Single-Nucleotide Polymorphisms in the 3' Untranslated Region of CORIN Associated With Cardiovascular Diseases in a Chinese Han Population: A Case–Control Study

Background: Corin is a transmembrane serine protease that activates pro-forms of atrial and brain natriuretic peptides. Numerous studies have indicated that corin played an important role in cardiovascular diseases (CVDs). However, there have been few studies about the correlation between single-nucleotide polymorphisms (SNPs) in the 3' untranslated region (3'UTR) of CORIN and CVDs. The aims of this study were to investigate the associations of three SNPs (rs3749585, rs4695253, and rs12641823) in the 3'UTR of CORIN with CVDs and to find the seed regions of microRNAs (miRNAs) that bind to SNPs of CORIN. Methods and Results: A case–control study (n = 3,537) was performed in a Han population of northeastern China. CVDs included essential hypertension (EH), atrial fibrillation (AF), heart failure (HF), and coronary artery disease (CAD). Genotyping was performed using high-resolution melt analysis. In the EH-control study, rs3749585T was significantly associated with the risk of EH after adjusting for sex and age in allelic (padj = 0.049; OR: 1.113) and dominant (padj = 0.015, OR: 1.233) models. Rs4695253T was significantly associated with the risk of EH in the recessive model after adjusting for sex and age (padj = 0.005, OR: 2.084). Rs3749585T was significantly and negatively associated with AF in the dominant and additive models after adjusting for sex, age, EH, HF, T2DM, and CAD (dominant: padj = 0.009, OR: 0.762; additive: padj = 0.048, OR: 0.873). In the HF-control study and CAD-control study, none of the three SNPs was associated with HF and CAD after adjusting for covariates in any models (padj > 0.05). The levels of high-density lipoprotein (HDL) in rs4695253CC+CT were lower than the levels of HDL in rs4695253TT (42.47 ± 10.30 vs. 48.0 ± 10.24 mg/dl, padj = 0.008). The levels of total cholesterol (TC) in rs4695253CC+CT were lower than the levels of TC in rs4695253TT (164.01 ± 49.15 vs. 180.81 ± 43.92 mg/dl, padj = 0.036). Luciferase assay revealed that the relative luciferase activity of rs3749585CC-transfected cells was significantly decreased by miR-494-3p, in comparison to cells transfected with rs3749585TT (p < 0.001). A significant decrease in the relative luciferase activity of rs3749585TT reporter was observed as compared with rs3749585CC reporter in the presence of miR-1323 or miR-548o-3p (p = 0.017 and 0.012, respectively). Conclusions: We found significant associations between rs3749585T and rs4695253T and EH, between rs4695253T and the levels of TC and HDL, and between rs3749585T and AF. Hsa-miR-494-3p may serve as a potential therapeutic target for EH and AF patients in the future.


