The Polymerase Associated Factor 1 complex (PAF1c) functions at the interface of epigenetics and gene transcription. The PAF1c is a multi-protein complex composed of Paf1, Cdc73, Leo1, Ctr9, Rtf1 and WDR61, which have all been shown to play a role in disease progression and different types of cancer. Previous reports demonstrated that the PAF1c is required for MLL-fusion driven acute myeloid leukemia. This is due, in part, to a direct interaction between the PAF1c and wild type MLL or MLL fusion proteins. Importantly, targeted disruption of the PAF1c-MLL interaction impairs the growth of MLL-fusion leukemic cells but is tolerated by normal hematopoietic stem cells. These data point to differential functions for the PAF1c in normal and malignant hematopoietic cells that may be exploited for therapeutic purposes. However, a detailed exploration of the PAF1c in normal hematopoiesis is currently lacking.
Here, we utilize a mouse genetic model to interrogate the role of the PAF1c subunit, Cdc73, in the development and sustenance of normal hematopoiesis. Using hematopoietic-specific constitutive and conditional drivers to express Cre recombinase, we efficiently excise floxed alleles of Cdc73 in hematopoietic cells. VavCre mediated excision of Cdc73 results in embryonic lethality due to hematopoietic failure. Characterization of the hematopoietic system demonstrated that cKit+ hematopoietic stem and progenitor cells (HSPC) are depleted due to Cdc73 knockout.
We next investigated the role of Cdc73 in adult hematopoiesis using Mx1Cre mediated excision. Conditional knockout of Cdc73 in the adult hematopoietic system leads to lethality within 15 days of Cdc73 excision while no phenotype was observed in heterozygous Cdc73fl/wt controls. Pathological examination of bones in these mice showed extensive bone marrow failure. Flow cytometry analysis revealed that cKit+ HSPCs in adult mice are ablated following loss of Cdc73. Bone marrow transplantation assays demonstrated a cell autonomous requirement of Cdc73 for HSC function in vivo.
To perform cellular characterization of HSPCs upon Cdc73 KO, we optimized excision conditions to capture cKit+ HSPCs with excised Cdc73 but before their exhaustion. Flow cytometry analysis demonstrated that Cdc73 KO leads to a cell cycle defect. Cdc73 excision leads to a 2.5 fold increase in the accumulation of HSPCs in the G0 phase of cell cycle with a reduction in the proliferative phases. This is accompanied with an increase in cellular death as indicated by Annexin V staining. Together, these data indicate that Cdc73 is required for cell cycle progression and HSPC survival.
To understand the molecular function of Cdc73, we performed RNAseq analysis to identify genes regulated by Cdc73 in HSPCs. We observed 390 genes are upregulated and 433 genes are downregulated upon loss of Cdc73. Specifically, Cdc73 excision results in upregulation of cell cycle inhibitor genes such as p21 and p57, consistent with the cell cycle defect observed following Cdc73 excision.
Further, when comparing our results to leukemic cells, we uncovered key differences in Cdc73 gene program regulation between ckit+ hematopoietic cells and MLL-AF9 AML cells. Loss of Cdc73 in leukemic cells leads to downregulation of genes associated with early hematopoietic progenitors and upregulation of myeloid differentiation genes consistent with previous studies. Interestingly, we observed a more even distribution of expression changes (non-directional) within these gene programs following Cdc73 inactivation in HSPCs. Most importantly, while loss of Cdc73 in MLL-AF9 AML cells leads to a profound downregulation of the Hoxa9/Meis1 gene program, excision of Cdc73 in HSPCs results in a modest non-directional change in expression of the Hoxa9/Meis1 gene program. This was attributed to no change in Hoxa9 and Meis1 expression in HSPCs following excision of Cdc73, in contrast to MLL-AF9 cells where these pro leukemic targets are significantly downregulated. Together, these data indicate an essential role for the PAF1c subunit Cdc73 in normal hematopoiesis but differential roles and context specific functions in normal and malignant hematopoiesis, which may be of therapeutic value for patients with AMLs expressing Hoxa9/Meis1 gene programs.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.