Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Aug 25.
Published in final edited form as: J Pharmacol Exp Ther. 2024 Nov 22;392(1):100034. doi: 10.1124/jpet.124.002424

Metformin in overcoming enzalutamide resistance in castration-resistant prostate cancer

Kendall Simpson 1, Derek B Allison 2,3, Daheng He 3, Jinpeng Liu 3,4, Chi Wang 3,4, Xiaoqi Liu 1,3,*
PMCID: PMC12371561  NIHMSID: NIHMS2102500  PMID: 39893002

Abstract

Androgen deprivation is the standard treatment for patients with prostate cancer. However, the disease eventually progresses as castration-resistant prostate cancer (CRPC). Enzalutamide, an androgen receptor inhibitor, is a typical drug for treating CRPC and with continuous reliance on the drug, can lead to enzalutamide resistance. This highlights the necessity for developing novel therapeutic targets to combat the gain of resistance. Metformin has been recently investigated for its potential antitumorigenic effects in many cancer types. In this study, we used enzalutamide and metformin in combination to explore the possible rescued efficacy of enzalutamide in the treatment of enzalutamide-resistant CRPC. We first tested the effects of this combination treatment on cell viability, drug synergy, and cell proliferation in enzalutamide-resistant CRPC cell lines. After combination treatment, we observed a decrease in cell proliferation and viability as well as a synergistic effect of both enzalutamide and metformin in vitro. Following these results, we sought to explore how combination treatment affected mitochondrial fitness using mitochondrial stress test analysis and mitochondrial membrane potential shifts due to metformin’s action in inhibiting complex I of oxidative phosphorylation. We employed 2 different strategies for in vivo testing using 22Rv1 and LuCaP35CR xenograft models. Finally, RNA sequencing revealed a potential link in the downregulation of rat sarcoma–mitogen-activated protein kinase signaling following combination treatment.

Keywords: Metformin, Prostate cancer, Enzalutamide resistance

1. Introduction

Prostate cancer (PCa) is responsible for the highest number of new cases in men in the United States with approximately 299,010 new cases in 2024 and is the second leading cause of cancer-related deaths with an estimated 35,250 deaths in 2024 (Siegel et al, 2024). Early stage patients with PCa who undergo localized therapies, radical prostatectomy, and hormone therapies will often experience cancer regression and symptom relief (Shoag et al, 2020). Hormone therapy, commonly referred to as androgen-deprivation therapy, is used to prevent androgen receptor (AR) signaling and therefore PCa progression; however, over time many patients will experience recurrence and are considered to have castration-resistant prostate cancer (CRPC) (Sharifi et al, 2005; Chandrasekar et al, 2015). Upon cancer recurrence, therapeutic options become more limited, and patients will often be treated with US Food and Drug Administration (FDA)–approved AR inhibitors such as abiraterone, enzalutamide, and darolutamide. In the case of metastatic CRPC, the only FDA-approved AR inhibitor available is enzalutamide; however, most patients being treated with enzalutamide over time will experience enzalutamide resistance (Antonarakis et al, 2014; Efstathiou et al, 2015; van Soest et al, 2015). This evidence demonstrates the critical need for the development of novel treatment strategies in advanced drug-resistant CRPC.

In recent years, metformin, the most commonly prescribed oral biguanide to treat type 2 diabetes, has gained traction with its implications for reduced cancer risk and potential use as a cancer treatment (Kasznicki et al, 2014; Zhao et al, 2019). Metformin has limited side effects and an excellent safety profile; hence, the investigation into possible drug-repurposing as a cancer therapy is an attractive option in many cancer types (Sleire et al, 2017). In addition to numerous mechanistic studies in PCa using metformin as a cancer therapy (Ben Sahra et al, 2008; Shao et al, 2015; Chen et al, 2016; Kong et al, 2020), there have been multiple clinical trials in recent years exploring this mechanism of cancer treatment (Murtola et al, 2008; Spratt et al, 2013). In particular, a phase II clinical trial in Switzerland used a combination treatment of enzalutamide and metformin in patients with CRPC who had never been exposed to enzalutamide and other endocrine agents (Rothermundt et al, 2014). Although there is validity in using enzalutamide and metformin in combination for patients with CRPC, there is little known about this combination treatment in drug-resistant CRPC.

In this study, we found that the combination treatment of enzalutamide and metformin in established drug-resistant CRPC lines demonstrated a synergistic antiproliferative effect in vitro. In addition, we investigated the effect of combination treatment on mitochondrial function using a mitochondrial stress test seahorse assay and measuring the mitochondrial membrane potential (MMP); however, we did not observe any significant effects on drug-resistant CRPC lines. To validate our synergy results in vivo, we employed 2 different xenograft models to determine the effects of combination treatment on tumor growth; however, we did not observe a difference in tumor growth between treatment groups. Finally, we treated drug-resistant CRPC lines with combination therapies for RNA sequencing to determine a mechanistic link. Together, these results highlight the importance of using robust models in cancer research to test novel treatment strategies.

2. Materials and methods

2.1. Cell culture, chemicals, and reagents

LNCaP, MR49F, C4–2, C4–2R, and 22Rv1 cell lines were used in this study. LNCaP cells are androgen-dependent cells; however, C4–2 cells were derived from LNCaP cells and are androgen independent. In a similar fashion, MR49F cells are also derived from LNCaP cells; however, MR49F cells are enzalutamide resistant. C4–2R cells are enzalutamide-resistant cells derived from C4–2 cells. C4–2 cells were obtained from the MD Anderson Cancer Center, whereas MR49F and C4–2R cells were kindly provided by Dr Amina Zoubeidi at the Vancouver Prostate Cancer Center and Dr Allen Gao at the University of California at Davis, respectively. LNCaP and 22Rv1 cells were purchased from ATCC. All cells were cultured in RMPI-1640 media supplemented with 10% (v/v) fetal bovine serum, 100 units/mL penicillin, and 100 units/mL streptomycin incubated at 37 °C and 5% CO2. MR49F and C4–2R were maintained in 10 μM and 20 μM enzalutamide solution, respectively, to maintain resistance. Enzalutamide was purchased from MedChemExpress (HY-70002). Metformin HCl, onvansertib (NMS-P937), and carbonyl cyanide-4-(trifluoromethoxy) phenylhydrazone (FCCP) were purchased from Selleckchem (S1950, S7255, S8276).

