Introduction

Metformin is one of the most widely described blood glucose-lowering- agents used to treat type 2 diabetes with a high safety profile (Foretz et al., 2014; Bailey, 2017). Over the years, studies have shown that the use of metformin reduced cancer incidence in various tumors (Evans et al., 2005). Since then, an increasing number of clinical studies have reported the anti-cancer effects of metformin in several cancer types (Suissa & Azoulay, 2012; Margel et al., 2013; Niraula et al., 2013; Tang et al., 2017). This is not surprising given the cytotoxic effect metformin has on cancer cells, however how the drug reduces tumor growth and the exact mechanism of action remains to be not fully understood (Algarni et al., 2020; Sarami et al., 2020).

 Recently, more attention has been directed toward understanding metformin’s anti-tumor effects, specifically metabolic reprogramming. It is known that energy metabolism is a crucial tumor hallmark that plays a role in cancer development and progression and thus is an attractive target for cancer therapeutics (Mohamed et al., 2020; Sun et al., 2021). Metformin is known to inhibit mitochondrial electron transport chain (ETC) complex 1, therefore, decreasing cellular respiration. This was seen in human cancer cells, and indeed metformin was able to induce cell death with glucose deprivation, implying that cancer cells would rely on glycolysis in the presence of metformin (Wheaton et al., 2014; Aleidi et al., 2022). Metformin’s metabolic-effect complexity was further seen in a study that showed that the tumor environment can alter cells’ sensitivity to metformin (Gui et al., 2016). For example, cells grown in DMEM without Pyruvate were more sensitive to metformin's anti-proliferative effect (Gui et al., 2016). Overall, this was attributed to the fact that environmental cues and metabolites can alter a cell's dependency on complex 1 and cellular respiration.

Despite some of the available metabolic studies for metformin, many questions remain unanswered including which metabolites can be affected by the drug, whether its effect is cell-type dependent, and perhaps what compensatory pathways induced by metformin treatment should be targeted simultaneously.

In this study, we were interested in studying the effect of metformin treatment on two different breast cancer cell lines, the triple-negative breast cancer cell line (TNBC) MDA-MB-231, and the estrogen-receptor-positive MCF-7, and to investigate the effect of the drug on their metabolites (Samir et al., 2019; Aloufi, 2022; Daivasigamani et al., 2022).

Materials and Methods

Drug Preparation

Metformin (Sigma-Aldrich, Zwijndrecht, The Netherlands) was dissolved in PBS to prepare a 1 M solution and stored at -20 °C until further use.

Cell Culture

 Human breast cancer cell lines MCF-7 (estrogen-receptor-positive) and MDA-MB-231(triple-negative) cells were obtained from the Regenerative Medicine Unit, King Fahad Medical Research Center, SA. All cells were maintained in DMEM (bio sera-France) medium supplemented with 10 % fetal bovine serum (FBS: bio sera-France) and 5% antibiotics (penicillin: bio sera-France), and incubated in 5% CO2 at 37˚C.

Viability Assay

The Cell Counting Kit-8 (CCK-8) assay kit (Dojindo, Japan) was used to assess cell viability. Briefly, cells were plated at a density between 4000 and 5000 cells/well in 96-well plates in 200 μl of DMEM supplemented with 10% FBS, and 5% antibiotics (penicillin). After 24 h, cells were treated with various metformin concentrations diluted with a culture medium. Five replicates were prepared for each concentration. Cells were then incubated at 37˚C for 24, 48, and 72 h, and then 5 μl of CCK-8 reagent were added for 2 h at 37˚C. Finally, the optical density was measured using a microplate reader (BioTek-ELx808) at 450 nm. The mean values of five readings were calculated, growth curves were drawn and IC50 was determined.

Apoptosis Assay

MCF-7 and MDA-MB-231 were either treated with the predetermined IC50s of metformin or fed with drug-free media as control cells, for 48 h. As a positive control, cells were treated with 40 μM of Thymoquinone (Bashmail et al., 2018). Cells were harvested and washed twice with PBS, and 3 µl of Annexin V-FITC/PI solution (Abcam Inc., Cambridge Science Park, Cambridge, UK) was added to the cells. Cells were then kept at room temperature in the dark for 5 mins. Three replicates were prepared for each condition. Following incubation, cells were injected via BD FACSAria ™ flow cytometer (BD Biosciences Inc., San Jose, CA, USA) and assessed for FITC and PI fluorescent signals

Stem Cell Detection

To study the effect of metformin on BCSC (CD44+/CD24−), cells were treated with the predetermined IC50s of metformin for 24h, while control cells were cultured with drug-free media.  After 24 h, cells were harvested and washed with ice-cold PBS. Cells were then incubated at 4 C for 30 min in the dark with 100μL of 1 μg/mL conjugated anti-CD44 and 1 μg/mL anti-CD24 antibodies (FITC labeled anti-CD44 (Cat. # AB27285) and APC/ Cy7 labeled anti-CD24 antibodies (Cat. #AB197137) (Abcam Inc., Cambridge Science Park, Cambridge, UK) or their isotype controls mouse IgG1, monoclonal [HyblgG1] (FITC)- isotype control (Cat. #Ab1264) and mouse IgG1, kappa monoclonal (APC/CY7)-isotype control (Cat. #AB46739). After incubation, cells were washed three times with ice-cold PBS. Finally, cells (10,000 events per sample) were assessed for FITC and APC/CY7 signals via BD FACSAria ™ flow cytometer (BD Biosciences Inc., San Jose, CA, USA) Cells quantity were analyzed by quadrant analysis using BD FACSDiva Software

 Untargeted Metabolomic Experiments

To study the effect of metformin on breast can cell metabolites, 3x106 cells were treated with the predetermined IC50s of metformin for 24h, while negative control cells were cultured with drug-free media. Cells were harvested and washed twice with PBS and frozen immediately as a pellet in liquid nitrogen.

