Widespread nuclear and cytoplasmic accumulation of mutant androgen rec
http://www.100md.com
《大脑学杂志》
1 Department of Neurology, Nagoya University Graduate School of Medicine, Nagoya,2 Central Research Laboratories, School of Medicine, Aichi Medical University
3 Department of Neuropathology, Institute for Medical Sciences of Aging, Aichi Medical University of Aichi, Japan
Summary
Spinal and bulbar muscular atrophy (SBMA) is an inherited adult onset motor neuron disease caused by the expansion of a polyglutamine (polyQ) tract within the androgen receptor (AR), affecting only males. The characteristic pathological finding is nuclear inclusions (NIs) consisting of mutant AR with an expanded polyQ in residual motor neurons, and in certain visceral organs. We immunohistochemically examined 11 SBMA patients at autopsy with 1C2, an antibody that specifically recognizes expanded polyQ. Our study demonstrated that diffuse nuclear accumulation of mutant AR was far more frequent and extensive than NIs being distributed in a wide array of CNS nuclei, and in more visceral organs than thus far believed. Mutant AR accumulation was also present in the cytoplasm, particularly in the Golgi apparatus; nuclear or cytoplasmic predominance of accumulation was tissue specific. Furthermore, the extent of diffuse nuclear accumulation of mutant AR in motor and sensory neurons of the spinal cord was closely related to CAG repeat length. Thus, diffuse nuclear accumulation of mutant AR apparently is a cardinal pathogenetic process underlying neurological manifestations, as in SBMA transgenic mice, while cytoplasmic accumulation may also contribute to SBMA pathophysiology.
Key Words: polyglutamine; spinal and bulbar muscular atrophy; diffuse nuclear accumulation; nuclear inclusion, cytoplasmic accumulation
Abbreviations: polyQ = polyglutamine; SBMA = spinal and bulbar muscular atrophy; AR = androgen receptor; NIs = nuclear inclusions; DRPLA = dentatorubral-pallidoluysian atrophy; CBP = CREB-binding protein
Introduction
Polyglutamine (polyQ) diseases are inherited neurodegenerative disorders caused by expansion of a trinucleotide CAG repeat in the causative genes. To date, nine polyQ diseases have been identified (Ross, 2002). Spinal and bulbar muscular atrophy (SBMA) is a polyQ disease involving mainly spinal and brainstem motor neurons (Kennedy et al., 1968; Sobue et al., 1989). In SBMA, a polymorphic CAG repeat ordinarily consisting of 14–32 CAGs is expanded to 40–62 CAGs in the first exon of the androgen receptor (AR) gene (La Spada et al., 1991; Tanaka et al., 1996), and shows somatic mosaicism (Tanaka et al., 1999). An inverse correlation exists between CAG repeat size and age at onset as well as disease severity in SBMA (Doyu et al., 1992; Igarashi et al., 1992; La Spada et al., 1992). SBMA patients develop premature muscular exhaustion, and subsequently slowly progressive muscular weakness, atrophy and fasciculations in bulbar and limb muscles (Kennedy et al., 1968; Sobue et al., 1989; Sperfeld et al., 2002). SBMA patients may also have mild sensory impairment, which usually remains subclinical (Sobue et al., 1989; Li et al., 1995; Mariotti et al., 2000). Besides these symptoms of neuronal degeneration, androgen insensitivity symptoms such as gynaecomastia, testicular atrophy and reduced fertility are common (Arbizu et al., 1983). Elevated serum creatine kinase concentrations, impaired glucose tolerance, hepatic dysfunction and hyperlipidaemia are frequent (Sobue et al., 1989; Li et al., 1995). These findings show that involvement of SBMA is not restricted to motor neurons, but extends to several visceral organs.
The cardinal pathological findings of SBMA are motor neuron loss in the spinal cord and brainstem (Sobue et al., 1989) and the presence of the nuclear inclusions (NIs), representing mutant AR, in residual motor neurons in brainstem motor nuclei, in spinal motor neurons (Li et al., 1998a) and in certain visceral organs (Li et al., 1998b). However, diffuse nuclear accumulation of the mutant protein has been detected in a more widespread distribution than NIs in a transgenic mouse model of SBMA (Katsuno et al., 2002, 2003; Adachi et al., 2003) and in models of other polyQ diseases (Schilling et al., 1999; Yvert et al., 2000; Lin et al., 2001). Such accumulation has been found to be relevant to neuronal dysfunction and eventual symptom appearance. Indeed, in dentatorubral-pallidoluysian atrophy (DRPLA), tissue distribution of diffuse nuclear accumulation of the responsible mutant protein was more widespread and more relevant to the disease severity and symptoms than that of NIs (Yamada et al., 2001a, b).
Recently we demonstrated in our transgenic mouse model that diffuse nuclear mutant AR accumulation can be prevented by reduction of circulating testosterone with castration or with an anti-androgenic agent such as leuprolerin; in treated animals, motor function and survival rate were dramatically improved (Katsuno et al., 2002, 2003), suggesting that disease manifestation in SBMA is highly testosterone dependent (Lieberman et al., 2002; Walcott and Merry, 2002a; Chevalier-Larsen et al., 2004). Indeed, a female carrier of SBMA, even if homozygous, does not express disease phenotypes (Sobue et al., 1993; Schmidt et al., 2002), presumably because circulating testosterone concentrations are low. These observations indicate that nuclear translocation and nuclear accumulation of mutant AR, detected as diffuse nuclear accumulation, is closely linked to the phenotypic expressions and that diffuse nuclear mutant AR accumulation is of major pathogenetic importance in neuronal dysfunction (Katsuno et al., 2002, 2003).
In this study, to understand better the pathophysiology of SBMA, we examined neural and non-neural tissue distributions of mutant AR accumulation in 11 SBMA patients at autopsy, using 1C2, an antibody specific for the expanded polyQ tract, as well as antibodies against AR. First, diffuse nuclear accumulation of mutant AR was far more extensive than that of NIs. Secondly, mutant AR accumulation was also present in cytoplasm, specifically in the Golgi apparatus, with predominance of nuclear or cytoplasmic accumulation being tissue specific. Thirdly, the extent of diffuse nuclear accumulation was closely related to CAG repeat length. Our present results strongly suggested that diffuse nuclear accumulation of mutant AR is of critical pathogenetic importance for motor symptoms as in the SBMA transgenic mouse model, although cytoplasmic accumulation may also contribute to the pathophysiology of SBMA.
Subjects and methods
Patients
Eleven patients with clinicopathologically and genetically confirmed SBMA (age at death, 51–84 years; mean, 66) were examined in this study (Table 1). These patients had been hospitalized and followed-up at Nagoya University Hospital and its affiliated hospitals during the past 25 years. Age at onset ranged between 20 and 75 years, and muscle weakness and bulbar symptoms had progressed for 6–53 years. Elevated serum creatine kinase and glucose was observed in many patients. Causes of death included respiratory failure related to pneumonia in seven patients, lung cancer and colon cancer in one patient each, and tuberculosis and suffocation in one patient each. At autopsy, the brain, spinal cord, dorsal root ganglia, thoracic sympathetic ganglia and various visceral organs were removed and fixed in 10% buffered formalin solution. CAG repeat length in the AR gene ranged between 40 and 50. Five other subjects (age 60–74 years, mean 67.3) who died of non-neurological diseases served as controls.
Tissue preparation and immunohistochemistry
We prepared 5 μm thick, formalin-fixed, paraffin-embedded sections of various portions of the cerebrum, brainstem, cerebellum, spinal cord, dorsal root ganglia, sympathetic ganglia, pituitary gland, peripheral nerve, muscle and non-neural visceral organs from SBMA and control subjects. Sections then were deparaffinized and rehydrated through a graded series of alcohol–water solutions. For the mutant AR immunohistochemical study, sections were pre-treated with immersion in 98% formic acid for 5 min and then with microwave oven heating for 10 min in 10 mM citrate buffer at pH 6.0. Sections were blocked with normal serum from the animal species in which each second antibody was raised (1 : 20), and then incubated with a mouse anti-expanded polyQ antibody (Trottier et al., 1995) (1C2; Chemicon, Temecula, CA; 1 : 10 000); a mouse anti-Golgi 58K protein antibody (Sigma, St. Louis, MO; 1 : 100); rabbit polyclonal antibody N-20 (Santa Cruz Biotechnology, Santa Cruz, CA; 1 : 200); rabbit polyclonal antibody PG-21 (Affinity BioReagents, Golden, CO; 1 : 200); rabbit polyclonal antibody H-280 (Santa Cruz; 1 : 200); rabbit polyclonal antibody C-19 (Santa Cruz; 1 : 200); or a mouse monoclonal antibody (Ab-1; Neomarkers, Fremont, CA; ready-to-use) against human AR protein. Then the sections were incubated with biotinylated IgG raised against the species used for each primary antibody (Vector Laboratories, Burlingame, CA). Immune complexes were visualized using streptavidin–horseradish peroxidase (Dako, Glostrup, Denmark) and 3,3'-diaminobenzidine (Dojindo, Kumamoto, Japan) substrate. Sections were counterstained with methyl green or Mayer's haematoxylin. As a negative control, primary antibodies were replaced with normal rabbit or mouse serum. The population of labelled neurons was analysed semi-quantitatively in all 11 SBMA patients, and non-neural visceral organs in nine patients (patients 1–8, and 11) by counting the positive and negative cells for labelling in the region of interest and graded as – to +++.
To assess the co-localization of cytoplasmic mutant AR accumulation and cell organelles, five selected SBMA patients (patients 1, 2, 6, 8 and 10) were analysed by double immunofluorescence staining. The sections were blocked with 5% normal serum and then sequentially incubated at 4°C overnight with any antibody to lysosomal markers, anti-cathepsin B antibody (Ab-3; Oncogene, Cambridge, MA; 1 : 20), anti-cathepsin D antibody (Ab-2; Oncogene; 1 : 20), anti-cathepsin K antibody (N-20; Santa Cruz Biotechnology; 1 : 50), anti-cathepsin L antibody (S-20; Santa Cruz Biotechnology; 1 : 50), antibody to Golgi apparatus, anti-human TGN46 antibody (Serotec, Oxford, UK; 1 : 1000), antibody to endoplasmic reticulum marker, anti-GRP78 antibody (N-20; Santa Cruz; 1 : 200), or antibody to mitochondria, anti-mitochondria antibody (Chemicon; 1 : 50), and 1C2 antibody (Chemicon; 1 : 10 000). Sections were incubated with Alexa 488-conjugated anti-mouse IgG (Molecular Probes, Leiden, The Netherlands; 1 : 1300) and Alexa 568-conjugated IgG raised against the species used for each primary antibody (Molecular Probes; 1 : 1000). For double immunofluorescence staining using anti-human TGN46 antibody, sections were incubated with biotinylated anti-sheep IgG (Vector Laboratories; 1 : 400) for 8 h at 4°C, the sections were incubated with Alexa 568-conjugated streptavidin (Molecular Probes; 1 : 1000) and Alexa 488-conjugated anti-mouse IgG (Molecular Probes; 1 : 1300) for 2 h at 4°C. Sections then were examined and photographed using a confocal laser scanning microscope (MRC 1024; Bio-Rad Laboratories, Hercules, CA).
