Apolipoprotein E Epsilon 4 Enhances the Association between the rs2910164 Polymorphism of miR-146a and Risk of Atherosclerotic Cerebral Infarction

Abstract

Aim: To analyse the relationship between two potentially functional single-nucleotide polymorphisms (SNPs) of the miR-146a gene (rs2910164 and rs57095329) and the risk of atherosclerotic cerebral infarction (ACI).

Methods: A total of 297 patients with ACI and 300 matched healthy individuals were enrolled in the study. The miR-146a polymorphism was detected using the polymerase chain reaction–restriction fragment length polymorphism method.

Results: A significant difference in the C allele frequency at rs2910164 (p = 0.028) was noted between patients with ACI and control subjects. In contrast, the genotype and allele frequencies of rs57095329 were not statistically associated with ACI. In addition, the decreased expression of miR-146a was significantly more frequent in ACI patients who were ApoEε4 (+) carriers (p = 0.0233), and rs2910164 G> C was intimately associated with the ApoEε4-containing genotype in patients compared with the ApoEε4 (−) carriers (p = 0.0323).

Conclusions: Our findings indicated that the C allele of rs2910164 miR-146a is an important risk factor for ACI, and ApoEε4 may function through attenuating miR-146a expression to enhance ACI susceptibility. This study provides new information about the possible relationship between miR-146a and ApoEε4 in the development of ACI, with potentially important therapeutic implications.

Introduction

Atherosclerotic cerebral infarction (ACI) is a complex disease caused by the combination of multiple risk factors, such as atherosclerosis, hypertension, diabetes, and dyslipidaemia, among which atherosclerosis is considered as the main factor involved in the pathogenesis of ACI¹⁾. Atherosclerosis is characterized by chronic local inflammation of the vascular wall, resulting in the accumulation of lipids and macrophage-derived foam cells in the subendothelial space²⁾. Although the inflammatory nature of atherosclerosis is widely accepted, the complex mechanisms underlying the atherogenic proinflammatory processes remain controversial. Recent findings have suggested that ACI has a significant genetic component^{3, 4)}. Considerable evidence gathered to date has allowed us to identify at least five potential important pathways that may be related to the genetic risk factors of atherothrombosis, including lipoprotein metabolism, inflammation, the renin-angiotensin-aldosterone system (RAAS), platelet biology and function, and blood coagulation and fibrinolysis, contributing to the risk of ACI⁵⁾.

MicroRNAs (miRNAs) are small noncoding RNAs that are 20–23 nucleotides in length and regulate gene expression in a sequence-specific manner⁶⁾. Accumulating evidence has implicated miRNAs as essential regulators of atherosclerosis by targeting important factors or key pathways^{7, 8)}. Abnormal miRNA expression has also been observed in various diseases that are associated with inflammatory and immune processes^9–11). MiR-146a is recognized for its ability to silence multiple inflammatory targets, including IRAK1 and TRAF6, and suppress TLRdriven NF-κB signaling in macrophages and other haematopoietic cells. Reduction of miR-146a levels in these cell lines could enhance the susceptibility to atherosclerosis and sepsis^12–14). The current evidence supports the hypothesis that miR-146a operates as an important epigenetic regulator in various pathways involved in atherosclerosis and stroke. Thus, we were prompted to explore the relationship between miR-146a and ACI^{15, 16)}. Certain single-nucleotide polymorphisms (SNPs) could influence the function of the target genes and are associated with diseases risk¹⁷⁾. Two SNPs in the miR-146a gene, rs2910164 and rs57095329, have functional importance in modifying the expression of mature miR-146a and have been reported to be associated with several inflammation-associated diseases^18–21).

