Inflammation and endothelial function‐related gene polymorphisms are associated with carotid atherosclerosis—A study of community population in Southwest China

Abstract Objectives To investigate the relationships between 18 single nucleotide polymorphisms with carotid atherosclerosis and whether interactions among these genes were associated with an increased risk of carotid atherosclerosis. Methods Face‐to‐face surveys were conducted with individuals aged 40 or older in eight communities. A total of 2377 individuals were included in the study. Ultrasound was used to detect carotid atherosclerosis in the included population. 18 loci of 10 genes associated with inflammation and endothelial function were detected. Gene‐gene interactions were analyzed using generalized multifactor dimensionality reduction (GMDR). Results Among the 2377 subjects, 445 (18.7%) subjects had increased intima‐media thickness in the common carotid artery (CCA‐IMT), and 398 (16.7%) subjects were detected with vulnerable plaque. In addition, NOS2A rs2297518 polymorphism was associated with increased CCA‐IMT, IL1A rs1609682, and HABP2 rs7923349 polymorphisms were associated with vulnerable plaque. Besides, GMDR analysis showed significant gene‐gene interactions among TNFSF4 rs1234313, IL1A rs1609682, TLR4 rs1927911, ITGA2 rs1991013, NOS2A rs2297518, IL6R rs4845625, ITGA2 rs4865756, HABP2 rs7923349, NOS2A rs8081248, HABP2 rs932650. Conclusion The prevalences of increased CCA‐IMT and vulnerable plaque were high in Southwestern China's high‐risk stroke population. Furthermore, inflammation and endothelial function‐related gene polymorphisms were associated with carotid atherosclerosis.

Stroke is caused by a combination of multiple factors, including carotid atherosclerosis, which is strongly associated with ischemic stroke (Ren et al., 2020).
Multiple factors promote the occurrence and development of carotid atherosclerosis, including interactions between traditional risk factors, lifestyle risk factors (age, sex, chronic disease, smoking habit, lack of exercise) and genetic risk factors (Amarenco et al., 2011;Goncalves et al., 2012;Hu et al., 2021;Huang et al., 1999;Lechner et al., 2020;Poznyak et al., 2020). Influence of traditional and lifestyle risk factors, gene-gene interactions may account for the complexity of carotid atherosclerosis (Yi et al., , 2017. Some studies have shown that some gene loci are closely associated with carotid atherosclerosis (Debette et al., 2006;Humphries & Morgan, 2004). Carotid intima-media thickening and vulnerable plaque are two types of carotid atherosclerosis (Forgo B et al., 2018). Intimamedia thickness in the common carotid artery is a subclinical marker of atherosclerosis and is associated with an increased risk of stroke (O'Leary et al., 1999). The rupture and detachment of vulnerable plaque are also leading causes of cerebral infarction (Katakami et al., 2010). Therefore, determining the genetic etiology of carotid intima-media thickened and vulnerable plaque is critical for preventing atherosclerosis and stroke (Fox et al., 2003;Pitchika et al., 2022).
Technological advancements have allowed for detection and identification of a large number of single nucleotide polymorphisms (SNPs) using matrix-assisted laser desorption/ionization-time flight mass spectrometry. Gene-gene interactions can be analyzed using the generalized multifactor dimensionality reduction (GMDR) approach (Lou et al., 2007). However, few studies have used GMDR to determine genetic factors associated with carotid atherosclerosis.
The risk factors related to carotid atherosclerosis of the high-risk stroke population in Southwest China were collected. Selected 18 candidate genes involved in inflammation and endothelial function, and investigated the relationships between these SNPs and carotid atherosclerosis (increased CCA-IMT, vulnerable plaque), GMDR was used to investigate whether interactions among these genes were associated with an increased risk of carotid intima-media thickening and vulnerable plaque (Lou et al., 2007). Furthermore, analyze whether the collected high-risk factors would influence the relationship between gene polymorphisms and carotid atherosclerosis. Informed consent was obtained from all participants.

