Lingering Mysteries of Ubiquitin-Chain Assembly
2006; Cell Press; Volume: 124; Issue: 1 Linguagem: Inglês
10.1016/j.cell.2005.12.025
ISSN1097-4172
Autores Tópico(s)Cancer-related Molecular Pathways
ResumoThe small protein ubiquitin is often linked to substrates as a polymer. Such polymers vary in both linkage and length, which has important consequences for their function. Surprisingly, the mechanisms of ubiquitin-chain assembly are still not known. Deciphering them will shed light on why substrates differ in the extent and timing of polyubiquitin modification and how ancillary ubiquitination factors function. The small protein ubiquitin is often linked to substrates as a polymer. Such polymers vary in both linkage and length, which has important consequences for their function. Surprisingly, the mechanisms of ubiquitin-chain assembly are still not known. Deciphering them will shed light on why substrates differ in the extent and timing of polyubiquitin modification and how ancillary ubiquitination factors function. As is true for other covalent protein modifications such as phosphorylation and acetylation, dynamic modification of proteins by ubiquitin or ubiquitin-like proteins (Ubls) enables reversible switches between different functional states (Pickart and Eddins, 2004Pickart C.M. Eddins M.J. Ubiquitin: structures, functions, mechanisms.Biochim. Biophys. Acta. 2004; 1695: 55-72Crossref PubMed Scopus (946) Google Scholar, Varshavsky, 2005Varshavsky A. Regulated protein degradation.Trends Biochem. Sci. 2005; 30: 283-286Abstract Full Text Full Text PDF PubMed Scopus (213) Google Scholar). Ubiquitination can also irreversibly inactivate a protein by targeting it to the 26S proteasome for degradation. The nature of ubiquitin-induced functional states depends on whether a single ubiquitin or multiple ubiquitin molecules are attached to the protein. Multiple ubiquitins can be attached to a protein either by monoubiquitin addition to separate substrate sites or, more commonly, in the form of a polyubiquitin chain attached at a single lysine. With the maturation of research in the ubiquitin field, a number of general principles have emerged about the enzymatic mechanisms of ubiquitin ligation to proteins. The discovery that ubiquitin molecules can be ligated to one another to form extended ubiquitin polymers on a protein substrate was an early milestone (Chau et al., 1989Chau V. Tobias J.W. Bachmair A. Marriott D. Ecker D.J. Gonda D.K. Varshavsky A. A multiubiquitin chain is confined to specific lysine in a targeted short-lived protein.Science. 1989; 243: 1576-1583Crossref PubMed Scopus (1072) Google Scholar). Ubiquitin polymers display specific amide (isopeptide) linkages between their ubiquitin units, which link the ɛ-amino group of a lysine of one ubiquitin to the C-terminal carboxyl group of the next ubiquitin in the chain. There are seven different lysines in ubiquitin that can potentially be used for ubiquitin-chain synthesis. Proteins bearing Lys48-linked chains of at least four ubiquitin molecules are generally good substrates for degradation by the proteasome (Thrower et al., 2000Thrower J.S. Hoffman L. Rechsteiner M. Pickart C.M. Recognition of the polyubiquitin proteolytic signal.EMBO J. 2000; 19: 94-102Crossref PubMed Scopus (1265) Google Scholar). Other ubiquitin-ubiquitin linkages have also been documented, and these different chain topologies correlate with different functional outcomes. The best-studied alternative ubiquitin-chain type is the Lys63-linked chain. Attachment to such chains activates specific proteins for DNA repair, signal transduction, and endocytosis, among other functions, but Lys63-linked polyubiquitin-protein conjugates are not targeted by the proteasome (Pickart and Fushman, 2004Pickart C.M. Fushman D. Polyubiquitin chains: polymeric protein signals.Curr. Opin. Chem. Biol. 2004; 8: 610-616Crossref PubMed Scopus (798) Google Scholar, Sun and Chen, 2004Sun L. Chen Z.J. The novel functions of ubiquitination in signaling.Curr. Opin. Cell Biol. 2004; 16: 119-126Crossref PubMed Scopus (361) Google Scholar). All this is fairly well established, as is the identity of the enzymatic machinery responsible for attaching ubiquitin to various protein substrates. The now well-known E1-E2-E3 trio