Metabolites were extracted from the cell lines (MCF-7 and MDA-MB-231) treated with metformin and control untreated cells by using the solvent in a ratio of methanol: acetonitrile: water (2:2:1 v/v). One mL of ice-cold solvent was added to the cells to form a cell lysate. The cell lysate was vortexed and then incubated for 1 h at 20° C. Following incubation cells were centrifuged at 13,000 rpm at 4 °C for 15 mins. The supernatant was saved, and the sample was dried in a vacuum concentrator. To remove insoluble debris, the dry extracts were resuspended in 100 µL of acetonitrile: water (1:1, v/v), vortexed, and then centrifuged at 13,000 rpm at 4 °C for 15 mins .10 µL of the sample was injected into an HPLC (Thermo Finnigan Surveyor Autosampler Plus) column (Hypersail gold; 150mm x 4.6 mm, 5 µm) with a flow rate of 0.250 mL/min. The mobile phase (A) of 0.1% Formic acid and (B) 99.9% ACN Formic acid (0.1%, v/v), using a linear gradient program where the component of solution was changed from 5% B to 100% B over 90 min, at a constant flow rate of 0.2 mL/min (95% A from 5% to 30% over 72 min, 30% to 100% over 10 min, and kept at 100% for 5 min at a flow rate of 0.25 mL/min). The column temperature was maintained at 30° C and metabolites were analyzed in LTQ XL linear ion trap instrument (Thermo Fisher Scientific). Full scan scope was chosen from 80 to 1000 m/z. The spray voltage was set at -3.0 kV, the capillary voltage was fixed at 4.0 V. The flow rate of nitrogen (sheath gas) was set at 40 arbitrary units. Further, helium was used as the buffer gas for run 187. The raw data were processed using (XC-MS) data processing software.

Statistical Analysis

Statistical analysis was performed using GraphPad Prism® for Windows, version 8.0 (GraphPad Software Inc., La Jolla, CA, USA). All assays were done in triplicates or more as mentioned. Data were presented as mean ± SD. The differences between groups were tested using a student t-test (unpaired) between two groups, or a one-way ANOVA test when more than two groups. P<0.05 was taken as a cut-off value for significance. Statistical analysis of metabolomic results and pathway analysis were performed using Metaboanalyst and metabolite identification using the human metabolome database (HMDB).

Results and Discussion

Effect of Metformin on  Cell Viability and Apoptosis of Breast Cancer Cell Lines

To study the anti-tumor effect of metformin on breast cancer cell lines, cell viability post-metformin treatment was determined using the CCK-8 assay. At 72 hours, metformin showed a decrease in cell viability in MCF-7 and MDA-MB-231 with IC50s of 14.21±1.33 mM and 12±5.8 mM respectively (Figure 1).

To determine the proportion of apoptotic cells induced by metformin, both cell lines were treated with the predetermined IC50s for 48 hr. Cells were assessed by Annexin- V/FITC apoptosis detection assay which showed that metformin-induced apoptosis in both cell lines MCF-7 and MDA-MB-231 (2.1±0.5% and 22.1±2.8%s respectively) (Figure 2).

Effect of Metformin on BCSC Population

To study the effect of metformin on tumor-associated stem cells, cells were treated with metformin, and CD44+/CD24- cell clone was assessed using flow cytometry. In MCF-7 cells treated with metformin for 24hr did not significantly alter the percentage of CD44+/CD24- cell clones compared with control untreated cells with a mean of 96.7 ± 0.5 % and 94.9± 1.0% respectively. In MDA-MB-231 cells, however, treatment with metformin appeared to increase CD44+/CD24- cell clone 91.3 ± 0.7% compared to control untreated cells 85.7± 2.8% (P<0.05). Expression levels of CD44 and CD24 were plotted as a percentage of total events (Figure 3).

Effect of Metformin on Breast Cancer Cell Lines Metabolites

To study the effect of metformin treatment on MCF-7 and MDA-MB-231 metabolites, both cell lines were either left untreated as control cells or treated with the predetermined IC50s of metformin for 24h. An untargeted metabolite analysis using LTQ XL linear ion trap mass spectrometer LC/MS-based was done. After analyzing the feature peaks (metabolites), 31 and 39 metabolites were detected in MCF-7 and MDA-MB-231, respectively (Bhavyasri et al., 2023; Choudhary et al., 2023; Nwankwo et al., 2023; Sarangi et al., 2023).

Two-dimensional principal component analysis (PCA) models score plots of all samples showed that all groups of samples were tightly clustered and revealed a significant difference between treated and untreated samples (Figure 4). The heat map further determined that metformin had strong effects, resulting in a metabolic variation on both cell lines used in our study (Figure 5).

Among the 31 metabolites that were detected in MCF-7, 12 metabolites were significantly different when compared to control untreated cells, 10 of them were upregulated while 2 were downregulated (Table 1). Interestingly, different metabolites were detected with MDA-MB-231 cells, and among the 39 detected metabolites, 30 metabolites were significantly different when compared to control untreated cells (Table 2). Interestingly all of those 30 metabolites were upregulated. Using metaboanalyst, the Enrichment Pathway Analysis of metabolites was obtained. The result showed that a total of 11 and 18 metabolic pathways were shown to be enriched in metformin-treated MCF-7 and MDA-MB-231 respectively. These pathways involved mainly carbohydrate metabolism in MCF-7 cells while lipid and amino acid metabolites appeared to be more enriched in MDA-MB-231 cells (Figure 5).