For electron microscopic immunohistochemistry, buffered formalin-fixed, paraffin-embedded tissue sections were deparaffinized, rehydrated, immunostained with 1C2 antibody (Chemicon, 1 : 10 000), and then incubated with biotinylated anti-mouse IgG (Vector Laboratories; 1 : 1300). Immunoreactivity in tissue sections was visualized using streptavidin–horseradish peroxidase (Dako) and 3,3'-diaminobenzidine substrate (Dojindo), fixed with 2% osmium tetroxide in 0.1 mol/l phosphate buffer at pH 7.4, dehydrated in graded alcohol–water solutions, and embedded in epoxy resin. Ultrathin sections then were cut for observation under an electron microscope (H-7100; Hitachi High-Technologies Corporation, Tokyo, Japan).
Quantification of diffuse nuclear- and NI-positive cell populations
For quantitative assessment, we prepared at least 100 transverse sections each from the cervical, thoracic and lumbar spinal cord for staining with 1C2 antibody as above. The numbers of 1C2-positive and -negative cells in the ventral and dorsal horn on both right and left sides were counted on every 10th section under the light microscope with a computer-assisted image analyser (Luzex FS; Nikon, Tokyo, Japan). For the purposes of counting, a cell was defined by the presence of its nucleus in a given 5 μm thick section. Diffuse nuclear staining and NI-positive neurons were assessed separately. Neurons showing both diffuse nuclear staining and NIs were counted in both categories. The area of the ventral and dorsal horn of each spinal cord section was determined as described previously (Terao et al., 1996; Adachi et al., 2001). Populations of 1C2-positive cells were expressed as percentages of the total neuronal count. For statistical analysis, mean values of these percentages in sections examined from each of the cervical, thoracic and lumbar spinal segments for each patient were obtained.
Statistical analysis
We analysed the data by Pearson's correlation coefficient and Spearman's rank correlation as appropriate using Statview software (version 5; Hulinks, Tokyo, Japan), considering P values <0.05 to be indicative of significance.
Results
Immunohistochemical localization of mutant androgen receptor in the neural tissues
In all 11 patients with SBMA, NIs were visualized clearly with 1C2 (Fig. 1). In addition to NIs, diffusely distributed staining with 1C2 was observed in neuronal nuclei (Fig. 1). Among nuclei with diffuse staining, some showed punctate, granular or web-like patterns, while others showed intense diffuse staining (Fig. 1). In some neurons, NIs and diffuse nuclear staining co-existed (Fig. 1). Moreover, occasional neurons showed granular or punctate 1C2-positive accumulation in the cytoplasm (Fig. 1). As reported previously (Li et al., 1998a, b), NIs were observed frequently in lower motor neurons, which are known to be affected in this disease. However, we found that neuronal nuclear and cytoplasmic accumulations extended to various regions of the nervous system previously reported to be spared (Li et al., 1998a, b), including the striatum, caudate nucleus, mammillary body, thalamus, hypothalamus, reticular formation, red nucleus, substantia nigra, locus coerulus, nucleus raphe pontis, pontine nuclei, cuneate nucleus, nucleus ambiguus, gracile nucleus, supraspinal nucleus, cerebellar dentate nucleus, Clarke's nucleus, posterior horn and intermediolateral nucleus of the spinal cord, dorsal root ganglia and sympathetic ganglia (Fig. 1, Table 2). Cytoplasmic inclusions were prominent in the dorsal root ganglia neurons, and some neurons in the mammillary body, hypothalamus and facial motor nucleus and anterior and posterior horns of the spinal cord showed a slight degree of cytoplasmic accumulation (Table 2). We detected both nuclear and cytoplasmic accumulations in some neurons (Fig. 1D). No significant difference in staining pattern was evident between regions previously reported to be affected and unaffected. Diffuse nuclear staining was seen more frequently than NIs in most regions (Table 2). Relative numbers of stained neurons varied between patients, but no staining was detected in cerebral cortex, hippocampus or cerebellar cortex. In contrast to neurons, NIs and diffuse nuclear staining were very rare in glial cells. NIs stained strongly with anti-ubiquitin antibody, while cytoplasmic accumulations did not (data not shown). Anti-human AR antibodies also recognized NIs (Fig. 1N), and occasionally stained diffuse nuclear accumulations (Fig. 1O). However, cytoplasmic accumulations were not seen with anti-AR antibodies.
Electron microscopic immunohistochemistry for 1C2 demonstrated granular dense aggregates without a limiting membrane corresponding to NIs and cytoplasmic accumulations, whereas amorphous aggregates corresponded to diffuse nuclear staining in neurons (Fig. 2). No filamentous structures such as those reported in Huntington's disease, DRPLA and Machado–Joseph disease were seen. Neural tissues from five control cases were also examined in the same manner as that for SBMA cases; in these, NIs, diffuse nuclear staining and cytoplasmic accumulations were not seen, indicating that the immunohistochemical procedure with the highly diluted condition of 1C2 applied in this study did not recognize the TATA-binding protein, a transcription factor containing a stretch of polyQ residues (Trottier et al., 1995), as previously demonstrated (Yamada et al., 2001a, 2002a).
Although we did not quantitatively examine neuronal populations in this study, the motor neurons in the spinal cord and brainstem showed the most conspicuous depletion, as expected. Neurons in the posterior horn of the spinal cord, where diffuse nuclear accumulations and NIs were present in relatively high frequency, also appeared to be depleted to some extent. Quantitative assessment of neuronal cell populations in regions newly showing mutant AR accumulation will be needed.
Co-localization of cytoplasmic organelles with mutant AR
We performed immunofluorescence with double staining using primary antibodies to recognize specifically various cytoplasmic cell organelles together with 1C2 in the dorsal root ganglia, where cytoplasmic mutant AR accumulation was most prominent (Table 2). TGN46 and 1C2 were co-localized (Fig. 3), indicating that mutant AR exists in the Golgi apparatus. Spinal dorsal root ganglia neurons exhibit some granular structures immunoreactive for another Golgi apparatus marker anti-Golgi 58K protein antibody in the cytoplasm (Fig. 1P and Q). Other organelle markers, including antibodies for lysosomes, endoplasmic reticulum and mitochondria, did not show co-localization with 1C2 (Fig. 3), indicating that expanded polyQ sequences were not detected in these organelles.
Correlation of diffuse nuclear accumulation and NIs with degree of CAG repeat expansion
We examined the correlation of diffuse nuclear accumulation and NIs with the degree of CAG repeat expansion in anterior and posterior horn spinal cord neurons. Averaged frequencies of diffuse nuclear accumulations and NIs in cervical, thoracic and lumbar spinal segments were evaluated for correlation with numbers of CAG repeats in the AR gene. The frequency of diffuse nuclear accumulation in anterior and posterior horn neurons correlated well with the degree of CAG repeat expansion (Fig. 4; r = 0.78, P < 0.05 and r = 0.69, P < 0.05, respectively). However, the frequency of NIs in motor neurons and posterior horn neurons did not show a significant correlation with number of CAG repeats (Fig. 4; r = 0.05, P = NS and r = –0.14, P = NS, respectively). These observations strongly suggest that diffuse nuclear accumulation of the mutant AR protein is more important pathogenetically than NIs.
Immunohistochemical localization of mutant AR in non-neural tissues
As in neural tissues, diffuse nuclear accumulations, NIs or cytoplasmic accumulations of mutant AR were observed in certain visceral organs and skin (Fig. 5, Table 3). Diffuse nuclear accumulations and NIs were detected in the liver, proximal tubules of the kidney, testis, prostate gland, and scrotal and other skin (Fig. 5, Table 3). Cytoplasmic accumulations were detected in the liver, pancreatic islets of Langerhans, testis and prostate gland (Fig. 5, Table 3). Nuclear labelling and cytoplasmic accumulation both were absent in the pituitary gland, heart, lung, intestine, spleen, thyroid, adrenal gland and skeletal muscles. Pancreatic islet cells showed exclusively cytoplasmic accumulations without detectable nuclear accumulations, suggesting that the impaired glucose tolerance frequently observed in our patients (Table 1) could be attributed to cytoplasmic mutant AR accumulation. Ubiquitin staining detected only NIs and, as observed in neural tissues, anti-human AR antibodies occasionally showed diffuse nuclear accumulation without cytoplasmic staining (Fig. 5J). The five control cases did not show any 1C2 immunoreactivity in viscera or skin.
Discussion
The present study clearly demonstrated that diffuse nuclear accumulation of mutant AR, detected by the antibody 1C2 which specifically recognizes the expanded polyQ tract, occurred more frequently than NIs in neural and non-neural tissues in SBMA patients. In neural tissues, diffuse nuclear mutant AR accumulation occurred in the basal ganglia, thalamus, hypothalamus, various midbrain, pontine and medullary nuclei, posterior horn, intermediolateral and Clarke's nuclei of the spinal cord and in sensory and sympathetic ganglion neurons, as well as brainstem and spinal cord motor neurons. NIs detected by 1C2 were similar in distribution to diffuse nuclear accumulation, but the frequency of NIs in each tissue was far less than for diffuse nuclear accumulation. We previously demonstrated that diffuse nuclear mutant AR protein accumulation was more extensive than NIs in male SBMA transgenic mice (Katsuno et al., 2002, 2003). Furthermore, expression and severity of motor dysfunction, and abatement of abnormalities when the mice were castrated or given leuprolerin, paralleled the extent of diffuse nuclear mutant AR accumulation rather than that of NIs (Katsuno et al., 2002, 2003). Accordingly, neuronal dysfunction appeared to be closely related to diffuse nuclear mutant AR accumulation. The key observation in the present study was a significant close correlation between frequency of diffuse nuclear mutant AR accumulation and length of CAG repeat expansion, while a similar correlation was not observed between frequency of NIs and CAG repeats. Diffuse nuclear mutant protein accumulation also has been demonstrated in the neural tissues affected by DRPLA (Yamada et al., 2001a, 2002a), Huntington's disease (Sapp et al., 1997) and Machado–Joseph disease (Yamada et al., 2001b) as well as corresponding transgenic mouse models (Schilling et al., 1999; Yvert et al., 2000; Lin et al., 2001); and, here too, diffuse nuclear mutant protein accumulation was more widespread and extensive than NIs in DRPLA patients. These observations are in good agreement with our present observation in SBMA patients; together, they suggest that diffuse nuclear accumulation of mutant proteins with an expanded polyQ tract is an early event prior to NI formation that is closely related to manifestation of neuronal dysfunction (Yamada et al., 2001a; Garden et al., 2002; Katsuno et al., 2002, 2003; Watase et al., 2002; Yoo et al., 2003). However, the molecular pathogenetic process by which diffuse nuclear mutant AR accumulation induces neuronal dysfunction still is unclear. Although considerable controversy recently surrounds the importance of NIs in the pathophysiology in polyQ diseases (Simeoni et al., 2000; Walcott and Merry, 2002b; Bates, 2003; Michalik and Van Broeckhoven, 2003; Ross et al., 2003), our data showed that diffuse mutant AR accumulation in nuclei could have potent cytotoxic effects inducing neuronal dysfunction through an active epitope of the expanded polyQ tract.
