Targeting the Dendritic Cell-Secreted Immunoregulatory Cytokine CCL22 Alleviates Radioresistance

Abstract

Purpose: Radiation-mediated immune suppression limits efficacy and is a barrier in cancer therapy. Radiation induces negative regulators of tumor immunity including regulatory T cells (Treg). Mechanisms underlying Treg infiltration after radiotherapy (RT) are poorly defined. Given that conventional dendritic cells (cDC) maintain Treg, we sought to identify and target cDC signaling to block Treg infiltration after radiation.

Experimental design: Transcriptomics and high dimensional flow cytometry revealed changes in murine tumor cDC that not only mediate Treg infiltration after RT but also associate with worse survival in human cancer datasets. Antibodies perturbing a cDC-CCL22-Treg axis were tested in syngeneic murine tumors. A prototype interferon-anti-epidermal growth factor receptor fusion protein (αEGFR-IFNα) was examined to block Treg infiltration and promote a CD8+ T cell response after RT.

Results: Radiation expands a population of mature cDC1 enriched in immunoregulatory markers that mediates Treg infiltration via the Treg-recruiting chemokine CCL22. Blocking CCL22 or Treg depletion both enhanced RT efficacy. αEGFR-IFNα blocked cDC1 CCL22 production while simultaneously inducing an antitumor CD8+ T cell response to enhance RT efficacy in multiple EGFR-expressing murine tumor models, including following systemic administration.

Conclusions: We identify a previously unappreciated cDC mechanism mediating Treg tumor infiltration after RT. Our findings suggest blocking the cDC1-CCL22-Treg axis augments RT efficacy. αEGFR-IFNα added to RT provided robust antitumor responses better than systemic free interferon administration and may overcome clinical limitations to interferon therapy. Our findings highlight the complex behavior of cDC after RT and provide novel therapeutic strategies for overcoming RT-driven immunosuppression to improve RT efficacy. See related commentary by Kalinski et al., p. 4260.

Figure 1.. Local irradiation expands a population of mature cDC enriched in regulatory markers within the tumor.

A, Heatmap of top 20 differentially expressed genes for each dendritic cell cluster identified. MC38 tumors were treated with 15 Gy ionizing radiation and live CD45⁺ cells sorted 4 days after treatment for scRNA-seq. B, Dendritic cell clusters identified from DC_Ccl22 and DC_Atox1 revealed 4 distinct dendritic cell populations (most closely resembling): DC_Ccl22 (mregDC), DC_Ly6a (cDC2), DC_Cd7 (pDC), and DC_Rab7b (cDC1). C, Stratification of cDC clusters using mregDC signature scores generated from gene lists provided in Maier, et al. D, Changes in cDC cell clusters after 15 Gy irradiation as a percentage of total cDC.

Figure 2.. Radiation increases tumor mregDC1.

A, UMAP of cDC1 and cDC2 from MC38 tumors (pooled from No Treatment, 3, 5 and 7 days after 20 Gy) using spectral flow cytometry. Unsupervised clustering with FlowSOM reveals 4 subpopulations of tumor cDC, including both resting cDC lineages (cDC1 and cDC2) and mregDC (mregDC1 and mregDC2). Color scale corresponds to either high (yellow) or low (dark blue) marker expression as measured by normalized MFI (scaled mean). B, UMAP of cDC from B16-OVA tumors (pooled from No Treatment and 5 days after 20 Gy) using spectral flow cytometry. Unsupervised clustering (FlowSOM) revealed 4 subpopulations of tumor cDC with similar marker expression patterns as observed in panel A. C, mregDC1 and mregDC2 clusters exhibit increased surface staining for CCR7 and H2Kb-SIINFEKL. D, Changes in cDC populations identified in panel b from B16-OVA tumors 5 days after treatment with 20 Gy. Data presented as mean ± SEM. Panel C: Mixed-effects analysis with Tukey’s multiple comparisons. Panel D: Unpaired t-test. ns: not significant, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 3.. Regulatory T cell-mediated immune suppression after radiotherapy is driven by mregDC1 production of CCL22.

A, Tumor Treg numbers in B16 tumors 5 days after treatment with either 8 Gy or 20 Gy as determined by flow cytometry. B, Treg in B16 tumors 5 days after treatment with 20 Gy in combination with αCCL22 (4 μg/mL_tumor was administered by intratumoral injection 1 day and 4 days after a single dose of 20 Gy). C, CCL22 protein in B16 tumors 3 days after treatment with 20 Gy. D, Quantification of intracellular CCL22 staining in splenic YFP⁺ CD24⁺ cDC1 or YFP⁺ CD172a⁺ cDC2 isolated from Zbtb46^cre Rosa^YFP mice cocultured in the presence of either non-irradiated or irradiated B16 tumor cells for 18 h. E, Treg in B16 tumors grown in either WT or Batf3^−/− mice isolated 5 days after treatment with 20 Gy analyzed by flow cytometry. F, Tumor growth of B16F1 in either WT or Batf3^−/− mice. 20 Gy was given on day 11. G-H, Tumor growth and survival of mice bearing B16 tumors treated with 20 Gy in combination with either αCCL22 or αCTLA4 clone 9H10. αCCL22 was administered intratumoral at a concentration of 4 μg/mL_tumor as determined by caliper measurement 3 days per week for 2 weeks beginning 1 day after RT (day 10). αCTLA4 9H10 was administered intraperitoneal (IP) as 200 μg on day 10 with RT and day 16. For image clarity, significant survival comparisons between No Treatment vs. 20 Gy + αCCL22 (P < 0.01) and No Treatment vs. 20 Gy + αCTLA4 (P < 0.01) were not shown on the figure. Data presented as mean ± SEM. Panel C: Mixed-effects analysis with Tukey’s multiple comparisons. Panels A, D, E: One-way ANOVA with Tukey’s multiple comparisons. Panels B, C, F (day 21): Unpaired t-test. Panel G (day 24, all groups): Mixed-effects model with Geisser-Greenhouse correction and Tukey’s multiple comparisons at day. Panel G (day 31, remaining groups): One-way ANOVA with Dunnett’s multiple comparisons. Panel H: Log-rank test. ns: not significant, *P < 0.05, **P < 0.01, ****P < 0.0001.

