Sensitivity of Mammalian Cells Expressing Mutant Ubiquitin to Protein-damaging Agents*

There is convincing evidence from studies in yeast that a functional ubiquitin/proteasome pathway is required to degrade misfolded or oxidatively damaged proteins but for technical reasons, it has been difficult to perform comparable studies in mammalian cells. To investigate the possibility that the ubiquitin/protea-some pathway is cytoprotective for mammalian cells, we have introduced epitope-tagged wild-type ubiquitin or dominant-negative mutant versions of ubiquitin into mouse HT4 neuroblastoma cells. Cells expressing mutant versions of ubiquitin were found to be sensitive to cadmium, an agent that causes oxidative damage to cellular components, and to canavanine, an amino acid analog that generates misfolded proteins. The greatest sensitivity to canavanine was observed in cells expressing a mutant version of ubiquitin unable to support the formation of Lys 48 linkages. Substrates of the proteasome were found to accumulate in these cells, suggest-ing a general deficit in proteolysis. Our data suggest that defects in the ubiquitin-mediated proteolytic system predispose mammalian cells to the toxic effects of abnormal protein. The ubiquitin-mediated proteolytic pathway mediates the rapid and selective degradation of many protein substrates in eukaryotic cells (reviewed in Ref. 1). The pathway is required both for normal homeostasis (regulating levels of specific proteins involved in signaling,

The ubiquitin-mediated proteolytic pathway mediates the rapid and selective degradation of many protein substrates in eukaryotic cells (reviewed in Ref. 1). The pathway is required both for normal homeostasis (regulating levels of specific proteins involved in signaling, cell cycle, differentiation programs, etc.) and under stress conditions, where it is called upon to remove proteins that are damaged or misfolded. Oxidative stress may contribute to the formation of the proteinaceous inclusions that have long been associated with neuropathological states (2) through the generation of aberrant protein structures with a propensity to aggregate (3). By eliminating oxidatively damaged or misfolded proteins, the ubiquitin/ proteasome pathway may play a cytoprotective role. In yeast, overexpression of ubiquitin has been shown to diminish the toxicity of the amino acid analog canavanine while increasing sensitivity to other agents (4). Expression of a dominant-neg-ative mutant form of ubiquitin (K48R) that interferes with proteolysis was found to increase the half-life of protein substrates in yeast, including canavanyl proteins (5). Yeast cells deficient in the UBC7 ubiquitin-conjugating enzyme were found to be sensitive to cadmium, an agent that generates oxidative damage to cellular components (6). Because of the technical impediments involved in the genetic manipulation of mammalian cells, less is known about the role of ubiquitin and ubiquitin metabolic enzymes in the response to oxidative damage. In this study, we adopted a dominant-negative strategy to test the hypothesis that cells in which the ubiquitin/proteasome pathway has been compromised will be sensitive to oxidative stress. The murine HT4 neuroblastoma cell line represents a well characterized system for the analysis of oxidative stress. Treatment of HT4 cells with cadmium has been previously shown to result in both a decrease in intracellular glutathione levels and an increase in mixed protein disulfides (7,8). Most important, cadmium exposure has been shown to induce the accumulation of high molecular weight ubiquitinated conjugates (7)(8)(9).
Ubiquitin functions as a molecular tag that targets substrates to the 26 S proteasome, an ATP-dependent multisubunit protease. The proteasome is thought to recognize threshold length ubiquitin chains assembled through Lys 48 -Gly 76 linkages (reviewed in Ref. 10) leading to the ATP-dependent proteolysis of the targeted protein. Our strategy, which is directly analogous to the mutational strategy previously employed in yeast (5,(11)(12)(13), was to engineer site-directed mutants of an epitope-tagged human ubiquitin that could be expressed in transfected HT4 cells. In our implementation of this strategy, a GFP 1 marker was also incorporated as a linear fusion with ubiquitin to identify cells in which the transgene was expressed. We have previously demonstrated that a ubiquitin/GFP fusion protein is processed by unidentified cellular ubiquitin carboxyl-terminal hydrolases to generate free ubiquitin monomers competent for conjugation to cellular substrates (14). Here we report that mammalian cells expressing mutant ubiquitin respond very differently to cells expressing only wild-type ubiquitin when faced with a burden of damaged protein and that particular mutations expose sensitivity to particular types of protein damage.

