Frequent Loss of IRF2 in Cancers Leads to Immune Evasion Through

Frequent Loss of IRF2 in Cancers Leads to Immune Evasion through Decreased MHC Class I Antigen Presentation and Increased PD-L1 Expression This information is current as of September 27, 2021. Barry A. Kriegsman, Pranitha Vangala, Benjamin J. Chen, Paul Meraner, Abraham L. Brass, Manuel Garber and Kenneth L. Rock J Immunol published online 30 August 2019 http://www.jimmunol.org/content/early/2019/08/30/jimmun Downloaded from ol.1900475

Supplementary http://www.jimmunol.org/content/suppl/2019/08/30/jimmunol.190047 http://www.jimmunol.org/ Material 5.DCSupplemental

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists by guest on September 27, 2021 • Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2019 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published August 30, 2019, doi:10.4049/jimmunol.1900475 The Journal of Immunology

Frequent Loss of IRF2 in Cancers Leads to Immune Evasion through Decreased MHC Class I Antigen Presentation and Increased PD-L1 Expression

Barry A. Kriegsman,* Pranitha Vangala,† Benjamin J. Chen,* Paul Meraner,‡ Abraham L. Brass,‡,x,{ Manuel Garber,† and Kenneth L. Rock*

To arise and progress, cancers need to evade immune elimination. Consequently, progressing tumors are often MHC class I (MHC-I) low and express immune inhibitory molecules, such as PD-L1, which allows them to avoid the main antitumor host defense, CD8+ T cells. The molecular mechanisms that led to these alterations were incompletely understood. In this study, we identify loss of the transcription factor IRF2 as a frequent underlying mechanism that leads to a tumor immune evasion phenotype in both humans

and mice. We identiﬁed IRF2 in a CRISPR-based forward genetic screen for genes that controlled MHC-I Ag presentation in Downloaded from HeLa cells. We then found that many primary human cancers, including lung, colon, breast, prostate, and others, frequently downregulated IRF2. Although IRF2 is generally known as a transcriptional repressor, we found that it was a transcriptional activator of many key components of the MHC-I pathway, including immunoproteasomes, TAP, and ERAP1, whose transcriptional control was previously poorly understood. Upon loss of IRF2, cytosol-to–endoplasmic reticulum peptide transport and N-terminal peptide trimming become rate limiting for Ag presentation. In addition, we found that IRF2 is a repressor of PD-L1.

Thus, by downregulating a single nonessential gene, tumors become harder to see (reduced Ag presentation), more inhibitory http://www.jimmunol.org/ (increased checkpoint inhibitor), and less susceptible to being killed by CD8+ T cells. Importantly, we found that the loss of Ag presentation caused by IRF2 downregulation could be reversed by IFN-stimulated induction of the transcription factor IRF1. The implication of these ﬁndings for tumor progression and immunotherapy are discussed. The Journal of Immunology, 2019, 203: 000–000.

he importance of adaptive immunity in preventing cancer mutagen-induced tumors (1–3). In addition, tumors derived from was revealed through studies in which immunodeficient such immunodeficient animals grew when transplanted into other animals, such as those lacking IFN-g, perforin, or RAG-2, immunodeficient hosts but were rejected when placed into im-

T by guest on September 27, 2021 werefoundtohaveamarkedincreaseinspontaneousand munocompetent hosts (3–5), providing further evidence that the immune system recognized such tumors and could reject them. In contrast, many tumors arising in immunocompetent animals *Department of Pathology, University of Massachusetts Medical School, Worcester, grew after being transplanted into immunocompetent hosts (3–5), MA 01655; †Department of Bioinformatics and Computational Biology, University of Massachusetts Medical School, Worcester, MA 01655; ‡Department of Microbi- thereby showing that cancers that arise and successfully progress ology and Physiological Systems, University of Massachusetts Medical School, in the face of the immune system have undergone immunoediting Worcester, MA 01655; xDepartment of Medicine, Gastroenterology Division, Uni- { to escape from immune control. This immunoediting process is versity of Massachusetts Medical School, Worcester, MA 01655; and Peak Gastro- enterology Associates, Colorado Springs, CO 80907 thought to be why many cancers express low levels of MHC class ORCIDs: 0000-0002-4146-2370 (B.A.K.); 0000-0003-4000-7891 (P.V.); 0000-0003- I (MHC-I) and upregulate certain inhibitory molecules (6). The 0153-4680 (P.M.). underlying molecular mechanisms responsible for these changes Received for publication April 29, 2019. Accepted for publication August 1, 2019. are poorly understood but have obvious potential impact on tumor This work was supported by National Institutes of Health Grants R01 AI114495 progression and immunotherapy (7, 8). (to K.L.R.), T32 AI095213 (to B.A.K.), and T32 GM107000 (to B.A.K.). This work The MHC-I presentation pathway is critical for immune rec- was funded by an Investigators in the Pathogenesis of Infectious Disease grant from ognition and elimination of tumors by CD8+ T cells. In this the Burroughs Wellcome Fund (to A.L.B.). A.L.B. is grateful to the Bill and Melinda Gates Foundation and to Gilead Sciences Inc. for their support. process, a fraction of peptides that are generated by proteaso- The RNA-sequencing data presented in this article have been deposited to the Gene mal degradation of cellular proteins are transported by the TAP Expression Omnibus under accession number GSE133089 (www.ncbi.nlm.nih.gov/ transporter into the endoplasmic reticulum (ER), wherein they can geo/query/acc.cgi?acc=GSE133089). be further trimmed by the aminopeptidase, ERAP1 (9–11). Sub- Address correspondence and reprint requests to Dr. Kenneth L. Rock, University of sequently, peptides of the correct length and sequence bind to Massachusetts Medical School, 55 Lake Avenue North, Worcester, MA 01655. E-mail address: [email protected] MHC-I molecules and these complexes are then transported to the + The online version of this article contains supplemental material. cell surface for display to CD8 T cells. This allows activated CD8+ T cells to identify and kill cells that are presenting tumor- Abbreviations used in this article: ChIP, chromatin immunoprecipitation; Ct, cycle threshold; DC, dendritic cell; ER, endoplasmic reticulum; EV, empty vector; IP, speciﬁc peptides (e.g., from mutant proteins) on their MHC-I (12). immunoprecipitation; IRF2-KO, IRF2-knockout; ISRE, IFN-stimulated response We performed an unbiased, forward genetic screen in human element; MFI, mean ﬂuorescence intensity; MHC-I, MHC class I; MHC-II, MHC class II; NEB, New England Biolabs; NSCLC, non–small cell lung cancer; PD-L1, cervical carcinoma HeLa H1 cells to identify genes whose loss programmed death-ligand 1; qPCR, quantitative PCR; RNA-seq, RNA sequencing; downregulated the MHC-I pathway. In this screen, the second RT, room temperature; sgRNA, short guide RNA; siRNA, small interfering RNA; strongest hit, second only to b2-microglobulin (the MHC-I WT, wild-type. L chain), was IRF2, an IFN regulatory transcription factor Copyright Ó 2019 by The American Association of Immunologists, Inc. 0022-1767/19/$37.50 that had not been previously recognized to positively regulate this

