Bromocriptine monotherapy overcomes prostate cancer chemoresistance in preclinical models

Highlights

• •Chemoresistance is a major obstacle in the management of advanced prostate cancer.
• •Bromocriptine has high specificity and potency in chemoresistant prostate cancer cells.
• •Bromocriptine may affect DRD2-dependent and -independent mechanisms.
• •Bromocriptine inhibits the skeletal growth of chemoresistant prostate cancer cells.
• •Bromocriptine could be repurposed to overcome prostate cancer chemoresistance.

Abstract

Chemoresistance is a major obstacle in the clinical management of metastatic, castration-resistant prostate cancer (PCa). It is imperative to develop novel strategies to overcome chemoresistance and improve clinical outcomes in patients who have failed chemotherapy. Using a two-tier phenotypic screening platform, we identified bromocriptine mesylate as a potent and selective inhibitor of chemoresistant PCa cells. Bromocriptine effectively induced cell cycle arrest and activated apoptosis in chemoresistant PCa cells but not in chemoresponsive PCa cells. RNA-seq analyses revealed that bromocriptine affected a subset of genes implicated in the regulation of the cell cycle, DNA repair, and cell death. Interestingly, approximately one-third (50/157) of the differentially expressed genes affected by bromocriptine overlapped with known p53-p21- retinoblastoma protein (RB) target genes. At the protein level, bromocriptine increased the expression of dopamine D2 receptor (DRD2) and affected several classical and non-classical dopamine receptor signal pathways in chemoresistant PCa cells, including adenosine monophosphate-activated protein kinase (AMPK), p38 mitogen-activated protein kinase (p38 MAPK), nuclear factor kappa B  (NF-κB), enhancer of zeste homolog 2 (EZH2), and survivin. As a monotherapy, bromocriptine treatment at 15 mg/kg, three times per week, via the intraperitoneal route significantly inhibited the skeletal growth of chemoresistant C4-2B-TaxR xenografts in athymic nude mice. In summary, these results provided the first preclinical evidence that bromocriptine is a selective and effective inhibitor of chemoresistant PCa. Due to its favorable clinical safety profiles, bromocriptine could be rapidly tested in PCa patients and repurposed as a novel subtype-specific treatment to overcome chemoresistance.

Introduction

Prostate cancer (PCa) is the most common type of cancer and the second leading cause of cancer-related death in American men. In 2023, an estimated 288,300 new cases will be diagnosed, and 34,700 patients will die, largely from metastatic castration-resistant PCa (mCRPC) [1]. Despite numerous therapeutic modalities, including radiation therapy (such as radium-223, lutetium-177–PSMA-617), chemotherapy (docetaxel, cabazitaxel), new-generation androgen deprivation therapy (ADT) (enzalutamide, abiraterone, apalutamide), and poly (adenosine diphosphate-ribose) polymerase (PARP) inhibitors (such as olaparib), mCRPC remains a dreadful progression and bears a poor prognosis for most patients who develop bone metastases [2,3]. Additionally, skeletal-related events from bone metastases can lead to severe pain, increased risk of death, increased healthcare costs, and reduced quality of life [4].

Docetaxel was approved in 2004 as the first-line chemotherapy to treat mCRPC [5,6]. As a semisynthetic taxane, docetaxel exhibits significant single-agent activity against PCa, mainly by inhibiting microtubules polymerization and inducing cell cycle arrest at the G2/M phase and subsequent apoptosis [7]. In recent trials, the combined use of docetaxel and ADT significantly improved the overall survival of patients with metastatic, hormone-sensitive PCa, particularly with high tumor volumes [8,9]. These promising data indicated that docetaxel would remain a mainstay for the clinical management of PCa [10]. Unfortunately, patients eventually develop docetaxel resistance, and there is currently no cure for PCa. It is imperative to develop novel therapeutic strategies to overcome chemoresistance [11].

Bromocriptine (2-Bromo-alpha-ergocryptine) is a semisynthetic ergot alkaloid derivative in the dopamine D2 agonist class of drugs [12,13]. Bromocriptine was patented in 1968 and approved in 1975 [14]. This drug has been used to treat multiple diseases and complications, including hyperprolactinemia-associated conditions, Parkinson's disease, acromegaly, prolactinomas, and other pituitary hormone-dependent adenomas. In 2009, bromocriptine was approved as the first dopaminergic medication to improve glycemic control in patients with type 2 diabetes. Bromocriptine is considered a safe and well-tolerated drug, even on long-term use [15]. Bromocriptine is usually administered via the oral route and experiences a significant first-pass effect, with a half-life of 2–8 h in circulation. The central effects indicated that bromocriptine could penetrate most areas of the brain and remain active. Bromocriptine is metabolized extensively, mainly by cytochrome P450 3A4 in the liver. Excretion of bromocriptine metabolites is mainly biliary [12,16].

Multiple lines of preclinical evidence suggested that aberrant dopamine signaling may play an important role in cancer progression [17]. Due to its excellent safety and tolerability in various disease conditions, bromocriptine has been pursued as an attractive drug candidate for cancer treatment [18]. Several small-size trials were conducted between the 1970s and 2000s to test the benefits of bromocriptine in patients with PCa and breast cancer. The drug was well tolerated in most trials, and bromocriptine treatment was associated with reduced prolactin levels. Although bromocriptine as monotherapy did not elicit significant clinical responses [19], [20], [21], [22], [23], [24], its combination with docetaxel increased the rates of partial response and stable disease compared with the docetaxel-only regimen in metastatic breast cancer pretreated with anthracyclines [25]. These observations are consistent with preclinical studies from our laboratory and others that bromocriptine had weak to moderate anticancer activities but could sensitize cancer cells to chemotherapeutics (such as docetaxel) [18,[26], [27], [28], [29]], suggesting bromocriptine could be used as an adjunct agent to treat chemoresistant cancers.

