GSK3-mediated MAF phosphorylation in multiple myeloma as a potential therapeutic target.

Multiple myeloma (MM) is an incurable haematological malignancy characterised by the proliferation of mature antibody-secreting plasma B cells in the bone marrow. MM can arise from initiating translocations, of which the musculoaponeurotic fibrosarcoma (MAF) family is implicated in ∼5%. MMs bearing Maf translocations are of poor prognosis. These translocations are associated with elevated Maf expression, including c-MAF, MAFB and MAFA, and with t(14;16) and t(14;20) translocations, involving c-MAF and MAFB, respectively. c-MAF is also overexpressed in MM through MEK/ERK activation, bringing the number of MMs driven by the deregulation of a Maf gene close to 50%. Here we demonstrate that MAFB and c-MAF are phosphorylated by the Ser/Thr kinase GSK3 in human MM cell lines. We show that LiCl-induced GSK3 inhibition targets these phosphorylations and specifically decreases proliferation and colony formation of Maf-expressing MM cell lines. Interestingly, bortezomib induced stabilisation of Maf phosphorylation, an observation that could explain, at least partially, the low efficacy of bortezomib for patients carrying Maf translocations. Thus, GSK3 inhibition could represent a new therapeutic approach for these patients.

Cell context reveals a dual role for Maf in oncogenesis.

Maf b-Zip transcription factors are involved in both terminal differentiation and oncogenesis. To investigate this apparent contradiction, we used two different primary cell types and performed an extensive analysis of transformation parameters induced by Maf proteins. We show that MafA and c-Maf are potent oncogenes in chicken embryo fibroblasts, while MafB appears weaker. We also provide the first evidence that MafA can confer growth factor independence and promote cell division at low density. Moreover, using MafA as a model, we identified several parameters that are critical for Maf transforming activities. Indeed, MafA ability to induce anchorage-independent cell growth was sensitive to culture conditions. In addition, the transforming activity of MafA was dependent on its phosphorylation state, since mutation on Ser65 impaired its ability to induce growth at low density and anchorage-independent growth. We next examined transforming activity of large Maf proteins in embryonic neuroretina cells, where they are known to induce differentiation. Unlike v-Jun, MafA, MafB and c-Maf did not show oncogenic activity in these cells. Moreover, they counteracted transformation induced by constitutive activation of the Ras/Raf/MEK pathway. Taken together, our results show that Maf proteins could display antagonistic functions in oncogenesis depending on the cellular context, and support a dual role for Maf as both oncogenes and tumor suppressor-like proteins.