Background
Triple-negative breast cancer (TNBC), which represents 10–20% of all breast cancers, is characterized by a lack of expression of the estrogen steroid receptor (ER), progesterone steroid receptor (PR), and tyrosine kinase human epidermal growth factor receptor 2 (HER2) [
1]. Compared to other cancer subtypes, TNBC tumors are more frequently diagnosed as aggressive, invasive, grade III, and lymph node-positive [
2]; however, no effective targeted therapy is currently available for the treatment of TNBC. Although approximately 50% of all patients with TNBC respond to conventional chemotherapies [
3,
4], the effectiveness of these treatments is limited by the development of drug resistance [
5,
6].
Cisplatin is widely used to treat solid tumors, including breast, testicular, and ovarian cancers [
7]. Cisplatin exerts its anticancer effects by inducing DNA double-strand breaks (DSBs) [
8,
9]. Despite a consistent initial response, cisplatin treatment results in the development of chemoresistance. For example, patients who initially respond to cisplatin therapy often develop resistance due to activation of the homologous recombination (HR) DNA repair mechanism [
10,
11]. Multiple mechanisms underlying the development of resistance include altered cellular accumulation [
12], increased drug inactivation [
13], and DNA repair [
14].
Homologous recombination is an error-free DNA repair mechanism for DSBs that is activated when cells are exposed to genotoxic stress [
15,
16]. RAD51 is a strand transferase that polymerizes into a nucleoprotein filament on single-stranded DNA and promotes DNA strand exchange with the undamaged homologous chromatid [
17]. Because RAD51 is an integral component of the cellular DNA damage response, its suppression sensitizes cancer cells to DNA-damaging drugs [
18,
19]. In contrast, high levels of RAD51 have been linked to elevated rates of DNA recombination and enhanced resistance to DNA-damaging chemotherapies and/or ionizing radiation [
20,
21]. In addition, RAD51 facilitates TNBC metastasis [
22], indicating that RAD51 is a therapeutic target for TNBC treatment.
Metformin (1,1-dimethylbiguanide hydrochloride), the most commonly prescribed oral antidiabetic medication, may be of benefit to diabetic cancer patients [
23]. Notably, the breast cancer risk has been shown to be lower in diabetic patients treated with metformin than in those treated with other antidiabetic medications [
24].. Metformin was shown to inhibit the DNA damage repair pathway in pancreatic cancer [
25], p53-deficient colorectal cancer [
26], and non-small cell lung cancer (NSCLC) cells [
27] by downregulating RAD51, indicating the anticancer effects of metformin. In addition, increased glucose concentrations reduced the efficacy of metformin [
28], implying that high glucose levels may negatively influence the anticancer efficacy of metformin. In our study, we also found that metformin decreased RAD51 expression more efficiently in culture conditions containing a normal glucose concentration (5 mM) than in conditions with high glucose concentrations (25 mM). Moreover, metformin also enhanced the therapeutic effect of cisplatin in ovarian cancer [
29], nasopharyngeal carcinoma cells [
30], lung tumors [
31], and oral squamous carcinoma cells [
32]. These observations led us to hypothesize that metformin may sensitize TNBC cells to cisplatin by downregulating RAD51 under physiological glucose concentrations. In the present study, we explored the therapeutic role of metformin and demonstrate that, in combination with cisplatin, metformin is effective TNBC treatment outcomes.
Methods
Reagents
Antibodies against RAD51 and phospho-H2AX (Ser139) were purchased from Abcam (Cambridge, UK). Antibodies against ubiquitin were purchased from Cell Signaling Technology (Danvers, MA, USA), while antibodies against β-actin were from Sigma-Aldrich (St. Louis, MO, USA). Anti-ERK1/2 and anti-phospho-ERK1/2 (Thr202/Tyr204) antibodies were procured from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG and goat anti-mouse IgG secondary antibodies were obtained from Enzo Life Sciences (Farmingdale, NY, USA). Cisplatin, metformin, MG132 (carbobenzoxy-Leu-Leu-leucinal), cycloheximide (CHX), PD98059, and lactacystin were obtained from Sigma-Aldrich. Protein A agarose beads were acquired from GE Healthcare (Piscataway, NJ, USA).
