Trp53-Loss in Hematopoietic Stem Cells Drives the Evolutionary Process of Leukemic Transformation in a Jak2^V617F-Driven Myeloproliferative Neoplasms Mouse Model

Myeloproliferative neoplasms (MPNs) are clonal disorders resulting from genetic lesions in hematopoietic stem cells (HSCs). The development of post-MPN AML is treatment resistant and associated with very poor outcomes. This leukemic transformation is due to TP53 inactivation in approximately 50% of cases. Existing mouse models with double mutant Jak2^V617F and Trp53 have demonstrated that Trp53-lossalone is sufficient to induce leukemic transformation in Jak2^V617F-driven MPN. These models were generated either by crossing Jak2^V617F transgenic mice with Trp53^R172H knock-in or Trp53 knock-out mice, or by retroviral expression of Jak2^V617F in Trp53 knock-out bone morrow (BM) cells. These germline mutated models do not accurately model the sequential somatic Trp53 mutations that arise in Jak2^V617F MPN, or track their clonal expansion in vivo. Here, we used CRISPR-Cas9 to induce Trp53-loss in HSCs of Jak2^V617F mice to generate a novel mouse model of high molecular risk MPN that transforms to AML.

We isolated lineage ^-/low Sca1 ⁺c-Kit ⁺ (LSK) cells from BM of donor mice expressing a conditional allele of mutant Jak2 from its endogenous locus ( Jak2^fl-V617F) in combination with an inducible Cas9-IRES-GFP allele and CreER-fusion allele, both from the Rosa26 locus ( Rosa26^{CreER/lsl-Cas9-IRES-GFP}). LSKs were transduced with lentivirus expressing a sgRNA targeting exon 4 of Trp53 (sgTrp53), located in the DNA-binding domain of Trp53. These transfected sgTrp53 LSKs, along with control groups transduced with either empty vector (EV) or a sgRNA targeting Catsper1 (sgCatsper1, off target cutting control), were transplanted into lethally irradiated recipients. Successful engraftment and CreER activity, induced by a 2-week period of Tamoxifen chow, were confirmed by flow cytometry.

Validation of Trp53-editing was performed using amplicon sequencing, revealing the most common mutations were frameshift mutations. A subset of these frameshift mutations exhibited an extremely high frequency, conferring a selective advantage within the cell populations in sgTrp53 mice. We further examined the levels of p53 induced by nutlin-3a (an MDM2 inhibitor) in edited BM cells from three groups of mice and found that Trp53 editing leads to a reduction in p53 levels and prevents the loss of viability in response to MDM2 inhibition. This indicates that CRISPR-Cas9 mediated Trp53 mutations result in a loss of function.

SgTrp53 mice recapitulated the evolution of leukemic transformation of MPN, showing a PV phenotype for 8 weeks post CreER activation, characterized by elevated hematocrit but normal leukocytes and platelet counts, consistent with the phenotype exhibited by the EV and sgCatsper1 recipients. Subsequently, after 12 weeks, this MPN phenotype transformed to acute leukemia. This disease progression was associated with significantly shortened survival of sgTrp53 recipients, compared to the sgCatsper1 or EV control groups. Histopathologic analysis further supported these findings, showing leukocytosis and the presence of immature blast cells in the PB, spleen and BM of sgTrp53 mice, whereas such cells were absent in the control groups. Notably, the frequency of blast cells was correlated with the frequency of Trp53 frameshift mutations, consistent with what has been reported in MPN patients.

Additionally, CRISPR-Cas9 mediated Trp53-loss primarily affected the erythroid lineage, driving expansion of erythroblasts, megakaryocyte-erythroid progenitors (MEP), and Lineage ^-/low Sca1 ^-c-Kit ⁺ in sgTrp53 mice. The erythroblasts showed aberrant co-expression of CD41 and CD71. Through secondary transplants with unfractionated spleen cells, whole BM cells, or sorted MEP cell populations, we confirmed that the leukemia driven by CRISPR-cas9 mediated Trp53-loss in the Jak2^V617F-driven PV is transplantable, and that the MEP compartment in leukemic sgTrp53 mice contains the leukemia initiation population.

In total, we used CRISPR-Cas9 to induce Trp53-loss in HSCs of Jak2^V617F mice, faithfully replicating the process of clonal expansion and evolution observed in the leukemic transformation of MPN. These mouse models of secondary AML are valuable for testing new therapies targeting disease initiating stem cells. Moreover, our models represent the distinct stages of MPN progression, offering a platform to evaluate preventive and therapeutic interventions at various stages of this disease.