The polyglutamine-expanded androgen receptor responsible for spinal and bulbar muscular atrophy inhibits the APC/CCdh1 ubiquitin ligase complex

Abstract

Polyglutamine expansion in the androgen receptor (AR) causes spinal and bulbar muscular atrophy (SBMA), an X-linked neuromuscular disease that is fully manifest only in males. It has been suggested that proteins with expanded polyglutamine tracts impair ubiquitin-dependent proteolysis due to their propensity to aggregate, but recent studies indicate that the overall activity of the ubiquitin-proteasome system is preserved in SBMA models. Here we report that AR selectively interferes with the function of the ubiquitin ligase anaphase-promoting complex/cyclosome (APC/C), which, together with its substrate adaptor Cdh1, is critical for cell cycle arrest and neuronal architecture. We show that both wild-type and mutant AR physically interact with the APC/C^Cdh1 complex in a ligand-dependent fashion without being targeted for proteasomal degradation. Inhibition of APC/C^Cdh1 by mutant but not wild-type AR in PC12 cells results in enhanced neurite outgrowth which is typically followed by rapid neurite retraction and mitotic entry. Our data indicate a role of AR in neuronal differentiation through regulation of APC/C^Cdh1 and suggest abnormal cell cycle reactivation as a pathogenic mechanism in SBMA.

The polyglutamine-expanded androgen receptor responsible for spinal and bulbar muscular atrophy inhibits the APC/C ubiquitin ligase complex

OPEN

R

Laura C. Bott,,, Florian A. Salomons, Dragan Maric, Yuhong Liu, Diane Merry, Kenneth H. Fischbeck & Nico P. Dantuma

Polyglutamine expansion in the androgen receptor (AR) causes spinal and bulbar muscular atrophy (SBMA), an X-linked neuromuscular disease that is fully manifest only in males. It has been suggested that proteins with expanded polyglutamine tracts impair ubiquitin-dependent proteolysis due totheir propensity to aggregate, but recent studies indicate that the overall activity of the ubiquitinproteasome system is preserved in SBMA models. Here we report that AR selectively interferes with the function of the ubiquitin ligase anaphase-promoting complex/cyclosome (APC/C), which, together both wild-type and mutant AR physically interact with the APC/C complex in a ligand-dependent fashion without being targeted for proteasomal degradation. Inhibition of APC/C by mutant but not regulation of APC/C and suggest abnormal cell cycle reactivation as a pathogenic mechanism in SBMA.

Spinal and bulbar muscular atrophy (SBMA) is an X-linked neuromuscular disease characterized by progressive loss of motor neurons in the brain stem and spinal cord, with atrophy and weakness of bulbar and extremity muscles1. It is caused by expansion of a CAG trinucleotide repeat in the androgen receptor (AR) gene, which encodes a polyglutamine (polyQ) tract in the AR protein2. PolyQ expansions in unrelated proteins are the underlying cause of eight other neurodegenerative disorders, including Huntingtons disease, dentatorubral-pallidoluysian atrophy, and six spinocerebellar ataxias3. These diseases share pathological features, such as intracellular accumulation of the mutant protein in inclusion bodies4. Expanded polyQ tracts confer a high propensity to aggregation and impose a demand on the proteostasis machinery for correct protein folding5.

PolyQ toxicity is associated with alterations in ubiquitin-dependent processes, which control a wide spectrum of cellular functions, including protein degradation via the ubiquitin-proteasome system (UPS). The UPS is a major pathway for the clearance of short-lived, misfolded, and damaged proteins in both the nucleus and cytoplasm6. It also has critical roles in cell cycle control, signaling, and apoptosis7, and a general impairment of this proteolytic system could therefore provide a mechanistic explanation for the inherent cytotoxic consequences of proteins with expanded polyQ tracts8. It has been suggested that polyQ proteins inhibit UPS function either directly, by blocking the proteasome, or indirectly, through sequestration of essential UPS components into inclusions9. However, although polyQ disease proteins can cause a general impairment of the UPS when

National Institute of Department of Biochemistry

SCIENTIFIC REPORTS

1

www.nature.com/scientificreports/

Figure 1. AR expression promotes androgen-dependent neurite outgrowth in PC12 cells. (A) Characterization of mCherry-AR protein expression in doxycycline (dox)-inducible PC12 cell lines. The cells were cultured in the absence of dox for the indicated times. (B) Eect of DHT on mCherry-AR protein levels. PC12 mCherry-AR lines were cultured in the presence or absence of 1g/ml doxycycline and 10nM DHT for 24hours. (C) Comparisonof the transcriptional activity of mCherry-AR25Q and AR107Q in the inducible PC12 cells using a luciferase reporter. The cells were incubated with 10nM DHT 24hours prior to the luciferase measurements. Data are expressed as mean SEM (n= 3). n.s. denotes non-signicant dierences between samples treated with DHT (Two-way ANOVA). (D) Representative images of PC12 mCherry-AR cells incubated in the presence or absence of 1ug/ml dox and 10nM DHT or ethanol vehicle for 72hours. Scale bar, 20m. (E) Quantitative assessmentof neurite outgrowth in PC12 mCherry-AR cell lines. Data are expressed as mean SEM (n=3). ***P<0.001 (Two-way ANOVA). (F) Quantication of neurite length in mCherry-AR-expressing cells in the presence of DHT. Each dot in the plot represents the measurement of a single cell with a horizontal line representing the median neurite length (n=250). **P< 0.01 (Mann-Whitney test).

SCIENTIFIC REPORTS

2

www.nature.com/scientificreports/

acutely overexpressed in cell lines10, studies in mouse models have shown that ubiquitin-dependent proteolysis is preserved in SBMA11 as well as other polyQ disorders1214.