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1900475 2 FREQUENT LOSS OF IRF2 IN CANCERS LEADS TO IMMUNE EVASION pathway. In this article, we show that under basal conditions, 59-GACTACCGGTATGCCGGTGGAAAGGATGCGCATG-39;humanIRF2 not only does IRF2 positively regulate the MHC-I pathway by sgRNA mut reverse: 59-GCCGTGCCTCGCTGCGTGCATCCAGGGGATCT- transcriptionally activating genes necessary for peptide transport GAAAAATCTTCTTTTCCTTG-39; human IRF2 sgRNA mut forward: 59-GATGCACGCAGCGAGGCACGGCTGGGATGTGGAAAAAGATGCA- and processing, but it also transcriptionally represses the ex- CCACTCTTTAGAAA-39; human IRF2 MluI reverse: 59-GATCACGCGTTTA- pression of programmed death-ligand1(PD-L1),animportant ACAGCTCTTGACGCGGGCCTGG-39; mouse IRF2 AgeI forward: immune checkpoint molecule. These results, together with our 59-GATCACCGGTATGCCGGTGGAACGGATGCGAATG-39;mouse findings that many human cancers have downregulated IRF2 IRF2 sgRNA mut reverse: 59-CCTTTTTTCGAGGGGCGCTCTGATAA- GGGCAGCATCCGGTAGACTCTGAAGGCG-39; mouse IRF2 sgRNA expression and that tumor cells lacking IRF2 are more difficult to mut forward: 59- CTTATCAGAGCGCCCCTCGAAAAAAGGAAAGA- kill, demonstrate that loss of IRF2 by cancers is a common immune AACCAAAGACAGAAAAAGAAGAGAG-39; and mouse IRF2 MluI re- evasion mechanism and this has obvious therapeutic implications. verse: 59-GATCACGCGTTTAACAGCTCTTGACACGGGCCTGG-39. Cell surface staining Materials and Methods Cells Where indicated, mouse cells were blocked with 2.4G2 and stained for surface MHC-I levels with anti-Kb–allophycocyanin (AF6-88.5.5.3; A E DC3.2 is a J2 virus-immortalized dendritic cell (DC) line (13). A particular eBioscience), MHC-II levels with anti–I /I -PECy7 (M5/114.15.2; Bio- DC3.2 clone (with Renilla luciferase) was used for all experiments in this Legend), PD-L1 levels with anti–PD-L1-PE (10F.9G2; BioLegend), or study, as this clone has very strong cross-presentation and MHC class II with isotype controls (eBioscience mouse IgG2a-allophycocyanin eBM2a, (MHC-II) presentation, as compared with other clones. RF33.70 is a T cell eBioscience rat IgG2b k-PE eB149) at 1:200 dilutions. Where indicated, hybridoma that recognizes the OVA peptide OVA257–264 in the context of human cells were stained for surface MHC-I levels with W6/32. W6/32 b staining was performed either by two-step labeling with W6/32 hybridoma H2-K (14). MF2.2D9 is a T cell hybridoma that recognizes OVA258–276 in Downloaded from the context of I-Ab (13). RF33.70 and MF2.2D9 were transduced with supernatant followed by 1:500 donkey–anti-mouse Alexa 647 (Life lentivirus containing NFAT-luciferase. NIH-3T3 cells were stably trans- Technologies) or by one-step labeling with 1:200 FITC-conjugated W6/32 fected with the mouse H2-Kb molecule. A549 and MCF7 were kindly (eBioscience). Where indicated, human cells were stained for surface provided by Leslie Shaw (University of Massachusetts), and the D53m and PD-L1 levels with 1:200 rabbit anti–PD-L1 (28-8; Abcam), followed by H50m mouse MCA-induced sarcoma lines were kindly provided by 1:500 donkey–anti-rabbit Alexa 647 (Life Technologies). Normalized R. Schreiber (Washington University in St. Louis). The MCA-induced mean fluorescence intensity (MFI) was computed by dividing the geo- sarcoma lines were grown in R10 medium, and all other cell lines were metric MFI of each knockout cell line by the geometric MFI of the WT grown in RPMI 1640 (Life Technologies) supplemented with 10% FBS (no sgRNA) cell line. http://www.jimmunol.org/ (Hyclone), 1% NEAA (Life Technologies), 1% HEPES (Life Technologies), 1% Antibiotic-Antimycotic (Life Technologies), and 5 3 1025 M 2-ME T cell hybridoma Ag presentation (Sigma-Aldrich). The MCF7 growth medium was also supplemented with Cross-presentation and MHC-II presentation were measured by coculturing 10 mg/ml insulin. Antibiotic selection for CRISPR-Cas9 knockout cells was DC3.2 lines in the presence of the indicated concentrations of OVA-coated done for 2 wk in medium containing 5 mg/ml blasticidin (InvivoGen). iron-oxide beads (Polysciences) and RF33.70-Luc CD8+ T cells or All cells were grown in a 10% CO2 atmosphere at 37˚C. MF2.2D9-Luc CD4+ T cells, respectively, for 16–18 h. Then, One-Glo Plasmids luciferase substrate (Promega) was added and luciferase activity quantified by an EnVision plate reader (Perkin Elmer). Rescue experiments The LentiCRISPRv2 plus blasticidin selection plasmid (15, 16) was acquired were performed by adding the OVA beads and RF33.70-Luc cells 48 h from Addgene (83480) and, unmodified, is the same as the “no short guide posttransduction of the DC3.2 lines. Data from a representative cross- by guest on September 27, 2021 RNA (sgRNA)” plasmid. The plasmids used to target mouse B2m, TAP1, presentation and MHC-II presentation experiment (of n = 4) are shown TAP2, ERAP1, IRF2, or IRF1 or to target human IRF2 were constructed by where points represent mean 6 SD of technical duplicates. Normalized + inserting the following sgRNA sequences, respectively, into the Lenti- CD8 T cell activation was calculated for rescue experiments by dividing + CRISPRv2 plasmid: mouse B2m: 59-AGTATACTCACGCCACCCAC-39; the CD8 T cell activation (relative luminometer units of luciferase) at + mouse TAP1: 59-ACTAATGGACTCGCACACGT-39;mouseTAP2:59- each point by the CD8 T cell activation of the DC3.2 no sgRNA line ATTACACGACCCGAATAGCG-39; mouse ERAP1: 59-TGCAGCATCCA- transduced with empty vector (EV) (WT + EV). GAGCATAAT-39;mouseIRF2:59-TCCGAACGACCTTCCAAGAA-39; mouse IRF1: 59- CTC AT CC GC AT TC GAGT GAT- 39; human IRF2: 59- Small interfering RNA transfections TGCATGCGGCTAGACATGGG-39. To insert these sgRNA sequences into 104 HEK293T or HeLa H1 cells were transfected with 10 nM Silencer LentiCRISPRv2, two primers were created for each sgRNA sequence: Select siRNA (Invitrogen) and 0.3 ml Lipofectamine RNAiMAX 1) a forward primer wherein CACCG preceded the aforementioned sgRNA (Invitrogen) per well in flat-bottom 96-well plates. Individual small inter- sequence; and 2) a reverse primer where the reverse complement of the fering RNAs (siRNAs) used were negative control no. 1 (4390843) human aforementioned sgRNA sequence was flanked by AAAC at the 59 end and C b2m (s1854), human TAP1 (s13778), human IRF2 no. 1 (s7506), and human at the 39 end. Then, these 100 mM primer sets were annealed and diluted IRF2 no. 2 (s7504). After 72 h, adherent cells were trypsinized, washed in 1:50. Three micrograms of LentiCRISPRv2 plasmid was digested for 3 h at PBS supplemented with 2% FBS, and stained with the surface MHC-I 55˚C with BsmBI (New England Biolabs [NEB]), and removal of the 2-kb pan–HLA-A/B/C Ab, W6/32. Normalized MFI was computed by divid- filler sequence was confirmed by gel electrophoresis. The larger m.w. band ing the geometric MFI of each experimental siRNA by the geometric MFI was gel extracted and quick ligated with the diluted annealed oligos of the negative control siRNA. according to the manufacturer’s instructions (NEB). Stable competent Escherichia coli (NEB) was then transformed with 2 ml of the quick ligation Minigene transfections product according to the manufacturer’s instructions and grown overnight at 4 30˚C on Luria broth + ampicillin (100 mg/ml) plates. Plasmids were isolated A total of 10 3T3-Kb cells were transfected with 100 ng of various (Clontech) from individual colonies and sequenced (Genewiz) using the pTracer-CMV2 plasmids (Invitrogen) containing SIINFEKL precursors m primer hU6-F: 59-GAGGGCCTATTTCCCATGATT-39 to confirm proper (18) and 0.4 l Lipofectamine 2000 (Invitrogen) in flat-bottom 96-well plates. After 72 h, cells were stained with 25-D1.16 (specific for the insertion of the sgRNA into LentiCRISPRv2. In addition, sgRNAs were combination of SIINFEKL and H-2Kb) (19), followed by donkey–anti- checked for high indel efficiencies in transduced cells by TIDE analysis (17). Rescue plasmids were constructed by inserting mouse IRF2, TAP2, or mouse Alexa 647 (Life Technologies), and analyzed by flow cytometry. ERAP1 cDNA or human IRF2 cDNA into the constitutive expression vector, Transfected cells were identified by GFP expression and the MFI of pCDH-CMV (Addgene), with a modified multiple cloning site. Overlapping 25-D1.16 staining was measured on the gated transfected cells. Trans- PCRs were run using the primers below on either mouse cDNA from DC3.2 fection efficiency, based on GFP expression, ranged from 5 to 20%, depending on the vector. The normalized MFI for each experiment (with cells or human cDNA from HeLa H1 cells to create IRF2 cDNA sequences technical duplicates) was calculated by dividing the MFI of each knockout containing six synonymous mutations within the IRF2 sgRNA target site. The mouse IRF2 K78R sequence was constructed by further mutating A to line by the MFI of the WT (no sgRNA) line. G at nt 233. The wild-type (WT) TAP2 and ERAP1 cDNA sequences were RNA sequencing of C57BL/6 origin. All plasmids were sequenced to confirm correct sequences and reading frames. The primer sets for cloning the rescue/ RNA was extracted using the RNeasy kit (Qiagen) after 2 h of stimulating overexpression constructs were as follows: human IRF2 AgeI forward: the DC3.2 lines with 5000 U/ml mouse IFN-a, 2 ng/ml mouse IFN-g,or The Journal of Immunology 3 medium alone. A standard library preparation protocol was used with washed in cell lysis buffer then nuclear lysis buffer. Chromatin was 50 ng of total RNA as starting material. Libraries were checked for ap- sheared by sonication for 20 min, and fragment size (∼300 bp) was de- propriate fragment size traces by Bioanalyzer (Agilent), and concentra- termined by gel electrophoresis. Immunoprecipitations (IP) of the sheared tions were determined to achieve similar sequencing depth per library. chromatin were done using 2 mg of primary Ab—normal rabbit IgG Libraries were run on NextSeq 500/550 high-output and mid-output kits (sc-2027; Santa Cruz Biotechnology), rabbit anti-IRF1 (ab186384; Abcam), (Illumina) and all libraries had at least 107 reads with single index paired- or rabbit anti-IRF2 (B-80 H53L46; Invitrogen)—by incubating for 1 h at end sequencing. Trimmomatic-0.32 (20) was used to remove 59 or 39 room temperature (RT) then overnight at 4˚C. The next day, 25 mlof stretches of bases having an average quality of ,20 in a window size prewashed Protein A/G beads (Pierce) were added to each of the samples of 10. Only reads longer than 36 bases were kept for further analy- and incubated at RT for 30 min, then 90 min at 4˚C. After washing se- sis. RSEM v1.2.28 (21) was used to estimate gene expression, with pa- quentially with low-salt buffer, high-salt buffer, LiCl buffer, and TE buffer, rameters -p 4–bowtie-e 70–bowtie-chunkmbs 100–strand-specific. Gene the DNA was eluted from the beads. All immunoprecipitated samples were quantification was run on the transcriptome (RefSeq v69 downloaded from treated with RNase A (Qiagen) and Proteinase K (Qiagen) and then column UCSC Table Browser) (22). Genes with more than 15 TPM in any time purified (Clontech). ChIP–quantitative PCR (qPCR) was performed in tripli- point were considered expressed, and genes that did not achieve this cate wells using SYBR Green (Bio Rad) and unique primer sets (below) threshold were removed from further analysis. Batch effects were observed flanking the IRF1/2-binding site within the gene’s promoter (26–28), according between samples from different replicates. We used the log-transformed to the manufacturer’s instructions. Data shown as fold enrichment (2^2DCt) TPM normalized expression values as input to ComBat (package sva over the normal rabbit IgG control IP (n = 2). The primer sets for ChIP-qPCR version 3.18.0) (23, 24) with default parameters and a model that specified were as follows: mouse TAP2 forward: 59-CAAATTGACAGGCGCCATCT- different replicates as batches. Corrected TPM values were transformed 39;mouseTAP2reverse:59-GCTTCTTCTCAAACTGGATCTCC-39;mouse back to read counts using the expected size of each transcript informed by ERAP1 forward: 59-CTTAGGCTTGCTCTCTTTTAGCG-39; mouse ERAP1 RSEM. We only considered genes with at least 15 TPMs in at least one reverse: 59-GACTCCTGCTCCCGATCCTC-39; mouse PD-L1 forward: replicate at any time point. The expressed gene list was filtered to include 59-CAAGAAAGCTAATGCAGGTTTCAC-39; and mouse PD-L1 reverse: only genes with homologs as defined by the previous step. We used the 59-CCTGCGGATGACTTTAGAGTC-39. Downloaded from batch corrected counts per gene to identify differentially expressed genes by at least 2-fold between unstimulated cells (time 0) and 2 h following Western blotting a g stimulation with IFN- or IFN- and whose change in expression was Whole-cell lysates were prepared in RIPA buffer with protease inhibitor p , DESeq2 significant ( -adjusted 0.05) according to the package (v1.10.1) (Pierce), protein concentrations were determined by BCA assay (Pierce), (25) in R (v3.5.1). Because of the large transcriptional changes observed in and 10 mg of denatured samples were run on 10% reducing gels (Gen- DESeq2 this system, we turned off the fold change shrinkage in with Script). After transfer, PVDF membranes (MilliporeSigma) were blocked betaPrior = FALSE and we added a pseudocount of 32 to all timepoints to with TBS–Tween 13 + 5% milk and then blotted with rabbit anti-IRF2 http://www.jimmunol.org/ avoid spurious large fold change estimates from lowly abundant genes. (ab124744; Abcam) or rabbit anti-IRF1 (ab186384; Abcam) in TBS– RNA sequencing (RNA-seq) data have been deposited to the Gene Ex- Tween 13 + 2% milk overnight at 4˚C. The following day, membranes pression Omnibus (accession number GSE133089; www.ncbi.nlm.nih.gov/ were washed three times with TBS–Tween 13, goat-anti-rabbit HRP geo/query/acc.cgi?acc=GSE133089). (MilliporeSigma) was added for 1 h at RT, membranes were washed 33, Human lung specimens and HRP substrate (MilliporeSigma) was added. Following exposure, membranes were stripped (MilliporeSigma), blocked, and reblotted with Twenty-seven non–small cell lung cancers (NSCLCs) were acquired from mouse anti–b-actin (sc-47778; Santa Cruz Biotechnology) in TBS–Tween the University of Massachusetts Pathology archive of formalin-fixed and 13 + 2% milk overnight at 4˚C. The following day, membranes were paraffin-embedded patient samples. PD-L1 expression was determined at prepared as above except anti-mouse HRP (Pierce) was used instead. the time of diagnosis by immunohistochemistry (22C3 pharmDx assay; Agilent) by a surgical pathologist. RNA was extracted from the archival In vitro cytotoxicity assays by guest on September 27, 2021 material and analyzed by RT-PCR (see below). OT-I CD8+ T cells were preactivated by coculturing them for 4–5 d with RT-PCR irradiated, SIINFEKL-pulsed WT LPS-stimulated B cell blasts in T cell medium containing 30 ng/ml IL-2. After this time, OT-I were checked RNA was extracted from cell lines using the RNeasy kit (Qiagen) or from for CD8 expression and upregulation of CD27 and CD44. WT or formalin-fixed paraffin-embedded NSCLCs using the RNeasy FFPE kit IRF2-null RMA cells were counted, subdivided into respective tubes, (Qiagen) and reverse transcribed to cDNA using EcoDry premix random SIINFEKL-pulsed at the indicated concentrations or kept unpulsed, and hexamers (Clontech). Fifty nanograms of cDNA was used per well (done in CFSE-labeled as high (1 mM) for cells pulsed with SIINFEKL and triplicate) with the indicated TaqMan probes (Applied Biosystems), CFSE-labeled as low (0.1 mM) for unpulsed cells. Pulsed and unpulsed according to the manufacturer’s instructions. The TaqMan probes used are cells were recounted and mixed 1:1, and 105 cells total were plated per as follows: mouse b-actin, 4352933; mouse b2m, Mm00437762_m1; mouse well in U-bottom 96-well plates. A total of 5 3 104 OT-I were added to the H2-K1, Mm01612247_m1; mouse H2-Ab1, Mm00439216_m1; mouse respective wells and incubated at 37˚C for 4 h. Live RMA cells were gated Tap1, Mm00443188_m1; mouse Tap2, Mm01277033_m1; mouse Erap1, by flow cytometry, and CFSE levels analyzed. Specific killing was deter- Mm00472842_m1; mouse ERp57, Mm00433130_m1; mouse tapasin, mined for each RMA line by calculating 100 3 (1 – [(% targets / % Mm00493417_m1; mouse calnexin, Mm00500330_m1; mouse calreticu- bystanders)/(% control targets / % control bystanders)]). lin, Mm00482936_m1; mouse Tapbpr, Mm00520408_m1; mouse Irap, Mm00555903_m1; mouse Psme1, Mm00650858_g1; mouse Psme2, Results Mm01702833_g1; mouse Psmb8, Mm00440207_m1; mouse Psmb9, Mm00479004_m1; mouse Psmb10, Mm00479052_g1; mouse PD-L1, Regulation of MHC-I presentation by IRF2 Mm03048248_m1; human GAPDH, Hs02758991_g1; human IRF2, When HeLa H1 cervical carcinoma cells were immunoselected for Hs01082884_m1; human TAP2, Hs00241060_m1; and human ERAP1, MHC-I low variants in a genome-wide CRISPR-Cas9 screen, IRF2 Hs00429970_m1. mRNA expression levels in the IRF2-knockout (IRF2-KO) DC3.2 were compared with those in the WT DC3.2 by first normalizing scored as the second most targeted gene, with six out of six in- to the mRNA expression of b-actin (mouse) in each sample (2^2DDCt, dependent IRF2 guide RNAs hitting. To validate this hit, we tested where Ct is the cycle threshold). Statistical analysis was done by com- whether IRF2 affects surface MHC-I levels in human cells by paring the expression of a given gene to that of H2-Ab1. Relative mRNA knocking out IRF2. HeLa H1 cells and HEK293T kidney cells were expression levels in each human lung tumor were measured by first normalizing to the mRNA expression of GAPDH (human) in each transduced with vectors expressing either Cas9 and a sgRNA specimen (2^2DCt). Results are displayed after further normalization to targeting human IRF2 or Cas9 alone, and surface MHC-I levels one of the tumors. Statistical analysis was done using two-tailed Mann– were checked by flow cytometry (Fig. 1A). The IRF2-KO HeLa H1 2 Whitney U tests and linear regression models with R for goodness of fit. and HEK293T had significantly lower surface MHC-I levels than Chromatin immunoprecipitations their WT controls (Fig. 1B). To further validate this finding with another technique, we silenced IRF2 expression with two inde- Chromatin immunoprecipitation (ChIP) procedure generally followed the Thermo Fisher Scientific ChIP protocol. In short, 107 DC3.2-no sgRNA pendent siRNAs and found that both siRNAs similarly decreased cells were stimulated for 2 h with 2 ng/ml IFN-g or medium alone, har- surface MHC-I levels in HeLa H1 and HEK293T (Fig. 1C), vested and fixed with 1% formaldehyde, quenched with glycine, and confirming the phenotype. The magnitude of the decrease in 4 FREQUENT LOSS OF IRF2 IN CANCERS LEADS TO IMMUNE EVASION Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 1. IRF2 positively regulates MHC-I presentation under basal conditions. (A) Representative histograms of surface MHC-I levels by W6/32 staining in HeLa H1 lines stably transduced with the LentiCRISPRv2 constructs; IRF2-KO (IRF2 sgRNA) or WT control (no sgRNA). Background is with secondary Ab staining only. (B)NormalizedMFIofsurfaceMHC-IlevelsonHeLaH1(left) or HEK293T (right) stable knockout lines. (C) Normalized MFI of surface MHC-I levels on HeLa H1 (left) or HEK293T (right) cells after 72 h silencing with 10 nM indicated siRNA. (D) Normalized MFI of surface MHC-I levels by AF6 staining on 3T3-Kb stable knockout lines (left), surface MHC-I levels on DC3.2 stable knockout lines (middle), or surface MHC-II levels on DC3.2 stable knockout lines (right). (E) Representative Western blot in DC3.2 lines IRF2-KO (IRF2 sgRNA) or WT control (no sgRNA) for protein expression of IRF2 or b-actin as control. (F) Representative cross-presentation experiment of OVA beads by DC3.2 lines to RF33.70-Luc CD8+ T cell hybridoma. (G) Representative MHC-II presentation experiment of OVA beads by DC3.2 lines to MF2.2D9-Luc CD4+ T cell hybridoma. (H) Normalized MFI of surface MHC-I levels on DC3.2 lines after transducing with pCDH expressing EV or WT IRF2 containing six synonymous mutations within the IRF2 sgRNA target site (IRF2). (I) Cross-presentation after transducing the DC3.2 lines with pCDH expressing EV, WT IRF2 containing six synonymous mutations within the IRF2 sgRNA target site (IRF2 WT), or mutant IRF2 (IRF2 K78R), which also contains the same six synonymous mutations as IRF2 WT. (B–D, H,andI) Bars represent mean + SEM (n $ 3). Statistical analysis by two-tailed ratio paired t tests; (F and G) points represent mean 6 SD of technical duplicates. *p , 0.05, **p , 0.01, ***p , 0.001. ns, not signiﬁcant.

