Oxidative Modifications and Down-regulation of Ubiquitin Carboxyl-terminal Hydrolase L1 Associated with Idiopathic Parkinson's and Alzheimer's Diseases*

Alzheimer's disease (AD) and Parkinson's disease (PD) are the two most common neurodegenerative diseases that occur either in relatively rare, familial forms or in common, sporadic forms. The genetic defects underlying several monogenic familial forms of AD and PD have recently been identified, however, the causes of other AD and PD cases, particularly sporadic cases, remain unclear. To gain insights into the pathogenic mechanisms involved in AD and PD, we used a proteomic approach to identify proteins with altered expression levels and/or oxidative modifications in idiopathic AD and PD brains. Here, we report that the protein level of ubiquitin carboxyl-terminal hydrolase L1 (UCH-L1), a neuronal de-ubiquitinating enzyme whose mutation has been linked to an early-onset familial PD, is down-regulated in idiopathic PD as well as AD brains. By using a combination of two-dimensional gel electrophoresis and mass spectrometry, we have identified three human brain UCH-L1 isoforms, a full-length form and two amino-terminally truncated forms. Our proteomic analyses reveal that the full-length UCH-L1 is a major target of oxidative damage in AD and PD brains, which is extensively modified by carbonyl formation, methionine oxidation, and cysteine oxidation. Furthermore, immunohistochemical studies show that prominent UCH-L1 immunostaining is associated with neurofibrillary tangles and that the level of soluble UCH-L1 protein is inversely proportional to the number of tangles in AD brains. Together, these results provide evidence supporting a direct link between oxidative damage to the neuronal ubiquitination/de-ubiquitination machinery and the pathogenesis of sporadic AD and PD.

Alzheimer's disease (AD) 1 and Parkinson's disease (PD) are the two most common neurodegenerative disorders in humans, however, the causes of AD and PD, particularly the sporadic cases, remain unclear. A prominent feature of AD and PD as well as other neurodegenerative diseases is the accumulation of insoluble proteinaceous deposits, such as senile plaques and neurofibrillary tangles in AD and Lewy bodies in PD (1). Although these deposits have different protein compositions, they all contain ubiquitin and ubiquitinated proteins (2). Interestingly, mutations in two enzymes of the ubiquitination/de-ubiquitination system, parkin and ubiquitin carboxyl-terminal hydrolase L1 (UCH-L1), have been identified as causative genetic defects for certain familial forms of PD (3). In familial AD patients, an aberrant form of ubiquitin resulting from a ϩ1 frameshift mutation in the ubiquitin-B gene has been detected (4). Moreover, impaired proteasome function has been reported in idiopathic AD and PD brains (2). Collectively, these different lines of evidence support a role for dysfunction of the ubiquitinproteasome pathway in the pathogenesis of AD and PD.
Oxidative stress is another important factor that has been implicated in the pathogenesis of a number of age-related neurodegenerative diseases, including AD and PD (5,6). Both AD and PD have been associated with increased production of reactive oxygen species (ROS), which could result from genetic predisposition and/or environmental factors, such as exposure to pesticides (5). Postmortem analyses reveal that the overall level of oxidative damage to proteins, lipids, and DNA is elevated in AD and PD brains (7,8). Although oxidative stress and ubiquitin-proteasome dysfunction have both been implicated in the pathogenesis of AD and PD, the interplay between these two processes is not understood.
Despite ample evidence indicating an elevation in overall protein oxidation in AD and PD, the specific protein targets of oxidative damage in these diseases are largely unknown. Furthermore, it remains to be determined whether the protein targets of oxidative damage are the same or are different in AD and PD. To address these issues, we performed a search for specific protein targets of oxidative damage in AD and PD brains by using a proteomic approach that combined two-dimensional gel electrophoresis and Western blot analysis with mass spectrometry. This search led to the identification of three UCH-L1 isoforms that are differentially expressed and oxidized in AD and PD compared with control brains. Our results suggest that oxidative damage to the ubiquitination/deubiquitination machinery may be critically involved in the etiology of AD and PD.

EXPERIMENTAL PROCEDURES
Human Brain Samples-Frontal cortex tissues from five PD, five AD, and five healthy non-demented control subjects (Table I) were obtained from the Emory Alzheimer's Disease Center brain bank. The neuropathological diagnosis of PD was based on the presence of nigral degeneration and Lewy bodies. The diagnosis of AD was established using CERAD (Consortium to Establish a Registry for Alzheimer's Disease) criteria (9). ApoE genotypes (Table I) were determined for all subjects as previously described (10).
