Address correspondence and reprint requests to Hiroaki Adachi, MD, PhD, or Gen Sobue, MD, PhD, Department of Neurology, Nagoya University Graduate School of Medicine, 65 Tsurumai-cho, Showa-ku, Nagoya 466-8550, Japan. E-mails: email@example.com or firstname.lastname@example.org
Spinal and bulbar muscular atrophy (SBMA) is an inherited motor neuron disease caused by the expansion of a polyglutamine (polyQ) tract within the androgen receptor (AR) gene. The pathologic features of SBMA are motor neuron loss in the spinal cord and brainstem, and diffuse nuclear accumulation and nuclear inclusions of mutant AR in residual motor neurons and certain visceral organs. AR-associated coregulator 70 (ARA70) was the first coregulator of AR to be identified, and it has been shown to interact with AR and increase its protein stability. Here, we report that genistein, an isoflavone found in soy, disrupts the interaction between AR and ARA70 and promotes the degradation of mutant AR in neuronal cells and transgenic mouse models of SBMA. We also demonstrate that dietary genistein ameliorates behavioral abnormalities, improves spinal cord and muscle pathology, and decreases the amounts of monomeric AR and high-molecular-weight mutant AR protein aggregates in SBMA transgenic mice. Thus, genistein treatment may be a potential therapeutic approach for alleviating the symptoms of SBMA by disrupting the interactions between AR and ARA70.
Spinal and bulbar muscular atrophy (SBMA) is a slowly progressive adult-onset motor neuron disease caused by a CAG repeat expansion within the androgen receptor (AR) gene, which results in an expanded polyglutamine (polyQ) tract (La Spada et al. 1991). The clinical symptoms of SBMA include proximal muscular atrophy, tremor, fasciculations, and bulbar involvement (Kennedy et al. 1968), and correlates with a loss of motor neurons and nuclear accumulation of the mutant AR in the spinal cord and brainstem and with obvious neurogenic changes in the skeletal muscle (Sobue et al. 1989; Adachi et al. 2007a). SBMA is a member of family of polyQ diseases that include Huntington's disease (HD), dentatorubral-pallidoluysian atrophy, and spinocerebellar ataxia (SCA) known as SCA1, SCA2, SCA3, SCA6, SCA7, and SCA17 (Orr and Zoghbi 2007). The expanded CAG repeats in exonic regions of DNA are thought to confer a toxic gain of function to the mutant protein. Because AR has specific natural ligands (i.e., testosterone and its more potent derivative dihydrotestosterone), the pathogenesis of SBMA is triggered by the binding of the mutant AR to its ligands. The successful anti-androgen therapy in SBMA mouse models has translated into clinical trials, which did not show definite amelioration on motor functions, although swallowing function improved in a subgroup of patients whose disease duration was less than 10 years (Katsuno et al. 2010; Fernandez-Rhodes et al. 2011).
The AR is a ligand-dependent nuclear transcription factor that belongs to the nuclear receptor superfamily. To initiate transcription, AR interacts with coregulators such as AR-associated coregulator 70 (ARA70) (Yeh and Chang 1996). ARA70, which was the first identified AR coregulator, has been shown to interact with and stabilize the AR protein (Hu et al. 2004). It has also been shown that ASCJ-9 disassociates AR and ARA70, which results in the increased degradation of AR, the suppression of pathogenic AR accumulation, and the amelioration of neuromuscular symptoms in a mouse model of SBMA (Yang et al. 2007). Thus, disrupting the interactions between coregulators, such as ARA70 and AR, could represent a novel therapeutic approach for the treatment of SBMA.
