GENOMICS OF COMPLEX CARDIOVASCULAR DISEASE
The genomics revolution, including the complete sequence of the human genome, has greatly advanced our understanding of the molecular basis of Mendelian disorders, which are largely the result of necessary and sufficient mutations that are individually rare. As fruitful as genome-targeted efforts have been in the field of single-gene disorders, progress in understanding the genetic architecture of complex traits has been slower and more tenuous. These traits are governed by the interplay between genes, epigenetic factors, and the environment. In this scenario, complex traits are likely to be modulated by multiple genes (and possibly by multiple genetic variants within a gene), which are individually neither necessary nor sufficient to determine a trait. In the field of cardiovascular genetics, the multitude of different outcomes assessed, often using slightly different definitions of the phenotype in question, add to the complexity. Thus, reproducibly identifying genetic variants that impact these traits was a daunting task until recently, when the popularization of microarray technology made it possible to survey large cohorts at a large number of loci in order to scan the genome for variants associated with disease. Genetic studies of cardiovascular disease (CVD) in particular are largely dichotomized into studies of genes that influence CVD risk factors, such as lipids and blood pressure, and studies that focus on clinical outcomes. In this chapter, we summarize the progress made in the search for genes affecting cardiovascular disease risk both by candidate-gene studies and genome-wide association studies (GWAS), while focusing on three specific CVD events (myocardial infarction, stroke, and sudden cardiac death), with the idea that genetic variants that directly affect clinical outcomes would not necessarily be identified only through studies of CVD risk factors, and often require a direct interrogation of the clinical outcome of interest, allowing for more accurate risk stratification. This is particularly important in sudden cardiac death, where two-thirds of victims do not have CVD risk profiles that would warrant intervention under current guidelines. The results of this search for genetic variants of interest are summarized in Table 21.1.
Table 21.1 COMMON GENETIC VARIANTS THAT INFLUENCE SUSCEPTIBILITY TO CARDIOVASCULAR EVENTS
*QRS interval is included as a subclinical phenotype for SCD, due to the limited direct studies of SCD
MYOCARDIAL INFARCTION (MI) AND CORONARY HEART DISEASE (CHD)
Myocardial infarction (MI) is an often fatal manifestation of coronary artery disease (CAD), with an annual incidence of approximately 935,000 cases in the United States, 172,000 of whom will die from the disease (Roger et al., 2012). The proximal cause of MI is believed to be a thrombosis event triggered by an atherosclerotic plaque rupture, which occludes a coronary artery and leads to necrosis of the myocardium. In a long-term follow-up study of ~21,000 monozygous (MZ) and dizygous (DZ) twins, heritability of fatal coronary events was 57% and 38%, respectively, for men and women (Zdravkovic et al., 2002), underlining the importance of genetic factors in susceptibility to MI. Several clinically relevant risk factors have been identified, many of which display significant heritability and thus are, at least in part, under genetic influence (Table 21.2). Indeed, rare Mendelian mutations that affect these risk factors can lead to premature coronary artery disease (Table 21.3) and have been signposts of important pathways that have directly led to successful drug therapies, including the cholesterol metabolism pathway that emerged from studies of familial hypercholsterolemia (FH).
Table 21.2 Risk FACTORS FOR MYOCARDIAL INFARCTION*
|Risk Factors with a Significant Genetic Component (Heritability) Total cholesterol (40–60%)
Total triglycerides (40–80%) Body mass index (25–60%) Blood pressure (50–70%)
Lp(a) levels (90%)
Homocysteine levels (45%)
Type 2 diabetes (40-80%)
C-reactive protein (20–50%) Gender
Environmental risk factors: Smoking
Diet Exercise Infection
Table 21.3 MENDELIAN DISEASES THAT EXHIBIT PREMATURE CORONARY ARTERY DISEASE*
|DISEASE||GENES||EFFECT OF MUTATIONS|
|Familial hypercholesterolemia||LDLR||Defective binding of LDL by receptor|
|Familial defective APOB||APOB||Reduced binding affinity of APOB to LDLR|
|Sitosterolemia||ABCG5, ABCG8||Increased absorption of plant sterols|
|Autosomal recessive hypercholesterolemia||ARH||Defective endocytosis of LDLR|
|APOA1 deficiency||APOA1||Deletion or loss-of-function mutations that lead to very low HDL|
|Tangier disease||ABCA1||Impaired cholesterol efflux in macrophages|
|Homocystinuria||CBS||Leads to increased thrombotic tendency|
The majority of research efforts prior to 2007 were focused on identifying genes that modulate common variation in traditional cardiovascular risk factors. A major breakthrough was made with the popularization of GWAS and the ability to genotype a large patient population for a million single- nucleotide polymorphisms (SNPs) or more. The efforts to identify common genetic variants associated with CAD and MI have produced a list of over 30 loci showing robust association that has been replicated across multiple studies and cohorts (Kathiresan et al., 2009; Samani et al., 2007; Schunkert et al., 2011).
We highlight the progress made in the search for genes associated with MI by both the candidate gene approach and the genome-wide association approach, by discussing four specific examples: apolipoprotein E (APOE), coagulation proteins, 9p21, and SORT1, in detail.
APOLIPOPROTEIN E (APOE)
Apolipoprotein E, a low-density lipoprotein (LDL) receptor ligand, is an important player in the metabolism of cholesterol and triglycerides, where it mediates the clearance of chylomicron and very low-density lipoprotein (VLDL) from plasma. Utermann and colleagues (Utermann et al., 1977) first described the effects of three common allelic variants of APOE (termed E2, E3, and E4, and with frequencies of 8%, 77%, and 15%, respectively, in Caucasian populations) on type III familial hyperlipoproteinemia, in which more than 95% of affected individuals were homozygous for the rare allele. The common genotype, E3/E3, is used as the reference group in most studies, and individuals who carry the E2 allele have ~14 mg/dl lower LDL levels, and E4 carriers have ~7 mg/dl higher LDL levels (Motulsky and Brunzell, 2002).
Numerous studies have examined the association between the E2/E3/E4 variants and coronary heart disease (CHD), including a large-scale meta- analysis incorporating 121 studies with 37,850 cases and 82,727 controls (Bennet et al., 2007). The meta-analysis, which was stratified by the number of participants in an individual study at a cutoff of at least 1,000 healthy controls and 500 cases, demonstrated a moderate increased risk for E4 carriers (odds ratio [OR] 1.06, 95% confidence interval [CI], 0.99–1.13), and a significant decreased risk for E2 carriers (OR 0.80, 95% CI 0.70–0.90) in the group with larger study size. In addition, the study also established a nearly linear relationship between APOE status and both LDL cholesterol levels and risk for CVD, putting forward the possibility that the differential risk associated with the different APOE isoforms may be mediated by LDL cholesterol, a well-established risk factor for CAD. Indeed, while most of the studies in the meta-analysis did not report odds ratios adjusted for lipids, and thus the meta-analysis did not determine whether the APOE variants influence risk for CHD independently of lipids, several studies that have examined the data for an association of the E4 allele after adjusting for lipid levels report conflicting results (Humphries et al., 2001; Lahoz et al., 2001; Volcik et al., 2006; Ward et al., 2009).
COAGULATION AND FIBRINOLYTIC GENES
The most common pathogenetic pathway of acute myocardial infarction is through thrombosis, generally triggered by atherosclerotic plaque rupture. Thus, a great number of candidate gene studies have involved the examination of genetic variants in genes involved in coagulation and fibrinolytic pathways. Ye and colleagues (Ye et al., 2006) performed a meta- analysis of 191 studies to determine the relationship between CAD and variants in seven genes involved in the thrombotic process: factor V Leiden, factor VII G10976A, prothrombin G20210A, plasminogen activator inhibitor-1 (PAI-1) [-675] 4G/5G, and three platelet glycoprotein (GP) variants (GPIa, C807T, GPIba T[-5]C, GPIIIa C1565T). In contrast to an earlier study (Boekholdt et al., 2001), which examined the association of several of the same variants with MI and found that associations for these genetic variants were either weak (PAI-1, fibrinogen) or absent (factor V, prothrombin), Ye and colleagues reported a mild association with factor V Leiden (OR 1.12, 95% CI 0.91–1.36) and prothrombin (OR 1.91, 95% CI 0.91–1.55). Despite this weak and conflicting association with MI, stronger, consistent associations are seen with stroke, and these variants are discussed in greater detail later in the chapter in that context.
Given the difficulty in identifying genetic variants that play a significant role in susceptibility to MI/IHD through a candidate gene approach, a problem observed for most complex diseases, family-based strategies have also been implemented. While many regions have been implicated through linkage analysis, these studies often result in large candidate regions, often comprising hundreds of genes. Therefore, their utility in identifying specific gene variants associated with disease is limited, though some successes have been reported, for instance in the association of genes ALOX5AP and PDE4D. with stroke (details in next section). With both a rapid reduction in genotyping costs and a vast increase in throughput, the focus has turned to genome-wide association studies.
Cardiovascular disease has proven to be an excellent example to illustrate the power of genome-wide association studies to find loci that contribute to common complex disorders. The high incidence of CVD, combined with the ready availability of large cohorts with detailed data on traditional cardiovascular risk factors, has led to several large association studies that have brought the list of loci associated with either CVD or MI to more than
- A majority of these loci come from three large multi-cohort Samani and colleagues used data from a combined 2801 cases and 4582 controls, replicated the association of the 9p21 region, and identified six additional loci associated with coronary artery disease (Samani et al., 2007). Four of these loci were also shown to be associated with myocardial infarction in a large-scale four-stage study by Kathiresan and colleagues (Kathiresan et al., 2009), who also reported three additional novel loci. In the third and largest association study for CAD, Schunkert and colleagues (Schunkert et al., 2011) carried out a meta-analysis of 14 studies to obtain a sample size of 22,233 cases and 64,762 controls of European descent. In addition to confirming the 10 previously associated loci, the study also identified 13 novel loci associated with CAD, only three of which showed any significant association with traditional CAD risk factors. From this long list of loci (see Table 21.4), we will focus on 9p21 and SORT1 as specific examples of GWAS discoveries resulting in interesting insights about biology and disease.
Table 21.4 CANDIDATE LOCI FROM GWAS OF EITHER MYOCARDIAL INFARCTION OR CORONARY ARTERY DISEASE
*Adapted from Kathiresan and Srivastava, 2012.
One of the most strongly associated signals from GWAS has been in the noncoding region on chromosome 9p21.3. After its initial discovery in a cohort of early-onset MI patients from the Ottawa Heart Study and deCode project (Helgadottir et al., 2007; McPherson et al., 2007), the region was fine- mapped to a 58kb locus with multiple tagged SNPs in tight linkage disequilibrium (LD), and it has shown consistent association with disease independent of traditional CAD risk factors, in multiple cohorts of various ethnic backgrounds (Cheng et al., 2011; Gori et al., 2010; IBC 50K CAD Consortium, 2011; Shen et al., 2008; Takeuchi et al., 2012; Xie et al., 2011). In a meta-analysis of 47 studies comprising 35,872 cases and 95,837 controls of either European or Asian descent, Palomaki and colleagues surveyed data from one of three common lead SNPs in the region (rs1333049, rs10757274, rs2382207) that are in tight LD, to confirm that the region had a small but significant effect on risk for MI (OR 1.25, 95% CI 1.21–1.29) (Palomaki et al., 2010). It is important to mention that while these results have been replicated in a number of European and Asian cohorts, the results from studies including African-Americans have been more conflicting (Beckie et al., 2011; Kral et al., 2011; Patel et al., 2011; Yamagishi et al., 2009). Although the risk locus is devoid of a protein-coding gene, it lies within a well-described noncoding RNA in the INK locus (ANRIL) (Pasmant et al., 2007) and overlaps with upstream cyclin-dependent kinase inhibitor genes CDKN2B and CDKN2A, which have been studied for their role as cell cycle regulators and tumor suppressors. While the genes lie in the same LD block, they are also approximately 100 kb upstream of the implicated risk region, and the direct evidence for their putative role in the modulation of cardiovascular disease did not emerge until Axel and colleagues developed a mouse model with a homozygous deletion of the orthologous region on the mouse chromosome 4, which showed reduced levels of cardiac expression of both genes (Visel et al., 2010). There have been numerous studies that have looked at the direct effect of expression of the CDKN2A, CDKN2B, and ANRIL (Cunnington et al., 2010; Folkersen et al., 2009; Harismendy et al., 2011; Jarinova et al., 2009; Liu et al., 2009) with conflicting results on the direction and magnitude of the effect of region in the development of CAD.
