UCLA

NK cells are known to limit growth and expansion of cancer stem cells and oral tumors by providing key cytokines such as IFN-γ and TNF-α, which drive differentiation of stem-like/poorly differentiated cancer stem cells. The detailed understanding of how and why their function is affected by the cancer microenvironment is very important for outlining the most efficient NK cell-based immunotherapies for cancer patients. In this regard, we first characterized NK cells from the cancer patients for their IFN-γ secretion, cytotoxic function against the cancer stem cells, and their surface receptors. In previous studies, we demonstrated that IFN-γ plays a dominant role in up-regulation of B7H1, CD54 and MHC1 surface receptor expression and inhibition of tumor growth by differentiating the cancer stem cells, and via inflammatory cytokine release when they are in contact with cancer stem cells. We also showed that induction of NK cell-mediated cytotoxicity resistance and differentiation in the stem cells correlated with the increased expression of CD54, B7H1, and MHCI, and mediated by the combination of membrane-bound or secreted IFN-γ. In this study, we demonstrated that NK media adjusted based cells from cancer patients secreted lower IFN-γ. When we used the NK cell condition media from the cancer patient and healthy NK cells to treat the cancer stem cells, the condition media from cancer patient NK cells was less capable of differentiating the cancer stem cells. The cancer stem cells treated with patient NK cells condition media, were more susceptible to NK cell-mediated cytotoxicity and displayed lower surface expression of CD54, MHC1, and B7H1 when compared to cancer stem cells treated with healthy IFN-γ.

We also demonstrated that among all the cells we characterized to design a novel strategy to expand highly potent NK cells, osteoclasts fit best based on the NK receptors ligand expression and secreted cytokines required for NK cells expansion and activation. In this regard, we tested the cancer patient osteoclasts (OCs) for their surface expression of MHC1 and MICA/B, both were showed lower expression levels on cancer patient OCs. So, not just the NK cells receptors required for tumor recognition are lower on cancer patient NK cells, but the ligands required to increase the expression of those receptors are also down-modulated in cancer patient OCs. Next, we cultured the cancer patient NK cells with healthy donor OCs, and we were able to increase the surface expression of NKG2D on patient NK cells but, again the cytotoxic activity and IFN-γ were lower in patient NK cells. We confirmed our finding by analyzing the OCs from KRAS mutated mice and hu-BLT mice injected with tumor.

Based on our data, we were not convinced in the use of autologous NK cells to treat the cancer patients, our strategy was to use expanded allogenic NK cells as cancer cell immunotherapy. Next, we tested the expanded allogeneic NK cells in tumor implanted hu-BLT mice. With just one NK cell injection (1.5 million cells), tumor growth was inhibited in the mice. On analyzing the immune cell compartments, NK cell-mediated cytotoxicity and IFN-γ from immune cells was improved significantly. Tumors dissected from animals that received NK therapy showed differentiation profile, high MHC1, CD54 and B7H1 on their surface and lower growth in the culture when compared to tumor dissected from hu-BLT mice without NK cell immunotherapy. It has been known that mature alloreactive NK cells can be safely infused into patients with no increased incidence of graft versus host disease (GvDH). To avoid any kind of risk of GvHD, we can also consider isolating the contaminating T cells from the super-charged NK cells, and injecting the high purity NK cells.