&p.1:Abstract Ubiquitin-mediated proteolysis is involved inthe turnover of many short-lived regulatory proteins.This pathway leads to the covalent attachment of one ormore multiubiquitin chains to target substrates which arethen degraded by the 26S multicatalytic proteasomecomplex. Multiple classes of regulatory enzymes havebeen identified that mediate either ubiquitin conjugationor ubiquitin deconjugation from target substrates. Timeddestruction of cellular regulators by the ubiquitin-protea-some pathway plays a critical role in ensuring normalcellular processes. This review provides multiple exam-ples of key growth regulatory proteins whose levels areregulated by ubiquitin-mediated proteolysis. Pharmaco-logical intervention which alters the half-lives of thesecellular proteins may have wide therapeutic potential.Specifically, prevention of p53 ubiquitination (and sub-sequent degradation) in human papilloma virus positivetumors, and perhaps all tumors retaining wild-type p53but lacking the retinoblastoma gene function, should leadto programmed cell death. Specific inhibitors of p27 andcyclin B ubiquitination are predicted to be potent anti-proliferative agents. Inhibitors of IκB ubiquitinationshould prevent NFκB activation and may have utility in avariety of autoimmune and inflammatory conditions. Fi-nally, we present a case for deubiquitination enzymes asnovel, potential drug targets.
&kwd:Key words Cancer · Inflammation · Autoimmunity ·Cancer therapy · Antiproliferative agents · Ubiquitination
The depth of current knowledge about the molecularmechanisms regulating cellular proteolysis, combinedwith the understanding of how impairment of such pro-cesses underlies certain pathological conditions, hasopened the way for a mechanism-based approach for thedevelopment of new drugs. We outline sequentially a va-riety of proteins whose intracellular levels have beendemonstrated to be controlled at least in part by theubiquitin-dependent proteolytic pathway, and where in-tervention in the ubiquitin transfer process could be ex-pected to have some therapeutic value in different clini-cal settings. Additionally we discuss the possibility of in-tervening in the ubiquitin transfer process by inhibitingspecific ubiquitin proteases that have been implicated asdominant oncogenes: in such cases knowledge of the rel-evant target proteins is lacking. Finally we briefly dis-cuss the possibility of exploiting enzymes in this path-way for the development of novel antifungal agents.
The multiubiquitinated proteins generated by thisubiquitin transfer cascade are degraded by the multicata-lytic 26S proteasome (see [105] for a recent review). Thisstep in the pathway affords little molecular specificity asall ubiquitinated proteins are degraded here. We feel thatpharmacological targeting of the proteasome is a less at-tractive strategy than targeting of the conjugating and de-ubiquitinating enzymes for this reason. However, there areproteasome specific pharmacological agents, such as pep-tide aldehydes [91, 102] and the antibiotic lactacystin [27]which are proving invaluable as research tools for definingnovel targets of the ubiquitin-proteasome pathway.
The ubiquitin pathway
Selective degradation of intracellular proteins is now rec-ognized as a major pathway for controlling the activity
M. Rolfe · M.I. Chiu · M. Pagano (✉)1
Mitotix Inc., One Kendall Square, Bldg. 600 Cambridge,MA 02139, USA
Present address:1 Department of Pathology, MSB 563, New York University,Medical Center, 560 First Ave., New York, NY 10016, USA&/fn-block:
&roles:Mark Rolfe · M. Isabel Chiu · Michele Pagano
The ubiquitin-mediated proteolytic pathway as a therapeutic area
&misc:Received: 14 May 1996 / Accepted: 11 July 1996
of a variety of cellular regulatory proteins (see [13, 18,46, 52, 55] for recent reviews). Because of the topic ofthis review we discuss specifically the ubiquitin pathwayof higher eukaryotes.
Within the past several years compelling evidence hasbeen presented that implicates ubiquitination in the turn-over of the tumor suppressor protein, p53 [49, 104,108–111], the cell cycle regulators cyclin A, cyclin B [34,45, 59, 121, 128], and p27 [90], the kinase c-Mos [86],the cystic fibrosis transmembrane conductance regulator[54, 130], the DNA repair protein O6-metthylguanine-DNA methyltransferase [116], the transcriptional coacti-vator p300 [3], the transcription factors c-jun, [126] c-fos[117, 127], and IκB/NFκB [89, 91, 112] (see also Ta-ble 1). Preliminary data suggest that ubiquitination mayalso be involved in the turnover of mammalian G1 regula-tory cyclins (J. Roberts, personal communication), thetranscription factors c-myc [14], DP1 [75], and E2F (R.Bernards and D. Livingston, personal communication),the regulatory subunit of cAMP-dependent protein kinase[41], the receptors for peptide growth factors [31, 82], theestradiol receptor [85], and the oncoprotein E1A ([14]and R. Bernards, personal communication).
The conjugation of ubiquitin to protein substrates is amultistep process (Fig. 1). In an initial ATP-requiring
step a thioester is formed between the C-terminus ofubiquitin and an internal cysteine residue in a ubiquitinactivating enzyme, E1. The activated ubiquitin is thentransferred to a specific cysteine on one of severalubiquitin-conjugating enzymes – known as Ubcs or E2s,with 11 so far identified in humans: Ubc2/Rad6 [60],Ubc3/Cdc34 [97], Ubc4/Ubc5B [53, 104], Ubc5/Ubc5A[53, 110], Ubc5C [53], Ubc6 [88], Ubc7 [88, 101], Ubc8[56], Ubc9 [61, 131], Ubc-epi [69], and Ubc-bendless(accession no. D83004). Ubcs share a common cysteineresidue used in the thioester bond formation with ubiqui-tin. Figure 2 shows the alignment of the human Ubc cat-alytic domains spanning the active cysteine. Finally,these Ubcs donate ubiquitin to protein substrates direct-ing their degradation by the multisubunit protease com-plex called the proteasome (see [105] for a review). Sub-strates are recognized either directly by a Ubc or by as-sociated recognition proteins, also known as the E3 pro-teins or ubiquitin ligases. Ubiquitin ligases are relativelypoorly characterized, but some of them have been shownto act enzymatically as the last member in the cascade ofthioester containing enzymes, ultimately donatingubiquitin to the protein that is destined for degradation.
It should be mentioned that ubiquitination not only di-rects protein degradation but in some cases seems to
6
Table 1 Cellular regulators that are substrates of the ubiquitin pathway in higher eukaryotes&/tbl.c:&tbl.b:
Substrate Function Ubiquitinating enzymes References
have a regulatory function. For example, ubiquitinationof IκBα kinase is a requirement for its activity [11].
Recently another class of enzymes in the ubiquitinpathway has begun to command attention. These en-zymes are known as ubiquitin C-terminal hydrolases(Uchs), deubiquitinating enzymes, isopeptidases orubiquitin-specific proteases. They are thiol proteases thatrecognize and hydrolyze the isopeptide bond betweenthe α-carboxy terminal glycine (G76) of ubiquitin andthe ε-amino group of a lysine residue in the attached pro-tein substrate, thereby removing the ubiquitin moietyfrom the protein. In the most simplistic sense their func-tions oppose the combined action of the conjugation cas-cade of E1, E2, and E3. Eukaryotic thiol proteases con-tain an active site cysteine. Generally catalysis of thiolproteases proceeds through a thioester intermediate andis facilitated by a nearby histidine side chain. In the caseof Uchs the thioester intermediate is formed with ubiqui-tin. Although there are two distinct families of Uchswhose members share no sequence similarity with oneanother, within each family there are conserved signaturesequences that include an active site cysteine and con-served histidine residues. The first family, Uch class I[134], comprises a small number of enzymes and areabout 30 kDa in size. The second family, Uch class II,more commonly known as Ubps (ubiquitin-specific pro-teases; see [46] for a review), has already added an in-triguing dimension to the field of ubiquitin-mediatedproteolysis. As a group the Ubps are larger than Uch
class I enzymes (about 60–120 kDa) and are postulatedto fulfill multiple physiological roles (Table 1). In princi-ple, depending on the specific substrate of the particularUbp enzyme, its activity may have stimulatory or inhibi-tory effects on ubiquitin-mediated proteolysis. For in-stance, the Ubps that cleave ubiquitin from either shortpeptides generated in the proteasome or branched poly-ubiquitin chains are contributing to the intracellular poolof free ubiquitin, thereby increasing the rate of ubiquitin-mediated proteolysis in general. On the other hand, Ubpenzymes that cleave ubiquitin chains from specific targetsubstrates are theoretically lowering the rate of proteoly-sis for such substrates.
