Discovery and refinement of loci associated with lipid levels

Disruption of ldlr causes increased LDL-c and vascular lipid accumulation in a zebrafish model of hypercholesterolemia

Hyperlipidemia and arterial cholesterol accumulation are primary causes of cardiovascular events. Monogenic forms of hyperlipidemia and recent genome-wide association studies indicate that genetics plays an important role. Zebrafish are a useful model for studying the genetic susceptibility to hyperlipidemia owing to conservation of many components of lipoprotein metabolism, including those related to LDL, ease of genetic manipulation, and in vivo observation of lipid transport and vascular calcification. We sought to develop a genetic model for lipid metabolism in zebrafish, capitalizing on one well-understood player in LDL cholesterol (LDL-c) transport, the LDL receptor (ldlr), and an established in vivo model of hypercholesterolemia. We report that morpholinos targeted against the gene encoding ldlr effectively suppressed its expression in embryos during the first 8 days of development. The ldlr morphants exhibited increased LDL-c levels that were exacerbated by feeding a high cholesterol diet. Increased LDL-c was ameliorated in morphants upon treatment with atorvastatin. Furthermore, we observed significant vascular and liver lipid accumulation, vascular leakage, and plaque oxidation in ldlr-deficient embryos. Finally, upon transcript analysis of several cholesterol-regulating genes, we observed changes similar to those seen in mammalian systems, suggesting that cholesterol regulation may be conserved in zebrafish. Taken together, these observations indicate conservation of ldlr function in zebrafish and demonstrate the utility of transient gene knockdown in embryos as a genetic model for hyperlipidemia.

Gene-environment dependence creates spurious gene-environment interaction

Gene-environment interactions have the potential to shed light on biological processes leading to disease and to improve the accuracy of epidemiological risk models. However, relatively few such interactions have yet been confirmed. In part this is because genetic markers such as tag SNPs are usually studied, rather than the causal variants themselves. Previous work has shown that this leads to substantial loss of power and increased sample size when gene and environment are independent. However, dependence between gene and environment can arise in several ways including mediation, pleiotropy, and confounding, and several examples of gene-environment interaction under gene-environment dependence have recently been published. Here we show that under gene-environment dependence, a statistical interaction can be present between a marker and environment even if there is no interaction between the causal variant and the environment. We give simple conditions under which there is no marker-environment interaction and note that they do not hold in general when there is gene-environment dependence. Furthermore, the gene-environment dependence applies to the causal variant and cannot be assessed from marker data. Gene-gene interactions are susceptible to the same problem if two causal variants are in linkage disequilibrium. In addition to existing concerns about mechanistic interpretations, we suggest further caution in reporting interactions for genetic markers.

Genetic predisposition to increased blood cholesterol and triglyceride lipid levels and risk of Alzheimer disease: a Mendelian randomization analysis

BACKGROUND

Although altered lipid metabolism has been extensively implicated in the pathogenesis of Alzheimer disease (AD) through cell biological, epidemiological, and genetic studies, the molecular mechanisms linking cholesterol and AD pathology are still not well understood and contradictory results have been reported. We have used a Mendelian randomization approach to dissect the causal nature of the association between circulating lipid levels and late onset AD (LOAD) and test the hypothesis that genetically raised lipid levels increase the risk of LOAD.

METHODS AND FINDINGS

We included 3,914 patients with LOAD, 1,675 older individuals without LOAD, and 4,989 individuals from the general population from six genome wide studies drawn from a white population (total n=10,578). We constructed weighted genotype risk scores (GRSs) for four blood lipid phenotypes (high-density lipoprotein cholesterol [HDL-c], low-density lipoprotein cholesterol [LDL-c], triglycerides, and total cholesterol) using well-established SNPs in 157 loci for blood lipids reported by Willer and colleagues (2013). Both full GRSs using all SNPs associated with each trait at p<5×10-8 and trait specific scores using SNPs associated exclusively with each trait at p<5 × 10-8 were developed. We used logistic regression to investigate whether the GRSs were associated with LOAD in each study and results were combined together by meta-analysis. We found no association between any of the full GRSs and LOAD (meta-analysis results: odds ratio [OR]=1.005, 95% CI 0.82-1.24, p = 0.962 per 1 unit increase in HDL-c; OR=0.901, 95% CI 0.65-1.25, p=0.530 per 1 unit increase in LDL-c; OR=1.104, 95% CI 0.89-1.37, p=0.362 per 1 unit increase in triglycerides; and OR=0.954, 95% CI 0.76-1.21, p=0.688 per 1 unit increase in total cholesterol). Results for the trait specific scores were similar; however, the trait specific scores explained much smaller phenotypic variance.

CONCLUSIONS

Genetic predisposition to increased blood cholesterol and triglyceride lipid levels is not associated with elevated LOAD risk. The observed epidemiological associations between abnormal lipid levels and LOAD risk could therefore be attributed to the result of biological pleiotropy or could be secondary to LOAD. Limitations of this study include the small proportion of lipid variance explained by the GRS, biases in case-control ascertainment, and the limitations implicit to Mendelian randomization studies. Future studies should focus on larger LOAD datasets with longitudinal sampled peripheral lipid measures and other markers of lipid metabolism, which have been shown to be altered in LOAD. Please see later in the article for the Editors' Summary.

Pleiotropic effects of lipid genes on plasma glucose, HbA1c, and HOMA-IR levels

Dyslipidemia is strongly associated with raised plasma glucose levels and insulin resistance (IR), and genome-wide association studies have identified 95 loci that explain a substantial proportion of the variance in blood lipids. However, the loci's effects on glucose-related traits are largely unknown. We have studied these lipid loci and tested their association collectively and individually with fasting plasma glucose (FPG), glycated hemoglobin (HbA1c), and IR in two independent cohorts: 10,995 subjects from LifeLines Cohort Study and 2,438 subjects from Prevention of Renal and Vascular Endstage Disease (PREVEND) study. In contrast to the positive relationship between dyslipidemia and glucose traits, the genetic predisposition to dyslipidemia showed a pleiotropic lowering effect on glucose traits. Specifically, the genetic risk score related to higher triglyceride level was correlated with lower levels of FPG (P = 9.6 × 10(-10) and P = 0.03 in LifeLines and PREVEND, respectively), HbA1c (P = 4.2 × 10(-7) in LifeLines), and HOMA of estimated IR (P = 6.2 × 10(-4) in PREVEND), after adjusting for blood lipid levels. At the single nucleotide polymorphism level, 15 lipid loci showed a pleiotropic association with glucose traits (P < 0.01), of which eight (CETP, MLXIPL, PLTP, GCKR, APOB, APOE-C1-C2, CYP7A1, and TIMD4) had opposite allelic directions of effect on dyslipidemia and glucose levels. Our findings suggest a complex genetic regulation and metabolic interplay between lipids and glucose.

Common genetic variants contribute to primary hypertriglyceridemia without differences between familial combined hyperlipidemia and isolated hypertriglyceridemia

BACKGROUND

The majority of hypertriglyceridemias are diagnosed as familial combined hyperlipidemia (FCHL) and primary isolated hypertriglyceridemias. The contribution of common genetic variants in primary hypertriglyceridemias and the genetic difference between FCHL and isolated hypertriglyceridemias have not been thoroughly examined.

