on the 25th day and there two of the plants (Wood lettuce and Water hyacinth) achieved the concentration before on the 25th day. In the treatment with Wood lettuce and Water hyacinth, concentrations of Hg in the research containers left only 0,004 mg/L and 0,001 mg/L or already under the limit of water quality for irrigation on the 25th day. Meanwhile, in the treatment with Water lilies, Hg concentrations left in the solution was right at the limit of water quality standards for agriculture (0.005 mg/L) on the 25th day. The reduction trend in heavy metal mercury (Hg) during the observation of the initial concentration of 0.06 mg/L can be seen in Figure 3. Figure 3, it can be seen that the length of time which is required by Water lilies, Wood lettuce and Water hyacinth to reduce the concentration of mercury (Hg) up to the quality standard limits for irrigation with an initial concentration is at 0.06 mg/L. In container of Water lilies, it takes 25 days to reach the quality standard concentrations of mercury (Hg) for irrigation (0.005 mg/L). In the container of Wood lettuce and Water hyacinth, it takes shorter than 15 days. Based on the interpolation method, the treatment with Wood lettuce takes 23.75 days to achieve water quality standards for irrigation, whereas treatment with Water hyacinth only takes 21.67 days. It means that the treatment with Water hyacinth on the initial concentration 0.06 mg/L is the best treatment in reducing the concentration of mercury (Hg). The percentage of reduction level of Mercury (Hg) concentration with an initial concentration of 0.06 mg/L can be seen in Figure 4. Figure 4, it can be seen that the reduction level percentage of Hg concentration with initial concentration of a solution of 0.06 mg/L in the containers with water liliess on the 25th day is 92% or 8% remains in solution. In the container with Wood lettuce on the 25th day, it reduces the concentration of mercury (Hg) by 94% and the treatment with Water hyacinth has the highest reduction percentage in heavy metals Hg, which on the 25th day the reduction percentage of Hg concentrations reaches 99% or only it is 1% that is left in solution [3][4][5][6].
Based on
The average concentrations of heavy metal Mercury (Hg) in the solution (research container) with an initial concentration of 0.1 mg/L during the observation can be seen in Table 3. Table 3: Levels of Hg in solution with initial concentration of solution 0.10 mg/L.
From Table 3, it can be seen that the average concentrations of Hg that exist in the research containers with an initial concentration of Hg is 0.1 mg/L or it is the highest concentration among other treatments. For all treatments, the quality standard is achieved before sampling on the 35th day. In the 25th day, Hg concentrations in container research left only 0.003 mg/L (water lilies), whereas for the treatment of wood lettuce and water hyacinth, concentrations of mercury (Hg) left only 0.002 mg/L in the solution. The reduction trend in heavy metal Mercury (Hg) during the observation of the initial concentration of 0.1 mg/L can be seen in Figure 5. Figure 5, it can be seen that the length of time which is required by water lilies, Wood lettuce and Water hyacinth to reduce the concentration of mercury (Hg) up to the quality standard limits for irrigation with initial concentration is at 0.1 mg/L. In all of the water plants treatment at the time of water sampling on the 35th day, the mercury concentration in the container has been below the quality standard of Mercury (Hg) concentrations for irrigation. Based on the interpolation method, the treatment with the water lilies takes 34.17 days to achieve water quality standards for irrigation, the treatment with wood lettuce takes 33.85 days, whereas the treatment with water hyacinth only takes 33.64 days. It means that the treatment with water hyacinth plants with an initial concentration of 0.1 mg/L is the best treatment to reduce the concentration of mercury (Hg) [7][8][9][10].
Based on
The percentage of reduction level of Mercury (Hg) concentration with an initial concentration of 0.1 mg/L can be seen in Figure 6. Figure 6, it can be seen that the reduction level percentage of Hg concentration with initial concentration of the solution 0.1 mg/L in the container with water liliess on the 35th day is 97% or remains 3% in solution. In the container with wood lettuce and water hyacinth on the 35th day can reduce the concentration of Mercury (Hg) by 98% or only 2% left in the solution.
Conclusion
Based on the results of the data and the analysis, it obtained some conclusions as follows : (1) The three water plants that were used namely water lilies (Salvinia molesta), wood lettuce (Pistia stratiotes) and water hyacinth (Eichhornia crassipes) could reduce the concentration of heavy metals Mercury (Hg) in achieving the water quality standards for irrigation. (2) Treatment with water hyacinth (Eichhornia crassipes) was the fastest plant in reducing the concentration Mercury to achieve water quality standards for irrigation at different level of initial concentration ranging from 0.02 mg/L to 0.1 mg/L, (3) For the treatment with initial concentration of 0.02 mg/L, the treatment by water hyacinth (Eichhornia crassipes) took about 11.67 days to achieve water quality standards for irrigation (0.005 mg/L), while for the highest initial concentrations of 0.1 mg/L, water quality standards for irrigation was achieved on 33.64th day. Then, it could be found that water hyacinth (Eichhornia crassipes) was the best plant to decrease the concentration of heavy metals mercury. | 2019-04-15T13:05:15.054Z | 2016-06-23T00:00:00.000 | {
"year": 2016,
"sha1": "a962771a329c3475d6107d5071a7a5d1bca13c16",
"oa_license": "CCBY",
"oa_url": "https://www.omicsonline.org/open-access/the-ability-of-water-plants-to-reduce-the-level-of-mercury-pollution-inwater-quality-in-irrigation-2252-5211-1000225.pdf",
"oa_status": "HYBRID",
"pdf_src": "MergedPDFExtraction",
"pdf_hash": "ae69c88ed72ae28c7ef9db4e78fea33ae8205aee",
"s2fieldsofstudy": [
"Agricultural And Food Sciences"
],
"extfieldsofstudy": [
"Environmental Science"
]
} |
12005655 | pes2o/s2orc | v3-fos-license | The effects of long-term dopaminergic treatment on locomotor behavior in rats
Long-term treatments with dopaminergic agents are associated with adverse effects, including augmentation. Augmentation consists of an exacerbation of restless legs syndrome (a sleep-related movement disorder) symptoms during treatment compared to those experienced during the period before therapy was initiated. The objective of this study was to examine locomotor activity in rats after long-term dopaminergic treatment and its relationship with expression of the D2 receptor, in addition to demonstrating possible evidence of augmentation. The rats were divided into control (CTRL) and drug (Pramipexole—PPX) groups that received daily saline vehicle and PPX treatments, respectively, for 71 days. The locomotor behavior of the animals was evaluated weekly in the Open Field test for 71 days. The expression of the dopamine D2 receptor was evaluated by Western Blot analysis. The animals that received the PPX demonstrated a significant reduction in locomotor activity from day 1 to day 57 and a significant increase in immobility time from day 1 to day 64 relative to baseline values, but these values had returned to baseline levels at 71 days. No changes in the expression of the D2 receptor were demonstrated after treatment with a dopaminergic agonist. This study suggests changes in locomotor activity in rats after long-term PPX treatment that include an immediate reduction of locomotion and an increase in immobilization, and after 64 days, these values returned to baseline levels without evidence of augmentation. In addition, it was not possible to demonstrate a relationship between locomotor activity and the expression of D2 receptors under these conditions.
b s t r a c t
Long-term treatments with dopaminergic agents are associated with adverse effects, including augmentation. Augmentation consists of an exacerbation of restless legs syndrome (a sleep-related movement disorder) symptoms during treatment compared to those experienced during the period before therapy was initiated. The objective of this study was to examine locomotor activity in rats after long-term dopaminergic treatment and its relationship with expression of the D2 receptor, in addition to demonstrating possible evidence of augmentation. The rats were divided into control (CTRL) and drug (Pramipexole-PPX) groups that received daily saline vehicle and PPX treatments, respectively, for 71 days. The locomotor behavior of the animals was evaluated weekly in the Open Field test for 71 days. The expression of the dopamine D2 receptor was evaluated by Western Blot analysis. The animals that received the PPX demonstrated a significant reduction in locomotor activity from day 1 to day 57 and a significant increase in immobility time from day 1 to day 64 relative to baseline values, but these values had returned to baseline levels at 71 days. No changes in the expression of the D2 receptor were demonstrated after treatment with a dopaminergic agonist. This study suggests changes in locomotor activity in rats after long-term PPX treatment that include an immediate reduction of locomotion and an increase in immobilization, and after 64 days, these values returned to baseline levels without evidence of augmentation. In addition, it was not possible to demonstrate a relationship between locomotor activity and the expression
Introduction
Long-term treatment with dopaminergic agents is associated with adverse effects, including daytime sleepiness and sleep attacks, impulse control disorder, addiction, augmentation and other effects [1].
Augmentation occurs in up to 60% of restless legs syndrome (RLS) patients treated with levodopa [2] and, to a lesser extent, dopaminergic agonists [3]. This effect was observed in patients as an increase in severity of RLS symptoms with increasing medication doses. Among the features of augmentation are early onset of symptoms, a decrease in latency to the onset of symptoms the during RLS inactivity, symptoms extending to other body parts (e.g., upper limbs and trunk), and reduced efficacy of medication [4]. There are five known factors that predispose patients to augmentation: drugs with a low half-life, dosage increases, long-term treatments [4], low serum ferritin levels [5], and an RLS-positive family history [6].
Previously, studies on augmentation were not directly based on animal models [5]. The vast majority of the research describes augmentation as a result of the chronic treatment of RLS. Animal models are important tools used to verify hypotheses and decipher the details of pathophysiological mechanisms, including the connections between genes, biology and disease. In this context, considering the effectiveness of PPX on the treatment of RLS and taking into account the results of Chernoloz et al. [7] that demonstrate the facilitatory effect of chronic PPX administration on DA neurotransmission, the assessment of the effect of long-term PPX treatment on locomotor activity in rats has been deemed relevant.
Thus, the objective of this research was to examine locomotor activity in rats after long-term dopaminergic treatment and its relationship with the expression of the D2 receptor and to demonstrate possible evidence of augmentation.
2.
Materials and methods
Animals
A total of 17 three-month-old male Wistar rats (initial and final weight: 270 g and 380 g, respectively) at the Center for Development of Experimental Models for Medicine and Biology (CEDEME-UNIFESP, São Paulo, Brazil) were used in the experiment. The rats were housed in standard polypropylene cages, maintained in a temperature-controlled room (2371 1C) with a 12:12-h light-dark cycle (lights on at 7 a.m.) and had access to food and water ad libitum. The experimental protocol was approved by the Ethics Committee of UNIFESP (CEP: 0881/11). The animals were divided into control group (n¼8) and drug groups (n¼ 9).
Experimental design
Initially, the animals underwent adaptation to the Open Field (10 min). After the adaptation period and baseline test, treatment was initiated, and the animals received a daily dose of either the drug or saline solution at the same time each day for 71 days. Tests were conducted in the Open Field on days 1, 8, 15, 22, 29, 36, 43, 50, 57, 64, and 71 of treatment. The treatment occurred between 7 a.m and 8 a.m on days when the animals did not participate in the Open Field test and between 7 a.m and 10:30 a.m on days when they were evaluated in the Open Field. After the last Open Field test on day 71 of treatment, the animals were euthanized, and the striatum was extracted.
Pharmacological treatment
The treatment consisted of a saline vehicle at 0.9% (0.1 mL/ 100 g) for the control group and pramipexole (0.1 mg/kg, dissolved in the saline solution at 0.9%) for the drug group and was always administered at the same time of day for 71 days [8]. On days when rats were evaluated in the Open Field, the dose was injected into the peritoneal cavity 10 min before the beginning of the test.
Open Field
The exposure to the Open Field [9] was performed between 9 a.m. and 11 a.m., and each animal was individually placed in the center of the apparatus and observed for 10 min. However, only the final 5 min was evaluated. The Open Field consisted of a circular wooden ring measuring 81 cm in diameter and enclosed by 41 cm-high white walls. The ceiling was open, and the ground was divided into 19 quadrants. In line with the circadian pattern of human RLS symptoms, which appear or worsen during the night, all tests were performed between 9 a.m. and 11 a.m. because rats are nocturnal animals [10]. During the experiment, the rats were evaluated for peripheral ambulation (number of quadrants that the animal stepped on with four paws near the walls of the apparatus), central ambulation (number of quadrants that the animal stepped on with four paws that were not close to the walls of the apparatus), total ambulation (the total number of quadrants that the animal stepped on with four paws) rearing (the number of times that the animal supported itself with both hind legs), total duration of grooming (the total amount of time that the animal put its mouth or paws on its body or head) and immobility time (the total amount of time that the animal remained perfectly still, moving only the vibrissae) [11].
Western Blot
After decapitation, the striatum (caudate and putamen) was rapidly removed, and the fragments were frozen. During the dissection, the equipment was kept at a low temperature, and the biological material was stored at À80 1C until used. D2 dopaminergic receptor and the dopamine transporter (DAT) expressions were evaluated by Western Blot analysis. The tissue was homogenized in lysis buffer (1% Triton X-100, 0.5% sodium deoxycholate, 100 mM Tris-HCl pH 8.0, 150 mM NaCl, 10 mM EDTA, 0.1% SDS, 10% glycerol, and Protease Inhibitor Cocktail (Sigma), incubated on ice for 10 min, centrifuged at 7000g for 5 min at 4 1C, and the supernatant was collected. Protein concentration was assayed using the bicinchoninic acid method (BCA, Pierce). Protein from each sample (20 mg) was subjected to SDS-PAGE and then transferred to a PVDF membrane. The membranes were incubated with blocking solution for 1 h (5% skimmed milk diluted in TBS-T (50 mM Tris pH 7.4, 150 mM NaCl, and 0.1% Tween-20), incubated with a primary antibody diluted in TBS-T (1:500) for 1 h, and washed 3 times with TBS-T for 5 min. After washing, membranes were incubated with the diluted anti-rabbit secondary antibody in TBS-T for 1 h (1:10,000) and washed five times with TBS-T for 5 min. Finally, the bands of interest were revealed with a chemiluminescent substrate (Pierce) and a digital imaging system (MF-ChemiBIS Imaging System). After image acquisition, quantification of proteins of interest was performed using the Totallab 100 (Nonlinear dynamics) software.
Statistical analysis
The experiments were evaluated with two-way analysis of variance (ANOVA) with repeated measures followed by posthoc Tukey's post-hoc test. The non-parametric Mann-Whitney test was used for the D2 receptor analyses. The results are expressed as the mean7standard error of the mean (SEM). The level of significance was set at po0.05. Fig. 1A shows that the drug group exhibited a significant increase in the immobility time compared to baseline from day 1 to 64 of the treatment. It also shows a significant increase in the immobility time for the drug group compared to baseline from days 15 to 36 of the treatment. Moreover, the drug group exhibited a significant increase relative to the control group between days 1 and 43 and on day 57 of the treatment. The control group exhibited a significant increase in immobility time relative to baseline from days 15 to 36 of the treatment.
Total ambulation
ANOVA for repeated measures revealed significant effects for the factor time (F (11.165) ¼ 2.7735, po0.001), and the interaction groups  time (F (11.165) ¼ 3.9991, po0.001). Fig. 1B shows that the drug group exhibited a significant reduction in total ambulation relative to baseline from days 1 to 57 of the treatment. In addition, the control group exhibited a significant increase in total ambulation compared to the drug group on days 8 and 43 of treatment.
Rearing
ANOVA for repeated measures revealed a significant effect for the factor time (F (11.165) ¼ 3.8419, po0.001). Fig. 2A shows that the drug group demonstrated a significant decrease in rearing relative to baseline from days 1 to 43 of the treatment. This figure also shows that the control group exhibited a significant decrease in rearing in relative to baseline from days 15 to 29 of the saline solution treatment. The control group exhibited a significant increase compared to the drug group on day 1 of the treatment.
Grooming (time)
ANOVA for repeated measures revealed significant effects for the factors groups (F (1.15) ¼ 6.441, po0.02273) and time (F (11.165) ¼ 6.767, po0.001). Fig. 2B shows that the drug group demonstrated a significant decrease in grooming relative to baseline from days 1 to 71 of the treatment. It is noteworthy that the control group showed a significant decrease in grooming relative to baseline from days 1-8 to days 22-71 of the treatment. There were also significant differences between the groups on day 15 of the treatment.
D2 receptor expression
The results showed no significant differences in the expression of the dopamine D2 receptor between the control and drug groups after 71 days of treatment (Fig. 3).
Discussion
This study aimed to assess the long-term effects of dopaminergic treatment (PPX) on locomotor activity and their relationship with the expression of the D2 receptor.
Pramipexole (PPX) is a non-ergot agonist of D2/D3 that is used for the treatment of RLS and Parkinson's disease [12]. The increased effectiveness of this medication has been demonstrated by numerous multicenter studies [13]. However, many patients have experienced residual symptoms during treatment, and from 12% to 25% of patients discontinue use due to either side effects or the loss of efficacy [1]. Ferini-Strambi [14] estimated that 8.3% of 60 patients treated for 6 months with dopamine agonists showed augmentation. In another study with a mean duration of 27.2 months, Silber et al. [15] found that 33% of the 49 patients developed augmentation after 13.8 months of treatment. García-Borreguero et al. [16] suggested that 20% of patients develop augmentation after 1 year of treatment, and the rate increases to 30% after 2 years.
Currently, the precise nature of augmentation is not fully understood. What is known is that it is caused by a disturbance of the dopaminergic system [5]. It is believed that the increased dopaminergic state is caused by excessive stimulation of D1 compared with D2 receptors, which leads to decreased sensitivity and a down-regulation of the D1 receptor. It is also believed that this hyper-dopaminergic state leads to a supersensitivity of the dopamine receptor, similar to a mechanism proposed for tardive dyskinesia. Another theoretical possibility based on excessive orexinergic stimulation states that this stimulation would lead to a hyper-motor syndrome via the large projections of orexinergic neurons to the monoaminergic system and the anterior horn cells of the spinal cord [4].
The results of this study showed that animals treated with PPX for 71 days initially exhibited reduced locomotor activity and increased immobilization, and after 64 days, these behavioral characteristics started to return to baseline levels.
Using PET, Cervenka et al. [17] demonstrated increased availability of D2 receptors in patients with RLS. This interpretation is consistent with the dopaminergic neurotransmission hypothesis in RLS, wherein the increased Field. * indicates a significant decrease in rearing in the drug group compared to the baseline level from days 1 to 43 of treatment (ANOVA, po0.05); @ indicates a significant decrease in rearing in the control group relative to baseline from days 15 to 29 of the treatment (ANOVA, po0.05); # indicates a same day difference between the groups on day 1 of treatment (ANOVA, po0.05). B assessment of grooming (time, seconds) in the Open Field in the control (saline) and drug (pramipexole) groups. * indicates a significant decrease in grooming in the drug group compared to baseline from days 1 to 71 of treatment (ANOVA, po0.05); @ indicates a significant decrease in grooming in the control group relative to baseline from days 1 to 8 and days 22 to 71 of treatment (ANOVA, po0.05); # indicates a same day difference between the groups on day 15 of treatment (ANOVA, po0.05). concentration of receptors may be due to the positive regulation of the receptor in response to low concentrations of dopamine. Interestingly, sustained treatment with PPX (which over stimulates somatodendritic D2/D3 receptors) can lead to a decrease in D2/D3 receptor responsiveness and a subsequent restoration of the mean firing rate of DA neurons [7]. The present study showed no significant difference in D2-like receptor expression due to long-term PPX administration (71 days). However, it is known that continuous stimulation during treatment down-regulates the number of receptors, and, by analyzing Fig. 3, there seems to be a tendency for a reduction of D2 receptor expression. This finding is consistent with results previously reported by Chernoloz et al. [7] that demonstrated a decreased sensitivity and density of D2-like receptors following chronic administration of the D2-like agonist PPX Studies have shown that the development of regular locomotion, requires a synergistic interaction between the dopaminergics D1 and D2 [18] or D1 and D3 receptors [19,20]. However, activation of the postsynaptic D2 receptors can increase locomotion. In the hyper-dopaminergic phase of RLS treatment, hyper-excitability of the dopaminergic system leads to a reduction of the flexor reflexes and an initial increase of the delayed flexor reflexes [5]. This result was also confirmed in the present study because during the initial phase of the PPX treatment, the rats showed a significant reduction in locomotion and a substantial increase in immobilization.
The present study has some limitations. Despite relatively long-term treatment, our animals did not demonstrate evidence of augmentation (by the Open Field test). This may be due to possible tolerance acquired during the long treatment period. Thus, an increase in the drug dose might have demonstrated the augmentation that usually occurs in humans.
In summary, the present results demonstrate the effects of long-term dopaminergic treatment on locomotor behavior in rats. Rats that received PPX treatment for 71 days showed an immediate reduction in locomotion and an increase in immobilization, and after 64 days, these values returned to baseline levels without evidence of augmentation. In addition, under these conditions it was not possible to demonstrate the relationship between locomotor activity and the expression of D2 receptors.
Thus, for future studies, it is important consider experiments using spontaneously hypertensive rats, an animal model of RLS previously reported by Esteves et al., [21], to investigate the underlying pathophysiology of augmentation. | 2016-10-14T09:01:21.650Z | 2014-11-12T00:00:00.000 | {
"year": 2014,
"sha1": "3ecf56ec8228b7bb4f81a88058bedb4e84fee008",
"oa_license": "CCBYNCND",
"oa_url": "https://doi.org/10.1016/j.slsci.2014.10.003",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "3ecf56ec8228b7bb4f81a88058bedb4e84fee008",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Psychology",
"Medicine"
]
} |
264059130 | pes2o/s2orc | v3-fos-license | Controlled Memristic Behavior of Metal-Organic Framework as a Promising Memory Device
Metal-organic frameworks (MOFs) have attracted considerable interests for sensing, electrochemical, and catalytic applications. Most significantly, MOFs with highly accessible sites on their surface have promising potential for applications in high-performance computing architecture. In this paper, Mg-MOF-74 (a MOF built of Mg(II) ions linked by 2,5-dioxido-1,4-benzenedicarboxylate (DOBDC) ligands) and graphene oxide composites (Mg-MOF-74@GO) were first used as an active layer to fabricate ternary memory devices. A comprehensive investigation of the multi-bit data storage performance for Mg-MOF-74@GO composites was discussed and summarized. Moreover, the structure change of Mg-MOF-74@GO after introducing GO was thoroughly studied. The as-fabricated resistive random access memory (RRAM) devices exhibit a ternary memristic behavior with low SET voltage, an RHRS/RIRS/RLRS ratio of 103:102:1, superior retention (>104 s), and reliability performance (>102 cycles). Herein, Mg-MOF-74@GO composite films in constructing memory devices were presented with GO-mediated ternary memristic properties, where the distinct resistance states were controlled to achieve multi-bit data storage. The hydrogen bonding system and defects of GO adsorbed in Mg-MOF-74 are the reason for the ternary memristic behavior. The charge trapping assisted hopping is proposed as the operation mechanism, which is further confirmed by XRD and Raman spectra. The GO-mediated Mg-MOF-74 memory device exhibits potential applications in ultrahigh-density information storage systems and in-memory computing paradigms.
Introduction
Resistive random access memories (RRAMs) have emerged to be readily on-chip integrated with microprocessors, whose essential physics comprise Pool-Frenkel emission [1], Schottky barrier [2], trap charging/discharging [3], space charge limited current (SCLC) [4], formation, and rupture of conductive paths (CPs) [5].They are subject to the requirements of low voltage, high endurance, and fast access speed.Currently, various hybrid materials have been increasingly applied in RRAM devices as active layers, like metal nanoparticles embedded in a polymer matrix [6], 2D nanomaterial-based composites [7], organic-inorganic hybrid perovskite (OIP) [8], and metal-organic frameworks (MOFs) [9].The memristor is a two-terminal RRAM device that displays memristic behaviors through field-dependent hysteresis.In contrast with traditional binary computing systems, multi-bit memristic behaviors not only enormously increase the information processing capability but contribute to the simple configuration with better anti-inference ability.This demonstrates that ternary memory and logic processing requires a device with reliable tristable memristic states and equal amplitudes of the set and reset voltages.However, the uncontrolled migration of the ionic species inside the oxide switching matrix leads to the random evolution of the metal-conductive filaments.As a matter of fact, the memristors commonly exhibit fluctuation in their programming voltages and memristance.Concerning this, efforts are made to optimize the chemical composition and morphology of the memristic materials.That is to say, improvement via material innovation remains desirable and challenging.
MOFs can be seen as a category of hybrid porous and crystalline materials and a favorable candidate to bridge the gap between inorganic and organic systems [10] on the basis of metal ions or clusters connected by organic linkers.They have the potential for fast ion transport, which is promising for advanced memory and logic concepts.Up to now, there have been some reports on binary MOF-based memory devices [9] but reports on ternary memristic behaviors are still rare.Upon this consideration, the focus in this work is on the development of a novel MOF's composite based on Mg-MOF-74@graphene oxide (Mg-MOF-74@GO), avoiding the issue that the typical high porosity of MOFs may easily bring about short-circuited devices.
Recently, a diversity of 2D layered nanomaterials has been used for developing RRAM devices with low power consumption and multiple resistance states, such as graphene [11], molybdenum disulfide (MoS 2 ) [12], hexagonal boron nitride (h-BN) [13], and black phosphorus [14].They were employed as the main component for floating gates, transistor channel materials in the flash memory, or as active layers for RRAMs in virtue of their unique band structure, high electron mobility, atomic-scale thickness, and outstanding photoluminescence properties [9].Elaborately, carboxylate groups in GO have the bridging dentate coordination ability to metal ions so that they will promote the strength of metal-ligand bonds and give rise to the adhesion with MOF onto GO.Therefore, MOF@GO composites will provide a valuable platform to connect the intermolecular interaction and electrical performance and realize the memristic properties.
For Mg-MOF-74 with excellent adsorption capacity [15], Mg(II) ions and the organic linker are linked to chains and are arranged in a parallel and hexagonal 1D pore, leading to a high surface area.Herein, the metal framework (Mg-MOF-74)-coated GO composites were used as an active layer in a memory device.The current-voltage (I-V) characteristics revealed ternary memristic behaviors with a low SET voltage, high R HRS /R IRS /R LRS ratio, long retention time (>10 4 s), and reliability performance (>10 2 cycles).To extend the functionality and adaptability of RRAM devices, Mg-MOF-74@GO composites should be promising and attractive when constructing functional synapses and smart devices.
Memory Device Fabrication
Mg-MOF-74 (molecular formula C 8 H 4 O 8 Mg 2 , molecular weight 276.72256) was purchased from XFNANO Inc., Jiangsu, Nan Jing, China, and used without any cleaning or treatment.Various amounts of GO (Tanfeng Tech.Inc., Suzhou, China) were dissolved in 20 mL of distilled water.Mg-MOF-74@GO composites were prepared by mixing the aqueous solution in the presence of GO.The mixtures of Mg-MOF-74@GO solutions (concentration of 5 mg/mL) were stirred until homogeneous aqueous solutions were achieved at a weight ratio of 2:1, 4:1, and 5:1, respectively.To prepare the Mg-MOF-74@GO devices, a glass substrate coated with an indium tin oxide (ITO) layer (South China Science & Technology Company Limited, Shenzhen, China) utilized as the bottom electrode (BE) was cleaned by ultrasonication in acetone (Kermel Chemical Reagents Co., Ltd., Tianjin, China), absolute methanol (Kermel Chemical Reagents Co., Ltd., Tianjin, China), and absolute alcohol (Tian in Fuyu Fine chemical Co., Ltd., Tianjin, China) sequentially for 30 min.In the following step, the substrate was dried on a hot plate (Bluepard Experimental Instrument Co., Ltd., Shanghai, China) at 40 • C. Next, the Mg-MOF-74@GO film was spin-coated onto ITO at 5000 rpm for 60 s.After that, the film coated onto ITO was left to anneal on the hot plate at 60 • C for 5 h.Finally, Ni electrodes used as the top electrodes (TEs) with an active area of 1 × 1 mm 2 were thermally evaporated onto the Mg-MOF-74@GO film by a ZZ-450A Vacuum Thermal Evaporator (Beiyi Innovation Vacuum Technology Co., Ltd., Beijing, China), using a shadow mask under 5 × 10 −4 torr.All of the memory devices were fabricated without packaging.
Mg-MOF-74@GO Composite Characterization
Mg-MOF-74 appeared as yellowish fine powder while Mg-MOF-74@GO composites featured a slightly darker tone.The crystalline structure of the samples was confirmed by means of powder X-ray diffraction (XRD).XRD patterns were recorded by an X'PERT X-ray diffractometer (Panalytical Analytical Instruments Company, Almelo, The Netherlands) with Cu Kα radiation (λ = 1.5406Å), a step size of 0.02 • in 2θ which ranged from 5 • to 70 • , and a scan rate of 1 • /min.
Fourier transform infrared (FTIR) spectroscopy and Raman spectroscopy were adopted to obtain the structure information of Mg-MOF-74@GO composites.FTIR spectra were recorded with a Foss DS 2500 Infrared Spectrometer (Foss NIRSystems Inc., Hillerød, Denmark) through KBr pellets in the transmission mode of in 4000-500 cm −1 region.Raman spectroscopy (Horiba Jobin Yvon Inc., Villeneuve-d'Ascq, France) was scanned from 100 cm −1 to 3200 cm −1 , with a 785 nm laser source, and a laser power of 50 mW.
Thermogravimetric analysis-differential thermal analysis (TGA-DTA) and differential scanning calorimetry (DSC) tests were performed to investigate the thermal stability of the samples.TGA-DTA was carried out with a TA Instrument (TA Instrument Inc., New Castle, DE, USA).The analyses detected from 40 • C to 600 • C were performed under a nitrogen atmosphere at a heating of 10 • C/min.DSC was implemented on a NETZSCH DSC 3500 (Netzsch Scientific Instruments Trading Ltd., Selb, Germany) to measure the glass transition temperature (T g ) of Mg-MOF-74@GO composites.The samples were sealed in aluminum pans and heated from 40 • C to 450 • C (heating/cooling rate of 10 • C/min).
Scanning electron microscopy (SEM) (Themoscientific Inc., Waltham, MA, USA) images were implemented at 20 kV and detected under the magnification of 120,000.Transmission electron microscopy (TEM) images of the Mg-MOF-74@GO composites were detected by a JEM-2100 TEM (JOEL Inc., Tokyo, Japan) operated at 200 kV.
A Keithley 4200-SCS semiconductor parameter analyzer (Tektronix Inc., Solon, OH, USA) was employed to characterize the electrical properties of the Mg-MOF-74@GO devices at room temperature.In the characterization progress, TEs were applied as voltage bias while BEs were the ground connection.
X-ray Diffraction (XRD) Spectroscopy
To confirm the obtained samples from their crystal structures, the XRD patterns (Figure 1) were analyzed for the prepared samples.The XRD pattern of Mg-MOF-74 describes sharp and robust diffraction peaks, which suggests the crystallinity and crystal structure of Mg-MOF-74 are perfect.Elaborately, there were three prominent diffraction peaks at angles 2θ of 6.8 • , 11.8 • , and 18.5 • corresponding to the crystallographic planes of (210), (300), and (510) of Mg-MOF-74 single crystal [16], respectively.The XRD patterns of the Mg-MOF-74@GO composites were similar to that of Mg-MOF-74.The featuring characteristic XRD signals of GO at 11.4 • could not be detected in Mg-MOF-74@GO composites by this analysis.No obvious characteristic diffraction peaks of GO appeared in the XRD spectrum patterns of Mg-MOF-74@GO composites because GO was encapsulated inside Mg-MOF-74 or the content was low [16].Moreover, a sharp peak of GO at 11.4 • indicates the characteristic (002) reflection with interlayer spacing of 7.70 Å.The interplanar spacing of Mg-MOF-74@GO(2:1), Mg-MOF-74@GO(4:1), and Mg-MOF-74@GO(5:1) composites was 7.45 Å, 7.51 Å and 7.48 Å, respectively.
The pristine Mg-MOF-74 and its composites exhibited sharp characteristic peaks at 6.8 , and 34.5 • .The well-defined sharp peaks of the XRD patterns for all samples confirm that Mg-MOF-74 and its composites have good crystallinity.However, qualitative comparison between the Mg-MOF-74 composites and the pristine material witnesses a light narrowing and an intensity diminishing in the principal peaks at 6.9 • and 11.9 • , indicating that the doping implies some structural defects [16].In contrast with Mg-MOF-74, the first Bragg diffraction peak of Mg-MOF-74@GO(2:1) has a slight shift from 6.8 • to 6.9 • , and the peak intensity ratio I 300 /I 210 increases from 1.0 to 1.2.These changes suggest that GO was successfully encapsulated into Mg-MOF-74.
Fourier Transform Infrared (FTIR) and Raman Spectroscopy
As a matter of fact, FTIR spectra (Figure 2a) can effectively detect the composition of the sample.The FTIR spectrum of Mg-MOF-74 can be divided into two different regions.On the one hand, most of the activation modes above the 700 cm −1 region belong to the organic ligands [17].In the spectrum of Mg-MOF-74, the peak at 1574 cm −1 is ascribed to the vibration of the C=C bond in the benzene ring.The peak at 1417 cm −1 is associated with -O-C-O in the carbonyl group and the peak at 1208 cm −1 conforms to the C-O vibration.On the other hand, the characteristic vibrations in the low wavenumber region (<700 cm −1 ) largely cover the ones that impact on the metal center [16,17].The vibration of Mg-O accounts for the vibration located at 591 cm −1 , proving the formation of the metal organic framework.Mg-MOF-74 is composed of Mg(II) ions and DOBDC, and forms a hexagonal honeycomb structure.In this structure, each Mg(II) ion is bonded to five oxygen atoms: three with carboxylate groups and two with hydroxy groups, leaving an open metal site.This open metal site provides a beneficial sorption site for the various guest molecules.After loading GO for the Mg-MOF-74@GO(2:1) composite, further investigation sees the vibration peak of Mg-O changed from 591 cm −1 to 587 cm −1 , and the peak of -O-C-O at 1417 cm −1 belonging to benzene ring simultaneously shift to 1420 cm −1 .The vibration deviations of the benzene ring and the coordination bond between the metal and O prove the successful loading of GO [17].
The most prominent Raman characteristics displayed by the Raman results (Figure 2b) are the D and G bands in GO.The D band stems from the breathing vibration of the sp 2 carbon atom in the ring, whereas the G band is derived from the stretching vibration of the sp 2 atom pair in the carbon chain and the ring [17].A D band will appear once the degree of disorder for GO increases.The G band at 1594 cm −1 and the D band at 1354 cm −1 can be clearly seen, proving that GO was in a disordered state.The ratio of D band in-
Fourier Transform Infrared (FTIR) and Raman Spectroscopy
As a matter of fact, FTIR spectra (Figure 2a) can effectively detect the composition of the sample.The FTIR spectrum of Mg-MOF-74 can be divided into two different regions.On the one hand, most of the activation modes above the 700 cm −1 region belong to the organic ligands [17].In the spectrum of Mg-MOF-74, the peak at 1574 cm −1 is ascribed to the vibration of the C=C bond in the benzene ring.The peak at 1417 cm −1 is associated with -O-C-O in the carbonyl group and the peak at 1208 cm −1 conforms to the C-O vibration.On the other hand, the characteristic vibrations in the low wavenumber region (<700 cm −1 ) largely cover the ones that impact on the metal center [16,17].The vibration of Mg-O accounts for the vibration located at 591 cm −1 , proving the formation of the metal organic framework.Mg-MOF-74 is composed of Mg(II) ions and DOBDC, and forms a hexagonal honeycomb structure.In this structure, each Mg(II) ion is bonded to five oxygen atoms: three with carboxylate groups and two with hydroxy groups, leaving an open metal site.This open metal site provides a beneficial sorption site for the various guest molecules.After loading GO for the Mg-MOF-74@GO(2:1) composite, further investigation sees the vibration peak of Mg-O changed from 591 cm −1 to 587 cm −1 , and the peak of -O-C-O at 1417 cm −1 belonging to benzene ring simultaneously shift to 1420 cm −1 .The vibration deviations of the benzene ring and the coordination bond between the metal and O prove the successful loading of GO [17].
The most prominent Raman characteristics displayed by the Raman results (Figure 2b) are the D and G bands in GO.The D band stems from the breathing vibration of the sp 2 carbon atom in the ring, whereas the G band is derived from the stretching vibration of the sp 2 atom pair in the carbon chain and the ring [17].A D band will appear once the degree of disorder for GO increases.The G band at 1594 cm −1 and the D band at 1354 cm −1 can be clearly seen, proving that GO was in a disordered state.The ratio of D band intensity to G band intensity, I D /I G , represents the defect degree of graphene [17].
Here, the I D /I G value of GO up to 0.89 indicates that GO was oxidized and that structural defects occurred.Likewise, three Mg-MOF-74@GO composites maintained the D and G bands of GO.Table 1 compares that the I D /I G values of the Mg-MOF-74@GO composites which all increased, indicating that MOF was successfully modified with different contents of GO.However, there are some differences between the Mg-MOF-74@GO composite and GO.New bands appear at 1507 cm −1 and 1462 cm −1 with the coexistence of Mg-MOF-74 and GO in the Mg-MOF-74@GO composites.The D+D' band was only observed in samples with significant amounts of defects, a band observed in all graphene composites accentuates in Mg-MOF-74@GO composites.The Mg-MOF-74@GO composites show a higher graphene stacking as observed in the growth of the ratio I 2D /I G, which is proportional to the number of graphene layers.
tensity to G band intensity, ID/IG, represents the defect degree of graphene [17].Here, the ID/IG value of GO up to 0.89 indicates that GO was oxidized and that structural defects occurred.Likewise, three Mg-MOF-74@GO composites maintained the D and G bands of GO.Table 1 compares that the ID/IG values of the Mg-MOF-74@GO composites which all increased, indicating that MOF was successfully modified with different contents of GO.However, there are some differences between the Mg-MOF-74@GO composite and GO.New bands appear at 1507 cm −1 and 1462 cm −1 with the coexistence of Mg-MOF-74 and GO in the Mg-MOF-74@GO composites.The D+D' band was only observed in samples with significant amounts of defects, a band observed in all graphene composites accentuates in Mg-MOF-74@GO composites.The Mg-MOF-74@GO composites show a higher graphene stacking as observed in the growth of the ratio I2D/IG, which is proportional to the number of graphene layers.
Thermal Analyses
Figure 3a exhibits the TGA-DTG curves of the Mg-MOF-74, GO, and Mg-MOF-74@GO composites.The thermal decomposition behavior of the Mg-MOF-74@GO composites is similar to that of the pristine Mg-MOF-74.The thermal decomposition of Mg-MOF-74 can be divided into two stages: the temperature ranges from 110 °C to 150 °C, and those from 380 °C to 420 °C.The first mass loss of about 10% occurs in the range from 110 °C to 150 °C and can be attributed to desorption of the H2O solvent trapped in the micropores of Mg-MOF-74 [18,19].The second mass loss of about 30% can be ascribed to the decomposition of the framework structure.This leads to the
Thermal Analyses
Figure 3a exhibits the TGA-DTG curves of the Mg-MOF-74, GO, and Mg-MOF-74@GO composites.The thermal decomposition behavior of the Mg-MOF-74@GO composites is similar to that of the pristine Mg-MOF-74.The thermal decomposition of Mg-MOF-74 can be divided into two stages: the temperature ranges from 110 • C to 150 • C, and those from 380 • C to 420 • C. The first mass loss of about 10% occurs in the range from 110 • C to 150 • C and can be attributed to desorption of the H 2 O solvent trapped in the micropores of Mg-MOF-74 [18,19].The second mass loss of about 30% can be ascribed to the decomposition of the framework structure.This leads to the production of carbon-containing gases like CO, CO 2 , and C x H y hydrocarbon mixture, as well as a small amount of H 2 [19].For the thermal degradation behaviors of the Mg-MOF-74@GO composites seen from the DTG curves, the degradation process displays three maxima of decomposition.The initial decomposition arises as the sample gains the 5% weight loss.The 10% weight loss (T 10 ) happens when the temperature of maximum decomposition rate acts as the maximum signal in the DTG curves.The temperatures of 5% (T 5 ) and 10% (T 10 ) weight losses (Table 2) illustrate that incorporating GO into Mg-MOF-74 can enhance the decomposition temperature (T d ) of the composites.Moreover, an increasing content of GO in the Mg-MOF-74@GO composites can give rise to a continuous increase in T d .This finding originated from the fine dispersion and hydrogen bonding interaction between Mg-MOF-74 and GO [18,19].The thermal stability of the Mg-MOF-74@GO composites is higher than that of Mg-MOF-74 during the first mass loss stage of Mg-MOF-74@GO composites.This illustrates that the wrapping of GO weakens the decomposition process.
production of carbon-containing gases like CO, CO2, and CxHy hydrocarbon mixture, as well as a small amount of H2 [19].For the thermal degradation behaviors of the Mg-MOF-74@GO composites seen from the DTG curves, the degradation process displays three maxima of decomposition.The initial decomposition arises as the sample gains the 5% weight loss.The 10% weight loss (T10) happens when the temperature of maximum decomposition rate acts as the maximum signal in the DTG curves.The temperatures of 5% (T5) and 10% (T10) weight losses (Table 2) illustrate that incorporating GO into Mg-MOF-74 can enhance the decomposition temperature (Td) of the composites.Moreover, an increasing content of GO in the Mg-MOF-74@GO composites can give rise to a continuous increase in Td.This finding originated from the fine dispersion and hydrogen bonding interaction between Mg-MOF-74 and GO [18,19].The thermal stability of the Mg-MOF-74@GO composites is higher than that of Mg-MOF-74 during the first mass loss stage of Mg-MOF-74@GO composites.This illustrates that the wrapping of GO weakens the decomposition process.DSC analyses (Figure 3b) were implemented to investigate the Tg of the Mg-MOF-74@GO composites.An increase in Tg could be owed to an effective attachment of Mg-MOF-74 to GO sheets, which constrains the segmental motion of the Mg-MOF-74 chains by hydrogen bonding and electrostatic attraction [19].The Tg value of 381.9 °C for Mg-MOF-74 is consistent with that reported in the literature [19].For the Mg-MOF-74@GO composites, a higher Tg is observed.Moreover, this value increased from 391.2 to 395.2 °C by loading more weight ratio between Mg-MOF-74 and GO.Such high Tg values describe the high affinity between Mg-MOF-74 and GO in account of the high compatibility and adhesion between their two phases, as previously observed.The thermograms of all the Mg-MOF-74@GO composites demonstrate a decrease in the intensity of peak corresponding to GO.The intensity of the sharp exothermic peak for GO (5.305 mW/mg) separately decreased to 0.1456 mW/mg, 0.2833 mW/mg, and 0.4343 mW/mg for the Mg-MOF-74@GO(2:1), Mg-MOF-74@GO(4:1), and Mg-MOF-74@GO(5:1) composites.In addition, the DSC of composites reveals that two endothermic peaks corresponding to Mg-MOF-74 and GO are shifted from 127.3 °C to 136.8 °C; and from 414.6 °C to 422.7 °C.The above observations prove the interaction between Mg-MOF-74 and GO. 1 Decomposition temperature at 5% and 10% weight loss (T 5 and T 10 ) recorded on DTG, respectively. 2Glasstransition temperature (T g ) measured by DSC. 3 Not observed.
DSC analyses (Figure 3b) were implemented to investigate the T g of the Mg-MOF-74@GO composites.An increase in T g could be owed to an effective attachment of Mg-MOF-74 to GO sheets, which constrains the segmental motion of the Mg-MOF-74 chains by hydrogen bonding and electrostatic attraction [19].The T g value of 381.9 • C for Mg-MOF-74 is consistent with that reported in the literature [19].For the Mg-MOF-74@GO composites, a higher T g is observed.Moreover, this value increased from 391.2 to 395.2 • C by loading more weight ratio between Mg-MOF-74 and GO.Such high T g values describe the high affinity between Mg-MOF-74 and GO in account of the high compatibility and adhesion between their two phases, as previously observed.The thermograms of all the Mg-MOF-74@GO composites demonstrate a decrease in the intensity of peak corresponding to GO.The intensity of the sharp exothermic peak for GO (5.305 mW/mg) separately decreased to 0.1456 mW/mg, 0.2833 mW/mg, and 0.4343 mW/mg for the Mg-MOF-74@GO(2:1), Mg-MOF-74@GO(4:1), and Mg-MOF-74@GO(5:1) composites.In addition, the DSC of composites reveals that two endothermic peaks corresponding to Mg-MOF-74 and GO are shifted from 127.3 • C to 136.8 • C; and from 414.6 • C to 422.7 • C. The above observations prove the interaction between Mg-MOF-74 and GO.
The high T d and T g of the Mg-MOF-74@GO composites particularly exhibits their applications for the long-term stability of device operation.The XRD, FTIR spectroscopy, and DSC studies were performed on the Mg-MOF-74@GO composites to identity the possible interaction between Mg-MOF-74 and GO and to confirm the formation of Mg-MOF-74@GO composites.
Cross-Sectional and Morphological Characterization
The cross-sectional profiles of the Mg-MOF-74@GO composite films were studied by SEM (Figure 4a-c).The Mg-MOF-74@GO composite films were synthesized using a solution-casting method, whose thickness was roughly 100 nm.TEM images with various Mg-MOF-74@GO loadings can be observed in Figure 4d-f.It can be observed that the dispersion of the Mg-MOF-74@GO composite looks like a lotus has a layered structure.
The high Td and Tg of the Mg-MOF-74@GO composites particularly exhibits their applications for the long-term stability of device operation.The XRD, FTIR spectroscopy, and DSC studies were performed on the Mg-MOF-74@GO composites to identity the possible interaction between Mg-MOF-74 and GO and to confirm the formation of Mg-MOF-74@GO composites.
Multi-Bit Memristic Behaviors of MOF-Based Memory Devices
In this paper, the Mg-MOF-74@GO composite was first utilized to fabricate a novel Ni/Mg-MOF-74@GO/ITO device (Figure 5) that displays multi-bit memristic characteristics, whose architecture is versatile and compatible with the current parallelism demands of integration.The MOF-based film was prepared by the solution-processable method on the ITO substrate without surface modification and complex equipment.
Multi-Bit Memristic Behaviors of MOF-Based Memory Devices
In this paper, the Mg-MOF-74@GO composite was first utilized to fabricate a novel Ni/Mg-MOF-74@GO/ITO device (Figure 5) that displays multi-bit memristic characteristics, whose architecture is versatile and compatible with the current parallelism demands of integration.The MOF-based film was prepared by the solution-processable method on the ITO substrate without surface modification and complex equipment.To explore the memristic behavior of the structure Ni/Mg-MOF-74@GO/ITO, the I-V curves of Ni/Mg-MOF-74@GO(5:1)/ITO (Figure 6a) were swept with the compliance current of ICC = 0.1 A. Originally, the device was at a high resistance state (HRS or OFF state).As the bias voltage (applied to TE) swept negatively, the current suddenly increased at the SET voltage VSET1 = −0.71V, indicating that the device switched to an intermediate resistance state (IRS or ON1 state).Then, the device was kept at a low resistance state (LRS or ON2 state) when swept, until the voltage was as large as VSET2 = −1.49V.As the voltage swept in the opposite direction, the device returned from LRS to IRS to HRS when the bias voltage separately reached 5.14 V and 5.57 V, denoted as the RESET voltage (VRESET1 and VRESET2).This observation demonstrates that the device exhibits ternary memristic behaviors.Moreover, the ratio of its memristance in HRS, IRS, and LRS (RHRS:RIRS:RLRS) is roughly 10 3 :10 2 :1.When the device suffered from one hundred of times for alternative cycle sweepings (Figure 6b), the central value of VSET1, VSET2, VRESET1, and VRESET2 (Figure 6c) was −0.85 V, −1.75 V, 4 V, and 5.25 V, respectively.As for the statistics from the results of repeated tests, the cycle-to-cycle distribution (Figure 6d) of RHRS, RIRS, and RLRS was relatively narrow.Moreover, the retention time of the tristable states (Figure 6e) was over 10 4 s at an absolute readout voltage (VREAD) of 0.1 V.All these results reveal that the fabricated devices have the advantages of superior stability and reliability.For comparison, the memory device based on pristine Mg-MOF-74 was fabricated but it exhibits no memristic behaviors.
In order to better understand the underlying physical mechanism, the different loading of Mg-MOF-74 to GO (4:1 and 2:1) was further explored.The electrical characteristics of the MOF-based memory device show that the ternary memristic behavior can be modulated with the presence of GO.Accordingly, the electronic properties of Ni/Mg-MOF-74@GO(4:1)/ITO (Figure 7a) were investigated.With the increased voltage bias from 0 to −6 V, the current separately jumps at VSET1 = −0.57V and VSET2 = −0.94V, indicating the switching from HRS to IRS to LRS.Opposite to that bias direction, two sudden decreasing currents appear at VRESET1 = 3.45 V and VRESET2 = 3.9 V.Under the cycle-to-cycle sweeping mode of the endurance properties (Figure 7b), the average values of VSET1, VSET2, VRESET1, and VRESET2 (Figure 7c) are −0.82V, −1.70 V, 4.48 V, and 5.30 V, re- To explore the memristic behavior of the structure Ni/Mg-MOF-74@GO/ITO, the I-V curves of Ni/Mg-MOF-74@GO(5:1)/ITO (Figure 6a) were swept with the compliance current of I CC = 0.1 A. Originally, the device was at a high resistance state (HRS or OFF state).As the bias voltage (applied to TE) swept negatively, the current suddenly increased at the SET voltage V SET1 = −0.71V, indicating that the device switched to an intermediate resistance state (IRS or ON1 state).Then, the device was kept at a low resistance state (LRS or ON2 state) when swept, until the voltage was as large as V SET2 = −1.49V.As the voltage swept in the opposite direction, the device returned from LRS to IRS to HRS when the bias voltage separately reached 5.14 V and 5.57 V, denoted as the RESET voltage (V RESET1 and V RESET2 ).This observation demonstrates that the device exhibits ternary memristic behaviors.Moreover, the ratio of its memristance in HRS, IRS, and LRS (R HRS :R IRS :R LRS ) is roughly 10 3 :10 2 :1.When the device suffered from one hundred of times for alternative cycle sweepings (Figure 6b), the central value of V SET1 , V SET2 , V RESET1 , and V RESET2 (Figure 6c) was −0.85 V, −1.75 V, 4 V, and 5.25 V, respectively.As for the statistics from the results of repeated tests, the cycle-to-cycle distribution (Figure 6d) of R HRS , R IRS , and R LRS was relatively narrow.Moreover, the retention time of the tristable states (Figure 6e) was over 10 4 s at an absolute readout voltage (V READ ) of 0.1 V.All these results reveal that the fabricated devices have the advantages of superior stability and reliability.For comparison, the memory device based on pristine Mg-MOF-74 was fabricated but it exhibits no memristic behaviors.
