This page has been archived and is no longer updated
Genetic susceptibility to coronary artery disease: from promise to progress
Author: Hugh Watkins
Keywords
Keywords for this Article
Add keywords to your Content
Save
|
Cancel
Share
|
Cancel
Revoke
|
Cancel
Rate & Certify
Rate Me...
Rate Me
!
Comment
Save
|
Cancel
Flag Inappropriate
The Content is
Objectionable
Explicit
Offensive
Inaccurate
Comment
Flag Content
|
Cancel
Delete Content
Reason
Delete
|
Cancel
Close
Full Screen
"� 2006 Nature Publishing Group Coronary artery disease (CAD) is the most common cause of death in industrialized countries and is rapidly increasing in prevalence in developing countries. Since 1990 more people have died worldwide from CAD than any other disease 1 . The age-adjusted mortality that is due to CAD in North America and Western Europe is falling each year, but as these populations age the prevalence of CAD is actually climbing, and this trend is expected to accelerate in the future owing to the burgeoning epi- demic of obesity. In developing countries, the reduction in childhood mortality is expected to change demo- graphic structures, with the result that more adults will be at risk from CAD. CAD has long been known to ?run in families?. A small proportion of cases can be attributed to rare, highly penetrant, monogenic effects, but most are multi- factorial in aetiology, involving numerous environ- mental and heritable risk factors. The identification of specific susceptibility genes will add to our knowledge of the molecular pathophysiology of CAD, refine the identification of high-risk individuals and suggest areas of research for drug discovery. Because modifiable lifestyle risk factors such as smoking make an important contribution in CAD, pharmacological intervention is only one part of any preventive strategy. Nevertheless, the demand for effec- tive drugs is considerable, and some important advances in preventive treatment have already been made. Most notable has been the advent of statins ? drugs that lower low density lipoprotein (LDL) cholesterol levels. It is important to acknowledge the contribution of human genetics to this discovery: the study of a Mendelian disorder, familial hypercholesterolaemia, implicated the LDL receptor (LDLR) in CAD, which advanced our understanding of the cholesterol metabolic pathways that have a key role in CAD pathogenesis. Despite this success, new drugs ? and therefore new drug targets ? are needed. In addition, available markers are not adequate for predicting individual risk of develop- ing CAD, which places an even greater importance on increasing our understanding of CAD pathophysiology. Currently, there is no convincing animal model of CAD in a genetically tractable species. Genetically manipulated mouse models and inbred strains are informative for aspects of atherosclerosis (a key component in the develop- ment of CAD) and for lipid intermediate phenotypes, but not for important aspects of human myocardial infarction, the main clinical outcome of CAD (see below). Therefore, new pathophysiological insights into this key area are likely to depend to a large extent on human genetics. Molecular genetic studies of rare Mendelian diseases have identified a few mutations that cause premature CAD. Genetic studies of intermediate phenotypes that are based on these findings are unravelling the heritable basis of several atherosclerosis risk factors, which are presumed to explain a fraction of an individuals? inher- ited risk for common, multifactorial CAD (see REF. 2 for reviews). However, there is still a large gap in our knowledge of the genes that are involved in susceptibility for most patients with multifactorial CAD. With this in mind, we review recent advances in genetic approaches to understanding CAD. These Department of Cardiovascular Medicine and Wellcome Trust Centre for Human Genetics, University of Oxford, Oxford OX3 9DU, UK. e-mail: hugh. watkins@cardiov.ox.ac.uk doi:10.1038/nrg1805 Published online 7 February 2006 Penetrance The conditional probability that carriers of a specific genotype are affected by an inherited disease. If the penetrance is 100%, all carriers will be affected. Low density lipoprotein (LDL) cholesterol Circulating complexes of apolipoprotein B and cholesterol constitute LDL particles. Raised LDL cholesterol level is predictive of coronary artery disease. Genetic susceptibility to coronary artery disease: from promise to progress Hugh Watkins and Martin Farrall Abstract | Family history is an important independent risk factor for coronary artery disease (CAD), and identification of susceptibility genes for this common, complex disease is a vital goal. Although there has been considerable success in identifying genetic variants that influence well-known risk factors, such as cholesterol levels, progress in unearthing novel CAD genes has been slow. However, advances are now being made through the application of large-scale, systematic, genome-wide approaches. Recent findings particularly highlight the link between CAD and inflammation and immunity, and highlight the biological insights to be gained from a genetic understanding of the world?s biggest killer. NATURE REVIEWS | GENETICS VOLUME 7 | MARCH 2006 | 163 REVIEWS � 2006 Nature Publishing Group Fibrous cap Lipid-rich necrotic core ? NO production Monocyte recruitment Thrombus Internal elastic lamina Proteolysis Apoptosis LDL Matrix proteolysis Process Endothelial dysfunction Inflammation Thrombosis Endothelial activation Apoptosis Oxidized LDL Foam cell Cell-adhesion molecules Endothelial cell Macrophage Cholesterol crystals Fibrin LDLSmooth muscle cell Collagen fibril Platelet Apoptotic cell Atherosclerosis Also known as arteriosclerosis, this is a ?narrowing? or ?hardening? of the arteries, and is a pathological process that involves the progressive expansion of lesions (plaques) that reduce blood flow in arteries. Intermediate phenotypes Biochemical or physiological traits that show quantitative genetic variation. These traits are thought to represent fundamental disturbances of normal cellular processes, thereby conferring risk for complex disease. Myocardial infarction Commonly known as heart attack. This is caused by late-stage atherosclerosis in coronary arteries, involving plaque rupture, coronary thrombosis and heart muscle necrosis. involve genome-wide linkage mapping and large-scale gene-association studies as core human genetics strategies, which are starting to produce exciting results. These have been complemented by comparative genomics stud- ies and mapping of quantitative traits in the mouse. The genes identified through these combined approaches have diverse functions: they include transcription fac- tors that are implicated in vasculogenesis, signalling molecules that are involved in inflammation, innate and adaptive immunity, and novel apolipoproteins, as well as genes for which the physiological roles are as yet unclear. Even with the benefit of hindsight, few of these genes are obvious candidates for CAD. The study of the genetic basis of CAD therefore emphasizes the power of genetics to generate new insights into the molecular basis of complex diseases. The pathobiology of CAD ? an overview The clinical manifestations of CAD are the main com- plications of atherosclerosis ? the underlying process that is also responsible for ischaemic stroke and peripheral vascular disease. Atherosclerosis is a degenerative disease that is characterized by the progressive deposition of lipids and fibrous matrix in the arterial wall (FIG. 1). The underlying pathology reflects a complex series of events, and much has been discovered about the molecular and cellular processes that are involved. The key steps are the loss of the normal barrier function of the endothelium, lipoprotein abnormalities that favour lipid entry (includ- ing high levels of LDL and/or low levels of high density lipoprotein (HDL) cholesterol), recruitment of monocytes and lymphocytes to the artery wall, and proliferation of smooth muscle cells (see REF. 3 for a review). Although this process reflects global changes in endothelial function, the clinically important lesions are focal and tend to occur at specific sites (FIG. 2). Early lesions appear as fatty streaks, which are characterized by lipid-containing foam cells. These later develop into atherosclerotic plaques, which consist of a lipid core and an overlying fibrous cap. Several variables govern the processes that precipitate clinical events and ultimately influence disease outcome: plaques can remain silent, can progressively narrow the lumen, which restricts flow (producing angina), or can precipitately occlude vessels through acute thrombosis, which leads to myocardial infarction. The clinical entity of CAD therefore results from a particularly complex disease process, with scope for a large degree of heterogeneity among patients who have apparently similar clinical presentations and, as a corollary, extensive genetic heterogeneity. Figure 1 | Cellular processes in the development of an atherosclerotic plaque. Low density lipoprotein (LDL) cholesterol enters dysfunctional endothelium (which is damaged by smoking or diabetes, for example, and this is reflected by decreased nitric oxide (NO) production) and is oxidized by macrophage and smooth muscle cells. Release of growth factors and cytokines, and upregulation of adhesion molecules, attracts further monocytes. Foam cells (arising from lipid-laden