METHODS AND RESULTS

This study involved 580 patients with hypertriglyceridemias and 403 controls. Of the 37 single nucleotide polymorphisms examined, 12 located in 10 genes showed allelic and genotype frequency differences between hypertriglyceridemias and controls. The minor alleles of APOE, APOA5, GALNTN2, and GCKR variants were positively correlated with plasma triglycerides, whereas minor alleles of ADIPOR2, ANGPTL3, LPL, and TRIB1 polymorphisms were inversely associated. Body mass index, glucose, sex, rs328 and rs7007797 in LPL, rs662799 and rs3135506 in APOA5, and rs1260326 in GCKR explained 36% of the variability in plasma triglycerides, 7.3% of which was attributable to the genetic variables. LPL, GCKR, and APOA5 polymorphisms fit dominant, recessive, and additive inheritance models, respectively. Variants more frequently identified in isolated hypertriglyceridemias were rs7412 in APOE and rs1800795 in IL6; rs2808607 in CYP7A1 and rs3812316 and rs17145738 in MLXIPL were more frequent in FCHL. The other 32 single nucleotide polymorphisms presented similar frequencies between isolated hypertriglyceridemias and FCHL.

CONCLUSIONS

Common genetic variants found in LPL, APOA5, and GCKR are associated with triglycerides levels in patients with primary hypertriglyceridemias. FCHL and isolated hypertriglyceridemias are probably trace to an accumulation of genetic variants predisposing to familial and sporadic hypertriglyceridemias or to hypertriglyceridemias and hypercholesterolemia in case of FCHL.

MR_predictor: a simulation engine for Mendelian Randomization studies

I present MR_predictor, a simulation engine designed to guide the development and interpretation of statistical tests of causality between phenotypes using genetic instruments. MR_predictor provides a framework to model either individual traits or complex scenarios where multiple phenotypes are correlated or dependent on each other. Crucially, MR_predictor can incorporate the effects of multiple biallelic loci (linked or unlinked) contributing genotypic variability to one or more simulated phenotypes. The software has a range of options for sample generation, and output files generated by MR_predictor port into commonly used analysis tools (e.g. PLINK, R), facilitating analyses germane for Mendelian Randomization studies. Benchmarks for speed and power calculations for summary statistic-based Mendelian Randomization analyses are presented and compared with analytical expectation.

AVAILABILITY AND IMPLEMENTATION

The simulation engine is implemented in PERL, and the associated scripts can be downloaded from github.com, and online documentation, tutorial and example datasets are available at http://coruscant.itmat.upenn.edu/mr_predictor.

LDL cholesterol: controversies and future therapeutic directions

Lifelong exposure to raised concentrations of LDL cholesterol increases cardiovascular event rates, and the use of statin therapy as an adjunct to diet, exercise, and smoking cessation has proven highly effective in reducing the population burden associated with hyperlipidaemia. Yet, despite consistent biological, genetic, and epidemiological data, and evidence from randomised trials, there is controversy among national guidelines and clinical practice with regard to LDL cholesterol, its measurement, the usefulness of population-based screening, the net benefit-to-risk ratio for different LDL-lowering drugs, the benefit of treatment targets, and whether aggressive lowering of LDL is safe. Several novel therapies have been introduced for the treatment of people with genetic defects that result in loss of function within the LDL receptor, a major determinant of inherited hyperlipidaemias. Moreover, the usefulness of monoclonal antibodies that extend the LDL-receptor lifecycle (and thus result in substantial lowering of LDL cholesterol below the levels achieved with statins alone) is being assessed in phase 3 trials that will enrol more than 60,000 at-risk patients worldwide. These trials represent an exceptionally rapid translation of genetic observations into clinical practice and will address core questions of how low LDL cholesterol can be safely reduced, whether the mechanism of LDL-cholesterol lowering matters, and whether ever more aggressive lipid-lowering provides a safe, long-term mechanism to prevent atherothrombotic complications.

A population genetic signal of polygenic adaptation

Adaptation in response to selection on polygenic phenotypes may occur via subtle allele frequencies shifts at many loci. Current population genomic techniques are not well posed to identify such signals. In the past decade, detailed knowledge about the specific loci underlying polygenic traits has begun to emerge from genome-wide association studies (GWAS). Here we combine this knowledge from GWAS with robust population genetic modeling to identify traits that may have been influenced by local adaptation. We exploit the fact that GWAS provide an estimate of the additive effect size of many loci to estimate the mean additive genetic value for a given phenotype across many populations as simple weighted sums of allele frequencies. We use a general model of neutral genetic value drift for an arbitrary number of populations with an arbitrary relatedness structure. Based on this model, we develop methods for detecting unusually strong correlations between genetic values and specific environmental variables, as well as a generalization of comparisons to test for over-dispersion of genetic values among populations. Finally we lay out a framework to identify the individual populations or groups of populations that contribute to the signal of overdispersion. These tests have considerably greater power than their single locus equivalents due to the fact that they look for positive covariance between like effect alleles, and also significantly outperform methods that do not account for population structure. We apply our tests to the Human Genome Diversity Panel (HGDP) dataset using GWAS data for height, skin pigmentation, type 2 diabetes, body mass index, and two inflammatory bowel disease datasets. This analysis uncovers a number of putative signals of local adaptation, and we discuss the biological interpretation and caveats of these results.

The process of adaptation is of fundamental importance in evolutionary biology. Within the last few decades, genotyping technologies and new statistical methods have given evolutionary biologists the ability to identify individual regions of the genome that are likely to have been important in this process. When adaptation occurs in traits that are underwritten by many genes, however, the genetic signals left behind are more diffuse, and no individual region of the genome is likely to show strong signatures of selection. Identifying this signature therefore requires a detailed annotation of sites associated with a particular phenotype. Here we develop and implement a suite of statistical methods to integrate this sort of annotation from genome wide association studies with allele frequency data from many populations, providing a powerful way to identify the signal of adaptation in polygenic traits. We apply our methods to test for the impact of selection on human height, skin pigmentation, body mass index, type 2 diabetes risk, and inflammatory bowel disease risk. We find relatively strong signals for height and skin pigmentation, moderate signals for inflammatory bowel disease, and comparatively little evidence for body mass index and type 2 diabetes risk.

Extension of GWAS results for lipid-related phenotypes to extreme obesity using electronic health record (EHR) data and the Metabochip

A variety of health-related data are commonly deposited into electronic health records (EHRs), including laboratory, diagnostic, and medication information. The digital nature of EHR data facilitates efficient extraction of these data for research studies, including genome-wide association studies (GWAS). Previous GWAS have identified numerous SNPs associated with variation in total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), and triglycerides (TG). These findings have led to the development of specialized genotyping platforms that can be used for fine-mapping and replication in other populations. We have combined the efficiency of EHR data and the economic advantages of the Illumina Metabochip, a custom designed SNP chip targeted to traits related to coronary artery disease, myocardial infarction, and type 2 diabetes, to conduct an array-wide analysis of lipid traits in a population with extreme obesity. Our analyses identified associations with 12 of 21 previously identified lipid-associated SNPs with effect sizes similar to prior results. Association analysis using several approaches to account for lipid-lowering medication use resulted in fewer and less strongly associated SNPs. The availability of phenotype data from the EHR and the economic efficiency of the specialized Metabochip can be exploited to conduct multi-faceted genetic association analyses.

