Minimal length requirement for proteasomal ... - Karolinska Institutet

Download PDF
19 downloads 0 Views 523KB Size Report
Comment
Aug 21, 2008 - Heessen, S., Masucci, M. G., and Dantuma, N. P. (2005) The UBA2 domain functions as an intrinsic stabilization signal that protects.
The FASEB Journal • Research Communication

Minimal length requirement for proteasomal degradation of ubiquitin-dependent substrates Lisette G. G. C. Verhoef,*,1 Christian Heinen,* Alexandra Selivanova,† Els F. Halff,* Florian A. Salomons,* and Nico P. Dantuma*,2 *Department of Cell and Molecular Biology, The Medical Nobel Institute, and †Department of Neurobiology, Health Care Sciences and Society, Karolinska Institutet, Stockholm, Sweden An erroneous transcriptional process, known as molecular misreading, gives rise to an alternative transcript of the ubiquitin B (UBB) gene. This transcript encodes the protein UBBⴙ1, which comprises a ubiquitin moiety and a 19-aa C-terminal extension. UBBⴙ1 is found in affected neurons in neurodegenerative diseases and behaves as an atypical ubiquitin fusion degradation (UFD) proteasome substrate that is poorly degraded and impedes the ubiquitin/proteasome system. Here, we show that the limited length of UBBⴙ1 is responsible for its inefficient degradation and inhibitory activity. Designed UFD substrates with an equally short 19-aa or a 20-aa C-terminal extension were also poorly degraded and had a general inhibitory activity on the ubiquitin/proteasome system in two unrelated cell lines. Extending the polypeptide to 25 aa sufficed to convert the protein into an efficiently degraded proteasome substrate that lacked inhibitory activity. A similar length dependency was found for degradation of two UFD substrates in Saccharomyces cerevisiae, which suggests that the mechanisms underlying this length constraint are highly conserved. Extending UBBⴙ1 also converted this protein into an efficient substrate of the proteasome. These observations provide an explanation for the accumulation of UBBⴙ1 in neurodegenerative disorders and offers new insights into the physical constraints determining proteasomal degradation.— Verhoef, L. G. G. C., Heinen, C., Selivanova, A., Halff, E. F., Salomons, F. A., Dantuma, N. P. Minimal length requirement for proteasomal degradation of ubiquitindependent substrates. FASEB J. 23, 123–133 (2009) ABSTRACT

Key Words: ubiquitin fusion degradation 䡠 UBB⫹1 䡠 neurodegeneration 䡠 conformational diseases 䡠 Alzheimer 䡠 protein degradation Although most polyubiquitylated proteins are rapidly degraded by the proteasome, some can resist proteasomal degradation (1). For example, the glycine-alanine repeat of the Epstein-Barr virus, which is responsible for infectious mononucleosis (2), and the expanded polyglutamine repeats found in some neurodegenerative diseases (3, 4) hinder proteasomal degradation. Moreover, the ubiquitin binding domains of polyubiquitin shuttling factor Rad23 (5) and the transcription factor Met4 (6) render these proteins resistant to proteasomal 0892-6638/09/0023-0123 © FASEB

degradation. Another protein that is poorly degraded despite the presence of a potentially efficient degradation signal is UBB⫹1, an aberrant product of the ubiquitin precursor gene UBB (7). As a consequence of a transcriptional erroneous process, known as molecular misreading, UBB transcripts with a dinucleotide deletion are generated at very low frequency (8). The open reading frame of these aberrant transcripts encodes the first ubiquitin monomer with the C-terminal glycine substituted for a tyrosine, followed by a 19-aa extension encoded by the frame-shifted open reading frame. Levels of UBB⫹1 are very low under normal conditions, but the protein tends to accumulate specifically in affected cells in several neurodegenerative diseases, including Alzheimer’s disease and Huntington’s disease (9, 10), as well as in several non-neuronal pathologies characterized by the presence of misfolded proteins such as liver pathologies (11, 12) and sporadic inclusion-body myositis muscle fibers (13). A recent study suggests that UBB⫹1 may be important for ␤-amyloid-induced neurotoxicity (14). UBB⫹1 structurally resembles a ubiquitin fusion degradation (UFD) substrate (7). This class of substrates is characterized by the presence of an N-terminal ubiquitin moiety that cannot be removed by general deubiquitylation enzymes. Since the lysine residues at positions 29 and 48 of the ubiquitin moiety function as acceptors for polyubiquitin chains, these fusions are normally rapidly degraded by the 26S proteasome (15). While UFD substrates have been generally engineered to target proteins of interest for proteasomal degradation (15–18), UBB⫹1 is to our knowledge the only identified naturally occurring UFD substrate at this time. It is noteworthy that proteasome substrates, which are subject to N-terminal ubiquitylation, form an intermediate that structurally resembles UFD substrates (19). Yet, unlike other UFD or N-terminal ubiquitylated proteasome substrates, UBB⫹1 tends to accumulate to 1 Current address: IFOM, The FIRC Institute for Molecular Oncology, Via Adamello 16, 20139 Milan, Italy. 2 Correspondence: Department of Cell and Molecular Biology, The Medical Nobel Institute, Karolinska Institutet, Von Eulers va¨g 3, S-17177 Stockholm, Sweden. E-mail: nico. [email protected] doi: 10.1096/fj.08-115055

123

high levels under stress conditions when the capacity of the ubiquitin/proteasome system is challenged (20). In line with this notion, at elevated levels UBB⫹1 is a stable protein that causes general impairment of the ubiquitin/proteasome system, resulting in cell cycle arrest and apoptosis (7, 21). It has been suggested that the presence of UBB⫹1 in human tissue may be a surrogate marker for impairment of the ubiquitin/proteasome system (8). Thus, cell populations that contain UBB⫹1 may do so as a consequence of ubiquitin/proteasome system dysfunction encountered in the course of the pathology (22). At the same time, it is clear that elevated UBB⫹1 levels inhibit the ubiquitin/proteasome system (7, 23), suggesting that a positive feedback loop may occur in which the already compromised ubiquitin/proteasome system is further inhibited due to the accumulated UBB⫹1. Although the molecular mechanism for UBB⫹1-mediated proteasome dysfunction remains unclear, an in vitro study suggests that ubiquitylated recombinant UBB⫹1 may resemble unanchored polyubiquitin chains (24). Indeed, in cell lines the inhibitory activity of UBB⫹1 depends on the presence of the lysine residues at positions 29 and 48, which are the two lysine residues in ubiquitin that can form polyubiquitin chains targeting for proteasomal degradation (7). In the present study, we investigated the cause for the unusual stability of UBB⫹1 and show that the C-terminal extension of UBB⫹1 is not sufficiently long to be adequately handled by the proteasome. These data give new insights into the molecular constraints for efficient proteasomal degradation and provide an attractive explanation for the unusual stability of UBB⫹1.

