Interferon Receptor Signaling Pathways Regulating PD-L1 and PD-L2 Expression

Abstract

PD-L1 and PD-L2 are ligands for the PD-1 immune inhibiting checkpoint that can be induced in tumors by interferon exposure, leading to immune evasion. This process is important for immunotherapy based on PD-1 blockade. We examined the specific molecules involved in interferon-induced signaling that regulates PD-L1 and PD-L2 expression in melanoma cells. These studies revealed that the interferon-gamma-JAK1/JAK2-STAT1/STAT2/STAT3-IRF1 axis primarily regulates PD-L1 expression, with IRF1 binding to its promoter. PD-L2 responded equally to interferon beta and gamma and is regulated through both IRF1 and STAT3, which bind to the PD-L2 promoter. Analysis of biopsy specimens from patients with melanoma confirmed interferon signature enrichment and upregulation of gene targets for STAT1/STAT2/STAT3 and IRF1 in anti-PD-1-responding tumors. Therefore, these studies map the signaling pathway of interferon-gamma-inducible PD-1 ligand expression.

Keywords: IRF1; JAK-STATs; PD-1; PD-L1; PD-L2; immunotherapy; interferon receptor signaling pathways; melanoma.

Figure 1. Induction of PD-L1 by Interferon Alpha, Beta, and Gamma

(A–C) Flow cytometry analysis of PD-L1 surface expression upon interferon treatment in the human melanoma cell lines M244 (A), M263 (B), and M381 (C) exposed to interferon alpha, beta, or gamma for 18 hr. Histograms represent changes in mean fluorescence intensity by flow cytometry compared to baseline. (D–F) Western blot analysis of interferon receptor signaling proteins in M244 (D), M263 (E), and M381 (F), including a set of proteins involved in interferon signaling pathways. Basal and activated (phosphorylated) states of the proteins are included to compare the induction trough these different mediators. The first lane of each panel represents untreated control cells. Each cell line was exposed 30 min or 18 hr with interferon alpha, beta, or gamma, respectively.

Figure 2. Effects on PD-L1 Reporter Expression upon shRNA Silencing of 33 Genes Involved in the Interferon Signaling Pathway

(A) Schematic representation of the PD-L1Prom-DSRed-FireflyLuciferase/Neo and EF1AProm-Renilla luciferase/RSV-BSD constructs used to generate the reporter melanoma cell lines. Reporter cells contain the PD-L1 promoter driving the expression of a DSRedDR-T2A-Firefly luciferase cassette and also an EF1-alpha promoter driving the Renilla luciferase gene used for normalization. (B–D) Normalized luciferase reporter expression of each cell line, M244 (B), M263 (C), and M381 (D), transduced with different sets of lentiviral shRNA hairpins. Black and checkered white bars represent the cells transduced with a lentiviral control containing no shRNA, with (black) or without (checkered) interferon gamma treatment. Gray bars represent the expression level of the cells transduced with a set of lentiviral shRNA hairpins with interferon gamma treatment to compare changes upon the interferon gamma induction of PD-L1 expression. Results are represented as a percentage of luciferase expression compared with the interferon-gamma-treated negative control. (E) Schematic representation of the interferon receptor signaling pathway depicting the hits taking into account the redundant siRNA/shRNA activity (RSA) score obtained for each factor in the three reporter cell lines. Color heatmap represents the summed rank score, with red indicating the greatest impact on interferon-gamma-induced PD-L1 reporter activity and dark blue the least impact. Spatial orientation places each gene in the context of its signaling pathway.

