Genetic Risk Factors of Intracranial Atherosclerosis

Abstract

Purpose of Review

Intracranial atherosclerosis (ICAS) is the most common cause of stroke throughout the world. It also increases the risk of recurrent stroke and dementia. As a complex and multifactorial disease, ICAS is influenced by multiple genetic, biological, and environmental factors. This review summarizes the candidate gene and genome-wide studies aimed at discovering genetic risk factors of ICAS.

Recent Findings

Numerous studies have focused on the association between single nucleotide polymorphisms (SNPs) of atherosclerosis-related genes and the risk of ICAS. Variants in adiponectin Q (ADIPOQ), ring finger protein 213 (RNF213), apolipoprotein E (APOE), phosphodiesterase 4D (PDE4D), methylenetetrahydrofolate reductase (MTHFR), lipoprotein lipase (LPL), α-adducin (ADD1) genes, angiotensin-converting enzyme (ACE), as well as other genes related to renin-angiotensin-aldosterone system have been associated with ICAS.

Summary

We review the available evidences on the candidate genes and SNPs associated with genetic susceptibility to ICAS, and point out future developments of this field. Genetic discoveries could have clinical implications for intracranial atherosclerotic disease.

Introduction

Intracranial atherosclerosis (ICAS) is the most common cause of stroke in the world and a common risk factor for dementia [1–4]. ICAS is a progressive disease characterized chiefly by the accumulation of lipids and cellular and fibrous elements in the large arteries, leading to changes ranging from eccentric wall thickening and substenotic plaques to hemodynamically significant luminal narrowing (Figure 1) [5–7]. ICAS can occur in isolation, or as part of coexistent atherosclerosis in systemic arteries such as the aorta, extracranial carotids, coronary, or lower extremity arteries [8]. The most common site of ICAS is the middle cerebral artery, followed by the basilar artery, internal carotid artery, and intracranial vertebral artery [9]. In addition, ICAS may coexist with other potential stroke etiologies, e.g. small vessel disease, in which ICAS confers a worse prognosis [10].

Figure 1. Pathological example of intracranial large artery atherosclerosis, a cholesterol-mediated disease.

Panel A demonstrates a typical example of atherosclerosis, defined by eccentric intima thickening with a necrotic core consisting of cholesterol crystals and necrotic tissue (panel 1). One may observe neoangiogenesis in the border zone between the media and the intima (panel b-c). One of the proposed mechanisms by which atherosclerosis may cause stroke is plaque rupture, with endothelial surface denudation, thrombosis and eventual artery-to-artery embolization or luminal occlusion due to an expanding in-situ thrombus (panel d-c). Images were obtained from the Brain Arterial Remodeling Study (BARS).

The most common risk factors for ICAS include age, hypertension, diabetes mellitus, and dyslipidemia [4, 11–14]. The prevalence of ICAS has been reported higher in women, particularly in women older than 63 years compared with men within the same age group [15]. Several studies suggest that race/ethnicity is a predisposing factor, especially in combination with acquired risk factors, such as lifestyle [1, 11, 13, 16–18]. Individuals of Hispanic [11, 16, 17], African-American [11, 16–18], and Asian (Chinese, Japanese, and Korean descent) ethnic backgrounds [12, 19–22] have a significantly higher incidence and prevalence of ICAS compare to non-Hispanic whites. The difference in the incidence of ICAS among different ethnicities could be caused by differences in certain genetic predisposition or morphological characteristics of the cerebral arteries [23]. Dyslipidemia, a risk factor that are linked to coronary atherosclerosis and myocardial infarction, is also associated with ICAS [24, 25]. Smoking has been associated with ICAS [4, 26], but not consistently across studies [27]. The variability in results may be due to heterogeneous populations and methods used to assess the outcome.

Certain genetic traits are associated with ICAS, either by predisposing to vascular risks such as hypertension or diabetes mellitus, or by a direct contribution to an established atherosclerosis mechanism [28, 29]. Genome-wide association studies (GWAS) have provided new approaches for detecting novel variants and gene loci without a specific hypothesis implicating a particular molecular pathway. In this article, we review genetic traits associated with ICAS, summarize candidate gene and genome-wide studies aimed at discovering and identifying genetic risk factors of ICAS and subclinical phenotypes, and outline the future developments and challenges of ICAS genetic research.