MATERIALS AND METHODS Construction of plasmids All open reading frames were expressed from a CMV promoter in the mammalian expression vectors pCMV-myc (Clontech, Palo Alto, CA, USA) or EGFP-N1 (Clontech). The ubiquitin open reading frame and C-terminal extensions were PCRamplified from UbG76V-GFP and cloned in frame in pCMV-myc. Nonfluorescent green fluorescent protein (nfGFP) was constructed by introducing the amino acid substitution Y67R in the chromophore of GFP (7). Yeast expression plasmids were generated by cloning the mycUb-20aa, mycUb-25aa, and myc UBB⫹1(NheI) open reading frames in the episomal yeast expressing plasmid pYES2 (Invitrogen, Carlsbad, CA, USA). The mycUb-(V5)15aa, mycUb-(V5)20aa, and mycUb–(V5)25aa fusions, which are based on the V5 epitope tag as well as the myc UBB⫹1-6aa, were generated by recombination cloning in Saccharomyces cerevisiae. The amino acid sequences of the extensions are as follows: myc Ub-19aa: mycUbG76V-VGKLGRQDPPVATMVSKGE myc Ub-20aa: mycUbG76V-VGKLGRQDPPVATMVSKGEE myc Ub-25aa: mycUbG76V-VGKLGRQDPPVATMVSKGEELFTGV myc Ub-30aa: mycUbG76V-VGKLGRQDPPVATMVSKGEELFTGVVPILV myc Ub-(V5)15aa: mycUbG76V-RPLESRGPFEGKPIP myc Ub-(V5)20aa: mycUbG76V-RPLESRGPFEGKPIPNPLLG 124

Vol. 23

January 2009

myc

Ub-(V5)25aa:

STR

myc

UbG76V-RPLESRGPFEGKPIPNPLLGLD-

UBB⫹1-6aa: mycUBB⫹1-ELFTGV

myc

Transfections and tissue culture The human cervical carcinoma HeLa cell lines and neuroblastoma SH-SY5Y cell lines were cultured in Dulbecco’s modified Eagle medium (Life Technologies, Inc., Gaithersburg, MD, USA) supplemented with 10% fetal calf serum (Life Technologies, Inc.), 10 U/ml penicillin, and 10 ␮g/ml streptomycin. Medium for SH-SY5Y cells was also supplemented with 5 ml nonessential amino acids (Life Technologies, Inc.) and 2 mM l-glutamine (Life Technologies, Inc.). Cells were transiently transfected with polyethylenimine (Polysciences, Warrington, PA, USA) or lipofectamine (Life Technologies) and analyzed at indicated time points. Where indicated, cells were incubated for 17 h with medium containing 1 ␮M of the highly specific proteasome inhibitor Z-Leu-Leu-Leu-B(OH)2 (MG262; Biomol, Plymouth Meeting, PA, USA). Western blot analysis Total lysates were separated by SDS-PAGE and transferred to polyvinylide difluoride (PVDF) filters (Millipore, Bedford, MA, USA). The filters were blocked in phosphate-buffered saline supplemented with 5% skim milk and 0.1% Tween-20 and incubated with a mouse monoclonal antibody specific to c-myc (9E10; Santa Cruz Biotechnology, Santa Cruz, CA, USA) in a 1:1000 dilution for 1 h, a polyclonal antibody against UBB⫹1 in a 1:1000 dilution for 2 h (Ubi-3 050897; gift from Dr. Fred van Leeuwen, Maastricht University, Maastricht, The Netherlands) (10), or a mixed monoclonal antibody against GFP (Roche, Mannheim, Germany) in a 1:2000 dilution for 1 h. For loading controls, blots were probed with a mouse monoclonal GAPDH (RDI Research Diagnostics, Concord, MA, USA) or ␤-actin antibody (Abcam, Cambridge, MA, USA). After subsequent washing steps and incubation with peroxidase-conjugated goat anti-mouse serum or a peroxidase-conjugated goat anti-rabbit serum (Amersham Biosciences, Piscataway, NJ, USA), the blots were developed by enhanced chemiluminescence (ECL; Amersham Pharmacia Biotech). Western blots were quantified with ImageJ software (National Institutes of Health, Bethesda, MD, USA). Fluorescence microscopy and flow cytometry For fluorescence microscopy, cells were fixed with 4% paraformaldehyde in phosphate-buffered saline solution. Cells were stained with a mouse monoclonal antibody specific to c-myc (9E10) in a 1:250 dilution followed by AlexaFluor 546 labeled anti-mouse immunoglobulins (Molecular Probes, Eugene, OR, USA) in a 1:1000 dilution. The cells were examined with a LSM510 META confocal laser scanning microscope (Zeiss, New York, NY). For analysis by flow cytometry, cells were harvested and fixed with cytofix/cytoperm (BD Biosciences, San Jose, CA, USA), followed by washing with perm/wash (BD Biosciences). The cells were either stained with the anti-c-myc mouse monoclonal antibody followed by APC-labeled goat anti-mouse immunoglobulins (BD Biosciences Pharmingen), or cotransfected with an expression plasmid of human CD8␣ (gift from Dr. J. Neefjes, Netherlands Cancer Institute, Amsterdam, The Netherlands), and stained with an APC-labeled anti-human CD8␣ antibody (BD Pharmingen). Flow cytometry was performed with a FACSort flow cytometer and analyzed with CellQuest software (Becton & Dickinson, Franklin Lakes, NJ, USA).

The FASEB Journal

VERHOEF ET AL.

