Ubiquitin, Proteasomes, and the Aging Brainsageke.sciencemag.org/cgi/reprint/2003/34/re6.pdf · Schematic representation of ubiquitin-mediated proteolysis.After the generation of

http//:sageke.sciencemag.org/cgi/content/full/sageke;2003/34/re6 Page 1

Ubiquitinated proteinaceous inclusions are the hall-mark of many neurodegenerative diseases. Ineffi-cient proteolysis might lead to the accumulation andultimate deposition of potentially toxic entities as in-clusions within neurons or glial cells. This hypothe-sis is supported by genetic evidence both from pa-tient populations and from engineered mutations ingenes that encode ubiquitin/proteasome compo-nents in mice. The appearance of similar inclusionsin the brains of elderly individuals of normal andsubclinical conditions begs the question of whetherthere is a general age-related decline in the ability ofthe ubiquitin/proteasome pathway (UPP) to recog-nize and eliminate abnormal proteins, and whethersuch a decline would be reflected by changes in theabundance or activity of some or all components ofthe UPP. Here we describe alterations in the agingmammalian brain that correlate with a decline in thefunction of the UPP and review the evidence for age-related changes in specific UPP components. Thesealterations are discussed within the context of preva-lent theories of aging.

IntroductionTheories of human aging must reconcile apparent inevitabilitywith apparent randomness. Barring some as-yet inconceivableintervention, there appears to be an upper limit for human lifespan, with no documented case exceeding 121 years. The endmay be sudden or protracted, and may be marked by the declineof one organ system or many. Somewhere in the system is astochastic generator breaking the test tubes in Medawar’s*memorable metaphor, and some of us will suffer early and mis-erably for it. In the end, all will succumb before we get too farinto our second century.

Among the more plausible stochastic theories of aging arethose involving the loss of quality control; over time, cells failto eliminate abnormal proteins generated by a lifetime’s worthof environmental damage, and this defect is superimposed oninherent biological sloppiness and any existing genetic predis-positions. Whether the source of these abnormal proteins is thesurprisingly high “normal” background of translational errorsreported in eukaryotic systems (1), elevated oxidation arisingfrom progressive mitochondrial derangement (2), the burdenimposed by expanded polyglutamine proteins [reviewed in (3)],or any number of environmental sources, there may ultimatelybe too much abnormal protein for a cell to deal with, and patho-logical consequences may ensue. One would expect the issue of

abnormal protein accumulation to be particularly acute in anonrenewing cell population, as typified by neurons in the cen-tral nervous system (CNS) of long-lived mammals like our-selves. If presented with a burden of accumulated protein, suchcells cannot be replaced and must soldier on. Morphological ev-idence for this can be found in the form of the age-dependentaccumulation of a pigmented granular lipoprotein byproduct ofcellular metabolism, lipofuscin, in neurons of the aging nervoussystem (4). The plight of such cells would be compoundedshould their proteolytic systems fail, perhaps precipitating acatastrophic deposition of toxic matter. Given that the CNS isthe right place to look for such accumulation (seehttp://sageke.sciencemag.org/literature/overviews/casestudies),we might use neurons as a model system with which to ask:What is the system whose failure underlies such a calamity, andwhat is the evidence that such a failure might occur?

The Ubiquitin/Proteasome PathwayEukaryotic cells have two major systems for the regulated de-struction of proteins: the relatively slow, vesicle-dependentlysosomal pathway, and the more rapid ubiquitin/proteasomepathway (UPP), which is operative in the cytosolic and nuclearcompartments. The mechanistic separation of these pathways isnot complete, with ubiquitin acting as a molecular “tag” for de-struction in both. The association of ubiquitin, the lysosomalpathway, and neurodegenerative disease has long been appreci-ated (5), and the role of ubiquitin in directing traffic within thelysosomal pathway has recently become the focus of consider-able research [reviewed in (6)]. The assembly of ubiquitin intochains for targeting substrates to the 26S proteasome (that is,the macromolecular cellular device charged with protein de-struction) is better understood [for a recent review see (7)]. Tosummarize, ubiquitin, a globular 76-amino acid protein, is gen-erally attached to substrate proteins through the formation ofcovalent isopeptide bonds involving the C-terminal glycineresidue of ubiquitin and the side chain of a lysine residue withinthe substrate. The initial ubiquitin molecule can serve as the tar-get of a second, whose C-terminus is attached to a lysine withinthe first, and so on until the ubiquitin chain is formed, anchoredto the substrate. When this chain reaches a threshold length offour ubiquitin subunits (8), the substrate is directed to the pro-teasome through affinity between the ubiquitin chain and pro-teasomal subunits (9). The substrate is unfolded and passesthrough the interior of the proteasome, where it is degraded by acombination of protease activities; the ubiquitin chain iscleaved from the residual peptide of the substrate, then disas-sembled to yield the monomeric form. (The enzymatic activitiesresponsible for the cleavage of ubiquitin are discussed in furtherdetail below.) The substrate specificity requirements of the UPPare truly remarkable in that the system must distinguish not on-ly the subset of normal proteins that under certain physiologicalconditions must be eliminated (short-lived transcription factors,