Apolipoprotein E (ApoE) plays an important role in lipid metabolism and cholesterol transport, which are considered to be candidate pathways with antiatherogenic functions^{22, 23)}. Several studies have reported that ApoE polymorphisms serve as a major genetic risk factor for a wide spectrum of inflammatory metabolic diseases, including atherosclerosis, cerebrovascular disorders, and hyperlipoproteinaemia, and they are also well recognized for their ability to suppress atherosclerosis^24–26). Interestingly, it has been recently reported that ApoE regulates the expression of the anti-inflammatory microRNA miR-146a in monocytes and macrophages¹⁴⁾. To the best of our knowledge, no study to date has examined the association of miR-146a polymorphisms with the risk of ACI. In this study, we conducted an association analysis to ascertain whether the two functional polymorphisms of miR-146a contribute to the risk of ACI and further studied the distribution of ApoE genotypes in these study populations.

Methods

Study Population

Our study consecutively recruited 297 patients with ACI (120 women and 177 men; mean age = 62.6 ± 8.63 years) from the Department of Cardiovascular Internal Medicine and the Department of Neurosurgery of the Affiliated second Hospital of Guangdong Medical University and 300 healthy controls (130 women and 170 men; mean age = 61.1 ± 9.58 years) from the medical examination center of the Affiliated Hospital of Guangdong Medical University. Patient diagnoses were verified with either computed tomography (CT) or magnetic resonance imaging (MRI), and all the patients with ACI were classified into subtypes according to the Trial of Org 10172 in Acute Stroke Treatment (TOAST) classification²⁷⁾. All subjects enrolled in our study were Chinese Han with similar dietary pattern, and the patients with histories of ischemic cerebrovascular diseases, cardiogenic cerebral infarctions, cerebral hemorrhage, coronary artery diseases (CAD), autoimmune diseases, systemic inflammatory diseases, blood diseases, and malignant tumors were excluded from the study. All experiments on human subjects were conducted in accordance with the Declaration of Helsinki, written informed consent was obtained from all the enrolled participants, and this study was approved by the Ethics Committee of the Guangdong Medical University.

DNA Isolation and Genotyping

Genomic DNA was isolated from peripheral blood samples using a blood genomic DNA extraction kit (Tiangen, China). The DNA was genotyped using the ABI PRISM SNapShot method (Applied Biosystems, Foster, CA). The polymerase chain reaction (PCR) primers used for the rs2910164 polymorphism were 5′-GAACTGAATTCCATGGGTTG-3′ and 5′-CACGATGACAGAGATATCCC-3′. The primers used for rs57095329 polymorphism were 5′-TCATTGGGCAGCCGATAAAG-3′ and 5′-AGGAAGTTCTGGTCAGGCG-3′. In brief, the SNapShot reaction was performed in a 10-ml final volume containing the SNapShot Multiplex Ready Mix (5 ml), primer mix (0.02–0.6 mmol/L), and templates (4 ml) consisting of the multiplex PCR products, which were purified using the QIAquick PCR Purification Kit (QIAGEN, Hilden, Germany). The cycling program included 25 cycles of 94 °C for 30 s, 57 °C for 30 s, and 72 °C for 40 s. Extension products were purified by a 15-min incubation with 1 U of shrimp alkaline phosphatase (Promega, Madison, WI) at 37°C and a subsequent 15-min incubation at 80°C to inactivate the enzyme. The purified products (0.5 ml) were mixed with 9 ml of formamide and 0.5 ml of GeneScan-120 LIZ size standard (Applied Biosystems) and separated by capillary electrophoresis (ABI PRISM 310 Genetic Analyzer; Applied Biosystems). The results were analyzed using GeneMapper 3.0 software (Applied Biosystems). For quality control, random duplicate samples (5%) were run for each sequence analysis.

