A major ubiquitin conjugation system in wheat germ extracts involves a 15-kDa ubiquitin-conjugating enzyme (E2) homologous to the yeast UBC4/UBC5 gene products.

In eukaryotes, conjugation of ubiquitin to proteins serves as a committed step for intracellular protein degradation. Formation of ubiquitin-protein conjugates involves the transfer of ubiquitin-conjugating enzyme (E2)-bound ubiquitin to the target proteins with or without the assistance of ubiquitin-protein ligase (E3). We report the isolation and characterization of an E2 purified from wheat germ that accounts for the majority of ubiquitin conjugation activity observed in vitro. This E2 is basic, has an apparent molecular mass of 15 kDa, and forms oligomers that dissociate upon treatment with sulfhydryl reducing agents. E(2)15kDa will not work alone in vitro but requires an additional factor putatively identified as an E3 for substrate recognition. This E3 is distinct from E3 alpha previously described to be required for N-terminal recognition of target proteins. Partial amino acid sequence analysis of E(2)15kDa revealed a substantial identity (approximately 80% in two peptide regions) with yeast E2s encoded by UBC4/UBC5 genes. This homology was confirmed by immunodetection of a 16-kDa yeast protein corresponding to the molecular mass of the UBC4/UBC5 proteins with E(2)15kDa antisera. The products of yeast UBC4 and UBC5 genes along with that of UBC1 gene constitute a subfamily of functionally overlapping E2s that mediate the selective degradation of short-lived and abnormal proteins in vivo. Considering the high degree of functional and structural similarity of wheat E(2)15kDa with that of yeast UBC4/UBC5, it is likely that yeast UBC4/UBC5 and their homologs from other eukaryotes exhibit the same E3 dependence in performing their roles in protein degradation.


A Major Ubiquitin Conjugation System in Wheat Germ Extracts
Involves a 15-kDa Ubiquitin-conjugating Enzyme (E2) Homologous  In eukaryotes, conjugation of ubiquitin to proteins serves as a committed step for intracellular protein degradation. Formation of ubiquitin-protein conjugates involves the transfer of ubiquitin-conjugating enzyme (E2)-bound ubiquitin to the target proteins with or without the assistance of ubiquitin-protein ligase (E3). We report the isolation and characterization of an E2 purified from wheat germ that accounts for the majority of ubiquitin conjugation activity observed in vitro. This E2 is basic, has an apparent molecular mass of 15 kDa, and forms oligomers that dissociate upon treatment with sulfhydryl reducing agents. E216kDa will not work alone in vitro but requires an additional factor putatively identified as an E3 for substrate recognition. This E3 is distinct from E3a previously described to be required for N-terminal recognition of target proteins. Partial amino acid sequence analysis of E216 kDa revealed a substantial identity (-80% in two peptide regions) with yeast E2s encoded by UBC4/UBC5 genes. This homology was confirmed by immunodetection of a 16-kDa yeast protein corresponding to the molecular mass of the UBC4/ UBC5 proteins with Ezl6kDa antisera. The products of yeast UBC4 and UBCS genes along with that of UBCl gene constitute a subfamily of functionally overlapping E2s that mediate the selective degradation of shortlived and abnormal proteins in vivo. Considering the high degree of functional and structural similarity of wheat E216k~,, with that of yeast UBC4/UBC5, it is likely that yeast UBC4/UBC5 and their homologs from other eukaryotes exhibit the same E3 dependence in performing their roles in protein degradation.
Selective degradation of proteins is essential for controlling the levels of key enzymes and regulatory proteins as well as preventing the accumulation of abnormal proteins in cells. In eukaryotes, the ubiquitin-dependent proteolytic system plays a n important role in this process (1)(2)(3). Mutational analysis has revealed that most abnormal and short-lived proteins are . I TO whom correspondence should be addressed: Dept. of Horticulture, University of Wisconsin-Madison, 1575 Linden Dr., Madison, WI 53706. Tel.: 608-262-8215; Fax: 608-262-4743. degraded by this mechanism (2,4,5). Examples of important cell regulatory proteins degraded, at least in part, by the ubiquitin pathway include phytochrome after photoconversion to the far red-light absorbing form (6), tumor suppressor p53 (7), cyclins (8), and MATcu2 repressor (9). In thispathway, ubiquitin functions by becoming covalently attached to proteins, committing them for breakdown. Attachment is through an isopeptide linkage between the C-terminal glycine of ubiquitin and the t-amino group of lysine residues within the target protein. Ubiquitin also becomes conjugated to itself forming multi-ubiquitin chains attached internally through ubiquitin residue Lys4' (10). The multi-ubiquitinated conjugates are then recognized by an ATP-requiring protease complex, which degrades the target proteins and releases ubiquitin in a free, functional form (10-12).
