Interleukin-17A Promotes Lung Tumor Progression through Neutrophil Attraction to Tumor Sites and Mediating Resistance to PD-1 Blockade

Abstract

Introduction: Proinflammatory cytokine interleukin-17A (IL-17A) is overexpressed in a subset of patients with lung cancer. We hypothesized that IL-17A promotes a protumorigenic inflammatory phenotype and inhibits antitumor immune responses.

Methods: We generated bitransgenic mice expressing a conditional IL-17A allele along with conditional Kras^G12D and performed immune phenotyping of mouse lungs, a survival analysis, and treatment studies with antibodies either blocking programmed cell death 1 (PD-1) or IL-6 or depleting neutrophils. To support the preclinical findings, we analyzed human gene expression data sets and immune profiled patient lung tumors.

Results: Tumors in IL-17:Kras^G12D mice grew more rapidly, resulting in a significantly shorter survival as compared with that of Kras^G12D mice. IL-6, granulocyte colony-stimulating factor (G-CSF), milk fat globule-EGF factor 8 protein, and C-X-C motif chemokine ligand 1 were increased in the lungs of IL17:Kras mice. Time course analysis revealed that levels of tumor-associated neutrophils were significantly increased, and lymphocyte recruitment was significantly reduced in IL17:Kras^G12D mice as compared with in Kras^G12D mice. In therapeutic studies PD-1 blockade was not effective in treating IL-17:Kras^G12D tumors. In contrast, blocking IL-6 or depleting neutrophils with an anti-Ly-6G antibody in the IL17:Kras^G12D tumors resulted in a clinical response associated with T-cell activation. In tumors from patients with lung cancer with KRAS mutation we found a correlation between higher levels of IL-17A and colony- stimulating factor 3 and a significant correlation among high neutrophil and lower T-cell numbers.

Conclusions: Here we have shown that an increase in a single cytokine, IL-17A, without additional mutations can promote lung cancer growth by promoting inflammation, which contributes to resistance to PD-1 blockade and sensitizes tumors to cytokine and neutrophil depletion.

Keywords: Cytokines; IL-17; MDSC; Neutrophils; PD-1; Resistance.

Figure 1. Expression of IL-17 in Kras mutant mice promotes lung cancer growth

A) Breeding scheme for the IL-17 and Kras mice. B) Hematoxylin-eosin (H&E) staining and IL-17 immunohistochemistry. Scale bars show: 1mm for the H&E panels and 50um for IL-17 IHC panels. C) Human IL-17A levels in the BAL fluid from the mice detected by ELISA. IL-17 (n=4), Kras (n=4) and IL-17:Kras (n=4). D) Total lung lesion counts for mice at 4 weeks (Kras:n=3, IL-17: Kras:n=4), 7 weeks (Kras:n=4, IL-17:Kras:n=4), and 12 weeks 4 weeks (Kras:n=6, IL-17:Kras:n=4) after Adeno-Cre induction. E) Top: Representative Ki67 immunohistochemistry and bottom: quantification of Ki67 staining. Scale bars show 50um for Ki67 panels. n=6 for Kras and n=5 for IL-17:Kras mice. 3 images have been quantified per mice. F) Kaplan Meier Survival curve for the IL-17, Kras, and IL-17:Kras mice. Mice were added to the graph either when they died or reached disease burden euthanasia criteria. N=10, 16, 13 for IL-17, Kras, and IL-17:Kras mice. Median survival for Kras: 23.2 for IL-17:Kras 11.2 p<0.0001.

Figure 2. IL-17 modulates the lung cytokine and immune cell profiles

A) Cytokine and chemokine (IL-6, G-CSF, CXCL1, and MFG-E8) analysis of BAL fluid at 4, 7, and 12 weeks in Kras and IL-17:Kras mice. n=4 at 4 weeks, n=4 at 7 weeks and n=5 at 12 weeks for Kras and IL-17:Kras mice both. B) Tumor associated macrophage (TAM) and tumor associated neutrophil (TAN) counts per mg of the lung tissue in Kras, IL-17:Kras mice. n=4 at 4 weeks, n=4 at 7 weeks and n=5 at 12 weeks for Kras and IL-17:Kras mice both.

Figure 3. T cell counts negatively correlate with disease burden in IL17 expressing mice and T cell express suppression markers

A) Correlations of CD4 and CD8 T cell counts and total lung weights for Kras and IL-17:Kras mice. Kras:n=14, IL-17:Kras:n=14. B) Immune checkpoint receptor expression in CD4 and CD8 T cells from the Kras vs IL-17:Kras mice at the same tumor burden (Supplemental figure 3A, Kras:n=5 and IL-17:Kras:n=5). C) PD-L1 expression in the tumor cells from Kras vs IL-17:Kras mice at the same tumor burden (Supplemental figure 3A, Kras:n=5 and IL-17:Kras:n=5)

Figure 4. IL-17 high tumors are resistant to PD-1 immune checkpoint blockade but sensitive to IL-6 blockade or neutrophil depletion

A) Left. Representative Magnetic resonance imaging (MRI) images and Right: Quantification of MR images of mice untreated or treated with PD-1, IL-6 or Ly-6G blocking antibodies for 2 weeks. B) Quantification of MRI for treatments in A. Each data point represents a different mouse. C) Ki67 IHC on the lung tissue from IL-17:Kras mice either untreated or treated with IL-6 antibody, and quantification of IHC. p=0.02 for Ki67 counts between untreated and IL-6 antibody treated IL-17:Kras tumors. Each data point is from a different mouse lung tissue. D) Levels of IL-6 and G-CSF in the Bronchoalveolar lavage fluid of IL-17:Kras mice untreated (n=4) or treated with IL-6 (n=5) or Ly-6G (n=5) antibodies for two weeks. E) CD8 T cell counts (left) and intracellular interferon gamma positive CD8 T cell counts per mg of lung for IL-17:Kras mice either untreated or treated with IL-6 or Ly-6G antibodies. Untreated (n=4), treated with anti-IL-6 treated:IL-6 ab (n=5) and anti-Ly-6G treated:Ly-6G ab (n=5).

Figure 5. Lung cancer patients have increased serum IL-17. Increase in neutrophils is associated with decreased lymphocytes

A) IL-17A levels in the sera of lung cancer patients and healthy controls n=10 for patients and 9 for control group. p=0.0003. B) Expression of IL17A, G-CSF (CSF3), and IL6 in EGFR and Kras mutant lung cancer samples from the TCGA dataset EGFR n=21, Kras n=75 samples . Expression of IL17A, CSF3, and IL6 were calculated from Log2 values of the RNA sequencing data. p=0.049, p=0.0103, and p=0.1546 for IL17A, G-CSF and IL-6 respectively. P values were calculated using Mann-Whitney test in graph-pad prism. C) Correlations of neutrophils (CD45+ CD66b+) vs T cells (CD45+ CD3+) in freshly resected patient NSCLC samples (Kras mutant) were calculated as % of immune cells in total CD45+ cells in all samples. p value was calculated by linear regression using graph-pad prism. n=35.