MHC-I was similar to that observed when the expression of the (3T3-Kb) and DC3.2 DCs (13) with vectors expressing either TAP transporter was silenced (Fig. 1C). To determine whether Cas9 and a sgRNA targeting mouse IRF2 or Cas9 alone. The IRF2 also affects surface MHC-I levels in mouse cells, we IRF2-KO mouse fibroblasts and DC cells also had significantly transduced NIH-3T3 fibroblasts stably transfected with H2-Kb reduced surface MHC-I levels (Fig. 1D). Disruption of the IRF2 The Journal of Immunology 5 gene was validated by TIDE analysis (17), and loss of IRF2 be responsible for producing the low MHC-I phenotype ob- expression was confirmed by Western blot (Fig. 1E). DC3.2 cells served in the IRF2-KO cells. RNA-seq was performed on the also express MHC-II molecules, and we found no change in WT and IRF2-KO DC3.2 lines (Fig. 2A, Supplemental Table I). surface MHC-II levels in the IRF2-KO cells (Fig. 1D), which Surprisingly, relatively few genes were differentially expressed demonstrates that IRF2 is selectively affecting the MHC-I by .2-fold (19 decreased and 33 increased). Of these 52 dif- pathway; further evidence supporting this conclusion will be ferentially expressed genes, we identified TAP2, ERAP1, and described below. the immunoproteasome subunit PSME1 as potential contribu- To investigate the functional consequence of this reduction in tors to the decreased MHC-I levels. To further confirm this MHC-I levels, we evaluated the importance of IRF2 for MHC-I result, we performed qPCR in these cell lines to check ex- cross-presentation (i.e., the presentation of peptides derived from pression of all MHC-I pathway genes (Fig. 2B), which showed exogenous Ag on MHC-I). The IRF2-KO DCs cross-presented that the mRNA levels of TAP2, ERAP1, TAPBPR, and PSME9 more poorly than their WT controls (Fig. 1F), demonstrating (another immunoproteasome subunit) were significantly that IRF2 positively regulates MHC-I Ag presentation. In the downregulated in the IRF2-KO DC3.2; PSME1 levels were same experiments, MHC-II presentation was unaffected, again reduced but this decrease did not achieve statistical signi- showing selectivity in IRF2 effects (Fig. 1G). Finally, to confirm ficance. Interestingly, the mRNA levels of the MHC H chain that IRF2 is responsible for these differences, we overexpressed (H2-K1) and MHC L chain (b2-microglobulin; b2m) were IRF2 in the IRF2-KO DC3.2 line and found that it completely unaffected (Fig. 2B), indicating that IRF2 is not required for restored surface MHC-I levels (Fig. 1H) and the ability of these synthesis of the MHC-I heterodimer. ChIP-qPCR experiments cells to cross-present Ag (Fig. 1I). Evidence that loss of IRF2 also confirmed that IRF2 regulates TAP2 and ERAP1 mRNA Downloaded from compromises the MHC-I presentation of endogenous cellular Ags expression by directly binding to their promoters (Fig. 2C). will be described below. To test whether functional IRF2 is needed for TAP2 and To determine whether IRF2 was exerting its function as a tran- ERAP1 mRNA expression, we transduced WT IRF2 or the scription factor, we introduced a lysine to arginine point mutation IRF2-K78R mutant in the DC3.2 IRF2-KO cells and found that at position 78, which prevents acetylation at this site and thereby the TAP2 and ERAP1 mRNA levels were increased in the cells