Anti-AR antibodies showed the ability to detect NIs, and some of them (H280, N-20 and Ab-1) occasionally stained diffuse nuclear accumulations. Diffuse nuclear accumulation had the appearance of amorphous aggregates of mutant AR as observed by electron microscopic immunohistochemistry using 1C2 (Fig. 2) (Katsuno et al., 2002). These observations suggest that the polyQ tract epitope can be detected by 1C2, while other AR protein epitopes may be protected by structural features of the aggregate state of the mutant AR. This view is supported by observations made by small-angle X-ray scattering and infrared spectroscopy carried out with myoglobin protein containing an inserted highly expanded polyQ tract, localizing the polyQ tract to the surface of aggregates, while other epitopes were sequestered within aggregates (Tanaka et al., 2001, 2003). These observations suggest that 1C2 can detect the amorphous aggregate state of mutant AR protein, making 1C2 a more sensitive histological and pathophysiological marker than anti-AR protein antibodies. On the other hand, cytoplasmic accumulations were not seen with anti-AR antibodies. This cytoplasmic mutant AR was not ubiquitinated, in contrast to nuclear accumulated mutant AR, particularly the heavily ubiquitinated NIs, suggesting that protein modification varies between the nucleus and cytoplasm. Different protein modification might mask other AR protein epitopes directly or through structural alterations of the aggregate state of the mutant AR in the cytoplasm.
Another important observation in our study was the occurrence of cytoplasmic mutant AR accumulation in neural and non-neural tissues. In neural tissues, cytoplasmic accumulation was restricted to certain neuronal populations such as dorsal root ganglia neurons, mammillary body, hypothalamus, facial motor nucleus, and anterior and posterior horn neurons. In non-neural tissues, cytoplasmic accumulation also occurred in certain organs. Cytoplasmic mutant AR accumulation co-localized with a Golgi apparatus marker. Co-localization of a polyQ-expanded mutant protein with the Golgi apparatus has also been reported for ataxin-2 (Huynh et al., 2003), although the significance of this localization remains unclear. Expression of polyQ-expanded mutant ataxin-2 disrupted the normal morphology of the Golgi complex and increased cell death (Huynh et al., 2003). On the other hand, the lysosomal occurrence of other mutant proteins with an expanded polyQ tract has been reported in DRPLA (Yamada et al., 2002b) and Huntington's disease (Sapp et al., 1997). The lysosomal localization of polyQ-expanded mutant proteins suggests a lysosomal autophagic degradation process acting independently of the ubiquitin–proteasome pathway in the polyQ diseases (Sapp et al., 1997). Additionally, the reason why neural tissues develop more nuclear than cytoplasmic accumulation while most involved visceral organs show equal or predominantly cytoplasmic accumulation is unknown. Differences in the predominant degradation pathway dealing with the mutant AR could influence the intracellular site of accumulation and eventual cell toxicity. One important question is whether cytoplasmic mutant AR accumulation exerts cytotoxicity in neural and non-neural tissues. Cytoplasmic mutant AR accumulation (Taylor et al., 2003) as well as other mutant protein accumulations (Kegel et al., 2000; Ravikumar et al., 2002; Huynh et al., 2003) involving an expanded polyQ tract in Golgi apparatus and lysosomes indeed has been found to induce cytotoxicity. Accumulation of mutant protein with expanded polyQ in the Golgi apparatus or lysosomes increases death of cultured cells through activation of apoptosis-related effectors such as caspase-3 (Ishisaka et al., 1998; Kegel et al., 2000; Huynh et al., 2003). One should note that histologically or immunohistochemically evident mutant protein accumulation is not necessarily cytotoxic, while microaggregates at the molecular level that are histologically undetectable can also exert cytotoxicity. Indeed, excessive accumulation of mutant AR in aggresomes was found to protect cells from a cytotoxic form of mutant AR (Taylor et al., 2003). However, our present study strongly suggests that these cytoplasmic mutant AR accumulations may be related to mutant AR-mediated cytotoxicity and eventual symptom manifestation. For instance, the pancreas showed only cytoplasmic mutant AR accumulation without obvious nuclear accumulation. Elevated serum glucose and impaired glucose tolerance were present in most of our patients, suggesting islet cell dysfunction in the pancreas. The frequency of cytoplasmic accumulations of mutant AR in pancreatic islet cells did not show a significant correlation with fasting blood glucose levels in the examined SBMA patients (data not shown), while certain symptoms and signs of SBMA apparently can be induced by cytoplasmic accumulation of mutant AR protein. Although further study of the significance of cytoplasmic mutant AR accumulation is needed, nuclear accumulation of the mutant AR protein appears to cause motor neuron dysfunction while cytoplasmic accumulation may underlie some visceral and possibly some neuronal dysfunction in SBMA. The pathological process is likely to differ between tissues, being more prominent in motor neuron nuclei, but mainly cytoplasmic in certain neuronal populations and visceral organs. We also need to clarify further which degradation process affecting mutant AR is most active in a given tissue, e.g. lysosomal in certain viscera versus via ubiquitination pathway in most neural tissues.
An important question here is why diffuse nuclear and possibly cytoplasmic accumulation of the mutant AR in the neuronal tissues beyond the major affected spinal and brainstem motor neurons has no apparent symptomatic involvement. First, the causative lesions for sensory impairment and essential-type tremor in SBMA patients have not yet been clearly substantiated. The novel lesion distribution of SBMA neurons shown in the present study, such as the posterior horn of the spinal cord, dorsal root ganglia, thalamus and cerebellum, might provide some explanations for these clinicopathological problems that have not been resolved. Since the cerebellothalamocortical pathway seems to be responsible for essential-type tremor (Pinto et al., 2003), these lesions might contribute to mostly subclinical but definite sensory impairment and essential-type tremor in SBMA. Secondly, the occurrence of neuronal nuclear and cytoplasmic abnormalities in both clinically affected and non-affected neural regions in SBMA suggests that this alteration does not always induce neuronal cell dysfunction or death. The selective neuronal loss and dysfunction in neural lesions that are characteristic of SBMA might depend on additional factors that are specific to neurons in these systems. Recent studies have demonstrated that CREB-binding protein (CBP) is sequestered in AR-positive NIs, resulting in a decrease in CBP-dependent transcription (McCampbell et al., 2000), and further histone acetylation is reduced in affected cells (McCampbell et al., 2001; Steffan et al., 2001; Minamiyama et al., 2004). These reports suggest that CBP-dependent transcriptional dysregulations may cause symptomatic neuronal dysfunction. Since CBP-dependent transcriptional control differs among neurons, this difference may show the lack of their symptomatic involvement in certain polyQ-containing neurons. Alternatively, the population of neurons with nuclear accumulation of mutant AR in the regions beyond the commonly affected lesions may not be simply enough to manifest the responsible symptoms. A precise neuronal cell count assay combined with assessment of nuclear mutant AR accumulation will be needed to clarify these clinicopathological problems.
Clearly, motor neuron impairment with nuclear accumulation of mutant AR is the major problem in SBMA. Thus, for a therapeutic strategy against motor neuron dysfunction in SBMA, nuclear accumulation of mutant AR should be the main target, as we demonstrated in transgenic mice treated with leuprolerin. Cytoplasmic accumulation of mutant AR, on the other hand, should be considered a therapeutic target with respect to certain symptoms in SBMA patients.
Acknowledgements
We wish to thank Dr Yasushi Iwasaki and Mrs Sugiko Yokoi for technical assistance. This work was supported by a Center of Excellence (COE) grant from the Ministry of Education, Culture, Sports, Science, and Technology of Japan, and grants from the Ministry of Health, Labor, and Welfare of Japan.
References
Adachi H, Kume A, Li M, Nakagomi Y, Niwa H, Do J, et al. Transgenic mice with an expanded CAG repeat controlled by the human AR promoter show polyglutamine nuclear inclusions and neuronal dysfunction without neuronal cell death. Hum Mol Genet 2001; 10: 1039–48.
Adachi H, Katsuno M, Minamiyama M, Sang C, Pagoulatos G, Angelidis C, et al. Heat shock protein 70 chaperone overexpression ameliorates phenotypes of the spinal and bulbar muscular atrophy transgenic mouse model by reducing nuclear-localized mutant androgen receptor protein. J Neurosci 2003; 23: 2203–11.
Arbizu T, Santamaria J, Gomez JM, Quilez A, Serra JP. A family with adult spinal and bulbar muscular atrophy, X-linked inheritance and associated testicular failure. J Neurol Sci 1983; 59: 371–82.
Bates G. Huntingtin aggregation and toxicity in Huntington's disease. Lancet 2003; 361: 1642–4.
Chevalier-Larsen ES, O'Brien CJ, Wang H, Jenkins SC, Holder L, Lieberman AP, et al. Castration restores function and neurofilament alterations of aged symptomatic males in a transgenic mouse model of spinal and bulbar muscular atrophy. J Neurosci 2004; 24: 4778–86.
Doyu M, Sobue G, Mukai E, Kachi T, Yasuda T, Mitsuma T, et al. Severity of X-linked recessive bulbospinal neuronopathy correlates with size of the tandem CAG repeat in androgen receptor gene. Ann Neurol 1992; 32: 707–10.
Garden GA, Libby RT, Fu YH, Kinoshita Y, Huang J, Possin DE, et al. Polyglutamine-expanded ataxin-7 promotes non-cell-autonomous Purkinje cell degeneration and displays proteolytic cleavage in ataxic transgenic mice. J Neurosci 2002; 22: 4897–905.
Huynh DP, Yang HT, Vakharia H, Nguyen D, Pulst SM. Expansion of the polyQ repeat in ataxin-2 alters its Golgi localization, disrupts the Golgi complex and causes cell death. Hum Mol Genet 2003; 12: 1485–96.