EXPERIMENTAL PROCEDURES
Plasmids-The construction of plasmid pDG268, the wild-type vector in which expression of a linear fusion of His 6 -tagged human ubiquitin and enhanced GFP is driven by the human UBC promoter, has been described elsewhere (14). Plasmid pDG279, containing the K48R mutant version of ubiquitin, was generated by a polymerase chain reaction-based site-directed mutagenesis strategy in which pDG268 served * The work was supported by Grant MT-15134 from the Canadian Institutes of Health Research (to D. A. G.). 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.
¶ Recipient of an Ontario graduate scholarship in science and technology.
ʈ Present address: Dept. of Radiation Oncology, University of Maastricht, 6200 MD Maastricht, The Netherlands.
‡ ‡ To whom correspondence should be addressed: Centre for Cancer Therapeutics, Ottawa Regional Cancer Centre, 503 Smyth Rd., Ottawa, Ontario K1H 1C4, Canada. Tel.: 613-737-7700 (ext. 6896); Fax: 613-247-3524; E-mail: Doug.Gray@orcc.on.ca. as the template. The entire plasmid was amplified using primers overlapping at a position corresponding to Lys 48 : upper strand primer, AAAGCCGGCAGCTGGAAGATGGCCGTACTC; and lower strand primer, AAAGCCGGCCTCAAAGATGAGCCTCTGC. These primers served to replace the lysine residue at position 48 with arginine and in the process created a novel NgoMIV site in the plasmid, which was unique. Following amplification, the linear product was purified from a gel and digested with NgoMIV and then ligated and transformed into bacteria. The resulting plasmid was sequenced to confirm that the site-directed mutation was present and that no other sequence alterations had occurred during polymerase chain reaction amplification. The plasmid encoding K63R mutant ubiquitin was generated from pDG268 in a similar fashion as described elsewhere. 2 Ubiquitin expression plasmids were cotransfected with a plasmid conferring resistance to puromycin (pgk-puro, the gift of Dr. Michael McBurney, Ottawa Regional Cancer Centre) and were grown in the presence of the drug to establish stable transfectants. Expression of the GFP marker was periodically confirmed by fluorescence microscopy.
Cell Culture and Oxidative Stress-Untransfected HT4 cells and cells stably transfected with wild-type ubiquitin/GFP, K48R mutant ubiquitin/GFP or K63R mutant ubiquitin/GFP were plated at a density of 40 ϫ 10 4 cells/60-mm dish and maintained at 32°C (HT4 cells express a temperature-sensitive SV40 large T antigen and partially differentiate at 37°C) (15) in ␣-minimal essential medium containing 10% non-heat-inactivated fetal calf serum. The cells were then either treated for 24 h with 5 M cadmium sulfate (Sigma) alone or pretreated with 0.2 mM buthionine sulfoximine (Sigma) prior to a 24-h incubation period with 5 M cadmium sulfate. Following the incubation period, cells were observed by light microscopy.
Canavanine-treated Cells-Untransfected HT4 cells and cells stably transfected with wild-type ubiquitin/GFP, K48R mutant ubiquitin/ GFP, or K63R mutant ubiquitin/GFP were treated with 20 mM canavanine (Sigma) for 30 h and then visualized by light microscopy. Toxicity was measured by trypan blue exclusion in attached cells and in cells dislodged by 5 min of rotation at 100 rpm on a platform shaker. The percentage of dead cells in floating versus adherent cells was determined independently. For Western blot analysis, untransfected HT4 cells and cell clones expressing wild-type ubiquitin/GFP or K48R mutant ubiquitin/GFP were treated with 10 mM canavanine for 24 h prior to harvesting.
