Ovarian Cancer-Intrinsic Fatty Acid Synthase Prevents Anti-tumor Immunity by Disrupting Tumor-Infiltrating Dendritic Cells

Abstract

Fatty acid synthase (FASN), the key metabolic enzyme of de novo lipogenesis, provides proliferative and metastatic capacity directly to cancer cells have been described. However, the impact of aberrant activation of this lipogenic enzyme on host anti-tumor immune milieu remains unknown. In this study, we depicted that elevated FASN expression presented in ovarian cancer with more advanced clinical phenotype and correlated with the immunosuppressive status, which characterized by the lower number and dysfunction of infiltrating T cells. Notably, in a mouse model, we showed that tumor cell-intrinsic FASN drove ovarian cancer (OvCa) progression by blunting anti-tumor immunity. Dendritic cells (DCs) are required to initiate and sustain T cell-dependent anti-tumor immunity. Here, our data showed that constitutive activation of FASN in ovarian cancer cell lead to abnormal lipid accumulation and subsequent inhibition of tumor-infiltrating DCs (TIDCs) capacity to support anti-tumor T cells. Mechanistically, FASN activation in ovarian cancer cell-induced the resulting increase of lipids present at high concentrations in the tumor microenvironment. Dendritic cells educated by FASN^high OvCa ascites are defective in their ability to present antigens and prime T cells. Accordingly, inhibiting FASN by FASN inhibitor can partly restore the immunostimulatory activity of TIDCs and extended tumor control by evoking protective anti-tumor immune responses. Therefore, our data provide a mechanism by which ovarian cancer-intrinsic FASN oncogenic pathway induce the impaired anti-tumor immune response through lipid accumulation in TIDCs and subsequently T-cells exclusion and dysfunction. These results could further indicate that targeting the FASN oncogenic pathway concomitantly enhance anti-tumor immunity, thus offering a unique approach to ovarian cancer immunotherapy.

Keywords: FASN; immune response; immunity; ovarian cancer; tumor-infiltrating dendritic cells.

Figure 1

FASN was upregulated in human ovarian cancer and negatively correlated anti-tumor functional T cell infiltration. (A) Heatmap depicted top genes list increased in different types of human ovarian cancer compared to the normal ovarian surface epithelium. 1. Ovarian Surface Epithelium (n = 5); 2. Ovarian Clear Cell Adenocarcinoma (n = 7); 3. Ovarian Endometrioid Adenocarcinoma (n = 9); 4. Ovarian Mucinous Adenocarcinoma (n = 9); 5. Ovarian Serous Adenocarcinoma (n = 20). (B) FASN expression was upregulated in a higher grade of human ovarian cancer. (Grade 1, n = 6; grade 2, n = 27; grade 3, n = 58). (C) The expression of FASN was increased in OvCa patients with metastasis (Primary patients, n = 166, metastatic patients, n = 66). (D) Correlation between the expression intensity of FASN and distribution of CD8⁺ T cells in human ovarian cancer biopsies. Fisher's exact test with n = 50 (FASN low, n = 25; FASN high, n = 25). (E) The correlation of FASN mRNA expression with tumor-infiltrating B cells, CD4⁺ T cells and CD8⁺ T cells in human ovarian cancer based on TIMER analysis from TCGA datasets. The scatterplots showed the purity-corrected partial Spearman's correlation and statistical significance. (F) The correlation of FASN mRNA expression with gene signature of CD8⁺ T cells (CD8A, IFNG, GZMB, GZMA, and PRF1) in human ovarian cancer based on TIMER analysis. The scatterplots showed the purity-corrected partial Spearman's correlation and statistical significance.

Figure 2

Ovarian cancer with elevated FASN expression showed T cell exclusion and defective tumor-infiltrating DCs. (A) Tumor growth in synergetic mice after the inoculation of ID8 cell transfected with vehicle control (shCtrl: FASN^high) or knocking-down FASN (shFASN: FASN^low). (B) Representative example out of five for flow cytometry analyzing tumor-infiltrating CD3⁺ T cells in ID8 tumor-bearing mice with high (FASN^high) or low expression of FASN (FASN^low) on day 15. Statistical bar graph showed CD3⁺ T cells depicted as percentage living CD45⁺ cells and absolute numbers per gram tumor. n = 10. (C) Representative color contour plots examining Eomes⁺T-bet⁻ exhausted CD8⁺ T cells from mice with shCtrl-ID8 or shFASN-ID8 tumor, and quantification of Eomes⁺T-bet⁻ cells in CD8⁺ T cells was shown. (D) Quantification of tumor infiltrated CD44⁺CD62L⁻ cell population in CD8⁺ T cell from tumor-bearing mice transplanted with shFASN-ID8 or shCtrl-ID8 (n = 10). (E) Analysis of cytokine IL-2 and IFN-r in the supernatant of sorted CD3⁺ T cells isolated from FASN^low or FASN^high-ID8 tumors with anti-CD3/CD28 stimulation (n = 10). (F,G) Flow cytometry analysis and quantification of MHC-II⁺CD11c⁺ dendritic cells in the tumor (F) and ascites (G) of mice with the inoculation of shCtrl-ID8 (FASN^low) or shFASN-ID8 (FASN^high) (n = 10). (H) mRNA expression of indicated DCs associated genes in TIDCs sorted from ID-8 transplanted mice with FASN^low or FASN^high. The results were normalized to the level of expression of 18sRNA. (I) Representative flow cytometry analysis and quantification of CD40 and CD86 on MHC-II⁺CD11c⁺ cells from tumor-bearing mice transplanted with FASN^low ID8 or FASN^high ID8 cells. (n = 10). (J) In vitro immunostimulatory activity of tumor-infiltrating CD11c⁺ cells purified from FASN^low or FASN^high ID-8 tumor-bearing mice. Representative histograms of CD8⁺ T cell proliferation at CD8⁺ to CD11c⁺ cell ratio 10:1 (left panel) and quantification of CD8⁺ T cell proliferation using CFSE dilution (right panel) (n = 3). Data presented as mean ± SEM; representative of at least 3 independent experiments; ns, no significance; ^*P < 0.05; ^**P < 0.01; ^***P < 0.001.

