Myeloma-derived macrophage inhibitory factor regulates bone marrow stromal cell-derived IL-6 via c-MYC

Despite significant recent advancements made in the treatment of multiple myeloma (MM), relapse remains inevitable and the disease presently remains incurable. This is attributable, in part, to the highly protective nature of the BM micro-environment niche in which the malignant plasma cells proliferate. Macrophage migratory inhibitory factor (MIF) is a cytokine associated with various roles [1] and is rapidly developing a pro-tumoral identity [2]. Elevated MIF levels are described in MM and have been implicated in MM bone marrow homing and chemotherapy resistance [3]; however, the adaptive effect that MM-derived MIF has on the tumor microenvironment is not yet defined. Here, we investigate the function of MM-derived MIF in the MM microenvironment by examining its effects on bone marrow stromal cells (BMSC).

We and others have found that MM cells have significantly elevated MIF gene expression and secreted protein levels [3] (Additional file 1: Figure S1). MIF KD resulted in lower MM proliferation in BMSC co-culture (Additional file 1: Figure S2C) alongside reduced tumor burden and improved overall survival in vivo (Fig. 1a, b). To investigate the effects of MIF secretion by MM on its microenvironment, we used cytokine arrays to establish if cytokine changes occur when MM cells are cultured with primary BMSC. Elevated levels of IL-6/8 were detected in co-culture experiments when compared with either BMSC or MM monoculture arrays (Fig. 1c, d); no MIF was detected in BMSC cultured alone. Cytokine array analysis of the supernatant from MIF-stimulated BMSCs confirmed this IL-6/IL-8 secretion (Fig. 1e, f) and was quantified via ELISA (Additional file 2) in Fig. 1g. Pre-treatment of BMSC with the MIF inhibitor ISO-1 significantly reduced the MIF induction of IL-6/IL-8 by BMSC (Fig. 1h). BMSC expressed all three known receptors for MIF (CXCR4, CXCR2 and CD74); however, only blocking CD74 inhibited MIF-Induced IL6/8 (Additional file 1: Figure S3A&B).

An inhibitor panel was used to screen for potential pathways responsible for MIF-induced BMSC-derived IL-6/IL-8 expression, and we found that JQ1 inhibited MIF induced IL-6 and IL-8 (Fig. 2a, b) mRNA in BMSC. We then determined if JQ1 could regulate BMSC pro-tumoral interleukin production in-vivo. Although BMSC are often cyto-protective in the context of anti-MM therapy, others have observed that the sensitivity of MM cell lines to JQ1 was unchanged by the presence of HS-5 [4]. Following MM engraftment (Fig. 2d), mice were randomized and treated for 5 days with I.P injections of 50 mg/kg JQ1 or alternatively vehicle control. Under 7 days of treatment with JQ1 was predicted to have no measurable effect on tumor burden [4] and was selected to control for the effects of tumor burden on MM-MIF secretion and subsequent BMSC IL-6 expression. We found no difference in MM burden between groups (Fig. 2d, f). Nevertheless, murine IL-6 was significantly reduced in the JQ1-treated animals (carrying human MM) compared to control animals (Fig. 2e).

Despite MIF’s association with hematological malignancies, previous work has focused on the effects of MIF on the malignant cells rather than the supportive cells of the microenvironment [3]. Here, we show that MM-derived MIF is pro-tumoral through induction of BMSC-derived IL-6/8. IL-6 is central to MM pathogenesis and primarily comes from the BMSC in the tumor micro-environment [5]. Here, we place IL-6 downstream of MIF-induced BMSC activation [6, 7]. IL-8 expression in BMSC, which has been shown to parallel MM disease progression [8] and positively influence osteoclastogenesis in MM [9], was increased in BMSC in response to MIF. Furthermore, the BET-bromodomain inhibitor JQ1 significantly decreased IL-6/8 secretion in MIF-stimulated BMSC. In vivo use of JQ1 significantly reduced levels of murine IL-6 in the serum [5, 10, 11]. Taken together, this suggests that JQ1 is exerting anti-MM activity, in part, through a direct effect on BMSC via the inhibition of BMSC IL-6 (and IL-8) synthesis. This in turn could explain why BMSC do not appear to offer MM protection from JQ1 therapy.