Genistein is a soy isoflavone that exerts many biological functions in living cells and has low oral toxicity in mammals (McClain et al. 2005; Michael Mc Clain et al. 2006). Genistein is a potent broad-spectrum tyrosine kinase inhibitor (Akiyama et al. 1987) that has been shown to attenuate the growth of cancer cells (Levitzki and Gazit 1995; Yan et al. 2010). Recent studies have indicated that this isoflavone may have therapeutic potential in genetic diseases for which there are currently no effective treatments, such as cystic fibrosis (CF) and mucopolysaccharidosis. Genistein has been shown to cross the blood–brain barrier in the rat (Tsai 2005) and mouse (Liu et al. 2008), and it has been shown to exert therapeutic effects on neurodegeneration and neuromuscular diseases (Trieu and Uckun 1999; Bang et al. 2004; Liu et al. 2008; Messina et al. 2011).
In this study, we examined the effects of genistein on cultured cells and transgenic mouse models of SBMA. Our results showed that genistein treatment promotes the dissociation of AR from ARA70 and thus induces AR protein degradation. Most importantly, through promoting mutant AR degradation, genistein inhibits neuronal nuclear accumulation of mutant AR in neurons and considerably ameliorates the motor phenotypes in mouse models of SBMA.
Materials and methods
DNA plasmid construction
cDNA expression plasmids encoding full-length AR were constructed by subcloning AR cDNA, which was derived from pSP64–AR97 (Kobayashi et al. 2000), into the pCR3.1 mammalian expression vector (Life Technologies, Carlsbad, CA, USA). VP16–ARA70 was constructed by replacing the SgfI–PmeI barnase fragment in the pFN10A (ACT) Flexi vector (Promega, Madison, WI, USA) with the SgfI–PmeI ARA70 fragment from the pf1KB-ARA70 vector (Promega). GAL4DBD–AR was constructed by replacing the SgfI–PmeI barnase fragment in the pFN11A (Bind) Flexi vector (Promega) with the SgfI–PmeI AR fragment from the pf1KB-AR vector (Promega). The GAL4DBD–AR S515D vector was generated using the KOD-Plus-Mutagenesis kit (Toyobo, Osaka, Japan) with the forward primer 5′-GATCCCACTTGTGTCAAAAGCGA-3′ and the reverse primer 5′-GGGATAGGGCACTCTGCTCACCAT-3′. We also constructed an (ARE)2-LUC reporter vector to analyze the transactivation activity of AR.
Reagents and cell culture
The following antibodies were used in this study: anti-AR (N-20 or H280; Santa Cruz Biotechnology, Santa Cruz, CA, USA); anti-ARA70 (Santa Cruz Biotechnology); anti-GR (Santa Cruz Biotechnology); anti- RXRα (Santa Cruz Biotechnology); anti-phospho-p44/42 MAPK (Cell Signaling Technology, Danvers, MA, USA); anti-p44/42 MAPK (Cell Signaling Technology); and anti-Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (Cell Signaling Technology). Genistein for cell culture was purchased from Sigma (Sigma, St. Louis, MO, USA). For cultured cells, we diluted a 100-mM stock solution of genistein in dimethylsulfoxide into fresh medium to achieve final concentrations of 25–100 μM. The HaloTag-ARA70 vector and the pRL-TK Renilla luciferase reporter vector were purchased from Promega. LNCaP cells were cultured in RPMI-1640 supplemented with 10% Fetal bovine serum (FBS), and Neuro2a cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% FBS.
Generation, maintenance, treatment, and behavioral assessment of transgenic mice
The generation and maintenance of SBMA transgenic mice have been described previously (Katsuno et al. 2002; Waza et al. 2005). We performed all animal experiments in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and those were approved by the Nagoya University Animal Experiment Committee. For mouse models, treatments with genistein (Tokyo Chemical Industry Co., Ltd., Tokyo, Japan) were initiated when mice reached 6 weeks of age and continued until the age of 25 weeks. SBMA transgenic mice received a soy-free diet (Central Institute for Experimental Animals Japan, Inc, Kawasaki, Japan) or an identical diet supplemented with genistein at a concentration of 2.5 g/kg of food. This diet resulted in an effective genistein dose of 250 mg/kg/d. The mouse rotarod task and grip strength were performed as described previously (Adachi et al. 2007b). The investigators who involved in the behavioral assessments were blinded to the treatment conditions.