Study design and participants
Eight communities of Sichuan province in southwestern China were randomly selected, and a cluster survey was performed from May 2015 to September 2015. Details on the study organization, survey, data processing procedures, and quality control were described previously (Yi et al., 2020) and summarized in Figure 1. We initially screened permanent residents aged 40 years or older who had lived in the area for more than 6 months in each community (community residents who lived in Southwest China for a long time had similar eating habits and living habits) using a structured face-to-face questionnaire  Figure 1).

Carotid ultrasonography measurements
Intima-media thickness in the common carotid artery and vulnerable plaque were evaluated using high-resolution B-mode ultrasound (type 512, Acuson Sequoia Apparatus, 7.5-MHz probe, Berlin, Germany) (Crouse et al., 1996). The intraobserver and interobserver coefficients were described in our previous study (Yi et al., , 2017. Carotid characteristics were evaluated independently by one ultrasound imaging doctor blinded to clinical status. The heads of the patients were turned to the opposite side of the examination before measurement, and vertical scans of the common carotid artery distal wall and the bilateral carotid arteries were taken. The CCA-IMT was measured at the common carotid artery along the long axis of the blood vessel. Carotid intima-media thickness was defined as the vertical distance from the tunica intima to the interface between the tunica media and adventitia. The measure was repeated 3 times at this site as well as 1 cm proximal and distal, and the average values were taken as CCA-IMT. Carotid intima-media thickening was defined as having an intima-media thickness ≥ 1.0 mm (Handa et al., 1990;Lorenz et al., 2010;Ma et al., 2016;Spence, 2006). The atherosclerotic plaque was defined as an endoluminal protrusion of at least 1.5 mm or a focal wall thickening >50% than the surrounding vessel wall (Rundek et al., 2008). Based on the plaque echogenicity and surface characteristics, a carotid plaque was further graded from class I to class IV as echolucent, predominantly echolucent, predominantly echogenic, and echogenic, respectively (Mathiesen et al., 2001). Plaque of class I or class II was defined as vulnerable plaque, and plaque of class III or class IV was defined as stable plaque (Yi et al., 2020).

Genes and SNPs selection
Single nucleotide polymorphisms in 10 genes involved in inflammation and endothelial function were selected from the NCBI database

SNP detection
Blood samples (3 mL, elbow vein) from the 2377 participants were drawn into sterile tubes containing ethylene diamine tetraacetic acid and were stored at −80 • C. DNA extraction and 18 SNPs detection were completed by Sangon Biotech Co, Ltd, Shanghai, China, and purified using the UNIQ-10 Kit (Sangon Biotech Co., Ltd. Shanghai, China).

Genotyping of the 18 SNPs was performed by investigators blinded
to the clinical status of the participants using matrix-assisted laser desorption/ionization-time of flight mass spectrometry (MassARRAY Analyzer software by Sangon Biotech Co, Ltd, Shanghai, China), as previously described (Yi et al., , 2017).

Statistical analysis
Statistical analysis was performed using SPSS 25.0 (SPSS Inc New York, New York, USA). The results were expressed as percentages for categorical variables, and continuous variables were expressed as the mean ± standard deviation. Baseline clinical characteristics and genotype distributions of the 18 variants were compared using χ 2 test (categorical variables) and Student's t-test (continuous variables) between subjects with and without carotid atherosclerosis (increased CCA-IMT, vulnerable plaque).
The allele frequencies for Hardy-Weinberg equilibrium were assessed using χ 2 test. Generalized multifactor dimensionality reduction software (β version 0.7, www.healthsystem.virginia.edu/internet/ addiction-genomics/Software) was used to assess gene-gene interactions among the 18 variants as previously reported (Lou et al., 2007;. Variables that were statistically significant (p < .05) in the univariate analysis were subjected to multivariate logistic regression analysis. All tests were two-sided, and p < .05 was considered statistically significant.