INTRODUCTION
Corin is a transmembrane serine protease made primarily in atrial and ventricular cardiomyocytes (1,2) and converts proatrial natriuretic peptide (pro-ANP) and pro-brain natriuretic peptide (pro-BNP) into biologically active ANP and BNP in cellbased studies (3), the hormones that regulate blood pressure by promoting natriuresis, diuresis, and vasodilatation (4, 5) and suppress renin and endothelin release (6). ANP is also known for its ability to modulate the electrical activities of the heart, its effects on specific ion channels (7,8), and its suppressive effect on cardiac fibroblast growth during cardiac hypertrophy (9). In humans, CORIN variants (I555/P568) are associated with hypertension and cardiac hypertrophy (10)(11)(12). Heart failure (HF) patients with the two variants had impaired pro-BNP processing and poor clinical outcomes (13). In mice, lack of corin prevented natriuretic peptide processing and resulted in hypertension (14); corin deficiency caused cardiac hypertrophy during pregnancy (15). Moreover, in many clinical studies, serum or plasma soluble corin concentrations were changed in the patients with cardiovascular diseases (CVDs) and the levels of soluble corin were associated with the risk and prognosis of CVDs. Plasma concentrations of corin were decreased in HF (16,17) and hypertensive patients (18). Patients with atrial fibrillation (AF) were found to have higher serum corin levels than those without AF (19). A large study involving 856 patients with acute myocardial infarction and 856 healthy subjects reported that serum corin concentrations were significantly decreased in the former (20). Plasma corin concentrations, assessed 2 days after reperfused ST-segment elevation myocardial infarction, correlated with infarct size as assessed by cardiac magnetic resonance (21). Furthermore, plasma corin can be a biomarker for the prognosis of acute myocardial infarction (22) and HF (23). The significance of corin seems more important in HF. HF is associated not only with reduced plasma corin levels (16,17,24) and decreased myocardial CORIN expression (25), but also with impaired cleavage of pro-ANP (17). In mice with progressive dilated or ischemic cardiomyopathy, restoration of suppressed cardiac corin expression levels reduces myocardial fibrosis, improves contractile function, reduces alveolar edema and congestion, and attenuates heart failure (26)(27)(28)(29).
Previous studies have demonstrated the association between CORIN variants and CVDs. Most of them mainly focused on the exons and the intron/exon boundaries (30-50 nucleotides into the intronic regions) (30,31). However, the association between single-nucleotide polymorphisms (SNPs) in the 3' untranslated regions (3' UTR) of CORIN and CVDs is unknown. Two studies reported the correlation between a SNP, rs3749585, which is located in the 3'UTR of CORIN, and hypertension, but the results were not conclusive (32,33).
Posttranscriptional regulation of gene expression by microRNAs (miRNAs) requires base-pairing between miRNAs and target mRNAs. Target recognition is crucially dependent on Watson-Crick base-pairing between the seed region of miRNA and miRNA recognition elements located in the 3' UTR of target mRNAs. The resulting translational inhibition or mRNA degradation contributes to the downregulation of the corresponding target protein (34,35). Cardiac-specific miRNAs have been shown to regulate a number of critical cellular functions in the developing and normal adult heart and CVDs. If there are allelic variants in the miRNA recognition elements of CORIN 3' UTR, different kinetics of CORIN downregulation may exist. The aims of this study are to evaluate the association between SNPs in the 3' UTR of CORIN and CVDs and to find the seed regions of miRNAs that bind to SNPs of CORIN.

Study Subjects
The study subjects are from the Gene ID population, which is a large current Chinese database with clinical data and blood samples from >10,000 Chinese northeastern patients and controls that can be used to identify susceptibility genes for various CVDs. Written informed consent was obtained from the participants when they agreed to be enrolled in the Gene ID database. All of the subjects were of Han ethnic origin by selfdescription. This study was approved by the Ethics Committee of The First Affiliated Hospital of Dalian Medical University and conformed to the guidelines set forth by the Declaration of Helsinki.
CVDs in the study included essential hypertension (EH), AF, HF, coronary artery disease (CAD), and left ventricular hypertrophy (LVH). The diagnostic criteria for EH were in accordance with 2018 Chinese Guidelines for Prevention and Treatment of Hypertension (36) and defined as a systolic blood pressure of ≥140 mmHg or a diastolic blood pressure of ≥90 mmHg. The diagnostic criteria for AF were in accordance with 2019 AHA/ACC/HRS Focused Update of the 2014 AHA/ACC/HRS guideline for the Management of Patients With Atrial Fibrillation (37) and diagnosis determined using ECG or Holter recordings, regardless of clinical symptoms. The diagnostic criteria for HF were in accordance with Chinese guidelines for the diagnosis and treatment of heart failure 2018 (38) and the etiologies of HF in the study included EH, AF, and CAD. The diagnostic criteria for CAD were from 2015 ESC guidelines for the management of acute coronary syndromes in patients and 2019 ESC guidelines for the diagnosis and management of chronic coronary syndromes (39,40).
LVH was diagnosed if the thickness of the left ventricular posterior wall (LVPW) and interventricular septum (IVS) was >12 mm and the ratio of IVS to LVPW was <1.3. Any patients with secondary hypertension, thyroid disorder, valvular heart disease (causing secondary AF), aortic valve stenosis and hypertrophic cardiomyopathy (causing ventricular hypertrophy), or HF etiologies other than EH, AF, and CAD were excluded. Each subject was evaluated by symptoms, blood pressure, BNP, electrocardiogram, echocardiogram, and computed tomography coronary angiography, and the subjects in the control group were free from the abovementioned CVDs.
MiRNAs mimics were used to upregulate the expression of miRNAs, and negative control (NC) was used as control. MiRNAs were purchased from GenePharma Limited Company (Shanghai, China).