Transient Transfections-Untransfected HT4 cells and cells stably transfected with wild-type ubiquitin/GFP or K48R mutant ubiquitin/ GFP were plated at a density of 2.5 ϫ 10 5 cells/60-mm dish and maintained at 32°C for 24 h prior to transfection. Following this incubation period, the cells were transiently transfected using FuGene 6 (Roche Molecular Biochemicals) with the pRc/CMV-HA-E2F-1 expression vector or with an expression vector coding for GFP U , a destabilized version of GFP (the generous gift of Dr. Ron Kopito, Stanford University). 24 h after transfection, lysates were prepared for Western analysis.
Extract Preparation for Western Blot Analysis-Following the indicated treatment, the cells were harvested in lysis buffer (20 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1% Nonidet P-40, 0.5 mM EDTA, and 20% glycerol with 1 mM phenylmethylsulfonyl fluoride, 5 g/ml leupeptin, 2 g/ml aprotinin, 200 M NaF, and 200 M sodium PP i ). The soluble fractions were recovered, and the proteins were quantified using the Bradford protein assay (Bio-Rad). Cytoplasmic protein extracts (20 g) were then resolved on a two-phase SDS-polyacrylamide gel (15 and 8%) and electroblotted onto a Hybond C nitrocellulose membrane (Amersham Pharmacia Biotech). The membrane was stained with Ponceau S (Sigma) prior to Western blotting with the indicated primary antibody. Proteins were visualized by a horseradish peroxidase method using the ECL kit from Kirkegaard & Perry Laboratories. Quantification of band intensities was performed using the Area Density Tool algorithm of the LabWorks software program analyzing images obtained with an Epi Chem II Darkroom instrument (UVP Inc.).
Antibodies-Western blot analysis was performed using both a mouse monoclonal antibody and a rabbit polyclonal antibody raised against the RGS-His epitope (QIAGEN Inc.) and ubiquitin (Dako), respectively. GFP was detected using a monoclonal antibody from Quantum Biotechnology. The anti-tubulin antibody was a gift from Dr. Michael McBurney. E2F-1 was detected using a rabbit polyclonal antibody (Santa Cruz Biotechnology).

Exogenous Variants of Ubiquitin Can Substitute for Endogenous Ubiquitin in Conjugation Reactions-In cells, monomeric
ubiquitin is produced from the cotranslational processing of either the polyubiquitin gene product or linear fusions of ubiquitin with ribosomal subunits (16,17). These fusions are processed to yield free monomeric ubiquitin that becomes available to tag cellular proteins and direct them to the proteasomal degradation apparatus. Callis and co-workers previously demonstrated that the addition of a hexahistidine tag does not interfere with the functionality of ubiquitin (18) and that K48R mutant ubiquitin can participate in the formation of conjugates in plant cells (19). We incorporated these findings in the design of our strategy to generate linear fusions of a His 6 -tagged wild-type or mutant human ubiquitin with GFP (Fig. 1A). These expression vectors were then used to generate stably expressing HT4 cell populations. We have previously reported that the wild-type ubiquitin/GFP fusion is processed to yield free ubiquitin in the same manner as the natural ubiquitin/ ribosomal subunit fusions, and we have demonstrated that epitope-tagged ubiquitin is conjugated to cellular proteins (14). We now demonstrate that K48R and K63R mutant ubiquitin/ GFP fusions were also processed in HT4 cells to give rise to monomeric ubiquitin that could be detected by Western blot analysis with an antibody directed against the His 6 epitope (Fig. 1). In addition to detecting the ubiquitin monomer, the Cytoplasmic cell extracts were prepared as described under "Experimental Procedures" were resolved on a two-phase SDS-polyacrylamide gel (15 and 8%). An antibody specific for the His 6 epitope detected the presence of high molecular weight ubiquitin conjugates (vertical bar) as well as the epitope-tagged ubiquitin monomer in lysates from cells stably transfected with the wild-type (wt) ubiquitin, K48R mutant ubiquitin, and K63R mutant ubiquitin constructs; these species were not detected in lysates from untransfected HT4 control cells. Monomeric ubiquitin (arrow) could be detected in all lysates using an anti-ubiquitin antibody (anti-Ub; which detected both endogenous and transgene-derived ubiquitin). The migration of markers is indicated on the left, with molecular mass in kilodaltons. The membrane was stripped and probed simultaneously for GFP and tubulin expression (boxed panel). The GFP-specific antibody detected the processed GFP marker protein and was indicative of the levels of transgene expression. An anti-tubulin antibody served as a loading control.