Ubiquitin, Proteasomes, and the Aging BrainDouglas A. Gray, Maria Tsirigotis, John Woulfe

(Published 27 August 2003)

R E V I E W

SAGE KESCIENCE OF AGING KNOWLEDGE ENVIRONMENT

D. A. Gray is at the Ottawa Regional Cancer Centre, Ottawa, Ontario,Canada K1H 1C4 and the Ottawa Health Research Institute, Ottawa,Ontario, Canada K1Y 4E9. M. Tsirigotis is at the Ottawa RegionalCancer Centre, Ottawa, Ontario, Canada K1H 1C4. J. Woulfe is atthe Ottawa Health Research Institute, Ottawa, Ontario, Canada K1Y4E9 and in the Department of Pathology, Ottawa Hospital, Ottawa,Canada K1Y 4E9. E-mail: [email protected] * http://sageke.sciencemag.org/cgi/content/full/2001/1/oa1


for example), but must also distinguish pristine and properlyfolded proteins from potentially toxic damaged or misfoldedproteins. The UPP accomplishes this astonishing feat through ahierarchical array of enzymes, including a very small number ofubiquitin-activating enzymes (E1 enzymes), whose job it is toelevate ubiquitin to an energetic state that allows covalent bondformation; a larger set of conjugating enzymes (E2 enzymes)charged with assembling the ubiquitin chains; and an extensiveset of ubiquitin ligases (E3 enzymes) that employ protein inter-action domains to impart substrate specificity. Countering theconstruction of ubiquitin chains on substrates are deubiquitinat-ing enzymes (DUBs) that may stabilize substrates by reducingthe length of ubiquitin chains below the threshold length or byremoving chains from substrates entirely [reviewed in (10)].

Polyubiquitinated proteins are generally targeted and recog-nized for degradation by the 26S proteasome. Each of theseself-compartmentalizing proteases is constructed from a barrel-shaped, 20S core particle in which the proteolytic activity is se-questered, and one or two 19S regulatory particles that cap theends of the 20S particle. The regulatory particles are believed torecognize, unfold, and translocate ubiquitinated substrates intothe central cavity of the 20S proteasome. The UPP is portrayedschematically in Fig. 1.

Age-Related Changes in UPP ProteinsSeveral investigators have examined the concentrations and ac-tivities of UPP components in various cell cultures, tissues fromhumans, and rodent models as a function of age, and it is fair tosay that for many elements of the system, the results have variedwidely. It is not our intention to review these findings in detail—that has been done admirably by other authors [for example, byGaczynska et al. (11)]; rather, the current review will focus onchanges that are of relevance to the CNS. There is no compellingevidence for an age-related increase in CNS concentrations of

monomeric ubiquitin; no such increase has been observed in thecerebella of mice up to 32 months of age, which is very near themaximum life span of laboratory strains (Fig. 2A). A marked in-crease in high-molecular-weight ubiquitin conjugates with in-creasing age was observed in our experiments and has been doc-umented by others (12). Where regional differences have beenexamined (12), the accumulation of conjugates was not uniformbut was more dramatic in the cerebellum and brain stem [the re-gions of the brain responsible for motor functions and vegetativefunctions (which operate below the level of consciousness), re-spectively] than in the cerebrum (the site of higher functions, in-cluding cognitive processing and sensorimotor integration). Thepotential significance of such conjugates is discussed below inthe context of proteasome activity. No changes in E1 enzymeconcentrations or activity have been reported in the aging brain,and we have not detected such changes in the brains of very oldmice (Fig. 2B). The literature reveals little about age-relatedchanges in the concentrations and/or activities of the E2 and E3enzymes. Because of their sheer numbers and widely divergentsubstrate specificities, any reasonably inclusive analysis of age-related changes in mammalian E2 or E3 enzymes would be anextremely large undertaking.

With regard to the proteasome, a consistent functional de-cline has been reported in several species and organ systemswhen samples from elderly rodents were compared to samplesfrom younger organisms. For example, the peptidylglutamylpeptide hydrolyzing (PGPH) activity of the proteasome [an ac-tivity resident in the β1 (Y) subunit of the 20S proteasome thatcleaves substrates C-terminal to acidic residues (13)] has beenshown to decrease by approximately one-half during aging, asevidenced by cleavage of an artificial fluorogenic substrate (14-17). CNS tissues were not included in these studies. In a studyby Keller et al., the chymotryptic activity of the proteasome,which is ascribed to the β5 or X subunit (18), was shown to be

R E V I E W


Fig. 1. Schematic representation of ubiquitin-mediated proteolysis. After the generation of a high-energy C-terminal thioester bond byan E1 enzyme, ubiquitin is covalently attached to a substrate protein (either a normal protein destined for degradation or a damagedor abnormally folded protein). Substrate recognition and subsequent chain assembly are mediated through the activities of E3 and E2enzymes, respectively. Upon reaching the threshold length, ubiquitin chains direct the substrate to the proteasome where it is unfolded(green line) and directed into the central channel of the proteasome. Proteolytic activities lining the inner rings of the proteasomecleave the substrate into short peptides (green bars). DUBs disassemble the ubiquitin chains to yield monomers and may also opposetrafficking to the proteasome by shortening ubiquitin chains or cleaving them off of substrates. A reduction in the efficiency of ubiquitin-mediated proteolysis at any step would lead to an accumulation of substrates, including potentially toxic proteins.