Ubiquitin ligation is accomplished by an ATP-dependent multi-step pathway consisting of ubiquitin-activating enzyme (El)' and members of the ubiquitin-conjugating enzyme (E2) family (1-3), Some E2s function alone, at least in vitro, in ligating ubiquitin, whereas others require the presence of a ubiquitin-protein ligase (E3) for protein substrate recognition. Multiple species of E2s have been isolated from yeast (2,13), mammals (14,15), and plants ( [16][17][18][19]; some have been grouped into subfamilies of enzymes with related functions but potentially different specificities (2). The best characterized subfamily includes yeast UBC1, UBC4, and UBCS, which are required for most of the ubiquitin-dependent degradation of abnormal and short-lived proteins in this organism (4,5). Disruption of all three genes is lethal, demonstrating that these E2s serve an essential function in yeast. A gene from Drosophila encoding an E2 that is structurally and functionally equivalent to UBC4/UBC5, has recently been reported, suggesting that this E2 class is present in all eukaryotes (20).
Among the E2s characterized thus far, only the E2 encoded by RAD6 from yeast and its homolog from rabbit reticulocytes have been shown to be involved in E3-mediated ubiquitination and degradation of proteins in vitro and in vivo (21)(22)(23). These E2s work in concert with a specific E3, designated E3a (also known as E3 type I and I1 or UBRI in yeast), in recognizing protein substrates bearing basic or bulky hydrophobic Nterminal residues (23)(24)(25)(26). In contrast, yeast UBCl does not function with E3a (2) and UBC4/UBC5 have been shown to be involved in degradation of protein substrates even in an ubr I null mutant (27). Another E3, designated E3@, also functions in N-terminal recognition, but its interaction with The abbreviations used are: E l , ubiquitin-activating enzyme; E2, ubiquitin-conjugating enzyme; E3, ubiquitin-prot.ein ligase; PAGE, polyacrylamide gel electrophoresis; DTE, dithioerythritol; PMSF, phenylmethylsulfonyl fluoride; IgG, immunoglobulin G; HPLC, high performance liquid chromatography; PVDF, polyvinylidene difluoride.

A Major
Ubiguitin Conjugation System in Wheat specific E2s is currently unclear (28).
In the present study, we report the isolation of a 15-kDa E2 from wheat that accounts for most of the conjugating activity in wheat germ extracts. It depends on an activity we propose is an E3 and shows a high degree of sequence identity to the yeast UBC4 and UBC5 gene products (4, 5 ) . Like the yeast enzymes, EZ15 kDa does not mediate E3a-dependent ubiquitination of proteins in vitro, but functions with distinct E3s. This strengthens the hypothesis that individual E3s interact specifically with different E2s in recognizing various types of protein substrates.

