Ubiquitin conjugation by the N-end rule pathway and mRNAs for its components increase in muscles of diabetic rats

Insulin deficiency (e.g., in acute diabetes or fasting) is associated with enhanced protein breakdown in skeletal muscle leading to muscle wasting. Because recent studies have suggested that this increased proteolysis is due to activation of the ubiquitin-proteasome (Ub-proteasome) pathway, we investigated whether diabetes is associated with an increased rate of Ub conjugation to muscle protein. Muscle extracts from streptozotocin-induced insulin-deficient rats contained greater amounts of Ub-conjugated proteins than extracts from control animals and also 40–50% greater rates of conjugation of ¹²⁵I-Ub to endogenous muscle proteins. This enhanced Ub-conjugation occurred mainly through the N-end rule pathway that involves E2_14k and E3α. A specific substrate of this pathway, α-lactalbumin, was ubiquitinated faster in the diabetic extracts, and a dominant negative form of E2_14k inhibited this increase in ubiquitination rates. Both E2_14k and E3α were shown to be rate-limiting for Ub conjugation because adding small amounts of either to extracts stimulated Ub conjugation. Furthermore, mRNA for E2_14k and E3α (but not E1) were elevated 2-fold in muscles from diabetic rats, although no significant increase in E2_14k and E3α content could be detected by immunoblot or activity assays. The simplest interpretation of these results is that small increases in both E2_14k and E3α in muscles of insulin-deficient animals together accelerate Ub conjugation and protein degradation by the N-end rule pathway, the same pathway activated in cancer cachexia, sepsis, and hyperthyroidism.

J. Clin. Invest.104:1411–1420 (1999).

A major feature of low-insulin states, such as diabetes mellitus and fasting, is a marked loss of muscle protein resulting in muscle atrophy (1, 2). In these insulin-deficient states, the amino acids released from muscle protein are used primarily for hepatic glucose production (2). Whereas some of the loss of muscle mass in acute diabetes is due to decreased protein synthesis, the rapid fall in tissue weight results primarily from increased protein degradation (2–5). Streptozotocin administration in rats leads to destruction of pancreatic beta cells and produces a useful model of acute insulin deficiency characterized by high blood glucose levels and ketoacidosis. Protein degradation in muscles isolated from such rats is also elevated (4, 5). Understanding the enzymatic adaptations that account for the enhanced muscle protein breakdown in uncontrolled diabetes and other catabolic diseases may help in the development of therapeutic regimens or pharmacological inhibitors of this excessive protein loss.

The bulk of proteolysis in skeletal muscle and most cultured cells occurs by the ubiquitin-proteasome (Ub-proteasome) pathway (6, 7), although degradation of membrane and organellar proteins within lysosomes also account for 10–20% of the overall protein breakdown (8). In the Ub-proteasome pathway, proteins to be degraded are first marked by covalent attachment of a chain of Ub molecules that targets them for destruction by the 26S proteasome (9). Ub molecules are linked to the protein substrate by a multienzyme pathway (10). The enzyme, E1, first activates Ub by forming a high-energy thiolester linkage between an internal Cys residue of E1 and the COOH-terminal Gly residue of Ub. Ub is transferred to a Ub-carrier protein (E2), which then transfers Ub to a lysine residue of the substrate in a reaction catalyzed by an E3 (Ub-protein ligase). Subsequently, E2 and E3 link additional Ub molecules onto the first Ub through isopeptide linkages to form a polyubiquitin chain. Whereas only a single E1 is present in cells (11), at least a dozen E2s and probably an even larger number of E3s exist (12, 13) to provide substrate specificity, because particular E2/E3 pairs appear to conjugate Ub to specific cell proteins or classes of proteins. Surprisingly, in rat, rabbit, and presumably in other mammals, the E2/E3 pair, E2_14k and E3α, accounts for most of the Ub conjugation activity to endogenous proteins in soluble extracts of skeletal muscle (6). Ub conjugation by this pair of enzymes has been termed the “N-end rule pathway” (14), because the Ub-protein ligase E3α recognizes protein substrates that contain basic and large hydrophobic NH₂-terminal amino acids (14, 15), catalyzing their rapid ubiquitination and degradation.

The enhanced muscle protein breakdown in catabolic conditions seems to occur through activation of the Ub-proteasome pathway (16, 17). For example, in muscles removed from acidotic or septic animals and incubated in vitro, the proteasome inhibitor MG132 lowered protein degradation to control levels (5, 18–20). Many studies have shown increases in mRNA for Ub (5, 8, 21–25), E2_14k (21, 26–29), and certain proteasomal subunits (5, 21, 23, 24, 27–29) in atrophying muscles. However, the increases in the amounts of these proteins in the muscles have not been demonstrated or even been measured in most of these cases. No studies have explored whether the increases in mRNA for Ub, E2_14k, or proteasomal subunits can, in fact, account for the enhanced protein breakdown in atrophying muscles, or if they are just associated adaptations. Perhaps the strongest indication that rates of ubiquitination are generally enhanced in atrophying muscles has been the finding of increased amounts of Ub conjugates in muscle from denervated, septic, and tumor-bearing rats (30–33). Recently, we demonstrated higher rates of Ub conjugation to muscle proteins in extracts from muscles of tumor-bearing, septic, and hyperthyroid animals (34). Surprisingly, enhanced Ub conjugation in these muscles again appeared to occur, in large part, by activation of the N-end rule system (34). However, the biochemical adaptations that activate this pathway have not been investigated.

In rats with acute insulin deficiency, overall protein degradation in muscle also appears to increase primarily through activation of the Ub-proteasome pathway, as suggested by greater sensitivity to proteasome inhibitors and increased mRNA for Ub and the proteasome (5). However, rates of Ub conjugation in diabetic muscles have not been investigated. We demonstrate here that the activation of proteolysis in insulin-deficient animals is associated with accelerated Ub conjugation and that activity of the N-end rule ubiquitination system is enhanced. To clarify how this system may influence proteolysis, we have examined whether its components, E2_14k or E3α, are rate limiting for Ub conjugation and have tested whether the expression of these proteins increases and accounts for the enhanced Ub conjugation.

