This was a fantastic international collaboration with the @mgarnett.bsky.social, Halim and Voest teams. Thank you and congrats to all the co-authors, especially Alex, @gabrielepicco.bsky.social, Vivien and Youhani.

Finally, in patients, tumors with lower CHD1 and MAP3K7 expression responded better to cancer immunotherapy, suggesting that these genes could serve as biomarkers of response to ICB.

In mice, knocking out these genes in tumour cells led to improved responses to immune-checkpoint blockade (ICB) and increased T cell numbers and activation in tumours.

CHD1 and MAP3K7 KO additively sensitises cancer cells to IFN-g, TNF-a and T cells and are known activators of NF-kB signalling. RNA-seq revealed they regulate the transcriptional response to cytokines involving the TF CDX2, overall switching the cell response from apoptosis to cell survival.

We also identified known regulators of cytokine signalling as hits in our cytokine and T cells screens across multiple cell contexts (e.g. TBK1, SOCS1, ADAR), allowing us to compare the against CHD1 and MAP3K7 KO and unpick the signalling pathways involved in each case.

Interestingly, CHD1 and MAP3K7 are co-lost in several common cancer types, such as prostate, thyroid and melanoma (~4 %), with higher rates of loss of either gene alone.

Through integrating these data, we identify CHD1 and MAP3K7 as key determinants of resistance to T cell attack. Our individual cytokine screens and the use of JAK1/2 KO isogenic cancer cell models allowed us to deconvolute different contributions of cytokines on T cell-mediated killing.

These T cells are derived from the same patient as the tumoroids - endogenous neoantigen expression drives reactivity. Special thanks to our collaborators from the Voest lab for pioneering this sophisticated system for studying anti-tumour immunity.

We used whole-genome CRISPR/Cas9 screens in autologous primary tumoroid T cell co-cultures to investigate genes that modulate cancer cell killing by T cells.

We generated a large-scale dataset of context-specific dependencies in 4 cancer cell models, across 4 cytokines and autologous tumour-reactive T cells, collectively encompassing data from >150 whole-genome CRISPR-Cas9 screen samples.

Deadline is 2nd of April