SYPRO Ruby Staining-Duplicate samples of brain proteins were subjected to two-dimensional gel electrophoresis as described above. Proteins in one gel were stained with SYPRO Ruby protein gel stain (Bio-Rad), whereas proteins in the other gel were electroblotted to polyvinylidene difluoride (PVDF) membranes using the Ettan-DALT system. For SYPRO Ruby staining, proteins were first fixed in the gel using 40% methanol/10% acetic acid (v/v) for 30 min. The gel was then incubated in SYPRO Ruby protein gel stain solution overnight and subsequently destained using 10% methanol/6% acetic acid (v/v) for 45 min. The SYPRO Ruby stained gel was placed in a light-tight cabinet directly on a transilluminator, and the fluorescence generated by excitation with UV light at 365 nm was recorded using a cooled, computerized charge-coupled device camera-based imaging system (Alpha Innotech, San Leandro, CA).
Immunoblotting and Image Analyses-The PVDF membranes were removed from the Ettan-DALT electroblotting apparatus and incubated for 1 h with phosphate-buffered saline containing 3% (v/v) Tween and 5% (w/v) non-fat dried skim milk (PBS-TM). Membranes were incubated overnight at 4°C with anti-DNP primary antibody (Molecular Probes, Eugene, OR) at 1:16,000 dilution as described previously (11). After extensive washing with PBS-TM, the membranes were then incubated for 1 h at 4°C with a 1:16,000 dilution of horseradish peroxidase-conjugated goat anti-rabbit secondary antibody (Sigma, St. Louis, MO). Immunostained proteins were detected by using a SuperSignal chemiluminescence kit (Pierce, Rockford, IL) and an Alpha Innotech imaging system. Digitized images from SYPRO Ruby-stained gels and immunoblots were analyzed using the two-dimensional electrophoresis gel analysis program PD Quest (Bio-Rad).
Mass Spectrometry-Spots of interest were excised from the gels and digested in situ with trypsin (modified, Promega, Madison, WI Immunohistochemistry-Brains were fixed in 4% paraformaldehyde, cryoprotected in sucrose, and serial sections (50 m) of cingulate cortex were cut on a cryostat. Sections were treated with 3% H 2 O 2 to eliminate endogenous peroxidase activity and blocked with a solution containing 8% goat serum, 10 g/ml avidin, 0.1% Triton X-100, 0.9% NaCl, and 50 mM Tris-HCl (pH 7.2). Sections were then incubated overnight at 4°C with anti-UCH-L1 (1:1000; Chemicon) followed by incubation with biotinylated secondary antibody (1:200). Bound antibodies were detected using a standard avidin-biotinylated peroxidase complex (ABC) method (Vectastain Elite ABC kit, Vector Laboratories). Sections were observed using a Leica DC500 microscope, and images were captured with a C4742-95 Hamamatsu digital camera.
Quantitation of Neurofibrillary Tangles-Eight-micron sections of paraformaldehyde-fixed, paraffin-embedded frontal cortex from the same five AD cases used for analysis of oxidation-sensitive proteins were deparaffinized and blocked with 0.5% nonfat dried skim milk. Sections were then incubated with rabbit polyclonal anti-tau antibody (Accurate Chemical and Scientific Inc.) at 1:100 for 1 h at 37°C followed by incubation with biotinylated secondary antibody. Bound antibodies were detected using the Vectastain Elite ABC method. Sections were counterstained with hematoxylin. Neurofibrillary tangles were counted in 10 random 20ϫ fields by an investigator blinded to the protein analysis results.

