1Universitätsklinikum Heidelberg Innere Medizin VIII, Institut für Experimentelle Kardiologie Heidelberg, Deutschland; 2Universitätsklinikum Heidelberg Klinik für Innere Med. III, Kardiologie, Angiologie u. Pneumologie Heidelberg, Deutschland; 3Deutsches Krebsforschungszentrum (DKFZ) Heidelberg, Deutschland; 4Medizinische Fakultät Mannheim der Universität Heidelberg NGS Core Facility Mannheim, Deutschland; 5Universität Köln Zentrum für Molekulare Medizin (CMMC) Köln, Deutschland; 6Philipps-Universität Marburg Pharmakologie und Klinische Pharmazie Marburg, Deutschland; 7Universitätsklinikum Heidelberg Dietmar-Hopp-Stoffwechselzentrum Heidelberg, Deutschland
Background. ACADVL deficiency (VLCADD) is a rare genetic cause for infantile heart failure. Acyl-CoA dehydrogenase, very long-chain specific (ACADVL) represents the rate-limiting enzyme of fatty acid oxidation in mitochondria. With a substrate specificity towards C14 to C18 it is particularly critical for the degradation of very long-chain fatty acids. Although downregulation of ACADVL is also frequently seen in common acquired heart diseases, its regulatory mechanisms and functional role are not understood.
Results. Using the Myocardial Applied Genomics Network, we found significantly reduced Acadvl expression in ventricular tissue of HFrEF patients compared to non-failing organs indicating decreased ACADVL levels as a common hallmark character across HFrEF as a heterogeneous entity. To identify potential regulators of Acadvl we analyzed a genome-wide CRISPRi perturbation sequencing. By this systematic approach Protein Arginine Methyltransferase 5 (PRMT5) was found to strongly correlate with Acadvl. PRMT5 is the main representative of type II PRMTs catalyzing symmetric dimethylation of arginine residues. Reported substrates include histones as well as non-genomic proteins such as RNA-binding proteins linking enzyme activity to transcriptional regulation and RNA metabolism. As a proof-of-principle we used a mouse model with inducible cardiomyocyte-specific deletion of Prmt5 (cKO) compared to Cre+ wildtypes (Ctrl.). Strikingly, induction of a cKO in adult mice led to progressive systolic dysfunction within 6 months (ejection fraction (EF), Ctrl.: 85.4±1.0%;n=12 vs cKO: 50.6±5.0%;n=10; p<0.0001) and upon β-adrenergic stress (EF, Ctrl.+Isoprenaline (ISO): 68.9±2.4%;n=15 vs cKO+ISO: 54.9±4.5%;n=11; p=0.0012) along with Acadvl downregulation, accumulation of very long-chain fatty acids and mitochondrial dysfunction. Since this cardiomyopathy phenocopies VLCADD, we hypothesized that ACADVL downregulation might be critical in the pathogenesis of Prmt5-cKO-induced HFrEF. Interestingly, maintained Acadvl expression via adenoviral-associated vectors was sufficient to prevent HFrEF unmasking ACADVL as critical mediator (EF, cKO+ISO+Ctrl-AAV: 59.0±2.3%;n=6 vs cKO+ISO+Acadvl-AAV: 79.8.2±1.4%;n=5; p=0.0062). Mechanistically, loss of PRMT5 leads to extensive changes in differential splicing with intron retention (IR) as predominating type in cardiomyocyte nuclei. Specifically, Acadvl was found to exhibit IR with a deltaPSI of 0,39 indicating that the splicing event has large effect size and, hence, mechanistical significance. Upon pharmacological PRMT5 inhibition dose-dependent accumulation of intron retaining Acadvl transcripts was exclusively found in nuclear fractions. Moreover, we unmasked that IR transcripts are absent in the cytoplasm pointing to IR-induced nuclear detention as causative mechanism for the global loss of functional ACADVL in vitro and in vivo.
Conclusion. Here, we provide evidence that loss of ACADVL is causative for HFrEF progression, expanding its pathophysiological significance from a rare genetic syndrome to common diseases. We further identify PRMT5 as a critical upstream regulator of ACADVL. Mechanistically, PRMT5 controls mRNA splicing of Acadvl: PRMT5 deficiency induces intron retention leading to nuclear detention and subsequent loss of functional ACADVL. Consequently, our findings suggest Acadvl gene therapy not only as novel causative treatment for VLCADD but for progressive HFrEF in general.