The activity of the ubiquitin-dependent proteolytic system in differentiated tissues under basal conditions remains poorly explored. We measured rates of ubiquitination in rat tissue extracts. Accumulation of ubiquitinated proteins increased in the presence of ubiquitin aldehyde, indicating that deubiquitinating enzymes can regulate ubiquitination. Rates of ubiquitination varied fourfold, with the highest rate in the testis. We tested whether ubiquitin-activating enzyme (E1) or ubiquitin-conjugating enzymes (E2s) could be limiting for conjugation. Immunodepletion of the E2s UBC2 or UBC4 lowered rates of conjugation similarly. Supplementation of extracts with excess UBC2 or UBC4, but not E1, stimulated conjugation. However, UBC2-stimulated rates of ubiquitination still differed among tissues, indicating that tissue differences in E3s or substrate availability may also be rate controlling. UBC2 and UBC4 stimulated conjugation half-maximally at concentrations of 10–50 and 28–44 nM, respectively. Endogenous tissue levels of UBC2, but not UBC4, appeared saturating for conjugation, suggesting that in vivo modulation of UBC4 levels can likely control ubiquitin conjugation. Thus the pool of ubiquitin conjugates and therefore the rate of degradation of proteins by this system may be controlled by E2s, E3s, and isopeptidases. The regulation of the ubiquitin pathway appears complex, but precise.
- deubiquitinating enzymes
the ubiquitin-dependent proteolytic pathway is the major nonlysosomal degradative system in the eukaryotic cell (10). In this proteolytic pathway, ubiquitin, an 8-kDa polypeptide, is ligated via its COOH-terminus to ε-amino groups of lysine residues on target proteins. Additional ubiquitin moieties can be coupled by similar isopeptide linkages to lysine residues of the previously attached ubiquitin moiety to form polyubiquitin chains linked to the substrate. This branched-chain polyubiquitination marks proteins for degradation by a 1,500-kDa multisubunit protease, the 26S proteasome (2). Conjugation of ubiquitin involves a sequence of reactions (11). Ubiquitin is first activated by ubiquitin-activating enzyme (E1) in an ATP-hydrolyzing reaction that initially results in formation of a ubiquitin adenylate intermediate bound to the enzyme. The ubiquitin is then transferred to the active site cysteine in the form of a thiolester linkage, and the AMP is released (7). E1 then transfers the activated ubiquitin to one of a family of ubiquitin-conjugating enzymes (E2), which also forms a thiolester linkage between the active site cysteine and the ubiquitin (8). E2s then support conjugation to substrates in conjunction with a third protein, ubiquitin-protein ligase (E3). E3s recognize specific substrates (21, 28). Individual E2s appear to act with specific E3s and therefore also have an important though indirect role in substrate selectivity. The E2s either transfer ubiquitin to the cysteine residue of specific E3s (24) or ubiquitinate substrate when bound to the E3 as an E2-E3-substrate complex (23). The ubiquitin-protein conjugate can either be deconjugated of ubiquitin by isopeptidases (29) or degraded.
Because only selected proteins are recognized by E3s and targeted for degradation in this pathway, this degradative system performs specific intracellular functions. Indeed, a growing body of literature has now identified numerous functions for this pathway, which include regulation of the cell cycle (14), DNA repair (13), antigen processing (17), and mediation of the inflammatory response (20). Many of these functions would take place in specialized cells under particular conditions. However, the activities of this pathway in differentiated tissues in basal states remain unclear. Furthermore, the relative contributions of different E2s to this basal activity of the ubiquitin system remain undefined.
