Lipid droplet-dependent fatty acid metabolism controls the immune suppressive phenotype of tumor-associated macrophages

Abstract

Tumor-associated macrophages (TAMs) promote tumor growth and metastasis by suppressing tumor immune surveillance. Herein, we provide evidence that the immunosuppressive phenotype of TAMs is controlled by long-chain fatty acid metabolism, specifically unsaturated fatty acids, here exemplified by oleate. Consequently, en-route enriched lipid droplets were identified as essential organelles, which represent effective targets for chemical inhibitors to block in vitro polarization of TAMs and tumor growth in vivo. In line, analysis of human tumors revealed that myeloid cells infiltrating colon cancer but not gastric cancer tissue indeed accumulate lipid droplets. Mechanistically, our data indicate that oleate-induced polarization of myeloid cells depends on the mammalian target of the rapamycin pathway. Thus, our findings reveal an alternative therapeutic strategy by targeting the pro-tumoral myeloid cells on a metabolic level.

Keywords: cancer immunotherapy; lipid droplets; lipid metabolism; tumor microenvironment; tumor-associated macrophage.

Figure 1. Oleate polarizes bone marrow‐derived myeloid cells into immune suppressive tumor‐associated macrophages

1. A

   Bone marrow cells were polarized in the presence of 40 ng/ml GM‐CSF and treated with 0.2 mM of the indicated compounds for 7 days. Gr1⁻CD11b⁺ cells were sorted and lysed for microarray. The hierarchical clustering was based on the different expression genes between BSA (Con) and oleate group.

2. B, C

   Signature genes involved in lipid metabolism, dendritic cell maturation, macrophage maturation, tumor‐associated macrophages (TAMs) phenotype, MHCII complex, and innate immune response are listed (B) and validated (C) via flow cytometry or catalytic activity assay. Data are expressed as mean ± SD from two to four independent experiments. Unpaired Student's two‐tailed t‐tests were performed to compare the expression level of indicated proteins in control and oleate groups. *P < 0.05; **P ≤ 0.01.

Figure 2. Oleate‐induced mitochondrial respiration regulates the suppressive phenotype of myeloid cells

1. A, B

   Gr1⁻CD11b⁺ cells, polarized by the indicated treatment, were purified for mitochondrial respiration detection. The oxygen consumption rate (OCR) of these cells was monitored after the addition of oligomycin (OA; 1 μM), carbonyl cyanide‐4‐(trifluoromethoxy) phenylhydrazone (FCCP, 1 μM), and the electron transport inhibitor rotenone and antimycin A (R/AA; 0.5 μM) at indicated time points. The basal OCR, basal extracellular acidification rate (ECAR), spare respiratory capacity, proton leak, ATP production, and maximal respiration based on the OCR value were quantified.

2. C, D

   Forty micromolar etomoxir was added starting from day 0 of bone marrow polarization, followed by mitochondrial respiration assay by using the same amount of polarized Gr1⁻CD11b⁺ myeloid cells on day 7. The mitochondrial respiration was monitored and analyzed via XFe96 Analyzer. The basal OCR, basal extracellular acidification rate (ECAR), spare respiratory capacity, proton leak, ATP production and maximal respiration based on the OCR value were quantified.

3. E, F

   Exemplary plots of the proportion of CD206⁺ cells, the mean fluorescence intensity (MFI) of CD38 and CD73 on polarized myeloid cells was determined via flow cytometry.

4. G–I

   T‐cell proliferation assay was performed via co‐culture with purified CD4⁺ T cells in the ratios of M (myeloid cells): T (T cells) = 1:30 (G, H). (I) Nitric oxide (NO) production from the co‐culture supernatant was quantified by Griess reaction.

Data information: Data are expressed as mean ± SD from two to four independent experiments and analyzed by either one‐way analysis of variance (ANOVA) (B) or two‐way ANOVA with Tukey's post hoc test (D, F, H, I). *P < 0.05; **P ≤ 0.01; ***P ≤ 0.001.Source data are available online for this figure.

Figure 3. Lipid droplet‐derived fatty acids facilitate mitochondrial respiration in myeloid cells

1. A

   The formation and utilization of lipid droplets in eukaryotes. Five micromolar combination of DGAT inhibitors (DGAT1 inhibitor A922500 and DGAT2 inhibitor PF‐06424439), 40 μM ATGL inhibitor atglistatin (Atg), or 5 μM MAGL inhibitor MJN110 was added to the bone marrow polarization system in the presence of 40 ng/ml GM‐CSF and indicated compounds for 7 days.

2. B

   The oxygen consumption rate (OCR) of 1 × 10⁵ purified Gr1⁻CD11b⁺ cells was monitored as described in Fig 2. The basal OCR, basal ECAR (extracellular acidification rate), spare respiratory capacity, proton leak, ATP production, and maximal respiration based on the OCR value were quantified.

3. C

   The polarization state was evaluated via the expression of CD206.

4. D

   T‐cell proliferation assay was performed employing co‐culture with purified CD4⁺ T cell in the variant ratios.

5. E

   The lipid droplet accumulation was determined by BODIPY staining. The percentage of divided cells and proliferation index was calculated.