To add to the confusion, 9p21 has turned out to be somewhat of a hot spot for GWAS hits for a number of conditions, including type 2 diabetes (T2D) (Diabetes Genetics Initiative of Broad Institute of Harvard and MIT, Lund University, and Novartis Institutes of BioMedical Research et al., 2007; Scott et al., 2007; Zeggini et al., 2007; Zeggini et al., 2008), intracranial aneurysm and abdominal aortic aneurysms (AAA) (Helgadottir et al., 2008), and a number of different cancers (Amos et al., 2011; Antoniou et al., 2012; Enciso-Mora et al., 2012; Rajaraman et al., 2012; Sherborne et al., 2010; Shete et al., 2009; Yang et al., 2010). While the link to cancer can be explained by the presence of CDKN2A/B, two well-established tumor suppressor genes, the presence of a strong association to T2D, a traditional CAD risk factor, paved the ground for the tempting possibility of a shared biological function of the region that would affect both conditions. In an elegant study, Helgadottir and colleagues showed that the region was in fact that tagged by two separate lead SNPs (rs10757278 for CAD, and rs10811661 for T2D) which were in adjoining LD blocks (Helgadottir et al., 2008). While the SNP associated with CAD also showed association with five other arterial diseases, with the strongest association being AAA (OR = 1.31, 95% CI 1.22–1.42) and intracranial aneurysms (OR = 1.29, 95% CI 1.16–1.43), rs10811661, the T2D SNP, showed no significant association with either CAD or the other arterial diseases. This result has since been replicated in several studies (Biros et al., 2010; Bown et al., 2008; Gori et al., 2010) and highlights the complexity of the genomic region and makes the process of going from an associated locus to a causal gene even more challenging. While the mechanism by which the risk “region” affects atherosclerotic processes remains to be determined, the presence of a consistent association in multiple cohorts makes 9p21 a prominent area of research efforts.
The elucidation of the role of Sortilin1, a multi-ligand sorting receptor in cholesterol metabolism, highlights the power of GWAS findings when put in the context of appropriate functional studies.
Samani and colleagues first implicated the 1p31 region in a GWAS for CAD in 2007 (OR 1.29, 95% CI 1.10–1.21) (Samani et al., 2007), which has since been replicated in a number of large-scale GWAS for both MI and CAD (Kathiresan et al., 2009; Schunkert et al., 2011). The SNPs tagged in these studies lie in a noncoding region, between genes PSRC1 and CELSR1, and in the same LD block as SORT1, making identification of the causal variant and the gene mediating the association challenging.
SORT1 was first implicated as the putative gene of interest by Linsel- Nitschke and colleagues (Linsel-Nitschke et al., 2010). This work identified an expression quantitative trait locus (eQTL), uncovering an association between one of the GWAS SNPs, rs599839, and SORT1 mRNA levels. In addition, they demonstrated a significant increase in LDL cholesterol uptake in HEK293 cells with over-expression of SORT1 and laid the foundation for the hypothesis that increased SORT1 expression could potentially have a protective role in CAD. The role of SORT1 in LDL cholesterol metabolism was further investigated by Musunuru and colleagues in a series of experiments that showed the direct effect of Sort1 knockdown and over- expression on plasma LDL cholesterol and lipoprotein levels in murine models of atherosclerosis. They also presented evidence to show that a common polymorphism, rs12740374, previously shown to be associated with CAD, creates a novel CCAAT-enhancer-binding protein (C/EBP) binding site and hence mediates an increased LDL cholesterol level via a hepatic secretory pathway. However, in direct contrast, Kjolby et al. showed that in a Sort1, Ldlr double knockout mouse model, plasma LDL cholesterol levels, and thus the atherosclerotic burden, are reduced when compared to a LDLR- null mouse (Kjolby et al., 2010). Considering the difference in mouse manipulations and timing of over-expression, the differences are not altogether surprising, but they definitely raise questions about the role of SORT1 in the biology of cholesterol metabolism and its subsequent effect on disease processes.
The success of GWAS is demonstrated by the fact that, in addition to identifying genes that are implicated as putative candidates because of their biological relevance in the disease process, we are now also able to identify a long list of genes that show a strong, definite association with CVD through yet-unknown mechanisms. Indeed, in a recent review, Kathiresan and Srivastava (Kathiresan and Srivastava, 2012) divided the list of loci mapped by GWAS into those known to be associated with an established risk factor like LDL-cholesterol or blood pressure, and those where the association is established but the mechanism is unknown.
Herein lies the challenge before us: to improve our ability to identify the causal variant, and thereby the underlying mechanism of action. Over the next couple of years, we will need to make a strong effort to determine the complete list of associated loci, and more importantly, to use this list to direct research efforts to better understand the biology of normal cardiovascular processes and myocardial infarction. Aside from identifying novel molecular pathways, the importance of these genetic variants from a clinical standpoint is often hard to interpret, given the small effect size of most GWAS hits and the ability to easily measure traditional cardiovascular risk factors. The argument in favor of using variants that have shown repeated association in several studies for either risk prediction has been shaky at best, with different groups reporting conflicting levels of success in using genetic data for risk prediction (Drenos et al., 2007; Kathiresan et al., 2008; Kathiresan et al., 2009; Paynter et al., 2010). In a large literature-survey-based test of a “genetic score” combining the effects of 101 SNPs with traditional cardiovascular risk factors, Paynter and colleagues showed no significant improvement in risk determination or classification in a cohort of 19,213 women (Paynter et al., 2010). However, it is important to mention that while these data suggest that genetic variants, at this point in time, have little value in cross-sectional measures of risk, there is strong evidence to suggest that genetic data might provide a better measure of lifetime risk than individual cross-sectional measures of risk factors. This has been successfully shown by Cohen and colleagues, who observed significantly reduced levels of plasma LDL cholesterol (~15%) and incidence of cardiovascular disease in carriers of the Arg47Leu allele of PCSK9 (hazard ratio 0.50, 95% CI 0.32–0.79) (Cohen et al., 2006). While the expected reduction of the CVD risk corresponding to the decrease in LDL-c levels that they observed is ~23%, analysis of the Atherosclerosis Risk in Communities (ARIC) prospective cohort showed a 47% decrease in R47L carriers, suggesting that there is an accumulated lifetime burden of reduced LDL-c levels that is not accurately captured by a single measure of lipid levels.
Stroke is one of the leading causes of death and disability in the developed world, with annual incidence of 795,000 (Roger et al., 2012). With limited treatment options available, focus has been on primary prevention, largely through modification of acquired risk factors (diabetes mellitus, smoking, high blood pressure, and atrial fibrillation) (Goldstein et al., 2001). However, as with many common diseases, rare monogenic conditions that cause stroke have been identified, and a great deal of progress has been made in identifying the underlying genetic defects. Indeed, the gene for cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL), which has served as a model for inherited ischemic stroke, was identified as NOTCH3 in 1996 (Joutel et al., 1996). More recently, substantial progress has been made in cerebral cavernous malformations (CCM), with genes for all three types of CCM now identified. CCM1 is caused by mutations in KRIT (Laberge-le Couteulx et al., 1999); CCM2 is caused by mutations in MGC4607 (malcavernin) (Denier et al., 2004); and CCM3 is caused by Programmed cell-death protein 10 (Bergametti et al., 2005). While these genetic defects are rare in the general population, they have high penetrance; therefore, being able to identify carriers has a significant clinical impact. Whether any of these genes has a role in common forms of stroke remains to be determined, though a study by Dong and colleagues, in which they screened individuals with lacunar stroke for coding mutations in NOTCH3, did not find any association (Dong et al., 2003). This does not, however, rule out the involvement of common variants, more likely to be found in noncoding conserved regions and involved in gene regulation.
While rare, cerebral venous thrombosis (CVT), which accounts for less than 1% of all strokes, is a complex disease, with numerous etiological risk factors (for review, see Agostoni et al., 2009), including genetic factors. Marjot and colleagues (Marjot et al., 2011) conducted a comprehensive meta- analysis of candidate gene studies for CVT, identifying 26 case-control studies covering six polymorphisms in six genes. With a sample size of 1183 CVT cases and 5189 controls, they demonstrated significant associations for three genes (discussed below), including factor V Leiden/G1691A (OR 2.40, 95% CI 1.75–3.30), prothrombin/G20210A (OR 5.48, 95% CI 3.88–7.74), and MTHFR/C677T (OR 2.30, 95% CI 1.20–4.42).
FACTOR V LEIDEN
Bertina and colleagues (Bertina et al., 1994) identified an arginine-to-glycine (R506Q) mutation (termed “Leiden allele”) in factor V in a family with activated protein C (APC) resistance and prone to thrombosis. APC limits clot formation by inactivation of factors Va and VIIIa, and the Leiden mutation is predicted to alter the amino acid at the APC cleavage site in factor V, causing factor V to be less efficiently degraded. Thus, individuals who carry factor V Leiden have increased thrombin generation and a hypercoaguable state, which could explain the increased risk for stroke associated with this allele (Dahlback, 1995).
Poort and colleagues (Poort et al., 1996) first identified the G20210A single- base-pair substitution in the 3′ untranslated region of the parent gene, coagulation factor II (F2), and demonstrated its association with elevated plasma prothrombin levels and increased risk for venous thrombosis. Subsequent studies have demonstrated that prothrombin levels were probably due to increased thrombin generation (Franco et al., 1999). More recently, the mechanism by which G20210A alters prothrombin levels has been established. Ceelie and colleagues (Ceelie et al., 2004) have demonstrated that the G20210A mutation results in a more effective poly (A) site (a poly [A] tail is required for mRNA to be efficiently exported from the nucleus and translated into protein in the cytoplasm), leading to elevated mRNA levels, resulting in increased prothrombin production and thrombin formation. Thus, like factor V Leiden, the G20210A mutation probably leads to a pro- coagulant state, thereby increasing risk of stroke.
METHYLENETETRAHYDROFOLATE REDUCTASE (MTHFR)
MTHFR catalyzes the conversion of 5,10- methylenetetrahydrofolate to 5- methyltetrahydrofolate, a cosubstrate for homocysteine remethylation to methionine. Frosst and colleagues (Frosst et al., 1995) identified a cytosine- to-thymine single-base-pair substitution at position 677 (C677T) that converts an alanine to a valine residue, and produces a thermolabile form of the protein. They demonstrated that this variant was associated with reduced enzyme activity and increased levels of homocysteine. Elevated levels of homocysteine are associated with increased risk for stroke (Wald et al., 2002), thus, the C677T variant is likely to contribute to increased risk of stroke directly due to its reduced ability to metabolize homocysteine.
More common forms of stroke can be divided into two major varieties: ischemic and hemorrhagic. The majority of strokes are ischemic (80–90%), and can be further subdivided into: 1) large-vessel occlusive disease, usually due to atherosclerosis and plaque formation; 2) small-vessel occlusive disease, which have involvement of small, perforating end-arteries in the brain; and 3) cardiogenic stroke, which is secondary to blood clots from a diseased heart. Traditionally, occlusive disease has been considered to be due to atherosclerosis, and cardiogenic stroke due to atrial fibrillation secondary to mitral-valve stenosis; however, arguments have been put forward that all three forms probably have a significant atherosclerotic component (Gulcher et al., 2005). Given the heterogeneous nature of the stroke phenotype, assumptions about the underlying etiology of disease can have a significant impact on the ability to identify genetic determinants of stroke susceptibility.