Differential Expression of Proteins in PD and AD Brains
Compared with Age-matched Controls-To investigate alterations in protein expression associated with PD and AD, we performed comparative high resolution, two-dimensional gel electrophoresis experiments on protein samples obtained from PD and AD brains and age-matched controls. Protein spots were visualized by staining with SYPRO Ruby, a fluorescent dye capable of achieving linear and sensitive staining of proteins. Fig. 1 shows partial images of SYPRO Ruby-stained two-dimensional gels of brain proteins from three individuals from the control, PD, and AD groups. These images illustrate a high degree of reproducibility of protein profiles among different subjects diagnosed with the same disease. We found that three distinct protein spots, #1, #2, and #3, which electrophoresed on the two-dimensional gel with apparent molecular mass/ isoelectric point (pI) values of 25 kDa/5.7, 19 kDa/6.1, and 17 kDa/6.1, respectively, showed differential expression among control, PD, and AD ( Fig. 1). Quantification of the intensities of these spots (Fig. 2) revealed that, in AD, the protein levels of all three spots were decreased by ϳ50% compared with agematched controls. In PD, the level of spot #1 was decreased by 30%, whereas the levels of spots #2 and #3 were virtually unaltered. Oxidative Modification of Specific Proteins in PD and AD Brains-A common oxidative modification of proteins induced by reactive oxygen species is the formation of protein carbonyls, which can be detected by reaction with 2,4-dinitrophenylhydrazine (DNPH). To identify protein targets of oxidative damage in PD and AD, brain extracts from PD, AD, and control subjects were resolved by isoelectric focusing on an immobilized pHgradient strip followed by in-strip DNP derivatization (reacting with protein carbonyls) and second-dimensional separation by SDS-PAGE. Oxidatively modified proteins in the two-dimensional gels were detected by immunoblotting with anti-DNP antibody. Interestingly, among the three protein spots described earlier (spots #1, #2, and #3), only spot #1 was found to exhibit a high degree of oxidation, as evidenced by the signifi-cant increase in DNP antibody reactivity of this spot in PD and AD brains (Fig. 3). Quantification results showed that the specific oxidation level of spot #1 was increased ϳ10-fold in PD and 8-fold in AD compared with age-matched control brains (Table II).
Identification of Protein Spots #1, #2, and #3 as UCH-L1 Isoforms by Mass Spectrometry-For identification of individual proteins of interest, the spots were excised from the twodimensional gel, digested with trypsin, and analyzed by mass spectrometry (Figs. 4 and 5). Three of the brain protein spots (spots #1, #2, and #3) shown in Figs. 1 and 3 were unambiguously identified as UCH-L1. Fig. 4A shows the peptide sequences of human UCH-L1 in spots #1, #2, and #3 that were detected by using a combination of MALDI-TOF/MS and HPLC-ESI/MS/MS. Data from the mass spectrometric analyses indicated that spot #1 represents the full-length form of human UCH-L1 (accession number NP_004172). This assignment is further strengthened by the agreement between the apparent molecular mass/pI values (25 kDa/5.7) of spot #1 and the predicted values of 24.8 kDa/5.3. By comparison of the mass spectral profiles, we found that the tryptic fragment of spot #1 containing the amino-terminal 15 amino acids of UCH-L1 was missing from the mass spectra of spots #2 and #3 (Fig. 4,  compare A with B). Furthermore, the apparent molecular mass of spot #2 (19 kDa) and spot #3 (17 kDa) were significantly lower than the predicted 24.8 kDa for full-length UCH-L1. Together, these data suggest that spots #2 and #3 represent UCH-L1 isoforms with amino termini that are shorter than the published full-length sequence for UCH-L1.
Identification of Oxidatively Modified Residues in Spots #1, #2, and #3 by Mass Spectrometry-MALDI-TOF/MS and HPLC-ESI/MS/MS analyses of tryptic digests obtained from spots #1, #2, and #3 were employed to characterize oxidative modifications in UCH-L1 isoforms. The results indicated that three methionine residues, Met-1, Met-6, and Met-12, at the amino terminus of UCH-L1 were oxidized to methionine sulfoxide (MetO) in spot #1 from PD and AD samples (Figs. 4 and  5A). In addition, Met-124 in spots #1 and #3 and Met-179 in spots #1 and #2 were also found to be MetO in PD and AD brains ( Fig. 4 and additional data not shown). Finally, Cys-220, a cysteine residue in the carboxyl-terminal region of UCH-L1, was found to be oxidized to cysteic acid (also known as cysteine sulfonic acid, Cys-SO 3 H) in all three spots (spots #1, #2, and #3) from PD and AD samples (Figs. 4 and 5B).
Confirmation of the Protein Identity of Spots #1, #2, and #3 by Western Blot Analysis-To confirm the protein identifications obtained by mass spectrometry, we performed immunoblot analysis of the same brain protein extracts using an antibody generated against a synthetic peptide composed of amino acids 61-78 of human UCH-L1. Because a peptide fragment matching amino acids 64 -78 of human UCH-L1 was positively identified in spots #1, #2, and #3 by mass spectrometry (Fig. 4), it was expected that this anti-UCH-L1 antibody should recognize all three UCH-L1 isoforms. Indeed, Western blot analysis (Fig. 6) demonstrated that all three protein spots (spots #1, #2, and #3) were immunoreactive to the anti-UCH-L1 antibody, providing further evidence in support of their identification as UCH-L1 isoforms.
