Type 1 conventional dendritic cells are systemically dysregulated early in pancreatic carcinogenesis

Abstract

Type 1 conventional dendritic cells (cDC1s) are typically thought to be dysregulated secondarily to invasive cancer. Here, we report that cDC1 dysfunction instead develops in the earliest stages of preinvasive pancreatic intraepithelial neoplasia (PanIN) in the KrasLSL-G12D/+ Trp53LSL-R172H/+ Pdx1-Cre-driven (KPC) mouse model of pancreatic cancer. cDC1 dysfunction is systemic and progressive, driven by increased apoptosis, and results in suboptimal up-regulation of T cell-polarizing cytokines during cDC1 maturation. The underlying mechanism is linked to elevated IL-6 concomitant with neoplasia. Neutralization of IL-6 in vivo ameliorates cDC1 apoptosis, rescuing cDC1 abundance in tumor-bearing mice. CD8+ T cell response to vaccination is impaired as a result of cDC1 dysregulation. Yet, combination therapy with CD40 agonist and Flt3 ligand restores cDC1 abundance to normal levels, decreases cDC1 apoptosis, and repairs cDC1 maturation to drive superior control of tumor outgrowth. Our study therefore reveals the unexpectedly early and systemic onset of cDC1 dysregulation during pancreatic carcinogenesis and suggests therapeutically tractable strategies toward cDC1 repair.

Graphical abstract

Figure 1.

cDC1 abundance declines systemically during pancreatic carcinogenesis. (A) H&E staining of healthy pancreas, PanIN-bearing pancreas, and PDA. Arrows highlight ducts featuring mucinous metaplasia without dysplasia characteristic of PanIN 1A. All images are taken at 10× magnification. Scale bars denote 300 µm. (B) Flow gating strategy for CD45⁺CD64⁻F4/80⁻Lin⁻MHC II⁺CD11c⁺ cDCs in a representative subcutaneously implanted KPC tumor. Lineage gate is composed of CD3, CD19, B220, NK1.1, and Gr-1. FSC-A, forward scatter-area; FSC-H, forward scatter-height; LD, live/dead. (C–F) Quantification of cDC1s in the (C) pancreas/tumor, (D) ppLNs, (E) iLN, and (F) spleen as a proportion of live cells and CD45⁺ cells. (G) Frequency of CD141⁺ cDC1s in peripheral blood of patients with untreated advanced PDA versus healthy volunteers (HV). Error bars indicate mean ± SD. ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P < 0.05 (one-way ANOVA with Tukey’s honest significant difference [HSD] post-test in C–F; Mann–Whitney test in G). Data shown in B–F are representative of at least three independent experiments with at least three mice per group.

Figure S1.

cDC1 abundance only declines based on cell fractions during pancreatic carcinogenesis. (A–D) Tissue weight, cDC1 number per organ, and cDC1 number per milligram tissue in the (A) pancreas/tumor, (B) ppLNs, (C) iLNs, and (D) spleen from healthy, PanIN-bearing, and tumor-bearing mice. Error bars indicate mean ± SEM. ****, P < 0.0001; *, P < 0.05 (one-way ANOVA with Tukey’s HSD post-test). Data shown are representative of one independent experiment.

Figure 2.

cDC1 maturation marker expression declines systemically during preinvasive neoplasia. Expression of maturation markers CD40, CD80, CD86, MHC II (I-A/I-E), and PD-L1 on cDC1s in the (A) pancreas/tumor, (B) ppLNs, (C) iLNs, and (D) spleen of healthy, PanIN-bearing, and tumor-bearing mice. Geometric mean fluorescence intensities (MFIs) shown. Samples were pooled across three to six mice per treatment group in B. Error bars indicate mean ± SD. ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P < 0.05 (one-way ANOVA with Tukey’s HSD post-test). Data shown are representative of four independent experiments with at least three mice per group.

Figure 3.

cDC1 maturation is progressively impaired during pancreatic oncogenesis. (A) Principal-component (PC) analysis of cDC1s collected from healthy, PanIN-draining, and tumor-draining ppLNs. (B) Heatmap of differentially expressed genes by z-score across samples. (C) Top hits from GSEA comparing cDC1s from tumor-draining versus healthy ppLNs. (D) Enrichment plots of proteasome degradation and T cell–polarizing cytokine gene sets in cDC1s from GSEA shown in C. FDR, false discovery rate; NES, normalized enrichment score. (E and F) Expression in tpm reads of genes encoding (E) inflammatory cytokines and (F) immune suppressive factors in cDC1s from healthy, PanIN-draining, and tumor-draining ppLNs. n = 3 samples per group. Each sample consists of total RNA collected from 10,000 sorted ppLN cDC1s pooled from three to six mice. Error bars indicate mean ± SD. ***, P < 0.001; *, P < 0.05 (one-way ANOVA with Tukey’s HSD post-test).