2.2. Clonogenic assay

MR49F, C4–2R, and 22Rv1 cells were seeded (3 × 103 to 6 × 103 per well) into 6-well plates with 3 mL of RPMI-1640. The following day, cells were treated with varying drugs as indicated and incubated at 37 °C. Cells were treated every other day for 10 days, then washed with ice-cold 1× PBS, fixed with ice-cold methanol for 10 minutes on ice, and stained with 0.5% crystal violet staining solution. The relative well intensity was calculated using ImageJ software.

2.3. Cell viability and synergy

MR49F, C4–2R, and 22Rv1 cells were seeded (6 × 103 per well) into 96-well plates with 100 μL of RMPI-1640. Twenty-four hours later, cells were treated with varying drugs at the indicated concentrations and allowed to incubate for 72 hours. To assess cell viability, AquaBluer solution (also known as Alamar Blue) was added to each well in a 1:100 ratio of AquaBluer solution: culture media, which monitors the reducing environment of the living cell. Cells were incubated for 4 hours at 37 °C before measuring the fluorescent intensity 540ex/590em via GloMax Discover microplate reader (Promega). Cells were seeded in quadruplicate for each drug concentration, and the readings were all normalized to average blank control wells without cells. The results are expressed as the percentage of viable cells with respect to the negative control (DMSO), which represents 100% viability shown above. Synergy scores were calculated using SynergyFinder.org.

2.4. Protein immunoblotting

Cells were previously treated with varying drug combinations for 48 hours before harvest. Cell lysis was achieved by 10% radioimmunoprecipitation assay (RIPA) solution with protease and phosphatase inhibitors followed by sonication. Protein concentration was measured by Pierce BCA Assay kit, and equal concentrations of protein lysate from each sample were mixed with SDS loading buffer, resolved on an SDS-PAGE, and transferred to either nitrocellulose or polyvinylidene difluoride membranes followed by blocking and incubation with primary and horseradish peroxidase–conjugated secondary antibodies. Electrochemiluminescence was used to induce chemiluminescence, and membranes were imaged using BioRad ChemiDoc MP. BioRad ImageLab software was used to analyze immunoblots.

2.5. Seahorse analysis

MR49F, C4–2R, and 22Rv1 cells were seeded (2 × 104 per well) into XFe96 cell culture microplates in RPMI-1640 culture medium and incubated for 24 hours. Cells were then treated with varying drug concentrations as indicated for 24 hours before analysis. Both the oxygen consumption rate (OCR) and the extracellular acidification rate were measured using the Seahorse XFe96 analyzer from Agilent. Mitochondrial stress test was performed by first measuring the initial OCR rate for cells, followed by 1 μM oligomycin, which inhibits complex V of oxidative phosphorylation (OXPHOS) (indicative of the ATP production rate). Next, 1–2 μM FCCP treatment was used to uncouple the proton gradient and determine maximum respiration (FCCP titration experiment to determine optimal FCCP dose was conducted before analysis). Following this, 1 μM of both rotenone and antimycin A were given, which inhibit complexes I and III, respectively.

2.6. Flow cytometric analysis

For measuring MMP , cells were seeded (between 5 × 105 and 1 × 106 per well) into 6-well plates with 3 mL of RPMI-1640 per sample and allowed to incubate for 24 hours at 37 °C. Cells were treated with various drugs as indicated with an incubation time of 24 hours. Forty-eight hours after initial seeding, cells were trypsinized, collected, and counted for a density of approximately 1 × 106/mL per sample. FCCP was used as a positive control (20 μM) and incubated for 15 minutes at 37°C prior to staining. All samples were then stained at 200 nM per sample with either tetramethylrhodamine ethyl ester (TMRE) reagent (Cayman chemical # 701310) or 5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraethylbenzimidazolylcarbocyanine iodide (JC-1) reagent (MedChemExpress # HY-15534) and incubated at 37 °C for 30 minutes. Cells were centrifuged at 2000 rpm for 3 minutes and resuspended in 300 μL fresh 1× PBS for analysis. Samples were analyzed using BD FACSymphony A5 Cell analyzer and FlowJo software.

2.7. RNA sequencing analysis

LNCaP, MR49F, C4–2, and C4–2R cells were previously treated with varying drug combinations for 48 hours prior to total RNA extraction. Extraction was achieved using Qiagen’s RNeasy Mini Kit (#74104) according to the manufacturer’s instructions. All samples were sent to Novogene Biotechnology Company for RNA quality assessment, library construction, Illumina sequencing, and data analysis. DEseq2 R package was used to analyze gene expression data normalization and differential expression. Significantly upregulated/downregulated genes were determined as a fold change of ≥2 and a q value of <0.05.

2.8. 22Rv1-derived xenograft mouse model

All animal experiments were approved by the Institutional Animal Care and Use Committee at the University of Kentucky. 22Rv1 cells were mixed with an equal volume of Matrigel and inoculated subcutaneously at 2.5 × 106 cells per mouse into the right flank of precastrated nude mice. After a week following inoculation, mice were randomized into 4 treatment groups. Treatments were started when the size of tumors reached 200 mm3. Enzalutamide (30 mg/kg) was dissolved in 10% DMSO and 90% corn oil, and metformin HCl (20 mg/kg) was dissolved in sterile water and administered through oral gavage daily for 4 weeks (Shao et al, 2015; Kong et al, 2020; Farah et al, 2022; Li et al, 2022; Liu et al, 2023). Tumors were measured every 3 days, and tumor volume was estimated using the following formula V = L × W2/2, where V is volume in cubic millimeters, L is length in millimeters, and W is width in millimeters.