prevents IRF2 from binding its DNA target sequences (29). expressing WT but not mutant IRF2 (Supplemental Fig. 1). http://www.jimmunol.org/ Overexpressing this mutant IRF2 did not restore function in the To more closely examine the functional effects of IRF2 on IRF2-KO DC3.2 cells (Fig. 1I). Therefore, IRF2 is necessary for peptide transport and trimming, 3T3-Kb IRF2-KO cells were optimal transcriptional regulation of MHC-I Ag presentation transfected with various pTracer plasmids, each of which contained pathways. a GFP reporter and a minigene encoding a peptide that could be processed down to the mature epitope (i.e., SIINFEKL/S8L). Transcriptional control of MHC-I pathway components The cells were surface stained with the Ab 25-D1.16, which by IRF2 recognizes H2-Kb-S8L complexes (19), and transfected cells Because IRF2 is functioning as a transcription factor, we next (GFP-positive) were analyzed. In each case, b2m-knockout cells sought to determine what genes are regulated by IRF2 and could were included to show MHC-I dependence and the maximum by guest on September 27, 2021

FIGURE 2. IRF2-mediated transcriptional regulation of MHC-I pathway under basal conditions. (A) Heatmap of genes (52) differentially expressed by .2-fold between the DC3.2 no sgRNA (WT) and IRF2 sgRNA (KO) lines. Columns represent independent duplicate RNA-seq runs in these two lines. Red → blue indicates high → low expression for a given row (gene). For clarity, only three of the downregulated genes in the KO line (Psme1, Tap2, and Erap1) are shown, and all are listed in Supplemental Table I. (B) mRNA expression levels by qPCR in DC3.2 IRF2-KO relative to those in DC3.2 WT, normalized to the mRNA expression of mouse b-actin in each sample (2^2DDCt). Values .1 indicate higher expression in DC3.2 IRF2-KO, and values ,1 indicate lower expression in the DC3.2 IRF2-KO. Bars represent mean + SEM mRNA expression (n $ 3). Statistical analysis by two-tailed unpaired t tests by comparing the expression of a given gene to that of H2-Ab1 (control). (C) ChIP-qPCR in DC3.2 WT for Tap2 (left) or Erap1 (right) DNA with rabbit anti-IRF2 IgG or normal rabbit IgG (control). Bars represent mean + SEM fold enrichment (2^2DCt) over the normal rabbit IgG control IP (n = 2). Statistical analysis by two-tailed ratio paired t tests. *p , 0.05, **p , 0.01, ***p , 0.001. ns, not significant. 6 FREQUENT LOSS OF IRF2 IN CANCERS LEADS TO IMMUNE EVASION inhibition of S8L presentation, and TAP1-, TAP2-, and ERAP1- also presented fewer H2-Kb-S8L complexes than WT cells when knockout cells were included as positive or negative controls, transfected with ss-N5-S8L (Fig. 3B). However, IRF2-deficient depending on the minigene examined. IRF2-KO cells presented cells were equally capable of presenting H2-Kb-S8L complexes fewer H2-Kb-S8L complexes than WT cells when given the when given a TAP-independent, ERAP1-independent peptide TAP-dependent, ERAP1-dependent Ags CD16-OVA (full-length (S8L with no extra N-terminal residues that was targeted into the OVA protein) (30), N25-S8L (a S8L precursor extended by ER via a colinear signal sequence; ss-S8L) (Fig. 3C), demon- 25 aa on the N terminus), or N5-S8L (a S8L precursor extended by strating that although IRF2 affects the transport and processing of 5 aa on the N terminus) (Fig. 3A). We also tested the presentation MHC-I epitopes, it does not affect the ability of such peptides to of a version of the precursor peptide with five extra N-terminal be loaded onto MHC-I. residues that was targeted into the ER by a colinear signal se- We also compared the MHC-I phenotype in IRF2-KO 3T3-Kb quence (ss-N5-S8L). Because the signal sequence allows this to that observed in the same cells with knockouts of TAP2 peptide to enter the ER through SEC61 instead of TAP, its pre- or ERAP1. The magnitude of the reduction in MHC-I levels on sentation is TAP independent but still dependent on ERAP1 to IRF2-KO cells was in between that of the TAP2- and ERAP1- remove the extra N-terminal residues (Fig. 3B). IRF2-KO cells knockout cells (Fig. 3D), which is consistent with our findings Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 3. IRF2 affects Ag transport and processing. (A–C) H-2Kb presentation of SIINFEKL derived from (A) the TAP-dependent, ERAP1-dependent Ags CD16-OVA (left), N25-S8L (middle), or N5-S8L (right); (B) the TAP-independent, ERAP1-dependent Ag ss-N5-S8L; and (C) the TAP-independent, ERAP1-independent Ag ss-S8L on 3T3-Kb knockout lines. 25-D1.16 staining analyzed on transfected (GFP+) cells. (D) Normalized MFI of surface MHC-I levels on 3T3-Kb (left) or DC3.2 (right) knockout lines. (E) Normalized MFI of surface MHC-I levels on DC3.2 lines 48 h after transduction with pCDH expression vectors containing EV, IRF2, TAP2, ERAP1, or dual transduction of TAP2 and ERAP1. (A–E) Bars represent the mean + SEM of the normalized MFI from independent experiments (n $ 3). Statistical analysis by two-tailed ratio paired t tests. *p , 0.05, ***p , 0.001. ns, not significant. The Journal of Immunology 7 that IRF2 positively regulates TAP2 and ERAP1 but their ex- IRF1 and IRF2 recognize the same IFN-stimulated response pression is not entirely lost in IRF2-KO cells. Finally, we per- element (ISRE) (33–35) and, whereas IRF2 is constitutively formed rescue experiments wherein we overexpressed TAP2 expressed and minimally affected by IFN induction, IRF1 and/or ERAP1 in the IRF2-KO DC3.2 and checked surface is significantly upregulated in response to IFN (Fig. 5B, MHC-I levels 2 d after transduction (Fig. 3E). Although the Supplemental Fig. 2) (36–38). Such upregulation of IRF1 double rescue partially restored MHC-I levels, it was not com- causes IRF1 to compete with IRF2 for binding to the TAP2, plete, suggesting that other genes regulated by IRF2 also con- ERAP1, and PD-L1 promoters, ultimately displacing it from tribute to surface MHC-I expression. them (Fig. 5C). Consistent with the literature that IRF1 positively regulates both MHC-I and PD-L1 expression under IFN- IRF2 and PD-L1 stimulated conditions (39–41), we found that knocking out One of the upregulated genes in the IRF2-KO cells was Cd274 IRF1 decreased both surface MHC-I and PD-L1 levels after (Supplemental Fig. 1), also known as PD-L1, which is often up- IFN-g stimulation (Fig. 5A). Interestingly, when we examined regulated in certain cancers (e.g., NSCLC) and functions as a the dual effects of IRF1 and IRF2 on surface MHC-I and PD-L1 checkpoint inhibitor to suppress Ag-specific CD8+ Tcelleffector levels using single or double knockout DC3.2 lines (Fig. 5A, function (31, 32). To validate our RNA-seq finding with an inde- Supplemental Fig. 2), we found that 1) IRF1/2 double knock- pendent technique, we performed qPCR on the IRF2-KO and WT outs have a larger reduction in surface MHC-I than is observed DC3.2 lines and found that the IRF2-KO cells expressed approxi- in either single knockout; and 2) IRF2-KOs have a larger effect mately twice as much PD-L1 mRNA as the WT controls (Fig. 4A). than the IRF1-knockouts on both MHC-I and PD-L1 expres-

