Treatment of prostate cancer with paclitaxel often fails due to the development of chemoresistance caused by downregulation of the tumor suppressor gene miR-34a. In this study, we demonstrate that codelivery of paclitaxel and 2'-hydroxy-2,4,4',5,6'-pentamethoxychalcone (termed rubone) drives upregulation of miR-34a and chemosensitizes paclitaxel-resistant prostate cancer cells, killing both cancer stem–like cells (CSC) and bulk tumor cells. Rubone upregulated miR-34a and reversed its downstream target genes in DU145-TXR and PC3-TXR cells. Paclitaxel and rubone combination therapy inhibited tumor cell growth, migration, and CSC population growth. We synthesized poly(ethylene glycol)-block-poly(2-methyl-2-carboxyl-propylene carbonate-graft-dodecanol; PEG-PCD) to prepare micelles. The drug-loading capacities were 9.70% ± 0.10% and 5.34% ± 0.02% for paclitaxel and rubone, respectively, controlling a drug release of 60.20% ± 2.67% and 60.62% ± 4.35% release of paclitaxel and rubone at 24 hours. Delivery of miR-34a and rubone decreased PC3-TXR cell viability with increasing paclitaxel concentration. Coincubation with a miR-34a inhibitor diminished the effect of rubone. Paclitaxel IC50 in PC3 and PC3-TXR cells was 55.6 and 2,580 nmol/L, respectively, but decreased to 49.8 and 93.2 nmol/L when treated in combination with rubone, demonstrating a reversal of paclitaxel resistance by rubone. Systemic administration of micelles carrying paclitaxel and rubone inhibited orthotopic prostate tumor growth in nude mice, compared with monotherapy, by reversing the expression of miR-34a, SIRT1, cyclin D1, and E-cadherin. In summary, our results showed how rubone acts as an efficient small-molecule modulator of miR-34a to reverse chemoresistance and further enhance the therapeutic efficacy of paclitaxel in paclitaxel-resistant prostate cancer. Cancer Res; 77(12); 3244–54. ©2017 AACR.

Most prostate cancers relapse within two years into hormone-refractory cancers due to the presence of tumor-initiating cells, known as cancer stem cells (CSC), which are involved in tumor progression and metastasis, but are resistant to chemotherapy. Recently, aberrant expression of miRNAs was critically implicated in the initiation, progression, migration, and chemoresistance of cancer (1, 2). Among these miRNAs, miR-34a is significantly downregulated in chemoresistant prostate cancer cell line (3) or CD44+ CSCs (4). As a tumor suppressor miRNA, miR-34a is responsible for promoting tumor cell apoptosis, inhibiting tumor metastasis (5) and chemoresistance (6). Thus, miR-34a replenishment might be a novel therapeutic method to reverse paclitaxel resistance for the treatment of chemoresistant prostate cancer. Kojima and colleagues reported that miR-34a reversed paclitaxel resistance by targeting the downstream genes including SIRT1 and Bcl-2 (7). Yao and colleagues reported that combination therapy using doxorubicin and miR-34a synergistically enhanced the antitumor property of doxorubicin and inhibited DU145 cell–formed tumor growth in vivo (8). Nonetheless, intrinsic challenges associated with oligonucleotide-based miRNA replenishment including off-target effects, poor cellular uptake, and in vivo instability hindered its clinical translation. Even though numerous miRNA delivery systems were developed, most of them were proven less effective or toxic for clinical use (9). Thus, to reverse the aberrant expression of miR-34a by small molecules might be a potent alternative method for the treatment for paclitaxel-resistant prostate cancer.

Natural and synthetic analogues of chalcones and isoflavones exhibit promising anticancer activity. However, only a few studies have focused on the role of chalcone derivatives on modulation of miRNAs. Rubone (10), isoliquiritigenin (11), and kuwanon V (12) modulate miR-34a, miR-25, miR-9, miR-29a, and miR-181a, respectively, with potent biological actions. Among these small molecules, Xiao and colleagues first reported rubone, a chalcone analogue, as a miR-34a modulator for the inhibition of hepatocellular carcinoma (HCC) growth (10). In their study, rubone upregulated miR-34a expression in a p53-dependent manner, downregulated the downstream target Bcl-2 and cyclin D1 expression, and suppressed HCC growth in vivo. However, the antitumor efficacy of rubone as a miR-34a modulator for treating paclitaxel-resistant prostate cancer and the underlining mechanisms remains largely unknown. Furthermore, poor aqueous solubility of paclitaxel and rubone (less than 50 mg/L) results in low and variable drug absorption. Thus, novel drug delivery systems for codelivery of both drugs are required for combination therapy against paclitaxel-resistant prostate cancer.

As the use of solubilizing agents and surfactants may cause organ and systemic toxicity, biodegradable polymers, which can self-assemble into nano-sized micelles, are gaining much attention. Polymeric micelles have spherical structures with a hydrophilic corona and hydrophobic core, which improves the aqueous solubility and stability of hydrophobic drugs (13). The stealth property of poly(ethylene glycol) (PEG) hydrophilic corona of micelles prevents their recognition by reticuloendothelial system (RES) and therefore minimizes their rapid elimination via the enhanced permeability and retention (EPR) effect. In our previous study, we designed and synthesized poly(ethylene glycol)-block-poly(2-methyl-2-carboxyl-propylene carbonate-graft-dodecanol; PEG-PCD), which significantly enhanced the aqueous solubility of embelin, an X-linked inhibitor of apoptosis protein (XIAP) inhibitor (14). Here, we determined the effect of rubone on miR-34a and its target genes and investigated whether it could chemosensitize paclitaxel-resistant prostate cancer cells and synergistically inhibited orthotopic prostate tumor in nude mice when administered intravenously as a micellar formulation with paclitaxel.