The question of specificity
There is a high degree of specificity in the ubiquitintransfer cascade the molecular details of which are cur-rently being worked out. Using primary sequence infor-mation from yeast, plants, flies, worms, and humans onecan align all the known Ubcs to identify important con-served domains. Additionally, the X-ray crystal struc-tures of two Ubcs have been published [15, 16]. If onesuperimposes the conserved residues onto the surface ofthe known structure (e.g., of yeast Ubc4), one finds ahighly conserved surface patch which presumably inter-acts with ubiquitin in the context of an E1-thioester (thisis an interaction which all Ubcs need to perform). Non-
7
Fig. 1 Schematic view of theubiquitin-proteasome pathway.In an ATP-dependent reaction,a molecule of ubiquitin isactivated at its C-terminus viaa thioester formation with E1.The ubiquitin is then trans-ferred to an E2 enzyme, againvia a thioester bond. Finally,the ubiquitin is transferred ontoa substrate, with or withoutthe help of an E3, via an iso-peptide bond. Substrates withpolyubiquitin chains are target-ed for degradation to the 26Sproteasome complex. Ubpscounteract the action of E2sand E3s. See text for details&/fig.c:
conserved surface patches on the Ubc are presumed tointeract with distinct ubiquitin ligases or directly withsubstrates.
Ubiquitin ligases are much less well characterizedthan Ubcs. However, one molecularly characterized E3has been extensively studied: the complex of human pap-illoma virus (HPV) E6 and E6AP (see later) involved inthe ubiquitination of p53 in virally infected cells [104,111]. E6AP is the archetypal member of a new family ofE3s [50] which appear to act catalytically in the ubiqui-tin transfer cascade; these enzymes are large (usually>100 kDa) and share limited homology with each otherat their C-termini. They have been named “hect” proteins(for homology to E6AP c-terminus). There are at least 12new human genes in currently available databases whichwould qualify as encoding hect proteins. Figure 3 showsthe alignment of the catalytic domain of human Hect
proteins. The fact that E6AP is specifically ubiquitinatedby Ubc4 and Ubc5 [104, 111] lends support to the notionthat different hect proteins may be ubiquitinated by dis-tinct Ubcs. On the other hand, Ubc4 is also involved inthe ubiquitination of cyclin B but for this substrate an-other E3, the complex of Cdc16/Cdc23/Cdc27, has beenimplicated (see below). There are therefore numerouspotential combinations of Ubc and E3 which furnish ahigh degree of specificity to these ubiquitination reac-tions.
The recent explosion of sequence information fromthe various organism-based genome efforts indicates thatthe class II Uch enzyme family, more commonly knownas Ubps, is remarkably large. A search in the combineddatabases readily identified upwards of 40 fully se-quenced family members, of which at least 9 are human,and at least 15 from the yeast Saccharomyces cerevisiae
8
Fig. 2 Alignment of catalyticdomain of human Ubcs.Sequences from ten humanUbcs from amino acid approx.30 to approx. 100 including thecatalytic cysteine residue areshown. Black, residues presentin at least four entries. Thealignment was performed usingthe Clustal method from Meg-Align software (DNASTAR).Ubc2 (M74525); Ubc3(L22005); Ubc4/Ubc5B(L40146); Ubc5/Ubc5A(X78140); Ubc6 (X92962);Ubc7 (X92963); Ubc8(Z29331); Ubc9 (D45050);Ubc-epi (M91670);Ubc-bendless (D83004)&/fig.c:
alone. At least twice as many are estimated from the ex-pressed sequence tag databases of human sequences.These high numbers raise the exciting possibility that, aswith E3s, Ubps may possess a considerable degree ofsubstrate specificity. As with other components of theubiquitin-proteolysis machinery, genes encoding Ubpsare found in yeasts, plants, flies, worms, and mammals.Accordingly, all family members share catalytic domainsencompassing an active site cysteine, and a conservedcarboxy-terminal His box sequence. Figure 4 shows thealignment of these conserved domains of human Ubps.
Another degree of control and specificity can be exer-cised at the level of phosphorylation. Indeed, there is ac-cumulating evidence that phosphorylation events are of-ten the trigger that leads to recognition by the ubiquitintransfer machinery and subsequent degradation (re-viewed by [17]). In addition to phosphorylation, there isan accumulating body of evidence suggesting that dis-crete degradation signals exist which may target a pro-
tein for degradation both in cis and, somewhat surpris-ingly, in trans. For instance, a stretch of nine amino ac-ids (RxxLxxIxN) followed by a lysine-rich domain in theN-terminal of cyclin B defines a “destruction box”whose deletion or mutation substantially stabilizes cyclinB [34, 44, 71]. Overexpression of the destruction box do-main can also inhibit cyclin B proteolysis [47]. Similarly,in c-Jun a 27 amino acid element named δ-domain hasbeen shown to be important for its ubiquitin-mediateddegradation [126]. Recently yet another novel signal hasbeen identified: a glycine-rich region which serves as a“stop signal” in the processing of NF-κB p105 leading tolimited proteolysis and release of p50 [66]. Stretches ofamino acids with high proline, glutamic acid, serine, andthreonine content (PEST sequences) have also been pos-tulated as signals which identify short lived proteins[103]. In addition to these cis-acting elements encodedwithin the target proteins themselves, there are examplesof one protein in a heterodimer pair targeting the second
9
Fig. 3 Alignment of catalyticdomain of human Hect pro-teins. A region of 70 aminoacids surrounding the catalyticcysteine was aligned using theClustal method from MegAlignsoftware (DNASTAR). Black,residues present in at least fourentries. E6AP (L07557); KG1C(D28476); ORFK (D25215);RSC (13635); 100 K (D45459);H23082 (H23082); URE-B1(R18042); RSP5 (T93069);NEDD4 (T74302); K1AAN(D42055); R20250 (R20250)&/fig.c:
protein of the heterodimer, in trans, for degradation. Thebest studied example of this is fos/jun where phosphory-lation of c-Jun subunit targets c-Fos for degradation [94].
p53
The p53 protein was discovered by virtue of its tight as-sociation with SV-40 large T antigen in SV-40 trans-formed cells [62, 68]. The normal cellular role of p53appears to be either in imposing a block in cell cycleprogression at G1 or in inducing apoptosis following ra-diation or chemically induced DNA damage (see [63,100] for recent reviews). The G1 arrest is thought to bemodulated in large part by p53 inducing expression ofthe cyclin-dependent kinase inhibitor p21 [22, 24]. Ap-optosis may be mediated by induction of the pro-apo-ptotic gene Bax [79, 80] and coincident repression of theanti-apoptotic gene Bcl2 [12]. Most of the effects of p53are thought to be mediated by this ability to modulatetranscription both positively and negatively. The geneproduct is completely dispensable for normal mouse de-velopment but does play a role in tumor suppressionsince p53 knockout mice succumb to a variety of tumorsearly in life [20]. The gene for p53 is mutated or deleted
in approx. 40% of a wide variety of human tumors, im-plicating p53 as the most significant tumor suppressorgene to date [6, 40].
Other DNA tumor viruses (in addition to SV-40) en-code proteins whose function appears to be abrogation ofp53 activity. Adenoviruses encode two E1b proteins, a55-kDa species which binds to and inactivates p53 [107]and a 19-kDa species which does not contact p53 direct-ly but can repress p53-mediated apoptosis in infectedcells [39, 106]. SV40 and adenovirus infection lead tohigh levels of inactivated p53 in infected cells. HPVs en-code a protein, E6, whose function is to abrogate p53 ac-tivity [132] and HPV-infected cells have vanishingly lowlevels of p53 protein [78].
