Tumor-infiltrating CCR2+ inflammatory monocytes counteract specific immunotherapy

Abstract

Tumor development and progression is shaped by the tumor microenvironment (TME), a heterogeneous assembly of infiltrating and resident host cells, their secreted mediators and intercellular matrix. In this context, tumors are infiltrated by various immune cells with either pro-tumoral or anti-tumoral functions. Recently, we published our non-invasive immunization platform DIVA suitable as a therapeutic vaccination method, further optimized by repeated application (DIVA²). In our present work, we revealed the therapeutic effect of DIVA² in an MC38 tumor model and specifically focused on the mechanisms induced in the TME after immunization. DIVA² resulted in transient tumor control followed by an immune evasion phase within three weeks after the initial tumor inoculation. High-dimensional flow cytometry analysis and single-cell mRNA-sequencing of tumor-infiltrating leukocytes revealed cytotoxic CD8⁺ T cells as key players in the immune control phase. In the immune evasion phase, inflammatory CCR2⁺ PDL-1⁺ monocytes with immunosuppressive properties were recruited into the tumor leading to suppression of DIVA²-induced tumor-reactive T cells. Depletion of CCR2⁺ cells with specific antibodies resulted in prolonged survival revealing CCR2⁺ monocytes as important for tumor immune escape in the TME. In summary, the present work provides a platform for generating a strong antigen-specific primary and memory T cell immune response using the optimized transcutaneous immunization method DIVA². This enables protection against tumors by therapeutic immune control of solid tumors and highlights the immunosuppressive influence of tumor infiltrating CCR2⁺ monocytes that need to be inactivated in addition for successful cancer immunotherapy.

Keywords: CCR2 monocytes +; cancer immunotherapy; immune evasion; transcutaneous immunization; tumor micro environment (TME).

Figure 1

Therapeutic DIVA² induces transient tumor control that turns into immune evasion. (A) Application pattern for DIVA² in a therapeutic tumor setting. Mice were immunized twice one and two weeks after tumor implantation using DIVA² or left untreated and the tumor volume was monitored three times per week. In this setting, two independent experiments were performed. (B) Tumor volumes were assessed three times per week until day 16 (green lines) or until day 27 (red lines). Every curve represents the tumor volume of one individual animal (n=11-15). (C) Tumor volumes during immune control phase on day 16 and immune evasion phase on day 22 were displayed. Visualized are individual data points, mean and SD. *p < 0.05 by two-way ANOVA with Sidak’s multiple comparisons test and one-way ANOVA with Kruskal-Wallis test, when sample numbers were different.

Figure 2

DIVA²-induced tumor-infiltrating CD8⁺ T cells exhibit an activated and functional phenotype. (A) Cell counts of tumor-infiltrating CD45⁺ cells, CD8⁺ T cells, (B) specific CD8⁺ T cells and frequencies of their PD-1, CLTA-4 and Lag-3 expression were assessed by flow cytometry during immune control (day 16) and immune evasion (day 22) (n=11-15). Visualized are individual data points, mean and SD. (C) Ex vivo tumor cell suspensions were restimulated for 20 h with OVA_257-264 or left unstimulated to determine IFN-γ production by ELISpot assay. p< 0.05 by one-way ANOVA with Kruskal-Wallis test. The Flow cytometric gating strategy is shown in Supplementary Figure 3 .

Figure 3

Single-cell RNA-sequencing analysis reveals monocytes to be absent during DIVA²-induced immune control. (A) Application pattern for DIVA² in a therapeutic tumor setting. Tumor cell suspensions were prepared during immune control (day 16) or immune evasion (day 20). Tumor-infiltrating leukocytes were prepared by MACS isolation of CD45⁺ cells. (B) Tumor volumes during immune control (day 16, green line) and immune evasion (day 20, red line). Visualized are the means and SD. *p<0.05 by two-way ANOVA with Sidak’s multiple comparisons test. Statistics were analyzed on day 16 and day 20, compared to the non-immunized control groups. (C) scRNA-seq-based t-SNE plots of tumor-infiltrating leukocytes, merged per condition (n=2-3). Cell types were predicted based on the immgen database annotation immgen main. (D) Quantitative distribution of tumor-infiltrating leukocytes per immune cell type and condition.

Figure 4

DIVA² treatment induced mainly cytotoxic CD8⁺ T cells with a mild exhaustion characteristic. (A) scRNA-seq-based t-SNE plots of tumor-infiltrating leukocytes showing signature score of cytotoxic gene signature. (B) Signature score of cytotoxic gene signature, split by cytotoxic lymphocyte subtype. (C) Quantitative distribution of cytotoxic lymphocyte subtypes. (D) scRNA-seq-based expression analysis of indicated exhaustion marker genes by CD8⁺ T cells. All samples are merged per condition (n=2-3).

Figure 5

Inflammatory CCR2⁺ Monocytes infiltrating the TME during immune evasion express immunosuppressive marker genes. (A) FlowSOM Map of CD45⁺ tumor-infiltrating immune cells and their predicted cell types. Cells were pre-gated on living cells, single cells, Lineage^- cells and CD45⁺ cells. Expression intensities were relatively set by the FlowSOM algorithm. t-SNE plots of CD45⁺ tumor-infiltrating immune cells, merged per condition (n=11-15). FACS Markers included in the t-SNE calculation are analogous to the markers in the FlowSOM map. For coloring, FlowSOM populations were applied onto the t-SNE plots. (B) scRNA-seq-based t-SNE plots of tumor-infiltrating leukocytes showing signature score of immunosuppressive gene signature. (C) Signature score of immunosuppressive gene signature, split by immune cell types. (D) scRNA-seq-based t-SNE plots of tumor-infiltrating leukocytes showing expression of indicated genes. All scRNA-seq samples are merged per condition (n=2-3). Flow cytometric gating strategies are shown in Supplementary Figure 3 . The flow cytometric gating strategy until gating of CD45⁺ lineage^- cells was performed according to the gating strategy of Figure 2 .

Figure 6

Depletion of CCR2⁺ Monocytes in a therapeutic tumor setting leads to a decreased tumor growth demonstrating their immunosuppressive capacity. (A) Schematic overview of the application pattern for Boost DIVA in a therapeutic tumor setting. DIVA²-treated or untreated mice were i.v. injected with anti-CCR2 antibody MC-21 from day 15-19 (20 µg daily) or left untreated (n=4-9). (B) Representative flow cytometry dot plots of LY6C⁺ CCR2⁺ peripheral blood cells of an untreated and anti-CCR2 treated mouse. (C) Tumor volumes were assessed three times per week. Every curve represents the tumor volume of one individual mouse. (D) Tumor volumes visualized as mean and SD per condition. (E) Kaplan-Meier survival curve. p < 0.05 by two-way ANOVA with Sidak’s multiple comparisons test and one-way ANOVA with Kruskal-Walli’s test, when sample numbers were different. Comparisons of survival curves were performed by Log-rank (Mantel-Cox) test.