Recently we established a high-throughput phenotypic screen that closely recapitulates the molecular characteristics and mechanisms of chemoresistant PCa [30]. Using this platform, we identified bromocriptine as a potent and selective inhibitor of chemoresistant PCa cells. We further investigated the mechanism of action of bromocriptine and evaluated the in vivo efficacy against mCRPC in experimental models. Here, we provide the first preclinical evidence supporting bromocriptine as a safe and effective monotherapy for chemoresistant PCa.

Materials and methods

Cell lines, chemicals, and reagents

ARCaP_E cells stably expressing short hairpin RNA (shRNA) against human epithelial protein lost in neoplasm (EPLIN) (ARCaP_E-shEPLIN) or control shRNA (ARCaP_E-shCtrl) were established and maintained as we have previously described [31]. The C4-2B parental cell line and its docetaxel-resistant subline C4-2B-TaxR cells were kindly provided by Dr. Allen C. Gao (University of California Davis, CA, USA), and maintained following the published procedures [32]. Docetaxel (LC Laboratories, Woburn, MA, USA) was used at 100 nM to maintain the chemoresistant feature of C4-2B-TaxR cells and was reduced to 5 nM in experimental assays.

Bromocriptine mesylate was purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). Dimethyl sulfoxide (DMSO) and propidium iodide (PI) were purchased from Sigma-Aldrich (St. Louis, MO, USA). RPMI1640 medium was purchased from Thermo Fisher Scientific (Waltham, MA, USA). Fetal bovine serum (FBS) was purchased from Atlanta Biologicals (Atlanta, GA, USA). Penicillin-streptomycin solution was purchased from Corning (Corning, NY, USA).

In vitro viability assay

Cell Counting Kit-8 (CCK-8; Dojindo Molecular Technologies, Inc., Rockville, MD, USA) was used to evaluate the in vitro cytotoxicity of bromocriptine and docetaxel following the manufacturer's instructions. SigmaPlot program (Systat Software Inc., San Jose, CA, USA) was used to calculate the half maximal inhibitory concentration (IC₅₀).

Cell cycle and apoptosis analyses

C4-2B-TaxR cells treated with vehicle control or different concentrations of bromocriptine for 24 h were stained with PI (50 µg/mL) at 37 °C for 2 h (h) for cell cycle assay, or stained by a PE Annexin V Apoptosis Detection Kit (BioLegend, San Diego, CA, USA) for 72 h for apoptosis assay, following the manufacturers’ instructions. Apoptotic cells or different cell cycle phases were evaluated by flow cytometry with a BD Accuri C6 Flow Cytometer and software (BD Biosciences, Bedford, MA, USA).

Transfection and reporter assay

Human survivin promotor reporters (pSVV-Luc1430 and pSVV-Luc230; kindly provided by Dr. Allen C. Gao) were transfected with pRL-TK (internal control; Promega, Madison, WI, USA) using Lipofectamine 2000 reagent, following the manufacturer's instructions (Life Technologies, Carlsbad, CA, USA). Luciferase activities were evaluated using a Dual-Luciferase reporter assay system (Promega). Relative luciferase activity was calculated by normalizing firefly luciferase intensity against Renilla luciferase intensity.

Western blot analysis

Total cell lysates were extracted using radioimmunoprecipitation (RIPA) buffer (Santa Cruz Biotechnology) supplemented with Halt™ protease inhibitor cocktail (Thermo Fisher Scientific). Western blot analyses followed the procedure we described previously [33]. The antibodies used are listed in Table S1. All Western blot analyses were repeated at least three times.

RNA-seq analysis

RNA samples were collected from C4-2B-TaxR cells treated with vehicle control or bromocriptine (1.3 μM) for 24 h in triplicates. RNA-seq analyses were performed by Omega Bioservices (Norcross, GA). Data were analyzed by the Rosalind® platform (Rosalind, Inc., San Diego, CA), Ingenuity Pathway Analysis (IPA, Qiagen, Germantown, MD), and Gene Set Enrichment Analysis (GSEA, University of California San Diego and Broad Institute).

In vivo intratibial xenografts

All animal experiments were in compliance with the National Institutes of Health guidelines and approved by Augusta University Institutional Animal Care and Use Committee (IACUC). Male athymic nude mice were purchased from Envigo RMS, Inc (Indianapolis, IN, USA). A total of 2.0  ×  10⁶ C4-2B-TaxR cells were injected into the bilateral tibia of each mouse (5 weeks old) following the procedures we described previously [33]. Following the confirmation of tumor formation as indicated by rising human prostate-specific antigen (PSA) in mouse sera (≥ 0.5 ng/ml) using an ELISA kit purchased from United Biotech, Inc (Mountain View, CA, USA), tumor-bearing mice were randomly divided into two groups and treated with vehicle (DMSO; n = 6) or bromocriptine (15 mg/kg, three times per week; n = 6) through the intraperitoneal (i.p) route. The vehicle control group of mice was the same as in a previous study [34]. Mice were weighed twice a week and serum PSA levels were detected once a week to monitor the tumor growth in mice. At the endpoint, X-ray radiography was performed using the MX-20 System (Faxitron, Tucson, Arizona, USA), mice were sacrificed by the standard procedure, major organs were examined for abnormalities, and tumor-bearing tibias were collected.

