CYP2E1 Degradation by in Vitro Reconstituted Systems: Role of the Molecular Chaperone hsp90

Archives of Biochemistry and Biophysics Vol. 379, No. 2, July 15, pp. 321–330, 2000 doi:10.1006/abbi.2000.1870, available online at http://www.idealibrary.com on

CYP2E1 Degradation by in Vitro Reconstituted Systems: Role of the Molecular Chaperone hsp90 Thierry Goasduff and Arthur I. Cederbaum 1 Department of Biochemistry and Molecular Biology, Mount Sinai School of Medicine, New York, New York 10029

Received February 10, 2000, and in revised form April 5, 2000

One major mode of regulation of cytochrome P450 2E1 (CYP2E1) is at the posttranscriptional level, since many low-molecular-weight compounds stabilize the enzyme against proteolysis by the proteasome complex. In an in vitro system containing human liver microsomes, degradation of CYP2E1 in the microsomes required addition of the human liver cytosol fraction in a reaction sensitive to inhibitors of the proteasome complex. It is not clear how CYP2E1 in the microsomal membrane becomes accessible to the cytosolic proteasome. Since molecular chaperones play a role in protein folding and degradation, the possible role of heat shock proteins in CYP2E1 degradation by this reconstituted system was evaluated. Degradation of CYP2E1 required ATP; ATP-␥S, a nonhydrolyzable analogue of ATP, did not catalyze CYP2E1 degradation by the cytosol fraction, indicating that ATP hydrolysis is required. Geldanamycin, a specific inhibitor of hsp90, inhibited the degradation of microsomal CYP2E1 by the cytosol fraction. Control experiments indicated that geldanamycin was not a substrate/ligand of CYP2E1 nor did it prevent microsomal lipid peroxidation, a process which increases CYP2E1 turnover. Inhibition by geldanamycin was prevented by molybdate. Both of these compounds have been shown to promote alterations in hsp90 structure and to modulate hsp90 –protein interactions. The proteasome activity in the cytosol, as assayed by the cleavage of a fluorogenic peptide, was enhanced when ATP was added and inhibited by 30 – 40% by geldanamycin, effects that are similar, although less pronounced, to the degradation of CYP2E1 by the cytosol. Purified 20S proteasome could catalyze degradation of CYP2E1; however, in an assay using equal peptidase activity, the cytosol fraction was much more effective than the 20S proteasome in promoting CYP2E1 degradation.

Immunodepletion of hsp90 from the cytosol resulted in prevention of the degradation of CYP2E1, a reaction that was reversed by the addition of pure hsp90 to this cytosol. These results suggest that in addition to the proteasome, the cytosol fraction contains other factors that modulate the efficiency of CYP2E1 degradation. The sensitivity to geldanamycin and molybdate and the immunodepletion experiments suggest that hsp90 is one of these factors that interact with CYP2E1 and/or with the proteasome to promote the degradation of this microsomal P450. © 2000 Academic Press Key Words: CYP2E1 degradation; proteasome; heat shock protein 90; molecular chaperone; geldanamycin; ATP.

Cytochrome P450 2E1 (CYP2E1) 2 is involved in the oxidation of acetaminophen, ethanol, benzene, and halogenated solvents (1). CYP2E1 is induced under a variety of conditions, and the understanding of this enzyme induction mechanism is important because of the ability of CYP2E1 to activate numerous carcinogens such as nitrosodimethylamine and industrial solvents (2, 3). The regulation of CYP2E1 expression is intriguing because unlike the majority of P450s, CYP2E1 induction appears to involve several different mechanisms. It may be regulated by transcriptional activation at birth, mRNA stabilization, increased translation of mRNA, or inhibition of protein degradation because of protein stabilization (4 –7). Enzyme stabilization by many low-molecular-weight substrates, such as ethanol and acetone, appears to be a common mechanism by which CYP2E1 is induced (8, 9). In the absence of substrate, CYP2E1 is degraded with a half-life of 6 –7 h followed by a slower secondary phase of about 37 h (9). In the presence of ethanol, only

1

To whom correspondence should be addressed at Department of Biochemistry and Molecular Biology, Box 1020, Mount Sinai School of Medicine, New York, NY 10029. Fax: (212) 996-7214. 0003-9861/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.

2 Abbreviations used: CYP2E1, cytochrome P450 2E1; GA, geldanamycin; PNP, p-nitrophenol.

321

322

GOASDUFF AND CEDERBAUM

the slower phase of CYP2E1 turnover is observed (10). Several studies have shown that this rapid CYP2E1 degradation process involves the ubiquitin-dependent or -independent proteasome complex system (11–14). CYP2E1 proteolysis by the 20S proteasome is enhanced by prior labilization by CCl 4, a suicide substrate of CYP2E1 (12), or by an NADPH-based mechanism, which appears to involve reactive oxygen species (14). A mechanism-based inactivation of liver microsomal cytochrome P450 3A, via heme modification of the protein, has been shown to result in enhanced proteolytic degradation of the enzyme by the ubiquitin–ATP-dependent 26S proteasome (15, 16). Common features needed in vitro for the degradation of several cytochrome P450 enzymes including CYP2E1 appear to be a protease system present in the cytosol (11–16), the need of ATP for efficient degradation of cytochrome P450 (14, 15), and enhancement of this degradation when the hemeprotein is modified (12, 16). It is not clear how membrane-bound substrates such as P450s interact with the largely cytosolic proteasome system or what the role of ATP is in the overall degradation process. Several studies have shown the involvement of molecular chaperones in protein degradation (reviewed in 17, 18). Chaperone proteins may function both in the refolding of misfolded proteins, in order to prevent their removal, and in the presentation of substrates to the proteolytic machinery for removal, if the substrate cannot be refolded properly (17). Proteolytic and chaperone systems are often associated with each other and can be coordinately regulated in prokaryotes and eukaryotes; e.g., a stress signal leads to the transcriptional activation of genes encoding components for proteolysis and chaperone activity (19). The family of heat shock proteins and molecular chaperones includes hsp100, hsp90, hsp70, hsp60, hsp40, and small hsps. hsp70 and hsp60 together with their respective cochaperones and ATP have the ability to refold nonnative intermediates (20). The hsp100 chaperone has the unique ability to disaggregate protein aggregates in an ATP-dependent manner (21). The biochemical mechanism of protein folding by hsp90 is poorly understood, and the direct involvement of ATP in this process has been particularly controversial. It has been shown in studies with rabbit reticulocyte lysates (22) and with purified components (23) that hsp90 can bind and hydrolyze ATP (24 –26) and that ATP is required for the association of hsp90 with its p23 ancillary protein. The N-terminal binding pocket of hsp90 is the site of ATP binding but also the site to which the benzoquinoid ansamycin geldanamycin binds (24). The binding of geldanamycin with hsp90 disrupts the function of numerous hsp-dependent proteins by altering their normal interactions with the hsp90 machinery, including p23 (23). Geldanamycin has been shown to increase the