Cell lines and culture condition

Prostate cancer cell lines LNCaP (obtained in 2016), C4-2 (obtained in 2011), DU145 (obtained in 2010), and PC3 (obtained in 2010) were purchased from the ATCC and cultured in RPMI1640 containing 10% FBS and 1% penicillin/streptomycin in a humidified 37°C incubator supplemented with 5% CO2. The paclitaxel-resistant version of DU145 and PC3 (DU145-TXR and PC3-TXR) were provided by Dr. Evan T. Keller from the University of Michigan (Ann Arbor, MI) in 2010. Normal prostate epithelial RWPE-1 cells were provided by Dr. Ming-Fong Lin from University of Nebraska Medical Center (Omaha, NE) in 2016 and cultured in complete keratinocyte growth medium, K-SFM (Life Technologies) supplemented with 50 μg/mL of bovine pituitary extract and 5 ng/mL of EGF. The approximate number of passages for all cell lines between collection and thawing is 15. All cell lines were authenticated using DNA fingerprint (last time performed in 2017) by the University of Arizona Genetics Core (Tucson, AZ).

RT-PCR and Western blot analysis

Following the treatment of different concentrations of rubone or paclitaxel and rubone combination therapy, total mRNA was isolated using RNeasy isolation kits (Qiagen) and 170 ng total RNA was converted to cDNA using miR-34a primer to determine miR-34a concentration. To determine protein concentration, cell protein was extracted using RIPA buffer after treatment with rubone at the doses of 5 and 10 μmol/L for 48 hours. The amount of protein was adjusted to the same concentration, transferred to polyvinylidene difluoride membrane, incubated with primary and secondary antibodies, followed by LI-COR Odyssey System Analysis (LI-COR Biosciences). The primary antibodies used in Western blot and IHC studies were the following: anti-E-cadherin (Abcam, ab15148), anti-SIRT1 (Santa Cruz Biotechnology, sc-15404), anti-cyclin D1 (Abcam, ab16663), anti-p53 (Santa Cruz Biotechnology, sc-6243), anti-Bax (Santa Cruz Biotechnology, sc-6236), anti-β-actin (Santa Cruz Biotechnology, sc-1616), anti-TAp73 (Santa Cruz Biotechnology, sc-7957), and anti-Elk-1 (Santa Cruz Biotechnology, sc-355).

Cell viability in 2D and 3D models

In a two-dimensional (2D) model, cell viability was determined by MTT assay after treating the cells with different concentrations of rubone or 5 μmol/L rubone plus different concentrations of paclitaxel. The antitumor effect of the combination therapy using paclitaxel and rubone was also determined using three-dimensional (3D) tumor model (15) including 3D on top and hanging-drop models. For 3D on top assays in 24-well plate, 200 μL of 100% Matrigel was used as the basement and 2 × 105 single cells were suspended in 300 μL 10% Matrigel in RPMI1640 as the growth medium. After culturing for 48 hours, growth medium was exchanged to growth medium containing paclitaxel and rubone at different concentrations, which was replaced with fresh media every two days for 2 weeks before analyzing the therapeutic effect. For the hanging-drop model, 40 μL medium containing 4,000 cells were added in each well of 3D 96-well hanging-drop plate (3D Biomatrix) and drug-containing medium was changed every two days for 3 weeks until sphere formation of the control group.

Cell invasion and migration

The effect of paclitaxel and rubone combination therapy on cell invasion and migration was determined using Transwell membrane filter inserts (pore size, 8 μm) in 6-well culture plates. For invasion assay, 200 μL Matrigel (BD Biosciences) was added to each Transwell insert where RPMI1640 without FBS was used as the cell culture medium, while RPMI1640 with 10% FBS was added in each well. Paclitaxel-resistant DU145-TXR and PC3-TXR cells were cultured for another 72 hours after drug treatment. The number of cells invading Matrigel was quantified after staining with crystal violet. For migration assay, 1 × 106 cells were seeded into each Transwell insert and cultured for another 72 hours after adding the drugs. The cell number was counted under a microscope in randomly selected three fields after crystal violet staining for 10 minutes.

Role of CSCs in chemoresistance

We further analyzed CSC population in DU145-TXR and PC3-TXR after treatment with paclitaxel, rubone, and their combination using AldeFluor reagent (Stem Cell Technologies) based flow cytometry. Cells were suspended in suspension media and stained by AldeFluor reagent, while a negative control comprising cells treated with ALDH-inhibitor diethylamino-benzaldehyde was included to gate the unspecific staining.

Polymer synthesis, micelle formulation, and characterization

PEG-PCD was synthesized and characterized by 1H NMR as described previously (14). Paclitaxel- and rubone-loaded micelles were prepared by film hydration with 10% theoretical drug loading. Chloroform was evaporated under vacuum, and resulting film was hydrated in 10 mL of PBS and sonicated for 10 minutes using Misonix ultrasonic liquid processor with an amplitude of 30, followed by removing the free drug at 5,000 rpm centrifugation for 5 minutes. Blank or drug-loaded micelles were characterized by measuring particle size using Malvern Zetasizer.

To determine the drug loading and encapsulation efficiency, drug-loaded micelles were dissolved in 1-mL mobile phase composed of 70:30 v/v of acetonitrile and water. Concentrations of paclitaxel and rubone were measured by reverse-phase high-performance liquid chromatography (RP-HPLC, Waters Milford) with a UV detector at 228.6 nm for paclitaxel and 324.3 nm for rubone using a reverse phase C18 column (250 mm × 4.6 mm, Inertsil ODS). We also compared the drug-loading capability of PEG-PCD with commercially available poly (ethylene glycol)-polylactide (PEG-PLA) and 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-(polyethylene glycol) (DSPE-PEG) of similar molecular weights. Paclitaxel and rubone release from PEG-PCD micelles was determined after dialysis (2,500–5,000 Da cutoff) against 50-mL PBS containing 20% ethanol as a cosolvent, which did not dissolve PEG-PCD and break the micellar structure at 37°C in a temperature controlled shaker at the speed of 100 rpm. Sample (1 mL) was taken at specific time points (1, 2, 3, 6, 12, 24, 48, and 96 hours) and replaced with 1-mL PBS containing 20% ethanol. The sample was dissolved with the mobile phase after removing the solvent using a rotary evaporator, followed by determining drug concentration using HPLC.