In order to better understand the underlying physical mechanism, the different loading of Mg-MOF-74 to GO (4:1 and 2:1) was further explored.The electrical characteristics of the MOF-based memory device show that the ternary memristic behavior can be modulated with the presence of GO.Accordingly, the electronic properties of Ni/Mg-MOF-74@GO(4:1)/ITO (Figure 7a) were investigated.With the increased voltage bias from 0 to −6 V, the current separately jumps at V SET1 = −0.57V and V SET2 = −0.94V, indicating the switching from HRS to IRS to LRS.Opposite to that bias direction, two sudden decreasing currents appear at V RESET1 = 3.45 V and V RESET2 = 3.9 V.Under the cycle-to-cycle sweeping mode of the endurance properties (Figure 7b), the average values of V SET1 , V SET2 , V RESET1 , and V RESET2 (Figure 7c) are −0.82V, −1.70 V, 4.48 V, and 5.30 V, respectively, and R HRS , R IRS , and R LRS (Figure 7d) range from 1.15 kΩ to 9.07 kΩ; from 0.061 kΩ to 1.76 kΩ; and from 30.56 Ω to 150.83 Ω, respectively.The retention ability (Figure 7e) was also investigated, in which the voltage bias of 0.1 V was carefully selected to read R HRS , R IRS , and R LRS .According to the results, the Mg-MOF-74-based device with excellent memory stability and reliability exhibited almost no fluctuation after either 10 4 s or 100 alternative cycle sweeping.
spectively, and RHRS, RIRS, and RLRS (Figure 7d) range from 1.15 kΩ to 9.07 kΩ; from 0.061 kΩ to 1.76 kΩ; and from 30.56 Ω to 150.83 Ω, respectively.The retention ability (Figure 7e) was also investigated, in which the voltage bias of 0.1 V was carefully selected to read RHRS, RIRS, and RLRS.According to the results, the Mg-MOF-74-based device with excellent memory stability and reliability exhibited almost no fluctuation after either 10 4 s or 100 alternative cycle sweeping.The multi-bit data storage performance of the as-fabricated memory device Ni/Mg-MOF-74@GO(2:1)/ITO (Figure 8a) demonstrates that the multilevel reversible resistance level can be implemented.The device characteristics are reasonably promising for a prototype: setting from HRS to IRS to LRS occurred at −2.44 V and −3.25 V, resetting from LRS to HRS at 3.57 V, and a RHRS:RIRS:RLRS ratio of 12:2:1.Figure 8b shows that the resistance values can be maintained well during a 100-cycle scanning.Investigation of the cycle-to-cycle distribution (Figure 8c,d), the SET and RESET voltage, as well as the As summarized in Tables 3 and 4, the devices with different chemical constituents of Mg-MOF-74 and GO endow a series of ternary memory behaviors.The statistical analysis of the SET and RESET voltages was assessed, as well as the memristance RHRS, RIRS, and RLRS for the Mg-MOF-74-based memory devices operated at ambient temperature.With the incremental increase in content of GO from 16.7wt% to 33.3wt%, RHRS and RIRS decline gradually while RLRS is well maintained, indicating that LRS is dominated by non-metallic or metallic CPs [9].To verify the composition of CPs, RLRS was examined at different temperatures (Figure 9), which were measured under 0.1 V. RLRS was tested under different temperatures ranging from 20 °C to 80 °C.Based on the data analyses, the temperature-independent RLRS can be verified, which may contradict the typical characteristic of the metallic CPs.As summarized in Tables 3 and 4, the devices with different chemical constituents of Mg-MOF-74 and GO endow a series of ternary memory behaviors.The statistical analysis of the SET and RESET voltages was assessed, as well as the memristance R HRS , R IRS , and R LRS for the Mg-MOF-74-based memory devices operated at ambient temperature.With the incremental increase in content of GO from 16.7wt% to 33.3wt%, R HRS and R IRS decline gradually while R LRS is well maintained, indicating that LRS is dominated by non-metallic or metallic CPs [9].To verify the composition of CPs, R LRS was examined at different temperatures (Figure 9), which were measured under 0.1 V. R LRS was tested under different temperatures ranging from 20 • C to 80 • C. Based on the data analyses, the temperatureindependent R LRS can be verified, which may contradict the typical characteristic of the metallic CPs.
Conclusions
In summary, Mg-MOF-74 was first employed to construct a novel RRAM device.To take into account the SCLC model obeying and the temperature-independent R LRS , the formation and rupture of metallic CPs may be excluded as the memristic mechanism.In this work, the ternary memristic behavior switching from HRS to IRS to LRS is attributed to the charge trapping assisted hopping, generating the multi-bit information storage state achieved during the SET and RESET process.The statistics-based analysis of the Mg-MOF-74@GO-based memory devices is applicable for multi-bit data storage with excellent retention and endurance properties.This work will be instructive in the study of Mg-MOF-74@GO-based multi-bit memory processes, which is expected to provide more opportunities in the construction of functional synapses and smart devices.
Figure 5 .
Figure 5. Schematic diagram of the device Ni/Mg-MOF-74@GO/ITO and structure figures of Mg-MOF-74 and GO.
Figure 5 .
Figure 5. Schematic diagram of the device Ni/Mg-MOF-74@GO/ITO and structure figures of Mg-MOF-74 and GO.
Table 1 .
Data for D, G, and 2D bands of GO and Mg-MOF-74@GO composites.
Table 1 .
Data for D, G, and 2D bands of GO and Mg-MOF-74@GO composites. | 2023-10-14T15:59:10.656Z | 2023-10-01T00:00:00.000 | {
"year": 2023,
"sha1": "bfa2613f6cc794ae721c97ddce2520087ea193f9",
"oa_license": "CCBY",
"oa_url": "https://www.mdpi.com/2079-4991/13/20/2736/pdf?version=1696917084",
"oa_status": "CLOSED",
"pdf_src": "PubMedCentral",
"pdf_hash": "47ee9458255b67c1d734960d7c6615f301a29f85",
"s2fieldsofstudy": [
"Materials Science"
],
"extfieldsofstudy": []
} |
265343295 | pes2o/s2orc | v3-fos-license | Treatment-seeking behavior for malaria among communities in Indonesia: A systematic review
Indonesia stands as one of the nine malaria-endemic countries in Southeast Asia with a total of 443,530 cases in 2022. Eastern Indonesia is listed as an area with high malaria endemicity and the Indonesian government has set a target of eliminating malaria by 2030. From 2010 to 2014, the number of malaria cases decreased but stagnated until 2020 and have continued to increase. Stagnation may occur as a result of many non-medical treatment-seeking behaviors. The aim of this systematic review was to provide a summary and overview of malaria treatment-seeking behavior among communities in several regions in Indonesia. The searches were conducted through four databases (Cochrane, PubMed, Google Scholar, and ScienceDirect) using medical subject headings (MeSH) “treatment-seeking behavior” OR “health-seeking behavior” AND “malaria” AND “Indonesia”. This systematic review was limited to studies conducted in Indonesia that were published between 2013 and 2023 using either a quantitative or qualitative approach. Out of 2831 studies, a total of thirteen studies were included. The pattern of seeking malaria treatment varied between doing nothing or no action, self-treatment (purchasing drugs at pharmacies and consuming leftover medicines), traditional medicine, and medical treatment (public health facilities or malaria control clinics). Those behaviors are attributed to education level, socioeconomic level, occupation, distance from home to health facilities, geographical conditions, and people’s perceptions of malaria and antimalarial medicines. There is still a range of malaria treatment-seeking behavior outside of recommended medical treatment in communities in several regions in Indonesia. The phenomenon of medical pluralism and syncretism requires approaches from various sectors in order to achieve a malaria-free Indonesia by 2030.
Introduction
Malaria is a parasitic infection of the genus Plasmodium sp. which is transmitted by female Anopheles mosquitoes [1,2].Based on data from the World Health Organization (WHO), Indonesia stands as one of nine malaria-endemic countries in Southeast Asia and accounts for 15.6% of cases reported from the entire region with a mortality rate of 22% [3].The prevalence of malaria in Indonesia in 2022 was 443,530 cases.The provinces of Papua, East Nusa Tenggara, West Papua, North Sumatra, Maluku and North Maluku are recorded as the provinces with the highest incidence of malaria [4].The annual parasite incidence (API) in Indonesia in 2022 was
Original Article
1.6 with a national target of API <1 [4].Between 2010-2014, the number of malaria cases decreased before finally stagnating [4].The Indonesian government, in collaboration with WHO, has targeted the elimination of malaria in Indonesia no later than 2030.Around 130 million people in Indonesia live in high-risk areas, but the geographic distribution of transmission is very heterogeneous [5].
In the human-mosquito life cycle, six species of the malaria parasite (Plasmodium falciparum, P. ovale walickeri, P. ovale wallickeri curtisi, P. vivax, P. malariae, and P. knowlesi) undergo ten or more morphological states and replicate from one to more than 10,000 cells.In human hosts, only a small number of morphological stages can cause humans to become infected and get symptoms such as fever, anemia, malaise, chills, myalgia, arthralgia, and others [6].Treatment of malaria includes combination therapy targeting the morphological stages of both erythrocytes and hepatocytes.Schizonticidal drugs, supportive care, hospitalization of at-risk patients, and 24-hour monitoring of patients receiving antimalarial treatment are needed to achieve recovery and avoid complications of this disease [7].The complications of malaria include cerebral malaria, anemia, respiratory failure, kidney failure, and even death [8].However, not all individuals in societies apply those recommended medical treatments in malaria treatmentseeking behavior.
Malaria treatment in Indonesia is sought through various approaches, including doing nothing or "no action", self-treatment, and traditional medicine.The stagnation of malaria cases in Indonesia can be attributed to such treatment-seeking behaviors, which is not limited to recommended medical treatments.Public perception regarding the etiologies of malaria and antimalarial drugs is closely related to the decisions made in selecting a treatment for malaria.Inappropriate treatments could result in life-threatening complications, anti-malarial resistance, and relapse.The aim of this study was to provide a summary and overview of malaria treatmentseeking behavior among communities in several regions in Indonesia.
Study setting and search strategy
A systematic review was conducted in accordance with the preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidelines.The systematic searches were conducted through four databases (Cochrane, PubMed, Google Scholar, and ScienceDirect) with medical subject headings (MeSH) "treatment-seeking behavior" OR "health-seeking behavior" AND "malaria" AND "Indonesia".
Eligibility criteria
This systematic review included studies within the last ten years (2013-2023) with both quantitative and qualitative approaches.The search was conducted in English and Indonesian as of August 8, 2023.This systematic review was limited to studies conducted in Indonesia that described the pattern of seeking malaria treatment in the community and related factors such as education level, socioeconomic level, occupation, distance from home to health facilities, geographical conditions, direct costs, and perceptions of disease.Study was excluded if did not focus on the pattern of treatment-seeking for malaria and the research was not conducted in Indonesia.
Evaluation of methodological quality
The quality assessment was carried out with the criteria of JBI Critical Appraisal Tool for quantitative studies [9] while quality of qualitative studies was assessed using the Critical Appraisal Skills Program (CASP) criteria [10].
Data extraction
Studies included were categorized based on the research method.Data extraction was carried out by two researchers with the following variables: title, author, year of publication, study location, number of samples, subject characteristics, conclusions from both quantitative and qualitative components.
Data synthesis
The data were synthesized qualitatively.The findings from the systematic reviews were presented in summary of malaria treatment-seeking behavior and factors associated.Data extracted from the studies were categorized into four distinct groups: no action, self-treatment, traditional medicine, and medical treatment.No action refers to the absence of any drug consumption despite the presence of malaria symptoms.Self-treatment encompasses the consumption of drugs acquired from pharmacies or local drug stores without requiring a prescription.Traditional medicine includes remedies provided by local traditional practitioners without formal medical qualifications.Lastly, medical treatment comprises treatments obtained from official healthcare facilities.Factors related to treatment decisions such as public perception of the causes of malaria and the perception of antimalarial drugs were also presented.
Results of study selection
The literature searches yielded a total of 2831 studies.After screening of titles and abstracts, 1,051 studies were excluded due to duplication, conducted outside Indonesia, review articles, not being written in Indonesian or English, and had no full articles.The PRISMA flow diagram is presented in Figure 1.A total of 23 studies were considered potentially relevant and underwent thorough assessment.Additional nine studies were excluded as they only discussed the names of drugs or herbs that were used without addressing treatment-seeking behavior in the community.Furthermore, one study was excluded as it did not clarify the definitions and categorize treatment-seeking behavior into "good" and "not good" categories.Following the assessment process, a total of 13 studies were included in this systematic review (Table 1).
Original Article
The overview of the characteristics of the included studies such as study year, research design, sample size, and factors associated with the patterns of malaria treatment-seeking behavior within the studied communities are presented in Table 1.
Study quality evaluation
The quality of the quantitative studies and qualitative studies were assessed using JBI Critical Appraisal Tool and Critical Appraisal Skills Program (CASP) checklist.The results are presented in Table 2 Were the study subjects and the setting described in detail? 10 (100) 0 (0) 0 (0) 0 (0) Was the exposure measured in a valid and reliable way?
No action
A study reported that a total of 1,104 people from 825 households in Mimika Regency, Papua who reported having a fever in the last 30 days had positive laboratory test results for malaria [13].Among them, 270 individuals (24%) chose not to consume any medications [13].According to the study conducted by Sidik et al., in 2022 at Muari Village, South Manokwari, 4 out of 8 village members opted to take no action when they experienced symptoms of malaria.The same
Original Article
symptoms of malaria, headaches, were reported by both studies, indicating that the illness was not severe and would resolve on its own with rest [19].
Self-treatment
The majority of urban residents choose to receive treatment from pharmacies (40.5%) and drugstores/stalls (30.2%) rather than formal health services (9.5%) [17].In contrast, rural communities prefer to receive treatment from stalls (44.1%) [17].Studies in East Kupang District reported that individuals chose to buy medicine from the nearest stall after noticing the initial symptoms (78.6%) and only went to a health facility if their symptoms worsened [12,13].The people of Muari Village, South Manokwari believe that treatment can be performed individually based on experience [19].Fever and headache were considered as symptoms indicating malaria, so they decided to buy paracetamol at the stall [11,14].Other studies found that people choose to purchase quinine or primaquine from pharmacies without a doctor's prescription or consume quinine leftover at home for previous malaria treatment [11,15].This is influenced by the distance between home and pharmacies, which is more easily accessible than the distance between home and health facilities [10, [16][17][18].The choice to seek treatment can also be affected by difficult geographical conditions and long distances [10, [16][17][18].Closed communities, who live in mountainous areas surrounded by forests and have not experienced environmental change or deforestation, tend not to seek treatment from neither health facilities nor health personnel [16,21].
Traditional medicine
A study conducted in East Kupang [11] reported that of 178 individuals with a history of malaria in the past year, 16.4% of those chose traditional treatment for malaria.The majority of these individuals were women with low educational levels that did not surpass elementary schools [11].They tend to perceive that the traditional treatments are safe and natural with fewer side effects.
A study found that individuals with low education levels are four times more likely to choose nonmedical treatment than medical treatment [18].A study found that group of married men who smoke, junior high school graduates, and work as farmers prefer traditional medicine to cure malaria [14].In addition, high transportation costs, difficulty of the terrain, and absence of access roads to health facilities also influence individual decisions to prefer traditional medicine as a treatment for malaria [16,19].A study in Namrole, South Buru, Maluku, found that closed community groups were three times as likely to use traditional medicine as modern medicine [16].Certain community groups in Aceh believe that malaria is a manifestation of malicious incantations contracted within work environments [23].Consequently, traditional healers, commonly referred to as 'battra', are perceived to possess the capability to effectively remedy this ailment through the utilization of written prayers and various religious objects [23].That study also found that the selection of traditional healers is community-driven, grounded upon individual qualities.In addition, community members exhibit a heightened sense of security with battra due to their perceived heightened familiarity with patients with congruent cultural background and beliefs.This can be attributed to the intimate bond between the traditional healer and the community, stemming from their shared cultural backgrounds and congruent belief systems.This is in contrast to the majority of Papuan people who use traditional medicine as an alternative treatment and not as the primary choice for malaria [10,15,20].
Medical treatment
A study conducted in 2022 across five islands in East Nusa Tenggara showed that 99.8% people chose to seek healthcare facilities after experiencing malaria symptoms [18].Among the population, 46.8% sought treatment at health care facilities within 24 hours of symptoms and 53.2% of the population sought treatment 24 hours after symptoms appeared.Deliberate delay in seeking treatment after 24 hours was also observed at the Mimika's society in Papua, as they believed that an early test might result in a negative result which can lead them to leave health care facilities without receiving any antimalarial treatment [20].The individuals who tend to postpone the anti-malaria treatment are dominated by individuals who have never been collegeeducated and have lower socioeconomic status [12,16].These people were occupied in the farming section, and the rest reside over three kilometers from the nearest health care facilities [12,18].In
Original Article
contrast to two studies found that individuals in Mimika, Papua, prefer to seek medical treatment after more than 48 hours of feeling a fever [10,20].The majority of Papuans and migrant workers in Kalimantan, Riau and Bengkulu preferred public health centers (Puskesmas) as the first choice in seeking medical treatment [15,21].Low socioeconomic level groups prefer the public sector compared to the private sector [15].The direct cost for treatment at public health facilities is estimated at around Rp61,000.00 while at clinics it is around Rp252,000.00 and the average expenditure for one episode of fever is Rp628,000.00[15].Around 22.7% of the population in Mimika, Papua, opt for malaria control clinics [20].A similar pattern was found among forestry workers in Aceh, where malaria control clinics were the first choice for treatment [23].This preference is due to the fact that malaria control clinics test all individuals, whether it is symptomatic or not, are less crowded, and have a higher likelihood of obtaining antimalarial medication [23].Additionally, the operating hours of malaria control clinics were much longer so that it did not interfere with someone's daily activities [14,15].
Public perception of malaria and its treatment
The Tetun ethnic community at South Halmahera in North Maluku thought that the main causes of malaria disease are sweet food and beverages such as sugarcane, ripe bananas, young coconut water, and young corn [22].In addition, they tend to think that working under intense sunlight or in damp and cold places can induce malaria.This idea is associated with a disturbance in the body's heat-cold equilibrium.Hence, the approach to malaria treatment involves avoiding its underlying causes such as reducing the consumption of sweet foods and beverages alongside the consumption of bitter-tasting herbal remedies like Carica papaya leaves.According to the Tetun community, malaria treatment can also involve utilizing two types of medicinal plants.Warm plants are considered to raise body temperature and stimulate sweat production (e.g., Acarus calamaus), while cool plants work to absorb heat from the body, resulting in a return to normal body temperature (e.g., Drynaria quercifolia).Recovery is marked by sweating, normal body temperature, increased appetite, no chills, and being able to get up to go to work [22].A qualitative study conducted in Topoyo village, West Sulawesi found that the decision making in seeking treatment was influenced by those closest to the patient [21].People who were related or friends with sando/tomanarang (traditional healer) tend to choose traditional medicine and use medical treatment as an alternative, while others who were related or friends with doctors, nurses, midwives, or other health workers tend to choose medical treatment [21].Moreover, people also have their own perceptions of antimalarials.According to the people of Mimika, Papua, administering the antimalarial combination of dihydroartemisinin-piperaquine or the "blue drug" is socially acceptable [20].However, giving primaquine or "chocolate medicine" is considered less effective and less important due to its size, so it is only considered as a supplement and is not consumed regularly and properly [20].
Discussion
The findings of this systematic review highlight the persistence of two prevalent patterns in malaria treatment in Indonesia: no action and self-treatment.Medications are easily accessible to the public through pharmacies and local stores [11][12][13]17].A study conducted in Ghana revealed that 57.7% of respondents adopted self-treatment practices.They acquired drugs by either using leftover medications stored at home (25%) or purchasing drugs from pharmacies (23.9%) [11].Shortages of medication in healthcare facilities, long queues, inability to afford healthcare service costs, and the freedom to purchase drugs independently are key factors driving communities toward self-treatment [11,20,23].Self-treatment practices often lead to delayed medical treatment seeking.A study identified specific groups most likely to delay medical treatment, including women, low-income individuals, married couples, and those lacking sufficient social support [24].Treatment delays can also be attributed to the distance between healthcare facilities and residences, as communities tend to postpone treatment when the nearest healthcare facility is more than three kilometers away [18].In Nigeria, the majority of the Aluu and Azikoro people prefer medical treatment for malaria [25].However, the use of traditional medicines is more prevalent among the Aluu people in rural areas compared to the Azikoro people, who have closer access to teaching hospitals [25].These findings
Original Article
underscore the critical role of accessible healthcare services in shaping malaria treatment choices within communities.
A segment of the community perceives traditional medicine as a safer treatment option [14,16].Traditional medicine, being the oldest method predating allopathic medicine, is often considered safer due to its reliance on natural materials [26].Traditional healers, known as 'battra' are chosen by the community based on their personal quality [16].Battra possess a deep understanding of their patients and share cultural backgrounds and belief systems, fostering trust [16].The high cost and limited accessibility of healthcare facilities present significant barriers to seeking medical care, leading the community to frequently turn to traditional medicine [14,16].This includes not only consultation and medication expenses but also the expenses associated with accessing healthcare facilities.Consequently, the total expenditure for medical treatment becomes impractical [27] Government subsidies for artemisinin combination therapy can help offset the high costs of healthcare services, encouraging the community to consider medical treatment as a viable option [28].
The severity of a disease is not only linked to the duration of pain but also to its intensity [21].Individuals who perceive their malaria as a serious condition are 12.5 times more likely to seek medical treatment compared to those who view it as less severe [29].Educational initiatives aimed at disseminating knowledge regarding the signs and symptoms of malaria play a pivotal role in raising awareness among rural populations.This, in turn, promotes timely treatmentseeking behaviors when malaria symptoms first appear.Delayed treatment beyond 24 hours can significantly increase the risk of severe complications, including severe malaria anemia, hospitalization, the need for blood transfusion, and higher mortality rates [18].All Plasmodium species causing infections in humans can lead to various levels of complications, some of which can be life-threatening.Furthermore, the long-term impacts on the quality of life can persist even after recovery [30].An intervention study in the Amazon region of French Guiana demonstrated that the distribution of 'Malakit', a rapid diagnostic tool for malaria, along with a regimen of artemether-lumefantrine and a single dose of primaquine, reduced malaria incidence by 42.9% in gold-mining communities [6].This highlights the importance of community education and access to effective tools in malaria prevention.
In addition to perceptions of disease severity, the community's attitude toward antimalarial medications can significantly influence behaviors related to malaria treatment.In Mimika, Papua, for example, the perception of primaquine's inefficacy has a substantial impact on malaria relapse prevention [20].Primaquine is the most widely accessible antimalarial medication in Indonesia, used to inhibit dormant Plasmodium in the liver and prevent relapses caused by P. vivax.However, suboptimal adherence to the recommended 14-day treatment regimen may limit its effectiveness [31].
The healthcare system is closely intertwined with the concept of medical pluralism, which involves the utilization of various medical modalities or traditions to achieve health and treat illnesses [33,34].This can manifest in different forms, such as the use of distinct treatment modalities, as seen in Namrole, Maluku, where a closed community prefers herbal treatments when dealing with malaria exposure [16].Additionally, it can be hierarchical, as observed in Topoyo Village, West Sulawesi, where people initially opt for self-medication at the onset of malaria symptoms, turning to medical treatment if the symptoms worsen [21].In some cases, individuals may simultaneously seek both traditional and biomedical treatments, often driven by a belief in mystical causes, a phenomenon referred to as medical syncretism in medical anthropology [34].These communities place high value on interpersonal interactions and trust in traditional healers.However, they also seek assurance through biomedical tests and diagnostic technology [33].Numerous community groups in Indonesia continue to work diligently towards seeking effective treatment through a variety of initiatives.In addition to comprehensive case management, rigorous monitoring and evaluation, and proactive malaria vector control efforts, effective communication strategies are crucial across various sectors beyond the medical field.The primary objective is to educate the populace about the significance of behavioral changes, ultimately aimed at diminishing the prevalence of malaria cases in Indonesia.
Conclusion
There is still a range of malaria treatment-seeking behavior outside of recommended medical treatment in communities in several regions in Indonesia.The pattern of seeking malaria treatment varies between doing nothing/no action, self-treatment, traditional medicine, and medical treatment.The decision making also depends on individual characteristics and perceptions on malaria and antimalarial drugs.The phenomenon of medical pluralism and syncretism requires approaches from various sectors in order to achieve a malaria-free Indonesia by 2030.
Figure 1 .
Figure 1.Preferred reporting items for systematic reviews and meta-analyses (PRISMA) flow diagram.
Table 1 .
Characteristics of the studies and factors associated with the seeking behavior for malaria Traditional medicine is favored by poorer households due to the high cost of transportation and difficult terrain • Traditional medicine is considered more natural, safer and has fewer side effects Devine et al.People chose to ignore the initial symptoms of malaria and treat them with analgesics/antipyretics.However, when the symptoms persisted, they bought chloroquine without undergoing a diagnostic test • Private clinics are considered more convenient because of their longer operating hours, efficiency, and ability to examine quickly • Public health care is thought to have higher transportation costs and only operate during working hours.
• In migration destination areas, 19.2% individuals received initial treatment at a hospital, 42.3% at a public health center, 3.8% at a midwife practice, 15.4% at a clinic, and 7.7% at a private medical practice • 34.6% individuals seek treatment after >3 days of symptoms • 26.9% individuals seek treatment after >7 days of symptoms Karyana et al. • 76% individuals opted for medical treatment (mostly people with high socioeconomic level and non-Papuan), 46% of them went to a private hospital and sought treatment after more than 2 days of symptoms • 24% individuals chose to do nothing (mostly people with low socioeconomic level) • 1 episode of fever cost 11% of the household's monthly income • The majority of individuals who choose traditional medicine are men, married, currently smoking, educated for <9 years, worked as a farmer, previously used traditional medicine, and are experiencing recurrent malaria • • Closed community groups were three times as likely to use traditional medicine as modern medicine • Traditional medicine is often associated with lack of health facilities, areas surrounded by forests and mountains, and no health personnel Original Article
Table 3 .
and Our assessments indicated that ten studies
Table 2 .
JBI Critical Appraisal checklist of the included quantitative (n=8) and mixed-methods studies (n=2) | 2023-11-22T16:36:47.718Z | 2023-11-13T00:00:00.000 | {
"year": 2023,
"sha1": "091529f18813eb017b8c5b6261cfb95e42dcf280",
"oa_license": null,
"oa_url": null,
"oa_status": null,
"pdf_src": "ScienceParsePlus",
"pdf_hash": "78912660e28dbcaaad95c1810d8bd4d3b04d5613",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": []
} |
214389602 | pes2o/s2orc | v3-fos-license | Seed production and dispersal limit treeline advance in the Pyrenees
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2019 The Authors. Journal of Vegetation Science published by John Wiley & Sons Ltd on behalf of International Association of Vegetation Science 1Institute of Botany and Landscape Ecology, University of Greifswald, Greifswald, Germany 2Department of Evolutionary Biology, Ecology and Environmental Sciences, Biodiversity Research Institute (IRBio), University of Barcelona, Barcelona, Spain 3Centre de Recerca Ecològica i Aplicacions Forestals (CREAF), Universitat Autònoma de Barcelona, Cerdanyola del Vallès, Spain
vancements in elevation may be explained by the decrease and even cessation of anthropogenic activities in the last century. In such cases, treelines are not expanding beyond their historical thermal limits, but recovering their past altitudinal distribution (Ameztegui et al., 2016;Cudlín et al., 2017). Regardless, the treeline advance, either through range expansion or through re-colonisation, may lead to changes in extant plant diversity and ecosystem function through displacement of alpine species and shifts in carbon and nutrient dynamics (Greenwood & Jump, 2014).
Production of viable seeds is the first step for a seed-based regeneration process at any treeline (Holtmeier & Broll, 2005). The next essential step is the dispersion of the seeds, which represents the first spatial demographic process that will determine the position of recruit establishment and potential range shifts (Nathan & Muller-Landau, 2000). Ultimately, any change in treeline position is thus related to successful seed production, viability, dispersal and germination, as well as seedling survival, at higher elevations (or northern latitudes). Some studies have reported a decrease in cone and seed production or seed germination with elevation and latitude (Brown et al., 2019;Juntunen & Neuvonen, 2006;Kruse et al., 2019;Šenfeldr & Maděra, 2011;Sirois, 2000), pointing to potential reproductive limitations at the treeline. These reproductive constraints seem exclusively related to changes in population density in some cases (Juntunen & Neuvonen, 2006), whereas in others they seem related to changes in the reproductive capacity of individual trees (Sirois, 2000), or a combination of both factors (Šenfeldr & Maděra, 2011). However, the reproductive ecology of treelines is still largely unknown (Körner, 2012), especially in relation to the dynamics of alpine treeline ecotones.
In this study, we aimed to address this gap by focusing on seed production, viability, dispersal, and germination dynamics of Pinus uncinata. Our main objective was to analyse whether seed production and dispersal can constrain the elevation advancement of the treeline in the Pyrenees, and to determine variations in seed germination dynamics between different provenances (i.e., elevation positions) along the elevation gradient under controlled environmental conditions. We hypothesised that: (a) viable seed production would decrease with elevation; (b) that seed rain and seed arrival would decrease along the elevation gradient; and (c) that seed germination rates and dynamics would mirror the provenance in the elevation gradient, with seeds from lower elevations (where temperatures are higher and growing seasons longer) showing higher germination success and a more extended germination period than seeds from trees growing at higher elevations.
| Study sites and species
Seed production was assessed at five sites in two protected areas of the Central Pyrenees, Catalonia, Spain. The study sites Gelada, Dellui, Son and Eixol are located in the Aigüestortes i Estany de Sant Maurici National Park, and the study site Capifonts is located in the Alt Pirineu Natural Park ( Figure 1, Table 1). A sixth study site, Amitges, was used for the seed rain experiment together with Gelada and Dellui ( Figure 1, Appendix S1). All the sites occupy NW-NE facing slopes. The anthropogenic alteration of forests in these areas has been minimal since the mid-twentieth century due to their protection status, as well as an ongoing reduction in livestock numbers and the abandonment of forest clearcutting (Domínguez Martín, 2001;Gracia, Meghelli, Comas, & Retana, 2011). Forest expansion has been noticeable from 1956 to the present, although this phenomenon has produced little displacement of the treeline in such near-natural treeline locations . The mean annual precipitation in climate stations located near the study sites for the period 2008-2014 is 1,340 mm, with a mean temperature of 9.8°C for the growing season months (JJA) and −3.6°C for the winter months (DJF) (Servei Català de Meteorologia). For detailed climate data, see Appendix S2.
The study species is Pinus uncinata, the dominant tree species in most of the treeline ecotones in the Pyrenees (Ninot et al., 2008). It forms dense subalpine forests between 1,700 m and 2,200 m a.s.l.
| Seed production along the elevation gradient
We carried out the sampling from mid-October to mid-November 2010. At each of the five study sites, we established a transect along the elevation gradient where we set four plots corresponding to different elevation positions and plant communities: dense forest at mid-elevation within the subalpine belt (DFM), dense forest at maximum elevation within the subalpine belt (DFH), scattered trees (ST), and krummholz zone (KZ) (Figure 1, Table 1). The three highest sampling plots were close to each other and encompass the treeline ecotone. At each plot we sampled 5-6 mature trees and collected 5-13 cones per tree. In total, we collected 635 cones across plots and study sites. The minimum distance between sampled trees was 50 m, and they were selected to include the local forest heterogeneity (e.g., tree size). To include variability linked to position within the canopy, cones were collected at different orientations across the canopy.
We applied a cold treatment to the collected cones. We placed them in growth chambers for one week at 4°C, one month at −18°C, and 2.5 months at temperatures ranging between −10°C and 9°C, to simulate natural conditions. We triggered cone opening with a moderate heat treatment that consisted of an initial temperature of 30°C that was gradually increased over 8 hr up to 55°C, which was maintained for 10 min. This temperature does not affect the viability of Pinus uncinata seeds (Escudero, Sanz, Pita, & Pérez-García, 1999). We mechanically opened those cones that were not completely open after the heat treatment to extract all their seeds. For each cone, we measured the maximum length (from base to apex) and assessed the number of empty seeds (without embryo) and the number and weight of fully developed seeds (hereafter full seeds).
Additional information on the assessment of cone production at two of the study sites can be found in Appendix S3.
| Germination experiment
We carried out a germination experiment to test the potential viability and germination dynamics of all full seeds under comparable environmental conditions. We used a growth chamber under con- ANADON-ROSELL Et AL. et al., 2002) and are close to those found in the subalpine belt in the most favourable period for germination in early summer (Ninot et al., 2008). We sowed all full seeds from the same cone in one pot (7.5 cm × 7.5 cm × 8 cm), or in two pots when cones had more than 50 full seeds, always at 1-cm depth. We used a sterile substrate of 1:1 perlite and vermiculite, watered weekly to saturation with deionised water and let the pots drain freely. We recorded seedling emergence and removed seedlings twice a week to avoid possible competition or inhibition. The germination trial lasted 121 days.
| Seed rain and seed dispersal along the altitudinal gradient
To explore in more detail seed production and dispersal of Pinus uncinata, we performed a seed rain experiment at three of the study sites, Gelada, Dellui and Amitges (Figure 1, Appendix S1). To account for the interannual variability in seed production, we carried out this experiment during three consecutive years. At each site we established one transect with four plots (elevation positions) along the elevation gradient. Plots corresponded to the subalpine dense forest (SDF), subalpine fragmented forest (SFF), krummholz zone (KZ), and alpine grassland (AG). Note that in Gelada and Dellui, seed rain plots do not correspond to the plots used in the seed production experiment (Appendix S1). In Amitges, due to its characteristics, the KZ and AG plots could not be established. Between 28 September and 5 October 2011 we installed six seed traps per plot at each site.
Seed traps were 1 m 2 and were made of artificial tuft grass. In each plot, the six traps were installed forming a circumference of 7-8 m in diameter. From 2012 to 2014 we collected all seeds in each trap during summer and counted them at the laboratory. Since dispersal distances of Pinus uncinata are frequently within 10 m from the parental tree (Camarero et al., 2005;Vitali et al., 2019), we established a circular area adding a supplementary radius of 10 m from where the circumference of seed traps was located. Within this circular area (seed trap area + external circumference), we counted all the trees present, we measured their diameter at breast height (DBH), and classified them as reproductive trees (DBH > 5 cm), potentially reproductive trees (non-reproducing sexually mature trees with DBH > 5 cm), non-reproductive trees (DBH < 5 cm), and dead trees.
We used these data to calculate reproductive tree density and the mean DBH of reproductive trees in the area covered by the seed traps at each study site and elevation position (see Appendix S4).
| Statistical analyses
We used linear and generalised linear mixed-effects models to analyse our data. For all the study variables, we performed model selection following the Akaike information criterion (AIC) to select the most suitable model (Zuur, Ieno, Walker, Saveliev, & Smith, 2009).
When two models did not differ in more than two AIC units, we applied a model average function (Bartoń, 2019). TA B L E 1 Altitude (m a.s.l.) and UTM coordinates (ED50, zone 31T) of each sampling plot at each study site for the seed production and germination analysis Note: For each study site, the complete toponym is given in parentheses.
Journal of Vegetation Science
ANADON-ROSELL Et AL.
We used generalised linear mixed-effects models to assess the effects of the elevation gradient on the number of seeds and number of full seeds per cone (fitted with a Poisson distribution) and on the proportion of full seeds per cone (fitted with a binomial distribution).
We used linear mixed-effects models fitted with the restricted maximum-likelihood method (REML) to test the effects of the elevation gradient on cone length and weight of full seeds. For all these variables, we used the elevation position as a fixed-effect factor and site and tree identity as random factors as a starting point for model selection. Additionally, we explored the relationship between cone length and number of full seeds per cone with Pearson correlation tests and exponential regression.
To analyse differences in the germination success (i.e., seed viability) at the end of the germination experiment (seeds germinated out of seeds sown), we used generalised linear mixed-effects models fitted with a binomial distribution. We included elevation position and site as fixed effects and tree identity as a random factor prior to model selection.
For the analysis of germination dynamics, we used meta-analytic random-effects models, which allow incorporating random factors in log-logistic models of germination data (Keshtkar, Mathiassen, Beffa, & Kudsk, 2017;Ritz, Pipper, & Streibig, 2013). For this, we carried a two-step approach. First, we fitted a three-parameter log-logistic model to the cumulative germination data according to Ritz et al. (2013): where F(t) is the cumulative seed germination at time t, d is the upper limit parameter denoting the proportion of seeds that germinated during the experiment out of the total number of seeds sown, t 50 is the time when 50% of the seeds that germinated during the experiment (d) have germinated, and b is the slope of F at time t = t 50 .
In the second step, we separately fitted the meta-analytic random-effects model to our parameters of interest. Therefore, we analysed the estimates of t 50 , b and d obtained with the event-time model (first step) with a linear mixed-effect model where tree was defined as a random factor and elevation position, site and the interaction elevation position × site were defined as fixed effects. Site was included here as a fixed effect to specifically investigate potential effects of the different provenances (i.e., elevation positions) of the seeds on the dynamics of germination. The estimated standard errors for the parameters of interest were also included in the models. We made pairwise comparisons to find significant differences between the factor levels.
We used a zero-inflated negative binomial (ZINB) model to assess the effects of the elevation gradient on seed rain per area.
We used these models to account for data overdispersion and zero inflation (their AIC values were lower than in generalised linear models fitted with Poisson and negative binomial distributions).
We used site, elevation position, study year and the interactions position × year and site × year as fixed-effect factors in the count part of the model. In the binomial part, we used the three factors without interactions following the model selection procedure.
We excluded the alpine grassland plots from all the analyses on seed rain because no seeds were found in these plots at any of the study sites and years.
We tested for normality and homoscedasticity of the residuals and applied data (sqrt-and log-) transformations when necessary to reach these assumptions in the linear regression models. We also used varIdent structure when the homogeneity of residuals was not reached (Zuur et al., 2009). When fixed terms were selected in the final model, we used post-hoc Tukey HSD tests to identify significant differences between fixed term levels.
For all statistical analyses we used r software v. 3.6 (R Core
| Seed production
The number of total seeds per cone decreased 65% along the elevation gradient from DFM to KZ. Values were significantly lower at KZ (16.4 ± 0.9) than in the other three elevation positions (DFM = 46.8 ± 1.8, DFH = 36.6 ± 1.7, ST = 33.3 ± 1.7, Tukey posthoc test p < 0.05). The number of full seeds per cone also decreased along the gradient, from 37.5 ± 1.8 full seeds per cone at DFM to 14.0 ± 0.8 at KZ (Table 3, Figure 2a), which represents a 63% reduction. Post-hoc tests showed significant differences between KZ and the forest sites (DFM and DFH, p < 0.05) but not between KZ and ST (p = 0.171). The elevation position showed no significant effect on the proportion of full seeds to empty seeds per cone and on the weight of the full seeds (mean across sites and elevation positions of 9.15 ± 0.2 mg/seed) ( Table 3).
The elevation gradient also had an effect on cone length, which decreased with increasing elevation (Table 3). Post-hoc tests showed that at the two higher elevations, KZ (3.92 ± 0.05 cm) and ST (4.35 ± 0.06 cm), cone length was significantly lower than in the two forest plots, and within these two, the lower subalpine forest DFM (4.79 ± 0.05 cm) showed longer cones than the higher subalpine forest DFH (4.43 ± 0.06 cm). Correlation analyses showed a significant exponential correlation between cone length and total full seeds per cone (no. viable seeds = exp (0.67 + 0.57 × cone length), p < 0.001 for both parameters;
| Seed germination
A mean of 65% of the sown seeds across the four assessed elevation positions germinated after 121 days (Figure 3, Appendix S6).
For the analysis of the germination success (proportion of germinated seeds at the end of the experiment) we selected two generalised linear mixed-effects models with a difference in AIC lower than two units. Elevation, study site, and their interaction were included in one model whereas only site was included in the other.
Model average only showed a significant effect of study site on the germination success (Table 4, Appendix S6), with the largest differences between Dellui and Eixol (post-hoc Tukey test run for each model, p < 0.01).
The meta-analytic random-effects models did not show consistent patterns in the germination dynamics between elevation plots across study sites for any of the estimated parameters ( Figure 3). However, when looking at the study sites independently we found significant had a lower t 50 than DFH (estimated difference of 26.6 ± 11.9 days, p = 0.025). At the study site Capifonts, the scattered trees (ST) had a higher t 50 than (KZ; estimated difference of 25.3 ± 9.7 days, p = 0.009) and higher than DFM (estimated difference 25.2 ± 11.6 days, p = 0.030). Finally, Dellui KZ had a lower t 50 than DFM (estimated difference 21.9 ± 11.2 days, p = 0.05, respectively).
| Seed rain and dispersal
Seed rain significantly decreased along the elevation gradient. The number of seeds recovered at each elevation position was significantly lower than at the elevation directly below within the elevation gradient, and it was zero at alpine grasslands (Table 5, Figure 4, p < 0.001 after Tukey post-hoc tests between each elevation plot).
The study site and the study year also had a significant influence on seed rain values.
Journal of Vegetation Science
ANADON-ROSELL Et AL.
| D ISCUSS I ON
Our study investigated the reproductive and dispersal patterns of the treeline-forming species Pinus uncinata along the treeline ecotone in the Pyrenees. Our findings on seed production, viability, and dispersal provide evidence of reproduction as a limiting factor for treeline advance.
| Seed production and viability at the treeline
Our results show a clear decrease in seed production of Pinus uncinata along the elevation gradient from the dense subalpine forest to the krummholz zone. Our analysis suggests that such decrease is directly related to the size of the cones.
Heat sum fluctuations, thermal limitations, and shorter growing seasons at high elevations may affect both cone production and development, and may have negative effects on fertilisation and embryo growth processes (Almqvist, Bergsten, Bondesson, & Eriksson, 1998;Sirois, 2000). Climate may act in synergy with other factors such as forest structure, resulting in lower pollen availability and an increased self-fertilisation (selfing) with the decrease in forest density along the elevation gradient (Iwasaki et al., 2013;Smith, Hamrick, & Kramer, 1988). We did not find a reduction in the proportion of full seeds with elevation, which would be the first negative outcome of selfing in other Pinus species (Iwasaki et al., 2013). Contrarily, our results suggest that seed development on the Pinus uncinata alpine treeline could be more influenced by low pollen supply at higher elevations (Iwaizumi & Takahashi, 2012). The decrease in the production of full seeds found along the gradient in our study sites is not accompanied by a reduction in the number of forest treeline transect in northern Québec, Canada. This author did not find differences in cone production along the transect (similar to our results), and suggested that the decrease in seed production and viability was related to a low viability of the pollen at the treeline.
TA B L E 3
Results of the statistical models for the parameters related to seed production Note: The estimated coefficient and standard error, z-or t-value, and p-value are shown for the fixed terms, including the interaction between two fixed terms when selected in the final model. The estimated variance and standard deviation are shown for the random effects in the mixed-effects models.
ANADON-ROSELL Et AL.
Šenfeldr and Maděra (2011) also found a clear decrease in the reproductive output of Norway spruce (Picea abies) in a former pastoral timberline ecotone in the Czech Republic, namely fewer seeds per cone, fewer cones per ha, and fewer cones per fertile specimen at the upper part of the ecotone than at the lower part. Another study in northern Finland found slight differences in seed anatomical maturity between timberline and treeline populations for Scots pine (Pinus sylvestris) but no differences in the number of cones per tree (Juntunen & Neuvonen, 2006). Overall, it seems a common pattern to find a decrease in reproductive output related to seed set of the treeline-forming species across elevation or latitudinal gradients, whereas patterns related to cone production and the proportion of viable seeds produced seem more variable among species.
Our findings strongly suggest that the production of viable seeds is the first limiting factor for forest expansion at the Pyrenean treeline. Even under the ongoing climate warming (IPCC, 2018), and unlike the situation in other treeline-forming species (Kullman, 2002), this suggests that the reproduction of Pinus uncinata at the treeline is still strongly limited by the present climate conditions. We acknowledge that fluctuations in the reproductive output due to inter-annual climate variations may occur (see section 4.2, Seed rain and dispersal, below), but we believe that the reported patterns and the relationship between the parameters studied (e.g., proportion Contrary to our expectations, we did not find a significant effect of elevation provenance on seed viability (i.e., germination success) and germination dynamics across sites, although the influence of elevation was marginally observed when assessing the sites separately. The lack of a consistent effect on the germination success by the elevation provenance reported here contrasts with other studies showing that germination is an adaptive trait that can vary largely with elevation (Castanha, Torn, Germino, Weibel, & Kueppers, 2013;Rehfeldt, 1989). Sirois (2000) showed that, 30 days after sowing, the percentage of total germination from seeds originating in trees closer to the treeline was lower than in those from the dense boreal forest. Results of Šenfeldr and Maděra (2011) reflected a similar pattern after 21 days, but they did not find significant differences. The different study species and the longer length of our germination trial (121 days) make it difficult to compare between these studies and our experiment. We did find, however, large differences in the germination success of seeds from
| Seed rain and dispersal
The seed rain experiment showed large differences between study sites and study years, emphasising the importance of local and inter-annual conditions for the reproduction dynamics of Pinus uncinata. However, despite the seed rain being much larger in 2012 than in the following years (2013 and 2014), we found a clear decrease in the number of seeds recovered in the seed traps along the elevation gradient that was constant through time. No seeds were found in the alpine grassland beyond the treeline, indicating that seed arrival, which is the first step for an eventual forest colonisation of alpine grasslands, is a limiting factor for treeline advance. Although the area covered by our seed traps is relatively small for concluding that no seeds arrive at alpine grasslands, our findings strongly suggest that important limitations for seed arrival beyond the krummholz zone exist. We acknowledge the possibility that seed predation influences our results, since it has been found to largely affect regeneration beyond the upper elevation limit in some species (Brown & Vellend, 2014;Castro, 1999 of dispersal constraints for Pinus uncinata seed (Camarero et al., 2005;Vitali et al., 2019).
The decrease in seed recovery with elevation was not significantly correlated with reproductive tree density, although the number of reproductive trees per area also decreased along the elevation gradient (Appendix S4). We believe these results need to be taken cautiously because there was a large variation in tree density among sites in the two subalpine forest positions (SFF and SDF), and the forest structure of the study sites was complex. There is, however, a clearly lower density of reproductive trees in the krummholz zone, which suggests that the lower number of seeds per area recovered there may be driven both by a lower reproductive capacity of individual trees and a decrease in tree density, as reported in other studies (Juntunen & Neuvonen, 2006;Šenfeldr & Maděra, 2011). Contrarily, the decrease in seed recovery from the subalpine dense forest (SDD) to the subalpine fragmented forest (SFF) may be mostly driven by a decrease in the reproductive capacity of individual trees, as shown by Sirois (2000). We found a positive correlation between the mean DBH of reproductive trees and the number of seeds recovered per area (Appendix S4), similarly to results found by Vitali, Camarero, Garbarino, Piermattei, and Urbinati (2017) on cone production. This reinforces the notion that the performance and characteristics of individual trees play an important role in explaining the reproductive output at treeline regardless of the forest structure.
We suggest that seed dispersal is of utmost importance in limiting treeline advance in the Pyrenees. Forest movement upslope will mainly depend on the establishment of seedlings germinating from nearby, scattered trees located at high elevation and rarely from long-distance dispersed seeds (Camarero et al., 2005;Kruse et al., 2019;Nathan, 2006). This may explain why, at the regional scale, ecotone densification (i.e., between the forest and the tree limit) is a more common response to global change in the Pyrenean treelines (Batllori et al., 2010; than treeline advance. Johnson, Gaddis, Cairns, and Krutovsky (2017) argued that seed dispersal was extensive enough in mountain hemlock (Tsuga mertensiana) in Alaska to ensure the advancement to higher elevations. However, this species has an average dispersal distance of 73 m, and has a long-distance dispersal capacity of up to 450 m, values much greater than those reported for Pinus uncinata (Camarero et al., 2005;Vitali et al., 2019). Therefore, the biology of the treeline-forming species may be of paramount importance in their reproductive dynamics, precluding any generalisation in terms of the reproduction-related limiting factors for treeline advance at the global scale. Additionally, seedling recruitment and the eventual treeline advance will not only depend on the arrival of the seeds, but also on the subsequent biotic and abiotic processes affecting seedling establishment and survival Nathan & Muller-Landau, 2000), such as the environmental harshness and scarcity of favourable sites for establishment (Smith, Germino,
TA B L E 5
Results of the zero-inflated negative binomial model for seed rain F I G U R E 4 Seeds collected in the seed traps along the elevation gradient (SDF = subalpine dense forest; SFF = subalpine fragmented forest; KZ = krummholz; AG = alpine grassland).