degradation of soluble proteins such as certain tyrosine kinases when added to cell lysates (27), presumably by preventing stabilization of the protein by hsp90. In an attempt to elucidate the role of ATP in CYP2E1 degradation and to further understand how membrane-bound CYP2E1 interacts with the proteasome complex, we hypothesized that molecular chaperones could be involved in the in vitro degradation of CYP2E1 by the proteasome. Use of a specific inhibitor of hsp90 function, geldanamycin, suggested that hsp90 is involved in the CYP2E1 degradation pathway, and that because of this hsp90 involvement, the proteasome activity of cytosolic extracts is enhanced by the addition of ATP and decreased by geldanamycin. MATERIALS AND METHODS Materials. ATP, ATP-␥S, ATP–Tris, MgCl 2, sodium molybdate, and Suc-Leu-Leu-Val-Tyr-7-amido-4-methylcoumarin (SLLVTAMC) were from Sigma Chemical Co. (St. Louis, MO). NADPH was from Boehringer-Mannheim (Indianapolis, IN). Geldanamycin was obtained from the Drug Synthesis and Chemistry Branch, Developmental Therapeutics Program, Division of Cancer Center Treatment, National Cancer Institute. Human liver was obtained from the Mount Sinai Liver Transplant program and was kindly provided by Dr. Dennis Feierman of the Department of Anesthesiology. The rabbit anti-human CYP2E1 polyclonal antibody was kindly provided by Dr. J. M. Lasker, Department of Biochemistry and Molecular Biology, Mount Sinai Medical Center (New York, NY). The purified 20S proteasome, purified hsp90, and anti-hsp90 IgG–Sepharose were kindly provided by Dr. Sherwin Wilk, Department of Pharmacology, Mount Sinai Medical Center. Microsomal and cytosolic fractions. Human liver microsomes and cytosol fractions (105,000g supernatant fraction) were prepared by differential centrifugation of the liver homogenate, which was prepared in 50 mM phosphate buffer, pH 7.4, containing 150 mM KCl and 250 mM sucrose as previously described (14). The microsomal pellet was resuspended in the same buffer. The cytosolic and microsomal fractions were divided into small aliquots and stored at ⫺80°C. In vitro assay for CYP2E1 degradation. The CYP2E1 degradation was followed in a reaction system containing 50 mM phosphate buffer, ph 7.4, 1.2 mM MgCl2, 0.6 mM ATP (or an ATP-regenerating system consisting of 8 mM creatine phosphate, 0.5 mM ATP, and 30 units/ml creatine phosphokinase), and 10 ␮g microsomal protein in the absence or presence of 40 ␮g protein of the cytosolic fraction in a total reaction volume of 0.05 ml. Microsomes were preincubated for 30 min at 37°C in the presence of NADPH (0.5 mM) prior to the addition of MgATP and cytosol (40 ␮g). A stock solution of geldanamycin was prepared as a sonicated solution in water. The incubation to study degradation of CYP2E1 by microsomes alone or microsomes plus cytosol was carried out at 37°C for 5 h, and the reaction was stopped by adding the SDS– PAGE running buffer followed by boiling at 100°C. Some tubes were kept in ice throughout the incubation and are considered 100% CYP2E1 concentration. The proteolytic activity toward CYP2E1 was assayed by Western blot analysis to detect the remaining content of CYP2E1. Immunoblot analysis. Western blot analysis was carried out on a 10-␮l aliquot of the boiled samples (equivalent to 0.66 ␮g microsomal protein) using an 8% running gel and a 4% stacking polyacrylamide gel. Electrophoresis and electrotransfer procedures were performed with the Bio-Rad Mini-Protean system as previously described (14). Nitrocellulose membranes with transferred proteins were incubated with anti human CYP2E1 polyclonal antibody as the first antibody followed by incubation with horseradish peroxidase-conjugated goat antirabbit antibody and developing by chemiluminescent detection (ECL Kit, Amersham, Arlington Heights, IL).

ROLE OF hsp90 IN THE DEGRADATION OF CYP2E1 Treatment of cytosol with anti-hsp90 IgG–Sepharose. Cytosol (50 ␮l) was incubated in the absence or presence of an excess of antihsp90 IgG-Sepharose (200 ␮l packed volume) on an orbital shaker in a 1.5-ml flat-bottom tube. After overnight shaking and centrifugation, the resultant supernatant was removed and used to assay CYP2E1 degradation as described above and for the remaining content of hsp90. Western blot analysis for hsp90 was carried out on 10 ␮g of either cytosolic protein or supernatant protein from anti-hsp90 IgG–Sepharose-treated cytosol, using an 8% running gel and a 4% stacking polyacrylamide gel. Anti-human hsp90␣ goat polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA) was used as the primary antibody and horseradish peroxidase-conjugated anti goat IgG as the secondary antibody (Santa Cruz Biotechnology). Assay of proteasome peptidase activity. Proteasome activity was determined by following the cleavage of the fluorogenic substrate SLLVT-AMC. Assays were performed at 37°C in a shaking water bath for 1 h using 0.1 M Tris–HCl, pH 8.0, 2.5 or 5 mM MgCl 2, 50 ␮M SLLVT-AMC, and 40 ␮g protein from human liver cytosol in a total volume of 100 ␮l. Experiments were carried out in the absence or presence of the indicated concentration of ATP–Tris. When ATP– Tris was present, geldanamycin (dissolved in DMSO) was added to the cytosol prior to the addition of the ATP. In some experiments, purified proteasome 20S (22 ng protein) was used instead of the human liver cytosol. Enzymatic reactions were terminated by adding 0.5 ml of absolute ethanol. Fluorescence of 4-aminomethylcoumarin was measured at an excitation wavelength of 390 nm and an emission wavelength of 440 nm using a Perkin–Elmer LS50B luminescence spectrometer (Perkin–Elmer Co., Norwalk, CT) and quantified by using standard concentrations of the AMC fluorescent product. Protein determination. Protein concentration was determined by using the Bio-Rad DC protein assay and bovine serum albumin as a standard. CYP2E1 catalytic activity: p-Nitrophenol oxidation. Human liver microsomes (50 ␮g) were incubated for 30 min at 37°C in 100 mM phosphate buffer, pH 7.4, plus 200 ␮M p-nitrophenol in a total volume of 100 ␮l. The reaction was started by adding 1 mM NADPH and stopped after 30 min by adding 30 ␮l of 20% TCA. After centrifugation, 10 ␮l of 10 N NaOH was added to the supernatant, and the optical density at 510 nm was immediately determined using a Shimadzu UV160U spectrophotometer. In some experiments, geldanamycin, prepared as a sonicated solution was added at 100 ␮M and diphenylene iodonium chloride, was added at 50 ␮M. Lipid peroxidation assay. Lipid peroxidation was assayed by measuring the production of TBA-reactive metabolites, which are the breakdown products of lipid peroxides. Briefly, microsomes (50 ␮g protein) or cytosol (160 ␮g protein) or both were incubated for 1 h at 37°C in 100 mM phosphate buffer, pH 7.4, containing 1.2 mM MgCl 2, 0.6 mM ATP, and 1 mM NADPH in the presence or absence of geldanamycin (20 ␮M) or EDTA (100 ␮M) in a total volume of 100 ␮l. The reaction was stopped by addition of 100 ␮l of 30% TCA. After addition of 100 ␮l of 0.68% TBA, the mixture was boiled for 30 min and then centrifuged, and the optical density at 532 nm of the supernatant was immediately determined. Statistical analysis. Where indicated, results are expressed as the mean ⫾ SD. Statistical significance was determined using the unpaired Student’s t test, with P ⬍ 0.05 considered significant (Microcal Origin 4.0, Microcal Software, Inc., Northampton, MA).

