Post-translational regulation in plants employing a ... - Semantic Scholar

Plant Signalling from Genes to Biochemistry

Post-translational regulation in plants employing a diverse set of polypeptide tags B. Downes and R.D. Vierstra1 Department of Genetics, 425-G Henry Mall, University of Wisconsin-Madison, Madison, WI 53706, U.S.A.

Abstract The concept that plants exploit polypeptides as post-translational modifiers is rapidly emerging as an important method to manipulate various cellular processes. The best known is Ub (ubiquitin) that serves as reusable tag for selective protein degradation by the 26 S proteasome and for endosomal trafficking. Genomic analyses indicate that Ub pathway alone comprises over 6% of the Arabidopsis proteome with thousands of proteins being targets. Consequently, this pathway influences much of plant biology. Others tags include RUB-1 (related to Ub-1; also known as NEDD8), SUMO (small Ub-like modifier), ATG-8 (autophagy-8) and ATG-12, UFM-1 (Ub-fold modifier-1) and HUB-1 (homology to Ub-1). Preliminary studies indicate that these tags have much more limited sets of targets and provide more specialized functions, including transcriptional regulation, protein localization, autophagic turnover and antagonizing the effects of Ub. On the basis of their widespread distribution and pervasive functions, peptide tags can now be considered as prime players in plant cell regulation.

Introduction All organisms use a variety of chemical modifiers for the posttranslational control of proteins that affect growth, development and homoeostasis. In addition to small molecules such as methane, acetate, fatty acids, sugars, nucleosides and phosphate, eukaryotes also employ polypeptides as modifiers. This concept of covalently attaching one polypeptide to another after the completion of translation was first realized ∼25 years ago with the Noble Prize-winning work of Hershko, Ciechanover, Rose and co-workers [1,2]. They discovered that the small protein Ub (ubiquitin) serves as an influential tag to target other proteins selectively for destruction. In recent years, a panoply of other polypeptide tags, ranging in size from 72 to 186 amino acids, has emerged. While similar in three-dimensional structure to Ub, they appear to impart different attributes to their targets. Some have affectionately called this class of modifiers ‘Darwin’s phosphate’. Since, unlike other modifiers such as phosphate that are immutable, the exploitation of polypeptides as modifiers allows organisms to harness evolution to generate new sequence variants as a facile way to expand functional diversity. In this review, we describe the current list of polypeptide tags evident in plants. Surprisingly, all can be found in animals and fungi as well, indicating that their appearance predated the evolutionarily split among these three eukaryotic kingdoms. In addition to Ub, the list includes RUB-1 (related to Ub-1; also known as NEDD8), SUMO (small Ub-like modifier), ATG-8 (autophagy-8) and ATG-12, UFM-1 (Ub-fold modiKey words: autophagy 8 (ATG8), homology to ubiquitin 1 (HUB1), polypeptide tag, related to ubiquitin 1 (RUB1), small Ub-like modifier (SUMO), ubiquitin-fold modifier (UFM). Abbreviations used: ATG, autophagy; CP, core protease; CSN, COP9 signalosome; DUB, deubiquitinating enzyme; Ub, ubiquitin; HUB, homology to Ub; RUB, related to Ub; SUMO, small Ub-like modifier; UFM, Ub-fold modifier; RP, regulatory particle; ULP, Ubl-specific peptidase. 1 To whom correspondence should be addressed (email [email protected]).

fier-1) and HUB-1 (homology to Ub-1). However, we predict that this list will probably expand as we continue to explore various plant genomes/proteomes for new types. As can be seen in Table 1 and Figure 1(A), some are highly similar to each other in amino acid sequence (Ub and RUB1), whereas others share little homology (e.g. Ub and ATG8). Despite this sequence diversity, structural determinations of Ub, RUB1/NEDD8, SUMO, ATG8, UFM1 and HUB1 indicate that most, if not all, of these tags adopt a similar threedimensional shape (Figure 1B). This structure, now called the Ub fold [3], consists of a globular core generated by a pocket of four interacting β-strands into which an α-helix fits diagonally. It is stabilized by numerous intramolecular interactions, which in turn impart remarkable stability against a variety of denaturants. Protruding from the Ub fold is a flexible C-terminal extension. The terminal amino acid, which is glycine in all cases but HUB1, provides the active-site carboxyl group. This group interacts with members of its conjugation cascade and ultimately participates in forming the isopeptide-type linkage that connects the tag to the target. In most cases, the binding site in the targets is the ε-amino group of an accessible lysine. However, for ATG8, the target is the lipid phosphatidylethanolamine, implying that primary amino groups from other types of nitrogen-containing compounds can be used. Although this recurring three-dimensional structure probably reflects a common progenitor, it may also reflect a universal property that optimizes its function. For example, the protruding C-terminal active site enables the tag to interact easily with the enzymes that promote processing, conjugation and deconjugation while minimizing binding to the target. The structural stability of the tag could further discourage inappropriate interactions with their targets and reduce the probability that the tags assume nonnative states that could prevent recycling.  C 2005