The mechanisms of regulation of this pathway also remain unclear. Previous studies have shown regulation of polyubiquitin genes and proteasome subunits (18), but the effects of these changes on actual flux through the pathway remain unknown. A simple model of regulation would be one in which degradation by the proteasome is dependent on the availability of its substrate, ubiquitinated proteins (4). The availability of ubiquitinated proteins is in turn dependent on the balance between the rate of conjugation and the rate of deubiquitination. We had previously hypothesized that the conjugation reaction mediated by E2s and E3s may be a rate-controlling step in this proteolytic pathway (31). In support of this, we have observed regulation of various E2s. For example, the 14-kDa E2 (E214k) gene expression in skeletal muscle increases when ATP-dependent proteolysis is activated (31) and can be regulated by insulin and insulin-like growth factor I (IGF-I; see Ref.32). The 20-kDa E2 (E220k) is regulated during erythroid development (27). Finally, there is induction of UBC4 isoforms at various stages of spermatid development (22). Both of these developmental processes are characterized by high levels of protein degradation as the cells mature to their highly specialized forms. E3s are less well characterized. Nonetheless, there is some evidence of regulation at this level. For example, phosphorylation of the cyclosome (anaphase-promoting complex) appears to be required for its activation to a form that can degrade mitotic cyclins (15). In addition, regulated expression of the F-box substrate recognition component of the SCF family of E3s permits temporal regulation of ubiquitination of substrates identified by these F-box proteins. The ability of isopeptidases to control levels of ubiquitinated proteins remains poorly explored. However, overexpression of a deubiquitinating enzyme can downregulate the size of the pool of ubiquitinated proteins (19), and inactivation of the gene encoding a deubiquitinating enzyme has increased the pool (16).
To explore the mechanisms for controlling ubiquitin conjugation more carefully, we have measured rates of ubiquitination of proteins in extracts prepared from different tissues. It has previously been shown that the E214k isoform of the UBC2 subfamily of E2s (5) and UBC4 (3, 35) are capable of supporting conjugation to endogenous proteins in extracts. Therefore, we have determined the relative roles that the UBC2 and UBC4 families of E2s play in this basal conjugation. Furthermore, we have determined whether E2 activities may be limiting for conjugation and have explored whether deubiquitinating activities can also play a role in regulation of steady-state levels of ubiquitinated proteins.
MATERIALS AND METHODS
Preparation of tissue extracts.
Male Sprague-Dawley rats (Charles River Laboratories) weighing 175–200 g were fed ad libitum. After death of the rats by decapitation under anesthesia, the indicated tissues were isolated quickly and homogenized using a Potter-Elvehjem homogenizer (Polytron tissue disruptor for heart and skeletal muscles) in 50 mM Tris, pH 7.5, 1 mM dithiothreitol (DTT), 0.25 M sucrose, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 10 μg/ml pepstatin A, and 10 μg/ml leupeptin at 4°C (5 ml/g tissue). The homogenates were centrifuged at 10,000 g for 10 min, and the supernatant was centrifuged at 100,000 g for 60 min. The final supernatants were stored at −80°C until use.
5′-Adenylylimidodiphosphate (AMPPNP) was obtained from Roche. Ubiquitin aldehyde was synthesized as previously described (30). All other reagents were obtained from Sigma. Ubiquitin was iodinated to a specific activity of >3,000 counts · min−1(cpm) · pmol−1 by the chloramine-T method, as previously described (1). E1 was purified from rabbit liver by ubiquitin affinity chromatography (5). Recombinant UBC4–11 was prepared and purified as previously described (35). MG-132 was a gift from Proscript.
Recombinant E214k isoform2 of UBC2 was expressed in Escherichia coli from a pET11d expression vector. Typically, 2.4 liters of E. coliBL21(DE3) bearing the plasmid were grown at 37°C until an optical density at 660 nm of 0.8, at which time the temperature was lowered to 30°C and 1 mM isopropyl β-d-thiogalactopyranoside was added to induce expression over 90 min. Cells were then harvested by centrifugation, and the cell pellets were frozen at −20°C. The lysate was prepared by resuspending the cell pellets in purification buffer [50 mM Tris, pH 7.5 (4°C), and 1 mM DTT] and 5 mM EDTA in 1/10 the volume of the original bacterial culture followed by sonication. The lysate was then clarified by centrifugation at 100,000 g for 1 h, and the supernatant was subjected to ammonium sulfate (ICN Ultrapure) fractionation. The 45–70% ammonium sulfate pellet fraction was resuspended in 1 vol of purification buffer containing 1.8 M ammonium sulfate and then was loaded on a phenyl-Sepharose (Waters ProteinPak Phenyl 5PW 0.8 × 7.5 cm) column equilibrated in the same solution. The column was eluted with a linear gradient of purification buffer containing 1.8–0 M ammonium sulfate over 90 min at a flow rate of 1 ml/min. Fractions containing activity (eluting at ∼0.9 M ammonium sulfate) were dialyzed against purification buffer and then resolved by quaternary amine anion exchange chromatography (Amersham Pharmacia Biotech MonoQ HR 5/5 column). The column was eluted with buffer containing a 0–0.5 M NaCl gradient over 50 min at a flow rate of 1 ml/min. Fractions containing activity (eluting at ∼0.2 M NaCl) were concentrated and applied to a gel filtration column (Amersham Pharmacia Biotech Superdex75 column) equilibrated and eluted with purification buffer containing 0.1 M NaCl. Fractions containing activity were found to contain E214k purified to homogeneity as assessed by SDS-PAGE followed by staining with Coomassie blue.