Association of the ST3GAL4 rs11220462 polymorphism and serum lipid levels in the Mulao and Han populations

Background

A previous genome-wide association study has displayed the association of the ST3 beta-galactoside alpha-2,3-sialytransferase 4 (ST3GAL4) gene variant and lipid traits in the individuals of European ancestry, but the reproducibility of this association has not been detected in the Chinese population. The present study was undertaken to detect the association of ST3GAL4 rs11220462 single nucleotide polymorphism (SNP) and several environmental factors with serum lipid profiles in the Mulao and Han populations.

Methods

A total of 700 unrelated individuals of Mulao nationality and 694 subjects of Han nationality were randomly selected from our previous stratified randomized samples. Genotypes of the SNP were determined via polymerase chain reaction and restriction fragment length polymorphism in combination with gel electrophoresis, and then verified by direct sequencing.

Results

Serum apolipoprotein (Apo) B levels were higher and the ApoAI/ApoB ratio was lower in Mulao than in Han (P < 0.05-0.01). There were no significant differences in the genotypic and allelic frequencies of the ST3GAL4 rs11220462 SNP between the two ethnic groups or between males and females. The A allele carriers in both Mulao males and females had higher total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), and ApoB levels than the A allele non-carriers (P < 0.05-0.01). The subjects with AA genotype in Han males but not in females had higher TC and triglyceride (TG) levels than the subjects with AG or GG genotype (P < 0.01 for each). Multiple linear regression analyses showed that the levels of TC, LDL-C and ApoB in Mulao females; TC and LDL-C in Mulao males; and TC in Han males were correlated with the genotypes (P < 0.05-0.001). Serum lipid parameters were also associated with several environmental factors in both ethnic groups (P < 0.05 -0.001).

Conclusions

The association of ST3GAL4 rs11220462 SNP and serum lipid levels was different between the Mulao and Han populations, suggesting that there may be a racial/ethnic-specific association, and/or sex-specific association between the ST3GAL4 rs11220462 SNP and serum lipid parameters in some ethnic groups.

Shared common variants in prostate cancer and blood lipids

BACKGROUND

Epidemiological and clinical studies suggest comorbidity between prostate cancer (PCA) and cardiovascular disease (CVD) risk factors. However, the relationship between these two phenotypes is still not well understood. Here we sought to identify shared genetic loci between PCA and CVD risk factors.

METHODS

We applied a genetic epidemiology method based on conjunction false discovery rate (FDR) that combines summary statistics from different genome-wide association studies (GWAS), and allows identification of genetic overlap between two phenotypes. We evaluated summary statistics from large, multi-centre GWA studies of PCA (n=50 000) and CVD risk factors (n=200 000) [triglycerides (TG), low-density lipoprotein (LDL) cholesterol and high-density lipoprotein (HDL) cholesterol, systolic blood pressure, body mass index, waist-hip ratio and type 2 diabetes (T2D)]. Enrichment of single nucleotide polymorphisms (SNPs) associated with PCA and CVD risk factors was assessed with conditional quantile-quantile plots and the Anderson-Darling test. Moreover, we pinpointed shared loci using conjunction FDR.

RESULTS

We found the strongest enrichment of P-values in PCA was conditional on LDL and conditional on TG. In contrast, we found only weak enrichment conditional on HDL or conditional on the other traits investigated. Conjunction FDR identified altogether 17 loci; 10 loci were associated with PCA and LDL, 3 loci were associated with PCA and TG and additionally 4 loci were associated with PCA, LDL and TG jointly (conjunction FDR <0.01). For T2D, we detected one locus adjacent to HNF1B.

CONCLUSIONS

We found polygenic overlap between PCA predisposition and blood lipids, in particular LDL and TG, and identified 17 pleiotropic gene loci between PCA and LDL, and PCA and TG, respectively. These findings provide novel pathobiological insights and may have implications for trials using targeting lipid-lowering agents in a prevention or cancer setting.

Biomarkers of coronary artery disease: the promise of the transcriptome

The last years have witnessed tremendous technical advances in the field of transcriptomics that enable the simultaneous assessment of nearly all transcripts expressed in a tissue at a given time. These advances harbor the potential to gain a better understanding of the complex biological systems and for the identification and development of novel biomarkers. This article will review the current knowledge of transcriptomics biomarkers in the cardiovascular field and will provide an overview about the promises and challenges of the transcriptomics approach for biomarker identification.

Genetic susceptibility to type 2 diabetes and obesity: from genome-wide association studies to rare variants and beyond

During the past 7 years, genome-wide association studies have shed light on the contribution of common genomic variants to the genetic architecture of type 2 diabetes, obesity and related intermediate phenotypes. The discoveries have firmly established more than 175 genomic loci associated with these phenotypes. Despite the tight correlation between type 2 diabetes and obesity, these conditions do not appear to share a common genetic background, since they have few genetic risk loci in common. The recent genetic discoveries do however highlight specific details of the interplay between the pathogenesis of type 2 diabetes, insulin resistance and obesity. The focus is currently shifting towards investigations of data from targeted array-based genotyping and exome and genome sequencing to study the individual and combined effect of low-frequency and rare variants in metabolic disease. Here we review recent progress as regards the concepts, methodologies and derived outcomes of studies of the genetics of type 2 diabetes and obesity, and discuss avenues to be investigated in the future within this research field.

Whole exome sequencing of familial hypercholesterolaemia patients negative for LDLR/APOB/PCSK9 mutations

BACKGROUND

Familial hypercholesterolaemia (FH) is an autosomal dominant disease of lipid metabolism, which leads to early coronary heart disease. Mutations in LDLR, APOB and PCSK9 can be detected in 80% of definite FH (DFH) patients. This study aimed to identify novel FH-causing genetic variants in patients with no detectable mutation.

METHODS AND RESULTS

Exomes of 125 unrelated DFH patients were sequenced, as part of the UK10K project. First, analysis of known FH genes identified 23 LDLR and two APOB mutations, and patients with explained causes of FH were excluded from further analysis. Second, common and rare variants in genes associated with low-density lipoprotein cholesterol (LDL-C) levels in genome-wide association study (GWAS) meta-analysis were examined. There was no clear rare variant association in LDL-C GWAS hits; however, there were 29 patients with a high LDL-C SNP score suggestive of polygenic hypercholesterolaemia. Finally, a gene-based burden test for an excess of rare (frequency <0.005) or novel variants in cases versus 1926 controls was performed, with variants with an unlikely functional effect (intronic, synonymous) filtered out.

CONCLUSIONS

No major novel locus for FH was detected, with no gene having a functional variant in more than three patients; however, an excess of novel variants was found in 18 genes, of which the strongest candidates included CH25H and INSIG2 (p<4.3×10(-4) and p<3.7×10(-3), respectively). This suggests that the genetic cause of FH in these unexplained cases is likely to be very heterogeneous, which complicates the diagnostic and novel gene discovery process.

Very low density lipoprotein cholesterol associates with coronary artery calcification in type 2 diabetes beyond circulating levels of triglycerides

OBJECTIVE

While recent genomic studies have focused attention on triglyceride (TG) rich lipoproteins in cardiovascular disease (CVD), little is known of very low-density lipoprotein cholesterol (VLDL-C) relationship with atherosclerosis and CVD. We examined, in a high-risk type-2 diabetic population, the association of plasma VLDL-C with coronary artery calcification (CAC).

METHODS

The Penn Diabetes Heart Study (PDHS) is a cross-sectional study of CVD risk factors in type-2 diabetics (n = 2118, mean age 59.1 years, 36.5% female, 34.1% Black). Plasma lipids including VLDL-C were calculated (n = 1879) after ultracentrifugation.

