P53 Pathway Activation Mediated High c-MAF Expression Is Associated with Overall and Post-Progression Survival in Multiple Myeloma

Background: Chromosomal abnormalities are strongly associated with prognosis of multiple myeloma (MM). Among them, t (14;16) resulting in high expression of c-MAF, and t (14;20) resulting in high expression of MAFB, lead to poor prognosis. However, clinical significance and mechanisms underlying high c-MAF and MAFB expression without these translocations have not been fully elucidated.

Methods: A total of 96 MM and 38 MGUS patients, 10 controls, and 9 MM cell lines were included in this study. RNA was extracted from purified CD138+ plasma cells from bone marrow (BM) mononuclear cells. c-MAF, MAFB, p53, and p21 RNA expressions were determined by RQ-PCR. Their expression levels were normalized against ACTB levels and calculated with 2^-ΔΔCt value. Inhibition studies using BRD4 inhibitor JQ1, and MYC inhibitor10058-F4, and MM cell lines expressing doxycycline-inducible p53 (Tet-on p53) or c-MAF knockdown (KD) with siRNA were used for in vitro studies.

Results: c-MAF expression in MM patients was higher than in MGUS patients and control (median level 0.043, 0.025, 0.002, p<0.001), but MAFB expression did not differ between MM, MGUS, and control (p=0.371). Although c-MAF expression level was significantly higher in MM with t (14;16) (p=0.0018), high c-MAF expression was observed in patients without the translocations, suggesting a role of c-MAF in MM progression. Other cytogenetic abnormalities such as del 17p, or t (4;14), did not affect c-MAF expression. Overall survival (OS) of MM patients with high c-MAF expression was significantly inferior compared to the patients with low c-MAF expression (HR 2.46, p=0.002, 2 years survival rate 42.9% vs. 72.7%, median 20.3 months vs. not reached), but progression-free survival (PFS) did not differ (p=0.551). Instead, post-progression survival (PPS) was significantly shorter in the patients with high c-MAF expression (HR 4.67, p<0.001, two years survival 0% vs. 60.2%, median 3.6 months vs. 49.3 months) suggesting that c-MAF confers drug resistance to residual disease. MAFB expression did not affect OS, PFS, and PPS. Reduction of cyclin D2 expression by c-MAF KD in vitro and a positive correlation between c-MAF and cyclin D2 expression (r=0.29, p<0.001) in vivo in patients, support previous reports describing that MM proliferation is induced by c-MAF via cyclin D2. c-MAF was not correlated with APOBEC3B expression. c-MAF expression was positively correlated with MDM2 and MYC (r=0.387, p=0.03 and r=0.221, p=0.019, respectively) but not with p53 and PUMA. A tendency of positive correlation was also observed with p21/CDK1NA (r=0.304, p=0.085). p53 overexpression from the Tet-on p53 and p53 stabilization by nutlin-3 increased c-MAF expression in three cell lines with t (14;16). c-MAF KD together with p53 overexpression significantly suppressed the proliferation, implying synthetic lethality. These findings suggest that c-MAF expression is upregulated in response to proliferation arrest induced by the p53 pathway; suppressing this responsive c-MAF expression abrogates cell growth more efficiently. JQ1 and CPI203 increased c-MAF expression in cell lines with t (14;16), suggesting that c-MAF expression is not controlled by super-enhancers.

Conclusion: Although the mechanism of high expression of c-MAF and MAFB has previously been reported in IgH translocation, the relationship between c-MAF and drug resistance remains to be determined. The induction of c-MAF expression by tumor suppressor gene p53 is suggested to be a mechanism underlying poor prognosis of MM with t (14;16), and suppressing c-MAF expression with chemotherapeutic drugs might prevent emergence of drug resistance in residual tumor cells.