The leap from the existence of genes for monogenic and rare forms of stroke to the likely existence of genes contributing to risk for common forms of stroke is bolstered by numerous studies that have shown that a genetic component to susceptibility to common forms of stroke probably exists. A comprehensive analysis of these studies, in which all genetic epidemiology studies of ischemic stroke from 1966 to 2003 were systematically reviewed, was conducted by Flossmann and colleagues (Flossmann et al., 2004). Based on twin (OR 1.65, 95% CI 1.2–2.3), case-control (OR 1.76, 95% CI 1.7–1.9), and cohort (OR 1.30, 95% CI 1.2–1.5) studies, they concluded that there is a modest but significant genetic component to the risk for ischemic stroke in the general population. Most genetic studies have focused on candidate genes under case-control designs, and Casas and colleagues (Casas et al., 2004) performed a meta-analysis of these studies on ischemic stroke, incorporating 32 genes studies across ~18,000 cases and ~58,000 controls. They identified significant associations for factor V Leiden (OR 1.33, 95% CI 1.12–1.58), methylenetetrahydrofolatereductase (MTHFR) C677T (OR = 1.24, 95% CI 1.08–1.42), prothrombin G20210A (OR = 1.44, 95% CI 1.11–1.86) (discussed above, in the context of CVT), as well angiotensin-converting enzyme (ACE) insertion/deletion (OR 1.21, 95% CI 1.08–1.35). A more recent meta-analysis conducted by Hamzi et al. (Hamzi et al., 2011) reviewed 300 manuscripts for five candidate genes among 152,797 individuals (45,433 cases and 107,634 controls) and confirmed the associations for prothrombin (OR 1.57, 95% CI 1.23–2.89) and ACE (OR 1.11, 95% CI 1.02–1.25), but did not find significant results for factor V Leiden or MTHFR for ischemic stroke. Therefore, they concluded that there are common variants in several genes involved in common forms of stroke, each with a modest effect. Meta-analyses of the ACE insertion/deletion variant (see below) in non–European descent individuals also revealed a significant association with ischemic stroke, indicating the importance of this variant across different ethnic groups (Ariyaratnam et al., 2007; Wang et al., 2012).
ANGIOTENSIN-CONVERTING ENZYME (ACE)
ACE plays an important role in blood pressure regulation and electrolyte balance, and ACE inhibitors have been at the forefront of therapy for treating hypertension and reducing risk for CVD. Indeed, ACE is an important regulator of the renin-angiotensin-aldosterone system through both its ability to hydrolyze angiotensin I into angiotensin II, a potent vasopressor, and its ability to inactivate bradykinin, a potent vasodilator that may stimulate nitric oxide production (Kim and Iwao, 2000). ACE is found on the surface of vascular endothelial cells and in circulating plasma, and animal studies have shown the importance of ACE in regulating blood pressure (Esther et al., 1997; Krege et al., 1995). In 1990, Rigat and colleagues (Rigat et al., 1990) identified an insertion/deletion polymorphism (I/D) that was responsible for up to 50% of the variation in circulating levels of ACE. While the molecular basis of how the I/D polymorphism affects circulating ACE levels is not entirely clear, a study using nearby polymorphisms to measure specific expression of the I and D alleles indicates that the D allele leads to higher expression of ACE mRNA (Suehiro et al., 2004).
Given the widespread use of ACE inhibitors in clinical treatment, and the high frequency of the I/D polymorphism in the general population (~30% in Caucasian populations), the ACE I/D polymorphism provides a prime target for testing the potential impact of a genetic variant on choice of drug therapy (pharmacogenetics). Arnett and colleagues (Arnett et al., 2005) have reported the results of a double-blind, active-controlled randomized trial of antihypertensive treatment in which they examined the impact of the ACE I/D polymorphism on response to four different medications (chlorthalidone, amlodipine, lisinopril, and doxazosin). The study included 37,939 participants ≥55 years of age with ≥1 risk factor for CVD. These individuals were followed up for four to eight years, with primary outcomes including fatal coronary heart disease (CHD) and/or nonfatal MI, and secondary outcomes including stroke, all-cause mortality, combined CHD, and combined cardiovascular disease. ACE I/D genotype was not predictive for CHD (though the risk for stroke was consistent with the meta-analysis of Casas et al., 2004), nor did it modify the response to treatment with the different antihypertensive medications. These results were surprising, as one would predict that those with the D/D genotype, and therefore higher ACE levels, would be more responsive to the ACE inhibitor therapy (lisinopril). Indeed, these results provide a warning for making the leap between genetics and treatment. Despite a functional variant in a gene whose product is a direct target of one of the therapies, the choice of therapy did not affect the outcome. There were however, some differences in outcome according to gender and diabetes status, but given the number of hypotheses tested, further follow-up is needed to verify any of these observations.
In addition to candidate-gene studies that led to the revelation of the association of the above-discussed genes with stroke, traditional genetic approaches have also focused on using large pedigrees in family-based linkage studies. While these studies have been at the forefront of identifying genes for Mendelian diseases, they are often difficult to implement for phenotypes that occur late in life, such as stroke, due to the difficulty in obtaining informative pedigrees. However, deCODE Genetics has leveraged the combination of extensive genealogical records and medical records in Iceland to be able to perform these types of studies.
PHOSPHODIESTERASE 4D (PDE4D)
Gretarsdottir and colleagues (Gretarsdottir et al., 2002) initially performed a genome-wide linkage scan in 476 patients with stroke within 179 extended pedigrees from Iceland, and identified a locus on 5q12 (Log of odds [LOD] score = 4.40), which they designated as “STRK1.” They employed a broad definition of stroke, including individuals with either ischemic or hemorrhagic stroke, as well as TIA (transient ischemic attack), which they considered an ischemic event, arguing that the same pathophysiological mechanisms are responsible for both. In a subsequent study, the same group fine-mapped this locus using a population-based case-control study composed of 864 Icelandic affected individuals and 908 controls, implicating PDE4D, a regulator of intracellular levels of cyclic adenosine monophosphate (cAMP) (Gretarsdottir et al., 2003). PDE4D mRNA is expressed in cardiac myocytes and may be involved in excitation–contraction coupling (Lehnart et al., 2005). However, this association was limited to ischemic stroke, and specifically to the combined cardiogenic and carotid forms (using Trial of Org 10172 in Acute Stroke Treatment [TOAST] subcategories). Subsequent studies have yielded ambiguous replication results, and two large meta-analyses, with the most recent containing >10,000 cases and >10,000 controls, have not demonstrated an association with ischemic stroke (Bevan et al., 2008; Lovkvist et al., 2012).
ARACHIDONATE 5-LIPOOXYGENASE-ACTIVATING PROTEIN (ALOX5AP)
Helgadottir and colleagues (Helgadottir et al., 2004) reported a finding of linkage and association with ALOX5AP and both stroke and MI in an Icelandic population. They identified a specific haplotype (HapA) that is relatively common and carried in 27% of patients with stroke (Relative risk [RR] 1.7, P <0.0001). The association with both MI and stroke was particularly intriguing, and in the same paper Helgadottir and colleagues identified another haploytpe (HapB) that was associated with MI in an independent British cohort. They went on to demonstrate that the synthesis of leukotriene B4 (LTB4) in ionomycin-stimulated neutrophils from patients with a history of MI is greater than from controls without MI and that this difference is largely accounted for by carriers of the HapA haplotype. LTB4, which is a biolipid inflammatory and vasoactive mediator, is produced from LTA4, which is produced from arachidonic acid by ALOX5AP. Thus, the association study is backed by functional evidence that points to a specific mechanism by which variants in ALOX5AP may increase risk for both stroke and MI: elevated levels of LTB4 might contribute to increased inflammation, a known risk factor for CVD events through the development and atherosclerosis and/or plaque instability. As with PDE4D, replication studies have yielded ambiguous results, and the largest meta-analysis to date, with>5,000 cases and >4,500 controls, does not show a significant association of the HapA haplotype and ischemic stroke (Zintzaras et al., 2009).
PITX2 AND ZFHX3
As noted above for MI, the advent of GWAS has begun to change the landscape of stroke genetics (see Table 21.5). The first successful GWAS for ischemic stroke (Gretarsdottir et al., 2008) identified a signal near PITX2 (paired-like homeodomain transcription factor 2), which had previously been associated with atrial fibrillation (AF) (Gudbjartsson et al., 2007). AF, one of the most common forms of electrical instability, is characterized by chaotic electrical activity of the atria, and plays a major role in cardioembolic stroke (Lip and Tse, 2007). While the association was initially also reported for non- cardiogenic stroke (Gretarsdottir et al., 2008), subsequent studies have validated the association with cardiogenic stroke (International Stroke Genetics Consortium [ISGC] et al., 2012), but not overall ischemic stroke (Carty et al., 2012). The utility of using AF as an endophenotype to identify cardiogenic stroke–associated variants is further validated by the observation of an association between stroke and SNPs at the ZFHX3 locus (Gudbjartsson et al., 2007; ISGC et al., 2012), which had previously been associated with AF (Benjamin et al., 2009; ISGC et al., 2012).
Table 21.5 GENETIC LOCI ASSOCIATED WITH STROKE
A. LOCI/GENES FROM CANDIDATE-GENE APPROACH
|ACE||17||. . .|
|APOA (LPA)||6||. . .|
|APOE||19||. . .|
|CYP4AII||1||. . .|
|DDAH1||1||. . .|
|Factor V Leiden||1||. . .|
|Fibrinogen||4||. . .|
|GP1BA||17||. . .|
|GPIIIa||17||. . .|
|IL-6||7||. . .|
|PAI-1 (Serpine)||7||. . .|
|Prothrombin||11||. . .|
|TNF-Alpha||10||. . .|
B. LOCI FROM GWAS FOR ISCHEMIC STROKE
*Adapted from Bevan et al., 2012.
In the first study of prospectively identified stroke in the general population, Ikram and colleagues identified a single locus, NINJ2 (Ikram et al., 2009). NINJ2 encodes ninjurin2, an adhesion molecule that is upregulated in response to nerve injury. However, this locus was not validated by subsequent studies (ISGC et al., 2012; ISGC and Wellcome Trust Case- Control Consortium 2, 2010). A more recent GWAS for ischemic stroke, conducted in >9,000 cases and >11,000 controls, replicated associations for cardioembolic stroke near PITX2 and ZFHX3, as well for the 9p21 locus (previously implicated in MI; see above) with large-vessel disease (Gschwendtner et al., 2009). They also reported a novel finding with the HDAC9 locus and large vessel stroke (OR 1.42, 95% CI 1.28–1.57). HDAC9 encodes histone deacetylase 9, an enzyme involved in regulating chromatin structure and gene transcription (Haberland et al., 2009). Using a Bayesian statistical framework to formally test whether different variants were associated with all subtypes of ischemic stroke or specific subtypes, they clearly demonstrate the importance of subtype classification and provide strong evidence of heterogeneity of genetic effects across stroke subtypes. The PITX2 and ZFHX3 variants appear to only affect risk for cardioembolic stroke, whereas the 9p21 variants appear to broadly influence ischemic stroke, and HDAC9 is specific for large-vessel stroke. Based on this specificity, the authors postulate that a mechanism for association of HDAC9 with stroke through accelerated atherosclerosis is possible, but note that this hypothesis is highly speculative.
In an example of population-specific findings, a gene-based association study using ~50,000 tag SNPs was conducted in ~1100 Japanese cerebral infarct cases, and identified a nonsynonymous SNP in PRKCH (V374I), which encodes protein kinase C eta (Kubo et al., 2007). The authors go on to show that the identified variant appears to be functional, resulting in higher autophosphorylation and kinase activity, which activates its downstream signaling pathway. The variant has a frequency of about 20% in Asian populations, but less than 1% in European ancestry individuals, and is not observed in the Yoruba from Ibadan (International HapMap Consortium, 2005). This finding has been validated in a Chinese population (Wu et al., 2009), as well as in a meta-analysis of >3,600 cases of ischemic stroke and>4,500 controls drawn from Chinese and Japanese populations (Li et al., 2012). PRKCH is a serine-theronine kinase that is mainly expressed in vascular endothelial cells and foamy macrophages (which play an important role in atherosclerosis), and is involved in regulation of cell differentiation, proliferation, and apoptosis. Indeed, increased expression of PRKCH was correlated with progression of coronary atherosclerotic lesion type, providing strong evidence for a mechanism by which altered PRKCH levels influence risk of stroke (Kubo et al., 2007).
