Activation of the PD-1 pathway contributes to immune escape in EGFR-driven lung tumors

Abstract

The success in lung cancer therapy with programmed death (PD)-1 blockade suggests that immune escape mechanisms contribute to lung tumor pathogenesis. We identified a correlation between EGF receptor (EGFR) pathway activation and a signature of immunosuppression manifested by upregulation of PD-1, PD-L1, CTL antigen-4 (CTLA-4), and multiple tumor-promoting inflammatory cytokines. We observed decreased CTLs and increased markers of T-cell exhaustion in mouse models of EGFR-driven lung cancer. PD-1 antibody blockade improved the survival of mice with EGFR-driven adenocarcinomas by enhancing effector T-cell function and lowering the levels of tumor-promoting cytokines. Expression of mutant EGFR in bronchial epithelial cells induced PD-L1, and PD-L1 expression was reduced by EGFR inhibitors in non-small cell lung cancer cell lines with activated EGFR. These data suggest that oncogenic EGFR signaling remodels the tumor microenvironment to trigger immune escape and mechanistically link treatment response to PD-1 inhibition.

Significance: We show that autochthonous EGFR-driven lung tumors inhibit antitumor immunity by activating the PD-1/PD-L1 pathway to suppress T-cell function and increase levels of proinflammatory cytokines. These findings indicate that EGFR functions as an oncogene through non-cell-autonomous mechanisms and raise the possibility that other oncogenes may drive immune escape.

Figure 1. Activation of the EGFR pathway in bronchial epithelial cells leads to an immunosuppressive lung microenvironment

(a) Microarray expression profiling analysis of lung tumors from mice with EGFR T790M, L858R (TL) or control lungs focusing on Pd-1, Ctla-4, Pd-l1, the EGFR ligands eregulin (Ereg), amphiregulin (Areg), and betacellulin (Btc), and the cytokines Tgf-β1, granulin (Grn), and interleukin-6 (Il6). 2 and 4 week time points indicate the time between the induction of the transgene with doxycycline and subsequent euthanasia. EGFR mutant vs WT for the gene set shown p=3×10⁻²⁰ (b) Left: Surface PD-L1 expression on CD45⁺ hematopoietic cell population and CD45⁻ human EGFR⁺ cells (tumor cells) was evaluated by FACS. PD-L1 and isotype control staining are shown with the clear black and gray filled lines respectively for normal lung (NL) and tumor bearing lung (TBL) with either microscopic disease or macroscopic nodules. Right: Representative images from the lungs of Del19, TD and TL mice stained for hematoxylin and eosin (H&E) and PD-L1. Scale bars show 100μm for all panels. (c) CD8⁺/CD4⁺ and CD8⁺/Foxp3⁺ ratios and PD-1 and Foxp3 positive frequencies in total CD3⁺ T cells from NL and tumor (T) from T790M L858R (TL) mice: n=4) *p<0.01. (d) Lung weights of control mice and mice carrying tumors driven by Del19, TD or TL. Quantitative analysis of PD-1 and Foxp3 positive T cells (con and TL: n=4, con and Del, con and TD: n=6) * p<0.05 (NL vs TBL for each group; PD-1⁺, PD-1⁺Foxp3⁺ and Foxp3⁺) (e) Co-expression of immune suppressive receptors; Foxp3, PD-1, LAG-3, and Tim-3 in CD3⁺ T cells. (f). Concentration of cytokines IL-6, TGF-b1, progranulin (PRGN), vascular endothelial growth factor (VEGF), granulocyte-macrophage colony stimulating factor (GM-CSF), and CCL2 in BALFs from NL (white bars) and TBL from TL mice (black bars) (con and TL: n=6). NL vs TBL for all cytokines, p<0.02 (g) Immune cell populations; T cell, B cell, NK cell, granulocytes (GR), alveolar macrophages (AM) and mixed populations (CD11b⁺F4/80⁺ population) (the method to identify each population was shown in supplementary methods) in NL and TBL (con and TL: n=4) * p<0.05

Figure 2. In vivo efficacy of PD-1 antibody blockade in EGFR mutant murine lung cancer models