RESULTS

In Tobit regression, VLDL-C levels were positively associated with increasing CAC after adjusting for age, race, gender, Framingham risk score, body mass index, C-reactive protein, exercise, medication and alcohol use, hemoglobin A1c, and diabetes duration [Tobit ratio (TR) and 95% confidence interval (CI) 0.38 (0.12-0.65), P = 0.005] and even after inclusion of apolipoprotein B data [TR 0.31 (0.03-0.58), P = 0.030]. Approximately 3-fold stronger effect was observed in women [TR 0.75 (0.16-1.34), P = 0.013] than men [TR 0.20 (-0.10-0.50), P = 0.189; gender interaction P = 0.034]. Plasma VLDL-C was related more strongly to CAC scores than TG levels (e.g., Akaike information criteria of 7263.65 vs. 7263.94) and had stronger CAC association in individuals with TGs >150 mg/dl (TR 0.80, P = 0.010) vs. those with TGs <150 mg/dl (TR 0.27, P = 0.185).

CONCLUSIONS

In PDHS, VLDL-C is associated with CAC independent of established CVD risk factors, particularly in women, and may have value even beyond apolipoprotein B levels and in patients with elevated TGs.

A systematic evaluation of short tandem repeats in lipid candidate genes: riding on the SNP-wave

Structural genetic variants as short tandem repeats (STRs) are not targeted in SNP-based association studies and thus, their possible association signals are missed. We systematically searched for STRs in gene regions known to contribute to total cholesterol, HDL cholesterol, LDL cholesterol and triglyceride levels in two independent studies (KORA F4, n = 2553 and SAPHIR, n = 1648), resulting in 16 STRs that were finally evaluated. In a combined dataset of both studies, the sum of STR alleles was regressed on each phenotype, adjusted for age and sex. The association analyses were repeated for SNPs in a 200 kb region surrounding the respective STRs in the KORA F4 Study. Three STRs were significantly associated with total cholesterol (within LDLR, the APOA1/C3/A4/A5/BUD13 gene region and ABCG5/8), five with HDL cholesterol (3 within CETP, one in LPL and one inAPOA1/C3/A4/A5/BUD13), three with LDL cholesterol (LDLR, ABCG5/8 and CETP) and two with triglycerides (APOA1/C3/A4/A5/BUD13 and LPL). None of the investigated STRs, however, showed a significant association after adjusting for the lead or adjacent SNPs within that gene region. The evaluated STRs were found to be well tagged by the lead SNP within the respective gene regions. Therefore, the STRs reflect the association signals based on surrounding SNPs. In conclusion, none of the STRs contributed additionally to the SNP-based association signals identified in GWAS on lipid traits.

Assessing multivariate gene-metabolome associations with rare variants using Bayesian reduced rank regression

MOTIVATION

A typical genome-wide association study searches for associations between single nucleotide polymorphisms (SNPs) and a univariate phenotype. However, there is a growing interest to investigate associations between genomics data and multivariate phenotypes, for example, in gene expression or metabolomics studies. A common approach is to perform a univariate test between each genotype-phenotype pair, and then to apply a stringent significance cutoff to account for the large number of tests performed. However, this approach has limited ability to uncover dependencies involving multiple variables. Another trend in the current genetics is the investigation of the impact of rare variants on the phenotype, where the standard methods often fail owing to lack of power when the minor allele is present in only a limited number of individuals.

RESULTS

We propose a new statistical approach based on Bayesian reduced rank regression to assess the impact of multiple SNPs on a high-dimensional phenotype. Because of the method's ability to combine information over multiple SNPs and phenotypes, it is particularly suitable for detecting associations involving rare variants. We demonstrate the potential of our method and compare it with alternatives using the Northern Finland Birth Cohort with 4702 individuals, for whom genome-wide SNP data along with lipoprotein profiles comprising 74 traits are available. We discovered two genes (XRCC4 and MTHFD2L) without previously reported associations, which replicated in a combined analysis of two additional cohorts: 2390 individuals from the Cardiovascular Risk in Young Finns study and 3659 individuals from the FINRISK study.

AVAILABILITY AND IMPLEMENTATION

R-code freely available for download at http://users.ics.aalto.fi/pemartti/gene_metabolome/.

Loss-of-function mutations in APOC3, triglycerides, and coronary disease

BACKGROUND

Plasma triglyceride levels are heritable and are correlated with the risk of coronary heart disease. Sequencing of the protein-coding regions of the human genome (the exome) has the potential to identify rare mutations that have a large effect on phenotype.

METHODS

We sequenced the protein-coding regions of 18,666 genes in each of 3734 participants of European or African ancestry in the Exome Sequencing Project. We conducted tests to determine whether rare mutations in coding sequence, individually or in aggregate within a gene, were associated with plasma triglyceride levels. For mutations associated with triglyceride levels, we subsequently evaluated their association with the risk of coronary heart disease in 110,970 persons.

RESULTS

An aggregate of rare mutations in the gene encoding apolipoprotein C3 (APOC3) was associated with lower plasma triglyceride levels. Among the four mutations that drove this result, three were loss-of-function mutations: a nonsense mutation (R19X) and two splice-site mutations (IVS2+1G→A and IVS3+1G→T). The fourth was a missense mutation (A43T). Approximately 1 in 150 persons in the study was a heterozygous carrier of at least one of these four mutations. Triglyceride levels in the carriers were 39% lower than levels in noncarriers (P<1×10(-20)), and circulating levels of APOC3 in carriers were 46% lower than levels in noncarriers (P=8×10(-10)). The risk of coronary heart disease among 498 carriers of any rare APOC3 mutation was 40% lower than the risk among 110,472 noncarriers (odds ratio, 0.60; 95% confidence interval, 0.47 to 0.75; P=4×10(-6)).

CONCLUSIONS

Rare mutations that disrupt APOC3 function were associated with lower levels of plasma triglycerides and APOC3. Carriers of these mutations were found to have a reduced risk of coronary heart disease. (Funded by the National Heart, Lung, and Blood Institute and others.).

Linkage analysis incorporating gene-age interactions identifies seven novel lipid loci: the Family Blood Pressure Program

OBJECTIVE

To detect novel loci with age-dependent effects on fasting (≥ 8 h) levels of total cholesterol, high-density lipoprotein, low-density lipoprotein, and triglycerides using 3600 African Americans, 1283 Asians, 3218 European Americans, and 2026 Mexican Americans from the Family Blood Pressure Program (FBPP).

METHODS

Within each subgroup (defined by network, race, and sex), we employed stepwise linear regression (retention p ≤ 0.05) to adjust lipid levels for age, age-squared, age-cubed, body-mass-index, current smoking status, current drinking status, field center, estrogen therapy (females only), as well as antidiabetic, antihypertensive, and antilipidemic medication use. For each trait, we pooled the standardized male and female residuals within each network and race and fit a generalized variance components model that incorporated gene-age interactions. We conducted FBPP-wide and race-specific meta-analyses by combining the p-values of each linkage marker across subgroups using a modified Fisher's method.

