Low Levels of MicroRNA‐21 Are a Marker of Reduced Arterial Stiffness in Well‐Controlled Hypertension

Abstract

MicroRNAs (miRNAs) play a crucial role in myocardial and vascular remodeling and have emerged as potential diagnostic and prognostic biomarkers or as therapeutic targets. The authors aimed to investigate the expression profile of selected miRNAs in the peripheral blood of patients with well‐controlled essential hypertension in relation to arterial stiffness. Expression levels of miRNAs miRNA‐1, miRNA‐133a, miRNA‐26b, miRNA‐208b, miRNA‐499, and miRNA‐21 in peripheral blood mononuclear cells were quantified by real‐time reverse transcription polymerase chain reaction. Carotid‐femoral pulse wave velocity (cfPWV) and carotid radial pulse wave velocity (crPWV) were evaluated at baseline and after 1 year of effective antihypertensive therapy. A total of 95 patients (50 men, mean age 62±9 years) with well‐controlled essential hypertension were included in the analysis. Only miRNA‐21 was independently correlated with changes in both cfPWV and crPWV, independently of blood pressure levels (r=−0.56 and r=−0.46, respectively; P<.001 for both). Low levels of miRNA‐21 are strongly associated with an improvement in arterial stiffness in patients with well‐controlled essential hypertension, independently of their blood pressure levels. These data highlight the significance of miRNA‐21 in vascular remodeling and its role as a potential prognostic marker and future therapeutic target.

Arterial stiffness is a hallmark of vascular dysfunction and has been proposed as an independent risk factor for fatal and nonfatal cardiovascular events in patients with hypertension.1 Blood pressure (BP) has been shown to be independently associated with pulse wave velocity (PWV), which is the most widely used examination for the reliable measurement of large artery stiffness.2 Numerous investigations have focused on the mechanisms underlying the vascular remodeling that leads to increased arterial stiffness. There is an emerging need to discover novel predictive markers and therapeutic approaches to limit its progression. Previous studies have shown that a simple reduction in brachial BP is not sufficient to ensure a corresponding improvement in aortic stiffness or arterial biomechanics; thus, there is inadequate protection of target organs with consequences for the patient's clinical outcome.3, 4 Muscular small arteries contribute minimally to the mechanical behavior of the large elastic arteries, whose stiffness is determined by the composition of the arterial wall, and mainly by elastin and collagen.5

In recent years, the role of microRNAs (miRNAs) in cardiac diseases has been increasingly recognized and they are considered to be of potential value, either as diagnostic and prognostic biomarkers or as a therapeutic target. miRNAs are short‐regulating RNA molecules that interfere with gene expression at the posttranscriptional level in a way that reduces or prevents protein synthesis.6 It has been suggested that an imbalance in the normal miRNA profile can be identified long before the onset of a disease.7 In addition, they seem to play a crucial role in the regulation of many of the steps leading to the development of cardiovascular diseases8 such as myocardial infarction,9 myocardial hypertrophy, and fibrosis.10

In the present study, we sought to identify miRNAs in hypertensive patients that could be of prognostic value and might be associated with changes in arterial stiffness after 1 year of effective treatment. We measured baseline miRNA‐1, miRNA‐133a, miRNA‐21, miRNA‐208b, miRNA‐499, and miRNA‐26b gene expression levels in peripheral blood mononuclear cells (PBMCs), cells that are known to play a significant part in the pathophysiology of target organ damage.11 These miRNAs were chosen because they have been implicated in vascular and heart remodeling and have a distinct expression profile in hypertension.12 For the measurement of arterial stiffness, we used PWV, which is widely accepted as the “gold standard” measure for this purpose.1

Methods

Study Population

The study was an observational and follow‐up study to identify miRNAs in hypertensive patients that could have prognostic value in improvement of arterial stiffness after 1 year of effective treatment. We included 122 consecutive patients with untreated essential hypertension (age 63.9±11.7 years) and no indications of other organic heart disease. The study population was recruited from a cardiology outpatient department. The diagnosis of hypertension was based on three outpatient measurements of BP >140/90 mm Hg at intervals of no longer than 2 weeks, in accordance with the recommendations of the European Society of Hypertension/European Society of Cardiology (ESH/ESC).13 Participants with BP >140/90 mm Hg on the final visit underwent 24‐hour ambulatory BP monitoring. To be eligible for inclusion in the study, a mean 24‐hour BP >130/80 mm Hg was required.