As reported previously (5), 3 days after streptozotocin treatment, serum glucose in rats averaged 288 ± 12 mg/dL compared with 95 ± 9 mg/dL in the control group (Table 1). The treated animals lost 23% of their initial body weight during these 3 days, and individual epitrochlearis muscles underwent marked atrophy, finally weighing, on average, 19% less than epitrochlearis muscles from control animals (Table 1). To ensure that changes occurring in the muscles were not due to differences in food intake between the 2 groups, the control and streptozotocin-treated animals were pair-fed and fasted for 12 hours before muscle harvest. When the epitrochlearis muscles were isolated and incubated in vitro, those from diabetic rats showed protein degradation rates 40% greater than the controls (Table 1). This increase in protein degradation was largely through a proteasome-dependent pathway because it was markedly inhibited by the proteasome inhibitor, MG132, as well as ATP depletion (5).

Table 1

Comparison of normal and diabetic rats

Ub conjugation is increased in muscle extracts from diabetic rats. To test whether the increased protein degradation in the muscles of streptozotocin-treated rats was associated with an increase in Ub conjugation generally, we measured the content of Ub conjugates in muscles from control and diabetic animals by subjecting soluble muscle extracts to SDS-PAGE and immunoblot analysis using an anti-Ub antibody. The diabetic muscle extract showed greater amounts of high-molecular-weight Ub-conjugated proteins than the control muscles (Figure 1a). Larger amounts of Ub-protein conjugates were also found when whole muscles from diabetic rats were directly solubilized in SDS (data not shown). Thus, the increased amount of Ub conjugates was not the result of extract preparation. Because the increased protein degradation in the diabetic muscles was sensitive to proteasome inhibitors (5), the flux of substrates through the proteasome in these atrophying muscles must be increased, and this accumulation of Ub conjugates cannot be due simply to a failure of the conjugates to be degraded. Therefore, the conjugation of Ub to muscle protein must also be more rapid. This build-up of conjugates in the face of increased protein breakdown suggests that in the streptozotocin-treated animals, proteasome function has become the rate-limiting step in the pathway.

Figure 1

Ub protein conjugates and rates of Ub conjugation are increased in muscles from diabetic rats. (a) Anti-Ub immunoblot. Muscle extract protein (35 μg crude extract) from control or diabetic rats was subjected to 15% SDS-PAGE and probed with a polyclonal anti-Ub antibody at a 1/500 dilution. An increased amount of high-molecular-weight Ub-immunoreactive material (Ub conjugates) was present in the muscles from diabetic animals. (b) ¹²⁵I-Ub conjugation to soluble muscle proteins. Muscle protein fraction II from either control (C) or diabetic (DM) rats was used as the source of both Ub-conjugating enzymes and ubiquitination substrates. Upon addition of AMPPNP and ¹²⁵I-Ub, ¹²⁵I-Ub protein conjugates were formed (lanes 2 and 3). Lanes 4 and 5 are the same as lanes 2 and 3, except for the addition of 2 μM C88S E2_14k. Lanes 6 and 7 are the same as lanes 2 and 3, except for the addition of 2 μM C114S E2-C. Lane 1 has ¹²⁵I-Ub alone. Note that higher amounts of Ub-protein conjugates were formed in the extracts from diabetic muscles and that the dominant negative E2_14k, but not the dominant negative E2-C, inhibited this increase in Ub conjugation. The intensely radioactive bands at ∼25 kD in lanes 4–7 are nonreducible ¹²⁵I-Ub-C88S E2_14k and ¹²⁵I-Ub-C114S E2-C complexes. (c) Ub conjugation to α-lactalbumin. Same as in a, except that ¹²⁵I-Ub was replaced by Ub and ¹²⁵I-α-lactalbumin. Lane 1: ¹²⁵I-α-lactalbumin alone. Increased ¹²⁵I-α-lactalbumin-Ub conjugates were formed in the extracts from diabetic muscles. The dominant negative E2_14k, but not the dominant negative E2-C, again inhibited the enhanced Ub conjugation in the diabetic extracts.

To compare directly rates of Ub conjugation in the muscles of normal and diabetic rats, soluble extracts were prepared and ubiquitination rates measured in vitro as the formation of conjugates between ¹²⁵I-Ub and endogenous extract proteins (6). For these studies we used preparations (fraction II) that contain E1 and many E2s and E3s, including those required for the N-end rule pathway, but this fraction lacks Ub (which does not bind to the column) (43–45). Extracts from the diabetic and control muscles supported Ub-conjugation to muscle proteins, as measured by the ATP-dependent incorporation of ¹²⁵I-Ub into high-molecular-weight protein conjugates (Figure 1b, lane 2). Ub conjugates were formed at a faster rate in the extracts from diabetic animals than in those from controls (Figure 1b, lanes 3 vs. 2). These findings suggest that the acceleration of Ub conjugation is a critical adaptation leading to the enhanced proteolysis in muscles from rats with uncontrolled diabetes.

To test further whether the increase in Ub conjugation to muscle proteins was due to greater activity of the Ub-conjugating enzymes and not to some alteration in the protein substrates in the extracts, we studied Ub conjugation to an exogenously added substrate, ¹²⁵I-human α-lactalbumin (46). This protein has a basic amino terminus (Lys) and therefore is conjugated to Ub by the N-end rule pathway. This ubiquitination system, which involves the enzymes E2_14k and E3α (46), selectively conjugates Ub to proteins with charged or large hydrophobic amino terminal residues (15), and our recent studies have indicated that this system plays a major role in Ub conjugation to proteins in skeletal muscle (6). Approximately twice as much of the ¹²⁵I-α-lactalbumin was conjugated to Ub in the extracts from diabetic rats than in the controls (Figure 1c, lanes 3 vs. 2). More rapid ubiquitination of this exogenous substrate suggests that the N-end rule pathway is activated in these atrophying muscles.