Histopathology

Haemotoxylin and eosin (H&E) staining and immunohistochemistry (IHC) using an anti-Ki67 antibody (Santa Cruz Biotechnology) in xenograft tissues were performed following standard procedures.

Statistical analysis

Unpaired t-test was used to evaluate the significance of differences between any two groups in all in vitro tests representing three or more experiments, and two-way ANOVA analysis was used to evaluate the significance of the overall difference in animals during the entire study period between the vehicle control and the bromocriptine treatment groups. Error bars represent standard error (SE) values of averaged results, and p < 0.05 represents statistical significance. GraphPad Prism program (GraphPad Software Inc., La Jolla, CA) was used to perform the statistical analyses.

Results

Bromocriptine exhibits high specificity and potency in chemoresistant PCa cells

Recently, we reported a two-tier phenotypic screen for discovering novel small-molecule inhibitors of chemoresistant PCa [30]. The platform consisted of two pairs of PCa cell derivatives. The first pair were ARCaP_E-shEPLIN and ARCaP_E-shCtrl [[31], [35]]. Compared with ARCaP_E-shCtrl cells, ARCaP_E-shEPLIN cells were resistant to chemotherapeutics, including docetaxel, thus representing the characteristics of the intrinsic chemoresistance [[31], [35]]. Potential hits selectively inhibiting ARCaP_E-shEPLIN but not ARCaP_E-shCtrl cells were further tested for their differential cytotoxicity in the second pair of PCa cells, i.e., docetaxel-resistant C4-2B-TaxR cells and chemoresponsive C4-2B parental cells [32]. Only the compounds demonstrating selective cytotoxicity in C4-2B-TaxR cells were investigated for the mechanism of action and in vivo anticancer efficacy.

Using this phenotypic screening platform, we identified bromocriptine as a potent and selective inhibitor of chemoresistant PCa cells (Fig. 1A). The half-minimal inhibitory concentration (IC₅₀) of bromocriptine was determined as 0.10 μM and > 60 μM in ARCaP_E-shEPLIN and ARCaP_E-shCtrl cells, respectively, with a selectivity index (SI) > 600, where the SI was defined as the ratio of the IC₅₀ in ARCaP_E-shCtrl cells and that in ARCaP_E-shEPLIN cells. Similarly, bromocriptine had an IC₅₀ of 1.41 μM in C4-2B-TaxR cells and > 60 μM in parental C4-2B cells, with an SI of > 42.55. Bromocriptine exhibited weak cytotoxicity (IC₅₀ values > 60 μM) in ARCaP_E-shCtrl, C4-2B cells, and other chemoresponsive PCa cell lines (including LNCaP, C4-2, CWR22Rv1, and PC-3), as we previously reported [26]. Intriguingly, bromocriptine exhibited higher potency and selectivity than the lead compound LG1980 in chemoresistant PCa cells, such as C4-2B-TaxR cells [30]. These results indicated that bromocriptine had high selectivity and potency against chemoresistant PCa cells.

Fig. 1.

Bromocriptine exhibits high selectivity and potency in chemoresistant PCa cells. (A) ARCaP_E-shCtrl and ARCaP_E-shEPLIN cells (left), C4-2B and C4-2B-TaxR cells (right) were treated with bromocriptine at varying concentrations for 72 h. In vitro viability was determined using CCK-8 assay. (B) Flow cytometry analysis of the cell cycle in C4-2B-TaxR cells treated with bromocriptine at the indicated concentrations (48 h). ** p < 0.01, *** p < 0.001. (C) Flow cytometry assay of Annexin V and 7-AAD in C4-2B-TaxR cells treated with bromocriptine at the indicated concentrations (72 h). ** p < 0.01, *** p < 0.001, **** p < 0.0001. (D) Western blot analysis on the expression of apoptotic markers in C4-2B-TaxR cells treated with bromocriptine (2.0 μM) at the indicated times.

Bromocriptine induces cell cycle arrest and apoptosis in chemoresistant PCa cells

Flow cytometry was performed to analyze the in vitro effect of bromocriptine on the cell cycle and apoptosis in C4-2B-TaxR cells. Bromocriptine treatment at 2.0 μM caused a significant increase in the mean percentage of cells in the G2/M phase and a slight increase in the S phase. The G2/M phase cell cycle arrest was accompanied by a decrease in the percentage of cells in the G0/G1 phase (Fig. 1B). The results indicated that bromocriptine inhibited the proliferation of C4-2B-TaxR cells primarily during the G2/M phase. Annexin V/7-AAD staining confirmed that compared with the vehicle control, bromocriptine more effectively induced early-stage apoptosis in a dose-dependent manner (Fig. 1C). Western blotting studies showed that bromocriptine treatment at 2.0 μM effectively caused the cleavage of poly (ADP-ribose) polymerase (PARP) and caspase-3 in C4-2B-TaxR cells in a time-dependent manner (Fig. 1D).

