Article Figures & Data
Figures
- Figure 1.
Knockout of IRF1 in tumor cells causes a loss of in vivo tumorigenicity. A–C, Cell growth rates of MC38, B16-F10, and CT26 WT and IRF1-KO in vitro. Cells were seeded and cultured in 10 wells of cell culture plates, and cell numbers were counted daily (n = 3). For each data point mean and SEM were plotted. D–F, Tumor growth rates of MC38 (n = 5), B16-F10 (n = 5), and CT26 (n = 5) WT and IRF1-KO cells in C57BL/6 and/or BALBc mice. Mice were subcutaneously or intradermally injected with 106 (MC38 and CT26) or 5 × 105 (B16-F10) cells, followed by tumor growth measurements. Representative results from at least twice repeated experiments are shown. PR, partial response; CR, complete response.
- Figure 2.
Host CD8+ T cells are required for the loss of tumorigenicity in IRF1-deficient tumor cells. A, Tumor growth of MC38 IRF1-KO cells in CD8+ T-cell depletion C57BL/6 mice. 106 of MC38 WT or IRF1-KO cells were subcutaneously injected into C57BL/6 mice (n = 8). IRF1-KO cell injection groups of mice were intraperitoneally injected with 250 μg of anti-CD8 or rat IgG2b isotype control 2 days before tumor cell injection, then followed by antibody or isotype control injections twice a week. A group of WT MC38 cell–injected mice were intraperitoneally injected with anti-CD8 as control. Representative results from at least twice-repeated experiments are shown. B, MC38 IRF1-KO cells induced WT tumor regression. Following complete regression of MC38 IRF-1 KO tumors (n = 8), mice were rechallenged with MC38 WT tumor cells on day 47 of postinjection. Seven of 8 mice showed complete tumor regression. Representative results from at least twice-repeated experiments are shown.
- Figure 3.
Loss of IRF1 in tumor does not influence the frequencies of tumor-infiltrating lymphocytes populations. A, A total of 5 × 105 of B16-F10 WT or IRF1-KO cells were intradermally injected into C57BL/6 mice (n = 13, 12). Tumors were collected around size 100–200 mm3 on days 12–14 postinjection. B, Representative flow cytogram (left) and tabulated (right) percentage of exhausted T cells and PD-1 intermediate expression CD8+ T cells (activated T cells) in B16-F10 WT and IRF1-KO tumor groups. C, Representative flow cytogram (left) and tabulated (right) percentage of Tregs in B16-F10 WT and IRF1-KO tumor groups. The ratio of CD8+ T cells: Tregs (D) and the frequencies of PD1+, Tim3+, LAG3+ in CD8+ T cells (E) in the two groups. For each column mean and SEM were plotted. Statistical significance was calculated by unpaired Student t test.
- Figure 4.
PD-L1 expression in vitro and in vivo is specifically affected by IRF1 loss. B16-F10 WT and IRF-1 KO cells were treated with mouse IFNγ for 0, 2, 4, 6, and 8 hours, and collected for the detection of expression of PD-L1. A and B, The mRNA expression of PD-L1 and ICAM1 were detected using TaqMan RT-PCR. C, Total protein expression of PD-L1 was examined by immunoblotting. D, The cell surface expression of PD-L1 was assessed by flow cytometry. E, A total of 5 × 105 of B16-F10 WT or IRF1-KO cells were intradermally injected into C57BL/6 mice (n = 5). Tumors were collected on day 12 of postinjection. The percentage of and geometric mean (MFI) of PD-L1high in CD45− cells (tumor cells) were tested by flow cytometry. In A and B, each data point represents mean and SEM from 3 independent replicates. Representative results from twice-repeated experiment are shown in C–E.
- Figure 5.
IRF1-deficient tumor cells affect infiltrating T-cell functions ex vivo and are more vulnerable to CD8+ T-cell cytotoxicity in vitro. A, Intracellular cytokine production of infiltrating CD8+ T cells in B16-F10 WT and IRF1-KO tumors (n = 9, 8). Single-cell suspension was prepared after tumor collection. The mixture of tumor cells and infiltrated lymphocytes was stimulated overnight with PMA and ionomycin, then cytokine secretion was blocked with GolgiPlug for 4 hours. Representative flow-cytogram for intracellular cytokine expression is shown. For each type of cytokine expression, cumulative mean and SEM were plotted from twice-repeated experiments, statistical significance was calculated by unpaired Student t test, and presented as *, P < 0.05. B, The percentage of dead tumor cells at different Pmel T: tumor cell ratios. A total of 1 × 104 tumor cells in each 96-well plate were cocultured with different amount of Pmel CD8+ T cells overnight. Cells were harvested and examined for viability using Zombie Aqua Fixable Viability Kit. The experiment was repeated 3 times. For each data point, mean and SEM from twice-repeated experiments were plotted, statistical significance calculated by two-way ANOVA with Sidak multiple comparison test, and presented as *, P < 0.05; **, P < 0.01.
- Figure 6.
Doxycycline (DOX)-induced PD-L1 expression rescues tumorigenicity of IRF1-KO cells in mice. B16-F10 WT cells were treated with mouse IFNγ for 4 hours for positive control. B16-F10 WT, IRF1-KO, and IRF1-KO pInducer20-PD-L1 cells were treated with doxycycline for 48 hours and collected for the detection of PD-L1expression. A, Total protein expression of PD-L1 was examined by immunoblotting. B, The cell surface expression of PD-L1 was assessed by flow cytometry. Representative histograms from twice-repeated experiments are shown. C, A total of 5 × 105 of B16-F10 WT, IRF1-KO, or IRF1-KO pInducer20-PD-L1 cells were intradermally injected into C57BL/6 mice (n = 5). One group of either B16-F10 IRF1-KO or IRF1-KO pInducer20-PD-L1 cell–injected mice were fed with doxycycline-containing food (200 mg/kg) from 3 days before tumor cell injection, and then continued being fed with doxycycline-containing food. Tumor growth measurements were taken with all groups of mice.
Volume 7, Issue 8
