Splicing factor-mutant clonal hematopoiesis associates with profoundly increased mortality and detrimental paracrine effects on different cardiac cells

Background: Over 50 somatic mutations in hematopoietic stem cells can result in clonal hematopoiesis of indeterminate potential (CHIP). This age-related phenomenon is characterized by the non-malignant clonal expansion of hematopoietic stem cells. CHIP is associated with poor outcomes for patients with cardiovascular disease. Less frequent high-risk mutations, i.e. in splicing factor genes SRSF2, SF3B1, and ZRSR2, were linked to greater myeloid malignancy, but may also contribute to increased cardiovascular mortality risk. However, the mechanisms underlying the crosstalk between mutation-bearing immune cells and cardiac tissue are yet to be discovered.

Purpose:  We aim to assess the paracrine interactions between macrophages with CHIP-driver mutations in splicing factor genes and cardiac cell populations.

Methods & Results: The incidence and relevance of CHIP-driver mutations in splicing factors were analyzed in 1018 patients with heart failure or aortic stenosis using targeted deep sequencing. CHIP-driver mutations in splicing factor genes SF3B1, SRSF2 or ZRSR2 were detected in 42 patients. Patients harboring mutations in the three splicing genes had significantly increased four-year mortality of 51% compared to patients with non-splicing factor CHIP (32%) or no CHIP at all (21%; p<0.01). To assess the impact of the individual CHIP-driver mutations, we studied the paracrine activity of CD14+-derived primary human macrophages, in which individual splicing regulator genes were silenced. Accordingly, we treated different cardiac cell types with macrophage conditioned medium. Silencing of distinct splicing genes differentially affected the paracrine effects of human macrophages, with depletion of ZRSR2 showing the most profound impact. Notably, supernatants from ZRSR2-depleted macrophages significantly reduced cardiomyocyte beating frequencies, induced hypertrophy and apoptosis (all p<0.05). Moreover, ZRSR2-silencing in macrophages increased monocyte-to-endothelial cell adhesion, inhibited fatty acids uptake and led to human endothelial cells activation as shown by upregulation of ICAM-1 and downregulation of CDH5 (all p<0.05). Finally, loss of splicing factor ZRSR2 in macrophages induced COL1A1 expression in humanized cardiac tissue mimetics (p<0.05). In contrast, siSF3B1-macrophages interfered with endothelial cell function by increasing monocyte-to-endothelial cell adhesion and activation. Conversely, depletion of SRSF2 in macrophages decreased endothelial cell attachment. First transcriptomic insights into macrophages with ZRSR2-silencing revealed a pro-fibrotic M2-like macrophage phenotype with upregulation of responsiveness to mechanical stimuli and downregulation of cell cycle processes. Moreover, upregulated genes encoding secretory proteins were implicated in pathways related to extracellular matrix (ECM) formation, collagen biosynthesis and degradation.

Conclusion & Clinical relevance: Splicing factor-linked CHIP-driver mutations are associated with profound increase in long-term mortality in patients with heart failure or aortic stenosis. Loss of splicing genes differentially affects the paracrine activity of human monocyte-derived macrophages. Particularly silencing of ZRSR2 in macrophages detrimentally affects cardiac cells and collagen deposition in humanized cardiac tissue mimetics. These findings hold promise to elucidate mechanisms underlying the poor prognosis of patients carrying CHIP-driver mutations in splicing factors.