Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
[Preprint]. 2023 Jun 1:rs.3.rs-2960521.
doi: 10.21203/rs.3.rs-2960521/v1.

Plasticity-induced repression of Irf6 underlies acquired resistance to cancer immunotherapy

Il-Kyu Kim  1 , Mark Diamond  1 , Salina Yuan  1 , Samantha Kemp  1 , Qinglan Li  1 , Jeffrey Lin  1 , Jinyang Li  1 , Robert Norgard  1 , Stacy Thomas  1 , Maria Merolle  1 , Takeshi Katsuda  1 , John Tobias  1 , Katerina Politi  2 , Robert Vonderheide  1 , Ben Stanger  1
Affiliations

Plasticity-induced repression of Irf6 underlies acquired resistance to cancer immunotherapy

Il-Kyu Kim et al. Res Sq. .

Update in

Abstract

Acquired resistance to immune checkpoint immunotherapy remains a critical yet incompletely understood biological mechanism. Here, using a mouse model of pancreatic ductal adenocarcinoma (PDAC) to study tumor relapse following immunotherapy-induced responses, we found that tumors underwent an epithelial-to-mesenchymal transition (EMT) that resulted in reduced sensitivity to T cell-mediated killing. EMT-transcription factors (EMT-TFs) ZEB1 and SNAIL function as master genetic and epigenetic regulators of this tumor-intrinsic effect. Acquired resistance was not due to immunosuppression in the tumor immune microenvironment, disruptions in the antigen presentation machinery, or altered expression of immune checkpoints. Rather, EMT was associated with epigenetic and transcriptional silencing of interferon regulatory factor 6 (Irf6), which renders tumor cells less sensitive to the pro-apoptotic effects of TNF-α. These findings show how resistance to immunotherapy in PDAC can be acquired through plasticity programs that render tumor cells impervious to T cell killing.

PubMed Disclaimer

Conflict of interest statement

Competing interests Dr. Stanger receives research funding from Boehringer-Ingelheim and previously served as a consultant to iTeos Therapeutics. Dr. Vonderheide reports receiving consulting fees from BMS, is an inventor on a licensed patents relating to cancer cellular immunotherapy and cancer vaccines and receives royalties from Children’s Hospital Boston for a licensed research-only monoclonal antibody.

