Characterizing the Genome Targeting Mechanisms of Caspase Activated DNase (CAD) During Cell Differentiation

Abstract

In skeletal muscle progenitor cells, caspase-3 activity engages the differentiation program, in part through the targeted activation of caspase activated DNase (CAD). Once activated, CAD nuclease induces transient DNA strand breaks, where CAD targeting of the promoter region of the cell cycle inhibitor p21, leads to p21 induction, which is itself a potent differentiation signal. However, the DNA strand breaks in differentiating muscle cells are numerous, suggesting that CAD may target multiple loci to influence cell fate. Here, we sought to identify the CAD targeted genome (and associated epigenetic alterations), in an effort to understand how this nuclease influences skeletal muscle cell differentiation. Using CUT & Tag (Cleavage Under Targets and Tagmentation), we mapped CAD genome wide association and identified two large cohorts of CAD binding preferences, promoter regions and to exon/intron boundaries. In particular, for the intragenic sites, we identified an enriched CAD binding to known translocation breakpoints in the Pax7 and Foxo1 genes. The resulting Pax7-Foxo1 fusion gene is the cause of alveolar rhabdosarcoma (aRMS) and our data show that Pax-7, Foxo1 gene breaks are detected during muscle cell differentiation, along with evidence of the fusion transcript. Pax7 expression must decline for the muscle cell differentiation program to proceed, and our data is consistent with a CAD mediated targeting of the Pax7 gene to disrupt its expression and secure the differentiation program. We also noted robust association of CAD to a large number of genes (in differentiating muscle cells) that are implicated in chromosomal translocations for both leukemia and lymphoma. The CAD genome targeting of breakpoint translocations appears to be a conserved phenomenon, as CAD CUT & Tag in differentiating T cells revealed a large overlap with the CAD targets in muscle cells, suggesting that the origin of many cancer causing breakpoint translocations may develop from CAD activity in differentiating progenitor cell populations. Our observations also demonstrated that CAD did not retain sequence specificity for genome targets, suggesting that other chromatin and/or epigenetic modifications may control CAD DNA targeting events. We reasoned that such epigenetic modifications may be repressive, as blocking CAD access to the genome would be a priority for the cell, as it would act to limit premature engagement of CAD induced differentiation. Enhancer of zeste homolog 2 (Ezh2) is a key component of the multiprotein polycomb repressive complex 2 (PRC2) that directs K27 lysine methylation to control gene expression, and CUT & Tag experiments revealed a significant overlap in the genome wide binding sites of Ezh2 and CAD in muscle cells. In a separate set of experiments, we also demonstrate that caspase-3 cleavage inactivates Ezh2, suggesting that caspase 3 protease may act to remove the epigenetic constraints that limit CAD genome targeting during muscle cell differentiation. Taken together, our observations suggest that CAD control of muscle cell differentiation is a complex genome reprogramming event, that may give rise to the origin of cancer causing break point translocations.