A Principal Role for the Proteasome in Endoplasmic Reticulum-associated Degradation of Misfolded Intracellular Cystic Fibrosis Transmembrane Conductance Regulator
2002; Elsevier BV; Volume: 277; Issue: 14 Linguagem: Inglês
10.1074/jbc.m111958200
ISSN1083-351X
AutoresMarina S. Gelman, Elisa Sandoval Kannegaard, Ron R. Kopito,
Tópico(s)Cellular transport and secretion
ResumoEndoplasmic reticulum-associated degradation of misfolded cystic fibrosis transmembrane conductance regulator (CFTR) protein is known to involve the ubiquitin-proteasome system. In addition, an ATP-independent proteolytic system has been suggested to operate in parallel with this pathway and become up-regulated when proteasomes are inhibited (Jensen, T. J., Loo, M. A., Pind, S., Williams, D. B., Goldberg, A. L., and Riordan, J. R. (1995) Cell 83, 129–135). In this study, we use two independent techniques, pulse-chase labeling and a noninvasive fluorescence cell-based assay, to investigate the proteolytic pathways underlying the degradation of misfolded CFTR. Here we report that only inhibitors of the proteasome have a significant effect on preventing the degradation of CFTR, whereas cell-permeable inhibitors of lysosomal degradation, autophagy, and several classes of protease had no measurable effect on CFTR degradation, when used either alone or in combination with the specific proteasome inhibitor carbobenzoxy-l-leucyl-leucyl-l-leucinal (MG132). Our results suggest that ubiquitin-proteasome-mediated degradation is the dominant pathway for disposal of misfolded CFTR in mammalian cells and provide new mechanistic insight into endoplasmic reticulum-associated degradation. Endoplasmic reticulum-associated degradation of misfolded cystic fibrosis transmembrane conductance regulator (CFTR) protein is known to involve the ubiquitin-proteasome system. In addition, an ATP-independent proteolytic system has been suggested to operate in parallel with this pathway and become up-regulated when proteasomes are inhibited (Jensen, T. J., Loo, M. A., Pind, S., Williams, D. B., Goldberg, A. L., and Riordan, J. R. (1995) Cell 83, 129–135). In this study, we use two independent techniques, pulse-chase labeling and a noninvasive fluorescence cell-based assay, to investigate the proteolytic pathways underlying the degradation of misfolded CFTR. Here we report that only inhibitors of the proteasome have a significant effect on preventing the degradation of CFTR, whereas cell-permeable inhibitors of lysosomal degradation, autophagy, and several classes of protease had no measurable effect on CFTR degradation, when used either alone or in combination with the specific proteasome inhibitor carbobenzoxy-l-leucyl-leucyl-l-leucinal (MG132). Our results suggest that ubiquitin-proteasome-mediated degradation is the dominant pathway for disposal of misfolded CFTR in mammalian cells and provide new mechanistic insight into endoplasmic reticulum-associated degradation. The endoplasmic reticulum (ER) 1The abbreviations used are: ERendoplasmic reticulumERADendoplasmic reticulum-associated degradationUPSubiquitin-proteasome systemTPCKtosylphenylalanyl chloromethyl ketoneCHOChinese hamster ovary, CFTR, cystic fibrosis transmembrane conductance regulatorGFPgreen fluorescent proteincl.2clone 2HMWhigh molecular weightUbubiquitinCT-Lchymotrypsin-like is the site of synthesis of membrane and secretory proteins and the site of their conformational maturation and assembly into correctly folded functional molecules. "Quality control" refers to the system that monitors the folding and assembly of proteins in the ER, ensuring that only correctly folded proteins can mature to later compartments of the secretory pathway (1.Helenius A. Philos. Trans. R. Soc. Lond. B Biol. Sci. 2001; 356: 147-150Crossref PubMed Scopus (49) Google Scholar, 2.Ellgaard L. Helenius A. Curr. Opin. Cell Biol. 2001; 13: 431-437Crossref PubMed Scopus (335) Google Scholar). Nascent proteins that fail to fold or assemble in the ER are degraded by a process known as ER-associated degradation (ERAD) (3.Brodsky J.L. McCracken A.A. Semin. Cell Dev. Biol. 1999; 10: 507-513Crossref PubMed Scopus (300) Google Scholar). Studies in mammalian cells and in yeast have led to the formulation of a general model for ERAD in which substrates are degraded by cytoplasmic proteasomes following retrotranslocation (i.e. "dislocation") across the ER membrane by a process that requires the Sec61 translocon and functional ubiquitylation machinery at the cytosolic face of the ER membrane (4.Kopito R.R. Cell. 1997; 88: 427-430Abstract Full Text Full Text PDF PubMed Scopus (483) Google Scholar, 5.Plemper R.K. Wolf D.H. Mol. Biol. Rep. 1999; 26: 125-130Crossref PubMed Google Scholar, 6.Johnson A.E. Haigh N.G. Cell. 2000; 102: 709-712Abstract Full Text Full Text PDF PubMed Scopus (69) Google Scholar). Dislocation of ERAD substrates across the ER membrane is kinetically coupled to their degradation by proteasomes, and in many cases, cytosolic intermediates of degradation are hard to detect (7.Yu H. Kopito R.R. J. Biol. Chem. 1999; 274: 36852-36858Abstract Full Text Full Text PDF PubMed Scopus (103) Google Scholar, 8.O'Hare T. Wiens G.D. Whitcomb E.A. Enns C.A. Rittenberg M.B. J. Immunol. 1999; 163: 11-14PubMed Google Scholar). Although a central role for proteasomes in ERAD is widely accepted, a number of studies have suggested that other, nonproteasomal, proteolytic systems may also contribute to the elimination of misfolded proteins from the ER (9.Jensen T.J. Loo M.A. Pind S. Williams D.B. Goldberg A.L. Riordan J.R. Cell. 1995; 83: 129-135Abstract Full Text PDF PubMed Scopus (775) Google Scholar, 10.Loayza D. Michaelis S. Mol. Cell. Biol. 1998; 18: 779-789Crossref PubMed Google Scholar, 11.Liao W. Chan L. Biochem. J. 2001; 353: 493-501Crossref PubMed Scopus (32) Google Scholar, 12.Umebayashi K. Fukuda R. Hirata A. Horiuchi H. Nakano A. Ohta A. Takagi M. J. Biol. Chem. 2001; 276: 41444-41454Abstract Full Text Full Text PDF PubMed Scopus (16) Google Scholar). endoplasmic reticulum endoplasmic reticulum-associated degradation ubiquitin-proteasome system tosylphenylalanyl chloromethyl ketone Chinese hamster ovary, CFTR, cystic fibrosis transmembrane conductance regulator green fluorescent protein clone 2 high molecular weight ubiquitin chymotrypsin-like The polytopic membrane protein cystic fibrosis transmembrane conductance regulator (CFTR) is a chloride channel expressed at the apical surface of polarized epithelial cells and one of the first integral membrane proteins demonstrated to be a substrate of the ubiquitin-proteasome system (9.Jensen T.J. Loo M.A. Pind S. Williams D.B. Goldberg A.L. Riordan J.R. Cell. 1995; 83: 129-135Abstract Full Text PDF PubMed Scopus (775) Google Scholar, 13.Ward C.L. Omura S. Kopito R.R. Cell. 1995; 83: 121-127Abstract Full Text PDF PubMed Scopus (1133) Google Scholar). Mutations in the CFTR gene cause the recessively inherited fatal disease cystic fibrosis. Although nearly 1000 mutations have been linked to cystic fibrosis, the majority of Caucasian cystic fibrosis patients have at least one copy of the ΔF508 mutation, a temperature-sensitive allele that is unable to fold correctly at physiological temperature. As a consequence, ΔF508 CFTR (ΔF508) is not deployed to the plasma membrane and is instead rapidly degraded without being processed in the Golgi apparatus (14.Kopito R.R. Physiol. Rev. 1999; 79: S167-S173Crossref PubMed Scopus (375) Google Scholar, 15.Bannykh S.I. Bannykh G.I. Fish K.N. Moyer B.D. Riordan J.R. Balch W.E. Traffic. 2000; 1: 852-870Crossref PubMed Scopus (56) Google Scholar). Several lines of evidence support a role for the ubiquitin-proteasome pathway in ΔF508 turnover. First, CFTR undergoes both co-translational and posttranslational ubiquitylation when expressed in a cell-free system (16.Sato S. Ward C.L. Kopito R.R. J. Biol. Chem. 1998; 273: 7189-7192Abstract Full Text Full Text PDF PubMed Scopus (119) Google Scholar, 17.Xiong X. Chong E. Skach W.R. J. Biol. Chem. 1999; 274: 2616-2624Abstract Full Text Full Text PDF PubMed Scopus (97) Google Scholar). Second, inhibition of proteasome function with chemical inhibitors or with a dominant negative form of ubiquitin leads to accumulation of multiubiquitylated, undegraded forms of CFTR and ΔF508 (13.Ward C.L. Omura S. Kopito R.R. Cell. 1995; 83: 121-127Abstract Full Text PDF PubMed Scopus (1133) Google Scholar). Third, degradation of ΔF508 in a cell-free system is sensitive to inhibitors of both the 19 S and 20 S subunits of the proteasome (17.Xiong X. Chong E. Skach W.R. J. Biol. Chem. 1999; 274: 2616-2624Abstract Full Text Full Text PDF PubMed Scopus (97) Google Scholar, 18.Oberdorf J. Carlson E.J. Skach W.R. Biochemistry. 