Learn more: PMC Disclaimer | PMC Copyright Notice
Effect of Fenofibrate Therapy and ABCA1 Polymorphisms on High Density Lipoprotein Subclasses in the Genetics of Lipid Lowering Drugs and Diet Network
Abstract
Background
Previous studies have shown that ATP-binding cassette transporter 1 (ABCA1) polymorphisms associated with increased ABCA1 expression result in increased small HDL (high density lipoprotein) subclass particle concentration. The present study examines the effect of treatment with fenofibrate, a drug known to bind peroxisome proliferator-activated receptor alpha (PPARα) which increases the expression of ABCA1 gene, on lipoprotein subclass profiles of individuals stratified by ABCA1 genotypes.
Methods
Participants of Genetics of Lipid-lowering Drugs and Diet Network (GOLDN) were treated with fenofibrate over a three week period. We analyzed six ABCA1 polymorphisms in 287 GOLDN participants with triglyceride concentrations ≥ 150 mg/dL and studied their associations with HDL subclass particle concentrations, as measured by nuclear magnetic resonance spectroscopy, before and after treatment.
Results
Fenofibrate treatment did not result in significant changes in small HDL subclass particle concentration. When changes in HDL subclasses were stratified by ABCA1 polymorphism genotypes, there were no statistically significant associations between ABCA1 genotypes and small HDL subclasses before fenofibrate treatment. However, after fenofibrate treatment the KK genotype of R1587K (mean 4.40 μmol/L; p = 0.004) and the RK genotype of R219K (mean 1.60 μmol/L; p = 0.02) polymorphisms were associated with significantly increased small HDL. The R1587K KK genotype (mean 4.80 μmol/L; p = 0.0002) and the R219K KK genotype (mean 2.50 μmol/L; p = 0.02) were also associated with increased HDL particle concentrations.
Conclusion
There is a synergistic effect between ABCA1 polymorphisms and fenofibrate. Thus our study indirectly confirms the role of fenofibrate and genotype in increasing cholesterol efflux, as evidenced by increased small HDL particles.
1. INTRODUCTION
Fenofibrate is one of the predominant drugs in the fibric acid family used clinically to lower serum triglycerides while raising HDL-cholesterol (HDL-C). In vivo, fenofibrate is converted to fenofibric acid, which binds to the nuclear peroxisome proliferator-activated receptor alpha (PPARα) [1,2]. Recent studies have demonstrated that binding of fibrates to PPARα not only induces genes related to fatty acid metabolism, but also increases expression of the ATP-binding cassette transporter 1 (ABCA1) gene [3-5].
A transmembrane protein present on peripheral tissue cells, ABCA1 is involved in cellular lipid efflux and HDL metabolism; specifically, it transports free cholesterol out of the cell where cholesterol can then bind to apoA1 to form preβ-HDL [6]. Confirming its importance in HDL formation, mutations in the ABCA1 gene are responsible for the rare genetic HDL deficiency syndrome known as Tangier disease. Further, subsequent studies have shown that at least 10% of individuals with low HDL-C in the general population are heterozygous for rare deleterious mutations of the gene [7].
Apart from rare mutations of the ABCA1 gene, common polymorphisms have also been shown to have moderate influences on HDL-C concentration; specifically, the K allele of the ABCA1 R219K polymorphism has been associated with increased HDL-C concentration [8-11] and a reduced incidence of coronary artery disease (CAD) events [8,9]. In a subcohort of the Multi-Ethnic Study of Atherosclerosis (MESA) we demonstrated that the K allele of the R219K polymorphism (designated as the 1051 A allele in MESA) is associated with both a lower prevalence of coronary artery calcium even when adjusted for HDL-C [12] and increased small HDL particle concentration [13].
In addition to the influence of genotype, fenofibrate treatment has also been documented to increase HDL-C, mainly due to an increase in the concentration of smaller HDL subclasses, including HDL3 [14-16]. In the Veterans Affairs High-Density Lipoprotein Intervention Trial (VA-HIT), participants were treated with the commonly-prescribed fibrate gemfibrozil; subsequent measurements of LDL and HDL particle subclasses by proton NMR spectroscopy showed that LDL particle numbers were reduced while total and small HDL particle numbers were increased in treated individuals [15]. Overall, the data suggest that that fenofibrate, as well as ABCA1 genotype, influence HDL particle concentrations and potentially the risk of CAD.
