Dietary fats suppress the peritoneal seeding of colorectal cancer cells through the TLR4/Cxcl10 axis in adipose tissue macrophages

Abstract

Peritoneal carcinomatosis (PC) of colorectal cancer (CRC) is a terminal phase of malignancy with no effective strategies for the prevention of this condition. Here we established PC models in mice by intraperitoneal engraftment of CRC cells and revealed an unexpected role for a high-fat diet (HFD) in preventing metastatic seeding in the visceral fat. Mechanistically, the HFD stimulated the activation of adipose tissue macrophages (ATMs) toward an M1-like phenotype and enhanced ATM tumor phagocytosis in a TLR4-dependent manner. Furthermore, the TLR4-Cxcl10 axis in ATMs promoted T cell recruitment, and M1-like macrophages stimulated T cell activation in tumor-seeded fats. The inhibitory effect of the HFD on tumor seeding was abolished with the ablation of macrophages, inactivation of T cells, or blockade of the TLR4-Cxcl10 axis in macrophages. Finally, we showed that a HFD and conventional chemotherapeutic agents (oxaliplatin or 5-fluorouracil) synergistically improved the survival of tumor-seeded mice. Collectively, our findings demonstrate that peritoneal seeding of CRC can be suppressed by short-term treatment with a HFD in the early phase, providing a novel concept for the management of these patients in the clinic.

Fig. 1

High-fat diet suppresses the metastatic seeding of CRC cells. a Metastatic seeding of CRC cells in the epididymal fat (eFat). Six-week-old male mice were intraperitoneally injected with GFP-tagged MC-38 cells (1.0 × 10⁶ cells in 100 µl PBS per mouse) or PBS, which was used as a control. The eFats were collected dynamically and stained with anti-GFP (green) or anti-perilipin (red) antibodies for immunofluorescence analysis. Nuclei were visualized by Hoechst staining (blue). Representative images are shown. Arrows indicate the cancer cells. Scale bars, 100 μm. b High-fat diet (HFD) feeding inhibits tumor seeding in the eFats. Six-week-old male mice were intraperitoneally inoculated with MC-38 cells (1.0 × 10⁶ cells in 100 µl PBS per mouse) and immediately fed a chow diet (CD) or HFD for 14 or 21 days. Then the tumor nodes in the eFats were evaluated. The blue and red arrows indicate the epididymal fat pads and tumor nodes, respectively. c Tumor nodes from the mice described above in b. Representative results for one mouse are shown. d Weights of the tumor nodes described in c. Data are shown as the mean ± s.e.m. (n = 5, ***P < 0.005; Student’s t test). e Dynamic measurement of the weights of the tumor-bearing mice described in b. Data are shown as the mean ± s.e.m. (n = 10, ***P < 0.005; two-way ANOVA). f Dynamic measurement of the food intake of the tumor-bearing mice described in b. Each plot represents the average food intake by one mouse in 1 day (n = 10; two-way ANOVA). g HFD consumption improves the survival of tumor-seeded mice. Tumor-seeded models were established as described in b. This experiment was repeated three times (n = 15, *P < 0.05; Gehan–Breslow–Wilcoxon test). h HFD inhibits the peritoneal seeding of CT-26 cells. Six-week-old male mice were intraperitoneally implanted with CT-26 cells (1.0 × 10⁶ cells in 100 µl PBS) and immediately fed a CD or HFD for 14 days. Then peritoneal tumor nodes were identified. i HFD consumption improves the survival of the tumor-seeded mice described in h. (n = 15, ***P < 0.005; Gehan–Breslow–Wilcoxon test)