Biochemical Society

393

394

Biochemical Society Transactions (2005) Volume 33, part 2

Table 1 Description of the seven types of polypeptide tags present in Arabidopsis

Tag

Similarity to Ub* (%)

Identity with Ub* (%)

Number of isoforms

Length (no. of amino acids)*

Ub

–

–

1

76

RUB1 HUB1 SUMO†

73 35 30

59 21 12

3 2 9

76 72 94

UFM1 ATG12‡ ATG8‡

30 30 21

14 8 5

1 2 9

86 96 116

*Length of the processed polypeptide. †Analysis of the SUMO1 isoform [25]. ‡Analysis of the a isoform [37].

In addition to their similar shapes, these seven polypeptide tags share mechanistically analogous multistep conjugation cascades (Figure 2). In ATP-dependent reactions catalysed by related E1s (or activating enzymes), the tags are activated, first by adenylation of the C-terminal carboxyl group followed by the formation of a high-energy intermediate in which the carboxyl group of the tag is attached to the E1 through an S-ester bond to an active-site cysteine. The activated tag is then transferred through transesterification to an active-site cysteine within an E2 (or conjugating enzyme). Finally, the tag is donated to the amino group of the target to generate the tag conjugate. For both Ub and SUMO tags (and maybe others), reiterative rounds of conjugation can occur [4,5]. These additional Ub/SUMO moieties are connected through lysine residues in the previously attached molecules to generate polymers bound to a single lysine in the target. For Ub, several of its lysine residues serve as acceptors with certain linkages imparting distinct functions to the polymer [4]. Akin to the phosphatases in the phosphorylation cycle, a unique protease also exists that can reverse the conjugation of the tags to their targets by cleaving only the isopeptide bond linking the two. In addition to being remarkably specific to the tag, these proteases probably also have some specificity to the target, thus allowing them to disassemble specific conjugates. Another unique feature for most of these tags (except ATG12) is that they are synthesized as larger precursors: these precursors are then processed by the tag-specific proteases to expose the functional C-terminal amino acid. Superficially, it is tempting to speculate that this processing helps to regulate the tags by controlling the amount of the functional form. However, it is also possible that the C-terminal appendages arose from genetic defects that remain silent simply because tag-specific proteases are available to repair these defects after translation.

Ub Ub was the first peptide tag to be identified and has been the most intensively studied. Its main function is to help selectively target proteins for degradation by the 26 S pro C 2005