All patients underwent a complete physical examination and routine laboratory tests. The patients had not previously taken any hypertensive medication and did not take any other drugs for 3 weeks before the study. Patients with any of the following characteristics were excluded: pregnant or lactating women; history of or medication for hypertension; grade 3 hypertension or secondary hypertension; tachyarrhythmias (such as atrial fibrillation) or bradyarrhythmia; coronary artery disease; cerebrovascular, liver, or renal disease; albumin excretion >300 mg/24 h; history of drug or alcohol abuse; any chronic inflammatory or other infectious disease during the past 6 months; thyroid gland disease; body mass index (BMI) >30 kg/m²; or a history of any hematological disease. Vascular or neoplastic conditions were ruled out in all participants by a careful examination of the history and routine laboratory tests. BMI was calculated as weight/height² (kg/m²). A full echocardiographic examination was performed in all participants. Following diagnosis of arterial hypertension, patients received antihypertensive treatment according to the physician's judgment and the recommendations of the ESH/ESC13 so that they reach target goals. The patients were followed up for 12 months via regular visits, in the first and second months and then every 2 months, or whenever they considered that BP was not well controlled or they experienced some clinical symptom. Patients who had achieved the target BP of ≤140/90 mm Hg by their second visit, as confirmed by home BP measurements according to the ESH/ESC guidelines, and maintained it throughout the study period, were included in the final analysis.13

Blood samples were taken at baseline for miRNAs gene expression levels assessment in peripheral mononuclear cells. More specifically, at the first visit in all patients, after a rest of 20 minutes, blood was drawn from a superficial brachial vein via a 21‐gauge needle with care to avoid stasis, hemolysis, and contamination by tissue fluids or exposure to glass.

The study was carried out in accordance with the Declaration of Helsinki and the protocol was approved by the local ethics committee. All patients gave informed consent to their inclusion in the study.

Measurement of PWV

PWV was measured at baseline and at 1‐year follow‐up. On the day of examination, all patients were asked to refrain from caffeine, alcohol, and smoking during the preceding 12 hours. The study was carried out between 8 am and 9 am in a quiet room at 22±1°C. Height and weight were measured. Patients were allowed a further 15‐minute supine rest before baseline measurements. Brachial BP was measured over the brachial artery three times at 5‐minute intervals. The mean of the last two measurements was recorded as representative of brachial BP. After brachial BP, the carotid, femoral, and radial arteries were palpated to find the location of the points with the most pronounced pulse pressure waves. Carotid‐femoral PWV (cfPWV) and carotid‐radial PWV (crPWV) artery waveforms were measured, and PWV (Complior SP, Alam Medical, Vincennes, France) was determined. The distances traveled by the pulse waves were assessed in triplicate over the surface of the body with a nonelastic tape measure. The same examiner, who was blinded to the patient's history, performed all measurements. The changes in cfPWV and crPWV (ΔcfPWV and ΔcrPWV) were estimated as the difference between the PWV values at baseline and at 1 year.

