Polyglutamine expansion, protein aggregation, proteasome activity, and neural survival.

Huntington's disease (HD) is one of eight established triplet repeat neurodegenerative disorders, which are collectively caused by the genetic expansion of polyglutamine repeats. While the mechanism(s) by which polyglutamine expansion causes neurodegeneration in each of these disorders is being intensely investigated, the underlying cause of polyglutamine toxicity has not been fully elucidated. A number of studies have focused on the potential role of protein aggregation and disruption of the proteasome proteolytic pathway in polyglutamine-mediated neurodegeneration. However, at present it is not clear whether polyglutamine-mediated protein aggregation is sufficient to induce cell death, nor has it been clearly determined whether proteasome inhibition precedes, coincides, or occurs as the result of the formation of polyglutamine-associated protein aggregation. To address these important components of polyglutamine toxicity, in the present study we utilized neural SH-SY5Y cells stably transfected with polyglutamine-green fluorescent protein constructs to examine the effects of polyglutamine expansion on protein aggregation, proteasome activity, and neural cell survival. Data from the present study demonstrate that polyglutamine expansion does not dramatically impair proteasome activity or elevate protein aggregate formation under basal conditions, but does significantly impair the ability of the proteasome to respond to stress, and increases stress-induced protein aggregation following stress, all in the absence of neural cell death.

A number of neurodegenerative disorders, including Huntington's disease (HD) 1 and the spinocerebellar ataxias, are caused by the genetic insertion of expanded and unstable glutamine repeats (1)(2)(3). Each of the polyglutamine disorders share a number of similarities including the existence of a polyglutamine length-dependent toxicity (1)(2)(3)(4), whereby the number of polyglutamine repeats is inversely correlated with the age of disease onset and disease severity, increased levels of protein aggregation (5,6), and selective neurodegeneration within the central nervous system (1)(2)(3)7). Numerous studies now indicate that polyglutamine expansion causes an as yet unidentified, toxic gain of function. For example, the various proteins involved in polyglutamine disorders share no homology outside of the polyglutamine repeat region, while the genetic transfer of polyglutamine repeats to nontoxic or truncated proteins is sufficient to induce HD-like neurochemical, neurophysiological, and neuropathological alterations (8 -12).
Recent reports suggest that the neurotoxic gain of function induced by polyglutamine expansion may involve disruptions in the proteasome proteolytic pathway (13)(14)(15). The proteasome is a large intracellular multicatalytic protease that is responsible for the majority of overall protein degradation, including the degradation of most oxidized, misfolded, and aggregated proteins (16 -18). The proteasome also appears to be particularly important in the degradation of polyglutamineenriched and readily aggregating proteins (19 -21). Because proteasome inhibition is sufficient to induce neuron death in vitro (22)(23)(24), and numerous reports have now demonstrated proteasome inhibition in a wide range of neurodegenerative conditions (25)(26)(27)(28), it is plausible that proteasome inhibition may play a role in polyglutamine disorders. However, at present the characterization of proteasome inhibition, and role of proteasome inhibition in polyglutamine disorders, has not been fully elucidated.
Protein aggregation is increased in polyglutamine disorders (1)(2)(3), through an as yet unidentified mechanism. At present, based on data from in vitro and in vivo studies, it is unclear what role aggregates play in polyglutamine toxicity (29 -34). Because the proteasome can become inhibited following exposure to excessively aggregated proteins (19,21,35,36), it has been suggested that elevated levels of protein aggregates in polyglutamine disorders may ultimately inhibit proteasome activity (14,19).
In the present study we analyzed polyglutamine aggregation, proteasome activity, and cell death in neural SH-SY5Y cells that were stably transfected with green fluorescent protein (GFP) containing 19, 56, or 80 glutamine repeats. Data presented in this report indicate that the ability to elevate proteasome activity is impaired in 56-and 80-glutamine GFPtransfected cells and precedes formation of stable GFP aggregates, all in the absence of neural cell death.

