Sustained inhibition of CSF1R signaling augments antitumor immunity through inhibiting tumor-associated macrophages

Abstract

Tumor-associated macrophages (TAMs) are one of the key immunosuppressive components in the tumor microenvironment (TME) and contribute to tumor development, progression, and resistance to cancer immunotherapy. Several reagents targeting TAMs have been tested in preclinical and clinical studies, but they have had limited success. Here, we show that a unique reagent, FF-10101, exhibited a sustained inhibitory effect against colony-stimulating factor 1 receptor by forming a covalent bond and reduced immunosuppressive TAMs in the TME, which led to strong antitumor immunity. In preclinical animal models, FF-10101 treatment significantly reduced immunosuppressive TAMs and increased antitumor TAMs in the TME. In addition, tumor antigen-specific CD8+ T cells were increased; consequently, tumor growth was significantly inhibited. Moreover, combination treatment with an anti-programmed cell death 1 (anti-PD-1) antibody and FF-10101 exhibited a far stronger antitumor effect than either treatment alone. In human cancer specimens, FF-10101 treatment reduced programmed cell death 1 ligand 1 (PD-L1) expression on TAMs, as observed in animal models. Thus, FF-10101 acts as an immunomodulatory agent that can reduce immunosuppressive TAMs and augment tumor antigen-specific T cell responses, thereby generating an immunostimulatory TME. We propose that FF-10101 is a potential candidate for successful combination cancer immunotherapy with immune checkpoint inhibitors, such as PD-1/PD-L1 blockade.

Keywords: Immunology; Macrophages.

Figure 1. Theoretical conformational prediction reveals a stable binding of FF-10101 to CSF1R.

(A) Chemical structure of FF-10101. (B) Estimated stabilized multimeric model structure of FF-10101 and CSF1R. (C) Predicted binding conformation between FF-10101 and CSF1R. FF-10101 is shown in ball and stick notation, and the protein structure is shown in cartoon and wire notation. The arrowhead indicates the formation of a covalent bond. (D) Binding interaction analysis for FF-10101 and CSF1R. The covalent bond between FF-10101 and CSF1R is shown as a black line.

Figure 2. FF-10101 harbors a strong and durable inhibitory activity against CSF1R.

(A) Western blot analyses showing the inhibition of CSF1R signaling molecules by FF-10101 treatment in murine bone marrow–derived macrophages. Murine bone marrow–derived macrophages were incubated with the indicated concentration of FF-10101. (B) Western blot analyses showing the persistence of CSF1R inhibitory activity of FF-10101. RAW264 cells were incubated with the indicated concentrations of FF-10101 or BLZ945. The persistence of CSF1R inhibitory activity of FF-10101 or BLZ945 was examined with (red) or without (blue) drug removal. The culture conditions are shown at the top.

Figure 3. FF-10101 treatment enhances the phosphorylation of STAT1 and inhibits STAT3 signaling.

(A) Experimental scheme. BMDMs generated with both GM-CSF and M-CSF and with or without FF-10101 were stimulated with 50 ng/mL IFN-γ for the last hour. (B) Western blot analyses showing the changes in protein expression between BMDMs generated with or without FF-10101 (100 nM). (C) Contour plots for I-A/I-E on CD45⁺CD11b⁺F4/80⁺ cells. Flow cytometry (FCM) analysis was performed 24 hours after IFN-γ stimulation.

Figure 4. FF-10101 inhibits tumor growth by polarizing TAMs toward M1-like macrophages.

(A) Experimental scheme. One million tumor cells (MCA205 or MC38) were inoculated into the mice on day 0, and FF-10101 was administered from day 1. (B) Tumor growth curves for MCA205 (left; n = 5 per group) and MC38 (right; n = 4 per group) models. The tumor volumes are shown as the means ± SDs and were compared using 2-way ANOVA with multiple t tests corrected with Bonferroni’s method. Adjusted P value: * < 0.05, ** < 0.01, *** < 0.001. (C–F) Tumors were collected on day 8 and subjected to bulk RNA-sequencing analysis (n = 3 per group). (C) Tumor bulk RNA sequencing was evaluated by CYBERSORTx. (D) GSEA plots of the tumoricidal macrophage, TNFα signaling, inflammatory response, and interferon-γ response gene sets for the FF-10101–treated group compared with the control group. NES, normalized enrichment score; FDR, false discovery rate. (E) Volcano plot of differentially expressed genes between the FF-10101 group and the control group. Molecules with significantly high and low expression in the FF-10101–treated group compared with the control group are shown in red and blue, respectively. FC, fold-change. (F) Heatmap of representative M1- and M2-related genes.

Figure 5. FF-10101 treatment effectively reduces immunosuppressive TAMs.