Additionally, ChIP-qPCR revealed that IRF2 regulates PD-L1 sion under basal conditions. This suggested that, although Downloaded from mRNA expression by directly binding the PD-L1 promoter these two IRFs recognize the same ISRE, the subset of genes (Fig. 4B). Therefore, IRF2 acts as a transcriptional repressor of they each primarily regulate differs and that a cell’s depen- PD-L1 in these cells. To evaluate the extent to which a roughly denceonIRF2versusIRF1foranygivengenemayvary 2-fold increase in PD-L1 mRNA translates to surface PD-L1 ex- depending on the cues (e.g., IFN), which that cell receives pression, we analyzed these cells by ﬂow cytometry (Fig. 4C). from its environment. To better characterize these expres-

This analysis reveals that the surface PD-L1 levels also increased sion changes more globally, we performed RNA-seq on the http://www.jimmunol.org/ by roughly 50% in the IRF2-KO cells (Fig. 4D). Taken together, single and double knockout DC3.2 lines under basal conditions these results demonstrate that IRF2 plays a role in repressing (Fig. 5D) or after adding IFN-g (Fig. 5E) or IFN-a (Supplemental PD-L1 expression under basal conditions. Fig. 2). From this analysis, several genes known to be important for Ag presentation and immune cell function were Contributions of IRF1 versus IRF2 identified (Supplemental Table I). Furthermore, this analysis Because IRF2 is an IFN regulatory transcription factor, we showed that genes influenced by IRF1 and IRF2 could be wanted to see how IFN induction would affect IRF2’s regula- grouped into multiple classes. Under basal conditions, there tion of the MHC-I pathway and PD-L1 expression. Interest- was a subset of Ag presentation–related genes whose expres- ingly, stimulation with either IFN-g or IFN-a restored the sion was activated by IRF2 (e.g., TAP2, ERAP1), others that by guest on September 27, 2021 surface MHC-I expression in the IRF2-KO DC3.2 (Fig. 5A, were repressed by IRF2 and remained so in the double Supplemental Fig. 2). This is an important finding because it knockouts (e.g., H2-T9), and yet others that were repressed indicates that impairment of the MHC-I pathway from loss of by IRF2 but did not remain so in the double knockouts (e.g., IRF2 is reversible. PSMB8, PSMB10) (Fig. 5D). After stimulating with IFN-g,

FIGURE 4. IRF2 represses PD-L1 expression under basal conditions. (A) H2-Ab1 and PD-L1 mRNA expression levels by qPCR in DC3.2 IRF2-KO relative to those in DC3.2 WT, normalized to the mRNA expression of mouse b-actin in each sample (2^2DDCt). Values .1 indicate higher expression in DC3.2 IRF2-KO. Bars represent mean + SEM mRNA expression (n = 3). Statistical analysis by two-tailed unpaired t test by comparing the expression of PD-L1 to that of H2-Ab1 (control). (B) ChIP-qPCR of DC3.2 WT for PD-L1 DNA with rabbit anti-IRF2 IgG or normal rabbit IgG (control); bars represent mean + SEM fold enrichment (2^2DCt) over the normal rabbit IgG control IP (n = 2). Statistical analysis by two-tailed ratio paired t test. (C) Representative histograms of surface PD-L1 levels by 10F.9G2 staining in DC3.2 lines stably transduced with the LentiCRISPRv2 constructs; IRF2-KO (IRF2 sgRNA) or WT control (no sgRNA). Background = isotype control staining. (D) Normalized MFI of surface PD-L1 levels on DC3.2 lines; bars represent mean + SEM (n = 6). Statistical analysis by two-tailed ratio paired t test. *p , 0.05, **p , 0.01. 8 FREQUENT LOSS OF IRF2 IN CANCERS LEADS TO IMMUNE EVASION Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 5. Contributions of IRF1 versus IRF2. (A) Normalized MFI of surface MHC-I and PD-L1 levels on the indicated single or double (IRF1 + IRF2, double knockout; DKO) DC3.2 knockout lines after overnight incubation with either medium alone or 2 ng/ml IFN-g; bars represent mean + SEM (n $ 3). (B) Western blot for IRF1 and IRF2 (and b-actin as loading control) in the DC3.2 knockout lines in the absence or presence of 2 ng/ml IFN-g for the durations indicated. (C) ChIP-qPCR of DC3.2 WT cells stimulated with 6 2 ng/ml IFN-g for 2 h for Tap2 (top), Erap1 (middle), or PD-L1 (bottom) DNA with rabbit anti-IRF2 IgG, rabbit anti-IRF1 IgG, or normal rabbit IgG (control). Bars represent mean + SEM fold enrichment (2^2DCt) over the normal rabbit IgG control IP (n = 2). (D and E) heatmap of genes differentially expressed between the DC knockout lines (D) at baseline or (E) after stimulation with 6 2 ng/ml IFN-g for 2 h. For clarity, only a few genes are shown, and all are listed in Supplemental Table I. Statistical analysis by two-tailed ratio paired t tests. *p , 0.05, **p , 0.01, ***p , 0.001, ****p , 0.0001. ns, not signiﬁcant. some genes were primarily activated by IRF1 (e.g., PSME1, studies demonstrate that whereas some genes are acted on an- PSME2, PSMB9), others were primarily repressed by IRF2 tagonistically by IRF1 and IRF2, other genes are regulated (e.g., PD-L1), and yet others were activated by both IRF1 synergistically by these two transcription factors and that the and IRF2 (e.g., TAP2, ERAP1) (Fig. 5E). Collectively, these relative contributions of IRF1 versus IRF2 in mediating these The Journal of Immunology 9 expression changes varies depending on the inﬂammatory state downregulation leads to immune evasion and that there are several of the cell. types of human cancers that may use this escape mechanism.

IRF2 in cancer Discussion Given that experimentally induced loss of IRF2 both compromises Our findings that IRF2 both positively regulates the MHC-I MHC-I presentation and increases PD-L1 expression, it was of pathway and negatively regulates PD-L1 expression have im- interest to see how often IRF2 is downregulated in primary cancers. plications for cancer progression and immunotherapy. Cancers We used the TIMER bioinformatics tool (42) to mine publicly need to evade the immune system and, if they are immunogenic, available databases for IRF2 expression in primary human can- require editing to escape and progress (6). One of the ways that cers. Remarkably, IRF2 was downregulated in several kinds of murine tumors can evade immune elimination is by down- human cancers and the overall reductions were highly statistically regulating the MHC-I pathway (3). Similar immunoediting significant (Fig. 6A). For each of the IRF2-low cancers, which of this pathway occurs in humans as it has been found that included invasive breast carcinoma, cholangiocarcinoma, colon tumors, which were predominantly MHC-I positive at early adenocarcinoma, liver hepatocellular carcinoma, lung adenocar- stages, subsequently become homogeneously MHC-I deficient cinoma and squamous cell carcinoma (NSCLCs), prostate, rectum (46, 47), and progressing cancers are frequently MHC-I defi- and stomach adenocarcinomas, and uterine corpus endometrial cient (48, 49). Another way tumors can evade immune elimi- carcinoma, a subset of patients had very low levels of IRF2. We nation is by expressing checkpoint inhibitors. Indeed, many chose one of the IRF2-low cancers, NSCLC, to determine whether human cancers upregulate PD-L1, and those that do tend to be

IRF2 levels were functionally limiting in primary cancers. At our more aggressive and have fewer T cells present in the tumor Downloaded from institution, NSCLCs are screened for PD-L1 expression by im- (50). Blocking PD-L1 or its receptor (PD-1) can lead to tumor munohistochemistry at the time of diagnosis, which enabled us to rejection in both mice and humans, proving that this is an randomly select tumors spanning a spectrum of PD-L1 expression. important immune evasion mechanism (51, 52). The molecular We extracted RNA from archival patient biopsy material and mechanisms that allow cancers to exploit these immune evasion quantiﬁed expression of IRF2, TAP2, and ERAP1 by qPCR. strategies are incompletely understood but are important to