Figure 3

Lipid accumulation in tumor-infiltrating DCs derived from the FASN^high ovarian cancer accounted for its defective function. (A) Staining of sorted tumor-infiltrating dendritic cells (TIDCs) from FASN^low or FASN^high ID-8 tumor-bearing mice with BODIPY 493/503. Representative images are shown; scale bar, 20 μm. Green, BODIPY 493/503; blue, DAPI. (B) Immunofluorescence intensity of BODIPY 493/503 in TIDCs was measured by Image J and quantification was shown (n = 5). (C) Flow cytometry analysis of lipid level in TIDC from FASN^low or FASN^high ID8 tumor-bearing mice. (D) Geometric mean fluorescence intensity (MFI) of BODIPY 493/503 in TIDCs was measured. Each group includes five mice. FASN^low and FASN^high ID8 tumor-bearing mice were evaluated in parallel in each experiment. (E) Quantification of lipid level in MHC-II⁺CD11c⁺ DCs from spleen, ascites, lymph node, and tumor in FASN^low or FASN^high ID-8 tumor-bearing mice. (F) Flow cytometry analysis of BODIPY 493/503 staining of TIDCs from an individual with ovarian cancer showed human TIDCs also have high lipid level. (G) Cumulative results of lipid levels in TIDCs from individuals with OvCa. Data = mean ± SEM; representative of at least 3 independent experiments; ^*P < 0.05; ^**P < 0.01.

Figure 4

The enrichment of lipids in the FASN-high ovarian tumor microenvironment. (A–C) The supernatants of malignant ascites were collected from ID-8 tumor-bearing mice after 3 weeks. The normal peritoneal fluid was collected from naive C57BL/6J mice. The amount of fatty acids and triacylglycerols were quantitatively analyzed via LC-MS. Depicted are total unsaturated fatty acids (A), saturated fatty acids (B), and triacylglycerols (C). (D–F) C57BL/6J mice were intraperitoneally injected with 5 × 10⁵ FASN^low or FASN^high ID8 cells, and supernatants of malignant ascites were collected after 3 weeks. Fatty acids and triacylglycerols were extracted and quantitatively analyzed by LC-MS. Depicted are total unsaturated fatty acids (D), saturated fatty acids (E), and triacylglycerols (F). Data = mean ± SEM; representative of 2 independent experiments; ^*P < 0.05; ^****P < 0.0001.

Figure 5

Educated DCs by ascites exhibited defective antigen presentation and impaired priming of T cell activation. (A) Flow cytometric staining with Bodipy 493/503 in BMDCs cultured in the presence of 20% ascites harvested from FASN^low or FASN^high ID-8 tumor-bearing mice for 24 h. MFI average is depicted. Means ± s.d. are shown. (B) mRNA expression of indicated DCs associated genes in BMDCs cultured in the presence of 20% ascites harvested from FASN^low or FASN^high ID-8 tumor-bearing mice. The results were normalized to the level of expression of 18sRNA. (C,D) Uptake of DQ-OVA by BMDCs in the presence of 20% ascites harvested from FASN^low or FASN^high ID-8 tumor-bearing mice. Quantification of uptaking DQ-OVA was shown. Means ± s.d. of three mice per group are shown. (E, F) In vitro immunostimulatory activity of BMDCs treated with ascites (E) or TCM (F) harvested from FASN^low or FASN^high ID8 tumor-bearing mice. Representative histograms of CD8⁺ T cell proliferation at CD8⁺ to BMDCs ratio 10:1 (left panel) and quantification of CD8⁺ T cell proliferation using CFSE dilution (right panel) (n = 3). Data = mean ± SEM; representative of at least 3 independent experiments; ns, no significance; ^*P < 0.05; ^**P < 0.01; ^***P < 0.001.

Figure 6

Inhibition of FASN suppress OvCa progression by inducing an anti-tumor immune response. (A) Mean tumor volume of the subcutaneous ID8 tumor with the treatment of vehicle or cerulenin (n = 10). The red arrow indicated the start of cerulenin treatment. (B) Representative flow cytometric analysis of CD8⁺ and CD4⁺ T cells populations in ID8 tumors with or without the treatment of cerulenin at 15 days post-implantation. (C) Quantification of CD8⁺ T cells population in ID8 tumors with or without the treatment of cerulenin at 15 days post-implantation (n = 5). (D) Representative flow cytometric analysis and quantification of Granzyme B expression on CD8⁺ T cells population in ID8 tumor with or without cerulenin treatment at 15 days post-implantation. (n = 10). (E) Representative flow cytometric analysis and quantification of lipid level in TIDCs sorted from the ID8 tumor with or without cerulenin treatment (n = 5). (F) In vitro immunostimulatory activity of TIDCs sorted from the ID8 tumor with or without the treatment of cerulenin (n = 5). Representative histograms and quantification of CD8⁺ T cell proliferation using CFSE dilution (right panel) were shown (n = 3). Data presented as mean ± SEM; ^**P < 0.01. (G) Summary of crosstalk between FASN^low or FASN^high tumor cells and DCs-mediated adaptive immune response as described in the text.