We are currently unaware of any previous study that examined the combined effect of fenofibrate and ABCA1 polymorphisms on changes in either HDL-C and/or HDL subclasses. The Genetics of Lipid Lowering Drugs and Diet Network (GOLDN) is a study aimed to characterize the role of genetic and dietary factors on an individual’s response to fenofibrate. Thus, this study provides us with a unique opportunity to study both the effect of fenofibrate and genetic variations on the distribution of HDL subclasses. Here, we report changes in HDL subclasses before and after fenofibrate treatment in a subcohort of GOLDN, stratified according to several ABCA1 polymorphisms, to determine whether fenofibrate treatment and ABCA1 gene polymorphisms interact in their influence on HDL subclasses.
2. MATERIALS AND METHODS
2.1 Study design
GOLDN is an NIH-funded family study consisting of 3-generation pedigrees in two genetically homogeneous centers (Minneapolis, MN and Salt Lake City, UT) with predominantly Caucasian populations of European ancestry. The GOLDN study is part of the Program for Genetic Interaction (PROGENI) Network, a group of NIH-funded family-intervention studies focusing on gene-environment interactions. The primary aim of the GOLDN study was to characterize the role of genetic and dietary factors on an individual’s response to fenofibrate. The detailed design and methodology of the study has been previously described [17,18].
Briefly, clinical and biochemical measurements were collected in all individuals before and after exposure to the drug fenofibrate. After a screening visit and before the start of fenofibrate therapy, the study participants were asked to suspend their use of lipid-lowering drugs for four weeks prior to study participation. Fasting baseline blood was collected 4 to 8 weeks later on two consecutive days. Participants were given a 28-day supply of open label fenofibrate as 160 mg tablets (TriCor®, Abbott Laboratories, Chicago, IL) and instructed to take one fenofibrate tablet with a breakfast meal once daily for a minimum of 21 days. After this time period of drug intervention, participants returned for blood draws on two consecutive days. Fasting lipid concentrations reported are the average of the two measurements on consecutive days, either before or after fenofibrate treatment.
2.2 Study population
The GOLDN study population consisted of 428 men and 433 women who completed the fenofibrate therapy protocol. At baseline, 574 individuals had triglyceride concentrations < 150 mg/dL (mean ± SD: 91 ± 30 mg/dL); 287 individuals had triglyceride concentrations ≥ 150 mg/dL (mean ± SD: 242 ± 111 mg/dL). Analyses were restricted to those 287 participants with triglycerides ≥ 150 mg/dL for two reasons: first, they are a group who are more likely to receive triglyceride lowering therapy and thus clinically more relevant; and second, baseline triglyceride levels are a key determinant of triglyceride lowering response to fibrates.
Because the participants in GOLDN are members of families, polymorphism frequencies were calculated using one person from each sibship in the entire GOLDN study, for a total of 431 subjects.
Written informed consent was obtained from each participant at the screening visit. The protocol was approved by the Institutional Review Boards at the University of Minnesota, the University of Utah, and Tufts University.
2.3 Genotyping
DNA was extracted from peripheral leukocytes isolated from packed cells of anticoagulated blood using commercially available reagents (Puregene, Gentra Systems, Minneapolis, MN).
The ABCA1 rs2297404, rs2578575, rs4149272, and rs2230806 (R219K) polymorphisms were genotyped by Applied Biosystems TaqMan SNP system. The rs2230808 (R1587K) and rs2066714 (M883I) polymorphisms were genotyped by Sequenom using the massEXTEND reaction. All polymorphisms were in Hardy-Weinberg equilibrium, as determined by gene counting and χ2 tests using unrelated individuals in the full GOLDN study. Minor allele frequencies were as follows: rs2230806, A allele, 27%; rs2066714, G allele, 13%; rs2230808, A allele, 23%; rs2575875, A allele, 34%; rs4149272, A allele, 33%; rs2297404, C allele, 7%.
