Drug resistance is a major obstacle to the treatment of most human tumors. In this study, we find that dual-specificity phosphatase 16 (DUSP16) regulates resistance to chemotherapy in nasopharyngeal carcinoma, colorectal cancer, gastric and breast cancer. Cancer cells expressing higher DUSP16 are intrinsically more resistant to chemotherapy-induced cell death than cells with lower DUSP16 expression. Overexpression of DUSP16 in cancer cells leads to increased resistance to cell death upon chemotherapy treatment. In contrast, knockdown of DUSP16 in cancer cells increases their sensitivity to treatment. Mechanistically, DUSP16 inhibits JNK and p38 activation, thereby reducing BAX accumulation in mitochondria to reduce apoptosis. Analysis of patient survival in head & neck cancer and breast cancer patient cohorts supports DUSP16 as a marker for sensitivity to chemotherapy and therapeutic outcome. This study therefore identifies DUSP16 as a prognostic marker for the efficacy of chemotherapy, and as a therapeutic target for overcoming chemoresistance in cancer.

No abstract

e14178 Background: Colorectal Cancer (CRC) is the third most commonly diagnosed cancer in the world. Yet, CRC has been difficult to treat with immunotherapy, as majority (85%) of CRC are Microsatellite stable (MSS), lowly immunogenic and do not respond well. To investigate immuno-cytotoxicity in MSS cancers, we set up an ex vivo co-culture model using patient-derived tumor epithelial cells and autologous tumor-infiltrating lymphocytes (TILs). Aims: To perform phenotypic and functional characterisation of the model, and show we are able to modulate immuno-cytotoxicity. Methods: Fresh tumor, adjacent normal and/or liver metastasis samples were collected and freshly dissociated from 29 colorectal cancer patients undergoing surgery. TILs and tumor epithelial cells were grown in culture. Weekly immunophenotyping of TILs was performed via flow cytometry and data was analysed to identify temporal changes in proportions of cell types. To modulate cytotoxicity, epithelial cell lines were co-cultured with autologous and allogeneic TILs at various effector:target ratios with and without anti-PD1 (Pembrolizumab). Caspase dye was used to detect apoptosis and measure cell death. Results: TILs from most patients were successfully expanded, with a theoretical potential of up to 10,000 times. CD4+ T cells were selectively expanded, while CD8+ T cells decreased in proportion over time. A select group of patients showed an opposite trend. Cell type proportions in metastases mirrored those of the primary tumour while NK cells expanded more in tumour samples than normal samples. Selection proceeded up to 3 weeks then decreased. In the cytotoxicity experiment, we show that cell death is increased at higher effector:target ratios in the presence of autologous TILs. Addition of Pembrolizumab further modulates cytotoxicity in vitro. Conclusions: We have developed and characterised an autologous T cell and epithelial cell co-culture system to evaluate immuno-cytotoxicity in colorectal cancer. This will allow screening of perturbations for improved immuno-cytotoxicity in colorectal cancer.