Type I interferon activates MHC class I-dressed CD11b+ conventional dendritic cells to promote protective anti-tumor CD8+ T cell immunity

Abstract

Tumor-infiltrating dendritic cells (DCs) assume varied functional states that impact anti-tumor immunity. To delineate the DC states associated with productive anti-tumor T cell immunity, we compared spontaneously regressing and progressing tumors. Tumor-reactive CD8⁺ T cell responses in Batf3^-/- mice lacking type 1 DCs (DC1s) were lost in progressor tumors but preserved in regressor tumors. Transcriptional profiling of intra-tumoral DCs within regressor tumors revealed an activation state of CD11b⁺ conventional DCs (DC2s) characterized by expression of interferon (IFN)-stimulated genes (ISGs) (ISG⁺ DCs). ISG⁺ DC-activated CD8⁺ T cells ex vivo comparably to DC1. Unlike cross-presenting DC1, ISG⁺ DCs acquired and presented intact tumor-derived peptide-major histocompatibility complex class I (MHC class I) complexes. Constitutive type I IFN production by regressor tumors drove the ISG⁺ DC state, and activation of MHC class I-dressed ISG⁺ DCs by exogenous IFN-β rescued anti-tumor immunity against progressor tumors in Batf3^-/- mice. The ISG⁺ DC gene signature is detectable in human tumors. Engaging this functional DC state may present an approach for the treatment of human disease.

Keywords: MHC class I dressing; T cell exhaustion; anti-tumor immunity; cross-presentation; dendritic cells; immunotherapy; optimal anti-tumor immunity; tumor regression; type 2 dendritic cells; type I interferon.

Figure 1.. The regression of MC57-SIY tumors is independent of Batf3-driven DC1

(A) Representative tumor outgrowth in WT mice (n = 3–4 mice/group; three independent repeats). (B and C) Representative flow plot (B) and quantification (C) of CD103⁺ DC1s and CD11b⁺ DC2s (pre-gated on live CD45⁺MHC class II⁺Ly6C⁻F4/80⁻CD11c⁺CD24^hi) in tumors at day 7 after tumor inoculation in WT mice. Data were pooled from two independent experiments (n = 3–4 mice/group). (D) Representative tumor outgrowth in Rag2^−/− mice (n = 3–5 mice/group; three independent repeats). (E and F) Representative flow plot (E) and quantification (F) of CD103⁺ DC1s and CD11b⁺ DC2s in tumors at day 15 after tumor inoculation in Rag2^−/− mice. Data were pooled from two independent experiments (n = 3 mice/group). (G) Representative tumor outgrowth in WT or Batf3^−/− mice (n = 3–4 mice/group; three independent repeats). (H) Quantification of SIY-specific CD8⁺ T cells in tumors at day 7 after tumor implantation in Batf3^−/− mice. Data were pooled from two independent experiments (n = 2–5 mice mice/group). (I) ELISpot quantification of IFN-γ-producing splenocytes at day 5 after tumor inoculation in WT and Batf3^−/− mice. Data were pooled from two independent experiments (n = 3–4 mice/group). Data are shown as mean ± SEM. *p < 0.05, **p < 0.01, ****p < 0.0001; ns, not significant: Mann-Whitney U (MWU) test (C, F, H, and I) or two-way ANOVA (A, D, and G).

Figure 2.. Functional assays and scRNA-seq identify a DC cluster characterized by an IFN-I gene signature in MC57-SIY tumors