2001; 40: 13397-13405Crossref PubMed Scopus (33) Google Scholar). Curiously, despite this compelling evidence in support of a role for proteasomes in the degradation of misfolded CFTR, the effect of proteasome inhibition on the disappearance of ΔF508 from pulse-chase studies is minimal (9.Jensen T.J. Loo M.A. Pind S. Williams D.B. Goldberg A.L. Riordan J.R. Cell. 1995; 83: 129-135Abstract Full Text PDF PubMed Scopus (775) Google Scholar, 13.Ward C.L. Omura S. Kopito R.R. Cell. 1995; 83: 121-127Abstract Full Text PDF PubMed Scopus (1133) Google Scholar). This observation has led some investigators to conclude that other proteolytic systems are primarily responsible for the ERAD of misfolded CFTR molecules (9.Jensen T.J. Loo M.A. Pind S. Williams D.B. Goldberg A.L. Riordan J.R. Cell. 1995; 83: 129-135Abstract Full Text PDF PubMed Scopus (775) Google Scholar). However, no other proteases or proteolytic systems have been convincingly demonstrated to contribute to the degradation of CFTR or other ERAD substrates. Therefore, the intracellular fate of ΔF508 and the identity of the proteases that participate in its degradation remain unresolved. In this study, we used a reporter (GFP-ΔF508) consisting of a fusion of GFP with ΔF508 (19.Moyer B.D. Loffing J. Schwiebert E.M. Loffing-Cueni D. Halpin P.A. Karlson K.H. Ismailov I.I. Guggino W.B. Langford G.M. Stanton B.A. J. Biol. Chem. 1998; 273: 21759-21768Abstract Full Text Full Text PDF PubMed Scopus (138) Google Scholar) to assess the participation of proteasomes and other proteolytic systems in ΔF508 degradation. Our data establish that proteolytic activities of the proteasome are responsible for vast majority of intracellular degradation of misfolded CFTR molecules. Most chemicals (with the exception of some proteasome inhibitors) were obtained from Sigma Chemical Co. Epoxomicin and carbobenzoxy-l-leucyl-leucyl-l-leucinal (MG132) were obtained from Calbiochem. YU102 was generously provided by Dr. Graig Crews (Yale University, New Haven, CT). Polyclonal anti-GFP antibody was a kind gift from Dr. Michael Rexach (Stanford University). pGFP-CFTR in a pEGFP-C1 vector (CLONTECH) was obtained from Dr. B. Stanton (Dartmouth Medical School, Hanover, NH) and subsequently subcloned into pcDNA3.1 vector (Invitrogen) using NheI and EcoRV. ΔF508 mutation was introduced by replacing the [Pml1-Blp1] cassette within the CFTR sequence. For selection of stable CHO cell lines, GFP-ΔF508/pcDNA3 plasmid was linearized and transfected into CHO cells by calcium phosphate precipitation. The cells were subjected to selection in G418-containing medium. Stable transfectants were expanded in culture and enriched for GFP-ΔF508-expressing cells by fluorescence-activated cell sorting of highly fluorescent cells after overnight incubation of these cells with the proteasome inhibitor N-acetyl-l-leucinyl-l-leucinyl-l-norleucinal. Cells were then cloned into 96-well dishes at a density of 1 cell/well. Clonal cell lines were expanded and analyzed for full-length GFP-ΔF508 expression by immunoblotting, and a single clone (clone 2) was selected for this study. Generation of human embryonic kidney HEK293 cells stably expressing GFPu was described elsewhere (20.Bence N.F. Sampat R.M. Kopito R.R. Science. 2001; 292: 1552-1555Crossref PubMed Scopus (1833) Google Scholar). CHO clone 2 cells were plated on 10 × 10-cm dishes and treated with 5 mm butyrate overnight to increase GFP-ΔF508 expression. The next day, cells were starved in cysteine/methionine-free media for 30 min and then labeled with [35S]cysteine/methionine (specific activity, 1 mCi/ml) for 30 min in the presence of 5 mm butyrate. Two 10-cm dishes were harvested immediately after the pulse (t = 0 h), whereas the remaining dishes were chased in the presence of the protein synthesis inhibitor emetine (75 μm) or emetine and MG132 (25 μm) together. At specified time intervals, cells were solubilized in buffer A (10 mm Tris, pH 7.5, 5 mm EDTA, 150 mm NaCl, 1% NP-40, 0.05% deoxycholate) containing a CompleteTM protease inhibitor mixture (Roche Diagnostics GmbH) for 30 min. Cell lysates were precleared with protein A-Sepharose beads and incubated with anti-GFP antibody and protein A-Sepharose. The immunoprecipitates were washed extensively with buffer A and fractionated by 4–15% gradient SDS-PAGE (Bio-Rad). Quantitation of radioactive bands was done using PhosphorImager SI and ImageQuant software (Molecular Dynamics). Analysis of GFP-ΔF508 CFTR fluorescence by flow cytometry was performed using Coulter® Epics® XL-MCL Flow Cytometer and EXPO v.2 cytometer software. Cells were typically incubated with 5 mm butyrate overnight to increase GFP-ΔF508 CFTR expression or with 10 μg/ml N-acetyl-l-leucinyl-l-leucinyl-l-norleucinal for 3 h to increase GFPu steady-state level. After the addition of emetine with or without protease inhibitors, cultures were incubated for different time intervals before analysis. Data from 10,000 cells were collected, and the mean fluorescence of the cell population was used as a measure of GFP-ΔF508 levels after each treatment. CHO cl.2 cells expressing GFP-ΔF508 and HEK293 cells expressing GFPu were ATP-depleted by incubating the cells in glucose-free medium (Dulbecco's modified Eagle's medium; Invitrogen) containing the metabolic inhibitors cyanide (5 mm) and 2-deoxy-d-glucose (5 mm) as described previously (21.Wang Y.H. Li F. Schwartz J.H. Flint P.J. Borkan S.C. Am. J. Physiol. Cell Physiol. 