EXPERIMENTAL PROCEDURES
Materials-Non-toasted wheat (Triticum aestiuum) germ was purchased from General Mills (Minneapolis, MN). Purified human ubiquitin and purified hen egg white lysozyme (Calbiochem) were radiolabeled with Na'*'I as described (29). El was isolated from wheat germ extracts by covalent affinity chromatography with ubiquitin-Sepharose using the method described in Ref. 30  Purification of E215kD,-Wheat germ (200 g) was ground in mortar and pestle at liquid Nz temperatures and homogenized in 1 liter of 50 mM HEPES-KOH (pH 6.8) containing 2 mM DTE (Buffer A). The homogenate was centrifuged a t 20,000 X g for 20 min a t 4 "C, and the resulting supernatant was filtered through Whatman No. 3MM paper and centrifuged a t 20,000 X g for another 20 min. The clarified supernatant was applied to a 50 ml S-Sepharose (Pharmacia) column equilibrated in Buffer A. The column was washed with 20 mM NaCl in Buffer A and proteins were eluted with 50 mM NaCI. The S-Sepharose eluate was made 1 M in NaCl and passed through a 10-ml phenyl-Sepharose (Pharmacia) column equilibrated in Buffer A + 1 M NaCI. The flow-through material was concentrated to 8 ml by ultrafiltration using an Amicon cell and subjected to gel filtration chromatography using a 100 X 2-cm Sephadex G-75 (Pharmacia) column equilibrated in Buffer A + 1 M NaCl and operated at a flow rate of 50 ml/h. The active fractions (as judged by their ability to restore ubiquitin-conjugating activity to the flow-through material from an S-Sepharose column) were pooled, dialyzed against Buffer A, and applied to a 10-ml S-Sepharose column. Proteins were eluted with a linear 0-75 mM NaCl gradient. Active fractions were pooled and dialyzed against Buffer A. The dialysate was made 2 mM PMSF, 2 mM ATP, 5 mM MgCI2, 10 mM creatine phosphate (pH adjusted to 7.0; 25 "C) and clarified by centrifugation. Twenty units of phophocreatine kinase and 50 pg of purified wheat El were added to the supernatant. The solution was applied a t room temperature to a 10ml column of bovine ubiquitin coupled to Affi-Gel 10 (Bio-Rad). El and E2 were eluted from the ubiquitin affinity column with 25 mM Tris-C1 (pH 8.5) containing 10 mM DTE. The eluate was collected on ice and concentrated by ultrafiltration. Proteins in the DTE eluate were resolved by HPLC on a Spherogel TSK SEC400 size exclusion column (Bio-Rad) equilibrated in Buffer A + 150 mM NaCl using a flow rate of 1 ml/min. The peak of E2 activity was concentrated by ultrafiltration with a Centricon-10 microconcentrator (Amicon) and the buffer exchanged for Buffer A.
Preparative SDS-PAGE of E215 k~~" 5 0 0 pg of purified E215 kDa were separated by SDS-PAGE on a 16% polyacrylamide gel pre-electrophoresed with 0.1 mM thioglycolic acid. The gel was immersed in 4 M sodium acetate to detect the resolved proteins. The band containing E21,5 kDa was excised, rinsed thoroughly with distilled water, and the protein was electroeluted. E215 kDa was concentrated with a Centricon-10 microconcentrator and the buffer exchanged for 50 mM Na2HP04 (pH 6.8) containing 2 mM DTE.
Peptide Sequencing--60 pg of SDS-PAGE purified E~I S L D * were resuspended in 50 pl of 50 mM Na2HPOr (pH 7.8) and cleaved by Staphylococcus aureus V8 protease for 24 h a t 37 "C with an enzyme t o protein ratio of 1:25 (w/w). Peptides were separated by HPLC using a Vydac C4 reversed-phase column and a linear 0-60% acetonitrile gradient in 0.1% trifluoroacetic acid over 45 min a t a flow rate of 0.8 ml/min. Peptides were collected and vacuum-dried. Two peptides were subjected to Edman degradation using an Applied Biosystems 470A Protein Sequencer followed by HPLC separation of the phenylthiohydantoin-derivatives.
Immunological Procedures-Antisera were generated in BALB/c mice against E215k~. purified by preparative SDS-PAGE. Twenty pg of protein dissolved in 50 mM phosphate buffer (pH 6.8) and emulsified with an equal volume of complete Freund's adjuvant were used for primary immunization of each mouse. Four weeks later, booster immunizations of 10 or 20 pg with incomplete Freund's adjuvant were carried out, followed by an additional booster immunization 3 weeks later. Sera were collected 2 weeks after the third injection.
For immunoblot analysis, proteins were fractionated by SDS-PAGE and electroblotted onto PVDF membranes. Membranes were probed with a 1/250 dilution of either preimmune or immune antiserum. Immunoreactive proteins were detected using alkaline phosphatase-linked goat anti-mouse IgGs in conjunction with the substrates nitro blue tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate as described (31).
For analysis of yeast proteins, Saccharomyces cerevisiae was grown for 24 h in YPD medium (32). Cells were harvested, resuspended in 100 mM Tris-C1 (pH 7.5), 10 mM Na,EDTA, 0.5 mM PMSF, and vortexed with an equal volume of glass beads for 5 min. The extract was clarified by centrifugation and the supernatant used as the source of proteins.
