Targeting Plasticity and Stemness in Multiple Myeloma Tumor Cells

Background: While there are many effective therapies that can induce remission, almost all Multiple Myeloma (MM) patients relapse. This has led to the hypothesis that a chemotherapy-resistant cancer stem cell (CSC) remains dormant and then undergoes self-renewal and differentiation to reestablish disease. To eliminate disease, novel therapies must be designed to preferentially target this CSC, either to maintain it in its dormant state, or alternatively to drive it into cycle where it will become sensitive to more traditional therapies. The MM CSC can be detected in cultured MM tissue culture models, including RPMI-8266 cells, and is defined by several parameters: 1) the lack of the plasma cell marker CD138, 2) lack of the differentiation markers Blimp1 and IRF4, 3) self-renewal as measured by the ability to form colonies in methylcellulose, 4) clinically relevant chemoresistance to Bortezomib and Lenalidomide, and 5) a G0 cell cycle state. Our finding is that the mature cancer cell (CD138+, Blimp+, IRF4+, rapidly proliferating and chemosensitive) has developmental plasticity; namely, the ability to dedifferentiate back into its own chemoresistant CSC progenitor, the CD138^-, Blimp-, IRF4-, quiescent pre-plasma cell. In our 2D culture system, we observe multiple cycles of sequential differentiation and dedifferentiation in the absence of niche or supportive accessory cells, suggesting that soluble cytokines secreted by the MM cells themselves are responsible for this bidirectional interconversion and that stemness and more importantly chemoresistance are dynamic characteristics that can be acquired or lost and thus are targetable.

Methods: Flow cytometric analysis of RPMI 8226, U266 and NIH-H929 demonstrates that 5-10% of these stable cells are consistently CD138-, and thus are a mixture of mature and precursor cells. Sorted >98% pure CD138- and CD138+ populations were cultured in tissue culture media, and then harvested at various times post plating and screened for CD138. Within 5 days post plating, the immature CD138- cells differentiated to produce progeny that were CD138+. When the mature CD138+ cells were plated post sorting, we found that within 5 days, a large population of CD138- cells appeared as well, suggesting that the mature CD138+ tumor cells had the capacity to dedifferentiate back into the more immature CD138- cells. While it has been shown that MM cells and other cell types are able to differentiate in the presence of a niche or accessory cells, our data suggest that the presence or absence of soluble factors secreted from the cells themselves alters their differentiation and dormancy status: the cell population has the plasticity required to maintain enough progenitor cells to provide a continuous supply of daughter cells. Using an array to examine cytokine secretion from sorted, 99% pure populations of either CD138-CSCs or CD138+ progeny from the RPMI-8226 line we identified that concomitant with interconversion, Macrophage Migration Inhibitory Factor (MIF-1) was secreted into the media. To determine if MIF-1 was responsible for interconversion, we treated pure populations of CD138+ cells with 4-IPP, a small molecule MIF-1 inhibitor. We saw that with increasing concentrations of inhibitor, the CD138+ cells rapidly accelerated dedifferentiation back into the CD138- progenitor and trapped these cells reversibly in the dormant, chemoresistant state. When we treated CD138- cells with 4-IPP, they remained CD138- and did not differentiate into CD138+ cells. This suggests that blocking MIF-1 forced and/or maintained cells in the chemoresistant CD138- state.

Conclusion: MIF-1 is an inflammatory cytokine, secreted by many cell types, and shown to increase survival and proliferation of several stem cell lineages. Our data suggest that MIF1 regulates interconversion and its levels, produced from either the MM cells themselves or in vivo from the bone marrow niche, regulate the balance between chemoresistant CSCs and chemosensitive progeny. Thus, targeting MIF-1 may be a viable way to alter chemosensitivity.