Biochemical Society

teasome [1,6]. This 76-amino-acid protein is synthesized by a large family of UBQ genes; for example 15 UBQ genes exist in the plant Arabidopsis thaliana [7]. They all synthesize Ub moieties in which the coding region is either concatenated in tandem arrays or fused in-frame to other proteins: these translational fusions are then processed to release functional Ub monomers. After processing, free Ub becomes attached to proteins through an E1 → E2 → E3 cascade as shown in Figure 2 [1,6]. The E3s provide much of the target specificity and accordingly this group of factors is the most numerous and complex. For example, whereas the Arabidopsis genome encodes only two types of E1s (UBA1–UBA2) and ∼37 different E2s (UBC1–UBC37), it potentially encodes more than 1400 distinct E3s [8]. The sheer number of E3s (∼5% of the Arabidopsis proteome) in turn implies that thousands of Ub targets are likely. While only a few Ub targets have been described empirically in plants thus far [6], the use of highthrough-put MS should expand this list considerably [9]. Plant E3s cluster into several categories defined by their subunit composition and mechanisms of action [6]. The HECT E3s are single long polypeptides that contain binding sites for the Ub-E2 and the target [10]. They are unique in that they form a transient E3-Ub intermediate that serves as the proximal Ub donor. The RING E3s are also single polypeptides. They use a diverse set of protein–protein interaction motifs to bind the target and a RING domain, consisting of an octet of cysteine and histidine residues that chelate two zinc atoms, to associate with the Ub-E2 intermediate. A related set of U-Box E3s employ electrostatic interactions instead of zinc chelation to generate a structurally similar U-Box pocket that binds Ub-E2. The Cullin-based E3s are multimeric complexes. They consist of a RING domaincontaining RBX1 subunit that binds the Ub-E2, a target recognition module, and a Cullin subunit that brings the two together in an appropriate orientation for transfer. The target recognition module has several distinct iterations. For example, in the BTB E3s, this module consists of a single BTB protein that binds both the target and the Cullin, whereas in the SCF E3s, it consists of an F-Box protein that binds the target and an SKP1 adaptor that tethers the F-Box to the Cullin. Through a simple exchange of BTB and F-Box subunits, related Cullin-based E3s can specifically ubiquitinate numerous targets. Two of the largest gene families in Arabidopsis encode the F-Box (∼700) and RING-containing proteins (469) [11,12]. Consistent with these large numbers, genetic analyses of individual RING and SCF E3s indicate that they influence most if not all aspects of plant biology, including embryogenesis, photomorphogenesis, floral homoeosis, hormone signalling, circadian rhythms, environmental sensing, disease resistance and senescence, to name a few [6,13]. Those targets modified with single Ubs have a variety of functions including roles in DNA repair, endocytosis and internalization and degradation of plasma membrane-bound proteins by the lysosome/vacuole [14,15]. Those targets bearing multiple Ubs are typically degraded by the 26 S proteasome. This 2 MDa protease is assembled from two

Plant Signalling from Genes to Biochemistry

Figure 1 Amino acid sequence and structure of the seven polypeptide tags found in plants (A) Amino acid sequence alignment of the tags from Arabidopsis, along with the Ub-like domains from RAD23 and the HECT E3 UPL5, and the Escherichia coli proteins ThiS and MoaD involved in sulphur metabolism. Sequences of the processed mature proteins were aligned using ClustalX v1.83 (http://www.igbmc.u-strasbg.fr/BioInfo/ClustalX/) and displayed with MACBOXSHADE (Institute of Animal Health, Pirbright, Surrey, U.K.). Identical and similar residues are highlighted with black and grey backgrounds respectively. The oval designates the rest of the UPL5 and RAD23 proteins. PDB (Protein Database) accession numbers for the sequences are as follows: Ub, At3g62250; RUB1, At2g35635; UPL5, At4g12570; RAD23c, At3g02540; HUBa, At5g42300; SUMO1, At4g26840; UFM1, At1g77710; ATG12a, At1g54210; ATG8a, At4g21980; ThiS, YP 026279; and MoaD, NP 992372. (B) Three-dimensional ribbon diagram of several peptide tags in comparison with that for E. coli ThiS. For SUMO1, UFM1 and ATG8, the structures of the full-length unprocessed proteins are shown. N and C identify the N- and C-terminal ends of the polypeptides. The PDB accession numbers and references (if available) for the structures are as follows: E. coli ThiS, 1F0Z [42]; Avena sativa Ub, 1D3Z [46]; A. thaliana RUB1, 1BT0 [47]; S. cerevisiae HUB1, 1M94 [48]; B. taurus ATG8, 1EO6 [49]; Caenorhabditis elegans UFM1, 1L7Y; and H. sapiens SUMO, 1AR5 [50].

 C 2005

Biochemical Society

395

396

Biochemical Society Transactions (2005) Volume 33, part 2

Figure 2 The proposed functions of the seven polypeptide tags found in plants and components required for their processing, conjugation and disassembly Question marks indicate factors or functions that are unknown at present. C, active-site cysteine. K, lysine acceptor site within the target that binds to the tag. PE, phosphatidylethanolamine. Whereas polymers of both Ub and SUMO can be attached to targets by reiterative rounds of conjugation, the other tags appear to be attached as monomers. For SUMO and RUB1, the E1 consists of a heterodimer of two proteins. For RUB1, ATG8, ATG12, UFM1 and HUB1 conjugation, it remains unclear if an E3 is required. For ATG8 and ATG12, only single targets have been identified, PE and the ATG5 protein respectively.

