Colorectal cancer is the second-leading cause of cancer death worldwide. Early-stage disease can be detected with preventative medical screening and is treatable with surgical resection. Sixty-percent of patients, however, are diagnosed with advanced-stage disease. With the current treatment options including mainly chemotherapeutic regimens and targeted therapy where applicable, five year survival for these patients is a mere 14%. Immunotherapeutic strategies that bolster the anti-tumor capabilities of the immune system have offered hope to patients, clinicians and researchers alike by rendering complete and/or durable responses in some cancer patients. High T cell infiltration is associated with favorable treatment outcomes (immunotherapy and chemotherapy inclusive) in human CRC. Because cytotoxic T cells rely on direct contact with their targets, treatments that potentiate or drive T cell recruitment into tumors represent an opportunity to improve treatment outcomes.

In this work, we investigate the tumor microenvironment (TME) of three genetic mouse models of colorectal cancer with accumulating mutations that reflect the oncogenic driver mutation of colorectal cancer, namely, APC, KRAS and p53. We therefore compared the TME of A-, AK- and AKP-mice by flow cytometry and single-cell RNA sequencing (scRNAseq). We found that constitutively-active KRAS tumors were associated with increased T cell presence within the tumors compared to tumors without the mutation. scRNAseq enabled the precise definition of tumor cell clusters and thus the direct comparison of the transcriptome of tumors with (AK- and AKP-mice) and without the KRAS activation (A-mice), enabling us to unravel the link between KRAS activation and T cells in the TME.

We found that Annexin A1 (Anxa1) was differentially expressed by Kras-mutant bearing tumors compared to tumors without. Furthermore, we show that ANXA1 expression is significantly associated with high immune cell infiltration in colorectal cancer patients. Deeper analysis into the function of Anxa1 and its role in T cell recruitment in the TME of CRC is required. Ultimately, a therapy that recruits T cells (perhaps inspired by Anxa1) could be combined with an immunotherapeutic regimen to increase its efficacy. Moreover, we validated that these mouse models respond modestly to aCTLA4 or aCD40 monoclonal antibody treatment, and thus could be advantageously combined with a T cell recruitment therapy.
Additionally, we defined how the TME of mouse models with differential tumor-driving mutations (A-, AK- and AKP-mice) changed over time. With a longitudinal study, were able to observe that T cell abundance is dynamic over time within all three models and that the moment of maximal T cell infiltration for each model was dependent on the oncogenic-driver mutations borne by the mouse model. Single-cell TCR-sequencing of AKP-tumors showed robust TCR expansion in response to tumor formation, suggesting that APC, KRAS and p53 mutations within the same model is sufficient to induce an adaptive immune cell response.