GAL4 reporter assay for AR/ARA70 binding
Neuro2a cells were seeded in 96-well plates in antibiotic-free Dulbecco's modified Eagle's medium supplemented with 10% FBS. Neuro2a cells were transiently cotransfected with VP16-ARA70 and GAL4DBD–AR or GAL4DBD–AR S515D. In this system, physical interactions between the two encoded proteins result in the activation of the cotransfected, GAL4-driven luciferase reporter (Promega), which is monitored using a dual-luciferase reporter assay system (Promega). The GAL4DBD–AR or GAL4DBD–AR S515D vector also encoding Renilla luciferase was used to normalize the transfection efficiencies. Twenty-four hours after transfection, the medium was changed, and the cells were treated with vehicle or genistein for 24 h in the presence of 10 nM dihydrotestosterone (DHT, Sigma).
HaloTag pull-down assay
Neuro2a cells were cotransfected with pCR3.1-AR-97Q and HaloTag-ARA70 full-length plasmids (Promega) using Lipofectamine 2000 (Life Technologies). After a 12-h treatment with genistein (or vehicle) in the presence of the proteasome inhibitor lactacystin (5 μM; Sigma) and DHT (10 nM), whole-cell extracts of the transfected Neuro2a cells were prepared using lysis buffer. To detect interactions between AR and ARA70, we incubated whole-cell extracts with Magnetic HaloTag Beads (Promega) in lysis buffer overnight at 4°C. After washing with lysis buffer, the bound protein fractions were digested with ProTEV protease (Promega). The eluted proteins were then subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis followed by immunoblot analysis using anti-ARA70 and anti-AR antibodies.
Protein expression analysis
Cells were lysed using CelLytic-M Mammalian Cell Lysis/Extraction Reagent (Sigma) and centrifuged at 15 000 g for 15 min at 4°C. Tissues from 12-week-old mice were homogenized in CelLytic-M (Sigma) and centrifuged at 2500 g for 15 min at 4°C. Western blotting and densitometric analyses were performed as described previously (Waza et al. 2005; Tokui et al. 2009).
Luciferase assays were performed using Neuro2a cells that were transiently transfected with the pCR3.1-AR-97Q construct together with the luciferase (ARE)2-LUC reporter vector and the pRL-TK Renilla luciferase reporter construct. AR-97Q transactivation was measured in the presence and absence of DHT using a luciferase assay and normalized to pRL-TK Renilla luciferase reporter activity.
Immunohistochemistry and histopathology
Tissues for immunohistochemistry and histopathology were prepared as previously described (Katsuno et al. 2002; Waza et al. 2005). Tissue sections were incubated with an expanded polyQ-specific antibody (1 : 10 000; 1C2; Merck Millipore, Darmstadt, Germany) and a Glial fibrillary acidic protein-specific antibody (1 : 1000; Roche, Mannheim, Germany). We air-dried 6-μm-thick paraffin-embedded sections of the gastrocnemius muscles and stained them with hematoxylin and eosin. To quantify the 1C2-positive cells, the 1C2-positive cells were counted within the quadriceps femoris muscle and thoracic spinal cord for each individual mouse as previously described (Adachi et al. 2001).
The data were analyzed using the unpaired t-tests for two-group comparisons, anovas with Dunnett's post hoc tests for multiple comparisons, and Kaplan–Meier and log-rank tests for analyses of the survival rate. We performed statistical analyses using Statview software (version 5, Hulinks Inc., Tokyo, Japan) and SPSS (Statistical Package for the Social Sciences, v16, IBM Corporation, Armonk, New York, USA) statistics for Windows. P values less than 0.05 were considered significant.