Purified E1 and E2 activities were quantitated either by burst assays in which the release of [32P]pyrophosphate is measured after incubation with ubiquitin and [γ-32P]ATP (9) or by thiolester assays in which the incorporation of125I-labeled ubiquitin into E1 and E2s is measured (11). Protein was measured by the Bradford assay using BSA as a standard. Rates of ubiquitination of tissue proteins were determined by incubation of extracts containing 50 μg of tissue protein with 50 mM Tris (pH 7.5), 1 mM DTT, 2 mM MgCl2, 2 mM AMPPNP, and 5 μM 125I-ubiquitin (>3,000 cpm/pmol) in a total volume of 20 μl at 37°C. Reaction times were 10 min for brain, lung, heart, liver, and testis extracts and 20 min for kidney, gastrocnemius, and tibialis anterior extracts. Pilot studies determined that conjugation rates were linear for these time periods and that the concentration of exogenous labeled ubiquitin was in significant excess of any endogenous ubiquitin (i.e., addition of larger amounts of radiolabeled ubiquitin did not increase rates of ubiquitination). In most reactions, the deubiquitinating enzyme inhibitor, ubiquitin aldehyde, was added, since data (Fig. 1) demonstrated that the ubiquitinated protein products in the reactions were being disassembled by these deubiquitinating enzymes. Pilot studies were done to determine the concentrations of ubiquitin aldehyde that provided maximal accumulation of ubiquitinated proteins (1.5 μM for the gastrocnemius and liver, 6 μM for the testis, and 3 μM for all other tissues). After incubation, conjugated ubiquitin was resolved from free ubiquitin by SDS-PAGE on 10% gels and visually confirmed by autoradiography. The gels were stained with Coomassie blue to precisely identify the lanes. After drying of the gel, lanes were cut out to measure the incorporated radioactivity by gamma counting.
Antibodies and immunoprecipitations.
Polyclonal anti-UBC2 and anti-UBC4 antibodies were prepared in rabbits using Freund's adjuvant and were affinity purified on agarose columns (Bio-Rad Affigel) to which purified E214k or the UBC4–1 isoform of the UBC4 family (22) had been coupled. To immunodeplete extracts of the indicated E2s, 2 ml of affinity-purified antibodies or an equivalent quantity (16 μg) of nonimmune IgG (control) was mixed with 30 μl of a 50% slurry of protein A-Sepharose (Pharmacia) and incubated overnight with mixing at 4°C. The next day, the Sepharose beads bound with antibodies were recovered by brief centrifugation at 10,000 g. The pellets were then mixed with 100 μl of tissue extract containing 370 μg protein and were mixed for 4 h at 4°C. After centrifugation at 10,000 g for 1 min, aliquots of the supernatants were assayed for ubiquitination activity as above or blotted for the E2s to confirm efficiency and specificity of the immunoprecipitation. Immunoblotting of the supernatants and pellets confirmed that the antibodies were specific for the particular E2s and were able to immunoprecipitate >95% of the E2s in the extracts. Quantitation of the UBC2 and UBC4 levels in the various tissues was by immunoblotting with the above antibodies. 125I-goat anti-rabbit IgG was used as a secondary antibody, and the signals were quantitated by densitometry. Various quantities of purified E214k and UBC4–1, whose concentrations had been determined by absorbance spectrophotometry, were loaded in parallel and used as standards for the assays. To estimate tissue concentrations of the E2s, the volumes of freshly isolated tissues were measured before fractionation, as described above. From quantitation of total protein in the soluble fraction, the above immunochemical quantitation of E2 levels in the soluble extracts (expressed per mg tissue protein), and estimation of cytoplasmic volume as 50% of total tissue volume, the tissue concentrations of E2s were determined. The small size of the tibialis anterior muscle did not permit accurate determination of tissue volume.