Dual-Luciferase Reporter Assay
HEK293 cells (2 × 10 5 cells/well) were maintained in 24well plates and co-transfected with luciferase reporter plasmids and mimics using Lipofectamine 2000 Transfection reagent (Thermo Fisher Scientific, Waltham, MA, USA). After 24 h of incubation at 37 • C with 5% CO 2 , the firefly luciferase activity was measured through the Dual-Glo luciferase report system (Beyotime Biotechnology, Shanghai, China) and calculated by normalization to the Renilla luciferase control.  different for each CVD. The EH group was corrected for sex and age; the AF group was corrected for sex, age, EH, HF, T2DM, and CAD; the HF group was corrected for sex, age, EH, CAD, AF, and T2DM; the CAD group was corrected for sex, age, EH, and T2DM. When the hypertensive subjects were divided into the LVH subgroup and the non-LVH subgroup, basic statistical methods were applied as described. Associations between SNPs and lipid levels under allelic and genetic models were analyzed by linear logistic regression; β and 95% CI were also calculated. In addition, we performed one-way ANOVA analysis to test the ratios of dual-luciferase activity in different groups. Two-tailed p < 0.05 was accepted as statistically significant.

Clinical Characteristics
There was no deviation from the HWE for rs3749585, rs4695253, and rs12641823 in either case or control groups (p > 0.001, Supplementary Table 2). Because rs3749584 deviated significantly from the HWE, we did not show the associations between rs3749584 and CVDs. The total included 3,537 subjects. In the EH group, there were 1,495 cases and 2,042 controls; in the AF group, there were 618 cases and 2,919 controls; in the HF group, there were 464 cases and 3,073 controls; in the CAD group, there were 715 cases and 2,822 controls. The mean age of the patients in each group was older than that of the controls (p < 0.001, Table 1). The proportion of T2DM in the case group was higher than that in the control group (p < 0.001, Table 1). The distribution of gender between cases and controls in each group was similar (p > 0.05, Table 1).

Analysis of Associations of SNPs in the 3' UTR of CORIN With EH
In the EH-control study, only rs3749585 T was significantly associated with the risk of EH [p adj = 0.049; OR: 1.113 (95% CI, 1.00-1.238);  Table 3]. SNP rs4695253 T was significantly associated with the risk of EH in the recessive model after adjusting for  Table 3]. SNP rs12641823 A failed to show any significant association with risk of EH in the study (p adj > 0.05, Tables 2, 3) under allelic and genotypic inheritance models. We divided the hypertensive group into the LVH subgroup (n = 590) and the non-LVH subgroup (n = 905) according to echocardiography. We found that all the three SNPs, rs3749585 T , rs4695253 T , and rs12641823 A , were unrelated to the risk of LVH in the hypertensive group after adjusting for sex and age both under allelic and genetic models (p adj > 0.05, Supplementary Tables 3, 4).

Analysis of Allelic and Genotypic Associations of SNPs in the 3' UTR of CORIN With AF, HF, and CAD
In the AF-control study, HF-control study, and CAD-control study, all the three SNPs, rs3749585 T , rs4695253 T , and rs12641823 A , were unrelated to the risk of AF, HF, and CAD after adjusting for covariates (p adj > 0.05, Supplementary Table 5).
In the HF-control study and CAD-control study, none of the three SNPs, rs3749585 T , rs4695253 T , and rs12641823 A , was associated with HF or CAD after adjusting for covariates in three inheritance models (p adj > 0.05, Table 4).