Figures

Figure 1
Figure 1. Recurrent PDAC develops acquired resistance to combination therapy.
a,b, Mice were injected subcutaneously (s.c.) with 4662 PDAC cells and treated intraperitoneally (i.p.) with control IgG (n = 17) or chemoimmunotherapy (n = 34) consisting of gemcitabine (G), nab-paclitaxel (A), αCD40 agonistic Ab (F), and checkpoint blockades αCTLA-4 (C) and αPD-1 (P) Abs. Tumor growth (a) and survival (b) were monitored. c, The proportion of non-responders and mice with partial response (PR), relapsed after complete response (CR), and durable CR upon treatment was depicted. d, Mice with recurrent tumors after CR or near CR were re-treated with GFCP (blue arrows) and tumor sizes were measured. e–g, Tumor cell lines were generated from s.c. tumors with control IgG (‘Ctrl’ lines, n = 4), early progressor (‘EP’ lines, n = 2), and the relapsed after CR (‘Esc’ lines, n = 8) upon therapy, as denoted by blue, green, and red lines in a, respectively. Naïve WT mice challenged s.c. with these cell lines were treated with control IgG (n = 3 or 4 per line) or GAFCP (n = 5 or 6 per line) and tumor growth (e) and survival (f) were monitored. Response rates (g) in mice received each cell line followed by treatment are shown.
Figure 2
Figure 2. EMT induces immunotherapy resistance in PDAC.
a, Principal component analysis (PCA) plot of RNA-seq data from parental, EP and Esc cell lines (triplicates) is depicted. b, Representative bright field (top) and H&E (bottom) images of cell lines and s.c. implanted tumors on day 18, respectively. Scale bars, 250 μm. c, GSEA of the EMT Hallmark (Molecular Signature Database) in 4662 parental vs. Esc cell lines. Normalized enrichment score (NES) and false discovery rate (FDR) are shown. d, GSEA plots of gene signatures derived from 4662 parental cells overexpressing Zeb1 and Snail (Zeb1/Snail OE) in 4662 parental vs. Esc lines. Gene signatures downregulated (left) and upregulated (right) with Zeb1/Snail OE are shown. e, GSEA plot of a gene signature derived from s.c. implanted Zeb1/Snail OE tumors in s.c. implanted parental vs. Esc tumors. f,g, Individual tumor growth (f) and survival (g) of mice bearing s.c. implanted 4662 parental empty vector (EV, left) and Zeb1/Snail OE (right) tumors treated with either control IgG or FCP (n = 10). h,i, Individual tumor growth (h) and survival (i) of mice bearing s.c. implanted 4662 Esc EV (left) and Zeb1−/−Snail−/− (right) tumors treated with either control IgG or FCP (n = 10). SD, stable disease. Clonal 4662 (C7 and C10) and derived Esc (C7.e1 and C10.e1) lines were used for genetic modification, and both lines showed a similar phenotype. Data represent two independent experiments.
Figure 3
Figure 3. EMT mediates cell-autonomous resistance to direct cytolytic T cell killing.
a,b, Flow cytometric analysis of immune populations in s.c. implanted 4662 parental (P, n = 6) vs. Esc (E1 and E2, n = 4 per line) (a) and 4662 parental EV (n = 5) vs. Zeb1/Snail OE (n = 5) tumors (b) on day 18 post inoculation. c, Representative αCD3 IHC images (left) and quantitation (right) from s.c. implanted parental and Esc tumors (n = 4 or 6). Scale bars, 250 μm. d,e, Mice with CR or near CR after therapy (black arrows) were treated with control IgG (n = 11), depleting αNK1.1 Ab (n = 10), or αCD4 and αCD8 Abs (n = 13), starting from day 50 (blue arrows) post tumor inoculation, and monitored for tumor recurrence (d). The corresponding survival curves are shown in e. n.s., non-significant. f, Tumor growth of s.c. inoculated OVA-transduced 4662 parental and Esc tumors in NOD/SCID mice, with or without adoptive transfer of activated OVA-specific CD8+ T cells (OT-I) on day 14 (arrow). g, OVA-tdTomato+ 4662 parental and Esc tumors were co-cultured with non-activated or activated OT-I by αCD3 and αCD28 Abs overnight, at indicated tumor to effector (T:E) ratios. Two days later, AnnexinV and 7-AAD expression on tumor cells were determined by flow cytometry. h, OVA+ 4662 parental and Esc tumors were additionally transduced with each fluorescence YFP or CFP plated separately or mixed, and co-cultured with activated OT-I. AnnexinV and 7-AAD on each tumor were measured 2 d after co-culture. ij, The percentages of 7-AAD+ cells on co-cultured OVA+ 4662 parental EV vs. Zeb1/Snail OE tumors (i) and 4662 Esc EV vs. Zeb1−/−Snail−/− tumors (j) with or without OT-I for 2 d. Data represent two independent experiments.
Figure 4
Figure 4. Transcriptional and chromatin profiling identifies Irf6as a potential regulator of acquired immunotherapy resistance.