Prevalence of carotid atherosclerosis (increased CCA-IMT, vulnerable plaque)
The clinical characteristics of the 2377 participants were summarized in Table 2. Among 2377 participants in the high stroke risk population, increased CCA-IMT was observed in 445 (18.7%) participants and the vulnerable plaque was observed in 398 (16.7%) participants. In the univariate analysis, increased CCA-IMT was associated with increased age, total cholesterol, low-density lipoprotein (LDL), decreased highdensity lipoprotein (HDL) levels and being male, history of stroke, antiplatelet, antihypertensive therapy, current smoking, lower educational background. Furthermore, the vulnerable plaque was associated with increased age, total cholesterol and being male, history of stroke, history of hypertension, antihypertensive therapy, current smoking, and regular alcohol consumption, lower educational background. In addition, we found that people without increased CCA-IMT and vulnerable plaque was more lacking in physical exercise and higher BMI, considering that since we included people over 40 years old, the elderly may rarely reach the amount of exercise required in the questionnaire and for metabolic reasons, they have a higher BMI.

Genotype distributions in subjects with and without carotid atherosclerosis (increased CCA-IMT, vulnerable plaque)
The genotype distributions of the 18 SNPs assessed in this study were consistent with the Hardy-Weinberg equilibrium (all p > .05). NOS2A rs2297518 SNP was significantly associated with increased CCA-IMT in the single SNPs analysis (Table 3), IL1A rs1609682, and HABP2 rs7923349 SNPs were associated with vulnerable plaque (Table 4).

Logistic regression analysis of risk factors for carotid atherosclerosis (increased CCA-IMT, vulnerable plaque)
Logistic regression analysis found that increased CCA-IMT was associ-  (Table 6). Meanwhile, multivariate logistic regression was performed to determine whether NOS2A rs2297518 SNP, IL1A rs1609682, and HABP2 rs7923349 SNPs increased the risk of carotid intima-media thickening and vulnerable plaque. Variables from the univariate analysis with p < .05 were included in the multivariate logistic regression analysis as covariates.