Each of the polyQ diseases has a distinct pathology with specic sets of neurons being aected3, indicating that cellular eects of the repeat expansion are highly dependent on the cell type and protein context. Among polyQ proteins, the physiological functions of the AR have been well characterized. AR is highly expressed in lower motor neurons in the spinal cord and brainstem15, a major site of toxicity in SBMA1, where it mediates gender dierences in neural organization and neuromuscular function during development16. Androgen signaling remains an important mediator of axon growth and regeneration during adulthood17,18. Studies in cell and animal models have shown that toxicity in SBMA requires androgen19 and nuclear localization of mutant AR20,21, which is consistent with the notion that normal functions of polyQ proteins may be critical for pathogenesis21,22.

While most AR functions have been attributed to its role as a transcription factor, there is also evidence for non-canonical functions of AR in cell cycle control and neurite outgrowth through direct interactions with signaling proteins and components of the cell cycle machinery23,24.

Results

AR-mediated neurite outgrowth is enhanced in a neuronal cell model of SBMA. To study the eects of AR expression in a neuronal cell line, we generated PC12 cell lines with inducible expression of mCherry-tagged full-length human AR and normal (AR25Q) or expanded (AR107Q) polyQ tracts under the control of a tetracycline transactivator. Western blot analysis of selected clones conrmed that removal of doxycycline caused a gradual increase in mCherry-AR25Q and AR107Q protein levels, reaching a maximum aer approximately 12 hours (Fig.1A). Treatment with the androgen dihydrotestosterone (DHT) further increased protein levels of mCherry-AR25Q and AR107Q (Fig.1B), consistent with earlier reports which showed that ligand extends the half-life of AR25. Cells expressing AR107Q formed nuclear inclusions that were positive for red uorescent signal at low frequency (approximately 5%) aer three days of DHT treatment (Supplementary Fig. S1). Next, we compared transactivation of a luciferase open reading frame under the control of androgen-responsive elements in these stable cell lines. We found DHT-dependent luciferase activity in AR-expressing cell lines, conrming that the mCherry-AR fusion proteins are functional in terms of ligand binding, nuclear translocation, and transcriptional activity (Fig.1C). Since PC12 cells are devoid of endogenous AR26, luciferase activity was absent when transgene expression was suppressed with doxycycline. Notably, we did not detect a signicant dierence in luciferase activity between the cell lines expressing mCherry-AR25Q and -AR107Q, indicating that the overall transcriptional output of the two AR variants in the assay is comparable in this cell system.

PC12 cells can be dierentiated in response to various soluble and genetic factors that cause cells to arrest in the G0/G1 phase and adopt a neuron-like phenotype27. Since AR is known to induce neurite outgrowth in neuronal cells18,28, we evaluated the eect of mCherry-AR expression on cell morphology in the PC12 cell lines. We found that both wild-type and mutant AR promote neurite outgrowth following treatment with DHT for 72hours (Fig.1D). This cellular phenotype was not evident aer DHT treatment in the presence of doxycycline, conrming that the AR transgene was required for the androgen-dependent dierentiation. Surprisingly, we found that the percentage of cells with neurites was approximately two-fold higher in DHT-treated cells expressing AR107Q compared to AR25Q (Fig.1E). Moreover, an eect of mutant AR on neurite outgrowth was further consolidated by the nding that the average length of neurites was signicantly higher in the mutant AR-expressing population compared to wild-type AR (Fig.1F). Our ndings support a role for AR in neuronal dierentiation and indicate that this function of AR may be altered by the polyQ expansion.

The polyQ expansion in AR reduces androgen-dependent cell cycle arrest. Aer seeing the differential eects of AR25Q and AR107Q on neurite outgrowth in PC12 cells, we next asked whether the wild-type and mutant receptors dier in their ability to induce and maintain cell cycle arrest. We found that DHT treatment signicantly increased the proportion of AR25Q-expressing cells in the G0/G1 phase and signicantly decreased the proportion of S phase cells (Fig.2A, Supplementary Fig. S2), conrming previous reports that AR induces cell cycle arrest in an androgen-dependent manner2931. A similar trend was observed with cells expressing AR107Q, however, the proportion of DHT-treated cells in the G0/G1 phase was lower compared to AR25Q and coincided with an increase in the G2/M population (Fig.2A). To test whether the increase in the G2/M population in AR107Q-expressing cells reects a G2 phase arrest, we analyzed the levels of phospho-histone H3, a mitotic marker32, in the PC12 cell lines. While the percentage of phospho-histone H3-positive cells was reduced aer DHT treatment in AR25Q-expressing cells, we found that it was unchanged in cells expressing AR107Q (Fig.2B, Supplementary Fig. S2). This nding indicates that cells did not arrest in the G2 phase but either remained mitotically active or arrested during mitosis.

Next, we tested the possibility that the altered cell cycle distribution caused by AR107Q may be a consequence of increased cell death. However, we did not detect a signicant eect of mutant AR expression on viability based on nuclear morphology aer staining of cells with Hoechst dye (Fig.2C). Our data conrm that expression of the AR gives rise to DHT-induced cell cycle arrest and suggest that the polyQ expansion alters this function of AR in cell cycle regulation.

AR physically interacts with the ubiquitin ligase APC/C . The ubiquitin ligase APC/C is a critical regulator of cell cycle progression and has important functions in postmitotic neurons. Together with its co-activator Cdh1, APC/C controls cell cycle exit and regulates axonal growth during neuronal dierentiation by targeting specic proteins for ubiquitin-dependent degradation33. We reasoned that APC/CCdh1 could be a target of AR because of the observed ligand-dependent eects of AR on both cell cycle progression and neurite outgrowth. We found that both AR25Q and AR107Q interact with Cdh1 in a DHT-dependent manner (Fig.3A), indicating a link between APC/CCdh1 activity and ligand-induced dierentiation of PC12 cells. Endogenous AR

SCIENTIFIC REPORTS

3

www.nature.com/scientificreports/

Figure 2. AR expression promotes androgen-dependent cell cycle arrest. (A) Cell cycle analysis of PC12 mCherry-AR cell lines. Cells were cultured in the presence or absence of 1g/ml dox and 10nM DHT for 36hours. Data are expressed as mean SEM (n=3). **P<0.01, ***P< 0.001 (Two-way ANOVA). (B)

Mitotic index based on phospho-histone H3 immunostaining of mCherry-AR-expressing cells analyzed by ow cytometry. Data are expressed as mean SEM (n=3). *P< 0.05 (One-way ANOVA). (C) Viability of PC12 cell lines following mCherry-AR expression and DHT treatment. Unxed cells were stained with Hoechst 33342 dye and assayed for the presence of pyknotic or fragmented nuclei. Data are expressed as mean SEM (n=3). No signicant dierences could be detected between the samples (Two-way ANOVA).

