In the chronic phase of chronic myeloid leukemia (CML-CP), leukemic stem cells do not necessarily depend on the BCR-ABL tyrosine kinase activity for their growth and survival and thus resistant to tyrosine kinase inhibitors (TKIs). In this study, we aimed to identify the initial progenitor population that is getting switched on BCR-ABL growth signaling and tried to elucidate the underlying molecular mechanisms of BCR-ABL dependent cell growth. We thus intensively analyzed the involvement status of CML clones in each developmental stage at diagnosis. To identify the hematopoietic stem or progenitor cell stage that is responsible for CML clone expansion, bone marrow cells from 13 newly-diagnosed CML-CP patients were analyzed by FACS, and purified stem/progenitor populations were tested for the t(9;22) involvement by FISH. Gene expression signature of each purified population was also evaluated by cDNA microarray.
CD34+CD38- HSC fraction was markedly diminished in all CML-CP patients compared to healthy volunteers (<1% of CD34+ cells in patients vs. ∼10% in volunteers), whereas CD34+CD38+ myeloid progenitors expanded. Interestingly, the t(9;22) positive ratios in the HSC fraction were greatly diversified among patients (
Figure 1). Of note, in 4 patients, the involvement of t(9;22) positive clone was less than 10%, suggesting that their leukemic stem cells have not outgrown normal HSCs. Among CD34+CD38+ myeloid progenitors, common myeloid progenitors (CMPs) robustly expanded and were composed more than 90% of t(9;22) positive clone in all cases. Downstream of CMPs, megakaryocyte/erythrocyte progenitors (MEPs) but not granulocyte/macrophage progenitors (GMPs) were dominantly involved in leukemia (t(9;22) positive ratio; 96.0+/-5.5% vs. 56.3+/-37%). The expression level of BCR-ABL is not different among these progenitor populations. These observations collectively suggest that in CMP-CP, BCR-ABL signaling becomes effective on cell proliferation especially at the CMP stage.
Gene expression analysis of stem/progenitor populations in CML patients revealed that IRF8 and GFI1, transcription factors playing critical roles in myeloid differentiation and cell proliferation, were down-regulated specifically in CMPs as compared with that in normal controls (
Figure 2). In order to substantiate the role of IRF8 and GFI1 in CML pathogenesis, we used a CML mouse model established by enforced retroviral expression of BCR-ABL. As in analysis of CML patients, BCR-ABL expressing CMPs but not stem/multipotent progenitor cells acquired growth advantage over normal counterparts. Importantly, the expression of IRF8 and GFI1 became undetectable after BCR-ABL transduction in the expanding CMPs.
Our observations revealed that, in CML-CP hematopoiesis, BCR-ABL dependent cell proliferation initiates at the CMP stage, and is accompanied with the down–regulation of IRF8 and GFI1. Because IRF8 knockout mice develop myeloproliferative disorders, and because CMPs expand in GFI1 null mice, the attenuation of these molecules could be downstream effector of BCR-ABL dependent myeloid cell growth. Taken together, the reactivation of these molecules might be useful to develop alternative therapeutic strategies for CML-CP, for example, with TKI-resistant BCR-ABL mutants.
Disclosures:
Miyamoto:Kyushu University Hospital: Employment.