of enzymes is responsible for activating and transferring ubiquitin to proteins (see Figure 1). Additionally, many high-resolution three-dimensional structures of different E2 and E3 enzymes have been determined, and these have yielded many insights into the mechanisms of ubiquitin-protein attachment. However, despite these advances, and contrary to most textbook depictions, the fundamental mechanism(s) of ubiquitin-chain assembly remains unknown. The textbook version—which has not been proven incorrect—holds that ubiquitin molecules are added one at a time, first to the substrate protein and then to the distal end of the growing ubiquitin chain. This "sequential addition model" is outlined in Figure 2A .Figure 2Models for Polyubiquitin-Chain SynthesisShow full caption(A) Sequential addition model (the "standard model"), in which ubiquitin molecules are added one at a time, first to a lysine on the substrate protein (S) and then to a specific lysine in the ubiquitin at the distal end of the growing ubiquitin chain. A RING E3 is depicted and is assumed to remain associated with the substrate through multiple rounds of ubiquitin addition.(B) Indexation model (Verdecia et al., 2003Verdecia M.A. Joazeiro C.A. Wells N.J. Ferrer J.L. Bowman M.E. Hunter T. Noel J.P. Conformational flexibility underlies ubiquitin ligation mediated by the WWP1 HECT domain E3 ligase.Mol. Cell. 2003; 11: 249-259Abstract Full Text Full Text PDF PubMed Scopus (208) Google Scholar). The ubiquitin chain is first built on the active-site cysteine in the HECT domain of the E3 ligase before ultimately being transferred to the substrate. A flexible hinge between two lobes of the HECT domain (not depicted) allows repositioning of the ubiquitin chain such that a lysine in the distal end of the chain is oriented for attack on the E2∼ubiquitin thioester. The chain is "indexed" to a limited length because of the physical constraints imposed by the E3 structure.(C) Seesaw model. Ubiquitin chains are built by a pair of E2s (either a homo- or heterodimer), which pass the growing chain back and forth between the two E2 active sites, before being transferred to the substrate. During chain assembly, the attacking lysine is always on a monomeric ubiquitin that is in thioester linkage to an E2. An E3 HECT domain could replace one of the E2s in this model. The order of ubiquitin addition is opposite to that in the other three models (compare colors).(D) Hybrid model. The ubiquitin chain is assembled prior to transfer to the substrate, but a noncovalent interaction between (poly)ubiquitin and a site in the E2 or E3 (the latter is depicted) positions it for nucleophilic attack on the E2∼ubiquitin thioester, much as in the sequential addition model. At some point, the free end of the ubiquitin chain must be activated by E1 and transferred to the E2 cysteine before the final transfer of the chain to the substrate.View Large Image Figure ViewerDownload Hi-res image Download (PPT) (A) Sequential addition model (the "standard model"), in which ubiquitin molecules are added one at a time, first to a lysine on the substrate protein (S) and then to a specific lysine in the ubiquitin at the distal end of the growing ubiquitin chain. A RING E3 is depicted and is assumed to remain associated with the substrate through multiple rounds of ubiquitin addition. (B) Indexation model (Verdecia et al., 2003Verdecia M.A. Joazeiro C.A. Wells N.J. Ferrer J.L. Bowman M.E. Hunter T. Noel J.P. Conformational flexibility underlies ubiquitin ligation mediated by the WWP1 HECT domain E3 ligase.Mol. Cell. 2003; 11: 249-259Abstract Full Text Full Text PDF PubMed Scopus (208) Google Scholar). The ubiquitin chain is first built on the active-site cysteine in the HECT domain of the E3 ligase before ultimately being transferred to the substrate. A flexible hinge between two lobes of the HECT domain (not depicted) allows repositioning of the ubiquitin chain such that a lysine in the distal end of the chain is oriented for attack on the E2∼ubiquitin thioester. The chain is "indexed" to a limited length because of the physical constraints imposed by the E3 structure. (C) Seesaw model. Ubiquitin chains are built by a pair of E2s (either a homo- or heterodimer), which pass the growing chain back and