Figure 3. AR interacts with APC/CCdh1 and is not targeted for proteasomal degradation by this complex.

(A) Co-immunoprecipitation of Cdh1 and AR in PC12 cells. Cells were transfected with FLAG-Cdh1 and induced to express mCherry-AR. Cdh1 was co-immunoprecipitated using AR antibodies following incubation with 10nM DHT for 36hours. (B) Endogenous AR was co-immunoprecipitated from MCF7 cells aer treatment with 10nM DHT to verify association with APC/CCdh1. (C) Co-immunoprecipitation of Cdh1 with mutant AR and other expanded polyQ proteins. HA-Cdh1 was transfected into PC12 cells together with the following GFP-tagged constructs: AR65Q, Ataxin-3 (Atx3-78Q), huntingtin exon 1 fragment (htt104Q), or an amino-terminal AR fragment (ARd112Q). Co-immunoprecipitation was performed with a polyclonal GFP antibody. (D) Eect of APC/C inhibition on AR protein levels. PC12 cells were co-transfected with either mCherry-AR25Q, -AR107Q, or Cdc25, and a dominant-negative form of Cdh1. Aer 24 hours, the cells were collected and analyzed by western blotting.

in MCF7 cells also co-immunoprecipitated with endogenous Cdh1 as well as Cdc27, a core component of APC/C (Fig.3B), demonstrating that AR and APC/CCdh1 indeed form a complex under physiological conditions.

The observation that wild-type AR also interacts with APC/C suggests that this binding is a normal function of this protein and not mediated by the polyQ expansion alone. Unlike AR, polyQ-expanded ataxin-3 and N-terminal huntingtin fragment did not interact with Cdh1 (Fig.3C). Moreover, an N-terminal fragment of the mutant AR truncated 52 amino acids C-terminal of the polyQ repeat that forms insoluble inclusions in cells34,35,

SCIENTIFIC REPORTS

4

www.nature.com/scientificreports/

Figure 4. Mutant AR expression causes ligand-dependent inhibition of APC/C activity in the absenceof global UPS impairment. (A) Eect of proteasome inhibition on the APC/C reporter Dbox-GFP in stably transfected PC12 mCherry-AR cell lines. Non-induced cells were treated with 100nM epoxomicin for 16hours. (B) Eect of neuronal dierentiation on steady-state levels of the Dbox-GFP reporter. Uninduced cells were treated with nerve growth factor (NGF) for 16hours. (C) Eect of wild-type or mutant AR on Dbox-GFP reporter levels. The stable PC12 cell lines were induced to express mCherry-AR25Q or -AR107Q and cultured in the presence or absence of 1g/ml doxycycline and 10nM DHT for 2 days. (D,E) PC12 mCherry-ARcells stably transfected with the UPS reporters UbG76V-GFP (D) or GFP-CL1 (E) were treated as in (C). As a reference, cells were treated in (D,E) with epoxomicin (100nM for 16h) to show the maximum level of reporter accumulation upon UPS impairment. Data in all graphs are expressed as mean SEM (n=3). ***P<0.001 (One or Two-way ANOVA).

was insufficient for the interaction with Cdh1. These ndings support the view that the association between AR and Cdh1 is not facilitated by the expanded polyQ tract alone but an inherent feature of wild-type AR.

Both wild-type and mutant AR are proteasome substrates and primarily cleared by the UPS36. Since APC/C has ubiquitin ligase activity, and Cdh1 is its substrate adaptor, we asked whether APC/CCdh1 targets AR for proteasomal degradation. For this, we analyzed the eect of expression of an HA-tagged dominant-negative Cdh1 (DN-Cdh1), consisting of the amino-terminal 125 amino acids, on the levels of mCherry-AR25Q, -AR107Q, and a known substrate of the ubiquitin ligase complex, Cdc2537. While DN-Cdh1 stabilized Cdc25, it did not aect AR protein levels in the presence or absence DHT (Fig.3D), indicating that AR is not targeted for proteasomal degradation by APC/CCdh1.

The mutant AR causes accumulation of an APC/C reporter substrate. It has previously been shown that APC/CCdh1 activity is critical for cell cycle exit and neuronal dierentiation through coordinated degradation of key factors that inhibit these processes in proliferating cells38. We therefore asked whether AR aects the function of this ubiquitin ligase complex, as dierences in APC/C activity may explain the observed eects of wild-type and mutant AR on the cell cycle distribution and neurite outgrowth in our cell system. To address this question, we generated a reporter consisting of an amino-terminal fragment from Xenopus cyclin B fused to the green uorescent protein (GFP). Cyclin B is a well-characterized proteasome substrate and subject to ubiquitylation by APC/C39. The fragment derived from cyclin B contains two destruction box (D-box) consensus motifs that are recognized by the APC/C complex. The resulting reporter construct, which we termed D-box-GFP, was stably transfected into the inducible mCherry-AR PC12 lines. We conrmed in live cells that the uorescence of the D-box-GFP reporter gradually increases as a function of cell cycle progression and disappears during mitosis at the transition of metaphase to anaphase, which inversely correlates with the predicted APC/C activity at these stages of the cell cycle (data not shown).

