Metformin-induced preferential killing of breast cancer initiating CD44+CD24−/low cells is sufficient to overcome primary resistance to trastuzumab in HER2+ human breast cancer xenografts

1De novo (i. e. primary) resistance to the monoclonal antibody trastuzumab (Herceptin) remains a prevalent challenge in the treatment of breast cancer patients whose tumors overexpress the human epidermal growth factor 2 (HER2). Evidence is mounting that the CD44+CD24−/low cell subpopulation, which is enriched with potential breast cancer stem cells (CSCs), could explain clinical resistance to HER2-targeted therapies. Therefore, future clinical trials should involve the integration of novel anti-breast cancer stem cells approaches to prevent and overcome the inherent unresponsiveness to trastuzumab across clinically important subgroups of HER2-positive breast cancer patients. Although in vitro studies have recently suggested that the anti-diabetic drug metformin can efficiently eliminate treatment-resistant stem/progenitor cell populations in heterogeneous breast cancer populations, it remained to be evaluated whether systemic metformin might overcome primary resistance to trastuzumab in HER2-gene amplified human breast cancer xenografts. Here we explored for the first time the in vivo sensitizing efficacy of metformin on trastuzumab therapy by using xenografts of the human breast cancer cell line JIMT-1, which was established from the pleural metastasis of a patient who was clinically resistant to trastuzumab ab initio. The JIMT-1 model is unique because of displaying at the same time several co-existing mechanisms of resistance to trastuzumab present at variable levels in other breast cancer cell lines, including moderate expression levels of HER2 (despite HER2 gene amplification), low expression of PTEN (phosphatase and tensin homolog), an activating mutation of the PIK3CA gene, high expression of NRG1 (neuregulin-1), and enhanced expression of mesenchymal markers, including a naturally enriched subpopulation of breast cancer initiating CD44+CD24−/low CSC-like cells.

2We first examined whether breast cancer initiating CD44+CD24−/low and non-CD44+CD24−/low cell subpopulations from the trastuzumab-refractory JIMT-1 cell line exhibited differential sensitivities to the growth inhibitory effects of metformin. We employed MTT-based cell viability assays to compare the degree of sensitivity to metformin of parental JIMT-1 cells and that of JIMT-1 cell subpopulations isolated and purified for breast cancer initiating CD44+CD24−/low and non-CD44+CD24−/low immunophenotypes by employing magnetic activated cell sorting (MACS; Fig. 1A, top panels). Cells were treated side-by-side on the following sorting day with graded concentrations of metformin for five days. While unsorted JIMT-1, CD44+CD24−/low, and CD44+CD24+ populations all showed an inhibitory response to metformin that augmented with the increase in metformin concentration (Fig. 1A, bottom left), each separate cell population differed in their extend of cytotoxic response to metformin. The breast cancer initiating CD44+CD24−/low subpopulation was significantly more sensitive than was the parental or CD44+CD24+ populations. CD44+CD24−/low cells had an estimated IC50 value of 1±0.2 mmol/L metformin whereas that in the corresponding non-CD44+CD24−/low population was 11±2 mmol/L metformin (Fig. 1A, bottom right). The IC50 of metformin in the unsorted JIMT-1 parental population was 8±2 mmol/L (Fig. 1A, bottom right). These findings, altogether, confirmed that the breast cancer initiating subpopulation within a heterogeneous population of trastuzumab-refractory, HER2-positive breast cancer cells notably exhibit greater sensitivity to the growth-inhibitory effects of metformin. Indeed, the growth of the breast cancer initiating CD44+CD24−/low population was more significantly affected by metformin at all tested concentrations. Because it has been shown that gradual loss of stem cell markers takes place in cells growing as a monolayer, these data may further underestimate hypersensitivity of the breast cancer initiating CD44+CD24−/low cells to metformin. Indeed, JIMT-1 cells sorted for the CD44+CD24−/low immunophenotype repopulated all the parental cell fractions after few days in monolayer culture (data not shown).

3The effects of metformin and trastuzumab alone and in combination on tumor growth were next studied in vivo using a JIMT-1 xenograft animal model (Fig. 1B). Compared to the control group after seven weeks of treatment (940±170 mm3), the trastuzumab-treated group likewise failed to exhibit significant reductions in mean tumor size (891±135 mm3). Compared to the mean xenograft tumor size in both the control and the trastuzumab group, the mean tumor size in the metformin group was significantly smaller (390±64 mm3), which confirmed that the inhibitory effect of metformin at the tested concentration (250 mg/kg/day) was notably stronger than that of trastuzumab (5 mg/kg/week) on JIMT-1 tumor growth. When the two drugs were combined, the xenograft tumor size (213±75 mm3) was smaller than those of the groups treated with trastuzumab or metformin alone, indicating that the combination of the drugs was much more effective at reducing tumor volume. Consequently, the days required for four-fold increase in tumor volume was 27±5 after treatment with trastuzumab whereas more than 50 days were required after the combined treatment with trastuzumab and metformin (Fig. 1B). No significant difference in body weight was observed in xenograft tumor-bearing mice between the treatment groups (data not shown).

4Our current findings reveal that: a.) Metformin as single agent is active against JIMT-1-derived tumor xenografts with primary resistance to trastuzumab; b.) metformin’s ability to distinctively kill breast cancer initiating CD44+CD24−/low mesenchymal subpopulations is sufficient to overcome in vivo primary resistance to trastuzumab in HER2-gene amplified breast tumors. A combination of trastuzumab and metformin may provide a valuable strategy for treatment of HER2-overexpressing breast cancer, and a phase II trial recently has opened in Spain to evaluate the efficacy of trastuzumab plus metformin as neo-adjuvant therapy for patients with HER2-positive breast cancer.