Pathogenic RUNX1 C-Terminal Truncation Mutations Contribute to Differentiation Block and Result in Unique Changes in Transcription and Chromatin Organization

Abstract

The transcription factor RUNX1 is one of the most frequently mutated genes in hematological disorders, including familial platelet disorder (FPD), predisposition to myeloid dysplastic syndromes (MDS) and acute myeloid leukemia (AML). In addition to a large number of mutations in the N-terminal DNA binding domain (RHD) of RUNX1, 27% of all hematopoietic mutations lie within the C-terminal region beyond the RHD. Previous studies have mainly focused on loss of function RHD mutations while C-terminal mutations are not well characterized and lack "hotspot" mutations. The majority of mutations in the C-terminal region are frameshift and nonsense mutations rather than missense. Some of these mutations are thought to be subject to degradation by nonsense-mediated decay (NMD), emulating loss of RUNX1. Here we demonstrate that RUNX1 C-terminal truncation mutations represent an alternative means of RUNX1 pathogenesis beyond loss of RUNX1 and RHD mutations. To examine the prevalence of RUNX1 C-terminal mutations we first analyzed publicly available data (COSMIC - Sanger Institute) and identified 247 RUNX1 nonsense/frameshift mutations C-terminal of the RHD in patients with blood disorders. Importantly, 76% of these mutations are predicted to escape Nonsense Mediated Decay (NMD) and produce protein products. To study this class of RUNX1 mutation, we introduced the pathogenic RUNX1R320* mutation into the endogenous RUNX1 locus of the human blood cell line K562 using CRISPR-Cas9. We experimentally validated that the RUNX1R320* transcript is not subject to NMD and is expressed at significantly higher levels than the wild-type at both the RNA and protein level. The resulting truncated protein retains the DNA binding RHD but lacks C-terminal domains for binding to other cofactors. To identify biological changes caused by the RUNX1R320* mutation we examined effects on proliferation, differentiation, and DNA damage. Although no significant change in proliferation was observed, megakaryocytic differentiation was impaired. After TPA treatment RUNX1R320* cells failed to fully differentiate and retained high levels of the erythroid marker CD235a. Furthermore, RUNX1R320* showed sensitivity to DNA damage after treatment with etoposide while cell death via apoptosis was not significantly affected. Together these data suggest that C-terminal truncation RUNX1 contributes to differentiation block and an increase in DNA damage sensitivity while evading apoptotic cell signaling pathways. As RUNX1 functions as a master transcriptional regulator of hematopoiesis, we examined how RUNX1 truncations alter transcription and if the effects emulate RUNX1 loss of function mutations by performing RNA-seq with parental, RUNX1 knockdown, and RUNX1R320* knock-in cells. We first identified 2880 significantly differentially expressed genes (DEGs) between wild-type and RUNX1R320*. When compared to RUNX1 knockdown data, we found that the majority of RUNX1R320* DEGs (2052) were unique to RUNX1R320*. This demonstrates that RUNX1R320* results in different transcriptional perturbations than the loss of RUNX1. We next asked whether changes in chromatin confirmation attributed to the observed transcriptional alterations. To study this the newly developed technology Global RNA Interactions with DNA by deep sequencing (GRID-seq) was performed on wild-type and RUNX1R320* cells. Combining these data with H3K27ac and H3K4me3 ChIP-seq data to identify active enhancers and promoters, respectively, we detected 2,891 significantly altered enhancer-promoter (E-P) pairs. Focusing on DEGs with GRID-seq detected enhancer RNA (eRNA), we conducted Gene Ontology (GO) analysis and found enrichment in pathways of chronic myeloid leukemia, regulation of hematopoietic stem cell differentiation, and cell division, revealing the effects of RUNX1R320* on key hematopoietic and myeloid pathways. In summary, we show that RUNX1 C-terminal truncation mutants can escape NMD, contribute to differentiation block, and increase sensitivity to DNA damage. Furthermore, RUNX1R320* affects transcription in a manner distinct from RUNX1 loss. In our GRID-seq analysis, we identify changes in enhancer-promoter interactions that affect key hematological pathways. Our findings demonstrate that pathogenic RUNX1 C-terminal mutations disrupt hematopoietic cell function, transcription, and chromatin in a distinctive fashion.