Data are plotted across the three study sites (Dellui, Amitges and Gelada) and the three study years (2012)(2013)(2014). Capital letters indicate significant differences between elevation positions. The asterisk for AG indicates that no seeds were found in any of the seed traps, and thus it was excluded from the analysis. See Appendix S9 for the number of seeds recovered in the seed traps at each elevation position per study year and site [Colour figure can be viewed at wileyonlinelibrary.com]
| CON CLUS IONS
Our study provides the first evidence of reproductive output as a limiting factor for the advance of the treeline in the Pyrenees. The decrease in seed production along the elevation gradient together with a lack of seed arrival at higher elevations (i.e., alpine grassland) strongly suggests that seed production and dispersal are constraining the ongoing rates of treeline advance in the Pyrenees. By contrast, the lack of clear differences in germination success and germination dynamics from the subalpine forests to the krummholz zone seems to indicate a limited role of genetic constraints at the site level, although genetic variability may have a more important role in treeline dynamics at the regional scale (differential site dynamics). Overall, despite potential positive effects of warming on seed production, our findings suggest that the poor dispersal capacity of Pinus uncinata seeds to the alpine grasslands beyond the treeline will slow down the upslope advancement of the forest. This, together with other climate change-related events such as increased droughts, which may imply higher seedling mortality rates, may reduce the expected climatic sensitivity of treeline position to warming climates.
ACK N OWLED G EM ENTS
We would like to thank Christian Ritz, from the University of Copenhagen, for his statistical advice on meta-analytic randomeffects models. We are grateful to Albert Ferré for providing the map and the picture in Figure 1. We also thank Oriol Grau, Francesc Talavera, Cecília Roma, Xavier Jou, Gerard Fuentes, Andreu Passola, Enric Cuadras and Xavier Trullàs for their help in the field. Finally, we are grateful to Matteo Garbarino and an anonymous reviewer for their helpful comments on the manuscript.
AUTH O R CO NTR I B UTI O N S
MT, JMN, EC and EB conceived the research idea. MT and JMN collected cones, seeds and data and performed the laboratory work.
AAR performed the statistical analyses. AAR wrote the manuscript with contributions from MT and EB. All the authors discussed the results and commented on the manuscript.
DATA AVA I L A B I L I T Y S TAT E M E N T
Data supporting this publication is stored at https ://github.com/ anadon-rosel l/Anadon-Rosell_et_al_2019_JVS. | 2019-12-12T10:22:53.502Z | 2020-02-27T00:00:00.000 | {
"year": 2020,
"sha1": "775370c326d9cff590a75ed26d9486ec829ce5e8",
"oa_license": "CCBY",
"oa_url": "https://onlinelibrary.wiley.com/doi/pdfdirect/10.1111/jvs.12849",
"oa_status": "HYBRID",
"pdf_src": "Adhoc",
"pdf_hash": "3dbca9d769ff41e381335da7365430634d98dab4",
"s2fieldsofstudy": [
"Environmental Science"
],
"extfieldsofstudy": [
"Biology"
]
} |
51894677 | pes2o/s2orc | v3-fos-license | Avian trichomonosis mortality events in band-tailed pigeons (Patagioenas fasciata) in California during winter 2014–2015
Avian trichomonosis is an upper digestive tract disease of birds typically caused by the protozoan parasite Trichomonas gallinae. In California (U.S.A), trichomonosis is known to cause periodic epidemics in the Pacific Coast band-tailed pigeon (Patagioenas fasciata monolis), a migratory upland game bird. We summarize the mortality events that occurred during winter 2014–2015 including the duration, estimated mortality, pathology, and genetic identity of infecting parasites. Increased mortality was reported from locations in 25 counties between November 2014 and June 2015. Based on reports, carcasses received, wildlife rehabilitation center admissions, site visits, and regular monitoring by local personnel, total mortality was estimated at 18,440. At necropsy, birds had multiple coalescing lesions in the oral cavity involving the upper palate and/or around the tongue and glottis, esophagus, crop, and/or proventriculus. Birds collected from Contra Costa (63.9%; 30/47); Marin (75.0%; 6/8), San Mateo (46.7%; 14/30), and Santa Clara (35.0%; 37/106) counties were more likely to have lesions extending into their head involving muscle, sinuses, ear canals, eye sockets, and bone (χ2 = 62.9; df = 11; P < 0.001). Histopathologic findings included pharyngitis, esophagitis, myositis, and air sacculitis of the pneumatic bone of the skull. Mixed bacterial colonies were found multifocally at the fronts of the necrosis in six of the eleven birds examined histologically. Infecting trichomonads included T. gallinae subtype A2 (n = 5), un-typed T. gallinae (n = 4), mixed infection with T. gallinae subtype A2 and T. stableri (n = 1), and mixed infection with un-typed T. gallinae and T. stableri (n = 1). The winter 2014–2015 epidemic was the largest on record in terms of duration, locations, and birds affected. Infection dynamics may have been exacerbated by the drought in California. Increased monitoring of band-tailed pigeons is needed to understand the long-term impacts of large-scale mortality events on their population.
Introduction
Avian trichomonosis is a disease caused by a protozoan parasite affecting a diverse array of species with columbids being especially susceptible (Forrester and Foster, 2008). The parasite most frequently isolated from infected birds is Trichomonas gallinae of which 15 different subtypes have been identified based on the internal transcribed spacer (ITS) region (Gerhold et al., 2008;Grabensteiner et al., 2010). Further genotyping of T. gallinae strains can be accomplished by sequencing the hydrogenosomal Fe-hydrogenase gene (Lawson et al., 2011;Chi et al., 2013). Infection can range from subclinical in carrier birds, to the development of caseonecrotic lesions in the upper digestive tract and death. Birds become infected by ingesting contaminated water or food while adult columbids can directly infect their chicks when feeding them crop milk (Forrester and Foster, 2008). Once infected, the parasite multiplies rapidly in the oral cavity invading the mucosa leading to parasite-mediated desquamation and development of lesions (Kietzmann, 1993). Infected birds typically become emaciated as the lesions block the passage of food and eventually die from starvation or suffocation if the lesions block the airway (Forrester and Foster, 2008). As such, sick birds are often weak and reluctant to fly, have difficulty swallowing and exhibit labored or open-mouth breathing with death occurring typically 10-14 days post-infection (Stabler, 1947;Perez-Mesa et al., 1961). While the death of an individual bird from trichomonosis may occur any time of the year, epidemics or mortality events tend to have a distinct seasonality within an avian species, typically coinciding with the period of closest contact between conspecifics often including an influx of immunologically naïve juveniles. In California, these events occur most frequently in mourning doves (Zeneida macroura) during the spring and summer (Stabler and Herman, 1951) and band-tailed pigeons (Patagioenas fasciata) during the winter (Rogers et al., 2016a).
Whereas mourning doves range across the continental United States, band-tailed pigeons have a more limited range with two distinct populations inhabiting the western U.S., the Interior (P. f. fasciata) and Pacific Coast (P. f. monolis) subspecies (Keppie and Braun, 2000). Interior band-tailed pigeons occur in the southwestern U.S. and Mexico, while Pacific Coast band-tailed pigeons occur primarily from British Columbia south through southern California. While once an important game bird in the western U.S., both populations have undergone drastic declines resulting in reduced hunting opportunities (Seamans, 2016). Periodic trichomonosis epidemics have been documented since the mid-1940s within the Pacific Coast population and are an important factor in their decline, with near annual mortality events occurring since the early 2000s (Rogers et al., 2016a). Estimated mortality during these events is highly variable from year to year (Stromberg et al., 2008;Rogers et al., 2016a), but has the potential to be more than 2-3 times higher than the annual harvest, which typically ranges between 6000-10,000 pigeons (Seamans, 2016). In contrast to most other game birds, such as doves, turkeys, and waterfowl, the reproductive rate for band-tailed pigeons is relatively low. A pair of band-tailed pigeons produce, on average, one chick per year (Keppie and Braun, 2000) which may result in slow recovery when the population experiences large losses of individual birds through trichomonosis epidemics and harvest.
In previous studies, Trichomonas gallinae Fe-hydrogenase subtype A2, was the only T. gallinae subtype isolated from band-tailed pigeons during epidemics in California (Girard et al., 2014b). Subtype A1 has also been isolated from band-tailed pigeons in California, however, at a much lower prevalence and only during non-epidemic time periods (Girard et al., 2014b). In Europe and Canada, subtype A1 is the most common subtype isolated from avian species including songbirds, raptors, and columbids (Lawson et al., 2011;Chi et al., 2013;McBurney et al., 2015;Stockdale et al., 2015). In addition to T. gallinae, bandtailed pigeons were also found to be infected with a newly described species, T. stableri, which is more closely related to the human pathogen T. vaginalis than to T. gallinae (Girard et al., 2014a). Trichomonas stableri has been isolated from band-tailed pigeons both during epidemics and non-epidemics and occurs at a lower prevalence than T. gallinae (Girard et al., 2014a); however, both parasites cause similar disease, highlighting the importance of molecular characterization of parasites when attributing mortality events to a specific etiology.
Historically, trichomonosis epidemics involving Pacific Coast bandtailed pigeons have only been documented during the winter when the majority of the population is overwintering in central and southern California (Rogers et al., 2016a). During this time, pigeons form large flocks increasing contact between individuals at communal water sources and facilitating parasite transmission between susceptible birds. Similar to infections in other columbids (Bunbury et al., 2007) the likelihood of an epidemic occurring during a given winter in bandtailed pigeons is correlated with weather conditions, namely warmer temperatures and lower precipitation, which may improve parasite viability while increasing contact between individual birds at fewer available water sources (Rogers et al., 2016a).
This pattern of winter-only trichomonosis mortality events changed in 2014, when events were documented for the first time during the summer in at least three central California locations (Rogers et al., 2016b). During summer, band-tailed pigeons are more dispersed for breeding activities compared to winter (Keppie and Braun, 2000) presumably leading to reduced contact between pigeons and reduced opportunity for parasite transmission. However, 2014 was the warmest year on record in California, followed by 2015, the second warmest year on record (http://ca.water.usgs.gov/data/drought/). By early 2015, California was entering a fifth year of drought with roughly 94% of the state classified in severe drought according to the National Drought Mitigation Center (http://droughtmonitor.unl.edu).
Here we describe the avian trichomonosis mortality events involving Pacific Coast band-tailed pigeons that took place during winter 2014-2015 in California. Our analysis includes a summary of the geographic range and duration of events, numbers of birds affected, histopathological analysis of diseased birds, and molecular characterization of infecting parasites. By evaluating the characteristics of these epidemics, we improve our understanding of the impacts of large-scale disease events on the population health of this declining upland game bird.
Mortality reports
The California Department of Fish and Wildlife's (CDFW) Wildlife Investigations Laboratory (WIL; Rancho Cordova, CA) is responsible for investigating causes of mortality in the state's wildlife. Incidents of sick and dead band-tailed pigeons were reported to WIL by department staff, the public, wildlife rehabilitation centers, and other government agencies by phone, e-mail, and online mortality reporting form (www. wildlife.ca.gov). Information compiled from these reports included date reported, location, number of sick and dead pigeons observed, and start and end dates of mortality, if applicable. When possible, a site visit was made to locations with reported mortality to confirm the bird species involved and to record numbers of sick and dead birds, habitat characteristics, and food and water sources.
Reports of avian mortality are also provided to WIL by the California Department of Public Health (CDPH; Richmond, CA) through their Dead Bird Hotline. The Dead Bird Hotline was established in 2003 for West Nile virus surveillance and enables the public to report dead birds to CDPH via a toll-free telephone number or online form (www. westnile.ca.gov). From these reports, incidences of band-tailed pigeon mortality were compiled.
Finally, wildlife rehabilitation centers in California are required to submit an annual report to CDFW of all wildlife admitted into the center under their permit. These reports total the number and final disposition (released, transferred, pending, euthanized, died, or DOA) of each species admitted to the center between January 1 and December 31 in a given year. While these reports do not denote date, location, or reason for intake for each individual animal, they do provide intake numbers for each species for the geographic area (e.g. county) in which the center is located. Band-tailed pigeon intake numbers were compiled by county from the 2015 annual reports for the 41 wildlife rehabilitation centers reporting band-tailed pigeon intake.
Post-mortem examination
Avian trichomonosis mortality events were defined as ≥5 birds dying in the same geographic location over several days to weeks (Rogers et al., 2016a). Once alerted to mortality, an effort was made to collect a subset of carcasses from outbreak locations for post-mortem examination and sampling for trichomonad isolation. Carcasses were received at WIL and stored in a freezer (−20°C) until the post-mortem examination. Prior to the examination, the carcass was thawed at 4°C for 24-48 h and gross findings were recorded including age, sex, presence and location of lesions, adipose deposition, condition of organs, and abnormalities (e.g. injuries). Birds were aged by plumage as juveniles (hatch year), subadults (second year), or adults (after second year) (Sanders and Braun, 2014). Nutritional condition was assessed by degree of adipose deposition and mass (grams). Adipose deposition was rated as none (no subcutaneous or internal adipose), trace (no subcutaneous and minimal internal adipose), and moderate to heavy (adequate subcutaneous and internal adipose).
Histopathology and trichomonad characterization
Of the carcasses received, eleven were selected for histopathology at the California Animal Health and Food Safety Laboratory (CAHFS; Davis, CA) and molecular characterization of infecting parasites was conducted at the One Health Institute, University of California, Davis. Tissue samples of oral mucosa with lesions, skull including the eyes and brain, muscle, thyroid/parathyroid glands, peripheral nerves, trachea, lung, heart, esophagus, crop, proventriculus, gizzard, pancreas, intestines, liver, spleen, adrenals, kidneys, and gonads were collected and immersed in 10% neutral buffered formalin, paraffin-embedded, sectioned at 4 μm, and stained with hematoxylin and eosin for histologic examination by light microscopy (Girard et al., 2014a). Lesion tissue was sampled by excision, placed in a cryovial, and stored at −80°C for molecular analysis. DNA was amplified and sequenced using primers targeting the ITS1/5.8S rRNA/ITS2 and Fe-hydrogenase loci and sequence analysis was performed by alignment to published trichomonad sequences in GenBank (Girard et al., 2014b). Amplicons whose sequences were difficult to interpret using Geneious Pro v. 5.34 sequence analysis software (http://www.geneious.com) (Kearse et al., 2012) were cloned using the TOPO ® TA cloning kit (Thermo Fisher Scientific, Waltham, MA), and 10 colonies per DNA isolate were chosen for further sequence analysis. All nucleotide sequences generated from isolates in this study were identical to isolates previously deposited in GenBank (Girard et al., 2014a, b).
Cause of death was assigned based on case history, clinical signs, if known, post-mortem findings (e.g. presence of lesions, trauma), and diagnostic testing for the eleven birds that underwent histopathology and trichomonad isolation.
Statistical analysis
We analyzed differences in prevalence using the chi-square (χ 2 ) test of independence and mean body mass was compared between event locations using the Mann-Whitney U test. Values reported are mean ± SE. Statistical analyses were performed using NCSS (Hintze, 2007) and P ≤ 0.05 were considered statistically significant. Maps were prepared using ArcMap (ESRI, Inc., 2014).
Wildlife rehabilitation center intake
Band-tailed pigeon intake was reported by 41 wildlife rehabilitation centers in 27 counties between 01 Jan and 31 Dec 2015. In total, 1215 pigeons were admitted to wildlife rehabilitation centers during 2015 (Fig. 1B). During the mortality event, rehabilitation centers reported increased intake of pigeons with visible lesions in the oral cavity or palpable lesions in the throat or crop, agonal breathing, exophthalmos often with discharge, otitis, and/or neurological symptoms including ataxia and torticollis.
Based on reports, carcasses received, rehabilitation center intake, site visits, and regular monitoring by local personnel during the duration of the events, the number of pigeons confirmed dead during these mortality events was 3467. Across all mortality event locations, the total estimated mortality was 18,440 birds, after considering the geographic range, duration of events, numbers of susceptible live pigeons present, and predator and scavenger activity. In total, deaths were reported from 33 of 58 California counties, with ≥5 deaths reported from locations in 25 counties between November 2014 and June 2015.
Histopathology
Due to the variation in lesion presentation for birds from certain geographic areas, histopathology was conducted on select birds collected from Santa Clara (n = 9) and San Mateo (n = 2) counties in January 2015 and included eight males and three females; all birds were aged as subadults with the exception of two males which were aged as adults. Histopathologic findings were similar among the birds and included pharyngitis (n = 10), esophagitis (n = 10), myositis (n = 10), cellulitis (n = 9), air sacculitis of the pneumatic bone of the skull (n = 6), sinusitis (n = 6), osteomyelitis (n = 4), tracheitis (n = 3), otitis (n = 2), hepatitis (n = 2) and/or splenitis (n = 1). Two birds had pneumonia due to aspiration of necrotic debris. A mixture of bacterial colonies including both Gram negative rods and Gram positive cocci were found multifocally at the fronts of the necrosis in six of the birds; culture and identification of bacteria was nonproductive. Large numbers of trichomonads were observed at the front of the lesions and in the necrotic debris (Fig. 5).
Trichomonad characterization
Trichomonad infection was confirmed by PCR in all eleven birds. Trichomonas gallinae subtype A2 (n = 5) was isolated in 4 birds from Santa Clara County and 1 bird from San Mateo County. Trichomonas gallinae that could not be fine-typed (n = 4) was isolated in 3 birds from Santa Clara County and 1 bird from San Mateo County. Mixed infection with T. gallinae subtype A2 and T. stableri (n = 1) and mixed infection with un-typed T. gallinae and T. stableri (n = 1) where isolated in birds from Santa Clara County.
Discussion
In this study, we document the most widespread and longest duration of an avian trichomonosis epidemic ever recorded for Pacific Coast band-tailed pigeons. Mortality events with ≥5 deaths were reported in 25 counties in California. This outbreak also was of the longest duration (Girard et al., 2014b). Total mortality for the 2011-2012 events was estimated around 9000 birds, while the total mortality for the 2014-2015 events reported here, likely exceeded 18,000 birds. These findings are significant given that the Pacific Coast band-tailed pigeon population has been declining since the 1960s (Seamans, 2016). Although legal harvest has been maintained at a restrictive level since 1990 because of population decline, low recruitment, and the potential for overharvest, the average annual harvest was still 9760 birds between 2011 and 2015 based on the Migratory Bird Harvest Information Program administered by the U.S. Fish and Wildlife Service (Seamans, 2016). As such, the estimated losses of band-tailed pigeons from this mortality event was approximately two times the amount removed during annual harvest.
Over 90% of infected birds submitted for necropsy were subadults, in their first winter, and adults, suggesting the timing of these events has serious consequences for future recruitment potential. Even when a pair of band-tailed pigeons breeds successfully, they produce on average only one chick per year (Keppie and Braun, 2000), as such, if disease events are frequent and/or severe, mortality will exceed recruitment resulting in population decline. Given the low reproductive rate of this species, high adult survivorship is required to grow the population. With a potentially large number of breeding birds dying due to disease, combined with losses during subsequent fall hunting seasons (Smith, 1968;Silovsky, 1969), adult survivorship likely drops well below management expectations during event years, hampering long-term recruitment potential (Jarvis and Passmore, 1992).
Altering or periodically withholding hunting seasons has been effectively utilized in the past when band-tailed pigeon populations dropped below sustainable levels (Pacific Flyway Council, 2010). Historically, a hunting moratorium was enforced between 1913 and 1931 as a result of overharvest that took place in the late 1800s to the early 1900s prior to the establishment of the Migratory Bird Treaty Act (Grinnell, 1913;Pacific Flyway Council, 2010). More recently, Washington (1991-2001) and British Columbia (1994-2001 have closed hunting seasons in response to the declining population (Seamans, 2016). These and other approaches may need to be considered rangewide to help recover the Pacific Coast population during consecutive years with high disease mortality to sustain the population. During the 2014-2015 mortality events the clinical symptoms and the pathology of infected birds from Contra Costa, Marin, San Mateo, and Santa Clara counties were different than birds from other locations and during previously investigated events (Girard et al., 2014b;Stromberg et al., 2008). Pigeons admitted to rehabilitation centers within these counties were characterized as having neurological deficits including vision and balance problems. The post-mortem examination of these birds often revealed caseous lesions involving the mucosa extending deep into underlying muscle and bone of the upper palate and into the choana to the sinuses, eye sockets and ear canals causing vestibular disease; manifestations that were not reported in previous studies (Girard et al., 2014b;Stromberg et al., 2008). In addition, roughly 60% of birds examined were in good nutritional condition at the time of death. In fact, birds from Contra Costa, Marin, San Mateo, and Santa Clara counties were on average 10 g heavier than birds collected from the remaining counties. This contrasts sharply with the typical pathology observed in band-tailed pigeons infected by Trichomonas spp., in which individuals become emaciated and die of starvation after lesions blocked the passage of food (Girard et al., 2014b;Rogers et al., 2016bStromberg et al., 2008. Birds with notable adipose stores at time of death likely succumbed to disease more quickly, perhaps corresponding to infection with a particular parasite strain with high virulence. Additionally, bacteria were observed multifocally at the fronts of necrosis in six of eleven birds examined histologically. Unfortunately, due to the volume of carcasses received, the carcasses had been frozen prior to necropsy and bacterial cultures were not productive limiting our ability to identify the bacteria present in the lesions. However, the bacteria was identified as a mixture of Gram negative rods and Gram positive cocci indicating their presence was likely secondary to the trichomonosis (Girard et al., 2014b;Stoute et al., 2009).
Consistent with previous surveillance data from the 2011-2012 mortality events (Girard et al., 2014b), the primary trichomonad identified in band-tailed pigeons in this study was T. gallinae subtype A2. While T. gallinae subtype A2 has occasionally been isolated from apparently healthy band-tailed pigeons (Girard et al., 2014b), it has been the only T. gallinae subtype isolated from band-tailed pigeons during mortality events. Interestingly, apart from band-tailed pigeons, the A2 subtype has only rarely been isolated, including from a mourning dove in California (Girard et al., 2014b) and a captive budgerigar (Melopsittacus undulates) in Scotland (Chi et al., 2013), both with oral lesions, and a lesion-free Madagascar turtle dove (Streptopelia picturata) in Seychelles (Lawson et al., 2011). Given the high pathogenicity of T. gallinae subtype A2 in band-tailed pigeons, and susceptibility of other bird species for the parasite (Rogers et al., 2016b), continued surveillance among band-tailed pigeons and conspecifics is warranted. Although fewer band-tailed pigeons in the present study were infected with the newly described parasite T. stableri (Girard et al., 2014a), its isolation during the 2014-2015 mortality events confirms this parasite continues to be an important pathogen for Pacific Coast band-tailed pigeons. Also consistent with the 2011-2012 mortality events (Girard et al., 2014b), some pigeons during the 2014-2015 mortality events were co-infected with T. gallinae and T. stableri.
Warmer temperatures and lower precipitation have been linked to increased Trichomonas spp. infection prevalence in endangered pink pigeons (Nesoenas mayeri) in Mauritius (Bunbury et al., 2007) and winter epidemics in band-tailed pigeons in California (Rogers et al., 2016a). Whether these conditions favor improved viability of trichomonads and/or increase transmission rates is unknown. However, the reduced availability of natural water sources during the drought in California (Hall et al., 2016, California Executive Order B-29-15), meant pigeons were relying more on artificial water sources such as bird baths, garden fountains, retention ponds, and horse troughs. Given the drought and statewide water restrictions (California Executive Order B-29-15), these sources of water may have contained higher amounts of organic debris possibly contributing to parasite persistence . In combination with warmer temperatures, pigeons may have increased their frequency of water consumption putting individuals at higher risk of infection (Swinnerton et al., 2005). In addition to influencing infection dynamics, the drought conditions may also have acted to increase an individual band-tailed pigeon's susceptibility to infection through physiological stress (Fair and Whitaker, 2008). Temperature extremes and drought are predicted to become more frequent in California as the climate changes (Diffenbaugh et al., 2015;Swain et al., 2016), in which case, large-scale mortality events may also become more common, threatening the long-term persistence of the band-tailed pigeon.
Given the increasing frequency of trichomonosis mortality events in combination with other sources of mortality (e.g. harvest, poaching, trauma), more research is needed to fully understand the dynamics of this disease in band-tailed pigeons and the future sustainability of the population. Evaluating pigeon migration and within-season movements in relation to food availability and habitat use may highlight risk factors for parasite transmission at communal sites and/or interaction with other animals. Continued studies of parasite molecular biology and virulence, coupled with host susceptibility data, will improve understanding of risk factors for widespread mortality in Pacific Coast bandtailed pigeons, and pave the way for species management interventions that may mitigate further population decline.
Conflicts of interest
The authors declare no conflict of interest. Haemotoxylin and eosin staining (left) of the oral tissue of a band-tailed pigeon (Patagioenas fasciata monilis) recovered during an avian trichomonosis mortality event showing a diffuse thick layer of necrosis extending through the submucosa and multifocally into the deeper soft tissue layers and skeletal muscle; scale bar is 200 μm. Immunohistochemical staining (right) of trichomonad antigen (red) of the same bird demonstrating large numbers of trichomonads in the oral tissue; scale bar is 50 μm. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) | 2018-08-14T13:49:22.607Z | 2018-06-30T00:00:00.000 | {
"year": 2018,
"sha1": "4f4c25a5575ccd45e7fc6f156e20c6b99a197cc7",
"oa_license": "CCBYNCND",
"oa_url": "https://doi.org/10.1016/j.ijppaw.2018.06.006",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "4f4c25a5575ccd45e7fc6f156e20c6b99a197cc7",
"s2fieldsofstudy": [
"Environmental Science",
"Biology"
],
"extfieldsofstudy": [
"Biology",
"Medicine"
]
} |
230629088 | pes2o/s2orc | v3-fos-license | A Novel Cosine-based Internal and External Validation metrics to assess twitter Data Clustering using Hybrid Topic Models
In document clustering labeled and unlabeled documents are organized into a desired number of coherent and meaningful sub-clusters. Topic models are useful in extracting cluster tendency from Twitter-based data document clusters. Evaluating cluster tendency and performance with a reliable metric is one of the unsolved problems in topic document clustering. In the previous study cluster validity metrics have been proposed under Euclidean distance measure, these metrics underperform in topic models when dealing with numerical databases and for large corpus datasets. In this paper, to assess Twitter data clustering a novel cosine based internal and external validity indices used by considering closeness between documents, the lexical similarity and cluster classification metrics for each topic separately using Hybrid topic models. Experimentally proved the effectiveness of cosine based internal and external validity metrics and results compared with Euclidean metrics using benchmark and Twitter-based datasets.
Introduction
In both textual and numerical data clustering, one challenging issue is to organize massive data into sub-clusters without or with prior knowledge.Topics of documents in Topic modeling found by using unsupervised classification [1] or by post clustering without prior information and supervised classification process or preclustering [2] in the presence of prior information. The final result consists of assigning a previously unknown or known category to each document relevant to the main topic. There sufficient literature in the field of topic modeling techniques [3][4][5][6], clustering methods and algorithms using hybrid topic models [7] in Twitter data document clustering. Evaluating cluster tendency and performance with a reliable metric is one of the unsolved problems in topic document clustering. Cluster validity measured with internal and external validity indices. The external validity indices [8][9][10][11] measure the correspondence between identified clusters and externally provided labels in social voting [28].The Internal validity indices [12][13][14][15][16][17] evaluate the goodness of cluster structure with partitioned data by considering compactness and separation of obtained partitioned structure. Internal validity indices are preferred in performance measure because in most of the cases prior information on the number of clusters will not be available. In previous literature, a wide variety of internal and external validity indices have been provided which will be useful in finding the number of topics but not in choosing an appropriate measure and metric to validate cluster and also not by considering the elements in the cluster are well classified are not. Most commonly used measure is especially using Euclidean distance, which shows poor results in high dimensionality document clustering. In this paper, a novel cosine based internal and external validity metrics proposed for internally evaluating the results of a document clustering by considering into account the peculiarity of textual data [18], closeness between documents [19], considering the lexical similarity [20] and also considered cluster classification metrics in classification of elements in the cluster are well classified ICRAEM 2020 IOP Conf. Series: Materials Science and Engineering 981 (2020) 022031 IOP Publishing doi: 10.1088/1757-899X/981/2/022031 2 or not. Experimentally effectiveness of proposed clusters validity metrics is evaluated with benchmark and Twitter-based datasets. Rest of the paper is organized as follows: Section2 presents the theoretical background of clustering in Topic Modelling; Section3 describes the process description and cluster validation; in Section4 preliminary experimental evaluation and performance of validation measures; Section5 a novel cosine based validity metrics for validating document clustering, comparison with validity metrics under Euclidean and cluster classification metrics in cluster validation. Finally, in Section6 the conclusion and future scope of the work is presented.
Theoretical Background Of Clustering In Topic Modeling
For the same dataset, using different algorithms gives different solutions by generating sub-clusters, different choice of input parameters produce different results for the same algorithm which affects the final result in finding the optimal number of topics or clusters in the given topic document. To assess cluster obtained by used algorithm, to decide which algorithm is most suitable for the specific application, and to provide reliability to results suitable evaluation criteria under suitable measure is still needed. In most of algorithms proximities, pairwise distances measured by using Euclidean distance metrics are considered which is suitable for the lower number of dimensionality, it loses its reliability and interpretability at an increase of dimensionality. Clustering algorithm deal with distance, and distance relates to similarity/dissimilarity. The complement to Euclidean metric is cosine based similarity metric in text classification problems which uses both magnitude and direction of vectors, which is non-negative, independent of document length and bounded between [0, 1]. One of the most interesting variations in K-means family is spherical k-means [21], [31] which is based on cosine based similarity used in information retrieval, in which the effect of different length of documents is reduced by normalization.Given two tweet documents di and djin a corpus, then cosine based distance similarity is given as (1) The cosine is 1 if the documents use the same words and 0 if they have no two terms in common. The effort of different length of documents is lessened by normalization.
Datasets Description
For the experiment, the datasets collected from Twitter on 20 topics of health-related [24], [27], [32] documents, TREC2014, TREC2015 Keyword Phrases Tweets collected from Twitter are used as described in [7] and Tweets extracted from Twitter related to 25 keyword phrases of TREC2018 [22] as described in Table 1 are also used. Experiments are implemented with Intel core i7processor @3.4 GHz, 8MB cache, 16GB RAM, 1TB HDD in IDLE (Python 3.8 64bit) environment on these four different datasets and results discussed in ensuing sections.
3.2Process Description
On each collected corpus as mentioned above the following steps are implemented: Step1: For each Twitter-based datasets collected preprocessing is performed by using Python Gensim library to prepare text documents for Document Clustering and classification. Step2: Programs implemented in Python to apply hybrid Topic models [7] under Cosine based and Euclidean distance-based measures. Step3: Document clustering and classification performed.
Step4: Assessment of document cluster with confusion matrices [23] and classification metrics by using novel cosine-based internal and external validity metrics. Step5: Results compared with Euclidean metrics with confusion matrices and classification metrics are done.
Performance of Cluster Validation
In Topic modelling selection of appropriate method for implementation and assessment of clustering quality in information retrieval [25], image processing applications [26], [33] is still open challenges. Since the number of topics or clusters is not known ahead, irrespective of the clustering model, the final results needed to be evaluated for cluster validation. To validate cluster, external validation indices and internal validation indices are used. For choosing an optimal clustering algorithm and to assess identified clusters corresponding to externally provided labels external validity indices are used. In addition to these internal validation indices evaluate cluster structure with partitioned data by considering compactness and separation of obtained partitioned structure. It measures intra-cluster homogeneity, inter-cluster separability or both. In the majority of the application, prior information of the number of clusters is not available in such scenarios internal validation indices are best suited for cluster validation. In this paper,both internal validity indices (C.A., NMI, Precision, Recall and F-Score) and internal validity indices (DB, SI, XI, PCI, PEI and SM) used for performance evaluation. In addition to these validity indices classification metrics are also used to check the elements in the cluster are well classified or not topic wise.
Preliminary Experimental Evaluation and Performance of Validation Measures
The experiment aims to compare the behaviour of cosine based internal and external validity indices with Euclidean based indices. To perform comparative study different benchmark and real-time Results of all datasets of external and internal validity indices under cosine and Euclidean are tabulated for four hybrid topic models. Some sample results are preseneted in tabular and graphical forms. In Table 2 External validity index (Cluster Accuracy) of 2 keyword phrases to 25 keyword phrases of TREC2018 datasets and in Table 3, all external and internal validity indices of TREC2014 dataset are shown. From these results interpreted that cosine based external and internal validity indices perform better than that of Euclidean in majority of keyword phrases. Particularly performs well when smaller keyword phrases, as keyword phrases size increases result values are decreasing under both metrics, but still consistency is maintained in case of cosine based metrics. Higher values of results are represented in bold format.
Document clusters validation by using Cosine based Measures
For evaluating compactness and separation of formed clusters usually Euclidean measure deployed in previous studies and for external validity indices in most of the cases. Using this measure may be inconsistent with the criterion for getting partition for a specific algorithm. With this motivation, in this paper novel cosine based derived metrics are usedin document clustering algorithms using hybrid topic models, hybrid framework [29], [30] and also in validating formed clusters by using these metrics. In addition to that clusters to have high cohesion and well distinguished, both compactness and separation are considered.
Consider corpus
,...... , 2 1 consists of n document vectors in 'p' terms space of dimension. With the help of hybrid clustering algorithm k number of clusters Cq (where q=1, 2,…,k) have been identified, such that each document has one of the labels identifying the k different ICRAEM 2020 IOP Conf. Series: Materials Science and Engineering 981 (2020) 022031 IOP Publishing doi:10.1088/1757-899X/981/2/022031 6 clusters.These clustering algorithm aims to maximize intra-cluster proximities and minimize intercluster proximities.Let di, di' and dj be three documents in a corpus X, with di, and di' belongs to the same cluster and djbelongs to other clusters. Compactness and separation can be calculated as follows: Compactness By considering cosine similarity Eq. (6) can be written as (7) By using these equations calculated values of validity indices are tabulated. Higher values are represented in bold form in the tables mentioned below. From these tabulated values to show comparison between cosine and Euclidean represented in graphical form in following sections. Figure 1(a) Accuracy index results for 2topics to 20 topics are shown, from this spiral graph interpreted that Visual NMF and Visual LSI algorithm perform well. At 7T, 8T, 11T and 12T visual NMF perform better than the other three methods. By observing NMI external index results shown in Figure 1(b) for most of the topics Visual LSI method perform well, whereas, for 7T, 8T, 11T and 13T Visual NMF performance is better than other methods. In Figure 1(c) precision values are shown, from thisinferred Visual NMF performs well in most of the topics except 3T to 6T, and 10T Visual LSI performs well.
Recall values are shown in Figure 1(d) from which conclusion drawn except for 3T to 6T, for rest of topics Visual NMF performance is good. In those topics, Visual LSI performs well. In Figure 1 In the case of this index, the minimum value will perform better clustering results. Form this graph on observation, for most of the keyword phrases visual LSI performs better than other methods. In case of 7keywords, 8keywords, 13keywords, 16keywords, 19keywords and 20keywords Visual NMF performed better than other methods.Silhouette index (SI) values range from -1 to +1. If this index value is nearer to +1 then cluster performance will be best. If values decrease from +1 to -1 its performance also decreases. From the bar graphshown in Figure 2(b) results can interpret Visual NMF under cosine performs well in all TREC2018 keyword phrases.Partition coefficient index (PCI) values lie between 0 and 1. Values nearer to 1 will be treated as best. From Figure 2(d), based on PCI values under cosine metric, in case of 7keywords, and 10 keywords Visual LSI perform better, for 11keyword phrases Visual NMF perform well and in rest of all keyword phrases Visual PLSI methods perform well. Figure 2(e) shows performance values of partition coefficient internal index values which range from 0 to logac. In this case its value range from 0 to 3 as indicated on the Y-axis line graph. The minimum value will be considered for higher performance in clustering. From this graph,Visual LSI method performs well for 7 to 10 keyword phrases, for 11 and 12 keyword phrases Visual NMF and for rest of keyword phrases Visual PLSI perform better than other methods.Separation Measure internal index value is smaller then it will have greater performance. In this case, its value ranges from 0 to 10 as represented on the Y-axis. From this line graph as shown in Fig. 3(f), 7keyword phrases, 8keywords, 11keywords and 13keywords Visual LSI perform well and in rest of keyword phrases, Visual PLSI under cosine metric performs better than other methods. Figure. 3(a) Accuracy index results for 2topics to 20 topics are shown, from this spiral graphVisual LSI algorithm under cosine based metric perform well. At 7T, 8T, 11T and 12T visual NMF under cosine performs better than the other three methods. By observing NMI external index results in Figure 3(b) for most of the topics Visual LSI under cosine metric performs well, whereas, for 7T, 8T, 11T and 12T Visual NMF under cosine performance is better than other methods. In Figure 3 In the case of this index, the minimum value will be considered for better clustering results. Form this graph on observation, 2keywords to 6keywords, 9keywords to 12keywords, 14keywords and 17keywords visual LSI under cosine perform well, for 18keyword phrases visual NMF perform best and rest of keyword phrases Visual NMF under Cosine performs better than other models. Silhouette index (SI) values range from -1 to +1. If this index value is nearer to +1 then cluster performance will be best. If values decrease from +1 to -1 its performance also decreases. From the line graph as shown in Figure 4( Figure 4(c) Xie-Beni index (XI) internal validity index value under cosine and Euclidean metric are represented. Its values range from 0 to 110 as represented on the Y-axis. The minimum value of this index will be considered as the best performance. From this line graph,in case of 2keyword phrases to 5keyword phrases visual LDA under Euclidean performs well, and rest of keyword phrases of TREC2018 datasets visual PLSI under Cosine metric performs better than other methods and also better than Euclidean distance metric.
(d) Partition Coefficient Index (PCI) Comparative Results
Partition coefficient index (PCI) values lie between 0 and 1. Maximum values will be considered as better performance, values nearer to 1 will be treated as best. From Figure 4(d), based on PCI comparative result values under Cosine and Euclidean metrics,interpret that 2keyword phrases to 6keyword phrases visual PLSI under cosine metricperformance are good, for 8keywords, 10keywords, 14keywords, and 17keyword phrases visual NMF under Euclidean metric and for rest of keyword phrases visual LSI under Cosine metric perform well.
(e) Partition Entropy Index (PEI) Comparative Results Figure 4(e) shows performance comparative values of partition coefficient internal index values which range from 0 to logac. In this case its value range from 0 to 3 as indicated on the Y-axis line graph. The minimum value will be considered for higher performance in clustering. From this graph, infer that for 2 to 4keywords, 6keywords visual PLSI under Cosine, for 5keywors, and 7keywords visual LSI under cosine, and for 8keywords, 10keywords, 14keywords, and 17keywords visual NMF under Euclidean and rest of keywords visual LSI under Euclidean perform better. In the case of this index, the minimum value will perform better clustering results. Form this graph on observation, inferred that visual NMF under cosine metric performs well when compared to the Euclidean metric for all models. Silhouette internal index (SI) values range from -1 to +1. If this index value is nearer to +1 then cluster performance will be best. If values decrease from +1 to -1 its performance also decreases. From the line graph as shown in Figure 6(b) Visual NMF under cosine metric performs well in all TREC2015 keyword phrases than that of Euclidean distance metric. In Figure 6(c) Xie-Beni index (XI) internal validity index value under cosine and Euclidean metric are represented. Its values range from 0 to 300 as represented on the Y-axis. The minimum value of this index will be considered as the best performance. From this line graph results interpreted for 2keyword phrases and 3keyword phrases visual NMF performs well, whereas for 4keyword phrases and 5keyword phrases of TREC2015 visual LDA performs well. In all cases performs is better under cosine based validity index than Euclidean metric based. Partition coefficient index (PCI) values lie between 0 and 1. Maximum values will be considered as better performance, values nearer to 1 will be treated as best. From Figure 6(d), based on PCI comparative result values under Cosine and Euclidean metrics, Visual NMF and Visual LSI both methods values are greater than that of other values under Cosine metric based validity indices. Figure 6(e) shows performance comparative values of partition coefficient internal index values which range from 0 to logac. In this case, its value range from 0 to 1.2 as the number of keywords considered are only four which is indicated on the Y-axis line graph. The minimum value will be considered for higher performance in clustering.From this graph, visual NMF values under cosine metric are greater than that of Euclidean distance-based metrics. Separation Measure internal index value is smaller then it will have greater performance.In this case, its value ranges from 0 to 1 as represented on the Y-axis. From this line graph as shown in Figure 6 In section 4.2 and 4.3 cluster validity indices are calculated based on confusion matrix and the number of clusters. In previous studies also cluster validation done using confusion matrices but not considered the elements in the cluster are well classified or not.In this paper, cluster validation done by considering both confusion matrices and classification metrics to see that elements in cluster are well classified or not. Cluster classification metrics are tabulated for all datasets for all four models under Cosine based and Euclidean metrics. Some sample results are represented from table 4 to table 9
Comparative Results of External validity indices based on Cluster Classification
In this paper, comparative study of external validity indices based on cluster classifications metrics also performed for different hybrid topic models under Cosine based and Euclidean based metrics. Experimental results are tabulated for all types of datasets mentioned in the datasets description section. Sample of comparative results of external validity indices based on cluster classification for 20 keyword phrases of TREC2018 datasets are mentioned in table 6 to table 9. Table 6. From these results inferred that both results are the same under two distance metrics for all twenty clusters. Table 7, comparative results of validity indices are shown under cosine and Euclidean based metrics for Visual LDA model. From these results interpreted Euclidean based metric on average for all clusters perform better than cosine based metric. Table 8. From these results, Euclidean results are better than that of cosine based on accuracy, Macro Average and Weighted Average. | 2020-12-10T09:04:22.999Z | 2020-12-05T00:00:00.000 | {
"year": 2020,
"sha1": "6b2ce1d9f9398c58537414e4e358f0acc9b68db9",
"oa_license": null,
"oa_url": "https://doi.org/10.1088/1757-899x/981/2/022031",
"oa_status": "GOLD",
"pdf_src": "IOP",
"pdf_hash": "58238a0963d7df0ef8e94387927567decbef84fa",
"s2fieldsofstudy": [
"Computer Science",
"Mathematics"
],
"extfieldsofstudy": [
"Physics",
"Computer Science"
]
} |
259088908 | pes2o/s2orc | v3-fos-license | DFormer: Diffusion-guided Transformer for Universal Image Segmentation
This paper introduces an approach, named DFormer, for universal image segmentation. The proposed DFormer views universal image segmentation task as a denoising process using a diffusion model. DFormer first adds various levels of Gaussian noise to ground-truth masks, and then learns a model to predict denoising masks from corrupted masks. Specifically, we take deep pixel-level features along with the noisy masks as inputs to generate mask features and attention masks, employing diffusion-based decoder to perform mask prediction gradually. At inference, our DFormer directly predicts the masks and corresponding categories from a set of randomly-generated masks. Extensive experiments reveal the merits of our proposed contributions on different image segmentation tasks: panoptic segmentation, instance segmentation, and semantic segmentation. Our DFormer outperforms the recent diffusion-based panoptic segmentation method Pix2Seq-D with a gain of 3.6% on MS COCO val2017 set. Further, DFormer achieves promising semantic segmentation performance outperforming the recent diffusion-based method by 2.2% on ADE20K val set. Our source code and models will be publicly on https://github.com/cp3wan/DFormer
Introduction
Image segmentation aims to group the pixels within an image into different units. There are different notations about the grouping, such as semantic categories or instances. Therefore, a variety of image segmentation tasks have been proposed, including semantic segmentation [1], instance segmentation [2], and panoptic segmentation [3]. Semantic segmentation groups the pixels into different semantic categories, whereas instance segmentation groups the pixels into different instances. On the other hand, panoptic segmentation not only groups the pixels of things into different instances, but also segments the stuffs into different semantic categories, which can be seen as a joint task of semantic segmentation and instance segmentation.
In the past years, researchers have proposed specialized architectures for different image segmentation tasks. For instance, different variants of fully-convolutional networks [1,4,5] are used for pixellevel classification in semantic segmentation. Detect-then-segment [2,6,7] and end-to-end [8][9][10] architectures are built for instance segmentation, whereas split-then-merge pipelines [11,12] are designed to combine semantic and instance segmentation for panoptic segmentation. While these specialized methods have achieved great success in each individual segmentation task, they typically struggle to effectively generalize to different image segmentation tasks.
To address the aforementioned issue, few existing works [13][14][15] attempt to develop a universal architecture that performs different image segmentation tasks through a unified framework. Inspired Method Backbone Pan. Inst. Sem. Pix2Seq-D [18] ResNet- 50 Figure 1: Left: Image segmentation results of our DFormer, using the same architecture, on three different tasks: panoptic segmentation (Column 1), instance segmentation (Column 2) and semantic segmentation (Column 3). Right: Comparison with other diffusion model-based segmentation methods on three tasks. 'Pan.' indicates PQ of panoptic segmentation on COCO val2017, 'Inst. ' indicates AP of instance segmentation on COCO val2017, 'Sem.' indicates mIoU of semantic segmentation on ADE20K val. Our DFormer achieves consistent improvements on all three tasks.
by the transformer-based object detector, DETR [16], these methods view image segmentation as an end-to-end set prediction problem. For instance, K-Net [13] employs a group of learnable kernels to dynamically segment instance and stuff, while Mask2Former [17] introduces a masked-attention mask transformer to perform mask classification and mask prediction. Recently, diffusion model-based methods have also been explored for image segmentation. For instance, Chen et al. [18] employ bit diffusion for panoptic segmentation, while Ji et al. [19] propose to concatenate noise and deep features for semantic segmentation. However, the diffusion model-based methods still lag behind existing universal image segmentation methods. In this work, we investigate the problem of designing an effectively diffusion model-based transformer approach that can achieve competitive universal image segmentation performance.
We propose a diffusion-guided transformer framework, named DFormer, for universal image segmentation. Our DFormer views image segmentation as a generative task from noisy masks. During training, we add various level of Gaussian noise to ground-truth masks to obtain noisy masks. Afterwards, we generate the attention mask with binary threshold, and aggregate noisy masks and deep pixel-level features to produce mask features. We then feed them to the transformer decoder to predict the ground-truth masks for each mask feature with masked attention. At inference, we first generate a set of noisy masks and then employ the diffusion-based decoder to predict masks and corresponding object categories.
We conduct experiments on three different image segmentation tasks: semantic segmentation, instance segmentation, and panoptic segmentation. Our experimental results show the merits of the proposed contributions achieving promising performance (see Fig. 1) on three segmentation tasks with a single architecture. With the backbone ResNet50, our DFormer achieves a PQ score of 51.1% on MS COCO val2017 set for panoptic segmentation, thereby outperforming the recent diffusion-based panoptic segmentation method Pix2Seq-D [18] with an absolute gain of 3.6% (see Fig.1). With the backbone Swin-T, our DFormer performs favorably against the recent diffusion-based semantic segmentation method DDP [19] by achieving mIoU score of 48.3% on ADE20K val set for semantic segmentation.
Related works
Semantic segmentation. Semantic segmentation aims to group the pixels within an image into different semantic categories (e.g., person, car, and road). With the advent of deep learning, semantic segmentation has achieved excellent progress. FCN [1] is one of the earliest works for semantic segmentation, which adopts fully convolutional networks for pixel-level classification. Afterwards, the researchers developed different FCN variants for improved semantic segmentation. For instance, some works focus on local contextual information aggregation with encoder-decoder structure [5,21,22] or spatial pyramid structure [23,4]. Further, several works exploit attention mechanism for non-local context aggregation [24][25][26], whereas other works aim at real-time design [27,28].
Instance segmentation. Instance segmentation classifies and segments different object instances (e.g., person and car), and have been addressed through two-stage, single-stage and end-to-end approaches. Two-stage approaches [2,34,35] are usually built on the two-stage object detectors, that first generate some candidate region proposals and then predict object masks within cropped region proposals.
Recently, with the success of end-to-end object detectors, such as DETR [16], several works have explored end-to-end instance segmentation. QueryInst [39] designs a dynamic mask head for querybased instance segmentation. SOLQ [40] builds a unified query representation for instance classes, locations, and masks. SOTR [41] develops a twin attention mechanism into transformer for instance segmentation. In addition, FastInst [42] and SparseInst [43] focus on efficient design.
Panoptic segmentation. Panoptic segmentation [3,11] combines instance segmentation and semantic segmentation into a unified task. At first, several works [44,45,12,46,47] explored to use different branches for two sub-tasks, and design a fusion module to combine the results of these branches for panoptic segmentation. For instance, Panoptic FPN [11] attaches two different head-networks on the shared feature pyramid, and employs a post-processing operation to combine the results of two head-networks. Though these methods have achieved promising results, they are relatively complex. Afterwards, several works [48][49][50][51] explore unifying thing and stuff segmentation into a single network. For instance, Panoptic FCN [48] first generates the kernels for both thing and stuff, and then employs dynamic convolutions for prediction. Panoptic SegFormer [50] introduces a query decoupling strategy, where thing and stuff predictions are based on different queries.
Universal image segmentation. Universal image segmentation aims to unify semantic, instance, and panoptic segmentation tasks into a single network. Inspired by the success of vision transformer for object detection [16] that treats detection as a set prediction problem, researchers [14,13] look into exploring a universal architecture for different image segmentation tasks. K-Net [13] learns a set of learnable kernels to predict both instance and semantic categories. MaskFormer [14] and Mask2Former [17] treat image segmentation as mask classification problem, where they predict mask embeddings and pixel-level embeddings, and use their dot product for mask prediction.
Diffusion models for perception tasks. Diffusion models [52] aim to recover the sample from noise with a parameterized Markov chain, that has achieved promising results on the image generation task. Recently, some works started to explore diffusion models for perception tasks. Chen et al. [53] introduce the diffusion model for object detection, where objects are recovered from noisy boxes with a progressive denoising process. Baranchuk [58] propose to sum the deep features and estimated segmentation map for progressive refinement. Wolleb et al. [59] and Ji et al. [19] propose to use diffusion model for improved segmentation. Chen et al. introduce bit diffusion [60] for discrete data generation and extended bit diffusion for panoptic segmentation [18]. However, we note that these diffusion models-based methods still lag behind universal image segmentation approaches on different image segmentation tasks. In this work, we set out to bridge this performance gap between diffusion models-based perception methods and universal image segmentation techniques.