In the asynchronously growing reporter lines, approximately 25% of cells had elevated levels of the D-box-GFP reporter at any given time. Treatment of the reporter cell lines with the proteasome inhibitor epoxomicin resulted in high D-box-GFP signal in almost all cells, conrming that this reporter is degraded by the UPS (Fig.4A). In contrast, incubation with nerve growth factor caused a marked reduction in the D-box-GFP-positive cell population, consistent with a high APC/CCdh1 activity associated with neuronal dierentiation (Fig.4B).

Expression of mCherry-AR reduced the percentage of D-box-GFP-positive cells in the presence of DHT, as would be expected for a condition that promotes G0/G1 arrest. Notably, the proportion of D-box-GFP-positive cells was approximately three-fold higher in the cell line expressing AR107Q compared to AR25Q (Fig.4C). This result shows that mutant AR stabilizes an APC/C-dependent reporter substrate in a ligand-dependent manner.

SCIENTIFIC REPORTS

5

www.nature.com/scientificreports/

Increased levels of the D-box-GFP reporter may reect changes in overall UPS functionality, which have been reported in various cellular models of polyQ toxicity10,14,40. Therefore, we next tested whether AR107Q aects the steady-state levels of two well-characterized UPS reporters: UbG76V-GFP, which has a ubiquitin-fusion degradation signal41, and GFP-CL1, which contains a peptide motif targeting for ubiquitin-dependent proteolysis10. These proteasome substrates are recognized by distinct, APC/C-independent ubiquitylation pathways. As with the D-box-GFP reporter, these constructs were stably transfected into the inducible mCherry-AR PC12 cell lines. Levels of the UbG76V-GFP reporter (Fig.4D) and GFP-CL1 reporter (Fig.4E) were not aected by AR25Q or AR107Q but dramatically increased upon chemical inhibition of the UPS with the proteasome inhibitor epoxomicin (Fig.4D,E). We conclude that stabilization of the APC/C-dependent substrate by polyQ-expanded AR is not a consequence of global UPS impairment.

Abnormal neurite outgrowth and cell cycle re-entry coincide in mutant AR-expressing cells.

We performed live-cell imaging to measure the dynamics of APC/C activity in AR-expressing cells over time. PC12 cell lines expressing mCherry-AR25Q or AR107Q were imaged continuously in 10min intervals over two days starting aer the addition of DHT or vehicle, and the relative numbers of D-box-GFP-positive cells were determined as a function of time. We found a steady increase in the number of D-box-GFP-positive cells in the vehicle-treated sample during the course of the experiment, which reects an increase in the total number of cells as a result of continuous proliferation. In the presence of DHT, the number of GFP-positive cells decreased over time relative to the value at the start of the experiment, indicating increased APC/C activity, and reached a plateau aer approximately 18hours (Fig.5A), which is consistent with our observation of androgen-induced cell cycle arrest and dierentiation of these cell lines. Importantly, the relative number of GFP-positive cells remained consistently higher in cells expressing AR107Q compared to AR25Q in the presence of DHT. The dierential eect of wild-type and mutant AR on reporter levels was clearly evident within 12 hours aer the addition of ligand, suggesting that altered APC/C activity may be an early event in this SBMA model.

Next, we asked whether a link exists between aberrant neurite outgrowth, cell cycle status, and APC/C inhibition in mutant AR107Q-expressing PC12 cells. Alternatively, these observations may be separate events and occur in distinct cellular populations. To address this, we monitored cell morphology as well as D-box-GFP and mCherry-AR uorescence signal intensities in single cells over 72 hours. We found that in the presence of DHT 75% of AR25Q-expressing cells did not accumulate the D-box-GFP reporter aer an initial cell division, indicating that they permanently exit the cell cycle in response to ligand. Only 26% of mutant AR-expressing cells underwent cell cycle exit aer an initial division by this measure (Fig.5B,C).

Our longitudinal analysis showed that AR25Q-expressing cells that formed neurites tended not to accumulate GFP uorescence over the duration of the recording (Fig.5D, Supplementary Movie 1). In contrast, cells expressing AR107Q displayed waves of elevated levels of D-box-GFP signal while developing neurites (Fig.5E, Supplementary Movie 2). As the reporter levels reach the maximum intensity, these cells typically retracted their neurites and this event was oen directly followed by cell division (Fig.5E). Thus abnormal neurite growth occurs in the same cells that display rapid neurite retraction and mitotic re-entry consistent with the model that both events are a direct consequence of reduced APC/C activity.

Activity of the APC/CCdh1 complex is regulated by phosphorylation, ubiquitylation, and association with protein inhibitors, which bind to the Cdh1 adaptor by acting as pseudo-substrates33. Since AR interacts with Cdh1 but does not behave as a substrate and instead reduces APC/C function, we hypothesized that mutant AR may act as such a pseudo-substrate inhibitor. If AR competes with other substrates for binding to APC/CCdh1, we

predicted that overexpression of Cdh1 should abrogate the cellular phenotype of the mutant protein. Indeed, overexpression of GFP-tagged Cdh1 reduced neurite outgrowth in PC12 cells expressing mCherry-AR25Q and mCherry-AR107Q (Fig.5F). Our data suggest that this inhibitory eect of the AR is enhanced by the long polyQ tract, resulting in abortive neuronal dierentiation that is followed by mitotic entry.