RESULTS

We identified seven novel loci with age-dependent effects; four total cholesterol loci from the meta-analysis of Mexican Americans (on chromosomes 2q24.1, 4q21.21, 8q22.2, and 12p11.23) and three high-density lipoprotein loci from the meta-analysis of all FBPP subgroups (on chromosomes 1p12, 14q11.2, and 21q21.1). These loci lacked significant genome-wide linkage or association evidence in the literature and had logarithm of odds (LOD) score ≥ 3 in the meta-analysis with LOD ≥ 1 in at least two network and race subgroups (exclusively of non-European descent).

CONCLUSION

Incorporating gene-age interactions into the analysis of lipids using multi-ethnic cohorts can enhance gene discovery. These interaction loci can guide the selection of families for sequencing studies of lipid-associated variants.

Genetic and environmental determinants of the susceptibility of Amerindian derived populations for having hypertriglyceridemia

Here, we discuss potential explanations for the higher prevalence of hypertriglyceridemia in populations with an Amerindian background. Although environmental factors are the triggers, the search for the ethnic related factors that explain the increased susceptibility of the Amerindians is a promising area for research. The study of the genetics of hypertriglyceridemia in Hispanic populations faces several challenges. Ethnicity could be a major confounding variable to prove genetic associations. Despite that, the study of hypertriglyceridemia in Hispanics has resulted in significant contributions. Two GWAS reports have exclusively included Mexican mestizos. Fifty percent of the associations reported in Caucasians could be generalized to the Mexicans, but in many cases the Mexican lead SNP was different than that reported in Europeans. Both reports included new associations with apo B or triglycerides concentrations. The frequency of susceptibility alleles in Mexicans is higher than that found in Europeans for several of the genes with the greatest effect on triglycerides levels. An example is the SNP rs964184 in APOA5. The same trend was observed for ANGPTL3 and TIMD4 variants. In summary, we postulate that the study of the genetic determinants of hypertriglyceridemia in Amerindian populations which have major changes in their lifestyle, may prove to be a great resource to identify new genes and pathways associated with hypertriglyceridemia.

Emerging directions in the genetics of atrial fibrillation

Atrial fibrillation (AF) is the most common arrhythmia and is associated with increased morbidity. As the population ages and the prevalence of AF continues to rise, the socioeconomic consequences of AF will become increasingly burdensome. Although there are well-defined clinical risk factors for AF, a significant heritable component is also recognized. To identify the molecular basis for the heritability of AF, investigators have used a combination of classical Mendelian genetics, candidate gene screening, and genome-wide association studies. However, these avenues have, as yet, failed to define the majority of the heritability of AF. The goal of this review is to describe the results from both candidate gene and genome-wide studies, as well as to outline potential future avenues for creating a more complete understanding of AF genetics. Ultimately, a more comprehensive view of the genetic underpinnings for AF will lead to the identification of novel molecular pathways and improved risk prediction of this complex arrhythmia.

The role of FADS1/2 polymorphisms on cardiometabolic markers and fatty acid profiles in young adults consuming fish oil supplements

Eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) are omega-3 (n-3) fatty acids (FAs) known to influence cardiometabolic markers of health. Evidence suggests that single nucleotide polymorphisms (SNPs) in the fatty acid desaturase 1 and 2 (FADS1/2) gene cluster may influence an individual's response to n-3 FAs. This study examined the impact of a moderate daily dose of EPA and DHA fish oil supplements on cardiometabolic markers, FA levels in serum and red blood cells (RBC), and whether these endpoints were influenced by SNPs in FADS1/2. Young adults consumed fish oil supplements (1.8 g total EPA/DHA per day) for 12 weeks followed by an 8-week washout period. Serum and RBC FA profiles were analyzed every two weeks by gas chromatography. Two SNPs were genotyped: rs174537 in FADS1 and rs174576 in FADS2. Participants had significantly reduced levels of blood triglycerides (-13%) and glucose (-11%) by week 12; however, these benefits were lost during the washout period. EPA and DHA levels increased significantly in serum (+250% and +51%, respectively) and RBCs (+132% and +18%, respectively) within the first two weeks of supplementation and remained elevated throughout the 12-week period. EPA and DHA levels in RBCs only (not serum) remained significantly elevated (+37% and +24%, respectively) after the washout period. Minor allele carriers for both SNPs experienced greater increases in RBC EPA levels during supplementation; suggesting that genetic variation at this locus can influence an individual's response to fish oil supplements.

Sex-specific association of the peptidase D gene rs731839 polymorphism and serum lipid levels in the Mulao and Han populations

Little is known about the association of peptidase D (PEPD) gene rs731839 single nucleotide polymorphism (SNP) and serum lipid profiles in the Chinese population. The objective of the present study was to detect the association of the PEPD rs731839 SNP and serum lipid levels in the Mulao and Han populations. Genotyping of the PEPD rs731839 SNP was performed in 751 subjects of Mulao and 762 subjects of Han using polymerase chain reaction and restriction fragment length polymorphism and then confirmed by direct sequencing. The A allele carriers had higher serum high-density lipoprotein cholesterol (HDL-C), apolipoprotein (Apo) AI levels and lower triglyceride (TG) levels in Mulao; and higher HDL-C, low-density lipoprotein cholesterol (LDL-C) and ApoAI levels in Han than the A allele non-carriers. Subgroup analyses showed that the A allele carriers had higher HDL-C, ApoAI levels and lower TG levels in Mulao males but not in females; higher total cholesterol (TC), HDL-C, LDL-C and ApoAI levels in Han males; and higher TG, HDL-C and ApoAI levels in Han females than the A allele non-carriers. Serum lipid parameters were also correlated with several environmental factors in Mulao and Han populations, or in males and females in both ethnic groups. The association of the PEPD rs731839 SNP and serum lipid levels was different between the Mulao and Han populations, and between males and females in the both ethnic groups. There may be an ethnic- and/or sex-specific association of the PEPD rs731839 SNP and serum lipid levels in our study populations.

Genetics of lipid traits: Genome-wide approaches yield new biology and clues to causality in coronary artery disease

A wealth of novel lipid loci have been identified through a variety of approaches focused on common and low-frequency variation and collaborative metaanalyses in multiethnic populations. Despite progress in identification of loci, the task of determining causal variants remains challenging. This work will undoubtedly be enhanced by improved understanding of regulatory DNA at a genomewide level as well as new methodologies for interrogating the relationships between noncoding SNPs and regulatory regions. Equally challenging is the identification of causal genes at novel loci. Some progress has been made for a handful of genes and comprehensive testing of candidate genes using multiple model systems is underway. Additional insights will be gleaned from focusing on low frequency and rare coding variation at candidate loci in large populations. This article is part of a Special Issue entitled: From Genome to Function.