Gene-gene interactions among the 18 variants and risk for carotid atherosclerosis (increased CCA-IMT, vulnerable plaque)
We evaluated the high-order interaction relationships for the 18 variants with carotid atherosclerosis (increased CCA-IMT, vulnerable plaque) using GMDR analysis. The results showed significant genegene interactions among genes involved in inflammation and endothe-lial function. After adjusting for confounding variables, the best interactive model for carotid atherosclerosis (increased CCA-IMT, vulnerable plaque) was the interaction among TNFSF4 rs1234313, IL1A   rs1609682, TLR4 rs1927911, ITGA2 rs1991013, NOS2A rs2297518,   IL6R rs4845625, ITGA2 rs4865756, HABP2 rs7923349, NOS2A rs8081248, HABP2 rs932650. The cross-validation consistency was 10/10-the sign test was 9 (increased CCA-IMT, p = .0107) (Table 7), and the cross-validation consistency was 10/10-the sign test was 10 (vulnerable plaque, p = .0010) ( Table 8). This indicated that the interactions among those variants strongly synergistically contributed to a higher risk for carotid atherosclerosis (increased CCA-IMT, vulnerable plaque) more than any individual variant alone.
Traditional risk factors for atherosclerosis, such as age, sex, history of cardiovascular disease, dyslipidemia, hypertension, and diabetes mellitus are known to predict atherosclerosis risk in the general population (Legge & Hanly, 2018).
With age, the phenomenon of clonal hematopoiesis of indeterminate potential in the bone marrow increases, the function of thread granules decreases, and IL-6 level in the vasculature increases, all of which promote the formation of atherosclerosis (Tyrrell & Goldstein, 2021).
A study of carotid endarterectomy specimens in men and women showed that the inflammatory infiltration of plaques was more pronounced in men than in age-matched women, and men had more CD68 staining (an inflammatory marker associated with macrophages) compared with women (Man et al., 2020).
Atherosclerosis is an inflammatory reaction accompanied by vas- The importance of lifestyle factors beyond traditional risk factors for the prevention of atherosclerosis is receiving increasing attention (Lechner et al., 2020).
We sought to assess the relationship between lifestyle and carotid atherosclerosis in the high-risk stroke population in Southwest China.
In univariate analysis, we found that smoking and alcohol consumption were closely related to carotid atherosclerosis, adjusting for confounding factors in a multifactor analysis found that smoking increased the risk of carotid atherosclerosis. Cigarette smoking is a strong risk factor for the development of certain chronic inflammatory diseases. Smoking can promote the progression of systemic inflammation and promote the occurrence and development of atherosclerosis (Doran et al., 2003;Chang et al., 2014;Papagoras et al., 2014;Gulati et al., 2016;Rodríguez Huerta et al., 2016). Buamina et al. found that there was an interaction between smoking and genes that may promote the development of carotid atherosclerosis (Maitusong et al., 2023). Reducing tobacco inhalation can reduce the incidence of carotid atherosclerosis.
In this study, we found that people without increased CCA-IMT and vulnerable plaque were more lacking in physical exercise and had higher BMI. But previous research had shown that high leisure-time physical activity as opposed to low physical activity were associated with reduced risk of incident atherosclerosis (Johansson & Acosta, 2020). We considered that it might be because there were a large number of elderly people in the included population, who lost weight due to diseases, aging, and other reasons that could not reach the exercise tolerance described in the questionnaire. Future age-stratified assessment investigations are needed to understand the role of these life factors in atherosclerosis.
The interesting finding was that after adjusting for covariates, lower education was associated with a significantly higher risk of vulnerable plaques. To our knowledge, very limited prior studies on the association between education level and atherosclerosis had been conducted in China. One prospective study showed that the incidence of stroke was significantly higher in those with lower education TA B L E 4 Genotype distribution in individuals with and without vulnerable plaque (%). levels than in those with higher education levels, which may be partly attributed to an insufficient understanding of the importance of proper management of cardiovascular risk factors in the population with low education level, as well as the living environment, living habits, and income level were different (Che et al., 2020). Larger prospective studies are needed to analyze the association between education and carotid atherosclerosis.
Genetic factors play an important role in the development of carotid atherosclerosis (Gardener et al., 2011;Grbić et al., 2022). Multiple risk factors may influence the relationship between the gene polymorphisms and carotid atherosclerosis (Li et al., 2022;Maitusong et al., 2023).
Previous studies showed that inflammation and endothelial function-related functional polymorphisms were associated with ischemic stroke (Cheng et al., 2016;Jiang et al., 2015). However, few studies have focused on carotid atherosclerosis, including carotid IMT and vulnerable plaque. Carotid IMT is an early surrogate marker of subclinical atherosclerosis  and is an important marker of acute ischemic cerebral infarct (Ravikanth, 2019). Besides, the rupture and detachment of vulnerable plaque are also a lead-ing cause of cerebral infarction (Katakami et al., 2010). Therefore, we investigated the association between variants in genes related to endothelial function and inflammatory processes and possible markers of carotid atherosclerosis (increased CCA-IMT, vulnerable plaque) in a population of individuals at high risk for stroke in China.
Inflammation induces the proliferation and migration of vascular smooth muscle cells and participates in the regulation of vascular tension to influence the progression and development of atherosclerosis (Gardener et al., 2011;Zhang et al., 2007). Immune cells congregate in intimal neovasculature and inhibit activation of urokinase, resulting in endothelial damage (Galkina & Ley, 2007;Willeit et al., 2003). Genetic polymorphisms associated with inflammation and endothelial function may directly or indirectly interact with vascular risk factors to influence the onset and progression of atherosclerosis (Galkina & Ley, 2007;Gardener et al., 2011;Zhang et al., 2007). Previous studies have shown that multiple risk factors have an impact on the relationship between genes and carotid atherosclerosis, so it is necessary to analyze the correlation between genes and carotid atherosclerosis by combining multiple factors (Li et al., 2022;Maitusong et al., 2023).