RESULTS

Requirement of ATP Hydrolysis for CYP2E1 Degradation by the Cytosol To determine the need for ATP and ATP hydrolysis for CYP2E1 degradation, we compared the effects of ATP to the nonhydrolysable ATP analogue ATP-␥S.

323

FIG. 1. Requirement of ATP hydrolysis for CYP2E1 degradation by the cytosol. (A) Human liver microsomes were incubated on ice (lane 1) or incubated at 37°C with 0.6 mM ATP (lane 2) or 0.6 mM ATP plus cytosol (lanes 6 to 8). In some samples, 0.6 mM ATP-␥S instead of ATP was added to microsomes (lanes 3 to 5) or microsomes plus cytosol (lanes 9 to 11). The reaction mixture was overlaid with mineral oil and incubated in a shaking water-bath at 37°C for 5 h. The reaction was stopped by addition of the SDS–PAGE loading buffer, and Western blot analysis was performed using an aliquot (10 ␮l) of the boiled samples. (B) Quantification of the scanned immunoblot from (A) using ImageQuant Software. Results are expressed as percentages of remaining CYP2E1 after the 5-h incubation and represent the means ⫾ SD of the three samples from (A) (a, significantly different from mic at *P ⬍ 0.05; b, significantly different from mic ⫹ cyt at *P ⬍ 0.05).

The incubation of the microsomal fraction alone at 37°C for 5 h did not result in significant degradation of CYP2E1 in the presence of either ATP or ATP-␥S (Fig. 1, lanes 2 to 5 compared to nonincubated lane 1). The addition of the cytosol fraction enhanced the CYP2E1 degradation in the presence of ATP (Fig. 1, lanes 6 to 8). When ATP-␥S was used in the reaction mixture instead of ATP, degradation of CYP2E1 did not occur (Fig. 1, lanes 9 to 11). The need for ATP hydrolysis for efficient CYP2E1 degradation could involve several proteolytic pathways described previously, including cAMP-dependent phosphorylation of CYP2E1, followed by degradation by serine proteases present in the endoplasmic reticulum (28 –30) or by ubiquitin-dependent proteasome degradation of CYP2E1 (11). The first possibility is not a likely explanation since the microsomes incubated alone did not show significant degradation of CYP2E1; addition of the cytosolic fraction with ATP is needed for CYP2E1 degradation. Concerning the need of ATP for ubiquitination, Roberts (12) found that the CCl 4-inactivated CYP2E1 of rat liver microsomes could be degraded by 20S proteasome without any evidence of ubiquitination. We have not

324

GOASDUFF AND CEDERBAUM

observed high molecular weight bands or smears when human liver microsomal CYP2E1 is incubated with cytosol or when HepG2 cells overexpressing CYP2E1 are treated with proteasome inhibitors such as lactacystin or MG132 (14). Although these results do not rule out the possibility that ubiquitination is required for efficient degradation of CYP2E1 and the need for ATP hydrolysis could reflect the ubiquitination– deubiquitination process, it may be interesting to consider the involvement of other cytosolic factors or processes modulated by MgATP that contribute to the ATP hydrolysis requirement for CYP2E1 degradation. One such process could involve molecular chaperones. Recent in vitro and in vivo studies have demonstrated an ATPase activity of hsp90, which is sensitive to inhibition by a specific inhibitor, geldanamycin (24 –26). Effect of Geldanamycin on Degradation of CYP2E1 by the Cytosol In order to show if the ATPase activity of hsp90 is involved in CYP2E1 degradation by the cytosol, the cytosol was preincubated in the presence of increasing concentrations of geldanamycin. Because many organic solvents, such as DMSO and ethanol, stabilize CYP2E1 and prevent its degradation, geldanamycin was prepared as a sonicated solution in water. As shown in Fig. 2, without the ATP-regeneration system, the cytosol did not degrade CYP2E1 (lane 2 compared to lane 1, no cytosol added), but with the ATP-regenerating system the cytosol is able to effectively degrade CYP2E1 (lane 3). When increasing concentrations of geldanamycin were added to the cytosol before the addition of the ATP-regenerating system, the degradation of CYP2E1 was strongly prevented. This inhibition of CYP2E1 degradation is complete at an added geldanamycin concentration in suspension of 20 ␮M (Figs. 2B and 2C). Many substrates of CYP2E1 stabilize the protein and prevent its degradation (7). To evaluate whether geldanamycin is a substrate or ligand of CYP2E1, pnitrophenol hydroxylation (PNP) specific enzymatic activity was measured using microsomes (mic) in the presence or absence of up to 100 ␮M geldanamycin. As shown in Fig. 3, 100 ␮M geldanamycin (GA) did not inhibit CYP2E1 catalytic activity (PNP hydroxylation rate in the absence of geldanamycin of 1590 ⫾ 257 pmol/min/mg protein compared to a rate in the presence of 100 ␮M geldanamycin of 1417 ⫾ 57 pmol/ min/mg protein). As a positive control for these experiments, diphenylene iodonium chloride (DPI), an NADPH P450 reductase inhibitor, at 50 ␮M strongly inhibited the PNP activity. CYP2E1 degradation can also be decreased by inhibitors of lipid peroxidation (14); the prior 30-min incubation of microsomes with NADPH promoted the degradation of CYP2E1 by labi-

