Serum microRNA-204 levels are associated with long-term cardiovascular disease risk based on the Framingham risk score in patients with type 2 diabetes: results from an observational study

Abstract

Background

Previous studies have demonstrated that microRNA-204 (miR-204) is involved in atherosclerosis and vascular calcification. However, the value of miR-204 as the predictive biomarker for cardiovascular disease (CVD) remains unclear. We aimed to evaluate the association between the circulating miR-204 level and ten-year CVD risk based on the Framingham risk score (FRS).

Methods

In this retrospective study, we enrolled 194 consecutive patients with type 2 diabetes mellitus (T2DM) without CVD in Beijing Anzhen Hospital between January 2015 and September 2016. We used the FRS to evaluate the risk of CVD for each patient. Circulating miR-204 levels were measured by quantitative real-time polymerase chain reaction.

Results

Circulating miR-204 levels were significantly lower in the group of patients (0.49 ± 0.13) at high risk of CVD (FRS > 20%) than in the low (FRS < 10%) and intermediate (FRS: 10%–20%) risk groups (0.87 ± 0.19 and 0.75 ± 0.25, respectively; P < 0.001). FRS was negatively correlated with miR-204 levels (r = –0.421, P < 0.001). According to multivariate logistic analyses, reduced miR-204 level was independently associated with an increased risk of CVD after adjusting for conventional risk factors (OR = 0.876, 95% CI: 0.807–0.950, P = 0.001). Receiver-operating characteristic curve analysis showed that the circulating miR-204 level can predict the high risk of CVD with higher specificity than the traditional risk factor of high systolic blood pressure or the protective factor of high-density lipoprotein cholesterol.

Conclusions

Our study demonstrated that patients with lower circulating miR-204 levels were at high risk for CVD. After adjustment for potential confounders, miR-204 was independently associated with CVD in patients with T2DM.

1. Introduction

MicroRNAs play important roles in physiological and pathological processes, including cell cycle regulation,[1] cell differentiation,[1] inflammatory responses,[2] apoptosis,[3] and extracellular matrix degradation.[4] Elevated serum microRNA-204 (miR-204) levels are associated with insulinoma,[5] acute lymphocytic leukemia,[6] human breast cancer,[7] and other tumors.[8] In addition, accumulating evidence suggests that the serum miR-204 level is associated with the increased cardiovascular event risk. Previously, one study had demonstrated the miR-204 attenuated aortic vascular smooth muscle cell (VSMC) phenotypic switch in rat models of type 2 diabetes mellitus (T2DM).[9]

Diabetes mellitus is a heterogeneous collection of metabolic abnormalities that are characterized by impaired insulin secretion, defective insulin action, or both, with hyperglycemia. Diabetic chronic hyperglycemia is associated with specific microvascular complications affecting the eyes, kidneys, and nerves in the long term, and with the particularly high risk for cardiovascular disease (CVD).[10]

CVD, the leading cause of mortality and morbidity worldwide, comprises atherosclerotic CVD and other diseases, such as coronary heart disease, stroke, peripheral vascular disease, and heart failure. The early prediction of risk plays a critical role in the evaluation and treatment of CVD.

The Framingham risk score (FRS) is widely recognized as a valuable preliminary screening tool for CVD. The Framingham Heart Study has been at the frontier of CVD epidemiology since its initiation in 1948.[11] The algorithm accounts for sex-related differences, and its parameters include age, total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), cigarette smoking status, and systolic blood pressure (SBP).[12] The FRS is the most convenient assessment for the long-term prediction of CVD. Since T2DM is a risk factor of cardiovascular events, the FRS is indispensable for predicting the risk of CVD in patients with T2DM. Our research focused on the association between the serum miR-204 level and FRS, with the aim of detecting populations at moderate and high risk of CVD among individuals with T2DM.