Discussion

We report that mutant AR inhibits the APC/C ubiquitin ligase in the absence of global UPS impairment in neuronal cells. This eect is likely to be contingent upon native functions of AR, such as ligand binding and nuclear translocation of the receptor, as well as androgen-dependent cellular dierentiation. Our data suggest that APC/C inhibition is a normal AR function that is amplied by the expanded polyQ tract. The observation that the normal AR functions are relevant for cellular toxicity21 also argues in favor for hypermorphic eects being responsible for the pathology as opposed to a more general gain-of-function toxicity caused by the propensity of the polyQ expanded protein to aggregate42. In support of this model, we found that overexpression of Cdh1 abrogated androgen-induced neurite outgrowth in cells expressing either wild-type or polyQ expanded AR. Since APC/ CCdh1 activity negatively regulates neurite outgrowth43,44, our observations suggest that AR inhibits the activity of the APC/CCdh1 complex, a mode of action reminiscent to pseudo-substrates, which inhibit APC/C activity by competing with substrates for binding to the Cdh1 substrate adaptor4547. It should be noted that it has been reported in an earlier study that mutant AR has a reduced capacity to induce DHT-dependent neuronal outgrowth24. This earlier work was performed with NSC34 cells and a possible explanation for this discrepancy may lie therefore in the dierent type of neuronal cell lines that have been used.

Our data suggest that the observed dierences on neuronal dierentiation observed with wild-type and mutant AR-expressing cells are not a direct consequence of dierences in transcriptional activation. Although our molecular data support a non-translational role of AR in regulating APC/CCdh1-dependent degradation, we cannot exclude the possibility that more quantitative dierences in the transcription prole may contribute to this phenomenon since transcriptional activation has been assessed with a single reporter substrate. Indeed a number of studies have indicated that the polyQ repeat expansion has an eect on the transcriptional activation of AR4851.

SCIENTIFIC REPORTS

6

www.nature.com/scientificreports/

Figure 5. Single-cell dynamics of APC activity, neurite outgrowth, and cell division. (A) Time course of Dbox-GFP reporter accumulation in PC12 cells expressing mCherry-AR in the presence or absence of 10nM DHT over 2 days. DHT was added to the cells at the 0-hour time point. (B) mCherry-AR expressing PC12cells were incubated with 10nM DHT and imaged for 72hours. The graph shows the percentage of cells that accumulate the Dbox-GFP reporter aer an initial cell division. (C) Dbox-GFP reporter accumulation in single mCherry-AR-expressing cells over time. Peaks indicate the maximal GFP intensity in individual cells (shownas stacked horizontal lines). (D,E) Representative micrographs of cells from (C). (F) PC12 mCherry-AR cells were transfected with GFP-Cdh1 or empty vector and incubated with 10nM DHT in the presence or absence of doxycycline (dox) for 72h. The percentage of cells with neurites was quantied by microscopy. Data are expressed as mean SEM (n= 3). Data are expressed as mean SEM (n=3). ***P< 0.001 (Two-way ANOVA).

SCIENTIFIC REPORTS

7

www.nature.com/scientificreports/

APC/C coordinates cell cycle transitions in proliferating cells and governs critical functions in postmitotic cells through ubiquitin-dependent degradation of a multitude of protein substrates52. Our observations are consistent with mutant AR-mediated inhibition of the APC/CCdh1 complex since cell cycle arrest and inhibition of neurite outgrowth were simultaneously compromised in our neuronal cell model. APC/CCdh1 dysregulation may also manifest in altered synaptic activity53 and oxidative stress, which have been reported in SBMA54,55. Although SBMA is primarily a motor neuron disease, recent studies suggest that mutant AR in the skeletal muscle may also play an important role in the pathology5659. Abnormal stabilization of Cdh1-dependent substrates by mutant AR may also be responsible for alterations in the activity and organization of the neuromuscular junction in SBMA mouse models60,61. Given the roles of APC/CCdh1 in establishing and maintaining communication between motor neurons and muscle6264, changes in the abundance of critical substrates can therefore have both cell-autonomous as well as cell-non-autonomous eects, and may contribute to motor neuron and muscle degeneration in SBMA.

Materials and Methods

Plasmid constructs and cell lines. Full length human androgen receptor with 25 or 107 glutamine repeats was tagged with mCherry uorescent protein at the amino terminus and cloned into the pTRE2hyg vector (Clontech). The APC-dependent reporter construct Dbox-GFP was generated by fusing amino acids 191 of Xenopus laevis cyclin B1 to EGFP. The UbG76V-GFP reporter has been described [41] and GFP-CL1 was generated by replacing the yellow uorescent protein in YFP-CL1 [61] with GFP.

The PC12 Tet-O cell line (Clontech) was maintained in high-glucose Dulbeccos Modied Eagle Medium supplemented with Glutamax, 10% heat-inactivated horse serum, and 5% fetal bovine serum at 37C and 5% CO2 (Life Technologies). The mCherry-AR constructs were transfected into PC12 cells with Lipofectamine 2000 (Life Technologies) according to manufacturers instructions and stable transformants were selected, isolated, and expanded in the presence of 1g/ml doxycycline and 100g/ml hygromycin. For induction of mCherry-AR, doxycycline was removed from the cells by repeated medium changes and exposed to either 10 nM DHT (Sigma-Aldrich) or ethanol vehicle in culture medium containing charcoal-treated serum (Life Technologies). Reporter cell lines were generated by stably expressing Dbox-GFP, UbG76V-GFP, and GFP-CL1 in PC12 Tet-o mCherry-AR lines.

Flow cytometry. For the cell cycle analysis, PC12 cells were pulse-labeled with 10M EdU for 2hours, xed, and stained with the Click-iT EdU Alexa Fluor 488 ow cytometry assay kit (Life Technologies) according to the manufacturers recommendations, phospho-histone H3 antibody (#9706, Cell Signaling Technology), and 1 ug/ml DAPI. Fluorescence intensity was assayed using the MoFlo Astrios cell sorter (Beckman Coulter) and Summit v6.2.6 soware for data acquisition. Data analysis was carried out with Kaluza v1.2 soware (Beckman Coulter) and standard gating methods.

