Metformin exerts antitumor activity via induction of multiple death pathways in tumor cells and activation of a protective immune response

Abstract

The antitumor effect of metformin has been demonstrated in several types of cancer; however, the mechanisms involved are incompletely understood. In this study, we showed that metformin acts directly on melanoma cells as well as on the tumor microenvironment, particularly in the context of the immune response. In vitro, metformin induces a complex interplay between apoptosis and autophagy in melanoma cells. The anti-metastatic activity of metformin in vivo was assessed in several mouse models challenged with B16F10 cells. Metformin's activity was, in part, immune system-dependent, whereas its antitumor properties were abrogated in immunodeficient (NSG) mice. Metformin treatment increased the number of lung CD8-effector-memory T and CD4⁺Foxp3⁺IL-10⁺ T cells in B16F10-transplanted mice. It also decreased the levels of Gr-1⁺CD11b⁺ and RORγ⁺ IL17⁺CD4⁺ cells in B16F10-injected mice and the anti-metastatic effect was impaired in RAG-1^-/- mice challenged with B16F10 cells, suggesting an important role for T cells in the protection induced by metformin. Finally, metformin in combination with the clinical metabolic agents rapamycin and sitagliptin showed a higher antitumor effect. The metformin/sitagliptin combination was effective in a BRAFV600E/PTEN tamoxifen-inducible murine melanoma model. Taken together, these results suggest that metformin has a pronounced effect on melanoma cells, including the induction of a strong protective immune response in the tumor microenvironment, leading to tumor growth control, and the combination with other metabolic agents may increase this effect.

Keywords: T cells; cell death; metformin; sitagliptin; tumor immunity.

Figure 1. Metformin effects in melanoma cells in vitro

(A) B16F10 cells (5×10³) were incubated for 24, 48, or 72 h with metformin (0–40 mM), and cell viability was measured using the MTT assay (^**p<0.002, ^****p<0.0001). (B) Different patient melanoma cells (Mel25, Mel28, and Mel11) were incubated with metformin (0–40 mM) for 72 h, and cell viability was measured using the MTT assay (^***p<0.003). (C) Patient Mel11 cell migration was determined by measuring wound width at 0, 3, and 7 h, and 24 h after incubation with 5 mM of metformin (^***p<0.003). (D) Oxygen consumption rate of B16F10 cells treated with metformin (10 mM) for 18 h (^*p<0.05). (E) B16F10 cells were incubated for 24 h with metformin (0–40 mM), and lactate dehydrogenase (LDH) levels were measured in the supernatant after a 24-h incubation (^****p<0.0001).

Figure 2. Metformin induces different cell death mechanisms in B16F10 melanoma cells

(A) Fold increase (>2) or decrease (<2) in gene expression after B16F10 cells were treated for 24 h with metformin (10 mM), measured using quantitative RT-PCR. Pie chart shows the distribution of genes related to a specific cell death process. (B) B16F10 cells treated for 72 h with metformin (10 Mm) were stained with annexin V and 7-AAD. (C) mRNA transcript quantification of HIF-1α, Socs-3, and Beclin-1 in B16F10 cells treated for 24 h with metformin (10 mM) (^**p<0.001). (D) B16F10 cells treated for 24 h with metformin (20 mM) in the presence of Z-VAD-FMK (20 μM), (E) Z-YVAD-FMK (20 μM), (F) necrostatin-1 (Nec-1, 100 μM). Cell viability was measured using the MTT assay. (G) B16F10 cells were incubated for 24 h with metformin (0–40 mM). Representative histograms and graphs of flow cytometry analyses showing mean fluorescent intensity of DCFH (left) or oxidized MitoSOX (right) following treatment.

Figure 3. Anti-metastatic activity of metformin requires an intact immune system

(A) C57BL/6 mice were intravenously challenged with B16F10 cells, and treated by gavage with 500 mg/kg metformin or phosphate-buffered saline (PBS) daily from day 3 after challenge. Quantification of the number of lung nodules in each group, n=12, ^***p<0.0002, mean ± standard deviation. (B) B16F10 cells (5×10⁵) were injected intravenously into C57BL/6 mice and NOD-SCID gamma null (NSG) mice. Three days after the challenge, the animals were treated with metformin (500 mg/kg) or PBS daily by gavage. Pulmonary nodules were counted on day 15. n=5 animals per group (^*p<0.05, ^**p<0.001, ^***p<0.0001, ^****p<0.001).

Figure 4. Tumor microenvironment and systemic profile of cytokines and hormones in mice treated with metformin

(A) Cytokine (IFN-γ, TNF-α, IL-10, IL-6, and IL-17) mRNA expression was determined in the lungs of C57BL/6 mice challenged intravenously with B16F10 cells and treated with metformin. Values are expressed relative to a control group. The cycle thresholds (Ct) for the gene and the internal control were determined for each sample. The relative mRNA expression was calculated as 2-ΔΔCT relative to HPRT. Levels of the same cytokines were determined in the lungs (B) and spleen (C) of these animals by ELISA. (D) TNF-α, IL-6, leptin, adiponectin, and GIP were measured in the sera of these mice by ELISA. n=5; values are expressed as mean ± standard error. ^*p<0.05. Images are representative of at least two independent assays.