Search strategy

We performed a MEDLINE/PubMed literature search using the MeSH term “intracranial arterial diseases” or “intracranial atherosclerosis” or “intracranial stenosis” or “cerebral atherosclerosis”. Articles published in English before October 2019 were screened for relevance to the genetic risk factor of ICAS. Studies of ≥200 participants were included. We selected studies with ICAS-related phenotypes represented as stroke caused by ICAS or studies related to ICAS physiopathology. The key findings are summarized in Table 1.

Table 1.

Genetic factors associated with ICAS.

Gene	Gene function	Polymorphism	Study population	Sample size	Main finding	References
ADIPOQ	Regulate adiponectin level	g.15734A>G	Chinese	602 (199 ICAS / 403 controls)	Increased risk of ICAS	[36]
			American	2847	Related to carotid and coronary atherosclerosis in African Americans	[35]
			Chinese	2212 (1105 T2DM /1107 controls)	Closely related to hypoadiponectinemia	[42]
		g.5320G>A	American	1200 (600 MI or IS / 600 controls)	Reduced of cerebral infarction risk level	[37]
			Chinese	476	Elevated plasma adrenomedullin levels	[43]
RNF213	Result in vascular fragility, lead to vessels more vulnerable to hemodynamic stress	c.14576G>A	Japanese	433	Significant association with ICAS	[47]
			Japanese	221	Associated with anterior ICAS	[48]
		c.14429G>A	Korean	532	Associated with vasculogenesis	[53]
APOE	Associate with high LDL cholesterol level	ɛ4	Finnish	700	Associated with larger coronary and aortic atherosclerotic lesion areas in men	[58]
			American	720		[59]
			Finnish	536 (237 IS / 362 controls)	There was a link with ICAS in men	[60]
			Thai	308	Associated with extracranial atherosclerosis	[61]
PDE4D	Has a major role in the degradation of cAMP	g.319027C>T	Indian	336	TT genotype was associated with ICAS	[72]
MTHFR	Reduce enzyme activity, decrease folate level and increase Hcy	g.14783C>T	Chinese	929	This variant and Hcy had interaction in the formation of cerebral atherosclerosis	[78]
			Korean	463 (267 atherosclerosis / 196 controls)	Hcy level and this variant did not contribute to the distribution of cerebral atherosclerosis	[79]
LPL	Involve in plasma lipoprotein metabolism and transportation	g.27496T>G	Indian	1025 (525 IS / 500 controls)	Associated with intracranial large artery atherosclerosis	[81]
		p.Ser447Stop	Japanese	354 (177 CVD / 177 controls)	Reduced risk of atherothrombotic cerebral infarction	[80]
		g.23608C>T, p.Ser447Ter	Chinese	371 (185 CI / 186 controls)	Closely associated with atherothrombotic cerebral infarction	[82]
ADD1	Associate with blood pressure levels	p.Gly460Trp	Dutch	6471 and 1018	Associated with atherosclerosis	[88]
			Indian	339	No significant difference between ICAS and extracranial atherosclerosis	[89]
ACE	Take part in vascular remodeling and the development of atherosclerosis	Insertion/Deletion	Thai	308	No association between this polymorphism and ICAS	[61]
			Chinese	210		[94]
			Chinese	708		[95]
CYP11B2	Involve in the aldosterone system	g.4660T>C	Indian	797 (403 stroke / 394 controls)	Associated with intracranial large artery atherosclerosis	[98]

CI: cerebral infarction; CVD: cerebrovascular disease; ICAS: intracranial atherosclerosis; IS: ischemic stroke; MI: myocardial infarction; T2DM: type 2 diabetes mellitus.