Bromocriptine significantly affects the transcriptome in chemoresistant PCa cells

To understand the molecular targets of bromocriptine in chemoresistant PCa cells, we performed an RNA-seq analysis in C4-2B-TaxR cells treated with bromocriptine (1.3 μM, 24 h) and the vehicle control (Fig. 2A). When a fold-change cutoff of ≥ 1.5 and an adjusted p-value (p-adj) < 0.05 were used, data analyses using the ROSALIND® platform identified 157 differentially expressed genes (DEGs) following treatment with bromocriptine (Fig. 2B, Table S2). Several DEGs with known functions in inhibiting the cell cycle, activating cell death, or suppressing aggressiveness, i.e., CHAC1, GDF15, SSEN2, CDKN1A (p21), FAS, ATF3, and CCNG2, were significantly increased. Interestingly, among the 89 downregulated genes, 41 belonged to the family of histone genes, with most in the replication-dependent histone clusters 1 (34 genes) and 2 (5 genes). CCNE2, the major E-cyclin within the histone locus bodies (HLBs) that may play a central role in activating histone transcription and causing genomic instability, was significantly reduced. The downregulated genes also included those implicated in the regulation of DNA replication and the cell cycle, for example, several members of the MCM and E2F families. Ingenuity Pathway Analysis (IPA) profiling further identified the top suppressed gene clusters as cell cycle control of chromosomal replication, estrogen-mediated S-phase entry, and cyclins and cell cycle regulation, and the top activated clusters as nucleotide excision repair (NER), p53 signaling, and senescence pathway (Fig. 2C, D). Interestingly, approximately one-third (50/157) of the DEGs affected by bromocriptine overlapped with known p53-p21-RB target genes [36] (Fig. 2E, Table S3). These results indicated that the short-term treatment with low-dose bromocriptine might cause stalled DNA replication and a rapid but unsuccessful effort to repair DNA damage, eventually leading to cell cycle arrest and cell death in chemoresistant PCa cells.

Fig. 2.

Bromocriptine treatment (1.3 μM, 24 h) significantly affects the transcriptome of chemoresistant PCa cells. (A) Volcano plot of RNA-seq results. The green dots represent significantly upregulated genes, the purple dots represent significantly downregulated genes (fold of change ≥ 1.5 and adjusted p-value < 0.05), and the black dots represent insignificantly differentially expressed genes. (B) Top ten upregulated and downregulated genes upon bromocriptine treatment. (C) Bubble chart plots of canonical pathways affected by bromocriptine treatment in IPA analysis. Y-axis: canonical pathway categories, X-axis: negative log of the Fisher's Exact Right-tailed p-value. Orange bubbles have positive values, and blue bubbles have negative values. (D) Most highly rated networks in IPA analysis. The genes shaded orange are upregulated, and those shaded blue are downregulated. Solid lines represent direct interactions between pairs of genes, and dotted lines represent indirect interactions. (E) Venn diagram showing 50 overlapped genes between p53-p21-RB target genes and bromocriptine target genes.

Bromocriptine affects multiple cell cycle regulators in chemoresistant PCa cells

We investigated the effect of bromocriptine on a panel of key regulators of the cell cycle progression in chemoresistant PCa cells using Western blot analyses as follows (Fig. 3A): (1) RB-E2F-c-Myc: Bromocriptine rapidly reduced the levels of phosphorylated retinoblastoma protein (p-RB) at serine 249, serine 807 and serine S811 [[37], [38], [39], [40]], and decreased the expression of E2F1 as early as 24 h. c-Myc, an E2F1-regulated gene that plays an important role in controlling the cell cycle and apoptosis [41], was also downregulated by bromocriptine treatment. These results indicated that inhibiting p-RB-E2F-c-Myc signaling may contribute to bromocriptine's effect on the cell cycle and apoptosis; (2) Mouse double minute 2 homolog (MDM2)-p53-p21 [42]. Bromocriptine treatment significantly decreased the phosphorylation of MDM2 at serine 166 and total MDM2 while increasing p53 expression as early as 24 h. Consistently, there was a significant increase in the p21 level that may contribute to the cell cycle arrest at the G₂/M transition; (3) S-phase kinase-associated protein 2 (Skp2)-p27 [43]: Bromocriptine significantly reduced Skp2 expression and increased p27 levels in a time-dependent manner. Interruption of the feedback loop of RB-E2F, Skp2, p27 and cyclin E-CDK2 may be an underlying mechanism by which bromocriptine caused the cell cycle arrest in chemoresistant PCa cells.

Fig. 3.

Bromocriptine affects multiple signaling pathways in chemoresistant PCa cells. (A) Western blot analysis on the expression of several key regulators of the cell cycle in C4-2B-TaxR cells treated with vehicle or bromocriptine (2.0 μM) at the indicated times. (B) Western blot analysis on the expression of DRD1 and DRD2 in parental C4-2B and C4-2B-TaxR cells (left) and in C4-2B-TaxR cells treated with vehicle or bromocriptine (2.0 μM) at the indicated times (right). (C) Western blot analysis on the expression of several known DRD2 downstream signaling pathways in C4-2B-TaxR cells treated with vehicle or bromocriptine (2.0 μM) at the indicated times. (D) Western blot analysis on the expression of newly identified downstream targets of bromocriptine in C4-2B-TaxR cells treated with vehicle or bromocriptine (2.0 μM) at the indicated times. (E) Relative luciferase activities of human survivin promoter reporters (pSVV-Luc1430 and pSVV-Luc230) in C4-2B-TaxR cells treated with bromocriptine (48 h) at the indicated concentrations. **: p < 0.01.

Bromocriptine affects DRD2 expression and downstream signaling in chemoresistant PCa cells

Bromocriptine exhibits potent agonist activity on the dopamine D2 receptor (DRD2), among several G protein-coupled receptors (GPCRs). Previously, we reported that DRD2 expression is significantly downregulated in PCa tissues with higher Gleason scores, which suggested an inverse association between DRD2 expression and PCa aggressiveness [26]. Western blotting analyses found that compared with that in parental C4-2B cells, the basal level of DRD2 protein was slightly reduced in C4-2B-TaxR cells whereas DRD1 expression appeared to be similar (Fig. 3B). Interestingly, bromocriptine treatment at low concentrations markedly increased DRD2 expression in a time-dependent manner, beginning at 48 h. In comparison, there was no significant change in the DRD1 level. These results indicated that bromocriptine could increase DRD2 expression in chemoresistant PCa cells.