antibody also recognized higher order conjugates in lysates from HT4 cells stably transfected with the wild-type or mutant ubiquitin constructs; these conjugates could not be detected in untransfected control lysates (Fig. 1B). Reprobing the same membrane with an anti-ubiquitin antibody revealed the presence of ubiquitinated proteins in these high molecular weight species. Western blotting with a GFP-specific antibody confirmed the efficient processing of the ubiquitin/GFP fusions and indicated that the level of expression of the K48R mutant ubiquitin/GFP fusion was slightly higher than that of the wildtype ubiquitin/GFP fusion, whereas the level of expression of K63 mutant ubiquitin/GFP was lower (Fig. 1B, boxed panel).
Sensitivity of Transfected Cells to Cadmium-Previous findings in yeast have demonstrated that a deficiency in the ubiquitin pathway renders yeast cells more sensitive to cadmiuminduced damage (6). Cadmium, by a poorly defined mechanism, has also been shown to induce the profound ubiquitination of proteins, resulting in the accumulation of high molecular weight conjugates in the HT4 neuroblastoma cell line (7)(8)(9). To test the hypothesis that cells expressing mutant versions of ubiquitin designed to interfere with normal chain formation would exhibit enhanced sensitivity to this potent cell toxin, we treated HT4 cells and pools of cells stably transfected with various versions of ubiquitin with 25 M Cd 2ϩ for up to 8 h. Western blot analysis of lysates from the stably transfected cells revealed that cadmium induced extensive ubiquitination of cellular proteins in all cell types in a time-dependent manner (Fig. 2). The accumulation of endogenous ubiquitinated proteins was detected by reprobing the membrane with an antibody directed against ubiquitin. The results of this Western blot analysis did not suggest gross differences in the levels of ubiquitination among the different cell lines at any given point. By light microscopy, we observed that the morphology of all cell types was affected, leading to cell rounding and loss of adhesion, and that these changes occurred in a time-and dose-dependent manner (Fig. 3 and data not shown). There was no profound difference in the percentage of dead to live cells among the various cell lines treated with 5 M cadmium for 24 h (data not shown); however, there were notable differences in the appearance of cells, with K48R and K63R mutant ubiquitin-expressing cells being more severely affected than untransfected cells or cells expressing exogenous wild-type ubiquitin (Fig. 3). It appears that under these conditions, many cells detached, but were still able to exclude the trypan blue dye. When glutathione was depleted by pretreatment with buthionine sulfoximine (thereby eliminating the predominant cellular defense against oxidation), the sensitivity to cadmium (as assayed by loss of adherence) was found to be increased in K48R pools, but most pronounced in cells expressing K63R mutant ubiquitin (Fig. 3), where very few attached cells remained. Treatment of cells with buthionine sulfoximine alone did not alter their morphology.