In these lung cancers, IRF2 mRNA levels and PD-L1 immuno- elucidate for understanding pathogenesis, how they affect http://www.jimmunol.org/ histochemistry status were significantly inversely correlated prognosis and immunotherapy, and how they might be reversed (Fig. 6B). Additionally, consistent with our cell line findings, to improve therapy. TAP2 and ERAP1 mRNA levels positively and significantly cor- Our studies have uncovered an immune evasion mechanism that related with IRF2 mRNA levels (Fig. 6C). To formally test cause is relatively frequent. Many cancers of diverse origins (NSCLC, and effect for these correlations, we analyzed a human NSCLC breast, colorectal, liver, stomach, prostate, uterine) downregulate cell line, A549, that is IRF2-low relative to other NSCLCs tested IRF2 expression, which is a molecule that is not essential for in the NCI-60 panel (43). A549 cells are also MHC-I low and viability in fully differentiated cells. In silico analysis through PD-L1 positive (44, 45). Eliminating the residual IRF2 in A549 cBioPortal (53, 54) revealed that IRF2 is rarely mutated in cancers cells by CRISPR-Cas9–mediated knockout further decreased and, therefore, the loss of expression must be epigenetic. As a by guest on September 27, 2021 surface MHC-I but did not further increase surface PD-L1 result of losing IRF2, tumor cells decrease their surface MHC-I (Fig. 6D). In contrast, restoring IRF2 by transfection repressed expression and increase their surface PD-L1 expression. Thus, surface PD-L1 expression and increased surface MHC-I expres- loss of a single gene can generate a “double whammy” for the sion (Fig. 6D). Stimulation of A549 with IFN-g augmented both immune system, as it enables tumor cells lacking IRF2 to be- surface MHC-I and PD-L1 expression, as expected, but transfec- come both harder to identify (loss of MHC-I Ag presentation) tion of IRF2 still had the same pattern of effects as without IFN and better able to suppress T cell–mediated elimination (in- (repressing PD-L1 and further increasing MHC-I expression) creased checkpoint inhibition). As shown in our in vitro cyto- (Fig. 6E). To further generalize these findings to other cancers, we toxicity assay, the loss of IRF2 renders such tumor cells more analyzed two human breast cancers (chosen because breast can- difficult for CD8+ T cells to kill. cers are often IRF2 low) (Fig. 6A) as well as two mouse sarcomas Analysis of IRF2-KO cells revealed that PD-L1 was highly (chosen because sarcomas are known to undergo immunoediting) upregulated under basal conditions. It is well established (3) and a mouse lymphoma (chosen for use in CD8+ T cell killing that IRF1 can promote IFN-g–inducible activation of PD-L1 assays) and found similar results (Fig. 6F, Supplemental Fig. 3). (40, 41). Recently, Dorand et al. (55) found that hyper- The effects of IRF2 loss on Ag presentation and checkpoint phosphorylation of IRF2BP2 (an IRF2-binding protein) could inhibition are predicted to make it harder for CD8+ T cells to kill lead to decreased PD-L1 expression after IFN-g stimulation. IRF2-low cells. To test this and quantify the magnitude of the Additionally, Wu et al. (56) recently discovered that a loss of effect, we analyzed mouse WT versus IRF2-KO lymphoma IRF2BP2 can lead to enhanced IRF2 binding to the PD-L1 (RMA) cell lines. RMA cells were chosen because they are known promoter and that IFN-g stimulation releases IRF2 from the to be very good targets for cytotoxic T cell killing assays (and are, PD-L1promoter.However,theroleofIRF2onPD-L1ex- therefore, a stringent test) and express H2-Kb, which allows the pression had not been directly examined. Our ChIP-qPCR data use of potent CD8+ T cell effectors from the H2-Kb-S8L–specific shows that IFN-g stimulation leads to both increased IRF1 TCR transgenic OT-I model. Preactivated OT-I effectors were binding and decreased IRF2 binding to the PD-L1 promoter. In cultured with pairs of WT or IRF2-KO cells that were S8L-pulsed addition, we show in this study for the first time, to our or not and labeled with different amounts of the dye CFSE. After knowledge, that, under basal conditions, a loss of IRF2 sig- 4 h, specific killing was quantified by flow cytometry and the nificantly increases PD-L1 mRNA and surface expression, and dose-response curve for the IRF2-KO cells was shifted up ∼3-fold IRF2 overexpression produces the opposite effects. Collec- (which equates to a decreased killing efficiency of ∼67%) in the tively, these results establish the repressive role of IRF2 on PD- IRF2-KO RMA, as compared with the WT RMA (Fig. 6G), L1 expression in the absence and presence of IFN-g and reveal demonstrating that tumor cells lacking IRF2 are harder for CD8+ that unstimulated cells can upregulate PD-L1 expression if T cells to eliminate. Taken together, these findings show that IRF2 upstream regulatory factors, such as IRF2, are defective/absent. 10 FREQUENT LOSS OF IRF2 IN CANCERS LEADS TO IMMUNE EVASION Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 6. IRF2 and cancer. (A) Differential IRF2 expression in tumor and normal tissue from patients with the indicated cancer types (TCGA abbreviations), as queried from TIMER (42). (B)IRF2mRNAexpressioninpatientNSCLCspecimens scored as PD-L1 low (1–50%) or high (.50%) by immunohistochemistry. IRF2 mRNA expression was normalized to GAPDH mRNA expression in each lung specimen, and then IRF2 expression across specimens was compared by calculating fold changes over the lowest IRF2-expressing specimen, which was set equal to 1. Mann–Whitney U test. (C) TAP2 (top) and ERAP1 (bottom) mRNA correlations with IRF2 mRNA in NSCLC specimens. Linear regression models shown with R2 for goodness of fit. (D and E)geometricMFIofsurfaceMHC-I(left) and PD-L1 (right) on (D) unstimulated or (E) overnight IFN-g–stimulated A549 lines knocked out for IRF2 (IRF2-KO) or control (WT) or overexpressing IRF2 or EV as control. Bars represent mean + SEM (n = 3). Two-tailed ratio paired t tests. (F) Geometric MFI of surface MHC-I on unstimulated or IFN-g–stimulated human breast carcinoma (BT20 and MCF7) lines knocked out for IRF2 (IRF2-KO) or control (WT). Bars represent mean + SEM (n = 3). Two-tailed ratio paired t tests. (G) Preactivated OT-I effectors were cultured with pairs of WT (WT RMA) or IRF2-KO (IRF2-KO RMA) cells that were S8L-pulsed or not and labeled with different amounts of the dye CFSE. After 4 h, specific killing was quantified by flow cytometry. Left, Percentage of specific killing dose-titration curve for one representative experiment using an E:T ratio of 1:2 and [S8L] indicated, points show mean 6 SD of technical triplicates. Right, Percentage of specific killing of IRF2-KO relative to that of WT RMA at 10 pM S8L. Bars represent mean + SEM (n = 4). Two-tailed ratio paired t test. *p , 0.05, **p , 0.01, ***p , 0.001. The Journal of Immunology 11

The transcriptional control of many MHC-I pathway compo- indications) or IFN-inducing agents could be used to restore nents, particularly under basal conditions, has not been well defined MHC-I Ag presentation in IRF2-low tumors. This would be pre- (57). In this study, we identified IRF2 as a novel positive regulator dicted to enhance the effects of immunotherapies, such as check- of MHC-I Ag presentation and show that, under basal conditions, point blockade, that are ultimately dependent on TCR recognition it is important for both classical MHC-I presentation and cross- of tumor MHC-I presentation. Currently, checkpoint therapy is ef- presentation. IRF2 binds the promoters of TAP2 and ERAP1 fective in only some patients and, based on the mechanisms we and transcriptionally activates their expression. In the absence of have uncovered, it is conceivable that reversing the IRF2 defects IRF2, cells express less TAP2 and ERAP1 and, because of the might increase the number of patients that can benefit. In addition, consequent defects in Ag transport and processing, present fewer because checkpoint blockade is an extremely expensive therapy peptide-MHC-I complexes at the cell surface. Additionally, IRF2 and one that can have serious side effects, there is a need for regulates the expression of many other genes, as demonstrated by good biomarkers to identify those patients that would be more others (58, 59) and shown in this study, such as immunoprotea- likely to benefit from this type of therapy. Therefore, it will be of some subunits, which likely also influence the MHC-I presentation interest to examine in future studies whether expression of IRF2 in IRF2-deficient cells (60). Interestingly, in one of these earlier and its downstream target genes (e.g., TAP2 and ERAP1) could studies, IRF2 was reported to bind to the promoter of TAP1 as be used as biomarkers to help identify checkpoint blockade- assessed by EMSAs (59); however, we found no evidence that responsive patients. IRF2 was needed for transcription of endogenous TAP1 by either RNA-seq or qPCR. Most likely, this discrepancy is due to the Acknowledgments nature of the different assays (e.g., extract binding to exogenous We thank Kate Fitzgerald for helpful discussions, Janice BelleIsle for tech- Downloaded from probes versus endogenous promoter activity). nical assistance, and Leslie Shaw and Robert Schreiber for sharing cells and In the DC line where we focused most of our studies and in reagents. other cell lines described in the literature, IRF2 is constitutively expressed, relatively stable with a t1/2 of 8 h, and minimally Disclosures affected by IFN (36–38). However, in these same sets of cells, The authors have no financial conflicts of interest.