Analysis of Associations of SNPs in the 3' UTR of CORIN With Lipid Levels
Lipid levels included TC, TG, LDL, and HDL. We found no association between any of the three SNPs, rs3749585 T , rs4695253 T , and rs12641823 A , and lipid levels in the allelic model (p adj > 0.05, Supplementary Table 6).

Hsa-miR-494-3p Targeting rs3749585 CC and Downregulating the Expression of Corin
We investigated whether miR-494-3p, miR-548o-3p, and miR-1323 directly target the rs3749585 of CORIN using a dual luciferase assay. The luciferase assay revealed that miR-494-3p mimics significantly decreased the luciferase activity of rs3749585 CC -transfected cells, in comparison to cells transfected with rs3749585 TT (p < 0.001, Figure 3). A significant decrease in the relative luciferase activity of rs3749585 TT reporter, compared with rs3749585 CC reporter, was observed in the presence of miR-1323 or miR-548o-3p (p = 0.017 and 0.012, respectively, Figure 3). The results demonstrated that miR-494-3p could target gain rs3749585 CC and reduce the expression of CORIN. MiR-1323 and miR-548o-3p could lose target with rs3749585 CC and had no effect on the expression of CORIN.

DISCUSSION
In the present study, we investigated the association between three SNPs in the 3' UTR of CORIN and CVDs (EH, AF, HF, and CAD) in a Chinese Han population. We also examined the relationship between the SNPs of CORIN and lipid levels. To the best of our knowledge, this is the first time that genetic association between SNPs in the 3'UTR of CORIN and CVDs or lipid levels has been observed. Our study demonstrated FIGURE 2 | The seed regions of miRNAs that bind to rs3749585 and rs4695253 of CORIN. (A) hsa-miR-494-3p gains target with rs3749585; (B,C) hsa-miR-1323 and hsa-548o-3p lose target with rs3749585; (D) hsa-miR-561-3p and hsa-miR-548at-5p gain target with rs4695253; (E) hsa-miR-550b-3p, hsa-miR-6507-3p, hsa-miR-1264, hsa-miR-550a-3p, hsa-miR-3184-3p and hsa-miR-200c-5p lose target with rs4695253. that rs3749585 T and rs4695253 T were associated with EH and that rs4695253 T was associated with blood levels of HDL and TC. In addition, rs3749585 T was significantly associated with protection from AF. We revealed that hsa-miR-494-3p could target rs3749585 CC and downregulate the expression of CORIN using dual-luciferase reporter assay.
Bioinformatic analysis showed that the three SNPs, rs3749585 T , rs4695253 T , and rs12641823 A , were functional SNPs, and rs3749585 T was also a tagSNP. In our study, rs3749585 T was significantly associated with the risk of EH after adjusting for covariates sex and age in allelic [p adj = 0.049; OR: 1.113 (95% CI, 1.00-1.238);  (32), which is consistent with our results. It should be noted that the populations of that study came from the Northeast FIGURE 3 | The relative luciferase activity of different miRNAs that gain or lose target with rs3749585 TT/CC . **A significant decrease in the relative luciferase activity of rs3749585 CC reporter was detected, compared with rs3749585 TT reporter in the presence of miR-494-3p (P < 0.001). A significant decrease in the relative luciferase activity of rs3749585 TT reporter was observed, compared with rs3749585 CC reporter in the presence of miR-1323 or miR-548o-3p (P = 0.017 and 0.012, respectively).
of China, which is the same geographic region from which the population in our study came. However, another study reported that rs3749585 was found to be unrelated with hypertension in 731 hypertensive patients and 731 controls from Jiangsu Province, located in the south of China [p = 0.852, OR: 0.98 (95% CI, 0.85, 1.14)] (33). Differing genetic backgrounds may underlie the discrepancy.
