Large-Scale Candidate Gene Analysis of HDL Particle Features

Citation: Kaess BM, Tomaszewski M, Braund PS, Stark K, Rafelt S, et al. ( Large-Scale Candidate Gene Analysis of HDL Particle Features Bernhard M. Kaess 0 Maciej Tomaszewski 0 Peter S. Braund 0 Klaus Stark 0 Suzanne Rafelt 0 Marcus Fischer 0 Robert Hardwick 0 Christopher P. Nelson 0 Radoslaw Debiec 0 Fritz Huber 0 Werner Kremer 0 Hans Robert Kalbitzer 0 Lynda M. Rose 0 Daniel I. Chasman 0 Jemma Hopewell 0 Robert Clarke 0 Paul R. Burton 0 Martin D. Tobin 0 Christian Hengstenberg 0 Nilesh J. Samani 0 Pieter H. Reitsma, Leiden University Medical Center, Netherlands 0 1 Department of Cardiovascular Science, University of Leicester , Leicester , United Kingdom , 2 Klinik und Poliklinik f u r Innere Medizin II, University of Regensburg , Regensburg, Germany, 3 LipoFIT Analytic GmbH, Regensburg, Germany , 4 Institut fu r Biophysik und physikalische Biochemie, University of Regensburg , Regensburg, Germany , 5 Department of Preventive Medicine, Brigham and Women's Hospital , Boston , Massachusetts, United States of America, 6 Clinical Trial Service Unit, University of Oxford , Oxford , United Kingdom , 7 Deptartment of Health Sciences, University of Leicester , Leicester , United Kingdom Background: HDL cholesterol (HDL-C) is an established marker of cardiovascular risk with significant genetic determination. However, HDL particles are not homogenous, and refined HDL phenotyping may improve insight into regulation of HDL metabolism. We therefore assessed HDL particles by NMR spectroscopy and conducted a large-scale candidate gene association analysis. Methodology/Principal Findings: We measured plasma HDL-C and determined mean HDL particle size and particle number by NMR spectroscopy in 2024 individuals from 512 British Caucasian families. Genotypes were 49,094 SNPs in .2,100 cardiometabolic candidate genes/loci as represented on the HumanCVD BeadChip version 2. False discovery rates (FDR) were calculated to account for multiple testing. Analyses on classical HDL-C revealed significant associations (FDR,0.05) only for CETP (cholesteryl ester transfer protein; lead SNP rs3764261: p = 5.6*10215) and SGCD (sarcoglycan delta; rs6877118: p = 8.6*1026). In contrast, analysis with HDL mean particle size yielded additional associations in LIPC (hepatic lipase; rs261332: p = 6.1*1029), PLTP (phospholipid transfer protein, rs4810479: p = 1.7*1028) and FBLN5 (fibulin-5; rs2246416: p = 6.2*1026). The associations of SGCD and Fibulin-5 with HDL particle size could not be replicated in PROCARDIS (n = 3,078) and/or the Women's Genome Health Study (n = 23,170). Conclusions: We show that refined HDL phenotyping by NMR spectroscopy can detect known genes of HDL metabolism better than analyses on HDL-C. - Funding: Recruitment and genotyping of the GRAPHIC cohort was funded by the British Heart Foundation. Phenotyping and BKs fellowship in Leicester were supported by grants from the National Genome Network (NGFNplus, 01GS0832) and the European Union-sponsored project Cardiogenics (LSH-2005-037593). NJS holds a British Heart Foundation Chair of Cardiology and MDT holds a Medical Research Council Clinician Scientist Fellowship (G0501942). NJS, MT and PRB are supported by the Leicester National Institute for Health Research (NIHR) Biomedical Research Unit in Cardiovascular Disease. NMR spectroscopy was performed at LipoFIT GmbH. WK and HRK are employed by the University of Regensburg and co-founders of LipoFIT GmbH; FH is co-founder and CEO of LipoFIT GmbH. With the exception of this, the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: WK and HRK are employed by the University of Regensburg and co-founders of LipoFIT GmbH; FH is co-founder and CEO of LipoFIT GmbH. However, direct financial benefits did not follow from this work. FH, WK and HRK contributed only to logistics, optimizing NMR spectroscopy and NMR data interpretation. They were not involved in the design of the study, genetical analyses or interpretation of the results. The involvement of LipoFIT GmbH and the affiliation of FH, WK and HKR to LipoFIT GmbH does not alter the authors adherence to all the PLoS ONE policies on sharing data and materials. . These authors contributed equally to this work. HDL cholesterol (HDL-C) is an established marker of cardiovascular risk. It reflects reverse cholesterol transport (RCT) and higher plasma HDL-C is associated with lower cardiovascular risk. However, the cholesterol content of HDL particles is mainly a historically founded and analytically feasible surrogate of RCT and is it not clear whether HDL-C is indeed the best RCT-related cardiovascular risk marker [1,2,3,4,5]. HDL particles are not a homogenous class but can be further divided into subclasses. Most studies suggest that large HDL particles are associated with a favourable outcome, while small HDL particles may even be positively correlated with CV risk [1,5,6]. Correspondingly, the lipid disturbances of the metabolic syndrome include a decreased mean HDL particle size [3,7]. Although HDL subclass measurements have been widely used for scientific purposes, practical clinical application is not well established yet, and there is an ongoing debate on whether lipid subclass measurements should be implemented into clinical routine [4,5,8]. The regulation of HDL metabolism and hence HDL particle features is not completely understood, but a strong genetic influence has been shown [9,10]. Recently, genome-wide association (GWA) studies have identified several loci that affect HDL-C [11,12,13]. However, only a small proportion of the heritability of HDL-C is explained by these loci. We hypothesized that a genetic analysis of distinct HDL particle traits may provide greater sensitivity and broaden our understanding of the genetic regulation of HDL metabolism. Accordingly, we conducted a large-scale candidate gene analysis on HDL particle traits. We genotyped 49,094 single nucleotide polymorphisms (SNPs) in .2,100 cardiometabolic candidate genes and assessed mean HDL particle size and HDL particle number by NMR spectroscopy. Ethics Statement The study was approved by the Leicestershire Research Ethics Committee, and all subjects provided written informed consent. The study conforms with the principles outlined in the Declaration of Helsinki, and all procedures followed were in accordance with institutional guidelines. Subjects Analyses were carried out in the GRAPHIC (Genetic Regulation of Arterial Pressure of Humans in the Community) cohort that has been previously described in detail [14]. In short, GRAPHIC contains 2,037 white European subjects in 520 nuclear families from the general population. Families were recruited by writing to women aged 40 to 69 registered with participating family practitioners in Leicestershire, UK, inviting them and their family to take part. Families were included if both parents aged 40 to 60 years and 2 of (...truncated)