2001; 281: C1667-C1675Crossref PubMed Google Scholar). This procedure reduces ATP content below 10% of the baseline value within 10 min and sustains this low level of ATP content (22.Wang Y.H. Borkan S.C. Am. J. Physiol. 1996; 270: F1057-F1065PubMed Google Scholar). Based on this determination, the initial fluorescence measurement in ATP-depleted cells (t = 0 h) was taken at 10 min after exposure of cells to metabolic inhibitors. To validate the use of GFP-ΔF508 as a model of ΔF508 degradation, we performed pulse-chase analysis in a clonal CHO cell line (cl.2) stably expressing this fusion protein. cl.2 cells were selected for low basal expression levels to avoid accumulation of GFP-ΔF508 in detergent-insoluble aggresomes, which are highly refractory to degradation (23.Johnston J.A. Ward C.L. Kopito R.R. J. Cell Biol. 1998; 143: 1883-1898Crossref PubMed Scopus (1786) Google Scholar). Before use, the cells were treated with the transcriptional activator sodium butyrate to increase GFP-ΔF508 expression. Under these conditions, aggregation of GFP-ΔF508 was undetectable by either immunoblot or fluorescence microscopy (data not shown). Immunoprecipitation of metabolically labeled cell lysates with anti-GFP antibody revealed a major Mr ∼170,000 band corresponding to core-glycosylated GFP-ΔF508 and high molecular weight (HMW) species migrating near the top of the gel (Fig. 1A). During the chase, the Mr 170,000 band disappeared with first-order kinetics (t½ = 50 min), consistent with previous determinations of ΔF508 (9.Jensen T.J. Loo M.A. Pind S. Williams D.B. Goldberg A.L. Riordan J.R. Cell. 1995; 83: 129-135Abstract Full Text PDF PubMed Scopus (775) Google Scholar, 24.Ward C.L. Kopito R.R. J. Biol. Chem. 1994; 269: 25710-25718Abstract Full Text PDF PubMed Google Scholar, 25.Lukacs G.L. Mohamed A. Kartner N. Chang X.B. Riordan J.R. Grinstein S. EMBO J. 1994; 13: 6076-6086Crossref PubMed Scopus (342) Google Scholar) and GFP-ΔF508 half-life (23.Johnston J.A. Ward C.L. Kopito R.R. J. Cell Biol. 1998; 143: 1883-1898Crossref PubMed Scopus (1786) Google Scholar). In the presence of MG132, a specific proteasome inhibitor (26.Rock K.L. Gramm C. Rothstein L. Clark K. Stein R. Dick L. Hwang D. Goldberg A.L. Cell. 1994; 78: 761-771Abstract Full Text PDF PubMed Scopus (2206) Google Scholar), the half-life of this species was only modestly extended (t½ = 1.5 h) (Fig. 1B, top left panel). A minor band migrating at Mr∼50,000 probably corresponds to an amino-terminal fragment of the GFP-Δ508 fusion protein comprising Mr∼20,000 of CFTR in addition to the GFP moiety because it reacted with antibody to GFP, but not with an antibody raised against the carboxyl terminus of CFTR (data not shown). Disappearance of the Mr 50,000 fragment paralleled that of the Mr 170,000 band (t½ = 1.1 h) and was also only modestly inhibited by MG132 (t½ = 1.5 h) (Fig. 1B, bottom left panel). HMW forms of GFP-ΔF508 were detected immediately after the pulse (top band in Fig. 1A). These complexes disappeared rapidly when the chase was conducted in the absence of proteasome inhibitor (t½ = 0.9 h), but in contrast to the core-glycosylated protein and the Mr 50,000 fragment, they were strongly stabilized by the presence of MG132 (t½ = 6.6 h) (Fig. 1B, top right panel). The HMW complexes could represent either polyubiquitylated forms of monomeric GFP-ΔF508 (16.Sato S. Ward C.L. Kopito R.R. J. Biol. Chem. 1998; 273: 7189-7192Abstract Full Text Full Text PDF PubMed Scopus (119) Google Scholar, 17.Xiong X. Chong E. Skach W.R. J. Biol. Chem. 1999; 274: 2616-2624Abstract Full Text Full Text PDF PubMed Scopus (97) Google Scholar) or detergent-insoluble aggregates of CFTR (23.Johnston J.A. Ward C.L. Kopito R.R. J. Cell Biol. 1998; 143: 1883-1898Crossref PubMed Scopus (1786) Google Scholar). To distinguish between these possibilities, we sought to determine whether the HMW material that accumulated during the chase