decreased in the cerebrum but not in the cerebellum of rats (19).In their review of proteasome alterations in the aging brain,Keller and co-workers point out that the loss of activity ob-served by various laboratories cannot be explained by a loss ofsubunit expression (20), because where such changes have beennoted, the loss of activity precedes gross changes in proteasomecomponents. We have observed that, in the mouse, at least somecomponents of the 20S proteasome core and 19S regulatorycaps increase with age (Fig. 2B). It is tempting to speculate thata compensatory mechanism exists wherein the loss of activity

leads to an increased abundance of proteasome subunits, an ul-timately futile effort to maintain cellular homeostasis. We cur-rently have no evidence to support this hypothesis.

Results from Microarray StudiesThe advent of DNA microarray technology has provided a pow-erful means of surveying the expression of thousands of genessimultaneously. In the context of aging, this methodology hasrevealed changing patterns of gene expression in the brain thatcan be clustered into functional systems. Comparing the brains

of 5-month-old versus 30-month-old mice, Lee et al. de-tected increased expression of inflammatory genes indica-tive of microglial activation and induction of the comple-ment cascade† (21). Also induced in the aged brains weregenes that encode lysosomal proteases and genes involvedin the stress response, including early response transcrip-tion factors and heat shock proteins. In this study, the levelof ubiquitin mRNA was found to be elevated, whereas thelevel of mRNAs encoding other UPP components was de-creased. This latter group of components included the ubiq-uitin ligase Nedd4, the proteasome β-2 subunit, and thedeubiquitinating enzyme Usp4 (previously designatedUnp). A recent microarray study of gene expression in theaging rat hippocampus also found evidence of increasedgene expression consistent with inflammation and identi-fied the proteasome β-2 subunit as being down-regulated atthe RNA level (22), but changes in other UPP transcriptswere not reported in this study. For the most part, the UPPtargets revealed by the microarray approach await valida-tion by Northern and Western blotting or proteomicsmethodologies, and it would be premature to draw conclu-sions based on the available evidence, but they do point toage-related dysregulation of the UPP at the transcriptionallevel. Intriguingly, caloric restriction‡ (at present, the best-characterized means of retarding aging in mammals) wasfound to at least partially attenuate age-related alterationsin gene expression in the rodent brain (21, 23).

The Peculiar Case of Ub+1

It should not be surprising that the assembly of ubiquitinchains through isopeptide bond formation places con-straints on the sequence of the C terminus of the protein.Indeed, the entire protein is highly conserved, being identi-cal in rodents and humans and differing by only three con-servative amino acid substitutions in the yeast versus thehuman protein. Van Leeuwen’s group has reported a re-markable perturbation of the C terminus of ubiquitin thatmay play a role in aging. It has been suggested that molec-ular misreading by RNA polymerase II of a dinucleotide re-peat in the ubiquitin-B gene (UBB, which encodes a stress-activated in-frame fusion of three tandem ubiquitin pro-teins) results in an aberrant ubiquitin C terminus that is in-capable of participating in conjugation. This aberrant formof ubiquitin is referred to as Ub+1. The neoepitope createdby the frameshift event was detected by immunohistochem-istry in the brains of Alzheimer’s§ and Down syndrome pa-tients (24) and has also been detected in the brains of non-

R E V I E W


Fig. 2. Western blot analysis of age-related UPP changes in themouse brain. Lysates were prepared from the brains of female BALB/cmice aged 12 months (12M) to 32 months (32M), obtained from theNational Institute on Aging, Bethesda, MD. Lysates were preparedfrom two mice at each time point (designated a and b). (A) Lysatesfrom cerebella were probed with a polyclonal antibody to ubiquitin(DAKO Diagnostics Canada). The position of substrates conjugatedwith ubiquitin chains and the position of the ubiquitin monomer are in-dicated by the schematic diagram on the right (the upper two imagesrepresent different exposures of distinct portions of the same mem-brane). Note the increase in the intensity of high-molecular-weightubiquitin conjugates in cerebella at 30 and 32 months of age. Themembrane was stripped and reprobed with an antibody to β-actin [thirdpanel in (A)] (Sigma). (B) Lysates of cerebra were probed with antibod-ies to the ubiquitin activating (E1) enzyme (Upstate), the α-1 subunit ofthe 20S multicatalytic protease-containing proteasome core (Cal-biochem), or the S6 subunit of the 19S substrate-binding and unfoldingcap structure (Affiniti Research Products). As in (A), the membraneswere reprobed with a β-actin antibody (Sigma) to control for loading.