One-way ANOVA was used to compare means of multiple samples. Apparent Michaelis constants (K m) of E2s were determined by measuring rates of conjugation at various concentrations of the relevant E2 (calculated from summing the endogenous and the supplemented amounts of the E2) and fitting the data to the function f = c +V m[S]/(K m + [S]), where c is the basal rate of conjugation in the absence of that particular E2, [S] is the concentration of the E2, and V m is the rate at saturating concentrations of [S].
Rates of ubiquitination in tissue extracts.
To determine whether different tissues manifest different rates of ubiquitination of proteins, soluble extracts were prepared and incubated with an excess of radiolabeled ubiquitin (Fig. 1). There was an approximately threefold difference among the tissues in rates of accumulation of ubiquitin-protein conjugates. Theoretically, steady-state levels of ubiquitinated proteins can be influenced not only by the rate of conjugation but also by the rate of loss of ubiquitinated proteins, either through degradation by the proteasome or deubiquitination by isopeptidases. Because in our assays we used AMPPNP, which can support activation of ubiquitin by E1 but not degradation by the proteasome (which hydrolyses ATP to ADP), levels of conjugates in our assays are unlikely to be influenced by proteasome activity. Indeed, addition of the proteasome inhibitor MG-132 to the extracts had no significant effect on the rate of accumulation of ubiquitinated proteins (data not shown). We therefore tested the effect of an inhibitor of deubiquitinating enzymes on the rates of ubiquitination in the different tissue extracts. Addition of ubiquitin aldehyde, an inhibitor of most deubiquitinating enzymes (12), resulted in accumulation of ubiquitinated proteins, indicating that isopeptidases can play a role in controlling levels of conjugates in these extracts. Thus, to more precisely quantitate the rate of conjugation in these extracts, all subsequent assays were done in the presence of ubiquitin aldehyde (Fig. 1). In the presence of ubiquitin aldehyde, the rate of ubiquitination was highest in the testis and approximately fourfold higher than in the other tissues examined except for the gastrocnemius muscle, which showed an intermediate rate of conjugation.
Relative contributions of UBC2 and UBC4 to the ubiquitination in different tissues.
Previous studies have demonstrated that the E214k isoform of UBC2 (5) and UBC4 (3, 35) are capable of supporting conjugation of ubiquitin in vitro to a broad spectrum of endogenous proteins. To test the relative importance of these two families of E2s in the conjugation of ubiquitin in these tissues, we depleted the extracts of either of the two E2 subfamilies by immunoprecipitation and remeasured the remaining ubiquitination activity in the supernatants (Fig. 2). Because of the limited amounts of purified antibody, four tissues were selected for study. Preliminary experiments showed that >95% of the E2s were specifically depleted by the immunoprecipitation (data not shown). Removal of either of the E2s resulted in similar degrees of inhibition of ubiquitination, suggesting that each of these E2 families contributes to a similar extent to the overall ubiquitination of proteins in these tissues.
Effects of supplementation of extracts with E1 and E2s on the rates of ubiquitination.
Because removal of the E2s lowered rates of ubiquitination, we tested whether the enzymes might be limiting by determining whether supplementing the extracts with enzymes involved in conjugation would increase rates of ubiquitination. The E1 or E214k isoform of UBC2 or UBC4–1 was added to the extracts to determine whether any of these enzymes are limiting for conjugation (Fig.3). Concentrations were chosen that were likely to be in excess based on previous studies. E1 has previously been shown to have K m values of ∼100 pM for E2s, and E214k has been found to have an apparentK m of 60 nM for E3-dependent conjugation in reticulocyte extracts (5). Supplementation of the extracts with 50 nM exogenous E1 did not increase rates of incorporation of ubiquitin, suggesting that E1 is not rate limiting for conjugation. However, addition of 250 nM E2s stimulated conjugation in almost all of the tissues. Supplementation with E214k stimulated conjugation modestly by approximately twofold in all tissues except for the testis. Supplementation with UBC4–1 stimulated conjugation in all tissues. However, UBC4–1 stimulated conjugation more potently than E214k, up to sixfold, and often generated extremely large ubiquitin-protein conjugates that remained in the stacking gel (Fig. 3).