(A) Experimental scheme. One million MCA205 cells were inoculated into the mice on day 0, and FF-10101 was administered from day 1. Immune cells were collected from tumors on day 8 and subjected to FCM analyses. (B) The frequency of the CD45⁺CD11b⁺F4/80⁺ fraction (TAMs) (n = 3 per group). (C) Representative contour plots of FRβ and CD204 on TAMs (left) and a summary of the frequency of FRβ⁺CD204⁺ TAMs (right; n = 3 per group). (D) Representative contour plots of PD-L1 and PD-L2 on TAMs (left) and summaries of the frequencies of PD-L1⁺ and PD-L2⁺ TAMs (right; n = 3 per group). (E) Representative contour plots of Arg-1 and iNOS on TAMs (left) and summaries of the frequencies of Arg-1⁺ and iNOS⁺ TAMs (right; n = 3 per group). The bar plots are shown as the means ± SDs and were compared by unpaired t tests. P values: NS ≥ 0.05, * < 0.05, ** < 0.01.

Figure 6. Antitumor T cell responses are induced by FF-10101 treatment.

(A) Experimental scheme. One million tumor cells (SIINFEKL-expressing MCA205 [MCA205-SIINFEKL] or MC38) were inoculated into the mice on day 0, and FF-10101 was administered from day 6. Some mice received intraperitoneal administration of anti-CD8β mAbs and anti-CD4 mAbs on days 4, 8, 12, and 16. To examine tumor antigen-specific CD8⁺ T cell responses, MCA205-SIINFEKL was employed. (B) Tumor growth curves for MCA205-SIINFEKL (left) and MC38 (right) models (n = 5 per group). The tumor volumes between the groups were compared using 2-way ANOVA with multiple t tests corrected with Bonferroni’s method. Adjusted P values: * < 0.05, ** < 0.01. (C) Experimental scheme. One million tumor cells were inoculated into the mice (wild-type or RAG2 KO) on day 0, and FF-10101 was administered from day 6. (D) Tumor growth curves for wild-type and RAG2 KO mice (n = 5 per group). The tumor volumes between the groups were compared using 2-way ANOVA with multiple t tests corrected with Bonferroni’s method. Adjusted P values: * < 0.05, ** < 0.01, *** < 0.001, **** < 0.0001. (E and F) Representative contour plots (left) and a summary (right; n = 3 per group) of the frequency of CD8⁺ T cells producing IFN-γ (E) and TNF-α (F) analyzed as in the experimental model shown in Figure 5A. (G) Representative contour plots for CD25 and Foxp3 in CD4⁺ T cells (left) and a summary of the frequency of Treg cells in CD3⁺ T cells (right; n = 7 per group). (H) The ratio of CD8⁺ T cells to Treg cells in tumor tissues (n = 7 per group). The bar plots are shown as the means ± SDs and were compared by unpaired t tests. P values: * < 0.05, ** < 0.01.

Figure 7. FF-10101 exhibits a long-lasting antitumor effect in vivo.

(A) Experimental scheme. One million MCA205 cells were subcutaneously inoculated into the mice on day 0, and drugs (FF-10101 or BLZ945) were administered from day 1. In a short-term study, FF-10101 or BLZ945 was only administered from day 1 to day 8. Tumor tissues and draining lymph nodes (dLNs) were extracted on day 10 and subjected to FCM analyses. (B) Tumor growth curves for long-term groups (solid lines) and short-term groups (dashed line) (n = 10 per group). The tumor volumes between the groups were compared using 2-way ANOVA with multiple t tests corrected with Bonferroni’s method. Adjusted P values: NS ≥ 0.05, * < 0.05, ** < 0.01, *** < 0.001. (C) Summaries of the frequency of FRβ⁺CD204⁺ TAMs in CD45⁺ cells (n = 3 per group). The data are shown as the means ± SDs and were compared by unpaired t test. P values: NS ≥ 0.05, ** < 0.01, *** < 0.001.

Figure 8. FF-10101 treatment exhibits antitumor activity by targeting TAMs in mice and humans.

(A) Experimental scheme. One million MCA205-SIINFEKL cells were inoculated into the mice on day 0. Some mice received FF-10101 treatment from day 6 and/or anti–PD-1 mAb treatment on days 6 and 9. (B) Tumor growth curves for control mice and mice that received FF-10101 treatment and/or anti–PD-1 mAb treatment (n = 12 per group). Comparisons between 2 groups were conducted by 2-way ANOVA with multiple t tests corrected with Bonferroni’s method. Adjusted P values: * < 0.05, **** < 0.0001. (C) Summaries of the frequency of tumor antigen-specific (SIINFEKL-tetramer⁺) CD8⁺ T cells in tumor tissues (left; n = 3 per group) and dLNs (right; n = 3 per group). (D) Summaries of the frequency of IFN-γ⁺TNF-α⁺CD8⁺ T cells in tumor tissues (left; n = 3 per group) and dLNs (right; n = 3 per group). The data are shown as the means ± SDs and were compared by unpaired t test. P values: * < 0.05. (E) Experimental scheme. The cells were extracted from primary tumor specimens and cultured with or without 10 nM FF-10101. (F) Reduction rates of the PD-L1⁺ TAMs (CD3^–CD11b⁺CD14⁺) in CD45⁺ cells (n = 9). Statistical analysis by paired t test; * P < 0.05. (G) Relative RNA expression of representative M1-related genes in TAMs was assessed by quantitative real-time PCR (n = 3).