Macrophage IL-10 blocks CD8+ T cell-dependent responses to chemotherapy by suppressing IL-12 expression in intratumoral dendritic cells

Abstract

Blockade of colony-stimulating factor-1 (CSF-1) limits macrophage infiltration and improves response of mammary carcinomas to chemotherapy. Herein we identify interleukin (IL)-10 expression by macrophages as the critical mediator of this phenotype. Infiltrating macrophages were the primary source of IL-10 within tumors, and therapeutic blockade of IL-10 receptor (IL-10R) was equivalent to CSF-1 neutralization in enhancing primary tumor response to paclitaxel and carboplatin. Improved response to chemotherapy was CD8(+) T cell-dependent, but IL-10 did not directly suppress CD8(+) T cells or alter macrophage polarization. Instead, IL-10R blockade increased intratumoral dendritic cell expression of IL-12, which was necessary for improved outcomes. In human breast cancer, expression of IL12A and cytotoxic effector molecules were predictive of pathological complete response rates to paclitaxel.

Copyright © 2014 Elsevier Inc. All rights reserved.

Figures

Figure 1. Macrophages are the primary source of interleukin-10 in mammary tumors

(A) Representative images of F4/80 immunoreactivity (left panels) in mammary tumors from MMTV-PyMT mice treated with IgG1 or αCSF-1 mAb in combination with paclitaxel (PTX). Quantitation of CD11b+F4/80+MHCII+Ly6C− macrophages by flow cytometry is shown to the right as a percent of total live cells. Significance was determined by an unpaired t-test with Welch’s correction with ***p < 0.001. (B) Fold change in whole tissue gene expression in tumors from PTX/αCSF-1 mAb treated mice determined by real-time PCR using a 96-gene immune array. Only significant (p<0.05; Mann-Whitney) changes are shown. n≥12 mice per group. (C) Il10 mRNA expression levels from FACS-sorted stromal populations isolated from untreated mice as determined by real time PCR. Data is normalized to Tbp expression and displayed as mean ± SEM with n=8 per cell type. MØ, macrophage; mono, monocyte; DC, dendritic cell. (D) Correlation between IL10 expression and various myeloid-associated genes in human breast cancer samples from the TCGA dataset (n=1161). (E) Detection of IL-10 in human breast cancer by immunohistochemistry. 14 CTX-naïve and 9 CTX-treated patient samples were evaluated. Representative images reflecting low and high staining are displayed. (F) Immunofluorescent staining for IL-10, CD163, and DNA using Hoescht 33342 in human breast cancer. Representative images from 1 of 3 patient samples are displayed. See also Figure S1.

Figure 2. IL-10 receptor blockade improves response to paclitaxel

(A) Relative tumor volume in MMTV-PyMT mice following treatment with IgG1, αCSF-1 mAb, αIL-10R mAb, PTX, or a combination thereof. Dosing strategy is displayed at top. Data is displayed as mean ± SEM with n≥9 mice per group. Significance was determined by two-way ANOVA with ***p < 0.001. (B) Survival of mice bearing orthotopic PyMT-derived tumor explant tumors following treatment with IgG1, αCSF-1 mAb, αIL-10R mAb, PTX, or a combination thereof. Significance was determined by log-rank with **p < 0.01, ***p < 0.001 with n≥12 mice per group. (C) Relative tumor volume in MMTV-PyMT mice following treatment with IgG1 or αIL-10R mAb in combination with carboplatin (CBDCA). Dosing strategy is displayed at top. Data is displayed as mean ± SEM with n≥5 mice per group. Significance was determined by two-way ANOVA with *p < 0.05. (D) Relative tumor volume in C3(1)-TAg mice following treatment with IgG1, αCSF-1 mAb, or αIL-10R mAb in combination with PTX. Dosing strategy is displayed at top. Data is displayed as mean ± SEM with n≥8 mice per group. Significance was determined by two-way ANOVA with ***p < 0.001. (E) Representative H&E sections of lungs taken from animals treated with PTX alone or in combination with αCSF-1 mAb or αIL-10R mAb. (F) The number of metastastic foci per lung in each treatment group from A. Data is displayed as mean ± SEM with n≥9 mice per group. Significance was determined by an unpaired t-test with Welch’s correction with *p < 0.05, **p < 0.01. (G) The average size of metastatic foci in each treatment group. Data is displayed as mean ± SEM with n≥9 mice per group. Significance was determined by an unpaired t-test with Welch’s correction with **p < 0.01. See also Figure S2.