We further estimated the in vivo stability of PEG-PCD micelles using time-dependent fluorescence resonance energy transfer (FRET) in the presence of 20% FBS. Fifty micrograms of 3,3'-dioctadecyloxacarbocyanine perchlorate (DiO) as a lipophilic fluorescent energy donor and 1,1'-dioctadecyl-3,3,3',3'-tetramethylindocarbocyanine perchlorate (DiI) as an acceptor, 1 mg of paclitaxel and rubone were loaded into 10 mg of PEG-PCD. Emission fluorescence spectra ranging from 490 to 590 nm was recorded at an excitation wavelength of 488 nm (donor excitation) and resulted in a strong emission at 565 nm (acceptor emission). We further compared the micelle stability of PEG-PCD with PEG-PLA and DSPE-PEG after drug loading.

In vivo tumor studies

All animal experiments were performed in accordance with the NIH animal use guideline and protocol approved by the Institutional Animal Care and Use Committee (IACUC) at the University of Nebraska Medical Center (Omaha, NE). To visualize and monitor tumor progression, we developed orthotopic prostate tumor using stably transfected prostate cancer cells with lentivirus encoding GFP and luciferase (LP-hLUC-Lv201-0200, Genecopoeia). A midline incision was made in the lower abdomen of 8 weeks old male nude mice to expose the dorsal prostate lobe, where 30-μL PBS containing 1 × 106 PC3-TXR cells expressing GFP and luciferase (PC3-TXR-GFP-Luc) were injected. Three weeks after tumor cell injection, mice with orthotopic tumor were randomly divided into six groups of 10 animals per group with administration of blank PEG-PCD micelles, paclitaxel (20 mg/kg) loaded PEG-PCD micelles, rubone-loaded PEG-PCD micelles (20 mg/kg), paclitaxel and rubone (10 mg/kg for each drug) along, paclitaxel and rubone–loaded DSPE-PEG micelles, and paclitaxel and rubone–loaded PEG-PCD micelles. These formulations were injected intravenously for five doses every other day. The body weight and tumor luminescence of the mice were recorded once a week. Followed by the last formulation injection, 4 mice in each group were sacrificed and tumors were excised for determining miR-34a expression by RT-PCR. The therapeutic effect of formulation was further determined by IHC for Bax, Ki-67, and miR-34a downstream targets including SIRT1, E-cadherin, and cyclin-D1. The side effects of each formulation were evaluated by hematoxylin and eosin (H&E) staining of the major organs including heart, liver, spleen, lung, and kidney. Other mice were monitored for another two weeks to further evaluate the anticancer efficacy of the formulation.

Statistical analysis

Results were presented as the mean ± SEM from three experiments for in vitro studies and six experiments for in vivo studies. The statistical difference between the two groups was calculated by unpaired Student t test, and a P < 0.05 was considered to be statistically significant.

Rubone upregulated miR-34a and reversed the expression of miR-34a downstream targets in paclitaxel-resistant prostate cancer cell lines

Our objective was to determine whether rubone could serve as a miR-34a modulator to reverse the miR-34a downstream tumor-associated gene expression. Thus, we first determined miR-34a expression in different prostate cancer cell lines. miR-34a expression was markedly downregulated in androgen-refractory DU145, PC3, and paclitaxel-resistant DU145-TXR and PC3-TXR cells compared with androgen-dependent LNCaP, LNCaP-developed C4-2, as well as normal prostate epithelial RWPE-1 cells (Fig. 1A), indicating the role of miR-34a in the initiation and progression of prostate cancer (16). Next, we determined the cytotoxicity of rubone in these cell lines. As shown in Fig. 1B, rubone exhibited significantly higher cytotoxicity in DU145-TXR and PC3-TXR cells, suggesting that rubone had stronger anticancer effect in advanced prostate cancer cells, which had lower miR-34a expression (Fig. 1A). However, rubone did not show obvious toxicity to RWPE-1, LNCaP, and C4-2 cells with high miR-34a expression, indicating rubone induced cytotoxicity through miR-34a-related pathways. Rubone upregulated miR-34a in paclitaxel-resistant DU145-TXR and PC3-TXR cell lines in a dose-dependent manner (Fig. 1C). After evaluating the anticancer effect of rubone, we determined miR-34a downstream target gene expression after rubone treatment. Paclitaxel-resistant cell lines showed more chemoresistance-related SIRT1 expression (17) and less metastasis-related E-cadherin expression (Fig. 1D and E; ref. 18). Rubone significantly reversed the expression of miR-34a downstream gene targets of DU145-TXR and PC3-TXR cell lines (Fig. 1D and E), including E-cadherin, SIRT1, and cyclin D1, whereas E-cadherin expression was not reversed in DU145-TXR cell line. Furthermore, rubone monotherapy promoted apoptosis determined by Bax expression in DU145-TXR and PC3-TXR cell lines. However, rubone showed less effect of reversing miR-34a downstream targets and inducing apoptosis in nonresistant DU145 and PC3 cell lines. These data indicated the downregulation of miR-34a in resistant prostate cancer and suggest that rubone might work as a specific miR-34a modulator to reverse miR-34a expression for treating androgen-refractory metastatic prostate cancer.