2. Methods

2.1. Study population

In this retrospective study, we enrolled 194 consecutive T2DM patients (average age: 60.6 ± 10.0 years, 61.9% male) without known CVD in Beijing Anzhen Hospital between January 2015 and September 2016. All patients underwent coronary angiography and did not have significant coronary stenosis. Patients with T2DM were identified by the fasting blood glucose (FBG) level of 126 mg/dL (7.0 mmol/L) or more and glycated hemoglobin A1c (HbA1c) levels > 6.0%, according to the World Health Organization guidelines, or the indication for insulin or anti-diabetic medications. Patients meeting any of the following criteria were excluded: (1) the history of heart failure or cardiomyopathy; (2) prior evidence of coronary artery disease (e.g., myocardial infarction, angina pectoris, or abnormal resting electrocardiogram); (3) kidney dysfunction; (4) peripheral artery disease; (5) hepatitis B infection, hepatitis C infection, or three times the normal range of liver transaminase levels; (6) hemolytic disease; (7) cancer; (8) thyroid disease; and (9) acute infection or inflammation. The study was approved by the Ethics Committee of Beijing Anzhen Hospital, Capital Medical University, Beijing, China (No.2016044X), and all patients provided informed written consent.

2.2. Plasma RNA extraction

The peripheral blood specimens were collected using EDTA-anticoagulative tubes and centrifuged at 4,000 rpm for 10 min at 4 °C, then centrifuged at 12,000 rpm for 15 min to completely remove cell debris. The serum was stored at –80 °C until analysis. We used TRIzol™ reagent (Invitrogen, Carlsbad, CA, USA) to extract total RNA from the prepared plasma samples according to the manufacturer's protocols. In brief, 250 µL plasma was mixed with TRIzol™ reagent and chloroform.[13] After centrifugation of the sample at 12,000 rpm at 4 °C for 10 min, the aqueous phase was transferred to a new centrifuge tube, and 0.8 × volumes of absolute isopropanol were added. After the sample was mixed and incubated at –20 °C for 15 min, it was centrifuged again at 12,000 rpm at 4 °C for 10 min. After removal of the supernatant, the precipitate was washed in 1.5 mL 75% ethanol. Then, the sample was centrifuged at 12,000 rpm at 4 °C for 5 min. The RNA precipitate was dissolved in 15µL RNase-free water. The concentration and purity of isolated RNA were detected with the NanoDrop™ 2000 system (Thermo Fisher Scientific, Waltham, MA, USA).

2.3. Quantitative real-time polymerase chain reaction

We used the GoScript™ Reverse Transcription System (Promega, Madison, WI, USA) to perform reverse transcription at 16 °C for 30 min and 42 °C for 60 min respectively, and then followed by incubation at 80 °C for 5 min to denature the enzyme. The expression levels of miR-204 were quantified with the CFX Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA). Each reaction consisted of 2.5 µL cDNA, 2.0 µL primer (7.5 µmol/L), 12.5 µL 2 × TaqMan^® Universal PCR Master Mix, and 8 µL sterile water. The PCR cycling comprised pre-denaturation for 10 min at 95 °C, followed by 40 cycles of 15 s at 95 °C and 60 s at 60 °C, followed by 0.5 °C incremental increases every 20 s from 75 °C to 95 °C.[14] As the endogenous U6 gene is widely used for the normalization of miRNA expression in tissues and cells, we calculated miR-204 expression relative to U6 expression by 2^−ΔΔCt method. The sequences of the quantitative real-time polymerase chain reaction primers were as follows: hsa-miR-204-5pRT forward: 5′-GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACAGGCATA-3′; hsa-miR-204-5pF forward: 5′-GCAGTTCCCTTTGTCATCCT-3′; Hu-U6-2QF forward: 5′-CTCGCTTCGGCAGCACA-3′; and Hu-U6-2QR forward: 5′-AACGCTTCACGAATTTGCGT-3′. The measurements were performed in triplicate, and the average values were reported as the detection levels. All protocols were carried out in accordance with the relevant guidelines and protocols.