The anti-tumor effects of anti-PD-1 antibodies in mouse models of EGFR driven lung cancers (a-e). (a) Tumor volume changes by MRI at varying time points; baseline, 2, and 4 weeks after treatment of the indicated genotypes of mice. H indicates location of the heart. (b) Quantification of tumor volume changes as compared to baseline tumor volumes in the mice that were treated with anti-PD1 antibody (aPD1 t.) or left untreated (Unt,). (c) Representative images of lung sections from tumor bearing mice (TD) that were either treated with anti-PD-1 antibody (aPD1 t.) for 1 week or left untreated. Sections were stained for H&E, TUNEL, and cleaved caspase 3. Scale bars represent 25 μm for all panels. (d) Quantification of TUNEL and caspase 3 stainings respectively. Data points indicate total positive signal per tumor field. For TUNEL: n=3 for untreated and n=4 for PD-1 treated mice; for cleaved caspase 3: n=6 for untreated and n=3 for PD-1 treated mice). (e) Kaplan Meier survival analysis of the anti-PD-1 antibody treated or untreated mice bearing EGFR driven tumors. Treatments were started after tumors were confirmed with MRI at the time points indicated by arrows for each of the mouse lines.

Figure 3. Anti-PD-1 antibody binds to activated T cells and improves effector function

(a) Schematic of the short term in vivo treatment of mice with anti-PD-1 antibodies after tumor burden was confirmed by MRI imaging. Each group was treated either with isotype control (untreated) or anti-PD-1 antibody on Days 0, 3, 5 and 8 (4 doses), and then at day 9 mice were sacrificed for analysis. (b) Representative flow cytometry results of PD-1⁺ or RatIgG2a⁺ (therapeutic anti-PD-1 antibody binding) in CD4⁺ and CD8⁺ T cells, anti-PD1 antibody treated mouse (+ aPD1), control antibody treated mouse (- aPD1) (c) Changes in total T cell (CD3), CD8⁺ T cells, and Tregs, and ratios of CD8/CD4 and CD8/Treg after PD-1 blockade. (d) Enhancement of effector T cell function (IFN-γ production) by PD-1 antibody blockade. (e) CD3 immunohistochemistry (top) and quantification of intra-tumoral CD3+ cells per high power field in untreated and PD-1 antibody treated tumors (bottom). Scale bars indicate 25 μm for all panels. Each point on the graph represents counts from single tumor nodule. For del19, N= 2 for untreated, n=5 for anti-PD-1 antibody treated mice. For TD, n=4 for untreated and n=5 for anti-PD-1 antibody treated mice. P=0.01 for both CD3 graphs. (e) Enhancement of T cell function represented as IFN-γ production by PD-1 antibody blockade. (f) Concentration of the cytokines IL-6, TGF-β1, and PRGN in BALFs. (g) Absolute number of alveolar macrophages in lungs from Del. For all bar graphs in this figure Del 19 (untreated and treated: n=6 and n=7) and TD (untreated and treated: n=6 and n=6) * p<0.05

Figure 4. EGFR pathway activation in human bronchial epithelial cells induces PD-L1 expression

(a) Microarray expression profiling analysis of established cell lines from human NSCLC patients. Black and red bars indicate identified Kras or EGFR mutations respectively. TGF-alpha, met proto-oncogene (MET), heparin-binding EGF-like growth factor (HBEGF), EREG and BTC are EGFR ligands. (b) PD-L1 up-regulation in BEAS-2B bronchial epithelial cell lines transduced with vectors encoding Kras mutation (G12V) or EGFR mutation (T790M-Del19), as assessed by qPCR and flow cytometry (c-e). Reduction of PD-L1 expression in NSCLC cell lines 72 hours after EGFR TKI treatment at the indicated concentrations (in the absence of drug-induced apoptosis) (c) EGFR-del19 mutant PC-9 and HCC827 NSCLCs (d), Gefitinib resistant H1975 NSCLC (e) EGFR wild type Kras mutant H358 NSCLC. Representative results from 3 independent experiments are shown. (f) Sections of formalin fixed patient tumors carrying EGFR mutations stained with H&E or PD-L1. Top panel, high expression on tumor cell membrane; middle panel, low expression on membrane; bottom panel, expression on macrophages. Scale bars show 100 μm.