Association of UCH-L1 with Neurofibrillary Tangles in AD Brains-Because UCH-L1 was identified as a significantly oxidized protein with altered expression in PD and AD brains, we used immunohistochemistry to investigate whether UCH-L1 is associated with neuropathological hallmarks of these diseases, such as neurofibrillary tangles, senile plaques, and Lewy bodies. In control human brains, UCH-L1 immunoreactivity was observed in neuronal cell bodies and processes throughout the neocortex (Fig. 7A). The UCH-L1 immunostaining patterns of PD brains were similar to the controls, although the intensity of the immunostaining was somewhat decreased in PD compared with the control group (Fig. 7B). No UCH-L1 immunoreactivity was detected in Lewy bodies of PD brains or in senile plaques of AD brains (data not shown). In contrast, prominent UCH-L1 immunostaining was associated with neurofibrillary tangles, a neuropathological lesion containing polymerized microtubule-associated protein tau (Fig. 7, C and D).
Correlation of UCH-L1 Levels with Neurofibrillary Tangle Numbers in AD-Next, we assessed the relationship between soluble UCH-L1 protein levels and the numbers of neurofibrillary tangles in AD brains. The number of neurofibrillary tangles in the frontal cortex of individual AD patients was quantified as described under "Experimental Procedures" and compared with the total levels of all three UCH-L1 isoforms in the corresponding protein samples (Table III). Linear regression analysis revealed that the quantities of soluble UCH-L1 protein were inversely proportional to tangle number (r ϭ Ϫ0.94997, p Յ 0.05 by Pearson's r).

DISCUSSION
Although UCH-L1 has been implicated in a rare familial form of PD (13), little is currently known about the role of UCH-L1 in the more common, sporadic forms of PD and in other age-related neurodegenerative diseases, such as AD. In this study, we found that human brain UCH-L1 exists as three different isoforms: a full-length form (isoform #1) and two amino-terminally truncated forms (isoforms #2 and #3). The truncated isoforms could be specific products of in vivo proteolytic processing or nonspecific cleavage fragments of postmortem degradation of UCH-L1. We favor the first interpretation, because the levels of isoforms #2 and #3 were very similar among samples of the control and PD groups with different postmortem intervals (Table I and Figs. 1 and 2). Furthermore, the levels of UCH-L1 isoforms #2 and #3 were decreased only in AD, but not in PD brains (Figs. 1 and 2), suggesting that there might be disease-specific alterations in amino-terminal processing of UCH-L1 protein. In contrast, a decreased level of isoform #1 was observed in both AD and PD brains, suggesting that down-regulation of full-length UCH-L1 could be a common pathologic feature in both diseases.
Ample evidence supports a crucial role for oxidative stress in causing cumulative damage to cellular macromolecules, thereby contributing to the pathogenesis of AD, PD, and other age-related neurodegenerative diseases (5,6). The most widely used marker for oxidative damage to proteins is the presence of carbonyl groups, which can be introduced into proteins by direct oxidation of Pro, Arg, Lys, and Thr side chains, or by Michael addition reactions with products of lipid peroxidation or glycooxidation (14). Elevation in the total level of protein carbonyls has been documented in both AD and PD (7,8). However, the identities of the oxidized proteins that have been altered by addition of carbonyl groups or other modifications

TABLE II Increased oxidation of protein spot # I in PD
and AD brains compared to controls The specific oxidation index of spot #1 in Fig. 3  remain largely unknown. By using a proteomic approach, we have identified UCH-L1 isoform #1, but not isoforms #2 and #3, as a significantly oxidized protein in PD and AD brains that is oxidative modified by the addition of carbonyls. A 10-fold in-crease in the specific oxidation level of full-length UCH-L1 was observed in idiopathic PD brains compared with age-matched controls, providing the first evidence linking oxidative modification of UCH-L1 to sporadic PD. In idiopathic AD brains, we found that the specific oxidation level of full-length UCH-L1 was increased 8-fold, which is in contrast to the 2-fold increase recently reported by Castegna et al. (15). This discrepancy might be due to differences in the methods used for detection of protein oxidation. In the study of Castegna et al. (15), brain samples were first reacted with DNPH, and reaction by-products were removed by several steps of precipitation and centrifugation before isoelectric focusing. This multistep process might lead to altered isoelectric points and loss of proteins. To avoid these problems, the present study used the in-strip DNP derivatization procedure that was recently developed (11,12), which allows proteins on the IPG strips to be derivatized directly by DNP after isoelectric focusing. We found that the in-strip DNP derivatization procedure is a very sensitive and reliable method for analyzing protein oxidation with high reproducibility.