EXPERIMENTAL PROCEDURES
Generation of Stable Cell Lines and Analysis of Cellular Proliferation-Neural SH-SY5Y cells were purchased from the American Tissue Culture Collection and grown in minimal essential medium containing 10% heat-inactivated fetal bovine serum and 1% penicillin/streptomycin. Neural SH-SY5Y cells in 60-mm dishes, at 60 -70% confluence, were transfected using LipofectAMINE reagent, following manufacturer's instructions (37). Cells were transfected with 2 g of empty * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ** Supported by the Huntington's Disease Society of America, American Health Assistance Foundation, American Heart Association, and a grant from the National Institutes of Health. To whom correspondence should be addressed: 101 Sanders-Brown Bldg., University of Kentucky, Lexington, KY 40536-0230. Tel.: 859-257-1412; Fax: 859-323-2866; E-mail: Jnkell0@pop.uky.edu. 1 The abbreviations used are: HD, Huntington's disease; GFP, green fluorescent protein; Hsp, heat shock protein; MCA, 7-amido-4-methylcoumarin; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide. pcDNA3.1 vector or pcDNA3.1 containing 19, 56, or 80 glutamines, which have been ligated to GFP, resulting in 19-, 56-, and 80-glutamine GFP, respectively. Following 24 h transfection, cells were placed under selection using G418 (800 g/ml) and individual clones of each cell line selected. Heat shock was induced by placing cells in a 43°C CO 2 incubator for the indicated period of time, with cells then immediately placed in adjacent 37°C incubator. For analysis of cellular proliferation cells were plated at 40% confluence (5 ϫ 10 4 ) and incubated overnight in minimal essential medium containing 1% serum (v/v). The following day the medium was replaced with normal growth medium, and following 96-h incubation cells were removed by mild trypsinization and counted. The percent increase in proliferation was then calculated by dividing the number of trypan blue-negative cells collected from the initial seeding concentration.
Analysis of Cell Survival and Aggregate Formation-Cell viability was determined by MTT conversion and propidium iodide staining as described previously (8,9,38). Viable cells contain an intact membrane that is impermeable to propidium iodide, thus allowing for the identification of viable cells. At least six cultures were utilized for each time point. Visible aggregate formation was determined using a Ziess fluorescence microscope equipped with a digital camera and Attofluor software. Living GFP-positive cells were counted in randomly chosen fields and the percentage of cells exhibiting aggregates calculated. In each experiment at least 30 cells were counted per dish, with a minimum of four dishes utilized for each time point. Nuclear aggregation was assessed by co-localization of aggregates with propidium iodide as described previously (9). Cells were pretreated with Triton X-100 prior to the addition of propidium iodide, to allow propidium iodide to stain all nuclei.
Analysis of protein aggregation by electron microscopy was conducted as described previously (14), with some minor modifications. Briefly, following experimental treatment cells were rinsed in ice-cold 0.1 M Sorenson's buffer (pH 7.4), followed by 30-min fixation in 3.5% glutaraldehyde, 0.1 M Sorenson's solution (v/v), followed by a 30-min incubation in 1% osmium tetroxide. Cells were then subject to dehydration via incubations in increasing concentrations of ethanol and embedded in Eponate 12 resin. The tissue was then sectioned (60 nm), placed into copper grids, and imaged using a Philips CM100 EM (Philips Electron Optics, Eindhoven, The Netherlands).
For analysis of protein aggregation in nondenaturing gels, cells were collected in ice-cold phosphate-buffered saline-containing protease inhibitor mixture (Invitrogen), following experimental treatment. Cells were then homogenized on ice and protein aliquots (50 g) separated within a 7.5% polyacrylamide gel using electrophoresis under nondenaturing conditions. Proteins were then transferred to a nitrocellulose membrane and immunoreacted with ␣-GFP antibody (1:500; Santa Cruz Biotechnology, Santa Cruz, CA), and developed using an ECL kit (Amersham Biosciences). For determinations of aggregate stability protein aliquots were incubated for 1 h in 1% SDS (v/v) and placed for 10 min in boiling water prior to Western blot analysis.
Northern Blot Analysis-Northern blot analysis was conducted as described previously (26,38). Briefly, RNA was extracted under RNasefree conditions using TRI-Reagent™ (Sigma). The integrity and quality of RNA were confirmed for each sample by gel analysis prior to beginning Northern procedure. A total of 8 g of RNA for each sample was size-fractioned by electrophoresis in a 1% formaldehyde-agarose gel. The RNA was then transferred to a Hybond ϩ nylon membrane, incubated with fluorescein isothiocyanate-labeled probe, and imaged using a Fuji Phosphoimager.
Statistical Analysis-Statistical significance was determined using a t test, with a maximum p value of Ͻ 0.05 required for significance.