Endogenous levels of UBC2 and UBC4 isoforms in different tissues.
Because these in vitro studies indicated that modulation of E2 levels can control ubiquitination, one possible mechanism for regulating the rate of conjugation is to regulate the levels of the E2s. To evaluate this possibility, we estimated the tissue concentrations of the E2s (Table 1) and the ranges of E2 concentrations capable of stimulating ubiquitination. Levels of UBC2 varied from 200 to 900 nM. Interestingly, levels of UBC4 were approximately less than one-half of those measured for UBC2. In addition, UBC4 levels were more variable, with the lowest levels seen in skeletal muscle (14 nM) and the highest levels (400 nM) seen in the liver. We characterized carefully the dependence of the rates of conjugation on the levels of each of these E2s in several tissue extracts and determined concentrations at which they were able to half-maximally stimulate conjugation in the extracts (Table2). These apparentK m values were similar in various tissues, varying from 9.8 to 50 nM for UBC2 and 28–44 nM for UBC4–1. Thus, in all tissues, UBC2 levels appear to be severalfold above the apparent K m values. However, UBC4 levels appear to be saturating in the testis and the liver, but they are below theK m in skeletal muscle and only approximately threefold higher than K m in the heart.
A simple model for regulation of the ubiquitin system has been proposed previously. In this model, flux through the pathway depends on the availability of ubiquitinated proteins for degradation by the 26S proteasome (4). The pool of ubiquitinated proteins available for degradation by the proteasome is determined by the balance between rates of ubiquitination and rates of deubiquitination. Rates of ubiquitination can be theoretically modulated by the availability of substrates or by the activities of the various enzymes involved in the ligation of ubiquitin to target proteins. In this study, we have, for the first time, explored rates of ubiquitination in extracts from various tissues, quantitated tissue levels of two major E2s, and evaluated the potential of various enzymes involved in ubiquitination to regulate these rates. In so doing, we have made interesting observations regarding potential mechanisms of regulation of ubiquitination.
First, we have observed that rates of ubiquitination were similar among most tissues examined. The obvious exception was the testis, where a threefold higher rate was observed. In contrast to the other tissues, which are terminally differentiated tissues, the testis contains cells that are actively proliferating and differentiating. The higher rates in the testis correlate well with the large number of proteins that are lost during the developmental maturation of haploid spermatids, which occurs in this tissue (33). Indeed, that the rate is related to the developmental maturation of spermatids is supported by our previous observations that the rate of ubiquitination is regulated, being lower in testis from very young animals in which the testis does not yet contain haploid spermatids (22).
Second, deubiquitinating enzymes are very active in cellular extracts, since the rate of accumulation of ubiquitinated proteins increased approximately twofold upon their inhibition. The ubiquitinated proteins that accumulated upon inhibition were unlikely to be polyubiquitin chain products arising after proteasome degradation, since our assays were done with AMPPNP, which does not support proteasome-mediated hydrolysis, and a proteasome inhibitor had little effect on the rate of accumulation of ubiquitinated proteins (data not shown). Thus isopeptidases can have an important role in modulating the steady-state levels of ubiquitinated proteins and thereby also influence the rate of degradation of those proteins.