That p53 is degraded in a ubiquitin-dependent reac-tion was demonstrated in studies on the mode of actionof the E6 viral oncoprotein encoded by the high-riskHPVs 16 and 18. These high-risk types of HPV arethought to be causally involved in the pathogenesis ofanogenital cancer, particularly in cancer of the cervix(reviewed by [26, 81]). The evidence includes the fol-lowing: (a) viral DNA is found in close to 90% of suchtumors; (b) most of the virus-positive tumors contain in-tegrated viral DNA; (c) the majority, if not all, of HPV-positive cancer biopsies and all HPV-containing celllines derived from individuals with cervical cancer revealspecific transcripts originating uniformly from two spe-cific open reading frames (E6 and E7) of the persistingHPV DNA. Furthermore, the E6 and E7 genes of high-risk types of HPV (but not of low-risk types) immortal-ize human foreskin, cervical keratinocytes, or epithelialbreast cells after in vitro transfection in tissue culture.Specific cellular proteins bind efficiently to the HPV E6and E7 oncoproteins of high-risk types of HPV. The E7
10
Fig. 4 Alignment of conserved domains of human Ubps. Align-ment of nine human Ubps showing the conserved cysteine (Cys)and histidine (His) domains. The name and position of each se-quence are also shown. Black, residues present in at least four en-tries. The alignment was performed using the Clustal method fromMegAlign software (DNASTAR). Tre213, orf2 (P35125); Unph(U20657); UnpX (P40818); TGT (U30888); Dub1 (A. D’Andrea,personal commun); Hsan (D38378); KIAA0190 (D80012); Iso-peptidase T (U47924), Uhx1 (GenPept U448839)
oncoprotein binds the protein product of the retinoblasto-ma gene, Rb [23], whereas the E6 protein binds to thep53 protein, promoting its ubiquitin-dependent degrada-tion [108].
We and others have identified human Ubc4 and Ubc5(93% identical at the protein level) as the ubiquitin con-jugating enzymes involved in the HPV E6-mediated ubi-quitination of p53 [104, 111]. However, Ubc4 and Ubc5do not directly ubiquitinate p53; rather, they are the cen-tral enzyme in a cascade of three thioester-forming en-zymes that begins with E1 and ends with E6AP. Thecomplex of ubiquitinated E6AP and E6 specifically me-diates transfer of ubiquitin onto p53. We have demon-strated using microinjection technology that the enzymesinvolved in ubiquitin transfer in vitro can be inhibited invivo to reverse E6-mediated p53 degradation.
In noninfected cells p53 seems to be ubiquitinatedand degraded by the proteasome [76]. It is not clear yetwhether E6AP is involved in the turnover of wild-typep53 in noninfected cells.
If a specific small-molecule inhibitor of Ubc-mediat-ed conjugation of ubiquitin to p53 could be identified,p53 proteolysis should be halted and intracellular levelsof p53 should be elevated. Such increased levels of p53should halt cellular growth and might also lead to pro-grammed cell death or apoptosis in cells with the appro-priate genetic background. It is clear now from severalpublished studies that increased levels of p53 can lead toapoptosis in many transformed cells if the Rb pathway isinactivated (see [133] for a review). This inactivation canbe a direct result of Rb gene inactivation or may involveother entities in the Rb pathway such as p16, cyclin D1,or E2F. Thus, inhibition of p53 degradation by a small-molecule inhibitor is expected to induce apoptosis, notonly in HPV16/18 infected tumors but in all tumors lack-ing Rb function and retaining wild-type p53.
p27
The eukaryotic cell cycle is regulated by a family of ser-ine/threonine protein kinases called cyclin-dependentkinases (Cdks) because their activity requires the associ-ation with regulatory subunits named cyclins (reviewedin [64, 74]). Cdks can also associate with inhibitory sub-units. So far, based on their sequence homology, twofamilies of these Cdk inhibitors (Ckis) have been identi-fied in mammalian cells: the Cip/Kip family, which in-cludes p21, p27, and p57, and the Ink family, which in-cludes p15, p16, p18, and p20 (see [115] for a review).
p27 is a powerful in vitro inhibitor of both Cdk4 andCdk2, two Cdks involved in the regulation of the G1phase of the cell cycle. As predicted, overexpression ofp27 arrests the cell division cycle in G1 [98, 125]. Com-pared to growing cells, human p27 levels are elevated inquiescent fibroblasts [42] and in resting primary T-lym-phocytes [29, 87]. In contrast, the levels of p27 mRNAare not different between quiescent and growing cells[42]. Indeed, p27 accumulation in quiescent cells is at
least partially due to a dramatic increase in the p27 half-life, which is six to eight times longer in quiescent cellsthan in growing cell [42, 90]. This difference can ac-count for the different p27 abundance in quiescent versusproliferating cells [29, 87, 90]. Finally, p27 containsPEST-like sequences in its carboxyl terminus (PEST-FIND score +10).
Based on these data we investigated the mechanismsthat control p27 degradation [90]. We found that in vivoinhibition of the chymotryptic active site of the protea-some with carbobenzoxyl-leucinyl-leucinyl-leucinal-H(z-LLL) induces an accumulation of p27 protein in bothtransformed and normal human cells and that, because oflack of proteasome activity, ubiquitinated forms of p27accumulate. Using cellular extracts as a source of ubiqui-tinating enzymes and proteasome subunits, we showedthat p27 is polyubiquitinated and degraded by the protea-some in vitro, and that human Ubc2 and Ubc3, the ho-mologs of S. cerevisiaeRad6 and Cdc34 gene products,respectively, are specifically involved in the in vitro ubi-quitination of p27.
Recombinant purified Ubc2 and Ubc3 can mono-ubiquitinate p27 in the presence of ATP and E1, theubiquitin-activating enzyme [90]. In contrast, p27 ispolyubiquitinated by incubation with human cellular ly-sates. The single ubiquitination of p27 in a purified invitro system is probably due to the lack of an E3, theubiquitin ligase, and/or of posttranslation modifications.Pertinent to the latter, it has been shown that two S. cere-visiae cyclins, Cln2 [19] and Cln3 [137], are substratesof Ubc3, and that phosphorylation of these cyclins by theyeast cdk, Cdc28, promotes their ubiquitination. Degra-dation of other proteins including c-fos [94], IκBα [91],and Rag2 [65] was also found to be stimulated by theirphosphorylation. It is possible that a protein kinase par-ticipates, in concert with Ubc2, Ubc3, and possibly a yetto be discovered E3, in the ubiquitination of p27 in hu-man cells.
Finally, we found that, compared to proliferatingcells, quiescent cells contain a far lower amount of p27ubiquitinating activity [90], despite the fact that the twosubpopulations contain similar amounts of Ubc2 andUbc3 (S. Tam and M.P., unpublished results), indicatingthat Ubc2 and Ubc3 abundance is not the limiting factorin p27 ubiquitination. We are currently exploring thepossibility that the abundance of a third component, theputative E3 enzyme, and/or posttranslational modifica-tions of p27 and of the ubiquitinating enzymes may beimportant in controlling p27 ubiquitination.
Cki functions are often inactivated during transforma-tion. Indeed, p15 and p16 genes are often deleted, mutat-ed, or inactivated by methylation in a large series of hu-man tumors (reviewed by [114]). In tumors where p53activity is missing, p21 levels are low, and p21 is notfound in association with Cdks [136]. p27 is also a targetof oncogenic events since the adenoviral E1A [77] onco-protein inactivates p27, dissociating it from cyclin-Cdkcomplexes. However, the p27 gene has never been foundaltered in tumors [58, 95, 99, 118]. We looked at the p27
11
protein expression by immunohistochemistry in colorec-tal carcinomas and found that p27 is prognostic in thesetumors [70]. In fact, absent/low p27 expression corre-lates with approximately three- and sixfold increase inrisk of death, respectively. Tumors expressing low p27levels contain more p27 specific degradation activity,suggesting that p27 protein is eliminated because of anenhancement of its degradation.
A specific small molecule inhibitor of p27 ubiquitina-tion would be expected to raise the intracellular levels ofp27 and should lead to a block in proliferation and hencea reversal in disease progression, especially in those tu-mors that have low to undetectable p27 protein levels asa result of an “activated” ubiquitination pathway.
