Atrial Fibrillation in Heart Failure Is Associated with High Levels of Circulating microRNA-199a-5p and 22-5p and a Defective Regulation of Intracellular Calcium and Cell-to-Cell Communication

Abstract

MicroRNAs (miRNAs) participate in atrial remodeling and atrial fibrillation (AF) promotion. We determined the circulating miRNA profile in patients with AF and heart failure with reduced ejection fraction (HFrEF), and its potential role in promoting the arrhythmia. In plasma of 98 patients with HFrEF (49 with AF and 49 in sinus rhythm, SR), differential miRNA expression was determined by high-throughput microarray analysis followed by replication of selected candidates. Validated miRNAs were determined in human atrial samples, and potential arrhythmogenic mechanisms studied in HL-1 cells. Circulating miR-199a-5p and miR-22-5p were significantly increased in HFrEF patients with AF versus those with HFrEF in SR. Both miRNAs, but particularly miR-199a-5p, were increased in atrial samples of patients with AF. Overexpression of both miRNAs in HL-1 cells resulted in decreased protein levels of L-type Ca²⁺ channel, NCX and connexin-40, leading to lower basal intracellular Ca²⁺ levels, fewer inward currents, a moderate reduction in Ca²⁺ buffering post-caffeine exposure, and a deficient cell-to-cell communication. In conclusion, circulating miR-199a-5p and miR-22-5p are higher in HFrEF patients with AF, with similar findings in human atrial samples of AF patients. Cells exposed to both miRNAs exhibited altered Ca²⁺ handling and defective cell-to-cell communication, both findings being potential arrhythmogenic mechanisms.

Keywords: HL-1 cells; L-type calcium channels; NCX1; atrial fibrillation; atrial remodeling; biomarkers; calcium regulation; connexin 40; heart failure; microRNA.

Figure 1

MicroRNA determination in human left and right atrial appendages. Expression levels of miR-199a-5p and miR-22-5p of AF patients (white columns) compared to those in sinus rhythm (no-AF, black columns). Scatter dot blots represent each individual patient, and the number within the column represents the total of independent experiments performed. * p < 0.05, RAA: right atrial appendage, LAA: left atrial appendage.

Figure 2

Gene and protein expression (A–C). Expression levels of CACNA1C, SLC8A1 and the connexin members, GJA5, GJA1 and GJC1 in HL-1 cells transfected with miR-NC (black columns) or miR-199a+miR-22 (white columns). Normalized to miR-NC (D–F). Protein expression levels of Cav1.2, NCX1 and Cx40, normalized to the endogenous control, tubulin (G). Representative WB experiments in HL-1 cells transfected with miR-22+199a and miR-NC. All proteins (Cav1.2, NCX1, connexin 40 and tubulin) were blotted on the same gel. This is a representative image of four independent experiments. Numbers within columns represent the number of independent experiments done. * p < 0.05, *** p < 0.0001, n.s.: non-significant.

Figure 3

Calcium entrance through LTCC and cytosolic recordings in HL-1 cells transfected with miR-199a+miR-22 (white) or miR-NC (black) (A). Left panel, fluctuations of intracellular calcium recorded in basal conditions in cells loaded with Fura-2AM. Right panel, calcium levels graphed and averaged in column bars (B). Ba²⁺ peak current through Cav1.2 channels at +20 mV recorded using patch-clamp technique in whole-cell configuration (C). Left panel, intracellular calcium fluctuations under depolarizing conditions induced by the application of an extracellular solution containing 75 mM KCl (arrow). Right panel, maximum calcium levels graphed and averaged in column bars. * p < 0.05, ** p < 0.005, *** p < 0.0001. Numbers within the columns represent the number of independent experiments performed with the total number of cells analyzed in parenthesis.

Figure 4

Caffeine-induced calcium response in HL-1 cells transfected with miR-199a+miR-22 (white) or miR-NC (black) (A). Normalized intracellular calcium fluctuations with the application of 10 mM caffeine (arrow). To assess caffeine-induced maximum peak, basal calcium levels in both transfection conditions were normalized to one (B). Maximum calcium levels post-caffeine graphed and averaged in column bars (C). Calcium decay expressed as the time constant, Tau. *** p < 0.0001. Number within the column represents the number of independent experiments performed with the total number of cells analyzed in parenthesis.

Figure 5

Cell-to-cell communication in HL-1 cells transfected with miR-199a+miR-22 (white) or miR-NC (black) (A). A representative experiment following the pass of Alexa 488 between two cells (pairs seen in the left bright field image with the injected cell marked with a red arrow, and the expected direction of Alexa flow is marked with a yellow arrow). Images were taken at four different time points over the course of 90 min (t1–t4). The passage through these cells was so fast that even at the beginning of the recording (t1) the fluorescent dye was already present in both cells (B). Percentage of cells that allowed (stripped) or did not allow (plain) passage of Alexa 488 to neighboring cell. Numbers within the columns represent the absolute value of cells, with a total of 23 in both cases. ** p < 0.005.

Figure 6

Study design. Of 495 patients with HFrEF, 18 (9 with permAF and 9 with SR) were selected for the discovery phase, and 80 (40 per group) for the replication phase. Matching variables for each phase are presented, with the accepted range in parenthesis. Abbreviations: HF-pEF, mrEF, and rEF: heart failure with preserved (pEF), mid-range (mrEF) and reduced ejection fraction (rEF), respectively, according to the ESC guidelines; (41) permAF: permanent AF; SR: permanent sinus rhythm (no history of past or present AF); BMI: body mass index; CKD: chronic kidney disease; Hb: hemoglobin; LVEF: left ventricular ejection fraction; LVEDD: left ventricular end-diastolic diameter; Log(proBNP): pro-BNP values at baseline in Log expression; * included matching for treatment with betablockers, angiotensin-converting-enzyme inhibitors (ACEI), angiotensin-receptor blockers (ARB), and diuretics.