Mononuclear Cell Isolation

In total, 30 ApoEε4 (+) carriers and 30 ApoEε4 (−) carriers from among 297 ACI patients and 30 ApoEε4 (+) carriers and 30 ApoEε4 (−) carriers from 300 healthy subjects were chosen randomly for the isolation of mononuclear cells. Peripheral blood mononuclear cells (PBMCs) were isolated using the density gradient centrifugation method with LymphoprepTM (Axis-Shield PoCAS, Oslo, Norway). In brief, blood samples were mixed with an equal volume of 0.9% NaCl. The diluted blood was then slowly added to tubes containing a Ficoll PREMIUM solution to layer the blood on the Ficoll. Samples were centrifuged at 800 × g for 30 min at room temperature. After centrifugation, the mononuclear cells form a distinct band at the medium interface. The cells were then transferred to other tubes using a Pasteur pipette without removing the upper layer and washed with 0.9% NaCl. The samples were then centrifuged again at 250 × g for 10 min. The mononuclear cells were harvested and stored at −80°C. Isolated PBMCs (2 × 10⁶ cells per well) were seeded in 24-well plates and cultured in RPMI medium 1640 supplemented with 10% fetal calf serum (FCS), 100 U/ml penicillin, and 100 mg/ml streptomycin.

RNA Extraction

Total cellular RNA was extracted from PBMCs using the RNAprep Pure Blood Kit (TianGen Biotech, Beijing, China) according to the manufacturer's instructions. In brief, PBMCs were pelleted and lysed using Trizol. The lysate was centrifuged at 13,400 × g for 2 min. The aqueous phase was transferred to a fresh microcentrifuge tube. Ethanol was added to the sample, and then the sample was transferred to the silica surface of a spin column. All contaminating genomic DNA was eliminated by performing RNase-free DNase I digestion on the silica membrane. Total RNA from all samples was further purified by performing the washing steps and then eluting the RNA from the membrane in water. Next, the integrity of the RNA samples was verified by agarose gel electrophoresis and stored at −80°C.

Real-Time PCR

MiR-146a expression was analyzed using the miScript SYBR Green PCR kit (Qiagen, Germany) according to the manufacturer's instructions (Qiagen, Germany). The assay was performed in triplicate. Human U6 was used as the internal control. The comparative threshold cycle (Ct) method was used to assess the relative changes in expression. In brief, 2^ΔDDCt represents the fold change in miR-146a expression between sample groups. A no-template control was run with the positive samples to assess the overall specificity of the reaction¹⁰⁾. The expression levels of IRAK-1 were analyzed using the miScript SYBR Green PCR kit. The IRAK-1 expression levels were analyzed in triplicate, and the expression was normalized to the level of glyceraldehyde 3-phosphate dehydrogenase (GAPDH), which was used as an internal control. The expression levels of IRAK-1 were also calculated using the 2^ΔDDCt method.

Plasma Lipid Measurements

Blood samples for lipid measurements were withdrawn from subjects after an overnight fast. The plasma triglycerides (TG), total cholesterol (TC), high-density lipoprotein (HDL) cholesterol, and low-density lipoprotein (LDL) cholesterol levels were measured using a Hitachi 7600 automatic analyser (Hitachi Instruments Corporation, Tokyo, Japan).

Statistical Analyses

Statistical analyses were performed using SPSS software, version 19.0 (IBM, Armonk, NY, USA). These data are presented as percentage frequencies or means ± standard deviation (SD). Allele frequencies were calculated from the genotypes of all subjects. The allele and genotype frequencies of miR-146a between the patients and control subjects were compared using Fisher's exact test or chi-squared test. Hardy–Weinberg equilibrium (HWE) was assessed using HWE software. Risk factors were screened using Student's t-test or chi-squared test. Analyses were repeated in subgroups stratified by miR-146a and ApoEε4 carrier status. Associations were expressed as odds ratios (ORs) or risk estimates with 95% confidence intervals (CIs). The relationships between different genotypes of miR-146a and ACI were evaluated using analysis of variance (ANOVA). A p value < 0.05 was considered to be statistically significant. Power analysis was performed using QUANTO 1.2 software.