ADIPOQ gene

Adiponectin Q (ADIPOQ) gene regulates a variety of metabolic processes and helps inhibit the biochemical pathways that lead to metabolic syndrome. It contributes to lipids and glucose regulation, with anti-inflammatory and anti-atherogenic effect [30–33]. ADIPOQ SNPs are likely to contribute to metabolic disorders, and consequently influence atherosclerosis [34]. The g.15734A>G genotypes are associated with carotid and coronary atherosclerosis in African Americans [35] and ICAS in Chinese [36], while g.5320G>A SNP is associated with reduced risk of cerbral infaction in Caucasian [37]. Among 602 subjects (199 cases with ICAS and 403 controls) included in the Chinese study [36], authors genotyped 10 selected tag ADIPOQ SNPs associated with adiponectin levels, atherosclerosis or cardiovascular events. After adjusting for conventional vascular risks, there was a modest association with ICAS in participants with the g.15734A>G AG/AA genotype (OR = 2.2, 95% CI: 1.1–4.9, p = 0.040) and the AG/GG genotype of the g.5320G>A (OR = 1.8, 95% CI: 1.1–2.9, p = 0.017). Moreover, the haploid analysis results indicated that the A-G haplotype prevalence of the g.15734A>G and g.5320G>A was higher in the ICAS group than in the control group (p = 0.026).

The g.5320G>A and g.15734A>G belong to the first and second intron of ADIPOQ gene, respectively, and are closely associated with reduction of adiponectin levels. Previous clinical studies suggested the ADIPOQ g.5320G>A was closely associated with reduced adiponectin levels in Asian, African and Caucasian populations [38–41]. Du et al. performed a case-control study using 1105 patients with type 2 diabetes mellitus and 1107 control subjects, and found patients with genotype AG/GG of g.15734A>G had lower levels of serum adiponectin than those with the genotype AA (p = 0.044), indicating g.15734A>G may be closely related to hypoadiponectinemia in the Chinese population [42]. Therefore, these ADIPOQ SNPs may play a role in proatherosclerotic blood signals mediated by adiponectin or adrenomedullin [43]. However, the findings require testing for replication and validation in larger sample size studies and experimentation. It is uncertain whether modulating these pathways may recue or modify the risk of ICAS and ICAS-related vascular events.

RNF213 gene

Ring finger protein 213 (RNF213) encodes a protein containing 5256 amino acids, containing a C3HC4-type RING (Really Interesting New Gene) finger domain and an AAA (Adenosine triphosphatase Associated with various cellular Activities) domain, which are involved in mediating protein-protein interactions and ATPase activity, respectively [44, 45]. RNF213 genetic variants may result in arterial fragility and susceptibility to hemodynamic stress, which may increase the risk of ICAS [46].

RNF213 was initially identified as a susceptibility gene for moyamoya disease (MMD) in a community‐based GWAS [45, 44]. In Japan, Miyawaki et al. compared the prevalence of c.14576G>A variant in each phenotype group (definite MMD, unilateral MMD, ICAS not diagnosed as MMD, extracranial carotid atherosclerosis, cerebral aneurysm, or intracerebral hemorrhage) with control subjects without cerebrovascular disease [47]. The authors found that the c.14576G>A variant was associated with definite MMD (OR=144.0, 95% CI: 26.7–775.9, p < 0.001), unilateral MMD (OR=54.0, 95% CI: 7.5–386.8, p < 0.001), and non-MMD ICAS (OR=16.8, 95% CI: 3.8–74.5, p <0.001) compared to controls without cerebrovascular disease. They also investigated whether c.14576G>A variant related to anterior or posterior circulation ICAS [48]. The investigators found an association with anterior circulation ICAS (OR = 14.8, 95% CI: 3.1–71.3, p < 0.001) but no association with posterior circulation ICAS. Studies in Korea and Japan found about one fifth of non-MMD ICAS patients carry RNF213 c.14576G>A variant [47, 49, 50]. It was estimated that about 16 million people carry this genetic polymorphism in East Asian countries [51]. Although the role of RNF213 c.14576G>A variants has mostly been studied in relationship to ICAS, another variants c.14429G>A may also play a role in vasculogenesis [52, 53].