Igarashi S, Tanno Y, Onodera O, Yamazaki M, Sato S, Ishikawa A, et al. Strong correlation between the number of CAG repeats in androgen receptor genes and the clinical onset of features of spinal and bulbar muscular atrophy. Neurology 1992; 42: 2300–2.
Ishisaka R, Utsumi T, Yabuki M, Kanno T, Furuno T, Inoue M, et al. Activation of caspase-3-like protease by digitonin-treated lysosomes. FEBS Lett 1998; 435: 233–6.
Katsuno M, Adachi H, Kume A, Li M, Nakagomi Y, Niwa H, et al. Testosterone reduction prevents phenotypic expression in a transgenic mouse model of spinal and bulbar muscular atrophy. Neuron 2002; 35: 843–54.
Katsuno M, Adachi H, Doyu M, Minamiyama M, Sang C, Kobayashi Y, et al. Leuprorelin rescues polyglutamine-dependent phenotypes in a transgenic mouse model of spinal and bulbar muscular atrophy. Nat Med 2003; 9: 768–73.
Kegel KB, Kim M, Sapp E, McIntyre C, Castano JG, Aronin N, et al. Huntingtin expression stimulates endosomal-lysosomal activity, endosome tubulation and autophagy. J Neurosci 2000; 20: 7268–78.
Kennedy WR, Alter M, Sung JH. Progressive proximal spinal and bulbar muscular atrophy of late onset. A sex-linked recessive trait. Neurology 1968; 18: 671–80.
La Spada AR, Wilson EM, Lubahn DB, Harding AE, Fischbeck KH. Androgen receptor gene mutations in X-linked spinal and bulbar muscular atrophy. Nature 1991; 352: 77–9.
La Spada AR, Roling DB, Harding AE, Warner CL, Spiegel R, Hausmanowa-Petrusewicz I, et al. Meiotic stability and genotype–phenotype correlation of the trinucleotide repeat in X-linked spinal and bulbar muscular atrophy. Nat Genet 1992; 2: 301–4.
Li M, Sobue G, Doyu M, Mukai E, Hashizume Y, Mitsuma T. Primary sensory neurons in X-linked recessive bulbospinal neuropathy: histopathology and androgen receptor gene expression. Muscle Nerve 1995; 18: 301–8.
Li M, Miwa S, Kobayashi Y, Merry DE, Yamamoto M, Tanaka F, et al. Nuclear inclusions of the androgen receptor protein in spinal and bulbar muscular atrophy. Ann Neurol 1998a; 44: 249–54.
Li M, Nakagomi Y, Kobayashi Y, Merry DE, Tanaka F, Doyu M, et al. Nonneural nuclear inclusions of androgen receptor protein in spinal and bulbar muscular atrophy. Am J Pathol 1998b; 153: 695–701.
Lieberman AP, Harmison G, Strand AD, Olson JM, Fischbeck KH. Altered transcriptional regulation in cells expressing the expanded polyglutamine androgen receptor. Hum Mol Genet 2002; 11: 1967–76.
Lin CH, Tallaksen-Greene S, Chien WM, Cearley JA, Jackson WS, Crouse AB, et al. Neurological abnormalities in a knock-in mouse model of Huntington's disease. Hum Mol Genet 2001; 10: 137–44.
Mariotti C, Castellotti B, Pareyson D, Testa D, Eoli M, Antozzi C, et al. Phenotypic manifestations associated with CAG-repeat expansion in the androgen receptor gene in male patients and heterozygous females: a clinical and molecular study of 30 families. Neuromuscul Disord 2000; 10: 391–7.
McCampbell A, Taylor JP, Taye AA, Robitschek J, Li M, Walcott J, et al. CREB-binding protein sequestration by expanded polyglutamine. Hum Mol Genet 2000; 9: 2197–202.
McCampbell A, Taye AA, Whitty L, Penney E, Steffan JS, Fischbeck KH. Histone deacetylase inhibitors reduce polyglutamine toxicity. Proc Natl Acad Sci USA 2001; 98: 15179–84.
Michalik A, Van Broeckhoven C. Pathogenesis of polyglutamine disorders: aggregation revisited. Hum Mol Genet 2003; 12: R173–86.
Minamiyama M, Katsuno M, Adachi H, Waza M, Sang C, Kobayashi Y, et al. Sodium butyrate ameliorates phenotypic expression in a transgenic mouse model of spinal and bulbar muscular atrophy. Hum Mol Genet 2004; 13: 1183–92.
Pinto AD, Lang AE, Chen R. The cerebellothalamocortical pathway in essential tremor. Neurology 2003; 60: 1985–7.
Ravikumar B, Duden R, Rubinsztein DC. Aggregate-prone proteins with polyglutamine and polyalanine expansions are degraded by autophagy. Hum Mol Genet 2002; 11: 1107–17.
Ross CA. Polyglutamine pathogenesis: emergence of unifying mechanisms for Huntington's disease and related disorders. Neuron 2002; 35: 819–22.
Ross CA, Poirier MA, Wanker EE, Amzel M. Polyglutamine fibrillogenesis: the pathway unfolds. Proc Natl Acad Sci USA 2003; 100: 1–3.
Sapp E, Schwarz C, Chase K, Bhide PG, Young AB, Penney J, et al. Huntingtin localization in brains of normal and Huntington's disease patients. Ann Neurol 1997; 42: 604–12.
Schilling G, Wood JD, Duan K, Slunt HH, Gonzales V, Yamada M, et al. Nuclear accumulation of truncated atrophin-1 fragments in a transgenic mouse model of DRPLA. Neuron 1999; 24: 275–86.
Schmidt BJ, Greenberg CR, Allingham–Hawkins DJ, Spriggs EL. Expression of X-linked bulbospinal muscular atrophy (Kennedy disease) in two homozygous women. Neurology 2002; 59: 770–2.
Simeoni S, Mancini MA, Stenoien DL, Marcelli M, Weigel NL, Zanisi M, et al. Motoneuronal cell death is not correlated with aggregate formation of androgen receptors containing an elongated polyglutamine tract. Hum Mol Genet 2000; 9: 133–44.
Sobue G, Hashizume Y, Mukai E, Hirayama M, Mitsuma T, Takahashi A. X-linked recessive bulbospinal neuronopathy. A clinicopathological study. Brain 1989; 12: 209–32.
Sobue G, Doyu M, Kachi T, Yasuda T, Mukai E, Kumagai T, et al. Subclinical phenotypic expressions in heterozygous females of X-linked recessive bulbospinal neuronopathy. J Neurol Sci 1993; 117: 74–8.
Sperfeld AD, Karitzky J, Brummer D, Schreiber H, Haussler J, Ludolph AC, et al. X-linked bulbospinal neuronopathy: Kennedy disease. Arch Neurol 2002; 59: 1921–6.
Steffan JS, Bodai L, Pallos J, Poelman M, McCampbell A, Apostol BL, et al. Histone deacetylase inhibitors arrest polyglutamine-dependent neurodegeneration in Drosophila. Nature 2001; 413: 739–43.
Tanaka F, Doyu M, Ito Y, Matsumoto M, Mitsuma T, Abe K, et al. Founder effect in spinal and bulbar muscular atrophy (SBMA). Hum Mol Genet 1996; 5: 1253–7.
Tanaka F, Reeves MF, Ito Y, Matsumoto M, Li M, Miwa S, et al. Tissue-specific somatic mosaicism in spinal and bulbar muscular atrophy (SBMA) is dependent on CAG repeat length and androgen receptor gene expression level. Am J Hum Genet 1999; 65: 966–73.
Tanaka M, Morishima I, Akagi T, Hashikawa T, Nukina N. Intra- and intermolecular beta-plated sheet formation in glutamine-repeat inserted myoglobin as a model for polyglutamine diseases. J Biol Chem 2001; 276: 45470–5.
Tanaka M, Machida Y, Nishikawa Y, Akagi T, Hashikawa T, Fujisawa T, et al. Expansion of polyglutamine induces the formation of quasi-aggregate in the early stage of protein fibrillization. J Biol Chem 2003; 278: 34717–24.
Taylor JP, Tanaka F, Robitschek J, Sandoval CM, Taye A, Markovic-Plese S, et al. Aggresomes protect cells by enhancing the degradation of toxic polyglutamine-containing protein. Hum Mol Genet 2003; 12: 749–57.
Terao S, Sobue G, Hashizume Y, Li M, Inagaki T, Mitsuma T. Age-related changes in human spinal ventral horn cells with special reference to the loss of small neurons in the intermediate zone: a quantitative analysis. Acta Neuropathol (Berl) 1996; 92: 109–14.
Trottier Y, Lutz Y, Stevanin G, Imbert G, Devys D, Cancel G, et al. Polyglutamine expansion as a pathological epitope in Huntingtin's disease and four dominant cerebellar ataxias. Nature 1995; 378: 403–6.
Walcott JL, Merry DE. Ligand promotes intranuclear inclusions in a novel cell model of spinal and bulbar muscular atrophy. J Biol Chem 2002a; 277: 50855–9.
Walcott JL, Merry DE. Trinucleotide repeat disease. The androgen receptor in spinal and bulbar muscular atrophy. Vitam Horm 2002b; 65: 127–47.
Watase K, Weeber EJ, Xu B, Antalffy B, Yuva-Paylor L, Hashimoto K, et al. A long CAG repeat in the mouse Sca1 locus replicates SCA1 features and reveals the impact of protein solubility on selective neurodegeneration. Neuron 2002; 34: 905–19.
Yamada M, Wood JD, Shimohata T, Hayashi S, Tsuji S, Ross CA, et al. Widespread occurrence of intranuclear atrophin-1 accumulation in the central nervous system neurons of patients with dentatorubral-pallidoluysian atrophy. Ann Neurol 2001a; 49: 14–23.
Yamada M, Hayashi S, Tsuji S, Takahashi H. Involvement of the cerebral cortex and autonomic ganglia in Machado–Joseph disease. Acta Neuropathol (Berl) 2001b; 101: 140–4.
Yamada M, Sato T, Tsuji S, Takahashi H. Oligodendrocytic polyglutamine pathology in dentatorubral-pallidoluysian atrophy. Ann Neurol 2002a; 52: 670–4.
Yamada M, Tsuji S, Takahashi H. Involvement of lysosomes in the pathogenesis of CAG repeat diseases. Ann Neurol 2002b; 52: 498–503.
Yoo SY, Pennesi ME, Weeber EJ, Xu B, Atkinson R, Chen S, et al. SCA7 knock in mice model human SCA7 and reveal gradual accumulation of mutant ataxin-7 in neurons and abnormalities in short-term plasticity. Neuron 2003; 37: 383–401.