To test further whether the increase in Ub conjugation in these muscles from insulin-deficient rats was due to greater activity of the N-end rule pathway, the effects of a specific inhibitor of this pathway was studied. A mutant form of E2_14k (Cys 88→Ser), which forms a stable ester bond with Ub instead of a thiolester linkage, can bind Ub but cannot transfer it to E3 or to substrates (47). This form of E2_14k acts as a dominant negative inhibitor of E3α (47). When added to the Ub-conjugation assay, it inhibited most of the Ub conjugation to muscle proteins from both groups (Fig. 1, b and c, lanes 4 and 5) and also decreased or eliminated the difference between the diabetic and control extracts in the rates of Ub conjugation to endogenous substrates and to α-lactalbumin. By contrast, another dominant negative E2, E2-C (Cys 114→Ser), which is not involved in the N-end rule pathway but inhibits ubiquitination of some cell cycle proteins (48, 49), had no effect on ubiquitination rates in extracts from either control or diabetic animals (Figure 1, b and c, lanes 6 and 7).

Is E2_14k or E3α rate limiting for Ub conjugation? Because Ub conjugation by the N-end rule pathway is enhanced in muscle extracts from the insulin-deficient animals, we investigated whether either of the enzymes of this pathway is rate limiting for ubiquitination in these extracts. Pure E2_14k, partially purified E3α, or a combination of both was added to rat muscle fraction II, and rates of Ub conjugation to ¹²⁵I-α-lactalbumin were measured (Figure 2). Addition of either E2_14k or E3α in amounts sufficient to at least double cellular levels consistently increased ubiquitination of ¹²⁵I-α-lactalbumin, but only by approximately 20% (Figure 2a). However, increasing the content of E2_14k and E3α together stimulated Ub conjugation of ¹²⁵I-α-lactalbumin by a larger amount than either alone (about 50%) (Figure 2a), as would be expected if the 2 proteins function together. Although increases in either E2_14k or E3α appear to enhance Ub conjugation, the quantitative effects of increasing these factors are difficult to estimate accurately, because it is unclear to what extent the recombinant E2_14k or the partially purified E3α have the same specific activity as the endogenous enzymes. Because purification often results in a partial loss of enzyme activity, the relatively small increases in Ub-conjugation rates observed in these in vitro experiments probably underestimate the actual effect of increased synthesis of E3α or E2_14k in the cell. In any case, these experiments suggest that even small increases in E2_14k or E3α production (e.g., < 50%) in the diabetic animals can cause a 40–50% increase in ubiquitination, as was seen after streptozotocin treatment.

Figure 2

The combination of E2_14k and E3α, when added to muscle extracts, increases the rate of Ub conjugation to an N-end rule substrate. Muscle extract fraction II (50 μg) was supplemented with recombinant E2_14k, E3α [as a 40% wt/vol (NH₄)₂SO₄ precipitate of crude muscle extract], or a combination of both proteins, and analyzed for Ub conjugation to ¹²⁵I-α-lactalbumin after addition of Ub (25 μM) and AMPPNP (2 mM). (a) Results are graphed as the amount of Ub conjugation measured in the extract supplemented with added E2_14k or E3α divided by the amount of Ub conjugation measured in the extract without added E2_14k or E3α × 100%. (b) The equivalent amount of fraction II with or without added E2_14k and E3α as was used in a was analyzed by SDS-PAGE and immunoblot to directly view the quantities of E2_14k and E3α added to the reactions. Densitometry analysis of the photographed immunoblots demonstrated that the reactions supplemented with E2_14k contained approximately 3.5 times the amount of E2_14k present in 50 μg of fraction II, and the reactions supplemented with E3α contained approximately 2.3 times the amount of enzyme present in 50 μg of fraction II.

Levels of E2_14k and E3α mRNA increase in the diabetic muscles. Because Ub conjugation by the N-end rule pathway was increased in muscle extracts from diabetic rats, we measured whether the genes for the components of the N-end rule pathway are induced in diabetic animals. As found previously (36), muscles from streptozotocin-treated rats contained twice as much of the E2_14k 1.2-kb transcript as did control muscles (Figure 3), whereas the abundance of the E2_14k 1.8-kb transcript was unchanged. A similar, specific increase in this 1.2-kb E2_14k transcript has been seen in muscles undergoing atrophy due to sepsis (29), cancer (21), and fasting (26).

Figure 3

Muscles from diabetic rats contain higher levels of E2_14k or E3α , but not E1 mRNA. (a) Northern blots. Total RNA (15 μg) pooled from 6 muscles of control or diabetic rats (for the E1 and E3α blots) or from a single control and diabetic rat (for E2_14k blot) were electrophoresed, transferred to nylon membranes, and hybridized with specific E1, E2_14k, or E3α probes as described in Methods. (b) Quantification of E2_14k or E3α transcripts in muscles from control and diabetic rats. E2_14k: Mean ± SD of mRNA content for the 1.2-kb E2_14k transcript in individual muscles from 4 diabetic and control rats. E3α: Mean ± SD of mRNA content for the single 6.0-kb E3α transcript in individual muscles from 6 diabetic and control rats. Quantities of E2_14k and E3α mRNA were normalized for gel-loading differences using the 18S rRNA band. The differences in E2_14k and E3α mRNA were statistically significant (P < 0.05) by a 2-tailed Student’s t test.

We also measured the mRNA levels of E3α, the Ub-protein ligase that functions together with E2_14k, and E1, the Ub-activating enzyme that charges E2 with Ub. The gene for E3α, also known as UBR1 in yeast, has recently been cloned from mouse and humans (39). In muscle it is expressed as a single 6.0-kb transcript. Using the 5′ 1 kb of the mouse UBR1 gene as a probe for E3α, we found that the E3α mRNA was also elevated about 2-fold in muscles from diabetic rats (Figure 3). These results demonstrate coordinate regulation of the 2 specific components of the N-end rule pathway. By contrast, mRNA levels for the Ub-activating enzyme, E1, which is required for all cellular ubiquitination reactions, was not increased in the muscles from diabetic rats (Figure 3a).

Preparation and characterization of anti-E3α antibody. To determine if increase in the level of E3α mRNA leads to an increase in E3α protein in muscles of diabetic rats, we prepared a polyclonal antibody to a 35-kD NH₂-terminal fragment of E3α that was expressed and purified in E. coli. A polyclonal anti-E3α IgG was obtained that reacted with a 200-kD protein in the rat muscle extract, in good agreement with the predicted molecular weight for the E3α gene product.