Figure 5. Effect of metformin on myeloid and T cells

(A) C57BL/6 mice intravenously challenged with B16F10 cells were treated with metformin or phosphate-buffered saline (PBS) and numbers of CD11b⁺F4/80⁺, CD11c⁺ or Gr-1⁺CD11b⁺ cells were assessed in spleens by fluorescence-activated cell sorting. Percentage of T cells in the lung, absolute count of T CD4⁺ and T CD8⁺ lung cells (B), percentage of CD8⁺CD69⁺, CD8⁺IFN-γ⁺, and CD8⁺granzyme B⁺ cells (C). Number of naïve T cells, CD4⁺CD44^hiCD62L^low, and CD8⁺CD44^hiCD62L^low cells (D). All quantities in absolute cell number or percentage ± standard deviation. ^*p<0.05.

Figure 6

(A) Anti-metastatic activity of metformin with T cell participation. B16F10 cells (5×10⁵) were injected intravenously into C57BL/6, RAG^−/−, CD4^−/−, or CD8^−/− mice, and the animals were treated with metformin or phosphate-buffered saline (PBS) (n=5–11, pooling at least two experiments). ^*p<0.05, ^**p<0.01. (B) B16F10 cells (5×10⁵) were injected intravenously into C57BL/6 mice treated with metformin or PBS; the animals were administered 200 μg of an anti-CD8 antibody intraperitoneally on days 3, 7, and 11. n=4–5 mice per group (^*p<0.05). (C) B16F10 cells (5×10⁵) were injected intravenously into C57BL/6 mice treated with metformin or PBS; the animals were administered 500 μg of an anti-CD4 antibody intraperitoneally on days 1 and 5. n=4–5 mice per group (^*p<0.05). Values are expressed as mean ± standard deviation.

Figure 7. Metformin treatment induces an increase in CD4+Foxp3+IL-10+ cells in vivo

(A) C57BL/6 mice were intravenously challenged with B16F10 cells, and treated by gavage with 500 mg/kg metformin or phosphate-buffered saline (PBS) daily from day 3 after challenge. Foxp3 mRNA expression in the lung tumor burden was measured by real-time PCR on day 15 after tumor challenge. Number of total Foxp3⁺ GFP⁺ cells from TDLN (B) or spleen (C) of Foxp3-gfp mice intravenously challenged with B16F10 cells and treated by gavage with metformin or phosphate-buffered saline (PBS). (D) Sorted Foxp3⁺ or Foxp3⁻ cells were stimulated with phorbol myristate acetate (PMA) and ionomycin (IOM) and then stained for intracellular IL-10 from the spleens of Foxp3-gfp knock-in mice challenged with B16F10 cells and treated by gavage with 500 mg/kg metformin or PBS for 15 days. (E) Mice were treated with intraperitoneal injection on days 1, 4, and 7 of anti-CD25 (500 μg/dose), followed by treatment with metformin (500 μg/mouse). (F) Number of CD4⁺RORγ⁺ T cells in the lungs of mice intravenously challenged with B16F10 cells and treated with metformin. (G) Naïve CD4⁺ T cells were sorted from C57BL/6J mice and cultivated in the presence of TGF-β (1 ng/mL), IL-6 (50 ng/mL), and anti-CD3/antiCD28 with or without 10 mM of metformin. Relative mRNA expression was calculated as 2-ΔΔCT relative to HPRT. n=3; values are expressed as mean ± standard error. ^*p<0.05, ^**p<0.01. Images are representative of at least two independent assays.

Figure 8. Anti-metastatic effect of metformin in combination with rapamycin or sitagliptin in a BRAF V600 murine model

(A) C57BL/6 mice intravenously challenged with B16F10 cells (5×10⁵) were treated with phosphate-buffered saline (PBS), metformin (500 mg/kg), rapamycin (rapa, 1.5 mg/kg), or metformin + rapa, starting 3 days after challenge (n=3–5). The number of lung nodules was evaluated 15 days after the challenge. (B) C57BL/6 mice intravenously challenged with B16F10 cells (5×10⁵) were treated with PBS, metformin, sitagliptin (50 mg/kg), or sitagliptin/metformin starting 3 days after the challenge (n=3–4). (C) Kaplan-Meier survival curves comparing Tyr:CreER; BrafCA/+ and Ptenlox/lox mice treated with sitagliptin/metformin 45 days after topical administration of 4-hydroxytamoxifen (4-HT); pool of two independent assays evaluated by the log-rank test (n=7–8). Animals are represented individually; mean and standard deviation are shown by horizontal lines. ^*p<0.05; ^**p<0.001; ^***p<0.0001; ns, not significant.