A series of GWAS have focused on intracranial aneurysm, a major cause of hemorrhagic stroke, and identified six loci using discovery and replication cohorts from Europe and Japan comprising >5,800 cases and >14,000 controls (Bilguvar et al., 2008; Yasuno et al., 2010; Yasuno et al., 2011). These loci include the 9p21 locus, which is implicated in CAD and large- vessel ischemic stroke (see above). Additional loci include SOX17, CNNM2, KL/STARD13, RBBP8, and EDNRA. The author note that a common pathway tying these genes together is cell cycle progression, and these genes may affect proliferation and senescence of progenitor-cell populations (Yasuno et al., 2010). Indeed, SOX17 is required for endothelial formation and maintenance, and Sox17-/- mice show vascular abnormalities (Sakamoto et al., 2007). EDNRA, which encodes a G protein-coupled receptor for endothelin, is a particularly intriguing candidate, as it mediates the vasoconstriction and mitogenic effects of EDN1 (Alberts et al., 1994; Suzuki et al., 1999). Yasuno and colleagues (Yasuno et al., 2011) note that the effects of EDNRA variants on IA risk may occur in two distinct ways, depending on whether the variant increases or decreases EDNRA-mediated signaling. An increase in signal might promote development of atherosclerosis, whereas a decreased signal might lead to an inability to adequately repair the vasculature after vascular injury. Understanding the specific nature of the risk variant may have important pharmacogenetic implications, as selective EDNRA antagonists are in clinical trials for treatment of subarachnoid hemorrhage (clazosentan) (Kramer and Fletcher, 2009; Macdonald et al., 2008; Macdonald et al., 2013; Vergouwen et al., 2012).
SUDDEN CARDIAC DEATH (SCD)
Sudden cardiac death continues to be one of the leading causes of death in the United States. According to the U.S. Centers for Disease Control and Prevention, about 462,000 of the 2,400,000 (19.3%) U.S. deaths in 1999 were classified as “sudden cardiac deaths” (SCDs) using their definition of SCD as including all deaths “due to cardiac disease that occurred out of hospital (~341,000) or in an emergency department, or one in which the decedent was reported ‘dead on arrival’” (Zheng et al., 2002). From the standpoint of preventive care, SCD poses a huge burden, since fewer than 10% of SCD victims survive, and approximately one-third of all victims manifest SCD as their first clinical event. Approximately two-thirds of SCD victims do not have clinical symptoms that would warrant preventive intervention. Therefore, the ability to identify individuals who are at high risk for SCD is crucial, and advances in genetics may fill this gap.
As for stroke and MI, a great deal of progress has been made identifying the genes involved in Mendelian forms of disease that contribute to increased risk of SCD. Mutations in coding sequences in at least seven cardiac sarcolemmal, sodium, potassium, and calcium ion channels subunit genes (i.e., KVLQT1 [KCNQ1], HERG [KCNH2], SCN5A, minK [KCNE1], RYR2, MiRP1 [KCNE2], and Kir2.1 [KCNJ2]) result in increased susceptibility to SCD (Priori and Napolitano, 2004). Electrophysiological dysfunction, manifested as delayed myocardial cell depolarization and repolarization, is caused by mutations in the proteins encoded by these genes, as was originally discovered by Keating, Schwartz, Moss, Priori, and others during the 1990s, and it is now known to underlie a whole family of related pro-arrhythmic conditions, exemplified by the Long QT syndrome (LQTS) and Brugada’s syndrome (Keating and Sanguinetti, 2001; Splawski et al., 2000). Mutations in these same genes also result in converse disorders, such as “short QT syndrome,” which also enhances SCD risk (Brugada et al., 2004; Gaita et al., 2003).
Multiple etiologies probably contribute to increased SCD risk, including susceptibilities that arise from genetic variations in sarcomeric proteins, such as beta-myosin heavy chain (MyHC), cardiac troponin T (cTnT), and myosin binding protein-C (MyBP-C), which underlie SCDs that occur in patients with inherited hypertrophic cardiomyopathies (Marian and Roberts, 2001). Another important factor is genetic changes that impact early patterning events during embryogenesis, and subsequently cause disturbances in cardiac electrical function from development through maturation. Chien and collaborators were among the first to report that genetic alterations affecting early transcription-factor expression may lead to enhanced arrhythmia susceptibility (Nguyen-Tran et al., 2000). Similar alterations in developmental factors were also implicated in rare familial vascular defects that appear to result in enhanced susceptibility to myocardial infarction (Wang et al., 2003).
The existence of rare inherited monogenic disorders, such as the LQTS and Brugada syndromes, demonstrate that mutations in ion-channel genes, structural proteins, and calcium-handling genes can increase susceptibility to lethal arrhythmias (Keating and Sanguinetti, 2001; Splawski et al., 2000). It is thus a small leap to propose that subtler variations in these same genes may predispose to more common forms of SCD. Two population-based studies have demonstrated that a family history of SCD increases the risk for SCD approximately 1.8-fold, independent of other traditional cardiovascular disease risk factors (Friedlander et al., 1998; Jouven et al., 1999). The first study, conducted in the United States by Friedlander and colleagues (Friedlander et al., 1998), analyzed associations with “primary cardiac events” in a cohort of men and women attended by first responders in King County, Washington (235 cases, 374 controls). The second, done in Paris by Jouven and colleagues (Jouven et al., 1999), analyzed deaths in a cohort of 7,746 asymptomatic middle-aged males, using retrospective autopsy and clinical data analyses to ascribe cardiac deaths to either SCD or MI. Multifactorial statistical analyses indicated that the occurrence of SCD in a parent results in a 1.6–1.8-fold increase in SCD susceptibility, despite controlling for conventional risk factors indicative of coronary disease (e.g., cholesterol sub-fractions, blood pressure, obesity, tobacco use, etc.). In a very limited number of cases in the Parisian study, where there was a history of both maternal and paternal SCD events (n = 19), the relative risk in offspring was ~9 (P = 0.01), indicating an additive genetic model. Elevated incidence of SCD in the Paris study segregated independently of elevated familial incidence of myocardial infarction, suggesting genetic factors that specifically associate with risk for SCD, rather than factors that may underlie overall CVD risk (e.g., atherosclerosis).
α NAV1.5 SUBUNIT (SCN5A)
The idea that ion-channel sequence variations that alter cardiac de- or re- polarizing currents in patients with rare inherited syndromes, like LQTS, may also contribute to enhanced SCD susceptibility seen in more common forms of cardiac disease represents an attractive hypothesis. Splawski and colleagues (Splawski et al., 2002) have given some insight into this paradigm. In their study, a single-nucleotide sequence variant in the SCN5A Na channel gene found in African Americans, S1102Y, was associated with a modest enhancement of arrhythmia risk. The aberrant allele was estimated to be present in up to 4.6 million African Americans: a level of prevalence far beyond all previously established SCD susceptibility alleles combined, and it was not identified in other ethnic populations sampled. In in vitro transfection experiments, the Y1102 allele accelerated channel activation, providing a plausible mechanism by which this variant may increase the likelihood of altered cardiac repolarization and arrhythmia. This finding was followed up in a series of 289 sudden deaths in blacks by Burke and colleagues (Burke et al., 2005). Individuals were classified into four categories: 1) controls, 2) cardiac deaths with clear anatomical substrate, 3) cardiac deaths with no anatomical substrate except mild to moderate cardiac hypertrophy, and 4) unexplained cardiac arrhythmias. The frequency of the Y1102 allele was significantly higher in those with SCD in the absence of a clear morphological abnormality (categories 3 and 4, combined n = 65). These findings strongly suggest that this allele is a risk factor for SCD in African Americans, but require confirmation in a larger cohort.
SCN5A variants have also been studied in non–African ancestry populations, yielding mixed results. An examination of all coding exons in 67 SCD cases with known coronary artery disease and 91 CAD controls in the Oregon Sudden Unexpected Death Study (primarily of European descent) found no association between coding variants and SCD risk (Stecker et al., 2006). In a Han Chinese population, the A1673G variant in SCN5A has been observed to modify risk for SCD (Chen et al., 2004; Fang et al., 2008), though this SNP has not been found to be associated in other populations with SCD (Doolan et al., 2008; Stecker et al., 2006). Finally, in a sequencing study of SCN5A and four potassium-channel genes (KCNQ1, KCNH2, KCNE1, and KCNE2), Albert and colleagues (Albert et al., 2008) identified a significantly higher proportion of variants in SCN5A in women who died suddenly. An examination of common variants in these same genes in a nested case-control analysis of 516 cases and 1,522 matched controls of European ancestry demonstrated two intronic SNPs significantly associated with SCD, one in SCN5A and one in KCNQ1 (Albert et al., 2010). While clearly requiring additional replication, together these results suggest that both common and rare variants in SCN5A may contribute to altered risk of SCD.
Thus, while several marginal associations have been reported for ion- channel genes, none of these results has been convincingly replicated in independent studies, and therefore they remain unproven. Indeed, SCD is likely to be the result of multiple pathways that contribute to increased susceptibility to arrhythmias, including atherosclerosis and thrombosis, electrogenesis and propagation, and initiating influences and triggers (Figure 21.1) (Spooner et al., 2001a; Spooner et al., 2001b). These pathways are likely to involve different genes, and thus extensive phenotyping of samples becomes important. In the absence of being able to distinguish SCD due to different underlying etiologies (e.g., structural defects vs. ion-channel defects), all samples will fall under the rubric of “SCD,” and the power to find genetic determinants is greatly reduced (Arking et al., 2004). Several studies are attempting to address this issue, including the Oregon Sudden Unexpected Death Study, which recruits samples through the emergency medical system and attempts to get electrocardiogram data on all SCDs and autopsy data when available (Chugh et al., 2003). There is also the need to obtain prospective data in order to assess attributable risk for SCD- susceptibility variants, and that effort is ongoing in the ARIC and CHS cohorts (ARIC-Investigators, 1989; Fried et al., 1991). In the absence of large, well-phenotyped SCD cohorts, a great deal of focus has been on subclinical phenotypes, including the QT interval.
Figure 21.1 Potential genetic contributors to sudden cardiac death (SCD). Potential and documented elements of susceptibility are suggested in three broad pathways: 1) those that lead to progressive atherosclerosis and frank coronary disease and the likelihood of an occlusive infarction and ischemic arrhythmias; 2) those involved in electrogenesis and intromyocardial conduction pathways; and 3) those that may influence the initiation of aberrant triggering events and the perpetuation of an arrhythmia. Adapted from Spooner et al., 2001, and Arking et al., 2004.
The QT interval is a measure of cardiac repolarization and is subject to the joint control of the depolarizing Na+ (INa) currents, Ca2+ (ICa) currents, and the repolarizing slow (IKs) and rapid (IKr) K+ currents. QT interval (corrected for heart rate) is a moderately heritable trait (25% to 52% heritability) (Busjahn et al., 1999; Carter et al., 2000; Newton-Cheh et al., 2005), and extremes of QT interval have been associated with increased risk for SCD in both Mendelian forms (LQTS and short QT syndrome [SQTS]), as well as in population-based settings (de Bruyne et al., 1999; Dekker et al., 1994; Dekker et al., 2004; Elming et al., 1998; Okin et al., 2000; Schouten et al., 1991; Sharp et al., 1998). Taken together, these observations suggest that QT interval is likely to have a genetic component (moderate heritability), and that genetic variants that modify QT interval may also modify risk of SCD.
Testing the utility of endophenotypes to identify disease-related genes was a major motivation behind one of the first successful GWAS, in which the investigators identified a common variant in the 5′ region of the NOS1AP gene associated with a 2–3 ms increase in QT interval per minor allele. NOS1AP encodes an adapter protein that physically bridges neuronal nitric oxide synthase with its targets and modulator proteins. Specifically how NOS1AP variants modulate QT interval is currently unknown, but over- expression of NOS1AP in guinea pig ventricular myocytes results in shortening of the cardiac action potential, a decrease in the L-type Ca2+ (ICa) current, and a smaller increase in the rapid delayed rectifier K+ current (IKr), with a resultant shortening of the QT interval (Chang et al., 2008). The genetic association has been extensively replicated (Arking et al., 2009; Eijgelsheim et al., 2009; Lehtinen et al., 2008; Post et al., 2007; Raitakari et al., 2008; Tobin et al., 2008). In a follow-up study of NOS1AP with SCD in the combined Atherosclerosis Risk in Communities Study and Cardiovascular Health Study cohorts (498 cases, 19,295 controls), Kao and colleagues (Kao et al., 2009) demonstrated that the NOS1AP SNP most strongly associated with QT interval, rs16847548, was associated with risk for SCD in white American adults, with the QT-prolonging variant associated with increased SCD risk (p = 0.002). Individuals homozygous for the risk allele were approximately 72% more likely to die of SCD than individuals homozygous for the non-risk genotype, even after adjusting for age, sex, and heart rate. It is important to note the risk allele is common, with 39% of the general white American population carrying one copy and 5% carrying two copies. These findings have been confirmed by a second independent study (Eijgelsheim et al., 2009).