RNA Isolation and miRNA Quantification

PBMCs were isolated by Ficoll‐Paque Plus gradient centrifugation (Stem Cell Technologies Inc., Vancouver, BC, Canada). Total RNA was isolated using the TRI‐Reagent (Ambion, Life Technologies, Carlsbad, CA) and reverse‐transcribed using the Mir‐X miRNA First‐Strand Synthesis kit (Clontech, Takara Bio Inc., Kisatsu, Shiga, Japan). Measurements of microRNA levels were performed by quantitative real‐time polymerase chain reaction (qPCR) using the Corbett Research 6000 detection system. qPCR assays were performed using the KAPA SYBR FAST qPCR Kit (Kapa Biosystems, Woburn, MA). Primers used were 5′‐TAG CTT ATC AGA CTG ATG TTG A‐3′ for miRNA‐21, 5′‐TTT GGT CCC CTT CAA CCA GCT G ‐3′ for miRNA‐133a, 5′‐TGG AAT GTA AAG AAG TAT GTA T‐3′ for miRNA‐1, TTC AAG TAA TTC AGG ATA GGT‐3′ for miRNA‐26b‐5p, 5′‐ATA AGA CGA ACA AAA GGT TTG T‐3′ for miRNA‐208b3p, 5′‐TTA AGA CTT GCA GTG ATG TTT‐3′ for miRNA‐499a‐5p, and 5′‐TAG CTT ATC AGA CTG ATG TTG A‐3′ for miRNA‐21. U6 expression was used as a normalization standard. All samples were run in duplicate. The standard curve method was used for absolute quantification of the amplification products and specificity was determined by performing a melting curve analysis.

Statistical Analysis

Summary descriptive statistics are presented as mean±(standard deviation) or frequency (percentage), as appropriate. Pearson correlation coefficients were computed between miRNA levels and the changes in cfPWV and crPWV. Stepwise linear regression analysis was used to assess which of the significant variables on univariate analysis were independently associated with cfPWV and crPWV. All statistical analyses were carried out with the statistical software SPSS version 21 (IBM, Armonk, NY) at the two‐sided 5% level of significance.

Results

Initially, we recruited 151 patients. Ten of them were excluded because of white‐coat hypertension. Two patients developed atrial fibrillation, six had BMI >30 kg/m², one had secondary hypertension, four had thyroid disease, three had renal failure, and four had rheumatoid arthritis. Nineteen patients were excluded because they had not achieved their BP target by the second visit and a further eight were excluded for the same reason during the course of the study. Thus, a total of 95 patients (50 men, mean age 62±9 years) were included in the final analysis. Their main demographic, echocardiographic, and clinical characteristics are reported in Table 1. The mean number of medications used was 1.9. Patients' BP decreased significantly in response to antihypertensive treatment (systolic BP before treatment 156±8 mm Hg and after treatment 137±6 mm Hg [P<.05]; diastolic BP before treatment 95±9 mm Hg and after treatment 83±7 mm Hg [P<.05]) (Table 2). In contrast to the BP decrease, neither cfPWV (10.5±2.3 m/s at baseline and 9.7±2.3 m/s after 1 year, P=.09) nor crPWV (9.7±2.04 m/s at baseline and 9.3±2.03 m/s, P=.3) changed significantly during the 1‐year study period. MiRNA expression levels were as follows: miRNA‐208b=23.5±28.4, miRNA‐133a=6.5±5.3, miRNA‐1=26.8±32, miRNA‐21=3.2±2.1, miRNA‐499=10.2±10.8, and miRNA‐26b=55.62±44.74.

Table 1.

Participants' Demographic, Clinical, and Laboratory Characteristics

Participants (N=95)
Age, y	62±9
Male/female, No.	48/47
Smokers, No.	54
Diabetics, No.	40
Ejection fraction, %	63±9
Left ventricular mass index, g/m2	105.9±25.2
Heart rate, beats per min	72±8
Body mass index, kg/m2	27.1±4
Serum creatinine, mg/dL	0.9±0.09
GFR, mL/min/1.73 m2	75.1±19
Uric acid, mg/dL	6.4±1.2
Fasting glucose, mg/dL	111±28
Total cholesterol, mg/dL	229±44
Triglycerides, mg/dL	222.5±77
HDL cholesterol, mg/dL	37.5±12
LDL cholesterol, mg/dL	115.5±49
Medication, %
Angiotensin‐converting enzyme inhibitors	31
Angiotensin receptor blockers	63
Calcium antagonist	33
Diuretics	40
β‐Blockers	28

Abbreviations: GFR, glomerular filtration rate; HDL, high‐density lipoprotein; LDL, low‐density lipoprotein. Data are expressed as mean±standard deviation unless otherwise indicated.