† http://sageke.sciencemag.org/cgi/content/full/2002/29/re3‡ http://sageke.sciencemag.org/cgi/content/full/sageke;2003/8/re2§ http://sageke.sciencemag.org/cgi/content/full/2001/1/oa2


demented elderly individuals (25, 26). The possibility exists thatthe aberrant ubiquitin may itself serve as a substrate for the for-mation of ubiquitin chains that are destined to remain unan-chored to other proteins. Through their affinity for subunitswithin the 19S proteasome cap structures, such unanchoredchains might exert a dominant negative effect on proteasome-mediated degradation of cellular substrates (27). There is alsoevidence that Ub+1 may be recognized as a ubiquitin fusiondegradation substrate (28) with an inherent ability to block theproteasome and induce neurotoxicity (29).

Predicted Consequences of UPP Dysregulation During AgingIt might reasonably be asserted that, for a long-lived nonrenew-ing cell such as the neuron, an efficient UPP is the key tohealthy living; conversely, an inefficient UPP is a recipe for dis-aster. If an intolerable burden of abnormal protein is presentedto the cell, or if one or more components of the UPP are al-lowed to become scarce, the system is vulnerable to spiralinginto a self-amplifying cycle of impairment. For example, themutant form of the huntingtin protein is inefficiently degradedby the proteasome, and when presented with mutant huntingtin,the proteasome is unable to efficiently degrade other substrates,which will then accumulate (30). If the backlog of ubiquitinatedsubstrates is not cleared from the cell, the polyubiquitin chainson these substrates, through binding to proteasome subunits,have the potential to further inhibit the proteasome’s ability torecognize and eliminate aberrant proteins. Thus, the accumula-tion of these proteins is likely to wreak further havoc. The trig-gering insult for the escalating cycle of impairment may comefrom the mutant protein (the product of germline or sporadicmutation) or from protein damaged by oxidation [reviewed in(31)]. The initial burden might also arise from endogenous neu-ronal proteins, such as α-synuclein, a protein with unusual fold-ing properties whose accumulation inhibits the proteasome(32). This model is consistent with the molecular evidence re-viewed above for the accumulation of polyubiquitin chains andincreasing inefficiency of proteasome activity with advancingage. It also predicts dire consequences at the histological level.What is the evidence that such consequences are manifest in theCNS of elderly but nondiseased individuals?

Histological Findings in the Brains of Elderly HumansIntracellular and extracellular protein aggregates are a unifyinghistopathological feature of several human neurodegenerative dis-orders [reviewed in (33)]. These proteins, and by extension the in-clusions they form, are ubiquitinated. Accordingly, dysfunction ofthe ubiquitin/proteasome system has been implicated in the patho-genesis of some of these neurodegenerative disorders, most no-tably Parkinson’s diseasell (PD) (see also Andersen Review¶ andDeadly Giveaway#) (34, 35). In some of these diseases, specificneuronal proteins present in the ubiquitinated inclusions are pri-mary candidates for pathogenic mediators. Proteins with polyglu-tamine expansions are found in ubiquitinated intranuclear inclu-sions in the CAG trinucleotide repeat disorders [reviewed in (36)].α-synuclein, convincingly implicated in PD pathogenesis, is aprominent component of Lewy bodies**, which are ubiquitinatedintraneuronal inclusions found in dopaminergic neurons of thesubstantia nigra [reviewed in (37)]. In other neurodegenerativediseases featuring ubiquitinated intraneuronal inclusions, the cul-prit pathogenic protein remains to be determined. For example,

spinal motor neurons in amyotrophic lateral sclerosis†† display in-tracytoplasmic ubiquitin-positive inclusions of a variety of mor-phologies (38). The number and identity of ubiquitinated proteinsin these structures are unknown.

The molecular mechanisms initiating protein accumulationand aggregation in many of these diseases remain to be defined.However, ubiquitination appears to be a common theme in thesedisorders, and it is tempting to speculate that altered UPP func-tion plays a role. A substantial proportion of inherited, early-on-set, autosomal recessive forms of PD are associated with muta-tions in the gene encoding parkin‡‡, an E3 ubiquitin ligase (39,40). Individuals with parkin mutations in only one allele maydevelop later-onset disease, suggesting that parkin-associatedabnormalities of the UPP may be involved in the much morecommon, later-onset, sporadic form of the disease, also knownas idiopathic PD (41).

Ubiquitinated inclusions are not confined to neurodegenerative

R E V I E W


Fig. 3. Ubiquitinated inclusions in neurodegenerative diseases andin normal aging. (A to C) Ubiquitinated inclusions in human neu-rodegenerative diseases. (A) Neurofibrillary tangles in cortical neu-rons in Alzheimer’s disease. (B) A pigmented dopaminergic neuronin the substantia nigra containing two intracytoplasmic Lewy bodiesin PD. (C) Glial cytoplasmic inclusions in oligodendroglial cells in thebasal ganglia in multiple system atrophy. (D to F) Ubiquitinatedstructures in nonpathological aging. (D) An intranuclear Marinescobody in a pigmented dopaminergic neuron of the substantia nigra.(E) A rodlike cytoplasmic inclusion in a large neuron of the caudatenucleus. (F) A granular cytoplasmic inclusion in a neuron of the infe-rior olivary nucleus. (G) Ubiquitin-immunoreactive granular bodies inthe neuropil of the granule cell layer of the dentate gyrus of the hip-pocampus. (H) Axonal swellings in the gracile nucleus (physiologi-cal neuroaxonal dystrophy).