forth between the two E2 active sites, before being transferred to the substrate. During chain assembly, the attacking lysine is always on a monomeric ubiquitin that is in thioester linkage to an E2. An E3 HECT domain could replace one of the E2s in this model. The order of ubiquitin addition is opposite to that in the other three models (compare colors). (D) Hybrid model. The ubiquitin chain is assembled prior to transfer to the substrate, but a noncovalent interaction between (poly)ubiquitin and a site in the E2 or E3 (the latter is depicted) positions it for nucleophilic attack on the E2∼ubiquitin thioester, much as in the sequential addition model. At some point, the free end of the ubiquitin chain must be activated by E1 and transferred to the E2 cysteine before the final transfer of the chain to the substrate. In this brief review, I will discuss variations of this model that have been proposed in recent years, and I will also outline more radical departures from the standard sequential addition model. In truth, there are few published results that allow a clear choice between these different schemes. Deciphering the exact mechanisms of ubiquitin-chain assembly will be necessary for understanding many key features of protein ubiquitination, such as the apparent processivity of protein polyubiquitination, the activity of so-called chain elongation factors, and the shielding of polyubiquitin-protein conjugates from premature deubiquitination by cellular deubiquitinating enzymes (DUBs). Attachment of ubiquitin to substrate proteins involves several enzymatic steps (Pickart and Eddins, 2004Pickart C.M. Eddins M.J. Ubiquitin: structures, functions, mechanisms.Biochim. Biophys. Acta. 2004; 1695: 55-72Crossref PubMed Scopus (946) Google Scholar) (Figure 1). The C terminus of ubiquitin is first activated by ubiquitin-activating enzyme (E1). Through an ATP-dependent mechanism, ubiquitin is coupled to a cysteine side chain in E1, yielding a reactive E1∼ubiquitin thioester intermediate. The activated ubiquitin is subsequently passed to one of a number of distinct ubiquitin-conjugating enzymes (E2s) by transthiolation to a conserved cysteine of the E2. The E2 proteins catalyze substrate ubiquitination in conjunction with a ubiquitin-protein ligase (E3). For one structural class of E3 proteins (the "HECT" E3s), the ubiquitin is first transferred to a conserved cysteine of the E3 before the final transfer to a substrate group (usually a lysine side chain; Figure 1). For most other ubiquitination reactions, the E3 appears to function as an adaptor that positions the substrate in close proximity to the reactive E2∼ubiquitin thioester bond. The majority of such E3s have a subunit or domain bearing a RING motif. In addition to substrate recognition, other roles for E3s in the catalytic cycle, such as allosteric activation of the E2, remain a distinct possibility (Wu et al., 2003Wu P.Y. Hanlon M. Eddins M. Tsui C. Rogers R.S. Jensen J.P. Matunis M.J. Weisman A.M. Wolberger C.P. Pickart C.M. A conserved catalytic residue in the ubiquitin-conjugating enzyme family.EMBO J. 2003; 22: 5241-5250Crossref PubMed Scopus (136) Google Scholar, Pickart and Eddins, 2004Pickart C.M. Eddins M.J. Ubiquitin: structures, functions, mechanisms.Biochim. Biophys. Acta. 2004; 1695: 55-72Crossref PubMed Scopus (946) Google Scholar). All E2s have an asparagine residue upstream of the active-site cysteine. This asparagine appears to help form the oxyanion hole that stabilizes the tetrahedral intermediate resulting from nucleophilic substrate attack on the activated ubiquitin carbonyl (Wu et al., 2003Wu P.Y. Hanlon M. Eddins M. Tsui C. Rogers R.S. Jensen J.P. Matunis M.J. Weisman A.M. Wolberger C.P. Pickart C.M. A conserved catalytic residue in the ubiquitin-conjugating enzyme family.EMBO J. 2003; 22: 5241-5250Crossref PubMed Scopus (136) Google Scholar). However, in the atomic structures of isolated E2 enzymes, the side chain of the asparagine is fully hydrogen bonded and oriented away from the active cysteine. E3 binding to the E2 or ubiquitin thioester formation on the E2 (or both) might trigger local structural changes that allow movement of the E2 asparagine side chain to a position where it can help generate a functional oxyanion hole (Reverter and Lima, 2005Reverter D. Lima C.D. Insights into E3 ligase activity revealed by a SUMO-RanGAP1-Ubc9-Nup358 complex.Nature. 