a, Venn diagram of HOMER de novo motifs identified in chromatin regions significantly enriched in EV vs. Zeb1/Snail OE (ZS) cell lines that have (right) or have not (left) been co-cultured with OT-1 cells. Significantly enriched chromatin regions were defined as |log2fold change| > 1.4 and p-value < 0.05 after DESeq2 analysis. b, Genome browser track showing ATAC-seq reads along the Irf6 gene. c, Aggregate plots comparing the average ATAC signal of EV (blue) and Zeb1/Snail OE (red) tumors around all putative Irf6 promoter sequences. More details are described in the Methods section. d, Transcripts per million (TPMs) of Irf6 in EV vs. Zeb1/SnailOE tumors that have (right) or have not (left) been co-cultured with OT-1 cells. Each dot represents biological replicates. e, Boxplots of log2fold changes in the expression of Irf6-associated genes with differentially open chromatin in parental EV (left, n=470 genes) vs. Zeb1/Snail OE (right, n=173 genes) tumors. f, GSEA plots of an Irf6-dependent gene signature (derived by comparing Irf6-expressing tumors to controls) in parental vs. Esc (left) and parental EV vs. Zeb1/Snail OE (right) tumors. g, GSEA of gene signatures derived from human PDAC cells that highly express IRF6 in parental vs. Esc (left) and EV vs Zeb1/Snail OE tumors (right). Negative normalized enrichment scores (NES) demonstrate enrichment in parental and EV tumors compared to Esc and Zeb1/Snail OE tumors, respectively.
Figure 5
Figure 5. Irf6 loss contributes to EMT-induced immunotherapy resistance.
a, OVA-tdTomato+ 4662 parental EV, Esc EV, and Esc Irf6 tumors were co-cultured with or without activated OT-I at indicated 4662 tumors to OT-I ratios for 2 d. 7-AAD expression on tumors was measured by flow cytometry. b, OVA-tdTomato+ 4662 parental EV and Irf6 KO tumors were used as target cells (1:5 ratio). c,d, Tumor growth (c) and survival (d) of mice bearing 4662 Esc EV and Esc Irf6 tumors treated with control IgG or FCP (n = 10). Data represent two independent experiments. e, TPMs of IRF6 in 7 treatment-paired NSCLC patient samples (Gettinger et al., 2017). All patients demonstrated initial response to combinatorial checkpoint blockade before relapsing. Lines are drawn from patient-matched early treatment (Pre-ICB) to immunotherapy recurrence (IR). Red lines indicate 3 patient samples demonstrating a decrease in IRF6 expression with recurrence. Blue lines indicate the others demonstrating unchanged or increased IRF6 expression with recurrence. f,g, GSEA plots of an Irf6-dependent gene signature derived from Irf6-expressing 4662 tumors (f) and the EMT Hallmark (g) in patient-matched Pre-ICB vs. IR samples separately assorted based on IRF6 expression as in e.
Figure 6
Figure 6. Irf6 promotes susceptibility to T cell killing by enhancing TNF-induced apoptosis.
a, Normalized viability of 4662 parental EV, Esc EV, and Esc Irf6 tumors with varying concentrations of TNF-α in the presence of IFN-γ (0.2 μg/ml) plus cycloheximide (1 μg/ml) for 48 h. IC50 values are 0.03472 ug/ml for parental EV, 0.6494 ug/ml for Esc Irf6, and not determined for Esc EV tumors. b, Mean fluorescence intensities (MFIs) of active caspase-3 in 4662 parental EV, Esc EV, and Esc Irf6 tumors treated with or without TNF-α (0.5 μg/ml) plus IFN-γ in the presence of cycloheximide for 48 h by flow cytometry. c, The percentages of cleaved caspase-3 among s.c. implanted YFP+ 4662 parental EV, Esc EV, and Esc Irf6 tumors with or without immunotherapy by IF staining. Tumors were prepared a week after treatment. d, Normalized viability of 4662 Esc EV and Esc Irf6 tumors, treated with vehicle or z-VAD (20 μM), in response to TNF-α plus IFN-γ in the presence of cycloheximide for 48 h. e, Left, immunoblots of IRF6 and TNF-related cell death mediators in 4662 Esc EV and Esc Irf6 tumors with or without ablation of each gene. Right, normalized viability of 4662 Esc EV and Esc Irf6 tumors with or without indicated gene ablation in response to TNF-α plus IFN-γ in the presence of cycloheximide for 48 h. f, OVA-tdTomato+ 4662 Esc EV and Esc Irf6tumors with or without indicated gene ablation were used as target cells for OT-I co-culture. g, OVA-tdTomato+ 4662 Esc EV and Esc Irf6 tumors were co-cultured with or without activated OT-I and TNF-α neutralizing Ab (5 μg/ml) for 2 d. 7-AAD expression on tumors was measured by flow cytometry. Data represent two independent experiments.
See this image and copyright information in PMC

Similar articles

See all similar articles

References

    1. Ribas A. & Wolchok J.D. Cancer immunotherapy using checkpoint blockade. Science 359, 1350–1355 (2018). - PMC - PubMed
    1. Sharma P et al. The Next Decade of Immune Checkpoint Therapy. Cancer discovery 11, 838–857 (2021). - PubMed
    1. Schachter J. et al. Pembrolizumab versus ipilimumab for advanced melanoma: final overall survival results of a multicentre, randomised, open-label phase 3 study (KEYNOTE-006). Lancet 390, 1853–1862 (2017). - PubMed
    1. Doroshow D.B. et al. Immunotherapy in Non-Small Cell Lung Cancer: Facts and Hopes. Clinical cancer research : an official journal of the American Association for Cancer Research 25, 4592–4602 (2019). - PMC - PubMed
    1. Yuan S., Norgard R.J. & Stanger B.Z. Cellular Plasticity in Cancer. Cancer Discov 9, 837–851 (2019). - PMC - PubMed

Publication types