The relationship between myocardial fibrosis and myocardial microRNAs in dilated cardiomyopathy: A link between mir‐133a and cardiovascular events

Abstract It is unknown whether fibrosis‐associated microRNAs: miR‐21, miR‐26, miR‐29, miR‐30 and miR‐133a are linked to cardiovascular (CV) outcome. The study evaluated the levels of extracellular matrix (ECM) fibrosis and the prevalence of particular microRNAs in patients with dilated cardiomyopathy (DCM) to investigate any correlation with CV events. Methods: Seventy DCM patients (48 ± 12 years, EF 24.4 ± 7.4%) underwent right ventricular biopsy. The control group was comprised of 7 patients with CAD who underwent CABG and intraoperative biopsy. MicroRNAs were measured in blood and myocardial tissue via qPCR. The end‐point was a combination of CV death and urgent HF hospitalization at the end of 12 months. There were differential levels of circulating and myocardial miR‐26 and miR‐29 as well as myocardial miR‐133a when the DCM and CABG groups were compared. Corresponding circulating and myocardial microRNAs did not correlate with one another. There was no correlation between microRNA and ECM fibrosis. By the end of the 12‐month period of the study, CV death had occurred in 6 patients, and a further 19 patients required urgent HF hospitalization. None of the circulating microRNAs was a predictor of the combined end‐point; however, myocardial miR‐133a was an independent predictor in unadjusted models (HR 1.53; 95% CI 1.14‐2.05; P < .004) and adjusted models (HR 1.57; 95% CI 1.14‐2.17; P < .005). The best cut‐off value for the miR‐133a level for the prediction of the combined end‐point was 0.74 ΔCq, with an AUC of 0.67. The absence of a correlation between the corresponding circulating and myocardial microRNAs calls into question their cellular source. This study sheds new light on the role of microRNAs in ECM fibrosis in DCM, which warrants further exploration.


| INTRODUCTION
Fibrosis of extracellular matrix (ECM) is a major pathology contributing to the progression from compensated dilated cardiomyopathy (DCM) to heart failure (HF). 1 Increasing evidence indicates that microRNAs are important factors in cardiac remodelling, including ECM fibrosis. Thus far, five microRNAs, namely miR-21, miR-26, miR-29, miR-30 and miR-133a, have been found to be implicated in ECM fibrosis, mostly in animal studies. [2][3][4] MicroRNAs function primarily inside cells and tissues, e.g. the myocardium; however, micro-RNAs have been recently detected in blood samples. 5,6 We examined: (i) the relationship between circulating and myocardial microRNAs, (ii) the relationship between myocardial microRNAs and ECM fibrosis, and (iii) the associations between microRNAs and cardiovascular (CV) outcome.

| Study population
Over a period of 14 months, 70 consecutive DCM patients were enrolled. DCM was diagnosed in line with currently accepted criteria. 7,8 The study complies with the World Medical Association Declaration of Helsinki, and the study protocol was approved by the institutional review board and the ethics committee. All patients gave written informed consent.
The control group was comprised of 7 patients with CAD who underwent CABG. During open-heart surgery, transmural needle biopsies of the anterior LV wall were performed. 2

| Statistical analysis
The combined end-point was composed of CV death and urgent HF hospitalization by the 12-month point. To examine any correlation between microRNAs and end-points unadjusted and adjusted for age, duration of disease, CVF, EF and NT-proBNP, Cox proportional hazard analyses were performed.

| The relationship between the corresponding circulating and myocardial microRNAs
Only miR-21 was more highly expressed in the myocardium (1.08 AE 0.97 DCq) compared to blood levels (0.14 AE 0.6 DCq). The remaining microRNAs were at much higher concentrations in the bloodstream compared to cardiac tissue: miR-26 (blood:

| MicroRNAs, fibrosis and CV outcome
No correlation was observed between myocardial microRNAs and fibrosis. There was no difference in any of the circulating microRNAs RUBI S ET AL.
| 2515 between patients with or without an event. However, tissue miR-133a was more highly expressed in patients with an event (  Figure 1).

| DISCUSSION
The biological role and cellular source of circulating microRNAs are largely unknown. 12 Thus, the issue of whether fibrosis-linked micro-RNAs are secreted from the cardiac tissue has yet to be resolved. In this study, corresponding microRNAs in blood and myocardial tissue were examined simultaneously, and the absence of any correlation between them was clearly demonstrated.
Preliminary studies showed that ECM fibrosis was associated with increased expression of pro-fibrotic microRNAs, including miR-21 and miR-208a and decreased expression of anti-fibrotic micro-RNAs, including miR-29, miR-30 and miR-133. [2][3][4] In accordance with these findings, we observed that myocardial miR-29 was up-regulated in our DCM cohort in comparison with CABG patients, whereas myocardial miR-26 and miR-133a were down-regulated in DCM. Thus, the aetiology of HF seems to be crucial with respect to molecular cardiac pathology, including the distribution of microRNAs.
We have previously reported relatively strong correlations between circulating miR-26 and miR-30 with CVF (r = .48, P < .01 and r = .72, P < .001, respectively). 11 Surprisingly, the corresponding myocardial microRNAs did not have any relationship with ECM fibrosis. In a predominantly inflammatory DCM cohort, Besler et al 13 found that increased myocardial expression of miR-133 was weakly associated with fibrosis. The aetiology seems to be of paramount importance as we did not observe any inflammatory infiltrations on biopsy, whereas Besler et al studied solely inflammatory DCM.
Bio-SHiFT investigators showed that of 7 circulating microRNAs previously linked to HF, only miR-22 was associated with the endpoint in a large cohort of unselected HF patients. 14

| Study limitations
The number of patients in the control cohort is small, which was as a result of a slow recruitment rate. Only selected microRNAs can be measured with qPCR. Although next-generation sequencing enables the analysis of virtually all microRNAs, we chose to use qPCR as our aim was to verify the role of the most "promising" microRNAs in relation to fibrosis. were found between myocardial microRNAs and fibrosis. Only myocardial miR-133a was found to be an independent predictor of prognosis. Although preliminary, this study sheds new light on the role of microRNAs in DCM pathology and warrants further exploration.