2.4 Biochemical assays
Cholesterol was measured using a cholesterol esterase, cholesterol oxidase reaction (Chol R1, Roche Diagnostics Corporation, Indianapolis, IN). This same reaction was used to measure HDL-C after precipitation of the non-HDL-C fractions with magnetic 50,000MW dextran sulfate and magnesium chloride. LDL-C was measured in plasma using a homogeneous direct assay manufactured by Genzyme Diagnostics (LDL Direct Liquid Select Reagent Kit, Equal Diagnostics, Exton, PA). Triglycerides were measured using a glycerol blanked enzymatic method (Trig/GB, Roche Diagnostics Corporation, Indianapolis, IN). All were determined on the Roche/Hitachi 911 Automatic Analyzer (Roche Diagnostics Corporation).
2.5 Lipoprotein subclass
Proton NMR spectroscopy was used to determine HDL subclass particle concentrations (Liposcience, Raleigh, NC). The concentrations of three HDL subclasses were measured: large HDL (defined as particles 8.8-13 nm), medium HDL (8.2-8.8 nm), and small HDL (7.3-8.2 nm).
2.6 Statistical analyses
For every lipid outcome, a linear mixed model was used to assess the effect of the ABCA1 polymorphisms on the baseline (before fenofibrate treatment) concentrations and the change (post minus pre-fenofibrate treatment) in concentrations. The model adjusted for gender, age, BMI, waist circumference, diabetes, and insulin concentration in comparison of baseline concentrations; pre-fenofibrate log-transformed triglyceride concentration and the pre-fenofibrate HDL subclass particle concentration of the subclass under comparison were added to this model for comparison of post minus pre-fenofibrate treatment. Log-transformation was used for medium HDL due to its skewed distribution. For log-transformed variables, the change is log of the post/pre-fenofibrate ratio. A random intercept model was used to account for familial structure.
3. RESULTS
Total cholesterol, LDL-C, and triglycerides were significantly lower after fenofibrate treatment and HDL-C was significantly higher (p < 0.0001 for all), as shown in Table 1. After fenofibrate treatment, mean HDL particle concentration and medium HDL particle subclass concentrations increased (p < 0.0001 for both) while mean HDL size was significantly reduced (p < 0.0001). Fenofibrate treatment had no effect on large or small HDL subclasses.
Table 1
Lipids and HDL subclass particle concentrations before and after fenofibrate treatment in 287 GOLDN participants with triglyceride ≥ 150 mg/dL.
| Before fenofibrate, mean ± SD | After fenofibrate, mean ± SD | p-value | |
|---|---|---|---|
| Cholesterol, mg/dL | 213.6 ± 38.7 | 188.9 ± 33.9 | < 0.0001 |
| HDL-C, mg/dL | 40.3 ± 10.6 | 44.0 ± 11.4 | < 0.0001 |
| LDL-C, mg/dL | 134.0 ± 32.0 | 125.0 ± 30.3 | < 0.0001 |
| Triglycerides, mg/dL | 212 (172, 280) a | 129 (100, 167) a | < 0.0001 |
| Small HDL particles, μmol/L | 23.5 ± 5.6 | 23.8 ± 7.1 | 0.31 |
| Medium HDL particles, μmol/L | 0.80 (0.00, 3.07) a | 2.37 (0.12, 7.17) a | < 0.0001 |
| Large HDL particles, μmol/L | 4.25 ± 3.16 | 4.34 ± 2.71 | 0.51 |
| HDL particles, μmol/L | 30.04 ± 6.42 | 32.57 ± 6.96 | < 0.0001 |
| HDL particle size, nm | 8.61 ± 0.38 | 8.52 ± 0.32 | < 0.0001 |
At baseline (before fenofibrate treatment), HDL size was significantly less in those individuals with the rs2297404 GC genotype, compared to the GG genotype (p = 0.024). Otherwise, there were no significant differences in mean HDL subclass particle concentrations between genotypes of any other ABCA1 polymorphisms, as shown in Table 2.
