A spatial profiling-based immune hub perturbation platform to improve immunotherapy in colon cancer - PROJECT SUMMARY Chimeric antigen receptor (CAR) T cell therapies and immunotherapies that boost endogenous T cell responses show limited success in most solid tumors including colorectal cancer (CRC). Potential new drug targets and next-generation cell therapies are rapidly emerging, but a preclinical model that faithfully recapitulates immune responses in human tumors and enables us to systematically test these candidate genes and therapies is lacking. Neither CAR T cells nor endogenous T cells execute their function in isolation. In our recent single cell RNA sequencing and spatial profiling studies in human CRC, we discovered several conserved, spatially organized networks between malignant, T and other immune cells in the tumor, namely inflammatory, tissue-remodeling, and anti-tumor hubs. Importantly, anti-tumor hubs which form as focal clusters of T cell chemokine-expressing myeloid and malignant cells together with activated T cells at the malignant/stroma interface, were associated with immunotherapy response in CRC and lung cancer. Thus, instead of optimizing for cell-autonomous effects in cancer or T cells, we propose to systematically prioritize candidate genes and therapies based on their potential to induce effective anti-tumor hubs, a task that requires a new experimental framework. In Aim 1, we will develop a genetic immune hub perturbation platform in mouse CRC organoid-based liver metastases. Our CRC organoids are engineered with defined driver mutations and, upon transplantation, generate 50-100 lesions with histologic, transcriptional, and immunologic features seen in our extensive human data sets. We propose to engineer these organoids with genetic perturbations that can be identified by spatial profiling alongside the perturbation-induced immune hub alterations. This will enable the simultaneous testing of several cancer cell-expressed and immune hub-associated candidate genes across multiple lesions per mouse. In Aim 2, we will engineer mouse CRC organoid lines and primary T cells for the testing of improved cell therapies, focusing on GCC19 CAR T cells. GCC19 CAR T cells showed remarkable efficacy in metastatic CRC according to a small Chinese trial and are currently being tested in a multi-center U.S. trial, including at UCSF. We will benchmark GCC19 CAR T cell efficacy in our model against the clinical results and develop a barcoded pooled CAR T testing platform with spatial readout, characterizing immune hub engagement of 4 new cell therapies. This proposal is innovative in concept and method because 1) our focus is on functional cell collectives instead of cancer or T cell-intrinsic properties, and 2) we combine advanced in vivo models based on the genetic engineering of CRC organoids and primary T cells with immune hub detection by single cell-resolved spatial profiling. We expect our work to be highly significant by transforming the utility of organoid-based models for preclinical research and to greatly accelerate the development of efficacious immunotherapies in metastatic CRC. Our work also sets the foundation for future efforts with organoid models from other epithelial tumors and patient-derived organoids transplanted into humanized mice.