2.9. LuCaP35CR xenograft mouse model

NOD scid gamma (NSG) mice bearing LuCaP35CR tumors were obtained by Dr Robert Vessella at the University of Washington. For tumor amplification, tumor sections were harvested and implanted subcutaneously into the flanks of precastrated NSG mice. When tumors reached a large size, tumors were harvested and sectioned into approximately 25 mm3 pieces. Tumor pieces were implanted into 40 precastrated NSG mice. Once the tumors reached approximately 200 mm3, mice were randomly assigned to 4 treatment groups. Enzalutamide (30 mg/kg) was dissolved in 10% DMSO and 90% corn oil, and metformin HCl (20 mg/kg) was dissolved in sterile water and administered through oral gavage daily for 4 weeks. Tumors were measured every 3 days and tumor volume was estimated using the following formula V = L × W2/2, where V is volume in cubic millimeters, L is length in millimeters, and W is width in millimeters.

2.10. Histology and immunohistochemistry

Xenograft tumors were fixed in 10% neutral-buffered formalin with rocking overnight and transferred into 70% ethanol the following day. Tumors were paraffin-embedded, sectioned to 5-mM sections, mounted, and processed using conventional H&E staining. Sections were also stained for the Ki67 proliferation marker and cleaved caspase 3.

2.11. Statistical analysis

Numerical data are represented as mean ± SD. The statistical significance of the results was analyzed by using an unpaired 2-tailed t test. The P values of <.05 indicate statistical significance.

3. Results

3.1. Metformin treatment exhibits a synergistic effect with enzalutamide in enzalutamide-resistant CRPC lines

To determine the optimal doses of either enzalutamide or metformin for the attenuation of PCa growth in vitro, we first used a cell viability assay and calculated IC50 values. For both isogenic lines, the enzalutamide-sensitive IC50 values of enzalutamide were lower than their enzalutamide-resistant counterparts when LNCaP and C4–2 IC50 values were 14.5 μM and 22 μM, respectively, and MR49F and C4–2R IC50 values were 26 μM and 35 μM, respectively (Fig. 1, A and C), which were consistent with our previous findings (Bai et al, 2019). The IC50 value for enzalutamide-resistant 22Rv1 cells was 110 μM (Fig. 1E). The same strategy was applied to measure the IC50 values of metformin when LNCaP was 2.1 mM, MR49F was 3.2 mM, C4–2 was 4.1 mM, C4–2R was 1.7 mM, and 22Rv1 was 15.4 mM (Fig. 1, B, D, and F). Using these doses, we sought to test whether metformin would enhance enzalutamide inhibition of cell growth using a clonogenic assay. All 3 enzalutamide-resistant (ENZ-r) lines were seeded at a low density and treated with DMSO as a control, 10 μM, 20 μM, or 30 μM of enzalutamide for MR49F, C4–2R, or 22Rv1, and 1 mM metformin alone or in combination with enzalutamide, respectively, for 14 days followed by crystal violet stain (Fig. 1, G–I). From the quantification, all 3 ENZ-r lines exhibited varying decreases in cell growth in mono treatment. However, combination treatment exhibited the greatest significant cell growth attenuation (P ≤ .001). To test whether metformin synergizes with enzalutamide to inhibit cell proliferation, we used a cell viability assay. Cells were treated with increasing combinations of enzalutamide or metformin and analyzed using the highest single-agent synergy model. We observed a strong synergistic effect in ENZ-r cells treated with a minimum of 1 mM metformin in combination with enzalutamide (Fig. 1, J–L). Together, these results suggest that enzalutamide and metformin have a synergistic effect on drug-resistant PCa growth in vitro.

Fig. 1.

Fig. 1.

Open in a new tab

Enzalutamide and metformin in combination synergistically inhibit the growth of enzalutamide-resistant CRPC in vitro. Cell viability assay of isogenic CRPC lines treated with either enzalutamide (A, C, E) or metformin (B, D, F) to compare IC50 values. Data are scaled into percentages and normalized to untreated groups, then shown as mean ± SD (n = 3). Clonogenic assay of MR49F (G), C4–2R (H), and 22Rv1 (I) treated with DMSO as control or drugs indicated for up to 14 days. Quantification of relative colony number is indicated below where *P ≤ .05; **P ≤ .01; ***P ≤ .001. Synergy scores were calculated for 22Rv1 (J), C4–2R (K), and MR49F (L) after treatment with varying doses of the indicated drugs. Scores ≤ −10 indicate an antagonistic interaction, scores between −10 and 10 indicate an additive effect, and scores ≥10 are considered synergistic.

3.2. Combination treatment results in metabolic reprogramming

Previous studies demonstrated the reliance on OXPHOS in advanced PCa (Cutruzzolà et al, 2017). We hypothesized that using metformin in combination with enzalutamide would subject the PCa cells to an energy crisis and vulnerability to apoptosis. To gain a better understanding of the effect of combination treatment on mitochondrial function, we treated ENZ-r cells followed by a mitochondrial stress test via seahorse analysis. We used the mitochondrial stress test to directly measure the OCR of cells after the injection of key modulators of cellular respiration to determine mitochondrial function (Ferrick et al, 2008; Horan et al, 2012). In MR49F cells, combination treatment lowered basal respiration, proton leak, ATP production, and spare respiratory capacity indicating an overall decline in OXPHOS (Fig. 2, A–C). Similarly, we confirmed these findings in C4–2R cells when the basal respiration, proton leak, and ATP production were decreased compared to control cells; however, C4–2R cells were markedly more sensitive to metformin treatment as the overall OCR was much lower than that for MR49F (Fig. 2, D–F). Interestingly, combination-treated C4–2R cells exhibited a higher spare respiratory capacity than metformin treatment alone indicating an increased capability of the cell to respond to energetic demand. Next, we sought to measure the MMP in response to combination treatment as an indicator of ATP production (Sukumar et al, 2016). As the MMP depolarizes, the membrane will become more permeable, allowing protons to diffuse out of the intermembrane space. Disruption of the proton gradient will inhibit ATP synthase, resulting in an overall inhibition of OXPHOS. We used TMRE, a fluorescent chemical indicating metabolic fitness, to stain ENZ-r cells following combination treatment followed by flow cytometric analysis. FCCP, an electron transport chain uncoupler, was used as a positive control for near-complete depolarization of the MMP. All samples were normalized to FCCP where TMRE-positive cells are indicative of MMP depolarization. In both MR49F and 22Rv1 cells, we did not observe any difference in MMP depolarization 24 hours after treatment (Fig. 2, G and I). Interestingly, we observed an increase in TMRE fluorescence in combination treatment compared to that for control, which indicates an overall increased metabolic fitness (Fig. 2H). These data are consistent with our results from the mitochondrial stress test, in that C4–2R cells seem to exhibit increased mitochondrial function in response to combination treatment. Finally, to confirm our results from TMRE, we employed a similar method of measuring MMP with the JC-1 chemical, which is considered more sensitive than TMRE. JC-1 differs from TMRE in that upon entrance into the mitochondria, the aggregate will emit a red color, indicative of a polarized and metabolically energetic MMP. Following depolarization of the MMP, JC-1 will manifest as monomers and diffuse out of the intermembrane space, emitting a green color. After flow cytometric analysis, we observed a similar ratio of red to green percentage cells between control, enzalutamide, and combination in MR49F cells with metformin solo treatment having the greatest effect at depolarization of the MMP (Fig. 2J). In C4–2R cells, metformin mono treatment and combination treatment remain similar to control in MMP. Due to the large error bar of the C4–2R enzalutamide mono treatment, further replicates need to be performed to generate conclusive data. 22Rv1 cells exhibited a similar effect on MMP across all treatment groups. Together, these data suggest that combination treatment may have a small effect on the metabolic fitness of the mitochondria; however, conclusive evidence is lacking at this time.