ll http://sageke.sciencemag.org/cgi/content/full/sageke;2001/7/dn4¶ http://sageke.sciencemag.org/cgi/content/full/sageke;2001/1/re1 # http://sageke.sciencemag.org/cgi/content/abstract/2002/42/nw145** http://sageke.sciencemag.org/cgi/content/full/sageke;2001/3/dn3†† http://sageke.sciencemag.org/cgi/content/full/sageke;2003/18/pe10‡‡ http://sageke.sciencemag.org/cgi/genedata/sagekeGdbGene;195


disorders, and a variety have been described in specific brain re-gions as a morphological consequence of apparently normal or“nonpathological” aging. By analogy with the appearance of ubiq-uitinated inclusions in neurodegenerative disorders, one can spec-ulate that these structures reflect a senescence-associated dysfunc-tion of the UPP. Examples of such inclusions are shown in Fig. 3.Some of these age-associated structures are reminiscent of patho-logical inclusions in neurodegenerative diseases. Perhaps most in-structive in this context is the Marinesco body, a common age-associated, round, eosinophilic intranuclear structure detectedmost commonly in neurons of the substantia nigra. These struc-tures are ubiquitin-immunoreactive (42), but their biochemicalcomposition is otherwise unknown. Marinesco bodies are thoughtto be formed in response to cellular stress. Morphologically, theyresemble the ubiquitinated intranuclear inclusions found in polyg-lutamine-repeat diseases. In this context, it is interesting that theyaccumulate nonexpanded ataxin 3 (43), a polyglutamine proteinthat, when expanded, causes spinocerebellar ataxia§§ type 3.

There are a variety of additional age-associated structural al-terations that affect neuronal cell bodies that stain for ubiquitin.These include ubiquitin-immunoreactive cytoplasmic rodlikeand skeinlike inclusions found in large (presumably choliner-gic) neurons in the neostriatum (a subcortical structure impor-tant for modulating motor, cognitive, and emotional informa-tion) (44, 45); in eosinophilic, granular inclusion bodies in thecytoplasm of neurons of the inferior olivary nucleus (a cerebel-lar relay nucleus important for motor coordination) (42); and in“colloid” or “hyaline” inclusions often found in hypoglossaland spinal motor neurons in the elderly (46). These neuronsprovide motor innervation to the muscles of the tongue and therest of the body, respectively.

Age-associated ubiquitination is not restricted to inclusionsin neuronal nuclei and cell bodies. The development of ubiqui-tin-positive axonal swellings has long been known to increase infrequency with age in certain parts of the CNS (42), most pre-dominantly in the cuneate and gracile nuclei, which are targetsof sensory information arising in the arms and legs, respective-ly. Similar age-associated changes are seen in sympathetic gan-glia and the cerebral cortex. The sympathetic ganglia containthe cell bodies of neurons that control autonomic functions, in-cluding the cardiovascular, gastrointestinal, and urinary sys-tems. The cerebral cortex controls higher functions, includingcognition and sensorimotor processing. What causes these ax-onal swellings is unknown. These changes, which may accountfor some of the neurological decline seen with aging, are re-ferred to as “physiological” neuroaxonal dystrophy, to distin-guish them from the changes seen in pathological neuroaxonaldystrophy, in which similar axonal swellings develop through-out the brain and spinal cord as part of a complex neurodegen-erative disorder presenting as infantile, late infantile, juvenile,and adult cases (47).

Ubiquitin-immunopositive granular bodies and dotlike struc-tures are also commonly encountered in the aging brain (42,48). The granular bodies are believed to be located in the distalaspects of dystrophic neurites. They are most commonly en-countered in the middle and upper cortical layers, especiallylayer II of the entorhinal cortex. This is potentially important,because these neurons form a crucial link in the circuit respon-sible for short-term memory. They are among the first to degen-

erate in Alzheimer’s disease and show neurofibrillary degenera-tive changes in normal aging. In contrast, dotlike structures areprominent in the white matter and appear to be located in oligo-dendroglia and myelin lamellae (42). The significance of thesestructures is unknown. Most of the inclusions described aboveare confined to certain regions of the nervous system, suggest-ing that specific subsets of neurons are differentially susceptibleto this process. The factors underlying this regionally selectivevulnerability are unknown. This feature recalls the neuropathol-ogy of neurodegenerative disorders, wherein ubiquitinated in-clusions show an anatomically selective pattern of distribution.Does an age-associated decline in the function of the UPP ac-count for the development of these protein deposits? Does thepresence of these inclusions in particular circuits correlate withage-associated decline in the function of these systems, as sug-gested for physiological neuroaxonal dystrophy? If so, then adelinquent UPP function is a prime suspect with respect to theforces culminating in neurological aging.

