Fig 1.
Generation of murine monoclonal tumor models of resistance to immune checkpoint inhibitors using CRISPR/Cas9.
(A) Frequency of gene alterations in human cancers across TCGA datasets. Examined gene lists (insets) are subdivided into antigen presentation machinery, checkpoint inhibitory ligands, immune-associated genes, IFNγ signaling pathway, and MHC class I. Genomic alterations include all non-silent single nucleotide polymorphisms (SNPs) and indels. (B) Schematic workflow for the development and validation of CRISPR gene knockout murine cell lines. (C) Representative gating strategy used to isolate transfected (GFP+) single cells by flow cytometry. Insets denote percentage of GFP+ cells. (D) B2m, Jak1 and LMP2 protein expression in 7–8 individual clones versus wild type EMT6 cells.
Fig 2.
Validation of CRISPR gene knockout single-cell clones.
(A) Oncoplot of mutations in selected genes in the generated CRISPR clones. (B) Corresponding western blot quantification of designated proteins in the B2m, Jak1, and LMP2 gene knockout (KO) clones previously incubated for 24 h in the presence or absence of murine IFNγ. Beta-actin was used as loading control. Blue boxes denote protein expression of the KO gene in each corresponding clone. All three KOs were performed once, independently of each other.
Table 1.
Summary table of the Sanger sequencing, whole exome sequencing, and western blotting results in EMT6 KO clones. Transcript and protein change columns describe the position and change according to the Human Genome Variation Society (HGVS) nomenclature.
Fig 3.
Effects of knocking out β2m, Jak1, and LMP2 on the expression of immune-related cell surface markers in vitro and in vivo.
(A) Representative histograms of in vitro cell surface expression of MHC class I haplotypes (H2-Kd and H2-Dd) and PD-L1 in EMT6 wild type (wt) cell line and corresponding KO cell lines on untreated cells compared to staining isotype controls and IFNγ-treated cells. (B) Representative histograms of cell surface expression of IFNγRI in vitro (top panel) on untreated cells compared to staining isotype controls and IFNγ-treated cells, and in vivo (lower panel) on untreated EMT6 wt and CRISPR KO clones compared to staining isotype controls. (C) Representative histograms of in vivo cell surface expression of MHC class I haplotypes (H2-Kd and H2-Dd) and PD-L1 in untreated EMT6 wt tumors and designated single gene-deleted EMT6 KO tumors compared to isotype controls. Insets denote percentage of positive cells and geometric mean fluorescence intensity (gMFI) for the designated protein. In vitro data are representative of 2 independent experiments with 3 replicates per experiment. In vivo data are representative of 2 independent experiments with 4–7 mice per experiment. KO cell lines: B2m KO clone #2 (442–6), Jak1 KO clone #1 (450–4), and Psmb9 KO clone #1 (454–5).
Fig 4.
Effects of β2m, Jak1, and LMP2 single-gene deletion on the intra-tumoral immune cell landscape.
Quantification of macrophages, myeloid derived suppressor cells (MDSC) of mononuclear (M-MDSC) or polymorphonuclear origin (PMN-MDSC), dendritic cells (DC), CD4+ T-lymphocytes (CD4), regulatory CD4+ T-lymphocytes (Tregs), CD8+ T-lymphocytes (CD8), and natural killer cells (NK) per milligram of individual EMT6 wt and designated gene KO tumors, identified by the designated phenotype. KO cell lines: B2m KO clone #2 (442–6), Jak1 KO clone #1 (450–4), and Psmb9 KO clone #1 (454–5).Data are representative of 2 independent experiments with 4–7 mice per experiment. Box plots denote values from each tumor, with median±SD. Tumors were analyzed when their volume reached ~1000 mm3. Two-tailed t-test, * = p < 0.05, ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001.
Fig 5.
Gene knockout tumor models are unresponsive to immune checkpoint inhibitor therapies.
Generic treatment schedule for all EMT6/EMT6 knockout (KO) (A) and MC38/MC38 B2m KO (C) tumor experiments. Arrows indicate days post tumor implant (day 0) when mice received αPD-1 or αPD-L1. Treatment was initiated when tumors reached an average volume of 50 mm3. (EMT6/EMT6 KO) or 60 mm3 (MC38/MC38 B2m KO). Graphs show (B) representative tumor growth curves (mean ± SEM) and (C) survival of EMT6 wt and gene KO tumors (n = 6-7/group) treated as depicted in the schematic in Fig 5A. Insets denote median overall survival (mOS) and number of mice without palpable tumors per treatment group at the time of study termination (day 50). (E) Graphs show tumor growth curves (mean ± SEM) of MC38 wt and B2m KO tumors (n = 5-7/group) treated as depicted in the schematic in Fig 5D. Inset denotes tumor growth in individual mice implanted with MC38 B2m KO cells and treated as depicted in Fig 5D. Black dashed lines depict days of αPD-1 or αPD-L1 doses. Tumor growth curves and corresponding survival of each KO tumor model were performed once independently of each other. The corresponding wt EMT6 data were performed three times independently of each other (data pooled). KO cell lines: EMT6 B2m KO clone #2 (442–6), EMT6 Jak1 KO clone #1 (450–4), EMT6 Psmb9 KO clone #1 (454–5), and MC38 B2m KO clone #3. Ordinary two-way ANOVA with Tukey’s multiple comparisons test (B), or two-tailed Log-rank (Mantel–Cox) (C). * = p < 0.05, ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001.