Third, we have explored the ability of various components of the conjugation system to control ubiquitination. Supplementation of tissue extracts with additional E1 did not enhance conjugation, arguing that this is not an important locus of regulation of conjugation. This is not surprising, since E1 is a common element in all pathways of conjugation and therefore would not permit regulation of specific functions of ubiquitin. Furthermore, E1 is a highly active enzyme, capable of charging excess amounts of E2s with ubiquitin (5). Instead, our data argue that a locus of regulation in the ubiquitin system lies at the E2 level. Manipulating levels of E2s in the extracts up- and downregulated rates of ubiquitination in a similar fashion (Figs. 2 and 3). Our data (Fig. 2) showed that UBC2 and UBC4 isoforms are responsible for similar amounts of ubiquitination in brain, heart, liver, and skeletal muscle. This is in contrast to our previous observations in the testis (22) and other observations in yeast (25) that indicate that UBC4 isoforms are responsible for most of the ubiquitination in those cells. Thus cells with higher rates of protein turnover, such as developing germ cells and proliferating yeast cells, appear to use more UBC4-dependent pathways of proteolysis. For the first time, we have estimated tissue levels of these two major E2 families and found that UBC2 levels are in the 200-800 nM range and that UBC4 levels are in the 10-400 nM range (Table 1). Our findings indicate that concentrations at which there is half-maximal stimulation by each of these E2s is generally in the 10-50 nM range. This would be consistent with the previously estimated apparent K m of E214k for reticulocyte E3 of ∼60 nM (6, 34). These findings would argue that UBC2 is probably saturating in vivo but that UBC4 levels can be subsaturating and can be involved in physiological modulation of rates of ubiquitin conjugation. High, potentially saturating levels of UBC4 were observed in the liver, kidney, and testis. However, our testis samples were derived from adult rats whose testes contain predominantly spermatids. In our previous work (22), we have shown that UBC4 levels are relatively low in spermatogonia and early spermatocytes but are markedly induced in spermatids. These changes in UBC4 levels correlate with changes in the rates of conjugation in the testis during development. Thus these observations would argue that the regulation we have observed in the in vitro extracts with UBC4 is likely to be relevant in vivo.
The maximal rates of ubiquitination seen upon UBC4 supplementation were similar between different tissues (Fig. 3). However, the maximal rates of ubiquitination upon UBC2 supplementation were still variable between tissues. This suggests that other factors can limit conjugation, such as the presence of different levels of substrate availability for UBC2-dependent E3s, or that regulation of conjugation can also take place at the level of the E3s. Possibly different tissues express alternative E3s with different affinities for charged E2s and/or different levels of E3s. Indeed, in muscle extracts, it has been observed that conjugation can be stimulated by supplementation of either E214k and/or E3α (26). As more E3s become identified and available as reagents, this possible involvement of E3s in controlling rates of ubiquitination will be testable on a wider range of E3s.
Ultimately, experiments in which activities of these components are deliberately manipulated in vivo by genetic methods will be required to confirm that modulation of these activities do regulate the pool size of ubiquitinated proteins and the rate of protein degradation. Initial reports are supportive of this. As described earlier, manipulation of levels of a deubiquitinating enzyme has led to differences in levels of ubiquitinated proteins (16, 19). Although the majority of E2 enzymes have likely been identified, the bulk of E3 isoforms and deubiquitinating enzymes likely remains to be discovered. When it is considered that each of these enzymes is a potential target of regulation and that controlling substrate availability is also clearly a mechanism of regulation, it becomes evident that modulation of the ubiquitin system is likely to be both extremely complex and precise.
We are grateful to Aaron Ciechanover for providing a protocol for the synthesis of ubiquitin aldehyde, to Arthur Haas for providing the plasmid expressing E214k and a protocol for purification of this E2, and to Proscript for a gift of MG-132.
This work was supported by the Canadian Institutes of Health Research (MT-12121) and initially by the Marvin M. Burke Grant from the Canadian Diabetes Association.
Present address for V. Rajapurohitam: Dept. of Pharmacology and Toxicology, University of Western Ontario, London, Ontario, Canada.
↵1 UBC4–1 (GenBank accession number U13177) was previously referred to as the 2E isoform of E217KB(35) but was renamed to conform to the current practice of naming E2s after their apparent yeast homolog.
Address for reprint requests and other correspondence: Polypeptide Laboratory, McGill Univ., Strathcona Anatomy and Dentistry Bldg, Rm. W315, 3640 University St., Montreal, Quebec, Canada H3A 2B2 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2002 the American Physiological Society