Yvert G, Lindenberg KS, Picaud S, Landwehrmeyer GB, Sahel JA, Mandel JL. Expanded polyglutamines induce neurodegeneration and trans-neuronal alterations in cerebellum and retina of SCA7 transgenic mice. Hum Mol Genet 2000; 9: 2491–506.(Hiroaki Adachi, Masahisa Katsuno, Makoto)
3 Department of Neuropathology, Institute for Medical Sciences of Aging, Aichi Medical University of Aichi, Japan
Summary
Spinal and bulbar muscular atrophy (SBMA) is an inherited adult onset motor neuron disease caused by the expansion of a polyglutamine (polyQ) tract within the androgen receptor (AR), affecting only males. The characteristic pathological finding is nuclear inclusions (NIs) consisting of mutant AR with an expanded polyQ in residual motor neurons, and in certain visceral organs. We immunohistochemically examined 11 SBMA patients at autopsy with 1C2, an antibody that specifically recognizes expanded polyQ. Our study demonstrated that diffuse nuclear accumulation of mutant AR was far more frequent and extensive than NIs being distributed in a wide array of CNS nuclei, and in more visceral organs than thus far believed. Mutant AR accumulation was also present in the cytoplasm, particularly in the Golgi apparatus; nuclear or cytoplasmic predominance of accumulation was tissue specific. Furthermore, the extent of diffuse nuclear accumulation of mutant AR in motor and sensory neurons of the spinal cord was closely related to CAG repeat length. Thus, diffuse nuclear accumulation of mutant AR apparently is a cardinal pathogenetic process underlying neurological manifestations, as in SBMA transgenic mice, while cytoplasmic accumulation may also contribute to SBMA pathophysiology.
Key Words: polyglutamine; spinal and bulbar muscular atrophy; diffuse nuclear accumulation; nuclear inclusion, cytoplasmic accumulation
Abbreviations: polyQ = polyglutamine; SBMA = spinal and bulbar muscular atrophy; AR = androgen receptor; NIs = nuclear inclusions; DRPLA = dentatorubral-pallidoluysian atrophy; CBP = CREB-binding protein
Introduction
Polyglutamine (polyQ) diseases are inherited neurodegenerative disorders caused by expansion of a trinucleotide CAG repeat in the causative genes. To date, nine polyQ diseases have been identified (Ross, 2002). Spinal and bulbar muscular atrophy (SBMA) is a polyQ disease involving mainly spinal and brainstem motor neurons (Kennedy et al., 1968; Sobue et al., 1989). In SBMA, a polymorphic CAG repeat ordinarily consisting of 14–32 CAGs is expanded to 40–62 CAGs in the first exon of the androgen receptor (AR) gene (La Spada et al., 1991; Tanaka et al., 1996), and shows somatic mosaicism (Tanaka et al., 1999). An inverse correlation exists between CAG repeat size and age at onset as well as disease severity in SBMA (Doyu et al., 1992; Igarashi et al., 1992; La Spada et al., 1992). SBMA patients develop premature muscular exhaustion, and subsequently slowly progressive muscular weakness, atrophy and fasciculations in bulbar and limb muscles (Kennedy et al., 1968; Sobue et al., 1989; Sperfeld et al., 2002). SBMA patients may also have mild sensory impairment, which usually remains subclinical (Sobue et al., 1989; Li et al., 1995; Mariotti et al., 2000). Besides these symptoms of neuronal degeneration, androgen insensitivity symptoms such as gynaecomastia, testicular atrophy and reduced fertility are common (Arbizu et al., 1983). Elevated serum creatine kinase concentrations, impaired glucose tolerance, hepatic dysfunction and hyperlipidaemia are frequent (Sobue et al., 1989; Li et al., 1995). These findings show that involvement of SBMA is not restricted to motor neurons, but extends to several visceral organs.
The cardinal pathological findings of SBMA are motor neuron loss in the spinal cord and brainstem (Sobue et al., 1989) and the presence of the nuclear inclusions (NIs), representing mutant AR, in residual motor neurons in brainstem motor nuclei, in spinal motor neurons (Li et al., 1998a) and in certain visceral organs (Li et al., 1998b). However, diffuse nuclear accumulation of the mutant protein has been detected in a more widespread distribution than NIs in a transgenic mouse model of SBMA (Katsuno et al., 2002, 2003; Adachi et al., 2003) and in models of other polyQ diseases (Schilling et al., 1999; Yvert et al., 2000; Lin et al., 2001). Such accumulation has been found to be relevant to neuronal dysfunction and eventual symptom appearance. Indeed, in dentatorubral-pallidoluysian atrophy (DRPLA), tissue distribution of diffuse nuclear accumulation of the responsible mutant protein was more widespread and more relevant to the disease severity and symptoms than that of NIs (Yamada et al., 2001a, b).
Recently we demonstrated in our transgenic mouse model that diffuse nuclear mutant AR accumulation can be prevented by reduction of circulating testosterone with castration or with an anti-androgenic agent such as leuprolerin; in treated animals, motor function and survival rate were dramatically improved (Katsuno et al., 2002, 2003), suggesting that disease manifestation in SBMA is highly testosterone dependent (Lieberman et al., 2002; Walcott and Merry, 2002a; Chevalier-Larsen et al., 2004). Indeed, a female carrier of SBMA, even if homozygous, does not express disease phenotypes (Sobue et al., 1993; Schmidt et al., 2002), presumably because circulating testosterone concentrations are low. These observations indicate that nuclear translocation and nuclear accumulation of mutant AR, detected as diffuse nuclear accumulation, is closely linked to the phenotypic expressions and that diffuse nuclear mutant AR accumulation is of major pathogenetic importance in neuronal dysfunction (Katsuno et al., 2002, 2003).
In this study, to understand better the pathophysiology of SBMA, we examined neural and non-neural tissue distributions of mutant AR accumulation in 11 SBMA patients at autopsy, using 1C2, an antibody specific for the expanded polyQ tract, as well as antibodies against AR. First, diffuse nuclear accumulation of mutant AR was far more extensive than that of NIs. Secondly, mutant AR accumulation was also present in cytoplasm, specifically in the Golgi apparatus, with predominance of nuclear or cytoplasmic accumulation being tissue specific. Thirdly, the extent of diffuse nuclear accumulation was closely related to CAG repeat length. Our present results strongly suggested that diffuse nuclear accumulation of mutant AR is of critical pathogenetic importance for motor symptoms as in the SBMA transgenic mouse model, although cytoplasmic accumulation may also contribute to the pathophysiology of SBMA.
Subjects and methods
Patients
Eleven patients with clinicopathologically and genetically confirmed SBMA (age at death, 51–84 years; mean, 66) were examined in this study (Table 1). These patients had been hospitalized and followed-up at Nagoya University Hospital and its affiliated hospitals during the past 25 years. Age at onset ranged between 20 and 75 years, and muscle weakness and bulbar symptoms had progressed for 6–53 years. Elevated serum creatine kinase and glucose was observed in many patients. Causes of death included respiratory failure related to pneumonia in seven patients, lung cancer and colon cancer in one patient each, and tuberculosis and suffocation in one patient each. At autopsy, the brain, spinal cord, dorsal root ganglia, thoracic sympathetic ganglia and various visceral organs were removed and fixed in 10% buffered formalin solution. CAG repeat length in the AR gene ranged between 40 and 50. Five other subjects (age 60–74 years, mean 67.3) who died of non-neurological diseases served as controls.
Tissue preparation and immunohistochemistry
We prepared 5 μm thick, formalin-fixed, paraffin-embedded sections of various portions of the cerebrum, brainstem, cerebellum, spinal cord, dorsal root ganglia, sympathetic ganglia, pituitary gland, peripheral nerve, muscle and non-neural visceral organs from SBMA and control subjects. Sections then were deparaffinized and rehydrated through a graded series of alcohol–water solutions. For the mutant AR immunohistochemical study, sections were pre-treated with immersion in 98% formic acid for 5 min and then with microwave oven heating for 10 min in 10 mM citrate buffer at pH 6.0. Sections were blocked with normal serum from the animal species in which each second antibody was raised (1 : 20), and then incubated with a mouse anti-expanded polyQ antibody (Trottier et al., 1995) (1C2; Chemicon, Temecula, CA; 1 : 10 000); a mouse anti-Golgi 58K protein antibody (Sigma, St. Louis, MO; 1 : 100); rabbit polyclonal antibody N-20 (Santa Cruz Biotechnology, Santa Cruz, CA; 1 : 200); rabbit polyclonal antibody PG-21 (Affinity BioReagents, Golden, CO; 1 : 200); rabbit polyclonal antibody H-280 (Santa Cruz; 1 : 200); rabbit polyclonal antibody C-19 (Santa Cruz; 1 : 200); or a mouse monoclonal antibody (Ab-1; Neomarkers, Fremont, CA; ready-to-use) against human AR protein. Then the sections were incubated with biotinylated IgG raised against the species used for each primary antibody (Vector Laboratories, Burlingame, CA). Immune complexes were visualized using streptavidin–horseradish peroxidase (Dako, Glostrup, Denmark) and 3,3'-diaminobenzidine (Dojindo, Kumamoto, Japan) substrate. Sections were counterstained with methyl green or Mayer's haematoxylin. As a negative control, primary antibodies were replaced with normal rabbit or mouse serum. The population of labelled neurons was analysed semi-quantitatively in all 11 SBMA patients, and non-neural visceral organs in nine patients (patients 1–8, and 11) by counting the positive and negative cells for labelling in the region of interest and graded as – to +++.
To assess the co-localization of cytoplasmic mutant AR accumulation and cell organelles, five selected SBMA patients (patients 1, 2, 6, 8 and 10) were analysed by double immunofluorescence staining. The sections were blocked with 5% normal serum and then sequentially incubated at 4°C overnight with any antibody to lysosomal markers, anti-cathepsin B antibody (Ab-3; Oncogene, Cambridge, MA; 1 : 20), anti-cathepsin D antibody (Ab-2; Oncogene; 1 : 20), anti-cathepsin K antibody (N-20; Santa Cruz Biotechnology; 1 : 50), anti-cathepsin L antibody (S-20; Santa Cruz Biotechnology; 1 : 50), antibody to Golgi apparatus, anti-human TGN46 antibody (Serotec, Oxford, UK; 1 : 1000), antibody to endoplasmic reticulum marker, anti-GRP78 antibody (N-20; Santa Cruz; 1 : 200), or antibody to mitochondria, anti-mitochondria antibody (Chemicon; 1 : 50), and 1C2 antibody (Chemicon; 1 : 10 000). Sections were incubated with Alexa 488-conjugated anti-mouse IgG (Molecular Probes, Leiden, The Netherlands; 1 : 1300) and Alexa 568-conjugated IgG raised against the species used for each primary antibody (Molecular Probes; 1 : 1000). For double immunofluorescence staining using anti-human TGN46 antibody, sections were incubated with biotinylated anti-sheep IgG (Vector Laboratories; 1 : 400) for 8 h at 4°C, the sections were incubated with Alexa 568-conjugated streptavidin (Molecular Probes; 1 : 1000) and Alexa 488-conjugated anti-mouse IgG (Molecular Probes; 1 : 1300) for 2 h at 4°C. Sections then were examined and photographed using a confocal laser scanning microscope (MRC 1024; Bio-Rad Laboratories, Hercules, CA).
