Dysregulation of Splicing Patterns in MDS Induced By the S100A9/Fto Axis

We previously reported that S100A9 promotes ineffective hematopoiesis and the development of MDS in a feed forward age-dependent fashion. Nonetheless, the precise mechanism by which S100A9 may foster DNA damage in MDS remains unclear. We recently showed that S100A9 directs overexpression of the fat-mass and obesity-associated gene (FTO) encoding an m⁶A RNA demethylase, which leads to nuclear exclusion of SRSF2. Removal of SRSF2 from its functional domain in the nucleosome leads to stalling of RNA polymerase II and formation of the nucleic acid R-loops, comprising DNA:RNA hybrids with the associated non-template single-stranded DNA. S100A9/FTO axis activation leads to SRSF2 deregulation through suppression of its main nuclear transport protein RanBP2, thereby stalling transcription machinery with resulting accumulation of nuclear R-loops and cytosolic/extracellular RNA:DNA hybrids. Persistent R-loops induce DNA damage while also compromising DNA repair. Here we identify an S100A9/FTO-regulated pathway responsible for induction of genomic instability through the accumulation of cytoplasmic RNA:DNA hybrids and modification of the spliceosomal patterns of aged S100A9Tg mice matching MDS hematopoietic stem and progenitor cells (HSPC). We first investigated which components of the S100A9/FTO axis are critical to hematopoiesis and those that are important for both the development of RNA:DNA hybrids and γH2AX activation. We analyzed the contribution of RanBP2 and the effects of elimination of R-loop formation via overexpression of RNAse H1, an enzyme that removes stalled R-loops in the nucleus by degrading DNA-hybridized RNA, thereby reducing the accumulation of cytoplasmic RNA:DNA hybrids. CRISPR knock-down of RanBP2 showed that the protein is critical for accumulation of yH2AX defined by double stranded breaks (DSB). Importantly, overexpression of RNAse H1 degraded R-loops and restored colony-forming capacity, indicating that RNA:DNA hybrids induced by the S100A9/FTO have profound effects on hematopoietic potential. However, while the FTO exclusion of SRSF2 from the nucleus explains the accumulation of γH2AX, it should potentially impact global RNA splicing. To investigate this, we performed a comparative RNAseq analysis on WT and S100A9Tg mice (young and old) to understand both changes induced through the normal aging process as well as those compounded by S100A9. We found that genes linked to splicing, RNA development, nucleotide excision repair and genomic instability and ribosome function were downregulated in aged S100A9Tg mice. Further analysis comparing splicing patterns of S100A9Tg and WT mice with human MDS BM HSPC led to ~200 common genes that were analyzed further. These genes showed that there are splicing changes enriched in spliceosomal assembly and mRNA splice selection site genes. We also found that the enriched genes affect the nucleolus and ribosome formation matching what is seen phenotypically with MDS. Dysregulation of these pathways are highly consistent with our observations of the pathways affected by the S100A9/FTO-induced inflammaging process, validating our hypothesis of S100A9 as a common initiator of dysfunction that can give rise to MDS. Importantly, our data demonstrates the potential for spliceosomal dysfunction regardless of the presence of spliceosomal mutations in MDS. We are currently in the process of performing both DRIPseq and m6A-seq of primary human MDS specimens and S100A9Tg mice to further assess the role of the S100A9/FTO pathway in the selection of sites for RNA/DNA hybrid formation and rise of genomic dysfunction that gives rise to MDS. We conclude that S100A9/FTO-induced nuclear exclusion of SRSF2 aids in the formation of RNA:DNA-hybrids that lead to genomic instability and the disruption of normal spliceosomal patterns in both human HSPC and the S100A9Tg MDS murine model, representing a previously uncharacterized mechanism contributing to MDS pathogenesis. Our studies provide evidence that targeting this cascade offers significant potential for development of novel, biologically rational therapeutics for MDS.