in the presence of MG132 could be degraded after washout of the inhibitor. Our results (data not shown) indicate that the HMW material was soluble in nonionic detergent and competent for degradation, strongly suggesting that the increase in molecular weight was due to covalent modification with Ub, as observed previously for CFTR in cell-free extracts (17.Xiong X. Chong E. Skach W.R. J. Biol. Chem. 1999; 274: 2616-2624Abstract Full Text Full Text PDF PubMed Scopus (97) Google Scholar). Quantification of the pulse-chase data indicates that MG132 was only marginally effective at stabilizing the Mr170,000 ΔF508 band. Even when the total amount of GFP-immunoreactive material is considered (the sum of all electrophoretic species, Fig. 1B, bottom right panel), only 40% of the initial amount of GFP-ΔF508 was stable at the end of a 4-h chase, despite proteasome inhibition. This observation suggests several plausible explanations. First, as suggested by Jensen et al. (9.Jensen T.J. Loo M.A. Pind S. Williams D.B. Goldberg A.L. Riordan J.R. Cell. 1995; 83: 129-135Abstract Full Text PDF PubMed Scopus (775) Google Scholar), in addition to proteasomes, other proteolytic systems could also contribute to ΔF508 ERAD. Alternatively, it is possible that some proteolytic activities within the proteasome remain uninhibited in the presence of MG132 and continue to degrade the substrate. Finally, immunoprecipitation of radiolabeled material from detergent extracts of pulse-labeled cells may underestimate the amount of stable ΔF508 if some of the protein becomes insoluble, if the epitope used for immunoprecipitation becomes inaccessible, or if the increased molecular weight (due to polyubiquitylation and/or aggregation) of the protein prevents it from being well resolved by SDS-PAGE. We reasoned that a noninvasive method of monitoring the fluorescence of GFP-ΔF508 in intact cells using fluorescence-activated cell-sorting analysis may be a more accurate way to measure protein stability because it does not rely upon efficient recovery of radiolabeled protein from cells or upon the detection of a resolvable electrophoretic species. This would in turn allow us to better evaluate the effect of protease inhibitors on prevention of CFTR degradation. To circumvent the limitations of the pulse-chase approach, we developed a fluorescence-based assay to measure degradation of GFP-ΔF508. Inhibition of GFP-ΔF508 synthesis in butyrate-treated cells with emetine, an inhibitor of translational elongation, caused a rapid decrease in mean fluorescence, as measured by flow cytometry (Fig. 2). Similar results were obtained with other translation inhibitors including cycloheximide and puromycin (data not shown). Emetine does not affect the degradation kinetics of CFTR or ΔF508 (13.Ward C.L. Omura S. Kopito R.R. Cell. 1995; 83: 121-127Abstract Full Text PDF PubMed Scopus (1133) Google Scholar). The decrease in GFP-ΔF508 fluorescence fitted a first-order exponential with a t½ = 60 min, in good agreement with the GFP-ΔF508 half-life measured in pulse-chase experiments (Fig. 2). In a control experiment, CHO cells expressing GFP alone were incubated with emetine for up to 10 h without any loss of total fluorescence (data not shown), consistent with GFP being a stable protein (20.Bence N.F. Sampat R.M. Kopito R.R. Science. 2001; 292: 1552-1555Crossref PubMed Scopus (1833) Google Scholar) and indicating that emetine treatment does not cause detectable cell lysis during the course of the experiment. Together, these results suggest that the decline in the fluorescence of cl.2 cells upon exposure to protein synthesis inhibitor faithfully reflects the degradation of ΔF508 and constitutes a valid and convenient method for characterization of this degradation process. To assess the effect of inhibition of different proteolytic systems on the degradation of mutant CFTR using the fluorescence assay, we evaluated the effect of a panel of protease inhibitors on the decline in GFP-ΔF508 fluorescence in butyrate-treated cl.2 cells in the presence of emetine (Fig. 3A). Among the inhibitors tested, only those that are known to inhibit the proteasome (N-acetyl-l-leucinyl-l-leucinyl-l-norleucinal, clasto-lactacystin β-lactone, and MG132) significantly slowed the degradation of GFP-ΔF508. In contrast, inhibitors of serine and cysteine proteases (TPCK, Nα-tosyl-Lys-chloromethyl ketone, and phenylmethylsulfonyl fluoride), calpains (EDTA), lysosomal cathepsins ((2S,3S)-trans-epoxysuccinyl-l-leucylamido-3-methylbutane ethyl