Paired line graphs a show geometric mean fluorescence intensity of IRF8 up to 40000, CD39 up to 80000, and LAG3 up to 15000 in P14 positive CD90.1 positive cells across IRF2, IRF4, and IRF8 knockout versus control conditions, where IRF8 geometric mean fluorescence intensity declines significantly across all three knockouts, CD39 declines significantly for IRF2 and IRF4 knockouts, and LAG3 declines significantly for IRF8 knockout only. Bar graphs b show relative quantification normalized to b2m for IRF2, IRF4, and IRF8 transcript levels across all four knockout and control conditions, confirming specific knockdown with each respective knockout showing the lowest expression for its target gene. Contour plots c show IRF2, IRF4, and IRF8 protein levels versus CD90.1 transduction marker for control and respective knockout gRNA conditions, with knockout plots showing a clear loss of the respective protein population. Schematic d shows B16-gp33 mice receiving P14 control and P14 IRF knockout cells at day 7, followed by hashtagging antibody multiplexing of individual tumors at day 14, pooling and sorting of CD90.1 positive cells into P14 control and IRF knockout populations, and subsequent single cell RNA sequencing 10X analysis with knockout and control pairing using multiplexing sequence. Flow cytometry plots e show the sequential gating strategy progressing through lymphocytes on SSC-A versus FSC-A, single cells on FSC-H versus FSC-A and SSC-H versus SSC-A, living CD8 positive cells on ZombieNIR versus CD8a, followed by CD90.1 positive selection on CD45.2 versus CD90.1 and final separation of control and knockout populations on CD45.1 versus CD45.2. Gene set enrichment analysis plots f show effector versus exhausted CD8 T cell downregulation scores for IRF8, IRF4, and IRF2 knockouts versus control, with normalized enrichment scores and adjusted p-values.
Comparison of IRF2, IRF4, and IRF8 KO in CD8 TILs. (a) Expression of IRF8, LAG3, and CD39 in IRF KO P14+CD90.1+ CD8 TILs compared with their co-transferred control-transduced counterpart. (b) IRF2, IRF4, and IRF8 KO validation by qRT-PCR performed in the different transduced CD90.1+ CD8 T cells after transduction and FACS sorting. (c) IRF2, IRF4, and IRF8 KO validation by FACS performed in the different transduced CD90.1+ CD8 T cells recovered from tumors. (d) Schematic representation of experimental strategy used for scRNA-seq on ex vivo co-transferred IRF KO P14+CD90.1+ CD8 TILs with multiplexed hashtagging before FACS sorting and sequencing for paired sample analysis. (e) Representative gating strategy used for FACS-sorting hashtagged IRF KO and control-transduced transferred P14 CD8 T cells from B16-gp33 tumors for scRNA-seq. (f) GSEA of pseudobulk from scRNA-seq for the indicated IRF KO versus control using gene set effector versus exhausted CD8 T cells down. GEO accession: GSE9650. Adjusted P value (adj. P val) and normalized enrichment score (NES) are shown for each comparison. *P < 0.05, **P < 0.01, and ***P < 0.001.