Induction of Paclitaxel Resistance by ERα Mediated Prohibitin Mitochondrial-Nuclear Shuttling

Paclitaxel is a drug within one of the most promising classes of anticancer agents. Unfortunately, clinical success of this drug has been limited by the insurgence of cellular resistance. To address this, Paclitaxel resistance was modeled in an in vitro system using estrogen treated prostate cancer cells. This study demonstrates that emerging resistance to clinically relevant doses of Paclitaxel is associated with 17-β-estradiol (E2) treatment in PC-3 cells, but not in LNCaP cells. We found that small interfering RNA mediated knockdown of ERα lead to a decrease in E2 induced Paclitaxel resistance in androgen-independent cells. We also showed that ERα mediated the effects of estrogen, thereby suppressing androgen-independent cell proliferation and mediating Paclitaxel resistance. Furthermore, E2 promoted Prohibitin (PHB) mitochondrial-nucleus translocation via directly mediation of ERα, leading to an inhibition of cellular proliferation by PHB. Additionally, restoration of Paclitaxel sensitivity by ERα knockdown could be overcome by PHB overexpression and, conversely, PHB knockdown decreased E2 induced Paclitaxel resistance. These findings demonstrate that PHB lies downstream of ERα and mediates estrogen-dependent Paclitaxel resistance signaling cascades.


Introduction
Prostate cancer is one of the leading causes of death among men in developed countries. The primary treatment for hormonerefractory prostate cancer is taxane-based chemotherapy, including Paclitaxel [1]. Paclitaxel functions by stabilizing microtubule assembly and inhibiting depolymerization, thus causing mitotic arrest or aberrant mitosis. Higher concentrations of Paclitaxel can induce mitotic phase cell death, thereby exerting antitumor effects [2]. Taxane-based therapy often improves patient survival, however, the cancer ultimately develops drug resistance in most patients, leading to recurrence of the cancer, distant metastasis and death [3].
Several pathways are involved in progression to androgen independence in cases of advanced prostate cancer treated with hormone deprivation [4], increasing evidence that estrogen signaling has a major role in prostate cancer development and progression, often associated with estrogen receptor (ER) signaling [5,6,7,8,9]. Genomic modifications of the ER gene have been described, including amplification [8,10] and mutation [11]. Highgrade, primary Gleason grade 4 and 5 tumors revealed ER protein expression in 43% and 62% of cases, respectively [8]. Significant ERa gene expression as measured by mRNA and protein levels was observed in hormone refractory tumors and metastatic lesions, including lymph node and bone metastases [8]. These studies suggest that estrogen can affect prostatic cancerogenesis and neoplastic progression through an ER-mediated process in human prostate tissue. However, the mechanisms underlying estrogen and estrogen receptor signaling in human prostate tissue remain poorly understood.
PHB is ubiquitously expressed in all tissues tested to date and has been shown to have significant effects on cell senescence, cell development and tumor cell suppression [12,13]. Data suggests that PHB can modulate the Rb-E2F transcription complex to repress E2F-mediated transcription and cell proliferation [14]. A significant correlation was found between low tumor cell proliferation and drug resistance. In non-Hodgkin's lymphomas, patients with tumor proliferation of less than 80% were significantly more likely than patients with rates of higher proliferation to be unresponsive to therapy or to fail to achieve a complete response, and tended to have a shorter period free of progression and lower overall survival [15]. Recently, Gregory-Bass et al. showed that repression of PHB in ovarian cancer cells increased their sensitivity to staurosporine [16]. Patel N et al. showed that stable and transient knockdown of PHB in a Paclitaxel-resistant lung cancer cell line or an uterine sarcoma cell line significantly improved sensitivity to Paclitaxel as well as to other chemotherapeutic agents in vitro and in vivo [17]. However, the mechanism underlying this PHB mediated Paclitaxel resistance remains unclear.
Our current work suggests that PHB is a mediator of E2-ERa induced Paclitaxel resistance. This resistance depends on the cellular localization of PHB, rather than on the absolute amount of the protein within the cell. These observations lead to the hypothesis that estrogen and PHB play a role in the development of drug resistance in prostate cancer.