Co-immunoprecipitation and western blotting. Cells were lysed in 150 mM NaCl, 50 mM Tris-HCl pH 7.5, 1mM EDTA, 10% glycerol, 0.2% NP-40 containing protease and phosphatase inhibitor cocktail (Thermo Scientic). Co-immunoprecipitation were performed with lysates of cells that had not been treated with DHT using indicated antibodies, followed by incubation with protein G-coupled sepharose beads (GE Healthcare). Protein samples were prepared in Laemmli buer, separated on 412% Tris-glycine polyacrylamide gels and transferred onto PVDF membranes (Life Technologies). Following incubation with relevant primary antibodies, blots were visualized with peroxidase-linked secondary antibodies (R&D Systems) and chemiluminescence rea-gent (Perkin Elmer). The following antibodies were used: AR (sc-816 and sc-7305, Santa Cruz Biotechnology), Cdc27 (C7104, Sigma-Aldrich), Cdh1 (K0085-3, MBL), FLAG (F1804, Sigma Aldrich; 2368, Cell Signaling Technology), GFP (ab290, Abcam), mCherry antibody (polyclonal antibody raised against mRFP; kindly provided by Jacques Neefjes, Netherlands Cancer Institute), HA tag (16B12, Covance), and -tubulin (T6199, Sigma-Aldrich).

Luciferase assay. For AR transactivation experiments, the plasmid (ARE)2TATA-luc expressing fire-y luciferase under the control of a minimal promoter with androgen-response elements (gi from Amilcar Flores-Morales, University of Copenhagen)65, and a plasmid encoding constitutively expressed Renilla luciferase (pGL4.74) were co-transfected into PC12 cells at a ratio of 10:1. Luciferase activity measurements were performed using the Dual luciferase assay kit (Promega) according to the manufacturers instructions.

Microscopy. Live-cell imaging was performed on the automated image acquisition uorescence microscope system ImageXpress (Molecular Devices) using a 20x magnication with DAPI, FITC, and Texas Red lter cubes at multiple positions. GFP reporter uorescence was analyzed using the MetaXpress soware and the multi-wavelength cell scoring application module (Molecular Devices). Long-term time-lapse imaging was performed on the uorescence microscope system DMI6000 (Leica) at 37 C and 5% CO2 using a 20x magnication with DIC, mCherry, and GFP lter sets at 4 positions per condition. Images were acquired at 10min intervals. Images were analyzed using Volocity soware (Perkin Elmer). For neurite quantication, a neurite was dened as a process extending from the soma by at least one cell diameter (10 m). Cell viability was determined based on the relative number of pyknotic or fragmented nuclei following incubation with Hoechst 33342 dye (Life Technologies).

Statistical analysis. Data sets were analyzed with Graphpad Prism soware (version 5) for statistical comparisons. P< 0.05 was considered as statistically signicant.

SCIENTIFIC REPORTS

8

www.nature.com/scientificreports/

References

1. Kennedy, W. R., Alter, M. & Sung, J. H. Progressive proximal spinal and bulbar muscular atrophy of late onset. A sex-linked recessive trait. Neurology 18, 671680 (1968).

2. La Spada, A. R., Wilson, E. M., Lubahn, D. B., Harding, A. E. & Fischbeck, K. H. Androgen receptor gene mutations in X-linked spinal and bulbar muscular atrophy. Nature 352, 7779 (1991).

3. Orr, H. T. & Zoghbi, H. Y. Trinucleotide repeat disorders. Annu. Rev. Neurosci. 30, 575621 (2007).4. Alves-Rodrigues, A., Gregori, L. & Figueiredo-Pereira, M. E. Ubiquitin, cellular inclusions and their role in neurodegeneration. Trends Neurosci. 21, 516520 (1998).

5. Perutz, M. F. Glutamine repeats and neurodegenerative diseases: molecular aspects. Trends Biochem. Sci 24, 5863 (1999).6. Sherman, M. Y. & Goldberg, A. L. Cellular defenses against unfolded proteins: a cell biologist thinks about neurodegenerative diseases. Neuron 29, 1532 (2001).

7. Hershko, A. & Ciechanover, A. The ubiquitin system. Annu. Rev. Biochem. 67, 425479 (1998).8. Ciechanover, A. & Brundin, P. The ubiquitin proteasome system in neurodegenerative diseases: sometimes the chicken, sometimes the egg. Neuron 40, 427446 (2003).

9. McKinnon, C. & Tabrizi, S. J. The ubiquitin-proteasome system in neurodegeneration. Antioxid. Redox Signal. 21, 23022321 (2014).10. Bence, N. F., Sampat, R. M. & Kopito, R. R. Impairment of the ubiquitin-proteasome system by protein aggregation. Science 292, 15521555 (2001).

11. Tokui, K. et al. 17-DMAG ameliorates polyglutamine-mediated motor neuron degeneration through well-preserved proteasome function in an SBMA model mouse. Hum. Mol. Genet. 18, 898910 (2009).

12. Bett, J. S., Cook, C., Petrucelli, L. & Bates, G. P. The ubiquitin-proteasome reporter GFPu does not accumulate in neurons of the R6/2 transgenic mouse model of Huntingtons disease. PLoS One 4, e5128 (2009).

13. Bowman, A. B., Yoo, S. Y., Dantuma, N. P. & Zoghbi, H. Y. Neuronal dysfunction in a polyglutamine disease model occurs in the absence of ubiquitin-proteasome system impairment and inversely correlates with the degree of nuclear inclusion formation. Hum Mol Genet 14, 679691 (2005).

14. Maynard, C. J. et al. Accumulation of ubiquitin conjugates in a polyglutamine disease model occurs without global ubiquitin/ proteasome system impairment. Proc. Natl. Acad. Sci. USA 106, 1398613991 (2009).

15. Matsumoto, A. Hormonally induced neuronal plasticity in the adult motoneurons. Brain Res. Bull. 44, 539547 (1997).16. Morris, J. A., Jordan, C. L. & Breedlove, S. M. Sexual dierentiation of the vertebrate nervous system. Nat. Neurosci. 7, 10341039 (2004).