For electron microscopic immunohistochemistry, buffered formalin-fixed, paraffin-embedded tissue sections were deparaffinized, rehydrated, immunostained with 1C2 antibody (Chemicon, 1 : 10 000), and then incubated with biotinylated anti-mouse IgG (Vector Laboratories; 1 : 1300). Immunoreactivity in tissue sections was visualized using streptavidin–horseradish peroxidase (Dako) and 3,3'-diaminobenzidine substrate (Dojindo), fixed with 2% osmium tetroxide in 0.1 mol/l phosphate buffer at pH 7.4, dehydrated in graded alcohol–water solutions, and embedded in epoxy resin. Ultrathin sections then were cut for observation under an electron microscope (H-7100; Hitachi High-Technologies Corporation, Tokyo, Japan).
Quantification of diffuse nuclear- and NI-positive cell populations
For quantitative assessment, we prepared at least 100 transverse sections each from the cervical, thoracic and lumbar spinal cord for staining with 1C2 antibody as above. The numbers of 1C2-positive and -negative cells in the ventral and dorsal horn on both right and left sides were counted on every 10th section under the light microscope with a computer-assisted image analyser (Luzex FS; Nikon, Tokyo, Japan). For the purposes of counting, a cell was defined by the presence of its nucleus in a given 5 μm thick section. Diffuse nuclear staining and NI-positive neurons were assessed separately. Neurons showing both diffuse nuclear staining and NIs were counted in both categories. The area of the ventral and dorsal horn of each spinal cord section was determined as described previously (Terao et al., 1996; Adachi et al., 2001). Populations of 1C2-positive cells were expressed as percentages of the total neuronal count. For statistical analysis, mean values of these percentages in sections examined from each of the cervical, thoracic and lumbar spinal segments for each patient were obtained.
Statistical analysis
We analysed the data by Pearson's correlation coefficient and Spearman's rank correlation as appropriate using Statview software (version 5; Hulinks, Tokyo, Japan), considering P values <0.05 to be indicative of significance.
Results
Immunohistochemical localization of mutant androgen receptor in the neural tissues
In all 11 patients with SBMA, NIs were visualized clearly with 1C2 (Fig. 1). In addition to NIs, diffusely distributed staining with 1C2 was observed in neuronal nuclei (Fig. 1). Among nuclei with diffuse staining, some showed punctate, granular or web-like patterns, while others showed intense diffuse staining (Fig. 1). In some neurons, NIs and diffuse nuclear staining co-existed (Fig. 1). Moreover, occasional neurons showed granular or punctate 1C2-positive accumulation in the cytoplasm (Fig. 1). As reported previously (Li et al., 1998a, b), NIs were observed frequently in lower motor neurons, which are known to be affected in this disease. However, we found that neuronal nuclear and cytoplasmic accumulations extended to various regions of the nervous system previously reported to be spared (Li et al., 1998a, b), including the striatum, caudate nucleus, mammillary body, thalamus, hypothalamus, reticular formation, red nucleus, substantia nigra, locus coerulus, nucleus raphe pontis, pontine nuclei, cuneate nucleus, nucleus ambiguus, gracile nucleus, supraspinal nucleus, cerebellar dentate nucleus, Clarke's nucleus, posterior horn and intermediolateral nucleus of the spinal cord, dorsal root ganglia and sympathetic ganglia (Fig. 1, Table 2). Cytoplasmic inclusions were prominent in the dorsal root ganglia neurons, and some neurons in the mammillary body, hypothalamus and facial motor nucleus and anterior and posterior horns of the spinal cord showed a slight degree of cytoplasmic accumulation (Table 2). We detected both nuclear and cytoplasmic accumulations in some neurons (Fig. 1D). No significant difference in staining pattern was evident between regions previously reported to be affected and unaffected. Diffuse nuclear staining was seen more frequently than NIs in most regions (Table 2). Relative numbers of stained neurons varied between patients, but no staining was detected in cerebral cortex, hippocampus or cerebellar cortex. In contrast to neurons, NIs and diffuse nuclear staining were very rare in glial cells. NIs stained strongly with anti-ubiquitin antibody, while cytoplasmic accumulations did not (data not shown). Anti-human AR antibodies also recognized NIs (Fig. 1N), and occasionally stained diffuse nuclear accumulations (Fig. 1O). However, cytoplasmic accumulations were not seen with anti-AR antibodies.
Electron microscopic immunohistochemistry for 1C2 demonstrated granular dense aggregates without a limiting membrane corresponding to NIs and cytoplasmic accumulations, whereas amorphous aggregates corresponded to diffuse nuclear staining in neurons (Fig. 2). No filamentous structures such as those reported in Huntington's disease, DRPLA and Machado–Joseph disease were seen. Neural tissues from five control cases were also examined in the same manner as that for SBMA cases; in these, NIs, diffuse nuclear staining and cytoplasmic accumulations were not seen, indicating that the immunohistochemical procedure with the highly diluted condition of 1C2 applied in this study did not recognize the TATA-binding protein, a transcription factor containing a stretch of polyQ residues (Trottier et al., 1995), as previously demonstrated (Yamada et al., 2001a, 2002a).
Although we did not quantitatively examine neuronal populations in this study, the motor neurons in the spinal cord and brainstem showed the most conspicuous depletion, as expected. Neurons in the posterior horn of the spinal cord, where diffuse nuclear accumulations and NIs were present in relatively high frequency, also appeared to be depleted to some extent. Quantitative assessment of neuronal cell populations in regions newly showing mutant AR accumulation will be needed.
Co-localization of cytoplasmic organelles with mutant AR
We performed immunofluorescence with double staining using primary antibodies to recognize specifically various cytoplasmic cell organelles together with 1C2 in the dorsal root ganglia, where cytoplasmic mutant AR accumulation was most prominent (Table 2). TGN46 and 1C2 were co-localized (Fig. 3), indicating that mutant AR exists in the Golgi apparatus. Spinal dorsal root ganglia neurons exhibit some granular structures immunoreactive for another Golgi apparatus marker anti-Golgi 58K protein antibody in the cytoplasm (Fig. 1P and Q). Other organelle markers, including antibodies for lysosomes, endoplasmic reticulum and mitochondria, did not show co-localization with 1C2 (Fig. 3), indicating that expanded polyQ sequences were not detected in these organelles.
Correlation of diffuse nuclear accumulation and NIs with degree of CAG repeat expansion
We examined the correlation of diffuse nuclear accumulation and NIs with the degree of CAG repeat expansion in anterior and posterior horn spinal cord neurons. Averaged frequencies of diffuse nuclear accumulations and NIs in cervical, thoracic and lumbar spinal segments were evaluated for correlation with numbers of CAG repeats in the AR gene. The frequency of diffuse nuclear accumulation in anterior and posterior horn neurons correlated well with the degree of CAG repeat expansion (Fig. 4; r = 0.78, P < 0.05 and r = 0.69, P < 0.05, respectively). However, the frequency of NIs in motor neurons and posterior horn neurons did not show a significant correlation with number of CAG repeats (Fig. 4; r = 0.05, P = NS and r = –0.14, P = NS, respectively). These observations strongly suggest that diffuse nuclear accumulation of the mutant AR protein is more important pathogenetically than NIs.
Immunohistochemical localization of mutant AR in non-neural tissues
As in neural tissues, diffuse nuclear accumulations, NIs or cytoplasmic accumulations of mutant AR were observed in certain visceral organs and skin (Fig. 5, Table 3). Diffuse nuclear accumulations and NIs were detected in the liver, proximal tubules of the kidney, testis, prostate gland, and scrotal and other skin (Fig. 5, Table 3). Cytoplasmic accumulations were detected in the liver, pancreatic islets of Langerhans, testis and prostate gland (Fig. 5, Table 3). Nuclear labelling and cytoplasmic accumulation both were absent in the pituitary gland, heart, lung, intestine, spleen, thyroid, adrenal gland and skeletal muscles. Pancreatic islet cells showed exclusively cytoplasmic accumulations without detectable nuclear accumulations, suggesting that the impaired glucose tolerance frequently observed in our patients (Table 1) could be attributed to cytoplasmic mutant AR accumulation. Ubiquitin staining detected only NIs and, as observed in neural tissues, anti-human AR antibodies occasionally showed diffuse nuclear accumulation without cytoplasmic staining (Fig. 5J). The five control cases did not show any 1C2 immunoreactivity in viscera or skin.
Discussion
The present study clearly demonstrated that diffuse nuclear accumulation of mutant AR, detected by the antibody 1C2 which specifically recognizes the expanded polyQ tract, occurred more frequently than NIs in neural and non-neural tissues in SBMA patients. In neural tissues, diffuse nuclear mutant AR accumulation occurred in the basal ganglia, thalamus, hypothalamus, various midbrain, pontine and medullary nuclei, posterior horn, intermediolateral and Clarke's nuclei of the spinal cord and in sensory and sympathetic ganglion neurons, as well as brainstem and spinal cord motor neurons. NIs detected by 1C2 were similar in distribution to diffuse nuclear accumulation, but the frequency of NIs in each tissue was far less than for diffuse nuclear accumulation. We previously demonstrated that diffuse nuclear mutant AR protein accumulation was more extensive than NIs in male SBMA transgenic mice (Katsuno et al., 2002, 2003). Furthermore, expression and severity of motor dysfunction, and abatement of abnormalities when the mice were castrated or given leuprolerin, paralleled the extent of diffuse nuclear mutant AR accumulation rather than that of NIs (Katsuno et al., 2002, 2003). Accordingly, neuronal dysfunction appeared to be closely related to diffuse nuclear mutant AR accumulation. The key observation in the present study was a significant close correlation between frequency of diffuse nuclear mutant AR accumulation and length of CAG repeat expansion, while a similar correlation was not observed between frequency of NIs and CAG repeats. Diffuse nuclear mutant protein accumulation also has been demonstrated in the neural tissues affected by DRPLA (Yamada et al., 2001a, 2002a), Huntington's disease (Sapp et al., 1997) and Machado–Joseph disease (Yamada et al., 2001b) as well as corresponding transgenic mouse models (Schilling et al., 1999; Yvert et al., 2000; Lin et al., 2001); and, here too, diffuse nuclear mutant protein accumulation was more widespread and extensive than NIs in DRPLA patients. These observations are in good agreement with our present observation in SBMA patients; together, they suggest that diffuse nuclear accumulation of mutant proteins with an expanded polyQ tract is an early event prior to NI formation that is closely related to manifestation of neuronal dysfunction (Yamada et al., 2001a; Garden et al., 2002; Katsuno et al., 2002, 2003; Watase et al., 2002; Yoo et al., 2003). However, the molecular pathogenetic process by which diffuse nuclear mutant AR accumulation induces neuronal dysfunction still is unclear. Although considerable controversy recently surrounds the importance of NIs in the pathophysiology in polyQ diseases (Simeoni et al., 2000; Walcott and Merry, 2002b; Bates, 2003; Michalik and Van Broeckhoven, 2003; Ross et al., 2003), our data showed that diffuse mutant AR accumulation in nuclei could have potent cytotoxic effects inducing neuronal dysfunction through an active epitope of the expanded polyQ tract.