Figure 3. Response to αIL-10R/PTX is CD8+ T cell-dependent

(A) Flow cytometric quantitation of CD3+CD8+ cytotoxic T cells (A), CD3+CD4+FoxP3− TH cells (B), and ratio of CD8+ to CD4+ TH cells (C) in orthotopic PyMT-derived tumors 2-days following the 2nd dose of PTX (day 7). Significance was determined by unpaired t-test with Welch’s correction relative to IgG1 or IgG1/PTX control groups, with *p < 0.05, **p < 0.01, ***p < 0.001. (D) Density of granzyme B (GZMB)-expressing cells in orthotopic PyMT-derived tumors on day 7 as determined by immunohistochemistry. Data is displayed as mean ± SEM with n≥12 mice per group. Significance was determined by an unpaired t-test with Welch’s correction relative to IgG1 or IgG1/PTX control groups, with *p < 0.05. Representative images are shown to the right. (E) Relative mammary tumor volume after 3 rounds of PTX in MMTV-PyMT transgenic mice following treatment with αIL-10R mAb and PTX. CD8+ T cells were depleted 5-days prior to the first dose of PTX. Data is displayed as mean ± SEM with n≥5 mice per group. Significance was determined by an unpaired t-test with ***p < 0.001. Polychromatic dot plots displaying CD8+ T cell depletion are shown to the right. See also Figure S3.

Figure 4. IL-10 does not influence macrophage recruitment, polarization or function

(A–D) Percent of total CD45+ cells in mammary tumors from A, comprised of CD11b+F4/80+MHCIIHILy6C− macrophages (A), CD11b+F4/80+MHCIIHILy6C− macrophages (B), CD11b+F480−Ly6G− monocytes (C), and CD11b+F4/80−LyG+ neutrophils (D). Data generated by flow cytometry. Significance determined by an unpaired t-test with Welch’s correction relative to IgG1 or IgG1/PTX control groups, with *p < 0.05, **p < 0.01, ***p < 0.001. (E) Representative histograms displaying cell trace fluorescence of CD8+ splenic T cells stimulated with αCD3/αCD28 for 60 hrs either alone (red) or co-cultured (blue) at a 1:1 ratio with FACS-sorted MHCIIHI or MHCIILO macrophages from mammary carcinomas. (F) Fold expansion of dividing CD8+ T cells (replication index) is displayed at various ratios of CD8+ T cells to macrophages as described in E. One of two experiments is shown, with samples assayed in triplicate. Data is displayed as mean ± SEM. (G) Detection of hydroxyprobe (pimonidazole) by intracellular staining of tumor-associated macrophages. Significance determined by unpaired t-test with Welch’s correction with ***p < 0.001. (H) Representative histograms displaying cell trace fluorescence of CD8+ splenic T cells as described in E, with addition of 10 μg/ml αIL-10R or αIL-10 mAb added at the start of the incubation period. One of two experiments is shown. (I) Socs3 expression in MHCIIHI and MHCIILO tumor macrophage subsets determined by real time PCR, with data analyzed by comparative threshold cycle method using Tbp as a reference gene. Significance determined by Mann-Whitney with *p < 0.05, **p < 0.01. See also Figure S4.

Figure 5. Expression of IL-10 receptor by tumor dendritic cells

(A) Flow cytometric analysis of IL-10R surface expression for EpCAM+ epithelial cells, CD31+ endothelial cells, PDGFRα+ fibroblasts, or CD45+ leukocytes in untreated MMTV-PyMT mice at end stage (>100 days). Fluorescence minus one (FMO) controls are shown in gray for each histogram. Representative data from 1 of 3 animals is shown. (B) Surface expression of IL-10R measured by mean fluorescence intensity (MFI) minus background in various leukocyte populations from mammary tumors from MMTV-PyMT mice as compared to similar populations in the spleens of non-tumor bearing animals. Data is displayed as mean ± SEM with n=3 mice per group. Significance determined by an unpaired t-test compared to the spleen, with *p < 0.05, **p < 0.01, ***p < 0.001. MØ, macrophage; mono, monocyte; DC, dendritic cell. (C) Flow cytometric analysis of CD11c+F4/80−MHCII+CD103−CD11b+ DCs (left panel) or CD11c+F4/80−MHCII+ CD11b−CD103+ DCs (right panel) in tumors from the MMTV-PyMT animals shown in 2A. Significance determined by unpaired t-test relative to IgG1 or IgG1/PTX control groups, with *p < 0.05, **p < 0.01, ***p < 0.001. (D) Localization of F4/80+ macrophages and CD103+ leukocytes within the tumor stroma of MMTV-PyMT tumors via immunofluorescent staining of serial sections. Representative images from 1 of 3 animals are displayed for each group. See also Figure S5.