Protein turnover analysis Transiently transfected HeLa cells were incubated in methionine/cysteine-free RPMI medium (Sigma, St. Louis, MO, USA) and then metabolically labeled with 40 ␮Ci [35S] methionine/cysteine (Redivue PRO-MIX 35S, Amersham Biosciences) for 30 min at 37°C. After the labeling period, cells were washed and incubated with RPMI medium supplemented with 10 mM methionine and 1 mM cysteine. Cells were harvested at the indicated time points in lysis buffer containing 150 mM NaCl, 1% Nonidet P-40, 0.5% deoxycholic acid, 0.1% SDS, 50 mM Tris-HCl pH 8, 20 mM N-ethylmaleimide (Sigma), and complete protease inhibitor cocktail (Roche) and were lysed for 1 h at 4°C. Lysates were centrifuged at 14,000 rpm for 10 min at 4°C, and the supernatant was used for immunoprecipitation. Myc-tagged proteins were immunoprecipitated for 1 h at 4°C with a mouse monoclonal antibody specific to c-myc (9E10, Santa Cruz Biotechnology) in a 1:50 dilution, followed by incubation with protein-G Sepharose beads (Amersham Bioscience) for 2 h at 4°C. Quantification of the pulse-chase experiments was performed with ImageJ software. Yeast experiments All experiments were performed with a haploid derivative of S. cerevisiae strain DF5 (lys2– 801, leu2–3, -112, ura3–52, his3⌬200, trp1-1, Mat-a). Yeast transformed with episomal pYES2 plasmids encoding the indicated UFD substrates was grown successively overnight in synthetic minimal medium with glucose, raffinose, or galactose as sole carbohydrate source. Protein turnover was determined by administration of 0.5 mg/ml cycloheximide to the yeast cultures in midlog phase and harvesting samples at the indicated times for Western blot analysis. Total protein extracts were obtained by precipitation in 12.5% trichloroacetic acid. For proteasome inhibition experiments, yeast was grown for 1 h in the presence of 50 ␮M proteasome inhibitor Z-Leu-Leu-Leu-CHO (MG132; Biomol). To allow uptake of the inhibitor, yeast was grown in synthetic medium with 0.1% proline as sole nitrogen source and proteasome inhibitor was added together with 0.003% SDS when the culture had reached an OD600 of 0.5 (25).

RESULTS The ⴙ1 extension does not function as a modular stabilization signal Since certain domains can counteract degradation signals and protect proteins from proteasomal degradation (2, 4 – 6), we investigated whether the 19-aa Cterminal extension of UBB⫹1 had intrinsic stabilizing properties that could account for the stable nature of UBB⫹1. For clarity we will refer hereafter to the Cterminal 19 aa as the “⫹1 extension” (Fig. 1A). Since the C-terminal part is translated from a frame-shifted open reading frame and likely to lack any higher-order structure, we considered it feasible that the ⫹1 extension rendered UBB⫹1 aggregation prone, which is a robust way to block proteasomal degradation (4) and impair the ubiquitin/proteasome system (26). To probe into this possibility, we fused the ⫹1 extension to GFP-based proteasome substrates carrying well-defined degradation signals (Fig. 1B). Insertion of MINIMAL LENGTH OF PROTEASOME SUBSTRATES

the ⫹1 extension at the C termini of either the UbG76VGFP or Ub-R-GFP proteasome substrates did not increase their steady-state levels nor did it inhibit their accumulation in response to the proteasome inhibitor MG262 as determined by flow cytometry (Fig. 1C, D) and Western blot analysis (data not shown). We conclude that the ⫹1 extension does not have intrinsic stabilizing properties. Stability of UBBⴙ1 is attributed to the length and not the primary amino acid sequence of the ⴙ1 extension Several full-sized proteins such as ␤-galactosidase (15), luciferase (17), and fluorescent proteins (16) have been successfully fused to uncleavable ubiquitin moieties, which resulted in short-lived UFD substrates. A striking difference between UBB⫹1 and these shortlived UFD substrates is the limited length of the Cterminal extension of UBB⫹1. We wondered, therefore, whether this feature could be the reason for the poor degradation of UBB⫹1. To address this issue, we truncated the short-lived UbG76V-GFP reporter substrate to obtain a UFD substrate that had an extension of the same length as the ⫹1 extension but an otherwise unrelated sequence (Fig. 2A). An N-terminal myc epitope tag was introduced in order to distinguish this ubiquitin fusion from the general pool of free or conjugated ubiquitin in the cells. We noted that the expression vector used for this purpose gave rise to higher expression levels than the construct used in our previous study (7), which was accompanied by increased steady-state levels and enhanced stability of UBB⫹1 (Supplemental Fig. 1). Since it has been previously shown that expression levels are important, as cells can turn over low levels of UBB⫹1 but fail to degrade slightly elevated levels of UBB⫹1 (27), we introduced all fusions in the same pCMV-Myc expression vector. HeLa cells were transiently transfected with mycUBB⫹1, myc Ub-19aa, and mycUbG76V- nfGFP (a nonfluorescent GFP was used in this context to allow analysis of its trans inhibitory effect in cell lines expressing GFP-based reporter substrates; see below). Flow cytometric analysis of transiently transfected HeLa cells showed a clear increase in the levels of the short-lived mycUbG76V-nfGFP protein after treatment with proteasome inhibitor, which confirmed that this fusion is turned over by proteasomal degradation (Fig. 2B). In contrast mycUBB⫹1 and mycUb19aa were relatively stable and did not accumulate on proteasome inhibitor administration, suggesting that they were not degraded by the proteasome (Fig. 2B). Thus, a UFD substrate with a 19-aa extension that was unrelated to the original extension derived from the ⫹1 frameshift was stable, which suggests that the unusual stability of UBB⫹1 is due to the length of the extension and not to its primary amino acid sequence. (For clarity, we will refer to both short-lived and long-lived ubiquitin fusions as “UFD substrates,” even though the latter are not degraded by the proteasome, and hence cannot be considered as true proteasome substrates.) 125

Figure 1. The ⫹1 extension does not function as a modular stabilization signal. A) Schematic representation of normal transcription and translation of the ubiquitin B gene resulting in 3 ubiquitin monomers, and molecular misreading giving rise to the aberrant ubiquitin UBB⫹1. The ubiquitin B gene encodes 3 head-to-tail ubiquitin molecules. The open reading frame is translated into a precursor protein that can be cleaved by deubiquitylation enzymes (DUBs) into 3 single ubiquitin molecules. Due to erroneous transcription of the GAGAGGT sequence, UBB transcripts with a dinucleotide deletion (GU) are generated at low frequency. The UBB⫹1 open reading frame encodes a ubiquitin molecule with an additional 19 aa at the C terminus. Due to substitution of the glycine at position 76 to a tyrosine, the 19 aa cannot be cleaved off by general DUBs. B) Schematic representation of the UbG76V-GFP reporter and the Ub-R-GFP constructs with or without the ⫹1 extension of UBB⫹1. C) Flow cytometric analysis of HeLa cells transiently transfected with UbG76V-GFP or UbG76V-GFP⫹1. One day post-transfection, the cells were treated with 1 ␮M proteasome inhibitor MG262 for 17 h or left untreated. Percentages of GFP-positive cells are indicated in the top right quadrant. One representative experiment of 3 is shown. D) Quantification of the experiment as shown in C. Ratios of the percentages of fluorescent cells in untreated cells vs. cells treated with proteasome inhibitor are plotted. Ratio of 1 indicates a stable protein; lower values indicate proteasomal degradation of the protein. Values are means ⫾ sd from 3 independent experiments. No significant differences were found between the GFP proteasome substrates without and with the ⫹1 extension (unpaired t test).