Studies have confirmed the abnormal involvement of CORIN in hypertension and ventricular hypertrophy. A genotypephenotype genetic association study demonstrated that CORIN T555I/Q568P allele is common in blacks and is associated with higher blood pressure and an increased risk for prevalent hypertension (10). A further study verified that the CORIN T555I/Q568P allele can be an independent predictor of left ventricular mass in subjects with elevated systolic blood pressure in self-identified blacks (11). A functional study demonstrated that T555I/Q568P mutant lacking frizzled-like domain 2 had 30 ± 5% activity compared to that of wild type, and the T555I/Q568P variant had a reduced pro-ANP and pro-BNP processing activity compared to that of wild-type. The zymogen activation of the corin variant was impaired significantly (12). Our study also examined the correlation between three SNPs in the 3'UTR of CORIN and LVH in the hypertensive group. However, no association was found though rs3749585 T and rs4695253 T are related to the risk of EH in different inheritance models (Supplementary Tables 3, 4). It is of note that the SNPs rs3749585 T and rs4695253 T are not located in the frizzled-like domain 2.
In addition to the roles of NPs in water and electrolyte balance, NPs also participate in lipid metabolism (42). The extracellular region of corin consists of two frizzled-like domains, eight LDL receptor (LDLR) repeats, a scavenger receptor-like domain, and a trypsin-like protease domain (43). LDLR repeats 1-4 were important structural elements for corin to recognize its physiological substrate, pro-ANP (44). It is tempting to ask, in addition to this role of LDLR repeats, whether LDLR, as part of the corin structure, participates in the lipid regulation. Wang et al. found that serum soluble corin was significantly and positively associated with dyslipidemia. The findings suggested that serum soluble corin may be a marker or risk factor for dyslipidemia (45). In our study, we investigated the relationship between SNPs of 3'UTR in CORIN and the blood levels of lipids. We found that rs4695253 T was significantly associated with the levels of HDL and TC in the recessive model after adjusting for covariates sex and age [HDL: p adj = 0.008, β: 5.053 (95% CI, 1.331-8.775); TC: p adj = 0.036, β: 15.944 (95% CI, 1.050-30.837); Table 5]. MiRNASNP v3 showed that rs4695253 and the two tag SNPs, rs13106975 and rs10938494 in ATP10D on chromosome 4p12, were located in the same linkage disequilibrium (LD) region (Figure 4). Both rs13106975 and rs10938494 were significantly associated with circulating sphingolipid levels (p cor = 4.45 × 10 −18 , p = 5.13 × 10 −4 , respectively) (46,47). Therefore, the LD region, including rs4695253, may be associated with lipid metabolism. Future investigations are needed to better understand how corin is involved in lipid metabolism.
A study suggested that miR-494-3p may be associated with AF and AF-related stroke (48). Another study showed that miR-494-3p targets JunD and miR-494-3p/JunD is a novel molecular axis involved in obesity-related metabolic cardiomyopathy (49). In our study, bioinformatic analysis predicted that target gain with rs3749585 was for hsa-miR-494-3p and allele C could bind the seed regions of miR-494-3p to regulate the expression of CORIN (Figure 2A). Therefore, dual-luciferase reporter assay was performed to verify that miR-494-3p targets rs3749585. The relative luciferase activity of pYr-MirTarget-rs3749585 CC +miR-493-3p was reduced compared to the activity of pYr-MirTarget-rs3749585 TT +miR-493-3p (p < 0.001, Figure 3). This is consistent with the analyses by bioinformatic software. Notably, allele T was significantly associated with the risk of EH and with the protection of AF. Zhang et al. showed that circulating soluble corin negatively correlated with EH (18). However, Chen et al. found that plasma corin levels positively correlated with AF (19). The inconsistent results suggested that the distinct pathway of corin could be involved in EH and AF. In this study, we did not examine the expression of miR-494-3p in EH and AF patients and cannot rule out the possibility that the expression of miR-494-3p was differentially modulated in the two diseases leading to the different effects of rs3749585 in EH and AF. Further study on the expression of hsa-miR-494-3p in patients with EH and AF is of great significance to understand the heterogeneity of rs3749585 in EH and AF. Target loss with rs3749585 was predicted for miR-1323 and miR-548o-3p by bioinformatic analysis. In this study, a decrease in the relative luciferase activity of rs3749585 TT reporter, compared with rs3749585 CC reporter, was observed in the presence of miR-1323 or miR-548o-3p (p = 0.017 and 0.012, respectively, Figure 3).