Results

Clinical Characteristics

The clinical characteristics of all the participants in the study are summarized in Table 1. Of the 597 participants, 297 were ACI patients and 300 were healthy controls. The mean age was 62.6 (± 8.63) years for the ACI patients and 61.1 (± 9.58) years for the control subjects, and the gender (male-to-female) ratio was 1:1.48 in the case group and 1:1.31 in the control group. No statistically significant difference was noted between the patients and controls in terms of sex (p = 0.468) or age (p = 0.064). The risk factors examined (e.g., hypertension and diabetes) were significantly more common in the ACI group compared with the control group. Moreover, compared with the controls, the patients with ACI exhibited increased TC (p = 0.028), HDL (p = 0.015), and LDL (p = 0.043). Individuals with hypertension (p < 0.001) and diabetes (p < 0.001) exhibited increased risk of ACI. However, patients with ACI exhibited a trend of increased total TG compared with controls, but no significant difference was noted (p = 0.075). Moreover, as expected, ApoEε4 allele frequencies were significantly elevated in ACI patients compared with controls (p < 0.001). Power analysis revealed that, with our sample size, we would have 93.2% power for rs2910164 and 78.1% power for rs57095329 to detect a genotype relative risk with an OR of 1.5 at a significance level of 0.05.

Table 1. Characteristic of Controls and ACI Patients.

Clinic data	ACI patients (297)	Controls (300)	P value
Mean ages (years)	62.6 ± 8.63	61.1 ± 9.58	0.064
Male/Female	177/120	170/130	0.468
BMI (kg/m2)	22.6 ± 3.41	23.35 ± 2.51	0.754
History of hypertension, n (%)	197 (66.33%)	66 (22.00%)	< 0.001*
History of diabetes, n (%)	62 (20.86%)	11 (3.67%)	< 0.001*
HDL (mmol/L)	1.19 ± 0.47	1.25 ± 0.31	0.015*
LDL (mmol/L)	2.85 ± 1.06	2.91 ± 0.86	0.043*
Total Cholesterol (mmol/L)	4.91 ± 0.97	4.59 ± 0.62	0.028*
Total Cholesterol (mmol/L)	1.84 ± 1.04	1.62 ± 0.63	0.075
ApoEε4, n (%)	94 (31.65%)	37 (12.33%)	< 0.001*

BMI, body mass index; HDL, high-density lipoprotein; LDL, low-density lipoprotein.

^*p < 0.05.

MiR-146a Polymorphisms in ACI Patients and Controls

All the enrolled participants were successfully genotyped for the rs2910164 and rs57095329 SNPs. All genotype distributions were in HWE in the ACI patients and controls (Supplement Table 1). The genotype and allele frequencies of miR-146a polymorphisms in the study are presented in Table 2. Although a trend was noted, no statistical association with the genotype frequencies of the rs2910164 SNP with ACI patients and controls was observed (p = 0.052). In contrast, a significant difference in the C allele frequency of the rs2910164 SNP was noted (p = 0.028). Moreover, a significant difference was observed in the ACI patients compared with the controls in a recessive model (GG+GC. vs. CC) (p = 0.015), which indicated that the rs2910164 polymorphism is a risk factor for ACI (Table 2). Regarding the rs57095329 polymorphism, no statistically significant difference was observed between the ACI patients and healthy controls for either the genotype or allele frequencies of the rs57095329 SNP (genotype frequencies p = 0.620, allele frequencies p = 0.354).

Supplement Table 1. The Hardy-Weinberg equilibrium assay for rs653765 and rs514049 genotypes in the ACI cases and controls.