By using a combination of MALDI-TOF/MS and HPLC-ESI/ MS/MS analyses, we identified five oxidatively modified methionine residues (Met-1, Met-6, Met-12, Met-124, and Met-179) in full-length UCH-L1 (isoform #1) in AD and PD brains. The oxidation of at least one methionine residue (Met-124 or Met-179) was also observed in UCH-L1 isoforms #2 and #3. Because of its reactive sulfur-containing side chain, methionine residues can be readily oxidized to MetO by a variety of reactive oxygen species, such as O 2 . , H 2 O 2 , ⅐ OH or peroxynitrite (16).
This modification can be reversed by peptide methionine sulfoxide reductase (MsrA), an enzyme that catalyzes the thioredoxin-dependent reduction of methionine sulfoxide to methionine (17). The reversible methionine oxidation/reduction has been suggested to act in a manner analogous to phosphorylation/dephosphorylation for regulating protein function and cellular processes (16). It is thus tempting to speculate that the methionine oxidation of UCH-L1 identified here represents a specific, reversible mechanism for regulation of UCH-L1 activity by intracellular redox status. In addition, it has been proposed that the reversible oxidation of methionine residues may serve as an important defense mechanism for scavenging ROS (18). Such a mechanism would suggest a novel antioxidant function for UCH-L1. Given its high abundance in the brain (1-2% of total soluble protein) (19), UCH-L1 might be an efficient scavenger of ROS that protects neurons from oxidative damage. For UCH-L1 to function as an antioxidant, MsrA would need to continually convert MetO to Met. Recently, it was reported that the level of MsrA activity is decreased in AD brains (20). Impaired MsrA function in combination with elevated ROS levels in AD and PD brains would lead to accumulation of oxidized methionine residues in specific proteins, such as UCH-L1 as reported here. In addition to the oxidized methionine residues, we found that the cysteine residue in the carboxyl-terminal region of all three UCH-L1 isoforms (Cys-220) was oxidatively modified to cysteic acid. Cysteine contains a reactive sulfhydryl group that can be reversibly oxidized to a sulfenic acid (Cys-SOH) under mild oxidative stress. The reversible oxidation of cysteine in some proteins has been shown to regulate protein function (21). However, cysteine sulfenic acid is generally unstable and can be easily oxidized to its irreversible sulfinic (Cys-SO 2 H) and eventually sulfonic (Cys-SO 3 H) form, particularly under strong oxidative stress (22).
The extensive oxidative modifications of UCH-L1 identified in this work demonstrate that UCH-L1 is a major target of oxidative damage in PD and AD brains. The observed oxidative modifications might cause irreversible alteration in the conformation and/or enzymatic activity of UCH-L1 and have deleterious effects on neuronal function and survival. UCH-L1 possesses a well characterized de-ubiquitinating activity that catalyzes the hydrolysis of carboxyl-terminal esters and amides of ubiquitin to generate monomeric ubiquitin (23,24). Such a hydrolase activity is believed to facilitate ubiquitin/protea- FIG. 6. Immunological confirmation of protein spots #1, #2, and #3 as UCH-L1 isoforms. Human brain proteins (350 g) were subjected to two-dimensional gel electrophoresis, followed by immunoblotting with anti-UCH-L1 antibody. All three protein spots identified by mass spectrometry as UCH-L1 isoforms were immunoreactive.

TABLE III Correlation between UCH-L1 levels and tangle numbers in AD brains
The numbers of neurofibrillary tangles in AD frontal cortex were counted under 20ϫ magnification and expressed as mean Ϯ S.D. of the results from ten independent determinations. The UCH-L1 protein level was expressed as a percentage of the level in AD sample #5. Values represent mean Ϯ S.E. of the results from three independent measurements.