Preliminaries
Diffusion model. Diffusion models [52] learn a series of state transitions to generate high-quality sample from the noise. During training, a forward diffusion process is first performed by gradually adding Gaussian noise to sample data, and a model is then learned to predict original sample data from noisy sample data. Specifically, the diffusion process generates noisy sample data x t at arbitrary timestep t as where, the timestep t uniformly ranges from 1 to T ,ᾱ t is a monotonically decreasing function from 1 to 0 with the increase of t. When the timestep t is equal to T ,ᾱ T is close to 0, and noisy data x T is close to I. The noise I is sampled with standard normal distribution. Afterwards, a network model is trained to predict sample data x 0 from x t at arbitrary timestep t, where the denoising loss can be written as Figure 2: Overall architecture of our DFormer for universal image segmentation. We employ pixel-level module, including backbone and pixel decoder, to extract multi-scale features. During training, we perform the diffusion process by adding Gaussian noise to ground-truth mask, and employ a transformer decoder to predict mask and corresponding category from the noisy mask.
where, f represents the learned model. During inference, the sample data x 0 is gradually recovered from the noisy data x T with a series of state transitions [52,61], where the learned model is repeatedly employed.
Universal image segmentation. Image segmentation is viewed as mask classification task, which predicts the masks and corresponding categories of existing stuffs and things in an image x, that can be written as M, C = f (x). Recently, the transformer-based image segmentation has achieved promising results, which can be represented as , where, f pixel indicates the pixel-level module for pixel-level feature extraction, f trans indicates the transformer module, and Q indicates the fixed size of learnable queries. In this paper, we aim to re-formulate image segmentation as a denoising process via diffusion model. Instead of using a set of learnable queries, we directly segment image from noisy masks as M, where I represents noise data, and f d represents diffusion-based decoder.
Architecture
Fig . 2 shows the overall architecture of our diffusion-guided transformer, named DFormer, which incorporates diffusion model for universal image segmentation. Following [17], our DFormer comprises a pixel-level module and replaces the standard decoder with the proposed diffusionbased decoder. The pixel-level module generates pyramid features F i p , i = 1, 2, 3 and pixel-level embeddings F pixel , while our diffusion-based decoder takes noisy masks M n , pyramid features F i p , i = 1, 2, 3, and pixel-level embeddings F pixel as inputs and predicts mask embeddings F mask . Finally, we generate the masks by dot product of pixel-level embeddings F pixel and mask embeddings F mask , and predict mask categories by feeding mask embeddings to an MLP layer.
Pixel-level module. Given an input image x ∈ R H×W ×3 , the pixel-level module extracts pyramid where S i represents the stride of feature map to input image, and C is the number of feature channels. Pixel-level module contains a backbone and a pixel decoder. Specifically, we first employ the backbone, such as ResNet [62] or Swin Transformer [63], to extract deep feature of low-resolution F low ∈ R H 64 × W 64 ×C . Afterwards, we use a pixel decoder to progressively upsample deep feature to generate pyramid features of various resolutions F i p , i = 1, 2, 3, 4. The first three successive pyramid features F i p , i = 1, 2, 3 are fed to successive transformer decoder layers to generate mask embeddings, and the last pyramid feature F 4 p is used as pixel-level embeddings F pixel .
Diffusion-based decoder. Based on the generated pyramid features F i p , i = 1, 2, 3 and noisy masks M n , our diffusion-based decoder aims to predict mask embeddings F mask . The diffusion-based decoder stacks L transformer decoder layers as in [17], where each decoder layer contains a maskedattention layer, a self-attention layer, and a FFN layer. The inputs to each decoder layer are attention masks M att , one of pyramid features F i p , and mask features F m . For the first decoder layer, the attention masks M att are generated by thresholding noisy masks as M att = M n > 0, and mask features F m are generated by aggregating pixel-level embeddings F pixel using noisy masks M n as weights, which can be written as where, f avg represents the global averaged pooling operation, and f norm represents the normalization operation. Except the first decoder layer, the output mask embeddings F mask of current decoder layer are used as input mask features F m of next decoder layer, and the predicted masks of current decoder layer are used as input attention masks M att for next decoder layer. The pyramid features F i p , i = 1, 2, 3 are alternately used as input image features of successive decoder layers for efficient training. With output mask embeddings F mask , we employ an MLP layer to predict mask classification scores, and employ dot product with pixel-level embeddings to generate predicted masks.
Training and Inference
During training, we perform the diffusion process to generate noisy masks from ground-truth masks, and train the model to recover ground-truth masks from noisy masks. During inference, we randomly generate noisy masks, and predict masks and categories with learned models.
Training
Algorithm 1 presents the training details of our DFormer, including pixel-level module, mask padding and encoding, mask corruption, diffusion-based decoder, and losses.
Mask padding and encoding. We first perform mask padding since the numbers of objects in different images vary. We pad ground-truth masks M gt to a fixed number of padded masks M pad . Specifically, we generate some random binary masks M rand and concatenate it with original groundtruth masks as M pad = cat(M gt , M rand ), where M pad ∈ R N ×H×W . Afterwards, we explore two different strategies to encode padded masks before adding noise: binary and random shuffling-based encoding strategy. The binary strategy directly encodes padded masks as binary inputs (0 or 1), while the random shuffling-based strategy encodes the padded mask from 0 to 1 where the object mask pixel randomly samples from 0.5 to 1, and the non-object pixel randomly samples from 0 to 0.5.
Mask corruption. Mask corruption aims to generate noisy masks M n at arbitrary timestep t by adding Gaussian noise to encoded masks M enc . We first re-scale the scale of encoded masks from [0,1] to [-b,b], where b is a scale factor. Afterwards, we add the Gaussian noise to encoded masks according to Eq. 1. The scale 1 −ᾱ t of added noise depends on timestep t, where the large t Loss formulation. With the N padded noisy masks, we can predict N masks and corresponding classification scores. Afterwards, we perform bipartite matching between the predictions and groundtruths via Hungarian-based algorithm considering both mask and classification matching cost. After that, we can calculate the overall training losses as where L cls is cross entropy loss for classification, L ce and L mask are binary cross-entropy loss and dice loss for mask prediction. λ 1 , λ 2 , λ 3 indicate the loss weights, which are equal to 2.0, 5.0, 5.0.
Inference
Algorithm 2 presents the inference details of our DFormer. We first generate noisy masks M n with Gaussian distribution, and then predict instance masks with multiple sampling steps. In each sampling step, we first predict instance masks according to noisy masks at previous timestep, and then add noise to the predicted instance masks to generate noisy masks at current timestep.
Datasets and metrics
We adopt the large-scale dataset MS COCO [65] for instance segmentation and panoptic segmentation, and the large-scale dataset ADE20K [66] for semantic segmentation. [65] is one of largest datasets for common vision tasks, such as object detection, instance segmentation, and panoptic segmentation. For object detection and instance segmentation, there are 80 object categories. For panoptic segmentation, there are 80 object (thing) categories and 53 stuff categories (i.e., totally 133 categories). It contains train2017, val2017, test-dev sets, where the train2017 set has 118k images, the val2017 set has 5k images, and the test-dev set has 20k images. We adopt AP, AP 50 , AP 75 , AP S , AP M , AP L as evaluation metrics of instance segmentation [65], and use PQ, PQ Th , PQ St as evaluation metrics of panoptic segmentation.
MS COCO. MS COCO
ADE20k. ADE20K [66] is one of largest datasets for semantic segmentation, which contains 150 semantic categories. The dataset contains train, val, test-dev sets, where the train set has 20k images, the val set has 2k images, and the test set has 3k images. We adopt mIoU as evaluation metric of semantic segmentation.
Implementation details
We adopt the deep model ResNet50 [62] or Swin-T [63] pre-trained on ImageNet as the backbone. The strides of feature maps F 1,2,3,4 p generated by pixel-level decoder are 32, 16, 8, and 4, and the number of feature channels is 256. In diffusion-based decoder, there are L = 9 cascaded decoder layers. During training, each decoder layer is supervised with an auxiliary loss, and the noisy masks are padded to 100. During inference, we only keep the predicted masks by last decoder layer. We adopt post-processing to obtain the expected output format for different segmentation tasks as in [14,17]. The implementation details on different image segmentation tasks are summarized below.
Instance and panoptic segmentation. Our method is trained on four NVIDIA RTX3090 GPUs with the batch size of 12. During training, we adopt AdamW as the optimizer and step learning rate strategy, and there are 50 epochs totally. The initial learning is set as 1 × 10 −4 , and the weigh decay is set as 0.05. The learning rate is decreased at 0.9 and 0.95 fractions of total epochs by a factor of 10.
We adopt large-scale jittering augmentation strategy as Mask2Former [17]. During inference, we adopt the standard single-scale test, where the image is resized with the short edge to 800 pixels and the longer edge up-to 1333 pixels.
Semantic segmentation. Our method is trained on four NVIDIA RTX3090 GPUs with the batch size of 16. During training, we adopt AdamW as the optimizer and poly learning rate schedule, and there are 160k iterations totally. The initial learning is set as 1 × 10 −4 , and the weigh decay is set as 1 × 10 −4 . We employ random scale jitter for data augmentation as Mask2Former [17] . During inference, we resize the images with the short edge to 512 pixels.
State-of-the-art comparison
Panoptic segmentation. We compare our DFormer with state-of-the-art methods for panoptic segmentation on MS COCO val2017 set in Table 1. All the methods use ResNet-50 as the backbone.
Compared to the recent diffusion-based method Pix2Seq-D [18], our DFormer achieves significant improvement while being efficient in terms of parameters. DFormer outperforms Pix2Seq-D with an absolute gain of 3.6% on PQ, 4.4% on PQ T h , and 2.5% on PQ St . In addition, our DFormer has fewer parameters and training epochs. Compared to other state-of-the-art methods, our DFormer achieves a competitive performance. Semantic segmentation. Lastly, we compare our DFormer with several state-of-the-art methods on ADE20K val set [66] in Table 3. We present the results using the backbone ResNet50 and Swin-T. When using the backbone Swin-T, our DFormer outperforms the recent diffusion-based method DDP [19] with an absolute gain of 2.2%. Moreover, our DFormer performs favorably against existing methods in literature. Compared to K-Net [13], DFormer achieves an absolute gain of 2.5% using the same backbone. Further, DFormer also obtains improved performance with AP of 48.3, compared to the recent Mask2Former [17], when using the same Swin-T backbone.
Ablation study
Decoder layers. Table 4a presents the impact of different number of decoder layers. It achieves a mask AP score of 41.1 when the number is equal to 3, whereas it obtains a mask AP score of 42.6 when the number is equal to 9. We therefore adopt 9 decoder layers in our final model.
Timesteps during inference. Table 4b presents the impact of different timesteps during inference. It almost achieves the best performance using 1-step forward process. The multi-step forward process provides a slight improvement in performance.
Different backbones on different tasks. Table 4c shows the impact of different models on different tasks. Swin-T achieves better performance than ResNet-50 on all three segmentation tasks. Number of noisy masks. Table 4d shows the impact of different numbers of noisy masks during inference. The default setting is using 100 which is same as the training. We observe that a larger number of noisy masks does not affect the performance.
Mask encoding and scale strategies. Table 4e shows the impact of different mask encoding and scale strategies. As discussed earlier, we introduce two different mask encoding strategies, including binary and random shuffling-based strategy. When mask scale b is equal to 1, the random shuffling-based encoding achieves a better performance than its binary counterpart. When the mask scale b is equal to 0.1, the random shuffling strategy is slightly inferior than the binary one. Therefore, we adopt the binary strategy with the scale as 0.1.
Qualitative results. Fig. 3 presents example qualitative results of our DFormer on the three image segmentation tasks. Our DFormer provides high-quality segmentation results on these different tasks under the challenging scenarios, such as segmenting occluded objects (e.g., row 2, column 2) and small-scale objects (e.g., row 3, column 4). Additional results are presented in the suppl. material.
Conclusion
We propose a diffusion-guided transformer for universal image segmentation. Our DFormer views universal image segmentation as diffusion-based mask classification. During training, we first perform diffusion process by adding Gaussian noise to ground-truth masks, and then learn a transformer-guided model to predict ground-truth mask from noisy masks. At inference, we directly predict masks and corresponding categories from Gaussian noise. Experiments are performed on panoptic, instance and semantic segmentation tasks to validate the effectiveness of our DFormer. Dformer significantly outperforms existing diffusion-based methods on all three image segmentation tasks.
Limitations and future directions. We observe that our proposed diffusion-guided method is slightly inferior to state-of-the-art methods on small objects on instance segmentation. A potential future direction is to explore improving performance on small objects with diffusion model.
A Visualization of diffusion process Fig. 4 shows visualization of diffusion process by adding Gaussian noise to the ground-truth mask. The total timestep T is equal to 1000. When timestep t is equal to T , the noisy mask will be Gaussian noise. With the decrease of t, the noisy mask become more clear. When timestep t is equal to T , the noisy mask will be the ground-truth mask. Our DFormer presents high-quality segmentation results on different tasks, using a single architecture.
B Extensions on video instance segmentation
We extend DFormer to video instance segmentation by adding a tracking branch. Table 5 shows the results on video instance segmentation. DFormer has also demonstrated competitive performance on video instance segmentation. | 2023-06-07T01:16:10.878Z | 2023-06-06T00:00:00.000 | {
"year": 2023,
"sha1": "74e3b4dd3f12d0eae1de836be288fe19c3f5e2db",
"oa_license": null,
"oa_url": null,
"oa_status": null,
"pdf_src": "Arxiv",
"pdf_hash": "74e3b4dd3f12d0eae1de836be288fe19c3f5e2db",
"s2fieldsofstudy": [
"Computer Science"
],
"extfieldsofstudy": [
"Computer Science"
]
} |
125892222 | pes2o/s2orc | v3-fos-license | Momentum distribution functions of strongly correlated systems of particles: Wigner approach and path integrals
The new numerical version of the Wigner approach to quantum mechanics for treatment thermodynamic properties of the strongly interacting systems of particles has been developed for extreme conditions, when there are no small physical parameters and analytical approximations used in different kind of perturbation theories can not be applied. The new path integral representation of the quantum Wigner function in the phase space has been developed for canonical ensemble. Explicit analytical expression of the Wigner function has been obtained in linear and harmonic approximations. The new quantum Monte-Carlo method for calculations of average values of arbitrary quantum operators has been proposed. Preliminary calculations of the momentum distribution function of the Coulomb systems of particles have been carried out. Comparison with classical Maxwell-Boltzmann distribution shows the significant influence of quantum effects on the high energy asymptotics (“tails’) of the calculated momentum distribution functions, which resulted in appearance of sharp oscillations.
Introduction
Quantum effects can affect the shape of the particle kinetic energy distribution function, as the interaction of a particle with its surroundings restricts the volume of configuration space, which, due to the uncertainty relation, results in an increase in the volume of the momentum space, i.e., in a rise in the fraction of particles with higher momenta. Allowing for quantum effects is important in kinetic consideration of such phenomena as the transition of combustion into detonation, flame propagation, vibrational relaxation, and even thermonuclear fusion at high pressure and low temperatures. Quantum effects are also important in treatment of transport properties of the strongly interacting systems of many particles [1][2][3][4].
The direct way to consider the influence of the quantum effects on the kinetic distribution functions is to use the Wigner formulation of quantum mechanics in phase space [5,6]. The quantum Wigner function is similar to distribution function in classical statistics in phase space. Thereby average values of arbitrary physical quantities can be calculated by formulas, similar to classical statistics. Methods for treatment of the single-particle quantum dynamics in Wigner approach in microcanonical ensemble were proposed in [7][8][9][10][11]. Recently a generalization of these methods to the many-body Wigner equation for distinguishable particles has been done in [12]. This approach is based on Monte Carlo approach for solving the discretized version of the integral Wigner-Liouville equation. Authors validate this approach by studying the timedependent evolution of a system containing only two charged particles (wave-packets) in an external potential. Interesting results have been obtained also for the system of two entangled wave-packets. However comparison with independent results has not been done.
In this work we are going to develop the new numerical version of the Wigner approach to quantum mechanics in the phase space for treatment of thermodynamic properties of the strongly interacting many particle systems under extreme conditions, when there are no small physical parameters and analytical approximations used in different kind of perturbation theories can not be applied [13][14][15]. The new path integral representation of the quantum Wigner function in the phase space has been developed for canonical ensemble. Explicit analytical expression of the Wigner function has been obtained in linear and harmonic approximations. We have developed new quantum Monte-Carlo method for calculations of average values of arbitrary quantum operators. Preliminary calculations of the momentum distribution function for Coulomb system of particles has been carried out. Comparison with classical Maxwell-Boltzmann distribution shows the significant influence of the quantum effects on the high energy asymptotics ("tails') of the calculated distribution functions resulted in appearance of sharp oscillations.
Wigner function for canonical ensemble
The Wigner function W (p, x; t) being the analogue of the classical distribution function in the phase space has a wide range of applications in quantum mechanics. The Wigner function of the many particle system in canonical ensemble is defined as a Fourier transform of the off-diagonal matrix element of density matrix in coordinate representation: Here we are going to obtain new representation of Wigner functions in the path integral form [16][17][18], which allows the numerical simulations of strongly coupled quantum systems of particles in canonical ensemble. Since operators of kinetic and potential energy in Hamiltonian do not commutate, the exact explicit analytical expression for Wigner function does not exist in general. To overcome this difficulty let us represent Wigner function similarly to path integral representation of the partition function [5,16,18]. As example of Coulomb system of particles, we consider a 3D two-component strongly asymmetric electron-hole plasma consisting of N e = N h = N electrons and heavy holes in equilibrium [19]. The hamiltonian of the system H =K +Û c contains kinetic energyK and Coulomb interaction energyÛ c =Û c hh +Û c ee +Û c eh contributions. The thermodynamic properties in the canonical ensemble with given temperature T and fixed volume V are fully described by the diagonal elements of the density operator ρ = e −βĤ /Z with the partition function (normalization constant) where β = 1/k B T , and ρ(x, σ; β) denotes the diagonal matrix elements of the density operator. In equation ( internal energy follows where α = L/L 0 is a length scaling parameter. Of course, the exact density matrix of interacting quantum systems is not known (particularly for low temperatures and high densities), but it can be constructed using a path integral approach [16,18] based on the operator identity e −βĤ = e −ǫĤ · e −ǫĤ . . . e −ǫĤ , where ǫ = β/M , which allows us to rewrite the integral in equation ( The spin gives rise to the spin part of the density matrix (S) with exchange effects accounted for by the permutation operatorsP e andP h acting on the electron and hole coordinates x M and spin projections σ ′ . The sum is over all permutations with parity κ Pe and κ P h . In equation (6) . In this paper we are going to consider strongly coupled non-degenerate plasma, so the main contribution in the sum over permutations comes from the identical permutations and each particle is represented by a closed trajectory consisting of a set of M coordinates ("beads"). So the whole configuration of the particles is represented by Figure 1 illustrates the representation of one (light) electron and one (heavy) hole. The circle around the electron beads symbolizes the region that mainly contributes to the partition function path integral. The size of this region is of the order of the thermal electron wavelength λ e (T ), while typical distances between electron beads are of the order of the electron wavelength taken at an M -times higher temperature. The same representation is valid for each hole but it is not shown since, due to the larger hole mass, the characteristic length scales are substantially smaller. However, in the simulations discussed below the holes are treated according to the full beads representation.
2.1. High-temperature asymptotics for the density matrix. Kelbg potential In this section we discuss approximations for the high-temperature density matrix which can be used for efficient direct path integral Monte Carlo (PIMC) simulations. It involves effective quantum pair potentials Φ ab , which are approximated by the Kelbg potential. Here, we closely follow earlier work [20] where details and further references can be found.
2.1.1. Pair approximation and Kelbg potential The N-particle high-temperature density matrix is expressed in terms of two-particle density matrices (higher order terms become negligible at sufficiently high temperature, i.e. for large number M of time slices) given by This results from factorization into kinetic and interaction parts, ρ ab ≈ ρ K 0 ρ U ab , which is exact in the classical case, i.e. at sufficiently high temperature. The error made at finite temperature vanishes with the number of time slices as M −2 [20]. The off-diagonal density matrix element (6) involves an effective pair interaction which is expressed approximately via its diagonal elements, , for which we use the familiar Kelbg potential [21,22] where is the reduced mass of the (ab)-pair of particles and the error function is defined by erf(x) = 2 √ π x 0 dte −t 2 . Note that the Kelbg potential is finite at zero distance which is a consequence of quantum effects. The validity of this potential as well as its the off-diagonal approximation is restricted to temperatures substantially higher than the exciton binding energy [23,24] which puts another lower bound on the number of time slices n. For a discussion of other effective potentials, we refer to [20,[23][24][25].
Summarizing the above approximations, we can conclude that with the approximations (6,7) each of the high-temperature factors on the r.h.s. of equation (6), carries an error of the order 1/M 2 . Within these approximations, we obtain the result where ρ (m) 0 is the kinetic density matrix, and U denotes the sum of all interaction energies, each consisting of the respective sum of pair interactions given by Kelbg potentials
Path integral representation of Wigner function
As in the previous section we represent statistical operator e −βĤ as product of large number M operators e −ǫĤ formally related to high temperature (ǫ = β/M ), so we can write Wigner function in form of multiple integral of the product of the high temperature density matrices (see equation (1)): For simplicity we have omitted spin variable. As before making use of the Kelbg potential the explicit expression of the high temperature density matrices can be calculated with accuracy up to O(ǫ 2 ) in the limit ǫ → 0 [18]: Here angle brackets means scalar product of two vectors, while here and further the symbol λ −2 ǫ a|b means that each term in scalar product a|b is normalized by the electron or hole thermal wavelength respectively and λ ǫ = 2π 2 ǫ m is taken at high temperature M · T . Now replacing variables of integration: we obtain Here C(M ) = M −M/2 depends only on number M . As it will be shown further, this constant is canceled in calculations of average values of operators. It is convenient to use dimensionless variablesξ andx 1 , . . . ,x M −1 : ξ = λξ, x m = λx m . Using designationǭ = 1/M , we obtain the next expression for Wigner function: Whenǭ → 0 this multiple integral turns to exact representation of the Wigner function W (p, x; β) in the form of path integral with continuous dimensionless 'time' τ , which corresponds toǭm in discrete case. Also set of independent variablesx m turns into closed trajectoryx(τ ). This trajectory starts and ends in point x/λ when τ = 0 and 1. In this limit expressionx m+1 −xm ǫ is nothing else than derivative dx dτ , which we denote asẋ(τ ). As a result, we have a new exact representation of Wigner function for canonical ensemble in form of path integral: where symbol |λa means that in vector |a each component is multiplied on related electron or hole thermal wave length λ = 2π 2 β m . Let us note that integration here relates to the integration over the Wiener measure of all closed trajectoriesx(τ ). Variableξ is integrand in Fourier transform. In fact, a particle is presented by the closed trajectory with characteristic size of order λ in coordinate space. This is manifestation of the uncertainty principle.
Harmonic approximation for Wigner function
The expression for Wigner function (13) is inconvenient practically because the momentum p is connected with other variables only through Fourier transform. Even in case of free particle (V (x) = 0) this expression does not explicitly contain the analytic expression of Maxwell-Boltzmann distribution. To obtain it explicitly, we have to integrate over variableξ. In general case this integral can not be calculated analytically. Exclusions are the linear or harmonic potentials, when power of variableξ is not more than two.
To do this integration analytically and to obtain explicit expression for Wigner function let us take the approximation for potential V (x) arising from the Taylor expansion up to the first or second order in theξ: Here the second term means scalar product of the vector proportional to ξ and the multidimensional gradient of pseudopotental energy, while third term means quadratic form with the matrix of the second derivatives. If we take into account all terms in expansion (14), the expression for Wigner function (13) takes the form of gaussian integral over variableξ and the final expression for Wigner function in harmonic approximation can be written in the following form: Wigner function in this expression is determined by integral over all closed trajectories z(τ ), which start and end in the origin x, as we have split trajectories on the "external" coordinate x and "internal" coordinates z(τ ) by the substitutionx(τ ) = x/λ + z(τ ). Here scalar, vector and matrix functionals L, J and χ are Note that the first term in exponent (15) looks explicitly like Maxwell distribution in classical statistics. The difference is in the matrix χ[x, z], which is reduced to unit matrix for free particle. Matrix χ[x, z] is also reduced to unit matrix in linear approximation accounting for only the first two terms in expansion (14).
Interpretation and applicability of harmonic approximation
The expression for Wigner function (15) is obtained under assumption, that potential V (x + λz(τ ) + λξ(τ − 1/2)) is expanded in Taylor series up to the second orderξ with good accuracy. In the same time we neglected terms with higher power ofξ. Thereby it is necessary to discuss the legality of such assumption. Primarily, note that variableξ is in three positions in the exponent (13). The first one is Fourier term i p|ξ / , which makes momentum correlated with other dynamical variables. The second one is gaussian term π ξ |ξ . The third one is integral term, whereξ is argument of potential 1 0 dτ βV (x + λz(τ )+ λξ(τ − 1/2)). Gaussian term provides rapid decay with increasingξ, so the main contribution comes from the region less or of order π −1/2 ≈ 0.6. Moreover the argument of potential function containsξ multiplied on factor τ −1/2, which is modulo less than 0.5. So using the mean value theorem, we can roughly estimate the high order terms of Taylor expansion in exponent (13) as where x 0 is certain point of trajectory. Numerical value of this integral rapidly decreases: when n = 2, 4, 6 it equals 1/24, 1/1920 and 1/322560. Thus we expect negligible contribution of high order Taylor terms in potential expansion. Numerical test calculations for a lot of potentials confirm this hypothesis [26]. Note that the first summand in exponent (15) is similar to Maxwell distribution in classical statistics. The difference is in coefficient χ −1 [x, z].
Average values of quantum operators
Average value of arbitrary quantum operator can be written as Weyl's symbol A(p, x), averaged over phase space with the Wigner function W (p, x; β) as weight: where the Weyl's symbol of operator is: Weyl's symbols for usual operators likep,x,p 2 ,x 2 ,Ĥ,Ĥ 2 etc. can be easily calculated directly from definition (19). For calculation of  we are going to use the harmonic approximation of the Wigner function in path integral representation (15). For making use of the Monte Carlo method (MC) [27,28] we have to represent path integrals in discrete form of multiple integral. As a result we obtain final expressions for MC calculations in the following form: Here brackets g(p, x, z 1 , . . . , z M −1 ) w denote averaging of any function g(p, x, z 1 , . . . , z M −1 ) with positive weight w(p, x, z 1 , . . . , z M −1 ): Note that denominator in (20) is equal to nominator with A(p, x) = 1, so C(M ) from (12) is canceled.
Results of numerical calculations
To obtain preliminary results and to test the developed approach we have carried out calculations of the path integral representation of Wigner function in the framework of the mentioned above linear approximation. Calculations have been done for two hundred particles each presented by twenty beads. Results have been obtained by averaging-out over one million particle configurations.
Calculations have been done for ideal and strongly coupled electron-hole plasmas with effective hole mass equal to 100m e at temperature T = 0.3Ha and density related to the Wigner-Seitz radius equal to two (r s = 2). Related coupling parameter is equal to 1.66.
We define the momentum distribution function for holes (a = h) and electrons (a = e) by the following expression: with normalization factor N equal to: N = Z(β) = V dpdx W (p, x; β) . The middle and bottom panels show the influence of the strong interparticle interaction on momentum distribution functions resulted in appearance of sharp oscillations. In the large momentum tails of electron distributions these sharp oscillations of one order of magnitude arise due to the interference of quantum effects, which are accounted for by the "cosine factor' in equation (15). For small momentums the "cosine factor' is close to unity and MC distribution functions approach practically the Maxwell-Boltzmann distribution.
On the whole presented normalized to unity distribution functions g a (|p a |) resemble the related Maxwell-Boltzmann function. However to conserve normalization factor the sharp peaks of the MC distribution functions have to be compensated by the pits below Maxwell-Boltzmann function.
The physical reasons of the around ten percent oscillations of the distribution functions in vicinity of zero momentum have to be investigated in later publications.
Conclusion
In our work we have derived the new path integral representation of Wigner function of the Coulomb system of particles for canonical ensemble. We have obtained explicit expression of Wigner function in linear and harmonic approximation resembling the Maxwell-Boltzmann distribution on momentum variables, but with quantum corrections depending on the second derivatives of interparticle potential. This approximation contains also the oscillatory multiplier describing quantum interference. We have developed new quantum Monte-Carlo method for calculations of average values of arbitrary quantum operators. Preliminary calculations of the momentum distribution function for non-ideal plasma has been done. Comparison with classical Maxwell-Boltzmann distribution shows the significant influence of the quantum effects on the high energy asymptotics ("tails') of the calculated momentum distribution functions resulted in appearance of sharp large scale oscillations. | 2019-04-22T13:04:06.740Z | 2016-11-01T00:00:00.000 | {
"year": 2016,
"sha1": "3620d83998b7abac59b2bf276ddb154f05910969",
"oa_license": null,
"oa_url": "https://doi.org/10.1088/1742-6596/774/1/012146",
"oa_status": "GOLD",
"pdf_src": "IOP",
"pdf_hash": "f1ad9ca74595eb1680930316b52404f9ab903712",
"s2fieldsofstudy": [
"Physics"
],
"extfieldsofstudy": [
"Mathematics",
"Physics"
]
} |
57189343 | pes2o/s2orc | v3-fos-license | Inositols in Insulin Signaling and Glucose Metabolism
In the past decades, both the importance of inositol for human health and the complex interaction between glucose and inositol have been the subject of increasing consideration. Glucose has been shown to interfere with cellular transmembrane transport of inositol, inhibiting, among others, its intestinal absorption. Moreover, intracellular glucose is required for de novo biosynthesis of inositol through the inositol-3-phosphate synthase 1 pathway, while a few glucose-related metabolites, like sorbitol, reduce intracellular levels of inositol. Furthermore, inositol, via its major isomers myo-inositol and D-chiro-inositol, and probably some of its phosphate intermediate metabolites and correlated enzymes (like inositol hexakisphosphate kinase) participate in both insulin signaling and glucose metabolism by influencing distinct pathways. Indeed, clinical data support the beneficial effects exerted by inositol by reducing glycaemia levels and hyperinsulinemia and buffering negative effects of sustained insulin stimulation upon the adipose tissue and the endocrine system. Due to these multiple effects, myoIns has become a reliable treatment option, as opposed to hormonal stimulation, for insulin-resistant PCOS patients.
Inositol and Insulin Signaling System
It is currently well known that diets providing high amounts of fibers have a protective role in the management (and prevention) of chronic diseases-namely those characterized by deregulated glucose and insulin metabolism [1,2]. While consumption of highly refined diets is frequently associated with increased risk for diabetes and cancer [3], the effect of high-fiber diets is usually ascribed to the presence of whole-grain cereal products and of a plethora of phytochemicals, which display an astonishing range of different biochemical-pharmacological effects, well documented in in vitro and in vivo studies [4]. Among the aforementioned constituents, inositol appears to occupy a prominent position. Indeed, several investigations have demonstrated that only fibers with high content in phytic acid or phytate (inositol hexakisphosphate, InsP6) and its derivative myo-inositol (myoIns) show negative correlation with colon cancer and diabetes, indicating that inositolderived compounds can suppress colon carcinogenesis [5] or improve glucose metabolism [6].
Indeed, several alterations in inositol metabolism have been observed in insulin-resistant and diabetic patients ( [7], reviewed in [8]). Namely, the decreased availability of inositol or inositol phosphoglycans (IPGs) can be primarily ascribed to increased urinary loss of myoIns in both animals and humans with insulin resistance. This effect is mostly due to the glucose-mediated inhibition of myoIns reabsorption by the kidney. Reduction in myoIns availability has a direct negative impact on the levels of its D-chiro-inositol (DCIns) isomer. In turn, plasma and intracellular depletion of myoIns and its isomers or IPG metabolites is likely to worsen insulin resistance. In fact, IPGs and probably myoIns itself are involved in the insulin signaling machinery. Conversely, altered profiles of plasma/urinary levels of inositol and its metabolites/ isomers, including DCIns, have been observed in diabetes, insulin resistance, or metabolic syndrome.
It is currently acknowledged that some of the effects ascribed to inositol can be explained by the ability of this molecule or its metabolites to modulate glucose metabolism, by transducing insulin effects downstream of insulinreceptor (IR) coupling. Yet, the complex interplay among insulin, glucose metabolism and inositol-related molecules is still poorly understood and only a few surveys have been hitherto performed to address such issue.
Inositol Availability and Dietary Supply
Inositol is chemically identified as hexahydroxycyclohexane and consists of a family of 9 stereoisomers. MyoIns is the most widely distributed representative of this family in nature, including animals and mammals [15]. A normal diet provides inositols, mainly present in cereals and legumes, as myoIns, for the most part, phosphatidylinositol (PI) and InsP6 [16]. Inositol has been shown to be essential for cell growth and has been considered for a long time as a component of the vitamin B family [17,18]. However, we now know that the human body (especially the liver and brain) [19] can produce up to 4 g/day inositol, and that assumption is no longer recognized.
Indirect estimates of the availability of myoIns in human diet, based on the consumption of phytate-rich food, reveal daily intakes not exceeding 500-700 mg/day for western countries, with higher consumption rates in Africa and Asia. In Italy, for example, available estimates of the mean value of phytic acid intake range from 219 to 293 mg/day [20,21]. In the USA and Canada, average phytic acid intake is of 538 mg/ day in adults, with relevant differences between males and females (608 mg vs 512 mg/day, respectively) [22], and of 170-390 mg/day in children [23].
Glucose and Inositol Biosynthesis
As anticipated, myoIns is actively synthetized by living cells. We do not know if this biosynthetic activity makes our body independent of dietary inositol supply, but an idea of how relevant inositol biosynthesis can be in some tissues comes from observations on the brain, which produces myoIns at concentrations reaching 10-to15-fold the values measured in circulating blood, with only limited "extracting" ability [24].
MyoIns biosynthesis varies among tissues depending on changing functional requirements. In yeast, Ino1 and IMPA-1 are induced by low inositol levels [26] and downregulated by high inositol levels [27]. Both Ino1 and Isyna1 can be epigenetically modified [28] and mammalian Isyna1 exhibits gender-and tissue-specific DNA methylation patterns [29] that modulate inositol biosynthesis.
In mammalian cells, however, intracellular myoIns availability seems ineffective in controlling Isyna1 activity [30], the expression of which is downregulated by inositol pyrophosphates (IP7), produced by inositol hexakisphosphate kinase (IP6K1) [31]. IP7 has been found to inhibit transcription of Isyna1 by increasing its methylation levels [32]. A key role in this pathway is played by phosphatidic acid (PA), which increases nuclear translocation of IP6K1, thus finally leading to decreased myoIns content. PA increases when higher energy demand imposes high rates of glucose catabolism [33], being synthesized from two glycolytic intermediates-dihydroxyacetone phosphate and glycerol-3-phosphate [34]. Furthermore, high glucose levels indirectly increase the activity of phospholipase D (PLD), a key enzyme for PA synthesis [35].
In mammalian cells, a positive regulator of Isyna1 transcription is glycogen synthase kinase 3 (GSK3), since loss of GSK3 activity causes myoIns depletion [36]. Furthermore, synthesis of inositol is positively regulated by Mck1, a GSK3 homolog required for normal activity of MIPS1, the rate-limiting enzyme in inositol synthesis [37].
Influence of Glucose on Inositol Uptake and Metabolism
Long ago, a negative association observed between phytic acid intake and the glycemic index of cereals and legumes consumed by humans led to a hypothesis of a correlation among inositol, phytic acid, and glucose metabolism [38]. This was supported by further studies, which revealed that removal of phytate from bean flour increases its glycemic index [39] and that myoIns significantly inhibits duodenal glucose absorption and reduces blood glucose rise, suggesting the existence of a competitive affinity for the same transporter system [40]. It now appears that a complex relationship between glucose and myoIns does exist, and glucose can interfere with myoIns metabolism at different levels.
In the first place, glucose significantly counteracts cellular uptake of inositol. Inositol is transported intracellularly by sodium ion-coupled transporters (SMIT1 and SMIT2) [41], which are posttranslationally regulated through phosphorylation by both protein kinase A and protein kinase C [42], and a proton-coupled transporter (HMIT1) [43].
In hepatocytes, inhibitors of the main sodium-glucose transporters (SGLT) 1/2 prevent both glucose and inositol uptake [44], suggesting that the two molecules share the transporter system(s). Inositol uptake is also decreased by glucose. MyoIns depletion in nervous tissues induced by hyperglycemia depends on competitive inhibition of sodium ion-coupled myoIns transporters [45]. Inositol uptake in cultured cells is also significantly inhibited by 20 mM glucose [46].
Secondly, glucose may induce myoIns depletion by activation of the glucose-sorbitol pathway, in which glucose is first converted to sorbitol by aldose reductase and then to fructose by sorbitol dehydrogenase, as observed in diabetic patients [47]. The increased production of sorbitol produces a harmful rise in intracellular osmolarity, an event counteracted by inhibition of the uptake of other relevant osmolytes, including inositol, via downregulated expression of their carriers [48]. Inhibiting aldose reductase in cultured cells restores myoIns levels counteracting the depleting effect of sorbitol [49].
Glucose-dependent depletion of normal levels of intracellular inositol is further confirmed by observations obtained on specific tissues, especially in those susceptible of developing diabetes complications, of hyperglycemic patients [50]. In addition, both hyperglycemia and insulin resistance have been found to modify the relative ratio in which different inositol isomers, especially myoIns and DCIns (approximately 100 : 1), are present in these tissues [51]. Changes in the ratio of plasma and urinary myoIns/ DCIns levels are so tightly linked to insulin abnormalities to be considered an early marker of hyperglycemia and insulin resistance [7]. This worsens insulin resistance and diabetes complications, impairs cellular redox and free radical defenses, and increases oxidative glycation stress [52,53]. Under these circumstances, myoIns supplementation reduces several diabetes symptoms and metabolic markers of different pathologies [54].
Both types of IPGs have been shown to exert an insulinmimetic activity, acting as second messengers downstream of insulin receptors [56]. When administered to normal or diabetic rats, they reduce hyperglycemia in a dose-dependent fashion and promote muscular glycogenesis [59]. Insulin stimulates glycosyl-phosphatidylinositol (GPI) hydrolysis through direct activation of phospholipase C (PLC) and PLD, with production of water-soluble IPSs [60,61]. Inositol phosphoglycans also inhibit AC and protein kinase A with antilipolytic and lipogenic effects [62].
Upon insulin receptor activation, IPGs are released outside the cell and reimported by an ATP-dependent inositol glycan transporter [63], activating cytosolic phosphoprotein phosphatase 2C-α (PP2Cα) and mitochondrial PDH. This enhances oxidative glucose metabolism along the tricarboxylic acid cycle. Activated PP2Cα stimulates glycogen synthase (GS) both directly and indirectly, through the phosphatidylinositide 3-kinase (PI3K)/Akt pathway [64]. Activation of Akt inactivates glycogen synthase kinase-3 (GSK-3), enhancing GS activity, and increases GLUT-4 translocation and glucose uptake. Finally, as observed in the rat, IPG-P inhibits the glucose-stimulated release of insulin from pancreatic β-cells, suggesting the existence of a negative feedback mechanism in this pathway [65]. On these bases, IPGs are supposed to exert a general antidiabetic function.
The proper ratio at which IPG-A and IPG-P display the most beneficial effects on insulin/glucose homeostasis is still unclear, besides the fact that their contemporary presence appears mandatory.
Hypotheses for Future Research
We would hypothesize that DCIns containing IPG is preferentially synthetized during metabolic stress, in response to increased insulin release. Indeed, upon insulin stimulation, myoIns is converted by a specific epimerase into its DCIns isomer. Epimerase activity is strictly dependent on insulin, and shows a tissue-dependent modulation [66]. Inositol isomerization occurs at the phospholipid level where ( 3 H)-myoIns phospholipid are converted into ( 3 H)-DCIns as soon as 15 minutes after insulin stimulation [67]. However, we cannot discard the possibility that epimerization also occurs at the free inositol level, but with the subsequent rapid incorporation of DCIns into phospholipids, as observed both in vitro and in vivo [68,69]. Conversely, myoIns epimerization is severely impaired when insulin-sensitive tissues (muscle, fat, and liver) become insulin resistant, and a reduced DCIns/myoIns ratio has been suggested to represent a measure of insulin resistance [70]. Accordingly, reduced DCIns levels are typical of urine and muscle tissue of type 2 diabetic patients [71,72]. Additionally, insulin resistance in both type 2 diabetic subjects with impaired glucose tolerance and healthy controls appears to be related linearly with urinary deficiency of DCIns [73]. Yet, insulin stimulates IPG-P release in normal subjects but not in insulin-resistant women with polycystic ovary syndrome (PCOS) [74]. These observations may be the result of a deficiency in membrane-bound IPG and/or insulin resistance with reduction of epimerase-dependent conversion of myoIns into DCIns.
Intriguingly, in both insulin-resistant animals and humans, the decreased availability of inositol or IPGs is associated with the increased urinary loss of myoIns [75], probably depending on glucose-mediated inhibition of renal myoIns reabsorption [76]. Ultimately, a shortage of myoIns impacts negatively on DCIns levels [77], worsening insulin resistance. Indeed, plasma and tissue levels of DCIns and IPGs are decreased in PCOS patients with insulin resistance and type 2 diabetic patients [76]. However, treatment with myoIns has not been found to lower insulin and glycaemia in all patients with metabolic syndrome [78], strongly suggesting that sensitivity to inositol supplementation of insulin-resistant subjects may depend on still unknown additional factors.
This picture is further complicated by an intriguing paradox that emerged in recent studies on PCOS patients displaying insulin resistance. While insulin-resistant tissues do suffer from a general paucity of DCIns due to the reduced epimerization of myoIns, organs and tissues that retain their insulin sensitivity, like the ovary [79], display increased levels of DCIns and reduced levels of myoIns, due to the enhanced epimerization of myoIns [80]. The impairment of ovarian function in PCOS patients has been specifically associated with the increase in the DCIns/myoIns ratio within the ovary [81].
Noticeably, the role of inositol-containing phosphoglycans in insulin regulation has been recently questioned by evidencing that several synthetic IPGs, opposite to natural IPGs, are deprived of insulin-mimetic activity [82]. Since no convincing explanation of such enigma has been proposed, it can be hypothesized that synthetic IPGs lack some critical cofactors required for the insulin-mimetic activity.
Other complications of insulin-inositol relationships come from the demonstration that in response to insulin stimulation of normal cells, myoIns may directly promote activation of insulin receptor substrate (IRS) and Akt [83]. In addition, myoIns and other inositol derivatives enhance GLUT-4 translocation through the cell membrane independently of insulin stimulation [84] or glucose uptake [85].
Overactivation of IP6K1, usually triggered by insulin stimulation, has been recently found to promote synthesis of IP7, which inhibits Akt activity by preventing its interaction with PI3K [86], thus reducing insulin sensitivity and protein synthesis via the GSK3β and mTOR signaling pathways, which are both associated with insulin resistance and weight gain [87]. On the contrary, IP6K1 knockout mice manifest insulin sensitivity and are resistant to obesity elicited by high-fat diet or aging.
To properly address such issues, we argue that a preliminary effort should be made by performing a thorough investigation of intracellular myoIns metabolism. Indeed, once uptaken by cells, myoIns undergoes several transformations into key active metabolites and macromolecular complexes, such as phosphoinositides and phospholipids. To date, despite widespread studies carried out to identify inositolbased effects on biological pathways, no comprehensive metabolomic analysis has been performed so far. Only an integrated metabolomic-genomic study would indeed provide the basic information required to identify the cellular fate of myoIns and its gene/protein targets.
Conclusions
Available data suggest that inositol, through its isomers (myoIns and DCIns) and probably some of its phosphate intermediate metabolites and correlated enzymes (like IP6K1), participates in insulin signaling and glucose metabolism, by influencing distinct pathways. Indeed, some preliminary clinical reports provided compelling evidence supporting the beneficial effects obtained with inositol in insulin-resistant patients. Inositol reduces glycemia levels and hyperinsulinemia, while buffering negative effects of sustained insulin stimulation upon the adipose tissue and the endocrine system [14,88]. As such, myoIns has become a reliable treatment option for insulin-resistant PCOS patients [89,90].
However, we are just unveiling the complex interplay between inositol(s) and insulin-related pathways. Namely, it is still unclear if insulin resistance depends (at least in part) on a reduced content of membrane-bound IPGs, on a primary defect in IPG release, or if impaired epimerase activation downstream of insulin stimulation would ultimately lead to reduced DCIns intracellular availability.
Anyhow, compelling evidence suggests that reduced DCIns levels are associated with impaired insulin transduction in insulin-sensitive tissues, where insulin-sensitive epimerase is unable to convert myoIns into DCIns. On the contrary, in insulin-insensitive tissues (like the ovary and placenta), insulin overstimulation induces an anomalous increased conversion of myoIns into DCIns [91]. The increased availability of DCIns in these tissues leads to several biological dysfunctions. Indeed, experimental evidence suggests that high DCIns levels in ovary and placenta cells can trigger paradoxical effects on insulin signaling. In preeclampsia-suffering women, an abnormal intracellular amount of DCIns within the placenta promotes a specific phosphorylation of insulin receptor substrate 1 (IRS-1) on serine 312 and leads to inhibition of the PI3K/Akt pathway [92]. Akt inhibition worsens insulin resistance and hampers several pathways downstream of insulin stimulation, as observed in PCOS women whose treatment with high doses of DCIns was shown to be detrimental [93]. Interestingly, these results are confirmed by analysis of an experimental PCOS mouse model in which administration of myoIns and DCIns at ratios of 5 : 1 and 20 : 1 is detrimental for ovarian morphology and function, whereas their administration at ratios of 40 : 1 and at a lesser extent 80 : 1 rapidly restores physiological ovarian parameters [94].
To sum up, the interplay among the different isomers of inositol and their glycan derivatives (IPG-A and IPG-P) is still unclear and in depth studies are warranted to fully appreciate the contribution of inositol and inositol glycans in insulin signaling.
Conflicts of Interest
Authors declare that there are no competing interests regarding the publication of this paper. | 2019-01-22T22:23:51.230Z | 2018-11-25T00:00:00.000 | {
"year": 2018,
"sha1": "7ad9ba63b66885571e39a2cca7e347d3a8787e64",
"oa_license": "CCBY",
"oa_url": "https://doi.org/10.1155/2018/1968450",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "7c5e8a770b7ef3d5a22e82b2abdfd82789c5114b",
"s2fieldsofstudy": [
"Biology",
"Computer Science"
],
"extfieldsofstudy": [
"Medicine"
]
} |
16443665 | pes2o/s2orc | v3-fos-license | A PARTIAL TEST AND DEVELOPMENT OF DELONE AND MCLEAN'S MODEL OF IS SUCCESS 3
DeLone and McLean's (1992) comprehensive review of different information system success measures concludes with a model of interrelationships between six IS Success constructs. This paper critically examines the meaning of four of these constructs and the evidence of relationships between them. It then provides results from empirical tests of these relationships. Tests are conducted using both conventional ordinary least squares regression path analysis and structural equation modeling - with substantially similar results. The empirical results provide substantial support for the "up stream" two thirds of DeLone and McLean's model. Three factors. System Quality, Information Quality, and Usefulness, are found to explain 75% of the variance in the overall User Satisfaction measure. The empirical results also provide substantial support for the use of Usefulness as an IS Success measure, and of the hitherto-unreported importance of "Importance of the task" in user perceptions of IS Usefulness.
INTRODUCTION
DeLone and McLean's (1992) comprehensive review of different information system success measures makes two important contributions to our understanding of Information System (IS) success.First, it provides a scheme for classifying the multitude of IS success measures that have been used in the literature into six categories.Second, it suggests a model of interdependencies between these categories.Commenting on their Figure 2 The l/S success model proposed in Figure 2 is an attempt to reflect the interdependent, process nature of I/S success.Rather than six independent success categories, there are six interdependent dimensions to US success.This success model clearly needs further development and validation before it could serve as a basis for the selection of appropriate I/S measures.
The last sentence in the above quotation provides the motivation for this paper.We do not address the two right-hand constructs4 ' but we do examine the meaning of, and relationships between, the other four.Usefulness, and Importance of the system has been added.) The meaning of the five constructs used in Figure 2, and the theory supporting the relationships between them, are discussed in the second section of this paper (the next section).To test the model, data were collected from 104 users of a university's departmental accounting system.Discussion of methodology used for testing the model, measurement of variables, and test results are presented in the third section of this paper.The model is analyzed using both conventional ordinary least squares (OLS) linear regression and using a LISREL-like structural equation modeling (Bollen 1989;Joreskog and Sorbom 1988) package called Amos 5 (Arbuckle, 1994a).Finally, the paper concludes with a brief discussion of the implications of these results for the measurement of IS success in future research.
Our conclusions are that (a) there is substantial support for the up-stream side of DeLone and McLean's model, (b) of the four measures examined, User Satisfaction is the most general individual-user perceptual measure of information system success, and (c) future researchers will need to control for Task Importance whenever they measure the Usefulness of an information system (because systems that perform more important tasks are perceived as more useful, irrespective of the quality of the actual system).
DEFINING CONSTRUCTS AND THEORIZING ABOUT THEIR RELATIONSHIPS
Replacing Use in Figure 1 by Usefulness in Figure 2 As reported by DeLone and McLean (1992), many researchers have used Use as an objective measure of system success.The implication is that if a system is used, it must be useful, and therefore successful.However, non-use does not necessarily mean a system is not useful, it may simply mean that the potential user has other more pressing things to be done.In addition, as DeLone and McLean point out, "usage, either Amos stands for "Analysis of Moment Structures".perceived or actual, is only pertinent when such use is voluntary" (p.68).When usage is compulsory, the number of hours a system is used conveys little information about system usefulness, and so success.
