In this issue of Blood, Xiao et al provide a direct link between the OTT protein and regulation of splicing of the RNA for c-MPL, which encodes the thrombopoietin receptor.1
OTT was first identified as part of the recurrent t(1;22) translocation in acute megakaryoblastic leukemia (AMKL) and was called OTT for “one twenty-two.” OTT1 is also known as RNA binding motif protein 15 (RBM15). Like other RNA binding motif proteins, RBM15 contains RNA recognition motifs (RRMs), 3 in total, which can bind to RNA and, at least for some RRMs, may also bind to DNA. Lifelong maintenance of hematopoietic stem cell (HSC) numbers is essential, and the genes for Ott and for Mpl are required for such maintenance. In prior studies from Dr Glen Raffel's laboratory at the University of Massachusetts and other laboratories, Rbm15 expression was shown to be critical for maintenance of HSCs in vivo in mice under stress.2 Loss of Rbm15 led to increased cycling and subsequent exhaustion of HSCs under stress conditions. In the current work, Xiao et al1 suggest a potential mechanism leading to this stem cell exhaustion.
Mpl, the receptor for thrombopoietin (Thpo), is expressed on hematopoietic stem and progenitor cells as well as on cells of the megakaryocyte lineage. Like Rbm15 knockout (KO) mice, mice lacking expression of either Mpl or Thpo develop exhaustion of the HSC pool. The full-length Mpl mRNA encodes the full-length transmembrane protein (Mpl-FL). However, a splice variant of Mpl that lacks exons 9 and 10 encodes a truncated Mpl protein (Mpl-TR), which lacks the transmembrane and intracellular domains.
In HSCs from Rbm15 KO mice, the ratio of Mpl-TR to Mpl-FL is greatly increased compared with control HSCs, suggesting that Rbm15 acts to promote the full-length Mpl transcript. To address whether expression of truncated Mpl affects the cellular response to Thpo, thereby explaining the stem cell defect in Rbm15 KO HSCs, Xiao et al1 tested the ability of HSCs to engraft recipient mice in control conditions or conditions in which they overexpress the truncated form of the Mpl protein. As predicted, overexpression of Mpl-TR decreased HSC function in a dominant-negative manner, most likely due to decreased signaling from Thpo in these cells.
Because Rbm15 is known to be a component of the spliceosome, this study tested the hypothesis that Rbm15 may regulate Mpl splicing. Xiao et al1 demonstrate that Rbm15 indeed binds Mpl pre-mRNA, suggesting that Rbm15 promotes exon retention in Mpl in HSCs. To dissect the mechanism by which Rbm15 affects exon inclusion, chromatin immunoprecipitation was performed. Confirming their hypothesis that cotranscriptional splicing (pre-mRNA being spliced during the process of transcription) underlies the splicing mechanisms conferred by Rbm15, the investigators found that in addition to binding Mpl pre-mRNA, Rbm15 also binds to regions of the Mpl genomic DNA in the vicinity of the spliced exons. An epigenetic (chromatin-modifying) function for Rbm15 had previously been shown, in that Rbm15 interacts with the Setd1b histone H3K4 methyltransferase (H3K4me).3 Consistent with this association, H3K4me levels differed on the Mpl locus in Rbm15 wild-type (WT) vs KO cells: in the absence of Rbm15, trimethylated H3K4 (H3K4me3) was markedly reduced in the region shown to bind Rbm15 in WT cells. In contrast, H4 acetylation was increased in the absence of Rbm15. Both such chromatin modifications, H4 acetylation and H3K4me3, are associated with an increased rate of transcription and spliceosome recruitment.4 The authors suggest that lack of H3K4me3 on the Mpl locus due to absence of Rbm15 results in faster transcription across the region of the alternatively spliced exons, resulting in their exclusion from the final transcript.
Conversely, in the presence of Rbm15 in WT cells, transcription across the Mpl gene is slower, and Mpl-FL, with inclusion of all the exons, is favored. The regulatory role of histone modifications on alternative splicing of Mpl RNA was corroborated by showing that addition of a histone deacetylase inhibitor to WT cells resulted in an increase in the Mpl-TR:Mpl-FL ratio. The study presented here reveals a novel level of regulation of stem cell function by linking transcription and splicing. This finding opens up new therapeutic avenues for stem cell regulation in vivo and in vitro, whereby epigenetic modifiers regulate alternative splicing events.
Now that Xiao et al1 have shown that RBM15 affects differential splicing of a gene critical for HSC maintenance, the path is cleared for future studies focused on several questions: What other RNAs are differentially spliced in the presence/absence of RBM15? What is the mechanism by which RBM15 regulates splicing? What is its interaction with the core spliceosome? How is RBM15’s association with specific genes regulated? What is the dominant-negative mechanism by which Mpl-TR acts to decrease HSC maintenance? Answers to these questions not only will elucidate critical aspects of the splicing machinery, but will be relevant for our understanding of the role of the RBM15-MKL1 fusion protein in AMKL.
Conflict-of-interest disclosure: The authors declare no competing financial interests.