We hypothesized that a 19-aa extension following a ubiquitin domain is too short to support efficient degradation rendering UBB⫹1 and mycUb-19aa as stable proteins. To test this hypothesis, we made a set of UFD substrates in which we extended the C terminus stepwise by adding 5 aa at a time (Fig. 2A). mycUBB⫹1 and the UFD substrates with extensions of 19 and 20 aa were not efficiently degraded by the proteasome (Fig. 2C). In contrast, UFD substrates with a 25- or 30-aa extension accumulated in the presence of proteasome inhibitor, suggesting that these ubiquitin fusions are subject to proteasomal degradation (Fig. 2C). Pulse-chase analysis with metabolically labeled cells confirmed that myc UBB⫹1, mycUb-19aa, and mycUb-20aa were stable during the 4 h chase period (Fig. 2D), whereas mycUb25aa was rapidly degraded, with a half-life of ⬃25 min 126

Vol. 23

January 2009

(Fig. 2E). Further extending the C-terminal part had only a modest effect on degradation rates. Half-lives of ⬃25 and 20 min were measured for mycUb-30aa and myc UbG76V-nfGFP, respectively (Fig. 2E). These results point to a threshold at around 25 aa: substrates with shorter extension are stable, whereas substrates with a longer extension are rapidly degraded by the proteasome. In order to study whether the length-dependent stability of ubiquitin fusions is a general phenomenon applying for different types of cells, we analyzed the stability of these UFD substrates in the human neuroblastoma cell line SH-SY5Y. The neuroblastoma cell lines were transiently transfected with plasmids encoding mycUBB⫹1, mycUb-20aa, and mycUb-25aa. We found that also in this cell line the mycUb-25aa was present at

The FASEB Journal

VERHOEF ET AL.

Figure 2. Minimal C-terminal extension for efficient degradation of ubiquitin fusions is ⬃25 aa. A) Schematic representation of the mycUbG76V-nfGFP, mycUb-19aa, mycUb-20aa, mycUb-25aa, and mycUb-30aa constructs. B) Quantification of flow cytometric analysis of myc-stained HeLa cells transiently transfected with mycUBB⫹1, mycUb-19aa, or mycUbG76V-nfGFP. Ratios of the percentages of fluorescent cells in untreated cells vs. cells treated with proteasome inhibitor are plotted. Ratio of 1 indicates a stable protein; lower values indicate proteasomal degradation of the protein. Values are means ⫾ sd from 3 independent experiments. C) Western blot analysis with a myc antibody of cell lysates from HeLa cells transiently transfected with mycUBB⫹1, myc Ub-19aa, mycUb-20aa, mycUb-25aa, or mycUb-30aa. One day post-transfection, cells were treated with 1 ␮M of the proteasome inhibitor MG262 for 17 h. GAPDH is shown as loading control. D) Half-lives were determined by pulse-chase analysis in HeLa cells transiently transfected with the mycUBB⫹1, mycUb-19aa, and mycUb-20aa constructs. Bottom panel shows densitometry quantification of the pulse-chase analyses. E) Half-lives were determined by pulse-chase analysis in HeLa cells transiently transfected with mycUb-25aa, mycUb-30aa, and mycUbG76V-nfGFP. Bottom panel shows densitometry quantification of the pulse-chase analyses. F) UbG76V-GFP SH-SY5Y cells were transiently transfected with mycUBB⫹1, mycUb-20aa, and a mycUb-25aaexpressing constructs or mock transfected. Two days post-transfection, cells were treated with 1 ␮M of the proteasome inhibitor MG262 for 17 h. Cell lysates were analyzed in Western blotting with an ␣-myc antibody. mycUBB⫹1, mycUb-20aa, mycUb-25aa, high-molecular-weight (HMW) products and the molecular weight marker are indicated. GAPDH is shown as loading control.

very low levels and accumulated readily in the presence of proteasome inhibitor, demonstrating that this UFD substrate is rapidly degraded in neuroblastoma cells (Fig. 2F). In contrast, steady-state levels of mycUBB⫹1 and mycUb-20aa were high and showed only a minor increase in the presence of proteasome inhibitor in line with poor proteasomal degradation of these substrates. Thus, the UFD substrates display a similar length dependency in human cervix carcinoma cells and neuroblastoma cells. Inefficient degradation of UFD substrates correlates with impairment of the ubiquitin/proteasome system The inhibitory effect of UBB⫹1 may arise from its unusual stability by engaging the proteasome or other components of the ubiquitin/proteasome system in a dead-end process of attempting to degrade a ubiquitylated protein that is inherently stable. This hypothesis predicts that any UFD substrate that is too short to be efficiently degraded should impair the ubiquitin/proMINIMAL LENGTH OF PROTEASOME SUBSTRATES

teasome system. To address this issue, we used a reporter cell line that stably expresses the UbG76V-GFP reporter substrate (16). UbG76V-GFP reporter cells were transfected with expression plasmids encoding the stable mycUb-20aa or the unstable mycUb-25aa fusions. In many cells the expression of mycUb-20aa was accompanied by elevated UbG76V-GFP levels, which suggests an impaired ubiquitin/proteasome system (Fig. 3A, top panel). The effect was similar to that previously found for ectopically expressed UBB⫹1 (ref. 7 and data not shown). In contrast, fewer myc-positive cells were observed when transfected with the mycUb-25aa-expressing plasmid, in line with the notion that this substrate is degraded by the proteasome. More important, we did not observe elevated UbG76V-GFP levels in cells that expressed mycUb-25aa, indicative of an operative ubiquitin/proteasome system (Fig. 3A, bottom panel). Also mycUb-30aa and mycUbG76V-nfGFP did not cause accumulation of reporter substrates (data not shown). To obtain more quantitative data on the inhibitory 127