Figure 4.

cDC1-mediated CD8⁺ T cell priming is impaired in PanIN- and tumor-bearing mice. (A) Generation of H-2Kb:SIINFEKL tetramer–positive splenic CD8⁺ T cells in healthy, PanIN-bearing, and tumor-bearing mice 7 d following subcutaneous implantation of 5 × 10⁵ cells from clonal OVA-expressing KPC cell line 4662.V6ova. (B) Quantification of H-2Kb:SIINFEKL tetramer–positive splenic CD8⁺ T cells from healthy, PanIN-bearing, and tumor-bearing mice 7 d following subcutaneous vaccination with 200 µg OVA + 10 µg CpG (OVA/CpG). (C) Activation/exhaustion marker expression in CD62L⁻CD44⁺ H-2Kb:SIINFEKL tetramer–positive CD8⁺ T cells following vaccination with OVA/CpG. gMFI, geometric mean fluorescence intensity. Error bars indicate mean ± SD. ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P < 0.05 (one-way ANOVA with Tukey’s HSD post-test). Data shown are representative of three independent experiments with at least three mice per group.

Figure S2.

cDC1 abundance and maturation are associated with increased cytolytic activity in human PDA. Correlation analyses of (A) XCR1 gene expression and cytolytic index (CTL), (B) CLEC9A gene expression and cytolytic index, (C) HLA-DRA gene expression and cytolytic index, (D) CD86 gene expression and cytolytic index, (E) HLA-DRA gene expression and IFNG gene expression, and (F) CD86 gene expression and IFNG gene expression in tumors of patients from the TCGA-PAAD. n = 182 total patients in TCGA-PAAD. Regression line, 95% confidence interval, Kendall’s τ rank correlation coefficient, and associated P value shown for all correlation analyses. Cytolytic index is calculated using the geometric mean of PRF1 and GZMA.

Figure S3.

Systemic cDC1 dysregulation requires neoplastic development. (A) H&E staining of pancreas from mice treated for 11 wk with PBS or cerulein. All images taken at 20× magnification. Scale bars denote 150 µm. (B and C) Enumeration of and expression of maturation markers CD40, CD80, CD86, MHC II, and PD-L1 on (B) iLN and (C) splenic cDC1s from PBS-treated and cerulein-treated mice. (D and E) Enumeration of and maturation marker expression on cDC1s from (D) pancreas and (E) ppLN cDC1s from 4-wk-old Cre/Cre (C/C) and KPC mice. Geometric MFIs are shown in E. (F) Proportions of CD11c^hiMHCII^int resident/resting versus CD11c^intMHCII^hi migratory/activated ppLN cDCs. (G and H) Proportion of cDC1s and cDC2s among (G) resident/resting and (H) migratory/activated ppLN cDCs shown in F. (I) Quantification of and Tbet and IFN-γ expression in H-2Kb:SIINFEKL tetramer–positive splenic CD8⁺ T cells 7 d following vaccination with 200 µg OVA + 10 µg CpG. Samples pooled across three mice per group in E. Error bars indicate mean ± SD. **, P < 0.01; *, P < 0.05 (two-tailed Student’s t test). Data shown in A-C are representative of one independent experiment. Data shown in D-E are representative of three independent experiments with at least three mice per group.

Figure S4.

Systemic cDC1 dysfunction does not occur in the KP mouse model of lung adenocarcinoma. (A) Enumeration of cDC1s as a proportion of live cells and CD45⁺ cells in the lung/tumor of AdFlp-treated controls and KP mice 8, 12, or 16 wk after inhalation of adenoviral Cre recombinase. (B–D) Enumeration of cDC1s as a proportion of CD45⁺ cells and total cDCs in the (B) mediastinal LN, (C) iLNs, and (D) spleen. (E and F) Expression of maturation markers CD40, CD80, CD86, MHC II, and PD-L1 on cDC1s from the (E) lung/tumor and (F) iLNs. (G) Serum levels of IL-6 and IL-1β as determined by cytokine bead array in the KP and KPC cancer mouse models, as well as cerulein-induced chronic pancreatitis. Samples pooled across four to seven mice per group in B. Error bars indicate mean ± SD. **, P < 0.01; *, P < 0.05 (one-way ANOVA with Tukey’s HSD post-test). Data shown are representative of one independent experiment.