Rubone enhanced the anticancer effect of paclitaxel in paclitaxel-resistant prostate cancer cell lines by reversing the expression of miR-34a downstream targets

To determine that miR-34a is an anticancer target for reversing chemoresistance of prostate cancer, we evaluated the gene regulation efficiency of miR-34a mimic and miR-34a inhibitor and their effect on PC3-TXR viability. Using Lipofectamine 2000 as a transfection reagent, miR-34a mimic and inhibitor could upregulate and suppress miR-34a expression in PC3-TXR cell line, respectively (Supplementary Fig. S1A). miR-34a inhibited PC3-TXR cell viability, whereas miR-34a inhibitor promoted cell growth (Supplementary Fig. S1B). We further demonstrated that miR-34a can enhance paclitaxel chemotherapy, whereas miR-34a inhibitor promoted cancer cell viability (Fig. 2A) with paclitaxel treatment in PC3-TXR cell line. Rubone did not enhance the anticancer effect of paclitaxel in chemosensitive DU145 and PC3 cell lines (Fig. 2B and D), but significantly reversed chemoresistance of DU145-TXR and PC3-TXR cell lines (Fig. 2C and E). miR-34a inhibitor reversed the effect of rubone on promoting paclitaxel cytotoxicity (Fig. 2F), indicating rubone promoted the effect of paclitaxel through upregulating the expression of miR-34a. To mimic the complexity of in vivo tumor environment, we determined the anticancer effect of paclitaxel and rubone in 3D tumor model. Paclitaxel and rubone combination therapy inhibited PC3-TXR cell growth and sphere formation in 3D model, including 3D on top (Fig. 2G) and hanging-drop model (Fig. 2H). Similar to rubone monotherapy, paclitaxel and rubone combination therapy more effectively reversed the expression of miR-34a downstream gene expression in DU145-TXR and PC3-TXR cell lines compared with nonresistant cell lines (Fig. 2I and J). Paclitaxel reduced the expression of E-cadherin (19) in DU145 and PC3 cell lines, whereas paclitaxel and rubone failed to reverse E-cadherin in DU145-TXR cell line (Fig. 2I). Thus, rubone could work as a miR-34a modulator to reverse paclitaxel resistance in prostate cancer and restore the expression of miR-34a targeted genes.

Rubone inhibited cell invasion, migration, and CSC population in a p53-independent pathway

Treatment of prostate cancer always fails due to the metastasis and the presence of CSCs, which is highly chemoresistant (20). Thus, we further determined the effect of paclitaxel and rubone combination therapy on the invasion and migration of DU145-TXR and PC3-TXR cells. Rubone alone or its combination with paclitaxel significantly inhibited DU145-TXR and PC3-TXR invasion (Fig. 3A and C) and migration (Fig. 3B and D). Furthermore, rubone or its combination with paclitaxel significantly downregulated aldehyde activity, which is a CSC marker (Fig. 3E). Collectively, our results demonstrated that combination therapy of paclitaxel and rubone significantly reversed chemoresistance, inhibited tumor cell migration and invasion, and decreased the CSC population of androgen-refractory prostate cancer cells.

Previously research claimed that miR-34a and p53 axis regulates miR-34a expression and tumor suppression (21, 22). However, our results indicated that there was no change in p53 expression after rubone treatment, even in PC3-TXR cell line (p53 null; Figs. 1E and 2J). To determine whether rubone upregulates miR-34a in p53-independent pathways including TAp73 (23, 24) and Elk-1 (25, 26), we determined TAp73 and Elk-1 expression after rubone alone or with paclitaxel. As shown in Fig. 3F and G, rubone monotherapy or paclitaxel and rubone combination therapy significantly enhanced TAp73 and Elk-1 expression, suggesting p53-independent pathway played a crucial role in miR-34a upregulation by rubone.

PEG-PCD micelles were effective drug delivery vehicles for paclitaxel and rubone

Polymeric micelles have been widely used for improving the solubility and enhancing the in vivo stability of hydrophobic drugs. In this study, we used PEG-PCD to form micelles for codelivery of paclitaxel and rubone. After chemical synthesis as described previously (14), PEG-PCD polymer was characterized by 1H NMR (Supplementary Fig. S2A) and the particle size distribution of micelles before and after drug loading was measured by dynamic light scattering (Supplementary Fig. S2B). PEG-PCD could form micelles with low polydispersity and drug loading did not affect the particle size. We further compared drug loading and micelle stability of PEG-PCD micelles with two commercially available polymers PEG-PLA and DSPE-PEG. After setting up standard methods of HPLC for measuring paclitaxel and rubone concentrations, we determined the drug loading and micelle stability of PEG-PCD, PEG-PLA, and DSPE-PEG. PEG-PCD had the highest drug loading of 9.70% ± 0.10% and 5.34% ± 0.02% for paclitaxel and rubone (Fig. 4A and B), respectively, compared with PEG-PLA, which showed 4.18% ± 0.03%, 1.51% ± 0.02%, and DSPE-PEG with 3.41% ± 0.36% and 3.58% ± 0.27%. We then determined paclitaxel and rubone release from PEG-PCD micelles. Interestingly, PEG-PCD micelles carrying both paclitaxel and rubone had a slower drug release profile compared with single drug-loaded micelles (Fig. 4C and D). At 12 hours, 54.95% ± 0.90% of paclitaxel and 61.06% ± 2.02% of rubone were released from PEG-PCD micelles with single drug loading, while 39.40% ± 0.22% of paclitaxel and 50.13% ± 3.31% of rubone were released from PEG-PCD micelles with both drugs. We evaluated the cytotoxicity of paclitaxel and rubone formulation on DU145-TXR and PC3-TXR cells. PEG-PCD micelles did not show cytotoxicity to each prostate cancer or normal prostate cell lines (data not shown). Micelle encapsulation decreased cytotoxicity of the combination therapy compared with free drug (Fig. 4E and F) in DU145-TXR and PC3-TXR cell lines, possibly due to slow drug release from the micelles. We also determined the micelle stability by FRET assay (Supplementary Fig. S3). Twenty-percent FBS promoted the degradation and dissociation of micellar structure according to the decreased maximum relative fluorescence unit (RFU) at the wavelength of 568 nm. Micelles formed by PEG-PCD polymer exhibited higher stability compared with DSPE-PEG at each time point, but less stable than PEG-PLA after 6 hours. To summarize the drug loading, release, and stability issues, PEG-PCD micelles might be an effective drug delivery system for in vivo paclitaxel and rubone delivery.