2005; 435: 687-692Crossref PubMed Scopus (361) Google Scholar). Not to be forgotten are the DUBs. These enzymes have a major impact on cellular amounts of polyubiquitin chains, both those that are attached to specific proteins and those that are apparently unanchored or "free." The mechanistic characteristics of chain assembly, such as the degree of processivity, will influence the susceptibility of ubiquitin-modified proteins to deubiquitination. There are at least five different structural classes of DUBs, and a wide range of substrate specificities and functions have been reported (Amerik and Hochstrasser, 2004Amerik A.Y. Hochstrasser M. Mechanism and function of deubiquitinating enzymes.Biochim. Biophys. Acta. 2004; 1695: 189-207Crossref PubMed Scopus (699) Google Scholar, Nijman et al., 2005Nijman S.M. Luna-Vargas M.P. Velds A. Brummelkamp T.R. Dirac A.M. Sixma T.K. Bernards R. A genomic and functional inventory of deubiquitinating enzymes.Cell. 2005; 123: 773-786Abstract Full Text Full Text PDF PubMed Scopus (1319) Google Scholar). When considering mechanisms of chain assembly, it is important to bear in mind that substrates are often adorned with very long ubiquitin polymers, sometimes containing over a dozen ubiquitin molecules (Varshavsky, 2005Varshavsky A. Regulated protein degradation.Trends Biochem. Sci. 2005; 30: 283-286Abstract Full Text Full Text PDF PubMed Scopus (213) Google Scholar). Another relevant fact is that free ubiquitin chains can be synthesized by some E2s without the aid of an E3. The earliest and most studied example is E2-25K, a mammalian E2 capable of assembling Lys48-ubiquitin chains (Chen et al., 1991Chen Z.J. Niles E.G. Pickart C.M. Isolation of a cDNA encoding a mammalian multiubiquitinating enzyme (E225K) and overexpression of the functional enzyme in Escherichia coli.J. Biol. Chem. 1991; 266: 15698-15704Abstract Full Text PDF PubMed Google Scholar). Interestingly, E2-25K from mammals and Ubc1 from yeast include C-terminal domains related in structure to the UBA domain, which forms a three-helix bundle that binds ubiquitin (Merkley et al., 2005Merkley N. Barber K.R. Shaw G.S. Ubiquitin manipulation by an E2 conjugating enzyme using a novel covalent intermediate.J. Biol. Chem. 2005; 280: 31732-31738Crossref PubMed Scopus (33) Google Scholar). In neither case is the UBA domain required for free polyubiquitin-chain synthesis; however, deletion of the UBA of Ubc1 alters the length and ubiquitin-ubiquitin linkage specificity of the chain (Pichler et al., 2005Pichler A. Knipscheer P. Oberhofer E. van Dijk W.J. Korner R. Olsen J.V. Jentsch S. Melchior F. Sixma T.K. SUMO modification of the ubiquitin-conjugating enzyme E2–25K.Nat. Struct. Mol. Biol. 2005; 12: 264-269Crossref PubMed Scopus (155) Google Scholar, Merkley et al., 2005Merkley N. Barber K.R. Shaw G.S. Ubiquitin manipulation by an E2 conjugating enzyme using a novel covalent intermediate.J. Biol. Chem. 2005; 280: 31732-31738Crossref PubMed Scopus (33) Google Scholar). Thus, noncovalent binding of a second ubiquitin molecule to a separate site in the E2 might dictate certain features of ubiquitin-chain assembly. In other pathways, the E3 might provide this additional noncovalent ubiquitin binding site. This could place a second ubiquitin molecule near the ubiquitin∼E2 thioester in a way that facilitates their linkage. Another way to juxtapose a pair of ubiquitin molecules for ubiquitin-ubiquitin formation might be through dimerization of E2s (Silver et al., 1992Silver E.T. Gwozd T.J. Ptak C. Goebl M. Ellison M.J. A chimeric ubiquitin conjugating enzyme that combines the cell cycle properties of CDC34 (UBC3) and the DNA repair properties of RAD6 (UBC2): implications for the structure, function and evolution of the E2s.EMBO J. 1992; 11: 3091-3098Crossref PubMed Scopus (83) Google Scholar, Chen et al., 1993Chen P. Johnson P. Sommer T. Jentsch S. Hochstrasser M. Multiple ubiquitin-conjugating enzymes participate in the in vivo degradation of the yeast MAT alpha 2 repressor.Cell. 1993; 74: 357-369Abstract Full Text PDF PubMed Scopus (351) Google Scholar, Gazdoiu et al., 2005Gazdoiu S. Yamoah K. Wu K. Escalante