645. T-Cell Activating Mesenchymal Stem Cells as a Biotherapeutic for HCC

Abstract

Background: The outcome for advanced stage hepatocellular carcinoma (HCC) remains poor, highlighting the need for novel therapies. Mesenchymal stem cells (MSCs) are actively being explored as cancer therapeutics due to their inherent ability to migrate to tumor sites and to deliver ‘toxic payloads’. We have recently shown that cells can be genetically engineered to secrete bispecific T-cell engagers (ENG), which activate T cells in an antigen-dependent fashion. The goal of this project was now to determine if it is feasible to genetically modify MSCs to secrete ENG molecules that are specific for the HCC-associated tumor antigen Glypican-3 (GPC3), and to evaluate their ability to redirect T cells to HCC. Methods: Bone-marrow derived MSCs were genetically modified with lentiviral vectors encoding GPC3-ENG or an irrelevant ENG (EGFRvIII-ENG) and GFP. We performed coculture assays with GPC3-ENG-secreting MSCs (GPC3-ENG.MSCs, +/- freshly isolated T cells, and GPC3+ (HUH7 and G401) or GPC3- (A549) tumor cells. MSCs expressing an irrelevant ENG (EGFRvIII-ENG) served as control. We used standard immunological techniques for measuring cytokine production, proliferation, and cytolytic activity of T cells. Results: Transduction of MSCs resulted in a median transduction efficiency of 95% (range 90-97%) as judged by GFP expression. GPC3-ENG. MSCs activated T cells only in the presence of GPC3+ targets as judged by IFNγ production, and their ability to T-cell proliferation and killing of GPC3+ targets. EGFRvIII-ENG.MSCs did not activate T cells, confirming specificity. However, GPC3-ENG.MSCs did not induce IL2 secretion by T cells indicating lack of costimulation. To evaluate if provision of costimulation could overcome this obstacle, ENG.MSCs were further genetically modified to express CD80, 41BBL, or both. GPC3-ENG.MSCs expressing CD80 and 41BBL induced robust IL2 production in the presence of GPC3+ tumor cells in contrast to GPC3-ENG.MSCs expressing only CD80 or 41BBL. Conclusions: We have successfully generated MSCs that secrete GPC3-ENG and express costimulatory molecules on their cell surface. These MSCs induced potent anti-HCC T-cell responses in vitro, warranting further exploration of our MSC-based cell therapy approach for HCC.