Fig. 2.

Fig. 2.

Open in a new tab

Combination treatment results in metabolic reprogramming. Representative traces of the OCR, when oligomycin, FCCP, antimycin A plus rotenone were injected into the assay XF96 plates for MR49F (A–C) and C4–2R (D–F) cells. Each data point is a mean ± SD (n = 6). Cells were treated with or without 10 μM enzalutamide, 1 mM metformin, or a combination of both for 12 hours. MMP was measured using TMRE. MR49F (G), C4–2R (H), and 22Rv1 (I) cells were treated with or without enzalutamide, metformin, or a combination of both for 12 hours and collected for flow cytometric analysis. FCCP was used as a positive control. (J) The MMP was measured with the same treatment as with TMRE in all 3 resistant lines, but the chemical JC-1 was used to visualize the membrane potential shift.

3.3. Combination treatment effect on 22Rv1-derived xenograft tumors

To investigate our findings in vitro, we evaluated the effect of enzalutamide and metformin alone or in combination with a 22Rv1-derived xenograft mouse model. 22Rv1 cells express the AR-V7 splice variant of AR, which harbors a truncated form of the ligand binding domain and prevents enzalutamide binding, making these cells intrinsically resistant to enzalutamide (Sobhani et al, 2021). Following 50 days of treatment, metformin alone exhibited a similar rate of tumor growth as in the controls, whereas enzalutamide alone and combination treatment groups had similar rates to each other (Fig. 3A). Similarly, the tumor weights for all 3 treatment groups after harvest were not significantly decreased compared to those of controls (Fig. 3B). Images from the harvested tumors confirm our results, ultimately showing no significant changes in tumor size among groups (Fig. 3C). There was no observable difference in body weight between groups, indicating a lack of treatment toxicity (Fig. 3D). Following harvest, we processed the tumors for immunohistochemistry analysis. H&E staining of tumor samples did not show any significant difference among different groups (Fig. 3E). To measure proliferation, we stained immunohistochemistry samples with proliferation marker Ki67 and observed what appears to be a general decrease in proliferation in combination-treated tumors; however, further analysis and confirmation by a pathologist would be required to make such a claim (Fig. 3F). Finally, cleaved caspase-3 staining of tumor samples indicated a similar level of apoptosis across samples, although further studies will be required to confirm these results (Fig. 3G). Interestingly, despite the observation of an inhibition of PCa cell growth in vitro, we did not observe the same effects on PCa growth in vivo following combination treatment.

Fig. 3.

Fig. 3.

Open in a new tab

Combination treatment of enzalutamide and metformin did not attenuate 22Rv1 xenograft tumor growth in vivo. (A) Tumor growth curves of 22Rv1-derived xenograft. Precastrated nude mice were inoculated subcutaneously with 22Rv1 cells (2.5 × 106 per mouse) and allowed to grow for 2 weeks. After 2 weeks, the mice were treated with various drugs as described in the Materials and methods section of this article. The sizes of the tumors in each group were measured every 3 days (mean ± SD; n = 13 mice per group). (B) Measurement of tumor weight immediately after harvest. (C) Images of 22Rv1-derived tumors at the end of the study. (D) Measurement of mouse body weight throughout the study. (E) Representative images of H&E staining on formaldehyde-fixed, paraffin-embedded, 22Rv1-derived tumor sections. (F) Representative images of anti-Ki67 immunohistochemistry staining of tumor sections. (G) Representative images of anticleaved caspase-3 immunohistochemistry staining of tumor sections.

3.4. Combination treatment effect on LuCaP35CR xenograft tumors

To further investigate the results of our in vitro work, we also employed a LuCaP35CR xenograft model, which is more closely related to patient samples, to determine the effect of combination treatment on tumor growth. Consistent with our previous results in the 22Rv1-derived xenograft experiment, LuCaP35CR did not exhibit any significant changes in tumor volume within the 4 treatment groups 50 days after the initial treatment (Fig. 4A). Immediately upon harvest, tumors were weighed and exhibited no significant changes among treatment group tumors (Fig. 4B) or between tumor size indicated by the tumor images in Fig. 4C. To determine toxicity, we measured body weight in the 4 groups throughout the study, and although there may be an observable difference between treatment groups and controls, this could be due to the small sample size for all groups (n = 5) (Fig. 4D). Taken together, our in vivo results indicate a lack of synergistic effect in vivo, in contrast to the phenotype we observed in vitro.

Fig. 4.

Fig. 4.

Open in a new tab

Combination treatment of enzalutamide and metformin did not attenuate LuCaP35CR tumor growth in vivo. (A) Tumor growth curves of LuCaP35CR xenografts (mean ± SD; n = 4 mice per group). (B) Measurement of tumor weight immediately after harvest. (C) Images of LuCaP35CR tumors at the end of the study. (D) Measurement of mouse body weight throughout the study.