6. F

   Nitric oxide (NO) concentration in the co‐culture supernatant was quantified by Griess reaction.

Data information: Shown is the mean ± SD from two to four independent experiments and analyzed by two‐way ANOVA with Tukey's post hoc test. *P < 0.05; **P ≤ 0.01; ***P ≤ 0.001.

Figure 4. Polarization of suppressive myeloid cells depends on mTOR signaling pathway

1. A

   Bone marrow‐derived myeloid cells were treated with BSA control (Con) or oleate in the presence or absence of 10 nM rapamycin (Rapa). The polarization of myeloid cells was detected via flow cytometry as indicated by CD11b and CD206 expression.

2. B

   Purified Gr1⁻CD11b⁺ myeloid cells were co‐cultured with CD4⁺ T cells for proliferation assay. The supernatant was collected for nitric oxide (NO) detection.

3. C

   The oxygen consumption rate (OCR) of 1 × 10⁵ differentiated myeloid cells were monitored after the addition of oligomycin (OA; 1 μM), the uncoupler carbonyl cyanide‐4‐(trifluoromethoxy) phenylhydrazone (FCCP, 1 μM), and the electron transport inhibitor rotenone and antimycin A (R/AA, 5 μM) at indicated time points. The basal OCR, basal extracellular acidification rate (ECAR), spare respiratory capacity, ATP production, and maximal respiration based on the OCR value were quantified.

4. D, E

   Mean fluorescence intensity (MFI) of intracellular phosphor‐mTOR (pS2448) in polarized myeloid cells as determined by (D) flow cytometry and (E) lysed for Western blot.

Data information: Shown is the mean ± SD from two to four independent experiments. Two‐way ANOVA with Tukey's post hoc test was performed to compare the effect of rapamycin in different groups. *P < 0.05; **P ≤ 0.01; ***P ≤ 0.001.

Figure 5. Disrupting lipid droplet‐derived fatty acids in myeloid cells impairs tumor growth

1. A

   Tumor‐infiltrating immune cells were isolated from either CT26 or MCA205 tumor‐bearing model and analyzed for the expression of BODIPY in CD206⁺ (black) and CD206⁻ (gray) cells.

2. B

   Balb/c mice were inoculated with 5 × 10⁵ CT26 tumor cells subcutaneously. Since day 6, vehicle or DGAT inhibitors (iDGAT) were administrated via intra‐tumoral injection once daily.

3. C

   At day 13, tumor‐infiltrated immune cells were isolated and stained for CD206 and CD11b.

4. D

   5 × 10⁵ MCA205 were injected subcutaneously into C57BL/6 mice. PBS, liposome control or iDGAT‐liposome was injected into the peritoneum every other day starting at day 7. Tumor size and body weight were measured every 7 days and the mice were sacrificed on day 31 for analysis.

5. E

   Immune cells from spleen, tumor‐draining lymph nodes (TDLN), or tumor were isolated and stained for CD11b and with BODIPY for lipid droplet quantification in myeloid cells.

6. F

   Tumor‐infiltrating CD8⁺ T cells (red) were stained together with nuclei (DAPI/blue) after sacrificing the tumor‐bearing mice. White bars represent 200 μm.

Data information: Shown is the mean ± SEM from two to four independent experiments. Two‐way ANOVA with Tukey's post hoc test was performed to compare the effect of iDGAT‐liposome in tumor model. One‐way ANOVA was performed to compare the effect of iDGAT‐encapsulated liposome in human CD14⁺ cells. *P < 0.05; **P ≤ 0.01; ***P ≤ 0.001.

Figure 6. Oleate induces polarization of human monocytes into CD206⁺ suppressive cells

1. A

   1 × 10⁶ CD14⁺ myeloid cells from healthy donors were polarized by 1 ng/ml GM‐CSF in the presence of BSA control (C), 0.2 mM oleate (O), or 0.2 mM stearate (S) over a 6‐day period. The expression of CD206, CD204, and CD38 was analyzed via flow cytometry.

2. B

   The level of lipid droplets was determined via BODIPY staining.

3. C

   The oxygen consumption rate (OCR) of differentiated myeloid cells was monitored as described in Materials and Methods.

4. D

   The purified CFSE‐labeled autologous CD4⁺ T cells were co‐cultured with polarized myeloid cells with indicated treatment for T‐cell proliferation assay.

5. E

   Tumor tissue (tumor) as well as corresponding non‐tumorous adjacent tissue (control) was collected from colorectal cancer patients and prepared for histopathology. The expression of CD68 (red), CD206 (green), and ADRP (blue) indicates the lipid droplets in tumor‐infiltrating myeloid cells. Nuclei (white) were stained using DAPI. The absolute number of positive cells was quantified in 5 high‐power fields (hpf; scale bar = 20 μm).

Data information: Shown is the mean ± SD from two to four independent experiments and analyzed by either one‐way ANOVA (A), Kruskal–Wallis test with Dunn's multiple comparisons test (D) or unpaired Student's two‐tailed t‐tests (E). *P < 0.05; **P ≤ 0.01; ***P ≤ 0.001.