To establish that this 200-kd immunoreactive band was indeed E3α, we immunodepleted E3α from a rabbit reticulocyte extract (Figure 4). After immunodepletion, the 200-kD E3α band was no longer present in the supernatant (Figure 4a). Furthermore, when the E3α activity in the immunodepleted extracts was measured by assaying Ub conjugation to ¹²⁵I-α-lactalbumin (Figure 4b) in the presence of a large excess of recombinant E1 and E2_14k, E3α activity was much lower than in the untreated extracts or in extracts treated with preimmune IgG and protein A (Figure 4b, lanes 3 vs. 1 and 2). As expected, there was no formation of α-lactalbumin Ub conjugates in reticulocyte extracts in the presence of the dipeptide LysAla, a competitive inhibitor of E3α, either before or after immunodepletion (6, 15) (Figure 4b, lanes 4–6). Thus, Ub conjugation to this substrate is by E3α rather than some other Ub-conjugating activity. It is also noteworthy that the immunoprecipitate retained E3α activity; i.e., it was capable of catalyzing Ub conjugation to α-lactalbumin in the presence of added Ub, AMPPNP, E1, and E2_14k (Figure 4c).

Figure 4

Depletion of E3α from rabbit reticulocyte extract by an anti E3α antiserum. Rabbit reticulocyte fraction II (50 μg) was mixed with buffer, preimmune IgG (3 μg), or anti-E3α IgG (3 μg), and rotated at 4°C for 30 hours. Protein A agarose (3 μL) was added to each sample and incubated an additional 3 hours at 4°C. After sedimentation of the agarose beads, the supernatant and washed beads were separately analyzed. (a) Anti-E3α immunoblot of postimmunoprecipitation supernatants. Supernatants of the above reactions were subjected to 8% SDS-PAGE, probed with anti E3α IgG at a dilution of 1:2000. Treatment of the extract with anti E3α IgG specifically removed a 200-kD immunoreactive band. (b) E3α activity was depleted from the postimmunoprecipitation supernatants after treatment with anti-E3α IgG. Supernatants of the above reactions were supplemented with AMPPNP (2 mM), Ub (25 μM), Ub aldehyde (1 μM), E1 (0.1 μM), E2_14k (5 μM), and ¹²⁵I-α-lactalbumin (∼0.75 μM). In the presence of LysAla (2 mM), a competitive inhibitor of E3α, no Ub-conjugation activity is seen (lanes 4–6). In the presence of the isomeric dipeptide, AlaLys (2 mM), which does not inhibit E3α, ¹²⁵I-α-lactalbumin-Ub conjugates were formed after buffer addition (lane 1) or after immunoprecipitation with preimmune IgG (lane 2). However, after immunoprecipitation with anti-E3α IgG, most of the E3α activity was abolished. (c) Pellets after anti-E3α IgG immunoprecipitation contained E3α activity. Washed protein A agarose beads from the above reactions (4-fold increase in amounts) were supplemented with AMPPNP (2 mM), Ub (25 μM), E1 (0.1 μM), E2_14k (10 μM), and ¹²⁵I-α-lactalbumin (∼0.75 μM). Only the beads that contained anti-E3α IgG supported formation of ¹²⁵I-α-lactalbumin-Ub conjugates.

Content and activity of E1, E2_14k, and E3α protein in muscle extracts. The anti-E3α antiserum, as well as specific antiserum to E1 and E2_14k, were used to compare the levels of these proteins in muscles from the insulin-deficient and control rats. No significant differences were detected in the content of E1, E2_14k, or E3α proteins in the extracts from diabetic muscles and controls (Figure 5a). To ensure that the immunoblot analysis was performed in a linear detection range, known amounts of E2_14k or partially purified E3α were detected with the same antibodies (Figure 5b). This standardization indicates that 2-fold changes in protein level should be easily detectable.

Figure 5

Protein content of E1, E2_14k, and E3α are similar in extracts from diabetic and control rats. (a) Immunoblots from control and diabetic muscle extracts. Muscle protein (50 μg of a crude extract for the anti-E2_14k blot, and 50 μg muscle extract fraction II for the anti-E1 and anti-E3α blots) from control or diabetic rats was subjected to SDS-PAGE (9% for the anti-E1 blot, 15% for the E2_14k blot, and 8% for anti-E3α blot) and probed with the appropriate polyclonal antibody. (b) Immunoblots of known quantities of E2_14k and E3α. Increasing quantities of E2_14k and E3α were subjected to immunoblot analysis as in a above.

Because no increase in the content of these proteins was demonstrable in the gastrocnemius from insulin-deficient rats, we determined if the activity of E1, E2_14k, or E3α, nevertheless, might be increased in these extracts through some posttranslational modification such as phosphorylation. E1 and E2 activity was assayed by measuring the formation of E1-Ub and E2-Ub thiolester adducts. These linkages are labile but can be seen by SDS-PAGE in the absence of reducing agents (50). The muscle extracts from diabetic and control rats showed similar abilities to activate Ub and to form E2 complexes; i.e., the patterns and intensities of E1-¹²⁵I-Ub and E2-¹²⁵I-Ub were indistinguishable (Figure 6a, lanes 2 vs. 3) and, specifically, the E2_14k-Ub complexes were of similar intensities (Figure 6a, arrowhead).

In addition, the activity of E3α was analyzed by measuring Ub-conjugation to α-lactalbumin using (NH₄)₂SO₄ precipitates (40%) of muscle extracts. Any difference in the rate of Ub conjugation to α-lactalbumin in the 2 extracts are due to E3α, because endogenous E2_14k had been removed (Figure 6b, lanes 2 and 3 vs. lanes 4 and 5) and equivalent, excess amounts of E2_14k, E1 and ATP were added. In control experiments the ubiquitination of α-lactalbumin by these preparations was completely inhibited by LysAla, an inhibitor of E3α (6, 15) (data not shown). In these assays no consistent difference in E3α activity was seen in the diabetic and control extracts (Figure 6b, lanes 4 vs. 5), in accord with the measurements of E3α protein.