2.4. Clinical and laboratory assessments

Each participant provided the medical history and underwent physical examination by a trained physician. Individuals who smoked regularly in the year before the examination were categorized as smokers. We measured the patient's height and weight, and calculated body mass index (BMI) by dividing the body weight in kilograms by the square of height in meters (kg/m²). Blood pressure was measured in the left arm by a physician using the mercury-column sphygmomanometer, the cuff of the appropriate size, and the standardized protocol. After the patient had been sitting for at least 5 min, the average of two blood pressure measurements was used for analysis. Biochemical parameters including serum TC, HDL-C, FBG, and HbA1c were measured with the enzymatic colorimetric method.

2.5. Assessment of cardiovascular risk

We used the FRS to study the risk of CVD. The FRS was calculated based on six risk factors of CVD: age, sex, smoking, TC, HDL-C, and SBP. The cutoff values for calculating the FRS were defined as: (1) TC < 160 mg/dL, 160–199 mg/dL, 200–239 mg/dL, 240–279 mg/dL, and ≥ 280 mg/dL; (2) SBP: < 120 mmHg, 120–129 mmHg, 130–139 mmHg, 140–159 mmHg, and ≥ 160 mmHg; and (3) HDL-C: < 40 mg/dL, 40–49 mg/dL, 50–59 mg/dL, and ≥ 60 mg/dL.[12] We applied the NCEP-ATPIII guidelines to the FRS,[12] calculated the corresponding scores for various parameters according to the algorithms for the different sexes, and added the values to obtain the total score, which we checked against the table to obtain the corresponding ten-year heart disease risk score. The absolute percentage of CVD risk over ten years was categorized as low (< 10%), moderate (10%–20%), or high (> 20%).[12]

2.6. Statistical analysis

For all statistical analysis, we used the SPSS 25.0 statistical software (IBM SPSS Inc., Chicago, IL, USA) and expressed as frequencies and percentages for categorical variables or the mean ± SD for continuous variables. We used the Student's t-test and analysis of variance to compare continuous variables with the normal distribution. Categorical variables were analyzed using the χ² test. The Spearman's ρ rank-order correlation test was used to examine the correlation between serum miR-204 level and other cardiovascular risk factors. We used the multivariate logistic regression analysis to examine the correlation between miR-204 levels and ten-year risk for CVD (based on low-intermediate risk and high risk, as determined by FRS) as the independent and dependent variables, respectively. For the different FRS-classified groups, we calculated the adjusted odds ratios and 95% confidence intervals (CIs). A P-value < 0.05 was considered to be statistically significant.

Additionally, the receiver operating characteristic (ROC) curve was plotted for the circulating miR-204 level, SBP, and HDL-C of patients with the high ten-year CVD risk, to evaluate the ability of each variable to classify the incidence of CVD. The area under the curve (AUC) and 95% CI of each ROC curve were calculated. A P-value < 0.05 (two-tailed) was considered to be statistically significant.

3. Results

3.1. Baseline characteristics

Table 1 shows the baseline characteristics of our patients. Among the potential participants, 194 patients met the inclusion criteria and were enrolled in the study (mean age: 60.6 ± 10.0 years, 61.9% male). According to FRS classification, 52.06%, 36.08%, and 11.86% of patients with T2DM were at low, intermediate, and high risk of CVD. Compared with those at low or moderate risk, patients with severe CVD risk were more likely to be older and male, and have the history of smoking. On average, participants with high risk score had higher SBP, FBG, and HbA1c levels; and lower levels of miR-204.

Table 1. Clinical characteristics of the study populations.