Immunodepletion of E 2 1 5 k D . from crude wheat germ extracts was achieved by repeated passages (10 times) of 0.5 ml of wheat germ crude extract through E2,, tDa IgG column. The column was prepared by coupling 400 &I of antiserum to 0.5 ml of protein A-Sepharose using bis(sulfosuccinimidy1) suberate (Pierce Chemical Co.) (33). Adsorbed proteins were washed with 150 mM NaCl in 20 mM Na2HP04 (pH 7.0), and eluted with 1 M NaCl in the same buffer. The eluant buffer was exchanged with Buffer A by ultrafiltration and concentrated to a volume of 500 p1. Control experiments used anti-AtUBCl IgGs (17) immobilized to protein A-Sepharose under the same conditions.
Preparation of E3-A crude extract containing -130 mg of protein from 5 g of wheat germ was prepared as described for the purification of E215k~a and applied to a 10-ml Q-Sepharose (Pharmacia LKB Biotechnology Inc.) column equilibrated in Buffer A. Proteins were eluted with a 30-ml linear 0-750 mM NaCl gradient in Buffer A. Twoml fractions were collected and concentrated by ultrafiltration (Centricon) to a volume of 100 p l and the buffer exchanged for Buffer A. A similar crude extract was applied to a 15-ml S-Sepharose column, and proteins were eluted according to the procedure described above for the anion exchange chromatography. Fractions stimulating Eland E215k~,.-dependent conjugation were pooled and further fractionated by HPLC on a Spherogel TSK SEC400 size exclusion column equilibrated in Buffer A + 150 mM NaCl using a flow rate of 1 ml/ min. Collected fractions were concentrated by ultrafiltration (Centricon) to a volume of 100 ~l and the buffer exchanged for Buffer A. The protein concentration was determined by Bradford assay. Two p1 of each fraction was used in conjugation assays.
Assay for Ubiquitin Thiol Ester and Conjugate Formation-Formation of thiol ester adducts between '251-ubiquitin and El (0.2 Gg) and the various E2s was carried out as described (30). The reactions were quenched by adding SDS-PAGE sample buffer with or without 2-mercaptoethanol and boiling the reaction mixtures for 3 min (30). Products of the reaction were separated by SDS-PAGE and visualized by autoradiography.
Conjugation of '251-ubiquitin to proteins was assayed in a 20-pl total reaction volume containing 0.35 pg of '251-ubiquitin, 10 mM creatine phosphate, 2 mM MgC12, 1 mM ATP, 2 mM DTE, and 0.1 unit of phosphocreatine kinase in 50 mM HEPES-KOH (pH 6.8). For 1251-lysozyme conjugation, reaction mixtures contained 5 pg of unlabeled ubiquitin (bovine) and 0.5 pg of 1Z51-lysozyme. The source of ubiquitin-conjugating enzymes used in the different assays were as follows. (i) For the crude extract, 4 pl of clarified crude extract was used. (ii) The reconstituted system with E215kDa contained 4 pl of flow-through material after passage of crude extract through the S-Sepharose column and 0.2 pg of E215k~a. (iii) The reconstituted system with E3 contained 0.2 pg of purified El, 0.2-0.5 pg of the respective purified E2s, and partially purified E3 used a t a volume equivalent to 190 p1 of crude extract (see "Preparation of E3"). (iv) The reconstituted system with E3n contained 0.2 pg of purified El, 0.2-0.5 pg of the respective purified E2s, and partially purified E3a at a concentration of 1.6 microunits (as defined in Ref. 28) per assay. Equivalent amounts of various E2s used in the assay were determined by using thiol ester activity. Reaction mixtures were incubated a t 30 "C for 90 min, terminated by boiling in SDS-PAGE sample buffer containing 2-mercaptoethanol, and subjected to SDS-PAGE (34). The gels were stained with Coomassie Blue and dried between cellophane. Radioactive protein bands were visualized by autoradiography. For quantitative evaluation of conjugate formation using either "'1-ubiquitin or 12SI-lysozyme, regions of the gels containing high molecular mass conjugates (>45 kDa) were excised and the radioactivity determined by scintillation spectroscopy.