In addition to QT interval, QRS (cardiac ventricular conduction) and RR (inverse heart rate) intervals have also been associated with cardiovascular mortality and SCD (Desai et al., 2006; Jouven et al., 2001). The publication of GWAS identifying numerous variants associated with these traits (Eijgelsheim et al., 2010; Sotoodehnia et al., 2010) has allowed a more comprehensive assessment of the role of electrocardiogram (ECG)-associated SNPs in SCD risk. In one such study, using data generated from an SCD GWAS composed of 1,283 SCD cases and >20,000 controls, Arking and colleagues (Arking et al., 2011) examined 49 SNPs associated (p < 5 × 10–8) with QRS, QT, and RR intervals. In a test looking at direction of the genetic effect, the ECG-trait-prolonging allele was significantly more often associated with increased risk of SCD (31 of 49, p = 0.03), with this effect almost entirely due to the QRS/QT-associated SNPs (28 of 40, p = 0.006). Three loci, including PLN, KCNQ1, and NOS1AP, showed nominal association with SCD, while a fourth locus, TKT/CACNA1D/PRKCD, was significant even after Bonferroni correction to account for the number of loci tested. The TKT/CACNA1D/PRKCD association is particularly intriguing due to the observation that the QRS-prolonging allele was protective for risk of SCD, which is counter to the effect observed with the measured trait (longer QRS duration is associated with increased risk of SCD). This result raises the possibility that the effect of the SNP variant on risk of SCD may not be mediated through its effect on QRS interval. A similar result was also seen in NOS1AP, where one of the alleles that were associated with SCD had no effect on QT interval (Kao et al., 2009).
While focusing on candidate genes and endophenotypes has yielded several compelling candidates, there is no doubt that additional genes play a role in SCD, and they are not likely to be identified through these approaches. Several GWAS with SCD as the phenotype of interest have been published, two of which have reported genome-wide significant findings. The AGNES (Arrhythmia Genetics in the Netherlands) cohort, which is composed of individuals with a first myocardial infarction and ventricular fibrillation (VF) who survived to hospital admission (n = 515) compared with individuals with myocardial infarction alone (n = 457), reported a SNP, rs2824292 in the 21q21 locus, with an OR of 1.78 (95% CI 1.47–2.13, p = 3.36 × 10–10), and an OR of 1.49 (95% CI 1.14–1.95, p = 0.004) in a replication sample of 146 out-of-hospital SCD cases and 391 controls (Bezzina et al., 2010). This SNP is a common variant (allele frequency of 47%) located in an intergenic region. The nearest gene, CXADR (~100 kb away), encodes a viral receptor implicated in viral myocarditis (Bowles et al., 1986), but it is not directly implicated by this study.
A second GWAS, comprising a total of 4,400 SCD cases and >30,000 controls, all of European ancestry, reported a significant signal at the 2q24.2 locus, with the strongest SNP, rs4665058 (p = 1.8 × 10–10), mapped to an intron of the BAZ2B gene. This locus contains three genes expressed in the heart but not previously known to play a role in cardiac biology (BAZ2B, WDSUB1, and TANC1). The risk allele had a study-size-weighted frequency of 1.4% and increased the risk for SCD by 1.92-fold per allele (95% CI 1.57– 2.34). Based on non-human primate sequence data, the risk allele is ancestral; thus its low frequency in European-ancestry populations suggests strong negative selection, as fewer than 0.8% of ancestral alleles have reached a frequency of 1.4% or lower. A search for missense/splice mutations correlated with rs4665058 (r2 > 0.8) using pilot 1 data from the 1000 Genomes Project (November 2010 release) (Durbin et al., 2010) was unsuccessful, indicating that the functional variant is probably regulatory in nature. The authors note that the meta-analysis consisted of both population- based and case-control studies, with some of the case-control studies using CAD controls as opposed to population-based controls. They thus suggest that the consistent results in studies with both CAD controls and population- based controls (see supplementary figure 2 from Arking et al., 2011) provide evidence that the risk associated with rs4665058 may be specific to SCD rather than a generic CAD risk factor. Somewhat surprisingly, this study did not replicate the 21q21 association seen in the AGNES cohort, despite adequate power. This lack of replication may reflect an association specific to the underlying population from which the AGNES cohort is drawn, or may be limited to the highly specific AGNES phenotype of individuals with a first myocardial infarction who survived a VF event, as opposed to a more broadly defined class of SCD observed in the general population.
MADS BOX TRANSCRIPTION ENHANCER FACTOR 2, POLYPEPTIDE A (MEF2A)
One approach to identifying genetic determinants of complex traits has been to identify families that exhibit monogenic forms of the phenotype of interest, with the idea that any genes identified are likely to be involved in complex forms of the phenotype as well. This approach has several merits, including less phenotypical heterogeneity, since presumably all affected individuals in the pedigree are manifesting the same phenotype. Additionally, one can use traditional linkage analysis, which does not require recruiting a control population: however, if not done carefully, this can lead to both false positives and false negative results. This approach was adopted by Wang and colleagues (Wang et al., 2003), who studied a large family that displayed an autosomal dominant form of CAD. They performed genome-wide linkage analysis, and identified a significant association on chromosome 15q26. With 93 genes in the associated region, they focused on the transcription factor MEF2A, largely due to the overall expression pattern; MEF2A was expressed in blood vessels during mouse early embryogenesis (Edmondson et al., 1994), and expression was similar to vascular endothelial growth factor receptor 2 (VEGFR2) and the Von Willebrand factor (Subramanian and Nadal-Ginard, 1996). These data led Wang and colleagues to speculate that MEF2A can be an early marker for vasculogenesis. They sequenced the gene in the affected individuals, and identified a 21-bp deletion in exon 11 (termed “Δ7aa”) that resulted in the loss of seven conserved amino acids and was present in all affected members in the family. They further went on to demonstrate altered cellular trafficking for the mutant protein, and that its ability to activate atrial natriuretic factor is reduced. In a subsequent study, the same group examined 207 CAD/MI patients for mutations in MEF2A, identifying three novel mutations in exon 7 in four patients, and no mutations in 191 controls (Bhagavatula et al., 2004). They demonstrated that these mutations significantly reduce the transcriptional activity of MEF2A, and suggest that “a significant percent of the CAD/MI population may carry mutations in MEF2A.”
The combination of both family- and population-based evidence would seem to present strong evidence for the involvement of MEF2A with CAD/MI. However, a subsequent study by Weng and colleagues (Weng et al., 2005) raised some doubts about the strength of the effect of MEF2A variants on susceptibility to CAD/MI. They sequenced MEF2A exons in 300 white individuals with documented CAD with onset before age 55 years (men) or 65 years (women), and in 300 elderly controls (men >60 yrs, women >70 yrs) who did not have signs or symptoms of CAD. Of five missense mutations identified, one was unique to the CAD individuals, one was unique to the controls, and three were common to both groups. They further observed the 21-bp deletion in three unrelated, unaffected individuals, and demonstrated that the deletion does not segregate with early CAD. They conclude that “these studies support that MEF2A mutations are not a common cause of CAD in white people and [we] argue strongly against a role for the MEF2A 21-bp deletion in autosomal dominant CAD.” A similar negative association was reported by Kajimoto and colleagues in a Japanese population (Kajimoto et al., 2005) and by Gonzalez and colleagues (Gonzalez et al., 2006) in a Spanish cohort for the Δ7aa allele, though Gonzalez and colleagues reported a positive association for a rare Pro279Leu mutation (OR 3.06, 95% CI 1.17–8.06).
In an accompanying commentary to the Weng and colleagues paper, Altshuler and Hirschhorn (Altshuler and Hirschhorn, 2005) concluded that the role of MEF2A variants in CAD has not been established, and used the MEF2A studies to illustrate a number of criteria that should be imposed when performing similar studies. They note that in a complex disease like CAD, it may not be particularly uncommon to find large pedigrees that appear to display monogenic forms of the disease. Thus, they propose looking for an unusual phenotype shared by affected individuals, such as early-onset or a syndromic form of disease. The phenotype also needs to be consistent across family members, which can be difficult with CAD, in which some individuals are “affected” by virtue of having had prior coronary events, while others may have angiographically defined disease in the absence of events. Given these concerns, linkage across multiple families with identical ascertainment criteria should be observed.
In the event that a linkage signal in a region is detected, similar concerns arise when trying to identify the underlying functional variant. The observation of rare, potentially functional variants segregating within a family is not uncommon, and any such variation observed in the linkage region will, by virtue of being in a disease-linked region, segregate with the disease. Thus, Altshuler and Hirschhorn propose that for a specific gene in a linked region to be associated with disease, it must meet one or more of the following criteria:
Multiple different mutations exist, each of which co-segregates with disease; There is confirmatory evidence for a particular allele in a case-control study; There are multiple rare variants that have been well ascertained in controls;
There is observation of a de novo mutation in an affected child (but not in his parent);
There is strong evidence of the effects of a human mutation in a model organism that recapitulates the human disease phenotype.
With the cost of genotyping decreasing and sample-collection increasing, the number of such studies being performed is on the rise; therefore, adopting rigorous criteria, as outlined above, for deeming a gene to be involved in complex disease is warranted.
Our understanding of the genetic landscape of cardiovascular disease has significantly changed in the last five years. We now have a long list of associated loci that have been robustly replicated in a number of different cohorts. However, the daunting challenge before us now is to uncover the underlying causal variant. There are many challenges that impede our ability to do so.
First, common genetic variants are of modest effect (OR <2), which has been a significant factor in the failure to identify and reproducibly demonstrate an effect for a given variant. Most studies have simply been underpowered, greatly contributing to the confusion in the literature. Indeed, as large studies with participant size soaring to over 200,000 become the norm, the list of reproducibly associated loci climbs steadily. Second, our ability to pick out genes involved in disease based on our limited understanding of biology is insufficient, especially given that half the genes in the genome are of largely unknown function. Third, most genes that have been studied have not been studied in a comprehensive way, limiting our ability to compare across studies and to truly exclude genes (as opposed to a specific variant) from association with disease. Fourth, functional variants are likely to be influenced by sex and/or environment (e.g., smoking, diet); thus, comprehensively collecting data on study populations is essential. With genotyping no longer an impediment to progress, the focus going forward is likely to shift from collecting and assessing larger populations to identify additional genetic variants, to finding ways to interrogate the function of the linked loci, identify the underlying causal genes, and use this information to better predict general population risk for cardiovascular disease.
- Agostoni, E., Aliprandi, A., and Longoni, M. (2009). Cerebral venous thrombosis. Expert Rev Neurother 9, 553–564.
- Albert, C.M., MacRae, C.A., Chasman, D.I., et al. (2010). Common variants in cardiac ion channel genes are associated with sudden cardiac death. Circ Arrhythm Electrophysiol 3, 222–229.
- Albert, C.M., Nam, E.G., Rimm, E.B., et al. (2008). Cardiac sodium channel gene variants and sudden cardiac death in women. Circulation 117, 16–23.
- Alberts, G.F., Peifley, K.A., Johns, A., Kleha, J.F., and Winkles, J.A. (1994). Constitutive endothelin-1 overexpression promotes smooth muscle cell proliferation via an external autocrine loop. J Biol Chem. 269, 10112–10118.
- Altshuler, D., and Hirschhorn, J.N. (2005). MEF2A sequence variants and coronary artery disease: a change of heart? J Clin Invest 115, 831–833.
- Amos, C.I., Wang, L.E., Lee, J.E., et al. (2011). Genome-wide association study identifies novel loci predisposing to cutaneous melanoma. Hum Mol Genet 20, 5012–5023.
- Antoniou, A.C., Kuchenbaecker, K.B., Soucy, P., et al. (2012). Common variants at 12p11, 12q24, 9p21, 9q31.2 and in ZNF365 are associated with breast cancer risk for BRCA1 and/or BRCA2 mutation carriers. Breast Cancer Res 14, R33.
- ARIC-Investigators. (1989). The Atherosclerosis Risk in Communities (ARIC) Study: design and objectives. Am J Epidemiol 129, 687–702.