The biochemical mechanisms by which skeletal muscles undergo rapid protein loss in response to endocrine signals and cytokines have been of intense interest, largely because of the debilitating effects of muscle wasting that accompanies numerous severe diseases (e.g., cancer, renal failure, sepsis). Many recent studies have suggested that the primary mechanism for this increased muscle-protein breakdown in diverse catabolic states is activation of the Ub-proteasome pathway (16). We demonstrate here that in muscles from acutely diabetic rats, both Ub-protein conjugates, the critical intermediates in this pathway, and rates of their formation are increased. These changes can account for the overall increase in proteolysis observed in these muscles (5). Recently, we also demonstrated that in 3 other experimentally induced catabolic states where muscle-protein breakdown is enhanced (i.e., septic rats, tumor-bearing rats, and rats treated with high doses of T₃ or T₄), there was also an increase in Ub conjugation in muscle extracts (34). In all these cases, this response occurs in large part through activation of the N-end rule pathway. This conclusion was based on our findings that a dominant inhibitor of E3α (C88S E2_14k) reduced the rate of Ub conjugation in the extract of diabetic muscle to below control levels and that mRNA for this E3 enzyme (E3α) and its protein cofactor, E2_14k, were elevated. Thus, activation of Ub conjugation appears to be a common adaptation in the development of muscle atrophy due to diverse causes. These findings strongly support the proposal (16, 17, 34) that there is a final, common cellular mechanism responsible for activating muscle protein degradation triggered by diverse physiological or pathological factors.

The physiological importance of the N-end rule pathway has long remained elusive because E3α recognizes selectively protein substrates with large, bulky, or charged NH₂-terminal residues, and more than 80% of intracellular proteins are either N-α-acetylated or begin with methionine (51). Deletion of the E3α homologue in yeast, UBR1, had no clear physiological consequences (52), and no physiological substrates with these abnormal amino termini have been identified in yeast or mammals. On the other hand, our recent experiments have demonstrated a major role for E2_14k and E3α in overall Ub conjugation and protein degradation in muscle extracts from normal rats, but not in cultured HeLa cells (6). Its special role in skeletal muscle may be related to a particular role in regulating overall proteolysis in muscle as a potential source of amino acids.

The finding that E2_14k and E3α are important in the enhanced protein degradation observed in these diverse forms of muscle atrophy raises the fundamental question of how a large fraction of muscle proteins can be substrates for the N-end rule pathway. All muscle proteins are synthesized with an NH₂-terminal methionine and then, in most cases, have their NH₂-termini acetylated; both should prevent recognition by this system. One possibility is that in these catabolic states, muscle proteins are modified to expose novel NH₂-terminal residues (e.g., perhaps by endoproteolytic cleavage or deacylation) that are recognized by E3α. Alternatively, E3α might recognize other properties of protein substrates. In fact, yeast UBR1 is able to recognize some substrates (e.g., Gα and CUP9), that contain an NH₂-terminal methionine, presumably through interaction with internal regions of the proteins (53, 54). An important goal of future work will be to identify the polypeptides that undergo ubiquitination by E2_14k and E3α in normal and atrophying muscle.

Before these studies, the rate-limiting component in the Ub-proteasome pathway, either in the degradation of specific polypeptides or tissue proteins, had not been systematically studied. We have found that both E2_14k and E3α are rate limiting for Ub conjugation in normal muscle, as had been suggested by earlier findings (6). E2_14k and E3α not only function together in the N-end rule pathway but their expression is coordinately regulated because mRNA for both enzymes increased in the diabetic muscles at a time when other genes (rRNA, E1) remained constant. Because E1 is required to activate ubiquitin before all its conjugation reactions, it is unlikely that E1 would be rate limiting for Ub conjugation, and, in fact, its addition to muscle extract did not augment protein degradation (6).

Surprisingly, in muscles from these insulin-deficient rats, despite the 2–3-fold increase in Ub protein conjugates, the 40–50% increase in the rate of Ub conjugation by the N-end rule pathway, and the 2-fold elevations in mRNA for E2_14k and E3α, no consistent differences in the corresponding levels or activities of these enzymes could be demonstrated. These results emphasize the potential danger in extrapolating from measurements of increases in mRNA levels to enhanced protein expression, as has often been done in studies of the Ub-proteasome pathway in atrophying muscle (5, 8, 21–29). Whereas it is possible that E2_14k or E3α protein levels might rise at a later time during the development of diabetes, at the time these experiments were performed (3 days after streptozotocin administration), the adaptations responsible for the major loss of muscle protein in diabetic animals are functioning because protein degradation in incubated muscles was elevated by approximately 40%, the animals had already lost more than 20% of their initial body weight, and Ub conjugation rates had already increased (36). Another example of an increase in mRNA without an increase in protein content is branched chain ketoacid dehydrogenase in the muscle of acidotic animals (55). Perhaps the high levels of mRNA for these enzymes may indicate an increased rate of their synthesis, which may be necessary just to maintain high levels of the enzymes in atrophying muscles in the face of accelerated protein breakdown.

The simplest explanation for the apparent discordance between our findings of an increase in the overall rate of Ub conjugation and both mRNA for E2_14k and E3α in the diabetic extracts and the lack of demonstrable increases in the amount of E2_14k or E3α protein may stem from the lack of sufficient sensitivity in the assays used to measure these proteins or their activities. Detection of mRNA is more sensitive and requires less manipulation of the tissue than the enzymatic assays used here. In fact, immunoblot analyses can easily miss 30% differences in protein content. In addition, there is evidence in mammals of an additional isoform of UBR1 of similar molecular weight (39). If this additional form exists and is recognized by our antibody, it would be even harder to demonstrate increases in one form of E3α if the other does not change. Finally, a failure to detect small increases in E2_14k and E3α seems quite possible in light of the data in Figure 2, which shows that small increases in E2_14k and E3α act in an additive or synergistic manner in promoting Ub conjugation. Whereas we may have been unable to detect small increases in E2_14k or E3α individually in muscle extracts from insulin-deficient animals, when coupled together in vivo, changes in overall Ub-conjugation rate might become detectable.