Does an Inefficient UPP Contribute to Aging?The presence of a variety of ubiquitinated inclusions indicatesthat protein accumulation, aggregation, and ubiquitination oc-cur during aging. By analogy with the involvement of UPP dys-function in some neurodegenerative diseases that feature ubiq-uitinated inclusions, it is tempting to speculate that the presenceof these inclusions in aging reflects an age-associated progres-sive failure of the UPP. Indeed, there are a number of nonneu-ronal diseases in which the UPP may also play a role [reviewedin (49)]; thus, the consequences of age-related changes in theUPP may not be confined to the nervous system. With regard tothe nervous system, it must be said that neuropathologists (thesleuths charged with interpreting the pathophysiological signifi-cance of ubiquitinated inclusions) are all too aware that in bothneurodegenerative disease and normal aging, ubiquitinated in-clusions represent mere fingerprints in a crime scene long aban-doned by additional culprits and accessories. Taken together, theimmunohistochemical and molecular findings implicate a sub-standard, perhaps deranged, UPP in the neural alterations typi-cal of the aged brain; but, to strain the metaphor, the evidencemay be too circumstantial to convict. We expect that more com-pelling evidence will emerge from targeted disruption of UPPgenes in mice. The implications of an acceleration or delay inaging would be profound and might amply reward the investiga-tor with the patience to perform life span studies in mice.

References

1. U. Schubert, L. C. Anton, J. Gibbs, C. C. Norbury, J. W. Yewdell, J. R. Ben-nink, Rapid degradation of a large fraction of newly synthesized proteinsby proteasomes. Nature 404, 770-774 (2000).

2. K. Hirai, G. Aliev, A. Nunomura, H. Fujioka, R. L. Russell, C. S. Atwood, A.B. Johnson, Y. Kress, H. V. Vinters, M. Tabaton, S. Shimohama, A. D. Cash,S. L. Siedlak, P. L. Harris, P. K. Jones, R. B. Petersen, G. Perry, M. A. Smith,Mitochondrial abnormalities in Alzheimer’s disease. J. Neurosci. 21, 3017-3023 (2001).

3. H. T. Orr, Beyond the Qs in the polyglutamine diseases. Genes Dev. 15,925-932 (2001).

4. R. E. Mrak, S. T. Griffin, D. I. Graham, Aging-associated changes in humanbrain. J. Neuropathol. Exp. Neurol. 56, 1269-1275 (1997).

5. R. J. Mayer, C. Tipler, J. Arnold, L. Laszlo, A. Al-Khedhairy, J. Lowe, M. Lan-don, Endosome-lysosomes, ubiquitin and neurodegeneration. Adv. Exp.Med. Biol. 389, 261-269 (1996).

6. P. D. Stahl, M. A. Barbieri, Multivesicular bodies and multivesicular endo-somes: the “ins and outs” of endosomal traffic. Sci. STKE 2002, pe32(2002).

7. M. H. Glickman, A. Ciechanover, The ubiquitin-proteasome proteolytic

R E V I E W


§§ http://sageke.sciencemag.org/cgi/content/full/2003/1/nw5


pathway: destruction for the sake of construction. Physiol. Rev. 82, 373-428(2002).

8. J. Piotrowski, R. Beal, L. Hoffman, K. D. Wilkinson, R. E. Cohen, C. M.Pickart, Inhibition of the 26 S proteasome by polyubiquitin chains synthe-sized to have defined lengths. J. Biol. Chem. 272, 23712-23721 (1997).

9. Y. A. Lam, T. G. Lawson, M. Velayutham, J. L. Zweier, C. M. Pickart, A pro-teasomal ATPase subunit recognizes the polyubiquitin degradation signal.Nature 416, 763-767 (2002).

10. S. S. Wing, Deubiquitinating enzymes—the importance of driving in reversealong the ubiquitin-proteasome pathway. Int. J. Biochem. Cell Biol. 35, 590-605 (2003).

11. M. Gaczynska, P. A. Osmulski, W. F. Ward, Caretaker or undertaker? Therole of the proteasome in aging. Mech. Ageing Dev. 122, 235-254 (2001).

12. H. Ohtsuka, R. Takahashi, S. Goto, Age-related accumulation of high-molecular-weight ubiquitin protein conjugates in mouse brains. J. Gerontol.A Biol. Sci. Med. Sci. 50, B277-B281 (1995).

13. C. Enenkel, H. Lehmann, J. Kipper, R. Guckel, W. Hilt, D. H. Wolf, PRE3,highly homologous to the human major histocompatibility complex-linkedLMP2 (RING12) gene, codes for a yeast proteasome subunit necessary forthe peptidylglutamyl-peptide hydrolyzing activity. FEBS Lett. 341, 193-196(1994).

14. T. Shibatani, M. Nazir, W. F. Ward, Alteration of rat liver 20S proteasomeactivities by age and food restriction. J. Gerontol. A Biol. Sci. Med. Sci. 51,B316-B322 (1996).

15. M. Conconi, L. I. Szweda, R. L. Levine, E. R. Stadtman, B. Friguet, Age-re-lated decline of rat liver multicatalytic proteinase activity and protectionfrom oxidative inactivation by heat-shock protein 90. Arch. Biochem. Bio-phys. 331, 232-240 (1996).

