Multiple myeloma (MM) is a hematological malignancy that is characterized by clonal proliferation of transformed plasma cells within the bone marrow (BM) and severe bone disease. Although MM cells are initially sensitive to many therapies, patients eventually relapse with refractory disease. Because MM cells show a dependency on the BM microenvironment for survival and proliferation, and BM adipocytes (BMAs) demonstrate a unique, endocrine signaling capacity and lipid composition, it is likely that there is cross-talk between MM cells and BMAs that leads to tumor support. Unlike the influences of osteoblasts and osteoclasts, the effect of BMAs on MM cells is poorly understood and few studies have investigated the relationship between marrow adipose tissue (MAT) and cancer. We predict there to be a strong link between MAT and MM, as obesity is a risk factor for MM and also correlates with increased MAT in humans and, as we have shown, in mice (Bornstein S et al. in press, Endocrinology). Moreover, MAT has been shown to support other tumors and correlates with bone diseases such as osteoporosis, anorexia, and aging. Thus, we explored here the novel hypothesis that BMAs induce drug resistance in MM cells through direct cell-cell contact signaling and/or secreted signaling molecules.

BMAs are distinct from white adipocytes, as we reviewed (McDonald M et al. 2016, CTI), but it is very difficult to culture BMAs, so we generated BMAs in vitro from BM- mesenchymal stem cells (MSCs) from mice (to create mMAT) or humans (to create hMAT) using adipogenic differentiation media. We directly or indirectly cultured these with luciferase-labeled MM cells, with or without anti-myeloma drugs (dexamethasone (dex), bortezomib (bort)), or adiponectin to assess cell adhesion, proliferation and drug resistance by bioluminescence imaging (BLI), cell counting, MTT and apoptosis assays. Adiponectin-KO and WT mMAT were compared as well. To enhance our findings, we developed a novel, tissue-engineered (TE) 3-D in vitro MAT model using silk scaffolds, formulated and imaged similarly to our publication (Abbott R et al. 2016, Adv Healthc Mat). These were seeded with hMSC or mMSCs, grown in adipogenic media, and validated to represent normal, healthy MAT using confocal imaging and a 22,000 gene microarray. MM cells (OPM2, 5TGM1, and MM1S) were cultured on TE-MAT for 2 weeks and examined with confocal microscopy for adhesion, proliferation, drug resistance.

We observed significant adhesion of MM cells to hMAT in 2D and in 3D TE-MAT cultures. In contrast to prior reports, we found that adiponectin does not induce MM apoptosis and that MAT does not typically affect MM proliferation. Interestingly, hMAT and mMAT CM and transwell membrane culture rescued MM1S cells from dex-induced apoptosis (7AAD/Annexin V flow cytometry (p<0.01) and BLI (p<0.01)). Moreover, OPM2 cells were significantly resistant to bort (p=0.0236, OPM+bort vs OPM2+hMAT+bort) and dex (p=0.021, OPM2+dex vs OPM+hMAT+dex) (Figure 1). OPM2 and MM1S cells also displayed a slight, but significant (p<0.05) resistance to bort when grown in indirect culture with 3T3L1 adipocytes. Interestingly, we observed "adipomimicry" in OPM2 cells when co-cultured with hBMAT (3 donors), as measured by increases in FABP4 gene expression. The 3D TE-MAT model provided a more realistic microenvironment with a less inflammatory profile than 2D MAT, and confirmed drug resistance in OPM2 in response to dex with hMAT (p<0.05).

Overall, MAT induced drug resistance in a subset of MM cells through mechanisms we are now fully elucidating. Our work may help identify biomarkers for patients at risk for MAT-driven resistance. Interestingly, we have recently published that anti-sclerostin Abs can prevent MM bone disease (McDonald M et al 2017, Blood), and also decrease BMA size and number (Fairfield H et al. 2017, JCP). Thus, the BMA is a promising, targetable cell that could provide a new avenue for attacking tumor-stromal drug resistance. Our data indicate that both cell-cell adhesion and soluble mediators, such as lipids, FABP4 and other adipokines, can induce MM drug resistance. Furthermore, targeting BMAs may reduce MM bone disease and improve bone health, as we and others have shown that MAT negatively affects bone strength. Thus, our findings highlight the potential clinical application of targeting the BMA or its factors, as novel treatment strategies and unveil a new mechanism of MM dependence on the BM.


Falank: BioPact Ventures, LLC: Research Funding. Farrell: BioPact Ventures, LLC: Research Funding. Reagan: BioPact Ventures, LLC: Research Funding.

Author notes


Asterisk with author names denotes non-ASH members.