Matrin3 is a Critical Splice Factor for Cardiac and Lymphvascular Development

Abstract

RNA binding proteins regulate post-transcriptional processes such as alternative splicing, thereby contributing to the overall proteome diversity. Cardiac alternative mRNA splicing is involved in the pathophysiology of heart failure in humans; however, the complexity of involved RNA-binding proteins remains unsolved. In the heart, extensive research revealed the function of the muscle-restricted RNA-binding protein RBM20 as an essential cardiac splice factor. In this study, the broadly expressed RBM20-paralogue MATR3 (Matrin3) was identified as an alternative splicing regulator, critical for cardiac function. Loss of MATR3 loss results in heart failure due to dilated cardiomyopathy, most likely due to defects in alternative splicing of cardiac-specific transcripts (identified by the TRIBE method). Absence of MATR3 results in mis-splicing or downregulation of physiological important target transcripts in vivo. Remarkably, paralogous RBM20 and MATR3 share a subset of commonly spliced transcripts and even local splice sites within targets such as Titin, Cypher, and Calcium/Calmodulin dependent Protein Kinases, revealing partial overlapping functions in respect to alternative splicing in cardiomyocytes. Titin was found to be a major pre-mRNA target of MATR3 and RBM20, since the loss of MATR3 leads to an elevated N2BA/N2B ratio, whereas loss of RBM20 results in a giant Titin variant as described previously. Furthermore, this study demonstrates that MATR3 is important for lymphangiogenesis, since lymphatic vessels become hyperplastic in response to loss of MATR3 in cell culture and in vivo, whereas blood vessels maintain their network and function. Consequently, constitutive and tissue-specific Matrin3 knockout mice models develop lymphedema and embryonic lethality due to dilated, malfunctional dermal lymphatic vessels and jugular lymph sacs. In addition to the function of MATR3 as a splicing factor within cardiomyocytes and the lymphatic endothelium, we shed light on the heterogeneous, cell-density-dependent expression and protein abundance of Matrin3 in cardiac and endothelial cells. Density-dependent changes in the localization of YAP correlate with the abundance of MATR3 protein. In addition, increased constitutive nuclear YAP activity in cardiomyocytes enhances Matr3 expression with concomitant upregulation of canonical targets, identifying MATR3 as a potential new downstream target of YAP. In conclusion, our results elucidate the essential role of MATR3-mediated alternative splicing for the cardiovascular system, since loss of MATR3 promotes pathophysiological alterations of the lymphatic network and the heart. In the future, we aim to link our current results with human disease conditions such as dilated cardiomyopathy and primary lymphedema.