C.R. Tappin I. Bermudez V. Aggarwal A.K. Hurwitz J. Pan Z.Q. Proximity-induced activation of human Cdc34 through heterologous dimerization.Proc. Natl. Acad. Sci. USA. 2005; 102: 15053-15058Crossref PubMed Scopus (33) Google Scholar). Most E2s behave in vitro as monomers, but both in vitro and in vivo analyses have suggested an ability to form dimers, particularly if the E2 is charged with ubiquitin. For the synthesis of Lys48-ubiquitin chains in vivo, E2 homodimers and heterodimers may be important in particular ubiquitination pathways. Homodimerization of the Cdc34 E2 is induced by ubiquitin thioester formation and is necessary for its function (Varelas et al., 2003Varelas X. Ptak C. Ellison M.J. Cdc34 self-association is facilitated by ubiquitin thiolester formation and is required for its catalytic activity.Mol. Cell. Biol. 2003; 23: 5388-5400Crossref PubMed Scopus (44) Google Scholar), and heterodimer formation between the Ubc6 and Ubc7 E2s was suggested by yeast two-hybrid analysis (Chen et al., 1993Chen P. Johnson P. Sommer T. Jentsch S. Hochstrasser M. Multiple ubiquitin-conjugating enzymes participate in the in vivo degradation of the yeast MAT alpha 2 repressor.Cell. 1993; 74: 357-369Abstract Full Text PDF PubMed Scopus (351) Google Scholar). Perhaps the clearest data on the initiation of ubiquitin-chain assembly comes from the structural analysis of the heterodimer formed between Ubc13 and UEV (ubiquitin-E2 variant) proteins, which specifically catalyzes Lys63-linked ubiquitin-chain synthesis (Pickart and Eddins, 2004Pickart C.M. Eddins M.J. Ubiquitin: structures, functions, mechanisms.Biochim. Biophys. Acta. 2004; 1695: 55-72Crossref PubMed Scopus (946) Google Scholar). The UEV proteins have the same overall tertiary structure as active E2 enzymes, but they lack the catalytic cysteine. Instead, the UEV protein positions a second ubiquitin molecule noncovalently in the Ubc13-UEV heterodimer such that Lys63 of the noncovalently bound ubiquitin can attack the thioester-linked ubiquitin on Ubc13. The structural data can neatly explain the topological specificity of ubiquitin dimer assembly by Ubc13-UEV heterodimers but do not clarify how longer chains are made. Another significant point to consider in relation to ubiquitin-chain assembly mechanisms is the discovery that E1 and E3 binding sites on the E2 overlap, and their binding to E2 is mutually exclusive (Eletr et al., 2005Eletr Z.M. Huang D.T. Duda D.M. Schulman B.A. Kuhlman B. E2 conjugating enzymes must disengage from their E1 enzymes before E3-dependent ubiquitin and ubiquitin-like transfer.Nat. Struct. Mol. Biol. 2005; 12: 933-934Crossref PubMed Scopus (116) Google Scholar). Therefore, if ubiquitin chains are assembled on substrates by a sequential addition mechanism, multiple cycles of E2-E3 binding and release are probably necessary. Potentially, if a stable E2 dimer is present in the E2-E3 complex, full release of the E3 from the E2 might be avoided if E1 and E3 bind to different E2 monomers in the dimer. Other mechanisms of ubiquitin-chain synthesis might also require cyclical release of E2 from E3 (see below). The sequential addition model (the standard model) for chain assembly is a logical extension of a monoubiquitination reaction in which a substrate lysine attacks the ubiquitin∼E2 (or ubiquitin∼HECT E3) thioester bond (see Figure 2A). In each ensuing cycle, the ubiquitin most distal in the chain from the substrate provides the attacking lysine. This model becomes less intuitively appealing when applied to substrates with long ubiquitin chains because the attacking ubiquitin group becomes structurally remote from the original substrate protein bound to E3. This could be accommodated by looping out the growing ubiquitin chain between the substrate and E2 binding sites on the E3. Because E1 and E3 binding to an E2 appears to be mutually exclusive, such looping would need to be re-established in each ubiquitin addition cycle if continuous E3-substrate binding is required for processive ubiquitination (Reiss et al., 1989Reiss Y. Heller H. Hershko A. Binding sites of ubiquitin-protein ligase. Binding of ubiquitin-protein conjugates and of ubiquitin-carrier protein.J. Biol. Chem. 1989; 264: 10378-10383Abstract Full Text PDF PubMed Google Scholar, Rape