17. Kujawa, K. A., Jacob, J. M. & Jones, K. J. Testosterone regulation of the regenerative properties of injured rat sciatic motor neurons. J. Neurosci. Res. 35, 268273 (1993).18. Marron, T. U. et al. Androgen-induced neurite outgrowth is mediated by neuritin in motor neurones. J. Neurochem. 92, 1020 (2005).

19. Katsuno, M. et al. Testosterone reduction prevents phenotypic expression in a transgenic mouse model of spinal and bulbar muscular atrophy. Neuron 35, 843854 (2002).

20. Montie, H. L. et al. Cytoplasmic retention of polyglutamine-expanded androgen receptor ameliorates disease via autophagy in a mouse model of spinal and bulbar muscular atrophy. Hum. Mol. Genet. 18, 19371950 (2009).

21. Nedelsky, N. B. et al. Native functions of the androgen receptor are essential to pathogenesis in a Drosophila model of spinobulbar muscular atrophy. Neuron 67, 936952 (2010).

22. Duvick, L. et al. SCA1-like disease in mice expressing wild-type ataxin-1 with a serine to aspartic acid replacement at residue 776. Neuron 67, 929935 (2010).

23. Balk, S. P. & Knudsen, K. E. AR, the cell cycle, and prostate cancer. Nucl Recept Signal 6, e001 (2008).24. Schindler, M. et al. Disruption of nongenomic testosterone signaling in a model of spinal and bulbar muscular atrophy. Mol. Endocrinol. 26, 11021116 (2012).

25. Palazzolo, I. et al. Akt blocks ligand binding and protects against expanded polyglutamine androgen receptor toxicity. Hum. Mol. Genet. 16, 15931603 (2007).

26. Walcott, J. L. & Merry, D. E. Ligand promotes intranuclear inclusions in a novel cell model of spinal and bulbar muscular atrophy. J. Biol. Chem. 277, 5085550859 (2002).27. Greene, L. A. Nerve growth factor prevents the death and stimulates the neuronal dierentiation of clonal PC12 pheochromocytoma cells in serum-free medium. J. Cell. Biol. 78, 747755 (1978).

28. Lustig, R. H., Hua, P., Smith, L. S., Wang, C. & Chang, C. An in vitro model for the eects of androgen on neurons employing androgen receptor-transfected PC12 cells. Mol. Cell. Neurosci. 5, 587596 (1994).

29. de Launoit, Y., Dauvois, S., Dufour, M., Simard, J. & Labrie, F. Inhibition of cell cycle kinetics and proliferation by the androgen 5 alpha-dihydrotestosterone and antiestrogen N,n-butyl-N-methyl-11-[16 alpha-chloro-3,17 beta-dihydroxy-estra-1,3,5-(10) triene-7 alpha-yl] undecanamide in human breast cancer ZR-75-1 cells. Cancer Res. 51, 27972802 (1991).

30. Greeve, M. A., Allan, R. K., Harvey, J. M. & Bentel, J. M. Inhibition of MCF-7 breast cancer cell proliferation by 5alpha-dihydrotestosterone; a role for p21(Cip1/Waf1). J. Mol. Endocrinol. 32, 793810 (2004).

31. Tsihlias, J. et al. Involvement of p27Kip1 in G1 arrest by high dose 5 alpha-dihydrotestosterone in LNCaP human prostate cancer cells. Oncogene 19, 670679 (2000).

32. Hendzel, M. J. et al. Mitosis-specic phosphorylation of histone H3 initiates primarily within pericentromeric heterochromatin during G2 and spreads in an ordered fashion coincident with mitotic chromosome condensation. Chromosoma 106, 348360 (1997).

33. Peters, J. M. The anaphase promoting complex/cyclosome: a machine designed to destroy. Nat. Rev. Mol. Cell. Biol. 7, 644656 (2006).

34. Merry, D. E., Kobayashi, Y., Bailey, C. K., Taye, A. A. & Fischbeck, K. H. Cleavage, aggregation and toxicity of the expanded androgen receptor in spinal and bulbar muscular atrophy. Hum. Mol. Genet. 7, 693701 (1998).

35. Taylor, J. P. et al. Aggresomes protect cells by enhancing the degradation of toxic polyglutamine-containing protein. Hum. Mol. Genet. 12, 749757 (2003).

36. Palazzolo, I. et al. Overexpression of IGF-1 in muscle attenuates disease in a mouse model of spinal and bulbar muscular atrophy. Neuron 63, 316328 (2009).

37. Donzelli, M. et al. Dual mode of degradation of Cdc25 A phosphatase. EMBO J. 21, 48754884 (2002).38. Harmey, D., Smith, A., Simanski, S., Moussa, C. Z. & Ayad, N. G. The anaphase promoting complex induces substrate degradation during neuronal dierentiation. J. Biol. Chem. 284, 43174323 (2009).

39. Glotzer, M., Murray, A. W. & Kirschner, M. W. Cyclin is degraded by the ubiquitin pathway. Nature 349, 132138 (1991).40. Bennett, E. J., Bence, N. F., Jayakumar, R. & Kopito, R. R. Global impairment of the ubiquitin-proteasome system by nuclear or cytoplasmic protein aggregates precedes inclusion body formation. Mol. Cell 17, 351365 (2005).

41. Dantuma, N. P., Lindsten, K., Glas, R., Jellne, M. & Masucci, M. G. Short-lived green uorescent proteins for quantifying ubiquitin/ proteasome-dependent proteolysis in living cells. Nat. Biotechnol. 18, 538543 (2000).

42. Gatchel, J. R. & Zoghbi, H. Y. Diseases of unstable repeat expansion: mechanisms and common principles. Nat. Rev. Genet. 6, 743755 (2005).

43. Konishi, Y., Stegmuller, J., Matsuda, T., Bonni, S. & Bonni, A. Cdh1-APC controls axonal growth and patterning in the mammalian brain. Science 303, 10261030 (2004).