Figure 3. Inefficient degradation of UFD substrates correlates with impairment of the ubiquitin/proteasome system. A) UbG76V-GFP HeLa cells were transiently transfected with a mycUb-20aa (top panel) or a mycUb-25aa (bottom panel) expressing construct. One day post-transfection, cells were stained with an antibody directed against the myc tag and analyzed for myc staining and native UbG76V-GFP fluorescence. Representative micrographs of the immunostaining (left panel), the UbG76V-GFP fluorescence (middle panel), and the merged pictures (right panel). B) Flow cytometric analysis of UbG76VGFP HeLa cells transiently cotransfected with mycUb-20aa or mycUb-25aa and the cell surface marker CD8␣. Cells were stained with an APC-conjugated antibody against CD8␣. Percentages of CD8␣-positive cells that were also positive for UbG76V-GFP fluorescence are indicated in the top right quadrant. One representative experiment of 3 is shown. C) Quantification of the experiment shown in B. Values are means ⫾ sd from a triplicate experiment. **P ⬍ 0.001 vs. myc UBB⫹1; unpaired t test. D) UbG76V-GFP SH-SY5Y neuroblastoma cells were transiently transfected with mycUBB⫹1 (top panel), mycUb-20aa (middle panel), and a mycUb-25aa (bottom panel) expression construct. One day post-transfection, cells were stained with an antibody directed against the myc tag and analyzed for myc staining and native UbG76V-GFP fluorescence. Representative micrographs of the immunostaining (left panel), the UbG76V-GFP fluorescence (middle panel), and the merged pictures (right panel). E) Quantification of the percentage of myc-positive cells with elevated levels of UbG76V-GFP of 3 independent experiments as shown in B. P values are mentioned (unpaired t test).

activity of these constructs, we turned to flow cytometry. We anticipated that many of the mycUb-25aa- but not myc Ub-20aa-expressing cells might have undetectable levels of the myc-tagged UFD substrate due to the rapid turnover of mycUb-25aa. To be able to detect transfected cells irrespective of the level of UFD substrate and make a fair comparison between mycUb-20aa- and myc Ub-25aa-expressing cells, we cotransfected cells with the cell surface marker CD8␣. Flow cytometry con128

Vol. 23

January 2009

firmed that in the case of mycUb-20aa a major fraction of the transfected cells (as identified by CD8␣ expression) had an impaired ubiquitin/proteasome system and accumulated the reporter (Fig. 3B, top panel). On the contrary, in cells transfected with mycUb-25aa/CD8␣, only a very small fraction of transfected cells had detectable levels of the reporter (Fig. 3B, bottom panel). Quantitative analysis of three independent experiments clearly showed that mycUBB⫹1 and mycUb-20aa had similar inhib-

The FASEB Journal

VERHOEF ET AL.

itory activities, whereas mycUb-25aa, mycUb-30aa and myc UbG76V-nfGFP did not block the ubiquitin/proteasome system (Fig. 3C). We tested the length-dependent inhibitory effect of ubiquitin fusions also in human neuroblastoma cell line SH-SY5Y stably expressing the UFD reporter substrate UbG76V-GFP or the N-end rule reporter substrate Ub-RGFP (16). Microscopic analysis showed that mycUBB⫹1 and mycUb-20aa impaired the ubiquitin/proteasome system in a large fraction of the transfected neuroblastoma cells, evidenced by the accumulation of reporter substrates, whereas the ubiquitin/proteasome system in cells expressing mycUb-25aa remained active (Fig. 3D). Quantitative analysis of 3 independent experiments confirmed a significant increase in cells with impaired ubiquitin/proteasome system in the mycUBB⫹1- and myc Ub-20aa-expressing cells (Fig. 3E). Similar results were obtained for the neuroblastoma cell line expressing the N-end rule reporter substrate Ub-R-GFP, suggesting that the inhibitory effect is not limited to the UFD pathway (Supplemental Fig. 2). We conclude that accumulation of a UFD substrate that is too short to be rapidly degraded can impair the ubiquitin/proteasome system. The length dependency of proteasomal degradation of UFD substrates is conserved in budding yeast Although the original goal of our study was to understand the unusual stability of UBB⫹1, and its relation to the inhibitory effect of UBB⫹1 in human cells, we felt that the length-dependent degradation of UFD fusions revealed in this study could shed also light on the molecular mechanisms underlying proteasomal degradation. To investigate whether the molecular mechanisms responsible for the length constraints of UFD substrates are conserved in eukaryotes, we turned to the budding yeast S. cerevisiae. Notably, UBB⫹1 or similar natural ubiquitin fusions have not been described for yeast. The mycUb-20aa and mycUb-25aa substrates were expressed in yeast from an episomal plasmid under the control of the GAL1 promoter. Western blot analysis revealed a striking difference in the steady-state levels of the two UFD substrates (Fig. 4A). Similar to what was observed in mammalian cells, the steady-state levels of myc Ub-20aa were much higher than those of mycUb25aa. Treatment with the proteasome inhibitor MG132 caused an increase in the mycUb-25aa levels, confirming that the low steady-state levels of this fusion are caused by proteasomal degradation (Fig. 4A). Analysis of their biological half-lives confirmed that mycUb-25aa was rapidly turned over, whereas the mycUb-20aa was stable over the analyzed 40-min timeframe (Fig. 4B, C). We next investigated whether the length dependency of UFD substrates is a general phenomenon by using an unrelated extension. For this purpose, we used the commonly used V5 epitope tag, which was introduced at the C terminus of the ubiquitinG76V moiety. We generated fusions with extensions of 15, 20, and 25 aa of the V5 epitope tag (Fig. 4D). Again we found a MINIMAL LENGTH OF PROTEASOME SUBSTRATES

profound difference in the steady state, with the levels of mycUb-(V5)15aa far exceeding the mycUb-(V5)20aa and mycUb-(V5)25aa levels, which suggests that the critical threshold for this extension is around 15 to 20 aa, slightly shorter than the extension used in the earlier experiments (Fig. 4E and data not shown). Moreover, the mycUb-(V5)-20aa levels increased in response to proteasome inhibitor, consistent with the notion that the fusion is degraded by the proteasome (Fig. 4E). In line with the low steady-state levels of the myc Ub-(V5)25aa and mycUb-(V5)20aa fusions, we found that these fusions were rapidly degraded, whereas the myc Ub-(V5)15aa was stable (Fig. 4F, G and data not shown). Together, these data show that the length dependency for degradation of UFD substrates is a general and conserved phenomenon. At the same time, these data suggest that the primary amino acid sequence of the extension may modulate the length required to accommodate degradation. A short extension converts UBBⴙ1 into an efficient proteasome substrate Our data predict that extending the tail of UBB⫹1 with a few amino acids will allow efficient proteasomal degradation (Fig. 5A). Indeed, we found that expression of mycUBB⫹1-6aa resulted in a low steady-state level of the fusion that was increased on administration of proteasome inhibitor, whereas mycUBB⫹1 was stable (Fig. 5B). Turnover analysis showed that addition of 6 aa converted mycUBB⫹1 from a very stable protein into a short-lived protein that was degraded within 10 min (Fig. 5C). These data confirms that the cause for the unusual stability of UBB⫹1 lies in the limited length of the extension.

