P-055: Conserved and unique pathways of Natural Killer cell resistance in multiple myeloma cell lines

Introduction: Multiple myeloma (MM) displays heterogeneity in gene expression profiles (GEP) at the cellular level that contributes to variable sensitivity to therapeutic agents. We used MM cell lines with variable sensitivity to natural killer (NK) cell-mediated cytotoxicity to determine if common pathways of resistance to cellular therapy could be identified. We utilized an induced pluripotent stem cell derived NK product expressing a chimeric antigen receptor (CAR) against B cell maturation antigen (BCMA) that is currently in Phase 2 trial for relapsed MM. In addition to engagement of the CAR, CAR-NK cells retain the ability to recognize MM cells through natural cytotoxicity or CD16 engagement.

Methods: MM cell lines (n=4) were stained with an antibody against BCMA. Flow cytometry showed greater than 85% expression of BCMA on U266, H929, and MM1S but only 40% expression on RPMI-8226. MM were incubated with CAR-NK cells at 1:1 ratio and samples at 0 and 24 hours used for genome wide transcriptome analysis. In parallel, fluorescently labelled MM cells were incubated with NK cells for 48 hours and killing was measured by live cell imaging (4 replicates per condition).

Results: RPMI-8226 were most resistant to CAR-NK cells (< 10% killing at 24 hours). H929, MM1S, and U266 showed 20-35% killing at 24 hours. Hierarchical clustering of the top genes with differential expression between pre-NK treatment and post-NK treatment showed 85 genes shared by all 4 cell lines that exhibited a 2-fold difference in gene expression after CAR-NK cell therapy. Genes highly expressed by all cell lines were HLA-B, TAP1, and TAP2, all of which suppress NK cell cytotoxicity through HLA engagement of inhibitory receptors. KEGG pathway analysis confirmed that genes in the NK cell cytotoxicity and cytokine receptor interaction pathways were enriched in the NK-resistant MM cells. Other genes upregulated in all 4 lines were PARP14, encoding a poly (ADP-ribose) polymerase and regulator of survival in MM cells as well as DTX3L, which encodes a ubiquitin ligase known to play a role in proliferation and drug resistance in MM.



Significantly more genes were altered in the most resistant cell line, RPMI-8226, compared to the more NK-sensitive cell lines. The gene encoding TRAIL, TNFSF10, was significantly upregulated and may directly modulate NK cell function through regulation of interferon-g signaling. The upregulated metalloproteases ADAM8 and MMP25 may promote RPMI-8226 resistance by clipping activating receptors from the NK cell surface.

Conclusions: Our data provide initial insights into the pathways that promote resistance to NK cell-mediated killing in MM. We are using human bone marrow samples before and after BCMA CAR-NK therapy to validate these findings in patients. These results will eventually identify novel targets that can improve cellular therapies and eliminate resistant MM.