PEG-PCD micellar formulation of paclitaxel and rubone suppressed paclitaxel-resistant prostate tumor growth in vivo

To monitor the tumor growth and metastasis in the orthotopic prostate cancer bearing nude mice, we first transduced PC3-TXR cell line with lentivirus expressing both GFP and luciferase. After injecting the cells into dorsal prostate lobe, tumor development was monitored by intraperitoneally injecting luciferin and recording body weight every week. We also evaluated the efficacy of PEG-PCD micelles by comparing the effect of PEG-PCD micelles with free drugs and DSPE-PEG–loaded drugs. The presence and location of prostate tumor was shown in Fig. 5A. The tumor-inhibitory effect was demonstrated by weekly monitoring the luminescence (Fig. 5B), which indicated the suppressed tumor growth in the micellar combination therapy group. However, PEG-PCD micelles have better therapeutic efficacy than DSPE-PEG micelles. This orthotopic prostate cancer mouse model is very aggressive so that we observed 20% body weight loss during the treatment, while paclitaxel and rubone formulated by PEG-PCD had little effect on body weight loss, suggesting the inhibition of tumor growth in combination therapy group (Fig. 5C). Finally, all the mice were sacrificed to isolate the tumor for measuring miR-34a expression and tumor size. Rubone monotherapy or combination therapy with paclitaxel significantly upregulated miR-34a expression in tumor (Fig. 5D). Tumor luminescence and size at 7th week in PEG-PCD micelle–loaded paclitaxel and rubone group were significantly lower than the other five groups (Fig. 5E). To further demonstrate the anticancer mechanism of rubone in vivo, we isolated the tumor and determined cell proliferation marker and miR-34a downstream targets expression (Fig. 6). Rubone alone or with paclitaxel significantly reversed E-cadherin, cyclin D1, and SIRT1 expression. Rubone monotherapy failed to suppress tumor cell proliferation as indicated by Ki-67 staining, whereas paclitaxel and rubone combination therapy significantly suppressed tumor cell growth compared with paclitaxel monotherapy. We further determined TAp73 and Elk-1 expression in tumor tissue. Rubone alone or combination therapy with paclitaxel significantly upregulated TAp73 and Elk-1 expression. These data indicated that rubone upregulated miR-34a expression in p53-independent pathways in vivo. Normal histology without necrosis was observed in heart and liver (Supplementary Fig. S4). There was significant decrease of hematopoiesis in spleen and acute tubular injury in kidney of paclitaxel-treated group, indicating the side effect of paclitaxel chemotherapy. However, this side effect was reversed in paclitaxel and rubone combination therapy group due to the decreased dose of paclitaxel. Collectively, our results demonstrated that rubone was a potent small-molecule miR-34a modulator to reverse the chemoresistance of advanced androgen-refractory prostate cancer and enhance the therapeutic effect of paclitaxel.

Drug resistance remains the major challenge to cancer chemotherapy even with the discovery of highly efficient anticancer compounds. Furthermore, the skeletal metastasis in advanced prostate cancer patients is the major cause of morbidity and mortality (27). Tumors are composed of bulk cancer cells and small population of CSCs, which are not responsive to most chemotherapeutic agents and result in chemoresistance and tumor recurrence (28). In recent years, miR-34a was found to inhibit CSC growth, metastasis, and chemoresistance by directly repressing the adhesion molecule CD44 (4). The downstream targets of miR-34a, including SIRT1 (29), LEF1 (30), TCF7 (31), AR, and Notch-1 (32), are crucial factors of proliferation, metastasis, and chemoresistance of advanced androgen-refractory prostate cancer. Furthermore, our data indicated that miR-34a was significantly downregulated in advanced prostate cancer, especially in paclitaxel-resistant cells (Fig. 1A). Thus, miR-34a replenishment by systemic delivery using nanoparticles can therefore be developed as a potent therapeutic strategy. However, the off-target effects, in vivo degradation, low efficacy, and high cytotoxicity associated with drug delivery systems of miRNA oligonucleotide still need to be overcome for miRNA-based clinical therapy.

Recently, several small molecules as oncogenic miRNA inhibitors (33, 34) and tumor suppressor miRNA modulators (10) were identified for the inhibition of tumor growth. Among these small molecules, retinoic acid (23), genistein (35), and rubone (10) have been reported to upregulate miR-34a expression in several types of cancer with the mechanism not well characterized. Traditional chemotherapy uses paclitaxel or docetaxel as a monotherapy for inhibiting cancer cell growth, which always fails due to the chemoresistance caused by downregulation of tumor suppressor miRNA. In this study, we present an alternative strategy for fighting paclitaxel-resistant prostate cancer through miR-34a upregulation by employing a combination therapy using rubone as a small-molecule miR-34a modulator. Our data suggest that rubone was nontoxic to normal prostate cells, but toxic to paclitaxel-resistant prostate cancer cells, which had low miR-34a expression (Fig. 1B). For combination therapy with PTX, rubone could reverse the chemoresistance of prostate cancer at low concentration (5 μmol/L). At this concentration, rubone significantly enhanced the cytotoxicity of paclitaxel in paclitaxel-resistant prostate cancer cell lines, whereas did not influence the anticancer effect of paclitaxel in nonresistant cell lines. Extracellular matrix is key regulator of homeostasis and tissue phenotype to form 3D culture assays (36), which allows the phenotypic discrimination between nonmalignant and malignant mammary cells. As some crucial signals are lost when cells are cultured in vitro on 2D plastic flasks (15), 3D model could better mimic the in vivo tumor environment and evaluate the anticancer effect of therapeutic agents. Thus, we determined the antitumor efficacy of paclitaxel and rubone combination therapy in 3D model (Fig. 2G and H), where 3D on top allows the tumor to grow on extracellular matrix (Matrigel) and hanging-drop model can help tumor cell form sphere-like structure without Matrigel. Paclitaxel and rubone combination therapy inhibited tumor cell growth and disturbed tumor morphology in 3D models. These data indicated that rubone could work as a nontoxic, highly specific miR-34a modulator to enhance the therapeutic effect of paclitaxel.