Figure 5.

cDC1 generation is unaffected by pancreatic neoplastic development. (A) Flow gating strategy for MDPs, CDPs, pre-cDCs, pre-cDC1s, and pre-cDC2s in RBC-lysed bone marrow suspension from a wild-type C57BL/6J mouse. Lineage gate consists of CD3, CD19, B220, NK1.1, and Gr-1. FSC-A, forward scatter-area; FSC-H, forward scatter-height; SSC-A, side scatter-area. (B) Enumeration of MDPs, CDPs, pre-cDCs, pre-cDC1s, and pre-cDC2s in the bone marrow of healthy, PanIN-bearing, and tumor-bearing mice. (C) Enumeration of pre-cDCs, pre-cDC1s, and pre-cDC2s in peripheral blood. (D) Expression of Ki67 in cDC1s from the mLN and iLN of healthy, PanIN-bearing, and tumor-bearing mice. Error bars indicate mean ± SD. ****, P < 0.0001; **, P < 0.01; *, P < 0.05 (one-way ANOVA with Tukey’s HSD post-test). Data shown are representative of at least two independent experiments with at least three mice per group. gMFI, geometric mean fluorescence intensity.

Figure 6.

Increased serum IL-6 drives cDC1 apoptosis systemically in tumor-bearing KPC mice. (A and B) Percentage of (A) ppLN and (B) iLN cDC1s positive for expression of active cleaved caspase-3 in healthy, PanIN-bearing, and tumor-bearing mice. (C) Enrichment plot of apoptosis gene set in cDC1s from PanIN-draining versus healthy ppLNs. FDR, false discovery rate; NES, normalized enrichment score. (D) Expression of select genes in tpm reads from the gene set shown in C. (E) Serum IL-6 levels as determined by cytokine bead array in healthy mice, tumor-bearing mice, and tumor-bearing mice treated with IL-6–neutralizing antibody (MP5-20F3). (F) Enumeration of cDC1s in the mLN and iLN. (G) Percentage of mLN and iLN cDC1s positive for expression of cleaved caspase-3. (H) Representative histogram of cleaved caspase-3 expression in mLN cDC1s from G. (I and J) Percentage of CD64⁺F4/80⁺ macrophages and CD64⁻CD11b⁺ myeloid cells positive for expression of cleaved caspase-3 in the (I) mLN and (J) iLN. (K and L) Quantification of cDC1s as a percentage of live CD45⁺ cells in (K) iLN cDC1s and (L) splenic cDC1s from tumor-bearing KPC mice treated with IL-1β blocking monoclonal antibody (AF-401-NA). Samples pooled across at least four mice per treatment group in A. (C and D) Each sample consists of total RNA collected from 10,000 sorted ppLN cDC1s pooled across three to six mice. Error bars indicate mean ± SD. ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P < 0.05 (one-way ANOVA with Tukey’s HSD post-test in B, D, E–G, I, and J; two-tailed Student’s t test in K and L). Data shown are representative of at least two independent experiments with at least three mice per group.

Figure 7.

CD40 activation repairs cDC1 maturation in KPC tumors. (A) Timeline of subcutaneous implantation of KPC cell line 6419.c5, administration of CD40 agonist (FGK45), and harvest of tissues for flow cytometric analysis. (B) Enumeration of cDC1s per live cells in subcutaneous KPC tumors from untreated and FGK45-treated mice. (C) Enumeration of CD11c^intMHCII^hi migratory/activated cDC1s in the TdLN. (D) Expression of Ccr7 in CD11c⁺ cells purified from the iLNs of healthy mice and TdLNs of untreated and FGK45-treated mice bearing subcutaneously implanted KPC tumors. (E) Expression of maturation markers CD40, CD80, CD86, MHC II (I-A/I-E), and PD-L1 on cDC1s from the tumors of untreated and FGK45-treated mice. MFI, mean fluorescence intensity. (F and G) Maturation marker expression on cDC1s from the (F) TdLN and (G) spleen of healthy mice, untreated tumor-bearing mice, and FGK45-treated tumor-bearing mice. (H) Enumeration of H-2Kb:SIINFEKL tetramer–positive splenic CD8⁺ T cells from healthy mice, untreated tumor-bearing KPC mice, and FGK45-treated tumor-bearing KPC mice 12 d following subcutaneous implantation of OVA-expressing clonal KPC cell line 4662.V6ova. 100 µg FGK45 was administered on day 9 after implantation. Error bars indicate mean ± SD. ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P < 0.05 (two-tailed Student’s t test in B, C, E; one-way ANOVA with Tukey’s HSD post-test in D, F, G, H). Data shown are representative of four independent experiments with at least three mice per group.