FIG. 2. Effect of geldanamycin on the degradation of microsomal CYP2E1 by human liver cytosol. (A) Human liver microsomes were incubated without cytosol (lane 1) or with cytosol plus ATP generating system (lane 3) or with cytosol without ATP generating system (lane 2). Geldanamycin (GA), prepared as a sonicated solution in water, was added at final concentrations of 2, 8, 16, and 20 ␮M in the cytosol (lanes 4 to 11) before adding the ATP-regenerating system (ATP⌺: 8 mM creatine phosphate, 30 U/ml creatine phosphokinase, 0.5 mM ATP). The reaction mixture was overlaid with mineral oil and incubated in a shaking water-bath at 37°C for 5 h. The reaction was stopped by addition of the SDS–PAGE loading buffer, and Western blot analysis was performed using an aliquot (10 ␮l) of the boiled samples. (B) Human liver microsomes were incubated on ice (lane 1) or incubated at 37°C without cytosol (lane 2), with cytosol (lanes 3 to 5), or with cytosol plus geldanamycin (lanes 6 – 8). Geldanamycin (GA), prepared as a sonicated solution in water, was added at a final concentration of 20 ␮M in the cytosol before addition of MgATP. Incubation and analysis was as described in (A). (C) Quantification using ImageQuant Software of the scanned immunoblot from (B). Results are expressed as a percentage of remaining CYP2E1 after incubation and represent the mean ⫾ SD of the three samples from (B) (a, significantly different from mic at *P ⬍ 0.05; b, significantly different from cyt at *P ⬍ 0.05).

lizing the enzyme by generating reactive oxygen species and causing microsomal lipid peroxidation to occur. To rule out a possible antioxidant action of geldanamycin in preventing CYP2E1 degradation, the effect of geldanamycin on microsomal lipid peroxidation was evaluated. Since geldanamycin was added to

ROLE OF hsp90 IN THE DEGRADATION OF CYP2E1

325

ure 5C shows the quantification of the blots from Fig. 5B. Generally, molybdate at concentrations of 10 –20 mM was sufficient to overcome the inhibition of CYP2E1 degradation by 20 ␮M geldanamycin. Effect of Geldanamycin and ATP on Proteasome Activity in the Cytosolic Fraction

FIG. 3. Geldanamycin does not inhibit CYP2E1 activity. CYP2E1 activity was determined by assaying p-nitrophenol hydroxylation by human liver microsomes (50 ␮g) as described under Materials and Methods. Geldanamycin or diphenylene iodonium chloride, prepared as sonicated solutions in water, were added at final concentrations of 100 or 50 ␮M, respectively to the reaction mixture. Results represent the mean ⫾ SD of three assays (ns, not significantly different from mic; *significantly different from mic at P ⬍ 0.01).

The proteasome activity in the cytosol was determined by assaying the cleavage of the fluorogenic substrate SLLVT-AMC in the presence or absence of ATP or of geldanamycin. By increasing the concentration of ATP (from 0.5 to 5 mM) at a MgCl 2 concentration of 2.5 mM, the proteasome activity was enhanced compared to the assay without ATP (Fig. 6A). At an ATP concentration of 1 mM, the proteasome peptidase activity of the cytosol was increased 2-fold. Addition of 10 or 20 ␮M geldanamycin to the cytosol in the absence of added ATP slightly inhibited the proteasome peptidase

the cytosol 30 min after the microsomes were incubated with NADPH, this compound is not likely to affect CYP2E1 degradation through an antioxidant mechanism. Nevertheless, we verified that geldanamycin did not inhibit lipid peroxidation using the thiobarbituric acid-reactive lipid peroxidation products assay. Results shown in Fig. 4 indicate that 20 ␮M geldanamycin did not inhibit microsomal-dependent lipid peroxidation in the absence or presence of the cytosol fraction, whereas 100 ␮M EDTA, a known inhibitor of microsomal lipid peroxidation (31), decreased lipid peroxidation by microsomes or microsomes plus cytosol. Effect of Molybdate on Geldanamycin Inhibition of CYP2E1 Degradation A previous report has shown that pretreatment of the rabbit reticulocyte lysate with molybdate could inhibit the subsequent ability of hsp90 to bind geldanamycin (32). To study if the geldanamycin effect on the ability of the cytosol to degrade CYP2E1 could be prevented by molybdate, the cytosol was preincubated with different concentrations of molybdate and MgATP followed by the addition of geldanamycin, and the effect on CYP2E1 degradation was assayed. In the absence of molybdate the addition of 20 ␮M geldanamycin to the cytosol inhibited the degradation of CYP2E1 (Figs. 5A and 5B, lanes ⫹GA compared to lanes cyt). However, if 20 mM molybdate is added before geldanamycin to the cytosol, this inhibition of CYP2E1 degradation by 20 ␮M geldanamycin is prevented (Fig. 5A, lanes ⫹Mo⫹GA compared to lanes ⫹GA). A molybdate concentration curve is shown in Fig. 5B; the inhibition by geldanamycin of CYP2E1 could be prevented by molybdate in a concentration-dependent manner. Fig-

FIG. 4. Geldanamycin does not inhibit lipid peroxidation. Lipid peroxidation was determined by assaying for thiobarbituric acid reactive lipid peroxidation products as described under Materials and Methods. Human liver microsomes (50 ␮g protein) or 160 ␮g cytosolic protein (cyt) were incubated with 1 mM NADPH for 1 h at 37°C and TBA reactive products were determined by assaying absorbance at 532 nm. Geldanamycin, from a stock solution prepared in DMSO, was added at 20 ␮M to the microsomes (mic⫹GA) and the microsomes⫹cytosol (mic⫹cyt⫹GA). EDTA was added at 100 ␮M to the microsomes (mic⫹EDTA) and to the mixture of microsomes⫹cytosol (mic⫹cyt⫹EDTA). Results represent the mean ⫾ SD of four assays and are expressed as OD at 532 nm (ns, not significantly different from mic; a, significantly different from mic at *P ⬍ 0.005; b, significantly different from mic⫹cyt at *P ⬍ 0.005).