Cell culture
MDA-MB-231 and Hs 578T human breast cancer cells (ATCC, Rockville, MD, USA) were maintained in Dulbecco’s high glucose (25 mM glucose) modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS) (Gibco, Carlsbad, CA, USA), 100 U/mL penicillin, and 100 μg/mL streptomycin. MCF10A cells were grown in DMEM/F-12 medium (Gibco) containing 5% horse serum (Gibco), 20 ng/mL EGF, 0.5 mg/mL hydrocortisone, 100 ng/mL cholera toxin, 10 μg/mL insulin, 100 U/mL penicillin, and 100 μg/mL streptomycin. All cells were cultured at 37 °C in a humidified incubator with 5% CO2.
MTT assay
Cell viability was measured by MTT assay. MDA-MB-231 and Hs 578T cells were seeded into 96-well plates at a density of 1 × 103 cells/mL. Growth medium was replaced with normal (5.5 mM) glucose medium 24 h prior to treatment. Subsequently, MTT (0.5 mg/mL) was added and the cells were incubated for 2 h at 37 °C. The cells were then lysed with DMSO, and the absorbance at 540 nm was measured using a microplate reader (Bio-Rad, Hercules, CA, USA).
Western blot analysis
The medium was removed, and cells were washed with ice-cold phosphate-buffered saline (PBS). The cells were then lysed in 100 μL of lysis buffer (50 mM Tris-HCl [pH 7.4], 1% Triton X-100, 0.25% sodium deoxycholate, 150 mM EDTA, 1 mM sodium orthovanadate [Na3VO4], 1 mM NaF, 1 mM phenylmethylsulfonyl fluoride [PMSF]). Proteins were resolved on 10% SDS-PAGE gels and transferred to nitrocellulose membranes. The membranes were blocked in 5% dry milk (w/v) for 1 h and then washed three times in TBST (Tris-buffered saline with Triton X-100). The membranes were incubated overnight at 4 °C with primary antibodies and then probed with an HRP-conjugated secondary antibody for 1 h. Blots were visualized using the Amersham Biosciences ECL Detection System (Amersham plc, GE Healthcare, Chicago, IL, USA).
siRNA transfection for RAD51 knockdown
MDA-MB-231 and Hs 578T human breast cancer cells were seeded in six-well plates and transfected at 60% confluence with RAD51-targeting siRNA duplexes or a negative control siRNA (L-003530-00-0005; UAUCAUCGCCCAUGCAUCA, CUAAUCAGGUGGUAGCUCA, GCAGUGAUGUCCUGGAUAA, and CCAACGAUGUGAAGAAAUU) purchased from Dharmacon (Lafayette, CO, USA). For transfection, 5 μL of siRNA targeting human RAD51 (CR536559) and 5 μL of Lipofectamine were each diluted in 95 μL of reduced serum medium (Opti-MEM, Invitrogen, Carlsbad, CA, USA). The mixtures were incubated for 15 min before being added dropwise to the culture wells containing 800 μL of Opti-MEM to achieve a final siRNA concentration of 50 nM.
Construction of pFLAG-RAD51
Human RAD51 was cloned into the BamHI and SalI sites of the pCMV-Tag 2C vector (Stratagene, San Diego, CA, USA). The cDNA from MDA-MB-231 cells was amplified by polymerase chain reaction (forward primer: 5′-CGGGATCCATGGCAATGCAGATGCAGC-3′; reverse primer: 5′-ACGGCGTCGACTCAGTCTTTGGCATCTCCCAC-3′), digested with BamHI and SacI, and then ligated to a linearized pCMV-Tag 2C vector. The construct was verified by DNA sequencing.
Wound healing assay
Confluent cells were serum-starved for 12 h, after which a standardized cell-free area was introduced by scraping the monolayer with a sterile tip. Cells were imaged using a phase-contrast microscope. After intensive washing, fresh medium supplemented with 10% FBS containing both metformin and cisplatin was added. After incubation for 36 h, three random areas of cells were imaged. Migrated cells were quantified by manual counting, and the inhibition ratios were expressed as percentages of control cells.