Figure 3. Transient Luciferase Reporter Assays and ChIP Analysis for the PD-L1 Promoter in M381 Melanoma Cells

(A) Sequence of the PD-L1 promoter showing the position of the most representative putative binding sites of the promoter, STAT1/STAT3, STAT2/STAT5, and IRF1. (B) PD-L1 promoter transient reporter assay including deletions of the putative binding sites. Results are represented as normalized relative luciferase units (RLUs). (C) ChIP assay in M381 cells at the PD-L1 promoter (gray), the HLA-B promoter as a positive control (white), and the human tRNA-Leu anti-codon (TAG) as irrelevant sequence for IRF1 binding (negative control). Results are represented as percent enrichment relative to input. Asterisks denote significance in an unpaired t test (*p < 0.005, **p < 0.001), and error bars denote SD.

Figure 4. Selected mRNA Expression Profiling of Interferon-Gamma-Induced Genes

(A–C) Changes in selected gene expression upon interferon gamma exposure in three melanoma cell lines: M244 (A), M263 (B), and M381 (C). Purple and blue dots represent the expression level of the transcripts at basal level in two different biological duplicates (untreated cells), and red and green dots represent the expression level of the transcripts after 3 hr of interferon gamma treatment. (D) Expression profile analysis of the good interferon gamma responder M233 cell line using the JAK2-specific inhibitor CEP33779. Dark blue dots represent untreated cells, green dots represent interferon-gamma-treated cells, and red dots represent cells treated with the JAK2 inhibitor CEP-33779 and interferon gamma. Data are presented as normalized counts (log2) of each transcript. (E) Same analysis of the JAK2 mutant cell line M368 with (red) or without (light blue) interferon gamma treatment. (F) Expression profile of the JAK1 mutant cell line M395 with (red) or without (light blue) interferon gamma treatment.

Figure 5. PD-L2 Expression and Promoter Function Analysis

(A–C) Flow cytometry analysis of PD-L2 surface expression upon interferon treatment in M244 (A), M263 (B), and M381 (C). (D) Sequence of the PD-L2 promoter and position of the putative transcription factor binding sites. (E) Transient reporter assay including deletions of STAT1/STAT3, IRF1, or double mutations for the putative binding sites. Analysis was performed in untreated cells (white bars) and under interferon beta (gray bars) and interferon gamma exposure (black bars). (F) ChIP assay using IRF1 antibody in interferon-gamma-treated cells, including primers for the PD-L2 promoter (gray), HLA-B promoter as positive control (white), and TAG gene (black) as an irrelevant sequence for IRF1 binding. (G) ChIP assay using STAT3 antibody in interferon-beta-treated cells, including primers for the C-FOS promoter as a positive control (white), the PD-L2 promoter (gray), the PD-L1 promoter (checkered), and TAG gene (black) as an irrelevant sequence for STAT3 binding. Asterisks denote significance in an unpaired t test (*p < 0.05, **p < 0.005, ***p < 0.001), and error bars denote SD.

Figure 6. RNA-Seq Analysis of TCGA Tumors and Anti-PD-1-Treated Biopsies

(A) Pearson correlations between log2 normalized mRNA expression levels (RPKM) of IRF1 and STAT1 versus PD-L1 in the TCGA skin cutaneous melanoma RNA-seq database. (B) Pearson correlations between log2 normalized mRNA expression levels (RPKM) of IRF1 and STAT1 versus PD-L2 in the TCGA skin cutaneous melanoma RNA-seq database. (C) Potential target genes of IRF1, STAT1, STAT3, and STAT1:STAT2 heterodimer among all up-expressed genes in anti-PD1 on-treatment tumor samples (upexpression is defined by fold change ≥1.5). Gray lines indicate the occurrence of the respective transcription factor binding motifs in the target genes. (D) Tilling of differential immune cell marker genes in on-treatment anti-PD1 tumors compared to their respective baselines (red and green, up- and down-regulation of at least 1.5 (log2) fold changes. (E) Tiling of differential IFNG expression (red and green, up- and down-expression) and interferon signature enrichment (orange and blue, positive and negative enrichments) in on-treatment tumors derived from patients receiving anti-PD1 treatment (red, patients who responded to treatment; black, patients who did not respond to treatment).