326

GOASDUFF AND CEDERBAUM

that the influence of ATP and geldanamycin on proteasome peptidase activity of the cytosol is similar to the effects of these agents on CYP2E1 degradation by the cytosol, although the stimulation effects of ATP or inhibitory effects of geldanamycin are more pronounced with the membrane-bound CYP2E1 substrate than with the soluble peptidase substrate. This could reflect a need for chaperones such as hsp90 for degradation of membrane-bound substrates such as CYP2E1 by the proteasome complex versus a stimulatory but not required effect by chaperones for the proteasome peptidase activity. Studies with Purified 20S Proteasome Degradation of SLLVT-AMC by purified 20S proteasome was not affected by ATP or by geldanamycin (data not shown), in contrast to the effects found for the

FIG. 5. Geldanamycin inhibition of CYP2E1 degradation is reversed by molybdate. (A) Human liver microsomes were incubated at 37°C without cytosol (lane 1; mic) or with cytosol (lanes 2 to 4: cyt). In lanes 5–7, geldanamycin (⫹GA), prepared as a sonicated solution in water was added at 20 ␮M in the cytosol before the MgATP was added. In lanes 8 to 10, a freshly prepared solution of molybdate (20 mM) was added to the cytosol with MgATP in the absence of geldanamycin (⫹Mo). In lanes 11 to 13, molybdate (20 mM) was added to the cytosol prior to the addition of geldanamycin (20 ␮M) and MgATP (⫹Mo⫹GA). The reaction mixture was overlaid with mineral oil and incubated in a shaking water bath at 37°C for 5 h. (B) Human liver microsomes were incubated on ice (lane 1: 0 h) or incubated at 37°C without cytosol (lane 2: mic) or with cytosol (lanes 3 to 5: cyt). In lanes 6 – 8 (⫹GA), geldanamycin prepared as a sonicated solution in water was added at 20 ␮M in the cytosol before the MgATP was added. Molybdate was added from a freshly prepared 0.25 M stock solution to the cytosol at final concentration of 1, 5, and 10 mM with MgATP in the absence of geldanamycin (lanes 9 to 11: ⫹Mo). In lanes 12–14 (⫹Mo⫹GA), molybdate (1, 5, 10 mM) was added to the cytosol before geldanamycin (20 ␮M) and MgATP were added. The reaction mixture was overlaid with mineral oil and incubated in a shaking water bath at 37°C for 5 h. (C) Quantification using ImageQuant Software of the scanned immunoblot from (B). Results are expressed as percentages of remaining CYP2E1 after incubation.

activity. However, the addition of 10 or 20 ␮M geldanamycin prior the addition of a stimulatory concentration of ATP, e.g., 1 mM decreased the proteasome activity by 30 to 50% (Fig. 6A). Figure 6B shows the same type of experiment, using 5 mM MgCl 2 instead of 2.5 mM in the reaction mixture. The proteasome activity was enhanced by ATP with a maximum 3-fold increase observed at 2.5 mM ATP. The addition of 20 ␮M geldanamycin prior to the addition of ATP decreased the proteasome activity by 20 to 30%. These results indicate

FIG. 6. Effect of geldanamycin and ATP on proteasome activity in the cytosol. Proteasome activity in the cytosol (40 ␮g protein) was determined using the fluorogenic peptide Suc-Leu-Leu-Val-Tyr-AMC as described under Materials and Methods. (A) ATP–Tris was added at final concentrations of 0.5, 1, 2, and 5 mM in the presence or absence of 10 or 20 ␮M geldanamycin. MgCl 2 was present at a final concentration of 2.5 mM. Data represent single determinations. (B) ATP–Tris was added at final concentrations of 1, 2.5, and 5 mM in the presence or absence of 20 ␮M geldanamycin. MgCl 2 was present at a final concentration of 5 mM. Results represent the mean ⫾ SD of three enzymatic determinations (ns, not significantly different from ⫺GA; a, significantly different from 0 mM ATP at **P ⬍ 0.005; b, significantly different from ⫺GA at *P ⬍ 0.01).

ROLE OF hsp90 IN THE DEGRADATION OF CYP2E1

peptidase activity of the cytosol. This likely reflects the presence of ATP- and geldanamycin-sensitive components present in the cytosol (e.g., hsp90) but absent from the purified 20S system. As geldanamycin is thought to be a specific inhibitor of hsp90 function, the inhibition of ATP-stimulated proteasome peptidase activity in the cytosol but not purified 20S by this compound could result from an activation effect of hsp90 on the proteasome activity. To examine this possibility, we determined the peptidase activity of the purified 20S in the presence or absence of purified hsp90. Results in Fig. 7A show that hsp90 actually inhibits the rate of cleavage of the fluorogenic substrate by the purified 20S. However, when increasing amounts of hsp90 were added to the cytosol, cleavage of the fluorogenic substrate by proteasome in the cytosol was not inhibited; a small increase by the added hsp90, relative to BSA added as a nonspecific protein control, was observed (Fig. 7B). We also determined the ability of the 20S proteasome to degrade CYP2E1. At identical peptidase activities against SLLVT-AMC, the cytosol was much more effective than 20S proteasome in degrading CYP2E1 (Fig. 8). Clearly, CYP2E1 degradation by the proteasome can be increased by modulating factors present in the cytosol. hsp90 Increases CYP2E1 Degradation by hsp90 Immunodepleted Cytosol The direct role of hsp90 in CYP2E1 degradation was determined by using anti-hsp90 IgG coupled to Sepharose to deplete the cytosol from hsp90 and evaluating the degradation of CYP2E1 by this hsp90 immunodepleted cytosol. The effect of adding back hsp90 to this hsp90 immunodepleted cytosol was also determined. Cytosol treated with these beads resulted in immunodepletion of hsp90 as shown in Fig. 9C; at least 80 to 90% of the content of hsp90 from the cytosol was removed (hsp90 remaining after treatment with antihsp90 IgG–Sepharose, lanes 3 and 4, compared to hsp90 present in the untreated cytosol, lanes 1 and 2 in Fig. 9C). This hsp90-depleted cytosol did not promote CYP2E1 degradation (Fig. 9A, lanes 3 and 4) in contrast to the untreated cytosol (Fig. 9A, lane 2). When purified hsp90 was added back to this hsp90 immunodepleted cytosol, CYP2E1 degradation was restored almost completely (Fig. 9A, lanes 5 and 6). Quantification of these results is shown in Fig. 9B. DISCUSSION

hsp90 is an abundant molecular chaperone, present at an amount equal to 1 to 2% of total cellular protein. Recent evidence suggests that hsp90 requires ATP to facilitate chaperone function (26). The ATP dependence in the mechanism of hsp90 function has emerged

