Immuno‐Reactive Cancer Organoid Models to Examine Microbiome Metabolite Effects on Immune Checkpoint Blockade Efficacy

Abstract

Introduction As the number of available immunotherapies for solid tumors increase, their prevalence in the clinic continues to rise as well. While the results are promising, a sizable percentage of patients are non-responders to all types of immunotherapy. Yet, there has been limited number of in vitro models to assess tumor immune-reactivity and enable directed study of the mechanisms of immunotherapy resistance in solid tumors. The goal of this project is develop a model in which we could validate cancer cell killing due to immune cell action in an easily modifiable environment. Towards that, we created 3D tumor tissue constructs (organoids) containing cancer cells paired with cytotoxic T-cells to model immunotherapy efficacy. Our choice of immune checkpoint inhibitor antibodies were a paired course of PD-1 and CTLA-4, which are used extensively in the clinic, or anti-CD47. Upon validation of our organoid model, they could address specific mechanisms behind certain tumor microenvironment elements, such as bacterial metabolites, on immunotherapy efficacy. Methods We created extracellular matrix like 3D structures using collagen/hyaluronic acid-based hydrogels into which tumor cell were mixed, along with T-cells. The cancer cells incorporated were either freshly-isolated CT-26 colon adenocarcinoma cells, 4T1 breast cancer cells, or cultured CT-26 cells. The T-cells were either derived from murine lymph nodes or were isolated from the murine tumor infiltrating mononuclear cell population. Organoids were treated with microbial metabolites and their response to therapeutic equivalent doses of anti- PD-1, CTLA-4 and CD47 antibodies was examined using cell viability assays, flow cytometry, RT-qPCR, and immunohistochemistry (IHC) staining. Results The results show that in our immune cell supplemented tumor organoid models the immune checkpoint inhibitor regimens stimulated internally localizing T-cells and inducing T-cell-mediated tumor cell killing. Checkpoint inhibitor treated samples showed proportionally greater loss of viability when compared to untreated controls (p<0.01). The results were corroborated by IHC, showing increased numbers of CD-4+ T-cells and cytotoxic proteins such as granzyme B in the stimulated samples. RT-qPCR results show an increase in key proteins for immune function when both immunotherapy and bacterial metabolites are administered simultaneously. Conclusion: We have created an ex-vivo immune-reactive tumor organoid model for studying immunotherapy. This model will allow us to modulate facets of the tumor cell component in the organoid system including cancer type and grade, tumor cell mutations, biochemical signals, and physical properties of the microenvironment. We can then observe the impacts of these changes on immunotherapy efficacy in order to identify factors that could be targeted to improve immunotherapy efficacy. In this study, we have focused on the effects of bacterial metabolites on immunotherapy efficacy. Preliminary data suggest that there may be a connection between these microbial factors and tumor sensitivity to immunotherapy.