Cyclin B
Cyclins are required for the timed activation of Cdks.They are also thought to determine the interaction withupstream regulatory enzymes, the subcellular localiza-tion, and the substrate specificity of Cdks. In mammaliancells at least eight different cyclin types have been iden-tified, acting at specific stages of the cell cycle (see [64]for a review). Human cyclin B was identified as the acti-vating regulatory subunit of Cdc2 (Cdk1), whose activitypeaks in mitosis [33, 96]. Cyclin B destruction is re-quired for exit from mitosis and entry into interphase ofthe next cell cycle [21, 72, 73, 83]. Deletion mutants ofcyclin B, lacking the destruction box (see above), arrestthe cell cycle in mitosis and maintain both cyclinB–Cdc2 kinase and protease in their active states [32, 34,73, 83]. In Xenopus, destruction of cyclin B2 but not B1requires binding to Cdc2 [120, 129]. Degradation of cy-clin B is regulated by the ubiquitin pathway [34, 71].The activity of the Ubcs (Ubc4 and Ubcx/E2-C) involvedin cyclin B degradation is cell cycle independent [2, 34,45, 59, 121, 138]. In contrast, the cyclin B specific E3 (amultiprotein complex including Cdc16, Cdc23, andCdc27) is only active during mitosis [59, 121].
A specific low molecular weight inhibitor of ubiquitintransfer onto cyclin B targeting the Ubc or the E3 in-volved in this process would prevent cyclin B destructionand would be expected to be a very strong cytostaticagent.
IκB
The signal transduction pathways involved in B- and T-cell activation in numerous immune and inflammatoryresponses are well characterized. One central transcrip-tion factor in these pathways is NFκB (for recent re-views, see [4, 123]) which is synthesized as a 105-kDaprecursor that is processed to a 50-kDa mature form(p50) in an ATP-, ubiquitin-, proteasome-dependent re-action [89, 91]. The C-terminal part of the molecule isdegraded during this process, leaving the N-terminal ac-tive region intact. p50 can associate with several related
regulators, such as p65, RelB and c-Rel to generate dis-tinct heterodimeric transcriptional activators. The tran-scriptional inhibitor IκBα can associate with each ofthese heterodimers and in so doing forms an inactive het-erotrimeric complex which is cytosolic. A variety of in-flammatory agents (mitogens, cytokines, viral proteins,antigens, UV light, etc.) activate a signal transductionpathway that ultimately leads to a tightly coupled phos-phorylation and ubiquitin-dependent degradation ofIκBα [1, 7, 8, 10, 11, 28, 67, 91, 112] which unmasks anuclear localization signal on NFκB. Now the active het-erodimer can be translocated to the nucleus to activatepro-inflammatory genes. Interestingly, the IκBα kinaserequires the presence of E1, Ubc4 or Ubc5, and ubiquitinfor its activity [11]. This is the first example of ubiquiti-nation-dependent protein kinase activity. Thus, threeubiquitin-dependent proteolytic processes, the activationof IκBα kinase, the complete destruction of IκBα, andthe limited posttranslational processing of p105 to p50are involved in the activation of NFκB. The identities ofthe E3 molecule/s involved in IκBα ubiquitination aswell as the kinase responsible for signal induced phos-phorylation are currently unknown but are undoubtedlythe subject of intense investigation in many laboratories.
Intervention in this signal induced ubiquitination anddegradation pathway with low molecular weight pharma-ceutical agents promises to provide novel agents with ahigh degree of specificity (and hopefully fewer adverseside effects) which could be used in several pathologicalconditions, such as inflammation, autoimmune disease,and viral infection.
As the above examples illustrate, manipulating or inhib-iting members of the enzymatic cascade responsible forconjugating ubiquitin to target substrates is a promisingnew strategy for therapeutic intervention. It is increas-ingly apparent that inhibition of enzymes that remove,rather than add, ubiquitin from protein substrates mayalso become a valuable therapeutic approach.
The most intriguing possibility for the large numbersof Ubps whose functions are unknown lies in their poten-tial role as key regulators of stability for individual sub-strates. Ubiquitin chains that assemble on various pro-teins are highly dynamic, with rapid addition and remov-al of ubiquitin units. With certain well-known excep-tions, for example histone 2A [57], the general rule hasbeen that attachment of ubiquitin to a susceptible sub-strate commits it to degradation, whereas removal ofubiquitin has a stabilizing effect. Ubps that act at thisstep in a substrate-specific manner should play a key rolein controlling the stability of that substrate, but wouldnot affect proteolysis in general. The in vivo half-life ofa given substrate would be determined by the balance be-tween the rate of ubiquitin conjugation on the one hand,and that of ubiquitin removal on the other.
12
Several lines of evidence suggest that such regulatoryUbps exist. The first, mentioned above, is their largenumbers, suggestive of a range of catalytic specificities.Second, the sequence similarity of the various Ubpsfrom a given species is confined to a few conserved mo-tifs including the active site cysteine and the histidinebox [93, 135]. Outside of these motifs the sequences aredissimilar, except of course among members of a sub-family, consistent with the notion that these enzymes arenonoverlapping both in their substrate specificities andin their physiological functions. Third, the fact that iso-peptidase T (see Table 2) is exquisitely sensitive to thechemical nature of the carboxyl-terminus of the ubiqui-tin chain (i.e., that it be free and unanchored [135]) atthe same end that is used in substrate attachment sug-gests that other members of this family may have built-in specificities for different substrates as well. Last, al-though the physiological substrates of these Ubps havenot yet been identified, it seems reasonable to assumethat Ubps effect their biological function as a conse-quence of their enzymatic activity on their cognate sub-strates (see Table 2). Mutation analyses of the firsthandful of Ubps suggest that Ubp substrates likely com-prise a diverse group, ranging from putative cellularproto-oncogenes (Unp [36, 37]) to tumor suppressors(Tre [84]), to extracellular factors that are essential forembryogenesis and cell fate (Faf, [30, 48]) and cyto-kine-inducible proteins with growth-suppressing activity(Dub1 [139]).
Most importantly, perturbation of different Ubps inhigher eukaryotes can lead to distinct and profoundphysiological alterations, consistent with the notion thatregulatory Ubps likely control the stability of proteinswhich themselves play pivotal regulatory roles in the cell
and could therefore offer the possibility of target-specifictherapeutic intervention.
In summary, we anticipate that unlike those Ubpswhich play a general role in ubiquitin-mediated proteoly-sis, the subset of Ubps which regulate the stability ofspecific substrates provide an ideal avenue for therapeu-tic intervention that conceptually complements manipu-lation of the E2 and E3 pathway. For a given target pro-tein that is subject to ubiquitin-mediated regulation, de-pending on its cellular function, one could choose to in-crease (e.g., tumor suppressor) or decrease (e.g., onco-gene) the cellular levels of the protein by designing smallmolecules that inhibit either the ubiquitinating or the de-ubiquitinating enzyme for that particular substrate. Al-though much remains to be discovered, the future for thisexciting area of research in pharmaceutical developmentseems promising indeed.
Antifungals
The emerging picture is that E2s, E3s, and Ubps com-prise three superfamilies whose members are found ineukaryotes as diverse as yeasts, plants, insects and mam-mals. The diverse members of each family, particularlythe E3s and Ubps, suggest that both the addition ofubiquitin to and the removal from target substrates areregulated by a certain degree of catalytic specificity.How much of this specificity is conserved across species,and how many family members encode redundant func-tions, remain to be seen. As more is known about the de-tails in each organism, it might be possible to targetmembers of the ubiquitin-mediated proteolysis pathwayfor antifungal therapy. The putative range of biological
13
Table 2 Examples of Ubps and their physiological functions&/tbl.c:&tbl.b:
Ubp References Biological functions
hs Tre [84] Originally cloned from genome of Ewing’s sarcoma cells as a fusion of sequences derived from chromo-somes 5-18-17, where Tre is truncated; expression of truncated version transforms nude mice, but it has nodeubiquitinating activity and is thought to act in a dominant negative manner over full length version.
hs Unp [35–37] Originally found during a survey of mouse genes near the Mpv 20 retroviral insertion site. Mouse Unpoverexpression transforms nude mice; human Unp maps to chromosome 3p21.3, a locus often rearrangedin lung cancers; human Unp mRNA overexpressed in small cell lung carcinoma; Unp protein has deubiquiti-nating activity; mutant version with active site cysteine mutated to alanine no longer transforms cells(R. Baker, pers. comm.).
hs Dub1 [139] Clone isolated as an interleukin-3 inducible immediate-early gene, by differential display; has deubiquiti-nating activity which is abolished when active site cysteine is changed to serine. Overexpression of Dub1induced growth arrest in G1 phase of the cell cycle in interleukin-3 responsive cells, but not in 3T3 cells.
hs Isopeptidase T [25, 38, First purified as a ubiquitin-binding protein from bovine reticulocytes by adsorption to a ubiquitin affinity119, 135] column; human version cloned by homology; disassembles branched polyubiquitin chains (glycine 76–lysine
48 linked) by a sequential exo mechanism, starting a the proximal end, which contains a free carboxyl-terminus. Enzyme has exquisite substrate specificity; activity is greatly reduced if the carboxyl-end of theubiquitin chain is modified or attached to another protein.
hs Uhx1 [122] cDNA clone isolated from a retina library. The 3.3-kb transcript is expressed in all tissues tested, but is aboutfivefold more abundant in the retina. UHX1 maps to human chromosome Xp21.2-p11.2, the site of severalinherited X-linked retinal disorders.
dm Faf [30, 48] Gene identified in a screen for mutations that affect eye development. Regulator of cell fate in eyedevelopment and o ogenesis; flies bearing mutant faf develop extra photoreceptors in their ocular facets;Faf is required outside of the photoreceptor cell. Mutations in the active site cysteine affect Faf function.