ester, chloroquine, and NH4Cl), and autophagy (3-methyladenine and wortmannin) had no measurable effect on the degradation process, either alone (Fig. 3A) or in combination with MG132 (25 μm) (data not shown). These data confirm a central role for proteasomes in the degradation of ΔF508 and suggest that other major proteolytic systems do not contribute significantly to this process. In the presence of MG132 (25 μm), the fluorescence of GFP-ΔF508 first increased, reaching almost 120% of the initial fluorescence after 1 h of emetine chase (Fig. 3B). This increase probably reflects slow fluorogenesis of the GFP moiety (t½ = 30–90 min; 27.Crameri A. Whitehorn E.A. Tate E. Stemmer W.P. Nat. Biotechnol. 1996; 14: 315-319Crossref PubMed Scopus (1062) Google Scholar, 28.Waldo G.S. Standish B.M. Berendzen J. Terwilliger T.C. Nat. Biotechnol. 1999; 17: 691-695Crossref PubMed Scopus (725) Google Scholar) in the absence of CFTR degradation. Indeed, this transient increase in fluorescence is also observed with GFPu, a GFP molecule that is targeted to the ubiquitin-proteasome system for degradation via a short carboxyl-terminal degron (20.Bence N.F. Sampat R.M. Kopito R.R. Science. 2001; 292: 1552-1555Crossref PubMed Scopus (1833) Google Scholar) (Fig. 3B). However, in contrast to GFPu, which was completely stabilized by 25 μm MG132, the fluorescence of GFP-ΔF508 started to decline after the first hour and dropped to about 70% of initial fluorescence after 4 h of emetine chase. The half-time of the decline of GFP-ΔF508 fluorescence in the presence of MG132 (25 μm) was 6.8 h, indicating that proteasome inhibition results in substantial stabilization of GFP-ΔF508. To determine whether the continued decline in GFP-ΔF508 fluorescence in the presence of MG132 was due to degradation by the proteasome, we examined the effect of increased inhibitor concentration on the fluorescence remaining at the 4 h chase point (Fig. 4A). These data suggest that GFP-ΔF508 degradation is inhibited by MG132 biphasically with a high-affinity component (IC50 ≤ 10 μm) and a low-affinity component that we were not able to saturate, even at concentrations as high as 300 μm. To further investigate the role of proteasomes in GFP-ΔF508 degradation, we performed a titration of epoxomicin, a highly specific and potent proteasome inhibitor (29.Meng L. Mohan R. Kwok B.H. Elofsson M. Sin N. Crews C.M. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 10403-10408Crossref PubMed Scopus (827) Google Scholar). Like MG132, epoxomicin inhibited GFP-ΔF508 fluorescence decay biphasically with a high-affinity component (IC50 < 1 μm) and a low-affinity component (Fig. 4B). Although epoxomicin can inhibit all three of the characterized activities of the proteasomes, it is 66- and 500- fold more effective at inhibiting the chymotrypsin-like (CT-L) activity than the trypsin-like or peptidyl-glutamyl hydrolyzing activities, respectively (30.Myung J. Kim K.B. Lindsten K. Dantuma N.P. Crews C.M. Mol. Cell. 2001; 7: 411-420Abstract Full Text Full Text PDF PubMed Scopus (106) Google Scholar). To investigate whether the low-affinity component of the MG132 and epoxomicin titrations was due to non-CT-L activities, we performed a titration of epoxomicin in the presence of 500 μm YU102, an α′,β′-epoxyketone inhibitor that is selective for the peptidyl-glutamyl hydrolyzing activity of the proteasome (30.Myung J. Kim K.B. Lindsten K. Dantuma N.P. Crews C.M. Mol. Cell. 2001; 7: 411-420Abstract Full Text Full Text PDF PubMed Scopus (106) Google Scholar). This treatment resulted in a small (10–15%) but significant increase in the degree of inhibition of GFP-ΔF508 fluorescence decay at all epoxomicin concentrations (Fig. 4C). These results suggest that the CT-L and peptidyl-glutamyl hydrolyzing activities of the proteasome may contribute independently to the degradation of GFP-ΔF508. A hallmark of proteolysis by the Ub-proteasome system is its dependence on ATP hydrolysis. ATP is required for at least two steps in Ub-dependent degradation: activation of Ub, and substrate unfolding (31.Hershko A. Ciechanover A. Annu. Rev. Biochem. 