Reflecting on the relevance of Use as an indicator of system success in some situations and its irrelevance in others, we conjectured that the underlying success construct that researchers have been trying to tap is Usefulness, not Use.Use is a good proxy for Usefulness in situations where a tool is used, and use is not mandatory.It then provides a simple objective measure of success.However, in cases where a system is not used during the period of study, or where usage is mandatory (as in the accounting system examined in this study), we argue that Usefulness continues to be a meaningful measure of success, even though Use does not.We therefore decided to measure Usefulness, not Use, in our test of DeLone and McLean's model (Figure 2)6 .
The First Four Hypotheses in Figure 2 (H1-H4) The first four (independent) hypotheses to be tested in this paper follow directly from There are four constructs involved.Information Quality is concerned with such issues as the timeliness, accuracy, relevance, and format of information generated by an information system.System Quality is concerned with whether or not there are "bugs" in the system, the consistency of the user interface, ease of use, response rates in interactive systems, documentation, and sometimes, quality and maintainability of the program code.Usefulness of an IS is "the degree to which a person believes that using a particular system would enhance his or her job performance" (Davis 1989).Satisfaction is "the result of the individual taking outcomes that have been received and evaluating them on a pleasant-unpleasant continuum" (Naylor et al. 1980).
There is quite strong support in the literature, both theoretical and empirical, for hypotheses H1-H4.A brief summary of seven relevant studies follows.
First, based on the work of Bailey and Pearson (1983), Ives et al. (1983) discuss the development of two instruments for measuring User Satisfaction.The longer of these instruments was factor analyzed into five factors: EDP Staff and Services, Information Product 1, Vendor Support, Information Product 2, and Knowledge or Participation 1 .The shorter instrument -later validated by Baroudi and Orlikowski (1988) factor analyzed into three factors: EDP Staff and Services, Information Product, and User Knowledge or Participation.A number of researchers have been critical of the use of these multi-factor measures of user satisfaction.Treacy (1985) described the factors as "imprecise and ambiguous".Galletta and Lederer (1989) argued that because of the heterogeneity of the items, results from the Ives et al. (1983) instruments should be "interpreted with caution".Our explanation of the factors that emerge from the Ives et al. instruments is that they measure the independent variables that Bailey and Pearson (1983) and Ives et al. (1983) thought were likely to cause Satisfaction, not User Satisfaction itself.In particular, we view the presence of so many questions about information quality in the Ives et al.User Satisfaction instruments as providing theoretical support for hypothesis 1 in Figure 2. Second, in a recent empirical study, Seddon and Yip (1992) These four questions (as they apply to DAS, the system for which data were collected for this study) are reproduced in Appendix 2 to this paper, Section C. 9 Such high correlations of multi-factor measures and overall satisfaction measures are not uncommon.Bailey and Pearson (1983, p.536) report a correlation of 0.79 between their normalized importance-weighted measure of user satisfaction (based on up to 39 questions) and their single-scale measure of Overall "Self-assessed" Satisfaction.Information Quality (t-statistic=7.48) is an important determinant of Satisfaction.This provides strong empirical support for hypothesis 1 in Figure 2. Third, Doll and Torkzadeh (1988) 2. Fourth, in their empirical study, Seddon and Yip (1992) report that Doll and Torkzadeh's EUCS explained over 70% of the variance in their four-item User Satisfaction measure, but that after Information Quality was included in the regression, Ease of Use was not significant (regression 7.2).This provides further empirical support for hypothesis 1, but not for hypothesis 2. Fifth, Davis (1989) has provided IS researchers with two highly reliable measures of Perceived Usefulness and Ease of Use.Davis was interested in explaining peoples' ex ante decisions to use information technology, but the measures seem equally applicable to ex post evaluations of information systems.Concerning Ease of Use (an important component of System Quality), Davis found significant correlations (p<0.001) between Ease of Use and Usefulness for three of four systems studied (Table 8, p.332), and suggests that "ease of use influences usage indirectly through its effect on usefulness" (p.330).Davis's work provides both theoretical and empirical support for hypothesis 3 in Figure 2 (i.e., that increased System Quality is associated with increased Usefulness).
Sixth, in support of hypothesis 4, Larcker and Lessig (1980, p.123), Franz and Robey (1986), and Kraemer, Danzinger, Dunkle, and King (1993, Q.2a, p. 133) have all argued that increased Information Quality will lead to increased Usefulness.Franz and Robey (1986) 2. Seventh, Kraemer et al. (1993) studied factors that influence the perceived usefulness of computer-based information (CBI).They used regression analysis to assess the relative importance of factors that affect usefulness.For data from 211 operations managers in public organizations, they report: "Finally, from the foregoing factors, CBI accessibility, CBI quality, and reliance on experts were found to have the most significant influence on the perceived usefulness of CBI" (p.139).This provides limited empirical support for hypothesis 4 10 .Something is useful if it provides future benefits".A car is useful for getting to work.A bicycle is less useful for the same task because it goes slower and you may get wet if it rains.A car that won't start is not useful for getting to work today, but it will be useful again in the future once someone has got it working.A bicycle is not '"Support is "limited" because in addition to asking questions similar to our questions Q.6, 7, 8, and 9 on Information Quality (see Appendix 2) under the heading "Information Quality", Kraemer et al. also asked whether "the computer makes new information available to me that was not previously available".This idea, which was absent from our study, appears to have been the most significant factor in Kraemer's regression (their Table 7, p. 141)." Davis (1989, p.320) expresses the same idea when he says something is useful if it is "capable of being used advantageously".useful for getting to work today if you don't know how to ride.Consistently, across many examples, we found that Usefulness is concerned only with the future benefits of performing some task.Costs are much less important.For instance, your new $30,000 automobile may be judged to be marginally more useful than the reliable old $6,000 automobile it replaced, but not by five times.The benchmark for judging usefulness of a tool is that the value of the benefits flowing from its use in some specific task must exceed zero.
Usefulness Causes User
The benchmark for judging satisfaction is different.User Satisfaction is the net feeling of pleasure or displeasure that results from aggregating all the benefits that a person hopes to receive from interaction with the information system.Each user has a set of expected benefits or aspirations for the information system.To the extent that the system meets or fails to meet each of these aspirations, the user is more or less satisfied.At a minimum, a tool is expected to be useful.Beyond that, the more useful the tool, the more likely the user is to be satisfied with it.However, satisfaction reflects a wider set of expected benefits or aspirations than mere usefulness.For instance, we are more likely to be satisfied with our new $30,000 automobile than with the old $6,000 one, even though the annual cost of ownership is much higher.Assuming the new car is only marginally more useful than the old, this example implies that we must be valuing other non-usefulness benefits (such as comfort and status) in determining satisfaction.
Considerations along these lines lead us to believe that increases or decreases in Usefulness will lead to increases or decreases in User Satisfaction with information systems, but not vice versa (because some increases in Satisfaction are unrelated to Usefulness).This is the basis for hypothesis 5 in Figure 2: H5: Increases in Usefulness will cause increases in User Satisfaction In effect, we regard DeLone and McLean's model as saying that User Satisfaction responds primarily to three types of aspirations that people have for information systems: people want their information systems to provide high quality information (HI), to be of high quality (H2), and to be useful in their jobs (H5).
What support is there in the literature for the hypothesis that Usefulness causes Satisfaction, but not vice versa?Whilst many researchers have studied the relationship between Use (a behaviour) and Satisfaction (an attitude), DeLone and McLean's model is concerned with Use as a measure of success, i.e., with Usefulness (a belief) 12 .Surprisingly few researchers have considered the relationship between Usefulness and Satisfaction.Apart from Bailey and Pearson (1983), who used a question about Utility (the relative balance between cost and usefulness) in their measure of Satisfaction, and Goodhue (1986), who speculated that there might be weak links in both directions between Satisfaction and Usefulness, the most helpful study we could find was by Franz and Robey (1986).Franz and Robey used Perceived Usefulness as the dependent variable in their study of user participation and organizational context.In discussing the choice of independent variable, they say: Our questions were designed to assess perceptions of usefulness rather than more-general attitudes [Robey 1979;Schewe 1976] or satisfaction [Bailey and Pearson 1983;Ives et al. 1983] (p.338).
It is clear from this sentence that they believe that Satisfaction is a more general concept than Usefulness.However, it is not clear how they think the two concepts are related.Since Franz and Robey's was the only study we could find that offered clear insights into the relationship between Usefulness and User Satisfaction, we were forced to rely on the "first-principles" analysis above to justify our interpretation of the arrows in DeLone and McLean's model, i.e., to justify hypothesis 5.
System Importance used to Explain Variance in Usefulness and User Satisfaction (H6 and H7)
Because DeLone and McLean's model was proposed as a way of interrelating various measures of IS success, it does not consider factors that might influence peoples' evaluations of success.However, as all four constructs in the dotted line box in Figure 2 are perceptual, it seems highly likely that users' opinions about the relevance of the system to their own goals and aspirations will influence their opinions about the value and so success of the system.For example, if what the system does is unimportant to the user, there seems little chance that the user will perceive the system as useful, no matter how well designed it is, or how easy it is to use.Conversely, if the task the system supports is perceived as very important, a poor system may be perceived as useful, even if it is quite user unfriendly.In this study of success measures, it therefore seems essential to consider the interests of the individuals being asked to evaluate the information system.How should users' interests be incorporated into Figure Ib?For help in this regard, we turned to Barki and Hartwick's notion of User Involvement.Barki and Hartwick's (1989) revolutionary paper presents a strong case for distinguishing between User Participation, which they define as "participation in the system development process" (p.53), and User Involvement, which they define as "the subjective psychological state" of the user "when he or she considers a system to be both important and personally relevant" (p.53).This User Involvement concept is very similar to task-relevance concept we had in mind, so we adopted the Importance component of User Involvement for our study.
Intuitively, we felt that people who believed that the task performed by the system was important would also perceive the system that performed it to be useful.The more important the task, the more useful the system.Thus H6 proposes that the greater the user's perception of the Importance of the task, the more useful the system will be perceived to be.Additional support for H6 comes from Larker and Lessig (1980, p.123), in the context of information usefulness.Larker and Lessig used the term "Perceived Importance" to refer to "the quality that causes a particular information set to acquire relevance to a decision maker".Consistent with our H6, they go on to argue that Perceived Importance will "tend to increase the perceived usefulness of the set".
The link from Importance of the task to Satisfaction is less clear.Provided the system works, the more important the task to the user, the more satisfied he or she is likely to be (Importance up, Satisfaction up).On the other hand, if the system does not work and the task is important, the user may be very dissatisfied (Importance up, Satisfaction down).Finally, if the task is unimportant, the user's threshold for satisfaction may be so low (i.e., the user would be indifferent to the system) that satisfaction scores might be moderate (Importance down, Satisfaction indeterminate).These last two scenarios would reduce correlations between Importance and User Satisfaction.On balance, since the system we proposed to test did work (though not very well) we expected the correlation to be positive.This is the basis for H7 in this study.Perceptions of System Quality and Information Quality are less likely to be coloured by perceptions of system Importance than perceptions of Usefulness and Satisfaction.For instance, twenty years ago, card-punch machines were the normal everyday way of communicating with computers.They were easy to use, and they were useful.Today, they are no less easy to use, but now that card decks are no longer used for communicating with computers, card-punch machines are considered useless.Based on this analogy (in which Usefulness changes as Importance changes but Ease of Use does not), we argue that System Quality (of which Ease of Use is a central component) is unlikely to be influenced by the importance of the system to the user, i.e., System Importance.
Similarly, we argue that Information Quality is unlikely to be influenced by System Importance.Information Quality is concerned with the timeliness, accuracy, relevance, and format of information generated by an information system.These are relatively objective qualities of information, and it seems unlikely that a user will judge Information Quality to be high simply because he or she thinks the task performed by the system is important.
We therefore expect System Importance will be positively associated with both Usefulness and Satisfaction but not with System Quality and Information Quality.These expectations are shown as hypotheses 6 and 7 in Figure 2. H6: Increases in System Importance will cause increases in Usefulness H7: Increases in System Importance will cause increases in User Satisfaction Surveying the literature for support for these hypotheses, we found two papers of particular interest.First, in the empirical study, Jackson, Chow, and Leitch (1993) studied factors affecting behavioural intention to use an information system.They used Zaichkowski's (1985) Involvement instrument to measure User Involvement, and found a highly significant relationship between User Involvement and Perceived Usefulness (t-statistic = 6.52).There are some difficulties with the use of all items in Zaichkowski's Involvement instrument in testing the relationship between User Involvement and Usefulness because some items in Zaichkowski's instrument actually measure Usefulness directly.Nonetheless, the large t-statistic in the Jackson et al. (1993) study provides considerable empirical support for the use of User Involvement, hence Importance™ to explain variance in Perceived Usefulness (H6).Second, Kappelman and McLean (1991), used the same Zaichkowski instrument to investigate the relationship between User Involvement and User Satisfaction.They report a highly significant association (p<0.001).This provides empirical support for our hypothesis 7.
TESTING THE MODEL OF USER SATISFACTION SHOWN IN FIGURE 2
The Questionnaire To test the model in Figure 2, we prepared a questionnaire based on a number of standard instruments.The questionnaire was to be administered to all users who had completed the training course for our university's recently-implemented Departmental Accounting System, DAS.Details of all questions used in measuring variables for this study are presented in Appendix 2. There are six groups of questions that measure seven variables as follows.First, the eight questions on System Quality are based on Doll and Torkzadeh's (1988) two questions on Ease of Use, four of Davis's (1989) questions on Perceived Ease of Use14 , plus three additional questions --all phrased in present (not future) tense15 .Second, the ten questions on Information Quality are all from Doll and Torkzadeh (1988).Third, the four questions on Overall Satisfaction are from Seddon and Yip (1992).Fourth, the six questions on Perceived Usefulness are from Davis (1989).
Fifth, the five questions on System Importance are from Zaichkowski (1985) via Kappelman and McLean (1991).Data for this study were collected before Barki and Hartwick (1994) published their instruments for measuring User Involvement, so we did not use their instrument for measuring Importance.However, because respondents answered all 14 questions from Kappelman and McLean (1991), it was possible to construct a measure very similar to Barki and Hartwick's.In fact, three of their five Importance questions and one of their four Personal Relevance questions (1994, Table 2a, p.70) are included in our five-item Importance measure.Finally, the last two variables used in this study are Use if not mandatory and 75 Use %16 .Single questions to measure these two variables were included to test the relationships between Use (the variable required in DeLone and McLean's model) and Usefulness (the variable that we argued earlier would be more relevant in situations where usage was mandatory).Use if not mandatory is a hypothetical intention-to-use measure because it asked users to pretend that usage of DAS was not mandatory.Based on tests of Davis's Technology Acceptance Model (TAM) (Davis (1989(Davis ( , 1993)), Davis et al. (1989), Adams et al. (1992), Segars and Grover (1993), and Igbaria et al. (1995)) one would expect quite a high correlation between Usefulness and Use if not mandatory.IS Use % measures self-reported actual use of DAS.Because some respondents were part time, IS Use % is measured as a percentage of the respondent's average working week.
Details of the System and Its Users
As noted above, the system we chose for testing the model in Figure 2 was the university's recentlyimplemented Departmental Accounting System, DAS.We chose this particular system because we knew there had been some difficulties with its implementation, and a wide range of Usefulness and Satisfaction scores was likely17 .A second attraction was that the same system was in use in all departments, and all users had been trained by the same trainers.There were therefore a smaller number of extraneous factors that could cause variance in the Usefulness and Satisfaction scores.
The goal of DAS is to keep track of all of each department's diverse range of income and expenditure, so that those who are managing various projects, research grants, etc., can know, at any time, how much money they have left to spend.Built in Ingres at a fully-implemented cost of over $500,000, DAS runs on a central computer and acts as a front end to the university's central accounting system.About half the implementation cost was for user training and support during the two-year phased implementation of the system.The typical DAS user is a relatively senior clerical officer in each department or faculty office who uses the system for about 4-5 hours per week for maintaining the department's accounting records.At the time of the survey, 169
Data Collection
After ten face-to-face trials, the questionnaire was mailed to the remaining 159 trained DAS users.134 questionnaires were returned, but only 94 were useful for data analysis.Non-useable responses were for the following reasons: some departments were not yet using DAS, some trained users were on leave, some were no longer responsible for DAS, and some had resigned.Combining the 10 initial responses with the 94 returned by mail gave 104 responses for statistical analysis18 .For each of the 35 questions asked, responses from early and late respondents (83 and 21 respondents, respectively) were compared using the Mann-Whitney U statistic (Siegel and Castellan 1988).None of the U statistics from these 35 questions indicated a significant difference, at the 5% level, between early and late respondents.Because of the high response rate and the lack of significant differences between early and late respondents, non-response bias is not considered a problem for this study.
Descriptive statistics for distributions of responses for each variable are given in Table 1.Principal components factor analysis (no further details in this paper) of the eight questions on System Quality showed that one question (question 7 in Appendix 2) had a loading of only 0.25 on the main factor.It was therefore deleted, and all statistics that follow are based on the remaining seven-item measure of System Quality.
Reliabilities (Cronbach alpha) and the Pearson correlation matrix for the variables used in the study are shown in Table 2.
Was it incorrect to argue (as we did in the previous section) that Usefulness should be used in place of Use in DeLone and McLean's model?The answer to this question proceeds in two steps.First, Table 2 shows that the correlation between Usefulness and hypothetical Use if not mandatory was 0.699.In other words, Usefulness explained approximately 50% of the variance in Use if not mandatory.This is entirely consistent with results from the TAM studies, and suggests that, in this mandatory-use setting, Usefulness is tapping the same construct that DeLone and McLean sought to measure with Use.Second, the correlation between Usefulness and self-reported /S Use % was only 0.295 19 .In addition, it is obvious from respondents' written comments (at the end of each questionnaire) that usage was determined by the needs of different departments, not by the desires of the users.
Results: Use versus Usefulness
In our opinion, the high correlation of Usefulness with Use if not mandatory, and the low correlations between the various perceptual success measures of IS Success and actual mandatory IS Use % provide quite strong empirical support for DeLone and McLean's argument that when measuring IS Success, "usage, either perceived or actual, is only pertinent when such use is voluntary" (1992, p.68).The remainder of this analysis therefore proceeds on the understanding that when usage is mandatory, Usefulness is the most appropriate usage-related measure of IS Success.
Results: Path Analysis
Results from path analysis of the model in Figure 2 are reported in Table 3. Path coefficients in column 4 were computed using two ordinary least squares (OLS) linear regressions, the first with Usefulness as the dependent variable (Hypotheses 6, 3, and 4), and the second with User Satisfaction as the dependent variable (Hypotheses 7, 2, 1, and 5).Path coefficients in column 5 were calculated using maximum likelihood estimation in the structural equation modelling (SEM) package called Amos20 .To avoid reliance on any assumptions of multivariate normality, standard errors in column 5 were calculated using bootstrap simulation with 1,000 replications21 .The path coefficients from the two methods of analysis are very similar, although the t-statistics for the SEM in column 5 are considerably lower than those for OLS regression.
The reason for including two sets of path coefficients (OLS and SEM) in Table 3 is that structural equation modeling is becoming much more common in the Information Systems literature (e.g., LISREL: Adams et al. 1992, Segars and Grover 1993, Chin and Todd 1995;PLS: Rivard and Huff 1988, Thompson et al. 1991& 1994, Igbaria et al. 1995), and many researchers probably wonder how the two/three methods of analysis compare.Most studies opt for just one method of analysis, a decision that leaves the reader without any basis for gaining an intuitive understanding of the relative merits of one method over another.Hence the two columns in Table 3.A detailed comparison of the techniques used for calculating the OLS and SEM path coefficients is presented in Appendix 1.Despite the differences in analysis techniques, it is apparent from Table 3 (and Figure 3, which presents the results graphically) that for this particular data set, the conclusions one would draw from either analysis technique are the same.The results in Table 3 provide quite strong support for five of the hypotheses in Figure 2. The t-statistics in columns 4 and 5 are so large that we have confidence in rejecting the null hypotheses (of no association between the constructs) for hypotheses 1, 2, 3, 5, and 6.Support for hypothesis 4 was weaker, however, and after controlling for the effect of Importance on Usefulness there is no support for hypothesis 7.With the benefit of hindsight, one can see that the lack of significance in hypothesis 7 is consistent with the difficulties with our theoretical analysis (see end of the hypothesis development section).Perhaps Kappelman and McLean's correlation was so highly significant (p<0.001) because the systems they studied were all working well -the "Importance up, Satisfaction up" case discussed in the development of hypothesis 7.
CONCLUSIONS
For the individual user of a specific information system, our results provide considerable support for the upstream side of DeLone and McLean's model of IS success.In effect, DeLone and McLean's model may be interpreted as saying that User Satisfaction is a response to three types of user aspirations for an information system: Information Quality (HI), System Quality (H2), and Usefulness (H5).As DeLone and McLean predicted, these three factors explained a large proportion of variance in User Satisfaction (over 70% in this study).To the best of our knowledge, no prior research has identified the importance of the Importance of.the task construct to a system's Perceived Usefulness and Use.Importance of the task is independent of the system, yet as shown in Figure 3, it has a major influence on Perceived Usefulness.Future research, both in the prediction of information system use (via Usefulness), and in the investigation of links between User Participation, User Involvement* 2 and system success (Barki andHartwick 1994, Hartwick andBarki 1994), will need to control for the Importance of task performed by the system.For instance, it is not level of User Involvement that matters in User Participation research.It is the change in Involvement (the change in the perceived importance of a system as a result of Participation) that matters.If researchers are looking for a simple, short measure of IS success, which one should they choose?DeLone and McLean argued that multiple indicators of success are required.To some extent, we agree.IS Success can mean many things to many people, and researchers must measure the IS success outcomes that are relevant to their own research goals.But as Satisfaction appears to be the most inclusive of the four perceptual measures investigated in this study, if a single measure is required, we recommend using User Satisfaction as the most general-purpose perceptual measure of system success.The four questions from Seddon and Yip (1992), which returned reliability coefficients (alpha) of 0.95 in their survey, and 0.92 in this study, may be all that is needed23 .Researchers wanting to understand why users are dissatisfied with their systems will probably find it helpful to measure at least the three causal constructs suggested by our slightly-modified DeLone and McLean model, namely, Information Quality, System Quality/Ease of Use, and Usefulness.When measuring Usefulness, it will of course be necessary to control for Importance of the task.To the non-statistician, the advantages of structural-equation models (SEM) over conventional factor analysis and ordinary-least-squares (OLS) regression techniques are unclear.Bollen (1989, page v) describes the advantages of LISREL-style SEM in the following rather vague terms: (Structural equation models)... are regression equations with less restrictive assumptions that allow measurement error in the explanatory as well as the dependent variables.They consist of factor analyses that permit direct and indirect effects between factors.They routinely include multiple indicators and latent variables.In brief, these models encompass and extend regression, econometric, and factor analysis procedures.
So, how much difference is there between OLS and LISREL-style SEM? Do these differences matter?In the case of the present data set, the principal differences between the two sets of path coefficients in columns 4 and 5 in Table 3 (in the main paper) are as follows: First, for the OLS regression, the score for each latent variable (theoretical construct) is the mean of responses to the questions used to measure the variable24 .For example, for any one respondent, the score for the latent variable called User Satisfaction is calculated as the mean of that respondent's responses to the four questions shown in the Appendix 2. The Cronbach alpha of 0.92 in Table 1 indicates that for the sample as a whole (100 observations), scores for these four questions tended to vary together very consistently.OLS then assumes these average latent variable scores for the independent variables are measured without error, and minimizes the squared difference between the dependent variable's average latent variable scores and the scores predicted by the regression equation.By contrast, in LISREL-like SEM, no scores are calculated for any latent variables, dependent or independent.Instead, scores on individual questions are used to calculate the covariance between each question and every other question.Amos then uses a combination of the researcher's path diagram and maximum likelihood estimation to minimize the difference between the sample covariance matrix and that implied by the path diagram.In Figure 4, the rectangles represent responses to individual questions, and the ovals represent the latent variables."Loadings" of each question (rectangle) on its underlying latent variable (oval) are the result of this covariance-matrix-fitting process25 .The chi-square coefficient reported by Amos (chi-sq.=880,df=454, chi-sq/df=1.94) is the extent of misfit between the sample and estimated covariance matrices.There is a large misfit.The adjusted goodness of fit index for this model is only 0.63 (root mean square residual = 0.19), but as we have no strong theoretical grounds for re-specifying the model, we have not attempted to "tune" the model to increase the goodness of fit.Second, for the OLS regression, missing values are handled either by the averaging process or else the entire response is deleted (listwise deletion).In this study, if one respondent only answers three of four questions, the score for the latent variable for that respondent is the average of the three (not four as with the other respondents).If the researcher judges that too few questions have been answered for the mean of responses to be meaningful, the score for that respondent is left undefined (missing), and the entire response from that respondent is deleted by SPSS during listwise deletion of missing variables during estimation of the regression equation.By contrast, in structural equation modeling, covariances from each question to every other question are required before model fitting can begin.Thus in SEM, the treatment of missing values is handled in during calculation of the sample covariance matrix.Amos provides a technique for such estimation (described in Arbuckle (1994b)), but when bootstrapping is used to obtain what Arbuckle describes as "more robust" estimates of standard errors (as in the data reported in Table 3, column 5) this missing-value treatment is not available.Thus for the data in Table 3, missing values for particular questions were replaced by means for the respective questions.Such mean replacement of missing values occurred for data from only 2 of the 100 respondents.(These two responses were also those deleted in SPSS during its listwise deletion phase.)A total of 17 missing values from 32*100 = 3,200 respondent answers (0.53%) were replaced.The effect on the covariance of replacing missing values by the mean of the remaining observations for a given question is entirely neutral.To see this, consider the formula for calculating covariance between responses to two questions Xi and X 2 : tv v , 2rf,-=I (*l« ~*lX*2,-"^2) , V i e r\ • • i i cov(X,,X 2 ) = --, where Xj = sample mean for Q. 7,7 = 1,2 N -I In this case, because 100 people responded, calculation of the covariance involves a summation of 100 terms of the form (X u -X,)(X 2l -X 2 ).Suppose the person completing the 88 th questionnaire did not answer question X 2 , and that the missing score for this question was replaced by the mean of all other respondents' scores for X 2 , i.e., by X 2 .This means the 88 th of the 100 terms being summed to calculate the covariance would be zero (because (X 288 -X 2 ) = ( X 2 -X 2 ) = 0).Thus the covariance is determined from the other 99 observations, and the mean-substituted 88 lh observation has no effect on the calculation of the numerator of the covariance strongly agree disagree 1. DAS is easy to use. 2. DAS is user friendly.3. Compared to other software, DAS is easy to learn.4. I find it easy to get DAS to do what I want it to do. 5.It is easy for me to become skilful at using DAS.6.I believe that DAS is cumbersome to use. 7. My using DAS require a lot of mental effort.Part B: Information Quality.For the system overall, 1. Do you think the output is presented in a useful format ? 2. Are you satisfied with the accuracy of the system ?3. Is the information clear ? 4. Is the system accurate ? 5. Does the system provide sufficient information ?6.Does the system provide up-to-date information ?7. Do you get the information you need in time ?8. Does the system provide reports that seem to be just about exactly what you need ?9. Does the system provide the precise information you need ?10.Does the information content meet your needs ? 1 2 3 4 5 6 7 never always 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 2 3 4 5 6 7 2 3 4 5 6 2 3 4 5 2 3 4 5 6 6 7 7 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 Part C: Overall Satisfaction.
, reproduced here as Figure la, DeLone and McLean say (p.88):
Figure 2 :
Figure la: DeLone and McLean's Model of IS Success (DeLone and McLean (1992), Figure 2, p.87) Satisfaction, Not Vice Versa (H5)DeLone and McLean's model (Figure la) suggests that there is a two-way causal relationship between Use and Satisfaction.A simple duplication of this two-way relationship in Figure2would imply a two-way causal relationship between Usefulness and Satisfaction.Is this valid, or does the change from the Use construct to Usefulness introduce a subtle change in the causality relationships?As we could think of no statistical test to determine whether Usefulness causes Satisfaction, or vice versa, we were forced to resort to semantics to try to answer this question.The conclusions from this back-to-first-principles analysis are presented in the next two paragraphs.
the DAS training course: 152 were women, 17 were men; 118 were accessing DAS through Apple Macintosh computers in their departmental offices, and 51 from IBM PCs.
Increases in Information Quality will cause increases in User Satisfaction H2: Increases in System Quality will cause increases in User Satisfaction H3: Increases in System Quality will cause increases in Usefulness H4: Increases in Information Quality will cause increases in Usefulness developed a measure of End User Computing Satisfaction (EUCS) that asked ten questions about Information Quality and two about Ease of Use.These questions factor analyzed into five factors.The first four factors relate to the construct that we call Information Quality: Information Content, Accuracy, Format, Timeliness.The fifth factor is Ease of Use, a component of System Quality.As with the Ives et al. instruments, we contend that Doll and Torkzadeh's instrument is actually measuring two variables that are causes of User Satisfaction.Thus as with the Ives et al. instrument, we view Doll and Torkzadeh's decision to include questions concerning Information Quality and System Quality in their measure of End User Computing Satisfaction as providing theoretical support for hypotheses 1 and 2 in Figure
Part D: Perceived Usefulness.
On the following scales, please circle the number which best reflects vour overall satisfaction with DAS.1.How adequately do you feel DAS meets the information processing needs of your area of responsibility ?On the following scales, please circle the number that best reflects how useful you perceive DAS to be.strongly agree disagree 1.Using DAS in my job enables me to accomplish my tasks more quickly. | 2015-07-20T21:03:00.000Z | 1996-11-01T00:00:00.000 | {
"year": 1996,
"sha1": "93f61d00e536c5264e33cb6c427a26996a6c0316",
"oa_license": "CCBYNC",
"oa_url": "https://journal.acs.org.au/index.php/ajis/article/download/379/345",
"oa_status": "GOLD",
"pdf_src": "Anansi",
"pdf_hash": "28d5d825f366b14418ba12b6f546bb5b2213cd09",
"s2fieldsofstudy": [
"Business",
"Computer Science"
],
"extfieldsofstudy": [
"Computer Science"
]
} |
219183026 | pes2o/s2orc | v3-fos-license | 31-year contrasting agricultural managements affect the distribution of organic carbon in aggregate-sized fractions of a Mollisol
Evaluation of soil organic carbon (SOC) dynamics is often limited by the complexity of soil matrix. Quantitative information on the distribution of SOC within aggregate hierarchy will help elucidate the carbon flow in soil matrix. However, this knowledge still needs to be documented. Soils were sampled from a surface Mollisol with plots under 100 years of continuous cropping, 31 years of simulated overgrazing to severely degraded bareland, and grassland restoration from cropped soil. A combined density and chemical fractionation procedure within water-stable aggregate was utilized to quantify the distribution of OC after long-term different land use patterns. Results showed that grassland significantly increased total SOC and mean aggregate associated OC compared to initial soil in 1985 with total SOC (g kg−1 soil) from 46.1 to 31.7 and mean aggregate associated OC (g kg−1 aggregate) from 31.6 to 44.7. Converting cropland to grassland also enhanced the formation of macroaggregates (>0.25 mm) (from 34.7% to 52.2%) and increased the OC concentrations in density and humic fractions by 48.3%-75.9% within aggregates. But the proportions of OC in density and humic fractions to SOC only increased in macroaggregates in grassland. Alternatively, converting cropland to bareland caused substantial depletion of total SOC, macroaggregates and their associated OC concentrations. The SOC (g kg−1 soil) and mean aggregate associated OC (g kg−1 aggregate) significantly decreased from 31.7 to 25.7 and from 31.6 to 26.2, respectively. While the OC concentration of density and humic fractions within aggregates in bareland did not show significant decreases. Principal component analysis demonstrated that the soils were developed by contrasting land use changes, with the grassland soil being more associated with labile OC fractions within macroaggregats and bareland soil more associated with recalcitrant OC fractions within microaggregates and silt-clay units. These findings highlighted the favorable preservation of plant-derived carbon within soil aggregates, particularly in the labile OC fractions within macroaggregates under high plant inputs with 31 years of grassland conversion. For the cropland and bareland soils without organic inputs, more OC was stabilized within fine aggregates via organo-mineral interactions, tending to be more recalcitrant.
The processes governing the formation, transformation, and stabilization of soil organic carbon (SOC) are closely tied to soil aggregate architecture 1 . Soil aggregates can physically protect SOC against decomposition and store SOC crucial to climate change mitigation 2 . Also, the location of organic carbon within aggregate is a key factor for the stabilization of SOC 3 due to spatial arrangement of soil constituents and soil microbes 1,4 . One important but poorly understood aspect of soil organic matter (SOM) decomposition and stabilization is the role of soil aggregates. Soil aggregates controlled the degree of hierarchical inaccessibility of SOC to decomposing organisms or catalytic enzymes 5,6 , the availability of SOC and nutrients, and their responses to land-use changes 7,8 . Better understanding of the aggregate-associated carbon, as well as the carbon allocation within aggregates, are crucial 31-year grassland significantly increased the OC concentration in all aggregate fractions, with the mean aggregate associated OC (g kg −1 aggregate) from 31.6 to 44.7, and higher increments were observed in 0.25-2 mm and 0.053-0.25 mm aggregate fractions, the increment was lowest in the silt-clay unit (P < 0.05) (Fig. 1b). Converting cropland to bareland caused substantial depletion of macroaggregates and their associated OC concentrations. The percentage of macroaggregate (%) and mean aggregate associated OC (g kg −1 aggregate) significantly decreased from 34.8 to 13.6, and from 31.6 to 26.2, respectively. Bareland only caused significant decrease of the OC concentration within 0.25-2 mm macroaggregate compared with soil in 1985 (P < 0.05). Cropland caused no significant changes of OC concentrations in all aggregate sizes, being higher in macro-and microaggregates, and lower in the silt-clay unit compared with soil in 1985. The aggregate-associated OC concentrations in all aggregates were significantly higher in grassland, followed by cropland, and lowest in bareland, significantly higher OC was also found in >0.25 mm macoaggregate in cropland than that in bareland (P < 0.05) (Fig. 1b).
Organic carbon of density and humic fractions within aggregates. The OC concentrations of different density and humic fractions differed obviously among soil aggregate hierarchy and land use patterns. Specifically, within all aggregate fractions, the OC concentrations of all density and humic fractions were highest in grassland, followed by cropland and bareland. Also, the OC concentration showed no significant differences in cropland and bareland soils among aggregates, with the exception of higher OC of free light (fLF) and occluded light (oLF) fractions within microaggregate in cropland than those in bareland soil (P < 0.05). The free light (fLF) and occluded light (oLF) fractions contained obviously higher OC concentrations than those in heavy fraction (HF) and humic fractions (Fig. 2). Among all the field treatments, the OC concentrations of density fractions were highest within 0.25-2 mm aggregates, and then decreased with the size of the aggregates, with the <0.053 mm silt-clay unit having the lowest OC concentrations (Fig. 2a-c). The OC concentrations of humic fractions within aggregates did not show the same changing treads as density fractions after 31-year land use changes. There were no significant differences of OC concentration of fulvic acids (FA) and humic acids (HA) among field soils, with the exception of higher OC concentrations of FA in >2 mm and 0.25-0.053 mm aggregate fractions in grassland than in cropland. The OC concentration of humin (HU) showed similar changes as HF, significantly higher OC concentrations were found within 0.25-2 mm and 0.053-0.25 mm aggregates in grassland (Fig. 2d-f).
Soils
Soil pH www.nature.com/scientificreports www.nature.com/scientificreports/ The proportions of density-and humic-associated OC in SOC differed obviously with OC concentration among land use patterns and soil aggregate hierarchy (Fig. 3). Among all the field treatments, the HF and HU fractions, containing lower OC concentrations, were the most abundant OC fractions relative to total SOC. The highest proportions were found within 0.25-2 mm aggregate in grassland and 0.053-0.25 mm aggregate in cropland and bareland (Fig. 3). Specifically, within >2 mm and 0.25-2 mm macroaggregate fractions, the proportions in SOC were significantly higher in all density and humic fractions in grassland than in the respective fractions of both cropland and bareland soils (P < 0.05). The bareland soil had the lowest OC proportions in density and humic fractions, with the exception of higher occluded light fractions (oLF) within >2 mm macroaggregate in bareland than that in cropland (P < 0.05). Within 0.053-0.25 mm and <0.053 mm aggregates, higher OC proportions were found in bareland soil, followed by cropland, and grassland had the lowest proportion except for higher OC proportion of fLF in grassland than that of cropland and bareland (Fig. 3).
Interrelations. Principal component analysis (PCA) including all aggregates in bulk soil, density, and humid fractions within all aggregate sizes, demonstrated significant effects of field treatment and soil aggregate hierarchy via both PC1 and PC2, which explained 61.08% and 28.92% of the total variance, respectively (Fig. 4). The grassland, cropland, and bareland soils separated from the soil in 1985 in different directions. It developed in two contrasting directions via PC1, the cropland and grassland were divided by the soil in 1985, with cropland more similar to the soil in 1985. The bareland soil was separated from soil in 1985, cropland and grassland soils via PC2. The separation towards grassland via PC1 was associated with SOC, TN, and SMBC, >0.25 mm macroaggregate and related density and humic fractions within macroaggregates, fLF and oLF fractions within microaggregates, and fLF within the silt-clay units. The separation towards cropland via PC1 was mostly driven by the mass distribution of microaggregate and the silt-clay unit, and OC in HF, FA, and HA fractions thereof. The separation towards bareland via PC2 was mostly driven by HU within microaggregates and the silt-clay units, and fLF and oLf within the silt-clay units (Fig. 4).
Discussion
Plant residue inputs: the driving factors linking SOC dynamics and land uses. The soil in 1985 had been used for more than a century for cropping, it was utilized to explain the changes of soil properties after 31-year of continuous cropping. Over 31-years continuous cropping, the selected soil properties did not show as much change as in natural grassland and bareland soils, with SOC and TN contents decreased compared with the soil in 1985 (Table 1). Such changes were consistent with previous results showing that long-term continuous cropping generally decreased SOC content 33 . The cropland soil here was considered as the "reference" soil when determining the C redistribution within soil aggregates under contrasting land use patterns. The changes in www.nature.com/scientificreports www.nature.com/scientificreports/ OC of density and humic fractions within aggregates were interpreted both on the OC concentrations per mass aggregate and on OC proportion of each fraction relative to the total SOC of the whole soil base by comparing the continuous cropping in cropland to the absence of vegetation in the bareland and conversion to natural grassland with high level of plant residue input in grassland. www.nature.com/scientificreports www.nature.com/scientificreports/ In natural grassland, the estimated shoot and root biomass were 4-9 t ha −1 and 12-43 t ha −1 , respectively. The highest amount of root biomass (i.e. root density) was also found in the upper 10 cm for the grass species of the present experiment (data not published). However, there were no, or only limited, OC inputs from root exudates in bareland and cropland. Higher amounts of plant residue inputs from natural grassland were theoretically the dominant factor influencing the SOC dynamics and OC redistribution patterns in soil matrixes as shown in our results 27 (Figs. 1-3). In addition, more significant increments of SOC and TN contents, macroaggregates, and associated OC concentrations were found for bulk surface soil in grassland than in bareland and cropland soils (Table 1; Fig. 1). This indicated that the vegetation type (natural perennial versus crops) and organic C inputs from above-and below-ground plants (only observed in grassland soil) could promote the accumulation of SOC and soil aggregation, as has been reported in Mollisol 31 and other soils 34 . It was the accumulation of relatively large amounts of plant residue in grassland, with the unique climate conditions in the Mollisol region, that contribute to the relatively high accumulation of SOC in grassland 29 (Table 1). Nevertheless, the SOC, TN, and SMBC contents decreased in bareland soil, indicating a sharp loss of labile OC in soil likely due to long-term substrate limitation and exhaustion with no OC amendments 35 . Soil aggregates response to land use patterns. Soil aggregate size distribution varied significantly in terms of land use types (Fig. 1a). The macroaggregates (>0.25 mm) distribution, which generally had an order of grassland > initial soil in 1985 >cropland >bareland, has been observed to decrease or disappear with sustained planting 36 . Converting cropland to perennial grassland 31 years ago caused increases in large macroaggregate and small macroaggregate of 269.7% and 164.5%, respectively, but decreases of 43.3% and 58.8% in microaggregate and silt-clay unit, respectively. In grassland soil, continuous plant residue inputs facilitated the aggregation of soil particles causing micro-aggregates to form as macro-aggregates 36 , thus increasing the larger sized aggregates and decreasing the smaller sized aggregates 24,37 . Additionally, the mass percentage of large and small macroaggregates dropped dramatically in bareland soil compared to grassland soil (Fig. 1). The disruption of soil aggregates may expose once-protected organic matter to decomposition 38 and induce organic C loss in soil, as has been observed here (Fig. 1). In addition, wind and water erosion may destroy soil macroaggregates, causing loss of aggregate-protected C since there are no plants growing in the bareland soil 39,40 . Continuous cropping also caused decrease of small macroaggregate, due to annual soil tillage at the surface 0-20 cm soil depth. Soil tillage could destroy soil macroaggregates and cause the redistribution of OC for the arable perturbed plot under maize-soya bean-wheat rotation 38 .
Soil aggregate-associated OC within different aggregate hierarchies can provide critical insight into C sequestration mechanisms and their contribution to SOC dynamics 1,36,41 . In line with previous studies, our results showed that the aggregate-associated C varied across aggregate sizes and land use patterns 9,42-44 . The increases of OC concentrations were observed in all aggregate fractions in grassland (Fig. 1b). Also, larger increments of OC concentrations were detected for fast-responding aggregate fractions, such as the large macroaggregate (>2 mm) and macroaggregate (0.25-2 mm) (Fig. 1b). This demonstrated that the plant-derived C entered into soil to form SOM, and macroaggregate played a more dominant role than microaggregate and silt-clay unit in C sequestration, as has been reported elsewhere 5,8,14 . The improvements were expected because the no-tilled perennial restoration led to aggregation of fine soil particles or the increase of root exudates by adding large amounts of plant residue into the soil 6,33,44 . The no-tilled perennials facilitated the protection of SOM from microbial degradation in different aggregate-size classes due to lower soil disturbance, which in turn favored the generation of physically stable large and small macroaggregates 43 . In addition, the new OC sources derived from recently photosynthesized products such as plant residue and root biomass play evidential roles in driving biological processes in soil 45 . Sufficient nutrient availability in grassland soil enhanced soil microbial processes and further caused SOC accumulation via soil microbial residue 30 . Indeed, there were more macro-aggregates and associated organic C in grassland than cropland and bareland (Fig. 1). Continuous cropping slightly increased the aggregate-associated OC concentrations in macro and microaggregates, but significantly decreased the OC concentration in the silt-clay unit (P < 0.05). Soil tillage in cropland induced soil C loss by acceleration of organic C decomposition 46 . The acceleration of organic matter decomposition was at least partly due to the disruption of the soil macroggregates, which exposed once-protected organic matter to decomposition 38 . In bareland, the OC concentrations decreased in all aggregate fractions compared with those in soil of 1985 (P < 0.05) (Fig. 1b). Because the bareland soil neither received OC amendments nor had plants grown for up to 31 years, this caused limited nutrient availability for soil microorganisms, reducing the physical protection capacity and causing decomposition of old C of soil aggregates 47 (Fig. 2).
Organic carbon allocation within aggregates under different land use patterns. Soil aggregates,
serving as "sieves", protected C by providing physical barriers between OC and OC decomposers, thus influencing the OC distribution and stabilization thereof 1,5,35,41,48 . In this study, the combined density and humic fractionation method of SOC could help quantify the allocation regimes of OC within aggregates and their underlying sequestrating mechanisms within soil matrixes 3,17 . The OC distribution varied within different sized aggregates under land use and agricultural managements due to changes of nutrient availability, soil biota, and their interactions with soil aggregates 48,49 . Compared with cropland and bareland soils, relatively higher OC concentrations were accumulated in all density and humic fractions within all aggregate sizes in grassland. Also, the proportions of OC in density and humic fractions to total SOC were higher in large and small macroagregates in grassland while the proportions of OC in density and humic fractions in microaggregates and silt-clay units were lower in grassland than those in cropland and bareland soil (Fig. 3). This suggested that not only plant residue (only found in grassland soil) were supposed to enhance C accumulation in macroaggregates, but also did so for vegetation type (perennial versus annual crops), reduction in tillage, or a combination. Previous studies reported that soil receiving higher organic inputs increased the coarse fraction within macroaggregate 14 www.nature.com/scientificreports www.nature.com/scientificreports/ fraction within macroaggregates may not only be lost temporally but also could be restored rapidly upon conversion of cropland into grassland, and the new steady status of OC can occur earlier in the coarse fraction than in bulk soil due partly to microbial residue 30,50 . In particular, prominent increases of OC proportions in the density and three humid fractions were observed in 0.25-2 mm macroaggregates in grassland soil compared to cropland and bareland soils (Figs. 2, 3). This 0.25-2 mm macroaggregates in grassland were considered to be the "preferential aggregate fractions" where most of the plant residue C was accumulated (Fig. 1b, 2 and 3) and can be used as an indicator of management effects 15,37 . Our results were in line with others 21,22 , identifying a gradient of OC incorporation into water-stable aggregates with faster response to land-use change by macroaggregate fractions compared with microaggregates and silt-clay units. The PCA analysis also confirmed these findings, in which the OC fractions within macroaggregates in grassland were shifted on the right plot from cropland and bareland soils (Fig. 4).
The proportions of OC in total SOC decreased in density and humic fractions within macroaggregates, and increased within microaggregates and silt-clay units in bareland soil without any fresh OC amendments and in cropland soil with only limited root exudates (Fig. 3). This indicated a predominant loss of labile and decomposable organic matter in the macroaggregate fraction likely due to long-term substrate limitation and soil erosion 51 . Also, the 0.25-0.053 mm microaggregates was considered as the "preferential aggregate" in cropland and bareland soils. The cropland and bareland soils lacked new plant-derived C for up to 31 years resulting in deficient C and nutrient sources for soil microbes and then preferential decomposition of microbial derived C 52 , consequently causing break-down of macroaggregates into microaggregates and decomposition of OC therein 22 . The OC within microaggregates and silt-clay units was supposed to have lower turnover rates due to stabilization by organo-mineral interactions in bareland and cropland 18 , which played a significant role on C stabilization 13 . Xie et al. 53 also found a similar change of aggregate associated C in loess soil when cropland was converted to fallow. Furthermore, the OC proportions of HF, FA, and HA fractions within microaggregate and silt-clay fractions were higher in cropland and bareland soils than those in grassland soil. This demonstrated greater C protection of aggregates in the fine fraction when substrates were limited as in bareland or cropland soil (Figs. 3,4). The reduced OC proportions within macroaggregate but increased OC proportions within microaggregate could be attributed to the lack of biological binding agents in cropland and bareland soils 6 .
Conclusion
The 31-year contrasting land use patterns caused C redistribution within soil matrixes. The 0.25-2 mm macroaggregates and 0.25-0.053 mm microaggregates were considered to be the "preferential fraction" largely contributing to the accumulation of labile and recalcitrant OC in grassland and bareland soils, respectively. Converting cropland to grassland caused relatively higher proportions of OC sequestered within macroaggregates, which highlights the importance of reducing soil disturbance and enhancing plant residue inputs and potential microbial contribution for SOC accumulation, particularly in heavy and humin fractions within macroaggregates. However, the absence of plant-derived C inputs in bareland limited new OC resources and subsequently caused the depletion of macroaggregates and their associated OC. Our results demonstrated that the soil aggregate matrix served as a "sieve" causing C redistribution thereof, with land use changes influencing the C redistribution processes further.
Materials and methods
Study site. The experiment was established at the National Observation Station of Hailun Agro-ecology System, Heilongjiang province, China (47° 27′N, 126° 55′E) (Hailun Station). The experimental site is located in the center region of the Mollisols in Northeast China. The study site was very flat, with the slope less than 2 degree. The region has a temperate continental monsoon climate characterizing by simultaneously hot and rainy period from June to September in the growing season. Then the temperature sharply declined below 0 °C since October. The mean annual precipitation is 550 mm and mean annual air temperature is 1.5 °C. The soil is classified as Phaeozems 54 , which was developed from Loess parent material. The soil texture is predominantly silt and clay with the total content higher than 70% 14 (Table 1). It is the unique climate condition, 2:1 clay nimerals (mainly vermiculite, smectite and illite), and the accumulation of relatively large amounts of organic matter in the soil, that forms the extremely fertile Mollisols with relatively high organic carbon content 28,29,55 . Experimental design. A long-term vegetation restoration experiment was established in 1985. The selected soil for this experiment was under crop cultivation for more than 100 years before simulated second restoration treatments began in 1985. The initial arable land was divided into three adjacent sections, simulated restored grassland, an overgrazed bareland, and a cropland under continuous soil tillage. The grassland (1120 m 2 in size) was naturally restored without any fertilizer or tillage. Leymus chinensis gradually became the dominant grass species. The estimated shoot and root biomass were about 4-9 t ha −1 and 12-43 t ha −1 , respectively (data not published). The bareland (1120 m 2 in size) plot was maintained by periodically removing plants by hand hoeing to simulate the effects of overgrazing or extreme land degradation. The cropland (1050 m 2 in size) was maintained as a continuously tilled arable soil under a 3-year soybean (Glycine max (L.) Merrill.) -maize (Zea mays L.) -wheat (Triticum aestivum L.) rotation since 1985 and was subject to a randomized block design using three fertilization management practices with four replicates (87 m 2 of each treatment plot). In this study, only the cropped treatment without chemical fertilizer and organic amendments was selected to compare with bareland and grassland, under the consideration of eliminating the effects of chemical fertilization on soil aggregate. In cropland, all crop residues were removed from the plot following harvest to simulate the common practice locally. The soil underwent conventional tilling three or four times per year, including spring disking before planting, harrowing during growing season, and autumn moldboard plowing to a depth of 0.2 m after crop harvest. www.nature.com/scientificreports www.nature.com/scientificreports/ Soil sampling and analysis. The bareland and grassland field blocks were divided equally into four randomly distributed plots representing four field treatment replicates. Eight randomized soil cores (2.64 cm in diameter) per replicate were taken from a depth of 0-10 cm to make a composite soil sample after harvest of soybean on October 3, 2016. Sampling was performed in such a way so as not to break soil aggregates and to minimize compression. The soils were put into cloth bags, transported to the laboratory, part of the fresh soil samples were stored at 4 °C for soil microbial biomass carbon (SMBC), and another part were air-dried. A total of 12 field soil samples were collected. First, the air-dried bulk soil samples were broken down and then passed through an 8-mm sieve after manual elimination of visible plant litter and root residues prior to fractionation and analysis. Considering the small mass of archive soil representing the initial soil characters of the experiment in 1985, the 1985 soil sample was only used for water-stable aggregate mass fractionation, without further SOC fractionations within aggregates. The selected soil properties of the field soils and initial soil in 1985 were represented in Table 1.