Figure 4.. Tumor-targeted type I interferon reprograms tumor myeloid populations after radiotherapy away from immune suppression.

A, Quantification of intracellular CCL22 staining in splenic cDC1 isolated from Zbtb46^cre Rosa^YFP mice and cocultured with irradiated B16 cells in the presence of IFNβ 10 μg/mL for 18 h. B, Structure of the αEGFR-IFNα fusion protein. C, Biological pathways enriched in CD11c⁺ cells isolated from tumors after treatment with the combination of 20 Gy and αEGFR-IFNα compared with RT alone. Mice were given 20 μg αEGFR-IFNα intratumoral on same day as RT and 10 μg 4 days later. Tumors were collected 7 days after RT and cells were sorted as Live CD45⁺ CD11c⁺ for RNA-sequencing. Chart produced using ShinyGO 0.76.1. D, Normalized gene expression in CD11c⁺ cells of the DC_Ccl22 cluster signature genes identified in Figure 1A after concurrent treatment with αEGFR-IFNα and RT. E, Relative expression of Ccl22 in Live CD45⁺ CD11c⁺ Ly6C⁻ cells isolated from B16-EGFR tumors 3 d after local RT and 20 μg intratumoral αEGFR-IFNα. Data presented as mean ± SEM. Panels A and E: One-way ANOVA with Tukey’s multiple comparisons. ns: not significant, *P < 0.05, **P < 0.01.

Figure 5.. Tumor-targeted type I interferon remodels the tumor cDC population and induces a population of ISG⁺ cDC2.

A, Changes in cDC populations by flow cytometry isolated from B16-EGFR tumors 5 days after 20 Gy as determined by conventional gating strategies. αEGFR-IFNα was administered intratumoral as 20 μg on the same day as RT. B-C, Identification of cDC cell clusters by flow cytometry 5 days after 20 Gy. αEGFR-IFNα was administered intratumoral as 20 μg on the same day as RT. Clusters were identified by unsupervised clustering with FlowSOM. Surface marker expression (B) and dimensionality reduction (C) were used to determine cDC populations and was consistent with induction of a recently described population of interferon-stimulated cDC2 (ISG⁺ cDC2). D, Changes in the proportions of cDC clusters identified in panel B and C across treatment groups. Data presented as mean ± SEM. Panels A and D: One-way ANOVA with Tukey’s multiple comparisons. Panel B: Two-way ANOVA with Tukey’s multiple comparisons. ns: not significant, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 6.. Tumor-targeted type I interferon alleviates CCL22-driven Treg immune suppression after radiotherapy to augment radiation efficacy.

A, CCL22 protein measurements in B16-EGFR tumors 3 days after treatment with 20 Gy with or without concurrent αEGFR-IFNα. Intratumoral αEGFR-IFNα was given as 20 μg on the day of RT. P = 0.06 between No Treatment and 20 Gy + αEGFR-IFNα. B, Treg as measured by flow cytometry in B16-EGFR tumors isolated 5 days after irradiation with 20 Gy. Intratumoral αEGFR-IFNα was administered as 20 μg on the day of irradiation and 10 μg four days later. C, CD8⁺ T cells as measured by flow cytometry in B16-EGFR tumors 7 days after 20 Gy with or without αEGFR-IFNα administered as described in panel B. D, Ex vivo IFNγ ELISPOT assay of CD8+ T cells isolated from OVA-expressing B16-EGFR tumors 7 days after local RT with or without αEGFR-IFNα and pulsed with the OVA peptide. αEGFR-IFNα administered as described in panel B. E-F, Tumor growth and survival of mice bearing B16-EGFR tumors treated with 20 Gy with or without intratumoral αEGFR-IFNα. αEGFR-IFNα was given as described in panel B. G-H, Tumor growth and survival of mice bearing B16-EGFR tumors treated with 20 Gy with or without intravenous (IV) αEGFR-IFNα. αEGFR-IFNα was given using the same schedule as in panel B. IFNα4 was given IV at a molar dose equivalent to αEGFR-IFNα using the same dosing schedule as panel B. Data presented as mean ± SEM. Panels A-D: One-way ANOVA with Tukey’s multiple comparisons. Panels E and G at day 21: Two-way ANOVA with Tukey’s multiple comparisons. Panels F and H: Log-rank test. ns: not significant, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.