We investigated the effect of bromocriptine on several classical and nonclassical DRD2 downstream pathways (Fig. 3C, 3D). (1) Protein kinase A (PKA)-cAMP response element-binding protein (CREB) [[44], [45], [46], [47]]: In contrast to the expected effect as a typical DRD2 agonist, bromocriptine increased the expression and phosphorylation of CREB as early as 24 h. There was also increased phosphorylation of cAMP-regulated phosphoprotein of molecular weight 32 kDa (DARPP-32) and type 1 protein phosphatase-α (PP1α), although PP1α phosphorylation was reduced to a lower level at 72 h. These results suggested that the PKA-CREB signaling was activated by bromocriptine in chemoresistant PCa cells; (2) Extracellular signal-regulated protein kinases 1 and 2 (ERK1/2): DRD2 activates ERK1/2 in G protein-dependent or -independent (e.g., via β-arrestin2 or the transactivation of receptor tyrosine kinases) manners. In line with this classical effect of D2 receptors, bromocriptine increased the phosphorylation of ERK1/2 in chemoresistant PCa cells; (3) Adenosine monophosphate-activated protein kinase (AMPK) [48]:  Interestingly, bromocriptine reduced the phosphorylation of calcium/calmodulin-dependent protein kinase kinase 2 (CaMKK2) at serine 511 in chemoresistant PCa cells, indicating the activation of CaMKK2 [49]. Consistently, AMPK phosphorylation was significantly increased following bromocriptine treatment; (4) p38 mitogen-activated protein kinase (p38 MAPK) [[50], [51], [52], [53]]: Compared with its effect on the induction of p-ERK1/2, bromocriptine more significantly increased the phosphorylation of p38 MAPK in chemoresistant PCa cells, suggesting that the activation of p38 MAPK signaling may contribute to the observed cell cycle arrest and apoptosis; (5) Nuclear factor kappa B  (NF-κB) [[54], [55], [56], [57]]: In chemoresistant PCa cells, bromocriptine effectively reduced the phosphorylation of NF-κB without altering the expression of total NF-κB protein; (6) Enhancer of zeste homolog 2 (EZH2)-Stat3: Our recent work demonstrated that the activation of noncanonical EZH2-Stat3-Skp2 signaling represents an important survival mechanism in chemoresistant PCa cells [30]. Bromocriptine effectively inhibited EZH2 expression and phosphorylation at serine 21. Consistently, phosphorylation of Stat3 at serine 727 and Skp2 expression was reduced in C4-2B-TaxR cells; (7) Survivin: survivin is a member of the inhibitor of apoptosis (IAP) family, and its transcription can be directly activated by the binding of Stat3 to three cis-elements within the region of -1,231 to -1,009 bp upstream from the translation initiation site [58]. In C4-2B-TaxR cells, bromocriptine significantly decreased survivin expression at the protein level. Dual-luciferase reporter assay was further performed using two human survivin reporters that contained a 1,430-bp region of the survivin promotor (pSVV-Luc1430) or a 230-bp truncated fragment (pSVV-Luc230). Bromocriptine more effectively inhibited the luciferase activity of pSVV-Luc1430 than pSVV-Luc230, suggesting a possible role of Stat3 in mediating the effect of bromocriptine on survivin transcription (Fig. 3E); (8) Androgen receptor (AR): AR plays an essential role in the development of mCRPC [59]. Our previous studies in chemoresponsive C4-2 cells found that bromocriptine treatment at 20 μM significantly reduced AR expression. In C4-2B-TaxR cells, however, bromocriptine at 2.0 μM did not affect AR expression (Fig. 3C).

Bromocriptine inhibits the skeletal growth of chemoresistant C4-2B-TaxR xenografts in male athymic nude mice

We evaluated the in vivo efficacy of bromocriptine in intratibial C4-2B-TaxR xenografts that closely mimicked the clinicopathology of bone metastatic, chemoresistant PCa [30,32,60]. A total of 12 male athymic nude mice inoculated with C4-2B-TaxR cells in both tibias were randomized and divided into 2 groups (n = 6 per group) and treated with the vehicle control (DMSO) or bromocriptine (15 mg/kg), respectively, three times per week, via intraperitoneal injection. Following an 11-week treatment, the average PSA level of each treatment group was 150.36 ± 61.29 ng/ml (control) and 66.86 ± 17.63 ng/ml (bromocriptine), respectively. Statistical analyses showed that the difference was significant (p < 0.001) (Fig. 4A). Bromocriptine treatment was associated with increased average body weight in mice, and no significant adverse effect was observed (Fig. 4B). X-ray radiography indicated that bromocriptine improved the skeletal architecture and reduced osteolytic and osteoblastic lesions in tumor-bearing bones (Fig. 4C). Immunohistochemical studies found that compared with the vehicle control, bromocriptine reduced the tissue expression of the proliferation marker Ki67 in bone tumors (Fig. 4D). These results indicated that as a monotherapy, bromocriptine effectively inhibited the in vivo growth of chemoresistant PCa in mouse bones.

Fig. 4.

