Fig 2: Forward genetic screen to identify regulators of the IFN? response.(A) Representative histograms of the three selected cell surface markers targeted in macrophage CRISPR screens: MHCII, CD40, and PD-L1. Blue histograms indicate expression of each marker in unstimulated macrophages, and alternatively colored histograms show expression following 24 hr stimulation with recombinant murine IFN? (10 ng/mL). Gates used for sorting ‘high’ and ‘low’ populations are shown. (B) Schematic of CRISPR screens. (C) Relative enrichment of select positive control (points) and all 1000 non-targeting control (NTC) sgRNAs (gray distribution) are plotted as a function of their log2 fold enrichment (‘high’ vs. ‘low’ bins). Data are from both replicate selections for each sgRNA (sgRNA denoted by shape). (D) Heatmap of ß scores from CRISPR analysis, ordered according to k-means clustering (k = 8) of the 5% most enriched or depleted genes in each screen. (E) Macrophages were stimulated for 24 hr with TNF (25 ng/mL), IFN? (10 ng/mL), or both TNF and IFN?. Mean fluorescence intensity (MFI) of CD40 and MHCII was quantified by flow cytometry. Data are mean ± standard deviation for three biological replicates. Representative scatter plot from two independent experiments is provided. (F) Macrophages transduced with sgRNA targeting Stat1, Ostc, Cnbp, or a NTC were cultured with or without IFN? for 24 hr, and cell surface expression of PD-L1 (MFI) was quantified by flow cytometry. For each genotype, data are the mean of cell lines with two independent sgRNAs ± standard deviation. Data are representative of three independent experiments. Statistical testing in panel (C) was performed with Tukey’s multiple comparisons test. Within each screen, the sgRNA effects for each gene were compared to the distribution of NTC sgRNAs. Statistical testing in panels (E) and (F) was performed by one-way ANOVA with Holm–Sidak multiple comparisons correction. p-Values of 0.05, 0.01, 0.001, and 0.001 are indicated by *, **, *** and ****, respectively. Figure 1—source data 1.Whole-genome profiling of macrophage MHCII, CD40, and PD-L1 expression. Figure 1—source data 2.k-means clustering of CRISPR-KO beta scores for regulators of the IFN? response.

Fig 3: SBT-100 decreases the expression of immune checkpoint inhibitors in various cancer cell lines. (A) Surface expression of checkpoint inhibitors, B7-H3, OX-2, and PD-L1, was determined by flow cytometry in SJSA-1 osteosarcoma cells in the presence or absence of SBT-100, upon stimulation with IFN-? (75 µg/mL). Fluorescent microscopy of PD-L1 expression in MDA-MB-231 cells without SBT-100 (B) and with SBT-100 (C). Confocal fluorescent microscopy of PD-L1 expression in MDA-MB-231 cells without SBT-100 (D) and with SBT-100 (E), both with the nuclei detection with DAPI. The scale bars in the lower-left corner of panels (B–E) represent 50 µm.

Fig 4: GSK3a/ß and Med16 function independently from mTORC1 to control IFN?-mediated MHCII expression.(A) NTC or Med16 KO cells were treated with DMSO or CHIR99021 then left untreated or stimulated with IFN? overnight. The following day MHC II cell surface expression was determined by flow cytometry. The quantification of the MFI of MHCII from four biological replicates is shown. **p < 0.001 by two-way ANOVA with multiple comparison correction. (B and C) NTC cells were treated with DMSO or 30 nM Torin2 for 2-hr then were stimulated with 6.25 ng/ml IFN? overnight. Eighteen hr later (B) MHCII expression and (C) PD-L1 expression were quantified by flow cytometry. Shown is the MFI of the indicated marker from three biological replicates and is representative of three independent experiments. (D and E) NTC, Gsk3b KO and Med16 KO cells were treated with DMSO or 10 uM CHIR99021 and/or the indicated Torin2 for 2 hours. Cells were then treated with IFN? and the surface expression of (D) MHCII and (E) PD-L1 were quantified by flow cytometry. Shown is the MFI of the indicated marker from three biological replicates and is representative of three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, **** p < 0.0001 by one or two-way ANOVA with correction for multiple comparisons.

Fig 5: Complex I is necessary for IFN?-induced MHCII and PD-L1 expression.Metabolic phenotypes in macrophage mutants were confirmed by measuring intracellular (IC) ATP abundance following culture in media containing only (A) glucose or (B) pyruvate. Values are normalized to the average respiratory capacity of non-targeting control macrophages (NTC) and are the mean ± standard deviation for four biological replicates. Statistical testing within each condition (with or without IFN? for 24 hr) was performed by one-way ANOVA with Dunnett’s multiple comparisons correction. (C–F) NTC, positive control (sgIfngr1 and sgPtpn2), and complex I mutant (sgNdufa1 and sgNdufa2) macrophages were stimulated for 24 hr with recombinant murine IFN?. Plotted values in (C) and (E) are the geometric mean fluorescence intensity (MFI) for a given mutant normalized to an internal control present in each well; for each gene, the data are the mean for two independent sgRNAs ± standard deviation. Representative histograms are provided in (D) and (F). Data are representative of >5 independent experiments. (G) MHCII MFI of macrophages stimulated with IFN? and treated with rotenone at the indicated concentrations for 24 hr. Mean ± standard deviation for two biological replicates are shown. Data are representative of four independent experiments. (H) Left: MHCII MFI on macrophages cultured in complete media (CM) and stimulated with IFN? and the indicated inhibitors for 24 hr. Right: MHCII MFI on macrophages cultured in CM or media containing only pyruvate (Pyr) or citrate (Cit) stimulated with IFN? for 24 hr. Mean ± standard deviation for two or three biological replicates is indicated. Data are representative of four independent experiments. Statistical testing was performed by one-way ANOVA with Tukey’s correction for multiple hypothesis testing. p-Values of 0.05, 0.01, 0.001, and 0.001 are indicated by *, **, ***, and ****, respectively.