Invasion assay
The upper chamber of a Transwell insert (8-μm pore size) was coated with 100 μL of Matrigel (BD Biosciences, Bedford, MA, USA) and PBS, followed by drying for 30 min at 37 °C. Cells were suspended in serum-free medium (100 μL; 4 × 105 cells/mL) and layered in the upper compartment of the chamber. The bottom chambers were supplemented with 500 μL of complete medium (10% FBS) containing the indicated concentrations of both metformin and cisplatin. After incubation for 24 h, the invading cells on the lower face were fixed in 4% paraformaldehyde and stained with crystal violet (Sigma-Aldrich). Random fields were counted, and representative images were obtained using an AxioCam HRC CCD camera (Carl Zeiss, Oberkochen, Germany).
Immunoprecipitation
Cellular protein (1 mg) was mixed with 1 μg of anti-RAD51 rabbit monoclonal antibodies and incubated at 4 °C for 24 h. Immune complexes were captured with protein A sepharose (Amersham, Uppsala, Sweden) for an additional 3 h. The precipitated immune complexes were washed three times with wash buffer, resuspended in SDS sample buffer (125 mM Tris-HCl [pH 6.8], 20% [v/v] glycerol, 4% [w/v] SDS, 100 mM dithiothreitol, and 0.1% [w/v] bromophenol blue), and heated at 95 °C for 5 min prior to electrophoresis.
Immunofluorescence staining
Cells were seeded in 12-well plates at a density of 5 × 104 cells/well on a sterile coverslip. After treatment with metformin or cisplatin, the cells were washed with PBS, fixed in 4% formaldehyde, and permeabilized with 0.2% Triton X-100 in PBS for 15 min. After blocking with 2% bovine serum albumin (BSA) in PBS for 1 h at room temperature, the cells were incubated overnight with primary antibodies against RAD51 and γ-H2AX in blocking buffer at 4 °C. The cells were then washed in PBS and incubated with Alexa Fluor 488-conjugated chicken anti-rabbit IgG secondary antibodies (1:500, Invitrogen) and Alexa Fluor 568-conjugated donkey anti-rabbit antibodies (1:500, Invitrogen) for 1 h at room temperature. The cells were counterstained with Hoechst 33342 for 10 min before the final wash. Images were captured using a confocal microscope (Zeiss LSM 700 Meta, Carl Zeiss) at × 10 magnification.
Experimental animals and tumor inoculation
Forty female BALB/c mice were randomly divided into four groups of 10 mice each. The mice in the control group were inoculated with 4T1 cells, while those in the metformin group were injected intraperitoneally with metformin (150 mg/kg body weight per day) for 21 days. The experiment was approved by the Korea University Institutional Animal Care and Use Committee (IACUC) and was performed according to the guidelines and regulations. The mice in the combination therapy group were injected intraperitoneally with metformin (150 mg/kg body weight per day) and cisplatin (3 mg/kg body weight once every 3 days) for 21 days. Mice in the cisplatin group were injected intraperitoneally with cisplatin (3 mg/kg body weight once every 3 days) starting from day 5 of tumor inoculation. The body weight of each mouse was determined daily during the entire experimental period. The 4T1 tumor cell suspension was diluted in PBS and injected subcutaneously (0.2 mL, 4 × 105 cells/mouse) and bilaterally into the fourth pair of mammary fat pads of each mouse. All injections were administered in a 0.15-mL volume. Tumor growth was determined by measuring the tumor diameter in two dimensions with a caliper every 3 days, and the tumor volumes ([width2 × length]/2) were calculated. Body weight was recorded to monitor the side effects of the drugs. Breast tumors and gonadal fat pads were either homogenized to prepare tissue lysates for western blot analysis, or formalin-fixed, embedded in paraffin, and cut into 5-μM sections for immunohistochemistry (IHC).