Samples
Tangle some-mediated protein degradation by recycling free ubiquitin (23)(24)(25). Interestingly, an I93M point mutation in human UCH-L1 that decreases the hydrolase activity has been linked to a rare, autosomal dominant form of familial PD (13). Furthermore, deletion of exons 7 and 8 containing the hydrolase catalytic residues of murine UCH-L1 causes gracile axonal dystrophy (gad), a recessively transmitted neurodegenerative disease characterized by progressive axonal degeneration (26). Our proteomic studies of idiopathic PD and AD brain samples suggest that effects similar to those of the UCH-L1 genetic mutations described above might be achieved by the identified oxidative modifications of UCH-L1. Consistent with this notion, a recent in vitro study showed that the hydrolase activity of recombinant UCH-L1 was decreased by treatment with 4-hydroxynonenal, a lipid peroxidation product that generates carbonyl groups in proteins via Michael addition reactions (27).
In addition to hydrolase activity, UCH-L1 was reported to exhibit a second, dimerization-dependent, ubiquitin-ubiquitin ligase activity that adds ubiquitin to ␣-synuclein-ubiquitin conjugates via a Lys-63 linkage (28). This ligase activity is thought to be at least partly pathogenic because Lys-63-linked polyubiquitination may inhibit Lys-48 ubiquitination-mediated ␣-synuclein degradation, leading to accumulation and aggregation of ␣-synuclein. In support of this idea, a S18Y polymorphic variant of UCH-L1 that has been associated with decreased PD risk has been shown to exhibit reduced ligase activity but normal hydrolase activity (28). The observed oxidative modifications of UCH-L1, particularly methionine oxidation in the amino-terminal region (Met-1, Met-6, and Met-12) close to the Ser-18 polymorphic site, may convert the enzyme into a conformation with enhanced ligase activity, thereby contributing to neurodegeneration.
Protein ubiquitination/de-ubiquitination has emerged as an important mechanism for regulating a variety of cellular processes, including protein degradation, synaptic function, and neuronal apoptosis (29). Aberrant ubiquitin hydrolase and/or ligase activity resulting from the identified oxidative modifications and down-regulation of UCH-L1 might lead to dysfunction of the neuronal ubiquitination/de-ubiquitination machinery, causing synaptic deterioration and neuronal degeneration in PD and AD brains. One of the major consequences of aberrant UCH-L1 activity is an impaired proteasome proteolytic system, which will lead to accumulation of damaged proteins and formation of protein aggregates, as exemplified by the gad mice (26, 30 -32). Oxidative modifications may render UCH-L1 itself more resistant to proteolysis and promote its aggregation into hallmark lesions of AD and PD brains. In support of this possibility, we found the presence of abundant UCH-L1 in neurofibrillary tangles in AD brains. Furthermore, the levels of soluble UCH-L1 protein were inversely proportional to tangle number. Although we were unable to detect UCH-L1 immunostaining in Lewy bodies in our PD samples, the previously reported association of UCH-L1 with Lewy bodies (33) is consistent with this hypothesis.
Neurofibrillary tangles are filamentous deposits consisting of ubiquitinated and hyperphosphorylated tau protein (34,35). Accumulation of neurofibrillary tangles is not only a hallmark of AD, but also the defining neuropathological characteristic of several other neurodegenerative disorders known as tauopathies, including sporadic corticobasal degeneration, progressive supranuclear palsy, hereditary frontotemporal dementia and parkinsonism linked to chromosome 17 (FTDP-17), and Pick's disease (34,36). Studies of these diseases, particularly the identification of mutations in the tau gene as the cause for FTDP-17, have implicated a critical role for neurofibrillary tangles in the onset and/or progression of neurodegeneration (36,37). Recently, it was reported that CHIP⅐Hsc70 complex ubiquitinates phosphorylated tau and promotes the aggregation of tau protein (38). The association of UCH-L1 with neurofibrillary tangles and the inverse correlation between UCH-L1 level and tangle number that we observed in AD brains raise a tantalizing possibility that UCH-L1 may play an important role in preventing neurofibrillary tangle formation by de-ubiquitination of phosphorylated tau. It would be interesting to determine whether UCH-L1 also associates with neurofibrillary tangles in other tauopathies. Further investigation of the relationship between neurodegeneration and oxidative modification of UCH-L1 as well as of other proteins should generate novel insights into the mechanisms of oxidative damage in the pathogenesis of PD and AD, and provide new opportunities for developing therapeutic strategies for treating these devastating diseases.