327

as a result of structural and biochemical studies (25, 26), which have demonstrated the presence of an ATP/ ADP binding site in the N-terminal domain of hsp90. The inherent ATPase activity of hsp90 is sensitive to inhibition by the specific inhibitor geldanamycin, which binds to the ATP pocket of hsp90. Two chaperone sites in hsp90 that differ in substrate specificity have been identified (33). The C-terminal fragment recognizes structured substrates and the N-terminal fragment preferentially binds unfolded (poly)peptides. Results in the current report suggest that hsp90 modulates the degradation of CYP2E1 in an in vitro system consisting of human liver microsomes and proteasome containing cytosol. The need for ATP for CYP2E1 degradation (besides possible ubiquitination) could reflect, in part, ATP binding to hsp90, while the requirement for ATP hydrolysis to efficiently degrade CYP2E1 could be linked to the weak ATPase activity of hsp90. Geldanamycin, which binds to the ATP binding pocket of hsp90, inhibits CYP2E1 degradation, suggesting that geldanamycin, by inhibiting the ATPase activity of hsp90, inhibits the CYP2E1 degradation by the cytosol. Control experiments indicate that geldanamycin is not a substrate/ligand for CYP2E1 since it did not competitively inhibit the oxidation of p-nitrophenol by the microsomes. Geldanamycin also did not inhibit microsomal lipid peroxidation, a process which labilizes CYP2E1 for rapid degradation by the proteasome (14). The inhibition by geldanamycin of CYP2E1 degradation is opposite to the increase by geldanamycin of the degradation of tyrosine kinase (27). Whether this reflects differences in the functions of hsp90 in the degradation of microsomal membrane-bound proteins versus soluble proteins is not known. Studies evaluating the effects of geldanamycin on turnover of other microsomal proteins, e.g., CYP3A4, are in progress. Previous experiments have shown that lactacystin, a specific inhibitor of the proteasome complex, prevents the degradation of CYP2E1 by this microsome-pluscytosol reconstituted system, indicating that the proteasome complex is the major proteolytic system responsible for CYP2E1 degradation by this system (14). We compared the cytosolic degradation potential toward microsomal CYP2E1 and toward the cleavage of a fluorogenic compound by cytosolic proteasome. CYP2E1 degradation by cytosol requires ATP and is strongly inhibited by geldanamycin, while the cleavage of fluorogenic peptide by the proteasome in the cytosol was stimulated by ATP and partially inhibited by geldanamycin. This suggests that the cytosol behaves similarly toward CYP2E1 degradation as well as for the cleavage of the fluorogenic peptide, with the more pronounced effects on the microsomal system perhaps reflecting a critical role for molecular chaperones such as hsp90 in the degradation of a membrane-bound protein compared to a soluble peptide.

328

GOASDUFF AND CEDERBAUM

Immunodepletion of hsp90 from the cytosol fraction resulted in the prevention of degradation of CYP2E1, a process which could be restored upon the addition of pure hsp90 to the immunodepleted cytosol. Near-complete restoration was observed with 1 ␮g of hsp90, an amount that approximates the content of hsp90 in the cytosol (assuming a content of hsp90 of 1 to 2% of total cytosolic protein, about 0.4 to 0.8 ␮g hsp90 would be present in the 40 ␮g cytosolic protein used in these assays). The immunodepletion experiments provide the most direct evidence for a role for hsp90 in CYP2E1 degradation by the human liver microsomal/cytosol reconstitution system. Future experiments will be required to extend this concept to intact cellular systems. Several reports have described a role for the proteasome in the degradation of CYP2E1 (12–14). The CYP2E1 degradation by the proteasome is enhanced when microsomes are preincubated with NADPH (14) or with a suicide substrate such as CCl 4 (12). A mechanism-based inactivation of CYP2E1, like that of other P450s involving protein modification, has been described (15, 16). We speculate that the preincubation of the microsomes with NADPH, which increases CYP2E1 degradation, labilizes CYP2E1 protein to such an extent that an unfolded structure is produced. This unfolded conformation may become a preferential substrate for interaction with hsp90, perhaps interacting with the N-terminal fragment, which binds unfolded polypeptides. This interaction of hsp90 with the labilized form of CYP2E1 requires ATP hydrolysis and is inhibited by geldanamycin. hsp90 could prevent protein aggregation and/or could maintain a CYP2E1 unfolded structure that is able to enter into the cavity of the 20S proteasome for proteolysis. Possible interactions between CYP2E1 and hsp90 are currently under investigation. Molybdate has been used to stabilize the hormone competence of receptor– hsp heterocomplexes (34). The stabilizing effect of molybdate on hsp–protein heterocomplexes has been described for a diverse group of proteins, suggesting that molybdate acts directly on hsp90 function rather than on the hsp90 substrate. Molybdate stabilizes the interaction of hsp90 with its p23 cohort (23, 35) and induces a change in the electrophoretic mobility of hsp90 (36). Thus, geldanamycin and molybdate alter the interaction of hsp90 with its substrate by altering the conformation of hsp90. However, these compounds have opposing effects on hsp90 structure and the mode by which hsp90 interacts with its substrates. The molybdate-locked form of hsp90 represents a chaperone conformation by which hsp90 binds to its substrate via hydrophobic interaction, while the geldanamycin-bound form of hsp90 represents a conformation of hsp90 with a weak affinity for substrate (32). The ability of molybdate to increase CYP2E1 degradation by the cytosol and to prevent the

FIG. 7. Effect of hsp90 on catalytic activity of the purified 20S proteasome and cytosolic proteasome. (A) Kinetics of the cleavage of the fluorogenic peptide Suc-Leu-Leu-Val-Tyr-AMC (SLLVT-AMC) by 22 ng/ml purified 20S proteasome (MPC) was determined in the presence or absence of 0.5 ␮g/ml of purified hsp90 in 0.1 M Tris–HCl pH 8.0, 5 mM MgCl 2, and 50 ␮M SLLVT-AMC. Assays were performed at 37°C in a shaking water bath for 0, 10, 20, 40, and 60 min after addition of SLLVT-AMC. (B) Proteasome activity in the cytosol (40 ␮g protein) was determined after addition of increasing amounts (from 1 to 10 ␮g protein) of hsp90 or bovine serum albumin (BSA) as a control. Assays were performed at 37°C in a shaking water bath for 1 h in 0.1 M Tris–HCl, pH 8.0, 5 mM MgCl 2, 50 ␮M SLLVT-AMC in a total volume of 100 ␮l.