Cell viability analysis
Cells were stained with Hoechst 33258 (5 mg/ml) to visualize nuclei and propidium iodide (PI) (0.2 mg/ml) to detect membrane damage. Cell death was quantified by scoring the number of PI positive cells relative to the total number of nuclei within the same visual field. Cells, 1000 cells per group at minimum, were counted in an unbiased manner and were scored blindly without knowledge of which treatment they had undergone.

Immunoblotting
Cells were harvested at 4uC in Laemmli lysis buffer. After determining the protein content of the cell lysates, the protein extracts were separated by 10% SDS-PAGE, transferred to a PVDF membrane and incubated with primary antibody (ERa, ERb, PHB and VDAC antibodies were from Santa Cruz Biotech, and Tubulin and Histone H1 antibodies were from Abcam). The signal was detected by ECL detection system (GE Healthcare).

DNA Growth Assay
Following treatment of cells, the media was discarded, cells were solubilized for 30 min at 37uC in 0.1% SDS and the amount of DNA was estimated using a Hoechst 33258 microassay, as extensively described previously [19].

Subcellular fractionation
Approximately 10 7 cells were harvested into 10 ml of isotonic fractionation buffer (250 mM sucrose, 0.5 mM EDTA, 20 mM Hepes, and 500 mM Na 3 VO 4 at pH 7.2) supplemented with protease inhibitor cocktail complete (Roche Molecular Biochemicals) and centrifuged at 900 g for 5 min. The pellet was then resuspended in 200 ml fractionation buffer, homogenized with a ball-bearing homogenizer and centrifuged at 900 g for 5 min to remove the nuclei. The post-nuclear supernatant was centrifuged at 20,000 g for 15 min to collect the heavy membrane fraction enriched in mitochondria.

Co-Immunoprecipitation
Cell extracts were prepared by solubilizing 10 7 cells in 1 ml of cell lysis buffer made of 1% Triton X-100, 150 mM NaCl, 20 mM Tris-Cl at pH 7.4, 1 mM EDTA, 1 mM EGTA, 1 mM Na3VO4, 2.5 mM pyrophosphate, 1 mM glycerol phosphate and protease inhibitor mixture for 10 min at 4suC. After brief sonication, the lysates were cleared by centrifugation at 15,000 g for 10 min at 4uC, the cell extract was immunoprecipitated with 6 mg of antibodies against ERa or PHB (antibodies were from Santa Cruz Biotech), and incubated with 100 ml of protein G plus protein A-agarose for 12h at 4uC by continuous inversion. Immunocomplexes were pelleted, washed 4 times, boiled in Laemmli buffer and analyzed by Western blot.

Constructs
The plasmids pCDNA3.1-Prohibitin and pCDNA3.1-ERa were made by inserting Human Prohibitin or ERa cDNA into a pCDNA3.1 expression vector. Constructs were transfected into cells using Lipofectamine 2000.

Statistical analysis
The statistics in the graphs represent the means with 6 S.E. bars of at least three independent experiments. Each group was compared to the control using Student's t test. The significance is indicated as follows: * denotes p,0.05, ** denotes p,0.01, and *** denotes p,0.001.

