PML-RARα Deregulates an Unexpectedly Small Number of Genes in Pre-Leukemic Promyelocytes

Acute Promyelocytic Leukemia (APL) is characterized by the accumulation in the blood and bone marrow of abnormal promyelocytes, which have the ability to transfer the disease to secondary recipients in animal models. The PML-RARα fusion protein is thought to be the primary abnormality implicated in the pathology, and is believed to prevent transcription of genes necessary for normal myeloid development and differentiation. Identifying PML-RARα targets is critical for understanding the road to leukemic transformation. However, such targets have so far been identified using cell line assays in vitro, murine cells differentiated into promyelocytes in vitro, or fully transformed murine or human leukemic cells. Focusing on the cell population in which the transforming potential is acquired, we describe here a novel strategy to identify the transcriptomic dysregulation induced by PML-RARα expression in maturing myeloid populations in vivo.

We utilize a murine model of human APL in which the human PML-RARα fusion gene is expressed under the control of the MRP8 promoter, driving its expression in maturing myeloid populations. Those animals can be described as pre-leukemic since they eventually develop leukemia when additional mutations occur. Fresh bone marrows from normal (Fvb/n) or pre-leukemic (PML-RARα) animals were harvested. Using an improved cell surface antigen staining strategy and fluorescence-activated cell sorting, three populations of increasingly differentiated myeloid populations have been sorted (Granulocyte Macrophage Progenitor, Early promyelocyte and Late promyelocyte). RNA was extracted and submitted for whole-genome microarray analysis. In addition, we are using a variety of bioinformatics approaches to decipher the network of novel interactions driven by PML-RARα expression.

Markers used in our sorting strategy were validated in the dataset, including CD34 and Gr1. In the normal samples, markers of neutrophil maturation increased, largely as expected, and a number of early transcription factors decreased in an expected manner including Hoxa9 and Meis1. One remarkable finding was that despite the previously described ability of PML-RARα to regulate transcription from multiple sites in the genome, only a small number of genes were differentially impacted by the expression of this protein. Surprisingly, well-known regulators of myeloid differentiation that have been implicated in the retinoic acid responsiveness of APL including Sfpi1 (PU.1) and Cebpa were not differentially expressed. However, in pre-leukemic samples PML-RARα did cause decreased expression of multiple neutrophilic granule genes including Ltf, Mmp9 and Ngp. The gene most upregulated in the pre-leukemic samples was Spp1 which encodes the osteopontin phosphoprotein. Of interest, we identified the myeloid tumor suppressor Irf8 to be downregulated 5 fold in the presence of PML-RARα. To investigate the importance of IRF8 levels in APL initiation, we transplanted Irf8+/+ PML-RARα or Irf8+/− PML-RARα bone marrow into irradiated recipients. Despite the potential for decreased expression of IRF8 to contribute to APL, we observed no difference. This result does not confirm a role for IRF8 in APL pathogenesis, but further investigations are needed to exclude such a role.

Bioinformatics studies highlighted enrichment in cell cycle-related genes upon PML-RARα expression, suggesting a possible difference in the proliferation capacity of the pre-leukemic cells, which is currently under investigation.

We found that in vivo the transcriptome was only modestly dysregulated by the presence of PML-RARα. These observations open up new questions on the role of the fusion protein in pathogenesis: How does PML-RARα prime pre-leukemic cells for full transformation? How do secondary events allow an initiated cell to advance to a fully transformed state? Such questions are currently being investigated, with a special interest on looking at the cooperation between PML-RARα and activated cytokine signaling in leukemia initiation.

No relevant conflicts of interest to declare.