Characteristics	Low risk (n = 101)	Intermediate risk (n = 70)	High risk (n = 23)	P-value
Age, yrs	56.4 ± 9.4	62.3 ± 7.9	73.5 ± 4.7	< 0.001
Male	37 (36.6%)	63 (90.0%)	13 (56.5%)	< 0.001
Body mass index, kg/m2	25.7 ± 2.8	26.2 ± 3.0	25.5 ± 2.13	0.338
Smoking	41 (40.6%)	41 (58.6%)	8 (34.8%)	0.034
Systolic blood pressure, mmHg	125.0 ± 15.0	131.8 ± 14.0	141.3 ± 15.9	< 0.001
Fasting blood glucose, mg/dL	131.5 ± 35.5	142.9 ± 47.6	154.3 ± 54.9	0.039
HbA1c, %	7.1 ± 1.0	7.6 ± 1.7	7.9 ± 1.7	0.009
Total cholesterol, mg/dL	76.7 ± 23.6	69.9 ± 18.9	76.5 ± 16.0	0.102
High-density lipoprotein cholesterol, mg/dL	21.2 ± 11.4	19.3 ± 5.1	21.4 ± 3.9	0.340
Low-density lipoprotein cholesterol, mg/dL	43.2 ± 17.5	40.5 ± 15.4	45.4 ± 14.2	0.371
miR-204	0.9 ± 0.2	0.8 ± 0.3	0.5 ± 0.1	< 0.001

Data are presented as means ± SD or n (%). HbA1c: glycated hemoglobin A1c.

3.2. Relationships between CVD risk factors and FRS

Table 2 summarizes the correlations between CVD risk factors and the FRS ten-year CVD risk as assessed by Spearman's ρ rank-order correlation analysis. Age, gender, SBP, and miR-204 were significantly correlated with FRS (P < 0.001), and HDL-C was associated with FRS (P = 0.020). However, no correlation was found between FRS and BMI, smoking, FBG, HbA1c, serum TC, or low-density lipoprotein cholesterol in our study population. There were significant positive correlations between FRS ten-year CVD risk and age, sex, and SBP; however, a significant negative correlation was found between miR-204 and FRS-defined risk in patients with T2DM (Figure 1). Figure 2 shows that the circulating miR-204 levels were negatively correlated with the FRS ten-year CVD risk (r = –0.421, P < 0.001).

Table 2. Correlations of FRS ten-year CVD risk with demographic and biochemical parameters.

Variables	R	P-value
Age	0.594	< 0.001
Sex	0.406	< 0.001
Body mass index	0.049	0.499
Smoking	0.096	0.184
Systolic blood pressure	0.432	< 0.001
Fasting blood glucose	0.066	0.358
HbA1c	0.122	0.090
Total cholesterol	–0.003	0.967
High-density lipoprotein cholesterol	0.167	0.020
Low-density lipoprotein cholesterol	0.028	0.700
miR-204	–0.421	< 0.001

CVD: cardiovascular disease; FRS: Framingham risk score; HbA1c: glycated hemoglobin A1c.

Figure 1. Association between FRS risk categories and miR-204 levels.

FRS: Framingham risk score.

Figure 2. Correlations of FRS ten-year CVD risk with miR-204 levels.

CVD: cardiovascular disease; FRS: Framingham risk score.

3.3. Prognostic values of CVD risk factors

ROC curves were generated to obtain the prognostic values and optimal cutoff values for miR-204 and the other CVD risk factors (Figure 3). The AUC of miR-204 was 0.884 (95% CI: 0.829–0.939; P < 0.001), demonstrating that the circulating miR-204 level can predict the high ten-year CVD risk by using the FRS. The specificity of miR-204 was higher than those of the conventional risk factors SBP and HDL-C (AUC = 0.746, 95% CI: 0.647–0.846, P < 0.001 and AUC = 0.629, 95% CI: 0.530–0.727, P < 0.05, respectively).

Figure 3. Prognostic values of miR-204 and other CVD risk factors.

AUC: area under the curve; CVD: cardiovascular disease; HDL-C: high-density lipoprotein cholesterol; SBP: systolic blood pressure.

3.4. Associations between risk factors and FRS risk categories

We performed multiple logistic regressions to evaluate the associations between risk factors of T2DM and FRS risk categories (Table 3). Our results revealed that age and circulating miR-204 levels were robust predictors of the high-risk FRS category versus the low-to-intermediate-risk category (P = 0.001).