RESULTS
Crude wheat germ extracts provide an active system for the ATP-dependent conjugation of ubiquitin to endogenous and purified substrates (30). However, upon fractionation of such crude extracts by cation-exchange chromatography using S-Sepharose, the conjugation activity of the unadsorbed material was severely impaired (Fig. l), even though it contained El and a multitude of E2s (Ref. 16 and data not shown). Conjugation of ubiquitin to endogenous protein and to ' *' Ilysozyme was reduced by 64 and 80%, respectively, indicating that a basic component essential for the in vitro ligation system (E2 or E3) was removed during this separation step (Fig. 3 B ) .
We attempted to purify this component based on its ability t o restore ubiquitin conjugation activity when added back to the flow-through material from crude extracts after passage through the S-Sepharose column. A combination of cation exchange and size exclusion chromatography, and covalent affinity purification using immobilized ubiquitin, resulted in the isolation of an active fraction containing a 15-kDa protein (Fig. 2 A ) . Conditions for binding (requirement of El and ATP) and elution (DTE) from the ubiquitin affinity column suggested that this protein was an E2. However, protein recovery from the affinity column was uncharacteristically low for an E2, even if the unadsorbed material was again passed through the ubiquitin column, suggesting that binding t o ubiquitin was weak and/or inefficient. Residual protein contaminants were removed from the ubiquitin affinity eluate by size exclusion chromatography. During this step, the 15-kDa protein eluted as a single species with a native molecular mass of -28-30 kDa, suggesting that it was a homodimer. This subunit association required low ionic strength buffers (5150 mM NaCl) and was easily dissociated a t higher  1 ) or absence (lane 2) of 2-mercaptoethanol and then subjected to SDS-PAGE, blotted on to PVDF membrane, and visualized using E2,,kDR antisera. mercaptoethanol in the SDS-PAGE sample buffer), oligomers of the protein up to pentamers were observed (Fig. 2C). This result suggested that denaturation in the absence of the reducing agent induces the formation of intermolecular disulfide bonds not present in the native enzyme.
The 15-kDa wheat protein was confirmed to be an E2 by its ability to form thiol-ester adducts with ubiquitin in an Eland ATP-dependent reaction. Several adducts ranging in size from 27 to 34 kDa were observed following SDS-PAGE under nonreducing conditions (Fig. 2B). Under reducing conditions, the adducts were absent demonstrating that attachment of ubiquitin was via a thiol ester bond (data not shown). The multiple thiol ester adducts observed here for EZIskDe were similar to those observed for several other E2s (16). Sullivan and Vierstra (17) showed that for wheat TuUBC1, these species probably result from artifactual migration of a single adduct in the nonreducing gel system used and not from the simultaneous attachment of multiple ubiquitin moieties.
When added back to the flow-through material of the S-Sepharose column, purified E21RkDa restored its capacity to conjugate ubiquitin to endogenous protein substrates and added lysozyme. The size distribution of conjugates was nearly indistinguishable from that of the crude extract. However, the amount of conjugates formed exceeded that of the original crude extract (Fig. 3, A and B ) . With respect to endogenous substrates, a 2-fold stimulation was obtained, saturating a t relatively low levels of added EZlSkDa. With respect to ' "1lysozyme, a -4-fold stimulation was obtained and was only saturable a t very high levels of E2,'kDa ( Fig. 3B and  munodepleted crude extracts lost the ability to ubiquitinate both endogenous and purified proteins with only a small fraction of the activity remaining in the unadsorbed material ( Fig. 3C and data not shown). This activity could be restored by subsequent addition of purified E215 kDn to the unadsorbed material.