The effects of genistein on androgen receptor expression in vitro
To address whether genistein promotes AR degradation, we treated LNCaP cells expressing the endogenous AR or Neuro2a cells transiently expressing mutant (AR-97Q) AR for 48 h with the indicated concentrations of genistein. Immunoblot analyses showed that genistein promotes AR protein degradation in a dose-dependent manner (Fig. 1a and b). To determine whether genistein promotes AR degradation through the proteasome pathway, we used lactacystin, a proteasome inhibitor. Lactacystin treatment resulted in a complete suppression of the reduction in AR induced by genistein, thus suggesting that the degradation of AR induced by genistein is dependent on the proteasome (Fig. 1c). AR is a client protein of Hsp90, and we have previously demonstrated that Hsp90 inhibition can promote the clearance of AR through the proteasome (Waza et al. 2005; Tokui et al. 2009). However, genistein had minimal effects on the expression levels of other Hsp90 substrates, such as the glucocorticoid receptor (GR) and the retinoid X receptor-α (RXRα; Fig. 1d), which indicates that genistein might promote AR protein clearance through an Hsp90-independent pathway.
Genistein disrupts the association of the androgen receptor with its coregulator, ARA70
Because ARA70 is known to associate with the AR protein and enhance its stability, treatments that disrupt the interaction of AR and ARA70 could induce AR clearance (Yang et al. 2007). We next examined the effects of genistein treatment on the status of the AR–ARA70 complex. We first used a GAL4 reporter assay for AR/ARA70 binding to demonstrate the physical interaction between AR and ARA70 in Neuro2a cells. The results from the GAL4 reporter assay revealed that genistein promotes the dissociation of AR from its coregulator, ARA70, in a dose-dependent manner (Fig. 2a). This interaction was confirmed using a HaloTag pull-down assay in cultured Neuro2a cells that were cotransfected with AR-97Q and the HaloTag-human ARA70 vector. Pull-down assays demonstrated that high concentrations of genistein decrease the association between AR and ARA70 (Fig. 2b). We were able to further demonstrate that genistein inhibits AR transactivation via its ability to dissociate AR from ARA70 in Neuro2a cells (Fig. 2c). Collectively, these results demonstrate that genistein can disrupt the interaction between AR and ARA70.
Genistein disrupts AR and ARA70 interactions through p44/42 MAP kinase inhibition
Previous work has shown that the association between AR and ARA70 can be disrupted through the inhibition of p44/42 MAP kinase (MAPK) (Yeh et al. 1999). Because genistein is a potent tyrosine kinase inhibitor, we hypothesized that genistein treatment results in the dissociation of AR from ARA70 via p44/42 MAPK inhibition. We first demonstrated that genistein inhibits p44/42 MAPK activation in Neuro2a/AR 97Q cells (Fig. 3a). We then demonstrated that the interaction between AR and ARA70 could be disrupted by treatment with PD0325901 (Fig. 3b), which suppresses p44/42 MAPK activation (Fig. 3c) by preventing the activation of its upstream activator MAPK kinase-1. p44/42 MAPK has been shown to phosphorylate AR at serine 515 (Yeh et al. 1999; Ponguta et al. 2008), and phosphorylation of this residue by p44/42 MAPK has been reported to facilitate AR–ARA70 interactions (Yeh et al. 1999).
To determine whether genistein promotes the dissociation of AR from ARA70 through the dephosphorylation of AR at serine 515, we examined the effects of genistein on AR variants in which serine 515 was substituted with aspartate, which mimics constitutive phosphorylation, in Neuro2a cells. Our results demonstrated that genistein treatment had no significant effects on the interaction between the AR phosphomimetic variant (S515D) and ARA70 using a GAL4 reporter assay for AR/ARA70 binding (Fig. 3d), suggesting that genistein disrupts AR and ARA70 interactions via the dephosphorylation of AR at p44/42 MAPK consensus sites. Together, these results demonstrate that genistein disrupts the interaction between AR and ARA70 via AR dephosphorylation by inhibiting p44/42 MAPK.