Best model
rs1609682TT and HABP2 rs7923349TT were associated with a higher risk for vulnerable plaque.
However, the interactions between the ten genes identified in this study and carotid atherosclerosis have not been characterized.
TA B L E 6 GMDR analysis of the best models, prediction accuracies, cross-validation consistencies, and p values for vulnerable plaque.  IL1A plays a vital role in immunomodulatory and inflammatory processes. Meanwhile, it promotes arterial plaque formation by regulating the expression of IL-1 in plaque macrophages (Kölsch et al., 2007). IL-1A rs1800587 SNP promotes irregular and carotid plaque formation (Gardener et al., 2011).
TLR4 is a lipopolysaccharide signaling receptor. Activated TLR4 associates with the tumor necrosis factor receptor and can activate NF-κB, extracellular signal-regulated kinases, SAPK/JNK, and p38 MAP kinases. Studies have shown that TLR4 promotes carotid atherosclerosis through increased expression of inflammatory cytokines (Methe et al., 2005).
ITGA2 regulates cell adhesion and cell surface-mediated signaling.
Studies have shown that ITGA2 polymorphism was associated with calcified plaques. Furthermore, ITGA2 polymorphism was associated with the risk of ischemic stroke, carotid intima-media thickening, and plaques in patients with type 2 diabetes (Gardener et al., 2011;Maeno et al., 2002;Matarin et al., 2008).
The NOS2A gene regulates inducible nitric oxide synthase, which produces nitric oxide. Nitric oxide is involved in vascular tone regulation, immune response, and neurotransmission. In humans, inducible nitric oxide synthase has been observed in the cores of carotid plaques, and inhibition of nitric oxide synthase slowed the progression of atherosclerosis in experimental rabbits (Gardener et al., 2011;Hayashi et al., 2006).

IL-6 is secreted by T lymphocytes and epithelial cells and can
promote endothelial cell activation and leukocyte adhesion (Hulkkonen et al., 2009). Studies have shown that IL-6 promoted increased carotid intima-media thickness (Rizza et al., 2020). Interleukin-6 signals through the IL-6 receptor (IL-6R). Therefore, the IL6R gene locus may be associated with carotid IMT (Hassan et al., 2020).
Vascular integrity is regulated by HABP2, which encodes a protein involved in cell adhesion. The haploid test showed that HABP2 rs7923349, rs2302373, and rs2240878 SNPs were associated with plaques (Gardener et al., 2011). HABP2 polymorphism is also associated with the progression of carotid artery stenosis (Willeit et al., 2003).
Therefore, these ten genes may promote the onset and progression of atherosclerosis because they encode and regulate enzymes related to inflammation and endothelial function, which are the main pathogenic mechanisms of atherosclerosis. Future studies will use knockout mice to characterize the molecular mechanisms of interaction among the ten gene variants identified in this study.
Our study was subject to the following limitations. First, the study only included people over 40 years and therefore does not represent the general population of Southwest China. Second, this study was a cross-sectional study and may be subject to recall bias. Third, this study analyzed the associations between gene polymorphism and thickening of the carotid intima-media, vulnerable plaque in a population at high risk for stroke, but did not simultaneously analyze the association between gene polymorphisms and carotid artery stenosis. Finally, our study only included 18 loci of 10 genes; there may be other gene loci related to carotid atherosclerosis. Future studies should include more genes to illustrate the relationships between genes and carotid atherosclerosis.

CONCLUSION
Despite the above limitations, the present study provides clear evidence that the prevalences of increased CCA-IMT and vulnerable plaque were very high in a population of individuals at high risk for stroke in southwestern China. Multivariate logistic regression showed that NOS2A rs2297518GG was associated with carotid intimamedia thickening; IL1A rs1609682TT and HABP2 rs7923349TT were associated with a higher risk for vulnerable plaque. Furthermore, GMDR analysis showed gene-gene interactions among TNFSF4 rs1234313, IL1A rs1609682, TLR4 rs1927911, ITGA2 rs1991013, NOS2A rs2297518, IL6R rs4845625, ITGA2 rs4865756, HABP2 rs7923349, NOS2A rs8081248, and HABP2 rs932650. Future studies are needed to confirm the significance of our findings.

ACKNOWLEDGMENTS
We thank the reviewers for their helpful comments.

CONFLICT OF INTEREST STATEMENT
The authors declare that they have no competing interests.

DATA AVAILABILITY STATEMENT
All data generated during the project will be made available upon the request from the corresponding author. There are no security, licensing, or ethical issues related to these data.