Bromocriptine monotherapy inhibits the skeletal growth of chemoresistant PCa cells in male athymic nude mice. (A) Serum PSA levels of mice inoculated with intratibial C4-2B-TaxR cells and treated with vehicle control (n = 6) and bromocriptine (15 mg/kg; n = 6), three times per week, via i.p injection. Two-way ANOVA analysis was used to evaluate the statistical difference in the PSA values between different treatment groups. ***: p < 0.001. (B) Left: average body weights of mice treated with vehicle or bromocriptine; right: percentage of change in the average body weight under different treatments. (C) X-ray radiography of tumor-bearing tibia under different treatments. (D) H&E staining of C4-2B-TaxR tumors and tissue expression of Ki67 in mice under different treatments. Scale bar: 100 µm.

Discussion

Previous studies have shown that as a single agent, bromocriptine only exhibited weak to moderate cytotoxicities in various experimental models of solid and hematological cancers, with IC₅₀ values ranging from 10 to > 60 μM [18,[26], [27], [28], [29]]. Bromocriptine monotherapy also failed to demonstrate significant responses in several pilot trials in PCa and breast cancer patients. The low efficacy of bromocriptine in these preclinical and clinical studies has questioned the promise of bromocriptine in cancer therapy. In the current study, for the first time, we identified bromocriptine as a potent and specific inhibitor of chemoresistant PCa. Unlike previous observations, we found that bromocriptine targeted chemoresistant PCa cells within a low micromolar range in a highly selective manner. In comparison, bromocriptine did not markedly affect the in vitro proliferation and survival of chemoresponsive PCa cells, even at much higher concentrations. The high specificity and potency of bromocriptine against chemoresistant PCa cells indicated a novel mechanism of action of the drug. We identified several key factors and pathways that could mediate the anticancer activities of bromocriptine in chemoresistant PCa cells. These molecular analyses suggested a working model that bromocriptine might exert its anticancer effect via DRD2-dependent and -independent pathways in chemoresistant PCa cells (Fig. 5). In animal models, bromocriptine monotherapy significantly retarded the skeletal growth of chemoresistant xenografts. These results revealed a new function of bromocriptine in chemoresistant PCa cells and provided important references to evaluate its clinical efficacy in patients with chemoresistant PCa.

Fig. 5.

Schematic diagram of possible mechanisms of action of bromocriptine in chemoresistant PCa cells. Bromocriptine may exert its anticancer functions in chemoresistant PCa cells via DRD2-dependent and -independent mechanisms. Activation of tumor suppressive signaling (for example, p53, p21, p27, and RB) and inhibition of oncogenic factors (including E2F, c-Myc, cyclin E, EZH2, Stat3, Skp2, and survivin) lead to cell cycle arrest and apoptosis.

Only a few possible mechanisms underlying the anticancer activities of bromocriptine have been proposed. (1) ATP binding cassette subfamily B member 1 (ABCB1): Orlowski et al. found that bromocriptine inhibited the basal and chemotherapeutic-induced ATPase activity of ABCB1 with inhibitory constant (Ki) values at approximately 0.07–0.30 μM [61]. In another study, the Ki of bromocriptine on ABCB1 activity was determined as 6.52 μM, and bromocriptine at 10 μM significantly sensitized doxorubicin- and vincristine-resistant, ABCB1-overexpressing human leukemia K562 cells [27]. These observations suggested that bromocriptine might be capable of reversing the ABCB1-mediated efflux of chemotherapeutics. Intraperitoneal injection of bromocriptine (6.25 mg/kg per day) slowed the orthotopic growth of multidrug-resistant human hepatocellular carcinoma HepG₂ xenografts, which was associated with an inhibition of ABCB1 expression and induction of apoptosis in xenograft tissues [28]; (2) NF-κB: Seo et al. reported that the cellular responses of the NCI-60 cell lines to bromocriptine were not associated with expression of ABC-transporters (such as ABCB1) and other drug resistance mechanisms (such as mutated p53). Instead, a significant correlation was found between bromocriptine treatment and NF-κB inhibition. Molecular docking further showed that bromocriptine had stronger binding to the NF-κB-DNA complex than with NF-κB alone, indicating that bromocriptine may inhibit DNA binding. However, high bromocriptine concentrations (20–40 μM) were required to inhibit NF-κB-dependent transcription [18].

Recent studies in experimental models have suggested a possible role of DRD2 signaling in cancer progression, although the clinical significance of these findings remains inconclusive and sometimes controversial. For example, it has been reported that DRD2 is frequently overexpressed in glioblastoma, gastric cancer, pancreatic cancer, acute myeloid leukemia and breast cancer, whereas the protein is downregulated in non-small cell lung cancers [56,[62], [63], [64], [65], [66], [67], [68], [69], [70]]. In a previous study, we demonstrated that DRD2 was increased in low-grade PCa compared with that in normal/benign prostatic tissues but was further downregulated in high-grade PCa, suggesting an inverse association between DRD2 expression and PCa aggressiveness. In chemoresponsive PCa cells (such as C4-2), bromocriptine treatment at 20 μM suppressed the expression of c-Myc, E2F-1, Skp2 and survivin, whereas it upregulated p21, p27 and p53, presumably via a DRD2-dependent mechanism. Interestingly, bromocriptine could also inhibit AR, a major driver of mCRPC. These bromocriptine-induced molecular changes might be responsible for the synergistic effect with docetaxel in suppressing cell cycle progression and causing apoptosis in chemoresponsive PCa cells [26].

