MicroRNAs (miRs) are short, non-coding RNA molecules that act in an RNA-induced silencing complex (miRISC) with Argonaute proteins to regulate gene expression at the post-transcriptional level. We had previously demonstrated that miR-21 is upregulated in cardiac disease and that in vivo silencing of miR-21 using a synthetic inhibitor (antimiR-21) attenuated cardiac fibrosis and dysfunction. Despite the well-established anti-fibrotic effect of antimiR-21 and its development until current phase 2 clinical testing (HERA trial), the underlying mechanism, i.e. the regulated targetome, has remained largely elusive. Here, we use an Argonaute2-ribonucleoprotein immunoprecipitation (AGO2-RIP) approach combined with single cell RNA sequencing and genetic deconvolution to elucidate the miR-21-mRNA network in failing myocardium.

8 weeks-old mice were subjected to a model of pressure overload-induced cardiac hypertrophy, and antimiR-21(10mg/kg) was injected for three days intravenously two weeks after surgery. Two days later, an AGO2-RIP was performed either from the total heart or from the main cardiac cell types (myocytes, fibroblasts, macrophages or endothelial cells). Next generation sequencing of AGO2-immunoprecipitated mRNA and total RNA followed by bioinformatic analysis identified the most de-enriched (RIP fold change >1, p-value<0.05) and most de-repressed targets (mRNA fold change >1, p-value<0.05) in the antimiR-21-treated group compared to the control group in the total heart and for the main distinct cardiac cell populations. Together with small RNA sequencing we detected a stronger antimiR-21 signal in the non-myocytes group compared to the myocytes. Genetic deconvolution identified macrophages as the primary cell type with regard to transcriptome regulation mediated by antimiR-21. Multicellular network analysis showed a strong and preferential intercellular communication between macrophages and fibroblasts after miR-21 inhibition. Single cell RNA  sequencing further delineated cell subtype-specific antimiR21-effects in the main cardiac cell types.

Combing biochemical and bioinformatic approaches allowed us to define the cell type-specific regulated targetome of microRNA-21 upon its therapeutic manipulation. Our data on miR-21 identified those cardiac cell types, where antimiR-21 takes its full effect and identified the cardiac cell-cell crosstalk mechanisms primarily affected by antimiR-21 treatment. These results provide a basis to develop cell type-specific antimiR-21 molecules with the goal to reduce off-target effects.