If E2_14k or E3α changes are not responsible for the enhanced Ub conjugation and protein degradation in diabetic rats, then another explanation for these changes must exist. Ubiquitin isopeptidase activity, which is known to disassemble polyubiquitin chains (56), could, in theory, be decreased in the muscles from diabetic rats. In fact, such regulation has been proposed to be involved in memory formation in Aplysia (57). Differences in isopeptidase activity do not appear to occur in extracts from diabetic rats, however, because the addition of the isopeptidase inhibitor, ubiquitin aldehyde (58), increased Ub conjugation rates to a similar extent in extracts from control or diabetic muscles (Stewart Lecker, unpublished results). Finally, an unidentified enzyme or cofactor could be more active or induced by diabetes to act with E2_14k and E3α to enhance Ub conjugation. Proteins that increase the efficiency and processivity of Ub-conjugation reactions (e.g., Rbx1 and homologues) have recently been found to enhance Ub conjugation to G1 cyclins (59), and a new factor termed E4 can act with E1, E2, and the Ub protein ligase to elongate Ub chains on particular protein substrates (60).

Finally, in the N-end rule pathway, there are enzymes that can modify certain polypeptides so they can be ubiquitinated by E3α. For example, arginyl tRNA protein transferase (ATE1) can add arginine to the NH₂-termini of polypeptides that bear Asp or Glu residues (61), and NH₂-terminal amidase (NTA1) can deamidate NH₂-terminal Asn and Gln residues to produce Asp and Glu, respectively (62). If these enzymes were more active or abundant in extracts from diabetic muscles, then higher rates of Ub conjugation might be seen. However, we found that α-lactalbumin, a protein that does not need further modification to be recognized by E3α (because of its NH₂-terminal lysyl residue) (46) was ubiquitinated at a faster rate in muscle extracts from diabetic rats, indicating that the enzymes that ubiquitinate α-lactalbumin (i.e., E2_14k and E3α) are the ones changing in the diabetic extracts.

In summary, even with the new experimental tools introduced here, it is difficult to dissect critical adaptations when the magnitude of the overall changes occurring is small (20% decrease in muscle weight over 3 days and 40% change in Ub-conjugation rate) and may be multifactorial, i.e., involving smaller changes in multiple proteins. Regardless, a 20% reduction in muscle mass can be highly debilitating and, if progressive, would cause an insurmountable loss of tissue protein in the organism as a whole.

This work is supported in part by grants from the National Space Biomedical Research Institute (to A.L. Goldberg), National Institutes of Health (GM-46147 to A.L. Goldberg, DK-37175 and HL-45317 to W.E. Mitch, and DK-50740 to S.R. Price), and the Muscular Dystrophy Foundation (to A.L. Goldberg). S.H. Lecker is a Physician Postdoctoral Fellow of the Howard Hughes Medical Institute. We are grateful to Jackie Pierce, Margaret Read, and Vincent Chau for providing the dominant negative E2s, Chikara Miyamoto for providing the E1 overproducer, Art Haas for providing the antibody to E2_14k, and Aaron Ciechanover for providing the antiserum to E1. We also thank O. Kandror and E. Tarcsa for their helpful discussions and critical reading of the manuscript.