Fig. 2

HFD potentiates T cell recruitment and activation in tumor-seeded eFats. a HFD consumption promotes the peritoneal metastasis of MC-38 cells in nude mice. Six-week-old male nude mice were intraperitoneally injected with MC-38 cells (1.0 × 10⁶ cells in 100 µl PBS) and immediately fed a CD or HFD for 14 days. Peritoneal tumor nodes were collected and weighed (n = 5). b–d HFD consumption enhances T cell infiltration into the eFats. Six-week-old male mice were intraperitoneally injected with MC-38 cells (1.0 × 10⁶ cells in 100 µl PBS) and immediately fed a CD or HFD. Then the stromal cells of the eFats were isolated dynamically. Total T cells (CD45⁺CD3⁺CD11b⁻CD45R⁻) (b), CD4⁺ T cells (CD45⁺CD3⁺CD11b⁻CD45R^-CD4⁺CD8⁻) (c), and CD8⁺ T cells (CD45⁺CD3⁺CD11b^-CD45R^-CD4^-CD8⁺) (d) were assessed by flow cytometry. Each tested sample was pooled from 4 individual samples (n = 3). e HFD consumption increases the frequencies of adipose tissue T cells in the s.c., in situ and Apc^min+/− tumor-seeded models described in “Methods.” Each tested sample was pooled from 2 individual samples (n = 3). f Frequencies of IFNγ⁺ cells among CD8⁺ T cells in the eFats of tumor-seeded mice treated with a CD or HFD for 5 days (n = 3). g–i Frequencies of IFNγ⁺ (g), IL17⁺ (h), and CD25⁺ (i) cells among CD4⁺ T cells in the eFats of tumor-seeded mice treated with a CD or HFD for 5 days (n = 3). j Six-week-old male C57BL/6 mice were intraperitoneally injected with MC-38 cells (1.0 × 10⁶ cells in 100 µl PBS) and immediately fed a CD or HFD for 14 days. Each mouse was intraperitoneally treated with antibodies (IgG, anti-CD4, or anti-CD8) 3 times (days 2, 4, and 6) at a dosage of 100 µg day⁻¹. Peritoneal tumor nodes were collected and weighed. (n = 3). All the data are shown as the mean ± s.e.m. The data in b–d were analyzed with two-way ANOVA. The data in other histograms were analyzed with Student’s t test (*P < 0.05, **P < 0.01, and ***P < 0.005)

Fig. 3

HFD stimulates Cxcl10-dependent T cell recruitment in tumor-seeded eFats. a The eFats of HFD-treated mice attracted T cells. The eFats were isolated from tumor-seeded mice fed a CD or HFD for 5 days. Mouse splenocytes were cultured in the upper chamber, and the eFats were cultured in the lower chamber of a Transwell system. Twenty-four hours later, the migrated T cells, CD4⁺ T cells, and CD8⁺ T cells were analyzed by flow cytometry (n = 3). b Proliferation rate of T cells cocultured with the eFats from tumor-seeded mice treated with a CD or HFD for 5 days (n = 5). c Heat map showing differentially expressed chemokines in the eFats. mRNA sequencing assays were performed on the eFats isolated from tumor-seeded mice fed a CD or HFD for 5 days. Each sample was pooled from three individual samples. d CXCL10 is required for HFD-induced T cell migration ex vivo. The eFats were isolated from tumor-seeded mice fed a CD or HFD for 5 days. Mouse splenocytes were cultured in the upper chamber, and the eFats were cultured in the lower chamber of a Transwell system. A neutralizing anti-CXCL10 antibody (100 ng ml⁻¹) was added into the lower chamber to block CXCL10 activity. Twenty-four hours later, the migrated T cells in the lower chamber were analyzed by flow cytometry. T cells were defined as CD45⁺CD11b⁻CD3⁺ (n = 5). e Relative mRNA levels of Cxcl10 in the eFats from tumor-seeded WT or Cxcl10 knockout (Cxcl10^−/−) mice fed a CD or HFD for 5 days. nd not detected (n = 4). f, g Infiltration of total T cells, CD4⁺ T cells, CD8⁺ T cells (f), and IFNγ⁺ T cells (g) into the eFats of WT or Cxcl10^−/− mice fed a CD or HFD for 5 days. Each sample was pooled from 3 individual samples (n = 3). h An MC-38-based tumor-seeded model was constructed in WT and Cxcl10^−/− mice. Peritoneal tumor nodes were collected on day 14 and weighed (n = 5). All data are shown as the mean ± s.e.m. (*P < 0.05, **P < 0.01 and ***P < 0.005; ns not significant; Student’s t test)

