Mutation or altered expression of key transcription factors resulting in aberrant myeloid differentiation is a critical step in the pathogenesis of acute myeloid leukemia (AML). The ETS-domain hematopoietic transcription factor PU.1 (SPI1) is an essential regulator of myeloid differentiation. While genetic mutation of PU.1 is rare, PU.1 is down-regulated or functionally repressed in about 50% of AML patients as a consequence of recurrent translocations (such as AML1-ETO and PML-RARα) or mutations (such as FLT3-ITD). While it is well-established that the inhibition of PU.1 function promotes AML development, our understanding of the downstream targets mediating its tumor-suppressor role is still incomplete.

In order to study PU.1-driven differentiation in AML, our laboratory has recently developed a mouse model of AML driven by doxycycline (Dox)-regulated knockdown of the PU.1 gene (shPU.1) on a p53-deficient background. This is an aggressive AML model that mimics high risk disease in the clinic, as loss of function mutations of p53 characterize a subset of AML patients with particularly poor prognosis. In this system, Dox treatment shuts off a GFP-linked shRNA specifically targeting PU.1. Restoration of PU.1 expression in the context of established leukemias in vivo, and in culture-adapted cell lines derived from them, triggered myeloid differentiation and induction of cell death. In order to study the dynamic transcriptional changes underpinning these effects, we performed genome-wide transcriptome analysis on shPU.1/p53-/- AML cells isolated from the bone marrow of multiple leukemic mice transplanted with 2 different primary AML (246 and 410) that were either untreated or Dox-treated for 2, 4, or 6 days. Dox-induced restoration of PU.1 expression led to dramatic transcriptional changes closely resembling those associated with the maturation of normal mouse granulocyte/macrophage progenitors (GMPs) into neutrophils. Among the differentially expressed genes at day 2 of Dox-treatment we identified 12 "early response" genes commonly induced in both 246 and 410 AML. This group of genes included the macrophage colony-stimulating factor receptor (Csf1r), a well-known target of PU.1, as well as other genes previously implicated in myeloid differentiation, cytokine signalling, survival and cancer. In order to explore the contribution of these "early-response" genes to the range of anti-leukemic effects exerted by PU.1 restoration, we have generated retroviral shRNA vectors to test whether specifically silencing each of these genes alters PU.1-induced differentiation, inhibition of proliferation and/or death of p53-/- AML cells in vitro.

In conclusion, in this work we have used a novel AML mouse model driven by reversible PU.1 inhibition to identify the PU.1-regulated transcriptome in AML, and to screen for key PU.1-response genes mediating its pro-differentiation and anti-leukemic effects in the context of this disease.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.