Our results in the current study indicated that bromocriptine might significantly increase DRD2 expression and affect certain aspects of DRD2 signaling in chemoresistant PCa cells, but at much lower concentrations (such as 2.0 μM) compared with its effect in chemoresponsive PCa cells. An examination of the classical DRD2 downstream signaling following bromocriptine treatment found that the drug could act similarly to dopamine or known DRD2 agonists, for example, by activating ERK1/2 and inhibiting the phosphorylation of NF-κB, which may be mediated via β-arrestin2/protein phosphatase 2A (PP2A)-dependent mechanisms [[54], [55], [56], [57]]. Similar to its effect in benign prolactinoma cells, bromocriptine increased the phosphorylation of AMPK at threonine 172 within the α subunit in C4-2B-TaxR cells. As a master regulator of the cellular energy metabolism [48], activation of AMPK by CaMKK2 or liver kinase B1 (LKB1) inhibits cancer cell growth and proliferation via several mechanisms, for example, inhibiting the mammalian target of rapamycin (mTOR) and activating the p53-dependent cellular senescence [48]. Another major downstream effector of AMPK is p38 MAPK, which can promote cell cycle arrest by activating p53-p21 and/or RB pathways, thereby exerting inhibitory effects during cancer progression [[50], [51], [52]]. Recent studies identified p38 MAPK as a key factor regulating apoptosis and dormancy-related signaling in PCa and other cancer types that may depend on the relative abundance of phosphorylated (active) p38 MAPK vs. phosphorylated ERK1/2 [53]. Interestingly, bromocriptine significantly activated p38 MAPK in chemoresistant PCa cells.

DRD2 activation in neurons usually leads to reduced production of cAMP and inactivation of PKA via coupling to Gα_i/o proteins [44], followed by dephosphorylation of PKA targets, such as CREB at serine 133 and DARPP-32 at threonine 34, and subsequently controlling the equilibrium between PKA and PP1 [45], a serine/threonine phosphatase whose phosphorylation is essential for G₁/S- and G₂/M transitions [46,47]. Although DRD2 activation was frequently associated with the inhibition of PKA-CREB signaling, certain DRD2 agonists, such as quinpirole, induced CREB phosphorylation via DARPP-32 [69]. It seemed that bromocriptine exhibited a similar effect in chemoresistant PCa cells. These observations suggested that bromocriptine might activate DRD2 signaling through both G protein-dependent and -independent (i.e., β-arrestin2-dependent) mechanisms (Fig. 5). An important note is that bromocriptine-induced DRD2 upregulation might lead to sustained activation of DRD2 signaling, thereby contributing to the anticancer effect of bromocriptine in chemoresistant PCa cells.

The RNA-seq results provided an unbiased view of the effect of bromocriptine on the transcriptome of chemoresistant PCa cells. Bromocriptine treatment at a low concentration rapidly affected several clusters of genes implicated in DNA replication and repair, cell cycle regulation and senescence, and cell death. Of particular interest, there was a high number of overlap between the bromocriptine DEGs and common p53-p21-RB genes (Table S3), indicating that the p53-p21-RB activation could be responsible for bromocriptine-induced cell cycle arrest and apoptosis. The RB protein is a central regulator of the cell cycle progression [71,72]. Cyclin-dependent kinase (CDK)-mediated RB phosphorylation at several serine residues, including serine 249, serine 807, and serine 811, dissociates RB from E2F transcriptional factors and promotes E2F-dependent genes required for the cell cycle progression [[37], [38], [39], [40]]. A shift from RB hyperphosphorylation to hypophosphorylation results in cell cycle arrest and apoptosis [[73], [74], [75], [76]]. p53 is a tumor suppressor with an essential function in the induction of cell cycle arrest and apoptosis. The CDKN1A gene, coding for the cyclin-dependent kinase inhibitor p21/WAF1/CIP1/CDKN1A, is a prominent p53 target and is transcriptionally activated by p53 [[77], [78], [79]]. The formation of p21-cyclin-CDKs complexes leads to RB hypophosphorylation and subsequent suppression of E2F-dependent transcription. Indeed, approximately half of the mostly downregulated DEGs (41/89) belonged to a subset of replication-dependent histone cluster genes implicated in the cell cycle regulation [80]. A key player could be cyclin E2, which is under the direct control of RB-E2F1 and is essential to histone gene transcription for DNA replication. Transcriptional inhibition of histones further requires functional p53 and p21 proteins [81]. Bromocriptine-mediated decrease in cyclin E2 and increase in p53/p21 might be responsible for inhibiting these histone gene clusters. RB can also exert E2F-independent functions by directly binding Skp2, a direct target of E2F transcriptional factors, and preventing Skp2-mediated ubiquitination and degradation of the CDK inhibitor p27 (KIP1, CDKN1B) [43]. Additionally, cyclin E2 reduction might lead to Skp2 inhibition and p27 stabilization and, consequently, a halted cell cycle (Fig. 5).