RESULTS
Characterization of GFP-transfected Cells-In the present report we conducted studies in neural SH-SY5Y cells that have been stably transfected with GFP containing 19, 56, or 80 glutamines as described previously (8). Cells transfected with 19-, 56-or 80-glutamine GFP demonstrated GFP fluorescence throughout the cytoplasm and had a similar shape (Fig. 1). All cell lines proliferated at similar rates ( Fig. 1), and based on Western blot analysis, expressed similar levels of GFP (data not shown). Under basal conditions the 19-glutamine GFPtransfected cells presented almost no GFP aggregates (Ͻ1%), while 56-and 80-glutamine GFP both exhibited low level aggregate formation (ϳ1-3% of cells).
Heat Shock Induces a Polyglutamine Length-dependent Susceptibility to Aggregation-In contrast to basal conditions, following increasing durations of heat shock (43°C), there was observed to be a significant increase in GFP aggregate formation in 56-and 80-glutamine GFP, but not 19-glutamine GFPtransfected cells (Fig. 1). Aggregates appeared to be primarily cytosolic, with multiple aggregates present in individual cells (Fig. 1). The effect of heat shock on protein aggregation was observed to be dependent on both the severity of heat shock and the amount of time cells were allowed to recover from heat shock. For example, short term heat shock (5 min) caused a rapid and reversible form of aggregate formation in 80-glutamine GFP-transfected cells, without inducing aggregation in 19-glutamine GFP-transfected cells (Fig. 2). Following heat shock of greater that 5 min, no early transitory elevations in protein aggregation was observed in 80-glutamine GFP-transfected cells (Fig. 2). Almost no aggregation was ever detected in 19-glutamine GFP-transfected cells under basal conditions or following heat shock, except for a slight transitory elevation observed 3 h following the most severe heat shock (120 min) (Fig. 2).
While no significant elevations in aggregates were detected in 80-glutamine GFP within the first 3 h of either 60 or 120 min of heat shock, significant elevations in protein aggregates were detected 24 h post-heat shock (Fig. 2). In contrast to the rapidly forming and transitory aggregates, aggregates present at 24 h were apparently sustained, with aggregation levels remaining elevated for at least 48 h following heat shock (Fig. 2). Results using 56-glutamine GFP were similar to that observed for 80-glutamine GFP (data not shown). Less than 6% of cells exhibited nuclear aggregates 48 h following even the most severe heat shock (Fig. 2).
To better elucidate the characteristics of aggregates formed following heat shock, we conducted electron microscopy and nondenaturing electrophoresis analysis. While most aggregates analyzed were less than 2 m in diameter, some of the aggregates formed following heat shock achieved sizes of 4 -6 m (Fig. 3). In most cases aggregates exhibited an irregular shape and appeared to have at least some fibrillar component. Analysis of aggregation using nondenaturing electrophoresis revealed the presence of a nonmigrating aggregate species in 80-glutamine GFP-transfected cells following heat shock (Fig.  3). This nonmigrating aggregate species appears highly stable, based on the inability of SDS and heat treatment to disrupt aggregate (Fig. 3). Similar results were obtained using 56glutamine GFP-transfected cells (data not shown). Under no circumstances, during basal conditions or following heat shock, was a nonmigrating aggregate species detected in 19-glutamine GFP-transfected cells (data not shown). Because no obvious increase in the level of GFP (summing of aggregated and nonaggregating species) is observed, these data suggest that increased aggregation of GFP following heat shock is likely not due to a gross elevation in the level of GFP protein.
The ability of aggregates to persist up to 48 h following heat shock suggested that cells may still be viable following heat shock. To elucidate whether heat shock induces neural cell death, we conducted cell survival analysis following the most severe heat shock conditions utilized in the present study (120-min heat shock). No significant elevation in cell death was observed between 19-glutamine GFP and 80-glutamine GFPtransfected cells, even up to 72 h following heat shock (Fig. 4).