3.5. RNA sequencing analysis of isogenic ENZ-r CRPC lines

To determine the mechanism by which CRPC lines respond to combination treatment, we performed RNA sequencing analysis with the isogenic sensitive and ENZ-r lines listed previously. After enzalutamide or metformin mono treatment or in combination, we compared gene lists to determine differences in RNA expression for genes that were specific to ENZ-r combination-treated samples (Fig. 5A). C4–2R combination-treated gene sets exhibit a significant decrease in genes in the Ras signaling pathway as well as phospholipase D signaling and genes related to the lipid and atherosclerosis pathway (Fig. 5B). Based on these results, we can speculate that a downregulation in Ras signaling specifically would inhibit cell growth and proliferation (Weber and Gioeli, 2004). In addition, phospholipase D signaling as well as lipid and atherosclerosis signaling play roles in cellular metabolism as well as cross-signaling with traditional oncogenic signaling pathways such as Ras, mTOR, and MAPK signaling, and we observe their downregulation in our samples (Weber and Gioeli, 2004; Rodríguez-Berriguete et al, 2012; Edlind and Hsieh, 2014; Shorning et al, 2020). We observed similar results in the significant downregulation of genes associated with MAPK signaling, lipid, and atherosclerosis pathways, and calcium signaling. As with C4–2R cells, these downregulated pathways foreshadow a shift in gene expression toward the inhibition of cell proliferation and cellular metabolism pathways (Fig. 5D). In contrast, we observed an upregulation of both cell cycle signaling proteins as well as proteins related to various DNA repair pathways in both C4–2R (Fig. 5C) and MR49F (Fig. 5E) following combination treatment. It is currently unclear how the upregulation of these genes may influence these ENZ-r PCa cells; further analysis is required to investigate these results.

Fig. 5.

Fig. 5.

Open in a new tab

RNA sequencing analysis of isogenic CRPC lines. (A) Schematic representation of gene comparisons for RNA sequencing result analysis. Dot plot analysis of significant C4–2R downregulated (B), upregulated (C), MR49F downregulated (D), and upregulated (E) pathways. Methods: The RNA-seq differential expression (DE) analysis involved the following comparisons and cell lines: (1) metformin versus controls in the C4–2 cell line; (2) metformin versus controls in the C4–2R cell line; and (3) Combo (combination) versus controls in the C4–2R cell line. Additionally, 3 more comparisons were performed in parallel with the above 3: (4) metformin versus controls in the LNCaP cell line; (5) metformin versus controls in the MR49F cell line; and (6) combo versus controls in the MR49F cell line. In this setup, the LNCaP cell line serves as a parallel to the C4–2 cell line, and the MR49F cell line serves as a parallel to the C4–2R cell line. For each comparison, RNA-seq DE analysis was conducted using the R package “edgeR,” with the control group chosen as the reference. In comparisons 1–3 and 4–6, respectively, significantly upregulated DE genes were identified by a log2(fold change) > 0.5 and a q-value < 0.05, whereas significantly downregulated DE genes were identified by a log2(fold change) < −0.5 and a q-value < 0.05. We then isolated the significant DE genes that were found in the combo versus control comparison but not in the metformin versus control comparison, identifying genes that responded exclusively to the combo treatment. Using this subset of upregulated or downregulated DE genes, we performed KEGG pathway enrichment analysis with the R package “clusterProfiler,” which uses Fisher’s exact test. The output of the enrichment analysis reveals the KEGG pathways that are enriched specifically in response to the combo treatment. Figure 5 illustrates the top 20 upregulated and downregulated KEGG pathways with the smallest P values for the cell lines C4–2/C4–2R and LNCaP/MR49F, respectively.

4. Discussion

Although treatments and therapies continue to be developed for various cancers at different stages, drug resistance remains a serious issue in advanced CRPC, and identifying novel treatment strategies is critical (Seruga et al, 2011; Amaral et al, 2012). Enzalutamide, as a competitive inhibitor of AR signaling, continues to be the only FDA-approved therapy for metastatic CRPC; however, resistance to treatment often occurs. In this study, we assessed whether a combination treatment of enzalutamide and metformin in enzalutamide-resistant PCa lines would induce an energy crisis and therefore induce vulnerability to apoptosis (Wang et al, 2021). Our results demonstrate that the combination treatment is synergistically compatible to inhibit drug-resistant PCa growth in vitro. Cell proliferation in ENZ-r was significantly inhibited following combination treatment (Fig. 1, G–I), and our results indicate that the 2 drugs tested act synergistically together using the highest single agent synergy model (Fig. 1, J–L) (Berenbaum, 1989). These results suggest a vulnerability in the metabolic signaling of ENZ-r PCa cells, which may have allowed for exploitation and ultimately cell death with enzalutamide treatment.

Although our combination treatment exhibited a similar growth inhibition phenotype across all 3 ENZ-r lines, we observed differences in mitochondrial function between these lines. It has been well documented that the mitochondrial stress test is a robust method for testing mitochondrial function (Ferrick et al, 2008). After combination treatment, MR49F cells exhibited more of a lack of mitochondrial function indicated by an overall lower basal consumption rate, lower spare respiratory capacity, and decreased ATP production (Fig. 2, A–C). In contrast, combination-treated C4–2R cells responded similarly to metformin mono treatment with a decreased basal respiration rate and ATP production; however, combination treatment may have a better capacity to respond to metabolic stress indicated by a higher respiratory capacity (Fig. 2, D–F). The observed difference in mitochondrial respiration between these 2 cell lines may be a result of the differences in metabolic gene expression. In addition, using an acute injection mitochondrial stress test and measuring the changes in OCR immediately after treatment may yield interesting changes in respiration because this experiment would capture the immediate responses to treatment. The MMP is maintained by the electron transport chain as a means of producing a proton gradient for ATP synthase to function; therefore, depolarization of the MMP is indicative of OXPHOS inhibition (Sukumar et al, 2016). Although we tested 2 different means in which the MMP can be measured for all 3 ENZ-r lines, we did not observe a significant difference in polarization among treatment groups compared to that for controls (Fig. 2, G–J). Although the MMP is indicative of OXPHOS inhibition, the MMP can stabilize quickly following challenge and may be best observed in an acute treatment experiment. To better observe the metabolic shift from OXPHOS to glycolysis, the glycolytic rate assay could be employed as a rigorous method in which rapid metabolic switches can be detected (Mookerjee and Brand, 2015). In addition, determining mitochondrial mass after treatment may be another method to measure the mitochondrial response of either fission or fusion (Westermann, 2012).