Estrogen inhibits Paclitaxel induced androgenindependent prostate cancer cell death
Human prostate cancer is considered a paradigm of an androgen-dependent tumor. However, the role of estrogen in malignant prostate cancer appears to be equally important. In animal model systems, estrogens, combined with androgens, appear to be required for the malignant transformation of prostate epithelial cells [20]. Although the mechanisms underlying the hormonal induction of prostate cancer in vivo remain uncertain, there is evidence to support that long term administration of androgens and estrogens results in an estrogenic environment in rat prostates and the ensuing development of cancer [20].
To examine whether estrogen is sufficient to regulate the progress of prostate cancer, we first examined the sensitivity of LNCaP cells (androgen-sensitive human prostate adenocarcinoma) and PC3 cells (androgen-independent prostate cancer) for Paclitaxel. We found that Paclitaxel induced the death of both LNCaP and PC3 cells ( Fig. 1A and C). E2 was used in this study as a representative of estrogen, because E2 is the most potent estrogen normally found in the circulation. Interestingly, we also found that E2 inhibited Paclitaxel induced PC3 cell death ( Fig. 1C and D), yet had no effect on Paclitaxel induced LNCaP cell death ( Fig. 1A and B). These results confirm that estrogen inhibits Paclitaxel induced cell death in androgen-independent prostate cancer cells.
ERa overexpression mediates the estrogen induced Paclitaxel resistance of PC3 cells Previously, we found that E2 treatment reduces the sensitivity of PC3 cells to Paclitaxel. Estrogens have been reported to suppress proliferation of cultured prostate cancer cells [21]. Two major estrogen receptor types, ERa and ERb, are expressed in both normal and diseased human prostate, albeit with differing cellular localization [22,23]. Since these two receptors also display differences in ligand binding, heterodimerization, transactivation and estrogen response element activity, it is likely that assessing and changing their expression may be critical to ultimately determine the effects of estrogen on prostate cancer cells [9].
First, the expression levels of ERa and ERb in LNCaP and PC3 cells were measured. Consistent with previous studies [23], mRNA and protein expression levels of ERb were equal in LNCaP and PC3 cells ( Fig. 2A and C), whereas ERa mRNA and protein expression levels were significantly higher in PC3 cells than LNCaP cells ( Fig. 2A and C). Furthermore, we confirmed that treatment of LNCaP and PC3 cells with E2 or Paclitaxel did not affect the expression levels of ERa and ERb ( Fig. 2B and D). Next, to identify the specific effect of ERa and ERb in Paclitaxel resistance, siRNAs that target human ERa or ERb were developed, and their efficacy verified by measuring endogenous ERa or ERb in PC3 cells following knockdown. Immunoblotting analysis revealed that ERa and ERb siRNAs specifically abolished the expression of endogenous ERa and ERb, respectively, in PC3 cells (Fig. 2E), demonstrating high selectivity and efficacy. Interestingly, we found ERa siRNA, but not ERb siRNA, significantly restored the sensitivity of PC3 cells to Paclitaxel induced death (Fig. 2F). To determine whether ERa is sufficient to induce resistance to Paclitaxel, the effect of ERa on LNCaP cells that had been treated with E2 and Paclitaxel was examined. It was found that overexpression of ERa in LNCaP cells resulted in E2mediated resistance to Paclitaxel induced cell death (Fig. 2G). Thus, ERa is both necessary and sufficient for E2-mediated Paclitaxel resistance.
Estrogen activates ERa to suppress PC3 cell proliferation and mediate resistance to Paclitaxel Evidence has been accumulating that suggests that the expression level of ERa affects the efficacy of chemotherapy. One clinical trial reported that the curative effect of Paclitaxel plus cyclophosphamide adriamycin chemotherapy was higher for ERa negative patients than ERa positive [24]. Furthermore, in breast cancer patients, MaeharaY et al. found that ERa negative breast  cancer is more sensitive to chemotherapy drugs than ERa positive breast cancer [25]. It is thought that Paclitaxel could induce mitotic phase death in cancer cells, thereby exerting an antitumor effect [2]. Based on these data, we hypothesize that estrogen activates ERa, which suppresses PC3 cell proliferation and thus mediates the cell's resistance to Paclitaxel.
To determine whether estrogen could inhibit the proliferation of prostate cancer cells, we treated LNCaP and PC3 cells with E2. We found that physiological E2 concentrations did not stimulate or inhibit growth of LNCaP (Fig. 3A). However, PC3 cells displayed a significant, E2 dose-dependent inhibition of growth with a maximal effect at 1 mM E2 (32.2% with respect to control) (Fig. 3B). The effect of E2 was also evident after treatment with concentrations higher than 0.5 nM E2 for 96 h (Fig. 3B).
To identify the role of ERa and ERb in E2 induced PC3 cell suppression, siRNAs targeting human ERa and ERb were used.
We found ERa siRNA, but not ERb siRNA, significantly inhibited the suppression of PC3 cell proliferation by E2 (Fig. 3C). These results suggest that estrogen suppresses PC3 cell proliferation and mediates Paclitaxel resistance through activation of ERa.

