Timing of the Loss of Pten Is Critical in Determining the Disease Phenotype in Mice- a Mouse Model for Pediatric Mixed MDS/MPN

Cancer is caused by accumulated genomic and epigenetic abnormalities during the development of an individual, particularly during the neonatal period, when developmental plasticity is actively occurring. Myeloid-specific deletion of pten in embryos or after 3 weeks of age causes acute monocytic or myeloid leukemia (AML) or acute lymphoblastic leukemia (ALL) following a transient myeloproliferative neoplasm (MPN) in adult mice, which can mimic the human diseases to varying degrees. However, it is not clear how the timing of genomic and epigenetic abnormalities contributes to the disease phenotype in a mouse that is of an age comparable to human children. We hypothesized that during the development/aging process, the timing of when the genomic abnormality or “hit” occurs, such as loss of Pten or Nf1, is a critical determinant of the disease phenotype. We tested this by investigating the effect of somatic deletion of Pten at an age of 8 days, one of the most vulnerable stages for malignancy development in mice with or without a germline mutant Nf1. Through crossbreeding, we generated mice with Pten^fl/flNf1^Fcr/+Mx1-Cre⁺ on a C57BL6/129 genetic background, and conditionally deleted Pten in a myeloid-specific manner by intraperitoneal injection of Poly(I:C). Mice with a pten deletion and mutant Nf1 (pten^koNf1^mut, hereafter referred as double mutant) showed signs of sickness at the end of the 2^nd week of life, and all died by age 3-5 weeks (equivalent to 1-3 years old in humans). The natural survival in double mutant mice (n=10) was significantly shorter than those with wild type pten and Nf1 (pten^wt; Nf1^wt, hereafter referred as WT, n=6, median 0.9 vs >14 months, p<0.001). At autopsy, the double mutant mice had obvious hepatosplenomegaly compared to the WT littermates, but no lymphadenopathy nor thymus enlargement was detected. We then analyzed 15 littermate pairs with/without mutations at age 3-4 weeks. The spleen and liver sizes in the double mutants were significantly increased compared to that in WT (median weights: spleen-198 vs 69mg, p< 0.001; liver-836 vs 532mg, p=0.013). Peripheral blood (PB) analysis showed no difference in WBC in both groups, but significant decreases in double mutant mice in RBC and Hgb (median RBC-5.6 vs 7.6x10⁶/ul, p=0.010; Hgb-10.1 vs 12.6g/ul, p=0.022), but with a normal MCV. HE and IHC stained tissue sections of formalin-fixed organs with antibodies to B220, CD3e, or F4/80 found increased cellularity in bone marrow (BM), and severe infiltration of mature myeloid cells in double mutant spleens, livers, and lungs. Flow cytometric analysis showed that double mutant mice had significantly increased cells marking for monocyte/macrophages(Mac-1⁺Gr-1^int, or Mac-1⁺CD115⁺) and granulocytes (Mac-1⁺ Gr-1⁺)[p≤0.001] in BM and PB; both T-cells (CD3e⁺) and B-cells (CD19) were significantly reduced in PB and spleen (p<0.01); platelets (CD41⁺Ter119^-) were reduced in PB. An erythroid differentiation block [indicative of myelodysplasia (MDS) or leukemia] was found in double mutant mice (significant increases in Ter119⁺CD71^high cells). Pten deleted mice without mutant Nf1had a similar phenotype, but survived longer than double mutant mice (median 1.2 vs 0.9 months, p<0.001). Colony formation assays showed no cytokine hypersensitivity in BM or spleen cells in double mutants in response to GM-CSF or IL-3. Based on the Bethesda proposals for classification of non-lymphoid hematopoietic neoplasms in mice, we believe that we have developed a mouse model of myeloproliferative disease with features of a pediatric unclassifiable mixed MDS/MPN. It also mimics to some degree human Juvenile Myelomonocytic Leukemia (JMML), in terms of age of onset and organ infiltration with myelomonocytic cells, but fails to mimic JMML with respect to GM-CSF hypersensitivity.

We demonstrate that deletion of pten at an early age in the mouse causes a fatal mixed MDS/MPN that is different from the previously reported mouse models, in which myeloid-specific deletions of Pten were induced in embryos or after 3 weeks of age, and these mice developed only a transient MPN, all eventually dying of AML or ALL at adult ages. To the best of our knowledge, this is the first model in young mice bearing a somatic mutation that resembles a human pediatric unclassifiable MDS/MPN and more closely represents the true genetic and epigenetic background, thus providing a novel tool to study these mechanisms underlying pediatric MDS/MPN.