Polyglutamine Expansion Alters the Ability of Cells to Upregulate Proteasome Activity and Cope with Proteasome Inhibition-Previous studies have suggested that polyglutamine expansion may impair the proteasome proteolytic pathway (13)(14)(15). We analyzed proteasome proteolytic activity in 19glutamine and 80-glutamine GFP-transfected cells under basal conditions. Under normal growth conditions there was no significant impairment in either chymotrypsin-like or postglutamyl peptidase activities of the proteasome (Fig. 5). Following heat shock there was observed to be a significant elevation in chymotrypsin-like proteasome activity in 19-glutamine GFPexpressing cells, which was significantly blunted in 80-glutamine GFP-expressing cells (Fig. 5). Interestingly, no significant elevations in post-glutamyl peptidase activity was evident following heat shock (Fig. 5). Results using 56-glutamine GFP were similar to 80-glutamine GFP (data not shown).
Because previous studies have suggested that impairments in proteasome activity may contribute to aggregate formation in polyglutamine disorders, we next sought to determine the effects of proteasome inhibition on protein aggregation. Application of low, nontoxic concentrations of the proteasome inhibitor MG-115, induced a slight time-dependent increase in GFP aggregation in 80-glutamine GFP-transfected cells, but not 19-glutamine GFP-expressing cells (Fig. 6). Increased levels of aggregation were evident within 3 h following administration of proteasome inhibitor (Fig. 6). Application of toxic concentrations of proteasome inhibitors induced higher levels of aggregate formation at 24 h (Fig. 6). Similar results were obtained with the proteasome inhibitor epoxomycin (data not shown).
Cells with 19-Glutamine, 56-Glutamine, and 80-Glutamine GFP Have Altered Expression of Proteasome Subunits-Because the response of 56-glutamine and 80-glutamine GFP to heat shock and proteasome inhibitors was so different from 19-glutamine GFP-transfected cells, we next sought to determine whether these observations could be due in part to altered levels of heat shock proteins (Hsp) or proteasome subunits. No significant difference in Hsp70 expression was evident between 19-glutamine, 56-glutamine, and 80-glutamine GFP-transfected cells under basal conditions or following heat shock (Fig.  7). In contrast to Hsp70, Northern blot analysis revealed that 80-glutamine GFP had higher basal levels of Hsp40 than 19glutamine GFP (Fig. 7). Interestingly, while 19-glutamine GFP-transfected cells exhibited increased Hsp40 expression following 120-min heat shock, 80-glutamine GFP exhibited no elevation in Hsp40 expression (Fig. 7). No difference in the expression of the C2 ␣-subunit of the 20 S proteasome was evident between 19-glutamine, 56-glutamine, or 80-glutamine GFP-transfected cells under basal conditions or following heat shock (Fig. 7). In contrast to expression of C2, cells expressing 56-glutamine and 80-glutamine GFP exhibited increased expression of the LMP2 ␤-subunit of the 20 S proteasome, as compared with 19-glutamine GFP-transfected cells (Fig. 7). Increased LMP2 expression in 56-glutamine and 80-glutamine

FIG. 3. Aggregates can be large and appear stable.
A series of increasingly high magnifications of the same aggregate by electron microscopy reveals that aggregates can be large (A) or irregular and amorphous (B and C). D, cell lysates from 80-glutamine GFP receiving no heat shock, 30-min heat shock, or 120-min heat shock were collected under nondenaturing conditions 24 h post-heat shock. One aliquot of 30-min and 120-min heat shock sample was incubated in SDS and heated as described under "Experimental Procedures" prior to loading. All lysates were subject to Western blot on nondenaturing gel and analyzed for ␣-GFP immunoreactivity. Band 1 represents aggregated GFP; band 2 represents nonspecific immunoreactivity; band 3 represents nonaggregated GFP. The molecular mass of nonaggregated GFP is ϳ36 kDa. Data are representative of results from two experiments. Bars in A-C indicate length and are 2 m, 500 nm, and 100 nm, respectively. GFP-transfected cells was evident under basal conditions as well as following heat shock. DISCUSSION In the present study we utilized neural SH-SY5Y cells that have been stably transfected with polyglutamine expansion constructs, to analyze aggregate formation and its effect on the proteasomal proteolytic pathway. Previous studies have demonstrated that neural cell lines stably transfected to express polyglutamine expansion can be established (42). In contrast to transient transfection or inducible polyglutamine expression (8,30,34), cells in the present study do not undergo a rapid form of cell death under basal conditions and thus have likely undergone genetic or biochemical changes, which allow them to cope with polyglutamine expansion. It is interesting to note that no known embryonic defects occur in humans or mice expressing mutant huntingtin (1-3), while additional studies have demonstrated that mutant huntingtin is capable of suppressing embryonic lethality in huntingtin knock-out mice (1)(2)(3), consistent with the ability of neuronal cells to withstand the existence of polyglutamine expansion for prolonged periods without apparent adverse effects. However, it is also important to note that cells stably expressing polyglutamine expansion have been demonstrated to be more vulnerable to stress (43) and were more prone to aggregate formation in the present study.