Although our in vitro results demonstrate a synergistic effect on PCa growth, the same effect was not observed in either the 22Rv1-derived xenograft model (Fig. 3) or the LuCaP35-CR xenograft model (Fig. 4). One potential reason for the significant difference between 22Rv1-derived xenograft response to combination treatment and the observable phenotype in vitro is the difference in the metabolic profile for 22Rv1 cells. As we observed with the MMP, 22Rv1 cells did not exhibit a difference among treatment groups, indicating that the ATP synthase remained functional due to the stable polarization of the mitochondrial membrane (Fig. 2I). 22Rv1 cells are typically used as the standard xenograft model in testing drug-resistant CRPC because they are intrinsically resistant to enzalutamide and account for the AR-V7 (Sarwar et al, 2016; Kregel et al, 2020); however, we observed that C4–2R and MR49F cells were more sensitive to changes in metabolism than 22Rv1 cells (Fig. 2, A–H). Another potential explanation for the difference in responses to combination treatment between in vitro and in vivo models could be the route of administration and treatment. In our study, we used an oral gavage technique for treatments at the concentrations listed in the methods section; however, utilization of an intraperitoneal technique might have yielded better results in the mice because this is a method of direct administration (Turner et al, 2011). In addition to changes in administration, metformin may be more sensitive to freeze/thaw than we anticipated. Future treatments with metformin in vivo may require dissolving smaller doses for treatment to avoid freeze/thawing effects. We are aware of a separate report on the combination of metformin and enzalutamide in PCa (Liu et al, 2017). In that report, the authors stated that metformin is capable of reversing enzalutamide resistance and restoring the sensitivity of 22RV1 xenografts to enzalutamide (Liu et al, 2017). Accordingly, we carefully compared the experimental conditions of the 2 studies. In our study, we used 10 μM enzalutamide and 1 mM metformin for cell culture experiments. In the previous work, the authors used 20 μM enzalutamide and 5 mM metformin for in vitro experiments. Thus, the metformin concentration in the previous study was 5 times higher than in our study. In vivo, we used 30 mg/kg enzalutamide and 20 mg/kg metformin, whereas the previous report used 25 mg/kg enzalutamide and 300 mg/kg metformin. The metformin dose used in the previous in vivo study was 15 times higher than in our study. Because of the very high concentration of metformin used, metformin alone was sufficient to completely inhibit tumor growth, and the combination of metformin and enzalutamide did not show a statistically significant difference compared to metformin alone (see Fig. 1e of previous study) (Liu et al, 2017). In conclusion, the 15-fold higher concentration of metformin used in the previous in vivo study is likely the main reason for the apparent discrepancy.

We acknowledge that multiple factors may affect the outcome of in vivo experiments. For example, we started the treatment when the tumors reached a size of 200 mm3, which may have been too late for optimal treatment. Early intervention could yield different results. Additionally, drug doses (as discussed earlier), treatment time points, and duration may also impact the outcomes of our in vivo experiments. However, we have carefully followed the well established lab protocol, as previously published (Liu et al, 2023). Furthermore, a bone metastatic model might be a better choice for testing the effectiveness of the combination, especially given that the osteoclast differentiation genes were shown to be downregulated (Fig. 5C). Future experiments will be designed to directly test this possibility.

RNA sequencing analysis identified an upregulation of both cell cycle and DNA repair pathways following metformin and enzalutamide treatment in enzalutamide-resistant cells (Fig. 5, C and E). These findings suggest potential new treatment strategies, because they indicate that enzalutamide-resistant PCa cells may rely on the activation of cell cycle or DNA repair pathways to survive metformin treatment. Cell cycle regulators such as polo-like kinase 1 and cyclin-dependent kinases (CDKs), as well as DNA repair regulators like ATR, play well-established roles in cancer progression (Kase et al, 2020). For instance, the CDK4 inhibitor palbociclib (Ibrance), which is approved for cancer treatment, is primarily used for hormone receptor–positive, HER2-negative breast cancer. Other CDK4/6 inhibitors, such as ribociclib and abemaciclib, are used in similar contexts. Several DNA damage repair inhibitors have also been approved for cancer treatment. Notable examples of poly (ADP-ribose) polymerase (PARP) inhibitors include olaparib (Lynparza), rucaparib (Rubraca), niraparib (Zejula), and talazoparib (Talzenna) (Taylor et al, 2023). Although palbociclib is primarily used in breast cancer, it is currently being investigated for PCa, particularly in combination with other therapies. Further experiments are needed to determine whether metformin can enhance the efficacy of palbociclib in treating CRPC (Kase et al, 2020). Regarding DNA damage repair inhibitors, olaparib has been approved for use in PCa with BRCA1 or BRCA2 mutations and is used in patients with metastatic CRPC who have been previously treated with other therapies (Taylor et al, 2023). It would be of great interest to test whether combining metformin with olaparib could enhance efficacy in CRPC patients. Additionally, rucaparib, another poly (ADP-ribose) polymerase inhibitor being explored for PCa, is particularly relevant for those with BRCA mutations. Several inhibitors targeting ATM and ATR are currently in clinical trials for PCa, investigating their efficacy either alone or in combination with other therapies. In the future, we will explore whether coadministration of metformin can enhance the efficacy of these various inhibitors in treating CRPC that no longer responds to enzalutamide.

In summary, the present study demonstrates the difficulty in treating drug-resistant CRPC because the combination treatment of enzalutamide and metformin in vitro demonstrated a positive attenuation of PCa growth; however, this effect was not observed in vivo. Our results highlight the importance of investigating different treatments in robust in vitro and in vivo models. Despite drug-resistant CRPC’s reliance on OXPHOS in energy metabolism, inhibition of OXPHOS with metformin did not produce an observable phenotype on ENZ-r lines.