macrophages) accumulate and smooth muscle cells proliferate, which results in the growth of the plaque. Inflammatory cell infiltrate, smooth muscle cell death through apoptosis, and matrix degradation through proteolysis (by matrix metalloproteinases ? MMPs) generate a vulnerable plaque with a thin fibrous cap and a lipid-rich necrotic core. Plaque rupture can cause thrombosis which might be sufficient to cause vessel occlusion. Modified with permission from Nature Reviews Drug Discovery REF. 84 ? (2004) Macmillan Publishers Ltd. REVIEWS 164 | MARCH 2006 | VOLUME 7 www.nature.com/reviews/genetics � 2006 Nature Publishing Group Normal artery ?Vulnerable? plaque ? Thin fibrous cap ? Large lipid pool ? Many inflammatory cells Endothelium Adventitia Smooth muscle cells Intima Media Plaque rupture and thrombosis Right coronary artery Lef_t coronary artery Acute myocardial infarction Thrombotic occlusion ?Stabilized? plaque Healed ruptured plaque; progressive vessel stenosis and angina Early atheroma Fibrous cap Linkage mapping A method for localizing genes that is based on the co-inheritance of genetic markers and phenotypes in families over several generations. Genetic basis of CAD as a complex trait Molecular genetic studies of rare, Mendelian forms of the disease have identified several mutations that cause premature CAD (TABLE 1), many of which affect levels of LDL and HDL cholesterol. However, common, multi- factorial CAD has also been shown to have a heritable basis. Data collated over 36 years in a cohort of 20,966 Swedish twins have shown that the heritability of fatal CAD events is 57% and 38% for men and women, respectively 4 . The influence of genes is evident across the age-range 36?86 years, declining slightly in older men (but not women), possibly owing to the increasing variance in environmental factors with age 5 . This has implications for the selection of affected individuals for genetic studies (BOX 1). The extent to which the familial clustering of CAD can be explained by heritable quantitative variation in classical risk factors ? such as LDL and HDL cholesterol and blood pressure (BOX 2) ? is contentious 6,7 . In the Framingham Heart Study, a large community-based prospective study, parental history of premature coronary, cerebrovascular or peripheral vascular disease (after adjustment for age and smoking) was a strong risk factor for CAD (with odds ratios (OR) of 2.4 for men and 2.2 for women). After adjustment for several classical risk factors, family history continued to predict an excess risk (OR = 2.0 for men and 1.7 for women) 8 . Similar conclusions were drawn from the PROCAM study, in which a family history of myocardial infarction was found to be an independent predictor of CAD risk 9 . The INTERHEART study 10 confirmed that family history of CAD (after adjustment for age, sex, smok- ing and geographical region) is a risk factor (OR = 1.55); after adjustment for nine classical risk factors, the risk was reduced only slightly (OR = 1.45). This indicates that other genetic factors have a role, as well as those that influence known risk factors such as cholesterol levels and blood pressure. These findings strengthen the arguments for genome- wide linkage and gene-association studies to identify CAD-susceptibility genes. These approaches do not depend on previous information about biochemical pathways or pathophysiology. They therefore comple- ment candidate-gene studies, which concentrate on well-known risk factors and biological pathways, such as cholesterol levels, metabolism and transport. Genetic architecture of CAD susceptibility Although the total genetic contribution to CAD risk can be quantified, the determination of the size and number of contributing effects is impossible without identifying all CAD-susceptibility genes. Nevertheless, the genetic complexity of CAD susceptibility is esti- mated to be considerable. First, the list of known risk factors is long and many of these traits have their own complex genetic basis. Second, CAD is rare before the age of 50 years, and is therefore confined to aged populations. Consequently, it can be considered to be an entirely modern disease that will not have been subject to direct selective pressure as it is unlikely to have an effect on reproductive success. Variants that confer either susceptibility or protection for CAD might therefore have evolved neutrally in the past, and so could be present at a wide range of frequencies 11 . Only if these variants have pleiotropic effects that affect susceptibility to more ?ancient? diseases (for example, infectious disease) would alleles with large effects on susceptibility to CAD be expected to be present at a detectable frequency. Within this anticipated complexity, both common and rare variants are likely to have a part in CAD Figure 2 | Stages in atherosclerosis and the clinical outcomes. Atheromatous plaques progress by sequential episodes of plaque rupture and subsequent healing. As a plaque encroaches on the luminal area, blood flow becomes limiting and flow cannot increase on demand; in the heart this manifests as chest pain on exertion, which is labelled as chronic stable angina. Alternatively, rupture or erosion of a vulnerable plaque (one that has a thin fibrous cap) can precipitate an acute occlusion owing to the formation of a thrombus or blood clot, which manifests clinically as an acute coronary syndrome. Partial or transient occlusion, often accompanied by downstream embolization, results in a partial-thickness myocardial infarction, which can be confirmed by the release of troponin into the blood. Complete occlusion results in a full-thickness myocardial infarction with subsequent scar formation. The equivalent event after a plaque rupture in the carotid artery is an ischaemic stroke. Modified with permission from Nature REF. 3 ? (2002) Macmillan Publishers Ltd. REVIEWS NATURE REVIEWS | GENETICS VOLUME 7 | MARCH 2006 | 165 � 2006 Nature Publishing Group Gene-association study An experimental design to test whether genetic markers predict the risk of developing a disease. Quantitative trait A biological trait that shows continuous variation (such as height) rather than falling into distinct categories (such as diabetic or healthy). The genetic basis of these traits generally involves the effects of multiple genes and gene? environment interactions. Examples of quantitative traits are body mass index, blood pressure and blood lipid levels. Apolipoproteins Proteins that when complexed with lipids form lipoprotein particles that are involved in the transport and metabolism of cholesterol and triglycerides. Ischaemic stroke A type of cerebrovascular accident that results from occlusion of cerebral arteries owing to atherosclerosis and leading to a neurological deficit or death. High density lipoprotein (HDL) cholesterol Cholesterol that is contained in HDL particles; it is involved in reverse cholesterol transport and is protective against coronary artery disease. susceptibility. However, as with any complex trait, there is some uncertainty as to which will be more important. A recent candidate-gene study to identify variants that are associated with low plasma levels of HDL choles- terol indicates that multiple rare alleles, each of which is associated with a large phenotypic effect, jointly make a substantial contribution to the population distribution of this quantitative trait 12 . This is at odds with the common disease/common variant (CDCV) model for the genetic architecture of common disease. In addition, a population study of consanguinity and the prevalence of complex diseases inferred an important role for del- eterious recessive alleles in CAD, a conclusion that is also in conflict with the CDCV model 13 . These findings raise potential concerns about the ability of association studies (which have low power to detect rare variants) to be effective for CAD. However, in contrast to this, a review of 25 reported associations supports the importance of common alleles in common diseases, including CAD 14 . Furthermore, several CAD genes have been discovered in genome-wide studies that are consistent with the CDCV model, as highlighted in the following sections. Positional cloning: linkage-based family studies A role for MEF2A? Families with Mendelian CAD with- out an associated dyslipidaemia syndrome are not com- monly described. Nevertheless, Wang and colleagues 15 reported the positional cloning of an autosomal dominant mutation with a role in CAD on the basis of a genome-wide linkage screen of a single multiplex family. Fine mapping of the locus was limited, as only 14 informative meioses (assuming complete penetrance) were available for study, and a linked region containing at least 93 genes was identified. The authors focused their attention on the myocyte enhancer factor 2A (MEF2A) gene, which encodes a transcription factor, as its embryonic expression pattern indicated a role in vasculogenesis. A 21-bp in-frame deletion co-segregated with disease in this family, and was absent in 119 indi- viduals who had normal angiograms. In vitro studies indicated that the deletion affects gene function. In a subsequent mutational survey of unrelated cases and normal controls, three rare variants that encode non- synonymous mutations of MEF2A were associated with CAD 16 . One of these variants, Pro279Leu, has also been associated with CAD risk in a Spanish cohort 17 . However, the link between MEF2A and CAD remains controversial. Weng