Previous report claimed that rubone inhibited HCC growth in a p53-dependent manner (10). In that research, rubone had no therapeutic effect in Hep3B cells, which did not express p53. Interestingly, our results showed that rubone significantly reversed miR-34a and its downstream target gene expression in p53-null PC3-TXR cells (Fig. 2E and J). Furthermore, rubone enhanced the therapeutic effect of paclitaxel, inhibited the metastasis, and decreased the population of CSCs in PC3-TXR cells (Fig. 3A–E), suggesting that rubone might upregulate miR-34a in a p53-independent pathway. Therefore, we analyzed TAp73 (23, 24) and Elk-1 (25, 26) expression, which were previously reported to be p53-independent miR-34a regulation pathway. Our data showed that TAp73 and Elk-1 were highly upregulated after rubone monotherapy or paclitaxel and rubone combination therapy (Fig. 3F and G). This discrepancy may be explained by the extremely low expression of TAp73 (37) and Elk-1 (38) in Hep3B cells compared with PC3-TXR cells, which means that all known miR-34a regulation pathways are blocked in Hep3B cells. Thus, we conclude that rubone might work as a miR-34a modulator for prostate cancer in a p53-independent manner.

Polymeric micelles can increase aqueous solubility of hydrophobic drugs thereby avoiding the use of toxic solubilizing agents, including DMSO and Cremophor EL. In this study, we synthesized PEG-PCD lipopolymer, which allowed the conjugation of multiple lipid chains to a polycarbonate backbone for the optimization of drug loading. The pendant lipid groups in the lipopolymers could increase the interaction of hydrophobic drugs with the core, improved in vivo micelle stability, and prolonged circulation half-life. Thus, we compared the drug delivery property of our PEG-PCD with two commercially available polymers PEG-PLA and DSPE-PEG. PEG-PCD had higher paclitaxel and rubone loading compared with PEG-PLA and DSPE-PEG, especially when loading both drugs (Fig. 4A and B). The decreased drug release was observed when loading both drugs, indicating drug–drug interaction in the same drug delivery platform could influence the drug delivery property. Although PEG-PCD showed less stability compared with PEG-PLA, PEG-PCD had high stability at the first 6 hours (Supplementary Fig. S3). By summarizing these results, PEG-PCD could be potent drug delivery platform for codelivery of paclitaxel and rubone.

The anticancer efficiency was evaluated in an orthotopic prostate tumor model to mimic the clinical condition and monitor tumor growth in a noninvasive manner. Tumor growth was significantly suppressed after systemic administration of paclitaxel and rubone formulation, including DSPE-PEG and PEG-PCD micelles, compared with other four groups according to the luminescence at each time point (Fig. 5B) and the tumor size at the end of the study (Fig. 5E). PEG-PCD micelles delivered paclitaxel and rubone more efficiently suppressed tumor growth and reversed miR-34a expression compared with free drug and DSPE-PEG micelles, probably because PEG-PCD micelles had better drug loading (Fig. 4A and B) and stability (Supplementary Fig. S3). This can be explained by the chemical structure of PEG-PCD with multiple dodecanal lipid chains attached to the polycarbonate backbone, which could enhance the hydrophobic interaction among the hydrophobic cores. However, DSPE-PEG only has two lipid chains per PEG molecule. Our data also indicated that paclitaxel and rubone combination therapy reversed the downstream target genes of miR-34a through TAp73 and Elk-1 pathways (Fig. 6). However, this orthotopic model using PC3-TXR cell line was very aggressive as we observed severe body weight loss in the progress of tumor (Fig. 5C) and few mice could survive for more than 7 weeks without treatment. Under this severe condition, paclitaxel and rubone combination therapy showed promising therapeutic effect by suppressing the tumor growth and avoid body weight loss. Furthermore, our formulations did not show severe organ toxicity in heart and liver, and combination therapy reversed the side effect of paclitaxel on hematopoiesis in spleen and acute tubular injury in kidney (Supplementary Fig. S4). In some liver and spleen slides of paclitaxel-treated groups, we also observed foamy histiocytes and Kupffer cells full of lipids, which were diminished in combination therapy group (Supplementary Fig. S4).

On the basis of our results, rubone could be a specific miR-34a regulator to reverse miR-34a and the downstream target gene expression for paclitaxel-resistant prostate cancer (Fig. 7). The replenished miR-34a enhanced the anticancer effect of paclitaxel on microtubule disarray, which promoted cell apoptosis and inhibits proliferation. Moreover, this miR-34a replenishment by rubone was in a p53-independent manner in DU145-TXR and PC3-TXR cell lines. Paclitaxel and rubone combination therapy formulated by PEG-PCD micelles could significantly suppress paclitaxel-resistant tumor growth in vivo. This study illustrated the therapeutic potency of rubone as a small molecule miR-34a modulator for the treatment of paclitaxel-resistant prostate cancer.

No potential conflicts of interest were disclosed.

Conception and design: D. Wen, Y. Peng, R.I. Mahato

Development of methodology: D. Wen, Y. Peng, R.I. Mahato

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): D. Wen, Y. Peng, F. Lin, R.I. Mahato

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): D. Wen, F. Lin, R.K. Singh

Writing, review, and/or revision of the manuscript: D. Wen, R.K. Singh, R.I. Mahato

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): D. Wen, F. Lin, R.I. Mahato

Study supervision: R.I. Mahato

We would like to appreciate the help from Dr. Geoffrey Talmon and Dr. Yuri Sheinin from the University of Nebraska Medical Center for analyzing the tissue slides.