&/tbl.b:
specificity accorded to E3s and Ubps because of their di-verse structures outside of the catalytic domains couldlikewise be exploited in the design of small moleculesthat can subtly differentiate enzyme homologs from dif-ferent species. A compound that differentially (or exclu-sively) affects the function of a given E3 or Ubp inpathogenic organisms such as Candida albicansoverthat of its human homolog could be a potent and specificantifungal agent which is also not toxic to host cells.
As with other superfamilies, one concern in the de-sign of drugs that target one member of the family is thepossibility of redundancy. In the pathogenic fungus ofinterest, can other enzymes of the family functionally re-place the targeted enzyme? The specific biology proba-bly varies from case to case, but there is good evidencefrom both budding and fission yeasts that although someenzymes in the ubiquitin-mediated proteolysis pathwayhave overlapping functions (Ubp1, 2, 3 [5, 124]), certainothers are functionally irreplaceable. Temperature-sensi-tive mutations in Ubc9 (a member of the E2 family) leadto growth arrest at the nonpermissive temperature [113],while deletion or mutations in Doa4 (a Ubp) result inmultiple defects [93], ranging from slow growth to a de-fect in DNA repair in response to UV irradiation. Tem-perature-sensitive mutations in Cdc16, 23, or 27 (compo-nents of the E3 for cyclin B degradation) each lead to le-thality at the restrictive temperature [51], while muta-tions in certain proteasome subunits lead to cell-cycle ar-rest at the G2/M phase. It seems reasonable therefore thatjudicious selection of components in this pathway couldlead to identification of effective antifungal candidatetargets and subsequent pharmaceutical development.
Conclusions
Although still partially speculative, the recent discoveriesin the area of ubiquitination have opened new possibilitiesin the treatment of some proliferative and autoimmunediseases and of inflammation. In fact, the high specificityof the ubiquitin pathway might allow the development ofnew classes of highly potent and selective low molecularweight enzyme inhibitors targeting particular members ofthe ubiquitin pathway that control the intracellular levelsof a wide range of important regulatory proteins.
&p.2:Acknowledgments We thank R. Baker, R. Bernards, D. Boh-oman, A. Ciechanover, A. D’Andrea, D. Finley, N. Heintz, P.Howley, E. Kipreos, D. Livingston, J. Magae, C. Maki, A. Musti,and J. Roberts for communicating data before publication; J. Eck-stein for discussion on the three-dimensional structure of Ubcs;and A. Ciechanover, A. D’Andrea, A. Hershko, and P. Worland forcritically reading the manuscript.
References
1. Alkalay K, Yaron A, Hatzubai A, Orian A, Ciechanover A,Ben-Neriah Y (1995) Stimulation-dependent IκBa phosphor-ylation marks the NFκB inhibitor for degradation via theubiquitin-proteasome pathway. Proc Natl Acad Sci USA 92:10599–10603
2. Aristarkhov A, Eytan E, Moghe A, Adom A, Hershko A,Ruderman J (1996) E2-C, a cyclin-selective ubiquitin carrierprotein required for the destruction of mitotic cyclins. ProcNatl Acad Sci 93:4294–4299
3. Avantaggiati M, Carbone M, Graessman A, Nakatani Y,Howard B, Levine A (1996) The SV40 large T antigen and ad-enovirus E1A oncoproteins interact with distinct isoforms ofthe transcriptional co-activator p300. EMBO J 15:2236–2248
4. Baeuerle P, Henkel T (1994) Function and activation ofNfKB in the immunosystem. Annu Rev Immunol 12:141–179
5. Baker R, Tobias JW, Varshavsky A (1992) Ubiquitin-specificproteases of Saccharomyces cerevisiae. Cloning of UBP2 andUBP3, and functional analysis of the UBP gene family. J BiolChem 267:23364–23375
6. Berns A (1994) Is p53 the only real tumor suppressor gene?Curr Biol 4:137–139
7. Britta-Mareen E, Heike T, Pahl H, Henkel T, Schmidt K,Wilk S, Baeuerle P (1995) Phosphorylation of human IκB-aon serine 32 and 36 controls proteolysis and NFκB activationin response to diverse stimuli. EMBO J 14:2876–2883
8. Brown K, Gersterberg S, Carlson S, Franzoso G, Siebenlist U(1995) Control of IκB-a proteolysis by site-specific, signal-induced phosphorylation. Science 267:1485–1487
9. Cenciarelli C, Hou D, Rellahan B, Smith H, Fried V, Weiss-man A (1992) Activation-induced ubiquitination of the T cellantigen receptor. Science 257:795–797
10. Chen Z, Hagler J, Palombella V, Melandri F, Scherer D, Bal-lard D, Maniatis T (1995) Signal-induced site specific phos-phorylation targets IκBα to the ubiquitin-proteasome path-way. Genes Dev 9:1586–1597
11. Chen Z, Parent L, Maniatis T (1996) Site-specific phosphory-lation of IκBα by a novel ubiquitination-dependent proteinkinase activity. Cell 84:853–862
12. Chiou S, Rao L, White E (1994) Bcl-2 blocks p53-dependentapoptosis. Mol Cell Biol 14:2556–2563
13. Ciechanover A (1994) The ubiquitin-proteasome proteolyticpathway. Cell 79:13–21
14. Ciechanover A, DiGiuseppe J, Bercovich B, Orian A, RichterJ, Schwartz A, Brodeur G (1991) Degradation of nuclearoncoproteins by the ubiquitin system in vitro. Proc Natl AcadSci USA 88:139–143
15. Cook W, Jeffrey L, Sullivan M, Vierstra R (1992) Three-di-mensional structure of a ubiquitin-conjugating enzyme (E2).J Biol Chem 267:15116–15121
16. Cook W, Jeffrey L, Xu Y, Chau V (1993) Tertiary structures ofclass I ubiquitin-conjugating enzymes are highly conserved:crystal structure of yeast Ubc4. Biochemistry 32:13809–17
17. Deshaies RJ (1995) Make it or break it: the role of ubiquitin-dependent proteolysis in cell regulation. Trends Cell Biol 10:428–434
18. Deshaies RJ (1995) The self-destructive personality of a cellcycle in transition. Curr Opin Cell Biol 7:781–789
19. Deshaies RJ, Chau V, Kirschner MW (1995) Ubiquitinationof the G1 cyclin Cln2p by a Cdc34p-dependent pathway.EMBO J 14:303–312
20. Donehower LA, Harvey M, Slagle BL, McArthur MJ, Mont-gomery CA, Butel J, Bradley A (1992) Mice deficient for p53are developmentally normal but susceptible to spontaneoustumours. Nature 356:215–221
21. Draetta G, Luca F, Westendorf J, Brizuela L, Ruderman J,Beach D (1989) cdc2 protein kinase is complexed with bothcyclin A and B: evidence for proteolytic inactivation of MPF.Cell 56:829–838
22. Dulic V, Kaufmann W, Wilson S, Tlsty T, Lees E, Harper W,Elledge S, Reed S (1994) p53-dependent inhibition of cyclin-dependent kinase activities in human fibroblasts during radia-tion-induced G1 arrest. Cell 76:1013–1023
23. Dyson N, Howley PM, Münger K, Harlow E (1989) The hu-man papilloma virus-16 E7 oncoprotein is able to bind to theretinoblastoma gene product. Science 242:934–937
24. El-Deiry WS, Tokino T, Velculescu VE, Levy DB, Parsons R,Trent J M, Lin D, Mercer WE, Kinzler KW, Vogelstein B
14
(1993) WAF1, a potential mediator of p53 tumor suppression.Cell 75:817–825