IRF1 is minimally expressed under basal conditions but strongly http://www.jimmunol.org/ induced in the presence of IFN and more short-lived than IRF2, References with a t1/2 of only 0.5 h (36). Because of these differences between IRF2 and IRF1, even though they both recognize and bind 1. Kaplan, D. H., V. Shankaran, A. S. Dighe, E. Stockert, M. Aguet, L. J. Old, and R. D. Schreiber. 1998. Demonstration of an interferon gamma-dependent tumor to the same ISRE (33–35), we hypothesized that the relative surveillance system in immunocompetent mice. Proc. Natl. Acad. Sci. USA 95: contributions of IRF2 and IRF1 to certain signaling pathways 7556–7561. 2. Street, S. E., E. Cretney, and M. J. Smyth. 2001. Perforin and interferon-gamma varies depending on the inflammatory state of the cell. Although activities independently control tumor initiation, growth, and metastasis. Blood IRF1 and IRF2 were originally characterized as an activator and 97: 192–197. repressor of IFN-a/b expression, respectively (35), several re- 3. Shankaran, V., H. Ikeda, A. T. Bruce, J. M. White, P. E. Swanson, L. J. Old, and R. D. Schreiber. 2001. IFNgamma and lymphocytes prevent primary tumour by guest on September 27, 2021 ports since then have highlighted that IRF1 and IRF2 do not development and shape tumour immunogenicity. Nature 410: 1107–1111. always act as such (37, 61, 62). Yet, to the best of our knowl- 4. Svane, I. M., A. M. Engel, M. B. Nielsen, H. G. Ljunggren, J. Rygaard, and edge, there have been no global differential expression analyses O. Werdelin. 1996. Chemically induced sarcomas from nude mice are more immunogenic than similar sarcomas from congenic normal mice. Eur. in IRF1- and/or IRF2-lacking cells of the same cell type to better J. Immunol. 26: 1844–1850. understand the synergistic versus antagonistic roles these tran- 5. Engel, A. M., I. M. Svane, J. Rygaard, and O. Werdelin. 1997. MCA sarcomas induced in scid mice are more immunogenic than MCA sarcomas induced in scription factors perform. Here, we conducted RNA-seq on congenic, immunocompetent mice. Scand. J. Immunol. 45: 463–470. IRF1- and/or IRF2-KO DCs in the absence or presence of IFN to 6. Mittal, D., M. M. Gubin, R. D. Schreiber, and M. J. Smyth. 2014. New insights help fill this void. Globally, we found that the expression profile into cancer immunoediting and its three component phases--elimination, equi- librium and escape. Curr. Opin. Immunol. 27: 16–25. of the double knockout DCs differed from that of either IRF 7. Sabbatino, F., V. Villani, J. H. Yearley, V. Deshpande, L. Cai, I. T. Konstantinidis, single knockout and that the relative contributions of IRF1 C. Moon, S. Nota, Y. Wang, A. Al-Sukaini, et al. 2016. PD-L1 and HLA class I versus IRF2, in terms of their ability to positively or negatively antigen expression and clinical course of the Disease in intrahepatic cholangiocarcinoma. Clin. Cancer Res. 22: 470–478. regulate certain genes, varied depending on the inflammatory 8. Gettinger, S., J. Choi, K. Hastings, A. Truini, I. Datar, R. Sowell, A. Wurtz, state of the cell. Transcriptional control of genes can be highly W. Dong, G. Cai, M. A. Melnick, et al. 2017. Impaired HLA class I antigen processing and presentation as a mechanism of acquired resistance to immune cell-type dependent, and therefore, it is possible that in other checkpoint inhibitors in lung cancer. Cancer Discov. 7: 1420–1435. cells, the contribution of IRF2 on surface MHC-I and PD-L1 9. Abele, R., and R. Tampe´. 2004. The ABCs of immunology: structure and expression may vary. However, our concordant observations in function of TAP, the transporter associated with antigen processing. Physiology (Bethesda) 19: 216–224. multiple mouse and human cell lines and primary tumors suggest 10. Saric, T., S. C. Chang, A. Hattori, I. A. York, S. Markant, K. L. Rock, that the number of cell types in which IRF2 regulates these genes M. Tsujimoto, and A. L. Goldberg. 2002. An IFN-gamma-induced aminopep- in this manner is quite large. tidase in the ER, ERAP1, trims precursors to MHC class I-presented peptides. Nat. Immunol. 3: 1169–1176. Immune evasion due to tumor PD-L1 upregulation can be re- 11. Serwold, T., F. Gonzalez, J. Kim, R. Jacob, and N. Shastri. 2002. ERAAP versed by blocking the PD-L1/PD-1 interaction, which is the basis customizes peptides for MHC class I molecules in the endoplasmic reticulum. Nature 419: 480–483. of targeted checkpoint blockade immunotherapy. However, any 12. Zamora, A. E., J. C. Crawford, and P. G. Thomas. 2018. Hitting the target: how + natural or invigorated (from checkpoint blockade) CD8 T cell T cells detect and eliminate tumors. J. Immunol. 200: 392–399. response to kill tumors that have downregulated their MHC-I 13. Shen, Z., G. Reznikoff, G. Dranoff, and K. L. Rock. 1997. Cloned dendritic cells can present exogenous antigens on both MHC class I and class II molecules. expression will continue to be impaired. In this context, it is of J. Immunol. 158: 2723–2730. considerable interest and potential importance that the downreg- 14. Rock, K. L., L. Rothstein, and S. Gamble. 1990. Generation of class I ulation of the MHC-I pathway from the loss of IRF2 is reversible. MHC-restricted T-T hybridomas. J. Immunol. 145: 804–811. 15. Sanjana, N. E., O. Shalem, and F. Zhang. 2014. Improved vectors and genome- When IRF2-deficient cells were treated with IFN, MHC-I levels wide libraries for CRISPR screening. Nat. Methods 11: 783–784. were restored, likely because of induction of IRF1 and possibly 16. Shalem, O., N. E. Sanjana, E. Hartenian, X. Shi, D. A. Scott, T. Mikkelson, D. Heckl, B. L. Ebert, D. E. Root, J. G. Doench, and F. Zhang. 2014. some other transcriptional activators. These findings suggest that Genome-scale CRISPR-Cas9 knockout screening in human cells. Science IFNs (which are Food and Drug Administration approved for other 343: 84–87. 12 FREQUENT LOSS OF IRF2 IN CANCERS LEADS TO IMMUNE EVASION

17. Brinkman, E. K., T. Chen, M. Amendola, and B. van Steensel. 2014. Easy Interferon receptor signaling pathways regulating PD-L1 and PD-L2 expression. quantitative assessment of genome editing by sequence trace decomposition. Cell Rep. 19: 1189–1201. Nucleic Acids Res. 42: e168. 42. Li, B., E. Severson, J. C. Pignon, H. Zhao, T. Li, J. Novak, P. Jiang, H. Shen, 18. Hearn, A., I. A. York, and K. L. Rock. 2009. The specificity of trimming of J. C. Aster, S. Rodig, et al. 2016. Comprehensive analyses of tumor immunity: MHC class I-presented peptides in the endoplasmic reticulum. J. Immunol. 183: implications for cancer immunotherapy. Genome Biol. 17: 174. 5526–5536. 43. Rajapakse, V. N., A. Luna, M. Yamade, L. Loman, S. Varma, M. Sunshine, 19. Porgador, A., J. W. Yewdell, Y. Deng, J. R. Bennink, and R. N. Germain. 1997. F. Iorio, F. G. Sousa, F. Elloumi, M. I. Aladjem, et al. 2018. CellMinerCDB for Localization, quantitation, and in situ detection of specific peptide-MHC class I integrative cross-database genomics and pharmacogenomics analyses of cancer complexes using a monoclonal antibody. Immunity 6: 715–726. cell lines. iScience 10: 247–264. 20. Bolger, A. M., M. Lohse, and B. Usadel. 2014. Trimmomatic: a flexible trimmer 44. Teo, J., A. Mirenska, M. Tan, Y. Lee, J. Oh, L. Z. Hong, R. Wnek, Y. S. Yap, for Illumina sequence data. Bioinformatics 30: 2114–2120. S. J. Shih, A. A. S Bhagat, et al. 2017. A preliminary study for the assessment of 21. Li, B., and C. N. Dewey. 2011. RSEM: accurate transcript quantification from PD-L1 and PD-L2 on circulating tumor cells by microfluidic-based chipcy- RNA-Seq data with or without a reference genome. BMC Bioinformatics 12: 323. tometry. Future Sci. OA 3: FSO244. 22. Pruitt, K. D., T. Tatusova, G. R. Brown, and D. R. Maglott. 2012. NCBI Ref- 45. Redondo, M., F. Ruiz-Cabello, A. Concha, T. Cabrera, M. Pe´rez-Ayala, erence Sequences (RefSeq): current status, new features and genome annotation M. R. Oliva, and F. Garrido. 1991. Altered HLA class I expression in non-small policy. Nucleic Acids Res. 40(Database issue): D130–D135. cell lung cancer is independent of c-myc activation. Cancer Res. 51: 2463–2468. 23. Johnson, W. E., C. Li, and A. Rabinovic. 2007. Adjusting batch effects 46. Carretero, R., J. M. Romero, F. Ruiz-Cabello, I. Maleno, F. Rodriguez, in microarray expression data using empirical Bayes methods. Biostatistics F. M. Camacho, L. M. Real, F. Garrido, and T. Cabrera. 2008. Analysis of HLA 8: 118–127. class I expression in progressing and regressing metastatic melanoma lesions 24. Leek, J. T., W. E. Johnson, H. S. Parker, A. E. Jaffe, and J. D. Storey. 2012. The after immunotherapy. Immunogenetics 60: 439–447. sva package for removing batch effects and other unwanted variation in high- 47. del Campo, A. B., J. A. Kyte, J. Carretero, S. Zinchencko, R. Meńdez, throughput experiments. Bioinformatics 28: 882–883. G. Gonza´lez-Aseguinolaza, F. Ruiz-Cabello, S. Aamdal, G. Gaudernack, 25. Love, M. I., W. Huber, and S. Anders. 2014. Moderated estimation of fold F. Garrido, and N. Aptsiauri. 2014. Immune escape of cancer cells with beta2- change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15: 550. microglobulin loss over the course of metastatic melanoma. Int. J. Cancer 134: 26. Mould, A. W., M. A. Morgan, A. C. Nelson, E. K. Bikoff, and E. J. Robertson. 102–113. 2015. Blimp1/Prdm1 functions in opposition to Irf1 to maintain neonatal toler- 48. Marincola, F. M., E. M. Jaffee, D. J. Hicklin, and S. Ferrone. 2000. Escape of Downloaded from ance during postnatal intestinal maturation. PLoS Genet. 11: e1005375. human solid tumors from T-cell recognition: molecular mechanisms and func- 27. Doody, G. M., S. Stephenson, C. McManamy, and R. M. Tooze. 2007. PRDM1/ tional significance. Adv. Immunol. 74: 181–273. BLIMP-1 modulates IFN-gamma-dependent control of the MHC class I antigen- 49. Hicklin, D. J., F. M. Marincola, and S. Ferrone. 1999. HLA class I antigen processing and peptide-loading pathway. J. Immunol. 179: 7614–7623. downregulation in human cancers: T-cell immunotherapy revives an old story. 28. Lu, C., P. S. Redd, J. R. Lee, N. Savage, and K. Liu. 2016. The expression Mol. Med. Today 5: 178–186. profiles and regulation of PD-L1 in tumor-induced myeloid-derived suppressor 50. Zou, W., and L. Chen. 2008. Inhibitory B7-family molecules in the tumour cells. OncoImmunology 5: e1247135. microenvironment. Nat. Rev. Immunol. 8: 467–477.

29. Masumi, A., Y. Yamakawa, H. Fukazawa, K. Ozato, and K. Komuro. 2003. 51. Pardoll, D. M. 2012. The blockade of immune checkpoints in cancer immuno- http://www.jimmunol.org/ Interferon regulatory factor-2 regulates cell growth through its acetylation. therapy. Nat. Rev. Cancer 12: 252–264. J. Biol. Chem. 278: 25401–25407. 52. Wang, X., Z. Bao, X. Zhang, F. Li, T. Lai, C. Cao, Z. Chen, W. Li, H. Shen, and 30. Fernandes, D. M., L. Vidard, and K. L. Rock. 2000. Characterization of MHC S. Ying. 2017. Effectiveness and safety of PD-1/PD-L1 inhibitors in the treat- class II-presented peptides generated from an antigen targeted to different ment of solid tumors: a systematic review and meta-analysis. Oncotarget 8: endocytic compartments. Eur. J. Immunol. 30: 2333–2343. 59901–59914. 31. Dong, H., S. E. Strome, D. R. Salomao, H. Tamura, F. Hirano, D. B. Flies, 53. Cerami, E., J. Gao, U. Dogrusoz, B. E. Gross, S. O. Sumer, B. A. Aksoy, P. C. Roche, J. Lu, G. Zhu, K. Tamada, et al. 2002. Tumor-associated B7-H1 A. Jacobsen, C. J. Byrne, M. L. Heuer, E. Larsson, et al. 2012. The cBio cancer promotes T-cell apoptosis: a potential mechanism of immune evasion. [Published genomics portal: an open platform for exploring multidimensional cancer ge- erratum appears in 2002 Nat. Med. 8: 1039.] Nat. Med. 8: 793–800. nomics data. Cancer Discov. 2: 401–404. 32. Yu, H., T. A. Boyle, C. Zhou, D. L. Rimm, and F. R. Hirsch. 2016. PD-L1 ex- 54. Gao, J., B. A. Aksoy, U. Dogrusoz, G. Dresdner, B. Gross, S. O. Sumer, Y. Sun, pression in lung cancer. [Published erratum appears in 2017 J. Thorac. Oncol. A. Jacobsen, R. Sinha, E. Larsson, et al. 2013. Integrative analysis of complex

