Endothelial SOX9 modulates cardiac aging by controlling fibroblast activation, immune cell recruitment and endothelial cell senescence

Introduction. Previously, we reported that the transcription factor SOX9 serves as master regulator of fibrotic remodeling in endothelial cells (ECs) in preclinical models of different heart failure entities by inducing both endothelial-to-mesenchymal activation and paracrine activation of fibroblasts. Whether endothelial SOX9 contributes to the development of cardiac failure during aging, has so far not been investigated.

Objective. Aim of the study was to investigate the impact of endothelial SOX9 on the development of fibrotic cardiac disease in aged mice.

Methods & Results. Aged mice (24 months old) displayed endothelial induction of SOX9 alongside a reduction of cardiac function, capillary rarefaction and myocardial fibrosis. Tamoxifen-induced, Cdh5-CreERT2 driven deletion of Sox9 (Sox9^EC-KO) in endothelial cells upheld cardiac function and reduced fibrotic lesions throughout 24 months. To investigate the underlying mechanisms, we performed whole-heart non-cardiomyocyte single-cell RNA sequencing. Unsupervised clustering identified 16 cell clusters, of which we focused on ECs, fibroblasts (FBs) and immune cells. Repeated cluster amplification of these cell types revealed quantitative shifts in cluster distribution, e.g. increased B cell (BC) and reduced neutrophil (NT) numbers in aged Sox9^EC-KO compared to aged Sox9^fl/fl control mice. Pathway analysis with CellChat v2 identified several upregulated immune cell-activating ligand-receptor relations from ECs to BCs in aged Sox9^EC-KO vs. control littermates, and profound anti-inflammatory and pro-apoptotic signaling from BCs to NTs, suggesting cardioprotection by regulatory BCs in aged Sox9^EC-KO. Additionally, FB stimulating signaling was upregulated in aged control mice, but less in Sox9^EC-KO mice, suggesting maturation from resident to rather secretory FB subtypes, which was confirmed by a pseudotime trajectory analysis and which was promoted by endothelial SOX9. Interestingly, these activated FBs signaled back towards ECs fostering anti-angiogenic cascades, which was confirmed by reduced capillary and nerve density in aged control mice via immunofluorescence staining. Lastly, EC senescence was remarkably reduced in aged Sox9^EC-KO vs. control mice, indicating that SOX9 promotes senescence in ECs.
In order to test a potential therapeutic feasibility of Sox9 inactivation, we applied Tamoxifen to so far unchallenged 18 months old Cdh5-CreERT2-Sox9^fl/fl mice. Interestingly, the delayed deletion of EC Sox9 restored cardiac function, capillary density, reduced fibrosis and endothelial senescence, and partially reverted cellular cluster composition.

Conclusions. Here, we present at least three mechanisms by which endothelial SOX9 promotes aging-induced cardiac dysfunction in a cell autonomous manner as well as well as in concert with fibroblasts and immune cells. Inactivation of endothelial Sox9 might be a promising therapeutic approach against cardiac dysfunction in the elderly.