Rs2910164	CC	GC	GG	P-value
ACI patients	141	128	28	0.892
Control	113	152	35	0.133
Rs57095329	GG	GA	AA
ACI patients	200	87	10	0.887
Control	211	82	7	0.770
ApoEε4(+)
Rs2910164	CC	GC	GG
ACI patients	59	28	7	0.170
Control	16	16	5	0.755
Rs57095329	GG	GA	AA
ACI patients	66	27	1	0.328
Control	25	10	2	0.469
ApoEε4(-)
Rs2910164	CC	GC	GG
ACI patients	82	100	21	0.237
Control	97	136	30	0.086
Rs57095329	GG	GA	AA
ACI patients	134	60	9	0.495
Control	186	72	5	0.514

Table 2. Genotype and allele Distributions of the miR-146a polymorphisms between controls and ACI Patients.

rs2910164 G > C	ACI patients (297), n (%)	Controls (300) n (%)	OR (95% CI)	P-value
CC	141 (47.47%)	113 (37.67%)		0.052
GC	128 (43.10%)	152 (50.66%)
GG	28 (9.43%)	35 (11.67%)
GC/GG	156 (52.52%)	187 (62.33%)	0.667 (0.482–0.927)	0.015*
GC/CC	269 (90.57%)	265 (88.33%)	1.27 (0.750–2.15)	0.373
C alleles	410 (69.02%)	378 (63.00%)	0.764 (0.601–0.972)	0.028*
G alleles	184 (30.98%)	222 (37.00%)
rs57095329 G > A
AA	200 (67.34%)	211 (70.33%)		0.620
GA	87 (29.29%)	82 (27.33%)
GG	10 (3.37%)	7 (2.33%)
GA/GG	97 (32.66%)	89 (29.66%)	1.15 (0.813–1.63)	0.480
GA/AA	287 (95.63%)	293 (97.67%)	0.686 (0.257–1.83)	0.472
A alleles	487 (81.99%)	504 (84.00%)	1.153 (0.852–1.561)	0.354
G alleles	107 (18.01%)	96 (16.00%)

Adjusted for age, sex, hypertension and diabetes.

^*p < 0.05.

MiR-146a Polymorphisms in ApoE Genotype Groups in ACI Patients and Controls

The prevalence of the ApoEε4-containing phenotypes was significantly increased in individuals with cardiovascular and cerebrovascular ailments, such as myocardial infarction, hypertension, coronary heart disease, and stroke. We examined whether a significant association exists between miR-146a and ApoE4 isoforms in ACI patients. As shown in Table 3, when these data were stratified for the presence or absence of the E4 isoform, the C allele of rs2910164 remained significantly different between ACI patients and controls in the ApoEε4(+) subgroup (p = 0.033) but not in the ApoEε4(−) groups (p = 0.472). In the recessive model (GG+GC. vs. CC), a significant difference was observed in ApoEε4(+) patients compared with ApoEε4(+) controls (p = 0.04). However, no significant difference was detected in the ApoEε4(−) subgroups (p = 0.440). Besides, neither the genotype nor the allele frequencies in rs57095329 exhibited significant differences in the ApoEε4 mutation subgroups [ApoEε4(+) genotype and allele frequencies: p = 0.326 and p = 0.492, respectively; ApoEε4(−) genotype and allele frequencies: p = 0.222 and p = 0.146, respectively] (Table 3).

Table 3. Genotype and allele distributions in ApoE genotype groups with ACI patients and controls.