Anti-AR antibodies showed the ability to detect NIs, and some of them (H280, N-20 and Ab-1) occasionally stained diffuse nuclear accumulations. Diffuse nuclear accumulation had the appearance of amorphous aggregates of mutant AR as observed by electron microscopic immunohistochemistry using 1C2 (Fig. 2) (Katsuno et al., 2002). These observations suggest that the polyQ tract epitope can be detected by 1C2, while other AR protein epitopes may be protected by structural features of the aggregate state of the mutant AR. This view is supported by observations made by small-angle X-ray scattering and infrared spectroscopy carried out with myoglobin protein containing an inserted highly expanded polyQ tract, localizing the polyQ tract to the surface of aggregates, while other epitopes were sequestered within aggregates (Tanaka et al., 2001, 2003). These observations suggest that 1C2 can detect the amorphous aggregate state of mutant AR protein, making 1C2 a more sensitive histological and pathophysiological marker than anti-AR protein antibodies. On the other hand, cytoplasmic accumulations were not seen with anti-AR antibodies. This cytoplasmic mutant AR was not ubiquitinated, in contrast to nuclear accumulated mutant AR, particularly the heavily ubiquitinated NIs, suggesting that protein modification varies between the nucleus and cytoplasm. Different protein modification might mask other AR protein epitopes directly or through structural alterations of the aggregate state of the mutant AR in the cytoplasm.
Another important observation in our study was the occurrence of cytoplasmic mutant AR accumulation in neural and non-neural tissues. In neural tissues, cytoplasmic accumulation was restricted to certain neuronal populations such as dorsal root ganglia neurons, mammillary body, hypothalamus, facial motor nucleus, and anterior and posterior horn neurons. In non-neural tissues, cytoplasmic accumulation also occurred in certain organs. Cytoplasmic mutant AR accumulation co-localized with a Golgi apparatus marker. Co-localization of a polyQ-expanded mutant protein with the Golgi apparatus has also been reported for ataxin-2 (Huynh et al., 2003), although the significance of this localization remains unclear. Expression of polyQ-expanded mutant ataxin-2 disrupted the normal morphology of the Golgi complex and increased cell death (Huynh et al., 2003). On the other hand, the lysosomal occurrence of other mutant proteins with an expanded polyQ tract has been reported in DRPLA (Yamada et al., 2002b) and Huntington's disease (Sapp et al., 1997). The lysosomal localization of polyQ-expanded mutant proteins suggests a lysosomal autophagic degradation process acting independently of the ubiquitin–proteasome pathway in the polyQ diseases (Sapp et al., 1997). Additionally, the reason why neural tissues develop more nuclear than cytoplasmic accumulation while most involved visceral organs show equal or predominantly cytoplasmic accumulation is unknown. Differences in the predominant degradation pathway dealing with the mutant AR could influence the intracellular site of accumulation and eventual cell toxicity. One important question is whether cytoplasmic mutant AR accumulation exerts cytotoxicity in neural and non-neural tissues. Cytoplasmic mutant AR accumulation (Taylor et al., 2003) as well as other mutant protein accumulations (Kegel et al., 2000; Ravikumar et al., 2002; Huynh et al., 2003) involving an expanded polyQ tract in Golgi apparatus and lysosomes indeed has been found to induce cytotoxicity. Accumulation of mutant protein with expanded polyQ in the Golgi apparatus or lysosomes increases death of cultured cells through activation of apoptosis-related effectors such as caspase-3 (Ishisaka et al., 1998; Kegel et al., 2000; Huynh et al., 2003). One should note that histologically or immunohistochemically evident mutant protein accumulation is not necessarily cytotoxic, while microaggregates at the molecular level that are histologically undetectable can also exert cytotoxicity. Indeed, excessive accumulation of mutant AR in aggresomes was found to protect cells from a cytotoxic form of mutant AR (Taylor et al., 2003). However, our present study strongly suggests that these cytoplasmic mutant AR accumulations may be related to mutant AR-mediated cytotoxicity and eventual symptom manifestation. For instance, the pancreas showed only cytoplasmic mutant AR accumulation without obvious nuclear accumulation. Elevated serum glucose and impaired glucose tolerance were present in most of our patients, suggesting islet cell dysfunction in the pancreas. The frequency of cytoplasmic accumulations of mutant AR in pancreatic islet cells did not show a significant correlation with fasting blood glucose levels in the examined SBMA patients (data not shown), while certain symptoms and signs of SBMA apparently can be induced by cytoplasmic accumulation of mutant AR protein. Although further study of the significance of cytoplasmic mutant AR accumulation is needed, nuclear accumulation of the mutant AR protein appears to cause motor neuron dysfunction while cytoplasmic accumulation may underlie some visceral and possibly some neuronal dysfunction in SBMA. The pathological process is likely to differ between tissues, being more prominent in motor neuron nuclei, but mainly cytoplasmic in certain neuronal populations and visceral organs. We also need to clarify further which degradation process affecting mutant AR is most active in a given tissue, e.g. lysosomal in certain viscera versus via ubiquitination pathway in most neural tissues.
An important question here is why diffuse nuclear and possibly cytoplasmic accumulation of the mutant AR in the neuronal tissues beyond the major affected spinal and brainstem motor neurons has no apparent symptomatic involvement. First, the causative lesions for sensory impairment and essential-type tremor in SBMA patients have not yet been clearly substantiated. The novel lesion distribution of SBMA neurons shown in the present study, such as the posterior horn of the spinal cord, dorsal root ganglia, thalamus and cerebellum, might provide some explanations for these clinicopathological problems that have not been resolved. Since the cerebellothalamocortical pathway seems to be responsible for essential-type tremor (Pinto et al., 2003), these lesions might contribute to mostly subclinical but definite sensory impairment and essential-type tremor in SBMA. Secondly, the occurrence of neuronal nuclear and cytoplasmic abnormalities in both clinically affected and non-affected neural regions in SBMA suggests that this alteration does not always induce neuronal cell dysfunction or death. The selective neuronal loss and dysfunction in neural lesions that are characteristic of SBMA might depend on additional factors that are specific to neurons in these systems. Recent studies have demonstrated that CREB-binding protein (CBP) is sequestered in AR-positive NIs, resulting in a decrease in CBP-dependent transcription (McCampbell et al., 2000), and further histone acetylation is reduced in affected cells (McCampbell et al., 2001; Steffan et al., 2001; Minamiyama et al., 2004). These reports suggest that CBP-dependent transcriptional dysregulations may cause symptomatic neuronal dysfunction. Since CBP-dependent transcriptional control differs among neurons, this difference may show the lack of their symptomatic involvement in certain polyQ-containing neurons. Alternatively, the population of neurons with nuclear accumulation of mutant AR in the regions beyond the commonly affected lesions may not be simply enough to manifest the responsible symptoms. A precise neuronal cell count assay combined with assessment of nuclear mutant AR accumulation will be needed to clarify these clinicopathological problems.
Clearly, motor neuron impairment with nuclear accumulation of mutant AR is the major problem in SBMA. Thus, for a therapeutic strategy against motor neuron dysfunction in SBMA, nuclear accumulation of mutant AR should be the main target, as we demonstrated in transgenic mice treated with leuprolerin. Cytoplasmic accumulation of mutant AR, on the other hand, should be considered a therapeutic target with respect to certain symptoms in SBMA patients.
Acknowledgements
We wish to thank Dr Yasushi Iwasaki and Mrs Sugiko Yokoi for technical assistance. This work was supported by a Center of Excellence (COE) grant from the Ministry of Education, Culture, Sports, Science, and Technology of Japan, and grants from the Ministry of Health, Labor, and Welfare of Japan.
References
Adachi H, Kume A, Li M, Nakagomi Y, Niwa H, Do J, et al. Transgenic mice with an expanded CAG repeat controlled by the human AR promoter show polyglutamine nuclear inclusions and neuronal dysfunction without neuronal cell death. Hum Mol Genet 2001; 10: 1039–48.
Adachi H, Katsuno M, Minamiyama M, Sang C, Pagoulatos G, Angelidis C, et al. Heat shock protein 70 chaperone overexpression ameliorates phenotypes of the spinal and bulbar muscular atrophy transgenic mouse model by reducing nuclear-localized mutant androgen receptor protein. J Neurosci 2003; 23: 2203–11.
Arbizu T, Santamaria J, Gomez JM, Quilez A, Serra JP. A family with adult spinal and bulbar muscular atrophy, X-linked inheritance and associated testicular failure. J Neurol Sci 1983; 59: 371–82.
Bates G. Huntingtin aggregation and toxicity in Huntington's disease. Lancet 2003; 361: 1642–4.
Chevalier-Larsen ES, O'Brien CJ, Wang H, Jenkins SC, Holder L, Lieberman AP, et al. Castration restores function and neurofilament alterations of aged symptomatic males in a transgenic mouse model of spinal and bulbar muscular atrophy. J Neurosci 2004; 24: 4778–86.
Doyu M, Sobue G, Mukai E, Kachi T, Yasuda T, Mitsuma T, et al. Severity of X-linked recessive bulbospinal neuronopathy correlates with size of the tandem CAG repeat in androgen receptor gene. Ann Neurol 1992; 32: 707–10.
Garden GA, Libby RT, Fu YH, Kinoshita Y, Huang J, Possin DE, et al. Polyglutamine-expanded ataxin-7 promotes non-cell-autonomous Purkinje cell degeneration and displays proteolytic cleavage in ataxic transgenic mice. J Neurosci 2002; 22: 4897–905.
Huynh DP, Yang HT, Vakharia H, Nguyen D, Pulst SM. Expansion of the polyQ repeat in ataxin-2 alters its Golgi localization, disrupts the Golgi complex and causes cell death. Hum Mol Genet 2003; 12: 1485–96.
Igarashi S, Tanno Y, Onodera O, Yamazaki M, Sato S, Ishikawa A, et al. Strong correlation between the number of CAG repeats in androgen receptor genes and the clinical onset of features of spinal and bulbar muscular atrophy. Neurology 1992; 42: 2300–2.