Significance Statement:

Increasing evidence suggests that oxidative phosphorylation might play a critical role in the development of resistance to cancer therapy. This study showed that targeting oxidative phosphorylation with metformin can enhance the efficacy of enzalutamide in castration-resistant prostate cancer in vitro.

Acknowledgments

We would like to express our sincere gratitude to Kandy Zhang, Yanquan Zhang, Jinghui Liu, and the other members of the Liu lab for their invaluable guidance and assistance in completing this project. The study was also supported by the Biospecimen Procurement & Translational Pathology, Biostatistics and Bioinformatic Shared Resources of the University of Kentucky Markey Cancer Center.

Financial support

This work was supported by the National Institutes of Health National Cancer Institute [R01 CA256893, R01 CA264652, R01 CA157429, R01 CA272483, P30 CA177558].

Abbreviations

AR

androgen receptor

CDK

cyclin-dependent kinase

CRPC

castration-resistant prostate cancer

ENZ-r

enzalutamide-resistant

FCCP

carbonyl cyanide-4-(trifluoromethoxy) phenylhydrazone

FDA

US Food and Drug Administration

JC-1

5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraethylbenzimidazolylcarbocyanine iodide

MMP

mitochondrial membrane potential

NSG

NOD scid gamma

OCR

oxygen consumption rate

OXPHOS

oxidative phosphorylation

PCa

prostate cancer

TMRE

tetramethylrhodamine ethyl ester

Footnotes

Conflict of interest

No author has an actual or perceived conflict of interest with the content of this article.

Data availability

All data/datasets are contained in the paper.