geldanamycin inhibition could represent a stabilizing effect on the hsp90 –protein (CYP2E1) interaction, resulting in a favorable conformation for CYP2E1 degradation by the proteasome. It has been shown that the 20S proteasome is more resistant to oxidative stress than the ATP- and ubiquitin-dependent 26S proteasome (37). Hsp90 can protect the 20S proteasome against oxidative inactivation (38) by interacting with certain subunits of the proteasome, as shown by two-dimensional gel electrophoresis patterns after metal-catalyzed oxidation of 20S in the absence or presence of hsp90. hsp90 exhibits a KEKE motif that could interact with the subunit C9 of the proteasome, which also contains a KEKE motif (38, 39). These results suggest that hsp90 may be implicated in proteasome function, as this chaperone can protect the proteasome activity and can interact with the 20S proteasome. We found an increase in the cyto-

ROLE OF hsp90 IN THE DEGRADATION OF CYP2E1

solic enzymatic activity of the proteasome (but not purified 20S activity) toward the fluorogenic peptide SLLVT-AMC by ATP. The ATP-independent cleavage of the fluorogenic peptide is believed to represent the 20S proteasome activity, whereas the ATP-dependent cleavage of the fluorogenic peptide may represent the 26S proteasome activity in vitro (37). The ATP effect on the cleavage of the fluoropeptide by the cytosol may reflect metabolism by the ATP-dependent 26S activity in addition to the ATP-independent 20S proteasome. The ATP enhancement of the cleavage of the fluoropeptide is partially inhibited by geldanamycin, implicating hsp90 in the interaction with the proteasome complex and in activation of its activity. This, however, is in contrast with the result showing that hsp90 inhibits the purified 20S peptidase activity. Tsubuki et al. (40) have demonstrated that hsp90 is an inhibitor of proteasome activity (assayed by the cleavage of the fluoropeptide N-Cbz-Leu-Leu-Leu-AMC) as catalyzed by the purified bovine brain multi-proteinase complexes (MPC). It should be pointed out that activity of the 20S proteasome is very sensitive to various treatments, e.g., it can be activated by binding with PA28, heating, freezing, storage in the absence of glycerol, or dialysis against water (41– 45). Purified 20S proteasome activity but not cytosolic proteasome activity can be inactivated by low concentrations of sodium (data not shown). The manner by which these treatments affect the purified 20S is in contrast to the activity of the proteasome in the cytosol. We did not observe inhibition of cytosolic proteasome activity when hsp90 was added. In this respect, degradation of CYP2E1 by the proteasome present in the cytosol was much more effective than by identical peptidase units of the purified 20S, indicating modulation of proteasome activity in the cytosol. These results suggest that hsp90 is involved in the overall mechanism of degradation of microsomal

FIG. 8. Comparison of CYP2E1 degradation by human liver cytosol versus purified 20S proteasome (MPC). Human liver microsomes (10 ␮g protein) were incubated on ice (lane 1: 0 h) or incubated at 37°C for 5 h without cytosol (lane 2: mic), with 40 ␮g cytosolic protein (lanes 3–5) or with 22 ng protein of purified 20S proteasome (lanes 6 – 8: 20S) in 50 mM phosphate buffer, pH 7.4, 1.2 mM MgCl 2 and 0.6 mM ATP–Tris (the peptidase activity as determined by the cleavage of the fluorogenic compound SLLVT-AMC was equivalent for both cytosolic protein and purified 20S proteasome protein content used in this study). Western blot was performed using an aliquot of the boiled sample (10 ␮l).

329

FIG. 9. Effect of addition of hsp90 on CYP2E1 degradation by hsp90-depleted cytosol. Immunodepletion of hsp90 was performed by incubating cytosol with anti-hsp90 IgG coupled to Sepharose as described in Materials and Methods. This hsp90 immunodepleted cytosol was used in the CYP2E1 degradation assay and for hsp90 immunoblot. (A) Human liver microsomes were incubated at 37°C without (lane 1) or with cytosol (lane 2) or with anti-hsp90 IgG– Sepharose treated cytosol (lanes 3 and 4) or with anti-hsp90 IgG– Sepharose-treated cytosol in which 1 ␮g of purified hsp90 was added (lanes 5 and 6). The reaction mixture was overlaid with mineral oil and incubated in a shaking water bath at 37°C for 5 h. The reaction was stopped by addition of SDS–PAGE loading buffer and Western blot was performed using an aliquot (10 ␮l) of the boiled samples (one representative immunoblot of three different experiments is shown). (B) Quantification of immunoblots using ImageQuant Software. Results are expressed as a percentage of remaining CYP2E1 after 5 h of incubation and represent the mean ⫾ SD of four different assays. (C) Hsp90 immunoblot analysis using 10 ␮g cytosolic protein (lanes 1 and 2) and 10 ␮g cytosolic protein after treatment with anti-hsp90 IgG–Sepharose (lane 3 and 4).

CYP2E1 in vitro, but whether hsp90 is involved by interacting with CYP2E1, the proteasome complex, or both remains to be elucidated. The intervention of hsp90 in the degradation process of CYP2E1 could be important to modulate the level of this potentially damaging enzyme. CYP2E1, a loosely coupled enzyme, is known to produce reactive oxygen species and to induce a state of cellular oxidative stress (46, 47). We have previously shown that overexpression of CYP2E1 is toxic to HepG2 cells (48) and speculated that one consequence of the increased labilization of this P450 by oxidative stress is to promote its rapid removal by proteasome-mediated degradation (14). Hsp90, which can be induced under diverse conditions of cellular

330

GOASDUFF AND CEDERBAUM

stress, could provide an important feedback mechanism to help remove this enzyme by proteasome-mediated proteolysis.

ACKNOWLEDGMENTS This work was supported by a grant from the Alcoholic Beverage Medical Research Foundation. We thank Dr. Avram Caplan, Department of Cell Biology, Mount Sinai Medical Center, for helpful discussion and advice.