Sensitivity of Transfected Cells to Canavanine-Unlike cadmium, which may directly or indirectly generate damage to nascent protein, canavanine is an amino acid analog that substitutes for arginine during translation, resulting in protein misfolding. In yeast, ubiquitin-mediated proteolysis is required to eliminate canavanyl proteins that would otherwise accumulate with deleterious effects (4,20). To determine whether mammalian cells expressing mutant ubiquitin would be sensitive to a burden of misfolded protein, we added canavanine to the culture media and assessed its toxicity to the various cell pools. Phase-contrast microscopy of cells treated with 20 mM canavanine for 30 h revealed that K48R mutant ubiquitin/ GFP-expressing cells exhibited the highest sensitivity to canavanine exposure as indicated by the dramatically altered cell morphology ( Fig. 4; representative data from four independent experiments). K63R mutant ubiquitin/GFP-expressing cells were found to be as sensitive to canavanine as the untrans- fected HT4 controls, whereas cells transfected with wild-type ubiquitin/GFP actually showed increased resistance to canavanine as evidenced by their morphology. The response to canavanine was quantified by counting live and dead cells, both adherent and detached. Our analysis revealed that among treated cells, there were significantly more detached cells in the pools of K48R mutant ubiquitin/GFP-expressing cells, and a higher proportion of these cells were dead compared with cells expressing other forms of ubiquitin (Fig. 5). No significant difference was observed between the untransfected cells and the K63R mutant ubiquitin/GFP-transfected cells. Western blot analysis of cell extracts from cells treated with 10 mM canavanine for 24 h revealed that canavanine could induce the accumulation of high molecular weight conjugates in all three cell types, with the greatest effects noted in cells expressing K48R mutant ubiquitin (Fig. 6A). The membrane was reprobed with an antibody directed against ubiquitin to confirm the presence of ubiquitin in the high molecular weight conjugates of the stably transfected cells as well as the ubiquitination and accumulation of endogenous substrates in the HT4 control cell extract (Fig. 6B). Densitometric analysis revealed that there was an ϳ3-fold increase in the level of ubiquitin conjugates in K48R mutant ubiquitin-expressing cells compared with cells expressing wild-type ubiquitin (Fig. 6C).
Accumulation of Proteasome Substrate Proteins-The sensitivity of cells expressing K48R mutant ubiquitin to agents that generate aberrant proteins suggested the possibility that such cells might have a general deficit in ubiquitin-mediated proteolysis, even in the absence of exogenous stressors. To determine whether this was the case, we examined the abundance of two unrelated substrates of the ubiquitin/proteasome pathway. One was the natural substrate protein E2F-1, whose degradation via the ubiquitin system is well documented (21)(22)(23). We have previously demonstrated the conjugation of epitopetagged ubiquitin to E2F-1 in cells cotransfected with epitopetagged wild-type ubiquitin and an E2F-1 expression vector (14). To analyze the effects of the K48R mutant ubiquitin mutation on E2F-1, the E2F-1 expression plasmid was transfected into HT4 cell pools stably expressing epitope-tagged wild-type or K48R mutant ubiquitin. E2F-1 protein levels in transfected cell lysates were determined relative to tubulin on blots simultaneously probed for both proteins. In cells expressing K48R mutant ubiquitin, the abundance of E2F-1 was found to be roughly twice that detected in cells expressing wild-type transgene-derived ubiquitin or only endogenous ubiquitin (Fig.  7A). An even greater effect was seen for a second, unrelated proteasome substrate. GFP U is a synthetic proteasome substrate generated through the addition of a destabilizing peptide sequence to GFP; GFP U has been shown to accumulate in cells presented with an overwhelming burden of aberrant protein or in cells treated with proteasome inhibitor (24). We observed a Ͼ5-fold increase in GFP U protein levels in cells stably expressing K48R mutant ubiquitin compared with cells expressing the