and colleagues 18 carried out a resequencing survey, focusing on exonic vari- ation in this gene among premature CAD cases. This identified five non-synonymous variants of which only one, Ser360Leu, was not found in asymptomatic control samples. Furthermore, the 21-bp deletion identified by Wang and colleagues 15 was identified in the control population and failed to co-segregate in a CAD family with a limited number of informa- tive meioses 18 . These data do not support the MEF2A deletion as a high-penetrance CAD allele, but might still be compatible with a role for MEF2A variants in susceptibility for CAD. Alternatively, as other com- pletely linked sequence variants must exist within the originally reported interval, an as yet unidentified gene could explain the original linkage finding. A success for linkage: evidence for the role of leukotrienes. Genetic linkage studies have been heavily criticized as a first step in the positional cloning of susceptibility genes for common disease, as their results are notori- ously difficult to replicate for low-penetrance mutations with small effects. Progress from ?linkage-to-locus? also remains a formidable technical challenge, despite the available genomic resources, and there are relatively few examples of complex disease loci that have been cloned in this way. However, one example of the suc- cessful use of such a strategy was the identification of ALOX5AP (which encodes 5-lipoxygenase activation protein (FLAP)) as a susceptibility gene for myocardial infarction and stroke 19 . In a linkage study of Icelandic families, moderate LOD scores (1.5?2.9) were obtained for a locus that was identified for myocardial infarction and stroke when the data were divided on the basis of sex and age-of- onset. However, none of these statistics fulfilled the stringent criteria for genome-wide significance (which is not unusual for linkage studies of complex diseases). Going against the current fashion of using SNPs, 120 microsatellite markers were then used to scan the 8-Mb linked region for evidence of a genetic association with myocardial infarction, leading to the identification of ALOX5AP. The FLAP protein activates arachidonate 5-lipoxyge- nase (ALOX5), which catalyses the biosynthesis of the leukotriene LTA4 ? a pathway that has been independ- ently implicated in the pathogenesis of atherosclerosis in humans 20 (FIG. 3). Furthermore, mouse alleles of Alox5 that are derived from certain inbred strains provide a high degree of protection from atherosclerosis on a high Table 1 | Mendelian diseases that involve premature coronary artery disease Disease Genes Effect of mutations Prevalence OMIM number Familial hypercholesterolaemia LDLR Defective binding of LDL by receptor 1/500 #143890 Familial defective APOB APOB Reduced binding affinity of APOB to LDLR 1/3,250 #144010 Sitosterolaemia ABCG5, ABCG8 Increased absorption of plant sterols Rare #210250 Autosomal recessive hypercholesterolaemia ARH Defective endocytosis of LDLR Rare #603813 APOA1 deficiency APOA1 Deletion or loss-of- function mutation that leads to very low HDL Rare #107680 Tangier disease ABCA1 Impaired cholesterol efflux in macrophages (foam cells) Rare #205400 Homocystinuria CBS Homocysteine increases thrombotic tendency Rare #236200 ABCG, ATP-binding cassette, subfamily G, member 5; APO, apolipoprotein; ARH, autosomal recessive hypercholesterolaemia protein; CBS, cystathionine ?-synthase; HDL, high density lipoprotein; LDLR, LDL receptor. REVIEWS 166 | MARCH 2006 | VOLUME 7 www.nature.com/reviews/genetics � 2006 Nature Publishing Group Monocytes Circulating phagocytic white blood cells that develop into macrophages when they enter tissues. Lymphocytes White blood cells that are responsible for humoral and cellular immunity. Foam cells Macrophages that are laden with lipid (predominantly cholesterol), and that have differentiated from circulating monocytes. Chronic stable angina Chest pain that is experienced on exertion and results from coronary artery stenosis. Prospective study design An experimental design that is used in epidemiology studies. A cohort of healthy subjects are followed up over a prolonged period to determine the association of risk factors with disease. Odds ratio A statistical measure of the strength of a risk factor. Pleiotropy The phenomenon in which a single gene is responsible for several phenotypic effects. Common disease/common variant model A theory of causation of common, complex disease in which ?ancient? common genetic variants, some of which might have been selected for, are associated with modest disease risks. Dyslipidaemia Abnormality in, or abnormal levels of, lipids and lipoproteins in the blood. Multiplex family A family in which multiple members are affected by an inherited disease. Angiogram A radiological investigation that is used to clinically evaluate the severity of coronary atherosclerosis. LOD score The base 10 logarithm of a likelihood ratio statistic; this is a preferred statistic for linkage analysis that indicates the strength of linkage. cholesterol (Ldlr ?/? knockout) background; similar find- ings were obtained when heterozygous Alox5 knockout mice were crossed with Ldlr ?/? animals 21 . Identifying the causal variant in ALOX5AP has been more difficult. In the Icelandic study 19 , a survey of SNPs in the affected families revealed many ALOX5AP variants that were in strong linkage disequilibrium with the disease phenotypes, but none of these had obvi- ous potential effects on function (for example, none caused non-synonymous substitutions). In the Icelandic cohorts, a 4-SNP haplotype (HapA) showed strong association with myocardial infarction, but this was not confirmed in a cohort of British patients with myocardial infarction; however, further analysis implicated a differ- ent 4-SNP risk haplotype (HapB) in the British cohort. These different risk haplotypes in populations with dif- fering genealogies might indicate allelic heterogeneity for CAD susceptibility. The Icelandic HapA risk haplotype was associated with increased neutrophil release of leukotriene B4 (LTB4), which is synthesized from LTA4, indicating that the ALOX5AP variants might be pro-inflamma- tory. LTB4 activates monocytes, which migrate through arterial endothelium and differentiate into macrophages that take up lipid to become foam cells, a key early event in atheroma development (FIG. 1). Blockade of the LTB4 receptor has been shown to reduce the amount of lipid accumulation, monocyte infiltration and lesion size in Ldlr ?/? or ApoE ?/? (apolipoprotein E) knockout mouse models of atherosclerosis (which are susceptible to atherosclerosis owing to dyslipidaemia) 22 . Therefore, the initial linkage studies that indicated a role for ALOX5AP in CAD susceptibility provided the first genetic evidence for the involvement of the leukotriene pathway in this disease, a finding that has more recently been supported by candidate-gene association studies 23 (see below). The outlook for further linkage studies. Several genome- wide linkage screens of varying sizes have been published for CAD, mostly using families that are chosen on the basis that they include survivors of early myocardial inf- arction (to enrich for a genetic component). Apart from the study that led to the identification of ALOX5AP, these screens ? including the larger ones ? have pro- duced loci but have not yet identified genes 27?30 . Putative loci have not been replicated, but this should not be seen as surprising (or indeed worrying) as none of the studies has had the statistical power to replicate a specific locus. Rather, the signal-to-noise ratio is such that these studies have low power to detect individual susceptibility genes, and so each will tend by chance to identify a different selection. For the same reason, little is likely to be gained by pooling or meta-analysis 31 of available studies unless larger samples are studied compared with those that initially identified specific loci. In the future, putative loci that are implicated from genome-wide linkage scans are likely to inform selection of regions for systematic large-scale gene-association studies. Candidate-gene association studies Candidate-gene studies build on existing knowledge by investigating variation in genes that are already implicated in the pathophysiology of a disease. Studies are motivated by various types of evidence (which can also be of varying quality). For example, studies of apolipoproteins, which have well-understood roles in Box 1 | Design considerations in genetic studies of coronary artery disease Genetic enrichment The multifactorial liability threshold model predicts that individuals who are affected by coronary artery disease (CAD) who have low risk profiles in terms of lifestyle factors carry greater genetic loads and therefore will be more informative in genetic studies. Some researchers have therefore biased their recruitment strategy towards younger patients. This ?best practice? is supported by a recent analysis of the Framingham Heart Study using a prospective study design to assess parents and offspring 8 . Parental premature coronary, cerebrovascular or peripheral vascular disease (in fathers aged <55 years and mothers <65 years) was found to be a stronger risk factor than non-premature disease (odds ratio (OR) = 2.0 versus 1.5 for younger versus older fathers and OR = 1.7 versus 1.1 for mothers). However, the magnitude of the genetic enrichment in young patients seems modest (particularly in men), so investigators should perhaps be wary of linkages and gene associations that seem to be confined to cohorts of young patients. Genetic enrichment can also be achieved in case?control studies by enrolling individuals with first-degree relatives who are affected as cases ? for example, members of an affected sibling pair. Clinical collections for association studies Case?control and family-based studies provide alternative experimental designs for detecting gene associations; the latter design is usually discounted for late-onset diseases, as parental DNA samples are expected to be impractical to collect in sufficient numbers. But case?control studies for late-onset diseases such as CAD are not without their technical problems. For reasons of genetic enrichment as described above, young cases are likely to be preferentially collected. As atherosclerosis is endemic to the Western world, all controls ? even in the absence of clinical signs or symptoms ? will carry some lesions. Screening of healthy controls using invasive diagnostic tests (for example, coronary angiography) is currently infeasible, so it might be tempting to collect aged controls who have no symptoms and will hopefully carry relatively low genetic loads. But comparison of young cases (for example, those who are <55 years old) with aged (for example, >85 years old) controls introduces potential confounding sources. Examples of these are differential survival that is due to genes that are unrelated to CAD, genetic drift and mismatching for potential covariates. Furthermore, even low levels of population stratification might be problematic when interpreting the results of case?control studies, despite strategies to reduce confounding 74 . It therefore seems prudent to re-examine the prospects for collecting families who are suitable for gene-association studies 50 to circumvent the problems of control misspecification and admixture. REVIEWS NATURE REVIEWS | GENETICS VOLUME 7 | MARCH 2006 | 167 � 2006 Nature Publishing Group Microsatellites Short (2?5 nucleotides) direct sequence repeats, which are often highly polymorphic and are therefore useful markers in linkage analysis. Leukotrienes Regulators of allergic and inflammatory reactions that are produced from arachidonic acid by the lipoxygenase pathway in white blood cells. Genetic drift The (stochastic) change in frequency of genetic markers across generations. atherosclerosis, have identified QTLs that are associated with CAD (for example, lipoprotein A (REF. 2)). However, human geneticists are often intrigued by less obvious connections, hoping that these will open up new oppor- tunities for therapy ? a half-way house between classical candidate-gene studies of validated biological pathways and genome-wide positional cloning projects that have the potential to identify new genes and pathways. An overlap in inflammatory responses that was observed in chronic infection and atherosclerosis prompted investigation into the possible role of infec- tious agents that might trigger plaque formation 32 . This led to interest in candidate genes for CAD that are involved in innate immunity. Among these, CD14, the membrane-bound glycoprotein receptor for bacterial lipopolysaccharide (LPS), was the first gene to be implicated in CAD susceptibility. A promoter poly- morphism (Cys260Thr) that has been proposed to be transcriptionally active 33 has been repeatedly studied in case?control studies 34 , although no consistent association has emerged. Another component of innate immunity signalling, toll-like receptor 4 (TLR4) ? which is an essential co-receptor for CD14 ? has also been implicated in CAD susceptibility 34,35 . TLR4 recognizes exogenous pathogen-associated molecular patterns (PAMPS) (for example, LPS), as well as endogenous ligands, and acti- vates inflammatory cells through the NF-?B pathway. A common non-synonymous polymorphism (Asp299Gly) in TLR4 has been associated with lower levels of sev- eral biomarkers of inflammation and a reduction in carotid-artery intima-media thickness (IMT) 36 . The reduction of CAD risk that is associated with this poly- morphism 37 was also evident in studies of acute coronary syndrome and in a subgroup of men who were treated with statins 38 . The hypothesis that TLR4 provides a link between innate immunity, inflammation and athero- sclerosis was further supported by a study of Tlr4/ApoE double-knockout mice 39 . These mice showed a modest reduction in the size of atherosclerotic plaques when compared with ApoE single-knockout animals, which are susceptible to atherosclerosis owing to dyslipidaemia. However, a large association study of this polymor- phism showed no difference in inflammatory markers in carriers of the TLR4 299Gly variant, and even showed a modest increased risk of myocardial infarction 40 . In addition, a study of coronary artery stenosis showed no evidence of association with this variant 41 . Such inconsistent results might be explained by differential pathogenic mechanisms in myocardial infarction (where plaque rupture leads to coronary occlusion) and other CAD phenotypes in which plaques can be large but rela- tively stable 40 . Alternatively, it might simply reflect the familiar life cycle of gene-association studies for complex diseases, in which small genetic effects at the limits of detection are difficult to replicate 14,42 . Two successes of candidate-gene association stud- ies in the identification of CAD susceptibility have provided further evidence for a role of components of the leukotriene pathway, following on from the link- age studies described above that identified the role of ALOX5AP. In the first of these, promoter variants in the ALOX5 gene were associated with increased carotid- artery IMT, a preclinical marker of atherosclerosis 23 . In this study, higher dietary levels of arachidonic acid (an omega-6 polyunsaturated fatty acid) increased carotid IMT, whereas fish-derived omega-3 fatty acids had the opposite effect, indicating a gene?environment interaction. Interestingly, C-reactive protein (CRP) levels were also associated with the ALOX5 variation in this study 23 . CRP is a non-specific biomarker of inflammation and a moderately strong predictor of CAD risk 24 (BOX 2), and is thought to have a direct role in plaque development 25 . A phase II clinical trial of a compound that inhibits FLAP is underway to assess its effects on biomarkers of inflammation in cardiovascular disease 26 . Other components of the leukotriene pathway might now be considered as plausible candidate genes for CAD. The LTA4H gene codes for the leukotriene A4 hydrolase protein, the metallopeptidase of which catalyses the rate- limiting step in LTB4 biosynthesis. Genetic variation in Box 2 | Intermediate phenotypes for coronary artery disease Classical risk factors for coronary artery disease Many of the known risk factors for development of coronary artery disease (CAD) are themselves genetically determined quantitative traits. For some, a causal pathological role can be considered proven: for example, both total and low density lipoprotein (LDL) cholesterol levels. High density lipoprotein (HDL) cholesterol levels are also strongly heritable, but confer protection from CAD. Other genetically influenced traits, which are themselves disease entities, also increase the risk of CAD, including high blood pressure, type II diabetes and metabolic syndrome. Measuring these risk factor phenotypes in family-based and case?control studies is helpful in understanding the genetic basis of these traits and exploring gene?environment interactions. These parameters are also important as covariants in analysis of risk. Furthermore, in hypothesis-free systematic linkage studies, measuring these phenotypes can help the prioritization of candidate genes within a linkage interval. ?New? plasma markers and markers of inflammation In addition to LDL and HDL cholesterol levels, a growing number of other plasma factors have been identified that are associated with CAD risk. However, there is often uncertainty whether these are causal links or associations that are attributable to unknown confounders. Many of these traits are also strongly genetically influenced: for example, lipoprotein Lp(a) levels are up to 90% heritable, and homocysteine and fibrinogen levels are also strongly genetically influenced. More recently, molecules that are involved in inflammation have been similarly implicated, notably C-reactive protein and circulating adhesion molecules. Combined genetic and intermediate phenotype analysis within single, large study groups provides a means of comparing the risk that is attributable to genetically determined life-long variations in the measured intermediate phenotype with the expected increment in CAD risk that is based on observational epidemiological studies. This approach, sometimes referred to as ?Mendelian randomization?, could potentially discriminate between causal relationships and those that are due to confounding or ?reverse causality? 