subcomplexes, the 20 S CP (core protease) and the 19 S RP (regulatory particle) [1,6]. The 28-subunit CP exists as a hollow cylinder that contains the protease active sites within a central chamber. The entrance to this chamber is sufficiently narrow and capped to prevent inadvertent proteolysis and to restrict breakdown to only proteins that are deliberately imported. The RP sits on top of either end of the CP and provides the specificity for Ub conjugates. Its 18 or so prin C 2005

Biochemical Society

cipal subunits have a number of activities, including detecting ubiquitinated substrates, unfolding the polypeptide, releasing the bound Ub moieties, opening the gate to the CP and importing the polypeptide. Additional loosely bound subunits also exist that may impart regulatory functions as yet to be discovered [16]. The Ub pathway also has numerous DUBs (deubiquitinating enzymes) that remove attached Ubs; at least 30 can be

Plant Signalling from Genes to Biochemistry

detected in the Arabidopsis genome [17,18]. The functions of most of these are unknown. Several DUBs appear to have roles in releasing Ub monomers from the initial translation product and in disassembling poly-Ub chains attached to various targets. Several DUBs are associated with the 26 S proteasome [16,19].

RUB1/NEDD8 We first detected RUB1/NEDD8 during a search of the Arabidopsis genome database for UBQ genes. One novel pair was discovered to encode a Ub monomer linked in-frame to a 76-amino-acid sequence 59% identical with Ub [7]. The discovery that RUB1 functioned as a distinct peptide modifier came from studies by Hellmann and Estelle [13] on various Arabidopsis mutants that affected auxin signalling. One locus required for auxin sensitivity, AXR1, was found to encode a protein related to the N-terminal half of the Ub E1s [20]. Whereas AXR1 would not interact with Ub, it did with RUB1. Ultimately, the RUB1 conjugation cascade was resolved and found to include a heterodimeric E1 consisting of AXR1 and ECR1 (related to the C-terminal half of UBA1) and the RCE1 E2 [21] Surprisingly, the targets of RUB1 are not overlapping with Ub and appeared to be much more limited. In fact, the only targets of RUB1 known thus far are members of the Cullin family using a conserved lysine in the Cullin for attachment [21]. RUB1 binding appears to regulate positively Cullinbased E3s, possibly by helping to stabilize the E3 complex with the Ub-E2 donor. Whereas no E3 has been identified for RUB1, it has been suggested that the RBX1 subunit of the Cullin–E3 complex can function in this capacity [22]. Similar to ubiquitination, rubination also can be reversed by isopeptidase activity(ies). One such activity is provided by subunit 5 of the CSN (COP9 signalosome) complex, an 8-subunit complex distantly related to the Lid subcomplex of the 26 S proteasome RP [23]. Through this derubinating activity, the CSN acts as a negative regulator of Cullin-based E3s. Given this integral role of RUB1 in Ub transfer, it is not surprising that defects in RUB1 conjugation/deconjugation have pleiotropic effects on plant growth and development [21].

SUMO SUMO was first discovered in animals during the analysis of RanGAP, a protein associated with the nuclear pore complex. On amino acid sequencing, RanGAP was discovered to have two N-termini but only one C-terminus, indicating that the protein contained a second polypeptide attached through an isopeptide bond [24]. The appended polypeptide was named SUMO based on its weak similarity to Ub. SUMOs are present in all eukaryotes; we found eight different SUMO isoforms in Arabidopsis that cluster into five subfamilies [25]. In addition to numerous amino acid substitutions within the Ub fold (only 12% identical with Ub), SUMOs contain a long flexible N-terminal extension, the function of which remains unclear (Figure 1). The corresponding conjugation