Figure 8.

CD40-driven cDC1 maturation is associated with an IFN-γ response signature. (A) Principal-component (PC) analysis of iLN cDC1 transcriptomes in the presence or absence of subcutaneously implanted KPC tumor, either treated or untreated with CD40 agonist (FGK45). (B) Heatmap comparing expression of differentially expressed genes across samples, scaled by z-score. (C) Top hits from GSEA of TdLN cDC1s from FGK45-treated versus untreated mice. (D) Enrichment plot of type II interferon response gene set from GSEA shown in C. FDR, false discovery rate; NES, normalized enrichment score. (E) Expression of Stat1 and Stat2 in tpm reads from gene set shown in D. n = 3 samples per group. Each sample consisted of total RNA collected from 10,000 sorted iLN cDC1s pooled from five mice per group. Error bars indicate mean ± SD. **, P < 0.01; *, P < 0.05 (one-way ANOVA with Tukey’s HSD post-test).

Figure 9.

Flt3L synergizes with CD40 activation to promote cDC1 survival and function. (A) Timeline of treatment of mice subcutaneously implanted with 3 × 10⁵ KPC cell line 4662.MD10 with CD40 agonist (FGK45) and Flt3L. Treatment was initiated 14 d after transplant. (B) Enumeration of cDC1s in the tumor microenvironment, TdLN, and spleen of untreated, Flt3L-treated, FGK45-treated, and combination-treated mice. (C) Expression of MHC II, CD80, and CD86 on TdLN cDC1s. gMFI, geometric mean fluorescence intensity. (D and E) Percentage of cDC1s positive for expression of active cleaved caspase-3 in the (D) ppLN (percentages in healthy and tumor-bearing mice are also reported in Fig. 6 A) and (E) iLNs of healthy mice, tumor-bearing KPC mice, and tumor-bearing KPC mice treated with FGK45 and Flt3L. (F) Enumeration of and IFN-γ expression in H-2Kb:SIINFEKL tetramer–positive splenic CD8⁺ T cells 7 d following subcutaneous vaccination with 200 µg OVA + 10 µg CpG in tumor-bearing KPC mice treated with FGK45 and Flt3L. Samples were pooled across at least four mice per treatment group in D. Error bars indicate mean ± SD. ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P < 0.05 (one-way ANOVA with Tukey’s HSD post-test). Data shown are representative of at least two independent experiments with at least three mice per group.

Figure 10.

Combination therapy with CD40 agonist and Flt3L results in superior T cell activation in the tumor-draining LN. (A) Enumeration of CD8⁺ T cells in the tumor microenvironment of untreated, Flt3L-treated, CD40 agonist (FGK45)-treated, or combination-treated subcutaneously implanted KPC tumors as shown in Fig. 9 A. (B) Enumeration of and IFN-γ production in CD8⁺ T cells from the TdLN. (C) Enumeration of FOXP3⁻ CD4⁺ T cells in the tumor microenvironment. (D) Enumeration of and IFN-γ production in FOXP3⁻ CD4⁺ T cells from the TdLN. (E) Enumeration of FOXP3⁺ CD4⁺ T cells in the tumor microenvironment. (F and G) Tumor growth (F) and survival curves (G) from mice subcutaneously implanted with 5 × 10⁵ KPC cell line 6419c5. Mice were treated with CD40 agonist and Flt3L beginning on day 12 after implantation using the treatment schedule shown in Fig. 9 A (corresponds to Fig. S5). n = 10 mice per group (F and G). Error bars indicate mean ± SEM. ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P < 0.05 (one-way ANOVA with Tukey’s HSD post-test in A–E; two-way ANOVA with Tukey’s HSD post-test in F; pairwise Kaplan–Meier survival log-rank test in G). Data shown are representative of three independent experiments with at least five mice per group.

Figure S5.

Tumor growth curves from subcutaneous implantation of 6419c5 and combination treatment with CD40 agonist and Flt3L. Individual tumor growth curves following subcutaneous implantation of 5 × 10⁵ T cell low KPC cell line 6419c5 in (A) untreated, (B) Flt3L-treated, (C) CD40 agonist-treated, and (D) combination-treated mice. CD40 agonist and Flt3L were administered beginning day 12 after implantation using the treatment schedule shown in Fig. 9 A. Data shown are representative of three independent experiments with at least five mice per group (corresponds to Fig. 10, F and G).