Table 2.

Changes in Office Blood Pressure Before and After Treatment

	Before Treatment	After Treatment	P Value
Systolic blood pressure, mm Hg	156±8	137±6	<.001
Diastolic blood pressure, mm Hg	95±9	83±7	<.001
Pulse pressure, mm Hg	62±10	56±8	<.001
Heart rate, beats per min	73±8	70±7	.025

Stepwise regression showed that miRNA‐21 was the only parameter that was independently correlated with changes in cfPWV and crPWV (Figure). More specifically, a strong negative association was found between the gene expression levels of miRNA‐21 in our patients' PBMCs and ΔcfPWV (r=−0.56, P<.001). A similar negative association was found for ΔcrPWV (r=−0.46, P<.001). Low levels of miRNA‐21 were strongly associated with the improvement in arterial stiffness in our patients with well‐controlled essential hypertension. Notably, BP levels before the start of treatment did not appear to have any effect on either of these significant relations, since the above correlations persisted even after controlling for BP levels. There were also no significant changes in the magnitude of the correlations between miRNA‐21 levels and ΔcfPWV or ΔcrPWV reported above, after controlling for the presence of diabetes.

Figure 1.

Correlation of baseline microRNA‐21 (miRNA‐21) gene expression levels with changes of (A) carotid‐femoral pulse wave velocity (ΔcfPWV) and (B) carotid‐radial pulse wave velocity (ΔcrPWV).

No significant correlation was found between gene expression levels of the other miRNAs we evaluated and ΔcfPWV (r=−0.04 [P=.67] for miRNA‐133a; r=−0.14 [P=.17] for miRNA‐1; and r=−0.09 [P=.37] for miRNA‐208b). A weak correlation was observed between ΔcfPWV and levels of miRNA‐499 (r=−0.28, P=.006) and of miRNA‐26b (r=–0.22, P=.03). Nor was any significant correlation found between the gene expression levels of the other miRNAs and ΔcrPWV (r=0.00 [P=.99] for miRNA‐133a; r=−0.08 [P=.4] for miRNA‐1; r=−0.15 [P=.12] for miRNA‐26b; r=−0.18 [P=.66] for miRNA‐499; and r=−0.14 [P=.17] for miRNA‐208b).

Examination of the miRNA expression levels relative to baseline values of PWV, showed positive correlations of miRNA‐21 levels with cfPWV before (r=0.23, P=.023) and crPWV before (r=0.28, P=.006) as well as with cfPWV after 1 year (r=0.50, P<.001) and crPWV after 1 year (r=0.46, P<.001). Weak correlations between miRNA‐1 levels and cfPWV after 1 year (r=0.25, P=.015) as well as crPWV after 1 year (r=0.21, P=.041) were also observed.

Examination of the miRNA expression levels relative to other parameters, such as BP, heart rate, or creatinine levels did not reveal any significant associations. We only observed a weak negative correlation between miRNA‐208b levels and serum creatinine levels (r=–0.21, P=.039) and a weak positive correlation between miRNA‐26b levels and total cholesterol levels (r=0.21, P=.043).

Discussion

This is the first study to examine the prognostic value and the association of different miRNAs in PBMCs with alterations in PWV in patients with well‐controlled essential hypertension before and after treatment. We found that miRNA‐21 showed a strong negative correlation with both ΔcfPWV and ΔcrPWV in patients after 1 year of effective antihypertensive therapy, independently of their BP levels.