Fig. 4

HFD potentiates macrophage-dependent T cell recruitment and activation in tumor-seeded eFats. a Relative Cxcl10 mRNA levels in T cells, B cells, macrophages, neutrophils, and adipocytes isolated from the eFats of tumor-seeded mice treated with a CD or HFD for 5 days. Each sample was pooled from 10 individuals (n = 3). b Adipose tissue macrophages (ATMs) in the eFats were depleted in Mac^−/− mice. Details are described in “Methods” (n = 3). c Relative mRNA levels of Cxcl10 in the eFats of tumor-seeded Mac^+/+ or Mac^−/− mice fed a CD or HFD for 5 days (n = 3). d HFD consumption stimulates macrophage-dependent T cell recruitment. Six-week-old male Mac^+/+ and Mac^−/− mice were intraperitoneally injected with MC-38 cells (1.0 × 10⁶ cells in 100 µl PBS) and immediately fed a CD or HFD for 5 days. Then the frequencies of total T cells, CD4⁺ T cells, and CD8⁺ T cells were calculated by flow cytometry. Each tested sample was pooled from 3 individual samples (n = 3). e Macrophage deletion prevents HFD-induced T cell activation. IFNγ⁺ T cells were counted in the eFats of tumor-seeded mice treated with a CD or HFD for 5 days. Each tested sample was pooled from 3 individual samples (n = 3). f Macrophage deletion prevents the HFD-mediated suppression of tumor seeding. Mac^+/+ and Mac^−/− mice were intraperitoneally injected with MC-38 cells (1.0 × 10⁶ cells in 100 µl PBS) and immediately fed a CD or HFD for 14 days. Peritoneal tumor nodes were collected and weighed (n = 5). All the data are shown as the mean ± s.e.m. (*P < 0.05 and ***P < 0.005; ns not significant; Student’s t test)

Fig. 5

Stearic acid stimulates macrophage activation in vitro via TLR4. a The Toll-like receptor signaling pathway was highly enriched in the eFats of tumor-seeded mice treated with a CD or HFD for 5 days. Six-week-old male mice were intraperitoneally injected with MC-38 cells (1.0 × 10⁶ cells in 100 µl PBS) and immediately fed a CD or HFD for 5 days. Then the eFats were collected for RNA sequencing assays, and altered gene expression was subjected to KEGG pathway enrichment analysis. b Fold changes in gene expression determined by comparing RNA sequencing results between the CD group and the HFD group. Red points indicate upregulated genes; blue points indicate downregulated genes; gray points show unchanged genes. Toll-like receptor 4 (TLR4) expression was obviously upregulated by HFD consumption. c The level of free fatty acid (FFA) in eFats of the mice treated with CD or HFD for 5 days. d Stearic acid and LPS stimulate TLR4-dependent phagocytosis in macrophages. Peritoneal macrophages from wild-type (WT) or TLR4 knockout (TLR4-ko) mice were pretreated with stearic acid (SA; 200 µM) and/or LPS (100 ng ml⁻¹) for 12 h and cocultured with GFP-tagged MC-38 cells for phagocytosis assays. This experiment was repeated twice. Representative results are displayed. e Statistical analysis of the results in d. Data are presented as the mean ± s.e.m. (n = 3; *P < 0.05 and **P < 0.01; Student’s t test). f–h A fatty acid stimulates TLR4-dependent proinflammatory cytokine expression. Peritoneal macrophages from WT or TLR4-ko mice were treated with SA (200 µM) or LPS (100 ng ml⁻¹) for 6 h, and then RNA was extracted for real-time PCR analysis of TNFα, IL-1β, and Cxcl10. Data are presented as the mean ± s.e.m. (n = 3; ***P < 0.005; Student’s t test)