Immunohistochemical analysis
Paraformaldehyde (4%)-fixed samples were gradually dehydrated in a graded ethanol series and cleared in xylene using a Leica AS300S tissue processor (Leica Microsystems GmbH, Wetzlar, Germany). The samples were then infiltrated with paraffin and cut into 5-μm sections using a Leica RM2255 rotary microtome (Leica Microsystems GmbH). Representative blocks of paraffin-embedded tissues were dewaxed and rehydrated. Briefly, sections were deparaffinized, rehydrated, and washed in PBS. To block nonspecific binding, sections were incubated in 4% BSA-dextran for 1 h at 4 °C. Sections were incubated with anti-RAD51 antibodies diluted 1:200 in 1% BSA and 0.1% Nonidet P-40 in PBS overnight at 4 °C. The Vectastain ABC kit (Vector Labs, Burlingame, CA, USA) was used to amplify the signal using the avidin-biotin complex (ABC) method according to the manufacturer’s instructions. Peroxidase activity was visualized with 3,3′-diaminobenzidine (DAB; Darko, Carpinteria, CA, USA). Sections were lightly counterstained with hematoxylin, dehydrated through an ethanol series to xylene, and mounted. Slides were visualized and imaged using a light microscope equipped with a computer-controlled digital camera.
Statistical analysis
Data are expressed as means ± SEM. One-way analysis of variance (ANOVA) was performed to compare multiple groups followed by Bonferroni’s post hoc test. A P value of 0.05 or lower was considered significant in all experiments. All analyses were performed using Sigma plot software (Systat Software Inc., San Jose, CA, USA). P values less than 0.05 were considered significant and were presented as #, ## vs. no treatment; #P < 0.05, ##P < 0.01, ###P < 0.001, *P < 0.05, **P < 0.01, **P < 0.001 by one-way ANOVA followed by Bonferroni’s post hoc test.
Discussion
Cisplatin resistance limits therapeutic options in patients diagnosed with TNBC. The main objectives of our study were to determine if metformin sensitized human TNBC cells to cisplatin and, if so, to identify the molecular signaling pathways involved. The principal findings of our study were that metformin acted as a cisplatin sensitizer in TNBC chemotherapy and that RAD51 played a critical role in the synergistic effect of metformin on cisplatin. Consequently, RAD51 represents a potential therapeutic target in TNBC patients.
Although single-agent therapy has yielded positive results in cell lines and preclinical models, it failed to show promising results in managing aggressive TNBC in clinical trials, likely due to therapy heterogeneity and potential for acquired drug resistance [
37]. Several studies have shown that combining metformin with cisplatin is effective in treating various cancers, including ovarian carcinoma [
29], human nasopharyngeal cell carcinoma [
30], lung carcinoma [
31], and oral squamous cell carcinoma [
32]. In addition, metformin reduces cisplatin-induced side effects like cognitive impairment, brain damage [
38], and peripheral neuropathy [
39] in mice. This is the first study exploring the chemosensitizing effect of metformin on cisplatin against TNBC cells through the regulation of DNA damage repair.
In this study, we found that metformin sensitized MDA-MB-231 and Hs 578T TNBC cells to cisplatin based on cell viability (Fig.
1c, d). Metformin also enhanced cisplatin-mediated inhibition of migration and invasion (Fig.
1e–h). Our results indicate that the anticancer effects of metformin under reduced glucose were more pronounced in MDA-MB-231 than HS-578T cells. Most in vitro studies have shown the efficacy of metformin as an anticancer agent using very high concentrations (> 5 mM), which may be due to the high glucose concentrations used in the culture of most cancer cell lines. The presence of glucose at high concentrations reduced the antineoplastic efficacy of metformin, indicating that investigations on the anticancer effects of metformin should be performed under physiologically relevant glucose concentrations. Metformin also exhibited significant biological activity in a 4T1 mouse breast cancer model in vivo. In mice with normal levels of glucose and insulin, combined metformin and cisplatin treatment decreased the tumor volume to a significantly greater extent than cisplatin treatment alone (Fig.
8c, d), suggesting that metformin has potential as a therapeutic agent against TNBC in combination with cisplatin.