Arterial remodeling is accompanied by changes in the arterial wall, including fragmentation and degeneration of elastin, increased collagen, and proliferation of vascular smooth muscle cells. It occurs in response to a variety of stimuli, such as pressure, flow, or injury. The remodeling is driven by cells that include endothelial cells, monocytes, and smooth muscle cells and involves the reconstruction of the extracellular matrix. Essentially, arterial stiffness is determined both by vascular tone and by the amount and composition of the extracellular matrix. Specific miRNAs may play key roles in determining tissue homeostasis and regulating the phenotype of cells in the vascular wall.

Arterial stiffening is a hallmark of vascular dysfunction and is an independent predictor of cardiovascular and cerebrovascular events.14 Stiffening of the aorta, as measured by the gold‐standard technique of aortic PWV, is independently associated with adverse cardiovascular outcomes across many different patient groups and in the general population.14 For this reason, we used it to evaluate the changes in arterial stiffness in hypertensive patients after 1 year of successful BP control. Quite often, good control of BP is not accompanied by an improvement in vascular function. Furthermore, in many cases, a reduction in arterial stiffness can be achieved indirectly, by lowering the mean pressure, whereas most therapies do not have a direct effect on the mechanical properties of large arteries. It has been shown that increases in arterial stiffness precede changes in BP and arterial morphology.15 If arterial stiffness is a precursor to hypertension, decreasing BP alone without increasing arterial elasticity might be not sufficient. An improvement in aortic stiffness during hypertension treatment may reduce cardiovascular morbidity and mortality. However, the evidence base for effective treatments still remains small.

On the other hand, there is increasing evidence that changes in miRNA expression play a significant role in cardiovascular diseases. Many miRNAs influence pathogenesis, via their effects on the mechanisms of apoptosis, hypertrophy, and fibrosis, and may be suitable as biomarkers.16, 17 Apart from cardiovascular diseases in general, miRNAs have been particularly associated with vascular remodeling18 and may play a pivotal role in regulating vascular cell functions and contributing to vascular calcification. It is noteworthy that each miRNA may regulate multiple genes, providing extensive translational regulation, and regulate cell–cell and cell–matrix interactions, while they are also sensitive to the mechanical loading and stretching that the vascular wall undergoes in conditions such as arterial hypertension.19, 20

We studied miRNA expression in the monocytes of our patients' peripheral blood, because PBMCs are of great importance in the cardiovascular complications of hypertension.21 They infiltrate into the vessel wall, leading to atherosclerosis via the development of lipid‐laden foam cells. Circulating monocytes adhere to intimal endothelial cells and begin the process of macrophage differentiation and migration.22 The selection of miRNAs was based on their characteristics and the upregulation or downregulation observed in myocardial and vascular remodeling. We chose to study miRNAs whose expression has been proved in clinical and experimental studies to be affected by hypertension and that have been shown to possess a distinct expression profile in arterial hypertension with target organ damage.12