In an attempt to identify the relationship of E 2 ] 5 k D n with other known E2s, a partial amino acid sequence of the protein was determined. This was achieved for two peptides generated from by s. aureus v8 protease digestion. Comparison of these sequences with those of all reported E2s revealed little homology with other wheat and Arabidopsis E2s sequenced to date (17)(18)(19), but a high homology with the yeast UBC4 and UBC.5 gene products (Fig. 4, upper panel; Refs. 4 and 5 ) . For the region encompassing the two peptides, the amino acid sequence of E 2 1 5 k D~ was -80% identical. A lesser FIG. 6. E3a-dependent ubiquitination of protein substrates by E2,, kDe and A. thaliana URC 1, respectively. EA, f;l, and the indicated E2s were incubated with native ubiquitin and "'I-lysozyme (lejt jive lanes) or '?'I-ubiquitin alone (right jive lanes). The conjugation assay contained ATP, 0.2 pg El, 1.6 microunits of fib, and 0.5 pg of AtUBCl or 0.2 pg of E215kDs. Conjugation was performed for 90 min a t 30 "C, and the ubiquitin conjugates were resolved by SDS-PAGE and detected by autoradiography. LYSO, "'I-lysozyme; URQ, "'I-ubiquitin. but significant homology was also observed with the yeast E2 encoded by UBCl (-45% identity). The second peptide aligned with the C terminus of yeast UBC4 and 5, but had an additional glycine as its C-terminal residue. That wheat E21skDa is a homolog of yeast UBC4 and UBC5 E2s was further supported by immunoblotting yeast crude extracts with E 2 1 s k~a antisera. The antisera cross-reacted with a 16-17-kDa protein corresponding to the molecular mass of UBC4 and UBC5 (Fig. 4, lower panel).
UBC4 and UBC5 are essential E2s for the ubiquitin-dependent degradation of abnormal and short-lived proteins in yeast (4,5). Although an E3 has also been proposed to be necessary for the breakdown of such proteolytic substrates (2), the cooperation of UBC4 and UBC5 with an E3 has not yet been demonstrated. With respect to E2,, kDa, we observed that whereas the crude extract reconstituted with E215 kDa did conjugate lysozyme, purified E215kD. and El alone did not recognize lysozyme as a substrate, nor did it catalyze the formation of poly-ubiquitin chains (Fig. 5A). Only monoubiquitin conjugates of El and of E215 kDa were observed. This indicated that another factor was required for lysozyme recognition and ubiquitination, in addition to El and E215 kDa.
To identify the additional component(s) of the E215k~a system required for lysozyme conjugation, the crude extract was fractionated on S-Sepharose and Q-Sepharose columns. T h e recovered protein fractions were assayed for their ability t o form ubiquitin conjugates with proteins endogenous to the fraction or with "51-lysozyme in an El-and E2,5ma-dependent manner. S-Sepharose chromatography resolved one peak of activit,y (designated S) capable of conjugating both endogenous proteins and '2611-lysozyme (data not shown and Fig.  5C). Q-Sepharose chromatography identified two fractions competent to form conjugates with endogenous proteins and two fractions competent to form conjugates with "'I-lysozyme (designated Q1 and Q2). The second Q-Sepharose fraction of each profile (Q2) was coincident and contained the highest E21skDa-dependent specific activity (Fig. 5, B and C). The pattern of lysozyme conjugates formed by E l and E215kDe in cooperation with the fractions S, Ql, and Q2 resembled that of the crude extract except that t.he proportion of intermediate molecular mass products was increased relative to high molecular mass forms (Fig. 5C). Only the Q2 fraction effectively catalyzed the formation of high molecular mass conjugates of endogenous proteins in the presence of E2,, kDa (Fig. 5c). This was particularly striking given that this fraction also had the lowest level of protein contaminants as potential substrates. Attempts to further purify the active factor in Q2 by size exclusion chromatography showed that it had a molecular mass in the 100-150-kDa range.
Based on the criteria used to define E3, i.e. a factor that participates with E1 and E2 in stimulating ubiquitin conjugation (3), it is possible that S, Q1, and Q2 contained an E3. In reactions containing lysozyme, all three fractions exhibited a strong preference for E 2 1 5 k D a ; conjugation was -10-fold higher than with any other E2 examined (data not shown). The possibility that E215k~s interacts with other E3(s) in addition to the putative E3s described above was tested by using E3n, isolated from rabbit reticulocytes (23)(24)(25)(26). E3a (and its yeast counterpart UBR1) is required for recognition of substrate proteins based on the nature of their N-terminal residue (23)(24)(25)(26). It works in concert with a 14-kDa reticulocyte E 2 homologous to the yeast E2 encoded by RAD6 gene (21-23). When used in combination with E3a, E215k~a was -6fold less effective in conjugating ubiquitin to lysozyme or to endogenous proteins (Fig. 6). In fact, many of the conjugates observed with endogenous proteins represented the E3a-independent conjugation of El and Ells kDa (compare reactions with or without E3a). Of all the remaining plant E2s, E3a functioned effectively only with A. thaliana kDa, encoded by the AtUBCl gene. Conjugation was observed not only to 12sII-lysozyme but also to endogenous proteins. Based on its E 2 dependence, we conclude that the putative wheat E3s described here are distinct from E3a. ' M. L. Sullivan and R. D. Vierstra, unpublished observation. DISCUSSION Ubiquitin-dependent proteolysis serves an essential role in eukaryotic intracellular protein degradation. The E2 enzymes encoded by the UBCl/4/5 gene family in yeast have been shown to be essential for eliminating abnormal and shortlived proteins by this pathway (2,4,5 ) . We have identified a 15-kDa E2 from wheat that is closely related to yeast UBC4 and UBC5, based on peptide sequence homology and antibody cross-reactivity. This high degree of structural conservation suggests that E215 kDa has a similar essential cellular function in plants, i.e. mediating the degradation of short-lived and abnormal proteins. In agreement with this, we showed that E21n kDa is necessary for most of the ubiquitin conjugation to endogenous proteins and added lysozyme in wheat germ extracts.