(A) Experimental design for (B) and (C). (B and C) Percentage of 2C T cell proliferation after co-culture with tumor-sorted APCs in WT (B) or Batf3^−/− (C) mice at day 5 after tumor inoculation (n = 5 mice/experiment; two independent repeats). (D) ELISpot of IFN-γ-producing splenocytes from DT-treated or PBS-treated Itgax-DTR BMC mice at day 5 after tumor inoculation. Data were pooled from two independent experiments (n = 3 mice/group). (E) Number of SIY-reactive tumor-infiltrating lymphocytes (TILs) in tumors from cDC-depleted (zDC-DTR) or non-depleted (WT) BMC mice at day 7 after tumor inoculation (n = 5 mice/group). (F) (Top) UMAP plot of cells from MC57-SIY tumors colored by expression module score of a DC signature. (Bottom) UMAP plot of cells contained in the highlighted DC cluster. (G) Heatmap of top 15 DEGs for each DC cluster identified in (F). (H) Tumor outgrowth (mm²) in WT or Ifnar1^−/− mice (n = 3–4 mice/group; three independent experiments). (I) Experimental design for (J). (J) ELISpot of IFN-γ-producing splenocytes from Itgax-DTR:Ifnar1^−/− mixed BMC mice (WT hosts, left; Ifnar1^−/− hosts, right) at day 7 after tumor inoculation. Data were pooled from two independent experiments (n = 2–3 mice/group). Data are shown as mean ± SEM. **p < 0.01, ****p < 0.0001: MWU test (B–E and J) or two-way ANOVA (H).

Figure 3.. ISG⁺ DCs are present in Batf3^–/– mice and comprise an activation state of CD11b⁺ DCs

(A) Violin plots of expression in DC clusters. (B and C) Flow gating strategy for APCs and DCs in WT (B) or Batf3^−/− (C) mice, pre-gated on live CD45⁺CD19⁻CD3e⁻NK1.1⁻ cells. (D–I) Quantification of DC subsets in tumors from WT (D and E) and Batf3^−/− (F and G) mice at day 7 after tumor inoculation, and in Rag2^−/− mice (H and I) at day 11 after tumor inoculation. Data were pooled from two independent experiments (n = 3–5 mice/group). (J) Experimental design for (K)–(M) and Figures S3B–S3D. (K) Feature UMAP plots of DC clusters. Each cell is colored by an expression module score. (L) Cluster-wise enrichment of cells scoring for the indicated bulk RNA-seq ISG⁺ DC signatures. Dotted line denotes significance threshold (p = 0.05, hypergeometric test). (M) GSEA plot showing highly significant enrichment of Rag2^−/− ISG⁺ DC signature in WT ISG⁺ DC signature. ES, enrichment score; NES, normalized enrichment score; FDR, false discovery rate. For (K)–(M), data are from two independent experiments (n = 5 mice/group). (N) Representative histograms showing myeloid marker expression on DC subsets in tumors at day 7 post-implantation in WT mice (n = 3 mice/group; two independent repeats). (O and P) Representative flow plot (O) and absolute numbers (P) of ISG⁺ DCs at day 7 after tumor implantation in Irf4^f/fxItgax^Cre or littermate control mice (n = 4–6 mice/group; two independent experiments). (Q) Cluster-wise enrichment of cells that scored highly for the scRNA-seq-derived ISG⁺ DC signature. Dotted line denotes significance threshold (p = 0.05, hypergeometric test). (R) UMAP plot of human tumor-infiltrating DC2 subsets. Each cell is colored by its expression module score of the ISG⁺ DC signature. Data are shown as mean ± SEM. **p < 0.01; MWU test (P).

Figure 4.. ISG⁺ DCs acquire and present tumor antigens by MHC class I dressing

(A) Experimental design for (B). (B) (Left) Replication index of 2C T cells after co-culture with tumor-sorted DCs from Rag2^−/− mice at day 11 after tumor inoculation. Data were pooled from three independent experiments (n = 5 mice/experiment). (Right) Representative histogram. (C) Experimental design for (D). (D) ELISpot quantification of IFN-γ-producing splenocytes at day 7 after tumor inoculation in WT or Batf3^−/− mice. Data were pooled from three independent experiments (n = 3–4 mice/group). (E) Experimental design for (F) and (G). (F and G) Quantification (F) and representative histograms (G) of H-2K^b expression on DC subsets infiltrating tumors in WT or B2M^−/− BMC mice. (H) ELISpot of IFN-γ-producing splenocytes from WT or B2M^−/− BMC mice at day 7 after tumor inoculation. For (F) and (H), data were pooled from two independent experiments (n = 4 mice/group). Data are shown as mean ± SEM. *p < 0.05, ***p < 0.001; ns, not significant: MWU test (B, D, F, and H).