75 . Vascular phenotypes Component phenotypes of atherosclerosis have been measured in an effort to define subsets of disease mechanisms. These can be expected to reflect the influence of a smaller number of genes that should therefore have a larger effect size and be easier to identify. These phenotypes also provide surrogate end-points in longitudinal studies and allow the classification of patients into more homogenous groups. Well-validated quantitative vascular intermediate phenotypes include carotid intima-medial thickness, which shows clear heritability. More recently proposed phenotypes aim to measure earlier components of the atherosclerotic process, and these traits also seem to have a substantial genetic component. Examples of these traits are non-invasive measures of endothelial function or arterial stiffness 76?77 . REVIEWS 168 | MARCH 2006 | VOLUME 7 www.nature.com/reviews/genetics � 2006 Nature Publishing Group Inflammation; primary cytokines: TNF? and IL1 Upregulation of cytokines and adhesion molecules in endothelium Leucocyte infiltration Atherosclerosis Thrombosis Known monogenic and QTL influences Known monogenic and QTL influences Lymphotoxin-? variants; galectin 2 variants that alter lymphotoxin-? levels OX40L co-stimulatory molecule Pro-atherogenic lipoprotein and lipid changes FLAP Putative innate immunity susceptibility alleles TLR4 and CD14 Leukotrienes Hyperlipidaemia-dependent inflammation and macrophage entry High LDL Low HDL Recently implicated genes that are involved in lipid metabolism: for example, APOA5 and USF1 Plaque rupture Prothrombotic state CRP release from liver ? Population stratification Populations that are composed of subgroups with different ancestries. If the frequency of disease varies across the subpopulations then spurious (false-positive) gene associations can be detected. Linkage disequilibrium The non-random association of tightly linked genetic markers, or of genetic markers with disease. Neutrophils White blood cells that are involved in chemotaxis and phagocytosis. Macrophages Mononuclear phagocytic cells that are found in tissues; they are derived from circulating monocytes. Lipopolysaccharide (Also known as endotoxin.) A molecule with lipid and complex sugar moieties that is used by gram-negative bacteria to evade host immune defences. LTA4H has recently been shown to be associated with CAD (REF. 83). The strength of the association was mod- est in patients of European descent, but much more striking in African-American patients, which indicates that factors that are specific to ancestry groups interact to confer LTA4H susceptibility. These results confirm the importance of genetic variation in inflammatory pathway genes for CAD pathogenesis. Altogether, despite some successes with replicated findings 14 , single-gene association studies remain prob- lematic, with most being poorly reproducible. Because only a small proportion of variants will have significant effect sizes, these studies have a low a priori likeli- hood of identifying the important variants. However, as small-scale studies are easy to carry out they have become numerous; some, by chance, yield a false- positive association and there is evidence of publication bias in the literature that might enrich for false posi- tives 43 . For these reasons the field is moving towards sys- tematic testing of large numbers of candidate genes or anonymous variants. Large-scale, systematic, gene-association studies Although a handful of positional cloning projects have successfully used linkage in first-pass mapping surveys of complex human diseases (for example, the study that identified ALOX5AP (REF. 19)), the ?smart money? is now being invested in large-scale gene-association studies 44?45 . Unlike in candidate-gene studies, where small numbers of variants that have been implicated in disease are tested for association, these studies look for association with an unbiased selection of variants that are located throughout the genome. This shift acknowledges the limitations of genetic linkage to reliably map genes that have only a modest effect on susceptibility 46 . Large-scale gene-association studies provide much greater power to identify these variants and, if enough markers can be typed at a sufficient density, this power can theoretically be harnessed to search systematically for susceptibility genes. These studies are then free from the biological assumptions that limit candidate-gene association stud- ies. The formidable practical problems of undertaking such genome-wide association studies are being eased by The SNP Consortium and The International HapMap Project, which aim to reduce the number of variants that need to be genotyped to cover the whole genome, and by advances in genotyping technologies. Both case?control and family-based studies can be exploited to search for association; case?control series are easier to assemble, but there are issues ? both general to complex traits and specific to CAD ? that confer some advantages to family-based studies (BOX 1). In a pioneering study of nearly 93,000 predominantly gene-based SNPs in a Japanese case?control cohort, poly- morphisms in the lymphotoxin-? (LTA) gene, which encodes a member of the TNF ligand family, were found to be associated with susceptibility to myocardial infarc- tion 47 . This association has been replicated in one of two other Japanese cohorts 48,49 , as well as a study of survivors of myocardial infarction that are of white European ancestry 50 . A non-synonymous SNP (Thr26Asn) in the LTA haplotype block that was associated with CAD risk in these studies increases the transcriptional induction of the adhesion molecules VCAM1 (vascular-cell adhesion Figure 3 | Lipid metabolism, inflammatory and thrombotic pathways that are involved in coronary artery disease. Possible points of action of recently implicated susceptibility genes and the proteins they encode are indicated (green indicates pathophysiological changes; orange indicates genetic influences). APOA5, apolipoprotein A5; CD14, the membrane-bound glycoprotein receptor for bacterial lipopolysaccharide; CRP, C-reactive protein; FLAP, 5-lipoxygenase activation protein; HDL, high density lipoprotein; IL1, interleukin 1; LDL, low density lipoprotein; OX40L, OX40 ligand; TLR4, toll-like receptor 4; TNF?, tumour necrosis factor ?; USF1, upstream transcription factor 1. REVIEWS NATURE REVIEWS | GENETICS VOLUME 7 | MARCH 2006 | 169 � 2006 Nature Publishing Group Metabolic syndrome The occurrence of hyperinsulinaemia, glucose intolerance, dyslipidaemia, hypertension and obesity in an individual. Pathogen-associated molecular patterns Molecules shared by multiple prokaryotic pathogens that are relatively invariant (for example, lipopolysaccharide and flagellin). NF-?B pathway The signal transduction pathway that is triggered by the inducible nuclear transcription factor NF-?B. It has a central role in immunological processes. Carotid artery The large artery in the neck that carries blood from the heart to the brain. molecule 1) and SELE (soluble selectin E), which have both been implicated in atherosclerosis and are believed to facilitate neutrophil recruitment (FIG. 1). A second, anonymous SNP in intron 1 (252A>G) was associated with increased in vitro transcriptional activation of LTA 47 , which is consistent with other expression stud- ies 51,52 . Both alleles are present on the haplotype that was associated with susceptibility to myocardial infarction, which might be mediated through a pro-inflammatory mechanism that leads to plaque rupture. LTA was shown to be expressed in fatty streak lesions in the aortic sinus of mice that were fed an atherogenic diet (a model of early atheromatous plaques) and LTA ?/? knockout animals showed smaller lesions 53 . Building on this success, Ozaki and colleagues 54 used the Escherichia coli two-hybrid system and phage display to identify proteins that interact with LTA. This identified galectin 2 (LGALS2), a ?-galactosidase-binding lectin that has no known physiological function. LGALS2 SNPs that were identified through resequencing were tested for disease association in a case?control study of myo- cardial infarction, and an intronic SNP (3279C) showed a convincing association with myocardial infarction. In vitro transcription assays indicated that this allele is roughly twice as active as the 3279T allele. LGALS2 was shown to be co-expressed with LTA in smooth muscle cells and macrophages in atherosclerotic plaques, and also to colocalize with ?- and ?