reaction requires a heterodimeric E1 consisting of SAE1 and SAE2, the E2 SCE1, and at least several types of E3s (Figure 2 and [5]). One E3 type contains a derivative of the RING motif that may bind the SUMO-E2 intermediate in binding [26]. The SUMO pathway also includes a large collection of isopeptidases, designated ULPs (Ubl-specific peptidases) that remove SUMO from targets and presumably help process the C-terminal end of the SUMO precursor. Both immunoblot analyses (e.g. [25]) and proteomic analyses by MS [27] suggest that numerous intracellular proteins are targets. In some cases, the SUMO is attached to a consensus SUMO-binding site KXE, where  is a hydrophobic residue, K the lysine for attachment, X is any residue and E stands for the glutamic acid [5]. A wide range of functions have been ascribed to SUMO, including helping proteins traverse the nuclear envelope, promoting localization within the nuclear matrix, activating/ deactivating transcription factors and blocking Ub-mediated events by becoming bound to the same lysine within targets [5]. In this last capacity for example, the inhibitory κB repressor of the nuclear factor κB transcriptional activator, can be modified by both SUMO and Ub to regulate its half-life [28]. SUMO may have a profound role during the response of plants to various environmental stresses. We found that when Arabidopsis is exposed to heat shock, H2 O2 , ethanol or the amino acid analogue canavanine, the levels of SUMO1/2 conjugates are significantly but reversibly increased [25]. The increase can be seen within 2 min of a heat stress, making it one of the most rapid heat shock responses found in plants. Arabidopsis plants overexpressing SUMO1 were subsequently shown to be less sensitive to the hormone abscisisc acid, further suggesting a role of sumolyation in stress signalling [29]. Several species of pathogenic bacteria for both plants and animals have been recently shown to secrete proteases related to ULPs into the host upon infection, with this secretion attenuating the accumulation of SUMO conjugates in planta [30,31]. An attractive possibility is that these bacteria interfere with stress responses directed by one or more SUMO conjugates by promoting their disassembly. Desumolyation of conjugates may also be important to floral induction. This connection was made recently with the discovery that the Arabidopsis ESD4 locus, which promotes early flowering in short days when mutated, encodes a nuclear-localized protein related to the ULP family of SUMO-specific protease [32].

ATG8 and ATG12 The polypeptide tags ATG8 and ATG12 were discovered during the genetic analysis of autophagy in the yeast Saccharomyces cerevisiae [33]. Autophagy is a degradative process whereby cytoplasm and organelles are engulfed in double membrane-bound vesicles. The outer membrane of these vesicles fuses with the tonoplast, which then delivers the interior vesicle and its cargo to the vacuolar lumen for breakdown. Unlike the Ub/26 S proteasome system, this degradation is for the most part non-selective and activated mainly during starvation and apoptosis when recycling of  C 2005

Biochemical Society

397

398

Biochemical Society Transactions (2005) Volume 33, part 2

nutrients is critical. Genetic dissection of the autophagic process identified a highly complex cascade of events, including an ATG8/12 conjugation pathway involving the ATG7 E1, that activates both tags in the presence of ATP, and the two E2s, ATG3 and ATG10, that work with ATG8 and ATG12 respectively (Figure 2 and [33]). Similar to RUB1, the targets appear to be very limited. Whereas ATG8 becomes attached to another protein ATG5, ATG8 become attached to the lipid phosphatidylethanolamine. During autophagy, ATG8 and ATG12 conjugation play important roles in forming the vesicles that encapsulate cytoplasm [34]. While their exact functions remain unclear, presumably the ATG8/12 conjugates interact with the cytoskeleton during vesiculation. Recent cell biological studies showed that some of the ATG8 pool binds to the autophagic vesicles, probably through the lipid moiety of the conjugate, and eventually is digested within the vacuole along with the inner vesicle (A.R. Thompson, J.H. Doelling, A. Suttangkakul and R.D. Vierstra, unpublished work and [36]). Arabidopsis null mutants in the conjugation cascade are viable, indicating that this autophagic process in not essential in plants (A.R. Thompson, J.H. Doelling, A. Suttangkakul and R.D. Vierstra, unpublished work and [36,37]). However, the mutant plants are hypersensitive to carbon- or nitrogenlimiting conditions and senesce early, indicating that this modification plays an important role in nutrient remobilization/recycling.