Genistein ameliorates SBMA-like phenotypes in AR-97Q mice
To examine whether genistein could alleviate polyQ-mediated motor dysfunction, genistein was orally administered to male SBMA transgenic mice (at the effective dose of 250 mg/kg/day) starting at the age of 6 weeks and continuing until the mice reached 25 weeks of age. We assessed motor impairments weekly using rotarod activity tests and grip strength tests. Disease progression was markedly ameliorated in AR-97Q mice treated with 250 mg/kg/d genistein (G-250) (Fig. 4 a–g). Untreated male transgenic mice (G-0) exhibited significant motor impairments, as assessed by the rotarod test, as early as 8 weeks after birth, whereas G-250 mice showed initial impairments 11 weeks after birth and exhibited less deterioration overall than G-0 mice (Fig. 4a). G-0 mice consistently performed poorer than G-250 mice in tests of grip strength (Fig. 4b). Genistein treatment also slowed the body weight loss of SBMA mice compared with that observed in untreated SBMA mice (Fig. 4c). Genistein treatment also significantly increased the survival rate of G-250 mice in comparison with untreated mice (p < 0.01; Fig. 4d). By 12 weeks, G-0 mice exhibited obvious differences in body size, muscular atrophy, and kyphosis compared with G-250 mice (Fig. 4e). In addition, G-250 mice walked with significantly longer steps than G-0 mice (Fig. 4f and g). In contrast, AR-24Q mice and normal littermates treated with genistein displayed no altered phenotypes (data not shown).
To evaluate genistein toxicity, we examined blood samples from 12-week-old mice treated with 250 mg/kg/day genistein. Measurements of aspartate aminotransferase (AST), alanine aminotransferase (ALT), blood urea nitrogen (BUN), and serum creatinine (Cre) revealed no evidence of infertility or liver or renal dysfunction in male AR-97Q mice following treatment with genistein at a dose of 250 mg/kg/day (data not shown). These results indicate that oral genistein treatment can significantly delay the onset and symptomatic progression of motor impairments in AR-97Q mice.
We examined mutant AR expression levels in mouse tissue using immunohistochemistry with the 1C2 antibody, which specifically recognizes expanded polyQ. We observed a marked reduction in the accumulation of 1C2-positive nuclei in spinal motor neurons and muscles of G-250 mice compared with G-0 mice (Fig. 5a and b). Quantitative assessments revealed significantly fewer 1C2-positive cells in the spinal cord (p < 0.01; Fig. 5c) and muscle (p < 0.01; Fig. 5d) of G-250 mice compared with G-0 mice. Glial fibrillary acidic protein-specific antibody staining in wild-type (C57BL/6J), G-0, and G-250 mice showed an apparent reduction in reactive astrogliosis in the anterior horn of the spinal cord in G-250 mice compared with G-0 mice (Fig. 5e). Muscle histology in wild-type (C57BL/6J), G-0, and G-250 mice also showed a marked amelioration of muscle atrophy in AR-97Q mice treated with genistein (Fig. 5f).
Genistein administration suppresses p44/42 MAPK phosphorylation and reduces mutant AR expression in vivo
After having demonstrated that genistein promotes mutant AR degradation in vitro, we examined the levels of AR in a mouse model of SBMA. Western blot analyses of lysates from the spinal cord and muscle tissue of AR-97Q mice revealed the presence of high-molecular-weight mutant AR protein aggregates retained in the stacking gel in addition to a band corresponding to the monomeric form of mutant AR. Treatment with genistein diminished both the high-molecular-weight complex and the mutant AR monomer in the spinal cord and muscle of the AR-97Q mice (Fig. 6a and b). Furthermore, western blots also showed that genistein treatment resulted in decreased p44/42 MAPK phosphorylation in the spinal cord and muscle (Fig. 6c and d).
In this study, we identified a novel genistein-mediated mechanism of ameliorating SBMA symptoms through the down-regulation of mutant AR. Motor neuron degeneration induced by mutant AR involves several distinct mechanisms, including transcription deregulation, aggregate formation, transglutaminase activation, and mitochondrial deficits (Parodi and Pennuto 2011). Although it remains unclear which of these mechanisms has the greatest impact on SBMA pathogenesis, the amount of mutant AR is a well-established indicator of degeneration (Adachi et al. 2005). Reducing the levels of the pathogenic AR protein in SBMA mice has been shown to attenuate the manifestations of this disease (Adachi et al. 2003; Katsuno et al. 2003; Waza et al. 2005; Yang et al. 2007; Miyazaki et al. 2012; Rinaldi et al. 2012), thus suggesting that the clearance of mutant AR may lead to a therapeutic benefit for SBMA. Our results demonstrate that mutant AR is sensitive to genistein treatment and indicate that genistein-mediated down-regulation of mutant AR can attenuate motor abnormalities in an animal model of SBMA.