D. Wen, Y. Peng, F. Lin, R.K. Singh, and R.I. Mahato were supported by the faculty start-up fund from the University of Nebraska Medical Center. R.I. Mahato is also partly supported by the National Institutes of Health (1R01EB017853 and R01GM113166).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Croce
CM.
Causes and consequences of microRNA dysregulation in cancer
.
Nat Rev Genet
2009
;
10
:
704
14
.
2.
Esquela-Kerscher
A
,
Slack
FJ
. 
Oncomirs - microRNAs with a role in cancer
.
Nat Rev Cancer
2006
;
6
:
259
69
.
3.
Singh
S
,
Chitkara
D
,
Mehrazin
R
,
Behrman
SW
,
Wake
RW
,
Mahato
RI
. 
Chemoresistance in prostate cancer cells is regulated by miRNAs and hedgehog pathway
.
PLoS One
2012
;
7
:
e40021
.
4.
Liu
C
,
Kelnar
K
,
Liu
B
,
Chen
X
,
Calhoun-Davis
T
,
Li
H
, et al
The microRNA miR-34a inhibits prostate cancer stem cells and metastasis by directly repressing CD44
.
Nat Med
2011
;
17
:
211
5
.
5.
Wang
G
,
Liu
G
,
Ye
Y
,
Fu
Y
,
Zhang
X
. 
Upregulation of miR-34a by diallyl disulfide suppresses invasion and induces apoptosis in SGC-7901 cells through inhibition of the PI3K/akt signaling pathway
.
Oncol Lett
2016
;
11
:
2661
7
.
6.
Fujita
Y
,
Kojima
K
,
Hamada
N
,
Ohhashi
R
,
Akao
Y
,
Nozawa
Y
, et al
Effects of miR-34a on cell growth and chemoresistance in prostate cancer PC3 cells
.
Biochem Biophys Res Commun
2008
;
377
:
114
9
.
7.
Kojima
K
,
Fujita
Y
,
Nozawa
Y
,
Deguchi
T
,
Ito
M
. 
MiR-34a attenuates paclitaxel-resistance of hormone-refractory prostate cancer PC3 cells through direct and indirect mechanisms
.
Prostate
2010
;
70
:
1501
12
.
8.
Yao
C
,
Liu
J
,
Wu
X
,
Tai
Z
,
Gao
Y
,
Zhu
Q
, et al
Reducible self-assembling cationic polypeptide-based micelles mediate co-delivery of doxorubicin and microRNA-34a for androgen-independent prostate cancer therapy
.
J Control Release
2016
;
232
:
203
14
.
9.
Wen
D
,
Danquah
M
,
Chaudhary
AK
,
Mahato
RI
. 
Small molecules targeting microRNA for cancer therapy: promises and obstacles
.
J Control Release
2015
;
219
:
237
47
.
10.
Xiao
Z
,
Li
CH
,
Chan
SL
,
Xu
F
,
Feng
L
,
Wang
Y
, et al
A small-molecule modulator of the tumor-suppressor miR34a inhibits the growth of hepatocellular carcinoma
.
Cancer Res
2014
;
74
:
6236
47
.
11.
Wang
Z
,
Wang
N
,
Liu
P
,
Chen
Q
,
Situ
H
,
Xie
T
, et al
MicroRNA-25 regulates chemoresistance-associated autophagy in breast cancer cells, a process modulated by the natural autophagy inducer isoliquiritigenin
.
Oncotarget
2014
;
5
:
7013
26
.
12.
Kong
SY
,
Park
MH
,
Lee
M
,
Kim
JO
,
Lee
HR
,
Han
BW
, et al
Kuwanon V inhibits proliferation, promotes cell survival and increases neurogenesis of neural stem cells
.
PLoS One
2015
;
10
:
e0118188
.
13.
Wen
D
,
Chitkara
D
,
Wu
H
,
Danquah
M
,
Patil
R
,
Miller
DD
, et al
LHRH-conjugated micelles for targeted delivery of antiandrogen to treat advanced prostate cancer
.
Pharm Res
2014
;
31
:
2784
95
.
14.
Li
F
,
Danquah
M
,
Mahato
RI
. 
Synthesis and characterization of amphiphilic lipopolymers for micellar drug delivery
.
Biomacromolecules
2010
;
11
:
2610
20
.
15.
Lee
GY
,
Kenny
PA
,
Lee
EH
,
Bissell
MJ
. 
Three-dimensional culture models of normal and malignant breast epithelial cells
.
Nat Methods
2007
;
4
:
359
65
.
16.
Kopczynska
E.
Role of microRNAs in the resistance of prostate cancer to docetaxel and paclitaxel
.
Contemp Oncol
2015
;
19
:
423
7
.
17.
Kojima
K
,
Ohhashi
R
,
Fujita
Y
,
Hamada
N
,
Akao
Y
,
Nozawa
Y
, et al
A role for SIRT1 in cell growth and chemoresistance in prostate cancer PC3 and DU145 cells
.
Biochem Biophys Res Commun
2008
;
373
:
423
8
.
18.
Fan
L
,
Wang
H
,
Xia
X
,
Rao
Y
,
Ma
X
,
Ma
D
, et al
Loss of E-cadherin promotes prostate cancer metastasis via upregulation of metastasis-associated gene 1 expression
.
Oncol Lett
2012
;
4
:
1225
33
.
19.
Lou
PJ
,
Chen
WP
,
Lin
CT
,
Chen
HC
,
Wu
JC
. 
Taxol reduces cytosolic E-cadherin and beta-catenin levels in nasopharyngeal carcinoma cell line TW-039: Cross-talk between the microtubule- and actin-based cytoskeletons