12: 157–159.] J. Thorac. Oncol. 11: 964–975. cancer genomics and clinical proﬁles using the cBioPortal. Sci. Signal. 6: pl1. by guest on September 27, 2021 33. Tanaka, N., T. Kawakami, and T. Taniguchi. 1993. Recognition DNA sequences 55. Dorand, R. D., J. Nthale, J. T. Myers, D. S. Barkauskas, S. Avril, of interferon regulatory factor 1 (IRF-1) and IRF-2, regulators of cell growth and S. M. Chirieleison, T. K. Pareek, D. W. Abbott, D. S. Stearns, J. J. Letterio, et al. the interferon system. Mol. Cell. Biol. 13: 4531–4538. 2016. Cdk5 disruption attenuates tumor PD-L1 expression and promotes anti- 34. Darnell, J. E., Jr., I. M. Kerr, and G. R. Stark. 1994. Jak-STAT pathways and tumor immunity. Science 353: 399–403. transcriptional activation in response to IFNs and other extracellular signaling 56. Wu, A., Q. Wu, Y. Deng, Y. Liu, J. Lu, L. Liu, X. Li, C. Liao, B. Zhao, and proteins. Science 264: 1415–1421. H. Song. 2019. Loss of VGLL4 suppresses tumor PD-L1 expression and immune 35. Harada, H., T. Fujita, M. Miyamoto, Y. Kimura, M. Maruyama, A. Furia, evasion. EMBO J. 38. T. Miyata, and T. Taniguchi. 1989. Structurally similar but functionally 57. Jongsma, M. L. M., G. Guarda, and R. M. Spaapen. 2017. The regulatory net- distinct factors, IRF-1 and IRF-2, bind to the same regulatory elements of work behind MHC class I expression. Mol. Immunol. DOI: 10.1016/j.molimm. IFN and IFN-inducible genes. Cell 58: 729–739. 2017.12.005. 36. Watanabe, N., J. Sakakibara, A. G. Hovanessian, T. Taniguchi, and T. Fujita. 58. Fairfax, B. P., P. Humburg, S. Makino, V. Naranbhai, D. Wong, E. Lau, 1991. Activation of IFN-beta element by IRF-1 requires a posttranslational event L. Jostins, K. Plant, R. Andrews, C. McGee, and J. C. Knight. 2014. Innate in addition to IRF-1 synthesis. Nucleic Acids Res. 19: 4421–4428. immune activity conditions the effect of regulatory variants upon monocyte gene 37. Ren, G., K. Cui, Z. Zhang, and K. Zhao. 2015. Division of labor between IRF1 expression. Science 343: 1246949. and IRF2 in regulating different stages of transcriptional activation in cellular 59. Rouyez, M. C., M. Lestingi, M. Charon, S. Fichelson, A. Buzyn, and I. Dusanter- antiviral activities. Cell Biosci. 5: 17. Fourt. 2005. IFN regulatory factor-2 cooperates with STAT1 to regulate trans- 38. Oshima, S., T. Nakamura, S. Namiki, E. Okada, K. Tsuchiya, R. Okamoto, porter associated with antigen processing-1 promoter activity. J. Immunol. 174: M. Yamazaki, T. Yokota, M. Aida, Y. Yamaguchi, et al. 2004. Interferon regu- 3948–3958. latory factor 1 (IRF-1) and IRF-2 distinctively up-regulate gene expression and 60. Kincaid, E. Z., J. W. Che, I. York, H. Escobar, E. Reyes-Vargas, J. C. Delgado, production of interleukin-7 in human intestinal epithelial cells. Mol. Cell. Biol. R. M. Welsh, M. L. Karow, A. J. Murphy, D. M. Valenzuela, et al. 2011. Mice 24: 6298–6310. completely lacking immunoproteasomes show major changes in antigen pre- 39. Hobart, M., V. Ramassar, N. Goes, J. Urmson, and P. F. Halloran. 1997. IFN sentation. Nat. Immunol. 13: 129–135. regulatory factor-1 plays a central role in the regulation of the expression of class 61. Jesse, T. L., R. LaChance, M. F. Iademarco, and D. C. Dean. 1998. Inter- I and II MHC genes in vivo. J. Immunol. 158: 4260–4269. feron regulatory factor-2 is a transcriptional activator in muscle where it 40. Lee, S. J., B. C. Jang, S. W. Lee, Y. I. Yang, S. I. Suh, Y. M. Park, S. Oh, regulates expression of vascular cell adhesion molecule-1. J. Cell Biol. 140: J. G. Shin, S. Yao, L. Chen, and I. H. Choi. 2006. Interferon regulatory factor-1 is 1265–1276. prerequisite to the constitutive expression and IFN-gamma-induced upregulation 62. Vaughan, P. S., F. Aziz, A. J. van Wijnen, S. Wu, H. Harada, T. Taniguchi, of B7-H1 (CD274). FEBS Lett. 580: 755–762. K. J. Soprano, J. L. Stein, and G. S. Stein. 1995. Activation of a cell-cycle- 41. Garcia-Diaz, A., D. S. Shin, B. H. Moreno, J. Saco, H. Escuin-Ordinas, regulated histone gene by the oncogenic transcription factor IRF-2. Nature 377: G. A. Rodriguez, J. M. Zaretsky, L. Sun, W. Hugo, X. Wang, et al. 2017. 362–365.

A B

Fig. S1. Effects of IRF2 deficiency on antigen presentation machinery and PD-L1 expression

(A) TAP1, TAP2, or ERAP1 mRNA expression levels by qPCR in DC3.2 IRF2 KO 24hrs after overexpressing pCDH wild-type IRF2 (IRF2 WT) or mutant IRF2 (IRF2 K78R) relative to overexpressing pCDH IRF2 K78R; normalized to the mRNA expression of mouse β-actin in each sample (2^-ΔΔCt). Values >1 indicate higher expression than overexpressing IRF2 K78R and values<1 indicate lower expression than overexpressing IRF2 K78R. Representative experiment shown; bars represent mean ± SEM of mRNA expression of duplicate technical replicates; (B) Geometric MFI of surface MHC-I on D53m (top) or H50m (bottom) mouse sarcoma lines transduced with empty vector (EV), wild-type IRF2 (IRF2 WT), or mutant IRF2 (IRF2 K78R). Representative experiment shown; bars represent mean ± SD of technical duplicates. (C) PD-L1 mRNA expression in the DC3.2 no sgRNA (“WT”) and DC3.2 IRF2 sgRNA (“IRF2 KO”) lines. Bars represent TPM from 3 independent RNA-seq replicates. A B DC3.2 IFN DC3.2 IFN Surface MHC-I Surface PD-L1

150 150 * )

) ns T

* T W

ns W

f * o

* o

100 100

(

I F

z i

50 i 50

N N 0 0 WT IRF1 IRF2 DKO WT IRF1 IRF2 DKO KO KO KO KO

Fig. S2. Contributions of IRF1 vs. IRF2 after IFN stimulation

(A) Normalized MFI of surface MHC I and PD-L1 levels on DC3.2 stable knockout lines after overnight incubation with 5,000U/mL IFNα; bars represent mean + SEM (N=3). Statistical analysis by two-tailed ratio paired t-tests; (B) Western blots of IRF1 and IRF2 (and β-actin as loading control) in DC3.2 KO lines in the absence or presence of 5,000U/mL IFNα for the durations indicated; (C) RNA-seq of DC3.2 no sgRNA (WT), IRF2 sgRNA (IRF2 KO), IRF1 sgRNA (IRF1 KO), or IRF1 + IRF2 sgRNAs (DKO) after stimulation for 2hrs with 5,000U/mL IFNα or media alone; (D-F) Second RNA-seq replicate of Fig. 5d, 5e, and Fig. S2c, respectively. (C-F) Genes from heatmaps listed in Table S1.

A B BT20 PD-L1 MCF7 PD-L1 RMA MHC-I RMA PD-L1 IFN IFN No IFN No IFN 3000 3000 15000 250 200

2000 I

2000 I I

I 10000

F F

F 150

M G

G 100 G 1000 1000 G 5000 50

0 0 0 0 WT IRF2 WT IRF2 WT IRF2 WT IRF2 KO KO KO KO

BT20 MHC-I MCF7 MHC-I D53m MHC-I H50m MHC-I C No IFN No IFN No IFN No IFN 40000 40000 2500 3000

30000 30000 2000

2000

I I

I 1500

F F

F M

M 20000 20000

G G G 1000 1000 10000 10000 500

0 0 0 0 pCDH pCDH pCDH pCDH pCDH pCDH pCDH pCDH EV IRF2 EV IRF2 EV IRF2 EV IRF2

D53m PD-L1 H50m PD-L1

D IFN IFN 1500 1500

1000 1000

I F

G G 500 500

0 0 pCDH pCDH pCDH pCDH EV IRF2 EV IRF2

Fig. S3. Effects of IRF2 knockout or overexpression on other cell lines

(A-D) Geometric MFI of (A) surface PD-L1 on IFNγ-stimulated BT20 and MCF7 knockout lines; (B) surface MHC-I (left) or surface PD-L1 (right) on unstimulated RMA lymphoma knockout lines; (C) surface MHC-I on (left) unstimulated BT20 and MCF7 lines or (right) unstimulated D53m and H50m mouse sarcoma lines overexpressing IRF2 (pCDH IRF2) or empty vector control (pCDH EV); and (D) surface PD-L1 on IFNγ-stimulated D53m and H50m overexpression lines. Representative experiments shown, error bars represent staining of technical replicates.