E215 kDa is a basic protein that exists as a homodimer in low salt conditions. When denatured in the absence of reducing agents, the protein freely forms multimers. Conventional protocols for the purification of E2s currently rely on adsorption to and elution from DEAE-columns (termed Fraction 11; see Ref. 3). We find that E216 kDs does not purify with the rest of the wheat E2 family characterized to date by this method, implying that additional E2s may be present in eukaryotes that are not isolated with this conventional E2 purification strategy. One such candidate may be the factor identified by Gonen et al. (36,37) as missing from Fraction I1 but essential for recognition and conjugation of N-acetylated proteins.
Previous studies have proposed that yeast UBC4 and UBC5 work in concert with an E3 (2, 4). In the present study, we provide evidence that E215kDa requires another factor for function that has E3-like activity. Three fractions containing such a factor(s) were partially purified from wheat germ by Sand Q-Sepharose chromatography (S, Q1, and Q2). They appear to be distinct from E3a (and its yeast counterpart UBR1; Refs. [23][24][25] previously identified in rabbit reticulocytes by their ability to work best with E215 kDa in vitro. They are also likely to be different from E30 because of their ability to ubiquitinate lysozyme (28). Based on these criteria, we designate the E3-like activity present in these fractions E3y. But the conclusion that these factors are indeed E3s will require a demonstration that they interact with target proteins (35). The functional and structural similarity of E 2 ] , 5 k D a with UBC4 and UBC5 suggests that all E215kDa homologs display E3 dependence. The E~Y -E~,~ kDa pair is likely to be a major component in the ubiquitin-dependent proteolytic pathway, based on the requirement of UBC4-and UBC5mediated proteolysis for the survival of yeast cells (4).
We do note that E215k~a is also able to function in an E3independent manner to catalyze the direct transfer of ubiquitin to a small group of basic hydrophobic wheat proteins (data not shown). Their molecular masses range from 23 to 27 kDa. They bind to S-Sepharose and elute along with E2],? kDa in the 20-50 mM NaCl fraction, but they can be resolved from E215 kDa by phenyl-Sepharose chromatography. The identity of these substrates is unknown, but it remains possible that their direct recognition by E215 kDa only occurs in vitro.
Recognition of protein substrates by the ubiquitin pathway based on the nature of their N termini (N-End Rule) was recently shown to require the E2 encoded by RAD6 and E3a (or UBRI) (23,25,26). With respect to plant E2s, E30 does not function with E215 kDa, but cooperates instead with another wheat E2, E 2 1 6 k D a encoded by AtUBCl, to form ubiquitin conjugates. Previously reported failure of E216 kDa to substitute for the DNA repair function of RAD6 in rud6 null mutants (17)  A model in which the selectivity of the uhiquitin-dependent proteolytic pathway is controlled at the level of substrate recognition would imply the existence of multiple E3s, each preferentially recognizing specific substrates. In contrast, our results suggest that a wide variety of different proteins are recognized by a limited number of constitutive E2-E3 ligation systems. The data presented here suggest that at least two distinct uhiquitin ligation systems with different specificities exist in wheat germ, one requiring E215kDa and the another requiring E216kDa. Each in turn may require a distinct E3 activity, E37 or E3a respectively, for substrate recognition.