Genetic determinants of long-term changes in blood lipid concentrations: 10-year follow-up of the GLACIER study

Recent genome-wide meta-analyses identified 157 loci associated with cross-sectional lipid traits. Here we tested whether these loci associate (singly and in trait-specific genetic risk scores [GRS]) with longitudinal changes in total cholesterol (TC) and triglyceride (TG) levels in a population-based prospective cohort from Northern Sweden (the GLACIER Study). We sought replication in a southern Swedish cohort (the MDC Study; N = 2,943). GLACIER Study participants (N = 6,064) were genotyped with the MetaboChip array. Up to 3,495 participants had 10-yr follow-up data available in the GLACIER Study. The TC- and TG-specific GRSs were strongly associated with change in lipid levels (β = 0.02 mmol/l per effect allele per decade follow-up, P = 2.0×10⁻¹¹ for TC; β = 0.02 mmol/l per effect allele per decade follow-up, P = 5.0×10⁻⁵ for TG). In individual SNP analysis, one TC locus, apolipoprotein E (APOE) rs4420638 (β = 0.12 mmol/l per effect allele per decade follow-up, P = 2.0×10⁻⁵), and two TG loci, tribbles pseudokinase 1 (TRIB1) rs2954029 (β = 0.09 mmol/l per effect allele per decade follow-up, P = 5.1×10⁻⁴) and apolipoprotein A-I (APOA1) rs6589564 (β = 0.31 mmol/l per effect allele per decade follow-up, P = 1.4×10⁻⁸), remained significantly associated with longitudinal changes for the respective traits after correction for multiple testing. An additional 12 loci were nominally associated with TC or TG changes. In replication analyses, the APOE rs4420638, TRIB1 rs2954029, and APOA1 rs6589564 associations were confirmed (P≤0.001). In summary, trait-specific GRSs are robustly associated with 10-yr changes in lipid levels and three individual SNPs were strongly associated with 10-yr changes in lipid levels.

Although large cross-sectional studies have proven highly successful in identifying gene variants related to lipid levels and other cardiometabolic traits, very few examples of well-designed longitudinal studies exist where associations between genotypes and long-term changes in lipids have been assessed. Here we undertook analyses in the GLACIER Study to determine whether the 157 previously identified lipid-associated genes variants associate with changes in blood lipid levels over 10-yr follow-up. We identified a variant in APOE that is robustly associated with total cholesterol change and two variants in TRIB1 and APOA1 respectively that are robustly associated with triglyceride change. We replicated these findings in a second Swedish cohort (the MDC Study). The identified genes had previously been associated with cardiovascular traits such as myocardial infarction or coronary heart disease; hence, these novel lipid associations provide additional insight into the pathogenesis of atherosclerotic heart and large vessel disease. By incorporating all 157 established variants into gene scores, we also observed strong associations with 10-yr lipid changes, illustrating the polygenic nature of blood lipid deterioration.

Epigenome-wide association study of fasting blood lipids in the Genetics of Lipid-lowering Drugs and Diet Network study

BACKGROUND

Genetic research regarding blood lipids has largely focused on DNA sequence variation; few studies have explored epigenetic effects. Genome-wide surveys of DNA methylation may uncover epigenetic factors influencing lipid metabolism.

METHODS AND RESULTS

To identify whether differential methylation of cytosine-(phosphate)-guanine dinucleotides (CpGs) correlated with lipid phenotypes, we isolated DNA from CD4+ T cells and quantified the proportion of sample methylation at >450 000 CpGs by using the Illumina Infinium HumanMethylation450 Beadchip in 991 participants of the Genetics of Lipid Lowering Drugs and Diet Network. We modeled the percentage of methylation at individual CpGs as a function of fasting very-low-density lipoprotein cholesterol and triglycerides (TGs) by using mixed linear regression adjusted for age, sex, study site, cell purity, and family structure. Four CpGs (cg00574958, cg17058475, cg01082498, and cg09737197) in intron 1 of carnitine palmitoyltransferase 1A (CPT1A) were strongly associated with very-low low-density lipoprotein cholesterol (P=1.8×10(-21) to 1.6×10(-8)) and TG (P=1.6×10(-26) to 1.5×10(-9)). Array findings were validated by bisulfite sequencing. We performed quantitative polymerase chain reaction experiments demonstrating that methylation of the top CpG (cg00574958) was correlated with CPT1A expression. The association of cg00574958 with TG and CPT1A expression were replicated in the Framingham Heart Study (P=4.1×10(-14) and 3.1×10(-13), respectively). DNA methylation at CPT1A cg00574958 explained 11.6% and 5.5% of the variation in TG in the discovery and replication cohorts, respectively.

CONCLUSIONS

This genome-wide epigenomic study identified CPT1A methylation as strongly and robustly associated with fasting very-low low-density lipoprotein cholesterol and TG. Identifying novel epigenetic contributions to lipid traits may inform future efforts to identify new treatment targets and biomarkers of disease risk.

Amerindian-specific regions under positive selection harbour new lipid variants in Latinos

Dyslipidemia and obesity are especially prevalent in populations with Amerindian backgrounds, such as Mexican-Americans, which predispose these populations to cardiovascular disease. Here we design an approach, known as the cross-population allele screen (CPAS), which we conduct prior to a genome-wide association study (GWAS) in 19,273 Europeans and Mexicans, in order to identify Amerindian risk genes in Mexicans. Utilizing CPAS to restrict the GWAS input variants to only those differing in frequency between the two populations, we identify novel Amerindian lipid genes, receptor-related orphan receptor alpha (RORA) and salt-inducible kinase 3 (SIK3), and three loci previously unassociated with dyslipidemia or obesity. We also detect lipoprotein lipase (LPL) and apolipoprotein A5 (APOA5) harbouring specific Amerindian signatures of risk variants and haplotypes. Notably, we observe that SIK3 and one novel lipid locus underwent positive selection in Mexicans. Furthermore, after a high-fat meal, the SIK3 risk variant carriers display high triglyceride levels. These findings suggest that Amerindian-specific genetic architecture leads to a higher incidence of dyslipidemia and obesity in modern Mexicans.

Whole-genome analyses of whole-brain data: working within an expanded search space

Large-scale comparisons of patients and healthy controls have unearthed genetic risk factors associated with a range of neurological and psychiatric illnesses. Meanwhile, brain imaging studies are increasing in size and scope, revealing disease and genetic effects on brain structure and function, and implicating neural pathways and causal mechanisms. With the advent of global neuroimaging consortia, imaging studies are now well powered to discover genetic variants that reliably affect the brain. Genetic analyses of brain measures from tens of thousands of people are being extended to test genetic associations with signals at millions of locations in the brain, and connectome-wide, genome-wide scans can jointly screen brain circuits and genomes; these analyses and others present new statistical challenges. There is a growing need for the community to establish and enforce standards in this developing field to ensure robust findings. Here we discuss how neuroimagers and geneticists have formed alliances to discover how genetic factors affect the brain. The field is rapidly advancing with ultra-high-resolution imaging and whole-genome sequencing. We recommend a rigorous approach to neuroimaging genomics that capitalizes on its recent successes and ensures the reliability of future discoveries.

The epigenetic drug 5-azacytidine interferes with cholesterol and lipid metabolism

DNA methylation and histone acetylation inhibitors are widely used to study the role of epigenetic marks in the regulation of gene expression. In addition, several of these molecules are being tested in clinical trials or already in use in the clinic. Antimetabolites, such as the DNA-hypomethylating agent 5-azacytidine (5-AzaC), have been shown to lower malignant progression to acute myeloid leukemia and to prolong survival in patients with myelodysplastic syndromes. Here we examined the effects of DNA methylation inhibitors on the expression of lipid biosynthetic and uptake genes. Our data demonstrate that, independently of DNA methylation, 5-AzaC selectively and very potently reduces expression of key genes involved in cholesterol and lipid metabolism (e.g. PCSK9, HMGCR, and FASN) in all tested cell lines and in vivo in mouse liver. Treatment with 5-AzaC disturbed subcellular cholesterol homeostasis, thereby impeding activation of sterol regulatory element-binding proteins (key regulators of lipid metabolism). Through inhibition of UMP synthase, 5-AzaC also strongly induced expression of 1-acylglycerol-3-phosphate O-acyltransferase 9 (AGPAT9) and promoted triacylglycerol synthesis and cytosolic lipid droplet formation. Remarkably, complete reversal was obtained by the co-addition of either UMP or cytidine. Therefore, this study provides the first evidence that inhibition of the de novo pyrimidine synthesis by 5-AzaC disturbs cholesterol and lipid homeostasis, probably through the glycerolipid biosynthesis pathway, which may contribute mechanistically to its beneficial cytostatic properties.

