Single Nuclei RNA Sequencing Reveals Cell Type Specific Senescence Signatures in Heart Disease (DGK-Forschungsstipendium)

Background: Aging remains a major, currently unmodifiable cardiovascular risk factor with underlying mechanisms that are incompletely understood. Cellular senescence is a complex and heterogeneous process impacting distinct cell populations in response to various stimuli. Senescent cells accumulate in different tissues during ageing and cardiovascular disease. However, their role in disease is strongly dependent on cell type and was shown to range from protective to detrimental. Therefore, a detailed characterization of the cell-type and disease-specific effects of senescence is essential for the development of targeted therapies.

Aim: To characterize cell-type and disease-specific transcriptomic signatures of senescent cells in myocardial samples of patients with heart failure with reduced ejection fraction (HFrEF) or calcific aortic valve stenosis (AVS).

Methods: To identify senescent cells in myocardial samples, we performed single-nucleus RNA sequencing (snRNA-Seq) on hearts of global Tert-KO mice to serve as a positive control of replicative senescence. Comparative differential gene expression analyses in aged, senolytic-treated and Tert-KO mice were conducted to identify cell-type-specific senescence marker genes. Using scoring approaches, we classified individual cells of snRNA-Seq datasets from human HFrEF and AVS patients as either senescent or non-senescent and further characterized their transcriptomic profiles.

Results: Established senescence-associated gene sets, and canonical markers, such as Cdkn2a, Cdkn1a and Trp53 did not effectively identify increased frequencies of senescent cells in snRNA-Seq datasets of mouse models for cellular senescence. Different cardiac cell types in Tert-KO, aged and senolytic-treated hearts exhibited distinct transcriptomic profiles. Overlap analyses identified Tead2, Mctp1, Bsg and Mat2a as commonly regulated genes in endothelial cells (EC), three of which are known to play roles in cellular senescence. Using the combination of these genes to identify senescent endothelial cells in human heart disease datasets revealed an increased frequency of such EC in AVS and HFrEF. High-scoring EC displayed distinct transcriptomic profiles between these diseases with genes associated with amyloid formation upregulated in AVS samples and genes linked to the ubiquitin-proteasome system and autophagy being upregulated in HFrEF samples. 

In fibroblasts (FB), we identified 13 commonly regulated genes across the senescence mouse models. Notably, many of these genes associated with complement system regulation, which might play a role in fibroblast metabolism and activation. However, these 13 regulated genes were not up-regulated in AVS or HFrEF, suggesting that FB senescence may not be augmented in human disease pathologies.

Conclusion & Outlook: snRNA-Seq and cardiac-specific senescence scores demonstrated different abundance and distinct transcriptomic profiles of cells, that exhibited high cell type-specific senescence-scores. Validation of these findings by spatial transcriptomics in human cardiac tissue sections and mechanistic in vitro experiments may provide for the development of targeted therapies.