- Ariyaratnam, R., Casas, J.P., Whittaker, J., Smeeth, L., Hingorani, A.D., and Sharma, P. (2007). Genetics of ischaemic stroke among persons of non-European descent: a meta-analysis of eight genes involving approximately 32,500 individuals. PLoS Med 4, e131.
- Arking, D.E., Chugh, S.S., Chakravarti, A., and Spooner, P.M. (2004). Genomics in sudden cardiac death. Circ Res 94, 712–723.
- Arking, D.E., Junttila, M.J., Goyette, P., et al. (2011). Identification of a sudden cardiac death susceptibility locus at 2q24.2 through genome-wide association in European ancestry individuals. PLoS Genet 7, e1002158.
- Arking, D.E., Khera, A., Xing, C., et al. (2009). Multiple independent genetic factors at NOS1AP modulate the QT interval in a multi-ethnic population. PLoS One 4, e4333.
- Arnett, D.K., Davis, B.R., Ford, C.E., et al. (2005). Pharmacogenetic association of the angiotensin- converting enzyme insertion/deletion polymorphism on blood pressure and cardiovascular risk in relation to antihypertensive treatment: the Genetics of Hypertension-Associated Treatment (GenHAT) study. Circulation 111, 3374–3383.
- Beckie, T.M., Beckstead, J.W., and Groer, M.W. (2011). The association between variants on chromosome 9p21 and inflammatory biomarkers in ethnically diverse women with coronary heart disease: a pilot study. Biol Res Nurs 13, 306–319.
- Benjamin, E.J., Rice, K.M., Arking, D.E., et al. (2009). Variants in ZFHX3 are associated with atrial fibrillation in individuals of European ancestry. Nat Genet 41, 879–881.
- Bennet, A.M., Di Angelantonio, E., Ye, Z., et al. (2007). Association of apolipoprotein E genotypes with lipid levels and coronary risk. JAMA 298, 1300–1311.
- Bergametti, F., Denier, C., Labauge, P., et al. (2005). Mutations within the programmed cell death 10 gene cause cerebral cavernous malformations. Am J Hum Genet 76, 42–51.
- Bertina, R.M., Koeleman, B.P., Koster, T., et al. (1994). Mutation in blood coagulation factor V associated with resistance to activated protein C. Nature 369, 64–67.
- Bevan, S., Dichgans, M., Gschwendtner, A., Kuhlenbaumer, G., Ringelstein, E.B., and Markus, H.S. (2008). Variation in the PDE4D gene and ischemic stroke risk: a systematic review and meta- analysis on 5200 cases and 6600 controls. Stroke 39, 1966–1971.
- Bevan, S., Traylor, M., Adib-Samii, P., et al. (2012). Genetic heritability of ischemic stroke and the contribution of previously reported candidate gene and genomewide associations. Stroke 43, 3161–3167.
- Bezzina, C.R., Pazoki, R., Bardai, A., et al. (2010). Genome-wide association study identifies a susceptibility locus at 21q21 for ventricular fibrillation in acute myocardial infarction. Nat Genet 42, 688–691.
- Bhagavatula, M.R., Fan, C., Shen, G.Q., et al. (2004). Transcription factor MEF2A mutations in patients with coronary artery disease. Hum Mol Genet 13, 3181–3188.
- Bilguvar, K., Yasuno, K., Niemela, M., et al. (2008). Susceptibility loci for intracranial aneurysm in European and Japanese populations. Nat Genet 40, 1472–1477.
- Biros, E., Cooper, M., Palmer, L.J., Walker, P.J., Norman, P.E., and Golledge, J. (2010). Association of an allele on chromosome 9 and abdominal aortic aneurysm. Atherosclerosis 212, 539–542.
- Boekholdt, S.M., Bijsterveld, N.R., Moons, A.H., Levi, M., Buller, H.R., and Peters, R.J. (2001). Genetic variation in coagulation and fibrinolytic proteins and their relation with acute myocardial infarction: a systematic review. Circulation 104, 3063–3068.
- Bowles, N.E., Richardson, P.J., Olsen, E.G., and Archard, L.C. (1986). Detection of Coxsackie-B- virus-specific RNA sequences in myocardial biopsy samples from patients with myocarditis and dilated cardiomyopathy. Lancet 1, 1120–1123.
- Bown, M.J., Braund, P.S., Thompson, J., London, N.J., Samani, N.J., and Sayers, R.D. (2008). Association between the coronary artery disease risk locus on chromosome 9p21.3 and abdominal aortic aneurysm. Circ Cardiovasc Genet 1, 39–42.
- Brugada, R., Hong, K., Dumaine, R., et al. (2004). Sudden death associated with short-QT syndrome linked to mutations in HERG. Circulation 109, 30–35.
- Burke, A., Creighton, W., Mont, E., et al. (2005). Role of SCN5A Y1102 polymorphism in sudden cardiac death in blacks. Circulation 112, 798–802.
- Busjahn, A., Knoblauch, H., Faulhaber, H.D., et al. (1999). QT interval is linked to 2 long-QT syndrome loci in normal subjects. Circulation 99, 3161–3164.
- Carter, N., Snieder, H., Jeffery, S., et al. (2000). QT interval in twins. J Hum Hypertens 14, 389–390.
- Carty, C.L., Buzková, P., Fornage, M., et al. (2012). Associations between incident ischemic stroke events and stroke and cardiovascular disease-related genome-wide association studies single nucleotide polymorphisms in the Population Architecture Using Genomics and Epidemiology study. Circ Cardiovasc Genet 5, 210–216.
- Casas, J.P., Hingorani, A.D., Bautista, L.E., and Sharma, P. (2004). Meta-analysis of genetic studies in ischemic stroke: thirty-two genes involving approximately 18,000 cases and 58,000 controls. Arch Neurol 61, 1652–1661.
- Ceelie, H., Spaargaren-van Riel, C.C., Bertina, R.M., and Vos, H.L. (2004). G20210A is a functional mutation in the prothrombin gene; effect on protein levels and 3′-end formation. J Thromb Haemost 2, 119–127.
- Chang, K.C., Barth, A.S., Sasano, T., et al. (2008). CAPON modulates cardiac repolarization via neuronal nitric oxide synthase signaling in the heart. Proc Natl Acad Sci U S A 105, 4477–4482.
- Chen, J., Xie, X., Zhu, J., Tao, Q., and Wang, X. (2004). Single-nucleotide polymorphisms in SCN5A gene in Chinese Han population and their correlation with cardiac arrhythmias. Genet Med 6, 159.
- Cheng, X., Shi, L., Nie, S., et al. (2011). The same chromosome 9p21.3 locus is associated with type 2 diabetes and coronary artery disease in a Chinese Han population. Diabetes 60, 680–684.
- Chugh, S.S., Dogra, V., Thompson, B., et al. (2003). Annual incidence of sudden cardiac death based on operative definition: a prospective, population-based, countywide study using multiple sources. Circulation 108, IV-1040.
- Cohen, J.C., Boerwinkle, E., H., M.T., Jr., and Hobbs, H.H. (2006). Sequence variations in PCSK9, low LDL, and protection against coronary heart disease. N Engl J Med 354, 1264–1272.
- Cunnington, M.S., Santibanez Koref, M., Mayosi, B.M., Burn, J., and Keavney, B. (2010). Chromosome 9p21 SNPs associated with multiple disease phenotypes correlate with ANRIL expression. PLoS Genet 6, e1000899.
- Dahlback, B. (1995). New molecular insights into the genetics of thrombophilia. Resistance to activated protein C caused by Arg506 to Gln mutation in factor V as a pathogenic risk factor for venous thrombosis. Thromb Haemost 74, 139–148.
- de Bruyne, M.C., Hoes, A.W., Kors, J.A., Hofman, A., van Bemmel, J.H., and Grobbee, D.E. (1999). Prolonged QT interval predicts cardiac and all-cause mortality in the elderly. The Rotterdam Study. Eur Heart J 20, 278–284.
- Dekker, J.M., Crow, R.S., Hannan, P.J., Schouten, E.G., and Folsom, A.R. (2004). Heart rate-corrected QT interval prolongation predicts risk of coronary heart disease in black and white middle-aged men and women: the ARIC study. J Am Coll Cardiol 43, 565–571.
- Dekker, J.M., Schouten, E.G., Klootwijk, P., Pool, J., and Kromhout, D. (1994). Association between QT interval and coronary heart disease in middle-aged and elderly men. The Zutphen Study. Circulation 90, 779–785.
- Denier, C., Goutagny, S., Labauge, P., et al. (2004). Mutations within the MGC4607 gene cause cerebral cavernous malformations. Am J Hum Genet 74, 326–337.
- Desai, A.D., Yaw, T.S., Yamazaki, T., Kaykha, A., Chun, S., and Froelicher, V.F. (2006). Prognostic significance of quantitative QRS duration. Am J Med 119, 600–606.
- Diabetes Genetics Initiative of Broad Institute of Harvard and MIT, Lund University, and Novartis Institutes of BioMedical Research, Saxena, R., Voight, B.F., Lyssenko, V., et al. (2007). Genome- wide association analysis identifies loci for type 2 diabetes and triglyceride levels. Science 316, 1331–1336.
- Dong, Y., Hassan, A., Zhang, Z., Huber, D., Dalageorgou, C., and Markus, H.S. (2003). Yield of screening for CADASIL mutations in lacunar stroke and leukoaraiosis. Stroke 34, 203–205.
- Doolan, A., Langlois, N., Chiu, C., Ingles, J., Lind, J.M., and Semsarian, C. (2008). Postmortem molecular analysis of KCNQ1 and SCN5A genes in sudden unexplained death in young Australians. Int J Cardiol 127, 138–141.
- Drenos, F., Whittaker, J.C., and Humphries, S.E. (2007). The use of meta-analysis risk estimates for candidate genes in combination to predict coronary heart disease risk. Ann Hum Genet 71, 611– 619.
- Durbin, R.M., Abecasis, G.R., Altshuler, D.L., et al. (2010). A map of human genome variation from population-scale sequencing. Nature 467, 1061–1073.
- Edmondson, D.G., Lyons, G.E., Martin, J.F., and Olson, E.N. (1994). MEF2 gene expression marks the cardiac and skeletal muscle lineages during mouse embryogenesis. Development 120, 1251–1263. Eijgelsheim, M., Newton-Cheh, C., Aarnoudse, A.L., et al. (2009). Genetic variation in NOS1AP is associated with sudden cardiac death: evidence from the Rotterdam Study. Hum Mol Genet 18,4213–4218.
- Eijgelsheim, M., Newton-Cheh, C., Sotoodehnia, N., et al. (2010). Genome-wide association analysis identifies multiple loci related to resting heart rate. Hum Mol Genet 19, 3885–3894.
- Elming, H., Holm, E., Jun, L., et al. (1998). The prognostic value of the QT interval and QT interval dispersion in all-cause and cardiac mortality and morbidity in a population of Danish citizens. Eur Heart J 19, 1391–1400.
- Enciso-Mora, V., Hosking, F.J., Sheridan, E., et al. (2012). Common genetic variation contributes significantly to the risk of childhood B-cell precursor acute lymphoblastic leukemia. Leukemia 26, 2212–2215.
- Esther, C.R., Marino, E.M., Howard, T.E., et al. (1997). The critical role of tissue angiotensin- converting enzyme as revealed by gene targeting in mice. J Clin Invest 99, 2375–2385.
- Fang, D.H., Wu, L.Q., Lu, L., et al. (2008). Association of human SCN5A polymorphisms with idiopathic ventricular arrhythmia in a Chinese Han cohort. Circ J 72, 592–597.
- Flossmann, E., Schulz, U.G., and Rothwell, P.M. (2004). Systematic review of methods and results of studies of the genetic epidemiology of ischemic stroke. Stroke 35, 212–227.
- Folkersen, L., Kyriakou, T., Goel, A., et al., and PROCARDIS Consortia. (2009). Relationship between CAD risk genotype in the chromosome 9p21 locus and gene expression. Identification of eight new ANRIL splice variants. PLoS One 4, e7677.
- Franco, R.F., Trip, M.D., ten Cate, H., et al. (1999). The 20210 G—>A mutation in the 3′-untranslated region of the prothrombin gene and the risk for arterial thrombotic disease. Br J Haematol 104, 50–54.
- Fried, L.P., Borhani, N.O., Enright, P., et al. (1991). The Cardiovascular Health Study: design and rationale. Ann Epidemiol 1, 263–276.