Selectivity in genetic association with sub-classified migraine in women

Migraine can be sub-classified not only according to presence of migraine aura (MA) or absence of migraine aura (MO), but also by additional features accompanying migraine attacks, e.g. photophobia, phonophobia, nausea, etc. all of which are formally recognized by the International Classification of Headache Disorders. It remains unclear how aura status and the other migraine features may be related to underlying migraine pathophysiology. Recent genome-wide association studies (GWAS) have identified 12 independent loci at which single nucleotide polymorphisms (SNPs) are associated with migraine. Using a likelihood framework, we explored the selective association of these SNPs with migraine, sub-classified according to aura status and the other features in a large population-based cohort of women including 3,003 active migraineurs and 18,108 free of migraine. Five loci met stringent significance for association with migraine, among which four were selective for sub-classified migraine, including rs11172113 (LRP1) for MO. The number of loci associated with migraine increased to 11 at suggestive significance thresholds, including five additional selective associations for MO but none for MA. No two SNPs showed similar patterns of selective association with migraine characteristics. At one extreme, SNPs rs6790925 (near TGFBR2) and rs2274316 (MEF2D) were not associated with migraine overall, MA, or MO but were selective for migraine sub-classified by the presence of one or more of the additional migraine features. In contrast, SNP rs7577262 (TRPM8) was associated with migraine overall and showed little or no selectivity for any of the migraine characteristics. The results emphasize the multivalent nature of migraine pathophysiology and suggest that a complete understanding of the genetic influence on migraine may benefit from analyses that stratify migraine according to both aura status and the additional diagnostic features used for clinical characterization of migraine.

Migraine is among the most common and debilitating neurological disorders. Diagnostic criteria for migraine recognize a variety of symptoms including a primary dichotomous classification for the presence or absence of aura, typically a visual disturbance phenomenon, as well as others such as sensitivity to light or sound, and nausea, etc. We explored whether any of 12 recently discovered genetic variants associated with common migraine might have selective association for migraine sub-classified by aura status or nine additional migraine features in a population of middle-aged women including 3,003 migraineurs and 18,180 non-migraineurs. Five of the 12 genetic variants met the most stringent significance criterion for association with migraine, among which four had selective association with sub-classified migraine, including one that was selective for migraine without aura. At suggestive significance, all of the remaining genetic variants were selective for sub-classifications of migraine although no two variants showed the same pattern of selectivity. The selectivity patterns suggest very different contributions to migraine pathophysiology among the 12 loci and their implicated genes. Further, the results suggest that future discovery efforts for new migraine susceptibility loci would benefit by considering associations with sub-classified migraine toward the ultimate goals of more specific diagnosis and personalized treatment.

Bayesian test for colocalisation between pairs of genetic association studies using summary statistics

Genetic association studies, in particular the genome-wide association study (GWAS) design, have provided a wealth of novel insights into the aetiology of a wide range of human diseases and traits, in particular cardiovascular diseases and lipid biomarkers. The next challenge consists of understanding the molecular basis of these associations. The integration of multiple association datasets, including gene expression datasets, can contribute to this goal. We have developed a novel statistical methodology to assess whether two association signals are consistent with a shared causal variant. An application is the integration of disease scans with expression quantitative trait locus (eQTL) studies, but any pair of GWAS datasets can be integrated in this framework. We demonstrate the value of the approach by re-analysing a gene expression dataset in 966 liver samples with a published meta-analysis of lipid traits including >100,000 individuals of European ancestry. Combining all lipid biomarkers, our re-analysis supported 26 out of 38 reported colocalisation results with eQTLs and identified 14 new colocalisation results, hence highlighting the value of a formal statistical test. In three cases of reported eQTL-lipid pairs (SYPL2, IFT172, TBKBP1) for which our analysis suggests that the eQTL pattern is not consistent with the lipid association, we identify alternative colocalisation results with SORT1, GCKR, and KPNB1, indicating that these genes are more likely to be causal in these genomic intervals. A key feature of the method is the ability to derive the output statistics from single SNP summary statistics, hence making it possible to perform systematic meta-analysis type comparisons across multiple GWAS datasets (implemented online at http://coloc.cs.ucl.ac.uk/coloc/). Our methodology provides information about candidate causal genes in associated intervals and has direct implications for the understanding of complex diseases as well as the design of drugs to target disease pathways.

Mining the genome for lipid genes

Mining of the genome for lipid genes has since the early 1970s helped to shape our understanding of how triglycerides are packaged (in chylomicrons), repackaged (in very low density lipoproteins; VLDL), and hydrolyzed, and also how remnant and low-density lipoproteins (LDL) are cleared from the circulation. Gene discoveries have also provided insights into high-density lipoprotein (HDL) biogenesis and remodeling. Interestingly, at least half of these key molecular genetic studies were initiated with the benefit of prior knowledge of relevant proteins. In addition, multiple important findings originated from studies in mouse, and from other types of non-genetic approaches. Although it appears by now that the main lipid pathways have been uncovered, and that only modulators or adaptor proteins such as those encoded by LDLRAP1, APOA5, ANGPLT3/4, and PCSK9 are currently being discovered, genome wide association studies (GWAS) in particular have implicated many new loci based on statistical analyses; these may prove to have equally large impacts on lipoprotein traits as gene products that are already known. On the other hand, since 2004 - and particularly since 2010 when massively parallel sequencing has become de rigeur - no major new insights into genes governing lipid metabolism have been reported. This is probably because the etiologies of true Mendelian lipid disorders with overt clinical complications have been largely resolved. In the meantime, it has become clear that proving the importance of new candidate genes is challenging. This could be due to very low frequencies of large impact variants in the population. It must further be emphasized that functional genetic studies, while necessary, are often difficult to accomplish, making it hazardous to upgrade a variant that is simply associated to being definitively causative. Also, it is clear that applying a monogenic approach to dissect complex lipid traits that are mostly of polygenic origin is the wrong way to proceed. The hope is that large-scale data acquisition combined with sophisticated computerized analyses will help to prioritize and select the most promising candidate genes for future research. We suggest that at this point in time, investment in sequence technology driven candidate gene discovery could be recalibrated by refocusing efforts on direct functional analysis of the genes that have already been discovered. This article is part of a Special Issue entitled: From Genome to Function.

Systems genetics: From GWAS to disease pathways

Most common diseases are complex, involving multiple genetic and environmental factors and their interactions. In the past decade, genome-wide association studies (GWAS) have successfully identified thousands of genetic variants underlying susceptibility to complex diseases. However, the results from these studies often do not provide evidence on how the variants affect downstream pathways and lead to the disease. Therefore, in the post-GWAS era the greatest challenge lies in combining GWAS findings with additional molecular data to functionally characterize the associations. The advances in various ~omics techniques have made it possible to investigate the effect of risk variants on intermediate molecular levels, such as gene expression, methylation, protein abundance or metabolite levels. As disease aetiology is complex, no single molecular analysis is expected to fully unravel the disease mechanism. Multiple molecular levels can interact and also show plasticity in different physiological conditions, cell types and disease stages. There is therefore a great need for new integrative approaches that can combine data from different molecular levels and can help construct the causal inference from genotype to phenotype. Systems genetics is such an approach; it is used to study genetic effects within the larger scope of systems biology by integrating genotype information with various ~omics datasets as well as with environmental and physiological variables. In this review, we describe this approach and discuss how it can help us unravel the molecular mechanisms through which genetic variation causes disease. This article is part of a Special Issue entitled: From Genome to Function.