SCIENTIFIC REPORTS

9

www.nature.com/scientificreports/

44. Lasorella, A. et al. Degradation of Id2 by the anaphase-promoting complex couples cell cycle exit and axonal growth. Nature 442, 471474 (2006).

45. Choi, E., Dial, J. M., Jeong, D. E. & Hall, M. C. Unique D box and KEN box sequences limit ubiquitination of Acm1 and promote pseudosubstrate inhibition of the anaphase-promoting complex. J. Biol. Chem. 283, 2370123710 (2008).

46. Enquist-Newman, M., Sullivan, M. & Morgan, D. O. Modulation of the mitotic regulatory network by APC-dependent destruction of the Cdh1 inhibitor Acm1. Mol. Cell 30, 437446 (2008).

47. Miller, J. J. et al. Emi1 stably binds and inhibits the anaphase-promoting complex/cyclosome as a pseudosubstrate inhibitor. Genes Dev. 20, 24102420 (2006).

48. Kazemi-Esfarjani, P., Triro, M. A. & Pinsky, L. Evidence for a repressive function of the long polyglutamine tract in the human androgen receptor: possible pathogenetic relevance for the (CAG)n-expanded neuronopathies. Hum. Mol. Genet. 4, 523527 (1995).

49. Mhatre, A. N. et al. Reduced transcriptional regulatory competence of the androgen receptor in X-linked spinal and bulbar muscular atrophy. Nat. Genet. 5, 184188 (1993).

50. Tut, T. G., Ghadessy, F. J., Triro, M. A., Pinsky, L. & Yong, E. L. Long polyglutamine tracts in the androgen receptor are associated with reduced trans-activation, impaired sperm production, and male infertility. J. Clin. Endocrinol. Metab. 82, 37773782 (1997).

51. Vismara, G. et al. Androgens inhibit androgen receptor promoter activation in motor neurons. Neurobiol. Dis. 33, 395404 (2009).52. Stegmuller, J. & Bonni, A. Moving past proliferation: new roles for Cdh1-APC in postmitotic neurons. Trends Neurosci. 28, 596601 (2005).53. Meriggioli, M. N. & Rowin, J. Fatigue and abnormal neuromuscular transmission in Kennedys disease. Muscle Nerve 27, 249251 (2003).

54. Ranganathan, S. et al. Mitochondrial abnormalities in spinal and bulbar muscular atrophy. Hum. Mol. Genet. 18, 2742 (2009).55. Su, S. et al. Mitochondrial DNA damage in spinal and bulbar muscular atrophy patients and carriers. Clin. Chim. Acta 411, 626630 (2010).56. Cortes, C. J. et al. Muscle expression of mutant androgen receptor accounts for systemic and motor neuron disease phenotypes in spinal and bulbar muscular atrophy. Neuron 82, 295307 (2014).

57. Lieberman, A. P. et al. Peripheral androgen receptor gene suppression rescues disease in mouse models of spinal and bulbar muscular atrophy. Cell Rep. 7, 774784 (2014).

58. Monks, D. A. et al. Overexpression of wild-type androgen receptor in muscle recapitulates polyglutamine disease. Proc. Natl. Acad. Sci. USA 104, 1825918264 (2007).

59. Soraru, G. et al. Spinal and bulbar muscular atrophy: skeletal muscle pathology in male patients and heterozygous females. J. Neurol. Sci. 264, 100105 (2008).

60. Kemp, M. Q. et al. Impaired motoneuronal retrograde transport in two models of SBMA implicates two sites of androgen action. Hum. Mol. Genet. 20, 44754490 (2011).

61. Oki, K. et al. Contractile dysfunction in muscle may underlie androgen-dependent motor dysfunction in spinal bulbar muscular atrophy. J. Appl. Physiol. (1985) 118, 941952 (2015).

62. Juo, P. & Kaplan, J. M. The anaphase-promoting complex regulates the abundance of GLR-1 glutamate receptors in the ventral nerve cord of C. elegans. Curr. Biol. 14, 20572062 (2004).

63. Li, L. et al. Anaphase-promoting complex/cyclosome controls HEC1 stability. Cell. Prolif. 44, 19 (2011).64. van Roessel, P., Elliott, D. A., Robinson, I. M., Prokop, A. & Brand, A. H. Independent regulation of synaptic size and activity by the anaphase-promoting complex. Cell 119, 707718 (2004).

65. Holterhus, P. M., Piefke, S. & Hiort, O. Anabolic steroids, testosterone-precursors and virilizing androgens induce distinct activation proles of androgen responsive promoter constructs. J. Steroid Biochem. Mol. Biol. 82, 269275 (2002).

Acknowledgements

We would like to thank Dr. Jacques Neefjes and Amilcar Flores-Morales for providing reagents. This work was supported by the Swedish Cancer Society (N.P.D.), the Swedish Research Council (N.P.D.), and intramural NINDS research funds (1ZIA NS003037; K.H.F.). L.C.B. was supported by the KI-NIH Graduate Partnership Program in Neuroscience.

Author Contributions

L.C.B., F.A.S., D.Me., K.H.F. and N.P.D. designed the research; L.C.B., F.A.S., D.Ma. and Y.L. performed the experiments; L.C.B., F.A.S., D.Ma., Y.L., D.Me., K.H.F. and N.P.D. analyzed the data; L.C.B. and N.P.D. wrote the paper with input from all authors.

Additional Information

Supplementary information accompanies this paper at http://www.nature.com/srep

Competing nancial interests: The authors declare no competing nancial interests.

How to cite this article: Bott, L. C. et al. The polyglutamine-expanded androgen receptor responsible for spinal and bulbar muscular atrophy inhibits the APC/CCdh1 ubiquitin ligase complex. Sci. Rep. 6, 27703; doi: 10.1038/ srep27703 (2016).

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the articles Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

SCIENTIFIC REPORTS

10