Table 3. Multiple logistic regression of the association of FRS risk categories with variables.

Variables	OR	95% CI	P-value
Age	2.316	1.412–3.799	0.001
Sex	1.267	0.938–1.712	0.123
Smoking	1.563	1.046–2.337	0.029
Systolic blood pressure	1.214	1.063–1.385	0.004
High-density lipoprotein cholesterol	0.850	0.679–1.064	0.156
Total cholesterol	1.052	0.979–1.129	0.166
miR-204	0.876	0.807–0.950	0.001

FRS: Framingham risk score.

4. Discussion

In the present study, we first evaluated the association between FRS and circulating miR-204 levels compared with other established risk factors. The FRS was significantly correlated between age and miR-204 levels. Moreover, older and low circulating miR-204 levels were associated with the greater risk of CVD in patients with T2DM. Our findings demonstrated that circulating miR-204 level can be used as a novel parameter to predict the incidence of CVD. Conversely, the high miR-204 level may be a protective factor against CVD.

4.1. T2DM is associated with the high risk of cardiovascular complications

It is widely accepted that individuals with T2DM have the significantly higher risk of macrovascular diseases than healthy individuals. Hyperglycemia has similar effects on endothelial cells and VSMC as hyperlipidemia.[15]^,[16] Previous studies had demonstrated that hyperglycemia was implicated in some vascular abnormalities, such as the increases in free fatty acids and insulin resistance that provoke decreased nitric oxide release, increased oxidative stress and subsequent inflammatory responses, disturbances of intracellular signal transduction, and the activation of receptors for advanced glycation end products that resulted in macrovascular dysfunction.[17]^–[22] Overall, T2DM is regarded as the major cause of cardiovascular events.

4.2. miR-204 has links to multiple diseases

MicroRNAs are small, noncoding RNAs that act as post-transcriptional regulators of gene expression and potent modulators of a variety of biological processes and pathologies.[23]^,[24] miR-204 is derived from the sixth intron of the transient receptor potential melastatin 3 gene. Its pathological functions have been observed in several diseases, such as pulmonary arterial hypertension, T2DM, and various types of cancers.

Many studies had focused on the roles of miR-204 in cancer. In the study by Lin, et al.,[25] miR-204-5p overexpression induced prostate cancer cell apoptosis by repressing BCL2 expression. Zhang, et al.[26] demonstrated that miR-204 as the tumor suppressor by directly targeting ATF2 in human non-small cell lung cancer. On the other hand, Turner, et al.[27] showed that miR-204 is more abundant in five tumor specimens than in healthy tissues. Similarly, Zanette, et al.[6] showed that miR-204 is one of the five most highly expressed miRNAs in acute lymphocytic leukemias. Therefore, miR-204 seems to be a double-edged sword in the development and progression of tumors. Furthermore, van Rensburg, et al.[28] reported that B-cell associated miRNAs, including miR-204-5p, are more highly expressed during the active tuberculosis.

A few studies had focused on the role of miR-204 in T2DM. Xu, et al.[29] observed that the novel TXNIP/miR-204/MAFA/insulin pathway that may promote T2DM progression. In their recent study, they identified serum miR-204 as an attractive novel biomarker of type 1 diabetes-associated beta-cell loss in humans.[30] In addition, Han, et al.[31] suggested that miR-204-3p may play a protective role in high glucose-induced apoptosis and dysfunction of podocytes through the down-regulation of Bdkrb2. With regard to diabetic complications, Mao, et al.[32] discovered that miR-204-5p promotes the development of diabetic retinopathy by down-regulating microtubule-associated protein 1 light chain 3. Fan, et al.[33] revealed that the up-regulation of circKMT2E may be implicated in the pathogenesis of diabetic cataracts due to sponging miR-204-5p. Similarly, Gao, et al.[34] described the role of miR-204 in the regulation of SIRT1 during the healing of diabetic corneal epithelial wounds.