APOE gene

Apolipoprotein E (APOE) is the principal cholesterol carrier in the brain [54]. It is mainly produced by astrocytes, and transports cholesterol to neurons through ApoE receptors [55]. There are six well-known common APOE genotypes (ɛ22, ɛ32, ɛ33, ɛ42, ɛ43, and ɛ44), which are generated by a combination of two genetic variants (g.7903T>C and g.8041C>T). APOE ɛ4 is associated with high LDL cholesterol level, which is an important factor in the early stage of atherosclerosis [56], and might interact with other risk factors to affect lipid metabolism and cellular repair mechanism [57]. Autopsy studies in men reveal the APOE ɛ4+ genotype is associated with larger coronary and aortic atherosclerotic lesion areas [58, 59]. Separately, another group of investigators tested whether APOE ɛ4 genetic variants related to atherosclerosis. Men with APOE ɛ4 allele had a higher prevalence of ICAS, but the association was not found in women [60]. The association between APOE ɛ4 allele and ICAS is conflicting, however, Chutinet et al. reported APOE ɛ4 allele was associated with extracranial atherosclerosis (ECAS) (OR=2.8, 95% CI:1.3–5.9, p < 0.05) but not with ICAS (OR=1.2, 95% CI: 0.5–2.7) [61]. Given the role of APOE ɛ4 in Alzheimer dementia [62], and the fact that Alzheimer dementia is associated with ICAS [63, 64], it is uncertain whether the association of APOE ɛ4 with arterial disease may contribute partially or have a role in the increased risk of dementia attributed to APOE ɛ4, independent of its effect in amyloid pathways.

PDE4D gene

Phosphodiesterase 4D (PDE4D) variants are associated with ischemic stroke, carotid atherosclerosis and coronary artery disease [65, 66], but it is less certain if PDE4D variants relate to ICAS. PDE4D encodes cyclic adenosine monophosphate (cAMP) -specific 3′,5′-cyclic phosphodiesterase 4D, which has a major role in the degradation of cAMP [67]. The proliferation and migration of vascular smooth muscle cells and macrophages, regulated partially by cAMP, is a crucial early stage in the development of atherosclerosis [68–71]. PDE4D TT genotype frequency of g.319027C>T polymorphism in ICAS patients was significantly higher than that in non-ICAS patients (20 vs 2%, p = 0.01) [72]. Furthermore, variant in the PDE4D gene may not only relate to the prevalence of ICAS, but it may also relate to the severity of the disease as suggested by a report of worse outcomes after ICAS-related stroke. It is uncertain how variant in PDE4D gene may alter outcome, but understanding these mechanisms may open opportunities for decreasing the risk of events related to ICAS.

MTHFR gene

Methylenetetrahydrofolate reductase (MTHFR) is the key enzyme for the metabolism of circulating homocysteine (Hcy), which have a critical role in the regulation of cell signaling and homeostasis [73]. MTHFR g.14783C>T, a common variant leading to a reduction in enzyme activity that results in decreased folate level and increased Hcy level, has been associated with increased risk of stroke [74–76]. MTHFR g.14783C>T polymorphism is associated with atherosclerosis by regulating genome methylation level [77]. For example, Hcy concentration is associated with ICAS and ECAS as demonstrated in a sample of 929 hypertensive patients without stroke in China. However, this association was modified by the g.14783C>T genotype, suggesting that there was an interaction between g.14783C>T genotype and Hcy in cerebral atherosclerosis [78]. Another study evaluated the effect of Hcy level and g.14783C>T polymorphism on determining the intracranial and extracranial locations of atherosclerosis, suggesting Hcy level and MTHFR g.14783C>T was not associated with intracranial as oppose to extracranial atherosclerosis [79].