Data from the present study demonstrate that heat shock can induce an apparently reversible form of aggregate formation in cells expressing polyglutamine expansion. Cells expressing longer polyglutamine stretches (56-glutamine and 80glutamine GFP) underwent apparently reversible increases in visible aggregate formation following very mild stress (5-min heat shock), while cells with 19-glutamine GFP underwent slight elevations in visible protein aggregation following the most severe stress (120 min heat shock), with aggregation in both paradigms being apparently reversible. It is likely that these rapidly forming visible aggregates are caused by heat shock-induced protein denaturing and protein misfolding, which increase the levels of nonspecific protein-protein interactions (39). It is unlikely that heat shock induced aggregation in GFP-expressing cells is due to intrinsic protein instabilities within the GFP protein, based on the fact that the majority of GFP remained functional, with GFP fluorescence detected throughout 19-glutamine, 56-glutamine, and 80-glutamine GFP-expressing cells following even the most severe heat shock (Fig. 1). Our data suggest that 19-glutamine GFP may be less promiscuous in forming nonspecific interactions than either 56-glutamine or 80-glutamine GFP, consistent with previous reports demonstrating the existence of a polyglutamine lengthdependent propensity for protein binding (1-3, 40, 41). In future studies it will be important to determine via immunohis- tochemical analysis exactly what proteins are present within heat shock-induced aggregates.
It is interesting to note that the reversible mode of visible aggregate formation occurs more rapidly (Ͻ3 h) than the formation of stable visible aggregates (which are not evident until 24 h post-heat shock). These data raise the possibility that different processes may be involved in the formation of visible aggregates which are reversible, as compared with those that are stable. For example, the formation of visible aggregates that are stable may require protein synthesis or a heterogeneous mixture of proteins to form and thereby is more dependent on time than reversible aggregate formation, which may form as the result of more spontaneous protein-protein interactions. However, it is important to point out that the apparent reversibility in visible aggregate formation may not necessarily reflect an actual decrease in aggregates, but may simply indicate that the type of aggregate created has been altered in some way. For example, analysis of aggregate formation using nondenaturing electrophoresis revealed that an SDS and heatinsoluble aggregate species is present (Fig. 3D) when no visible aggregates are detectable (Fig. 2B).
In the present study, cells remained viable following the formation of aggregates. In most cases multiple aggregates were present in individual cells, and in some cases even large aggregates (4 -6 m) were observed, without an apparent induction of cytotoxicity. These data indicate that the presence of intracellular aggregates, and even large intracellular aggregates, do not necessarily serve as an indicator of cytotoxicity, consistent with previous reports (1)(2)(3)(4)(5)(6). Very few of the cells in the present study demonstrated nuclear aggregates (Fig. 2E), possibly indicating that aggregation within the nucleus may be particularly important in the initiation of cytotoxicity. Although no increase in cell death was observed following heat shock, data in the present study do not rule out the possibility that cells may have compromised viability. It will be important in future studies to determine whether protein aggregation in this paradigm alters gene expression or signal transduction pathways that may be important in the neurotoxicity observed in polyglutamine disorders.
To survive in response to genetic and environmental stressors, cells rely on the function Hsp family members, which function in a coordinated manner to aid in protein degradation, protein folding, and protein trafficking. Previous studies have demonstrated that increased levels of Hsp family members can attenuate polyglutamine toxicity (1)(2)(3), suggesting that alterations in the expression of Hsp family members may play a role in maintaining cellular homeostasis in our cells stably transfected with polyglutamine expression. In the present study no differences in the expression levels of Hsp70 were evident by Northern blot analysis, or Western blot analysis (data not shown), between the 19-glutamine, 56-glutamine, and 80-glutamine GFP-transfected cells (Fig. 7). However, 19-glutamine GFP-expressing cells demonstrated an enhanced up-regulation of Hsp40 following heat shock than 80-glutamine GFP-transfected cells (Fig. 7). It is interesting to note that 19-glutamine, 56-glutamine, and 80-glutamine GFP all had higher levels of Hsp expression than wild-type SH-SY5Y cells (data not shown). Taken together, these data suggest that the presence of polyglutamine expansion may have a direct effect on the intracellular Hsp pathway and affect Hsp function in such a manner that is potentially of direct importance to polyglutamine neuropathology. For example, the presence of polyglutamine expansion alone may place such a burden on the Hsp pathway that cells are ultimately unable to up-regulate the Hsp pathway further in response to subsequent stressors.