wild-type ubiquitin transgene (Fig. 7B). DISCUSSION Much of what is known about the ubiquitin/proteasome pathway has been determined using the yeast model system. The use of epitope-tagged ubiquitin to study the fate of wild-type or mutant ubiquitin in conjugation and proteasomal targeting is well established in yeast (11). More recently, epitope-tagged ubiquitin has proved of great utility in transfected cells of higher eukaryotes (25,26), and transgenic plants have been generated in which tagged ubiquitin can be used to retrieve ubiquitinated substrates (19). We have reported that an epitope-tagged ubiquitin/GFP fusion protein is efficiently processed in mammalian cells and in transgenic mice and that the ubiquitin moiety so generated is competent for conjugation to substrate proteins (14). Here we report that dominant-negative mutant versions of ubiquitin, when stably introduced into murine HT4 neuroblastoma cells, were also processed and that conjugates could be detected by Western blot analysis (Fig. 1). Our objective in creating K48R mutant ubiquitin was to investigate the role of the ubiquitin/proteasome pathway in response to oxidative damage in mammalian cells. It has been previously reported that yeast cells deficient in the ubiquitin-conjugating enzyme Ubc7 are hypersensitive to cadmium-induced protein damage (6). Mammalian cells respond to cadmium-mediated oxidative damage by triggering a ubiquitin/proteasome response, and previous studies in the HT4 line have demonstrated the accumulation of ubiquitin-protein conjugates in a time-and dose-dependent manner (7). Cadmium was found to deplete cellular pools of glutathione (the cell's primary defense against oxidation), resulting in an accumulation of mixed protein thiols (8). We have observed the accumulation of high molecular weight ubiquitin conjugates in cadmium-treated cells, detectable with the epitope tag-specific antibody or with a ubiquitin-specific antibody (Fig. 2). These data suggest that transgene-derived ubiquitin becomes incorporated into ubiquitin chains in response to oxidative stress. The cadmiuminduced ubiquitination of proteins was found to occur in all of our transfected HT4 cell lines (Fig. 2). The differences in intensity observed between HT4 cells expressing wild-type versus K63R mutant ubiquitin are attributable to the lower level of K63R mutant ubiquitin, and we speculate that higher level expression of the K63R mutant ubiquitin/GFP fusion would result in comparable levels of the high molecular weight smear of ubiquitinated proteins detected by Western blot analysis. Although the differences in expression levels among the lines were not great, there were profound differences in the sensitivity of the lines to cadmium as evidenced by the striking morphological changes observed under light microscopy. The cells expressing K48R mutant ubiquitin were more severely affected than the cells expressing exogenous wild-type ubiquitin or untransfected control cells, but the most pronounced effects were observed in the K63R mutant ubiquitin-expressing cells (Fig. 3). Cadmium is known to damage multiple targets in cells, including protein, lipids, and DNA (27). Our data suggest that in neuroblastoma cells exposed to cadmium, DNA may be a more important target than protein, given the relative sensitivity of K63R versus K48R mutant ubiquitin-expressing cells. There is clear evidence from yeast (28) and more recently from mammalian cells (29) demonstrating a role for Lys 63linked ubiquitin chains in DNA repair, particularly in the pathway that mediates replication bypass through lesion-containing DNA. It is our supposition that HT4 cells expressing K63R mutant ubiquitin have a compromised ability to assemble the Lys 63 -linked chains that are required for the repair of cadmium-induced DNA damage. K63R mutant ubiquitin has been found to sensitize human lung cancer cells to other DNAdamaging agents. 2 Lys 63 -linked chains have been reported to be involved in other pathways in the cell, including ribosome assembly (30) and IB degradation (31), and we cannot exclude contributions to the cadmium toxicity we have observed in K63R mutant ubiquitin-expressing cells from mechanisms unrelated to DNA repair.