- Friedlander, Y., Siscovick, D.S., Weinmann, S., et al. (1998). Family history as a risk factor for primary cardiac arrest. Circulation 97, 155–160.
- Frosst, P., Blom, H.J., Milos, R., et al. (1995). A candidate genetic risk factor for vascular disease: a common mutation in methylenetetrahydrofolate reductase. Nat Genet 10, 111–113.
- Gaita, F., Giustetto, C., Bianchi, F., et al. (2003). Short QT syndrome: a familial cause of sudden death. Circulation 108, 965–970.
- Goldstein, L.B., Adams, R., Becker, K., et al. (2001). Primary prevention of ischemic stroke: a statement for healthcare professionals from the Stroke Council of the American Heart Association. Circulation 103, 163–182.
- Gonzalez, P., Garcia-Castro, M., Reguero, J.R., et al. (2006). The Pro279Leu variant in the transcription factor MEF2A is associated with myocardial infarction. J Med Genet 43, 167–169.
- Gori, F., Specchia, C., Pietri, S., et al., GISSI Prevenzione Investigators, and SIBioC-GISSI Prevenzione Group. (2010). Common genetic variants on chromosome 9p21 are associated with myocardial infarction and type 2 diabetes in an Italian population. BMC Med Genet 11, 60.
- Gretarsdottir, S., Sveinbjornsdottir, S., Jonsson, H.H., et al. (2002). Localization of a susceptibility gene for common forms of stroke to 5q12. Am J Hum Genet 70, 593–603.
- Gretarsdottir, S., Thorleifsson, G., Manolescu, A., et al. (2008). Risk variants for atrial fibrillation on chromosome 4q25 associate with ischemic stroke. Ann Neurol 64, 402–409.
- Gretarsdottir, S., Thorleifsson, G., Reynisdottir, S.T., et al. (2003). The gene encoding phosphodiesterase 4D confers risk of ischemic stroke. Nat Genet 35, 131–138.
- Gschwendtner, A., Bevan, S., Cole, J.W., et al. (2009). Sequence variants on chromosome 9p21.3 confer risk for atherosclerotic stroke. Ann Neurol 65, 531–539.
- Gudbjartsson, D.F., Arnar, D.O., Helgadottir, A., et al. (2007). Variants conferring risk of atrial fibrillation on chromosome 4q25. Nature 448, 353–357.
- Gulcher, J.R., Gretarsdottir, S., Helgadottir, A., and Stefansson, K. (2005). Genes contributing to risk for common forms of stroke. Trends Mol Med 11, 217–224.
- Haberland, M., Montgomery, R.L., and Olson, E.N. (2009). The many roles of histone deacetylases in development and physiology: implications for disease and therapy. Nat Rev Genet 10, 32–42.
- Hamzi, K., Tazzite, A., and Nadifi, S. (2011). Large-scale meta-analysis of genetic studies in ischemic stroke: five genes involving 152,797 individuals. Indian J Hum Genet 17, 212–217.
- Harismendy, O., Notani, D., Song, X., et al. (2011). 9p21 DNA variants associated with coronary artery disease impair interferon-gamma signalling response. Nature 470, 264–268.
- Helgadottir, A., Manolescu, A., Thorleifsson, G., et al. (2004). The gene encoding 5-lipoxygenase activating protein confers risk of myocardial infarction and stroke. Nat Genet 36, 233–239.
- Helgadottir, A., Thorleifsson, G., Magnusson, K.P., et al. (2008). The same sequence variant on 9p21 associates with myocardial infarction, abdominal aortic aneurysm and intracranial aneurysm. Nat Genet 40, 217–224.
- Helgadottir, A., Thorleifsson, G., Manolescu, A., et al. (2007). A common variant on chromosome 9p21 affects the risk of myocardial infarction. Science 316, 1491–1493.
- Humphries, S.E., Talmud, P.J., Hawe, E., Bolla, M., Day, I.N., and Miller, G.J. (2001). Apolipoprotein E4 and coronary heart disease in middle-aged men who smoke: a prospective study. Lancet 358, 115–119.
- IBC 50K CAD Consortium. (2011). Large-scale gene-centric analysis identifies novel variants for coronary artery disease. PLoS Genet 7, e1002260.
- Ikram, M.A., Seshadri, S., Bis, J.C., et al. (2009). Genomewide association studies of stroke. N Engl J Med 360, 1718–1728.
- International HapMap Consortium. (2005). A haplotype map of the human genome. Nature 437, 1299– 1320.
- International Stroke Genetics Consortium (ISGC), Wellcome Trust Case Control Consortium 2 (WTCCC2), Bellenguez, C., Bevan, S., Gschwendtner, A., et al. (2012). Genome-wide association study identifies a variant in HDAC9 associated with large vessel ischemic stroke. Nat Genet 44, 328–333.
- International Stroke Genetics Consortium, and Wellcome Trust Case-Control Consortium 2. (2010). Failure to validate association between 12p13 variants and ischemic stroke. N Engl J Med 362, 1547–1550.
- Jarinova, O., Stewart, A.F., Roberts, R., et al. (2009). Functional analysis of the chromosome 9p21.3 coronary artery disease risk locus. Arterioscler Thromb Vasc Biol 29, 1671–1677.
- Joutel, A., Corpechot, C., Ducros, A., et al. (1996). NOTCH3 mutations in CADASIL, a hereditary adult-onset condition causing stroke and dementia. Nature 383, 707–710.
- Jouven, X., Desnos, M., Guerot, C., and Ducimetiere, P. (1999). Predicting sudden death in the population: the Paris Prospective Study I. Circulation 99, 1978–1983.
- Jouven, X., Zureik, M., Desnos, M., Guerot, C., and Ducimetiere, P. (2001). Resting heart rate as a predictive risk factor for sudden death in middle-aged men. Cardiovasc Res 50, 373–378.
- Kajimoto, K., Shioji, K., Tago, N., et al. (2005). Assessment of MEF2A mutations in myocardial infarction in Japanese patients. Circ J 69, 1192–1195.
- Kao, W.H., Arking, D.E., Post, W., et al. (2009). Genetic variations in nitric oxide synthase 1 adaptor protein are associated with sudden cardiac death in US white community-based populations. Circulation 119, 940–951.
- Kathiresan, S., Melander, O., Anevski, D., et al. (2008). Polymorphisms associated with cholesterol and risk of cardiovascular events. N Engl J Med 358, 1240–1249.
- Kathiresan, S., and Srivastava, D. (2012). Genetics of human cardiovascular disease. Cell 148, 1242– 1257.
- Kathiresan, S., Voight, B.,F., Purcell, S., et al. (2009). Genome-wide association of early-onset myocardial infarction with single nucleotide polymorphisms and copy number variants. Nat Genet 41, 334–341.
- Keating, M.T., and Sanguinetti, M.C. (2001). Molecular and cellular mechanisms of cardiac arrhythmias. Cell 104, 569–580.
- Kim, S., and Iwao, H. (2000). Molecular and cellular mechanisms of angiotensin II-mediated cardiovascular and renal diseases. Pharmacol Rev 52, 11–34.
- Kjolby, M., Andersen, O.M., Breiderhoff, T., et al. (2010). SORT1, encoded by the cardiovascular risk locus 1p13.3, is a regulator of hepatic lipoprotein export. Cell Metab 12, 213–223.
- Kral, B.G., Mathias, R.A., Suktitipat, B., et al. (2011). A common variant in the CDKN2B gene on chromosome 9p21 protects against coronary artery disease in Americans of African ancestry. J Hum Genet 56, 224–229.
- Kramer, A., and Fletcher, J. (2009). Do endothelin-receptor antagonists prevent delayed neurological deficits and poor outcomes after aneurysmal subarachnoid hemorrhage? A meta-analysis. Stroke 40, 3403–3406.
- Krege, J.H., John, S.W., Langenbach, L.L., et al. (1995). Male-female differences in fertility and blood pressure in ACE-deficient mice. Nature 375, 146–148.
- Kubo, M., Hata, J., Ninomiya, T., et al. (2007). A nonsynonymous SNP in PRKCH (protein kinase C eta) increases the risk of cerebral infarction. Nat Genet 39, 212–217.
- Laberge-le Couteulx, S., Jung, H.H., Labauge, P., et al. (1999). Truncating mutations in CCM1, encoding KRIT1, cause hereditary cavernous angiomas. Nat Genet 23, 189–193.
- Lahoz, C., Schaefer, E.J., Cupples, L.A., et al. (2001). Apolipoprotein E genotype and cardiovascular disease in the Framingham Heart Study. Atherosclerosis 154, 529–537.
- Lehnart, S.E., Wehrens, X.H., Reiken, S., et al. (2005). Phosphodiesterase 4D deficiency in the ryanodine-receptor complex promotes heart failure and arrhythmias. Cell 123, 25–35.
- Lehtinen, A.B., Newton-Cheh, C., Ziegler, J.T., et al. (2008). Association of NOS1AP genetic variants with QT interval duration in families from the Diabetes Heart Study. Diabetes 57, 1108–1114.
- Li, J., Luo, M., Xu, X., and Sheng, W. (2012). Association between 1425G/A SNP in PRKCH and ischemic stroke among Chinese and Japanese populations: a meta-analysis including 3686 cases and 4589 controls. Neurosci Lett 506, 55–58.
- Linsel-Nitschke, P., Heeren, J., Aherrahrou, Z., et al. (2010). Genetic variation at chromosome 1p13.3 affects sortilin mRNA expression, cellular LDL-uptake and serum LDL levels which translates to the risk of coronary artery disease. Atherosclerosis 208, 183–189.
- Lip, G.Y., and Tse, H.F. (2007). Management of atrial fibrillation. Lancet 370, 604–618.
- Liu, Y., Sanoff, H.K., Cho, H., et al. (2009). INK4/ARF transcript expression is associated with chromosome 9p21 variants linked to atherosclerosis. PLoS One 4, e5027.
- Lovkvist, H., Olsson, S., Hoglund, P., et al. (2012). A large-sample assessment of possible association between ischaemic stroke and rs12188950 in the PDE4D gene. Eur J Hum Genet 20, 783–789.
- Lusis, A.J., Mar, R., Pajukanta, P. (2004). Genetics of atherosclerosis. Annu Rev Genomics Hum Genet 5, 189–218.
- Macdonald, R.L., Higashida, R.T., Keller, E., et al. (2013). Randomised trial of Clazosentan, an endothelin receptor antagonist, in patients with aneurysmal subarachnoid hemorrhage undergoing surgical clipping (CONSCIOUS-2). Acta Neurochir Suppl 115, 27–31.
- Macdonald, R.L., Kassell, N.F., Mayer, S., et al., and CONSCIOUS-1 Investigators. (2008). Clazosentan to overcome neurological ischemia and infarction occurring after subarachnoid hemorrhage (CONSCIOUS-1): randomized, double-blind, placebo-controlled phase 2 dose- finding trial. Stroke 39, 3015–3021.
- Marian, A.J., and Roberts, R. (2001). The molecular genetic basis for hypertrophic cardiomyopathy. J Mol Cell Cardiol 33, 655–670.
- Marjot, T., Yadav, S., Hasan, N., Bentley, P., and Sharma, P. (2011). Genes associated with adult cerebral venous thrombosis. Stroke 42, 913–918.
- McPherson, R., Pertsemlidis, A., Kavaslar, N., et al. (2007). A common allele on chromosome 9 associated with coronary heart disease. Science 316, 1488–1491.
- Motulsky, A.G., and Brunzell, J.D. (2002). Genetics of coronary atherosclerosis. In The Genetic Basis of Common Diseases (pp 105–126). King, R.A., Rotter, J.I. and Motulsky, A.G., eds. (Oxford, UK: Oxford University Press).
- Newton-Cheh, C., Larson, M.G., Corey, D.C., et al. (2005). QT interval is a heritable quantitative trait with evidence of linkage to chromosome 3 in a genome-wide linkage analysis: the Framingham Heart Study. Heart Rhythm 2, 277–284.
- Nguyen-Tran, V.T., Kubalak, S.W., Minamisawa, S., et al. (2000). A novel genetic pathway for sudden cardiac death via defects in the transition between ventricular and conduction system cell lineages. Cell 102, 671–682.
- Okin, P.M., Devereux, R.B., Howard, B.V., Fabsitz, R.R., Lee, E.T., and Welty, T.K. (2000). Assessment of QT interval and QT dispersion for prediction of all-cause and cardiovascular mortality in American Indians: The Strong Heart Study. Circulation 101, 61–66.
