We previously showed that highly activated natural killer cells can be generated in large numbers from the peripheral blood mononuclear cells (PBMC) of healthy donors and multiple myeloma (MM) patients by co-culturing with K562 stimulator cells modified to express 41BBL and membrane bound IL15 (Garg et al. Haematologica, 2012; in press). We are utilizing these expanded NK (ENK) cells in a clinical trial to treat genomically-defined high-risk relapsed MM. In this study we investigated possible mechanisms of MM escape from ENK cell control. First we determined whether it is feasible to induce ENK cell resistance by co-culturing 11 MM cell lines with ENK cells from the same donor in 6 well-plates at E:T ratios of 10:1. Fresh ENK cells were added every 3–4 weeks. The MM cell lines were tested for emergence of resistance at regular intervals using standard chromium release cytotoxicity assays. In seven of 11 cell lines tested all MM cells died upon adding ENK cells. However, from 4 MM cell lines (OPM2, JJN3, ANBL-6, INA-6), subpopulations emerged which were resistant to ENK cell-mediated lysis (Figure 1A). These cell lines retained resistance even when passaged for ≥3 months in the absence of ENK cells. Flow cytometry showed down regulation of Tumor Necrosis Factor-Related Apoptosis Inducing Ligand-Receptors (TRAIL-R) –R1 and -R2 on all ENK resistant cell lines. Further experiments with OPM2 showed that the resistant cell line was less sensitive to recombinant TRAIL-induced apoptosis. Bortezomib treatment of both ENK-resistant and-sensitive OPM2 cell lines increased TRAIL-R1 and –R2 expression and susceptibility to recombinant TRAIL. Interestingly, gene expression profiling of CD138 positive MM patient cells failed to demonstrate a statistically significant difference in the level of TRAIL-R1 and-R2 expression among molecular subgroups, high- versus low-risk 70 gene score, diagnosis versus relapse, or baseline versus 48 hours post bortezomib. However, we also observed that TRAIL-R1 gene expression was down regulated in only 2 of the 4 resistant cell lines suggesting that there might not always be a tight correlation for TRAIL-R expression between GEP and flow cytometry. Next, we demonstrated that supernatants derived from both ENK cell-sensitive and -resistant MM cell lines could down modulate cell surface molecules either by ligand-shedding or reducing expression of receptors critical for activation such as CD54, NKp30, NKp80, NKG2D, CD26, CD69, CD70 and DNAM-1. Furthermore, the cytolytic activity of ENK cells was adversely affected, suggesting an additional NK cell suppressive effect mediated by soluble factors. The latter could be reversed by re-incubation of ENK cells in fresh medium with IL2 (Figure 1B). In summary, down regulation of the TRAIL-R deserves further exploration as a potential mechanism of resistance to ENK cell mediated lysis of MM cells and could be examined pre-and post-ENK cell therapy. MM cells may also evade ENK cell-mediated lysis via the production of suppressive factors. However, both mechanisms of resistance may be surmountable by NK cell sensitizing agents such as bortezomib or by generating the optimal cytokine milieu for infused ENK cells.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.