rs2910164 (n)	Genotypes n (%)
ApoEε4(−)	CC	GC	GG	p-value
ACI (94)	59 (62.77%)	28 (29.79%)	7 (7.45%)	0.119
Controls (37)	16 (43.24%)	16 (43.24%)	5 (13.51%)
ApoEε4(−)
ACI (203)	82 (40.39%)	100 (49.26%)	21 (10.34%)	0.733
Controls (263)	97 (36.88%)	136 (51.71%)	30 (11.41%)
Rs57095329 (n)	Genotypes n (%)
ApoEε4(−)	AA	GA	GG	p-value
ACI (94)	66 (70.21%)	27 (28.72%)	1 (1.06%)	0.326
Controls (37)	25 (67.57%)	10 (27.03%)	2 (5.41%)
ApoEε4(−)
ACI (203)	134 (66.01%)	60 (29.56%)	9 (4.43%)	0.222
Controls (263)	186 (70.72%)	72 (27.38%)	5 (1.90%)
rs2910164 (n)	Alleles n (%)
ApoEε4(−)	G	C	p-value	OR (95% CI)
ACI (94)	42 (22.24%)	146 (77.66%)	0.033*	0.531 (0.295–0.956)
Controls (37)	26 (35.14%)	48 (64.87%)
ApoEε4(−)
ACI (203)	142 (34.98%)	264 (65.03%)	0.472	0.906 (0.691–1.186)
Controls (263)	196 (37.26%)	330 (62.74%)
Rs57095329 (n)	Alleles n (%)
ApoEε4(−)	G	A	p-value	OR (95% CI)
ACI (94)	29 (15.43%)	159 (84.57%)	0.492	0.782 (0.387–1.580)
Controls (37)	14 (18.92%)	60 (81.08%)
ApoEε4(−)
ACI (203)	78 (19.21%)	328 (80.79%)	0.146	1.288 (0.915–1.811)
Controls (263)	82 (15.59%)	444 (84.41%)
rs2910164 (n)	Recessive model
ApoEε4(−)	GG−GC .vs. CC	p-value		OR (95% CI)
ACI (94)	35 (37.24%)	0.04*		0.452 (0.209–0.980)
Controls (37)	21 (56.75%)
ApoEε4(−)
ACI (203)	121 (59.61%)	0.440		0.862 (0.5920–1.256)
Controls (263)	166 (63.12%)
Rs57095329 (n)	Recessive model
ApoEε4(−)	GG−GA. vs. AA	p-value		OR (95% CI)
ACI (94)	28 (29.897%)	0.767		0.884 (0.390–2.00)
Controls (37)	12 (32.44%)
ApoEε4(−)
ACI (203)	69 (33.99%)	0.277		1.24 (0.8391–1.844)
Controls (263)	77 (29.28%)

Adjusted for age, sex, hypertension and diabetes.

^*p < 0.05.

The Influence of ApoEε4 on miR-146a Expression in ACI Patients

From the case-control study above, we found that rs2910164 of miR-146a exhibited a high-risk genotype in ApoEε4 patients with ACI. To further explore the influence of ApoEε4 on the expression of mature miR-146a, we examined the expression levels of the rs2910164 miR-146a genotypes in PBMCs that were obtained from the ACI patients who were E4 allele carriers. According to the data obtained, the mean level of miR-146a mRNA expression was significantly reduced in ApoEε4(+) patients with ACI compared with ApoEε4(−) patients (p = 0.0233), and the mean level of miR-146a mRNA was also significantly reduced in ApoEε4(+) healthy control compared with ApoEε4(−) control (p = 0.039) (Fig. 1A). Moreover, regarding ACI patients with ApoEε4(+), individuals with the CC genotype for rs2910164 exhibited significantly lower levels of miR-146a than those with the GG or GC genotype (p = 0.0323) (Fig. 1B). However, the mean level of miR-146a revealed no significant difference between the CC and GG+GC genotypes for rs2910164 in ACI patients who were ApoEε4(−) carriers, suggesting that the CC genotype for rs2910164 has a stronger influence on mature miR-146a expression in ACI patients with ApoEε4(+) but not ApoEε4 (−). Furthermore, patients with the CC genotype expressed significantly higher relative expression of IRAK-1 in ApoEε4(+) patients (p = 0.048; compared with GG/GC, Fig. 1C); however, no statistically significant difference was observed in ApoEε4(−) subjects. The comparison of genotype distributions between the ACI subjects revealed a statistical association between the rs2910164 polymorphism of miR-146a and risk of ACI. Furthermore, ApoEε4 may alter the effect of C allele mutations in rs2910164 on decreasing miR-146a expression, further contributing to ACI susceptibility.