Soil bulk density (SBD) was measured after drying soil cores at 105 °C for 48 h. SMBC was determined using the fumigation -extraction method 56 . The OC from the fumigated (24 h) and non-fumigated (control) soil were measured using a C analyzer (Shimadzu Model TOC-V). The SMB-C was calculated using a k EC factor of 0.45 57 .
SOC fractionation. The air-dried soil samples were used to physically fractionate the soil aggregates by wet sieving 14,23 . First, 100-g aliquots of soil samples that had been dried at 40 °C overnight were placed on the top of a column of three sieves measuring 2, 0.25, and 0.053 mm. Then, the sieves were immersed in deionized water for 3 min and gently moved up and down by hand for a total of 50 cycles over 2 min. The remains on the sieves and those passing through the 0.053 mm sieve were collected separately. The sieving process was repeated to obtain sufficient material for density and humic substance fractionation analysis for each aggregate size fraction. The final sieving was performed to determine aggregate size distribution, total organic C, and total N in each fraction after drying in an oven at 40 °C overnight, then weighed to determine the mass distribution among the aggregate size classes, namely, large macroaggregate (>2 mm), small macroaggregate (0.25-2 mm), microaggregate (0.053-0.25 mm), and soil silt-clay unit (<0.053 mm).
Density and humic fractionation of OC within different aggregate sizes were carried out according to the method described by Golchen et al. 58 , Yamashita et al. 3 , and Dou et al. 59 . First, a 10-g soil aggregate sample was put into a 100-mL centrifuge tube with 50 mL of sodium iodide (NaI) solution (d = 1.8 g cm −3 ) and left at 20 °C overnight. The following day, after centrifugation for 15 min at 3500 revolutions·min −1 , the supernatant was passed through a membrane filter (0.45 μm) into a millipore vacuum unit. The fraction recovered on the filter was washed with a 50 mL 0.01 M calcium chloride (CaCl 2 ) solution and 100 mL distilled water and then moved to a pre-weighed 50-mL beaker. The obtained fraction was dried at 50 °C to a constant weight and used as the free light fraction (fLF). Then, the residual soil was mixed with 50 mL of NaI solution, placed in an ice bath, and sonicated at 300 J·mL −1 using a probe-type ultrasonic disintegrator. After separation, the procedure was repeated. The obtained fraction was dried at 50 °C to a constant weight and used as the occluded light fraction (oLF). The leftover soil in the centrifuge tube was washed with distilled water until the water became clear, then oven dried at 50 °C to a constant weight and used as the heavy fraction (HF).
To attain humic substances within aggregates, a 2-mm-sieved aggregate sample was extracted using a solution of 0.1 M sodium pyrophosphate (Na 4 P 2 O 7 ) and 0.1 M sodium hydroxide (NaOH) mixture (pH 13). The soil to extractant ratio was 1:6. The mixture was shaken in a thermostatic water bath oscillator for 1 h. The supernatant and particles were separated through centrifugation for 15 min at 3500 revolutions·min −1 to extract HA and FA. The extraction supernatant was filtered into a volumetric flask and diluted to 50-mL volume using distilled water. The supernatant was decanted, then acidified with 1 M sulfuric acid (H 2 SO 4 ) to pH 1 and left for 24 h at room temperature to precipitate the HA. The HA fraction was dissolved by 0.05 M H 2 SO 4 and 0.05 M NaOH. Titration was used to analyze HA and FA fractions. The residual soil particles (Humin) were analyzed with the VarioEL CHN elemental analyzer.
Total organic C contents in bulk soil, soil aggregates, and all OC fractions within different sized aggregates were analyzed using a VarioEL CHN elemental analyzer (Heraeus Elementar VarioEL, Hanau, Germany). The Mollisol is free of carbonates. where, S i is the average diameter (mm) between (i-1)th and ith aggregate size and P i is mass fraction of ith aggregate size.
The organic carbon content of different carbon fraction within each aggregate was calculated using the Eq. (2): where C i is the organic C content of ith carbon fraction (g kg −1 ), C ci is the organic C concentration of ith C fraction (g kg −1 ), P i is the proportion of ith C fraction within each aggregate. The ratio of OC content in each C fraction within aggregate to total SOC was calculated as the proportion of OC in total SOC (Fig. 3).
One-way analysis of variance with Tukey honestly significant difference (HSD) as a post hoc was used for means separation of SOC and total nitrogen (TN) contents, aggregate fractions, density, and humic fractions among the three field treatments and the initial soil in 1985 at P < 0.05. Principal component analysis (PCA) of | 2020-06-03T14:31:10.302Z | 2020-06-03T00:00:00.000 | {
"year": 2020,
"sha1": "3d44c0b802763a20c4f965d50581ea70fa4e586e",
"oa_license": "CCBY",
"oa_url": "https://www.nature.com/articles/s41598-020-66038-1.pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "3d44c0b802763a20c4f965d50581ea70fa4e586e",
"s2fieldsofstudy": [
"Agricultural And Food Sciences"
],
"extfieldsofstudy": [
"Chemistry",
"Medicine"
]
} |
258428438 | pes2o/s2orc | v3-fos-license | CYP4F2 is a human-specific determinant of circulating N-acyl amino acid levels
N-acyl amino acids are a large family of circulating lipid metabolites that modulate energy expenditure and fat mass in rodents. However, little is known about the regulation and potential cardiometabolic functions of N-acyl amino acids in humans. Here, we analyze the cardiometabolic phenotype associations and genomic associations of four plasma N-acyl amino acids (N-oleoyl-leucine, N-oleoyl-phenylalanine, N-oleoyl-serine, and N-oleoyl-glycine) in 2351 individuals from the Jackson Heart Study. We find that plasma levels of specific N-acyl amino acids are associated with cardiometabolic disease endpoints independent of free amino acid plasma levels and in patterns according to the amino acid head group. By integrating whole genome sequencing data with N-acyl amino acid levels, we identify that the genetic determinants of N-acyl amino acid levels also cluster according to the amino acid head group. Furthermore, we identify the CYP4F2 locus as a genetic determinant of plasma N-oleoyl-leucine and N-oleoyl-phenylalanine levels in human plasma. In experimental studies, we demonstrate that CYP4F2-mediated hydroxylation of N-oleoyl-leucine and N-oleoyl-phenylalanine results in metabolic diversification and production of many previously unknown lipid metabolites with varying characteristics of the fatty acid tail group, including several that structurally resemble fatty acid hydroxy fatty acids. These studies provide a structural framework for understanding the regulation and disease associations of N-acyl amino acids in humans and identify that the diversity of this lipid signaling family can be significantly expanded through CYP4F-mediated ω-hydroxylation.
N-acyl amino acids are a large family of circulating lipid metabolites that modulate energy expenditure and fat mass in rodents. However, little is known about the regulation and potential cardiometabolic functions of N-acyl amino acids in humans. Here, we analyze the cardiometabolic phenotype associations and genomic associations of four plasma N-acyl amino acids (N-oleoyl-leucine, N-oleoyl-phenylalanine, Noleoyl-serine, and N-oleoyl-glycine) in 2351 individuals from the Jackson Heart Study. We find that plasma levels of specific N-acyl amino acids are associated with cardiometabolic disease endpoints independent of free amino acid plasma levels and in patterns according to the amino acid head group. By integrating whole genome sequencing data with N-acyl amino acid levels, we identify that the genetic determinants of N-acyl amino acid levels also cluster according to the amino acid head group. Furthermore, we identify the CYP4F2 locus as a genetic determinant of plasma N-oleoyl-leucine and N-oleoyl-phenylalanine levels in human plasma. In experimental studies, we demonstrate that CYP4F2-mediated hydroxylation of N-oleoylleucine and N-oleoyl-phenylalanine results in metabolic diversification and production of many previously unknown lipid metabolites with varying characteristics of the fatty acid tail group, including several that structurally resemble fatty acid hydroxy fatty acids. These studies provide a structural framework for understanding the regulation and disease associations of N-acyl amino acids in humans and identify that the diversity of this lipid signaling family can be significantly expanded through CYP4F-mediated ω-hydroxylation.
The N-acyl amino acids are a large and structurally diverse family of circulating lipid signaling molecules. These lipid metabolites are unusual peptide conjugates of fatty acids and amino acids. In mice, levels of specific circulating N-acyl amino acids are tightly controlled by the action of two enzymes, PM20D1 and fatty acid amide hydrolase (FAAH), which catalyze bidirectional N-acyl amino acid synthesis from and hydrolysis to free fatty acids and free amino acids (1)(2)(3)(4). Pharmacological, genetic, and mechanistic studies in rodents suggest that certain members of the N-acyl amino acids stimulate mitochondrial respiration and whole-body energy expenditure (1). Other complementary studies have also established roles for N-acyl amino acids in glucose homeostasis (2), adipogenesis (5), vascular tone (6,7), and bone homeostasis (8,9). Importantly, the functional consequence and enzymatic regulation of each N-acyl amino acid are highly dependent on the structural properties of the fatty acid tail group and amino acid head group. For example, even subtle modification of the lipid chain or amino acid moieties can affect N-acyl-amino acid substrate specificity for PM20D1 and ligand potency in mitochondrial respiration (1).
Despite the considerable body of rodent literature on N-acyl amino acids, our knowledge of the regulation and clinical associations of these lipid metabolites in humans is still lacking. Two small studies have previously investigated the role of PM20D1 in the regulation of N-acyl amino acids in human plasma. In one report, plasma N-oleoyl-leucine and N-oleoylphenylalanine levels were found to be associated with circulating PM20D1 protein levels and correlated positively with adiposity and several parameters of glucose homeostasis (10). However, a second study failed to identify any association of serum levels of two N-acyl amino acids, N-oleoyl-leucine, or N-oleoyl-glutamine with genetic variants in the PM20D1 locus though in a very small cohort (11). The conflicting human data, as well as the scope of phenotypes and the limited number of individuals examined in the previous two studies, therefore, motivate additional investigations as to the clinical relevance and genetic underpinnings of this class of molecules.
Our group recently measured circulating levels of four N-acyl-amino acids (N-oleoyl-glycine, N-oleoyl-serine, N-oleoyl-leucine, and N-oleoyl-phenylalanine) in plasma samples from 2351 participants of the Jackson Heart Study (JHS) using liquid chromatography-mass spectrometry (LC-MS) (12). In two broad metabolomic-wide association studies in this population, we identified that levels of N-oleoyl serine and N-oleoyl leucine were among the top metabolites in human plasma associated with future risk of coronary heart disease (13) and heart failure (12), respectively. These observations extend prior studies in murine models and suggest that additional investigations of N-acyl amino acid biology may provide important new insights into human metabolism and cardiometabolic disease.
Here, we provide a detailed analysis of the role of the oleic fatty acid tail group and amino acid head group in driving associations with cardiometabolic disease. We use genomewide association studies (GWAS) to distinguish between different molecular species of N-acyl amino acids according to the amino acid head group. Further, this analysis identifies a previously unknown enzymatic regulator of N-oleoyl-leucine and N-oleoyl-phenylalanine, CYP4F2, in humans. Finally, we use untargeted metabolomics in live cells to map the fate of Noleoyl-leucine and N-oleoyl-phenylalanine downstream of CYP4F2 hydroxylation and identify numerous previously unknown lipid metabolites with varying characteristics of the fatty acid tail group, including several that structurally resemble fatty acid hydroxy fatty acids (FAHFAs), a recentlydiscovered family of bioactive lipids that signal through specific G protein-coupled receptors to improve glucose-insulin homeostasis and block inflammatory cytokine production (14). Our data integrating mass spectrometry, human phenotyping, genetics, and model systems provide new insights into this emerging class of molecules.
Results
N-acyl amino acids are associated with cardiometabolic disease independent of free amino acid plasma levels and according to the amino acid head group We recently measured circulating levels of N-oleoyl-glycine, N-oleoyl-serine, N-oleoyl-leucine, and N-oleoyl-phenylalanine in plasma samples from 2351 participants of the JHS using LC-MS, as described (12). The baseline characteristics of the JHS study participants are detailed in Table S1. Age-and sexadjusted analyses identified N-oleoyl-serine and N-oleoylleucine as top metabolites associated with the future risk of coronary heart disease (13) and heart failure (12), respectively. However, prior work did not examine whether the associations of the intact N-acyl oleic acid conjugate of each amino acid were independent of the corresponding free amino acid (Fig. 1A). Using age-, sex-, and batch-adjusted Cox regression models, we determined that the association between N-oleoylserine and the future risk of coronary heart disease remained unaffected when further adjusted for free serine (hazard ratio [HR] 0.81 per 1 SD increment in transformed and normalized metabolite level; 95% CI, 0.69-0.94; p-value = 7.3 × 10 −3 ; median follow-up 11.6 years). Similarly, the association between N-oleoyl-leucine and future heart failure remained highly significant when further adjusted for free leucine (HR 0.78; 95% CI, 0.66-0.91; p-value = 2.1 × 10 −3 ; median followup 12.6 years). These observations may point towards additional biological mechanisms conferred by the oleic acid tail group of N-acyl-serine and N-acyl-leucine.
Genetic loci associated with levels of N-acyl amino acids in human plasma
In order to determine if the genetic determinants of N-acyl amino acid plasma levels also differ according to amino acid head group, we leveraged available whole genome sequencing data (NHGRI-EBI GWAS Catalog study GCP000239 (15)) in the JHS to compare genome-wide associations between oleic acid N-acyl conjugates of serine, glycine, leucine, and phenylalanine, as described (16). As shown in Figure 2A, we identified a strong association between the FAAH genetic locus and plasma levels of N-oleoyl-glycine (sentinel variant rs324420, p-value = 5.7 × 10 −64 , beta 0.54, n = 2463) and Noleoyl-serine (sentinel variant rs324420, p-value = 3.6 × 10 −32 , beta 0.38, n = 2101). These findings are consistent with studies in cell-and murine-based systems that have identified FAAH as an intracellular N-acyl amino acid synthase/hydrolase. Interestingly, the sentinel variant for both metabolites was a common (observed MAF = 0.21 in JHS and global MAF 0.21 as reported in the NCBI Allele Frequency Aggregator), missense variant (chr1:46405089C > A; Pro129Thr) that has previously been shown to reduce FAAH stability and enzymatic activity in cell-based model systems and has been associated with overweight and obesity in human studies (17,18).
Taken together, these human genetic data point to specific pathways that regulate circulating N-acyl amino acid levels according to the amino acid head group. In particular, the association of the CYP4F2 locus with circulating N-oleoyl- Figure 1. N-acyl amino acids are associated with cardiometabolic disease independent of free amino acid plasma levels and according to the amino acid head group. A, Age-and sex-adjusted associations between plasma levels of N-oleoyl-serine and future risk of coronary heart disease (13) (top) and N-oleoyl-leucine and future risk of heart failure (12) (bottom) in JHS were unaffected by further adjustment for circulating levels of the corresponding free amino acid using cox regression models. The relationship between plasma levels of each N-acyl amino acid (log transformed and standardized) and cardiometabolic traits (standardized) was analyzed using age-and sex-adjusted mixed linear regression models for continuous variables measured in the JHS (B) and age-and sex-adjusted logistic regression models for categorical variables in the JHS (C). Estimated β coefficients are represented by a color scale from red to blue (red represents positive and blue represents negative association). The area of each node corresponds to −log10(p-values). Bonferroniadjusted p-value = 6.0 × 10 −4 (0.05/21 tested traits/4 analytes).
leucine and N-oleoyl-phenylalanine was unexpected, and the role of CYP4F2 in N-acyl amino acid metabolism has not been explored.
CYP4F2-mediated omega-hydroxylation of N-oleoyl-leucine and N-oleoyl-phenylalanine
Canonically, CYP4F2 catalyzes the omega (e.g., terminal) hydroxylation of specific lipophilic substrates, including fatty acids (e.g., arachidonic acid) (19) and vitamins (e.g., vitamin E and K) (20,21). The identification of CYP4F2 as a potential enzymatic regulator of N-oleoyl-leucine/phenylalanine was surprising as this enzyme had not been previously implicated in the metabolism of fatty acid-amino acid conjugates. Nevertheless, we reasoned that CYP4F2 might also catalyze the omega hydroxylation of N-oleoyl-leucine and N-oleoylphenylalanine (Fig. 3A). To directly test this hypothesis in vitro, we first generated lysates overexpressing CYP4F2 by transfection of plasmids expressing human CYP4F2 into HEK293T cells (Fig. 3B, left). Transfected whole-cell lysates were then assayed for the ability to catalyze N-oleoyl-leucine and N-oleoyl-phenylalanine hydroxylation in the presence of NADPH. As shown in Figure 3B (right), CYP4F2-transfected cell lysates exhibited a robust hydroxylation activity on both N-acyl amino acids, leading to the production of two previously unknown lipid species in vitro, hydroxy-oleoyl-leucine and hydroxy-oleoyl-phenylalanine. The absolute rates of CYP4F2-mediated hydroxylation of these N-acyl amino acids were similar to that of a canonical CYP4F2 substrate, arachidonic acid (Fig. S1A). Additionally, we observed increasing formation rates of hydroxy-oleoyl-leucine and hydroxy-oleoylphenylalanine across a range of N-oleoyl-leucine/phenylalanine substrate concentrations (Fig. 3C). Notably, this range of low-mid μM concentrations of N-oleoyl-leucine/phenylalanine correlate with prior studies that detected functional effects of these N-acyl amino acids on energy homeostasis at similar murine model plasma levels and cell-culture treatment doses (1). The use of CYP4F2-transfected lysates precluded the ability to calculate the catalytic parameters, selectivity, or specificity of this reaction. These data demonstrate that CYP4F2 catalyzes the hydroxylation of N-oleoyl-leucine and N-oleoyl-phenylalanine in vitro and provide a biochemical CYP4F2 regulates N-acyl amino acids explanation for the observed association between the CYP4F2 gene locus and circulating N-oleoyl-leucine and N-oleoylphenylalanine levels in humans.
The sentinel variant in the CYP4F2 locus that was associated with N-oleoyl-leucine and N-oleoyl-phenylalanine plasma levels in our human studies was a chr19:15879621C > T CYP4F2 regulates N-acyl amino acids missense variant that results in a Val433Met substitution (Fig. 2B). To examine the consequences of the specific CYP4F2 (V433M) variant on N-oleoyl-leucine and N-oleoyl-phenylalanine hydroxylation activity, we performed similar in vitro hydroxylation activity assays with CYP4F2 (V433V) or CYP4F2 (V433M) enzymes. Equivalent transfection for both alleles was confirmed by equivalent mRNA levels (Fig. 3D). Under these conditions, both the CYP4F2 protein as well as N-oleoylleucine, N-oleoyl-phenylalanine, and arachidonic acid hydroxylation activities of CYP4F2(V433M) were lower than that of CYP4F2(V433V) (Figs. 3D and S1B). To determine the relative activity of CYP4F2 (V433V) or CYP4F2 (V433M) under conditions of equal protein loading, we normalized levels of CYP4F2 (V433M) and CYP4F2 (V433V) protein as assessed by Western blotting (Fig. 3E). CYP4F2 (V433M) exhibited reduced hydroxylation of N-oleoyl-leucine, N-oleoyl-phenylalanine, and arachidonic acid compared to CYP4F2 (V433V) (Figs. 3E and S1C). These data demonstrate that CYP4F2 (V433M) has both lower catalytic efficiency and reduced protein levels relative to CYP4F2 (V433V).
N-oleoyl-leucine and N-oleoyl-phenylalanine are competitive inhibitors of CYP4F2
The ability of CYP4F2 to hydroxylate multiple endogenous lipid substrates raises the possibility that active site competition might modulate the relative flux of each substrate through CYP4F2 (Fig. 4A). To mechanistically examine whether such substrate competition might occur, we used our in vitro CYP4F2 enzyme activity assay to directly measure substrate competition between N-oleoyl-leucine and N-oleoyl-phenylalanine on the canonical CYP4F2 substrate arachidonic acid. At a 1:1 M ratio, N-oleoyl-leucine efficiently inhibited CYP4F2-mediated conversion of arachidonic acid to 20-HETE (86% suppression, Fig. 4B). Similar results were observed with N-oleoyl-phenylalanine (76% suppression, Fig. 4B). A doseresponse curve demonstrated that either N-oleoyl-leucine or N-oleoyl-phenylalanine could inhibit CYP4F2-mediated arachidonic acid hydroxylation even at substoichiometric levels (10 mol%, Fig. 4B). To examine the potential generality of N-oleoyl-leucine and N-oleoyl-phenylalanine inhibition of CYP4F2, we performed similar in vitro competition experiments with two additional CYP4F2 substrates, docosanoic acid, and 8-HETE. As shown in Figure 4, C and D, both Noleoyl-leucine and N-oleoyl-phenylalanine competed CYP4F2mediated hydroxylation of both 8-HETE and docosanoic acid but with differences in potency. For instance, even at substoichiometric levels (10 mol%), both N-oleoyl-leucine and Noleoyl-phenylalanine strongly suppressed CYP4F2-mediated docosanoic acid hydroxylation (87% inhibition by N-oleoylleucine and 81% inhibition by N-oleoyl-phenylalanine, Fig. 4C). In contrast, superstoichiometric levels (10-fold excess) of either N-acyl amino acid were required to observe similar levels of competition of 8-HETE hydroxylation (94% inhibition by N-oleoyl-leucine and 84% inhibition by N-oleoylphenylalanine, Fig. 4D).
Conversely, we investigated if the canonical substrates arachidonic acid, docosanoic acid, and 8-HETE might also compete CYP4F2-catalyzed N-acyl amino acid hydroxylation. While arachidonic acid was able to compete N-oleoyl-leucine and N-oleoyl-phenylalanine hydroxylation by CYP4F2 (Fig. 4E), neither docosanoic acid nor 8-HETE exhibited strong inhibition of either N-oleoyl-leucine or N-oleoyl-phenylalanine, even at superstoichiometric levels (10:1 competitor:substrate, Fig. 4, F and G). We therefore conclude that N-oleoyl-leucine and Noleoyl-phenylalanine and other canonical substrates engage in bidirectional competition to regulate hydroxylation flux at the CYP4F2 enzyme node. The different dose responses observed for competition between each substrate pair potentially reflect differences in effective concentrations and/or substrate affinities at the CYP4F2 active site.
Metabolic diversification of N-acyl amino acids downstream of CYP4F2
CYP4F2-catalyzed substrate hydroxylation can result in either metabolic diversification (e.g., arachidonic acid to 20-HETE) or oxidative degradation (e.g., vitamin E). To determine which of these metabolic outcomes was the relevant pathway for the N-acyl amino acids, we used untargeted lipidomics in live cells to map the fate of N-oleoyl-leucine and Noleoyl-phenylalanine downstream of CYP4F2. For these studies, we initially focused on N-oleoyl-leucine. First, CYP4F2 or mocktransfected HEK293T cells were treated with N-oleoyl-leucine for 4 h. Total intracellular lipids were extracted by the Folch method and differential peaks were identified using XCMS (22) (Fig. 5A). Two statistically significant peaks of high fold change (p < 0.05, >20-fold) were enriched in cells overexpressing CYP4F2. As expected, the top-scoring peak of m/z = 410.3271 matched ω-hydroxy-oleoyl-leucine (expected m/z = 410.3276) (Fig. 5B). The second peak of mass m/z = 674.5718 was also highly enriched in CYP4F2-transfected cells, but its chemical structure remained initially unknown. The mass difference observed between the two peaks (264.2442) as well as the later retention time of the second peak was consistent with an additional lipidation by oleate (+C 18 H 32 O, expected +264.2453). We therefore hypothesized that this second peak represented an unusual and previously unknown very long chain fatty acid, oleic acid-hydroxy-oleoyl-leucine (OAHOL). We used chemical synthesis to generate authentic standards for both hydroxyoleoyl-leucine and OAHOL (see Experimental Procedures). As shown in Figure 5, C and D, fragmentation of both synthetic hydroxy-oleoyl-leucine and the endogenous m/z = 410.33 peak gave an identical daughter ion at m/z = 130.087, which matched the mass of the leucine anion. In addition to demonstrating the existence of these metabolites in cell lysates, we confirmed the presence of endogenous hydroxy-oleoyl-leucine in human plasma as well (Fig. 5C, lower panel). Similarly, fragmentation of both synthetic OAHOL and the endogenous m/z = 674.57 mass gave daughter ions at m/z = 130.087 (leucine anion) and m/z = 281.249 (oleate). We therefore conclude that m/z = 410.33 and m/z = 674.57 correspond to hydroxy-oleoyl-leucine and OAHOL, respectively.
OAHOL is a novel lipid metabolite that is not found in any public databases (23,24). This lipid metabolite is presumably formed by enzymatic acylation of hydroxy-oleoyl-leucine with oleoyl-CoAs. Based on this metabolic pathway hypothesis, we manually examined our dataset for additional fatty acyl-derivatives of ω-hydroxy-oleoyl-leucine. Beyond OAHOL, we also identified PAHOL (palmitic acid-hydroxy-oleoyl-leucine) and SAHOL (stearic acid-hydroxy-oleoyl-leucine) dramatically elevated in CYP4F2-transfected but not mock-transfected cells (Figs. 5E and S3, A and B). To determine if these very long-chain lipid derivatives could also be formed downstream of CYP4F2-mediated hydroxylation of N-oleoyl-phenylalanine, we performed similar untargeted experiments in live CYP4F2-transfected cells with Noleoyl-phenylalanine as a substrate. As shown in Figure 5F, addition of N-oleoyl-phenylalanine also resulted in the downstream production of ω-hydroxy-oleoyl-phenylalanine and oleoyl-and palmitoyl-conjugates of hydroxy-oleoyl-phenylalanine (oleic acidhydroxy-oleoyl-phenylalanine and palmitic acid-hydroxy-oleoylphenylalanine, respectively). For N-oleoyl-phenylalanine, the stearoyl acylation product was not detected. Taken together, we conclude that CYP4F2-catalyzed hydroxylation of N-acyl amino acids leads to the metabolic diversification of a wide array of previously unknown lipids, including several very long chain lipids that structurally resemble previously reported anti-diabetic FAHFAs (14), Fig. S3C).
Discussion
By integrating genetic and clinical data with circulating levels of plasma N-acyl amino acids from a large human CYP4F2 regulates N-acyl amino acids cohort, our study provides several insights of potential importance into the regulation of N-acyl amino acids in humans. First, individual plasma N-acyl amino acid levels are associated with cardiometabolic disease endpoints independent of free amino acid plasma levels and in patterns according to the amino acid head group. Second, the underlying genetic architecture and biological pathways that may regulate plasma levels of N-acyl amino acids also differ according to the amino acid head group. Finally, CYP4F2 functions as a humanspecific enzyme node that catalyzes the metabolic diversification of N-acyl amino acids into a much larger family of lipids with varying characteristics of the fatty acid tail group, including several that structurally resemble FAHFAs.
Our large-scale human study revealed novel associations between the N-acyl amino acids and specific cardiometabolic diseases in participants of the JHS. Notably, we identified that the associations between fasting baseline levels of N-oleoylserine and future coronary heart disease, as well as N-oleoylleucine and future heart failure, were independent of the plasma levels of free serine and leucine, respectively. This suggests that these associations are driven by the intact N-acyl oleic acid conjugate of each amino acid, rather than levels of the free amino acid themselves. This may suggest that N-acyl amino acids affect cardiometabolic outcomes through distinct biological mechanisms from those mediated by circulating levels of the corresponding free amino acid.
Interestingly, when analyzing associations of N-acyl-amino acids with the prevalent cardiometabolic disease at baseline, we detected a bifurcation in associations according to the amino acid head group, with N-oleoyl-glycine and serine associated with cardiometabolically "advantageous" traits (such as prevalent diabetes), and N-oleoyl-leucine and phenylalanine associated with "disadvantageous" cardiometabolic traits (such as obesity). This division mirrors positive associations between branched chain (e.g. leucine) and aromatic (e.g. phenylalanine) free amino acids with obesity and insulin resistance, and inverse associations between glycine and serine with impaired glucose tolerance and risk of diabetes (25)(26)(27)(28)(29)(30)(31)(32)(33). The mechanisms linking these free amino acid species to cardiometabolic traits and risk of the disease remain to be fully elucidated but may involve various opposing effects on regulatory pathways upstream of pancreatic islet β-cell insulin secretion (25,(34)(35)(36)(37)(38). Whether the N-oleoyl species of these amino acids-as well as the CYP4F2-mediated derivatives of these metabolites-act through the same or different pathways adds a potential additional layer of complexity to this regulatory balance.
Our studies of human N-acyl amino acid metabolism highlight the utility of integrating metabolomic profiling data with genetic data for pathway discovery. In mice, FAAH and PM20D1 are two major enzymes that catalyze the bidirectional N-acyl amino acid synthesis and hydrolysis from free amino acids and free fatty acids. Human-based studies have established FAAH as the primary degradative enzyme of the structurally-related N-acyl ethanolamines (39). Therefore, the association between the FAAH locus and plasma levels of N-oleoyl-glycine and N-oleoyl-serine in participants of the JHS was expected. However, the absence of an association of N-acyl amino acids to the human PM20D1 locus, as well as the identification of the CYP4F2 gene as a novel and humanspecific association, were both unexpected. We validated the novel genetic association between N-oleoyl-leucine and Noleoyl-phenylalanine with the CYP4F2 locus by demonstrating that these two N-acyl amino acids are both bona fide substrates and inhibitors of the human CYP4F2 enzyme in vitro and in cell-based models. The top association for both N-oleoyl-leucine and N-oleoyl-phenylalanine was with a missense variant (chr19:15879621C > T; Val433Met) that has previously been shown to result in reduced CYP4F2 protein levels and enzymatic activity in cell-based model systems, but that has not previously been tied to N-acyl amino acid metabolism (20,40,41). This variant is common with a global MAF of 0.29 and a subgroup MAF of 0.09 in African Americans, as reported in the NCBI Allele Frequency Aggregator.
While it is possible that our study did not detect an association between the PM20D1 locus and N-acyl amino acids due to cohort-specific characteristics or sample size, we note that none of the measured N-acyl amino acids demonstrated even a modest association (nominal p-value ≤ 0.001) with a genetic variant between the PM20D1 transcriptional start and end sites. This finding will require validation in additional cohorts, however, may suggest that human PM20D1 controls local, and possibly tissue-specific extracellular paracrine pools of N-acyl amino acids that do not interact with the levels found in the circulation.
It is notable that we detected very strong associations between the CYP4F2 locus on chromosome 19 and N-oleoylleucine and N-oleoyl-phenylalanine with no appreciable association signals between this locus and either N-oleoylglycine or N-oleoyl-serine (p > 0.001 for all variants between the CYP4F2 transcriptional start and end sites). Importantly, the biochemical basis for this genetic specificity remains the subject of future study. We have previously found that subcellular location, and even organ-specific substrate accessibility can contribute to the regulation of specific N-acyl amino acids by FAAH and PM20D1 (4). Future detailed enzymology studies of CYP4F2 may uncover additional fundamental biochemical characteristics of this enzyme that underlie these specific genetic associations.
Our data suggest that CYP4F2 appears to be a key enzymatic node at the center of a complex lipid network that contains classical bioactive lipids, such as arachidonate and 20-HETE, with more novel lipid species, including N-acyl amino acids, hydroxylated N-acyl amino acids, and very long chain fatty acyl derivatives of N-acyl amino acids that structurally resemble FAHFAs. Notably, we detect that examples of these, including hydroxylated oleoyl-leucine, are present in human plasma.
Currently, the precise biochemical and functional relationship between all of these lipid species, especially in a complex tissue environment such as a human liver, remains unknown. For instance, it may be possible that hydroxy N-acyl amino acids exhibit similar vasoactive effects as 20-HETE, and fatty acylated hydroxy-N-acyl amino acid derivatives might also function similarly to that of the anti-diabetic FAHFAs.
Alternatively, N-acyl amino acids may indirectly modulate the levels of other CYP4F2-regulated lipids via substrate competition. An important area of future work will be to understand the relative fluxes of each lipid class through CYP4F2 as well as the potential functional roles of these new N-acyl amino acid derivatives.
The integration of human functional genomics with untargeted lipidomics studies of cultured cells further allowed for the discovery of a novel class of very long-chain lipids. These compounds structurally resemble FAHFAs, although instead of containing a characteristic branched hydroxyl linkage, contain a terminal ester linkage between a fatty acid and a hydroxy N-acyl amino acid. A similar terminal linkage has been described in O-acyl-ω-hydroxy fatty acids (ωOAHFAs); however, these unusual species have only recently been detected in human skin (42) and meibum (43), and not to our knowledge in human plasma. Further, we are not aware of a previous report of fatty acyl derivatives of hydroxy N-acylamino acids. FAHFAs are a diverse family of bioactive lipids that have recently been shown to signal through specific G protein-coupled receptors and other pathways to improve glucose-insulin homeostasis and block inflammatory cytokine production (14). Our data provide new insights into this emerging class of molecules and raise the intriguing possibility that ω-hydroxy-oleoyl amino acid species may also serve as signaling effectors in human metabolism.
Experimental procedures
Human cohort JHS is a prospective population-based observational study designed to investigate risk factors for cardiovascular disease (CVD) in Black individuals, as previously described (44). From 2000 to 2004, 5306 Black individuals from the Jackson, Mississippi tri-county area (Hinds, Rankin and Madison counties), were recruited for a baseline examination. Of the original cohort, 2351 individuals had metabolomic profiling of N-acyl amino acids performed from baseline samples and were included in the analyses. Details regarding the collection and calculation of clinical data included in Table S1 and Figure 1 have been previously described (13). All clinical data were collected during the baseline exam (Exam 1), except visceral adipose tissue (Exam 2), subcutaneous adipose tissue (Exam 2), coronary artery calcium score (Exam 2), and abdominal aortoiliac calcium score (Exam 2).
Study approval
The Institutional Review Boards of Beth Israel Deaconess Medical Center and the University of Mississippi Medical Center approved the human study protocols, and all participants provided written informed consent. All procedures involving human participants were in accordance with the ethical standards of the 1964 Helsinki Declaration and its later amendments.
N-acyl amino acid metabolite profiling in human plasma
Methods for metabolomics profiling in human plasma have been described (12). In brief, to measure N-oleoyl-leucine/ phenylalanine/glycine/and serine, chromatography was performed using an Agilent 1290 infinity LC system equipped with a Waters XBridge Amide column, coupled to an Agilent 6490 triple quadrupole mass spectrometer. To measure endogenous hydroxy-oleoyl-leucine, chromatography was performed using a Waters UPLC BEH Amide (1.7um, 1.0 × 150 mm) column. Metabolite transitions were assayed using a dynamic multiple reaction monitoring systems. LC-MS data were analyzed with Agilent Masshunter QQQ Quantitative analysis software. Isotope-labeled internal standards were monitored in each sample to ensure proper MS sensitivity for quality control. Pooled plasma samples were interspersed at intervals of ten participant samples to enable correction of drift in instrument sensitivity over time and to scale data between batches. A linear scaling approach was used to the nearest pooled plasma sample in the queue.
Genotyping
Whole genome sequencing (WGS) in JHS has been described (45). Briefly, participant samples underwent >30 × WGS through the Trans-Omics for Precision Medicine project at the Northwest Genome Center at University of Washington and joint genotype calling with participants in Freeze 6; genotype calling was performed by the Informatics Resource Center at the University of Michigan.
Whole genome association studies
Summary statistics from whole genome association studies of metabolite levels in plasma samples of participants of the JHS are available on the NHGRI-EBI GWAS Catalogue (Accession GCP000239) (15) and GWAS methods have been previously described (16). Briefly, metabolite LC-MS peak areas were log-transformed and scaled to a mean of zero and standard deviation of 1 and subsequently residualized on age, sex, batch, and principal components (PCs) of ancestry 1 to 10 CYP4F2 regulates N-acyl amino acids as determined by the GENetic EStimation and Inference in Structured samples (GENESIS) (46), and subsequently inverse normalized. The association between these values and genetic variants was tested using linear mixed effects models adjusted for age, sex, the genetic relationship matrix, and PCs 1 to 10 using the fastGWA model implemented in the GCTA software package (47). Variants with a minor allele count less than five in JHS were excluded from the analysis.
Cell line cultures
All cell lines were grown at 37 C with 5% CO 2 . HEK293T cells were obtained from ATCC and grown in Dulbecco's modified Eagle's medium with 10% fetal bovine serum and 1% penicillin/streptomycin (pen/strep).
Untargeted measurements of metabolites by LC-MS
Untargeted metabolomics measurements were performed on an Agilent 6520 Quadrupole Time-of-Flight (Q-TOF) LC/MS. Mass spectrometry analysis was performed using electrospray ionization (ESI) in negative mode. The dual ESI source parameters were set as follows, the gas temperature was set at 250 C with a drying gas flow of 12 l/min and the nebulizer pressure at 20 psi. The capillary voltage was set to 3500 V and the fragmentor voltage was set to 100 V. Separation of metabolites was conducted on a Luna 5 μm C5 100 Å, LC Column 100 × 4.6 mm (Phenomenex 00D-4043-E0) with normal phase chromatography. Mobile phases were as follows: Buffer A, 95:5 water/ methanol; Buffer B, 60:35:5 isopropanol/methanol/water with 0.1% ammonium hydroxide in both Buffer A and B for negative ionization mode. For 10 min runs, the LC gradient started at 95% A with a flow rate of 0.6 ml/min from 0 to 1 min. The gradient was then increased linearly to 95% B at a flow rate of 0.6 ml/min from 1 to 8 min. From 8 to 10 min, the gradient was maintained at 95% A at a flow rate of 0.6 ml/min. For 30 min runs, the LC gradient started at 95% A with a flow rate of 0.6 ml/min from 0 to 3 min. The gradient was then increased linearly to 95% B at a flow rate of 0.6 ml/min from 3 to 25 min. From 25 to 30 min, the gradient was maintained at 95% A at a flow rate of 0.6 ml/min.
Targeted metabolomics in cell culture samples
Targeted measurements were performed on an Agilent 6470 Triple Quadrupole (QQQ) LC/MS. Mass spectrometry analysis was performed using ESI in negative mode. The AJS ESI source parameters were set as follows, the gas temperature was set at 250 C with a gas flow of 12 l/min and the nebulizer pressure at 25 psi. The sheath gas temperature was set to 300 C with the sheath gas flow set at 12 l/min. The capillary voltage was set to 3500 V. Separation of metabolites was performed as described above in the untargeted metabolomics section. Multiple reaction monitoring (MRM) was performed for the indicated metabolites with the listed dwell times, fragmentor voltage, collision energies, cell accelerator voltages, and polarities.
Synthesis of ω-hydroxy-oleoyl-leucine 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC-HCl, 1.05 eq) was added to a cold mixture (ice bath) of ω-hydroxy-oleic acid (A2B Chem BG26794, 1 eq) and 1-hydroxybenzo-triazole monohydrate (HOBt H 2 O, 2 eq) in N,N-dimethyformamide (DMF). Stirring was continued for 10 min before a mixture of L-leucine ethyl ester hydrochloride (5 eq) and N,N-diiso-propylethylamine (DIPEA, 3 eq) in DMF was added. The ice bath was removed, and the reaction mixture was stirred for 16 h at room temperature. After removal of the solvent, the residue was taken up in ethyl acetate (EtOAc); washed with KHSO4, brine, NaHCO3, and CYP4F2 regulates N-acyl amino acids brine; and then concentrated under reduced pressure. The crude ester was dissolved in tetrahydrofuran, and 2 N of LiOH was added. The resulting mixture was stirred at room temperature for 4 h. The reaction was acidified to pH 2 to 3 by the addition of 1 M of HCl and then extracted with DCM and concentrated. The resulting crude product was used directly for LC-MS/MS analysis.
Synthesis of oleic acid-hydroxy-oleoyl-leucine
To a solution of oleic acid in c (DCM) was added oxalyl chloride (one drop) and one drop of DMF at 0 C. The mixture was stirred at room temperature for 2 h. The mixture was concentrated and dissolved in DCM then added to a suspension of ω-hydroxy-oleoyl-leucine and DIPEA (1 drop) in DCM. The reaction mixture was stirred at room temperature for 16 h and then acidified to pH 4 with 1 M HCl. The resulting mixture was extracted with DCM and washed with brine then the solvent was removed under reduced pressure. The crude product was analyzed directly by LC-MS/MS.
In vitro CYP42 assays HEK293T cells were co-transfected with CYP4F2 WT/ V433 M, POR, and CYB5R1. CYP4F2 and mock-transfected cells were harvested in PBS, lysed by sonication, and centrifuged (10 min at 15,000 rpm) to remove debris. In vitro enzymatic reactions were conducted in 96-well plates and initiated with 1 mM NADPH. The final reaction conditions were 100 μM substrate (N-oleoyl-leucine, Cayman Chemical 20064, N-oleoyl-phenylalanine, Cayman Chemical 28921, or arachidonic acid, Cayman Chemical 90010) and 50 μg protein in 50 μl PBS. After 1 h at 37 C, reactions were transferred to glass vials, quenched with 150 μl 2:1 v/v chloroform: methanol, and vortexed. Reaction vials were centrifuged (10 min at 1000 rpm) and the organic layer was transferred to a mass spec vial and analyzed by LC-MS as described above. For competition assays, cell lysates were incubated for 5 min with competitor (Arachidonic acid, docosanoic acid, Cayman Chemical 9000338, or 8-HETE, Cayman Chemical 34340) at 37 C before the addition of other substrates and initiation with NADPH.
Kinetic enzymatic assays CYP4F2 lysates were harvested from transiently transfected HEK293T cells as described above. In vitro enzymatic reactions were conducted in 96-well plates and initiated with 1 mM NADPH. The final reaction conditions were: 1 μM, 4 μM, 20 μM, and 100 μM of substrate and 0.46 μg CYP4F2 enzyme in 50 μl PBS. After 10 min at 37 C, reactions were transferred to glass vials quenched with 150 μl 2:1 v/v chloroform:methanol, and vortexed. Reaction vials were centrifuged (10 min at 1000 rpm) and the organic layer was transferred to a mass spec vial and analyzed by LC-MS as described above.
Live cell tracing experiments
Transiently transfected (CYP4F2 or mock) HEK293Ts were washed twice with PBS and then harvested by scraping. Cells were spun down (5 min at 1000 rpm) and resuspended in serum-free media then aliquoted into a 96-well plate (1.2 million cells per well). The final reaction conditions were 10 μM N-oleoyl-leucine or N-oleoyl-phenylalanine. Reactions were initiated with 1 mM NADPH. After 4 h at 37 C, reactions were transferred to glass vials, quenched with 2:1 v/v chloroform:methanol, and vortexed. After centrifuging vials (10 min at 1000 rpm), the organic layers were analyzed by LC-MS or LC-MS/MS as described above. LC-MS data were uploaded to Scripps XCMS Online to identify significantly changed metabolites.
Western blot analysis
Cells were collected and lysed by sonication in PBS. Cell lysates were centrifuged at 4 C for 10 min at 15,000 rpm to remove residual cell debris. Protein concentrations of the supernatant were normalized using the Pierce BCA protein assay kit and combined with 4 x NuPAGE LDS Sample Buffer with 10 mM DTT. Samples were then boiled for 10 min at 95 C. Prepared samples were run on a NuPAGE 4 to 12% Bis-Tris gel and then transferred to nitrocellulose membranes. Blots were blocked for 30 min at room temperature in the Odyssey blocking buffer. Primary antibodies (mouse anti-FLAG and rabbit anti-Beta-actin) were added to Odyssey blocking buffer at a ratio of 1:1000. Blots were incubated in the indicated primary antibodies overnight while shaking at 4 C. The following day, blots were washed three times with PBS-T, 10 min each before staining with the secondary antibody for 1 h at room temperature. The secondary antibodies used were goat anti-rabbit and goat anti-mouse antibodies diluted in blocking buffer to a ratio of 1:10,000. Following secondary antibody staining, the blot was washed 3 times with PBS-T before being imaged with the Odyssey CLx Imaging System.
Statistics
All data were expressed as mean ± SEM unless otherwise specified. A student's t-test was used for pairwise comparisons. Unless otherwise specified, statistical significance was set as p < 0.05.
Data availability
Summary statistics from metabolomics studies in plasma samples of participants of the JHS that were analyzed in this study have previously been uploaded to the JHS database of Genotypes and Phenotypes (dbGaP) repository and are available upon request from the respective study cohorts, which can be facilitated by the corresponding author. Summary statistics from whole genome association studies of metabolite levels in plasma samples of participants of the JHS have previously been made publicly available on the NHGRI-EBI GWAS Catalogue (Accession GCP000239) (15). All other results, analytic methods, and details of study materials are available within the manuscript. Noncommercial study materials will be made available to other researchers for the purposes of reproducing the results or replication of the procedure, as respective IRB and Material Transfer Agreements permit.
Supporting information-This article contains supporting information. | 2023-04-30T15:07:32.183Z | 2023-04-01T00:00:00.000 | {
"year": 2023,
"sha1": "6cd8fcb5cdabadd96d4d3c701be53b0eff63e280",
"oa_license": "CCBY",
"oa_url": "https://doi.org/10.1016/j.jbc.2023.104764",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "2723097f4e9d9f51fd3dee01acd256fd803f3a59",
"s2fieldsofstudy": [
"Biology",
"Chemistry"
],
"extfieldsofstudy": [
"Medicine"
]
} |
53719898 | pes2o/s2orc | v3-fos-license | Glycerol Monolaurate (GML) and a Nonaqueous Five-Percent GML Gel Kill Bacillus and Clostridium Spores
Bacillus and Clostridium spores are known to be highly resistant to killing, persisting on environmental and human body surfaces for long periods of time. In favorable environments, these spores may germinate and cause human diseases. It is thus important to identify agents that can be used on both environmental and human skin and mucosal surfaces and that are effective in killing spores. We previously showed that the fatty acid monoester glycerol monolaurate (GML) kills stationary-phase cultures of Bacillus anthracis. Since such cultures are likely to contain spores, it is possible that GML and a human-use-approved GML nonaqueous gel would kill Bacillus and Clostridium spores. The significance of our studies is that we have identified GML, and, to a greater extent, GML solubilized in a nonaqueous gel, as effective in killing spores from both bacterial genera.
B acillus and Clostridium species produce spores as nutrient levels become limiting; these spores are highly resistant to changes in environment conditions and can withstand both heat and drying. For example, Bacillus subtilis is nonpathogenic, but its spores may contaminate environmental surfaces (1), including workbenches in laboratories and ventilation systems. In contrast, B. anthracis and B. cereus may be environmental contaminants, but these organisms are also important causes of human diseases, such as anthrax (2)(3)(4) and necrotizing fasciitis (5,6), respectively. Similarly, there are large numbers of clostridial species, the majority of which can cause human diseases if introduced to traumatized tissue, for example, Clostridium perfringens (7,8). Also, treatment of humans with antimicrobials that disrupt the normal microbiota can allow germination and growth of pathogens such as Clostridium (Clostridioides) difficile (9)(10)(11). Spores from C. difficile may in turn contaminate the environment and skin and clothing after elimination from the infected host.
Glycerol monolaurate (GML) is a broad-spectrum antimicrobial with large numbers of bacterial targets (12)(13)(14). For example, Staphylococcus aureus has 16 two-component systems, all of which appear to be targeted for inactivation by GML (13). GML likely inserts into the plasma membranes of bacteria, with the net effect of preventing structural changes in membrane proteins required for their activity (13). The final effect may be to reduce the potential difference across the plasma membrane, comparably to another broadly antimicrobial molecule, reutericyclin (15). As with reutericyclin, bacteria with an gene conferring immunity to reutericyclin, such as some lactobacilli, are resistant to GML, and, indeed, GML serves as a growth stimulant for such microbes (14).
GML alone is active against most Gram-positive bacteria, such as streptococci and staphylococci, but the molecule is completely inactive against Enterobacteriaceae and Pseudomonas aeruginosa, due to the presence of the intact lipopolysaccharide (13,16). Gram-negative bacteria with lipo-oligosaccharide, such as Neisseria, are susceptible to killing by GML (13). However, all Gram-negative bacteria, as well as most Gram-positive organisms, are highly susceptible to GML when the compound is solubilized in a nonaqueous gel (12,14,17,18). The nonaqueous gel has been used extensively on human, animal, and environmental surfaces (12,14,17,18). A large number of in vitro and in vivo experiments have shown that 5% GML plus a nonaqueous gel (5% [50,000 g/ml] GML gel) is both effective at killing bacteria and safe for use on human and animal mucosal and skin surfaces (12,14,17,18).