High-density lipoproteins in the prevention of cardiovascular disease: changing the paradigm

High-density-lipoprotein cholesterol (HDL-C) has been identified in population studies as an independent inverse predictor of cardiovascular events. Although the causal nature of this association has been questioned, HDL and its major protein, apolipoprotein (apo)A1, have been shown to prevent and reverse atherosclerosis in animal models. In addition, HDL and apoA1 have several putatively atheroprotective functions, such as the ability to promote efflux of cholesterol from macrophages in the artery wall, inhibit vascular inflammation, and enhance endothelial function. Therefore, HDL-C and apoA1 have been investigated as therapeutic targets for coronary heart disease. However, recent clinical trials with drugs that raise HDL-C, such as niacin and inhibitors of cholesteryl ester transfer protein, have been disappointing. Here, we review the current state of the science regarding HDL as a therapeutic target.

Reading and language disorders: the importance of both quantity and quality

Reading and language disorders are common childhood conditions that often co-occur with each other and with other neurodevelopmental impairments. There is strong evidence that disorders, such as dyslexia and Specific Language Impairment (SLI), have a genetic basis, but we expect the contributing genetic factors to be complex in nature. To date, only a few genes have been implicated in these traits. Their functional characterization has provided novel insight into the biology of neurodevelopmental disorders. However, the lack of biological markers and clear diagnostic criteria have prevented the collection of the large sample sizes required for well-powered genome-wide screens. One of the main challenges of the field will be to combine careful clinical assessment with high throughput genetic technologies within multidisciplinary collaborations.

Joint analysis of functional genomic data and genome-wide association studies of 18 human traits

Annotations of gene structures and regulatory elements can inform genome-wide association studies (GWASs). However, choosing the relevant annotations for interpreting an association study of a given trait remains challenging. I describe a statistical model that uses association statistics computed across the genome to identify classes of genomic elements that are enriched with or depleted of loci influencing a trait. The model naturally incorporates multiple types of annotations. I applied the model to GWASs of 18 human traits, including red blood cell traits, platelet traits, glucose levels, lipid levels, height, body mass index, and Crohn disease. For each trait, I used the model to evaluate the relevance of 450 different genomic annotations, including protein-coding genes, enhancers, and DNase-I hypersensitive sites in over 100 tissues and cell lines. The fraction of phenotype-associated SNPs influencing protein sequence ranged from around 2% (for platelet volume) up to around 20% (for low-density lipoprotein cholesterol), repressed chromatin was significantly depleted for SNPs associated with several traits, and cell-type-specific DNase-I hypersensitive sites were enriched with SNPs associated with several traits (for example, the spleen in platelet volume). Finally, reweighting each GWAS by using information from functional genomics increased the number of loci with high-confidence associations by around 5%.

Systematic evaluation of coding variation identifies a candidate causal variant in TM6SF2 influencing total cholesterol and myocardial infarction risk

Blood lipid levels are heritable, treatable risk factors for cardiovascular disease. We systematically assessed genome-wide coding variation to identify new genes influencing lipid traits, fine map known lipid loci and evaluate whether low-frequency variants with large effects exist for these traits. Using an exome array, we genotyped 80,137 coding variants in 5,643 Norwegians. We followed up 18 variants in 4,666 Norwegians and identified ten loci with coding variants associated with a lipid trait (P < 5 × 10(-8)). One variant in TM6SF2 (encoding p.Glu167Lys), residing in a known genome-wide association study locus for lipid traits, influences total cholesterol levels and is associated with myocardial infarction. Transient TM6SF2 overexpression or knockdown of Tm6sf2 in mice alters serum lipid profiles, consistent with the association observed in humans, identifying TM6SF2 as a functional gene within a locus previously known as NCAN-CILP2-PBX4 or 19p13. This study demonstrates that systematic assessment of coding variation can quickly point to a candidate causal gene.

Epidemiological studies of CHD and the evolution of preventive cardiology

Cardiovascular diseases (CVDs) cause nearly one-third of all deaths worldwide. Coronary heart disease (CHD) accounts for the greatest proportion of CVDs, and risk factors such as hypertension, cigarette smoking, diabetes mellitus or elevated glucose level, elevated cholesterol levels, and obesity or being overweight are the top six causes of death globally. Ecological and population-based longitudinal studies, conducted globally or within individual countries, have established the role of traditional and novel risk factors and measures of subclinical disease in the prediction of CHD. Risk assessment with short-term or long-term risk prediction algorithms can help to identify individuals who would benefit most from risk-factor interventions. Evaluation of novel risk factors and screening for subclinical atherosclerosis can also help to identify individuals at highest cardiovascular risk. Prevention of CHD focuses on identifying and managing risk factors at both the population and individual levels through primordial, primary, and secondary prevention. Epidemiological studies have provided the hypotheses for subsequent clinical trials that have documented the efficacy of risk-factor interventions, which are the basis of preventive cardiology. Future research efforts will determine the screening and intervention strategies that have the greatest effect on CHD prevention.

CELSR2-PSRC1-SORT1 gene expression and association with coronary artery disease and plasma lipid levels in an Asian Indian cohort

BACKGROUND

Genetic regulation of plasma lipids has been shown to influence the risk of coronary artery disease (CAD). We analyzed the relationship between rs599839 and rs646776 single nucleotide polymorphisms (SNPs) present in the CELSR2-PSRC1-SORT1 gene cluster, candidate gene expression, and their association with CAD and circulating lipid levels in a representative cohort of Asian Indians selected from the Indian Atherosclerosis Research Study.

METHODS

SNPs rs599839 and rs646776 were genotyped by Taqman assay in 1034 CAD patients (cases) and 1034 age- and gender-matched controls. Expression of CELSR2, PSRC1, and SORT1 genes was measured in 100 cases and 100 controls. Plasma levels of total cholesterol (TC), triglycerides, high-density lipoprotein-cholesterol, and low-density lipoprotein-cholesterol (LDL-c) were measured by enzymatic assay.

RESULTS

Both rs646776 and rs599839 were in strong linkage disequilibrium (r = 0.98) and showed significant protective association with CAD (OR = 0.315, 95% CI 0.136-0.728, p<0.007 and OR = 0.422, 95% CI 0.181-0.981, p = 0.045, respectively). Haplotype TA showed 72% frequency and was associated with CAD (OR 0.77, 95% CI 0.67-0.88, p = 0.0002). PSRC1 gene expression was lower in the cases than in the controls (0.75 ± 0.405 versus 1.04 ± 0.622, p = 2.26 × 10(-4)). The homozygous variant and heterozygous genotypes showed 30% and 15% higher PSRC1 expression, respectively. Correspondingly, the minor alleles were associated with lower plasma TC and LDL-c levels.

CONCLUSION

PSRC1 in the cholesterol gene cluster shows a significant association with CAD by virtue of the two SNPs, rs646776 and rs599839 that also regulate plasma cholesterol levels.