Several oxidative stress-responsive genes were among the top-upregulated DEGs upon the short-term treatment with bromocriptine. CHAC1 encodes ChaC glutathione specific γ-glutamylcyclotransferase 1, a component of the unfolded protein response (UPR) pathway that plays an important role in cellular oxidative homeostasis. A recent study found that overexpression of CHAC1 in PCa cells inhibited cell viability and sensitized the cells to docetaxel, at least partially through the induction of endoplasmic reticulum (ER) stress [82]. DNA-damage-inducible transcript 4 (DDIT4, also known as REDD1 and RTP801) is rapidly induced by various stresses to repress the mechanistic target of rapamycin complex I (mTORC1), and the loss of DDIT4 leads to dysfunctional mitochondria with increased reactive oxygen species (ROS) and impaired oxidative phosphorylation [83]. DDIT4 was identified as a molecular target of metformin via a p53-dependent mechanism, resulting in mTOR inhibition and cell-cycle arrest in LNCaP cells [84]. Sestrin2 (SESN2), a stress-inducible protein involved in homeostatic regulation via the inhibition of mTORC1 and ROS, was also markedly increased. Like DDIT3, SESN2 is a transcriptional target of p53 and may mediate the tumor suppressor activities of p53. SESN2 overexpression suppressed cancer cell proliferation, migration, and epithelial-mesenchymal transition [85]. SLC7A11 (or xCT) encodes the transporter subunit of the heterodimeric amino acid transport system x_c⁻ and is the major transporter of extracellular cystine. Overexpression of SLC7A11 promotes the reduction of lipid hydroperoxides to lipid alcohols using reduced glutathione (GSH), thereby protecting cells against membrane lipid peroxidation. Collectively, the upregulation of these antioxidant genes indicated a rapid response of the cancer cells to control bromocriptine-induced oxidative damage and, consequently, cell death.

It remains to be addressed whether DRD2 is the only major, direct target of bromocriptine in chemoresistant PCa cells. Our pilot screen on the ARCaP_E-shEPLIN/ARCaP_E-shCtrl platform included 72 dopamine receptor modulators (agonists, antagonists, inhibitors, dopamine precursors, and metabolites). When used at 12.3 μM, bromocriptine and dihydroergocristine methanesulfonate were the only two compounds that demonstrated high potency (≥ 90% inhibition) and selectivity (selectivity index over ARCaP_E-shCtrl cells ≥ 10.0) in ARCaP_E-shEPLIN cells (Table S4). Whereas bromocriptine and dihydroergocristine are known DRD2 agonists, other selective and potent DRD2 agonists, such as ropinirole, quinelorane, (±)-PPHT, did not exhibit cytotoxicity in ARCaP_E-shEPLIN cells. On the other hand, several DRD2 antagonists, including cortexolone and (±)-octoclothepin, had relatively high cytotoxicity and selectivity against ARCaP_E-shEPLIN cells. These observations raised interesting questions about whether interfering with DRD2 signaling is solely responsible for the unique anticancer activity of bromocriptine in chemoresistant PCa cells. In fact, due to the high homology of dopamine receptors, most dopaminergic drugs are well known for their highly polypharmacologic nature and capabilities of co-targeting multiple dopamine receptors and other aminergic GPCRs. For example, bromocriptine exhibits agonist activity on 5-hydroxytryptamine (5-HT)_1D and DRD3 and antagonist activity on α_2A-adrenergic receptors and DRD1, among the known receptors of bromocriptine. One or more of these GPCRs could be involved in the anticancer effect of bromocriptine. It also remains to be elucidated how bromocriptine reduced the phosphorylation of EZH2 in chemoresistant PCa cells. Further investigation is needed to elucidate the direct molecular target(s) of bromocriptine in chemoresistant cancer cells.

Bromocriptine has been extensively used to manage various benign diseases with minimal or no adverse effects on primary physiological functions, and long-term treatment with bromocriptine has been associated with additional benefits, such as reduced cardiovascular risks [15]. Bromocriptine has a relatively wide range of safe dosages (for example, up to 100 mg/kg via oral dosing) and flexible routes of administration. These favorable features of bromocriptine make it an ideal candidate to be repurposed for cancer treatment. However, the lack of clinical benefits of bromocriptine in several pilot trials raised questions about its potential application in cancer treatment. Our current study demonstrated that bromocriptine exhibited high specificity and potency against chemoresistant PCa cells, providing preclinical evidence for developing bromocriptine as a subtype-specific targeted therapy. Bromocriptine could be rapidly tested in future clinical trials as a monotherapy in PCa patients who have failed docetaxel chemotherapy. As a repurposed drug, bromocriptine could provide an effective and affordable treatment to overcome chemoresistance in PCa and other cancer types.

Financial Support

This work was supported by National Cancer Institute (NCI) grants R01CA256058 and R42CA217491, National Institute on Minority Health and Health Disparities (NIMHD) Research Center in Minority Institute grant 5U54MD007590, Department of Education Title III Program (D. Wu), and NIMHD 5U54MD007590 Pilot Project Grant (X. Li).

Data availability

The data generated in this study are available upon request from the corresponding authors.

CRediT authorship contribution statement

Lijuan Bai: Data curation, Formal analysis, Investigation, Methodology, Visualization. Xin Li: Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Visualization, Writing – original draft. Yang Yang: Data curation, Investigation. Rui Zhao: Data curation, Investigation. Elshaddai Z. White: Data curation, Formal analysis, Investigation, Visualization. Alira Danaher: Data curation, Formal analysis, Investigation. Nathan J. Bowen: Data curation, Formal analysis, Methodology, Visualization, Writing – review & editing. Cimona V. Hinton: Formal analysis, Writing – review & editing. Nicholas Cook: Investigation. Dehong Li: Investigation. Alyssa Y. Wu: Formal analysis, Visualization, Writing – review & editing. Min Qui: Investigation, Methodology. Yuhong Du: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources. Haian Fu: Conceptualization, Project administration, Resources. Omer Kucuk: Conceptualization, Writing – review & editing. Daqing Wu: Conceptualization, Data curation, Formal analysis, Funding acquisition, Project administration, Resources, Supervision, Writing – original draft, Writing – review & editing.

Declaration of Competing Interest

The authors declare that there is no conflict of interest.