25. Falquet L, Paquet N, Frutiger S, Hughes GJ, Hoang-Van K,Jaton J C (1995) cDNA cloning of a human 100 kDa de-ubiquitinating enzyme: the 100 kDa human de-ubiquitinase be-longs to the ubiquitin C-terminal hydrolase family 2 (UCH2).FEBS Letters 376:233–237
26. Farthing A, Vousden KH (1994) Functions of human papillo-mavirus E6 and E7 oncoproteins. Trends Microbiol 2:170–175
27. Fenteany G, Standaert R, Lane W, Choi S, Corey E, Schrei-ber S (1995) Inhibition of proteasome activities and subunit-specific amino-terminal threonine modification by lactacys-tin. Science 268:726–728
28. Finco T, Beg A, Baldwin A (1994) Inducible phosphorylationof IκBa is not sufficient for for its dissociation from NFκBand is inhibited by protease inhibitors. Proc Natl Acad SciUSA 91:11884–11888
29. Firpo EJ, Koff A, Solomon M, Roberts J (1994) Inactivationof a Cdk2 inhibitor during interleukin 2-induced proliferationof human T lymphocytes. Mol Cell Biol 14:4889–4901
30. Fischer-Vize JA, Rubin GM, Lehmann R (1992) The fat fac-ets gene is required for Drosophilaeye and embryo develop-ment. Development 116:985–1000
31. Galcheva-Gargova Z, Theroux S, Davis R (1995) The epider-mal growth factor receptor is covalently linked to ubiquitin.Oncogene 11:2649–2655
32. Gallant P, Nigg E (1992) Cyclin B2 undergoes cell cycle de-pendent nuclear translocation and, when expressed as a nondestructable mutant, causes mitotic arrest in HeLa cells. JCell Biol 117:213–224
33. Giordano A, Whyte P, Harlow E, Franza BR, Beach D, Dra-etta G (1989) A 60 kd cdc2-associated polypeptide complex-es with the E1A proteins in adenovirus-infected cells. Cell58:981–990
34. Glotzer M, Murray A, Kirschner M (1991) Cyclin is degrad-ed by the ubiquitin pathway. Nature 349:132–138
35. Gray, DA, Inazawa J, Gupta K, Wong A, Ueda R, TakahashiT (1995) Elevated expression of Unph, a proto-oncogene at3p21.3, in human lung tumors. Oncogene 10:2179–2183
36. Gupta K, Copeland NG, Gilbert DJ, Jenkins NA, Gray DA(1993) Unp, a mouse gene related to the tre oncogene. Onco-gene 8:2307–2310
37. Gupta K, Chevrette M, Gray DA (1994) The Unp proto-onco-gene encodes a nuclear protein. Oncogene 9:1729–1731
38. Hadari T, Warms JVB, Rose IA, Hershko A (1992) A ubiqui-tin C-terminal isopeptidase that acts on polyubiquitin chains.Role in protein degradation. J Biol Chem 267:719–727
39. Han J, Sabbatini P, Perez D, Rao L, Modha D, White E(1996) The E1B 19 K protein blocks apoptosis by interactingwith and inhibiting the p53-inducible and death-promotingBax protein. Genes Dev 10:461–477
40. Harris C, Hollstein M (1993) Clinical implications of the p53tumor-suppressor gene. N Eng J Med 329:1318–27
41. Hedge A, Goldberg A, Schwartz J (1993) Regulatory subunitsof cAMP-dependent protein kinases degraded after conjugationto ubiquitin: a molecular mechanism underlying lomg-termsynaptic plasticity. Proc Natl Acad Sci USA 90:7436–7440
42. Hengst L, Reed S (1996) Translation control of p27Kip1 accu-mulation during the cell cycle. Science 271:1861–1864
43. Hermida-Matsumoto M, Boon-Chock P, Curran T, Yang D(1996) Ubiquitinylation of transcription factors c-Jun and c-Fos using reconstituted ubiquitinylating enzymes. J Biol Chem271:4930–4936
44. Hershko A, Ganoth D, Pehrson J, Palazzo R, Cohen L (1991)Methylated ubiquitin inhibits cyclin degradation in clam em-bryo extract. J Biol Chem 266:16376–16379
45. Hershko A, Ganoth D, Sudakin V, Dahan A, Cohen L, LucaF, Ruderman J, Eytan E (1994) Components of a system thatligates cyclin to ubiquitin and their regulation by the proteinkinase cdc2. J Biol Chem 269:4940–4946
46. Hochstrasser M (1995) Ubiquitin, proteasomes, and the regu-lation of intracellular protein degradation. Curr Opin Cell Biol7:215–223
47. Holloway S, Glotzer M, King R, Murray A (1993) Anaphaseis initiated by proteolysis rather than by inactivation of matu-ration-promoting factor. Cell 73:1393–1402
48. Huang Y, Baker RT, Fischer-Vize JA (1995) Control of cellfate by a deubiquitinating enzyme encoded by the fat facetsgene. Science 270:1828–1831
49. Huibregtse J, Scheffner M, Howley P (1991) A cellular pro-tein mediates association of p53 with the E6 oncoprotein ofhuman papillomavirus types 16 or 18. EMBO J 10:4129–4135
50. Huibregtse J, Scheffner M, Beaudenon S, Howley P (1995) Afamily of proteins structurally and functionally related to E6-AP ubiquitin-protein ligase. Proc Natl Acad Sci USA 92:2563–2567
51. Irniger S, Piatti S, Michaelis C, Nasmyth K (1995) Genes in-volved in sister chromatid separation are needed for B-typecyclin proteolysis in budding yeast. Cell 81:269–277
52. Jennissen HP (1995) Ubiquitin and the enigma of intracellu-lar protein degradation. Eur J Biochem 231:1–30
53. Jensen J, Bates P, Yang M, Vierstra R, Weissman A (1995)Identification of a family of closely related human ubiquitinconjugating enzymes. J Biol Chem 270:30408–30414
54. Jensen T, Loo M, Pind S, Williams D, Goldberg A, Riordan J(1995) Multiple proteolytic systems including the protea-some, contribute to CTRF processing. Cell 83:129–135
55. Jentsch S, Schlenker S (1995) Selective protein degradation:a journey’s end within the proteasome. Cell 82:881–884
56. Kaiser P, Seufert W, Hoefferer L, Kofler B, Sachsenmair C,Herzog H, Jentsch S, Schweiger M, Schneider R (1994) Ahuman ubiquitin-conjugating enzyme homologous to yeastUBC8. J Biol Chem 269:8797–8802
57. Kanda F, Sykes DE, Yasuda H, Sandberg AA, Matsui S(1986) Substrate recognition of isopeptidase: specific cleav-age of the epsilon-(alpha-glycyl) lysine linkage in ubiquitin-protein conjugates. Biochim Biophys Acta 870:64–75
58. Kawamat AN, Morosetti R, Miller C, Park D, Spirin K, Nak-amaki T, Takeuchi S, Hatta Y, Simpson J, Wilczynski S, LeeY, Bartram C, Koeffler H (1995) Molecular analysis of thecyclin-dependent kinase inhibitor gene p27/kip1 in humanmalignancies. Cancer Res 55:2266–2269
59. King R, Peters J, Tugendreich S, Rolfe M, Hieter P, Kirsch-ner M (1995) A 20S complex containing Cdc27 and Cdc16catakyzes the mitosis-specific conjugation of ubiquitin to cy-clin B. Cell 81:279–288
60. Koken M, Reynolds P, Jaspers-Dekker I, Prakash L, PrakashS, Bootsma D, Hoeijmakers J (1991) Structural and function-al conservation of two human homologs of the yeast DNA re-pair gene RAD6. Proc Natl Acad Sci USA 88:8865–8869
61. Kovalenko O, Plug A, Haaf T, Gonda D, Ashley T, Ward D,Radding C, Golub E (1996) Mammalian ubiquitin-conjugat-ing enzyme Ubc9 interacts with Rad51 recombination proteinand localizzes in synaptonemal complexes. Proc Natl AcadSci USA 93:2958–2963
62. Lane DP, Crawford LV (1979) T-antigen is bound to host pro-tein in SV40-transformed cells. Nature 278:261–263
63. Lane DP, Lu X, Hupp T, Hall PA (1994) The role of the p53protein in the apoptotic response. Philos Trans R Soc Lond BBiol Sci 345:277–80
