Crosstalk between Regulatory T Cells and Tumor-Associated Dendritic Cells Negates Anti-tumor Immunity in Pancreatic Cancer

Abstract

Regulatory T (Treg) cell infiltration constitutes a prominent feature of pancreatic ductal adenocarcinoma (PDA). However, the immunomodulatory function of Treg cells in PDA is poorly understood. Here, we demonstrate that Treg cell ablation is sufficient to evoke effective anti-tumor immune response in early and advanced pancreatic tumorigenesis in mice. This response is dependent on interferon-γ (IFN-γ)-producing cytotoxic CD8⁺ T cells. We show that Treg cells engage in extended interactions with tumor-associated CD11c⁺ dendritic cells (DCs) and restrain their immunogenic function by suppressing the expression of costimulatory ligands necessary for CD8⁺ T cell activation. Consequently, tumor-associated CD8⁺ T cells fail to display effector activities when Treg cell ablation is combined with DC depletion. We propose that tumor-infiltrating Treg cells can promote immune tolerance by suppressing tumor-associated DC immunogenicity. The therapeutic manipulation of this axis might provide an effective approach for the targeting of PDA.

Keywords: anti-tumor immunity; pancreatic ductal adenocarcinoma; regulatory T cells; tolerance; tumor-associated dendritic cells.

Graphical abstract

Figure 1

Intratumoral Treg Cells Promote Pancreatic Neoplastic Growth GFP-Kras^G12D-PDECs were implanted (day 0) into the pancreata of syngeneic mice. Pancreata were analyzed at the time points indicated. (A) Representative images of immunofluorescence staining for Foxp3 at 3 weeks post-implantation. Boxed region is magnified below (scale bar, 200 μm). (B) Flow cytometric plots (left) and quantification (right) of the percentage and number of CD4⁺ Foxp3⁺ Treg cells (n = 3–4 mice). (C) Flow cytometric plots (left) and quantification (right) of the expression of CD44 on Foxp3⁺ Treg cells (n = 3–4 mice). (D) Flow cytometric plots (left) and quantification (right) of the expression of CTLA-4 in Foxp3⁺ Treg cells in pancreata and tumor-draining Pan LNs 5 weeks post-implantation (n = 3 mice). (E–H) Schematic of the experimental design. The arrows indicate either PBS (control) or DT (50 μg/kg) was injected i.p. on days 7, 9, 11, and 13 after implantation of GFP-Kras^G12D-PDECs into pancreata of Foxp3^DTR mice. Analyses in (F) – (H) were performed 14 days after orthotopic implantation. (F) Flow cytometric plots (left) and quantification (right) of CD4⁺ Foxp3⁺ Treg cells (n = 4 mice). (G) Orthotopic pancreatic tumors (left) and quantification of tumor volume (right) (n = 4 mice). (H) Representative images of H&E staining (top) and GFP immunohistochemical staining (bottom) on sections from orthotopic pancreatic grafts. Scale bar, 1 mm. (I and J) Schematic of the experimental design. Mice were analyzed on day 25 (I) or when moribund (J) after implantation of KPC cells. (I) Orthotopic pancreatic tumors (left) and quantification of tumor volume (right) from mice treated as shown (n = 4 mice). (J) Kaplan-Meier survival curves (^∗∗∗∗p < 0.0001, log-rank test) of mice implanted with KPC cells (n = 7–8 mice). Data are representative of two or three independent experiments and are presented as mean ± SEM. ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001. See also Figures S1 and S2.