To reduce the complicating effects of DNA damage and to increase the component of toxicity due to protein damage, we treated cells with canavanine, an analog of arginine whose incorporation during protein translation leads to misfolding. Interference with the ubiquitin proteolytic pathway is known to sensitize yeast to canavanine (32), whereas overexpression of wild-type ubiquitin enhances the survival of yeast in the presence of this arginine analog (4). Yeast cells expressing K48R mutant ubiquitin have an impaired ability to degrade canavanyl proteins (5), which is probably due to their diminished ability to assemble the Lys 48 -linked chains that are the primary signal for proteasomal targeting (10). In accordance with our prediction, HT4 cells expressing K48R mutant ubiquitin showed greater sensitivity to canavanine than cells expressing K63R mutant ubiquitin (Fig. 4). As had been previously reported in yeast (4), overexpression of wild-type ubiquitin conferred some level of resistance, as evidenced by the viability of cells transfected with the wild-type expression vector compared with untransfected control cells. The relative sensitivity of K48R mutant ubiquitin-expressing cells cannot be explained by higher levels of transgene expression in K48R mutant ubiquitin expressors versus the other cell lines because expression of wild-type ubiquitin and K63R mutant ubiquitin conferred resistance, not sensitivity ( Figs. 1 and  5). It is likely that K48R mutant ubiquitin interferes with the assembly of proteasomal targeting signals on canavanyl proteins. We have witnessed an accumulation of high molecular weight ubiquitin conjugates in cells expressing K48R mutant ubiquitin, even in the absence of additional stress (Fig. 6). This phenomenon may also occur in the transgenic Arabidopsis model and may contribute to the enhanced recovery of ubiquitinated proteins by nickel chromatography that was reported for K48R mutant versus wild-type ubiquitin (19). Experiments performed in our laboratory on clonal populations (four different clones were tested, one of which is shown in Fig. 6) revealed that lysates of K48R mutant ubiquitin-expressing cells consistently contained elevated levels of high molecular weight conjugates relative to cells expressing the wild-type ubiquitin transgene; GFP levels were equivalent as assessed by Western blotting with an antibody directed against GFP or by flow cytometry (data not shown). The data suggest that even in the absence of any stress, K48R mutant ubiquitin either caps the growing ubiquitin chains, resulting in stable subthreshold length chains, or directs the formation of unconventional linkages not competent to target proteins for proteolysis. Our Western blot data do not allow us to discriminate between these possibilities; and given the complexity of potential substrates, it is possible that both types of chains are represented in the high molecular weight smear. The assembly of nonfunctional chains would have the effect of depleting ubiquitin pools; and in some experiments, we noted a decrease in the intensity of monomeric ubiquitin in cells expressing K48R mutant ubiquitin. Depletion of ubiquitin pools would compromise the ability of cells to cope with the burden of aberrant protein presented by agents like canavanine and may be the basis for the increased detachment and cell death that we observed in K48R mutant ubiquitin-expressing cells treated with canavanine (Fig. 5). All of these mechanisms (formation of subthreshold chains, assembly of chains with unusual linkages, and depletion of monomeric ubiquitin pools) would have global effects on ubiquitin-mediated proteolysis, and we have documented such effects using two unrelated proteasome substrates (Fig. 7). The present data do not allow us to specify which of the inhibitory mechanisms is operative in cells expressing K48R mutant ubiquitin. Further experiments will be required to determine this.
Oxidative damage has been associated with aging as well as with the onset or progression of a variety of neuropathological states (reviewed in (33). It has been proposed that oxidation of cellular proteins may induce the formation of aggregates that, if not eliminated, can compromise neuronal viability. Proteins that misfold as a consequence of genetic mutations may suffer a similar fate. Neurodegenerative disorders are characterized by the presence of abnormal cytoplasmic or intranuclear protein aggregates that are immunoreactive for ubiquitin and components of the ubiquitin machinery (2, 34 -44), and it is reasonable to assume that the ubiquitin pathway, normally charged with the responsibility of clearing abnormal proteins, can be overwhelmed under some circumstances. The work presented herein demonstrates that ubiquitin is cytoprotective for mammalian cells of neural origin and that dominant-negative mutants of ubiquitin inhibit proteolysis and confer sensitivity to agents that damage protein. We are extending these investigations to the in vivo setting of transgenic mouse models.