Cytotoxic Exposure to Both Hematopoietic Stem and Progenitor Cells and the Bone Marrow Niche Cooperate to Promote Leukemogenesis in a Mouse Model of Therapy-Related Myeloid Neoplasms with a Del(5q)

A del(5q) is frequently noted in MDS, AML, and therapy-related myeloid neoplasms (t-MN) following alkylating agent therapy. Mutation/loss of the TP53 is gene found in 80% of t-MNs with a del(5q). Recent studies suggest that TP53 mutations are found at a low frequency in hematopoietic stem/progenitor cells (HSPCs) in adults (Xie et al., Nat Med 20:1472, 2014; Genovese et al., NEJM 371:2477, 2014; Jaiswal et al., NEJM 371:2488, 2014), and chemotherapy confers a selective growth advantage to these rare clones (Wong et al., Nature 518:552, 2015). We previously established a mouse model for t-MN with a del(5q) and showed that haploinsufficiency of two del(5q) genes, Egr1 and Apc, cooperate with loss of function of the Trp53 (p53) gene to induce myeloid neoplasms in mice. Specifically, transplantation of Egr1^+/-, Apc^del/+ bone marrow (BM) cells transduced with p53 shRNAs into wild type (WT) recipients resulted in the development of a transplantable AML, characterized by a complex karyotype and genetic instability, in 17% of mice.

There is growing evidence that microenvironment perturbations play a major role in the malignant process; however, the effect of cytotoxic therapy on HSPCs as well as the BM niche is not well understood. Using our mouse model of t-MN with a del(5q), we explored the effects of ENU, an alkylating agent, on both HSPCs and the BM microenvironment by exposing both donor and recipient mice to ENU (Panel 1 in the figure). In mice transplanted with Egr1^+/-, Apc^del/+, p53 shRNA HSPCs, exposure to ENU strikingly decreased survival (median survival: 200d vs. not reached) and increased the incidence of AML or MDS with multilineage dysplasia (73% vs. 17%). In the absence of p53 knockdown (i.e., control shRNA), mice survived longer (370d vs. 200d, P = 0.0014); however, 100% of mice developed MDS with dyserythropoiesis. None developed AML, suggesting that loss of p53 function is critical for leukemic transformation (Panel 2). Loss of both del(5q) genes, EGR1 and APC, was necessary to develop AML. Compared to mice transplanted with Egr1^+/-, Apc^del/+, p53 shRNA HSPCs, mice transplanted with Egr1^+/-, p53 shRNA HSPCs survived longer (369 d vs. 200 d, P = 0.0117) and only 40% of mice developed MDS with dysgranulopoiesis and/or dyserythropoiesis (Panel 3). None developed AML. Thus, severity of disease increases with loss of more than one del(5q) gene. Finally, to determine the separate effects of alklating agent therapy on HSPCs vs. the niche, we treated either the recipient or donor mice with ENU. Whereas ENU exposure to both donor and recipient resulted in a profound expansion of p53 shRNA⁺ cells and the development of MDS/AML in 73% of mice, ENU exposure of either donor or recipient led to only modest expansion of p53 shRNA⁺ cells and none of the mice developed MDS or AML. This suggests that the clonal expansion of cells with loss of multiple 5q genes and p53 is likely promoted by cytotoxic exposure to the cells themselves, as well as exposure to the surrounding niche cells.  

t-MN patients with a del(5q) typically present with trilineage dysplasia implicating all three hematopoietic cell lineages (erythroid, myeloid, and megakaryocytic) in the dysplastic process. Our mouse models shed light on some of the key genes on 5q, as well as the environmental exposure, that contributes to trilineage dysplasia in patients. Finally, our data suggests that t-MN is a "disease of the tissue", and the expansion of mutant HSPCs, e.g., with TP53 mutations, likely results from the combined effects of cytotoxic therapy on the hematopoietic cells themselves, as well as the BM microenvironment that supports hematopoiesis