References

  1. Amaral TM, Macedo D, Fernandes I, and Costa L (2012) Castration-resistant prostate cancer: mechanisms, targets, and treatment. Prostate Cancer 2012:327253. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Antonarakis ES, Lu C, Wang H, Luber B, Nakazawa M, Roeser JC, Chen Y, Mohammad TA, Chen Y, and Fedor HL, et al. (2014) AR-V7 and resistance to enzalutamide and abiraterone in prostate cancer. N Engl J Med 371:1028–1038. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bai Y, Zhang Z, Cheng L, Wang R, Chen X, Kong Y, Feng F, Ahmad N, Li L, and Liu X (2019) Inhibition of enhancer of zeste homolog 2 (EZH2) overcomes enzalutamide resistance in castration-resistant prostate cancer. J Biol Chem 294:9911–9923. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Ben Sahra I, Laurent K, Loubat A, Giorgetti-Peraldi S, Colosetti P, Auberger P, Tanti JF, Le Marchand-Brustel Y, and Bost F (2008) The antidiabetic drug metformin exerts an antitumoral effect in vitro and in vivo through a decrease of cyclin D1 level. Oncogene 27:3576–3586. [DOI] [PubMed] [Google Scholar]
  5. Berenbaum MC (1989) What is synergy? Pharmacol Rev 41:93–141. [PubMed] [Google Scholar]
  6. Chandrasekar T, Yang JC, Gao AC, and Evans CP (2015) Mechanisms of resistance in castration-resistant prostate cancer (CRPC). Transl Androl Urol 4:365–380. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Chen L, Ahmad N, and Liu X (2016) Combining p53 stabilizers with metformin induces synergistic apoptosis through regulation of energy metabolism in castration-resistant prostate cancer. Cell Cycle 15:840–849. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cutruzzolà F, Giardina G, Marani M, Macone A, Paiardini A, Rinaldo S, and Paone A (2017) Glucose metabolism in the progression of prostate cancer. Front Physiol 8:97. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Edlind MP and Hsieh AC (2014) PI3K-AKT-mTOR signaling in prostate cancer progression and androgen deprivation therapy resistance. Asian J Androl 16:378–386. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Efstathiou E, Titus M, Wen S, Hoang A, Karlou M, Ashe R, Tu SM, Aparicio A, Troncoso P, and Mohler J, et al. (2015) Molecular characterization of enzalutamide-treated bone metastatic castration-resistant prostate cancer. Eur Urol 67:53–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Farah E, Zhang Z, Utturkar SM, Liu J, Ratliff TL, and Liu X (2022) Targeting DNMTs to overcome enzalutamide resistance in prostate cancer. Mol Cancer Ther 21:193–205. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Ferrick DA, Neilson A, and Beeson C (2008) Advances in measuring cellular bioenergetics using extracellular flux. Drug Discov Today 13:268–274. [DOI] [PubMed] [Google Scholar]
  13. Horan MP, Pichaud N, and Ballard JW (2012) Review: quantifying mitochondrial dysfunctionin complex diseases of aging. J Gerontol A Biol Sci MedSci 67:1022–1035. [DOI] [PubMed] [Google Scholar]
  14. Kase AM, Copland Iii JA, and Tan W (2020) Novel therapeutic strategies for CDK4/6 inhibitors in metastatic castrate-resistant prostate cancer. Onco Targets Ther 13:10499–10513. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Kasznicki J, Sliwinska A, and Drzewoski J (2014) Metformin in cancer prevention and therapy. Ann Transl Med 2:57. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Kong Y, Zhang Y, Mao F, Zhang Z, Li Z, Wang R, Liu J, and Liu X (2020) Inhibition of EZH2 enhances the antitumor efficacy of metformin in prostate cancer. Mol Cancer Ther 19:2490–2501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Kregel S, Wang C, Han X, Xiao L, Fernandez-Salas E, Bawa P, McCollum BL, Wilder-Romans K, Apel IJ, and Cao X, et al. (2020) Androgen receptor degraders overcome common resistance mechanisms developed during prostate cancer treatment. Neoplasia 22:111–119. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Li C, Liu J, He D, Mao F, Rao X, Zhao Y, Lanman NA, Kazemian M, Farah E, and Liu J, et al. (2022) GSTM2 is a key molecular determinant of resistance to SG-ARIs. Oncogene 41:4498–4511. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Liu J, Zhao Y, He D, Jones KM, Tang S, Allison DB, Zhang Y, Chen J, Zhang Q, and Wang X, et al. (2023) A kinome-wide CRISPR screen identifies CK1alpha as a target to overcome enzalutamide resistance of prostate cancer. Cell Rep Med 4:101015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Liu Q, Tong D, Liu G, Xu J, Do K, Geary K, Zhang D, Zhang J, Zhang Y, and Li Y, et al. (2017) Metformin reverses prostate cancer resistance to enzalutamide by targeting TGF-beta1/STAT3 axis-regulated EMT. Cell Death Dis 8:e3007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Mookerjee SA and Brand MD (2015) Measurement and analysis of extracellular acid production to determine glycolytic rate. J Vis Exp 106:e53464. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Murtola TJ, Tammela TL, Lahtela J, and Auvinen A (2008) Antidiabetic medication and prostate cancer risk: a population-based case-control study. Am J Epidemiol 168:925–931. [DOI] [PubMed] [Google Scholar]
  23. Rodríguez-Berriguete G, Fraile B, Martinez-Onsurbe P, Olmedilla G, Paniagua R, and Royuela M (2012) MAP kinases and prostate cancer. J Signal Transduct 2012: 169170. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Rothermundt C, Hayoz S, Templeton AJ, Winterhalder R, Strebel RT, Bärtschi D, Pollak M, Lui L, Endt K, and Schiess R, et al. (2014) Metformin in chemotherapynaive castration-resistant prostate cancer: a multicenter phase 2 trial (SAKK 08/09). Eur Urol 66:468–474. [DOI] [PubMed] [Google Scholar]
  25. Sarwar M, Semenas J, Miftakhova R, Simoulis A, Robinson B, Gjorloff Wingren A, Mongan NP, Heery DM, Johnsson H, and Abrahamsson PA, et al. (2016) Targeted suppression of AR-V7 using PIP5K1alpha inhibitor overcomes enzalutamide resistance in prostate cancer cells. Oncotarget 7:63065–63081. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Seruga B, Ocana A, and Tannock IF (2011) Drug resistance in metastatic castration-resistant prostate cancer. Nat Rev Clin Oncol 8:12–23. [DOI] [PubMed] [Google Scholar]
  27. Shao C, Ahmad N, Hodges K, Kuang S, Ratliff T, and Liu X (2015) Inhibition of polo-like kinase 1 (Plk1) enhances the antineoplastic activity of metformin in prostate cancer. J Biol Chem 290:2024–2033. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Sharifi N, Gulley JL, and Dahut WL (2005) Androgen deprivation therapy for prostate cancer. JAMA 294:238–244. [DOI] [PubMed] [Google Scholar]
  29. Shoag JE, Nyame YA, Gulati R, Etzioni R, and Hu JC (2020) Reconsidering the trade-offs of prostate cancer screening. N Engl J Med 382:2465–2468. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Shorning BY, Dass MS, Smalley MJ, and Pearson HB (2020) The PI3K-AKT-mTOR pathway and prostate cancer: at the crossroads of AR, MAPK, and WNT signaling. Int J Mol Sci 21:4507. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Siegel RL, Giaquinto AN, and Jemal A (2024) Cancer statistics, 2024. CA Cancer J Clin 74:12–49. [DOI] [PubMed] [Google Scholar]
  32. Sleire L, Førde HE, Netland IA, Leiss L, Skeie BS, and Enger PO (2017) Drug repurposing in cancer. Pharmacol Res 124:74–91. [DOI] [PubMed] [Google Scholar]
  33. Sobhani N, Neeli PK, D’Angelo A, Pittacolo M, Sirico M, Galli IC, Roviello G, and Nesi G (2021) AR-V7 in metastatic prostate cancer: a strategy beyond redemption. Int J Mol Sci 22:5515. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Spratt DE, Zhang C, Zumsteg ZS, Pei X, Zhang Z, and Zelefsky MJ (2013) Metformin and prostate cancer: reduced development of castration-resistant disease and prostate cancer mortality. Eur Urol 63:709–716. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Sukumar M, Liu J, Mehta GU, Patel SJ, Roychoudhuri R, Crompton JG, Klebanoff CA, Ji Y, Li P, and Yu Z, et al. (2016) Mitochondrial membrane potential identifies cells with enhanced stemness for cellular therapy. Cell Metab 23:63–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Taylor AK, Kosoff D, Emamekhoo H, Lang JM, and Kyriakopoulos CE (2023) PARP inhibitors in metastatic prostate cancer. Front Oncol 13:1159557. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Turner PV, Brabb T, Pekow C, and Vasbinder MA (2011) Administration of substances to laboratory animals: routes of administration and factors to consider. J Am Assoc Lab Anim Sci 50:600–613. [PMC free article] [PubMed] [Google Scholar]
  38. van Soest RJ, de Morrée ES, Kweldam CF, de Ridder CMA, Wiemer EAC, Mathijssen RHJ, de Wit R, and van Weerden WM (2015) Targeting the androgen receptor confers in vivo cross-resistance between enzalutamide and docetaxel, but not cabazitaxel, in castration-resistant prostate cancer. Eur Urol 67:981–985. [DOI] [PubMed] [Google Scholar]
  39. Wang Y, Chen J, Wu Z, Ding W, Gao S, Gao Y, and Xu C (2021) Mechanisms of enzalutamide resistance in castration-resistant prostate cancer and therapeutic strategies to overcome it. Br J Pharmacol 178:239–261. [DOI] [PubMed] [Google Scholar]
  40. Weber MJ and Gioeli D (2004) Ras signaling in prostate cancer progression. J Cell Biochem 91:13–25. [DOI] [PubMed] [Google Scholar]
  41. Westermann B (2012) Bioenergetic role of mitochondrial fusion and fission. Biochim Biophys Acta 1817:1833–1838. [DOI] [PubMed] [Google Scholar]
  42. Zhao Y, Zeng X, Tang H, Ye D, and Liu J (2019) Combination of metformin and paclitaxel suppresses proliferation and induces apoptosis of human prostate cancer cells via oxidative stress and targeting the mitochondria-dependent pathway. Oncol Lett 17:4277–4284. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

All data/datasets are contained in the paper.

RESOURCES