To date, no studies have assessed the effectiveness of GML alone or of 5% (50,000 g/ml) GML gel against Bacillus or Clostridium spores. The goal of this study was to assess the effectiveness of GML in reducing and eliminating spores. Our studies showed that GML alone was effective in killing vegetative cells of B. subtilis, B. anthracis, C. perfringens, and C. difficile. Additionally, GML alone also killed spores by these same organisms but was not as effective as GML gel. GML gel was effective in killing both vegetative cells and spores. Because of its safety record, 5% (50,000 g/ml) GML gel may be useful in environmental and human surface contamination with bacterial spores.
RESULTS
Effect of GML alone on Bacillus and Clostridium vegetative cells and spores. GML is broadly antimicrobial and prevents exotoxin synthesis, with greatest effect on Gram-positive bacterial species and on Gram-negative species without an intact Enterobacteriaceae lipopolysaccharide, for example, Neisseria (13).
We first evaluated the effect of a range of GML concentrations on growth of B. subtilis and B. cereus, noting that stock cultures likely contained both vegetative cells and spores, as we initiated experimentation from 24-h subcultures. Over the 24-h test period, GML alone was bactericidal at the GML concentration of 100 g/ml for both B. subtilis and B. cereus (Fig. 1). This concentration of GML alone is at the solubility limit of GML in aqueous solutions. GML was bacteriostatic at the 50 g/ml concentration and did not inhibit growth at lower concentrations. These data indicate that vegetative cells were killed but also that indicate the spores which were likely present in the starting inoculum were either killed or prevented from germinating. The same assay had been previously published for B. anthracis Sterne (19) and so was not repeated.
A similar assay was performed on C. perfringens. For this organism, the minimum bactericidal concentration was 1 g/ml ( Fig. 2A), 2 logs below that for all three Bacillus species. At a 10-fold GML concentration below the growth inhibition level (0.1 g/ml), lecithinase/hemolysin production by C. perfringens was completely inhibited as tested at the 24-h culture time point (Fig. 2B). At concentrations of GML of Յ0.01 g/ml, inhibition of exotoxin production was partially lost, and at concentrations of GML of Ͻ0.01 g/ml, inhibition of exotoxin production was completely lost.
We did not perform a time course experiment to examine the effect of GML on C. difficile; however, we showed that the minimum bactericidal concentration of GML for this organism was 12.5 g/ml, well below that for all three Bacillus species but closer to that corresponding to the greater sensitivity of C. perfringens (Fig. 3).
We also tested C. difficile for effects of GML on production of Clostridium difficile toxin A (TcdA) with the use of a fluorescent reporter (red fluorescent protein [RFP]). In this experiment, GML did not interfere with overnight growth at 10 and 5 g/ml for strains R20291 and 630Δerm, respectively; higher GML concentrations were inhibitory to growth. Production of TcdA was assessed in the two strains at the indicated GML concentrations (Fig. 4). There was an approximately 2-fold reduction in production of TcdA, seen only at the last GML concentration that did not affect C. difficile growth at 10 g/ml for strain R20291 and 5 g/ml for strain 630Δerm.
In prior studies, accelerants of activity, for example, a nonaqueous gel as used in the present studies or acidic pH or EDTA, could be added to GML to increase its antimicrobial activity through synergism and/or increasing solubility (13). We have performed many in vitro and in vivo studies (nonhuman primate and human studies) with 5% (50,000 g/ml) GML gel (12,14,17,18,20).
With this as background, we evaluated the ability of the 5% (50,000 g/ml) GML gel to kill spores of two of the three Bacillus species and spores of C. difficile (Fig. 5). The GML gel was bactericidal for three representative spore formers as follows: B. subtilis by 1 h postinoculation, B. anthracis by 8 h postinoculation, and C. difficile by 4 h postinoculation. We know the three Bacillus strains tested in the studies described above contained spores by the late stationary phase as demonstrated by the results of heat resistance analysis and microscopy with simple spore staining and by the fact that we purchased anthrax spores from Colorado Serum Company. The data corresponding to the figures described above thus imply that GML alone may also be effective in killing spores. This possibility was tested (Fig. 6). GML alone (100 g/ml) was bactericidal for the B. subtilis spores by 4 h postinoculation. However, the B. anthracis spores were quite resistant to killing, with a log drop of approximately 1 over 24 h.
Our prior studies suggested that one major effect of GML is that of interfering with plasma membrane proteins, such as two-component systems, in vegetative cells, with the final killing effect resulting from dissipation of potential differences across plasma membranes (13). We have also shown that GML prevents biofilm formation and removes established biofilms (13). These are unlikely to be the mechanisms of action in killing spores, as we performed the experiments described above in both nongrowth medium (5% [50,000 g/ml] GML gel) and growth medium (GML alone added to spores in Todd-Hewitt medium). With this as background, we began studies in attempt to identify why spores are killed by GML. We first treated B. subtilis spores with GML alone (100 g/ml) for 2 h and then examined the spores microscopically after staining. We could not see visible signs of spore disintegration compared to non-GML-treated spores as assessed by simple spore staining.
In a second set of studies, we hypothesized that GML may have been coating the spores, with killing occurring upon subsequent germination. Thus, we treated B. subtilis spores with 5% (50,000 g/ml) GML gel for 2 h and then washed the spores with absolute ethanol or phosphate-buffered saline (PBS) to solubilize and wash away any surface-attached GML and nonaqueous gel. The starting inoculum of spores was 6.0 ϫ 10 5 /ml. After incubation with GML gel, there was no difference in spore viability (as measured by germination) whether the spores were not rinsed or were rinsed with absolute ethanol or were rinsed with PBS (Fig. 7). Thus, spore killing did not appear to be due to GML gel coating of the spores and killing upon germination.
DISCUSSION
As noted above, GML alone was bactericidal for vegetative cells of all strains tested. Some of those strains (for example, Bacillus subtilis) contained spores in high concentrations in the stationary phase, suggesting that GML alone and solubilized in a nonaqueous, glycol-based gel might kill resistant spores. Our studies showed that GML alone was bactericidal for vegetative cells of all five organisms, including 3 aerobes and 2 anaerobes. However, 5% (50,000 g/ml) GML solubilized in the nonaqueous gel (5% [50,000 g/ml] GML gel) was more effective in killing spores of these same organisms than GML alone. These studies are important since both GML and the nonaqueous gel have been shown to be safe for chronic use in nonhuman primates (12,17,18) and human mucosal surfaces (14,20). This suggests in particular that the 5% (50,000 g/ml) GML gel could be used to decontaminate environmental surfaces and human skin and mucosal surfaces of C. difficile to help reduce spread of the pathogen in hospitals.
B. anthracis has been used as an agent of bioterrorism, notably as spread through the mail. Our studies suggest that 5% (50,000 g/ml) GML gel could be used to decontaminate environmental surfaces contaminated with anthrax spores. Our studies showed that these spores were killed only slowly by GML alone at concentrations at the solubility limit in aqueous solutions, unlike spores from B. subtilis. However, the same organisms were completely killed by use of 5% (50,000 g/ml) GML gel. It is clear from our prior studies that the nonaqueous, glycol-based gel alone, used to solubilize 5% (50,000 g/ml) GML, has antimicrobial properties (13). Additionally, our prior studies with use of GML alone at concentrations that exceed the solubility limit at 37°C (100 g/ml) showed that GML has antimicrobial activity for bacteria not killed at the 100 g/ml solubility limit (16). The latter study suggested that soluble GML may become embedded in the bacterial membrane, thereby removing GML from solution and causing the previously insoluble GML to become soluble, with added membrane insertion, until antimicrobial activity is achieved. In our studies, it is therefore likely that it was the combination of the nonaqueous gel and the greatly increased amount of soluble GML in the gel that allowed greater killing by 5% (50,000 g/ml) GML gel than by GML alone. Also, other researchers have shown that GML can function in concert with other antimicrobials to increase efficacy (21). It is possible that addition of other agents to GML alone or increased GML amounts in the GML gel would further increase the rate of killing of B. anthracis spores. At the same time, it is worth considering that environmental surfaces are unlikely to have such high concentrations of spores as were present per milliliter in our studies, increasing the likelihood that GML gel alone could be used.
It is unclear why the spores of all organisms were killed by 5% (50,000 g/ml) GML gel, with Bacillus subtilis alone being highly susceptible to killing by GML without added nonaqueous gel. We performed preliminary studies to test for visible alterations in B. subtilis spores to determine if GML was killing B. subtilis upon germination. The results of both of these studies were negative. Although GML is often thought of as a surfactant, prior studies showed that GML stabilizes rather than destabilizes mammalian cells and bacteria (22,23). Those prior studies also showed resistance of red blood cells to lysis in the presence of hypotonic solutions containing GML (22) and interference with lipid raft mobility in immune cells (24,25). It is also possible that GML in some way further stabilizes the spore coats to prevent regeneration.
Previously, it was shown that GML prevents exotoxin production by Gram-positive bacteria at concentrations that do not kill toxin-producing bacteria (16,19,23). This was also observed in our studies, in which GML alone at concentrations below those that inhibit growth prevented hemolysin/lecithinase production by C. perfringens. Previously, it was shown that this effect of prevention of exotoxin production resulted in part from GML plasma membrane effects on two-component systems in interference with transcription (23). We do not know the precise effect of GML with respect to prevention of exotoxin production by C. perfringens, but the explanation is most likely to involve effects that increase the rigidity of the plasma membranes and interfere with signal transduction as shown previously. Unlike the results seen with other Gram-positive microbes, we saw only minimal (approximately 2-fold) effects of GML on exotoxin production by C. difficile that were independent of effects on microbial growth.
In sum, our studies have shown that the use of GML alone was effective in killing Gram-positive spore-forming bacteria, with subinhibitory concentrations preventing exotoxin production. A gel composed of 5% (50,000 g/ml) GML solubilized in glycols and approved for use in humans was effective in killing spores produced by the same microbes.
MATERIALS AND METHODS
Bacteria. B. subtilis, B. cereus, and C. perfringens, kindly provided originally by Dennis W. Watson, University of Minnesota, Minneapolis, MN (now deceased), were cultured from lyophilized Schlievert laboratory stocks. B. anthracis Sterne was obtained from the Colorado Serum Company, Denver, CO. C. difficile R20291 is of ribotype 027 and is maintained in the Ellermeier laboratory (26). The Bacillus and C. perfringens strains were cultured in Todd-Hewitt broth or on Todd-Hewitt agar plates. C. difficile was cultured in or on TY broth/agar and plated on CCFA (Anaerobe Systems, CA) plates and incubated at 37°C anaerobically for 24 h before enumeration. Bacillus strains were cultured, and plates were incubated aerobically in a standard incubator at 37°C. Clostridium difficile was cultured, and plates were incubated to the stationary phase in a Coy anaerobic chamber at 37°C. C. perfringens was cultured to the stationary phase in GasPak jars (Carolina Biological Supply Company, Burlington, NC) at 37°C, and plates were incubated in the same jars.
For the purposes of this study and as often defined in diagnostic microbiology, we define bactericidal results as represented by a Ն3 log 10 reduction in CFU compared to non-GML control results and bacteriostatic results as represented by CFU counts within 1 log 10 in similar comparisons. Our experience with GML and 5% (50,000 g/ml) GML gel is that the minimum bactericidal concentration and MIC are typically close to each other (13).
Production of spores. B. anthracis spores were used as purchased from the Colorado Serum Company. Spores from B. subtilis, B. cereus, and C. perfringens were prepared as follows. The organisms were cultured overnight in 25 ml Todd-Hewitt broth at 37°C. Cells were then collected by centrifugation (4,000 ϫ g, 15 min), suspended in 5 ml phosphate-buffered saline (PBS; 0.001 M sodium phosphate [pH 7.2], 0.15 M NaCl), and heated to 80°C for 10 min. The following method (27) was used to produce C. difficile spores. Strain R20291 was grown overnight in tryptone-yeast extract broth at 37°C in a Coy anaerobic chamber with 10% hydrogen, 5% carbon dioxide, and 85% nitrogen. Approximately 0.2 ml of overnight culture was spread on each of 2 SMC plates [9% Bacto peptone, 0.5% proteose peptone, 0.1% (NH 4 ) 2 SO 4 , 0.15% Tris base, 1.5% agar] using sterile beads and incubated at 37°C anaerobically for 4 days. Plates were then removed from the anaerobic chamber, and each plate was flooded with 5 ml of sterile PBS. Using a disposable inoculating loop, the bacterial lawns were resuspended. The resuspensions from each plate were pooled and washed in sterile PBS. After centrifugation (3,000 ϫ g, 15 min), the pellet was resuspended in 10 ml of 95% ethanol and incubated for 1 h at room temperature. After incubation, spore preparations were washed 3 times in PBS. After the final wash, spore preparations were resuspended in 1 ml of PBS; these preparations were incubated at 70°C for 20 min to kill vegetative cells. Spores were then diluted in PBS and transferred to the anaerobic chamber. Dilutions were plated on CCFA (Anaerobe Systems, CA) plates and incubated at 37°C anaerobically for 24 h before enumeration.
Experimentation with GML alone. Bacillus species were cultured in 25-ml flasks for designated time periods in the presence of various concentrations of GML from a 100,000 g/ml stock solution. Culture conditions were aerobic, with 200 rpm shaking at 37°C. Samples were removed at indicated times for serial 10-fold dilution plate counts and plated onto Todd-Hewitt agar plates (Difco, Detroit, MI). C. perfringens was cultured to the stationary phase in GasPak jars in 25 ml of Todd-Hewitt broth at 37°C for indicated times. Samples were removed for serial 10-fold dilution plate counts, and broths and plates were immediately returned to the GasPak jars. Samples of Todd-Hewitt broth cultures were also centrifuged (1,000 ϫ g, 15 min) and sterilized by filtration (Millex-GS; Merck Millipore Ltd., Tullagreen, Carrigtwohill, County Cork, Ireland) (0.22 m pore size). Filtrates were used for determination of rabbit red blood cell lysis on 0.85% agarose slides (28). C. difficile was cultured in the presence of indicated concentrations of GML for 24 h in an anaerobic chamber. At that time, samples were removed for plate counts in the same anaerobic chamber. For measurement of the effect of GML on exotoxin production by C. difficile, the following procedures were used. Overnight cultures were subcultured 1:100 in TY medium supplemented with thiamphenicol at 10 g/ml and with various concentrations of GML. Cultures were grown overnight and fixed as previously described (29). Fluorescence was measured at the Flow Cytometry Facility at the University of Iowa on an LSR II instrument (Becton, Dickinson). Strains and plasmid used for these studies, as described by Ransom et al. (30), included the following: RAN925, which is constructed as R20291/pRAN737(P tcdA ::rfpcat), and GMK134, which is constructed as CD630Δerm/ pRAN737(P tcdA ::rfp cat). Plasmid pRAN737 was a pDSW1728 derivative with P tcdA ::rfp.
Experimentation with GML gel. Spores (1/10 final volume of spores) were added to 5% (50,000 g/ ml) GML gel at 37°C for indicated periods of time. Samples were removed, and serial dilution plate counts (colony formation) were used for determination of spore germination into vegetative cells. Spore levels in initial cultures were determined by plate counts prior to addition to the GML gel. Experiments with Bacillus species were performed aerobically, whereas those with C. difficile were performed in an anaerobic chamber. The GML gel was prereduced 24 h prior to experimentation.
We considered that GML might adhere to spores and in that way kill spores upon germination. To test this, B. subtilis spores, which were the most easily killed by GML gel, were treated with GML gel for 2 h at 37°C. The suspension was then centrifuged (4,000 ϫ g, 15 min), and the pellets were resuspended in 1 ml of absolute ethanol to solubilize any GML adherent to spores or in 1 ml of PBS as a negative control. The preparations were again centrifuged, and the pellets were dried and then suspended in Todd-Hewitt broth for plate count determination.
Statistics. For some experimental data, standard deviations of the means are presented. Where standard deviations are not presented, experiments were repeated with similar results. | 2018-12-02T16:19:45.833Z | 2018-11-21T00:00:00.000 | {
"year": 2018,
"sha1": "823f7b2a36266a663cc079ecdcf779835ea16a1b",
"oa_license": "CCBY",
"oa_url": "https://msphere.asm.org/content/msph/3/6/e00597-18.full.pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "1535e8ee9bf5a814cee076efeba4a8d673a0614a",
"s2fieldsofstudy": [
"Biology"
],
"extfieldsofstudy": [
"Chemistry",
"Medicine"
]
} |
425 | pes2o/s2orc | v3-fos-license | Amazing geometry of genetic space or are genetic algorithms convergent?
There is no proof yet of convergence of Genetic Algorithms. We do not supply it too. Instead, we present some thoughts and arguments to convince the Reader, that Genetic Algorithms are essentially bound for success. For this purpose, we consider only the crossover operators, single- or multiple-point, together with selection procedure. We also give a proof that the soft selection is superior to other selection schemes.
I. Introduction
We are using Genetic Algorithms (GA's) for solving hard global optimization problems for at least three reasons: • they are easy to implement in many computer languages, • they are applicable to problems, which cannot be easily, if at all, specified analytically as a set of closed-form formulas, • we believe, that their inherent intelligence will automatically, i.e. with almost no programmer's effort, find the way to solve, "sufficiently well", our difficult problems. The very idea of GA's, to simply mimic the Nature [1], belongs mostly to the sphere of intuition, and is almost lacking a solid mathematical background. Indeed, numerical values of many important "tuning parameters" (mutation rate, probability of selection for reproduction, etc.) are largely selected on the base of experience of other people solving problems similar to ours. The hypothesis of "building blocks" appeared false. Other investigations of GA's and their inner working are rare. We simply believe, that following the Nature's paths cannot be wrong. But are we right? And, if so, why?
II. Distance between parents and offsprings
It is easy to see, that after the crossover operation, the distance between resulting offsprings is identical to the distance between their parents [2]. Consider a pair of parents, p a = p 1 a , p 2 a , . . . p N a and p b = p 1 b , p 2 b , . . . p N b consisting of N genes each. The distance between them may be calculated in many ways, depending on metrics in use. In the simplest case, when each gene is just a binary digit, the Hamming distance (d H (·, ·)) is perhaps the most natural choice. This simply counts the number of bits differing on the corresponding positions in the two given bit-strings. It is obvious, that in this case When individual genes are more complex, i.e. when they consist of more bits (or, more often, are the symbols drawn from finite size alphabet(s)), or even when they are just the real numbers, the same remains true in any metrics induced by L p norms. Indeed, the expression: has to be equal to d p (o a , o b ), as the numerical components of the sum, shown above, are identical in both cases; even their order is preserved. The property (1) holds also for less frequently used norm L ∞ : ||x|| ∞ = max k |x k |.
Consider now a triangle in genetic space defined by vortices: two parent chromosomes, p a and p b and any other fixed, but otherwise arbitrary, reference point r. We will apply the crossover operator to the pair (p a , p b ), obtaining another pair of chromosomes (o a , o b ), as shown in Fig. 1. It is easy to verify (think of Hamming distance between chromosomes consisting of 1-bit genes), that and, after adding together two first rows of the above and comparing the result with the sum of the two last rows of Eq. 3, that In remaining cases, with discrete or continuous genes, other measures of distance between them may be used. Looking again at Fig. 1, we can conclude, that in general the following equality takes place: where p is positive and finite 1 integer, and d p is a distance induced by L p norm. The relation (5) may be extended even further, just for elegance, by adding to the left-hand side the p-th power of the distance between parents and -to the r.h.sp-th power of the distance between offsprings, since, by virtue of (1), they are equal to each other. Calling the sum of lengths of the triangle's edges, first raised to the fixed integer power p, the generalized circumference, we may express our result shortly as: The generalized circumference of the triangle made of three chromosomes, remains unchanged when any two of them are replaced by their offsprings.
III. "Geometric" conclusion and discussion
Recall that the chromosome r was chosen arbitrarily. One may wonder what happens, if r = x ⋆ , i.e. when r is the searched, still unknown, optimal chromosome -possibly one of many -in a genetic space. If this is the case, then our finding may be expressed as follows: Since the sum of p-th powers of distances between parent chromosomes and the desired solution is conserved by the crossover operator, then the offspring chromosomes cannot relocate too far from the optimal solution.
Proof:
Let the parents' distances from the solution be equal to d 1 and d 2 , both strictly positive, and the offsprings' distancesd a and d b , respectively. We can write: Depending on relation between d a and d 1 the value of the expression appearing in the square bracket may be lower than 1, higher than 1, or exactly equal to one. Respectively, we have In words: if one of the offsprings moves further apart from the solution than one of its parents, then the second one gets closer to the solution then their other parent.
Comment: In the degenerate case, when exactly one of the parents is already an optimal solution, it may happen that both offspring chromosomes will be closer to the solution than the second parent, with none of them being the optimal point, which quite unexpectedly appears to be a repeller, rather than an attractor! In conclusion: the outcome of the crossover operation may vary. One thing, however, is sure. It may never happen than both offspring chromosomes are located more far away from the solution than their more distant parent. On the other hand, there is no guarantee, that at least one of them gets closer to the desired solution, than its "better" parent. Nevertheless, at least one of newly created individuals is equally or less distant from the solution than its "worse" parent. We use quotation marks, since in reality the chromosomes located closer to the optimal one (the "better") need not to be better fitted. This is most easily seen in cases, when genotypes arbitrarily close to the solution are unacceptable at all, for example due to violation of constraints. So the really important question is: Which one of the two offsprings is closer to the solution? Judgment based on their fitness alone cannot be regarded as reliable or conclusive. We shortly summarize all interesting outcomes of the crossover operation in Tab. I. Analyzing its contents we see, that the symbol "o" can be found at advantageous position exactly 6 times, while only twice on disadvantageous one. Does it mean, that the odds for selecting "proper" offsprings, i.e. to improve at least one trial solution, are 6 : 2?
The answer would be positive, if the events 2-5 occurred with equal probability, what is unlikely. On the other hand, if only the case 5 (the worst) occurs again and again, then the random, unbiased selection of one of the offsprings, would give us exactly (!) 50% chance to move closer to the reference chromosome r. This means, that in practice, the chances for improvement can be even higher than 1 2 ; we will prove that, rigorously, in the following section.
Important note: We have to carefully distinguish between continuous and discrete case. In discrete genetic space the only convergent sequences are constant sequences. This is because there are no elements of discrete genetic space, which would be located arbitrarily close to any existing chromosome, the optimal one in particular. Therefore the notion of convergence is sensible and usable only in continuous cases.
On the other hand, since r is arbitrary, then it may have nothing to do with the optimal solution. That is why the entire evolutionary process may not converge at all without additional driving forces, other than the actions of crossover operators.
As we will show now, the key to the question of convergence is the selection process -the practical realization of the Darwinian rule of evolution, survival of the fittests, understood in a probabilistic sense rather than an absolute rule.
IV. Chances for success
The following text is based on the problem stated and solved by Lata la in Delta [3] -a popular Polish monthly on mathematics, physics and astronomy, targeted mainly at high-school students. The problem and its solution are freely rephrased by the current author.
Problem:
Find the winning strategy in the following game: Looking at an integer number, randomly chosen from two such numbers written down by our opponent, guess whether the other (unknown) number is higher or lower. The two numbers in question are distinct. We win, when our guess is correct, otherwise we loose.
Solution:
Use arbitrary, strictly increasing, sequence of numbers {c k } ∞ k=−∞ , each belonging to the (open) interval (0, 1), for example c k = 1 2 + arc tan k π . When the selected (known) number is equal to k, then with probability c k "guess", that the other (unknown) number is lower, or, with probability 1 − c k , that it is higher.
It is obvious, that this strategy should work equally well not only for unknown integer numbers, but also when the numbers are drawn from any countable subset of reals. But why does it work at all?
Let p m,n denotes the probability, that the numbers chosen by our opponent are m and n, and that m > n. The probability, that our guess is correct, may be written as m,n∈Z, m>n where Z is the set of integers. First " 1 2 " in the r.h.s. of (6) comes from the fact, that p m,n = 1. The second component is strictly positive, since c m > c n for arbitrary m > n (as the sequence {c k } is strictly increasing) and at least one of p m,n is greater than zero. In conclusion: our chances to win always exceed 50%. This wouldn't be so, if the sequence {c k } was not strictly increasing -in such circumstances our chances to win could be estimated only as not less than 1 2 . Let us note, that nothing certain can be said about how much our chances to win exceed 50%. They will peak, if all the differences c m − c n are maximized, at least those of them, which "meet" non-zero p m,n 's in formula (6). Unfortunately, we know nothing ahead about probabilities p m,n 's.
How is the above problem related to Genetic Algorithms? Quite simply: the sequence {c k } should be regarded as a tool to convert the value of fitness to probability of selection. The superiority of the soft selection, realized with such a sequence {c k }, over the hard selection schemes, is evident. In the case of soft selection, our chances to win (i.e. to improve the objective by selecting a better offspring for further processing) are always higher than chances for failure. On the contrary, the hard selection 2 implies that, in the unlucky event, both chances can be equal to each other.
The hard selection scheme can be considerably improved to become comparable with the soft selection. It is enough to select the number n 0 (see footnote) as any average of fitnesses of all individuals present in the previous generation. This trick should work best in cases, when our opponent -the objective function -produces only a few discrete values.
The difference seems rather subtle: sharp versus not sharp inequality. But let us recall the brutal practice of citizens of an ancient Greek city of Sparta. In strive to have only excellent warriors as their descendants, they used to physically eliminate all "defective" newborns. Did they succeed?
V. More on selection
Consider the objective function with many local extrema of very similar fitness value, yet having exactly one global optimum. The evolving population will sooner or later split into many loosely connected clusters, concentrated around those extrema. To discover the true, global optimum, we need the ability to correctly rank the individuals with very close values of their fitness. Only then the "useless" individuals, located around local extrema, would be extinct. Therefore, in particular computer implementation, not every kind of average used as threshold for hard, stepwise selection, is equally good. To increase our chances for success, and accelerate the convergence as well, we should apply the sequence {c k }, or its continuous counterpart -which may be selected individually for each new generation -in such a way, that it changes most significantly around majority of fitness values across the population. When searching for maximum, the following simple and numerically plausible transformations from f itness to probability of selection, often called scaling of the fitness function, satisfy this requirement: with the first choice being definitely softer. The subscripted constants f α denote respective quantiles (more precisely: quartiles) of the fitness distribution across the current population, with f 1/2 being the median. Put unity 3 into the denominator when f 3/4 − f 1/4 = 0.
Replace summation in (7) with subtraction, when searching for minimum.
It is clear, that GA's can be most effective for objectives, for which only a very limited information is available, namely nothing but fitness values computed for every member of the population, usually only the last one. Their ability to quickly detect and then to concentrate in the interesting parts of the search space makes them clearly superior to generic Monte Carlo approach, which waste time for uniform and fruitless exploration of other regions. The above is certainly true for objectives, which are at least piecewise continuous and have no singularities. For such a broad class of problems, with chromosomes coded in a natural way 4 , the stopping rules should be based on compactness of the evolving population, paying only little attention to the behavior of fitness. The evolutionary process should be continued as long as the volume of the search space occupied by "better half" of the population still decreases. One should be aware, however, that this strategy will fail for objectives with more than one global optimum or when the unique extremum is degenerate (flat, improper), i.e. consists of more than a single point, either in reality or due to roundings. If this is the case, then careful analysis of the last generation may be helpful.
For discrete problems (with integer and maybe boolean variables present) the notion of continuity does not apply, so the task is to efficiently find the acceptable solution without performing exhaustive search. It can be shown [4], that for purely discrete problems we need O N 3 2 ln N evaluations of the objective instead of 2 N , as necessary and required by the exhaustive search. N is the number of bits, not unknowns, in a single chromosome. To be precise: after N 3 2 ln N evaluations of the objective, the chance that the best so far chromosome is separated no more than 1 unit of distance (in Hamming sense) from the optimal one, are higher than 50%. No well justified stopping rules can be given for discrete case.
Mixed problems, involving real and integer unknowns, are even harder to analyze. From the formal point of view, such problems may be regarded as large, but finite, sets of continuous problems.
VI. Summary
We have shown, that Genetic Algorithms are bound for success. The chances for improvement are always higher than for lack of it, if the selection of parents is performed either in a soft, or hard but adaptive, manner. This is a very general result, completely independent on the optimization problem under study. It applies equally well to discrete, continuous and mixed optimization problems.
As we can see, the quite high chances of Genetic Algorithms for success are strictly related to their property of not rejecting nor ignoring the "bad" trial points in the search space. Contrary, the rigorous, deterministic search methods are simply unable to "jump over" the barrier surrounding even the single global minimum, if started too far from the solution.
Our result is of stochastic nature rather than deterministic. This may mean in practice, that we may be unable at all to improve the already known, approximate solution of our particular problem. Nothing can be said how quickly we will arrive at any improvement. This may be significantly influenced by other components of GA's: mutation operators, population size and numerical values of various tuning parameters, not on the selection scheme or crossover mechanism alone. | 2005-12-05T03:12:42.000Z | 2005-12-05T00:00:00.000 | {
"year": 2005,
"sha1": "4b3548fe9604bf234908e04a031e37eabd7e99eb",
"oa_license": null,
"oa_url": null,
"oa_status": null,
"pdf_src": "Arxiv",
"pdf_hash": "4b3548fe9604bf234908e04a031e37eabd7e99eb",
"s2fieldsofstudy": [
"Computer Science",
"Mathematics"
],
"extfieldsofstudy": [
"Computer Science"
]
} |
27952622 | pes2o/s2orc | v3-fos-license | Combination of Carbon Dioxide Angiography and Outback® Elite for Revascularization of a Patient with Renal Insufficiency with Bilateral Femoropopliteal Chronic Total Occlusions
A new reentry device (Outback Elite) system has been available in Japan since June 2016. This new device enables easier treatment of chronic total occlusion (CTO) in the lower extremities. We report a case of a woman in her 70s who underwent revascularization using this new device twice to treat both of her femoropopliteal CTO lesions. She was referred to our hospital complaining of intermittent claudication in both legs. She had a long history of diabetes mellitus complicated with severe chronic kidney disease. Her estimated glomerular filtration rate was <20. She refused surgical revascularization; therefore, we performed our treatment without iodine contrast medium. First, magnetic resonance imaging was performed to confirm that the CTO lesions had caused severe claudication before intervention. Subsequently, the Outback Elite device and carbon dioxide (CO2) angiography made it possible to revascularize both of her legs without iodine contrast medium. At 6 months after the procedures, we did not observe exacerbation of claudication in her legs.
Introduction
Iodine contrast medium is considered as the gold standard for endovascular treatment (EVT) because it provides much information needed to select a treatment strategy. However, the problem of contrast-induced nephropathy (CIN) remains associated with EVT. In high-risk patients, such as those with chronic renal impairment, diabetes mellitus, and congestive heart failure and the elderly, the incidence of CIN is >20%-30% [1]. Carbon dioxide (CO 2 ) angiography is an alternative method that has been used since the 1970s for patients with renal impairment or hypersensitivity to iodine [2]. Subintimal angioplasty (SIA) with recanalization of SFA-CTO is an acceptable first-line therapy that was first described by Bolia et al. [3]. However, the reentry procedure often becomes difficult when the wire crosses the distal true lumen after excessive reach beyond the subintimal space. The Outback Elite is a new device that can lead to successful outcomes of revascularization in SFA-CTO. Here we report a patient with chronic renal impairment who underwent revascularization using the reentry device with CO 2 angiography guidance in bilateral SFA-CTO.
Case Report
A 76-year-old woman with a long history of diabetes mellitus was referred to our hospital with complaints of severe claudication in bilateral lower limbs. She had noticed claudication symptoms approximately 10 years earlier. Recently, she could not walk 100 m without needing to rest. Her long history of diabetes mellitus required that she take multiple oral medications but not insulin. Unfortunately, her condition was complicated by diabetic nephropathy. Laboratory studies revealed increased HbA1c level [6.7% (NGSP); reference value (RV) 4.6%-6.2%], blood urea nitrogen (45.3 mg/dL; RV 7-22 mg/dL), creatinine (2.07 mg/dL; RV 0.4-0.8 mg/dL), and a decreased estimated glomerular filtration rate of 19.16. The ankle-brachial pressure index (ABI) was 0.72 on the right side and 0.74 on the left side. We gave her the option of one of two revascularization methods: femoropopliteal bypass surgery or EVT. Both treatments have benefits and disadvantages for patients. After careful consideration, she decided to undergo revascularization by EVT. Before EVT, we needed more information about the status of her arteries in both legs; therefore, we performed magnetic resonance imaging (MRI). The MRI results confirmed that both of her SFAs were occluded from bifurcation of the deep femoral artery ( Figure 1). Figure 2). Initially, we performed EVT for the left SFA-CTO. A 6 Fr Destination5 catheter (Terumo Corporation, Tokyo, Japan) was inserted from the right femoral artery into the left common femoral artery as a contralateral antegrade wiring approach. A 0.018-inch Treasure5 guidewire (Asahi Intecc Co., Ltd., Nagoya, Japan) with a NaviCross5 microcatheter (Terumo Corporation, Tokyo, Japan) was able to penetrate the proximal fibrous cap of the CTO. After advancing the microcatheter into the CTO site a little, we exchanged the 0.018-inch wire for a hydrophilic 0.035-inch, 1.5-mm, J-type, shaped-tip wire. A knuckle-shaped hydrophilic 0.035-inch wire was used to easily cross the middle of the CTO site and to eventually capture the distal site of the CTO. The 0.035-inch wire was not in the true lumen, however, so we subsequently used a reentry device called an Outback Elite catheter (Cordis, Florida, USA) to attempt to gain access to the true lumen. The device was delivered to the distal subintimal space adjacent to the reconstructed area of the distal true lumen, and some effort was needed to successfully penetrate using a hollow needle. Two orthogonal angiographic views were taken to confirm the correct direction: an L-shaped fluoroscopic marker enabled orientation of the tip toward the reentry site and, to attain the appropriate position, we needed to fine-tune the positioning at the site and used a T-shaped fluoroscopic marker to confirm the desired alignment. Next, a 22 G nitinol reentry cannula was plunged into the distal true lumen in the distal SFA. A 0.014-inch Chevalier Floppy5 extrasupport wire (Cordis, Florida, USA) was advanced into the true lumen. Then, the reentry device was removed, and we used a SABER X5 5 × 60 mm (Cordis, Florida, USA) to perform balloon dilatation in the CTO lesion. Next, two 6.0 × 100 mm SMART5 stents (Cordis, Florida, USA) were implanted from the proximal to middle portion of the left SFA. We performed postballoon dilatation using the same balloon. All procedures were performed without iodine contrast medium. The patient's left ABI improved from 0.74 to 0.98. (Figure 3). One month after the procedure for the left SFA-CTO, we performed EVT for the right SFA-CTO. A 6 Fr Parent Plus606 sheath was inserted from the right femoral artery as an ipsilateral antegrade wiring approach. A 0.018-inch Treasure guidewire with a 4.0 Fr CXI6 support catheter (Cook Medical, Bloomington, IN, USA) was used to penetrate the fibrous cap of the CTO. Advancing the microcatheter into the CTO site, we exchanged the 0.018-inch wire for a hydrophilic 0.035inch, 1.5-mm, J-type, shaped-tip wire. A knuckle-shaped hydrophilic 0.035-inch wire was used to reach the distal site of the CTO. The 0.035-inch wire was intended to be placed in the false lumen, but subsequent attempts to gain access to the true lumen were performed using an Outback Elite catheter. We performed CO 2 angiography to guide exact positioning of the needle and stents. A 0.014-inch extrasupport wire was advanced into the true lumen. Then, the reentry device was removed, and we performed balloon dilatation using a 5 × 60 mm balloon in the CTO lesion. Subsequently, two SMART stents (7.0 × 100 mm and 6.0 × 150 mm) were implanted from the proximal to middle portion of the right SFA. We performed postballoon dilatation using the same balloon. All procedures were performed without iodine contrast medium. Subsequent examinations showed that the patient's left ABI improved from 0.72 to 0.91.
Discussion
Generally speaking, expansion of a subintimal stent is less than that of an intraplaque stent and contributes to longterm outcomes, particularly in coronary arteries. For treating CTO, intraplaque angioplasty (IPA) is preferable over SIA to maintain chronic patency for CTO lesions. However, some studies have found no statistically significant differences between IPA and SIA for treating SFA-CTO [4][5][6]. Therefore, we have frequently adopted SIA using a 0.035-inch knuckledwire technique. This technique is feasible and easy to perform except for reentering the distal true lumen at the CTO site [4,5]. However, the guidewire may not pass the subintimal lumen in all cases. Otherwise, the intraluminal procedure using a 0.014-or 0.018-inch guide wire does not always succeed in capturing the distal true lumen. If the distal true lumen cannot be captured, retrograde access may be a useful and feasible option [7]. However, the retrograde technique requires a certain amount of experience with EVT. Thus, a reentry device was developed to overcome procedural failures and improve success rates after subintimal passage through the CTO. Recent studies have reported Figure 2: Initially, we performed EVT to treat the occlusion of her left SFA. We used a new reentry device called an Outback Elite and the contralateral approach. Two orthogonal angiographic views are essential to confirm the right direction: an L-shaped fluoroscopic marker provides the orientation of the tip toward the reentry site, and fine tuning of the positioning is needed when using a T-shaped fluoroscopic marker to confirm the desired alignment. Next, a 22 G nitinol reentry cannula was plunged into the distal true lumen in the distal SFA. A 0.014-inch extrasupport wire was advanced into the true lumen. EVT: endovascular therapy; SFA: superficial femoral artery.
>95% procedural success rates in achieving revascularization using the Outback catheter [8]. In contrast, Shin et al. found that significant calcification at the reentry site was a strong predictor of procedural failure after use of this reentry device [9]. Our patient underwent EVT for CTOs in both SFAs, and these revascularizations intentionally needed the reentry device in both of her legs. The first session used the contralateral approach, and the second session used the ipsilateral approach. On the basis of our experience, if using a reentry device, the ipsilateral approach would be preferred to the contralateral approach. The contralateral approach makes it more difficult for a reentry device to cross over the aortic bifurcation; therefore, in this approach, previous dilatation of the subintimal lumen is required to insert such a bulky device. In our 2 experiences, use of a reentry device required more time in the contralateral approach than in the ipsilateral approach. Therefore, we recommend using and/or being prepared to use the ipsilateral approach with a 6 Fr system as a standby if this new reentry device is used. Lastly, operators should be aware of characteristics of CO 2 angiography. CO 2 angiography is particularly advantageous for patients at risk of CIN and allergic reactions against iodine contrast medium (ICM) [2]. Moreover, we consider CO 2 as effective as ICM within the femoropopliteal lesions [10]. Complications following the CO 2 angiography have been reported, including vapor-lock, which occurs upon entrapment of CO 2 at the origin of a vessel, thereby impairing flow and potentially resulting in bowel ischemia especially in EVT for aortoiliac or renal lesions [11].
Conclusions
To the best of our knowledge, this is the first report to use an Outback Elite catheter under CO 2 angiography. Six months after the procedure, we did not observe exacerbation of claudication in our patient's legs. We found that CO 2 angiography for treating SFA-CTO lesions was useful and effective. EVT for SFA-CTO using SIA or IPA would be expected to give the same results. This combination therapeutic technique could potentially lead to a new intervention era in which contrast dye is not needed. Figure 3: We subsequently performed EVT to treat the occlusion of the right SFA. We used an Outback Elite and the ipsilateral approach. This new reentry device enabled successful performance of EVT for this tough CTO lesion. The ipsilateral approach is better than the contralateral approach to bring the reentry device into the subintimal space. EVT: endovascular therapy; SFA: superficial femoral artery; CTO: chronic total occlusion. | 2018-04-03T03:47:04.873Z | 2017-07-09T00:00:00.000 | {
"year": 2017,
"sha1": "6f194a33615c08c4d96d4ed8c8bec7ff1a74467c",
"oa_license": "CCBY",
"oa_url": "http://downloads.hindawi.com/journals/cric/2017/8632747.pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "06b7a7109d3aec394e1832267ca4a060ed92a518",
"s2fieldsofstudy": [
"Engineering",
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
265275083 | pes2o/s2orc | v3-fos-license | Songorine inhibits oxidative stress-related inflammation through PI3K/AKT/NRF2 signaling pathway to alleviate lipopolysaccharide-induced septic acute lung injury
Abstract Objective The present study aimed to investigate the protective action and mechanism of songorine on sepsis-induced acute lung injury (ALI). Methods The sepsis-induced ALI mouse and cell models were established by lipopolysaccharide (LPS) induction. Lung injury was assayed by hematoxylin and eosin staining, lung injury score, and lung wet-to-dry (W/D) weight ratio. Apoptosis in lung tissues was evaluated by TUNEL assay, and the expression of apoptosis-related markers (Bcl2, Bax, and caspase-3) was measured by western blotting. Levels of pro-inflammatory factors and oxidative stress markers in the bronchoalveolar lavage fluid (BALF) of mice were measured by ELISA and RT-qPCR. The expression of PI3K/AKT/NRF2 pathway-related proteins was analyzed by western blotting. Results Songorine treatment at 40 mg/kg mitigated sepsis-induced ALI, characterized by improved histopathology, lung injury score, and lung W/D weight ratio (p < 0.05). Moreover, songorine markedly attenuated sepsis-induced apoptosis in lung tissues; this was evidenced by an increase in Bcl2 levels and a decrease in Bax and caspase-3 levels (p < 0.01). Also, songorine reduced levels of proinflammatory cytokines (TNF-α, IL-6, IL-1β and MPO) and oxidative stress regulators (SOD and GSH) in the BALF of LPS-induced sepsis mice and RAW264.7 cells (p < 0.05). In addition, songorine upregulated the PI3K/AKT/NRF2 pathway-related proteins in LPS-induced sepsis mice and RAW264.7 cells (p < 0.05). Furthermore, LY294002 (a PI3K inhibitor) treatment reversed the protective effect of songorine on sepsis-induced ALI. Conclusion Songorine inhibits oxidative stress-related inflammation in sepsis-induced ALI via the activation of the PI3K/AKT/NRF2 signaling pathway.
Introduction
Sepsis is joint systemic inflammation in response to severe infection and trauma, with high mortality and long-term morbidity [1,2].The lung is highly vulnerable during sepsis, owing to its inherent susceptibility to infection and inflammation; therefore, acute lung injury (ALI) is one of the most common and severe complications [3].ALI is initially characterized by pulmonary inflammation and microvascular permeability and eventually progresses to acute respiratory distress syndrome [4].To our knowledge, there is no effective drug and therapeutic approach for sepsis-induced ALI.
Oxidative stress and the resulting inflammation are closely implicated in the pathogenesis of ALI.The imbalance between the production of reactive oxygen species (ROS) and the antioxidants results in oxidative stress and further promotes inflammation [5].Songorine is a C 20 -diterpene alkaloid from the root of Aconitum carmichaelii, exhibiting anti-inflammatory effects and low toxicity [6,7].Songorine can inhibit the generation of inflammatory cytokines, including IL-6, IL-1β, and TNF-α [8].Unlike nonsteroidal anti-inflammatory drugs, songorine exerted antioxidative activity without ulcerogenic effect [6].A previous study has shown that songorine prevents lipopolysaccharide (LPS)-evoked mitochondrial ROS production in septic heart injury [9]-Songorine is a critical bioactive ingredient of Shenfu injection for the treatment of sepsis [10].These pieces of evidence suggest a potential therapeutic avenue of songorine in sepsis-induced ALI that warrants further exploration.
Nuclear factor erythroid 2-related factor 2 (Nrf2) is a central controller of oxidative stress and inflammatory response [11].A previous study indicated that songorine fosters cardiac mitochondrial biogenesis through the induction of Nrf2 during sepsis [9].Moreover, Li et al. demonstrated that obacunone mitigates ferroptosis in the context of LPS-induced ALI through the enhancement of Nrf2-dependent antioxidant responses [12].PI3K/Akt signaling can activate Nrf2, and crosstalk between them protects cells against inflammatory and oxidative damage [13].A previous study has demonstrated that PI3K/Akt signaling via Nrf2 protects against inflammation in hyperoxia-induced ALI [14].Moreover, boosting PI3K/Akt/mTOR and Nrf2/HO-1 signaling alleviates oxidative stress to prevent cyclophosphamide-induced ALI [15].For these reasons, PI3K/ Akt/Nrf2 signaling pathway may be a potential therapeutic target for songorine against sepsis-induced ALI by suppressing oxidative stress and inflammation.
The current study was designed to explore the effect of songorine attenuating LPS-induced septic ALI.We aimed to confirm the therapeutic efficacy of songorine on LPS-induced septic ALI by inhibiting oxidative stress and inflammation.Additionally, we revealed the underlying mechanism through which songorine treats LPS-induced septic ALI, involving regulation of the PI3K/Akt/Nrf2 signaling pathway.This research elucidates the therapeutic potential of songorine in addressing sepsis-induced ALI through its modulation of the PI3K/ Akt/Nrf2 signaling pathway, highlighting its prospective application in lung injury therapy.
Animal model establishment and treatment
Animal experiments conformed to the Guide for the Care and Use of Laboratory Animals and were approved by the Ethics Committee of The First Affiliated Hospital of Ningbo University.Healthy male C57BL/6 mice (6-8 weeks old) were obtained from Hubei Province Experimental Animal Center (Wuhan, China).The choice of male C57BL/6 mice is informed by their well-documented immune responses, genetic consistency, and susceptibility to LPS-induced septic lung injury [16].Mice were housed under specific pathogen-free conditions.All mice were randomly divided into four groups (n = 6 for each group): sham, LPS, LPS + Son (20), and LPS + Son (40) groups.In the LPS group, mice were intraperitoneally instilled with 10 mg/kg LPS for establishing a septic ALI model.For the LPS + Son (20) and LPS + Son (40) groups, mice were pretreated with 20 mg/ kg and 40 mg/kg songorine by intraperitoneal injection before 2 h of stimulation with LPS.We referenced the study by Li et al. as a guideline for the dose selection of songorine, wherein similar doses of songorine were employed to achieve desired pharmacological outcomes [9].After 12 h, lung tissues were removed from mice for subsequent experiments.
Hematoxylin and eosin (HE) staining
HE staining for lung tissues of mice was performed as previously described [17].The fixed lung tissues with 10% neutral buffered formalin were embedded in paraffin and sectioned into 5-µm slices.The sections were stained with hematoxylin and eosin for observation under a microscope.
Lung injury score
The severity of lung injury was assessed by a semiquantitative scoring standard [18].Four indicators (alveolar septal thickening, inflammation, hemorrhage, and edema) reflecting the severity of lung injury were blindly scored.Alveolar septal thickening, hemorrhage, and edema were characterized as follows: absent (score = 0), mild (score = 1), moderate (score = 2), severe (score = 3), and extensive (score = 4).Inflammation was quantified based on the total number of inflammatory cells per ×100 field view.The overall pathological scores of lung injury were analyzed using the non-parametric Kruskal-Wallis test.
Lung wet-to-dry (W/D) weight ratio
The Lung W/D ratio was used to evaluate the severity of pulmonary edema.Lung tissues were weighed immediately after removal from mice (wet weight) and then placed into a 60 °C oven for drying.After 48 h, the dried lung was re-weighed as dry weight.The ratio of wet to dry weight was calculated.
TUNEL assay
The cell apoptosis in lung tissues was measured by TUNEL staining assay.Lung tissues were fixed in 10% phosphate-buffered formalin and embedded into paraffin for cutting into 5-μm sections.The sections were incubated with a cocktail containing terminal deoxynucleotidyl transferase enzyme, and DAPI was used for nuclear counterstaining.The fluorescence was detected using a Leica DMI-3000B phase-contrast fluorescence microscope.
Enzyme-linked immunosorbent assay (ELISA)
Bronchoalveolar lavage fluid (BALF) was collected from the right lung of mice by lavage with 500 µL of phosphate-buffered saline (PBS) immediately after the end of ventilation.Levels of inflammatory cytokines (TNF-α, IL-6, IL-1β, and MPO) in bronchoalveolar lavage fluid (BALF) were measured using commercial ELISA kits.Levels of glutathione (GSH), superoxide dismutase (SOD), and malondialdehyde (MDA) in lung tissues were also measured by corresponding commercial ELISA kits (Jiancheng Bioengineering Institute, China).
Immunohistochemistry (IHC)
After dewaxing and rehydration, lung tissues were blocked by 3% H 2 O 2 for 30 min and subjected to citric acid retrieval at 100 °C for 20 min.Then, the tissues were blocked by bovine serum albumin and incubated with the primary antibodies anti-CD68 (1:100; ZM-0060, China) overnight and further incubated with the secondary antibody for 1 h.Next, tissues were stained by 3,3′-Diaminobenzidine and hematoxylin, followed by observation under a microscope.medium containing 10% fetal bovine serum, 100 units/mL penicillin, and 100 μg/mL streptomycin sulfate in a humidified atmosphere with 5% CO 2 .Cells were incubated with gradient concentrations of songorine (1-60 μM) for 4 h.Then, cells were treated with 1 μg/mL LPS for 24 h.Some cells were treated with songorine and/or LY294002 (a PI3K inhibitor), followed by 1 μg/mL LPS for 24 h.
Cell viability
RAW264.7 macrophages were seeded into a 96-well plate with 100 μL culture medium.Various concentrations of songorine (1-60 μM) were used to treat macrophages with 1 μg/mL LPS stimulation.Next, the CCK8 reagent (#CK04, Dojindo, Japan) was added into each well to incubate for 2 h.Cell viability was reflected by the optical density value using a microplate reader at the wavelength of 450 nm.The IC50 value for songorine treating RAW264.7 macrophages was calculated and used for subsequent cell experiments.
Western blotting
Total proteins were extracted from lung tissues and RAW264.7 macrophages using ice-cold RIPA buffer containing protease inhibitors.Protein concentration was measured by the bicinchoninic acid method, and 30 μg of protein was loaded into 10% SDS-PAGE gel for electrophoresis.Afterward, proteins were transferred to PVDF membranes and blocked with 5% nonfat milk for 1 h.Then, the PVDF membranes were incubated with primary antibodies (Table S1) overnight at 4 °C.After being washed with TBST, membranes were incubated with HRP-conjugated secondary antibodies for 1 h.After being washed thrice with TBST, protein bands were visualized by FluorChem Q and quantified by ImageJ software (National Institutes of Health, Bethesda, MD, USA).β-actin was used as a reference protein to calculate the relative expression of target proteins.