However, for successful clinical application, a few limitations should be considered. First, it is still unknown whether the anticancer effects of metformin are replicated in clinical models. Therefore, studies are necessary to determine the most appropriate dose and establish the safety of metformin in patients with TNBC. Second, although metformin is used as the first-line treatment for type 2 diabetes, the appropriate range for its therapeutic concentration is still confounding. According to previous studies, a range of approximately 5 mM metformin was effective in breast cancer cell lines [
40,
41]. Moreover, metformin accumulated and reached tissue concentrations substantially higher than those found in the plasma [
42], implying that the therapeutic metformin plasma concentration might be lower than that for tissue. Therefore, the metformin concentration (5 mM) used in the present study seems appropriate and is considered relevant for use in vitro studies.
Elevated expression of RAD51 is associated with tumor aggressiveness and is known to confer treatment resistance in a variety of tumors, including ovarian cancer [
43], breast cancer [
44], lung tumors [
45], pancreatic adenocarcinomas [
46], and malignant gliomas [
47]. Furthermore, downregulation of RAD51 protein levels by antisense oligonucleotides, RNA interference [
48], aptamers [
49], or small-molecule inhibitors can be used to sensitize tumors to chemotherapy or radiation. In this study, we found that RAD51 expression increased in a dose- and time-dependent manner following cisplatin treatment, whereas it decreased in a dose- and time-dependent manner with metformin treatment (Fig.
2a–d). Interestingly, metformin inhibited cisplatin-mediated RAD51 upregulation (Fig.
2e), indicating that the metformin-mediated downregulation of RAD51 may inhibit resistance to cisplatin in TNBC cells. We further investigated the effect of metformin on the normal breast epithelial cells, MCF10A. Metformin decreased the expression of RAD51 and inhibited the cisplatin-mediated RAD51 expression in MCF10A (Fig.
2f). Previous reports showed that extracellular vesicles (EVs) from triple-negative breast cancer cells promoted proliferation and drug resistance in MCF-10A [
50,
51], implying that TNBC-mediated EVs (TNBC-EVs) may induce tumorigenic potentiality in normal cells. Combined with the result of Fig.
2f, metformin may reduce cisplatin resistance induced by TNBC-EVs in normal tissues via RAD51. In addition, it was reported that metformin selectively targeted cancer stem cells and also induced apoptosis in human breast carcinoma cell line MCF-7 with minimal toxicity to MCF10A [
52,
53]. Furthermore, metformin prevented normal cell apoptosis against cisplatin-induced ototoxicity and nephrotoxicity in auditory cell and tubular cell [
54]. Together, these findings indicate that metformin may be a potentially adjuvant therapy drug to combine with cisplatin. In the future, in-depth studies are necessary to determine appropriate modes of combination therapy of metformin and cisplatin.
Moreover, we confirmed the effect of RAD51 on the metformin-induced inhibition of migration and invasion after knock down or overexpression of RAD51 using RAD51 siRNA and RAD51-flag. As expected, RAD51 overexpression blocked metformin-mediated inhibition of migration and invasion while its downregulation enhanced the effect of metformin (Fig.
7e, f). This suggests that RAD51 is a potential therapeutic target for TNBC treatment. In support of our findings, studies have shown that RAD51 overexpression contributes to chemoresistance in human soft tissue sarcoma cells [
55] and rescues radiation sensitivity in BRCA2-defective cancer cells [
56].
Double-strand breaks represent one of the most important types of cisplatin-induced DNA damage. In response to DSBs, histone H2AX is rapidly activated and phosphorylated, generating γ-H2AX. In this study, metformin enhanced the cisplatin-mediated phosphorylation of γ-H2AX (Fig.
6b, c), suggesting that metformin prolongs the process of cisplatin-induced DSB repair and regulates the γ-H2AX-RAD51 axis to overcome resistance to cisplatin.
Reduced food intake and weight loss are serious health concerns in patients undergoing cisplatin therapy [
57]. In this study, cisplatin treatment resulted in progressive weight loss. Interestingly, however, metformin and cisplatin combination treatment attenuated the cisplatin-mediated weight loss (Fig.
8b). Our data demonstrated that metformin attenuates cisplatin-induced side effects and potentiates cisplatin-mediated anticancer effects.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.