REFERENCES 1. Ronis, M. J. J., Lindros, K. O., and Ingelman-Sundberg, M. (1996) in Cytochrome P450: Metabolic and Toxicological Aspects (Ionnides, C., Ed.), pp. 211–239. CRC Press, New York. 2. Koop, D. (1992) FASEB J. 6, 724 –729. 3. Guengerich, F. P. (1991) Chem. Res. Toxicol. 4, 391– 407. 4. Hu, Y., Ingelman-Sundberg, M., and Lindros, K. O. (1995) Biochem. Pharmacol. 50, 155–161. 5. Song, B. J., Matsunaga, T., Hardwick, J. P., Park, S. S., Veech, R. L., Yang, C. S., Gelboin, H. V., and Gonzalez, F. J. (1987) Mol. Endocrinol. 1, 542–547. 6. Guengerich, F. P., Kim, D. H., and Iwasaki, M. (1991) Chem. Res. Toxicol. 4, 168 –179. 7. Koop, D. R., and Tierney, D. J. (1990) BioEssays 12, 429 – 435. 8. Eliasson, E., Johansson, I., and Ingelman-Sundberg, M. (1988) Biochem. Biophys. Res. Commun. 150, 436 – 443. 9. Song, B. J., Veech, R. L., Park, S. S., Gelboin, H. V., and Gonzalez, F. J. (1989) J. Biol. Chem. 264, 3568 –3572. 10. Roberts, B. J., Song, B. J., Soh, Y., Park, S. S., and Shoaf, J. E. (1995) J. Biol. Chem. 270, 29632–29635. 11. Tierney, D. J., Haas, A. L., and Koop, D. R. (1992) Arch. Biochem. Biophys. 293, 9 –16. 12. Roberts, B. J. (1997) J. Biol. Chem. 272, 9771–9778. 13. Yang, M. X., and Cederbaum, A. I. (1996) Biochem. Biophys. Res. Commun. 226, 711–716. 14. Goasduff, T., and Cederbaum, A. I. (1999) Arch. Biochem. Biophys. 370, 258 –270. 15. Korsmeyer, K. K., Davoll, S., Figueiredo-Pereira, M. E., and Correia, M. A. (1999) Arch. Biochem. Biophys. 365, 31– 44. 16. Wang, H. F., Figueiredo-Pereira, M. E., and Correia, M. A. (1999) Arch. Biochem. Biophys. 365, 45–53. 17. Mathew, A., and Morimoto, R. I. (1998) Ann. NY Acad. Sci. 851, 99 –111. 18. Hayes, S. A., and Dice, F. J. (1996) J. Cell Biol. 132, 255–258. 19. Sherman, M. Y. S., and Goldberg, L. A. (1996) in Stress-Inducible Cellular Responses (Feige, U., Morimoto, R. I., Yahara, I., and Polla, B., Eds.), pp. 57–78. Birkhauser Verlag, Boston. 20. Gething, M. J., and Sambrook, J. (1992) Nature 355, 33– 45. 21. Schirmer, E. C., Glover, J. R., Singer, M. A., and Lidquist, S. (1996) Trends Biochem. Sci. 19, 87– 89. 22. Johnson, J. L., and Toft, D. O. (1994) J. Biol. Chem. 269, 24989 – 24993.

23. Sullivan, W., Stensgard, B., Caucutt, G., Bartha, B., McMahon, N., Alnemri, E. S., Litwack, G., and Toft, D. O. (1997) J. Biol. Chem. 272, 8007– 8012. 24. Grenert, J. P., Sullivan, W. P., Fadden, P., Haystead, T. A. J., Clark, J., Mimnaugh, E., Krutzsch, H., Oche, H.-J., Schulte, T. W., Sausville, E., Neckers, L. M., and Toft, D. O. (1997) J. Biol. Chem. 272, 23843–23850. 25. Prodromou, C., Roe, S. M., O’Brien, R., Ladbury, J. E., Piper, P. W., and Pearl, L. H. (1997) Cell 90, 65–75. 26. Panaretou, B., Prodromou, C., Roe, S. M., O’Brien, R., Ladbury, J. E., Piper, P. W., and Pearl, L. H. (1998) EMBO J. 17, 4829 – 4836. 27. Sakagami, M., Morrison, P., and Welch, W. J. (1999) Cell Stress Chaperones 4, 19 –28. 28. Eliasson, E., Mkrtchian, S., and Ingelman-Sundberg, M. (1992) J. Biol. Chem. 267, 15765–15769. 29. Eliasson, E., Johansson, I., and Ingelman-Sundberg, M. (1990) Proc. Natl. Acad. Sci. USA 87, 3225–3227. 30. Zhukov, A., Werlinder, V., and Ingelman-Sundberg, M. (1993) Biochem. Biophys. Res. Commun. 197, 221–228. 31. Aust, S. D., Morehouse, L. A., and Thomas, C. E. (1995) Free Radical Biol. Med. 1, 3–25. 32. Hartson, S. D., Thulasiraman, V., Huang, W., Whitesell, L., and Matts, R. L. (1999) Biochemistry 38, 3837–3849. 33. Scheibel, T., Weikl, T., and Buchner, J. (1998) Proc. Natl. Acad. Sci. U.S.A. 95, 1495–1499. 34. Pratt, W. B., and Toft, D. O. (1997) Endocrinol. Rev. 18, 306 – 360. 35. Johnson, J. L., and Toft, D. O. (1995) Mol. Endocrinol. 9, 670 – 678. 36. Soti, C., Radics, L., Yahara, I., and Csermely, P. (1998) Eur. J. Biochem. 255, 611– 617. 37. Reinheckel, T., Sitte, N., Ullrich, O., Kuckelkorn, U., Davies, K. J. A., and Grune, T. (1998) Biochem. J. 335, 637– 642. 38. Conconi, M., Petropoulos, I., Emod, I., Turlin, E., Biville, F., and Friguet, B. (1998) Biochem. J. 333, 407– 415. 39. Realini, C., and Rechsteiner, M. (1995) J. Biol. Chem. 207, 29664 –29667. 40. Tsubuki, S., Yumiko, S., and Kawashima, S. (1994) FEBS Lett. 344, 229 –233. 41. Chu-Ping, M., Vu, J. H., Proske, R. J., Slaughter, C. A., and DeMartino, G. N. (1994) J. Biol. Chem. 269, 3539 –3547. 42. Dubiel, W., Pratt, G., Ferell, K., and Rechsteiner, M. (1992) J. Biol. Chem. 267, 22369 –22377. 43. Tanaka, K., Li, K., Ichihara, A., Waxman, L., and Goldberg, A. L. (1986) J. Biol. Chem. 261, 15197–15203. 44. Tokumoto, T., and Ichikawa, K. (1993) Biochem. Biophys. Res. Commun. 192, 1106 –1114. 45. McGuire, M. J. M., Mc Cullough, M. L., Croal, D. E., and DeMartino, G. N. (1989) Biochim. Biophys. Acta 995, 181– 186. 46. Ekstrom, G., and Ingelman-Sundberg, M. (1989) Biochem. Pharmacol. 38, 1313–1318. 47. Gorsky, L. D., Koop, D. R., and Coon, M. J. (1984) J. Biol. Chem. 259, 6812– 6817. 48. Chen, Q., and Cederbaum, A. I. (1998) Mol. Pharmacol. 53, 638 – 648.