16. F. Bardag-Gorce, L. Farout, C. Veyrat-Durebex, Y. Briand, M. Briand,Changes in 20S proteasome activity during ageing of the LOU rat. Mol. Bi-ol. Rep. 26, 89-93 (1999).

17. B. Anselmi, M. Conconi, C. Veyrat-Durebex, E. Turlin, F. Biville, J. Alliot, B.Friguet, Dietary self-selection can compensate an age-related decrease ofrat liver 20 S proteasome activity observed with standard diet. J. Gerontol.A Biol. Sci. Med. Sci. 53, B173-B179 (1998).

18. W. Heinemeyer, A. Gruhler, V. Mohrle, Y. Mahe, D. H. Wolf, PRE2, highly ho-mologous to the human major histocompatibility complex-linked RING10gene, codes for a yeast proteasome subunit necessary for chrymotrypticactivity and degradation of ubiquitinated proteins. J. Biol. Chem. 268, 5115-5120 (1993).

19. J. N. Keller, K. B. Hanni, W. R. Markesbery, Possible involvement of protea-some inhibition in aging: implications for oxidative stress. Mech. AgeingDev. 113, 61-70 (2000).

20. J. N. Keller, J. Gee, Q. Ding, The proteasome in brain aging. Ageing Res.Rev. 1, 279-293 (2002).

21. C. K. Lee, R. Weindruch, T. A. Prolla, Gene-expression profile of the ageingbrain in mice. Nat. Genet. 25, 294-297 (2000).

22. E. M. Blalock, K. C. Chen, K. Sharrow, J. P. Herman, N. M. Porter, T. C. Fos-ter, P. W. Landfield, Gene microarrays in hippocampal aging: statistical pro-filing identifies novel processes correlated with cognitive impairment. J.Neurosci. 23, 3807-3819 (2003).

23. C. K. Lee, R. G. Klopp, R. Weindruch, T. A. Prolla, Gene expression profileof aging and its retardation by caloric restriction. Science 285, 1390-1393(1999).

24. F. W. van Leeuwen, D. P. de Kleijn, H. H. van den Hurk, A. Neubauer, M. A.Sonnemans, J. A. Sluijs, S. Koycu, R. D. Ramdjielal, A. Salehi, G. J.Martens, F. G. Grosveld, J. Peter, H. Burbach, E. M. Hol, Frameshift mu-tants of beta amyloid precursor protein and ubiquitin-B in Alzheimer’s andDown patients. Science 279, 242-247 (1998).

25. F. W. van Leeuwen, D. F. Fischer, D. Kamel, J. A. Sluijs, M. A. Sonnemans,R. Benne, D. F. Swaab, A. Salehi, E. M. Hol, Molecular misreading: a newtype of transcript mutation expressed during aging. Neurobiol. Aging 21,879-891 (2000).

26. F. W. van Leeuwen, L. Gerez, R. Benne, E. M. Hol, +1 Proteins and aging,Int. J. Biochem. Cell Biol. 34, 1502-1505 (2002).

27. Y. A. Lam, C. M. Pickart, A. Alban, M. Landon, C. Jamieson, R. Ramage, R.J. Mayer, R. Layfield, Inhibition of the ubiquitin-proteasome system inAlzheimer’s disease. Proc. Natl. Acad. Sci. U.S.A. 97, 9902-9906 (2000).

28. K. Lindsten, F. M. de Vrij, L. G. Verhoef, D. F. Fischer, F. W. van Leeuwen, E.M. Hol, M. G. Masucci, N. P. Dantuma, Mutant ubiquitin found in neurode-

generative disorders is a ubiquitin fusion degradation substrate that blocksproteasomal degradation. J. Cell Biol. 157, 417-427 (2002).

29. F. M. De Vrij, J. A. Sluijs, L. Gregori, D. F. Fischer, W. T. Hermens, D.Goldgaber, J. Verhaagen, F. W. Van Leeuwen, E. M. Hol, Mutant ubiquitinexpressed in Alzheimer’s disease causes neuronal death. FASEB J. 15,2680-2688 (2001).

30. N. F. Bence, R. M. Sampat, R. R. Kopito, Impairment of the ubiquitin-pro-teasome system by protein aggregation. Science 292, 1552-1555 (2001).

31. P. A. Szweda, B. Friguet, L. I. Szweda, Proteolysis, free radicals, and aging.Free Radi. Biol. Med. 33, 29-36 (2002).

32. H. Snyder, K. Mensah, C. Theisler, J. Lee, A. Matouschek, B. Wolozin, Ag-gregated and monomeric alpha-synuclein bind to the S6′ proteasomal pro-tein and inhibit proteasomal function. J. Biol. Chem. 278, 11753-11759(2003).

33. J. B. Schulz, J. Dichgans, Molecular pathogenesis of movement disorders:are protein aggregates a common link in neuronal degeneration? Curr.Opin. Neurol. 12, 433-439 (1999).

34. E. Leroy, R. Boyer, G. Auburger, B. Leube, G. Ulm, E. Mezey, G. Harta, M.J. Brownstein, S. Jonnalagada, T. Chernova, A. Dehejia, C. Lavedan, T.Gasser, P. J. Steinbach, K. D. Wilkinson, M. H. Polymeropoulos, The ubiqui-tin pathway in Parkinson’s Disease. Nature 395, 451-452 (1998).