1998; 67: 425-479Crossref PubMed Scopus (6959) Google Scholar, 32.Ciechanover A. EMBO J. 1998; 17: 7151-7160Crossref PubMed Scopus (1200) Google Scholar). However, previous studies suggested that the fraction of CFTR degradation that is insensitive to proteasome inhibitors is independent of ATP (9.Jensen T.J. Loo M.A. Pind S. Williams D.B. Goldberg A.L. Riordan J.R. Cell. 1995; 83: 129-135Abstract Full Text PDF PubMed Scopus (775) Google Scholar). To assess the dependence of GFP-ΔF508 degradation on ATP, we assayed GFP-ΔF508 fluorescence in ATP-depleted cl.2 cells (Fig. 5A). These data demonstrate that GFP-ΔF508 degradation is absolutely dependent on the presence of cellular ATP. The degradation of GFPu was similarly dependent upon ATP (Fig. 5B). ER-associated degradation is a central element of the quality control pathways that ensure that only correctly folded and assembled proteins are deployed in the secretory pathway of eukaryotic cells. Although many studies have confirmed a primary role of the Ub-proteasome system in ERAD, the apparently minimal effect of highly potent proteasome inhibitors on the degradation of CFTR has remained an enigma and has suggested that other ATP-independent proteolytic systems also contribute to ERAD. In this study, we have used a fluorescent substrate to re-evaluate the role of the UPS in ERAD. Our data establish the primacy of the UPS in the degradation of ER-associated misfolded CFTR and suggest that the different proteolytic functionalities of the proteasome, associated with the three active β-subunits, may operate independently in the degradation of an ERAD substrate (GFP-ΔF508), but apparently not in the degradation of a cytoplasmic substrate (GFPu). Three lines of evidence support our conclusion that the UPS is the principal pathway for degradation of misfolded CFTR. First, proteasome inhibitors, but not inhibitors of serine or cysteine proteases or inhibitors of lysosomal or autophagic pathways, strongly stabilize GFP-ΔF508 fluorescence. Second, GFP-ΔF508 degradation was inhibited by MG132 and epoxomicin in a dose-dependent fashion, with biphasic apparent IC50 values in the expected range for CT-L and non-CT-L activities (30.Myung J. Kim K.B. Lindsten K. Dantuma N.P. Crews C.M. Mol. Cell. 2001; 7: 411-420Abstract Full Text Full Text PDF PubMed Scopus (106) Google Scholar). Third, GFP-ΔF508 degradation was strictly dependent on ATP, a hallmark of UPS-mediated proteolysis (31.Hershko A. Ciechanover A. Annu. Rev. Biochem. 1998; 67: 425-479Crossref PubMed Scopus (6959) Google Scholar, 32.Ciechanover A. EMBO J. 1998; 17: 7151-7160Crossref PubMed Scopus (1200) Google Scholar). The possibility that other proteases initiate the clipping of the CFTR molecule while it is situated in the ER membrane, followed by proteasome-mediated proteolysis, is also unlikely because simultaneous addition of protease and proteasome inhibitors did not protect CFTR from degradation. Although our results do not rule out participation in CFTR degradation of protease classes for which cell-permeable inhibitors are not available (i.e. aspartic proteases, aminopeptidases, and so forth) or of substrate-specific proteases (e.g. colligin is a specific inhibitor of procollagen degradation (33.Jain N. Brickenden A. Ball E.H. Sanwal B.D. Arch. Biochem. Biophys. 1994; 314: 23-30Crossref PubMed Scopus (19) Google Scholar)), these data establish that inhibition of proteolytic activities of the proteasome alone is sufficient to prevent the degradation of the majority of ΔF508 molecules in living cells. In previous studies using metabolic pulse-chase labeling and immunoprecipitation, proteasome inhibitors including MG132 and lactacystin had a very modest effect at inhibiting the decay of the electrophoretic species corresponding to core-glycosylated ΔF508 degradation (9.Jensen T.J. Loo M.A. Pind S. Williams D.B. Goldberg A.L. Riordan J.R. Cell. 1995; 83: 129-135Abstract Full Text PDF PubMed Scopus (775) Google Scholar, 13.Ward C.L. Omura S. Kopito R.R. Cell. 1995; 83: 121-127Abstract Full Text PDF PubMed Scopus (1133) Google Scholar). In the present work, we show that at least part of radiolabel lost from core-glycosylated GFP-ΔF508 could be recovered as high molecular weight material. Because this HMW material remains competent for degradation, our data suggest that in the absence of proteasome function, a significant fraction of radiolabeled GFP-ΔF508 molecules are converted to multiubiquitylated forms. Accumulation of HMW forms in the presence of proteasome inhibition is not observed for most other ERAD substrates such as the α-subunit of the T-cell receptor (7.Yu H. Kopito R.R. J. Biol. Chem. 1999; 274: 36852-36858Abstract Full Text Full Text PDF PubMed Scopus (103) Google Scholar, 34.Yu H. Kaung G. Kobayashi S. Kopito R.R. J. Biol. Chem. 1997; 272: 20800-20804Abstract Full Text Full Text PDF PubMed Scopus (203) Google Scholar) and soluble proteins such as IgG light chains (8.O'Hare T. Wiens G.D. Whitcomb E.A. Enns C.A. Rittenberg M.B. J. Immunol. 