HUB1 HUB1 was first discovered in yeast from sequence searches of its complete genome for Ub-related proteins [38]. Although bearing weak similarity to Ub (21% identity), its threedimensional structure is nearly identical with Ub (Figure 1). However, the processed HUB1 is notably distinct from other polypeptide tags by ending in a C-terminal tyrosine instead of a glycine. Other elements of its conjugation cascade are not yet known; given the use of a different C-terminus, it is possible that the HUB1 pathway is mechanistically distinct from the other tags. Of the numerous possible targets in yeast detectable with epitope-tagged HUB1, only two have been identified thus far, Sph1 and Hbt1, both of which help form projections during mating [38]. In hub1∆ mutants, the localization of Sph1 and Hbt1 are defective and mating is impaired, suggesting that their modification by HUB1 is critical to appropriate cell polarity during mating. Using yeast HUB1 as the query, related proteins have been found in plants and animals [38]. A pair of HUB1 genes is present in Arabidopsis, but the role(s) of this tag in plants is currently unknown.

UFM1 UFM1 is the most recently identified member of the Ub-like polypeptide tags. It was discovered indirectly in yeast during the analysis of the E1-related protein UBA5. Using UBA5 as a bait, Komatsu et al. [39] discovered UFM1 as the protein that would form an S-ester adduct with this E1 in the presence of ATP. Although being only 14% identical with Ub, UFM1  C 2005

Biochemical Society

assumes a similar Ub fold (Figure 1). And like Ub, the UFM1 precursor is processed by a yet to be described protease to expose a C-terminal glycine that is essential for its conjugation. In addition to UBA5, the conjugation cascade employs the E2 UFC1 [39]. Whether an E3 is also required is unknown. At present, the function(s) of UFM1 are not understood. Several UFM1 conjugates have been detected immunologically in mammalian cell cultures but their identification awaits further proteomic analyses. Both UFM1 and UBA5 can be easily identified in both the Arabidopsis and rice genomes, indicating that a similar conjugation pathway exists in plants.

Translationally fused polypeptide tags In addition to being added post-translationally, numerous proteins exist in plants and animals that bear a Ub-like domain that is translationally linked. Whereas some of these Ub-like domains are released from the protein after translation, a number remain permanently attached [6]. Examples of the latter in plants include RAD23 and the HECT E3 UPL5 that bear a Ub-like domain at their N-terminal ends [6,10]. Given their similarity to Ub, these Ub fusions are expected to assume the same Ub-fold within the context of the full-length protein (Figure 1). However, they are typically missing the C-terminal glycine, which prohibits their processing by DUB-like proteases. Where studied, the translationally fused Ub-like domains appear to help the parent protein interact with various receptors within the Ub pathway [40]. RAD23 for example, uses the Ub-like domain to associate with the 26 S proteasome through interaction with RPN1, a Ubbinding subunit. RAD23 also contains two UBA domains that are capable of binding Ub polymers. One possibility is that the combination of the Ub-like and UBA domains allows RAD23 to chaperone ubiquitinated cargo to the 26 S proteasome [40].

The evolutionary origins of polypeptide tags Both their sequence similarity and widespread distribution among plants, fungi and animals indicate that polypeptide tags appeared early during the evolution of eukaryotes. However, their apparent absence in prokaryotes implies that the first tag was not of prokaryotic origin. A possible prokaryotic ancestor could be the sulphur carrier proteins ThiS and MoaD that participate in thiamin and molybdenum cofactor biosynthesis respectively [41,42]. These short proteins contain the Ub-fold and, like most polypeptide tags, they end in a flexible protruding C-terminal glycine (Figure 1). The terminal glycine residues of each become activated in an ATP-dependent reaction by the E1-like proteins, ThiF and MoeB, to form a high-energy acyl-adenylate, and subsequently form an S-ester intermediate. The bound sulphur atoms are then incorporated into the thiamin and molybdenum cofactors. Given the catalytic similarity of these reactions to the conjugate cascades used by polypeptide tags, one can imagine that they provided the ancestral activation

Plant Signalling from Genes to Biochemistry

scheme. In the case of the polypeptide tags, the energy in the S-ester bond was directed towards forming the isopeptide bond to available amino groups and not towards sulphur chemistry.