Figure 6. Response to PTX is dependent on IL-12 expression by dendritic cells

(A) Il12a or Il12b mRNA expression levels in FACS-sorted leukocytes from mammary tumors of MMTV-PyMT mice as determined by real time PCR. Data is normalized to Tbp expression and displayed as mean ± SEM with n=8 per cell type. MØ, macrophage; mono, monocyte; DC, dendritic cell. (B) Detection of IL-12p40 by intracellular flow cytometry in tumor leukocytes following in vivo administration of brefeldin A. Representative data from 1 of 2 animals is shown. (C) Production of IL-12p70 by BMDCs as measured by ELISA following 24-hour stimulation with IFN-γLPS, IFN-γ/αCD40 or αCD40/LPS. Cells were pre-treated for 24 hours with 1 to 10 U/ml of IL-10 prior to stimulation. (D–E) Expression of Il12a (D) or Il12b (E) mRNA by BMDCs as measured by RT-PCR. Cells were pre-treated for 24 hours with 10 U/ml of IL-10 prior to stimulation. For C–E, data is displayed as mean ± SEM and samples were assayed in quadruplicate and one of two representative experiments is shown. Significance determined by unpaired t-test relative the control group, with *p < 0.05, **p < 0.01, ***p < 0.001. (F) Il12a or Il12b mRNA expression levels in FACS-sorted macrophage (MØ) and DC populations from mammary tumors of MMTV-PyMT mice 2-days following the 2nd dose of PTX (day 7) in combination with αIL-10R mAb. Data is normalized to Tbp expression and displayed as mean ± SEM with n≥4 mice per cell type. Significance determined by Mann-Whitney with *p < 0.05, **p < 0.01. (G) Relative tumor volume after 3 rounds of PTX in MMTV-PyMT transgenic mice following combination therapy with αCSF-1 mAb or αIL-10R mAb. IL-12 neutralizing mAb (αIL-12p40 or αIL-12p70) administered concurrently every 5 days. Data is displayed as mean ± SEM with n≥5 mice per group. Significance determined by unpaired t-test with *p < 0.05, **p < 0.01, ***p < 0.001. See also Figure S6.

Figure 7. IL12A expression correlates with pathologic complete response rates in breast cancer

(A) Linear regression analysis between IL12A or IL12B expression and various DC-associated genes (CCR7, IRF8, CIITA) in human breast cancer samples from the TCGA dataset (n=1161). (B) Linear regression analysis between IL12A or IL12B expression and various cytotoxic lymphocyte-associated genes (GZMB, CD8A, IFNG) in human breast cancer samples from the TCGA dataset. (C) Frequency of pathologic complete response (pCR) in patients separated by median expression for genes associated with a cytotoxic T cell response (GZMB, CD8A, IFNG) or DCs (IL12A, CIITA, CCR7, IRF8). Data reflects a cohort of 379 patients constructed from 2 independent datasets. Significance was determined by chi-square.

Figure 8. Model of IL-10 suppressing IL-12 production by dendritic cells and limiting the CD8+ T cell-dependent response to CTX in mammary carcinomas

CD11b+ monocytes are recruited into mammary tumors through increased CSF-1 gradients largely derived from epithelial cells. Once in tissue/tumor parenchyma, differentiated macrophages stimulate invasion and metastasis through EGF, which is induced in response to CSF-1, in combination with IL-4/IL-13 derived from CD4+ T cells. IL-4/IL-13 also induces expression and/or activity of cathepsin proteases, directly promoting resistance to CTX-induced cell death. Macrophages further suppress CTX efficacy by indirectly limiting a cytotoxic T cell response. Macrophage-derived IL-10 suppresses the ability of dendritic cells to produce IL-12 during CTX, likely in response to damage associated molecular patterns (DAMPs). In the absence of macrophages or IL-10, IL-12 is able to promote a productive CD8+ T cell response, thereby enhancing response to chemotherapy.