E2 promotes PHB mitochondrial-nucleus translocation, thus inhibiting cell proliferation
The androgen receptor is currently the major hormonal target for prostate cancer treatment. However, increasing evidence suggests that estrogen signaling also has an important role in tumor development and progression. Some variants of genes involved in estrogen metabolism, including PHB [26] and estrogen receptors [9], are associated with an increased risk of prostate cancer. Previously, we reported that PHB is an important regulator of transit through the cell cycle [18]. While delineating were treated with 100 nM of E2 for 96h, then transfected with vector or ERa expression plasmids. Twenty four hours post-transfection, 50 nM of Paclitaxel was added to the media for 24h and the level of cell death was quantified (left) as in Figure 1 and then the expression of ERa was monitored using Western blotting (right). The values represent the mean 6 S.E. of at least three independent experiments. * denotes p,0.05, ** denotes p,0.01, and *** denotes p,0.001. doi:10.1371/journal.pone.0083519.g002 how estrogen inhibits the proliferation of PC3 cells, we hypothesized that it is through PHB that estrogen mediates its effects. Work by our lab, as well as others, has found that in prostate cancer cells, there is an increased PHB expression in response to stimulation by cholesterol [18]. In this study, we found PHB protein levels remained constant in PC3 and LNCaP cells during E2 treatment (Fig. 4A). Using PHB siRNA, significant knockdown of PHB expression was achieved (Fig. 4B) and this PHB knockdown inhibited the E2 induced suppression of PC3 cell proliferation (Fig. 4C). Furthermore, knockdown of PHB significantly restored the sensitivity of PC3 cells to Paclitaxel induced death (Fig. 4D), suggesting a critical role for PHB in Paclitaxel resistance of PC3 cells.
To determine whether PHB is sufficient to induce Paclitaxel resistance, the effect of PHB on LNCaP cells treated with E2 and Paclitaxel was examined. Forced overexpression of PHB in LNCaP cells could generate E2-mediated Paclitaxel resistance to cell death (Fig. 4E). Thus, PHB itself was sufficient for E2mediated Paclitaxel resistance.
PHB has been suggested to be localized in the nucleus, to modulate transcriptional activity by interacting with various transcription factors, including nuclear receptors, and to suppress cell proliferation [27,28,29]. Also, it has been suggested that PHB may be able to translocate between the mitochondria and the nucleus [30,31]. However, the mechanisms of PHB function in PC3 prostate cancer cells have not been delineated. Here, we found that the majority of PHB localized in the mitochondrial fraction in both PC3 and LNCaP cells (Fig. 4F). Following E2 treatment, PHB levels were elevated in the nuclear fractions and decreased in the mitochondrial fractions as compared to untreated PC3 cells (Fig. 4F left). However, E2 did not affect the localization of PHB in LNCaP cells (Fig. 4F right). These results indicate that E2 promotes PHB mitochondrial-nuclear translocation, thus inhibiting cell proliferation.

ERa directly mediates PHB mitochondrial-nuclear shuttling
Previous studies have shown that PHB is mainly localized and functions in the mitochondria, and that mitochondrial PHB translocates to the nucleus in the presence of ERa [30,31]. However, it remains unclear how PHB is delivered to the nucleus in prostate cancer cells. We hypothesized that the nuclear redistribution of PHB is driven by ERa.
To investigate whether ERa mediated the mitochondrialnuclear translocation of PHB, PC3 cells were treated with or without E2, as well as with ERa or ERb siRNA. PHB localization was qualitatively assessed to determine if treatment induced translocation to the nucleus, rather than remaining in the mitochondria. We found that E2 induced the translocation of PHB from the mitochondria to the nucleus, and that this translocation was inhibited by ERa siRNA (Fig. 5A). By contrast, ERb siRNA did not affect the translocation of PHB (Fig. 5B), suggesting that PHB mitochondrial-nuclear shuttling occurs in an ERa-dependent, but ERb-independent manner.
To further confirm whether ERa could directly mediate the translocation of PHB, we analyzed the immunoprecipitated pellet of endogenous PHB for the presence of estrogen receptors. In PC3 cells, immunoblot analysis of immunoprecipitated PHB detected the presence of ERa (Fig. 5C) but no ERb (data not shown). Also, PHB was present following a reciprocal immunoprecipitate using the antibody against ERa (Fig. 5C). Furthermore, we found that PHB interacted with ERa in both the mitochondrial and nuclear fractions upon E2 treatment ( Figure 5D and 5E). In LNCaP cells, no interaction was detected between PHB and ERa (because the expression level of ERa in LNCaP cell is very low) or ERb (data not shown). These results indicate that PHB could physically interact with ERa, and ERa could directly mediate the PHB mitochondrial-nuclear shuttling.