Table 2
Differences in HDL subclass particle concentrations between genotypes of ABCA1 polymorphisms, as compared to reference genotypes, at baseline (pre-fenofibrate treatment) in 287 GOLDN participants.
| small HDL, μmol/L | medium HDL, μmol/Lc | large HDL, μmol/L | HDL particles, μmol/L | HDL size, nm | ||||
|---|---|---|---|---|---|---|---|---|
|
| ||||||||
| Polymorphism (reference genotype) | na | Comparison genotype | n | mean (lower 95% CI, upper 95% CI)b | mean (lower 95% CI, upper 95% CI) | mean (lower 95% CI, upper 95% CI) | mean (lower 95% CI, upper 95% CI) | mean (lower 95% CI, upper 95% CI) |
| R1587K (RR) | 152 | RK | 84 | 1.00 (−0.50, 2.60) | −0.04 (−0.74, 0.67) | −0.29 (−1.00, 0.46) | 0.32 (−1.30, 2.00) | 0.03 (−0.07, 0.12) |
| KK | 14 | −0.53 (−3.70, 2.70) | −0.72 (−2.20, 0.74 | 0-0.67 (−2.20, 0.89) | −2.50 (−5.80, 0.89) | −0.02 (−0.22, 0.18) | ||
| R219K (RR) | 155 | RK | 113 | −0.10 (−1.70, 1.50) | −0.04 (−0.77, 0.68) | 0.00 (−0.77, 0.78) | 0.50 (−1.10, 2.10) | −0.01 (−0.11, 0.09) |
| KK | 19 | −0.01 (−2.90, 2.80) | −0.82 (−2.10, 0.48) | 0.41 (−0.98, 1.80) | 0.47 (−2.50, 3.40) | −0.05 (−0.23, 0.13) | ||
| M883I (II) | 191 | IM | 57 | −0.28 (−1.90, 1.40) | −0.06 (−0.82, 0.70) | −0.05 (−0.85, 0.75) | −0.32 (−2.10, 1.4) | −0.03 (−0.13, 0.08) |
| MM | 1 | 9.50 (−1.70, 21.00) | −2.50 (−7.60, 2.60) | −0.09 (−5.50, 5.30) | 9.30 (−2.40, 21.00) | 0.02 (−0.68, 0.720 | ||
| rs4149272 (GG) | 128 | GA | 128 | 0.15 (−7.70, 8.00) | −2.10 (−5.70, 1.50) | 1.60 (−2.20, 5.50) | 0.58 (−7.60, 8.80) | 0.01 (−0.48, 0.50) |
| AA | 31 | −7.90 (− 22.00, 5.90) | 2.40 (−3.90, 8.70) | 3.70 (−3.00, 10.00) | −1.60 (−16.00, 13.00) | 0.47 (−0.39, 1.30) | ||
| rs2297404 (GG) | 252 | GC | 34 | 1.80 (−0.49, 4.10) | −0.36 (−1.40, 0.68) | −0.31 (−1.40, 0.81) | 1.60 (−0.70, 3.90) |
−0.16 (−0.31, −
0.02) d |
| CC | 1 | −1.50 (− 12.00, 9.50) | 2.10 (−2.90, 7.10) | 1.40 (−4.00, 6.70) | 0.77 (−11.00, 12.00) | 0.06 (−0.63, 0.75) | ||
| rs2575875 (GG) | 126 | GA | 129 | 0.24 (−7.70, 8.20) | 1.90 (−1.70, 5.60) | −2.10 (−6.00, 1.80) | −1.30 (−9.60, 7.00) | −0.05 (−0.55, 0.44) |
| AA | 32 | 8.60 (−5.10, 22.00) | −2.60 (−8.90, 3.60) | −5.10 (−12.00, 1.60) | 0.23 (−14.00, 15.00) | −0.67 (−1.50, 0.19) | ||
Table 3 shows that quantitative changes in HDL subclass particle concentrations from baseline to after fenofibrate treatment are significant for the ABCA1 R1587K and R219K polymorphisms. When genotype and fenofibrate treatment were considered together, the KK genotype of the R1587K polymorphism was associated with 4.4-fold higher small HDL (p = 0.004) and 4.8-fold higher HDL particles (p = 0.0002), compared to the RR genotype. The R219K RK (p = 0.02) genotype was associated with 1.6-fold higher small HDL, compared to the RR genotype. In addition, HDL particle concentration was 2.5-fold higher in the KK genotype of the R219K polymorphism, compared to the RR genotype (p = 0.02). The rs2297404, rs2578575, rs4149272, and rs2066714 polymorphisms did not show significant associations with change in HDL particle concentrations.