Conclusions Clearly, the continuously expanding number of polypeptide tags implies that this type of modification plays prominent and diverse roles in plant biology. Their main effect is directed not at perturbing the structure or activity of the target protein but towards imbuing the target with new cellular interactions. The binding of ubiquitinated proteins to the 26 S proteasome is a prime example. Regardless of their origins and functions, several tags converge to regulate similar processes. For example, Ub, ATG8 and ATG12 promote intracellular recycling by the 26 S proteasome and the vacuole. RUB1 works as an accessory modification to promote ubiquitination of targets by Cullin-based E3s. And SUMO can work as an antagonist to Ub by competing for the same lysine in the target. One central question is the identity of the targets. Here, the application of MS coupled with methods to enrich for tagged species will be invaluable [9,27]. In addition to the seven tags known so far in plants, it is probable that others exist. For example, several other tags related in sequence to Ub have been identified in animals but remain to be confirmed in plants (e.g. URM1p, ISG15 and FAT10 [43–45]). Both ISG15 and FAT10 are unique in that they resemble translationally fused Ub dimers. Cursory searches of the Arabidopsis genome database have detected several unclassified proteins with Ub-folds as well as orphan E1-like proteins that could promote their activation. Whether these proteins represent new types of polypeptide tags await further investigations. We thank the U.S. National Science Foundation Arabidopsis 2010 program and a NIH-NRSA Postdoctoral Fellowship for support.

References 1 Hershko, A. and Ciechanover, A. (1998) Annu. Rev. Biochem. 67, 425–479 2 Vierstra, R.D. (1999) in Discoveries in Plant Biology, vol. 3, Chapter 16 (Kung, S.D. and Yang, S.F., eds.), pp. 305–324, World Scientific Publishing, Hong Kong 3 Larsen, C.N. and Wang, H. (2002) J. Proteome Res. 1, 411–419 4 Pickart, C.M. (2000) Trends Biochem. Sci. 25, 544–548 5 Johnson, E.S. (2004) Annu. Rev. Biochem. 73, 355–382 6 Smalle, J. and Vierstra, R.D. (2004) Annu. Rev. Plant Biol. 55, 555–590 7 Callis, J., Carpenter, T., Sun, C.W. and Vierstra, R.D. (1995) Genetics 139, 921–939 8 Vierstra, R.D. (2003) Trends Plant Sci. 8, 135–142 9 Peng, J., Schwartz, D., Elias, J.E., Thoreen, C.C., Cheng, D., Marsischky, G., Roelofs, J., Finley, D. and Gygi, S.P. (2003) Nat. Biotechnol. 21, 921–926 10 Downes, B.P., Stupar, R.M., Gingerich, D.J. and Vierstra, R.D. (2003) Plant J. 35, 729–742 11 Gagne, J.M., Downes, B.P., Shiu, S.H., Durski, A.M. and Vierstra, R.D. (2002) Proc. Natl. Acad. Sci. U.S.A. 99, 11519–11524 12 Stone, S.L., Hauksdottir, H., Troy, A., Herschleb, J., Kraft, E. and Callis, J. (2005) Plant Physiol. 137, 13–30

