The transcription factor RUNX1 is a key regulator of both Acute Myeloid and Acute Lymphoid Leukemia. Work by our lab and others has shown that RUNX1 binds Histone Deacetylase 1 (HDAC1) and that the two factors co-localize at the promoters of actively transcribed RUNX1 target genes in leukemia. However, the role of HDAC1 in this complex is poorly understood. To test whether HDAC1 activity regulates the interaction of RUNX1 with other co-factors, we performed a mass spectrometry screen for RUNX1 binding partners in the presence or absence of the HDAC1 inhibitor, entinostat. From this screen, we identified Polypyrimidine Tract Binding Protein 1 (PTBP1) as a potential HDAC1-dependent interacting protein of RUNX1. PTBP1 is a heterogenous nuclear ribonucleoprotein that regulates pre-mRNA processing, mRNA metabolism and transport. PTBP1 is required for normal hematopoiesis, including stem cell maintenance, erythroid differentiation, and B-cell selection. However, little is known about the role of PTBP1 in leukemia.
We confirmed the PTBP1/RUNX1 association via immunoprecipitation and western blot in a mouse model of AML and in a panel of human AML (Molm-13, MV411, Me-1, Kasumi-1) and ALL (Jurkat, MOLT-4, REH, SEM) cell lines, indicating that this interaction occurs in multiple leukemia subtypes. Using proximity ligation assay (PLA) we found that this association occurred exclusively in the nucleus of both mouse and human leukemia cells. Using PLA, we also confirm that HDAC1 inhibition significantly reduced association of these two proteins in Molm-13 and REH cells.
Interestingly, we found that the RUNX1/PTBP1 interaction is significantly increased in mouse leukemic cells compared to lineage depleted (lin-) mouse bone marrow cells from healthy mice. In more stem/early progenitor enriched human CD34+ cells, we saw similar levels of PLA puncta as in leukemia cells. These results imply that the increased interaction between RUNX1 and PTBP1 observed in leukemia cells may be due to their arrest at an early stage of hematopoietic differentiation.
PTBP1 has been primarily studied for its role in mRNA splicing after transcription. However, the interaction with RUNX1 implies that PTBP1 also binds DNA. To test this, we performed CUT and Tag to determine the chromatin binding sites for PTBP1 and RUNX1 in Molm-13 cells. We observed extensive overlap of PTBP1 and RUNX1 binding sites, with PTBP1 present at almost 70% of RUNX1 sites. Using publicly available data we determined that both PTBP1 and RUNX1 bind H3K4me3 rich DNA regions indicating a strong bias towards transcriptionally active gene loci and regions associated with transcription start sites (TSSs). We observe lesser overlap with H3K27ac and H3K9ac, implying that this interaction is not very enriched at enhancers, and negligible binding in regions associated with the repressive histone mark, H3K27me3.
To address the role of PTBP1 in leukemia cells, we generated inducible PTBP1 shRNA knockdown (KD) Molm-13 and Kasumi-1 cells. We found no change in the survival of knockdown cells, but significantly increased expression of the differentiation marker CD15 and increased staining of the DNA damage marker gamma H2A.X. To determine if knockdown of PTBP1 induced changes in gene expression and splicing, we performed long-read RNA sequencing of control and PTBP1 KD Molm-13 cells. One of the most highly expressed genes with the loss of PTBP1 is the related splicing factor PTBP2, indicating a possible compensatory mechanism. We also found decreased expression of genes associated with glycolysis and pyruvate metabolism, implying that PTBP1 is required to maintain the increased metabolic rate characteristic of leukemia cells. Consistent with this finding, we found that PTBP1 KD cells showed decreased expression of the glucose importer, GLUT3. We also found that loss of PTBP1 induced changes in splicing, with the most common changes involving use of alternative transcriptional start sites (25%), termination sites (23%), and skipped exons (24%).
Collectively, our results demonstrate that PTBP1 binds RUNX1 in an HDAC1 dependent manner and that this interaction is present in the nucleus of both AML and ALL cells. We also show that PTBP1 co-localizes with RUNX1 at the promoters of transcriptionally active genes and regulates key metabolic genes. Collectively, our results suggest that PTBP1 cooperates with RUNX1 to promote expression and splicing of key leukemogenic targets.
No relevant conflicts of interest to declare.
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal