Type III interferon primes pDCs for TLR7 activation and antagonizes immune suppression mediated by TGF-β and PGE2

Abstract

Conventional dendritic cell and plasmacytoid dendritic cell (pDC) subsets have specialized functions that can be modulated by the tumor microenvironment, and produce different interferons that are central to antitumor immune responses. While the function of type I interferons in tumor immunity is well characterized, that of type III interferons produced by type 1 conventional dendritic cells in the tumor microenvironment remains unclear. Here we demonstrate in vitro that type III interferons orchestrate pDC survival, activation and TLR7 expression in the blood, thereby enhancing pDC responses to a TLR7 ligand. Moreover, we show that tumor-associated pDCs express the highest level of IFNLR1, and that these immune cell subsets are the most responsive to IFN-III. Importantly, type III interferons prevent the inhibition of pDCs induced by TGF-β or PGE2 in tumor soluble milieu from patients to restores production of IFN-α in pDCs. With TGF-β or PGE2 having pleotropic functions in immune regulation, our results thus implicate IFN-III-mediated immune modulation to have broad impact on various pathological situations.

Fig. 1. pDCs are the only dendritic cells that respond to IFN-III in blood.

A, B Normalized counts were retrieved from public data available from the Human Cell Atlas (HCA) to assess receptor expression of healthy peripheral blood mononuclear cells (PBMCs). A IFNLR1 expression or (B) IL-10Rβ expression. C Blood cDC1s, cDC2s, and pDCs were sorted from 5 blood PBMCs (#6, #7, #8, #9 and #10), RNA was then sequenced to retrieve the normalized counts of IL-28Rα and IL-10Rβ genes. D pSTAT1 levels were measured in cDC1s, cDC2s and pDCs by flow cytometry after 45 min stimulation with 100 ng/mL IFN-λ1 (purple) or 1000 U/mL IFN-β (green) (n = 6 independent donors). Mean values ± s.e.m are shown. Statistical significance was determined using the two-sided Friedman test on the mean fluorescent intensity (MFI). Source data are provided as a Source Data file. E FACS analysis of pSTAT1 level from a representative donor.

Fig. 2. Increased expression of activation markers PD-L1 and ICOS-L on pDCs stimulated by IFN-III.

A–E pDCs were treated 24 h at 37 °C with (purple) or without (gray) 100 ng/mL of IFN-λ1 (without IL-3). Cells were stained and protein expression was assessed by flow cytometry (n = 8 independent donors). Source data are provided as a Source Data file. A Quantification of the mean fluorescence intensity (MFI) of the activation markers CD86, CD80, HLA-DR, CD123. B pDCs expressing ICOS-L from a representative healthy donor. C Percentage of pDCs expressing ICOS-L at their surface (n = 8 independent donors). Source data are provided as a Source Data file. D pDCs expressing PD-L1 from a representative healthy donor. E Percentage of pDCs expressing PD-L1 at their surface (n = 10 independent donors). Source data are provided as a Source Data file. Statistical significance was determined using a two-sided Wilcoxon test.

Fig. 3. IFN-III induces TLR7 and pathways involved in IFN-I production.

RNA sequencing was performed on purified pDCs treated with or without 100 ng/mL IFN-λ1 for 12 h at 37 °C (n = 4 healthy donors; without IL-3). A PCA of the normalized RNAseq transcripts per million (TPM) of purified pDCs treated or not with IFN-III. B Volcano plot representation of differentially expressed genes (DEGs) in IFN-III-treated pDCs versus untreated pDCs. The x-axis shows log2 fold changes in expression and the y-axis the -log10 adjusted p-value of each DEG obtained with DEseq2 R package using the Wald test. p-values obtained by the Wald test are corrected for multiple testing using the Benjamini and Hochberg method.C GSEA Hallmark analysis showing enriched gene sets. The x-axis represents the normalized enriched score indicating the hallmark enrichment in IFN-III-treated pDCs versus untreated pDCs. FDR q values are represented in a gradient of blue-to-red and the size of each dot represents the percentage of genes enriched for each hallmark. D Heatmap of z-scores of RNAseq expression using all 3129 DEGs. Each column corresponds to a sample and each row corresponds to a specific gene.

Fig. 4. IFN-III pre-treatment sensitize pDCs to low doses of Imiquimod.

A TLR7 normalized counts from RNA sequencing of pDCs treated with (purple) or without (gray) 100 ng/mL IFN-λ1 for 12 h (n = 4 independent donors). Statistical significance was addressed with a two-sided Wilcoxon test. Source data are provided as a Source Data file. B, C TLR7 protein expression was evaluated by flow cytometry on purified pDCs treated (purple) or not (gray) with 100 ng/mL IFN-λ1 for 18 h. Source data are provided as a Source Data file. B TLR7 expression from a representative donor C) Bar plot with individual values of TLR7 expression from healthy donors (n = 6 independent donors). Statistical significance was addressed with a two-sided Wilcoxon test. D PD-L1-positive, E ICOS-L-positive pDCs and F HLA-DR MFI were quantified by flow cytometry after different combinations of treatment with or without IFN-λ1 (100 ng/mL) for 18 h followed by 50 or 250 ng/mL Imiquimod (IMQ) or no Imiquimod (“No stim”) for 24 h. D–F Without IFN-λ1 pre-treatment, cells were kept at 4 °C (Medium) before IMQ stimulation (n=minimum 3 independent donors). G IFN-α2a was quantified by electrochemiluminescence multiplex assay (MSD) in the supernatants of pDCs cultured after different combinations of pre-treatment for 18 h followed by 250 ng/mL or 50 ng/mL of Imiquimod, or no Imiquimod (“No stim”) for 24 h. All culture conditions were performed without IL-3 on minimum n = 3 independent donors. Mean values ± s.e.m are shown. Source data are provided as a Source Data file. Statistical significance was determined using a two-sided Mann-Whitney test.

Fig. 5. IFN-III-stimulated pDCs acquire an IFN-I secretion phenotype upon TLR7 stimulation and favors the secretion of numerous cytokines and chemokines.

A Schematic representation of the experimental protocol. B–C Percentage analysis of pDC subpopulations P1, P2, and P3 according to their expression of PD-L1 and CD80 after pre-treatment and IMQ exposure assessed by flow cytometry. B One representative FACS plots of the different experimental conditions C P1, P2, and P3 distribution following different pre-treatments and IMQ activation (n = 10 independent donors). Mean values ± s.e.m are shown. Source data are provided as a Source Data file. D IFN-α2a and (E) IFN-β, TNF-α, IL-6, MIP1-β, CXCL10, CCL19 were quantified by multiplex assay in the supernatants of pDCs pre-treated with IFN-III or IL-3 for 18 h before being stimulated 24 h with 250 ng/mL IMQ (n = 7 independent donors). Without pre-treatment, cells were kept at 4 °C (Medium) before IMQ stimulation. Statistical significance was determined using a two-sided Kruskal-Wallis test (C) and two-sided Friedman test (D–E). Each point represents a separate donor (C–E). Mean values ± s.e.m are shown (D, E). Source data are provided as a Source Data file.

Fig. 6. cDC1s and pDCs interact through type I and III IFNs axis and are in close contact or in the same discrete area of the TME.

A Visualization of CD8 + T cells, pDCs, cancer cells, cDC1s, respectively by CD8 (red), BDCA2 (green), cytokeratin (white), XCR1 (yellow) using multiplex IF (mIF). Nuclei were stained with DAPI. The white square focuses on a close contact between a cDC1 with a pDC observed in 7 out of 10 triple negative breast tumors. Magnitude x50 (B) 2D projection of cDC1s (purple) and pDCs (green) present in a whole tumor slide of triple negative breast cancer. Blue squares represent a close contact between cDC1 and pDC (distance from nuclei center < 15 µm). C Percentage of XCR1 + cDC1 in close-contact with pDCs in 10 triple negative breast tumors. Mean values ± s.e.m are shown. D The number of pDCs present in each cDC1 area (radius = 100, 70, 30 µm) was calculated in the 10 whole tumor slides. Mean values ± s.e.m are shown. Source data are provided as a Source Data file. E Percentage of PD-L1 positive pDCs after 24 h with supernantants of FACS sorted human in-vitro generated cDC1 stimulated with 10 µg/ml Poly(I:C) 24 h. Before stimulation, pDCs were incubated 1 h with [IFNLR1] or / and [IFNAR2] blocking antibodies (5 µg/ml). [Mouse IgG1] & [mouse IgG2a] (5 µg/ml) were used as isotype controls. Each point represents a separate donor and there is no technical replicate. Mean values ± s.e.m are shown (minimum n = 4 independent donors).

Fig. 7. Tumor-associated pDCs strongly respond to IFN-III.

A, B Normalized counts were retrieved from public data available on the Tumor Immune Single-cell Hub (TISCH) to assess receptor expression of human tumor-infiltrating immune cells. A IFNLR1 expression or (B) IL-10Rβ expression. C Tumor-associated cDC1s, cDC2s, and pDCs were sorted from 4 or 5 breast tumors (patients #1, #2, #3, #4 and #5), RNA was then sequenced to retrieve the normalized counts of IL-28Rα and IL-10Rβ genes. D, E Tumor-associated immune cells were cultured with IFN-I (green) or IFN-λ1 (purple) for 45 min at 37 °C and pSTAT1 was analyzed. D Representative histograms of pSTAT1 in sorted immune cells from an ovarian tumor sample. E pSTAT1 from breast tumor (n = 2: square), ovarian tumors (n = 4: triangle), lung tumors (n = 7: circle) infiltrating cells. Source data are provided as a Source Data file. Statistical significance was determined using a two-sided Friedman test on mean fluorescence intensity (MFI) values.

Fig. 8. IFN-III prevents pDCs inhibition induced by exogeneous immunosuppressive cytokines.

A TGF-β, TNF-α and IL-10 concentrations were quantified using a multiplex immune assay kit in supernatants of primary dilacerated breast (square) and ovarian (triangle) tumors (SN-DIL). B TGFβ-RI normalized counts from RNA sequencing of pDCs treated with (purple) or without (gray) 100 ng/mL IFN-λ1 for 12 h (n = 4 independent donors). Statistical significance was addressed with a two-sided Wilcoxon test. C Schematic diagram of the experimental procedure using [TGF-β] = 2 ng/ml, [PGE2] = 10 µM, [IFN-λ1] = 100 ng/ml. All conditions were in the presence of 20 ng/mL IL-3 to preserve good pDC viability under all conditions. D–H Percentage of PD-L1 positive pDCs and CD123 MFI on pDCs (E–I) after 48 h treatment at 37 °C in different conditions with 250 ng/ml IMQ. Mean values ± s.e.m are shown. Source data are provided as a Source Data file. Minimum n = 8 independent donors. G PGE2 was quantified using PGE2 Elisa assay kit in the supernatants of primary dilacerated breast (n = 7) and ovarian tumors (n = 6). F, J Quantification of IFN-α2 (pg/mL or ng/mL) secreted by pDCs after 48 h treatment at 37 °C in differents conditions with 250 ng/ml IMQ. Mean values ± s.e.m are shown. Source data are provided as a Source Data file. Minimum n = 6 independent donors. Statistical significance was determined using a two-sided Friedman test. K Viability of treated pDC in culture conditions described in panel C. Mean values ± s.e.m are shown (n = 9 independent donors). Source data are provided as a Source Data file.

Fig. 9. IFN-III prevents inhibition of pDCs induced by TME immunosuppressive cytokines such as TGF-β and PGE2.

A Schematic diagram of the experimental procedure. All conditions were in the presence of 20 ng/mL IL-3 to preserve good pDC viability under all conditions and stimulated with high dose of IMQ (250 ng/ml). Mean values ± s.e.m are shown. Source data are provided as a Source Data file. B Percentage of PD-L1 positive pDCs, C Quantification of IFN-α2 (ng/mL) secreted by pDCs in different conditions. Mean values ± s.e.m are shown. Source data are provided as a Source Data file. D Percentage of PD-L1 positive pDCs and (E) Quantification of IFN-α2 (ng/mL) secreted by pDCs in the different conditions. Before pre-treatments, pDCs were incubated 1 h with TGFβ and PGE2 inhibitors: TGFBRi = Galunisertib (5 µM); EPi = EP2 + EP4 inhibitors (10 µM). Mean values ± s.e.m are shown. Source data are provided as a Source Data file and n=minimum 3 independent donors/condition. TGF-β and PGE2 (pg/mL) in SNDIL were quantified by MSD (ND = value not determined). F Quantification of IFN-α (pg/ml) by ECLIA multiplex assay in breast (square) and ovarian (triangle) SNDIL containing IFNL1 or not (n = 80 breast & 25 ovarian tumors). TGFβ-low and TGF β -high groups were determined using the median value of 475.3 pg/mL (n = 71 breast and 22 ovarian tumors). Mean values ± s.e.m are shown. Statistical significance was determined using a two-sided t test. G GSEA Hallmark analysis showing enriched gene sets in tumor associated pDCs versus blood pDCs RNA sequencing. Tumor-associated pDCs and blood pDCs were sorted from n = 4 breast tumors or 5 healthy PBMCs. Enriched gene sets obtained with the fGSEA R-package were filtered with a padj < 0.05.