RNA-binding proteins in hematopoiesis and hematological malignancy

Abstract

RNA-binding proteins (RBPs) regulate fundamental processes, such as differentiation and self-renewal, by enabling the dynamic control of protein abundance or isoforms or through the regulation of noncoding RNA. RBPs are increasingly appreciated as being essential for normal hematopoiesis, and they are understood to play fundamental roles in hematological malignancies by acting as oncogenes or tumor suppressors. Alternative splicing has been shown to play roles in the development of specific hematopoietic lineages, and sequence-specific mutations in RBPs lead to dysregulated splicing in myeloid and lymphoid leukemias. RBPs that regulate translation contribute to the development and function of hematological lineages, act as nodes for the action of multiple signaling pathways, and contribute to hematological malignancies. These insights broaden our mechanistic understanding of the molecular regulation of hematopoiesis and offer opportunities to develop disease biomarkers and new therapeutic modalities.

Figure 1.

SF3B1, SRSF2, and U2AF function in splicing. (A) The U1 small nuclear ribonucleoprotein (snRNP) and U2AF initially bind to the 5′′ and 3′′ splice sites, respectively. This is followed by binding of the SF3B-containing U2 snRNP and, subsequently, assembly of a multiprotein complex (including U4, U5, and U6 snRNPs) known as the “spliceosome,” which then leads to excision of the intervening intron. Further sequences within exons and introns act as splicing enhancer or silencer elements and are bound by proteins, such as hnRNPs and SR proteins (eg, SRSF2). These RBPs allow splicing to be controlled in a tissue-developmental stage– and stimulus-specific manner. (B) SRSF2 binds equally to GGNG and CCNG exonic splicing enhancers (ESE) to allow expression of EZH2 in healthy HSPCs. In MDS/chronic myelomonocytic leukemia, the P95H mutation of SRSF2 has preferential binding to the CCNG ESE, giving rise to a splice variant of EZH2 including an exon with a premature stop codon that is degraded by NMD.

Figure 2.

Signaling to cap-dependent translation initiation. Cap-dependent translation can be controlled through activation of the PI3K-mTOR and MAPK pathways. Binding of eIF4E and eIF4G is required for eIF4F function and translation of many mRNAs; however, this can be inhibited by eIF4E-binding protein (4E-BP). mTORC1 controls the binding of 4E-BP to eIF4E through phosphorylation of 4E-BP. Furthermore, mTORC1 can control the availability of eIF4A through activation of S6K1/2, which phosphorylates PDCD4, releasing eIF4A. Mitogen-activated protein kinase-interacting kinase 1/2, which is bound by eIF4G, can also regulate translation by phosphorylating eIF4E. The PI3K-mTOR and MAPK pathways converge to phosphorylate eIF4B, a cofactor of eIF4A, leading to increased eIF4A activity.

Figure 3.

m6A mRNA methylation control and dysregulation in AML. (A) The amount of m6A mRNA methylation in a cell is determined by the activities of methyltransferases (METTL3 and METTL14) and demethylases (eg, FTO and ALKBH5). (B) In healthy HSPCs, knockout of METLL3 reduces methylation, increases differentiation, and reduces cell growth, whereas increased METTL3 has the opposite effect. AML cells frequently have increased METTL3 and increased methylation. In AML, knockout of METTL3 leads to increased differentiation and apoptosis.