Fig. 1.

The relative expression level of miR-146a with rs2910164 genotypes in ACI patients and healthy controls. (A) The mature miR-146a expression in PBMCS was analyzed in ACI patients with ApoEε4 (+/−); (B) miR146a expression level with rs2910164 genotypes in ACI patients with ApoEε4 genotypes; (C) the relative mRNA expression of IRAK1 with rs2910164 genotypes in ACI patients with ApoEε4 genotypes. The horizontal line indicates the mean expression level.

To further explore whether rs2910164 SNP may influence lipid metabolism and cholesterol transport that has been proved to be associated with ApoE, we also evaluated the possible association between the rs2910164 SNP of miR-146a and TC/total TG levels in patients with or without ApoEε4(+). However, neither the TC nor the total TG levels exhibited a statistically significant difference between the ApoEε4(+) and ApoEε4(−) groups. In addition, no significant difference was observed between the CC and GG+ GC genotypes for rs2910164 miR-146a (Fig. 2).

Fig. 2.

The levels of total cholesterol (A) and total triglycerides (B) with rs2910164 genotypes were analyzed in ACI patients. N presents the numbers of the patients of ACI.

Discussion

MiR-146a is a powerful innate immune and proinflammatory-related regulator of immune and inflammatory responses. MiR-146a drives PBMCs toward Th1 polarization, and the levels of this particular miRNA are increased in patients with diseases such as acute coronary syndrome²⁸⁾. MiR-146a is upregulated in THP-1 cells, macro-phages, Langerhans cells, and monocytes after induction by lipopolysaccharide (LPS), TNF-α, IL-1β, or TGF-β1^{29, 30)}. Furthermore, miR-146a is a novel regulator of vascular smooth muscle cell (VSMC) fate, which is crucial in the pathogenesis of proliferative cardiovas-cular diseases, including atherosclerosis and post-angioplasty restenosis³¹⁾. These data suggest that miR-146a is associated with the development of atherosclerosis, which is one of the important risk factors for ACI^{32, 33)}. However, few studies to date have examined the relationship between miR-146a and ACI. Here we focused on identifying the possible association between miR-146a functional polymorphisms and other risk factors of ACI in a case-control study.

SNPs present in precursor and mature miRNAs influence the levels of mature miRNAs and are associated with various diseases^34–36). In our study, two functional SNPs in miR-146a were chosen to evaluate the association between miR-146a polymorphisms and ACI risk. The two SNPs in miRNA-146a, rs2910164 and rs57095329, have been reported to influence the expression level of mature miR-146a; thus, these SNPs may ultimately affect an individual's susceptibility to disease^{10, 37, 38)}. Our results revealed that the rs2910164 polymorphism was associated with an increased risk of ACI; the C allele of rs2910164 was more frequent in patients with ACI compared with the G allele, indicating that the C allele of rs2910164 is associated with an increased risk of ACI susceptibility. The SNP rs2910164 G/C is located in the stem region, and the C allele creates a mismatch in the stem structure of pre-miR-146a. This mismatch results in a decrease in the total levels of mature miR-146a and influences the transcription of target genes, ultimately leading to disease pathogenesis^{17, 18)}. MiR-146a is assumed to be protective against atherosclerosis given its ability to suppress NF-κB signaling, the TLR4 signaling pathway, and the apparent activation of endothelial cells. miR-146a levels were elevated in human atherosclerotic plaques as well as in the circulation and plaques of atherosclerotic mice because of the strong activation of NF-κB signaling in plaques, implying that miR-146a is induced as part of a negative feedback loop^{32, 39, 40)}. Xiong et al. stated that the GC and CC genotypes of the miRNA-146a rs2910164 polymorphism are associated with an increased risk of CAD⁴¹⁾, whose risk subsequently increases with the presence of metabolic syndrome induced by inflammation^{42, 43)}. In this study, we confirmed that patients with reduced miR-146a expression exhibit an increased risk of ACI. In this study, significantly reduced miR-146a mRNA levels were observed in ACI subjects. Considering the supporting evidence that the decreased miR-146a levels are also associated with the development of cardiovascular diseases^{32, 33, 41)}, we speculate that the C allele of rs2910164 may reduce miR-146a expression and subsequently weaken anti-inflammatory action in the pathogenesis of ACI, thus contributing to the pathological process and increased risk of ACI.

