Phenformin as prophylaxis and therapy in breast cancer xenografts

Abstract

Background: Observations that diabetics treated with biguanide drugs have a reduced risk of developing cancer have prompted an enthusiasm for these agents as anti-cancer therapies. We sought to determine the efficacy of the biguanide phenformin in the chemoprophylaxis and in the treatment of oestrogen receptor (ER)-positive MCF7 and receptor triple-negative MDAMB231 xenografts in immunocompromised mice. We also compared the efficacy of phenformin and metformin in the treatment of MDAMB231.

Methods: Immunocompromised mice were divided into groups: (1) phenformin administered for 2 weeks prior to cell injection; (2) established tumours treated with phenformin; (3) established tumours treated with metformin (only for MDAMB231 tumours); (4) untreated controls. Post-treatment tumours, liver and spleen were harvested for further analysis.

Results: Phenformin significantly inhibited both the development and growth of MCF7 and MDAMB231 tumours, and for MDAMB231 at greater efficacy than metformin without murine toxicity. The number of mitotic figures was significantly fewer in xenografts treated with phenformin compared with controls. Results suggested that the mechanism of action of phenformin in vivo is consistent with AMPK activation.

Conclusion: Phenformin has clinical potential as an antineoplastic agent and should be considered for clinical trials both in ER-positive and triple-negative breast cancer.

Figure 1

Effect of 300 mg kg⁻¹ phenformin on human breast MCF7 (A) and MDAMB231 (B), and 300 mg kg⁻¹ metformin on MDAMB231 (C) tumour xenografts. For the phenformin experiment, mice were divided into three groups. Controls received no phenformin. Pre-treatment mice were given phenformin (300 mg kg⁻¹) in 5% sucrose instead of normal drinking water for 2 weeks prior to injection of MCF7 or MDAMB231 cells. The phenformin group received normal drinking water until tumours reached ⩾30 mm³, after which drinking water was replaced with 5% sucrose containing phenformin (300 mg kg⁻¹). Control group also had water replaced with 5% sucrose. For the metformin experiment, mice were divided into two groups. Controls received no metformin and mice with established tumours received metformin (300 mg kg⁻¹) in water. MCF7 tumours pre-treated or treated with phenformin had statistically significant inhibition of tumour growth of 88% relative to the control group (P<0.05). Animals injected with MDAMB231 cells and treated prophylactically with phenformin showed small lumps 6 weeks after inoculation, which remained static for the rest of the experiment. Established MDAMB231 tumours treated with phenformin demonstrated statistically significant inhibition of tumour growth of 60% relative to the control group (P<0.05). There were no statistically significant differences between control mice and mice treated with metformin.

Figure 2

Connective tissue histological analysis of MCF7 and MDAMB231 xenografts treated with phenformin using Van Gieson stain. MCF7 tumours treated with phenformin showed substantial increase of connective tissue, demonstrated as red/pink staining, which replaces epithelial tumour cells (A and B). No differences were observed in MDAMB231 tumours (C and D), × 5 magnified image captured with an Aperio ScanScope XT, Aperio Technologies, Vista, CA, USA.

Figure 3

Phospho-histone H3 immunohistochemistry analysis of MCF7 and MDAMB231 xenografts treated with phenformin. Tumours were harvested and immunohistochemistry analysis performed as described in Materials and methods. No significant differences were observed for phenformin (B and D) compared with untreated MCF7 and MDAMB231 tumours (A and C), × 5 magnified image captured with an Aperio ScanScope XT, Aperio Technologies.

Figure 4

Cleaved PARP analysis of MCF7 and MDAMB231 xenografts treated with phenformin. Tumours were harvested and immunohistochemistry analysis performed as described in Materials and methods. No significant differences were observed for phenformin (B and D) compared with untreated MCF7 and MDAMB231 tumours (A and C), × 5 magnified image captured with an Aperio ScanScope XT, Aperio Technologies.

Figure 5

Ki67 immunohistochemistry analysis of MCF7 and MDAMB231 xenografts treated with phenformin. Tumours were harvested and immunohistochemistry analysis performed as described in Materials and methods. No significant differences were observed for MCF7 tumour treated with phenformin compared with control (A and B). Similar results were found for MDAMB231 tumours (C and D), × 5 magnified image captured with an Aperio ScanScope XT, Aperio Technologies.

Figure 6

AMPK activation in liver and spleen of mice. Liver and spleen of control mice and mice that were pre-treated or treated after xenograft establishment with 300 mg kg⁻¹ phenformin were processed and western blots produced as described in Materials and methods. Lanes 1–4 show the results for control mice, lanes 5–9 show the results for mice pre-treated with phenformin and lanes 10–13 show the results for mice treated with phenformin. Phosphorylation of the activation loop of AMPK (T172) was enhanced in liver and spleen from mice pre-treated or treated (lanes 10–13) with phenformin compared with untreated control mice.

Figure 7

AMPK activation in MDAMB231 xenograft tumours. Tumours of control mice and mice that were pre-treated or treated with 300 mg kg⁻¹ phenformin were processed and western blots produced as described in Materials and methods. Lanes 1–5 show the results for control mice, lanes 6–9 for mice pre-treated with phenformin and lanes 10–13 for mice treated with phenformin. Phosphorylation of the activation loop of AMPK (T172) was enhanced in most of the tumours from mice pre-treated or treated with phenformin compared with untreated control mice.