PHB acts downstream of ERa to mediate resistance to Paclitaxel
Based on the importance of ERa and PHB in E2 induced Paclitaxel resistance of PC3 cells, we hypothesized that PHB might mediate the resistance induced by estrogen and ERa. ERa and PHB were manipulated using a combination of knockdown and overexpression approaches to investigate their functional relationship. The ERa siRNAs, known to abolish ERa protein levels in PC3 cells, significantly blocked E2 induced Paclitaxel resistance of PC3 cells (Fig. 2D). Similarly, the knockdown of endogenous PHB also markedly inhibited E2 induced Paclitaxel resistance of PC3 cells (Fig. 4D and Fig. 6) and overexpression of ERa failed to overcome the effect of PHB siRNAs (Fig. 6). Furthermore, overexpression of PHB, combined with transfection of ERa siRNAs, reversed the phenotype normally seen following suppression of ERa (Fig. 6). These findings demonstrate that PHB lies downstream of the E2/ERa-dependent Paclitaxel resistance signaling cascade.

Discussion
Patients with castration-resistant prostate cancer are at a high risk of death. Treatment of these cancers includes second-line hormones, novel agents and chemotherapy with taxanes, such as Paclitaxel and docetaxel. Paclitaxel is a widely used, effective agent in the treatment of a variety of human cancers. As with many other chemotherapeutic agents, however, resistance to Paclitaxel remains a limiting factor for its clinical efficacy. Despite this limitation, Paclitaxel remains at the frontline of cancer therapy and has stimulated a concerted effort to understand the molecular mechanisms of Paclitaxel resistance [32]. In the current study, we demonstrate Paclitaxel resistance is associated with estrogen treatment in androgen-independent PC3 cells, but not in androgen-sensitive LNCaP cells. We also show that estrogen activates ERa to suppress PC3 cell proliferation and mediate Paclitaxel resistance. Moreover, ERa directly mediates the mitochondrial-nuclear shuttling of PHB and inhibits cell proliferation. Combined, these findings demonstrate that estrogen activates ERa and promotes PHB mitochondrial-nuclear translocation, leading to resistance of androgen-independent prostate cancer cells to Paclitaxel.
Experimental androgen-deprivation therapy (ADT) for prostate cancer in the form of estrogen treatment was first reported nearly 70 years ago [33]. The results of these initial studies were translated to the clinic, where ADT was shown to slow the inexorable progression of prostate cancer [34]. Unfortunately, however, the development of so-called castrate-resistant prostate cancer limits the effects of ADT [35]. In fact, it is controversial as to whether estrogen has inhibitory effects on prostate cancer. On one hand, estrogen may be effective as a second line hormonal treatment for patients with androgen-independent prostate cancer and may improve patient survival [36,37]. On the other hand, increasing evidence shows that estrogen signaling has a major role in prostate cancer development and progression [6,7,8]. At present, there is no clear mechanism that explains these two disparate conclusions. We demonstrate here that estrogen, through repression of cell proliferation, induces Paclitaxel resistance in androgen-independent PC3 cells, but not androgensensitive LNCaP cells, in an ERa/PHB-dependent pathway. siRNA or NC RNA. Twenty four hours post-transfection, Paclitaxel was added to the media at the indicated concentrations for 24h, and the level of cell death was quantified as described in Figure 1. (E) 100 nM of E2 was added to the media for 96h, and then LNCaP cells were transfected with either vector or PHB expression plasmids. Twenty four hours post-transfection, 50 nM of Paclitaxel was added to the media for 24h and the level of cell death was quantified as described in Figure 1. (F) E2 promoted PHB mitochondrial-nucleus translocation. 