LPL gene

Lipoprotein lipase (LPL) has an essential role in plasma lipoprotein metabolism and transportation. Several studies have reported a relationship between g.27496T>G, g.23608C>T, p.Ser447Ter polymorphisms of LPL and cerebrovascular diseases [80–82]. In a case-control study of 1025 subjects (525 ischemic stroke patients and 500 control cases), the g.27496T>G variant of LPL gene was associated with ICAS in Indian population (OR=3.7, 95% CI: 1.9–7.2, p < 0.001) [81]. In another case-control study carried out in Japan, p.Ser447Stop polymorphism was associated with intracranial large artery atherosclerosis-related stroke [80]. Moreover, Xu et al. indicated that PvuII and p.Ser447Ter polymorphisms, were closely associated with atherothrombotic cerebral infarction in Chinese population [82]. A meta-analysis of 4681 ischemic stroke cases and 8516 controls from 13 studies found that the p.Ser447Ter polymorphism was more important in association with the reduced risk for ischemic stroke (OR = 0.79, 95% CI: 0.68–0.93, p = 0.005), especially atherosclerotic stroke (OR = 0.44, 95% CI: 0.32–0.62, p < 0.001), indicating p.Ser447Ter might be the protective factors for atherosclerotic stroke [83].

ADD1 gene

Adducin is a cytoskeletal protein consisting of an α- and a β-subunit. The α-adducin (ADD1) p.Gly460Trp polymorphism has been associated with blood pressure levels, salt sensitivity and the risk factor for cardiovascular events [84–87]. van Rijn et al. studied the p.Gly460Trp polymorphism in relation to atherosclerosis, myocardial infarction, and cerebrovascular disease within two cohorts, involving in 6471 and 1018 subjects respectively, revealing this ADD1 polymorphism was significantly associated with atherosclerosis, but data related to ICAS was not reported in this study [88]. Other investigators evaluated the relationship between ADD1 gene variants and ICAS versus ECAS. But according to further comparison of this variant’s effect in ICAS and ECAS, there was no significant difference between ICAS and ECAS [89].

ACE gene

Angiotensin-converting enzyme (ACE) is a major component of renin-angiotensin system, which takes part in vascular remodeling and the development of atherosclerosis [90]. Previous evidence suggests that ACE is an important candidate gene for hereditary susceptibility to vascular diseases [91, 92]. Several studies provided evidence of the association between the ACE insertion/deletion (I/D) polymorphism and stroke risk [93]. The linkage between ACE I/D polymorphism and ICAS has been studies before, but no association between I/D polymorphism and ICAS has been found. [61, 94, 95].

RAAS related genes

The renin-angiotensin-aldosterone system (RAAS) regulates blood pressure, sodium and water balance, and cardiovascular and renal homeostasis. RAAS related genes role in these processes relate them to atherosclerosis [96, 97]. Polymorphisms of RAAS related genes have high frequency in Asians. Among these genes, g.4660T>C polymorphism of aldosterone synthase (CYP11B2) was associated with ICAS in a case-control study using 403 stroke patients and 394 sex and age matched controls in India (OR=3.0, 95% CI: 1.5–5.9, p < 0.001) [98]. A case-control study of 332 patients and 250 controls investigated the association of CYP11B2 polymorphism genotype with ischemic stroke and subtypes in Chinese Han population, suggesting TT genotype of g.4660T>C polymorphism has association with ischemic stroke (OR = 1.6, 95% CI: 1.1–2.3; p = 0.014), large artery atherosclerosis (OR = 1.7, 95% CI: 1.2–2.6, p = 0.005) and small vessel disease (OR = 1.7, 95% CI: 1.1–2.7, p = 0.012) [99]. Since genetic variations in CYP11B2 gene are associated with the progression of carotid atherosclerotic plaque size, CYP11B2 polymorphism might act through carotid atherosclerosis to enhance susceptibility to the large artery atherosclerosis [100].

Shared genetic variants between large artery atherosclerosis and small vessel disease

