The therapeutic success of imatinib mesylate (IM) in chronic myeloid leukemia (CML) is impaired by persistence of malignant hematopoietic stem and progenitor cells (HSPC). The bone marrow microenvironment regulates the self-renewal, proliferation and differentiation of HSPC. We investigated the role of microenvironmental interactions in resistance of CML HSPC to elimination by BCR-ABL tyrosine kinase inhibitors (TKI). CML CD34+CD38− primitive progenitor cells and CD34+CD38+ committed progenitor cells were cultured for 96 hours with IM (5μM), nilotinib (5μM) and dasatinib(150nM), in medium supplemented with low concentrations of growth factors, with and without irradiated primary human marrow stromal cells (immortalized by ectopic telomerase expression) followed by an assessment of apoptosis and proliferation. Culture with stroma did not result in significant alteration of apoptosis in the absence of TKI treatment (3.1±0.7% apoptosis for primitive progenitors with stroma and 2.7±0.9% without stroma, 3.7±0.2% for committed progenitors with stroma and 4.7±2.1% without stroma). Coculture with stroma completely protected CML primitive and committed progenitors from TKI-induced apoptosis. CML CD34+CD38− cells demonstrated 20±6% apoptosis following culture with IM in the absence of stroma, but only 3.8±1% apoptosis in the presence of stroma (p=0.04, n=4). Similarly, apoptosis with nilotinib decreased from 12.5±1.8% without stroma to 2.9±0.3% with stroma (p=0.033), and apoptosis with dasatinib decreased from 7.1±0.04% without stroma to 2.7±0.2% with stroma (p=0.001). Apoptosis of CML CD34+CD38+ cells also significantly decreased following TKI treatment with 12.9±4.0%, 10.6±3.2%, 8.4±2.3% apoptosis observed after IM, nilotinib and dasatinib treatment respectively without stroma and 7.1±1.2%, 4.8±1.0%, 3.7±0.4% with stroma, (p=0.04, p=0.03 and p=0.02 respectively, n=4). Culture with stroma resulted in mild reduction in CML progenitor proliferation in the absence of TKI treatment, but TKI treatment resulted in similar degrees of inhibition of proliferation regardless of the presence of stroma. Culture of CML CD34+ cells in a Transwell insert with 0.45μm pores, allowing free diffusion of stromal factors but preventing direct contact with stroma, was associated with reduction in the protective effect of stroma coculture (32.2% apoptosis without stroma, 14.7% with stroma, and 24.6% with Transwell insert). Addition of blocking antibodies to a4 integrin and N-cadherin did not affect survival of CML CD34+ cells in the absence of IM, but resulted in enhanced apoptosis of CML CD34+ cells cocultured with stroma after addition of IM (20.4% apoptosis without antibody, 28.9% with anti-N-cadherin, and 29.8% with anti-integrin antibody). We conclude that the bone marrow stromal microenvironment protects CML primitive and committed progenitors from pro-apoptotic effects of BCR-ABL TKI treatment. Direct contact-mediated interactions, likely through VLA-4 and N-Cadherin, play an important role in protecting CML CD34+ cells from TKI-mediated apoptosis. These observations indicate that measures aimed at interfering with the protective effects of stroma could be of benefit for the eradication of residual malignant progenitors in CML patients receiving BCR-ABL TKI treatment.

Author notes

Disclosure: No relevant conflicts of interest to declare.