13 Hellmann, H. and Estelle, M. (2002) Science 297, 793–797 14 Hoege, C., Pfander, B., Moldovan, G.L., Pyrowolakis, G. and Jentsch, S. (2002) Nature (London) 419, 135–141 15 Bonifacino, J.S. and Traub, L.M. (2003) Annu. Rev. Biochem. 72, 395–447 16 Leggett, D.S., Hanna, J., Borodovsky, A., Crosas, B., Schmidt, M., Baker, R.T., Walz, T., Ploegh, H. and Finley, D. (2002) Mol. Cell 10, 495–507 17 Doelling, J.H., Yan, N., Kurepa, J., Walker, J. and Vierstra, R.D. (2001) Plant J. 27, 393–405 18 Yan, N., Doelling, J.H., Falbel, T.G., Durski, A.M. and Vierstra, R.D. (2000) Plant Physiol. 124, 1828–1843 19 Verma, R., Aravind, L., Oania, R., McDonald, W.H., Yates, III, J.R., Koonin, E.V. and Deshaies, R.J. (2002) Science 298, 611–615 20 Leyser, H.M., Lincoln, C.A., Timpte, C., Lammer, D., Turner, J. and Estelle, M. (1993) Nature (London) 364, 161–164 21 Parry, G. and Estelle, M. (2004) Semin. Cell Dev. Biol. 15, 221–229 22 Gray, W.M., Hellmann, H., Dharmasiri, S. and Estelle, M. (2002) Plant Cell 14, 2137–2144 23 Cope, G.A., Suh, G.S.B., Aravind, L., Schwarz, S.E., Zipursky, S.L., Koonin, E.V. and Deshaies, R.J. (2002) Science 298, 608–611 24 Matunis, M.J., Coutavas, E. and Blobel, G. (1996) J. Cell Biol. 135, 1457–1470 25 Kurepa, J., Walker, J.M., Smalle, J., Gosink, M.M., Davis, S.J., Durham, T.L., Sung, D.Y. and Vierstra, R.D. (2003) J. Biol. Chem. 278, 6862–6872 26 Johnson, E.S. and Gupta, A.A. (2001) Cell (Cambridge, Mass.) 106, 735–744 27 Wohlschlegel, J.A., Johnson, E.S., Reed, S.I. and Yates, J.R. (2004) J. Biol. Chem. 279, 45662–45668 28 Desterro, J.M., Rodriguez, M.S. and Hay, R.T. (1998) Mol. Cell 2, 233–239 29 Lois, L.M., Lima, C.D. and Chua, N.H. (2003) Plant Cell 15, 1347–1359 30 Orth, K., Xu, Z., Mudgett, M.B., Bao, Z.Q., Palmer, L.E., Bliska, J.B., Mangel, W.F., Staskawicz, B. and Dixon, J.E. (2000) Science 290, 1594–1597 31 Hotson, A., Chosed, R., Shu, H.J., Orth, K. and Mudgett, M.B. (2003) Mol. Microbiol. 50, 377–389 32 Murtas, G., Reeves, P.H., Fu, Y.F., Bancroft, I., Dean, C. and Coupland, G. (2003) Plant Cell 15, 2308–2319 33 Ohsumi, Y. and Mizushima, N. (2004) Semin. Cell Dev. Biol. 15, 231–236 34 Thompson, A.R. and Vierstra, R.D. (2005) Curr. Opin. Plant Biol., in the press 35 Reference deleted 36 Yoshimoto, K., Hanaoka, H., Sato, S., Kato, T., Tabata, S., Noda, T. and Ohsumi, Y. (2004) Plant Cell 16, 2967–2983 37 Doelling, J.H., Walker, J.M., Friedman, E.M., Thompson, A.R. and Vierstra, R.D. (2002) J. Biol. Chem. 277, 33105–33114 38 Dittmar, G.A., Wilkinson, C.R., Jedrzejewski, P.T. and Finley, D. (2002) Science 295, 2442–2446 39 Komatsu, M., Chiba, T., Tatsumi, K., Iemura, S., Tanida, I., Okazaki, N., Ueno, T., Kominami, E., Natsume, T. and Tanaka, K. (2004) EMBO J. 23, 1977–1986 40 Hartmann-Petersen, R. and Gordon, C. (2004) Semin. Cell Dev. Biol. 15, 247–259 41 Lake, M.W., Wuebbens, M.M., Rajagopalan, K.V. and Schindelin, H. (2001) Nature (London) 414, 325–329 42 Wang, C., Xi, J., Begley, T.P. and Nicholson, L.K. (2001) Nat. Struct. Biol. 8, 47–51 43 Goehring, A.S., Rivers, D.M. and Sprague, Jr, G.F. (2003) Mol. Biol. Cell 14, 4329–4341 44 Ritchie, K.J. and Zhang, D.E. (2004) Semin. Cell Dev. Biol. 15, 237–246 45 Raasi, S., Schmidtke, G. and Groettrup, M. (2001) J. Biol. Chem. 276, 35334–35343 46 Vijay-Kumar, S., Bugg, C.E., Wilkinson, K.D., Vierstra, R.D., Hatfield, P.M. and Cook, W.J. (1987) J. Biol. Chem. 262, 6396–6399 47 Rao-Naik, C., delaCruz, W., Laplaza, J.M., Tan, S., Callis, J. and Fisher, A.J. (1998) J. Biol. Chem. 273, 34976–34982 48 Ramelot, T., Cort, J., Yee, A., Semesi, A., Edwards, A., Arrowsmith, C. and Kennedy, M. (2003) J. Struct. Funct. Genomics 4, 25–30 49 Paz, Y., Elazar, Z. and Fass, D. (2000) J. Biol. Chem. 275, 25445–25450 50 Bayer, P., Arndt, A., Metzger, S., Mahajan, R., Melchior, F., Jaenicke, R. and Becker, J. (1998) J. Mol. Biol. 280, 275–286

Received 10 January 2005

 C 2005

Biochemical Society

399