Large artery atherosclerosis (LAA) and small vessel disease (SVD) are major subtypes of ischemic stroke. Several studies have shown a relationship between SVD and LAA [101–103]. For example, treatments for LAA (e.g. antihypertensive therapy) are also effective in SVD patients [104]. Recent studies discovered common variants shared between LAA and SVD. A meta-analysis of combined LAA and SVD phenotype was performed to detect the shared genetic SNPs, observing three SNPs (g.153898623T>A, g.153899292A>T and g.153900493C>A) at chromosome 6q25.2, ~100kb upstream of the opioid receptor μ1 (OPRM1) gene [105]. Previous study had shown that the OPRM1 gene may be associated with coronary heart disease, which commonly has an atherosclerotic etiology [106]. Another case-control study in China identified the association of two tumor necrosis factor super family member 4 (TNFSF4) SNPs, g.15225T>C and g.4236G>A, with the risk of LAA and SVD [107]. TNFSF4 encodes the costimulatory molecule, OX40 ligand (OX40L), which is involved in T cell activation and the formation of atherosclerosis [108, 109]. In addition, TNFSF4 expression is associated with an increased risk of atherosclerosis [110]. However, the biological mechanisms of these SNPs are still not clear, and more attention and targeted efforts are needed to better define the clinical significance of these SNPs.

Future developments

As a complex, multifactorial disease that is influenced by multiple pathophysiologic, genetic and environmental factors, ICAS does not have a single genetics cause, but is likely associated with the effects of multiple genes (polygenic) in combination with lifestyle and environmental factors. Environmental factors such as dietary habits have a significant role in development of atherosclerosis, while genetic factors represent consequential ICAS determinants [111]. GWAS use large case-control cohorts, including familial and sporadic cases, categorical trait definitions and up to half a million common SNPs, which could provide identification of novel genomic associations with complex diseases. Even though GWAS has initial successes, the variance of interpretation for these genetic associations is limited. The complexity of genes makes it possible to have thousands of unknown variants, each of which may provide a modest contribution to disease.

With the development of life science and technology, research on the analysis and detection of disease-related genes has advanced to clinical levels and provides a scientific basis for diagnosis and differential diagnosis of diseases. The novel technologies could help identify multiple causal variants as well as rare variants [112]. Advanced imaging technologies and genetic data can improve diagnostic accuracy and distinguish cerebral atherosclerotic from non-atherosclerotic intracranial diseases. After diagnosis, prospective genetic studies could be performed in patients with diverse ethnicities, which will facilitate the understanding of different ICAS phenotypes across ethnic and racial groups. Future studies should focus also on sample size and the synergistic effect of multiple ICAS candidate genes to further elucidate the pathogenesis of ICAS-related gene polymorphisms and cerebrovascular disease, and to provide target sites for gene techniques. The comprehensive understanding of the pathogenesis of ICAS is essential for preventing and treating it accordingly.

The next challenge is how to translate the genetic information into an improved therapeutic approach to decrease ICAS-related events. A first step may be variant assessment. It is necessary to determine variants of the reported ICAS phenotypes and how these variants affect the risk of ICAS-related events, especially because different types of variants in the same gene may be associated with distinct phenotypes or inheritance patterns [113]. The studies presented here use various definitions of ICAS, and homogenizing the phenotype and methods may allow for larger meta-analyses and reproducibility of the findings. The advent of high-resolution intracranial wall imaging to evaluate eccentric intima thickening or for a necrotic core among people with ICAS may prove valuable to increase the specificity of the atherosclerotic phenotype and distinguish atherosclerosis from other non-atherosclerotic phenotypes that may cause arterial stenosis.

Conclusions

We present an overview of the current genetic knowledge of ICAS, as it relates to variants related to ADIPOQ, RNF213, APOE, PDE4D, MTHFR, LPL, ADD1, ACE, and RAAS genes. The results discussed here should be contextualized in the setting of limitation related to the data reviewed. Most of clinical studies are cross-sectional, and therefore the quality and quantity of acquired risk factors might not be adequately studied. In addition, due to the heterogeneity in risk factor definitions, diagnostic methods, and characteristics of study subjects (such as ethnicity) or cerebral vessels (such as middle cerebral artery vs. anterior cerebral artery; asymptomatic vs. symptomatic), admixture of non-atherosclerotic intracranial arterial diseases may be partly responsible for the heterogeneity of the results. Additionally, ICAS is a slow and complex multifocal arterial disease. Consequently, polygenic interactions might not be readily apparent in a study of individual genes. Nonetheless, this review provides clues to gain a deeper understanding of the genetic aspects of ICAS, and future paths to continuing gaining understanding of one of the most common causes of stroke in the world.