4.3. miR-204 may play a critical role in CVD

Cardiovascular complications are the major characteristic of chronic inflammatory disorders, such as chronic kidney disease, T2DM, and atherosclerosis, which are associated with clinically serious morbidity and mortality.[35]^–[37] In recent years, emerging studies had shown that miR-204 may play an essential role in the development and progression of cardiovascular events.

In a recent study, Yu, et al.[38] showed that silencing lncRNA AK139328 was significantly increased miR-204-3p expression and inhibited cardiomyocyte autophagy, thereby reducing myocardial ischemia/reperfusion injury in diabetic mice. Similarly, Xue, et al.[39] discovered that up-regulation of miR-204 could alleviate ventricular remodeling and improve cardiac function in mice after myocardial ischemia/reperfusion injury via regulation of SOCS2. Torella, et al.[9] reported that miR-204 down-regulation in T2DM inhibits VSMC proliferation by targeting CAV1 in vitro and in vivo. Yu, et al.[40] elucidated the interaction between TUG1 and miR-204-5p in calcific aortic valve disease (CAVD), and revealed that TUG1 actively regulates post-transcriptional expression of Runx2 by sponging miR-204-5p in CAVD. Similarly, Wang, et al.[41] demonstrated that PARP1 promotes osteogenesis and calcification of VSMCs by reducing the translational inhibition of miR-204, thereby increasing Runx2 protein levels. Xiao, et al.[42] reported that overexpression of miR-204 efficiently reversed the MALAT1-induced up-regulation of Smad4, which prevented osteogenic differentiation of human aortic valve interstitial cells. Moreover, Vikram, et al.[43] showed that the microbiome remotely regulates the expression of vascular miR-204 and impairs endothelial function by targeting Sirt1. Meloche, et al.[44] demonstrated in patients with pulmonary arterial hypertension that coronary artery remodeling is due, in part, to miR-223/DNA damage/Poly [ADP-ribose] polymerase 1/miR-204 axis activation and subsequent BRD4 overexpression.

In conclusion, we primarily evaluated serum miR-204 levels in the context of cardiovascular epidemiology. It remains challenging to diagnose and treat cardiovascular complications in patients with T2DM in a timely manner. Our results suggest that circulating miR-204 detection may be a novel approach to predict CVD risk.

4.4. Limitations

One limitation of this study is the small sample size, which may restrict the generalizability of the results. Due to the relatively small sample size of this study, it has not been powered to demonstrate the cause and effect between miR-204 and CVD risk. Our risk equation was derived from individuals aged 20 to 69 years, so our findings may not be generalizable to older or younger persons. Serum miR-204 detection tends to be more expensive at present, which leds to the small sample size and limits clinical use. Moreover, this study is a cross-sectional study with endpoints that are not the precise incidence of CVD or major adverse cardiovascular events. Further studies are needed to validate our findings with the help of the basic research models and the larger, prospective, randomized clinical cohorts.

4.5. Conclusions

We demonstrated the association between miR-204 and CVD risk, which may help physicians estimate the long-term risk of cardiovascular events in patients with T2DM. In our study, the population of patients with lower circulating miR-204 levels was at high risk of developing CVD. Our findings suggest a method that may help physicians evaluate the long-term risk of cardiovascular complications in patients with T2DM earlier than is possible with the current methods. In addition, it may facilitate earlier therapeutic intervention and long-term follow-up for patients in the intermediate and high risk categories, which could reduce the incidence of cardiovascular events in patients with T2DM. Further advances in the early diagnosis, prevention, and treatment of CVD are needed. The miR-204-mediated regulation of the calcification pathway may be a potential therapeutic target. However, the exact link between miR-204 and CVD remains unclear. Additional studies are needed to confirm the relationship between miR-204 and CVD, and determine the mechanistic link. The study of miRNAs and its functions in CVD will help us better understand the pathological mechanism and develop miRNA drugs that target specific genes for the treatment of CVD.