Atrial fibrillation, the most common sustained cardiac arrhythmia, has long resisted one of its most stubborn underlying drivers: the scarring of heart tissue known as atrial fibrosis. The findings, published in Nature Cardiovascular Research, offer a proof of concept for an RNA-based strategy aimed directly at the structural remodeling that makes atrial fibrillation so difficult to treat.
The research began with a puzzle. Patients with persistent atrial fibrillation carry atrial fibroblasts—scar-forming cells—in which the calcitonin receptor, encoded by the CALCR gene, is suppressed. Searching for the upstream culprit, the investigators used computational prediction tools to screen microRNAs reported to change in atrial fibrillation for ones capable of binding the CALCR messenger RNA. Four candidates emerged, but only one survived experimental scrutiny: miR-31-5p, an aging-associated microRNA enriched in the atria.
In a luciferase reporter assay in HEK293 cells, a miR-31-5p mimic cut the reporter signal by roughly 60 percent, an effect fully reversed by a miR-31-5p inhibitor, confirming direct binding to the CALCR 3′ untranslated region. The other three candidate microRNAs showed no binding at all. Fibroblasts from both the right and left atrial appendages of fibrillation patients carried about 2.7-fold and 2.5-fold more miR-31-5p, respectively, and those levels correlated negatively with CALCR expression.
Functional tests in cultured human atrial fibroblasts sharpened the picture. Overexpressing miR-31-5p in cells from patients with normal rhythm halved CTR messenger RNA and protein, while inhibiting the microRNA in fibrillation patient cells increased CTR messenger RNA by about 30 percent and nearly doubled the protein. An RNA immunoprecipitation assay confirmed that both CALCR messenger RNA and miR-31-5p sit together in the RNA-induced silencing complex, the cellular machinery that degrades or silences target transcripts. Blocking the specific binding site with a locked nucleic acid target site blocker reduced CALCR loading into that complex by roughly 60 percent without touching miR-31-5p itself, and actinomycin D chase experiments showed that miR-31-5p accelerates degradation of the CALCR transcript.
To test causality in living animals, the researchers engineered mice in which miR-31 overexpression could be switched on selectively in fibroblasts using a tamoxifen-inducible Cre system. These animals developed roughly double the atrial fibrosis of controls, measured by both Masson’s trichrome staining and hydroxyproline assay, along with about 40 percent increases in atrial collagen-1, collagen-3 and α-smooth muscle actin protein. Ventricular fibrosis and ventricular function remained unchanged, underscoring the atrial specificity of the effect.
The structural changes translated into electrical ones. Burst pacing induced atrial fibrillation in twice as many transgenic mice as controls, and induced episodes lasted longer and occurred more frequently. Notably, these arrhythmogenic changes arose without left atrial dilation or any measurable decline in ventricular systolic or diastolic function, suggesting that fibroblast-driven fibrosis alone can build a substrate capable of sustaining the arrhythmia. Sex-stratified analyses hinted that the prolongation of arrhythmic episodes was more pronounced in females, though the study was not powered to resolve sex-specific mechanisms.
The therapeutic pivot came from a locked nucleic acid target site blocker, an antisense oligonucleotide designed to bind precisely the microRNA recognition element on the CALCR messenger RNA, shielding it from miR-31-5p without globally suppressing the microRNA. The researchers tested it in the Lkb1-aKD mouse, a model of spontaneous atrial fibrillation that recapitulates the human pattern of doubled miR-31-5p and reduced calcitonin receptor expression. In eight-week-old mice treated for three weeks, the blocker rescued atrial CTR messenger RNA and protein, suppressed atrial collagen-1 expression by about 40 percent and collagen-3 by about 25 percent, prevented fibrosis development, and cut arrhythmia duration by roughly 45 percent and burden by about 32 percent. Other validated miR-31 targets—Nos1, dystrophin, Stk40 and Bap1—were untouched, confirming the approach’s selectivity.
The decisive experiment addressed established disease. After three weeks of treatment with the target site blocker at 25 milligrams per kilogram, three times weekly, magnetic resonance imaging using a collagen-binding contrast agent and confirmatory histology showed fibrosis falling about 30 percent below the pre-treatment baseline. In control mice receiving a scrambled blocker, fibrosis instead progressed about 46 percent above baseline. Treated animals also showed reduced atrial messenger RNA for collagen-1, fibronectin and α-smooth muscle actin, and their arrhythmia burden, tracked by continuous electrocardiographic telemetry, dropped by approximately 16 percent.
Parallel experiments in human atrial fibroblasts from fibrillation patients reinforced the translational relevance. Applying the same target site blocker to these cells roughly doubled CTR protein and suppressed collagen-1 messenger RNA and protein by about 68 percent and 51 percent, respectively, while reducing collagen-3 protein by 56 percent. A hybridization assay confirmed the blocker reached the atrial myocardium in mice, and the treatment produced no changes in blood pressure, heart rate or cardiac contractility. The authors caution, however, that the Lkb1-aKD model reflects a specific genetic mechanism dominated by diffuse interstitial fibrosis, so whether the strategy can reverse replacement fibrosis or scar-related remodeling remains unknown, and the durability of the anti-fibrotic effect after treatment stops has yet to be determined.
The study positions the miR-31-5p−calcitonin receptor axis as a promising therapeutic target and, importantly, demonstrates that targeting the binding site rather than the microRNA itself can spare the microRNA’s many other gene targets. Because miR-31-5p is barely detectable in the ventricular myocardium, such an approach could in principle avoid the ventricular side effects that limit many current anti-arrhythmic drugs. The authors note that a relatively modest 16 percent reduction in arrhythmia burden accompanied the substantial fibrosis regression, possibly reflecting the short treatment window or the progressive loss of the receptor’s ligand, calcitonin, in the model. Clinical trials in patients with atrial fibrillation will be needed to establish whether this RNA-based strategy can translate from mouse atria to human ones.
Subject of Research: Targeted RNA blockade of the miR-31-5p−calcitonin receptor interaction as a therapy to reverse atrial fibrosis in atrial fibrillation
Article Title: Selective blockade of microRNA-31-5p−calcitonin receptor interaction reverses established atrial fibrosis and atrial arrhythmia burden
Article References: Trompf, J., Cox, K., Hulsurkar, M., Moreira, L. M., Yiu, C. H. K., Johnston, A. M., Lahiri, S. K., Zhao, S., Robinson, P., Hiram, R., Mehdizadeh, M., Perez Carrillo, L., Sayeed, R., Krasopoulos, G., Srivastava, V., Walcot, N., Farid, S., Kourliouros, A., Sastry, P., … Reilly, S. (2026). Selective blockade of microRNA-31-5p−calcitonin receptor interaction reverses established atrial fibrosis and atrial arrhythmia burden. Nature Cardiovascular Research. https://doi.org/10.1038/s44161-026-00897-1
Image Credits: AI Generated
DOI: 10.1038/s44161-026-00897-1
Keywords: atrial fibrillation, atrial fibrosis, microRNA-31-5p, calcitonin receptor, locked nucleic acid, target site blocker, cardiac fibroblasts, RNA therapeutics, structural remodeling, arrhythmia burden, Lkb1-aKD mouse model, fibrosis regression
News Source: Drew Townsend. (November 5, 2026). Blocking a microRNA−receptor interaction reverses atrial fibrosis in mice. Scienmag.