Table 3
Differences in HDL subclass particle concentrations between genotypes of ABCA1 polymorphisms, as compared to reference genotypes, post-fenofibrate treatment in 287 GOLDN participants.
| small HDL, μmol/L | medium HDL, μmol/Lb | large HDL, μmol/L | HDL particles, μmol/L | HDL size, nm | ||
|---|---|---|---|---|---|---|
|
| ||||||
| Polymorphism (reference genotype) | Comparison genotype | mean (lower 95% CI, upper 95% CI)a | mean (lower 95% CI, upper 95% CI) | mean (lower 95% CI, upper 95% CI) | mean (lower 95% CI, upper 95% CI) | mean (lower 95% CI, upper 95% CI) |
| R1587K (RR) | RK | −0.48 (−1.90, 0.92) | −0.41 (−0.96, 0.14) | 0.27 (−0.19, 0.73) | −0.33 (−1.50, 0.85) | 0.03 (−0.04, 0.10) |
| KK | 4.40 (1.50, 7.40) c | −0.73 (−1.90, 0.44) | 0.28 (−0.69, 1.30) | 4.80 (2.30, 7.40) d | −0.04 (−0.18, 0.10) | |
| R219K (RR) | RK | 1.60 (0.20, 3.00) e | −0.26 (−0.81, 0.30) | −0.20 (−0.65, 0.26) | 0.61 (−0.59, 1.80) | −0.06 (−0.13, 0.01) |
| KK | 2.30 (−0.20, 4.80) | 0.54 (−0.45, 1.50) | 0.74 (−0.08, 1.60) | 2.50 (0.33, 4.60) e | 0.03 (−0.08, 0.15) | |
| M883I (II) | IM | −0.24 (−1.70, 1.20) | −0.23 (−0.81, 0.36) | 0.48 (−0.00, 0.97) | 0.49 (−0.76, 1.70) | 0.04 (−0.03, 0.11) |
| MM | −2.40 (−12.00, 7.50) | 3.80 (−0.11, 7.70) | −1.80 (−5.10, 1.40) | 0.48 (−8.00, 8.90) | −0.17 (−0.63, 0.29) | |
| rs4149272 (GG) | GA | 3.10 (−3.90, 10.00) | 1.60 (−1.20, 4.40) | −1.30 (−3.60, 0.96) | 2.80 (−3.20, 8.70) | −0.17 (−0.50, 0.16) |
| AA | 8.80 (−3.50, 21.00) | 4.10 (−0.77, 9.00) | −1.00 (−5.10, 3.00) | 9.30 (−1.10, 20.00) | −0.15 (−0.73, 0.42) | |
| rs2297404 (GG) | GC | 1.00 (−0.93, 3.00) | 0.51 (−0.27, 1.30) | −0.07 (−0.71, 0.56) | 1.30 (−0.38, 3.10) | 0.01 (−0.08, 0.11) |
| CC | −4.00 (−14.00, 5.80) | −0.90 (−4.70, 2.90) | 1.40 (−1.80, 4.60) | −6.50 (−15.00, 1.80) | 0.28 (−0.18, 0.74) | |
| rs2575875 (GG) | GA | −4.10 (−11.00, 2.90) | −1.60 (−4.40, 1.10) | 1.10 (−1.20, 3.40) | −3.20 (−9.20, 2.80) | 0.14 (−0.19, 0.47) |
| AA | −8.90 (−21.00, 3.30) | −4.10 (−8.90, 0.75) | 0.24 (−3.80, 4.30) | −9.10 (−19.00, 1.30) | 0.09 (−0.48, 0.67) | |
4. DISCUSSION
The current study examined the interaction of fenofibrate treatment with ABCA1 genotypes and their potential influences on lipoprotein subclasses. Specifically, we were interested in whether fenofibrate therapy worked synergistically with certain ABCA1 genotypes resulting in an enhanced pharmacological outcome of the drug, i.e. increased small particle HDL concentrations.
Mechanistically, fenofibrate has been shown to activate transcription of a number of genes involved in lipid metabolism via PPARα and its downstream peroxisome receptor response elements [1, 2]. PPARα has additionally been shown to induce expression of the ABCA1 gene [3-5]; functionally, ABCA1 regulates cholesterol efflux from cells and the subsequent conversion to apoAI, forming discoidal pre-beta-migrating HDL or α-nascent apoAI-containing particles, which are rapidly converted to HDL3-like particles [16,21]—particles associated with a lower risk of CAD. We speculate that our finding of increased small HDL particles is due to this rapid conversion of pre-beta HDL particles; direct measurement of pre-beta HDL particles would be needed for confirmation. In addition, it is not clear whether the increase in small particle HDL concentration we observed is entirely due to peripheral cell cholesterol efflux, given the roles of both liver and intestinal ABCA1 in HDL and cholesterol metabolism.