-tubulin in a pattern that is consistent with a role in intracellular protein traffick- ing, and so potentially relating this allele to altered LTA secretion. These intriguing findings of a possible role of LGALS2 in CAD susceptibility show how large-scale association studies and functional genomics approaches can be used together to indicate entirely unanticipated molecular mechanisms for complex disease. Mouse genetic studies As discussed above, although dietary and transgenic manipulation can produce a mouse model of accelerated atherosclerosis, there are not yet any convincing models of the fulminating event of plaque rupture with throm- botic occlusion 55 . However, researchers have gained insights by studying mouse models of atherosclerosis susceptibility, or by applying comparative genomic strat- egies that depend on the conservation of physiological and biochemical pathways in mammalian species. A further link to immunological pathways. Athero- sclerosis 1 (Ath1) is a QTL on mouse chromosome 1 that renders C57BL/6 mice susceptible to diet-induced atherosclerosis, whereas it has the opposite effect in C3H/He mice 56 . One of the 11 known genes in the linked region ? Tnfsf4 (tumour necrosis factor (ligand) superfamily, member 4), which encodes the ligand of the OX40 receptor ? was further implicated by knock- out of this gene, which resulted in smaller atheroscle- rotic lesions than those seen in control mice. The OX40 ligand is expressed on activated B cells, endothelial cells and macrophages. It generates co-stimulatory signals in these cells by interacting with OX40 on T lymphocytes to enhance their proliferation and differentiation, and might therefore contribute to the accumulation of these cells in atherosclerotic lesions. In two independent, moderately sized patient populations, the less com- mon allele of one TNFSF4 SNP was associated with an increased risk of myocardial infarction in women but not in men, which indicates a gender-specific effect 56 . Although the human data are still preliminary, these find- ings are notable because they indicate a candidate gene for human disease that might be particularly amenable as a drug target through receptor blockade. New insights into dyslipidaemia. A high level of circulat- ing triglycerides is recognized as a moderate risk factor for CAD 57 and has long been associated with genetic variation in the APOA1?APOC3?APOA4 gene cluster 58 . APOC3 is known to downregulate triglyceride hydroly- sis, but no genetic variation that has been observed in APOC3 has explained the quantitative variation seen 59 . Research in this field was boosted by a comparative sequence analysis of the mouse and human genomes, which revealed an undiscovered member of this cluster, APOA5 (REF. 60). Knockout and transgenic experiments showed that APOA5 expression and triglyceride levels were inversely correlated in mice 60 . Box 3 | Combined genetic and genomic approaches Much excitement surrounds genomic tools that provide systematic surveys of an intact biological system, such as expression profiling, proteomics or metabolomics. Nonetheless, these tools are simply descriptors ? albeit efficient and comprehensive ones ? without the power of genetics to determine causality. However, recent data from crosses in model organisms show that genetic and genomic approaches can be combined ? for example in studies of the genetics of gene expression. In this approach, measurements from genomic analyses ? such as transcript levels ? are treated as quantitative traits to be analysed genetically. So, when evidence is found for segregation with anonymous genetic markers, heritability can be inferred and the genetic factors that determine the genomic trait in question can be identified 78,79 . Therefore, instead of merely describing changes in expression that might be independent or reactive, causal associations between gene expression and disease are defined. In addition, genetic linkages that are associated with levels of particular transcripts (?expression QTLs?) are probably closer to the underlying metabolic and regulatory pathways than would be any conventional quantitative trait (which might reflect the summation of several of these pathways). Significant recent progress has been achieved in rodent models ? for example, in exploring previously mapped conventional QTLs such as a known locus for cholesterol level 81 . Similarly, variation in 5-lipoxygenase (the gene implicated in human CAD by positional cloning 19 ) has recently been shown to also underlie previously identified mouse QTLs for body fat, lipid levels and bone density 82 . The fundamental molecular insights that have been obtained have also been enlightening; for example, the expected modulation of single genes by adjacent cis-acting variation has been shown to be overshadowed by more extensive trans-acting loci 80 . Many of the latter modulate many genes and, surprisingly, reflect many diverse classes of genes, rather than transcription factors alone. Moreover, these trans-acting expression QTLs seem to cluster in ?hubs? at given chromosomal loci. Combined genetic and genomic approaches, particularly mapping of expression QTLs, show great promise for studies of complex traits. However, these approaches have yet to be validated in natural outbred populations. If genetical genomics can similarly be exploited in human populations, then a surge of new information about quantitative heritable variation should accelerate research into the genetic basis of complex traits and diseases. But at present this remains uncertain. These approaches also tend to be limited to genomic analyses carried out on readily accessible samples that are available in large numbers of individuals; success stories so far relate to studies on blood or blood cells. For this reason also, it is not yet clear how informative these approaches will be for understanding atherosclerosis in humans. REVIEWS 170 | MARCH 2006 | VOLUME 7 www.nature.com/reviews/genetics � 2006 Nature Publishing Group Acute coronary syndrome A group of clinical symptoms that is associated with acute myocardial ischaemia following plaque rupture. Coronary artery stenosis The chronic narrowing of the coronary artery lumen. TNF ligand family Tumour necrosis factors are pro-inflammatory cytokines; the superfamily of TNF ligands have a wide range of functions, including apoptosis, B-cell and T-cell co-stimulation, and bone metabolism. Aortic sinus One of the anatomical dilations of the ascending aorta (the main artery that carries blood from the heart to the body), which occurs at the aortic root, just above the aortic valve. Two-hybrid system An experimental approach that is used to clone genes for which the protein products interact with another protein of interest. Phage display An experimental cloning system in which proteins are displayed on the surface of bacteriophages after fusion with coat proteins to allow the identification of novel protein interactions. Lectin A molecule with multiple sugar-binding sites that agglutinates cells. These findings prompted a series of human studies that examined how genetic variation in APOA5 ? probably in combination with APOC3 ? influences tri glyceride levels 61 . Data that directly map CAD susceptibility to the cluster come from the study of familial combined hyperlipidaemia (FCHL), an inherited dyslipidaemia that shows complex patterns of hypercholesterolaemia and hyper triglyceridaemia 62 . A prospective study showed that 20-year cardiovascular mortality is roughly doubled in families that show FCHL 63 . The first hint of a con- nection with the APOA1?APOC3?APOA4 gene cluster came from a combined linkage and association analysis for FCHL, which implicated APOC3 variants in suscep- tibility 64 , although this finding was refractory to replica- tion. However, several recent association studies support the link between the APOA1?APOC3?APOA4?APOA5 cluster and FCHL, probably through a mechanism that involves alleles in the APOA5 and APOC3 loci 61 . The information that is inferred from comparative genomics is not, however, infallible. The spontaneous mouse mutant Hyplip1 shows several metabolic attributes of FCHL, including raised levels of triglyceride, choles- terol and APOB. The mutation maps to mouse chromo- some 3 (REF. 65), and a nonsense mutation in the gene that encodes thioredoxin interacting protein (Txnip) has been identified by positional cloning 66 . Although sev- eral cellular functions had been ascribed to this gene, a putative role in the regulation of lipid metabolism was unforeseen. The human orthologue, TXNIP, maps to chromosome 1q21.2 within a linked region identified in Finnish families that show FCHL 67 . However, no association of TXNIP with FCHL or triglyceride levels has been detected 67,68 . Instead, another positional candidate gene USF1 (upstream transcription factor 1) that lies within 15 Mb of TXNIP showed evidence of association with several FCHL phenotypes 67 . This gene encodes a transcription factor that is implicated in the regulation of several apolipoproteins and other proteins with roles in lipid and glucose metabolism. No easily interpretable muta- tion (for example, non-synonymous SNP) was found in USF1, and a subsequent screen in British patients with FCHL also failed to detect any coding variant 69 . However, a 60-bp highly conserved element within intron 7 was identified that is thought to influence transcription of the gene in vitro 67 ; this regulatory func- tion has since been experimentally refined to a 20-bp element 70 . The importance of quantitative variation in sites that are implicated in regulation of gene transcrip- tion is increasingly being recognized 71 and involves both cis and trans variation 72 , so any variants in this regula- tory region are strong candidates for further analysis. Finally, it is presumably an evolutionary coincidence that USF1 and TXNIP are syntenic (they are both on human chromosome 1), that Usf1 and Txnip are not syntenic (being on mouse chromosomes 3 and 1, respectively), and that mutations in Txnip and USF1 can both lead to FCHL phenotypes. These studies show that there can be substan- tial benefits in studying genes that underlie mouse models of atherosclerosis. Sometimes, genetic variation in orthologous genes causes similar pathologies and related phenotypes in humans. Even if the human gene confers little or no susceptibility, information from mouse genetics can still direct drug-discovery programmes by identifying new biological pathways. Conclusions CAD genetics is a field that is gathering momentum. Although complex trait analyses for other diseases have longer histories, the public-health importance of CAD has ensured that it is now the focus of some of the largest human genetics programmes (for exam- ple, see REF. 47). Genome-wide linkage analyses indi- cate that although there are