65. Lin W, Desiderio S (1993) Regulation of V(D)J recombina-tion activator protein RAG-2 by phosphorylation. Science260:953–959
66. Lin L, Ghosh S (1996) A glycine-reach region in NF-κBp105 functions as a processing signal for the generetion ofthe p50 subunit. Mol Cell Biol 16:2248–2254
67. Lin Y, Bown K, Siebenlist U (1995) Activation of NFκB re-quires proteolysis of the inhibitor IκB-a: signal-incducedphosphorylation of IκB-a alone does not release activeNFκB. Proc Natl Acad Sci USA 92:552–556
68. Linzer D, Levine A (1979) Characterization of a 54 K daltoncellular SV40 tumor antigen present in SV40-transformedcells and uninfected embryonal carcinoma cells. Cell 17:43–52
15
69. Liu Z, Diaz L, Haas A, Giudice G (1992) cDNA cloning of anovel human ubiquitin carrier protein. J Biol Chem 267:15829–15835
70. Loda M, Cukon B, Tam S, Lavin P, Fiorentino M, Draetta G,Jessup J, Pagano M (1996) p27 is an independent prognosticmarker in colorectal carcinoma. (submitted)
71. Lorca T, Galas S, Fesquet D, Devault A, Cavadore J, DoreeM (1991) Degradation of the proto-oncogene product p39mos
is not necessary for cyclin proteolysis and exit from meioticmetaphase: requirement for a Ca2+-calmodulin dependentevent. EMBO J 10:2087–2093
72. Luca FC, Ruderman JV (1989) Control of programmed cyclindestruction in a cell-free system. J Cell Biol 109:1895–1909
73. Luca F, Shibuya E, Dohrmann C, Ruderman J (1991) Bothcyclin A∆60 and B∆97 are stable and arrest cells in M-phase,but only cyclin B∆97 turns on cyclin destruction. EMBO J10:4311–4320
74. MacLachlan T, Sang N, Giordano A (1995) Cyclins, cyclin-dependent kinases and Cdk inhibitors: implications in cell cy-cle control and cancer. Crit Rev Eukaryotic Gene Expression5:127–156
75. Magae J, Illenye S, Wells J, Heintz NH (1996) Deregulated ex-pression of E2F-1 and DP-1 reveals distinct effects on G1-spe-cific gene expression and cell cycle progression. (submitted)
76. Maki C, Hulbregtse J, Howley P (1996) In vivo ubiquitina-tion targets wild-type p53 for degradation by the proteasome.(submitted)
77. Mal A, Poon R, Howe P, Toyoshima H, Hunter T, Harter M(1996) Inactivation of p27kip1 by the viral E1A oncoproteinin TGFβ-treated cells. Nature 380:262–265
78. Matlashewski G, Banks L, Pim D, Crawford L (1986) Analy-sis of human p53 proteins and mRNA levels in normal andtransformed cells. Eur J Biochem 154:665–672
79. Miyashita T, Reed J (1995) Tumor suppressor p53 is a directtranscriptional activator of the human bax gene. Cell 80:293–299
80. Miyashita T, Krajewski S, Krajewska M, Wang H, Lin H,Liebermann D, Hoffman B, Reed J (1994) Tumor suppressorp53 is a regulator of bcl-2 and bax gene expression in vitroand in vivo. Oncogene 9:1799–1805
81. Moran E (1993) DNA tumor virus transforming proteins andthe cell cycle. Curr Opin Gen Dev 3:63–70
82. Mori S, Tanaka K, Omura S, Saito Y (1995) Degradationprocess of ligand-stimulated platelet-derived growth factor β-receptor involves ubiquitin-proteasome proteolytic pathway. JBiol Chem 270:29447–29452
83. Murray AW, Solomon MJ, Kirschner MW (1989) The role ofcyclin synthesis and degradation in the control of maturationpromoting factor activity. Nature 339:280–286
84. Nakamura T, Hillova J, Mariage-Samson R, Onno M, Hueb-ner K, Cannizzaro LA, Boghosian-Sell L, Croce CM, Hill M(1992) A novel transcriptional unit of the tre oncogene wide-ly expressed in human cancer cells. Oncogene 7:733–741
85. Nirmala P, Thampan R (1995) Ubiquitination of the rat uter-ine estrogen receptor: dependence on estradiol. Biochem Bio-phys Res Commun 213:24–31
86. Nishizawa M, Okazaki K, Furuno N, Watanabe N, Sagata N(1992) The 2nd codon rule and autophosphorylation governthe stability and activity of Mos during the meiotic cell cyclein Xenopusoocyte. EMBO J 11:2433–2447
87. Nourse J, Firpo E, Flanagan M, Coats S, Polyak C, Lee M,Massague J, Crabtree G, Roberts J (1994) Interleukin-2-me-diated elimination of p27Kip1 cyclin-dependent kinase inhib-itor prevented by rapamycin. Nature 372:570–573
88. Nuber U, Schwarz S, Kaiser P, Schneider R, Scheffner M(1996) Cloning of human ubiquitin-conjugating enzymesUbcH6 and UbcH7 (E2-F1) and characterization of their in-teraction with E6-AP and RSP5. J Biol Chem 271:2795–2800
89. Orian A, Whitedside S, Isral A, Stancovski I, Schwartz A,Ciechanover A (1995) Ubiquitin-mediated processingof NF-kB transcreptional activator precursor p105. J Biol Chem270:21707–21714
90. Pagano M, Tam SW, Theodoras AM, Romero-Beer P, Del SalG, Chau V, Yew R, Draetta G, Rolfe M (1995) Role of theubiquitin-proteasome pathway in regulating abundance of thecyclin-dependent kinase inhibitor p27. Science 269:682–685
91. Palombella V, Rando O, Goldberg A, Maniatis T (1994) Theubiquitin-proteasome pathway is required for processing theNF-κB1 precursor protein and the activation of NF-κB. Cell78:773–785
92. Paolini R, Kinet J (1993) Cell surface control of the multiubi-quitination and deubiquitination of high-affinity immuno-globulin E receptors. EMBO J 12:779–786
93. Papa F, Hochstrasser M (1993) The yeast DOA4 gene en-codes a deubiquitinating enzyme related to a product of thehuman tre-2 oncogene. Nature 366:313–319
94. Papavassiliou A, Treier M, Chavrier C, Bohmann D (1992)Targeted degradation of c-Fos, but not v-Fos, by a phosphor-ylation signal on c-Jun. Science 258:1941–1944
95. Pietenpol J, Bohlander S, Sato Y, Papadopoulos N, Liu B,Friedman C, Trask B, Roberts J, Kinzler K, Rowley J, Vogel-stein B (1995) Assignment of the human p27Kip1 gene to 12p13and its analysis in leukemias. Cancer Res 55:1206–1210
96. Pines J, Hunter T (1989) Isolation of a human cyclin cDNA:evidence for cyclin mRNA and protein regulation in the cellcycle and for interaction with p34cdc2. Cell 58:833–846
97. Plon S, Leppig K, Do H, Groudine M (1993) Cloning of thehuman homolog of the CDC34 cell cycle gene by comple-mentation in yeast. Proc Natl Acad Sci USA 90:10484–10488
98. Polyak K, Lee M, Erdjement-Bromage H, Koff A, Roberts J,Tempst P, Massague J (1994) Cloning of p27kip1, a cyclin-de-pendent kinase inhibitor and a potential mediator of extracel-lular antimitogenic signals. Cell 79:59–66
99. Ponce-Castañeda V, Lee M, Latres E, Polyak K, Lacombe L,Montgomery K, Mathew S, Krauter K, Sheinfeld J, MassagueJ, Cordon-Cardo C (1995) p27Kip1: Chromosomal mapping to12p12-12p13.1 and absence of mutations in human tumors.Cancer Res 55:1211–1214
100. Prives C (1993) Doing the right thing: feedback control andp53. Curr Opin Cell Biol 5:214–218
101. Robinson P, Leek J, Thompson J, Carr I, Bailey A, MoynihanT, Coletta P, Lench N, Markham A (1995) A human ubiquitinconjugating enzyme, L-UBC, maps in the Alzheimer’s diseaselocus on chromosome 14q24.3. Mammal Genome 6:725–731