100 nM of E2 was added to the media for 96h, then PC3 or LNCaP cell mitochondria (M) and nuclei (N) were separated and analyzed by Western blot using PHB, Histone H1 (nucleus marker), VDAC (mitochondrial marker) and Tubulin (cytoplasm marker) antibodies. T (total cell lysates). Results are representative of three independent experiments. The values represent the mean 6 S.E. of at least three independent experiments. * denotes p,0.05; ** denotes p,0.01; *** denotes p,0.001. doi:10.1371/journal.pone.0083519.g004 Figure 5. ERa directly mediated PHB mitochondrial-nuclear shuttling. (A and B) ERa mediated PHB mitochondrial-nucleus translocation. PC3 cells were transfected with siRNAs specific to (A) era or (B) erb, or NC RNA. Twenty four hours post-transfection, 100 nM of E2 was added to the media for 96h. The PC3 cell mitochondria (M) and nuclei (N) were separated and analyzed by Western blot using PHB, ERa Histone H1 (nucleus marker), VDAC (mitochondrial marker) and Tubulin (cytoplasm marker) antibodies. T (total cell lysates). (C) ERa directly associates with PHB. 100 nM of E2 was added to the media for 96h, then PC3 cell lysates were immunoprecipitated (IP) using PHB antibody and analyzed by Western blot (WB) using the indicated antibodies (left panel). PC3 lysates were immunoprecipitated with ERa antibody, and PHB and ERa levels were analyzed by Western blot (middle panel). Equal amounts of total input PHB and ERa (Input) were used for immunoprecipitations for each condition (right). (D) ERa directly associates with PHB in mitochondria. PC3 cells were treated as in C, then the PC3 cell mitochondria were separated and immunoprecipitated as in C. (E) ERa directly associates with PHB in nucleus. PC3 cells were treated as in C, then PC3 cell nuclei were separated and immunoprecipitated as in C. Results are representative of three independent experiments. doi:10.1371/journal.pone.0083519.g005 Several studies have shown that repression of PHB in ovarian cancer cells and lung carcinoma cells improves their sensitivity to staurosporine and Paclitaxel [17,32]. Here, we show that PHB knockdown restored Paclitaxel sensitivity to resistant PC3 cells. This restored sensitivity to Paclitaxel following PHB knockdown results from increased proliferation of PC3 cells. Interestingly, we observed a recovery in Paclitaxel sensitivity after repression of PHB in estrogen induced Paclitaxel resistant cells, suggesting that PHB may sufficient for the onset and maintenance of estrogen induced Paclitaxel resistance. Together, our results suggest that PHB reduction can improve Paclitaxel sensitivity in androgenindependent and taxane-resistant prostate cancer cells.
Our data establishes for the first time that estrogen induces taxane-resistance of androgen-independent prostate cancer cells in an ERa/PHB-dependent mechanism. Of particular interest is how this mechanism relates to ADT. In cases where castrate-resistant prostate cancer limits the effects of ADT, estrogen could be used as a hormonal treatment for prostate cancer, while simultaneously repressing the activity of PHB to avoid the induction of the drug resistant effect of estrogen. This method may recover the efficacy of taxane in the clinic. Furthermore, alterations in PHB levels or interference with PHB localization may also result in increased sensitivity of other types of tumors to taxane treatment.  . PHB acts downstream of ERa to mediate resistance to Paclitaxel. 100 nM of E2 was added to the media for 96h, and then PC3 cells were transfected with the indicated siRNAs, siRNAs plus ERa or PHB expression plasmids, or NC RNA. Twenty four hours post-transfection, Paclitaxel was added to the media at the indicated concentrations for 24h, and the level of cell death was quantified as described in Figure 1. The values represent the mean 6 S.E. of at least three independent experiments. * denotes p,0.05; ** denotes p,0.01; *** denotes p,0.001. doi:10.1371/journal.pone.0083519.g006