Figure 5.. MHC class I-dressed ISG⁺ DCs can induce protective systemic anti-tumor T cell immunity

(A) Experimental design for (B) and (C). (B and C) Percentage (B) and representative histogram (C) of H-2K^b:SIIN expression on DC subsets in BALB/c mice at day 5 after tumor implantation. Data were pooled from three independent experiments (n = 3–4 mice/experiment). (D) Experimental design for (E) and (F). (E and F) Replication index (E) and representative histogram (F) of 2C T cell proliferation after co-culture with tumor-sorted DCs from BALB/c mice at day 5 after tumor inoculation. Data were pooled from three independent experiments (n = 5 mice/experiment). (G) Experimental design for (H). (H) Tumor outgrowth of MC38-SIY in Batf3^−/− mice that were pre-inoculated with MC57-SIY or PBS on the contralateral flank. Numbers in parentheses indicate tumor-free mice. Data were pooled from five independent experiments (n = 2–5 mice/group). (I) Experimental design for (J). (J) Tumor outgrowth of MC38-SIY in Batf3^−/− mice that were pre-inoculated with MC57-SIY-B2M^−/− or PBS on the contralateral flank. Numbers in parentheses indicate tumor-free mice. Data were pooled from two independent experiments (n = 4 mice/group). Data are shown as mean ± SEM. *p < 0.05, **p < 0.01: MWU test (B and E).

Figure 6.. IFNAR signaling in the MC57-SIY TME drives ISG⁺ DC activation

(A) Number of ISG⁺ DCs in tumors at day 11 following implantation in Rag2^−/− mice. Data were pooled from two independent experiments (n = 2–3 mice/group). (B) Experimental design for (C). (C) (Left) Replication index of 2C T cells after co-culture with tumor-sorted DCs from Rag2^−/− mice at day 11 after tumor inoculation. Data were pooled from two independent experiments (n = 5 mice/experiment). (Right) Representative example. (D) Representative tumor outgrowth (mm²) in WT or Sting1^−/− mice (n = 3–5 mice/group; three independent experiments). (E) Relative expression of IFNβ1, Irf7, and Isg15 in tumor cells. Data were pooled from IFNβ1 (n = 6), Irf7 (n = 7), and Isg15 (n = 3) independent experiments. (F) Expression level of IFNβ1, Irf7, and Isg15 in BM-DCs that were unstimulated or cultured with tumor-conditioned media. Data were pooled from two independent experiments. (G) Representative tumor outgrowth in Batf3^−/− mice (n = 3–4 mice/group; three independent repeats). (H) (Left) Geometric mean fluorescence intensity (gMFI) and (right) representative histogram of CD86 expressed by DC subsets from Rag2^−/− mice at day 11 after tumor implantation (n = 4–5 mice/group; two independent repeats). (I) Experimental design for (J). (J) Ratio of the gMFI values of CD86 expressed by CD45.1⁺:CD45.2⁺ DC subsets. Data were pooled from two independent experiments (n = 5 mice/experiment). (K and L) Number of SIY-specific TILs (K) and ELISpot quantification of IFN-γ-producing splenocytes (L) from WT and Batf3^−/− mice at day 7 after tumor implantation. Data were pooled from three independent experiments (n = 3 mice/group). Data are shown as mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001: MWU test (A, C, E, F, H, and J–L) or two-way ANOVA (D and G).

Figure 7.. Exogenous addition of IFN-β to progressor tumors restores anti-tumor T cell responses in Batf3^–/– mice via activation of MHC class I-dressed ISG⁺ DCs

(A and B) Representative expression level of IFNβ1 in murine (A) and human (B) tumor cell lines (two independent experiments). (C) Experimental design for (D). (D) ELISpot quantification of IFN-γ-producing splenocytes from WT and Batf3^−/− mice at day 7 after tumor implantation. Data were pooled from seven independent experiments (n = 3–5 mice/group). (E) Experimental design for (F). (F) ELISpot quantification of IFN-γ-producing splenocytes from WT or Batf3^−/− mice at day 7 after tumor implantation. Data were pooled from three independent experiments (n = 3–5 mice/group). Data are shown as mean ± SEM. ***p < 0.001, ****p < 0.0001: MWU test (D and F).