.
J Cell Biochem
2000
;
79
:
542
56
.
20.
Li
F
,
Mahato
RI
. 
MicroRNAs and drug resistance in prostate cancers
.
Mol Pharm
2014
;
11
:
2539
52
.
21.
Okada
N
,
Lin
CP
,
Ribeiro
MC
,
Biton
A
,
Lai
G
,
He
X
, et al
A positive feedback between p53 and miR-34 miRNAs mediates tumor suppression
.
Genes Dev
2014
;
28
:
438
50
.
22.
Menges
CW
,
Kadariya
Y
,
Altomare
D
,
Talarchek
J
,
Neumann-Domer
E
,
Wu
Y
, et al
Tumor suppressor alterations cooperate to drive aggressive mesotheliomas with enriched cancer stem cells via a p53-miR-34a-c-met axis
.
Cancer Res
2014
;
74
:
1261
71
.
23.
Agostini
M
,
Tucci
P
,
Killick
R
,
Candi
E
,
Sayan
BS
,
Rivetti di Val Cervo
P
, et al
Neuronal differentiation by TAp73 is mediated by microRNA-34a regulation of synaptic protein targets
.
Proc Natl Acad Sci U S A
2011
;
108
:
21093
8
.
24.
Busuttil
V
,
Droin
N
,
McCormick
L
,
Bernassola
F
,
Candi
E
,
Melino
G
, et al
NF-kappaB inhibits T-cell activation-induced, p73-dependent cell death by induction of MDM2
.
Proc Natl Acad Sci U S A
2010
;
107
:
18061
6
.
25.
Frigo
DE
,
Duong
BN
,
Melnik
LI
,
Schief
LS
,
Collins-Burow
BM
,
Pace
DK
, et al
Flavonoid phytochemicals regulate activator protein-1 signal transduction pathways in endometrial and kidney stable cell lines
.
J Nutr
2002
;
132
:
1848
53
.
26.
Christoffersen
NR
,
Shalgi
R
,
Frankel
LB
,
Leucci
E
,
Lees
M
,
Klausen
M
, et al
p53-independent upregulation of miR-34a during oncogene-induced senescence represses MYC
.
Cell Death Differ
2010
;
17
:
236
45
.
27.
Nandana
S
,
Chung
LW
. 
Prostate cancer progression and metastasis: Potential regulatory pathways for therapeutic targeting
.
Am J Clin Exp Urol
2014
;
2
:
92
101
.
28.
Collins
AT
,
Berry
PA
,
Hyde
C
,
Stower
MJ
,
Maitland
NJ
. 
Prospective identification of tumorigenic prostate cancer stem cells
.
Cancer Res
2005
;
65
:
10946
51
.
29.
Duan
K
,
Ge
YC
,
Zhang
XP
,
Wu
SY
,
Feng
JS
,
Chen
SL
, et al
miR-34a inhibits cell proliferation in prostate cancer by downregulation of SIRT1 expression
.
Oncol Lett
2015
;
10
:
3223
7
.
30.
Liang
J
,
Li
Y
,
Daniels
G
,
Sfanos
K
,
De Marzo
A
,
Wei
J
, et al
LEF1 targeting EMT in prostate cancer invasion is regulated by miR-34a
.
Mol Cancer Res
2015
;
13
:
681
8
.
31.
Chen
WY
,
Liu
SY
,
Chang
YS
,
Yin
JJ
,
Yeh
HL
,
Mouhieddine
TH
, et al
MicroRNA-34a regulates WNT/TCF7 signaling and inhibits bone metastasis in ras-activated prostate cancer
.
Oncotarget
2015
;
6
:
441
57
.
32.
Kashat
M
,
Azzouz
L
,
Sarkar
SH
,
Kong
D
,
Li
Y
,
Sarkar
FH
. 
Inactivation of AR and notch-1 signaling by miR-34a attenuates prostate cancer aggressiveness
.
Am J Transl Res
2012
;
4
:
432
42
.
33.
Bose
D
,
Jayaraj
G
,
Suryawanshi
H
,
Agarwala
P
,
Pore
SK
,
Banerjee
R
, et al
The tuberculosis drug streptomycin as a potential cancer therapeutic: inhibition of miR-21 function by directly targeting its precursor
.
Angew Chem Int Ed Engl
2012
;
51
:
1019
23
.
34.
Vo
DD
,
Staedel
C
,
Zehnacker
L
,
Benhida
R
,
Darfeuille
F
,
Duca
M
. 
Targeting the production of oncogenic microRNAs with multimodal synthetic small molecules
.
ACS Chem Biol
2014
;
9
:
711
21
.
35.
Chiyomaru
T
,
Yamamura
S
,
Fukuhara
S
,
Yoshino
H
,
Kinoshita
T
,
Majid
S
, et al
Genistein inhibits prostate cancer cell growth by targeting miR-34a and oncogenic HOTAIR
.
PLoS One
2013
;
8
:
e70372
.
36.
Bissell
MJ
,
Radisky
DC
,
Rizki
A
,
Weaver
VM
,
Petersen
OW
. 
The organizing principle: microenvironmental influences in the normal and malignant breast
.
Differentiation
2002
;
70
:
537
46
.
37.
Wang
J
,
Xie
H
,
Gao
F
,
Zhao
T
,
Yang
H
,
Kang
B
. 
Curcumin induces apoptosis in p53-null Hep3B cells through a TAp73/DNp73-dependent pathway
.
Tumour Biol
2016
;
37
:
4203
12
.
38.
Yue
CH
,
Huang
CY
,
Tsai
JH
,
Hsu
CW
,
Hsieh
YH
,
Lin
H
, et al
MZF-1/elk-1 complex binds to protein kinase calpha promoter and is involved in hepatocellular carcinoma
.
PLoS One
2015
;
10
:
e0127420
.