Multilineage Expansion of Lymphoid and Myeloid Cells in Mice Expressing PML-RARα under the Control of the Murine Cathepsin G Locus.

In one of our laboratory’s models of acute promyelocytic leukemia (APL), the PML-RARα fusion cDNA is knocked into the 5′ UT of the azurophil granule protease, cathepsin G. Nearly all mCG-PML-RARα mice develop lethal leukemia with promyelocytic features following a 150–400 day latent period. We originally chose the cathepsin G gene for the targeting locus because its expression was believed to be restricted to promyelocytes. However, we have now observed high levels of cathepsin G expression with gene expression profiling of leukemic bone marrow samples from patients with AML FAB M1 or M2, i.e. blasts with minimal or no differentiation (M1: n=21, mean raw expression value for cathepsin G=23,885 ± 31,737; M2: n=22, mean=24,088 ± 26,585; M3: n=14, mean 116,029 ± 47,017). To examine whether cathepsin G is normally expressed in murine hematopoietic progenitor cells, we performed gene expression arrays with highly purified Sca-1⁺/lin⁻ bone marrow cells; cathepsin G mRNA was easily detected in these cells (n=4, mean cathepsin G raw expression value=9,324 ± 5,082; array average normalized to 1,500). We therefore decided to determine whether the early progenitor compartment in mCG-PML-RARα mice was expanded due to unexpected expression of the transgene in KLS (c-kit⁺, Lin⁻, Sca-1⁺) cells; however, no difference was detected in the frequency of marrow-derived KLS cells between mCG-PML-RARα and WT mice that were age, strain, and gender matched (0.074% vs 0.071%, p=0.89). The GMP compartment (Lin⁻, c-kit⁺, Sca1⁻, CD34⁺, FcRγ⁺) showed a tendency towards expansion in mCG-PML-RARα mice (0.10% vs 0.032%) but the difference was not significant (p=0.067). To better assess stem cell function in these mice, we performed a competitive repopulation study using marrow derived from Ly 5.2/mCG-PML-RARα mice mixed with Ly5.1/WT marrow cells at various ratios (1:1, 9:1, and 1:9) that were transplanted into genetically compatible hosts. Lineage markers in the peripheral blood were tested at 6, 12, and 24 weeks to assess the contribution of mCG-PML-RARα-derived progenitor and stem cells to the B, T, and myeloid lineages. As expected, we observed a highly reproducible increase in the proportion of Gr-1⁺ cells derived from mCG-PML-RARα donors at all time points (78.8% ± 11% donor-derived cells at a 1:1 donor: recipient ratio at 6 months, p=0.0006). Surprisingly, we also observed a reproducible increase in B220⁺/CD19⁺ cells (66.7% ± 4%, p<0.0001) and CD3⁺ cells from the mCG-PML-RARα donors (60.4% ± 3% p=0.0010) at 6 months, suggesting that mCG-PML-RARα also confers a growth or survival advantage to B and T cells. To determine whether cathepsin G was expressed in these compartments, we purified CD19⁺ B and CD3⁺ T cells by flow cytometry (>99% purity) and did not detect cathepsin G mRNA using a sensitive, knockout-proven qRT-PCR analysis. These data strongly suggest that both the WT cathepsin G gene and the mutant mCG-PML-RARα allele are unexpectedly expressed in the stem cell compartment, with both myeloid and lymphoid progeny of these cells displaying a growth advantage in vivo. However, lymphoid malignancies have not been detected in these mice (n>400), suggesting that PML-RARα is unable to initiate transformation in lymphocytes. Our data imply that the PML-RARα gene is activated in a multipotent compartment in this mouse model, raising the possibility that PML-RARα may not actually ‘reprogram’ progenitor cells to undergo self-renewal, but may rather initiate transformation in pluripotent cells with intrinsic self-renewal capabilities.