35. T. Kitada, S. Asakawa, N. Hattori, H. Matsumine, Y. Yamamura, S. Minoshi-ma, M. Yokochi, Y. Mizuno, N. Shimizu, Mutations in the parkin gene causeautosomal recessive juvenile parkinsonism. Nature 392, 605-608 (1998).

36. M. F. Perutz, Glutamine repeats and neurodegenerative diseases: molecu-lar aspects. Trends Biochem. Sci. 24, 58-63 (1999).

37. J. E. Galvin, V. M. Lee, J. Q. Trojanowski, Synucleinopathies: clinical andpathological implications. Arch. Neurol. 58, 186-190 (2001).

38. P. N. Leigh, H. Whitwell, O. Garofalo, J. Buller, M. Swash, J. E. Martin, J. M.Gallo, R. O. Weller, B. H. Anderton, Ubiquitin-immunoreactive intraneuronalinclusions in amyotrophic lateral sclerosis. Morphology, distribution, andspecificity, Brain 114 (Pt. 2), 775-788 (1991).

39. C. B. Lucking, A. Durr, V. Bonifati, J. Vaughan, G. De Michele, T. Gasser, B.S. Harhangi, G. Meco, P. Denefle, N. W. Wood, Y. Agid, A. Brice, Associa-tion between early-onset Parkinson’s disease and mutations in the parkingene. N. Engl. J. Med. 342, 1560-1567 (2000).

40. M. Periquet, M. Latouche, E. Lohmann, N. Rawal, G. De Michele, S. Ricard,H. Teive, V. Fraix, M. Vidailhet, D. Nicholl, P. Barone, N. W. Wood, S. Raskin,J. F. Deleuze, Y. Agid, A. Durr, A. Brice, Parkin mutations are frequent in pa-tients with isolated early-onset parkinsonism. Brain 126, 1271-1278 (2003).

41. S. A. Oliveira, W. K. Scott, E. R. Martin, M. A. Nance, R. L. Watts, J. P. Hub-ble, W. C. Koller, R. Pahwa, M. B. Stern, B. C. Hiner, W. G. Ondo, F. H. Allen,Jr., B. L. Scott, C. G. Goetz, G. W. Small, F. Mastaglia, J .M. Stajich, F.Zhang, M. W. Booze, M. P. Winn, L. T. Middleton, J. L. Haines, M. A. Peri-cak-Vance, J. M. Vance, Parkin mutations and susceptibility alleles in late-onset Parkinson’s disease. Ann. Neurol. 53, 624-629 (2003).

42. D. W. Dickson, A. Wertkin, Y. Kress, H. Ksiezak-Reding, S. H. Yen, Ubiquitinimmunoreactive structures in normal human brains. Distribution and devel-opmental aspects. Lab. Invest. 63, 87-99 (1990).

43. H. Fujigasaki, T. Uchihara, J. Takahashi, H. Matsushita, A. Nakamura, S. Koy-ano, K. Iwabuchi, S. Hirai, H. Mizusawa, Preferential recruitment of ataxin-3 in-dependent of expanded polyglutamine: an immunohistochemical study onMarinesco bodies. J. Neurol. Neurosurg. Psychiatry 71, 518-520 (2001).

44. K. Oyanagi, H. Ogawa, T. Nakajima, Rod-like intracytoplasmic inclusions inlarge neurons of the caudate nucleus: frequent appearance in myotonicdystrophy. Clin. Neuropathol. 13, 134-138 (1994).

45. T. Kawashima, A. Furuta, K. Doh-ura, H. Kikuchi, T. Iwaki, Ubiquitin-im-munoreactive skein-like inclusions in the neostriatum are not restricted toamyotrophic lateral sclerosis, but are rather aging-related structures. ActaNeuropathol. (Berlin) 100, 43-49 (2000).

46. J. R. Mendell, W .R. Markesbery, Neuronal intracytoplasmic hyaline inclu-sions. J. Neuropathol. Exp. Neurol. 30, 233-239 (1971).

47. F. Seitelberger, in Extrapyramidal Disorders, P. Vinken, G. Bruyn, H.Klawans, Eds. (Elsevier, Amsterdam, 1986), pp. 391-415.

48. D. W. Dickson, H. A. Crystal, L. A. Mattiace, D. M. Masur, A. D. Blau, P.Davies, S. H. Yen, M. K. Aronson, Identification of normal and pathologicalaging in prospectively studied nondemented elderly humans. Neurobiol.Aging 13, 179-189 (1992).

49. R. J. Mayer, M. Landon, J. Lowe, in Ubiquitin and the Biology of the Cell, J.-M.Peters, J. R. Harris, D. Finley, Eds. (Plenum, New York, 1998), pp. 429-462.

R E V I E W


Documents

Ubiquitin, Proteasomes, and the Aging Brainsageke.sciencemag.org/cgi/reprint/2003/34/re6.pdf · Schematic representation of ubiquitin-mediated proteolysis.After the generation of