Ishisaka R, Utsumi T, Yabuki M, Kanno T, Furuno T, Inoue M, et al. Activation of caspase-3-like protease by digitonin-treated lysosomes. FEBS Lett 1998; 435: 233–6.
Katsuno M, Adachi H, Kume A, Li M, Nakagomi Y, Niwa H, et al. Testosterone reduction prevents phenotypic expression in a transgenic mouse model of spinal and bulbar muscular atrophy. Neuron 2002; 35: 843–54.
Katsuno M, Adachi H, Doyu M, Minamiyama M, Sang C, Kobayashi Y, et al. Leuprorelin rescues polyglutamine-dependent phenotypes in a transgenic mouse model of spinal and bulbar muscular atrophy. Nat Med 2003; 9: 768–73.
Kegel KB, Kim M, Sapp E, McIntyre C, Castano JG, Aronin N, et al. Huntingtin expression stimulates endosomal-lysosomal activity, endosome tubulation and autophagy. J Neurosci 2000; 20: 7268–78.
Kennedy WR, Alter M, Sung JH. Progressive proximal spinal and bulbar muscular atrophy of late onset. A sex-linked recessive trait. Neurology 1968; 18: 671–80.
La Spada AR, Wilson EM, Lubahn DB, Harding AE, Fischbeck KH. Androgen receptor gene mutations in X-linked spinal and bulbar muscular atrophy. Nature 1991; 352: 77–9.
La Spada AR, Roling DB, Harding AE, Warner CL, Spiegel R, Hausmanowa-Petrusewicz I, et al. Meiotic stability and genotype–phenotype correlation of the trinucleotide repeat in X-linked spinal and bulbar muscular atrophy. Nat Genet 1992; 2: 301–4.
Li M, Sobue G, Doyu M, Mukai E, Hashizume Y, Mitsuma T. Primary sensory neurons in X-linked recessive bulbospinal neuropathy: histopathology and androgen receptor gene expression. Muscle Nerve 1995; 18: 301–8.
Li M, Miwa S, Kobayashi Y, Merry DE, Yamamoto M, Tanaka F, et al. Nuclear inclusions of the androgen receptor protein in spinal and bulbar muscular atrophy. Ann Neurol 1998a; 44: 249–54.
Li M, Nakagomi Y, Kobayashi Y, Merry DE, Tanaka F, Doyu M, et al. Nonneural nuclear inclusions of androgen receptor protein in spinal and bulbar muscular atrophy. Am J Pathol 1998b; 153: 695–701.
Lieberman AP, Harmison G, Strand AD, Olson JM, Fischbeck KH. Altered transcriptional regulation in cells expressing the expanded polyglutamine androgen receptor. Hum Mol Genet 2002; 11: 1967–76.
Lin CH, Tallaksen-Greene S, Chien WM, Cearley JA, Jackson WS, Crouse AB, et al. Neurological abnormalities in a knock-in mouse model of Huntington's disease. Hum Mol Genet 2001; 10: 137–44.
Mariotti C, Castellotti B, Pareyson D, Testa D, Eoli M, Antozzi C, et al. Phenotypic manifestations associated with CAG-repeat expansion in the androgen receptor gene in male patients and heterozygous females: a clinical and molecular study of 30 families. Neuromuscul Disord 2000; 10: 391–7.
McCampbell A, Taylor JP, Taye AA, Robitschek J, Li M, Walcott J, et al. CREB-binding protein sequestration by expanded polyglutamine. Hum Mol Genet 2000; 9: 2197–202.
McCampbell A, Taye AA, Whitty L, Penney E, Steffan JS, Fischbeck KH. Histone deacetylase inhibitors reduce polyglutamine toxicity. Proc Natl Acad Sci USA 2001; 98: 15179–84.
Michalik A, Van Broeckhoven C. Pathogenesis of polyglutamine disorders: aggregation revisited. Hum Mol Genet 2003; 12: R173–86.
Minamiyama M, Katsuno M, Adachi H, Waza M, Sang C, Kobayashi Y, et al. Sodium butyrate ameliorates phenotypic expression in a transgenic mouse model of spinal and bulbar muscular atrophy. Hum Mol Genet 2004; 13: 1183–92.
Pinto AD, Lang AE, Chen R. The cerebellothalamocortical pathway in essential tremor. Neurology 2003; 60: 1985–7.
Ravikumar B, Duden R, Rubinsztein DC. Aggregate-prone proteins with polyglutamine and polyalanine expansions are degraded by autophagy. Hum Mol Genet 2002; 11: 1107–17.
Ross CA. Polyglutamine pathogenesis: emergence of unifying mechanisms for Huntington's disease and related disorders. Neuron 2002; 35: 819–22.
Ross CA, Poirier MA, Wanker EE, Amzel M. Polyglutamine fibrillogenesis: the pathway unfolds. Proc Natl Acad Sci USA 2003; 100: 1–3.
Sapp E, Schwarz C, Chase K, Bhide PG, Young AB, Penney J, et al. Huntingtin localization in brains of normal and Huntington's disease patients. Ann Neurol 1997; 42: 604–12.
Schilling G, Wood JD, Duan K, Slunt HH, Gonzales V, Yamada M, et al. Nuclear accumulation of truncated atrophin-1 fragments in a transgenic mouse model of DRPLA. Neuron 1999; 24: 275–86.
Schmidt BJ, Greenberg CR, Allingham–Hawkins DJ, Spriggs EL. Expression of X-linked bulbospinal muscular atrophy (Kennedy disease) in two homozygous women. Neurology 2002; 59: 770–2.
Simeoni S, Mancini MA, Stenoien DL, Marcelli M, Weigel NL, Zanisi M, et al. Motoneuronal cell death is not correlated with aggregate formation of androgen receptors containing an elongated polyglutamine tract. Hum Mol Genet 2000; 9: 133–44.
Sobue G, Hashizume Y, Mukai E, Hirayama M, Mitsuma T, Takahashi A. X-linked recessive bulbospinal neuronopathy. A clinicopathological study. Brain 1989; 12: 209–32.
Sobue G, Doyu M, Kachi T, Yasuda T, Mukai E, Kumagai T, et al. Subclinical phenotypic expressions in heterozygous females of X-linked recessive bulbospinal neuronopathy. J Neurol Sci 1993; 117: 74–8.
Sperfeld AD, Karitzky J, Brummer D, Schreiber H, Haussler J, Ludolph AC, et al. X-linked bulbospinal neuronopathy: Kennedy disease. Arch Neurol 2002; 59: 1921–6.
Steffan JS, Bodai L, Pallos J, Poelman M, McCampbell A, Apostol BL, et al. Histone deacetylase inhibitors arrest polyglutamine-dependent neurodegeneration in Drosophila. Nature 2001; 413: 739–43.
Tanaka F, Doyu M, Ito Y, Matsumoto M, Mitsuma T, Abe K, et al. Founder effect in spinal and bulbar muscular atrophy (SBMA). Hum Mol Genet 1996; 5: 1253–7.
Tanaka F, Reeves MF, Ito Y, Matsumoto M, Li M, Miwa S, et al. Tissue-specific somatic mosaicism in spinal and bulbar muscular atrophy (SBMA) is dependent on CAG repeat length and androgen receptor gene expression level. Am J Hum Genet 1999; 65: 966–73.
Tanaka M, Morishima I, Akagi T, Hashikawa T, Nukina N. Intra- and intermolecular beta-plated sheet formation in glutamine-repeat inserted myoglobin as a model for polyglutamine diseases. J Biol Chem 2001; 276: 45470–5.
Tanaka M, Machida Y, Nishikawa Y, Akagi T, Hashikawa T, Fujisawa T, et al. Expansion of polyglutamine induces the formation of quasi-aggregate in the early stage of protein fibrillization. J Biol Chem 2003; 278: 34717–24.
Taylor JP, Tanaka F, Robitschek J, Sandoval CM, Taye A, Markovic-Plese S, et al. Aggresomes protect cells by enhancing the degradation of toxic polyglutamine-containing protein. Hum Mol Genet 2003; 12: 749–57.
Terao S, Sobue G, Hashizume Y, Li M, Inagaki T, Mitsuma T. Age-related changes in human spinal ventral horn cells with special reference to the loss of small neurons in the intermediate zone: a quantitative analysis. Acta Neuropathol (Berl) 1996; 92: 109–14.
Trottier Y, Lutz Y, Stevanin G, Imbert G, Devys D, Cancel G, et al. Polyglutamine expansion as a pathological epitope in Huntingtin's disease and four dominant cerebellar ataxias. Nature 1995; 378: 403–6.
Walcott JL, Merry DE. Ligand promotes intranuclear inclusions in a novel cell model of spinal and bulbar muscular atrophy. J Biol Chem 2002a; 277: 50855–9.
Walcott JL, Merry DE. Trinucleotide repeat disease. The androgen receptor in spinal and bulbar muscular atrophy. Vitam Horm 2002b; 65: 127–47.
Watase K, Weeber EJ, Xu B, Antalffy B, Yuva-Paylor L, Hashimoto K, et al. A long CAG repeat in the mouse Sca1 locus replicates SCA1 features and reveals the impact of protein solubility on selective neurodegeneration. Neuron 2002; 34: 905–19.
Yamada M, Wood JD, Shimohata T, Hayashi S, Tsuji S, Ross CA, et al. Widespread occurrence of intranuclear atrophin-1 accumulation in the central nervous system neurons of patients with dentatorubral-pallidoluysian atrophy. Ann Neurol 2001a; 49: 14–23.
Yamada M, Hayashi S, Tsuji S, Takahashi H. Involvement of the cerebral cortex and autonomic ganglia in Machado–Joseph disease. Acta Neuropathol (Berl) 2001b; 101: 140–4.
Yamada M, Sato T, Tsuji S, Takahashi H. Oligodendrocytic polyglutamine pathology in dentatorubral-pallidoluysian atrophy. Ann Neurol 2002a; 52: 670–4.
Yamada M, Tsuji S, Takahashi H. Involvement of lysosomes in the pathogenesis of CAG repeat diseases. Ann Neurol 2002b; 52: 498–503.
Yoo SY, Pennesi ME, Weeber EJ, Xu B, Atkinson R, Chen S, et al. SCA7 knock in mice model human SCA7 and reveal gradual accumulation of mutant ataxin-7 in neurons and abnormalities in short-term plasticity. Neuron 2003; 37: 383–401.
Yvert G, Lindenberg KS, Picaud S, Landwehrmeyer GB, Sahel JA, Mandel JL. Expanded polyglutamines induce neurodegeneration and trans-neuronal alterations in cerebellum and retina of SCA7 transgenic mice. Hum Mol Genet 2000; 9: 2491–506.(Hiroaki Adachi, Masahisa Katsuno, Makoto)