- Palomaki, G.E., Melillo, S., and Bradley, L.A. (2010). Association between 9p21 genomic markers and heart disease: a meta-analysis. JAMA 303, 648–656.
- Pasmant, E., Laurendeau, I., Héron, D., Vidaud, M., Vidaud, D., and Bièche, I. (2007). Characterization of a germ-line deletion, including the entire INK4/ARF locus, in a melanoma-neural system tumor family: identification of ANRIL, an antisense noncoding RNA whose expression coclusters with ARF. Cancer Research 67, 3963–3969.
- Patel, R.S., Eapen, D.J., Zafari, A.M., Vaccarino, V., and Quyyumi, A.A. (2011). Letter by Patel et al. regarding article “Chromosome 9p21 haplotypes and prognosis in white and black patients with coronary artery disease.” Circ Cardiovasc Genet 4, e11; author reply e12.
- Paynter, N.P., Chasman, D.I., Pare, G., et al. (2010). Association between a literature-based genetic risk score and cardiovascular events in women. JAMA 303, 631–637.
- Poort, S.R., Rosendaal, F.R., Reitsma, P.H., and Bertina, R.M. (1996). A common genetic variation in the 3′-untranslated region of the prothrombin gene is associated with elevated plasma prothrombin levels and an increase in venous thrombosis. Blood 88, 3698–3703.
- Post, W., Shen, H., Damcott, C., et al. (2007). Associations between genetic variants in the NOS1AP (CAPON) gene and cardiac repolarization in the Old Order Amish. Hum Hered 64, 214–219.
- Priori, S.G., and Napolitano, C. (2004). Genetics of cardiac arrhythmias and sudden cardiac death. Ann N Y Acad Sci 1015, 96–110.
- Raitakari, O.T., Blom-Nyholm, J., Koskinen, T.A., Kahonen, M., Viikari, J.S., and Lehtimaki, T. (2009). Common variation in NOS1AP and KCNH2 genes and QT interval duration in young adults. The Cardiovascular Risk in Young Finns Study. Ann Med 41, 144–151.
- Rajaraman, P., Melin, B.S., Wang, Z., et al. (2012). Genome-wide association study of glioma and meta-analysis. Hum Genet 131, 1877–1888.
- Rigat, B., Hubert, C., Alhenc-Gelas, F., Cambien, F., Corvol, P., and Soubrier, F. (1990). An insertion/deletion polymorphism in the angiotensin I-converting enzyme gene accounting for half the variance of serum enzyme levels. J Clin Invest 86, 1343–1346.
- Roger, V.L., Go, A.S., Lloyd-Jones, D.M., et al. (2012). Heart disease and stroke statistics—2012 update: a report from the American Heart Association. Circulation 125, e2–e220.
- Sakamoto, Y., Hara, K., Kanai-Azuma, M., et al. (2007). Redundant roles of Sox17 and Sox18 in early cardiovascular development of mouse embryos. Biochem Biophys Res Commun 360, 539–544.
- Samani, N.J., Erdmann, J., Hall, A.S., et al. (2007). Genomewide association analysis of coronary artery disease. N Engl J Med 357, 443–453.
- Schouten, E.G., Dekker, J.M., Meppelink, P., Kok, F.J., Vandenbroucke, J.P., and Pool, J. (1991). QT interval prolongation predicts cardiovascular mortality in an apparently healthy population. Circulation 84, 1516–1523.
- Schunkert, H., König, I. R., Kathiresan, S., et al. (2011). Large-scale association analysis identifies 13 new susceptibility loci for coronary artery disease. Nat Genet 43, 333–338.
- Scott, L.J., Mohlke, K.L., Bonnycastle, L.L., et al. (2007). A genome-wide association study of type 2 diabetes in Finns detects multiple susceptibility variants. Science 316, 1341–1345.
- Sharp, D.S., Masaki, K., Burchfiel, C.M., Yano, K., and Schatz, I.J. (1998). Prolonged QTc interval, impaired pulmonary function, and a very lean body mass jointly predict all-cause mortality in elderly men. Ann Epidemiol 8, 99–106.
- Shen, G.Q., Rao, S., Martinelli, N., et al. (2008). Association between four SNPs on chromosome 9p21 and myocardial infarction is replicated in an Italian population. J Hum Genet 53, 144–150.
- Sherborne, A.L., Hosking, F.J., Prasad, R.B., et al. (2010). Variation in CDKN2A at 9p21.3 influences childhood acute lymphoblastic leukemia risk. Nat Genet 42, 492–494.
- Shete, S., Hosking, F.J., Robertson, L.B., et al. (2009). Genome-wide association study identifies five susceptibility loci for glioma. Nat Genet 41, 899–904.
- Sotoodehnia, N., Isaacs, A., de Bakker, P.I., et al. (2010). Common variants in 22 loci are associated with QRS duration and cardiac ventricular conduction. Nat Genet 42, 1068–1076.
- Splawski, I., Shen, J., Timothy, K.W., et al. (2000). Spectrum of mutations in long-QT syndrome genes, KVLQT1, HERG, SCN5A, KCNE1, and KCNE2. Circulation 102, 1178–1185.
- Splawski, I., Timothy, K.W., Tateyama, M., Clancy, C.E., and Malhotra, A. (2002). Variant of SCN5A sodium channel implicated in risk of cardiac arrhythmia. Science 297, 1333.
- Spooner, P.M., Albert, C., Benjamin, E.J., et al. (2001a). Sudden cardiac death, genes, and arrhythmogenesis: consideration of new population and mechanistic approaches from a National Heart, Lung, and Blood Institute workshop, Part I. Circulation 103, 2361–2364.
- Spooner, P.M., Albert, C., Benjamin, E.J., et al. (2001b). Sudden cardiac death, genes, and arrhythmogenesis: consideration of new population and mechanistic approaches from a National Heart, Lung, and Blood Institute workshop, Part II. Circulation 103, 2447–2452.
- Stecker, E.C., Sono, M., Wallace, E., Gunson, K., Jui, J., and Chugh, S.S. (2006). Allelic variants of SCN5A and risk of sudden cardiac arrest in patients with coronary artery disease. Heart Rhythm 3, 697–700.
- Subramanian, S.V., and Nadal-Ginard, B. (1996). Early expression of the different isoforms of the myocyte enhancer factor-2 (MEF2) protein in myogenic as well as non-myogenic cell lineages during mouse embryogenesis. Mech Dev 57, 103–112.
- Suehiro, T., Morita, T., Inoue, M., Kumon, Y., Ikeda, Y., and Hashimoto, K. (2004). Increased amount of the angiotensin-converting enzyme (ACE) mRNA originating from the ACE allele with deletion. Hum Genet 115, 91–96.
- Suzuki, E., Nagata, D., Kakoki, M., et al. (1999). Molecular mechanisms of endothelin-1-induced cell- cycle progression: involvement of extracellular signal-regulated kinase, protein kinase C, and phosphatidylinositol 3-kinase at distinct points. Circ Res 84, 611–619.
- Takeuchi, F., Yokota, M., Yamamoto, K., et al. (2012). Genome-wide association study of coronary artery disease in the Japanese. Eur J Hum Genet 20, 333–340.
- Tobin, M.D., Kahonen, M., Braund, P., et al. (2008). Gender and effects of a common genetic variant in the NOS1 regulator NOS1AP on cardiac repolarization in 3761 individuals from two independent populations. Int J Epidemiol 37, 1132–1141.
- Utermann, G., Hees, M., and Steinmetz, A. (1977). Polymorphism of apolipoprotein E and occurrence of dysbetalipoproteinaemia in man. Nature 269, 604–607.
- Vergouwen, M.D., Algra, A., and Rinkel, G.J. (2012). Endothelin receptor antagonists for aneurysmal subarachnoid hemorrhage: a systematic review and meta-analysis update. Stroke 43, 2671–2676.
- Visel, A., Zhu, Y., May, D., et al. (2010). Targeted deletion of the 9p21 non-coding coronary artery disease risk interval in mice. Nature 464, 409–412.
- Volcik, K.A., Barkley, R.A., Hutchinson, R.G., et al. (2006). Apolipoprotein E polymorphisms predict low density lipoprotein cholesterol levels and carotid artery wall thickness but not incident coronary heart disease in 12,491 ARIC study participants. Am J Epidemiol 164, 342–348.
- Wald, D.S., Law, M., and Morris, J.K. (2002). Homocysteine and cardiovascular disease: evidence on causality from a meta-analysis. BMJ 325, 1202.
- Wang, B., Guo, Q., Peng, Y., Lu, J., Singh, B., and Hua, B. (2012). Association of AGT M235T and ACE I/D polymorphisms with the risk of ischemic stroke: meta-analysis in Han Chinese population. J Neurol Sci 320, 79–84.
- Wang, L., Fan, C., Topol, S.E., Topol, E.J., and Wang, Q. (2003). Mutation of MEF2A in an inherited disorder with features of coronary artery disease. Science 302, 1578–1581.
- Ward, H., Mitrou, P.N., Bowman, R., et al. (2009). APOE genotype, lipids, and coronary heart disease risk: a prospective population study. Arch Intern Med 169, 1424–1429.
- Watkins, H., Farrall, M. (2006). Genetic susceptibility to coronary artery disease: from promise to progress. Nat Rev Genet 7, 163–173.
- Weng, L., Kavaslar, N., Ustaszewska, A., et al. (2005). Lack of MEF2A mutations in coronary artery disease. J Clin Invest 115, 1016–1020.
- Wu, L., Shen, Y., Liu, X., et al. (2009). The 1425G/A SNP in PRKCH is associated with ischemic stroke and cerebral hemorrhage in a Chinese population. Stroke 40, 2973–2976.
- Xie, F., Chu, X., Wu, H., et al. (2011). Replication of putative susceptibility loci from genome-wide association studies associated with coronary atherosclerosis in Chinese Han population. PLoS One 6, e20833.
- Yamagishi, K., Folsom, A.R., Rosamond, W.D., Boerwinkle, E., and ARIC Investigators. (2009). A genetic variant on chromosome 9p21 and incident heart failure in the ARIC study. Eur Heart J 30, 1222–1228.
- Yang, X.R., Liang, X., Pfeiffer, R.M., et al. (2010). Associations of 9p21 variants with cutaneous malignant melanoma, nevi, and pigmentation phenotypes in melanoma-prone families with and without CDKN2A mutations. Fam Cancer 9, 625–633.
- Yasuno, K., Bakircioglu, M., Low, S.K., et al. (2011). Common variant near the endothelin receptor type A (EDNRA) gene is associated with intracranial aneurysm risk. Proc Natl Acad Sci U S A 108, 19707–19712.
- Yasuno, K., Bilguvar, K., Bijlenga, P., et al. (2010). Genome-wide association study of intracranial aneurysm identifies three new risk loci. Nat Genet 42, 420–425.
- Ye, Z., Liu, E.H., Higgins, J.P., et al. (2006). Seven haemostatic gene polymorphisms in coronary disease: meta-analysis of 66,155 cases and 91,307 controls. Lancet 367, 651–658.
- Zdravkovic, S., Wienke, A., Pedersen, N.L., Marenberg, M.E., Yashin, A.I., and De Faire, U. (2002). Heritability of death from coronary heart disease: a 36-year follow-up of 20,966 Swedish twins. J Intern Med 252, 247–254.
- Zeggini, E., Scott, L.J., Saxena, R., et al. (2008). Meta-analysis of genome-wide association data and large-scale replication identifies additional susceptibility loci for type 2 diabetes. Nat Genet 40, 638–645.
- Zeggini, E., Weedon, M.N., Lindgren, C.M., et al. (2007). Replication of genome-wide association signals in UK samples reveals risk loci for type 2 diabetes. Science 316, 1336–1341.
- Zheng, Z.J., Croft, J.B., Giles, W.H., et al. (2002). State-specific mortality from sudden cardiac death— United States, 1999. MMWR 51, 123–126.
- Zintzaras, E., Rodopoulou, P., and Sakellaridis, N. (2009). Variants of the arachidonate 5- lipoxygenase-activating protein (ALOX5AP) gene and risk of stroke: a HuGE gene-disease association review and meta-analysis. Am J Epidemiol 169, 523–532.