1999; 163: 11-14PubMed Google Scholar), which tend to accumulate as core-glycosylated detergent-soluble monomers in the ER after proteasome inhibition. It is likely that this difference is a consequence of the topology of CFTR, in which most of the mass of the protein is accessible to the cytoplasmic face of the ER and hence to components of the ubiquitin conjugation system. A recent study reported that inhibition of proteasome function leads to accumulation of a substantial fraction of CFTR in reconstituted cell-free extracts as high molecular weight multiubiquitylated protein (17.Xiong X. Chong E. Skach W.R. J. Biol. Chem. 1999; 274: 2616-2624Abstract Full Text Full Text PDF PubMed Scopus (97) Google Scholar). In our studies, no proteasome inhibitor or combination thereof was able to completely suppress the decay of GFP-ΔF508 fluorescence. It is unlikely that this apparent degradation (a ∼20–30% decrease of fluorescence after a 4-h chase in the presence of proteasome inhibitors) is the result of cell death due to toxic effects of simultaneous exposure to emetine and proteasome inhibitor because >98% of cells were able to exclude propidium iodide staining at the end of a 4-h chase. Alternatively, the decrease in GFP-ΔF508 fluorescence could, in principle, be due to self-quenching of GFP fluorescence in closely packed aggregates of GFP-ΔF508, which are known to form upon chronic exposure to proteasome inhibitors (23.Johnston J.A. Ward C.L. Kopito R.R. J. Cell Biol. 1998; 143: 1883-1898Crossref PubMed Scopus (1786) Google Scholar). Such self-quenching is a form of fluorescence resonance energy transfer (fluorescence resonance energy transfer homotransfer). Because we do not observe heterotransfer between the compatible fluorescence resonance energy transfer heterotransfer pair CFP-ΔF508 and YFP-ΔF508, even under conditions designed to maximize aggregation, 2R. Rajan and R. R. Kopito, unpublished data. it is unlikely that homotransfer between adjacent GFP-ΔF508 molecules could account for the observed decrease in fluorescence. It is also formally possible that the small decrease in GFP-ΔF508 fluorescence could be due to partial unfolding of the GFP moiety by the ATPase activity of the 19 S proteasome cap, without actual degradation of CFTR itself. However, the simplest explanation may be that none of the proteasome inhibitors used, alone or in combination, are able to fully inhibit all of the three proteasomal β-subunits. This explanation is consistent with conversion of a similar fraction (∼30%) of labeled GFP-ΔF508 as high molecular weight multiubiquitin conjugates in the presence of MG132 in both our pulse-labeling experiments (Fig. 1) and in the cell-free experiments of Oberdorf et al. (18.Oberdorf J. Carlson E.J. Skach W.R. Biochemistry. 2001; 40: 13397-13405Crossref PubMed Scopus (33) Google Scholar). It is likely, therefore, that the proteolytic activities of the proteasome associated with individual β-subunits do not act in a strictly cooperative fashion in ERAD. Our findings with GFP-ΔF508 in vivo are consistent with those of Oberdorf et al. (18.Oberdorf J. Carlson E.J. Skach W.R. Biochemistry. 2001; 40: 13397-13405Crossref PubMed Scopus (33) Google Scholar), who found that complete inhibition of the β5-subunit (associated with CT-L activity) led to only a 40% reduction in cell-free CFTR degradation, and with those of Myung et al. (30.Myung J. Kim K.B. Lindsten K. Dantuma N.P. Crews C.M. Mol. Cell. 2001; 7: 411-420Abstract Full Text Full Text PDF PubMed Scopus (106) Google Scholar), who reported recently that selective peptidyl-glutamyl hydrolyzing inhibition is insufficient to inhibit protein degradation in vivo. These data all strongly indicate that the catalytic sites of the proteasome function independently. It is not clear why, in our studies, the degradation of GFPu, a small soluble substrate of the UPS, is completely inhibited by MG132, whereas degradation of GFP-ΔF508 is not. Perhaps this discrepancy is a reflection of functional differences in the mechanisms by which proteasomes engage substrates delivered from the cytosol and those that are dislocated across the ER membrane. We thank Neil Bence for providing HEK293 cells expressing GFPu and Karim Helmy for help with epoxomicin and ATP depletion experiments. We are grateful to Dr. Craig Crews for a kind gift of YU102. We also thank members of the Kopito laboratory for helpful comments.
Referência(s)