Reverse transcription-quantitative polymerase chain reaction (RT-qPCR)
Total RNAs in BALF, lung tissues and RAW264.7 macrophages were extracted with the TRIeasyTM Total RNA Extraction Reagent (Yeasen, China).Then, extracted RNA was reverse transcribed into cDNA with the Hifair® III 1st Strand cDNA Synthesis.Next, RT-qPCR was performed with the SYBR Green Real-Time PCR Master Mix (Yeasen, China) and detected on an Applied Biosystems.The relative mRNA levels were calculated using the 2 −ΔΔCt formula [19].GAPDH was applied as the internal reference gene.Primer sequences are listed in Table S2.
Statistical analysis
All experiments were performed at least three times.Prior to conducting parametric statistical tests, the normality of data distribution was confirmed using the Kolmogorov-Smirnov test.The data were analyzed by GraphPad Prism (software version 7.0) and are expressed as mean ± standard deviation.For two-group comparisons, an unpaired Student's t-test was used.One-way ANOVA followed by the Tukey-Kramer post hoc test was employed for multiple group comparisons.p < 0.05 was significantly different.
Songorine alleviates sepsis-induced ALI
A mouse model of septic ALI was established through LPS induction and was treated with either 20 or 40 mg/kg of songorine.HE staining showed severe lung injury in the LPS group, whereas treatment with songorine alleviated the lung injury in model mice (Figure 1(A)).Lung injury scores presented the same results with HE staining (p < 0.05; Figure 1(B)).Meanwhile, LPS-induced model mice had a higher lung W/D ratio than sham mice (p < 0.01).Treatment with songorine significantly reduced the lung W/D ratio of model mice (p < 0.05; Figure 1(C)).Moreover, cell apoptosis was detected in the lung tissues of mice.
Observations revealed a pronounced increase in TUNEL-positive cells in LPS-treated mice, indicating heightened apoptosis (p < 0.01).However, upon treatment with songorine, a marked decline in TUNEL-positive cells was evident (p < 0.05; Figure 1(D)).Additionally, western blotting revealed that the expression of anti-apoptotic protein Bcl2 was downregulated, whereas pro-apoptotic proteins (Bax and caspase-3) were upregulated in LPS-induced model mice (p < 0.01).Treatment with songorine reduced apoptosis in the lung tissues of the model mice, as evidenced by increased Bcl2 expression and decreased levels of Bax and caspase-3 (p < 0.01; Figure 1(E)).Of note, 40 mg/ kg songorine treatment showed better mitigative efficacy on septic ALI than 20 mg/kg songorine (p < 0.05; Figure 1(A-E)).
Songorine treatment inhibits inflammation and oxidative stress in septic ALI mice
Further, the effects of songorine on inflammation and oxidative stress in septic ALI were investigated.ELISA and RT-qPCR analyses showed that the levels of inflammatory factors (TNF-α, IL-6, IL-1β, and MPO) were higher in the BALF of LPS-induced model mice compared to the sham mice (p < 0.01).Treatment with songorine significantly reduced levels of these inflammatory factors in LPS-induced model mice (p < 0.05; Figure 2(A,B)).Also, CD68 is a histochemical/cytochemical marker of inflammation [20], which showed higher expression in LPS-induced model mice than in sham mice.After treatment with songorine, the CD68 expression was reduced in the model mice (Figure 2(C)).Furthermore, oxidative stress in the lung tissues of mice was detected.ELISA showed that SOD and GSH level was decreased, and MDA (a lipid peroxidation biomarker) level was increased in LPS-induced model mice compared to sham mice (p < 0.01).Treatment with songorine increased the SOD and GSH levels but decreased the MDA level in model mice (p < 0.05; Figure 2(D)).Additionally, the expression of antioxidant reactive elements (SOD2 and CAT) was downregulated in LPS-induced model mice, which was reversed by the treatment with songorine (p < 0.05; Figure 2(E)).Also, 40 mg/ kg songorine treatment showed better mitigative efficacy on septic ALI-induced inflammation and oxidative stress than 20 mg/kg songorine (p < 0.05; Figure 2(A-E)).
Songorine suppresses inflammation and oxidative stress in LPS-induced RAW264.7 cells
The mitigative effects of songorine on inflammation and oxidative stress induced by septic ALI were validated in vitro.The mouse monocyte cell line RAW 264.7 was treated with 1-60 μM songorine and then stimulated by 1 μg/mL LPS.Using the CCK-8 assay, we determined that the IC50 value for songorine treatment in RAW 264.7 cells was 18.97 μM (Figure 4(A)).Subsequently, cells were treated with 15 μM songorine to detect inflammation and oxidative stress.As shown in Figure 4(B-D), levels of MDA, TNF-α, IL-1β, IL-6, and MPO have increased in LPS-induced RAW 264.7 cells, significantly decreased by treatment with songorine (p < 0.01; Figure 4(B,C)).For oxidative stress, we found that SOD2, GSH, and CAT expression levels were reduced in LPS-induced RAW 264.7 cells, which was reversed by treatment with songorine (p < 0.01; Figure 4(B-D)).
Songorine represses inflammation and oxidative stress by activating the PI3K/AKT/NRF2 signaling pathway in LPS-induced RAW264.7 cells
Further, the mechanism of songorine protecting against septic ALI involved in the PI3K/AKT/NRF2 signaling pathway was verified in vitro.Consistent with the results in vivo, the expression of p-PI3K/PI3K, p-AKT/AKT, HO-1, NQO1, and NRF2 in the nucleus was downregulated in LPS-induced RAW 264.7 cells, while the expression of NRF2 was upregulated, which was reversed by treatment with songorine (p < 0.01; Figure 5(A)).Moreover, LY294002 (a PI3K inhibitor) addition offset the stimulatory effect of songorine on the PI3K/AKT/NRF2 signaling pathway in LPS-induced RAW264.7 cells (p < 0.01; Figure 5(B)).Functionally, the addition of LY294002 weakened the inhibitory effects on inflammation and oxidative stress in LPS-induced cells, as evidenced by increased levels of MDA, TNF-α, IL-1β, IL-6, and MPO, along with decreased levels of SOD2, GSH, and CAT (p < 0.01; Figure 5(C-E)).
Discussion
ALI is a critical indicator of mortality risk in patients with sepsis [21].Numerous studies have concentrated on ALI in sepsis.However, there remains no effective strategy for preventing or treating lung function impairment in septic conditions.This challenge primarily arises from the complex pathogenesis of sepsis [22].Therefore, the need for novel treatments for sepsis-induced ALI is pressing.Songorine has been shown to promote mitochondrial biogenesis and enhance antioxidant defense through NRF2 induction in sepsis [9].A previous study has demonstrated that inhibiting PI3K/AKT and CYP2E/NRF2/ROS signaling alleviates LPS-induced sepsis by suppressing oxidative stress-related inflammation [23].The current study found that songorine attenuated sepsis-induced ALI in mice via its anti-apoptotic, anti-inflammatory, and anti-oxidative effects through PI3K/AKT/ NRF2 signaling pathway.The key findings are as follows: first, songorine significantly alleviated lung injury in LPS-induced sepsis mice, as evidenced by the changes in histopathology, the decreased lung injury score, and W/D ratio; second, songorine reduced cell apoptosis in lung tissues of LPS-induced sepsis mice; third, the levels of proinflammatory cytokines in the BALF in LPS-induced sepsis mice and RAW264.7 cells were decreased by songorine treatment; fourth, songorine alleviated oxidative stress in LPS-induced sepsis mice and RAW264.7 cells; finally, songorine activated the PI3K/AKT/NRF2 signaling pathway in LPS-induced sepsis mice and RAW264.7 cells.
During the inflammatory response, the NRF2 pathway can be activated by multiple signal kinases.PI3K/AKT pathway is an essential upstream regulator of Nrf-2/HO-1 [11].The interaction of the PI3K/Akt and NRF2 signaling pathways can safeguard cells from inflammatory and oxidative harm.Also, the PI3K/AKT signaling pathway could mediate cell survival in many cell types, and many studies have shown that the PI3K/ AKT signaling pathway plays a vital role in treating ALI [24,25].These studies suggest that the PI3K/AKT/NRF2 signaling pathway is an essential regulator for ALI.
Studies have shown that songorine possesses antiinflammatory and antioxidant properties [7].Compared to other agents like buformin for treating sepsis-induced acute lung injury, the natural compound songorine offers higher safety, fewer side effects, and easier accessibility [9,26].Zyuz'kov et al. observed that songorine could stimulate the proliferation and differentiation of mesenchymal progenitor cells by activating PI3K signaling [27].However, the protective effect and mechanism involving PI3K signaling of songorine against ALI have yet to be elucidated.The current study used LPS to induce sepsis and ALI in mice.LPS-induced sepsis is a commonly used animal model for the pharmacological and pathological studies of sepsis treatment.LPS induction leads to an intense inflammatory response, culminating in the significant production and release of pro-inflammatory cytokines including IL-6, TNF-α, and IL-1β, which are directly implicated in the onset of septic ALI [28].Our study found that severe ALI was present after LPS induction, including pathological injury in lung tissue, increased lung W/D ratio, cell apoptosis, and inflammatory response.However, songorine administration markedly reduced LPS-induced lung injury, apoptosis, inflammation, and oxidative stress.The levels of pro-inflammatory cytokines TNF-α, IL-6, and IL-1β were decreased in LPS-induced sepsis mice treatment with songorine.Similar effects of songorine were also presented in LPS-induced sepsis mice and RAW264.7 cells.Meanwhile, we found that songorine upregulated the expression of PI3K/AKT/NRF2 signaling pathway-related proteins.Furthermore, LY294002 (a PI3K inhibitor) treatment reversed the protective effect of songorine on sepsis-induced ALI.These results indicated that songorine protects against sepsis-induced ALI via activating the PI3K/AKT/NRF2 signaling pathway.
Conclusion
Our study demonstrated that songorine significantly alleviated lung injury, inflammation, and oxidative stress in LPS-induced septic ALI.Moreover, our findings further delineate the activation of the PI3K/AKT/NRF2 signaling pathway by songorine, pointing toward its role in curtailing oxidative stress-induced inflammation.Despite these promising results, it is important to maintain a circumspective approach, considering that the research does not comprehensively address the possible adverse effects associated with songorine use.This aspect forms a crucial frontier for further research, seeking to ensure a balanced understanding of both its therapeutic efficacy and safety profile.Besides, we used only male mice to control for gender-related variations in the results.Future studies will consider including female mice to offer more comprehensive insights.This research serves as a foundational step in elucidating the protective roles orchestrated by songorine in the context of sepsis-induced ALI, highlighting a promising therapeutic avenue.In integrating these insights with existing literature, we pave the way for future investigations targeting the PI3K/AKT/NRF2 signaling axis in ALI.The upstream and downstream components of PI3K/AKT/ NRF2 pathway affected by songorine will be further explored.Consequently, songorine emerges as a promising therapeutic candidate for sepsis-induced ALI.
RAW264. 7
macrophages were obtained from Procell Life Science and Technology Co., Ltd.(China) with STR profile and cultured in Roswell Park Memorial Institute-(RPMI-) 1640
Figure 2 .
Figure 2. Songorine (Son) treatment inhibits inflammation and oxidative stress in septic acute lung injury mice.(a,B) Enzyme-linked immunosorbent assay (ElISa) and rt-qpcr for detecting the levels of inflammatory cytokines (tnF-α, Il-6, Il-1β, and mpo) in bronchoalveolar lavage fluid of mice.(c) Immunohistochemical staining revealing cd68 expression, a marker of inflammation, in lung tissues (scale bar = 200 μm).(d) ElISa for detecting oxidative stress-related biomarkers (Sod, gSh, and mda) in lung tissues of mice.(E) rt-qpcr analysis evaluated the relative mrna expression of key antioxidant genes, Sod2 and cat, in mice.the data are presented as means ± standard deviation (n = 6/group).two-group differences assessed by student's t-test; multiple group comparisons used one-way anoVa with tukey-Kramer post hoc.**p < 0.01 versus sham group; #p < 0.05 and ##p < 0.01 versus lpS group; ^p < 0.05 and ^^p < 0.01 versus lpS + Son (20) group. | 2023-11-19T06:18:23.049Z | 2023-11-17T00:00:00.000 | {
"year": 2024,
"sha1": "bd337c9396917d4ec05b22e5753ebbc4d8f0c709",
"oa_license": "CCBY",
"oa_url": "https://figshare.com/articles/journal_contribution/Songorine_inhibits_oxidative_stress-related_inflammation_through_PI3K_AKT_NRF2_signaling_pathway_to_alleviate_lipopolysaccharide-induced_septic_acute_lung_injury/24585437/1/files/43189898.pdf",
"oa_status": "GREEN",
"pdf_src": "TaylorAndFrancis",
"pdf_hash": "c8480ba72e85ff0da6e068bd45b0e564ada3f617",
"s2fieldsofstudy": [
"Medicine",
"Environmental Science"
],
"extfieldsofstudy": [
"Medicine"
]
} |
264988091 | pes2o/s2orc | v3-fos-license | Sexual dimorphism and dietary composition of the sandfi sh lizard Scincus scincus (Linnaeus, 1758) of southeastern Algeria
In this study, we present the fi rst data on the sexual dimorphism and diet of a typical desert lizard species, Scincus scincus , in the region of El-Oued, southeastern Algeria. The objective was to characterize the types of prey in the diet and determine whether there is a correlation between prey size and body size in males and females. The results obtained reveal that out of the 115 individuals captured (43 males and 72 females), sexual dimorphism is observed. Males tend to be larger than females, with males exhibiting relatively larger snout-vent length, head height, head length, and jaw length compared to females. However, no diff erence is recorded in terms of head width. The analysis of stomach contents allowed us to identify 485 prey items distributed among 9 prey categories, all attributed to insects. Coleoptera was the most commonly ingested prey category, accounting for 61.54% of stomachs, 59.79% of total prey items, and 62.41% of total volume. A similar diet was observed between females and males of S. scincus , with a high diet overlap ( O = 0.99) and a low diversity of prey types ingested by both males ( Ba = 0.17) and females ( Ba = 0.19). Coleoptera represented the most dominant order in the skink‘s diet, followed by the Hymenoptera order. Generally, the remaining seven taxa were consumed in low proportions by both sexes. Additionally, we observed no signifi cant diff erence between the number of prey consumed and the volume of prey between the sexes, and no correlation was found between morphometric characteristics, diet composition
Introduction
In the Algerian Sahara, the sandfi sh lizard, Scincus scincus, plays important ecological roles and is also valued for its dietary preferences among the local inhabitants.This reptile is highly regarded by the local population as a source of protein, serving as an alternative to meat and fi sh (Toumi et al., 2022).The analysis of the ecological niche and more particularly its trophic dimension constitutes an essential step in understanding the functioning of populations and ecosystems (Mori et al., 2018).In ectotherms species, digestive performance, energy budget, and energy allocation for diff erent activities including reproduction depend on the quality of the diet (Brewster et al., 2021).Indeed the energy needs and the search for food of an animal depend according to certain intrinsic factors: behavior of search for food, the body size, the sex and the ontogeny, and the availability of food resources (Vitt, 2000;Sales et al., 2012).The quantitative and qualitative variations of the trophic resources in the environment are likely to modify the diet and the dietary strategy of individuals (Millon et al., 2009).The animal has behavioral and physiological plasticity that allows them to adjust their food search eff ort according to the available trophic resources to increase their chances of survival (Guariento et al., 2018), and more particularly for animals living in arid and desert areas characterized by relatively limited trophic resources (Pianka, 2017).The sandfi sh lizard S. scincus is a diurnal Acta fytotechn zootechn, 26, 2023(3): 263-272 http://www.acta.fapz.uniag.skskink with a wide distribution from North Africa to the Middle East.This specie is strictly subservient to the desert environment; it makes the exception of Squamata by its ability to adapt to life subharenal, through a particular physiological mechanism (Vihar et al., 2015).However, few studies have addressed the trophic ecology and diet of Skinks S. scincus (Al-Sadoon et al., 1999).Our objective is to study sexual size dimorphism and diet composition in S. scincus, and to examine whether the observed size dimorphism can qualitatively and quantitatively influence the food spectrum of the lizards.We also aim to investigate whether there is a relationship between the size of ingested prey and the size of individuals (both males and females).
Study area
Fieldwork was conducted in the region of El Oued (33° 21' 21.9" N, 6° 51' 47.5" E), which is situated approximately 700 km southeast of Algiers (Figure 1).It is considered one of the primary oases in the northern Sahara of Algeria, located in the eastern Erg.The average annual temperature is around 26 °C, with summer temperatures reaching up to 45 °C and occasionally dropping to freezing point in winter.Precipitation is low, sporadic, and variable, with an average annual rainfall not exceeding 70 mm and rarely reaching 100 mm (Khezzani et al., 2016).
Lizards' collection and processing
The lizard specimens were collected by hand between October 2017 and November 2018, the capture process was the search for the place likely to find the lizard, through the footprints left on the surface of the ground surface to facilitate capture.The sex is determined by direct observation; the adult males differ from the females by the presence of dark spots on the back, while the females have a duller color which is similar to the juvenile livery.For this we have carried out verification through the gonads to avoid confusion.Once lizards are captured they are transported to the laboratory, where they are euthanized using formaldehyde (5%).The following measurements are taken using a digital caliper (0.01mm accuracy).The length of the nasal opening (SVL) from the tip of the nose to the anterior end of the cloacal cavity.Head length from the posterior margin of the tympanic membrane to the tip of the nose (HL).Head width (HW) at the widest part of the skull.Head height (HH) at maximum skull height.and jaw length (JL) (Figure 2).where: fi -the number of stomachs containing the prey category (i); (F) -the total number of stomachs non-empty stomachs where: ni -the number of prey of categories (i); (N) -the total number of prey where: vi -the volume of prey of categories (i); (V) -the total volume of prey The contribution of each prey category to the diet is quantified using the Relative Importance Index (% IRI) calculated by the following equation for males, females, and the total sample (Mesquita & Colli, 2003): Trophic niche diversity is calculated by the standardized Levin's index (Ba): where: (n) -the number of prey categories; (B) -the Levin's index of niche breadth: where: pi -the proportion of each category (i) This index describes the breadth of the trophic niche where a value of '1' indicates a generalized diet and a value of '0' indicates a specialized tendency (Krebs, 1999).
To compare dietary overlap between females and male, we used the formula: where: pij -the male proportion prey category (i); pikthe female proportion prey category (i), which has a range value from 0 (no similarity) to 1 (complete similarity The S. scincus diet is developed based on stomach content analysis, being the most accurate method (Arab & Doumandji, 2003;Paray et al., 2018).After dissection, the stomach contents of each specimen is then extracted and placed in a petri dish containing 75° alcohol.We analyzed the stomach contents under a binocular dissecting scope equipped with a micrometer eyepiece so as to, determine, count and measure the size (length and width) of the prey items so long as it is not too fragmented.Prey is identified at the low taxonomic level possible (order); the volume is estimated by the formula for a prolate spheroid: Three parameters are calculated: percentage of occurrence (%F), numerical percentage (%N) and Acta fytotechn zootechn, 26, 2023(3): 263-272 http://www.acta.fapz.uniag.sk For each lizard, we determine maximum and minimum prey size by considering the items with the largest and smallest volumes in each stomach, respectively.Lizards that ingested less than two prey items are excluded from the prey size analysis because estimates of the maximum and minimum prey sizes were inconsistent (Palmeira et al., 2021).
Data analyses
Firstly, we applied the Shapiro-wilk test to confirm the normality of all data (the Gaussian distribution) of all measurement parameters.Student's test (t-test) is used to verify the presence of differences between males and females in (SVL, HH, HL, and JL).The Mann-Whitney U test is applied to verify the existence of sexual differences in HW, diet for the number of food items ingested (Nprey) and the volumetric sum of all prey items inside the stomach (Vprey), and the minimum and maximum prey volume (Preymin and Preymax).The Person's correlation was useful to verify the existence of sexual correlation between SVL and HH, Hl, and JL.The correlation of Spearman is also to verify the existence of sexual correlation between the measurement of the body and the width of the head of S. scincus with the number and volume of prey consumed (SVL, HW, Nprey,Vprey, Prey max and Prey min).Statistical test and graphs are performed using SPSS 19.0 for Windows.Mean values are expressed as mean ±SE and P <0.05 is considered as statistically significant.
Morphometric parameters and Sexual size dimorphism
The 116, p = 0.000).With the T-test, the difference is clearly significant between the two sexes (p = 0.000) for the three parameters HH, HL and JL.Table 1 illustrates the corresponding data.
A nonparametric Mann-Whitney test is used for the parameter HW and it revealed that the difference between the two sexes is well reported for this parameters (p = 0.000) but the five parameters HW, Nprey, Vprey, Prey min and prey max revealed that only the difference between the two sexes is well signaled in the HW parameters (p = 0.000).However, the four parameters (Nprey, Vprey, Preymin and Preymax) related to the diet studied in both sexes, the differences are not significantly apparent according to Table 1.
The linear regression between SVL and HH, HL and JL shows a well determined correlation between these four parameters, according to the calculated significance (P = 0.000).In males, the strongest correlation is reported between SVL and HH with correlation coefficient of Pearson (R 2 = 68.80%),followed by SVL/JL, SVL/HW, SVL/ HL with a correlation coefficient of Pearson respectively (56.7%; 50%; and 47.9%) (Figure 3 and 4).In females, a correlation coefficient R 2 of 67.3% between the SVL and HL considered to be the most important and shows a positive correlation between body length and head width (HW).About SVL/HH, SVL/HW and SVL/JL are also correlated respectively with the correlation coefficient of R 2 60.5%, 50.5% and 37.6%.
Food composition
The comparison between the volume and the number of prey consumed by males and females of S. scincus according to the Mann-Whitney U test shown that there is no significant difference (p >0.05) for these parameters, and the same remark about the minimum and maximum volume of prey consumed of both sexes and SVL (Table 1).
A weak correlation is recorded between the SVL with the number and volume of prey consumed for both sexes, males and females, with a correlation coefficient (R 2 ) not exceeding 10%.
The food spectrum of S. scincus was composed mainly of Arthropods ( the presence of fragments of two scorpion species: Anthroctunus sp and Buthidae sp, and two plant species: Spartidium saharae and Retama retam are diagnosed in the stomach contents.The analysis of diet between both sexes shows that, males consumed 08 prey categories, and females 07.In addition, the Levi's index indicated a low diversity of ingested prey types in males (Ba = 0.17) and for females it is (Ba = 0.19).Pianka's indices showed high similarity between the two nutrient spectra (O = 0.99).The strong contribution of Coleoptera in their food spectrum is clearly visible (IM = 57% and IF = 65.74%)respectively in male and female, several identifiable genera and species, they are essentially represented in the male by Pimilia sp (12.60%) and Ptinus sp (10.08%).The same species of this prey category with approximately the same proportion is found in the diet of females, which is further distinguished by the consumption of a species of Cymindis sp with (17.54%) while only 3.36% which is represented in the diet of males.The Hymenoptera present a food supply in the lizards of (IM = 26.1%,IF = 22.59%) respectively in males and females.The dominant families of this order are: Formicidea, which is exclusively represented by many ant species,of which Messor sp is the most consumed ant species in females with (32.90%), against (8.7%) quantified in male, meanwhile, there is a slight consumption trend of Pheidole pallidula in females which is (11.11%) compared to males (28.98%).
It has been reported that the order of Lipidoptera does not appear in the food composition of males, also the order of Dermaptera and Orthoptera are absent from the food menu of females.
From the previous observations it is known that males generally have relatively larger body size and head dimensions than do females in many Lacertid lizards, as in other lizard families (Cox et al., 2003;Kaliontzopoulou et al., 2007;Zhao &Liu, 2013;Liang et al., 2021) This pattern is also confirmed in this study, in which a significant difference is observed between male and female of S. scincus which affects body size (SVL) and also in favors of the males with a larger snout vent length (SVL) of 94.75 mm.Hence, these results are in agreement with the results obtained by Babelhadj et al. (Babelhadj et al., 2021), though an average of (122 mm ±4.4) is mentioned by Toumi et al. (Toumi et al., 2022).It is under the effect of sexual selection that favors a large size for the male which confers an advantage in intrasexual behavior (Cox et al., 2003), and natural selection for large female size is related to a fecundity advantage of larger females (Cox et al., 2003;Du et al., 2005;Pincheira Donoso & Hunt, 2017).Indeed, the dimension of the head which affects the length, the height of the head also the length of the jaw is relatively more important in the males than that of the females Scincus, our results are confirmed by those obtained by Babelhadj et al. (Babelhadj et al., 2021).In fact, the studies carried out in this direction clearly show this intersexual differentiation, in many lizard families for exemple Agamidae (Withers &Thompson, 2005); Lacertidae (Kaliontzopoulou et al., 2008) and Scincidae (Ciracì et al., 2022).According to Brannelly et al. this is maybe due to sexual selection because larger head in males allows for higher bite force than that in females, which is important during intersexual competition (Brannelly et al., 2019).Moreover, because of larger bite forces, animals with larger heads can attack larger and harder prey (Arcos et al., 2017).feeding behavior in male reptiles depends on jaw length (JL) (Pianka & Vitt, 2003).
For example the food spectrum of S. scincus is very broad and consists primarily of terrestrial wide variety of invertebrates available in the study area, most of which are arthropods, especially insects.These diets have been documented in most lizard species, some herbivorous has been noted (Rastegar-Pouyani & Mohammad, 2016).However, food selectivity is noticed in our lizards specimens, this agrees with other previous study on the diet of Scincus genus (Al-Sadoon et al., 1999;Paray et al., 2018;Kadry, 2019).Indeed, some of the invertebrate preys are not used as a food source by S. scincus, the study by (Paray et al., 2018).Some studies have shown that Coleoptera and Hymenoptera are two of the most common orders of insects in the stomach contents of some lizards species (Puga & Colmenares et al., 2019).The strong representation of Coleoptera in the diet of S. scincus observed is 60%, this predominant consumption of this order has also been observed in sandfish lizards from arid zones, such as S. scincus from Egypt (Kadry, 2019), S. hemprichii and S. mitranus of Saudi Arabia (Al-Sadoon et al., 1999;Paray et al., 2018).The contribution of ants diet is in second position in the menu of Scincus but a low consumption, results also mentioned in the same species in Egypt (Kadry, 2019).Carretero reports that myrmecophagy is associated with poor environments with few nutritional resources (Carretero, 2004) like in arid and desert areas (Zhao & Liu, 2013).According to Pianka, desert habitats are characterized by low and changeable product, Coleoptera, Formicidae and Isoptera become the main prey of this ecosystem (Pianka, 2017).Therefore, it seems that these predators tend to prefer catching easier prey such as ants and beetles, and avoid relatively agile prey such as spiders and flies and consquently they are found in small proportions in the sampled stomachs of the Scincus.However, some studies, which that have focused on the diet of lizards, have revealed that their choice of prey is affected by the availability of such prey and also by the edaphic characteristics of the surrounding environment (de Fraga et al., 2022).Unfortunately, we did not sample potential prey in the habitat to verify whether their abundance in this lizard's diet reflected their abundance in the environment.This would have made it possible to highlight any selection phenomena.The results obtained reveal the presence of scorpion fragments in the stomach contents of the Scincus, which has not been mentioned before for the Scincus of the same family (Al-Sadoon et al., 1999;Paray et al., 2018), especially for the same species of lizards (Kadry, 2019), with the exception of the study by (Mouane et al., 2022).According to several authors (Pianka, 2017) the diet can be varied in the same genus of lizards even they cohabit the same habitat.
Plant consumption is also negligible.This can be explained by the fact that compared to animal prey, plant material is generally less energetic and difficult to assimilate (Cooper Jr &Vitt, 2002;Bombi et al., 2005).
Regarding the trophic diet of males and females of sand fish, there is a similarity in the composition, quantity of different categories of prey ingested, mainly Coleoptera order followed by Hymenopteran, like the case of the Agama, Phrynocephalus przewalskii of the Tengger desert of China, in which both sexes consumed similar prey types and ingested a similar amount of food items, while a size discrepancy is observed in males and females (Zhao & Liu, 2013).Sexual differences in the diet composition are usually attributed to sexual dimorphism in the body and the head dimensions that allows a differential consumption of prey between sexes, with the larger sex having the potential to consume larger prey (Shine et al., 2002), in addition, the head divergence between males and females could reduce both intraspecific and interspecific competition for food resources (Nel et al., 2015).The sexual dimorphism in S. scincus is well established, and depends on the size of the body and the different dimensions of the head, however, no relationship is observed between, the quantity and the size of the prey (volume).Taverne et al. and Zhao and Liu have found that prey size does not correlate with the degree of head dimorphism (Zhao & Liu, 2013;Taverne et al., 2019).
Conclusion
In conclusion, lizards play a significant role in the food chain of arid ecosystems.Our study on S. scincus has revealed morphological differences between males and females, with males being larger and having wider jaws.
There is a strong correlation between body size, head size, and jaw size.However, we did not find a correlation between body size and the prey consumed by S. scincus in terms of number and volume.Interestingly, both males and females exhibited a similar diet composition, showing a preference for consuming Coleoptera and Hymenoptera.For future research, it is important to further investigate why this species favors certain prey over others and to delve deeper into the feeding behavior of this lizard.Understanding whether this variability in diet is influenced by the natural distribution of prey, selective preferences, or other factors will contribute to a more comprehensive understanding of S. scincus and its ecological role in arid ecosystems.Advancing our knowledge of S. scincus and its ecological dynamics in arid ecosystems will provide valuable insights into population dynamics and interactions within these specific habitats.This information can contribute to improved conservation and management strategies for arid ecosystems, highlighting the importance of lizards in these fragile environments.
Figure 1
Figure 1Carte of the study area
Figure 5
Figure 5Difference in Prey number (Nprey) (left) and Volume of Prey (Vprey) (right) in both sexes application of the normality test is done on all the parameters studied in both sexes to know the normality of the natural distribution, by applying the test of Shapiro-Wilk.The four SVL, JL, HH and HL parameters have a p-value greater than 0.05 (p >0.05), which shows the significance of the normality of these parameters, whereas, the HW, Nprey, Vprey, Prey min and Prey max indicate a p-value less than 0.05 (p <0.000), with outlier values and non-normal distribution.So we apply the non-parametric tests according to this normality test.
A total of 115 adults of S. scincus (43 males and 72 females) are measured in this study.The largest male and female are 119.27mm and 109.70 mm, respectively.However the smallest male and female are 74.82 and 66.44 mm, respectively.Males are significantly larger than females, the mean SVL is different between sexes (males: 101.69 ±11.33 mm, females: 90.77 ±10.62 mm; t = 5.19, df =
Table 2
The Hymenoptra, is the order represented in the second position with (29.28%) of the diet.The proportions of the other categories of prey, namely: Arenae, Isopoda, Blattoptera, Dermaptera, Diptera, Orthoptera and Lepidoptera, do not exceed (5.36%).Additionally,Figure 4Linear regressions of head width (HW), jaw length (JL), and head higher (HH), on head length (HL), in S. scincus of 43 males and 72 females
Table 1
Descriptive statistics of morphological traits and consumed prey for females (n = 73) and males (n = 43) of 115 specimens of S. scincus
Table 2
Descriptors parameters of the taxonomic diet of S. scincus %F -frequency of occurrence; %N -relative abundance; %V -volumetric percentage; I -index of relative Importance; T -total sample; M -sample of adult males; F -sample of adult female; -Indicates no individuals of that prey category were found) http://www.acta.fapz.uniag.skSlovak University of Agriculture in Nitra Faculty of Agrobiology and Food Resources | 2023-11-04T15:16:08.842Z | 2023-10-16T00:00:00.000 | {
"year": 2023,
"sha1": "721c31480c2e6a6a3b967f320bf0aeffb9f54e15",
"oa_license": "CCBY",
"oa_url": "http://acta.fapz.uniag.sk/journal/article/download/291/94",
"oa_status": "GOLD",
"pdf_src": "ScienceParsePlus",
"pdf_hash": "a5539b4ff80149a36f30d6a305042f20f09eb18e",
"s2fieldsofstudy": [
"Biology",
"Environmental Science"
],
"extfieldsofstudy": []
} |
222112014 | pes2o/s2orc | v3-fos-license | Comparison of Different Methods of Automated Landform Classification at the Drainage Basin Scale: Examples from the Southern Italy
In this work, we tested the reliability of two different methods of automated landform classification (ACL) in three geological domains of the southern Italian chain with contrasting morphological features. ACL maps deriving from the TPI-based (topographic position index) algorithm are strictly dependent to the search input parameters and they are not able to fully capture landforms of different size. Geomorphons-based classification has shown a higher potential and can represent a powerful method of ACL, although it should be improved with the introduction of additional DEM-based parameters for the extraction of landform classes.
Introduction
In the last years, several factors such as availability of high-resolution DEMs, advances in computer power and the growing ability of research group to produce GIS-aided tools have promoted the development of many procedures and algorithms of automated classification/extraction of landforms [1]. Many works demonstrated that landform maps based on unsupervised classification represent a key tool in different research fields such as geomorphology, geology, archaeology, soil science and seismology [2][3][4][5][6][7]. Automatic classification of landform (ACL) provides several advantages than the traditional methods of geomorphological analysis, such as photo-interpretation and field surveys. Firstly, the use of an appropriate algorithm of landform classification overcomes the issue of the subjective interpretation of the user and the low reproducibility of "hand-made" geomorphological maps. Moreover, the full coverage of a study area can be useful in different applications, aimed at the investigation of the relationships between landforms and other parameters [8]. Although these factors should promote a fast growth of the application of ACL methods, the maps deriving by such an approach are frequently not able to fully define the spatial pattern of landforms and are affected by high-level noise. However, a relevant limitation of the method is due to its high scale-dependence. Landforms pertained to different geological landscapes have peculiar dimensions and in a first step the ACL should include the hierarchical definition of type and size of landforms occurring at different scales [1]. Most of the ACL approaches are focused on the characterization of mesoscale landforms (size-scale of hundred meters) and are strictly dependent of the search window size. Finally, this kind of approach does not provide any information about the time and origin of the geomorphic features [2], thus implying additional steps of "experts" in map interpretation.
Due to these limitations, the time-consuming traditional approaches of landform recognition and mapping are still the preferred ones by geomorphologists whereas extensive validation studies on the ACL are lacking.
In this paper, we try to fill this gap through the comparison of ACL maps coming from two of the most promising methods of automatic/semiautomatic landform extraction. The first method is the well-known TPI-based classification [9], and the second one is the geomorphons algorithm, recently proposed by [10]. Results of the ACL obtained by the two methods have been investigated in three different landscapes of southern Italy, featured by contrasting geological and geomorphological characters ( Fig. 1). A comparison between two different ACL methods was carried out in the three test areas and the robustness and reliability of the two unsupervised classification methods have been discussed.
Methods
The existing methods of ACL are mainly based on the extraction of first and second orders DEM-derivative attributes (slope and curvature). Local surface shape is frequently combined to the relative slope position of the landform elements thus to group the landscape sectors with similar statistical parameters and within a selected moving windows [1,11]. Available algorithms are strictly dependent of the moving windows size and clustering can be cell-based or object-based [12,13].
In this work, we test the accuracy of two different procedures of ACL in three study areas where also an interpretation of the "real" landforms based on an "expert" geomorphological analysis was realized. Application of the two algorithms of automated landform classification has been carried out using high-resolution DEMs (i.e. spatial resolution of 5 m for the axial and foredeep sectors and 8 m for the foreland area) deriving from Airborne LIDAR surveys. The DEMs are freely available.
The first one is a well-known algorithm that classifies landform units through a combination of TPI value and slope thresholds. TPI-based slope classification has been performed using an ArcGIS tool developed by Jenness Enterprises [9]. The algorithm evaluates the elevation difference between a cell and the average elevation around it within a predetermined radius. Positive TPI values indicate that the pixel has an elevation value higher than the average ones of the neighborhood searching window, whereas negative values are representative of a pixel with a lower elevation than the average of the search radius. TPI is strongly influenced by the search radius dimension: in particular, a large search radius highlights major landscape units, whereas the detection of minor landforms should be performed by using a small radius [4,6,14]. For this reason, in the Landform Classification Tool two search radii are inserted, a smaller and a larger neighborhood, in order to try to capture different sizes or landforms. After the calculation of the TPI index, a slope position index (i.e. SPI) classification was produced, where ten discrete classes (landforms) are obtained by combining the degree to which the TPI is lower or higher than the average and some slope classes set by default in the tool.
The second method used to realize a landform map and named as geomorphon, was proposed by [10] and implemented in a Grass GIS tool. It differs from the other existing algorithms of ACL because it does not use classic map algebra methods to calculate elevation differences inside the neighborhood. Conversely, it uses a computer vision approach, a pattern-based classification that self-adapts to the local topography. This technique utilizes the line-of-sight principle to evaluate a D quantity in each surrounding; more specifically, such an approach takes in count not only the elevation differences but also the zenith and nadir angles in the profiles and the lookup distance and should ensure identification of a landform at their most appropriate spatial scale [10]. Landform classification is derived from the extraction of Local Ternary Patterns (LTP). LTP are the basic micro-structure that constitute each existing type of landform and are named geomorphons; through the combination of geomorphon, landforms are extracted, in particular the first ten classes given back by the algorithm constitute the most frequent existing landform elements [10]. Also in this method, there are an inner and an outer search radius as input parameters, but the analysis is more independent from them compared to the TPI-based method. In fact, the neighborhood of the geomorphon method is self-adapting. The input radii instead have the aim to simplify the algorithm execution and bound it in the inserted areas, because otherwise, for each pixel, every time it would be achieved for all the raster. Finally, in the geomorphon method, also a flatness threshold (similar to the slope classification used in the TPIbased landform classification, but limited to areas considered flat) should be defined.
Regional Geological Framework
The comparison of the two methods of automatic extraction of landforms was carried out in three study areas belonging to different morphotectonic domains of the southern Italian chain. It is a northeast-verging fold-and-thrust belt deriving from the deformation of Mesozoic-Cenozoic shallow-water and deep-sea sedimentary succession pertained to the western border of the African-Apulian plate, and related Neogene-Pleistocene foredeep and satellite-basin deposits [15]. The foreland area is represented by the Apulian carbonate platform and represents the lower and subducting plate of the orogenic system. Starting from the Late Oligocene, the migration of the thrust front toward the north-east promoted the progressive deformation of pre-orogenic and syntectonic deposits from the inner (i.e. south-western) to the outer sectors of the belt. Pliocene-Pleistocene tectonic evolution of the chain was related to the activity of strikeslip and extensional faulting, which dismembered the contractional structure and promoted both the formation of many continental intermontane basins and the deep vertical incision of the fluvial net [16].
The landscape evolution of the southern Apennine chain was strictly related to its tectonic evolution, which in turn controlled also the distribution of the following NW-SE-trending main geological and geomorphological domains: the inner and axial zone of the chain (south-west sector), the outer zone of the chain (front of the chain/foredeep) and the foreland area (Fig. 1).
The axial zone of the chain is mainly constituted by impressive mountain ridges, bordered by steep slopes, frequently related to the activity of high-angle faults ( [16] and references therein). These fault-related mountain blocks are mainly carved by erosional processes in Mesozoic shallow-water carbonate and in deep-sea sedimentary rocks of the same age. Quaternary faulting and base level lowering promoted the creation and evolution of tectonic and morphological depressions, which are crossed by longitudinal V-shaped valleys with thalwegs generally placed between 500 and 700 m of elevation a.s.l. Another peculiar feature of this sector is the presence of several orders of lowangle erosional surfaces and fluvial terraces, which are arranged in a staircase geometry between 500 and 2000 m of elevation a.s.l. ( [16] and references therein).
The outer-zone and the front of the chain are featured by a NW-SE-trending thrust sheet system producing the same trend of morphostructural ridges, which are mainly carved by erosional processes affecting Cretaceous-to-Miocene pelagic deposits [15]. These units overlap the Pliocene and Pleistocene clastic deposits of the Bradano Foredeep, represented by a thick regressive succession of marine clay, and marine-totransitional sands and gravels. The top of the foredeep succession is formed by alluvial units corresponding to large and gently-dipping plain-surfaces deeply dissected by the main rivers. These rivers exhibit a low slope gradient of longitudinal profiles in the middle/lower reaches where flow as wide braided-and meander-type fluvial patterns [17][18][19]. Widespread badlands also occurred in sectors where marine clay deposits crop out [20,21].
The Foreland area is formed by an asymmetric NW-SE trending horst and graben system bordered by normal and transtensive faults inherited from late Cretaceous faulting activity. The Adriatic side (i.e.: the north-easternmost sector) of the foreland corresponds to a north-east gently-dipping carbonate landscape, which is also featured by karstic landforms and by the presence of several generations of Middle to Late Pleistocene marine terraces [22]. A well-developed drainage network cut both the karst landscape and the sequence of marine terraces: it is formed by regularly spaced and N-25-30°-trending flat-bottomed valleys with steep scarps [14].
ACL Map of the Study Areas
High-resolution DEMs of the three test areas have been processed in order to automatically extract the main landforms of some of the geological domains of the southern Italian Apennines (Fig. 1). In order to select the most appropriate search radii in the TPI-based classification, we compared TPI values for different search radii with landforms detected through geomorphological analysis. We chose a neighborhood with an annulus shape. The dimensions of the radii chosen are reported in Table 1: Similarly, for the geomorphon-based method different input parameters where inserted and chosen after a preliminary analysis of the main landforms of the study areas carried by an "expert" geomorphological analysis. For all the three study areas it was highlighted that the best radius is of 10 cells and the flatness threshold is 1.
TPI-derived maps showed a high level of influence from the search windows and the selection of the most suitable ACL map was done through a visual inspection of "manual" landform mapping and the support of "expert" geomorphological analyses. On the contrary, ACL maps based on geomorphon highlight a low level of dependency to the input parameters and neighborhood sizes, as supposed by the study of the algorithm. Axial Zone of the Chain Comparison between the TPI and geomorphon classifications was carried out in a peculiar landscape of the axial zone of the chain (Fig. 1). From a geomorphological viewpoint, the test area shows a heterogeneous landscape featured by an alternation of impressive ridges with steep slopes and flat depressions. Fluvial net is well developed and V-shaped valleys are prevalent. Moreover, a main braided river flowing into a wide floodplain (i.e. the Basento River) is placed in the north-western sector. Several orders of low-relief erosional land surfaces also occurred at the top and along the main slopes. In Fig. 2 results outcoming from the ACL analysis using the two above mentioned algorithms are shown.
The TPI-based ACL map discriminates 10 landform classes, which are distributed as follow: open slope and upper slope classes were the larger ones with a percentage of about 42% and 21%, respectively; fluvial landforms are mainly classified as U-shaped valleys (i.e. 18.7% of the total area) with a general overestimation of the "real" valleys size. Flat areas such as small intermontane basins and floodplains of main rivers are also well recognized.
The ACL map deriving from the geomorphon classification, using a line of sight of 50 m and a threshold for the flat area of 1°portrays an articulated landscape featured by a well-developed drainage network that dissects a number of ridges and slopes with different orientations. The slope class is the most representative landform whereas the flat areas exhibit a high level of noise. The map is able to delineate the fluvial net of the study area (i.e. "valley" class, Fig. 2) with a higher level of accuracy than the SPI map. Moreover, the mountain tops and ridges and the main watersheds appear well delineated.
Outer Front and Foredeep Area
The test area of this sector (Fig. 3) is a relatively «simple» landscape, since it is featured by a NE-dipping gently terraced surfaces moderately dissected by wide fluvial valleys and by the tributary fluvial net. The two ACL maps showed contrasting results (Fig. 3). In the TPI-based map, the most represented class is represented by flat areas, with a percentage of about 50.0% of the total area. Other relevant landform classes are open and upper slopes, which cover 23% and 12% of the total area, respectively. Fluvial net was largely identified as U-shaped valleys. This map shows a good potential to detect the sub-horizontal flat top terraced surface as well as the main valley flanks and their thalwegs.
The geomorphons-based map was extracted using the same search windows used for the southern Apennines axial zone and the flat threshold has been set to 1°. Slope class is the largest one with a percentage of 53% of the total area whereas the flat areas cover only 11% of the entire territory. This latter class can be associated to the undissected remnants of the terraced surfaces whereas the terrace edges are frequently classified as slope areas.
In this case, the geomorphons-based method is not able to differentiate valley flanks and gently-dipping surfaces, thus it could be useful to introduce a slope threshold, able to discriminate gently-dipping landform and steeper slopes.
Foreland Area
The third study area is a small catchment of the central sector of the Apulia foreland and corresponds to a karstic landscape carved in Cretaceous shallow-water limestones (Fig. 4). The test area is a low-relief carbonate landscape with typical karst features and a well-defined drainage network, known as the Murge. The latter is formed by regularly spaced and N-25-30°-trending flat-bottomed valleys with steep scarps [14].
These landforms have a subtle topographic expression and are hard to recognize by classical geomorphological analyses. Recently, the application of the TPI-based landform classification at a catchment scale has been already tested as a fast and effective approach to delineate the main landforms of the Murge area [14].
TPI-based ACL map (Fig. 4) is dominated by flat areas, which represent about 80% of the total area. Flat-bottomed valleys (U-shaped valley class, see Fig. 4) and their steep flanks are discontinuously delineated.
Geomorphons-based classification was performed using the same parameters adopted for axial and foredeep areas. The map showed a better delineation of the drainage network than the TPI-based classification and related SPI map and portrayed the main fluvio-karstic landforms of the study area. Due to the general low-angle dip of the landscape, slope class is the largest one in the map but local flat areas also occurred. These sectors coincide with NW-SE-elongated sub-horizontal areas where fluvial landforms disappeared (see for example the north-westernmost sector of the map) and can be related to marine terraces related to past ancient base levels.
Discussion and Concluding Remarks
Advantages and limitations of two different algorithms of ACL have been investigated in three sectors of southern Italy, characterized by contrasting geological and geomorphological characters. The first element that emerges from the study is that the comparison between the two ACL methods, in order to identify the best method, is not an easy task. In fact, an extensive field control was required to verify the full suitability of the results of ACL at a large scale. Nevertheless, preliminary geomorphological analysis based on "expert" users suggested that the two algorithms could represent a basic tool to recognize the main geomorphological elements of the study areas.
Visual comparison of the ACL maps in the study areas highlights a strong difference between the two algorithms, which provided different landform classifications mainly in the foredeep and foreland sectors. Although a robust assessment of the accuracy degree of landform extraction from the two algorithms should include a detailed geomorphological mapping and a quantitative comparison from "expert" delineation of landforms and automatic classification, we provide an attempt to quantitatively compare the results coming from the two ACL methods through the grouping of landform classes of the two methods with similar features. In particular, we argued that eight of ten classes extracted by the two algorithms can be reasonably considered as similar landforms whereas two classes (i.e. footslope and hollow in the geomorphons-method; U-shaped valleys and upland drainage in the TPI classification) does not have a direct correspondence. The similarity between each class of the two methods is highlighted using the same colours in Fig. 2, 3 and 4 and allowed us to quantitatively compare the two classification methods. For example, geomorphon class "named" as plain has been interpreted as similar to the class "flat" of the TPI-method, thus we represented both with light grey.
In order to investigate the obtained results from a quantitative viewpoint, we have extracted the total number of pixels showing the similar landform in the two methods. Results highlight that differences are apparently not so high since they vary from the 7 to 17% (diagrams in Fig. 5).
In particular, inequalities are greater than equalities in the axial zone of the chain and in the foreland area. Instead in the outer front and foredeep areas we have more similar pixel than with different landforms. However, despite to this measure, if we look at the difference/equalities maps (Fig. 5), we can see how much the results of the two methods differ and which are the hot spot of these differences: in the axial zone of the chain, geomorphons-method is able to delineate the fluvial net of the study area (i.e. "valley" class, Fig. 2) with a higher level of accuracy than the TPI map but tend to underestimate several sub-horizontal landforms of the study area such as alluvial plain and top of intermontane basins. ACL maps of the foredeep and foreland areas show more pronounced differences. In the foredeep area, the main geomorphological element (i.e. the gently-dipping terraced surfaces) is classified differently and TPI-based map is not able to detect minor fluvial channels and incisions. The geomorphons-based map portrays the undissected remnants of the terraced surfaces in a more satisfactory manner but it is not able to differentiate between the gently-dipping terrace surfaces and steeper flanks of the main valleys. A similar result can be observed in the foreland area where gemorphons-based provided a more detailed delineation of the main landforms than the TPI-based method. Also in this case, the simple introduction of a slope threshold in the geomorphons-method can help to improve the landform classification and to discriminate between peculiar geomorphological elements such as gentlydipping landsurfaces and steeper slopes related to fluvial deepening.
Our preliminary analyses suggested that geomorphons-based classification is a promising tool for automatic landform classification and can provide better results than more consolidated algorithms such as the TPI-based classification. Of course, the method needs to be deeply tested in other landscapes and/or at a wider scale but the encouraging preliminary results suggest a high potential of the proposed approach although the introduction of a slope threshold to differentiate gently-dipping and steeper slopes is required. | 2020-10-03T13:12:07.536Z | 2020-08-19T00:00:00.000 | {
"year": 2020,
"sha1": "da56c8b1840764456fbf42652b0b840b2970b0ab",
"oa_license": null,
"oa_url": null,
"oa_status": "CLOSED",
"pdf_src": "PubMedCentral",
"pdf_hash": "93dcd47b0d2f289943f0707c5560952a581aa003",
"s2fieldsofstudy": [
"Environmental Science",
"Geography",
"Geology"
],
"extfieldsofstudy": [
"Computer Science",
"Geology"
]
} |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.