Another novel finding of this study was that ACI patients with the ApoEε4 allele exhibited reduced miR-146a expression compared with controls. More interestingly, CC genotype carriers exhibited reduced miR-146a expression compared with GG or GC genotype carriers. The ApoEε4(+) group contained 2.5-fold more ACI patients than controls, whereas the ApoEε4(−) group included 1.4-fold more controls than ACI patients. It appears that carrying the E4 allele attenuates the level of miR-146a, which increases the risk of ACI. Moreover, the CC genotype of the miRNA-146a rs2910164 polymorphism exhibited a more significant effect in the ApoEε4(+) group compared with the ApoEε4(−) group. ApoE is an important regulator of plasma lipoprotein metabolism and cholesterol homeostasis, and variants of ApoE are major genetic risk factors for a wide spectrum of inflammatory metabolic diseases^{44, 45)}. Individuals who carry at least one copy of the ApoEε4 allele have an increased risk of developing atherosclerosis, which is an accumulation of fatty deposits and scar-like tissue in the lining of the arteries that narrow the arteries, thus increasing the risk of ACI⁴⁶⁾. MiR-146a has been reported to negatively regulate multiple inflammatory targets, including IRAK1 and TRAF6, and to suppress NF-κB-driven TNF-α expression^{47, 48)}. As shown in our results, the reduction of miR-146a expression increased IRAK1 levels significantly in CC genotypes in ApoEε4(+) carries. Reduction of miR-146a expression can reverse the effect of NF-κB activity inhibition⁴⁹⁾. ApoE controls inflammation by suppressing inflammatory NF-κB signaling²⁴⁾. Mechanistically, ApoE increases the expression of the transcription factor PU.1, which increases miR-146 levels to suppress NF-κB signaling. Li et al. demonstrated that ApoE regulates the expression of miR-146a in monocytes and macrophages to repress NF-κB signaling in these cells, and intravascular delivery of miR-146a mimetics can inhibit atherogenesis in mouse models¹³⁾. However, the ε4 gene polymorphism causes strong dysfunction of ApoE, which suppresses the level of miR-146a^{13, 38, 50)}. Based on the finding that ApoEε4 induced decreasing expression of miR-146a in ACI patients, we suggest that the downregulation of NF-κB-dependent pathways by miR-146a through a critical negative feedback regulatory loop was attenuated by the effects of ApoEε4. Furthermore, considering the important roles of ApoE in lipid and cholesterol metabolism, we explored whether rs2910164 influences TC and TT levels. Although no significant association was observed, we speculate that the rs2910164 polymorphism, with its synergistic effect with ApoE, may contribute to the risk of ACI but not through the lipid and cholesterol pathway. Nevertheless, the mechanisms of the interaction of ApoEε4 dysfunction with the CC genotype or C allele of rs2910164 should be further examined.

Conclusions

This is the first study to identify a significant association between the rs2910164 polymorphism of miR-146a and risk of ACI in the south Chinese Han population. The C allele of rs2910164 was associated with the E4-containing phenotype of ApoE, which plays an important role in enhancing ACI susceptibility. This study provides new information about the relationship between miR-146a and ApoE in the development of ACI. However, more work is still required to shed light on the role of miR-146a and ApoE in the pathogenesis of ACI and further clarify its prognostic and therapeutic potential.