1. Nair, KS, Ford, GC, Ekberg, K, Fernqvist-Forbes, E, Wahren, J. Protein dynamics in whole body and in splanchnic and leg tissues in type I diabetic patients. J Clin Invest 1995. 95:2926-2937.
   View this article via: JCI PubMed Google Scholar
2. Kettelhut, IC, Wing, SS, Goldberg, AL. Endocrine regulation of protein breakdown in skeletal muscle. Diabetes Metab Rev 1988. 4:751-772.
   View this article via: PubMed Google Scholar
3. Nair, KS, Schwenk, WF. Factors controlling muscle protein synthesis and degradation. Curr Opin Neurol 1994. 7:471-474.
   View this article via: PubMed Google Scholar
4. Pepato, MT, Migliorini, RH, Goldberg, AL, Kettelhut, IC. Role of different proteolytic pathways in degradation of muscle protein from streptozotocin-diabetic rats. Am J Physiol 1996. 271:E340-E347.
   View this article via: PubMed Google Scholar
5. Price, SR, et al. Muscle wasting in insulinopenic rats results from activation of the ATP-dependent, ubiquitin-proteasome proteolytic pathway by a mechanism including gene transcription. J Clin Invest 1996. 98:1703-1708.
   View this article via: JCI PubMed Google Scholar
6. Solomon, V, Lecker, SH, Goldberg, AL. The N-end rule pathway catalyzes a major fraction of the protein degradation in skeletal muscle. J Biol Chem 1998. 273:25216-25222.
   View this article via: PubMed CrossRef Google Scholar
7. Rock, KL, et al. Inhibitors of the proteasome block the degradation of most cell proteins and the generation of peptides presented on MHC class 1 molecules. Cell 1994. 78:761-771.
   View this article via: PubMed CrossRef Google Scholar
8. Wing, SS, Goldberg, AL. Glucocorticoids activate the ATP-ubiquitin-dependent proteolytic system in skeletal muscle during fasting. Am J Physiol 1993. 264:E668-E676.
   View this article via: PubMed Google Scholar
9. Coux, O, Tanaka, K, Goldberg, AL. Structure and functions of the 20S and 26S proteasomes. Annu Rev Biochem 1996. 65:801-847.
   View this article via: PubMed CrossRef Google Scholar
10. Hershko, A, Ciechanover, A. The ubiquitin system. Annu Rev Biochem 1998. 67:425-479.
    View this article via: PubMed CrossRef Google Scholar
11. Handley, PM, Mueckler, M, Siegel, NR, Ciechanover, A, Schwartz, AL. Molecular cloning, sequence, and tissue distribution of the human ubiquitin-activating enzyme E1. Proc Natl Acad Sci USA 1991. 88:258-262.
    View this article via: PubMed CrossRef Google Scholar
12. Huibregtse, JM, Scheffner, M, Beaudenon, S, Howley, PM. A family of proteins structurally and functionally related to the E6-AP ubiquitin protein ligase. Proc Natl Acad Sci USA 1995. 92:2563-2567.
    View this article via: PubMed CrossRef Google Scholar
13. Patton, EE, Willems, AR, Tyers, M. Combinatorial control in ubiquitin-dependent proteolysis: don’t Skp the F-box hypothesis. Trends Genet 1998. 14:236-243.
    View this article via: PubMed CrossRef Google Scholar
14. Varshavsky, A. The N-end rule: functions, mysteries, uses. Proc Natl Acad Sci USA 1996. 93:12142-12149.
    View this article via: PubMed CrossRef Google Scholar
15. Reiss, Y, Kaim, D, Hershko, A. Specificity of binding of NH₂-terminal residue of proteins to ubiquitin-protein ligase. Use of amino acid derivatives to characterize specific binding sites. J Biol Chem 1988. 263:2693-2698.
    View this article via: PubMed Google Scholar
16. Mitch, WE, Goldberg, AL. Mechanisms of muscle wasting. The role of the ubiquitin-proteasome pathway. N Engl J Med 1996. 335:1897-1905.
    View this article via: PubMed CrossRef Google Scholar
17. Lecker, SH, Solomon, V, Mitch, WE, Goldberg, AL. Muscle protein breakdown and the critical role of the ubiquitin-proteasome pathway in normal and disease states. J Nutr 1999. 129:227S-237S.
    View this article via: PubMed Google Scholar
18. Bailey, JL, et al. The acidosis of chronic renal failure activates muscle proteolysis in rats by augmenting transcription of genes encoding proteins of the ATP-dependent ubiquitin-proteasome pathway. J Clin Invest 1996. 97:1447-1453.
    View this article via: JCI PubMed Google Scholar
19. Tawa, NE, Odessey, R, Goldberg, AL. Inhibitors of the proteasome reduce the accelerated proteolysis in atrophying rat skeletal muscles. J Clin Invest 1997. 100:197-203.
    View this article via: JCI PubMed Google Scholar
20. Hobler, SC, Tiao, G, Fischer, JE, Monaco, J, Hasselgren, PO. Sepsis-induced increase in muscle proteolysis is blocked by specific proteasome inhibitors. Am J Physiol 1998. 274:R30-R37.
    View this article via: PubMed Google Scholar
21. Temparis, S, et al. Increased ATP-ubiquitin-dependent proteolysis in skeletal muscles of tumor-bearing rats. Cancer Res 1994. 54:5568-5573.
    View this article via: PubMed Google Scholar
22. Tiao, G, et al. Sepsis stimulates nonlysosomal, energy-dependent proteolysis and increases ubiquitin mRNA levels in rat skeletal muscle. J Clin Invest 1994. 94:2255-2264.
    View this article via: JCI PubMed Google Scholar
23. Mitch, WE, et al. Metabolic acidosis stimulates muscle protein degradation by activating the adenosine triphosphate-dependent pathway involving ubiquitin and proteasomes. J Clin Invest 1994. 93:2127-2133.
    View this article via: JCI PubMed Google Scholar
24. Medina, R, Wing, SS, Goldberg, AL. Increase in levels of polyubiquitin and proteasome mRNA in skeletal muscle during starvation and denervation atrophy. Biochem J 1995. 307:631-637.
    View this article via: PubMed Google Scholar
25. Fang, CH, et al. Burn injury stimulates multiple proteolytic pathways in skeletal muscle, including the ubiquitin-energy-dependent pathway. J Am Coll Surg 1995. 180:161-170.
    View this article via: PubMed Google Scholar
26. Wing, SS, Banville, D. 14-kDa ubiquitin-conjugating enzyme: structure of the rat gene and regulation upon fasting and by insulin. Am J Physiol 1994. 267:E39-E48.
    View this article via: PubMed Google Scholar
27. Dardevet, D, et al. Sensitivity and protein turnover response to glucocorticoids are different in skeletal muscle from adult and old rats. Lack of regulation of the ubiquitin-proteasome proteolytic pathway in aging. J Clin Invest 1995. 96:2113-2119.
    View this article via: JCI PubMed Google Scholar
28. Taillandier, D, et al. Coordinate activation of lysosomal, Ca²⁺-activated and ATP-ubiquitin-dependent proteinases in the unweighted rat soleus muscle. Biochem J 1996. 316:65-72.
    View this article via: PubMed Google Scholar
29. Voisin, L, et al. Muscle wasting in a rat model of long-lasting sepsis results from the activation of lysosomal, Ca²⁺-activated, and ubiquitin-proteasome proteolytic pathways. J Clin Invest 1996. 97:1610-1617.
    View this article via: JCI PubMed Google Scholar
30. Wing, SS, Haas, AL, Goldberg, AL. Increase in ubiquitin-protein conjugates concomitant with the increase in proteolysis in rat skeletal muscle during starvation and atrophy denervation. Biochem J 1995. 307:639-645.
    View this article via: PubMed Google Scholar