We found that, among those we examined, the only miRNA to show a strong correlation with PWV improvement was miRNA‐21. Vascular remodeling and increased arterial stiffness are caused by diverse etiologies, but miRNA‐21 may play an important pathological role, not only because it is known to regulate important signal transduction pathways connected to fibrosis, but also because it is implicated in various cell homeostasis processes of endothelial and vascular smooth muscle cells. It is a highly expressed miRNA in the cardiovascular system and its behavior is altered in many cardiovascular diseases.23 miRNA‐21 was the first miRNA demonstrated to be involved in the regulation of vascular smooth muscle cell phenotype; it is highly expressed in both vascular smooth muscle and endothelial cells and it promotes their differentiation, proliferation, and apoptosis.24 Inhibition of miRNA‐21 decreases the proliferation of smooth muscle cells and increases their apoptosis.24 miRNA‐21 is upregulated in proliferative vessels, as was confirmed in a mouse ligation model.25 Moreover, increased shear stress, as in arterial hypertension, induces the expression of miRNA‐21 at the transcriptional level in cultured human umbilical vein endothelial cells via an increased binding of c‐Jun to the promoter region of miRNA‐21.26 Most important, however, is that miRNA‐21 is an important regulator of fibrosis, which is probably the main reason for its correlation with arterial stiffness. Recent studies have demonstrated that elevated expression of miRNA‐21 may play a vital role in the development of fibrosis by promoting the proliferation of interstitial fibroblasts and increasing the abnormal deposition of extracellular matrix. miRNA‐21 is essential in the activation of fibrosis related to angiotensin II. The use of a locked nucleic acid targeting miRNA‐21 ameliorates angiotensin II‐induced cardiac fibrosis.27 In the same study, pretreatment with losartan was able to block miRNA‐21 expression in cultured fibroblasts. A positive correlation was also found between miRNA‐21 and fibrosis stage in a CCl4 mouse fibrosis model.28 In addition, by activating the transforming growth factor β signaling pathway, miRNA‐21 suppresses endothelial progenitor cell proliferation, which plays an important role in the maintenance of endothelial homeostasis and the integrity of arterial wall.29 It should be noted that the elastic properties of the aorta and brachial artery are different. In the aorta, elastin is the dominant component, and, in the brachial artery, collagen and smooth muscle cells dominate. The higher contribution of the smooth muscle cells in the arterial compliance of brachial artery may explain the different level of correlation between miRNA‐21 levels and the changes in cfPWV and crPWV.

Finally, miRNA‐21 seems to be an important modulator of macrophage polarization and inflammation mediated by macrophages that might penetrate into the arterial wall.30 All the aforementioned mechanisms may provide insights into the vascular processes by which miRNA‐21 affects the vascular wall and promotes dysfunction in the elastic properties of the arteries.

Study Limitations

Some limitations of this study should be noted. We used PWV for the estimation of arterial stiffness. PWV is not the only measure of arterial stiffness, nor is it adequate for characterizing the pathophysiology of arterial stiffening; however, it is widely accepted and is the recommended method for this purpose.13

This was an observational, not interventional, study and it suffered from the known limitations of that type of investigation. We cannot exclude the possibility that different classes of antihypertensive drugs might have an impact on our findings. Unfortunately, the small size of our population did not permit us to analyze the impact of different antihypertensive drugs on arterial stiffness.

In addition, we did not investigate the changes of microRNA levels after treatment. These data could give us stronger indications of whether we could modify arterial properties by interfering in the level of microRNA. However, our preliminary results appeared to be the first on this subject. They are indicative of a significant role of miRNA‐21 in vascular remodeling and this provides motivation for further research in this direction. Since the existing knowledge is limited, additional clinical and experimental studies are needed in order to provide mechanistic insight into how expression levels of miRNA‐21 were regulated during the antihypertensive treatment and what vascular processes and changes were subsequently modulated.

Conclusions

Our study showed that low levels of miRNA‐21 are strongly associated with an improvement in arterial stiffness in patients with well‐controlled essential hypertension, independently of their BP levels. Our data could provide valuable data for the improvement of risk stratification in patients with essential hypertension, and might possibly indicate populations that need more aggressive intervention as regards their risk factors for cardiovascular diseases. Our findings offer a fresh perspective on the development of a new generation of biomarkers for the better monitoring of vascular remodeling and dysfunction in hypertension. In particular, they highlight miRNA‐21 as a potential diagnostic and prognostic marker and a future therapeutic target in those patients.

Disclosure

The authors report no specific funding in relation to this research and no conflicts of interest to disclose.

J Clin Hypertens (Greenwich). 2017;19:235–240. 10.1111/jch.12900 © 2016. Wiley Periodicals, Inc.