DISCUSSION In this study, we investigated the molecular mechanisms underlying the unusual stability of the natural occurring UFD substrate UBB⫹1, an aberrant ubiquitin fusion found in neurodegenerative disorders (9, 10), as well as a number of non-neuronal pathologies (11–13). Our analysis not only shed light on the reason for the stability of UBB⫹1 but revealed fundamental insights into how substrates are handled by the proteasome. Most important, we found that UFD substrates require a minimal length in order to be efficiently degraded by proteasomes. The mechanisms responsible for this length dependency appear to be conserved in eukaryotes, as the critical length for efficient proteasomal degradation was similar for budding yeast and human cells. Our data show that UFD substrates that do not fulfill this length requirement accumulate and have a general inhibitory effect on ubiquitin-dependent proteasomal degradation. Our results suggest that proteasomes require a short polypeptide stretch protruding from the UFD signal in order to process the substrate. Johnson et al. (28) 129

Figure 4. Length dependency of proteasomal degradation of UFD substrates is conserved in budding yeast. A) Western blot analysis of S. cerevisiae transformed with a control plasmid pYES2 or UFD substrate expression plasmids pYES2-mycUb-20aa and pYES2-mycUb-25aa. Expression was induced by growing the yeast with galactose as sole carbon source. Yeast was left untreated or incubated with the proteasome inhibitor MG132. Molecular weight markers, fusion constructs, and HMW adducts are indicated. ␤-Actin is shown as loading control. A short exposure (left panel) and long exposure (right panel) of the same blot are shown. B) Intracellular turnover of mycUb-20aa and mycUb-25aa in S. cerevisiae. Synthesis of mycUb-20aa and mycUb-25aa was shut off by administration of cycloheximide and degradation of the fusions was followed by Western blotting with an ␣-myc antibody. A short exposure (top panel) and long exposure (bottom panel) of the same blot are shown. Molecular weight markers, fusion constructs, and HMW adducts are indicated. C) Densitometry quantification of Western blots of 3 independent experiments. Short and long exposures were used for quantification of the stable mycUb-20aa and unstable mycUb-25aa, respectively. D) Amino acid sequences of the mycUb-(V5)15aa, (V5)20aa, and (V5)25aa constructs. E) Steady-state levels of myc Ub-(V5)15aa and mycUb-(V5)20aa in the absence or presence of proteasome inhibitor. Experiment was performed as in A. F) Intracellular turnover of mycUb-(V5)15aa and mycUb-(V5)20aa. mycUb-(V5)15aa, mycUb-(V5)20aa and HMW are indicated. Experiment was performed as in B. G) Densitometry quantification of Western blots of 3 independent experiments.

130

Vol. 23

January 2009

The FASEB Journal

VERHOEF ET AL.

Figure 5. A short extension converts UBB⫹1 into an efficient proteasome substrate. A) Amino acid sequence of ⫹1 tail of UBB⫹1 without and with the additional 6 aa. B) Steady-state levels of mycUBB⫹1 and myc UBB⫹1-6aa in the absence and presence of proteasome inhibitor. ␤-Actin is shown as loading control. Experiment was performed as in Fig. 4A. Short exposure (left panel) and long exposure (right panel) of the same blot are shown. C) Intracellular turnover of mycUBB⫹1 and myc UBB⫹1-6aa. Experiment was performed as in Fig. 4B. Molecular weight markers, fusion constructs and HMW adducts are indicated. Asterisks mark truncated mycUBB⫹1 product.

reported earlier that ubiquitin carrying a short 16-aa C-terminal tag was ubiquitylated, similar to authentic UFD substrates, but nevertheless stable in yeast. Recently, it was reported that ubiquitin is poorly degraded in vitro unless it is provided with a C-terminal extension consisting of a combined histidine and hemagglutinin tag of 25 aa (29). These observations are in line with our finding that two unrelated UFD substrates require a minimal length of 20 and 25 aa, respectively, to accommodate proteasomal degradation. The unfolding of substrates at the proteasome is accomplished by the AAA-ATPases residing in the base of the 19S regulatory particle (30) and is completed prior to substrate translocation and peptide hydrolysis (31, 32). It has been shown that a short unfolded structure is required for initiation of proteasomal degradation (33). Since it is not only the eukaryotic proteasome that requires an unfolded structure for degradation, but also remotely related bacterial compartmentalized proteases (34), this length limitation may reflect a fundamental physical constraint of such proteases independent of the targeting system that is used. An important question is whether the length dependency found for UFD substrates also applies to other ubiquitin-dependent proteasome substrates. Unfortunately, it is technically impossible at this point to directly address this question. The UFD substrates are unique in the sense that the first ubiquitin in the polyubiquitin chain is directly fused to the substrate (28). Because of this feature short substrates consisting of only 20 –30 aa can be expressed in the context of a full-sized UFD protein. If a similar approach would be used for other types of substrates, this would require expression of a substrate of no longer than 20 –30 aa. Peptides of this length are known to be inherently unstable and rapidly degraded by tripeptidyl peptidase II and amino peptidases (35) and hence cannot be MINIMAL LENGTH OF PROTEASOME SUBSTRATES