Previous studies have shown that increased ABCA1 expression and the subsequent increase in cholesterol efflux may result in an increase in HDL3-like small HDL particles [13,15]. In accordance with this observation, our laboratory demonstrated that the KK genotype of the R219K polymorphism of ABCA1 was associated with an increase in small HDL particles [13]; the finding suggests that the KK genotype is associated with increased ABCA1 activity, resulting in increased cholesterol efflux. Furthermore, studies such as VA-HIT [15] and Bezafibrate Coronary Atherosclerosis Intervention Trial (BECAIT) [14] have demonstrated that treatment with either gemfibrozil or bezafibrate can also result in increases in small HDL particles—again, potentially due to increased cholesterol efflux. Given the above reports that fenofibrate may mediate ABCA1 transcription resulting in increased cholesterol efflux, we tested the hypothesis that fenofibrate works in concert with certain ABCA1 genotypes to produce an enhanced result in those individuals.
In GOLDN, participants were treated with fenofibrate and their ABCA1 genotypes were determined. After three weeks of fenofibrate treatment, lipid profiles of the participants improved: cholesterol, LDL-C, and triglycerides significantly decreased and HDL-C significantly increased. Also, HDL particle concentration and medium HDL significantly increased while HDL size decreased. However, treatment with fenofibrate did not result in a significant increase in small HDL particles in this subcohort of GOLDN participants. While we previously demonstrated that the ABCA1 1051 A allele (aka R219K K allele) was associated with increased small HDL in the MESA population [13], in the current study there was no increase in small HDL subclass when stratified by ABCA1 R219K genotype. Rather, we show that the K alleles of R219K and R1587K are associated 1.6- and 4.4-fold increases in small HDL after fenofibrate treatment.
There are several factors which may be responsible for our inability to duplicate the findings from VA-HIT, BECAIT and MESA studies. Potentially the most important factor is the limited power of the current study based on the relatively small number of participants (n = 287) compared to MESA (n = 999). In addition, the VA-HIT study enrolled only men with coronary heart disease and low HDL, in contrast to the relatively healthy participants of the GOLDN study. Finally, to expand on previous work that examined the ABCA1 polymorphism R219K, five additional SNPs were examined in the current study—of these, R1587K was found to have a significant effect on small HDL concentration (p = 0.004). However, adjustments for multiple comparisons were not made, making additional studies necessary to confirm this finding.
Despite the limited statistical power of a small number of participants, the current study demonstrates that after fenofibrate treatment, individuals with certain ABCA1 genotypes showed increased small HDL particle concentrations. Though the mechanism of this phenomenon is still unclear, our finding does agree with previous studies showing that increased ABCA1-mediated cholesterol efflux, whether due to genotypic differences [13] or increased ABCA1 gene expression through PPARα activation, results in increased small HDL particles in plasma [14,15]. Further studies are needed to determine whether the fibrate class of drugs directly increases small HDL particle concentrations through a transcriptionally mediated increase in ABCA1 activity, as well as the clinical significance in terms of disease development and progression.
Acknowledgements
This study was funding by the National Heart, Lung, and Blood Institute grant no U01-HL072524, Genetic and Environmental Determinants of Triglycerides. We are grateful to the staff of the GOLDN study for the assistance in data collection and management.
Abbreviations
| HDL-C | HDL-cholesterol |
| PPARα | peroxisome proliferator-activated receptor alpha |
| ABCA1 | ATP-binding cassette transporter 1 |
| CAD | coronary artery disease |
| MESA | Multi-Ethnic Study of Atherosclerosis |
| VA-HIT | Veterans Affairs High-Density Lipoprotein Intervention Trial |
| GOLDN | Genetics of Lipid Lowering Drugs and Diet Network |
| BECAIT | Bezafibrate Coronary Atherosclerosis Intervention Trial |
| PROGENI | Program for Genetic Interaction |
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.