Figure 2

Anti-tumor Effect of Treg Cell Ablation Is Dependent on IFN-γ-Producing CD8⁺ T Cells For (A)–(D), either PBS or DT was injected as described in Figure 1E after implantation of GFP-Kras^G12D-PDECs into pancreata of Foxp3^DTR mice. Analyses were performed 14 days after orthotopic implantation. (A) Quantification of the percentage and number of CD8⁺ T cells (n = 4–7 mice). (B–D) Flow cytometric plots (left) and quantification (right) of CD44- (B), granzyme B- (C), and IFN-γ- (D) expressing CD8⁺ T cells in orthotopic pancreatic grafts (n = 4–7 mice). (E) Either PBS or DT was injected i.p. 7 and 9 days post-implantation, and 250 μg rat IgG2a isotype or anti-CD8 (αCD8) antibody was injected 8 and 11 days after implantation of GFP-Kras^G12D-PDECs into pancreata of WT or Foxp3^DTR mice. Analysis was performed 15 days after orthotopic implantation. Orthotopic pancreatic tumors (left) and quantification of tumor volume (right) from mice treated as shown (n = 4–5 mice). (F–I) Either PBS or DT was injected i.p. 7 and 9 days post-implantation, and 1 mg rat IgG1 isotype or anti-IFN-γ (αIFNγ) antibody was injected i.p. 9 days after implantation of GFP-Kras^G12D-PDECs into pancreata of WT or Foxp3^DTR mice. Analyses were performed 17 days after implantation. (F) Orthotopic pancreatic tumors (left) and quantification of tumor volume (right) from mice treated as shown (n = 4–5 mice). (G) Representative images of H&E staining on sections from orthotopic pancreatic grafts from Foxp3^DTR mice (left) with boxed regions shown at higher magnification (right). Scale bar, 100 μm. (H) Representative images of GFP immunohistochemical staining (left) and quantification of the percentage of GFP-positive area (right) from orthotopic pancreatic grafts from Foxp3^DTR mice. Scale bar, 100 μm. (I) Representative images of cleaved (cl.) caspase-3 immunohistochemical staining (left) and quantification of the number of cleaved-caspase-3-positive cells per field (right) in orthotopic pancreatic grafts from Foxp3^DTR mice. Scale bar, 50 μm. Data are representative of two independent experiments and are presented as mean ± SEM. ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001. See also Figure S3.

Figure 3

Dynamics of the Interaction between Treg Cells and Tumor-Associated CD11c⁺ Cells in the Pancreatic TME (A) Representative still images of CD11c⁺ cells from intravital imaging from pancreata of CD11c-EYFP mice 4 weeks after injection of sham (Matrigel/PBS, 1:1, control) (left) and implantation of GFP-Kras^G12D -PDECs (middle). Vasculature (gray) was visualized by intravenously injection of Evan blue. Representative image of H&E staining on sections from orthotopic pancreatic grafts of GFP-Kras^G12D-PDEC implants (right). Scale bar, 50 μm. (B) Flow cytometric analysis of YFP⁺ population in orthotopic pancreatic grafts 10 days after implantation. (C) The schematic illustrates the experimental system. Representative in vivo time-lapse images of contacts between Foxp3⁺ and CD11c⁺ cells in the pancreas of a Foxp3-GFP;CD11c-YFP mouse 3 weeks after implantation of GFP-Kras^G12D-PDECs. Scale bar, 50 μm. (D) Analysis of contact duration between Foxp3⁺ Treg cells and CD11c⁺ cells (n = 3 mice). (E) The schematic illustrates the experimental system. Representative in vivo time-lapse images of contacts between CD8⁺ T and CD11c⁺ cells in the pancreas of a CD8a-Cre;Rosa^tdtomato;CD11c-EYFP mouse 3 weeks after implantation of GFP-Kras^G12D-PDEC. Scale bar, 50 μm. (F) Analysis of contact duration between CD8⁺ T cells and CD11c⁺ cells (n = 3 mice). (G) Average contact duration of CD11c⁺ cells with Foxp3⁺ Treg cells or CD8⁺ T cells, respectively (n = 3 mice). (H and I) Representative images of immunofluorescence staining of Foxp3⁺ and CD11c⁺ cells from pancreata of WT mice 2 weeks after implantation of GFP-Kras^G12D-PDECs or KPC cells (H) and from a p48-Cre; Kras^G12D; Rosa^tdTomato (KCT) mouse (I). Boxed regions are magnified (right) with arrows indicating close proximity between cells expressing Foxp3 and CD11c. Scale bar, 200 μm. Data are presented as mean ± SEM. ^∗∗p < 0.01. See also Figures S4 and S5 and Movies S1, S2, S3, and S4.