used to address this question. Notably, several conceptual insights into proteasomal degradation have been made with UFD substrates, such as the minimal ubiquitin chain required for degradation (36), suggesting that UFD substrates are not very different from more conventional substrates. Also, when it comes to the minimal length requirement found in this study, it seems unlikely that this would reflect a UFD-specific feature. This notion is strongly supported by a recent study on ubiquitin-independent proteasomal degradation of ornithine decarboxylase (ODC) fusions (37). Takeuchi et al. (37) reported that ODC fusions need a loosely unstructured region of 20 –30 aa, which is very similar to the minimal length for UFD substrates. It is noteworthy that these substrates apply different targeting mechanisms and that their degradation pathways merge only at the level of the proteasome (38). The fact that the length dependency is shared between proteasome substrates as diverse as UFD- and ODC-based substrate strongly argues that it reflects a general prerequisite for proteasomal degradation that also applies to other proteasome substrates irrespective of their targeting mechanism. A number of proteins associate with the proteasome through the presence of an N-terminal ubiquitin-like (UbL) domain (39). The UbL domain of Rad23 can be functionally replaced by an N-terminal uncleavable ubiquitin moiety (40). Vice versa, some UbL domains can target proteins for proteasomal degradation (5, 41). Thus, the functional difference between the uncleavable N-terminal ubiquitin moieties present in UFD substrates and UbL domains may be rather subtle, with other factors determining whether it functions as a proteasome binding motif or instead targets for proteasomal degradation. A number of UbL-containing proteins have been reported to be able to decelerate proteasomal degradation. Recently, Finley and co-workers (42) reported that 131

the deubiquitylation enzyme Ubp6 delays ubiquitin-dependent proteasomal degradation independent of its catalytic activity. Also, the UbL-containing protein Rad23 can impair the ubiquitin/proteasome system under in vitro conditions (43). Rad23 is safeguarded from proteasomal degradation through a ubiquitin associated domain (UBA) domain, which functions as an intrinsic stabilization signal (5). High levels of another UbL-UBA-containing protein, Dsk2, also lead to severe cellular defects and inhibition of proteolysis (44). We propose that the inhibitory potential observed in these studies may be a common characteristic of proteins that interact with the proteasome by a UbL or UFD signal but resist proteasomal degradation. The unusual stability of these fusions is unlikely to be due to poor proteasomal targeting since enhancing targeting of UBB⫹1 for proteasomal degradation by insertion of additional UFD signals (18) does not result in accelerated degradation, but on the contrary, further enhanced its inhibitory effect on ubiquitin-dependent proteolysis (7). Based on these observations, we postulated that UBB⫹1 is a difficult substrate that can frustrate the ubiquitin/proteasome system (7). Although our data show that proteasomal degradation of UBB⫹1 is problematic, it is not impossible, as cells manage to clear low levels of the protein by proteasomal degradation (7). However, elevated UBB⫹1 levels (7, 21, 27) or moderately reduced activity of the ubiquitin/proteasome system (20) triggers a rapid accumulation and stabilization of UBB⫹1, which in turn causes general dysfunction of the ubiquitin/proteasome system (7, 23, 24). Even though our data explain the unusually slow turnover of UBB⫹1 and show that the exceptional stability of UBB⫹1 is crucial to its inhibitory effect on the ubiquitin/proteasome system, the exact mechanism responsible for its inhibitory activity remains open. While many questions remain unanswered regarding the precise role of UBB⫹1 in neurodegeneration, this atypical UFD substrate has revealed new insights into the prerequisites for efficient proteasomal degradation. We demonstrated that the inhibitory activity attributed to this protein can be recapitulated by expression of an unrelated equally short UFD substrate. It is striking and ironic that, although N-terminal ubiquitin fusions have been routinely used in experimental settings as a robust way to target recombinant proteins for ubiquitin-dependent proteasomal degradation, the only presently identified natural occurring UFD substrate, i.e., UBB⫹1, is not only a poor proteasome substrate but its presence has a general inhibitory effect on the ubiquitin/proteasome system. We thank Drs. Michael Glickman and Dasha Krutauz for stimulating discussions and critical reading of the manuscript, Dr. J. Neefjes (Netherlands Cancer Institute, Amsterdam) for the CD8␣ expression plasmid, Dr. F. van Leeuwen (Maastricht University, Maastricht, The Netherlands) for the UBB⫹1 antibody, and the members of the Dantuma lab for helpful suggestions. The work in the Dantuma lab was supported by the Swedish Research Council, Swedish Cancer Society, the Wallenberg foundation, the Hereditary Disease 132

Vol. 23

January 2009

Foundation, the Nordic Center of Excellence Neurodegeneration, the Marie Curie Research Training Network (MRTNCT-2004-512585), and the Karolinska Institute. N.P.D. is supported by the Swedish Research Council. E.F.H. received financial support from the Prinses Beatrix Fonds.

REFERENCES 1. 2.

3.