However, the roles of the two miRNAs in EH and AF were unclear.
Target gain with rs4695253 was predicted for miR-561-3p and miR-548at-5p. The latter is associated with Parkinson's disease (50), but the roles of the two miRNAs in CVDs have not been studied. Target loss with rs4695253 was predicted for six miRNAs: MiR-1264 was upregulated in the circulation 3 h after ischemia (51) and critical limb ischemia in T2DM (52), and downregulated in neuroglioma (53). MiR-550a-3p was downregulated in myelodysplastic syndromes (54). MiR-550a-3p also plays its tumor-suppressor role by directly repressing ERK1 and ERK2 protein expression, thereby suppressing the oncogenic ERK/RSK cascade to reduce breast cancer cell viability, survival, migration, invasion, tumorigenesis, and metastasis (55). MiR-3184-3p may help to stabilize miR-423-5p in the pericardial fluid in patients undergoing cardiac surgery (56). MiR-200c-5p inhibition could reduce the effects of SnHG12 downregulation on cell viability and apoptosis, without affecting SnHG12 expression levels in renal cell carcinoma (57). However, the underlying mechanisms by which the six miRNAs participate in the development of CVDs through binding to rs4695253 remain to be elucidated.
Further investigation is required to understand the biological function of the SNPs of the 3'UTR of CORIN in CVDs. First, the relationship between SNPs in the 3'UTR of CORIN and CVDs in the Chinese Han population is the focus of this study; the results need to be validated in other ethnic cohorts. Second, we only tested the associations between SNPs in the 3'UTR of CORIN and CVDs. The SNPs in all the exons and exonintron boundaries of CORIN were not tested in our studies. Third, hsa-miR-494-3p, hsa-miR-1323, and hsa-miR-548o-3p were validated to target rs3749585 and regulate the expression of CORIN by dual-luciferase reporter assay. However, in view of the different roles of rs3749585 in EH and AF, the expressions of the three miRNAs need to be probed in EH and AF patients. Fourth, the mechanisms whereby the miRNAs participate in the development of EH and lipid metabolism through targeting rs4695253 need to be further studied.
In conclusion, we found significant associations: rs3749585 T and rs4695253 T and EH, rs4695253 T and the levels of TC and HDL, and rs3749585 T and AF. No correlation was found between the three SNPs in the 3'UTR of CORIN and CAD, HF, or LVH in hypertensive patients. Our work provides an important understanding of the roles of SNPs in the 3'UTR of CORIN in susceptibility to EH and AF in the Chinese Han population. Luciferase assay revealed that the relative luciferase activity of rs3749585 CC -transfected cells was significantly decreased by miR-494-3p, in comparison to cells transfected with rs3749585 TT . Further study on the expression of hsa-miR-494-3p in patients with EH and AF is of great significance to understand the heterogeneity of rs3749585 in EH and AF. Hsa-miR-494-3p may serve as a potential therapeutic target for EH and AF patients in the future.

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by Ethics Committee of The First Affiliated Hospital of Dalian Medical University. The patients/participants provided their written informed consent to participate in this study.

AUTHOR CONTRIBUTIONS
YL and FC contributed to the conception of the study and manuscript writing. YiZ contributed to the management of blood samples, DNA extraction, and dual-luciferase reporter assay. XY contributed to SNP genotyping. YaZ contributed to the clinical data collection. YutZ and SZ contributed to data entry and organization. SL and YuaZ contributed to data processing and statistical analysis. WZ and JL contributed to bioinformatics analysis. YY and YX proposed critical revision to the manuscript. All authors contributed to the article and approved the submitted version.