et al., 2006Rape M. Reddy S.K. Kirschner M.W. The processivity of multiubiquitination by the APC determines the order of substrate degradation.Cell. 2006; 124 (this issue)Abstract Full Text Full Text PDF PubMed Scopus (226) Google Scholar). Several variations of the sequential addition model have been put forward. One idea, called the "hit-and-run" hypothesis, proposes that the ubiquitin∼E2 thioester dissociates from the E3 and diffuses to the distal end of the growing ubiquitin chain, where ubiquitin transfer then occurs (Deffenbaugh et al., 2003Deffenbaugh A.E. Scaglione K.M. Zhang L. Moore J.M. Buranda T. Sklar L.A. Skowyra D. Release of ubiquitin-charged Cdc34-S - Ub from the RING domain is essential for ubiquitination of the SCF(Cdc4)-bound substrate Sic1.Cell. 2003; 114: 611-622Abstract Full Text Full Text PDF PubMed Scopus (79) Google Scholar). This model is based on the observation that the E2 (Cdc34) must be released from the E3 (SCFCdc4) for efficient substrate polyubiquitination. However, this requirement might simply reflect the competition between E1 and E3 for overlapping binding sites on the E2. On the other hand, if the E3 triggers an activating conformational change in the ubiquitin∼E2 complex prior to release, the hit-and-run model could still potentially explain how an E3 stimulates polyubiquitin-chain extension on a substrate. The Lys48 linkage specificity of these chains would have to derive from an intrinsic property of the Cdc34 E2. Another hypothesis that has been proposed to help explain the synthesis of long ubiquitin chains in the context of the sequential addition model invokes chain elongation factors (sometimes called "E4s") (Koegl et al., 1999Koegl M. Hoppe T. Schlenker S. Ulrich H.D. Mayer T.U. Jentsch S. A novel ubiquitination factor, E4, is involved in multiubiquitin chain assembly.Cell. 1999; 96: 635-644Abstract Full Text Full Text PDF PubMed Scopus (811) Google Scholar). These elongation factors still require an E3 of the HECT class for chain assembly. Although the exact mechanism by which they enhance long ubiquitin-chain assembly remains obscure, the proteins proposed to be E4s can themselves function as E3s in other assays (Hatakeyama et al., 2001Hatakeyama S. Yada M. Matsumoto M. Ishida N. Nakayama K.I. U box proteins as a new family of ubiquitin-protein ligases.J. Biol. Chem. 2001; 276: 33111-33120Crossref PubMed Scopus (454) Google Scholar). The E4 proteins contain a U box motif, which turns out to have a tertiary structure very similar to that of the RING domain (Ohi et al., 2003Ohi M.D. Vander Kooi C.W. Rosenberg J.A. Chazin W.J. Gould K.L. Structural insights into the U-box, a domain associated with multi-ubiquitination.Nat. Struct. Biol. 2003; 10: 250-255Crossref PubMed Scopus (211) Google Scholar). It is therefore plausible that E4 elongation factors are actually E3s that can use the ubiquitin thioester-linked HECT domain of their cognate E3s in much the same way that other E3s use a cognate ubiquitin∼E2 thioester for chain elongation. As will be discussed below, chain "elongation factors" can be incorporated into other chain-assembly models as well. What are the alternatives to the sequential addition model? Surprisingly few explicit discussions of polyubiquitin-chain-assembly mechanisms exist in the literature. The few exceptions usually only consider formation of the first ubiquitin-ubiquitin bond, which allows the more complicated gymnastics of long ubiquitin-chain synthesis to be side-stepped. One early idea, developed from the finding that unanchored ubiquitin chains can be synthesized de novo and can serve as donors for transfer to a (nonubiquitin) substrate (Chen et al., 1991Chen Z.J. Niles E.G. Pickart C.M. Isolation of a cDNA encoding a mammalian multiubiquitinating enzyme (E225K) and overexpression of the functional enzyme in Escherichia coli.J. Biol. Chem. 1991; 266: 15698-15704Abstract Full Text PDF PubMed Google Scholar, Van Nocker and Vierstra, 1993Van Nocker S. Vierstra R.D. Multiubiquitin chains linked through lysine 48 are abundant in vivo and are competent intermediates in the ubiquitin proteolytic pathway.J. Biol. Chem. 1993; 268: 24766-24773Abstract Full Text PDF PubMed Google Scholar), was that ubiquitin chains could be plucked from solution