no loci for CAD risk that have very large effects, there are mappable loci with effects that are comparable to those in other diseases for which gene identification has been successful. Indeed, the first positionally cloned CAD genes are now emerging 19,47,67 . Progress will accelerate now that genome- wide association studies, using hundreds of thousands of SNPs, are feasible. Similarly, combined genetic and genomic approaches have the potential to reveal new realms of quantitative heritable variation that influence the biological processes underlying CAD (BOX 3). Large studies, involving thousands of cases and controls, will be needed to provide robust evidence of novel suscepti- bility genes in CAD because effect sizes will be modest and gene?environment interactions will undoubtedly be important. However, even a gene of modest effect can be of enormous significance if it identifies a new therapeutic target or biological pathway. At this stage, one can only speculate on the prob- able biological roles of the susceptibility genes that will emerge. Many of the genes that have been implicated until now in CAD are involved in innate and adaptive immunity. Inflammation is increasingly recognized as a crucial component in the development of atheromatous plaques and thrombosis 3 , so these results are not unex- pected. However, the predominance of inflammatory response genes among those that have been identified might also indicate that it will only be in the context of previous selection pressure that CAD variants will be easily discernable in genome-wide studies. If so, this could constrain the types of genetic influence that will be identified; however, it could also yield unexpected syner- gies across fields. If selection pressures predict an overlap between genes that affect the risk of CAD and the risk of ?ancient? infectious diseases, one would also expect an overlap with susceptibility for other modern diseases that are governed by innate and adaptive immune responses. Of note, a polymorphism in the type III promoter of the MHC class II transactivator (MHC2TA; also known as CIITA) has recently been associated with increased susceptibility to rheumatoid arthritis, multiple sclerosis and myocardial infarction 73 . As genetic data accumulate across a range of diseases that involve inflammation, we should expect to see more connections of this type being identified and ?added value? being gained from disparate mapping studies. This is another reason to be optimistic that the substantial investment in complex trait analysis in human disease will produce the desired returns. REVIEWS NATURE REVIEWS | GENETICS VOLUME 7 | MARCH 2006 | 171 � 2006 Nature Publishing Group 1. Mackay, J. & Mensah, G. (eds) The Atlas of Heart Disease and Stroke (WHO, Geneva, Switzerland, 2004). 2. Lusis, A. J., Mar, R. & Pajukanta, P. Genetics of atherosclerosis. Annu. Rev. Genomics Hum. Genet. 5, 189?218 (2004). 3. Libby, P. Inflammation in atherosclerosis. Nature 420, 868?874 (2002). 4. Zdravkovic, S. et al. Heritability of death from coronary heart disease: a 36-year follow-up of 20,966 Swedish twins. J. Intern. Med. 252, 247?254 (2002). 5. Wright, A., Charlesworth, B., Rudan, I., Carothers, A. & Campbell, H. A polygenic basis for late-onset disease. Trends Genet. 19, 97?106 (2003). 6. Myers, R. H., Kiely, D. K., Cupples, L. A. & Kannel, W. B. Parental history is an independent risk factor for coronary artery disease:the Framingham study. Am. Heart J. 120, 963?969 (1990). 7. Wilson, P. W. et al. Prediction of coronary heart disease using risk factor categories. Circulation 97, 1837?1847 (1998). 8. Lloyd-Jones, D. M. et al. Parental cardiovascular disease as a risk factor for cardiovascular disease in middle-aged adults: a prospective study of parents and offspring. JAMA 291, 2204?2211 (2004). 9. Assmann, G., Cullen, P. & Schulte, H. Simple scoring scheme for calculating the risk of acute coronary events based on the 10-year follow-up of the prospective cardiovascular Munster (PROCAM) study. Circulation 105, 310?315 (2002). 10. Yusuf, S. et al. Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study): case?control study. Lancet 364, 937?952 (2004). 11. Weiss, K. M. & Terwilliger, J. D. How many diseases does it take to map a gene with SNPs? Nature Genet. 26, 151?157 (2000). 12. Cohen, J. C. et al. Multiple rare alleles contribute to low plasma levels of HDL cholesterol. Science 305, 869?872 (2004). 13. Rudan, I. et al. Inbreeding and risk of late onset complex disease. J. Med. Genet. 40, 925?932 (2003). 14. Lohmueller, K. E., Pearce, C. L., Pike, M., Lander, E. S. & Hirschhorn, J. N. Meta-analysis of genetic association studies supports a contribution of common variants to susceptibility to common disease. Nature Genet. 33, 177?182 (2003). 15. Wang, L., Fan, C., Topol, S. E., Topol, E. J. & Wang, Q. Mutation of MEF2A in an inherited disorder with features of coronary artery disease. Science 302, 1578?1581 (2003). 16. Bhagavatula, M. R. et al. Transcription factor MEF2A mutations in patients with coronary artery disease. Hum. Mol. Genet. 13, 3181?3188 (2004). 17. Gonzalez, P. et al. The Pro279Leu variant in the transcription factor MEF2A is associated with myocardial infarction. J. Med. Genet. 15 Jun 2005 [epub ahead of print]. 18. Weng, L. et al. Lack of MEF2A mutations in coronary artery disease. J. Clin. Invest. 115, 1016?1020 (2005). 19. Helgadottir, A. et al. The gene encoding 5-lipoxygenase activating protein confers risk of myocardial infarction and stroke. Nature Genet. 36, 233?239 (2004). A landmark positional cloning study, as it was the first to use linkage, linkage-disequilibrium mapping and functional studies to identify a CAD-susceptibility gene. 20. Spanbroek, R. et al. Expanding expression of the 5-lipoxygenase pathway within the arterial wall during human atherogenesis. Proc. Natl Acad. Sci. USA 100, 1238?1243 (2003). 21. Mehrabian, M. et al. Identification of 5-lipoxygenase as a major gene contributing to atherosclerosis susceptibility in mice. Circ. Res. 91, 120?126 (2002). 22. Aiello, R. J. et al. Leukotriene B4 receptor antagonism reduces monocytic foam cells in mice. Arterioscler. Thromb. Vasc. Biol. 22, 443?449 (2001). 23. Dwyer, J. H. et al. Arachidonate 5-lipoxygenase promoter genotype, dietary arachidonic acid, and atherosclerosis. N. Engl. J. Med. 350, 29?37 (2004). 24. Danesh, J. et al. C-reactive protein and other circulating markers of inflammation in the prediction of coronary heart disease. N. Engl. J. Med. 350, 1387?1397 (2004). 25. Paul, A. et al. C-reactive protein accelerates the progression of atherosclerosis in apolipoprotein E-deficient mice. Circulation 109, 647?655 (2004). 26. Fletcher, L. Genomics companies shop around for chemical expertise. Nature Biotechnol. 22, 137?138 (2004). 27. Broeckel, U. et al. A comprehensive linkage analysis for myocardial infarction and its related risk factors. Nature Genet. 30, 210?214 (2002). 28. Wang, Q. et al. Premature myocardial infarction novel susceptibility locus on chromosome 1P34?36 identified by genomewide linkage analysis. Am. J. Hum. Genet. 74, 262?271 (2004). 29. Hauser, E. R. et al. A genomewide scan for early-onset coronary artery disease in 438 families: the GENECARD Study. Am. J. Hum. Genet. 75, 436?447 (2004). 30. Harrap, S. B. et al. Genome-wide linkage analysis of the acute coronary syndrome suggests a locus on chromosome 2. Arterioscler. Thromb. Vasc. Biol. 22, 874?878 (2002). 31. Chiodini, B. D. & Lewis, C. M. Meta-analysis of 4 coronary heart disease genome-wide linkage studies confirms a susceptibility locus on chromosome 3q. Arterioscler. Thromb. Vasc. Biol. 23, 1863?1868 (2003). 32. Campbell, L. A. & Kuo, C. C. Chlamydia pneumoniae ? an infectious risk factor for atherosclerosis? Nature Rev. Microbiol. 2, 23?32 (2004). 33. LeVan, T. D. et al. A common single nucleotide polymorphism in the CD14 promoter decreases the affinity of Sp protein binding and enhances transcriptional activity. J. Immunol. 167, 5838?5844 (2001). 34. Arroyo-Espliguero, R., Avanzas, P., Jeffery, S. & Kaski, J. C. CD14 and toll-like receptor 4: a link between infection and acute coronary events? Heart 90, 983?988 (2004). 35. Pasterkamp, G. J., van Keulen, K. & de Kleijn, D. P. V. Role of Toll-like receptor 4 in the initiation and progression of atherosclerotic disease. Eur. J. Clin. Invest. 34, 328?334 (2004). 36. Kiechl, S. et al. Toll-like receptor 4 polymorphisms and atherogenesis. N. Engl. J. Med. 47, 185?192 (2002). 37. Ameziane, N. et al. Association of the Toll-like receptor 4 gene Asp299Gly polymorphism with acute coronary events. Arterioscler. Thromb. Vasc. Biol. 23, e61?e64 (2003). 38. Boekholdt, S. M. et al. Variants of toll-like receptor 4 modify the efficacy of statin therapy and the risk of cardiovascular events. Circulation 107, 2416?2132 (2003). 39. Michelsen, K. S. et al. Lack of Toll-like receptor 4 or myeloid differentiation factor 88 reduces atherosclerosis and alters plaque phenotype in mice deficient in apolipoprotein E. Proc. Natl Acad. Sci. USA. 101, 10679?10684 (2004). 40. Edfeldt, K. et al. Association of hypo-responsive toll- like receptor 4 variants with risk of myocardial infarction. Eur. Heart J. 25, 1447?1453 (2004). 41. Yang, I. A., Holloway, J. W. & Ye, S. TLR4 Asp299Gly polymorphism is not associated with coronary artery stenosis. Atherosclerosis 170, 187?190 (2003). 42. Colhoun, H. M., McKeigue, P. M. & Davey Smith, G. Problems of reporting genetic associations with complex outcomes. Lancet 361, 865?872 (2003). 