Fig. 6

TLR4 is required for HFD-regulated macrophage activation and T cell recruitment and activation as well as peritoneal metastasis. a HFD consumption activates M1-like macrophages in the visceral fat via TLR4. Six-week-old male fl/fl and macrophage-specific TLR4 knockout (TLR4-cko) mice were intraperitoneally injected with MC-38 cells (1.0 × 10⁶ cells in 100 µl PBS) and immediately fed a CD or HFD for 5 days. Then total macrophages, M1 macrophages and M2 macrophages in the eFats were analyzed. Each tested sample was pooled from 3 individual samples (n = 3). b HFD consumption potentiates T cell recruitment via macrophage TLR4. Frequencies of total T cells, CD4⁺ T cells, and CD8⁺ T cells in the eFats of the mice described in a (n = 3). c HFD consumption stimulates tumor phagocytosis by macrophages via TLR4. fl/fl and TLR4-cko mice were injected with GFP-tagged MC-38 cells (1.0 × 10⁶ cells in 100 µl PBS). Fourteen days later, the in vivo phagocytosis rates of tumor-associated macrophages were measured (n = 3). d IFNγ⁺ T cells in the eFats of fl/fl or TLR4-cko mice fed a CD or HFD for 5 days. Each sample was pooled from 3 individual samples (n = 3). e HFD consumption suppresses tumor seeding via TLR4. MC-38-based tumor-seeded models were constructed in fl/fl and TLR4-cko mice. Fourteen days later, peritoneal tumor nodes were collected and weighed (n = 5). The data are shown as the mean ± s.e.m. (*P < 0.05, **P < 0.01, and ***P < 0.005; ns not significant; Student’s t test)

Fig. 7

HFD synergistically potentiates the preventive effect of chemotherapeutic drugs on peritoneal metastasis. a Schematic diagram for the time points of diet and drug treatments. After tumor inoculation, mice were organized into six groups with different diets and oxaliplatin treatments. CD group: mice were treated with a CD the entire time; HFD group: mice were treated with a HFD the entire time; HFD(7) group: mice were treated with the HFD from day 1 to day 7; CD + OXP group: mice were treated with the CD the entire time and oxaliplatin (5.0 mg kg⁻¹) on days 1, 5, and 9 after tumor inoculation; HFD + OXP group: mice were treated with the HFD the entire time and oxaliplatin (5.0 mg kg⁻¹) on days 1, 5, and 9 after tumor injection; and HFD(7)+OXP group: mice were treated with the HFD from day 1 to day 7 and oxaliplatin (5.0 mg kg⁻¹) on days 1, 5, and 9 after tumor injection. b Long-term HFD consumption synergistically potentiates the survival of oxaliplatin-treated tumor-seeded mice. Mice were grouped and treated as described in a. The survival times of these tumor-bearing mice were recorded. c Seven days of HFD feeding synergistically potentiated the survival of oxaliplatin-treated tumor-seeded mice. Mice were treated as described in a. d Schematic diagram for the time points of diet and drug treatments. After tumor inoculation on day 0, mice were organized into 6 groups with different diets and drug treatments. CD group: mice were treated with a CD the entire time; HFD(7) group: mice were treated with a HFD from day 1 to day 7; CD + OXP group: mice were treated with the CD the entire time and oxaliplatin (2.5 mg kg⁻¹) on days 3, 7, and 11; CD + 5FU group: mice were treated with the CD the entire time and 5-fluorouracil (25 mg kg⁻¹) on days 3, 7, and 11; HFD(7) + OXP group: mice were treated with the HFD from day 1 to day 7 and oxaliplatin (2.5 mg kg⁻¹) on days 3, 7, and 11; and HFD(7)+5FU group: mice were treated with the HFD from day 1 to day 7 and 5-fluorouracil (25 mg kg⁻¹) on days 3, 7, and 11. e HFD feeding and oxaliplatin synergistically potentiate the survival of the tumor-seeded mice described in d. f HFD and 5-fluorouracil synergistically potentiate the survival of the tumor-seeded mice described in d. All the data are shown as the mean ± s.e.m. (n = 15; *P < 0.05, **P < 0.01, and ***P < 0.005; Gehan–Breslow–Wilcoxon test)

Fig. 8

Proposed hypothesis for the prevention of colorectal peritoneal metastasis by a HFD. Disseminated colorectal cancer cells preferentially target the visceral fat in early metastasis. HFD consumption stimulates TLR4-dependent M1 macrophage activation and phagocytosis in adipose tissue macrophages (ATMs), which further enhance the recruitment (by Cxcl10) and activation (by M1 cytokines) of CD4+ and CD8+ T cells in the visceral fat and ultimately prevents the metastatic seeding of colorectal cancer