In addition to the effects of genistein in cell culture systems, we also observed in vivo effects of genistein treatment in SBMA transgenic mice. Genistein treatment improved the motor performance and survival rates of AR-97Q mice and ameliorated their weight loss and muscle pathology. These results indicate that long-term oral treatment of genistein can delay the progression of SBMA in vivo.
ARA70 was the first identified coregulator of AR, and it has been shown to interact with AR and increase its protein stability (Hu et al. 2004). Our results showed that genistein treatment can disrupt the association between AR and ARA70. AR is a phosphorylated protein with multiple phosphorylation sites. The direct phosphorylation of AR has been shown to influence its ability to interact with ARA70 (Heinlein and Chang 2002). p44/42 MAPK can phosphorylate AR at serine residue 515 (Yeh et al. 1999; Ponguta et al. 2008), an effect that enhances AR transactivation and promotes the AR–ARA70 interaction (Yeh et al. 1999). Dephosphorylation of AR at serine 515 increases the degradation of AR protein through the ubiquitin–proteasome system (Chymkowitch et al. 2011).
In addition, phosphorylation of polyQ AR by p44/42 MAPK at serine 515 is associated with increased toxicity in cellular models of SBMA, and inhibition of the p44/42 MAPK pathway reduces the cell toxicity induced by the polyQ-expanded mutant AR (LaFevre-Bernt and Ellerby 2003). These observations indicate that dephosphorylation of AR at serine 515 via p44/42 MAPK inhibition disrupts the AR–ARA70 interaction and promotes AR protein clearance, which could attenuate polyQ toxicity. Genistein is a broad-spectrum tyrosine kinase inhibitor (Akiyama et al. 1987; Yan et al. 2010) that inhibits p44/42 MAPK activation via the Ras-Raf-MAPK pathway. Our experiments demonstrated that genistein promotes the dissociation of AR from ARA70 by leading to the dephosphorylation of AR at serine 515.
AR is also a direct target of Akt, a kinase that is a member of the PI3K signal transduction pathway. AR phosphorylation by Akt has been shown to decrease the interaction between AR and ARA70 (Lin et al. 2001), and the phosphorylation of AR at serine 215 and serine 792 by Akt induces the proteasomal degradation of AR (Lin et al. 2002). Palazzolo et al. demonstrated that activation of Akt by insulin-like growth factor-1 (IGF-1) phosphorylates AR and induces the clearance of mutant AR through the ubiquitin–proteasome system. In addition, IGF-1 treatment has been shown to attenuate the motor phenotype in mouse models of SBMA (Rinaldi et al.; Palazzolo et al. 2009). Based on our observation that genistein treatment can disrupt AR–ARA70 interactions via the dephosphorylation of AR on serine 515, it is possible that post-translational modifications of mutant AR could represent a novel potential therapeutic approach for the treatment of SBMA.
We thank Miwa Ito and Kazuko Matsuba in the Department of Neurology; Noboru Ogiso, Yasutaka Ohya, and Kumiko Yano in the Division for Research of Laboratory Animals at the Center for Research of Laboratory Animals and Medical Research Engineering for their technical assistance; and Gary S. Goldberg in the Molecular Biology Department of the University of Medicine and Dentistry of New Jersey for helpful suggestions. This work was supported by a Center-of-Excellence (COE) grant from the Ministry of Education, Culture, Sports, Science, and Technology of Japan, and KAKENHI from the Ministry of Education, Culture, Sports, Science, and Technology of Japan, and a scholarship from the China Scholarship Council to Qiang Qiang. The authors declare no competing financial interests.