Figure 4

Tumor-Associated CD11c⁺ DCs Display Reduced Local Migration and Bring Tumor Antigen to the Tumor-Draining LN (A) Flow cytometric analysis of YFP⁺ population 10 days after orthotopic pancreatic implantation of GFP-Kras^G12D-PDECs into CD11c-EYFP mice. After gating on the CD45⁺ CD3⁻ CD19⁻ cell population, YFP⁺ cells were analyzed using the cell-surface markers CD11c, MHC class II, CD135, and F4/80 and compared to respective isotype controls (gray). (B) Flow cytometric analysis of CD11c⁺F4/80⁺ cells in the YFP⁺ population. (C) Representative still images of CD11c-YFP cells from sham injected (left) or GFP-Kras^G12D-PDECs implanted (right) 3 weeks post-implantation. The blue lines of each CD11c-YFP cell correspond to the displacement of CD11c⁺ cells during a 10-min time lapse. Scale bar, 50 μm. (D) Quantitative mean velocity of CD11c⁺ cells in control, or proximal (within 50 μm) or distal (>50 μm) to the tumor border. Each dot represents an individual CD11c⁺ cell (n = 3 mice). (E) Displacement tracks of individual CD11c⁺ cells in sham or proximal or distal to the tumor border. Each track represents the displacement of a cell from its starting point during a 10-min time lapse. Data are representative of at least three movies in three independent experiments. (F) Representative still images from intravital imaging from pancreata of KCT chimeric mice generated by transplantation of bone marrow cells isolated from CD11c-EYFP mice. y-z and x-z cross sections are shown to the immediate right and above, respectively. White arrows indicate phagocytic tdTomato particles inside CD11c⁺ cells. The boxed regions labeled a–c show CD11c⁺ cells in which phagocytosis or co-expressed tdTomato particles are magnified (far right). Scale bar, 50 μm. (G and H) Flow cytometric analysis of CD45⁺tdTomato⁺ cells (G) and expression of CD11c and MHC class II after gating on the CD45⁺tdTomato⁺ population compared to respective isotype controls (gray) (H) in the pancreas and Pan LNs of CT (control) or KCT mice. Data are representative of three independent experiments and are presented as mean ± SEM. ^∗∗∗p < 0.001.

Figure 5

Tumor-Associated CD11c⁺ DCs Display Reduced Maturation Markers and IDO Expression (A) Relative surface expression levels of MHC class II, CD40, CD80, and CD86 on CD11c⁺ cells from orthotopic pancreatic grafts at 1 and 5 weeks after implantation of GFP-Kras^G12D-PDECs compared to respective isotype controls (gray (n = 5 mice). (B) Representative images of immunofluorescence staining of indoleamine 2,3-dioxygenase (IDO) and CD11c expression in pancreata from 2 weeks post-implantation of GFP-Kras^G12D- PDECs and KPC cells into WT and 4–6 month-old KC mice. Quantification of the percentage of IDO⁺ cells in CD11c⁺ cells per field is indicated in the bottom right-hand corner of the images. The boxed region is shown at higher magnification (insets). Scale bar, 100 μm. (C) Representative images of immunofluorescence staining of IDO and CD11c expression in human PDA. Quantification of the percentage of IDO⁺ cells in CD11c⁺ cells per field is indicated in the bottom right-hand corner of the image. The boxed region is shown at higher magnification (inset) Scale bar, 100 μm. Data are presented as mean ± SEM. ^∗∗p < 0.01. See also Figure S6.

Figure 6

Treg Cell Ablation Changes the Frequency and Phenotype of Tumor-Associated CD11c⁺ DCs in the Pancreatic TME (A) Schematic of the experimental design. GFP-Kras^G12D-PDECs were implanted into the pancreata of syngeneic Foxp3^DTR mice, and either PBS or DT was injected as described in Figure 1E. Analyses were performed 14 days after orthotopic implantation. (B) Flow cytometric plots (left) and quantification (right) of the percentage and number of CD11c⁺ cells out of total CD45⁺ cells (n = 5 mice). (C) Relative surface expression levels of CD40, CD80, and CD86 on CD11c⁺MHC class II⁺ cells from orthotopic pancreatic grafts of Foxp3^DTR mice treated with either PBS (red) or DT (green) compared to respective isotype controls (gray) (n = 4 mice). (D and E) Relative intracellular expression levels (left) and quantification of the percentage (right) of IDO in CD11c⁺MHC class II⁺ cells from pancreata (Pan) and Pan LNs of Foxp3^DTR mice treated with either PBS (red) or DT (green) compared to respective isotype controls (gray) (n = 5–8 mice). Data are presented as mean ± SEM. ^∗p < 0.05. See also Figure S6.

Figure 7

CD8⁺ T Cell Activation following Treg Cell Ablation Is Mediated by CD11c⁺ DCs (A) Schematic of the experimental design. Orthotopic pancreatic grafts were analyzed by flow cytometry 14 days after implantation. Orthotopic pancreatic tumors (left) and quantification of tumor volume (right) (n = 5 mice). (B and C) Flow cytometric plots and quantification (right) of the percentage and number of CD4⁺Foxp3⁺ Treg cells (B) and CD11c⁺ cells (C) (n = 4–5 mice). (D and E) Quantification of expression of CD44 and IFN-γ in CD4⁺ Foxp3⁻ cells by flow cytometry (D) (n = 4–5 mice) and CD44, granzyme B, and IFN-γ in CD8⁺ T cells by flow cytometry (E) (n = 5–7 mice). Data are presented as mean ± SEM. ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001. NS, not significant. See also Figure S7.