102. Rock K, Gramm C, Rothstein L, Clark K, Stein R, Dick L,Hwang D, Goldberg A (1994) Inhibitors of the proteasomeblock the degradation of most cell proteins and the generationof peptides presented on MHC class I molecules. Cell 78:761–771
103. Rogers S, Wells R, Rechsteiner M (1986) Amino acid se-quences common to rapidly degraded proteins: the PEST hy-pothesis. Science 234:364–368
104. Rolfe M, Romero P, Glass S, Eckstein J, Berdo I, TheodorasA, Pagano M, Draetta G (1995) Reconstitution of p53-ubiquitinylation reaction from purified components: the roleof human UBC4 and E6AP. Proc Natl Acad Sci USA 92:3264–3268
105. Rubin D, Finley D (1995) The proteasome: a protein-degrad-ing organelle? Curr Biol 5:854–858
106. Sabbatini P, Chiou S, Rao L, White E (1995) Modulation ofp53-mediated transcriptional repression and apoptosis by theadenovirus E1B 19 K protein. Mol Cell Biol 15:1060–1070
107. Sarnow P, Ho Y, Williams J, Levine A (1982) AdenovirusE1b 58KD antigen and SV-40 large tumor antigen are physi-cally associated to the same 54 KD cellular protein in trans-formed cells. Cell 28:387–394
108. Scheffner M, Werness B, Huibregtse J, Levine A, Howley P(1990) The E6 oncoprotein encoded by human papillomavi-rus types 16 and 18 promotes the degradation of p53. Cell63:1129–1136
109. Scheffner M, Huibregtse J, Vierstra R, Howley P (1993) TheHPV-16 E6 and E6-AP complex functions as a ubiquitin-pro-tein ligase in the ubiquitination of p53. Cell 75:495–505
110. Scheffner M, Huibregtse J, Howley P (1994) Identification ofa human ubiquitin-conjugating enzyme that mediates the E6-
16
AP-dependent ubiquitination of p53. Proc Natl Acad SciUSA 91:8797–8801
111. Scheffner M, Nuber U, Huibregtse J (1995) Protein ubiquiti-nation involving an E1-E2-E3 enzyme ubiquitin thioester cas-cade. Nature 373:81–83
112. Scherer D, Brockman J, Chen Z, Maniatis T, Ballard D (1995)Signal-induced degradation of IκBa requires site-specific ubi-quitination. Proc Natl Acad Sci USA 92:11259–11263
113. Seufert W, Futcher B, Jentsch S (1995) Role of a ubiquitin-conjugating enzyme in degradation of S- and M-phase cy-clins. Nature 373:78–81
114. Sheaff RJ, Roberts JM (1995) Lesson in p16 from phylumFalconium. Curr Biol 5:28–31
115. Sherr C, Roberts J (1995) Inhibitors of mammalian G1 cy-clin-dependent kinases. Genes Dev 9:1149–1163
116. Srivenugopal KU, Friedman H, Ali-Osman F (1996) Ubiquiti-nation-dependent proteolysis of O6-metthylguanine-DNAmethyltransferase in human and murine tumor cells followinginactivation with O6-benzylguanine or 1,3-bis(2-chloroethyl)-1-nitrosourea. Biochemistry 35:1328–1334
117. Stancovski I, Gonen H, Orian A, Schwartz A, Ciechanover A(1995) Degradation of the proto-oncogene c-Fos by theubiquitin proteolytic system in vivo and in vitro: identifica-tion and characterization of the conjugating enzymes. MolCell Biol 15:7106–7116
118. Stegmaier K, Takeuchi S, Golub T, Bhohlander S, Bartram C,Koeffler P, Gilliland G (1996) Mutational analysis of the can-didate tumor suppressor genes Tel and Kip1 in childhoodacute lymphoblastic leukemia. Cancer Res 56:1413–1417
119. Stein R, Chen Z, Melandri F (1995) Kinetic studies of iso-peptidase T: modulation of peptidase activity by ubiquitin.Biochemistry 34:12616–12623
120. Stewart E, Kobayashi H, Harrison D, Hunt T (1994) Destruc-tion of Xenopuscyclins A and B2, but not B1, requires bind-ing to p34Cdc2. EMBO J 13:584–594
121. Sudakin V, Ganoth D, Dahan A, Heller H, Hershko J, Luca F,Ruderman J, Hershko A (1995) The cyclosome, a large com-plex containing cyclin selective ubiquitin ligase activity, tar-gets cyclins for destruction at the end of mitosis. Mol BiolCell 6:185–198
122. Swanson D, Freund C, Ploder L, Mclnnes R, Valle D (1996)A ubiquitin C-terminal hydrolase gene on the proximal shortarm of the chromosome: inplication for X-linked retinal dis-order. Human Mol Gen 5:533–538
123. Thanos D, Maniatis T (1995) NF-κB: a lesson in family val-ues. Cell 80:529–532
124. Tobias J, Varshavsky A (1991) Cloning and functional analy-sis of the ubiquitin-specific protease gene UBP1 or Saccharo-myces cerevisiae. J Biol Chem 266:12021–12028
125. Toyoshima, Hunter T (1994) p27, a novel inhibitor of G1-cy-clin-cdk protein kinase activity, is related to p21. Cell 78:67–74
126. Treier M, Staszewski L, Bohmann D (1994) Ubiquitin-depen-dent c-Jun degradation in vivo is mediated by the ∂ domain.Cell 78:787–798
127. Tsurumi C, Ishida N, Tamura T, Kakizuka A, Nishida E, Ok-umura E, Kishimoto T, Inagaki M, Okazaki K, Sagata N,Ichihara A, Tanaka K (1995) Degradation of c-Fos by the 26Sproteasome is accelerated by c-Jun and multiple protein kin-ases. Mol Cell Biol 15:5682–5687
128. Tugendreich S, Tomkiel J, Earnshaw W, Hieter P (1995)Cdc27Hs colocalizes with Cdc16Hs to the centrosome andmitotic spindle and is essential for the metaphase to anaphasetransition. Cell 81:261–268
129. van der Velden H, Lohka M (1994) Cell cycle-regulated deg-radation of Xenopuscyclin B2 requires binding to p34cdc2.Mol Biol Cell 5:713–724
130. Ward C, Omura S, Kopito R (1995) Degradation of CFTR bythe ubiquitin-proteasome pathway. Cell 83:121–127
131. Watanabe T, Fujiwara T, Kawai A, Shimizu F, Takami S,Hirano H, Okuno S, Ozaki K, Takeda S, Shimada Y, NagataM, Takaichi A, Takahashi E, Nakamura Y, Shin S (1996)Cloning, expression, and mapping of UBE2I, a novel geneencoding a human homologue of yeast ubiquitin-conjugatingenzymes which are critical for regulating the cell cycle. Cyt-ogen Cell Gen 72:86–89
132. Werness B, Levine A, Howley P (1990) Association of hu-man papilloma virus types 16 and 18 E6 protein with p53.Science 248:76–79
133. White E (1994) p53, guardan of Rb. Nature 371:21–22134. Wilkinson, KD, Lee K, Deshpande S, Duerksen-Hughes P,
Boss JM, Pohl J (1989) The neuron-specific protein PGP9.5 isa ubiquitin carboxyl-terminal hydrolase. Science 246:670–673
135. Wilkinson, KD, Tashayev VL, O’Connor LB, Larsen CN,Kasperek E, Pickart CM (1995) Metabolism of the poly-ubiquitin degradation signal: structure, mechanism, and roleof isopeptidase T. Biochemistry 34:14535–14546
136. Xiong Y, Zhang H, Beach D (1993) Subunit rearrangement ofthe cyclin dependent kinases is associated with cellular trans-formation. Genes Dev 7:1572–1583
137. Yaglom J, Linskens M, Sadis S, Rubin D, Futcher B, FinleyD (1995) p34Cdc28-mediated control of Cln3 cyclin degrada-tion. Mol Cell Biol 15:731–741
138. Yu H, King R, Peters J, Kirschner M (1996) Identification ofa novel ubiquitin-conjugating enzyme involved in mitotic cy-clin degradation. Curr Biol 6:455–466
139. Zhu Y, Carroll M, Papa F, Hochstrasser M, D’Andrea A(1996) Dub-1, a deubiquitinating enzyme with growth sup-pressing activity. Proc Natl Acad Sci USA 93:3275–3279