31. Tiao, G, et al. Energy-ubiquitin-dependent muscle proteolysis during sepsis in rats is regulated by glucocorticoids. J Clin Invest 1996. 97:339-348.
    View this article via: JCI PubMed Google Scholar
32. Baracos, VE, DeVivo, C, Hoyle, DH, Goldberg, AL. Activation of the ATP-ubiquitin-proteasome pathway in skeletal muscle of cachectic rats bearing a hepatoma. Am J Physiol 1995. 268:E996-E1006.
    View this article via: PubMed Google Scholar
33. Llovera, M, et al. Ubiquitin gene expression is increased in skeletal muscle of tumour-bearing rats. FEBS Lett 1994. 338:311-318.
    View this article via: PubMed CrossRef Google Scholar
34. Solomon, V, Baracos, V, Sarraf, P, Goldberg, AL. Rates of ubiquitin conjugation increase when muscles atrophy, largely through activation of the N-end rule pathway. Proc Natl Acad Sci USA 1998. 95:12602-12607.
    View this article via: PubMed CrossRef Google Scholar
35. Solomon, V, Goldberg, AL. Importance of the ATP-ubiquitin-proteasome pathway in the degradation of soluble and myofibrillar proteins in rabbit muscle extracts. J Biol Chem 1996. 271:26690-26697.
    View this article via: PubMed CrossRef Google Scholar
36. Mitch, WE, et al. Evaluation of signals activating the ubiquitin-proteasome proteolysis in a model of muscle wasting. Am J Physiol 1999. 276:C1132-C1138.
    View this article via: PubMed Google Scholar
37. Johnston, NL, Cohen, RE. Uncoupling ubiquitin-protein conjugation from ubiquitin-dependent proteolysis by use of beta, gamma-nonhydrolyzable ATP analogues. Biochemistry 1991. 30:7514-7522.
    View this article via: PubMed CrossRef Google Scholar
38. Laemmli, UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970. 227:680-685.
    View this article via: PubMed CrossRef Google Scholar
39. Kwon, YT, et al. The mouse and human genes encoding the recognition component of the N-end rule pathway. Proc Natl Acad Sci USA 1998. 95:7898-7903.
    View this article via: PubMed CrossRef Google Scholar
40. Harlow, E., and Lane, D. 1988. Antibodies: a laboratory manual. Cold Spring Harbor Laboratory Press. Cold Spring Harbor, NY.
    View this article via: PubMed Google Scholar
41. Joyner, AL, Kornberg, T, Coleman, KG, Cox, DR, Martin, GR. Expression during embryogenesis of a mouse gene with sequence homology to the Drosophila engrailed gene. Cell 1985. 43:29-37.
    View this article via: PubMed CrossRef Google Scholar
42. Church, GM, Gilbert, W. Genomic sequencing. Proc Natl Acad Sci USA 1984. 81:1991-1995.
    View this article via: PubMed CrossRef Google Scholar
43. Wing, SS, Dumas, F, Banville, D. A rabbit reticulocyte ubiquitin carrier protein that supports ubiquitin-dependent proteolysis (E2_14k) is homologous to the yeast DNA repair gene RAD6. J Biol Chem 1992. 267:6495-6501.
    View this article via: PubMed Google Scholar
44. Fagan, JM, Waxman, L, Goldberg, AL. Skeletal muscle and liver contain a soluble ATP + ubiquitin-dependent proteolytic system. Biochem J 1987. 243:335-343.
    View this article via: PubMed Google Scholar
45. Ciechanover, A, Elias, S, Heller, H, Hershko, A. “Covalent affinity” purification of ubiquitin-activating enzyme. J Biol Chem 1982. 257:2537-2542.
    View this article via: PubMed Google Scholar
46. Reiss, Y, Hershko, A. Affinity purification of ubiquitin-protein ligase on immobilized protein substrates. Evidence for the existence of separate NH₂-terminal binding sites on a single enzyme. J Biol Chem 1990. 265:3685-3690.
    View this article via: PubMed Google Scholar
47. Sung, P, Prakash, S, Prakash, L. Stable ester conjugate between the S. cerevisiae RAD6 protein and ubiquitin has no biological activity. J Mol Biol 1991. 221:745-749.
    View this article via: PubMed CrossRef Google Scholar
48. Townsley, FM, Aristarkhov, A, Beck, S, Hershko, A, Ruderman, JV. Dominant-negative cyclin-selective ubiquitin carrier protein E2-C/UbcH10 blocks cells in metaphase. Proc Natl Acad Sci USA 1997. 94:2362-2367.
    View this article via: PubMed CrossRef Google Scholar
49. Aristarkhov, A, et al. E2-C, a cyclin-selective ubiquitin carrier protein required for the destruction of mitotic cyclins. Proc Natl Acad Sci USA 1996. 93:4294-4299.
    View this article via: PubMed CrossRef Google Scholar
50. Haas, AL, Bright, PM. The resolution and characterization of putative ubiquitin carrier protein isozymes from rabbit reticulocytes. J Biol Chem 1988. 263:13258-13267.
    View this article via: PubMed Google Scholar
51. Brown, JL, Roberts, WK. Evidence that approximately eighty percent of the soluble proteins from Ehrlich ascites cells are N-alpha acetylated. J Biol Chem 1976. 251:1009-1014.
    View this article via: PubMed Google Scholar
52. Bartel, B, Wunning, I, Varshavsky, A. The recognition component of the N-end rule pathway. EMBO J 1990. 9:3179-3189.
    View this article via: PubMed Google Scholar
53. Madura, K, Varshavsky, A. Degradation of G alpha by the N-end rule pathway. Science 1994. 265:1454-1458.
    View this article via: PubMed CrossRef Google Scholar
54. Byrd, C, Turner, GC, Varshavsky, A. The N-end rule pathway controls the import of peptides through degradation of a transcriptional repressor. EMBO J 1998. 17:269-277.
    View this article via: PubMed CrossRef Google Scholar
55. England, BK, et al. Rat muscle branched-chain ketoacid dehydrogenase activity and mRNAs increase with extracellular acidemia. Am J Physiol 1995. 268:C1395-C1400.
    View this article via: PubMed Google Scholar
56. Wilkinson, KD, et al. Metabolism of the polyubiquitin degradation signal - structure, mechanism, and role of isopeptidase T. Biochemistry 1995. 34:14535-14546.
    View this article via: PubMed CrossRef Google Scholar
57. Hegde, AN, et al. Ubiquitin C-terminal hydrolase is an immediate-early gene essential for long-term facilitation in Aplysia. Cell 1997. 89:115-126.
    View this article via: PubMed CrossRef Google Scholar
58. Hershko, A, Rose, IA. Ubiquitin-aldehyde: a general inhibitor of ubiquitin-recycling processes. Proc Natl Acad Sci USA 1987. 84:1829-1833.
    View this article via: PubMed CrossRef Google Scholar
59. Skoyra, D, et al. Reconstitution of G1 cyclin ubiquitination with complexes containing SCF-GRR1 and Rbx1. Science 1999. 284:562-565.
    View this article via: CrossRef Google Scholar
60. Koegl, M, et al. A novel ubiquitination factor, E4, is involved in multiubiquitin chain assembly. Cell 1999. 96:635-644.
    View this article via: PubMed CrossRef Google Scholar
61. Ciechanover, A, et al. Purification and characterization of arginyl-tRNA-protein transferase from rabbit reticulocytes. Its involvement in post-translational modification and degradation of acidic NH₂ termini substrates of the ubiquitin pathway. J Biol Chem 1988. 263:11155-11167.
    View this article via: PubMed Google Scholar
62. Grigoryev, S, et al. A mouse amidase specific for N-terminal asparagine. The gene, the enzyme, and their function in the N-end rule pathway. J Biol Chem 1996. 271:28521-28532.
    View this article via: PubMed CrossRef Google Scholar