43. Keavney, B. et al. Large-scale test of hypothesised associations between the angiotensin-converting- enzyme insertion/deletion polymorphism and myocardial infarction in about 5000 cases and 6000 controls. International Studies of Infarct Survival (ISIS) Collaborators. Lancet 355, 434?442 (2000). 44. Hirschhorn, J. N. & Daly, M. J. Genome-wide association studies for common diseases and complex traits. Nature Rev. Genet. 6, 95?108 (2005). 45. Farrall, M. & Morris, A. P. Gearing up for genome-wide gene-association studies. Hum. Mol. Genet. 14 (Suppl. 2), R157?R162 (2005). 46. Risch, N. & Merikangas, K. The future of genetic studies of complex human diseases. Science 273, 1516?1517 (1996). 47. Ozaki, K. et al. Functional SNPs in the lymphotoxin-? gene that are associated with susceptibility to myocardial infarction. Nature Genet. 32, 650?654 (2002). A pioneering genome-scale gene-association study for susceptibility to myocardial infarction that identified causal variants in the LTA gene. 48. Iwanaga, Y. et al. Association analysis between polymorphisms of the lymphotoxin-? gene and myocardial infarction in a Japanese population. Atherosclerosis 172, 197?198 (2004). 49. Yamada, A. et al. Lack of association of polymorphisms of the lymphotoxin ? gene with myocardial infarction in Japanese. J. Mol. Med. 82, 477?483 (2004). 50. The PROCARDIS Consortium. A trio family study showing association of the lymphotoxin-? N26 (804A) allele with coronary artery disease. Eur. J. Hum. Genet. 12, 770?774 (2004). 51. Knight, J. C., Keating, B. J., Rockett, K. A. & Kwiatkowski, D. P. In vivo characterization of regulatory polymorphisms by allele-specific quantification of RNA polymerase loading. Nature Genet. 33, 469?475 (2003). 52. Knight, J. C., Keating, B. J. & Kwiatkowski, D. P. Allele- specific repression of lymphotoxin-? by activated B cell factor-1. Nature Genet. 36, 394?399 (2004). 53. Schreyer, S. A., Vick, C. M. & LeBoeuf, R. C. Loss of lymphotoxin-? but not tumor necrosis factor-? reduces atherosclerosis in mice. J. Biol. Chem. 277, 12364?12368 (2002). 54. Ozaki, K. et al. Functional variation in LGALS2 confers risk of myocardial infarction and regulates lymphotoxin-? secretion in vitro. Nature 429, 72?75 (2004). 55. Cullen, P. et al. Rupture of the atherosclerotic plaque: does a good animal model exist? Arterioscler. Thromb. Vasc. Biol. 23, 535?542 (2003). 56. Wang, X. et al. Positional identification of TNFSF4, encoding OX40 ligand, as a gene that influences atherosclerosis susceptibility. Nature Genet. 37, 365?372 (2005). 57. Austin, M. A., Hokanson, J. E. & Edwards, K. L. Hypertriglyceridemia as a cardiovascular risk factor. Am. J. Cardiol. 81, 7B?12B (1998). 58. Rees, A., Shoulders, C. C., Stocks, J., Galton, D. J. & Baralle, F. E. DNA polymorphism adjacent to the human apolipoprotein A1 gene: relationship to hypertriglyceridemia. Lancet 1, 444?446 (1983). 59. van Dijk, K. W., Rensen, P. C., Voshol, P. J. & Havekes, L. M. The role and mode of action of apolipoproteins CIII and AV: synergistic actors in triglyceride metabolism? Curr. Opin. Lipidol. 15, 239?246 (2004). 60. Pennacchio, L. A. et al. An apolipoprotein influencing triglycerides in humans and mice revealed by comparative sequencing. Science 294, 169?173 (2001). An exemplary comparative genomics study that identified and characterized a novel apolipoprotein which was later implicated in human dyslipidaemia. 61. Shoulders, C. C., Jones, E. L. & Naoumova, R. P. Genetics of familial combined hyperlipidemia and risk of coronary heart disease. Hum. Mol. Genet. 13, R149?R160 (2004). 62. Goldstein, J. L., Schrott, H. G., Hazzard, W. R., Bierman, E. L. & Motulsky, A. G. Hyperlipidemia in coronary heart disease. II. Genetic analysis of lipid levels in 176 families and delineation of a new inherited disorder, combined hyperlipidemia. J. Clin. Invest. 52, 1544?1568 (1973). 63. Austin, M. A. et al. Cardiovascular disease mortality in familial forms of hypertriglyceridemia: A 20-year prospective study. Circulation 101, 2777?2782 (2000). 64. Wojciechowski, A. P. et al. Familial combined hyperlipidaemia linked to the apolipoprotein AI-CII-AIV gene cluster on chromosome 11q23?q24. Nature 349, 161?164 (1991). 65. Castellani, L. W. et al. Mapping a gene for combined hyperlipidaemia in a mutant mouse strain. Nature Genet. 18, 374?377 (1998). 66. Bodnar, J. S. et al. Positional cloning of the combined hyperlipidemia gene Hyplip1. Nature Genet. 30, 110?116 (2002). 67. Pajukanta, P. et al. Familial combined hyperlipidemia is associated with upstream transcription factor 1 (USF1). Nature Genet. 36, 371?376 (2004). Following up on earlier linkage to chromosome 1q, gene-association studies excluded a candidate gene (TXNIP) that was implicated in mouse studies and instead identified a novel susceptibility gene that has a conserved intronic regulatory element. 68. Coon, H. et al. TXNIP gene not associated with familial combined hyperlipidemia in the NHLBI Family Heart Study. Atherosclerosis. 174, 357?362 (2004). 69. Putt, W. et al. Variation in USF1 shows haplotype effects, gene:gene and gene:environment associations with glucose and lipid parameters in the European Atherosclerosis Research Study II. Hum. Mol. Genet. 13, 1587?1597 (2004). REVIEWS 172 | MARCH 2006 | VOLUME 7 www.nature.com/reviews/genetics � 2006 Nature Publishing Group 70. Naukkarinen, J. et al. USF1 and dyslipidemias: converging evidence for a functional intronic variant. Hum. Mol. Genet. 14, 2595?2605 (2005). 71. Farrall, M. Quantitative genetic variation: a post- modern view. Hum. Mol. Genet. 13, R1?R7 (2004). 72. Morley, M. et al. Genetic analysis of genome-wide variation in human gene expression. Nature 430, 743?747 (2004). 73. Swanberg, M. et al. MHC2TA is associated with differential MHC molecule expression and susceptibility to rheumatoid arthritis, multiple sclerosis and myocardial infarction. Nature Genet. 37, 486?494 (2005). 74. Marchini, J., Cardon, L. R., Phillips, M. S. & Donnelly, P. The effects of human population structure on large genetic association studies. Nature Genet. 36, 512?517 (2004). 75. Keavney, B. Commentary: Katan?s remarkable foresight: genes and causality 18 years on. Int. J. Epidemiol. 33, 11?14 (2004). 76. Benjamin, E. J. et al. Clinical correlates and heritability of flow-mediated dilation in the community: the Framingham Heart Study. Circulation 109, 613?619 (2004). 77. Mitchell, G. F. et al. Heritability and a genome-wide linkage scan for arterial stiffness, wave reflection, and mean arterial pressure: the Framingham Heart Study. Circulation 112, 194?199 (2005). 78. Jansen, R. C. & Nap, J. P. Genetical genomics: the added value from segregation. Trends Genet. 17, 388?391 (2001). 79. Schadt, E. E. et al. An integrative genomics approach to infer causal associations between gene expression and disease. Nature Genet. 37, 710?717 (2005). 80. Jansen, R. C. & Nap, J. P. Regulating gene expression: surprises still in store. Trends Genet. 20, 223?225 (2004). 81. Cervino, A.C. et al. Integrating QTL and high-density SNP analyses in mice to identify Insig2 as a susceptibility gene for plasma cholesterol levels. Genomics. 86, 505?517 (2005). 82. Mehrabian, M. et al. Integrating genotypic and expression data in a segregating mouse population to identify 5-lipoxygenase as a susceptibility gene for obesity and bone traits. Nature Genet. 37, 1224?1233 (2005). 83. Helgadottir, A. et al. A variant of the gene encoding leukotriene A4 hydrolase confers ethnicity-specific risk of myocardial infarction. Nature Genet. 38, 68?74 (2006). 84. Choudhury, R. P. Fuster, V. & Fayad, Z. A. Molecular, cellular and functional imaging of atherothrombosis. Nature Rev. Drug Discov. 3, 913?925 (2004). Competing interests statement The authors declare no competing financial interests. DATABASES The following terms in this article are linked online to: Entrez Gene: http://www.ncbi.nlm.nih.gov/entrez/query. fcgi?db=gene ALOX5 | ALOX5AP | LGALS2 | LTA | LTA4H | MEF2A | MHC2TA | SELE | TLR4 | Tnfsf4 | Txnip | VCAM1 OMIM: http://www.ncbi.nlm.nih.gov/entrez/query. fcgi?db=OMIM familial combined hyperlipidaemia | familial hypercholesterolaemia FURTHER INFORMATION The International HapMap Project: http://www.hapmap.org The SNP Consortium: http://snp.cshl.org WHO atlas of heart disease and stroke: http://www.who.int/ cardiovascular_diseases/resources/atlas/en Access to this interactive links box is free online. REVIEWS NATURE REVIEWS | GENETICS VOLUME 7 | MARCH 2006 | 173 "
Add Content to Group
|
Bookmark
|
Keywords
|
Flag Inappropriate
share
Close
Digg
Facebook
MySpace
Google+
Comments
Close
Please Post Your Comment
*
The Comment you have entered exceeds the maximum length.
Submit
|
Cancel
*
Required
Comments
Please Post Your Comment
No comments yet.
Save Note
Note
View
Public
Private
Friends & Groups
Friends
Groups
Save
|
Cancel
|
Delete
Please provide your notes.
Next
|
Prev
|
Close
|
Edit
|
Delete
Genetics
Gene Inheritance and Transmission
Gene Expression and Regulation
Nucleic Acid Structure and Function
Chromosomes and Cytogenetics
Evolutionary Genetics
Population and Quantitative Genetics
Genomics
Genes and Disease
Genetics and Society
Cell Biology
Cell Origins and Metabolism
Proteins and Gene Expression
Subcellular Compartments
Cell Communication
Cell Cycle and Cell Division
Scientific Communication
Career Planning
Loading ...
Scitable Chat
Register
|
Sign In
Visual Browse
Close
Comments
CloseComments
Please Post Your Comment