A dramatic finding in the present study is the increased expression of the LMP2 ␤-subunit of the 20 S proteasome (Fig.  7), which is one of the proteasome subunits responsible for proteasomal chymotrypsin-like activity (16), in 56-glutamine and 80-glutamine GFP-transfected cells. No alteration in the C2 ␣-subunit of the 20 S proteasome was observed between 19-glutamine, 56-glutamine, or 80-glutamine GFP-transfected cells in the present study (Fig. 7). These data are consistent with recent reports and strongly suggest that cells can undergo specific alterations in proteasome subunit expression in response to polyglutamine expansion. For example, recent studies have demonstrated polyglutamine expansion can up-regulate specific components of the lysosomal proteolytic pathway (44). Alterations in proteasome subunits expression are likely responsible for mediating the ability of the proteasome to exhibit normal levels of activity, even in the presence of polyglutamine expansion. Additionally, these data indicate that LMP2 may be a particularly important proteasome subunit in conditions where protein aggregation occurs. Because there are over 20 individual genes encoding for the subunits for the core 20 S proteasome, and an additional 30ϩ genes encoding for subunits of the 19 S and 11 S caps of the 20 S proteasome (16), these data highlight the importance of elucidating the specific effects of polyglutamine expansion on proteasome subunit expression. Such basic information is critical for the understanding of how alterations in proteasome biology contribute to polyglutamine neurotoxicity.
The greatest effect of polyglutamine expansion on proteasome activity was not on basal proteasome activity, but on the activity of the proteasome following stress. Following heat shock, cells expressing longer lengths of polyglutamine expansion (80-glutamine GFP) were unable to increase proteasome activity as well as 19-glutamine GFP-transfected cells (Fig. 5). These data suggest that the deleterious effect of polyglutamine expansion on the proteasome-proteolytic pathway may ultimately be on its ability to suppress elevations in proteasome activity in response to stress and not because of the impairment of basal levels of proteasome activity per se. Additionally, polyglutamine expansion may have important effects on proteasome localization. For example, only 10 -15% of proteasome activity in the brain is nuclear (16); therefore small changes in the localization or function of the proteasome in the nucleus may have profound effects on cellular homeostasis. It is likely that proteasome activity is up-regulated in response to the presence of increased levels of damaged and misfolded proteins following heat shock (39). Consistent with such a paradigm, previous studies have demonstrated that proteasome activity can be rapidly elevated in reponse to mild oxidative stress (45,46).
Application of proteasome inhibitors increased protein aggregation similar to heat shock, in that it occurred only in 80-glutamine GFP-expressing cells (Fig. 6). At least 20% proteasome inhibition was required for the formation of protein aggregates (data not shown), which was not overtly toxic with 24 h of application (Fig. 6D). It is important to note that application of proteasome inhibitors induces aggregates rapidly (ϳ 3 h) (Fig. 6), as compared with heat shock (Fig. 2). Previous studies have utilized proteasome inhibitors to elucidate the role of the proteasome in degradation of various substrates, including GFP-and polyglutamine-containing proteins (14,19,41,47). However, it is important to point out that proteasome inhibitors are highly toxic to many cells, including neurons and neural cell lines (22)(23)(24), and that the effects of proteasome inhibition may ultimately be due to induction of cell death process and not proteasome inhibition per se.
Previous studies have demonstrated a gross loss of proteasome activity in Alzheimer's disease, Parkinson's disease, and ischemia-reperfusion injury (25)(26)(27)(28). Data from the present study suggest that while basal levels of proteasome activity may only be slightly impaired in polyglutamine disorders, the ability of the proteasome to respond to stress may be significantly compromised. Such a compromise may contribute to protein aggregation and a loss of neuronal homeostasis. It is likely that developing a better understanding of the role of proteasome activity and aggregate formation in polyglutamine disorders may enhance our understanding of neurotoxicity in general and ultimately lead to the discovery of effective interventions.