by an E1, and the available C-terminal carboxyl group at the base of the chain could then be activated and conjugated to a substrate. Most cell types have significant amounts of what appear to be "free" ubiquitin chains, so this idea is not entirely implausible (Van Nocker and Vierstra, 1993Van Nocker S. Vierstra R.D. Multiubiquitin chains linked through lysine 48 are abundant in vivo and are competent intermediates in the ubiquitin proteolytic pathway.J. Biol. Chem. 1993; 268: 24766-24773Abstract Full Text PDF PubMed Google Scholar). Noel and colleagues recently proposed an "indexation model" for ubiquitin-chain synthesis (Verdecia et al., 2003Verdecia M.A. Joazeiro C.A. Wells N.J. Ferrer J.L. Bowman M.E. Hunter T. Noel J.P. Conformational flexibility underlies ubiquitin ligation mediated by the WWP1 HECT domain E3 ligase.Mol. Cell. 2003; 11: 249-259Abstract Full Text Full Text PDF PubMed Scopus (208) Google Scholar). Although highly speculative, this model was the first to address how a longer ubiquitin chain might be built on an enzyme active site prior to direct transfer to a substrate lysine (see Figure 2B). The model was inspired by the observation that in HECT E3s, the HECT domain has two lobes connected by a flexible linker; flexibility in this hinge is essential for E3 activity. In the indexation model, sequential transfer of ubiquitin units from the E2 active site to the distal end of a ubiquitin chain tethered to the E3 cysteine would be marked by a progressive opening of the hinge between the two HECT domain lobes. A length limit to the chain is implied by this mechanism, and once a ubiquitin tetramer formed on the HECT active site, it was proposed that a substrate lysine would attack the thioester and acquire the ubiquitin chain. The existing literature, however, offers no evidence for such a chain length limit. Alternatives to both the sequential addition and indexation models can be proposed. In what I will call the "seesaw model," a polyubiquitin chain is also assembled by a pair of active sites prior to transfer to substrate; these sites are assumed to arise from an E2 dimer, although an E2-HECT domain pairing is also possible (see Figure 2C). The ubiquitin chain is built by a back-and-forth transfer of the growing ubiquitin chain from one active-site cysteine to the other. Chain transfer (as shown in Figure 2C) is caused by nucleophilic attack on the thioester-linked carbonyl of one ubiquitin by a lysine side chain of the other thioester-linked ubiquitin. Another possible step in the cycle (not depicted in Figure 2C) would be a transfer of the extended ubiquitin chain back to the other E2 cysteine by transthiolation. This would mean that only a single E2 in the dimer is charged by the E1 in all the ubiquitin addition cycles. At some point, a substrate lysine will attack the thioester at the base of the chain, resulting in substrate polyubiquitination. If substrate bound E3 specifically stimulated the ubiquitin-chain-forming activity of the E2, then ubiquitin chains would be assembled preferentially when a substrate is available for subsequent modification by the chain. This might also limit E2 autoubiquitination and inactivation. The seesaw and indexation mechanisms differ in several key regards. First and most importantly, in the seesaw mechanism, the latest ubiquitin added is always at the base of the chain rather than the distal end. Second, there is no obvious length limit imposed by this mechanism. Finally, if the ubiquitin chain is not returned to one of the E2s by transthiolation in each ubiquitin addition cycle, then E1 must alternate between charging different E2 monomers with ubiquitin, whereas in the indexation model, a single E2 suffices. (If an E2-HECT E3 dimer were the basis of the seesaw mechanism, either another E2 would be needed to charge the E3 or the ubiquitin chain would have to be moved by transthiolation to the E3 cysteine in each cycle.) Of all the models discussed in Figure 2, only the seesaw mechanism always places the most recently added ubiquitin at the base of the chain, a feature that can be tested experimentally. A final model to consider here is the "hybrid model" diagrammed in Figure 2D. In this scheme, a second, noncovalent ubiquitin binding site is
Referência(s)