4. 5. 6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

Dantuma, N. P., and Masucci, M. G. (2002) Stabilization signals: a novel regulatory mechanism in the ubiquitin/proteasome system. FEBS Lett. 529, 22–26 Levitskaya, J., Sharipo, A., Leonchiks, A., Ciechanover, A., and Masucci, M. G. (1997) Inhibition of ubiquitin/proteasomedependent protein degradation by the Gly-Ala repeat domain of the Epstein-Barr virus nuclear antigen 1. Proc. Natl. Acad. Sci. U. S. A. 94, 12616 –12621 Venkatraman, P., Wetzel, R., Tanaka, M., Nukina, N., and Goldberg, A. L. (2004) Eukaryotic proteasomes cannot digest polyglutamine sequences and release them during degradation of polyglutamine-containing proteins. Mol. Cell. 14, 95–104 Verhoef, L. G., Lindsten, K., Masucci, M. G., and Dantuma, N. P. (2002) Aggregate formation inhibits proteasomal degradation of polyglutamine proteins. Hum. Mol. Genet. 11, 2689 –2700 Heessen, S., Masucci, M. G., and Dantuma, N. P. (2005) The UBA2 domain functions as an intrinsic stabilization signal that protects Rad23 from proteasomal degradation. Mol. Cell 18, 225–235 Flick, K., Raasi, S., Zhang, H., Yen, J. L., and Kaiser, P. (2006) A ubiquitin-interacting motif protects polyubiquitinated Met4 from degradation by the 26S proteasome. Nat. Cell Biol. 8, 509 –515 Lindsten, K., de Vrij, F. M., Verhoef, L. G., Fischer, D. F., van Leeuwen, F. W., Hol, E. M., Masucci, M. G., and Dantuma, N. P. (2002) Mutant ubiquitin found in neurodegenerative disorders is a ubiquitin fusion degradation substrate that blocks proteasomal degradation. J. Cell Biol. 157, 417– 427 Fischer, D. F., De Vos, R. A., Van Dijk, R., De Vrij, F. M., Proper, E. A., Sonnemans, M. A., Verhage, M. C., Sluijs, J. A., Hobo, B., Zouambia, M., Steur, E. N., Kamphorst, W., Hol, E. M., and Van Leeuwen, F. W. (2003) Disease-specific accumulation of mutant ubiquitin as a marker for proteasomal dysfunction in the brain. FASEB J. 17, 2014 –2024 De Pril, R., Fischer, D. F., Maat-Schieman, M. L., Hobo, B., de Vos, R. A., Brunt, E. R., Hol, E. M., Roos, R. A., and van Leeuwen, F. W. (2004) Accumulation of aberrant ubiquitin induces aggregate formation and cell death in polyglutamine diseases. Hum. Mol. Genet. 13, 1803–1813 Van Leeuwen, F. W., de Kleijn, D. P., van den Hurk, H. H., Neubauer, A., Sonnemans, M. A., Sluijs, J. A., Koycu, S., Ramdjielal, R. D., Salehi, A., Martens, G. J., Grosveld, F. G., Peter, J., Burbach, H., and Hol, E. M. (1998) Frameshift mutants of beta amyloid precursor protein and ubiquitin-B in Alzheimer’s and Down patients. Science 279, 242–247 Wu, S. S., de Chadarevian, J. P., McPhaul, L., Riley, N. E., van Leeuwen, F. W., and French, S. W. (2002) Coexpression and accumulation of ubiquitin ⫹1 and ZZ proteins in livers of children with alpha(1)-antitrypsin deficiency. Pediatr. Dev. Pathol. 5, 293–298 French, B. A., van Leeuwen, F., Riley, N. E., Yuan, Q. X., Bardag-Gorce, F., Gaal, K., Lue, Y. H., Marceau, N., and French, S. W. (2001) Aggresome formation in liver cells in response to different toxic mechanisms: role of the ubiquitin-proteasome pathway and the frameshift mutant of ubiquitin. Exp. Mol. Pathol. 71, 241–246 Fratta, P., Engel, W. K., Van Leeuwen, F. W., Hol, E. M., Vattemi, G., and Askanas, V. (2004) Mutant ubiquitin UBB⫹1 is accumulated in sporadic inclusion-body myositis muscle fibers. Neurology 63, 1114 –1117 Song, S., Kim, S. Y., Hong, Y. M., Jo, D. G., Lee, J. Y., Shim, S. M., Chung, C. W., Seo, S. J., Yoo, Y. J., Koh, J. Y., Lee, M. C., Yates, A. J., Ichijo, H., and Jung, Y. K. (2003) Essential role of E2–25K/Hip-2 in mediating amyloid-beta neurotoxicity. Mol. Cell 12, 553–563 Johnson, E. S., Ma, P. C., Ota, I. M., and Varshavsky, A. (1995) A proteolytic pathway that recognizes ubiquitin as a degradation signal. J. Biol. Chem. 270, 17442–17456

The FASEB Journal

VERHOEF ET AL.

16.

17. 18. 19.

20.

21.

22.

23.

24.

25. 26. 27.

28. 29. 30.

Dantuma, N. P., Lindsten, K., Glas, R., Jellne, M., and Masucci, M. G. (2000) Short-lived green fluorescent proteins for quantifying ubiquitin/proteasome-dependent proteolysis in living cells. Nat. Biotechnol. 18, 538 –543 Luker, G. D., Pica, C. M., Song, J., Luker, K. E., and PiwnicaWorms, D. (2003) Imaging 26S proteasome activity and inhibition in living mice. Nat. Med. 9, 969 –973 Stack, J. H., Whitney, M., Rodems, S. M., and Pollok, B. A. (2000) A ubiquitin-based tagging system for controlled modulation of protein stability. Nat. Biotechnol. 18, 1298 –1302 Breitschopf, K., Bengal, E., Ziv, T., Admon, A., and Ciechanover, A. (1998) A novel site for ubiquitination: the N-terminal residue, and not internal lysines of MyoD, is essential for conjugation and degradation of the protein. EMBO J. 17, 5964 –5973 Menendez-Benito, V., Verhoef, L. G., Masucci, M. G., and Dantuma, N. P. (2005) Endoplasmic reticulum stress compromises the ubiquitin-proteasome system. Hum. Mol. Genet. 14, 2787–2799. De Vrij, F. M. S., Sluijs, J. A., Gregori, L., Fischer, D. F., Hermens, W. T. J. M. C., Goldgaber, D., Verhaagen, J., van Leeuwen, F. W., and Hol, E. M. (2001) Mutant ubiquitin expressed in Alzheimer’s disease causes neuronal death. FASEB J. 15, 2680 –2688 Van Leeuwen, F. W., Kros, J. M., Kamphorst, W., van Schravendijk, C., and de Vos, R. A. (2006) Molecular misreading: the occurrence of frameshift proteins in different diseases. Biochem. Soc. Trans. 34, 738 –742 Hope, A. D., de Silva, R., Fischer, D. F., Hol, E. M., van Leeuwen, F. W., and Lees, A. J. (2003) Alzheimer’s associated variant ubiquitin causes inhibition of the 26S proteasome and chaperone expression. J. Neurochem. 86, 394 – 404 Lam, Y. A., Pickart, C. M., Alban, A., Landon, M., Jamieson, C., Ramage, R., Mayer, R. J., and Layfield, R. (2000) Inhibition of the ubiquitin-proteasome system in Alzheimer’s disease. Proc. Natl. Acad. Sci. U. S. A. 97, 9902–9906 Liu, C., Apodaca, J., Davis, L. E., and Rao, H. (2007) Proteasome inhibition in wild-type yeast Saccharomyces cerevisiae cells. BioTechniques 42, 158, 160, 162 Bence, N. F., Sampat, R. M., and Kopito, R. R. (2001) Impairment of the ubiquitin-proteasome system by protein aggregation. Science 292, 1552–1555 Van Tijn, P., de Vrij, F. M., Schuurman, K. G., Dantuma, N. P., Fischer, D. F., van Leeuwen, F. W., and Hol, E. M. (2007) Dose-dependent inhibition of proteasome activity by a mutant ubiquitin associated with neurodegenerative disease. J. Cell Sci. 120, 1615–1623 Johnson, E. S., Bartel, B., Seufert, W., and Varshavsky, A. (1992) Ubiquitin as a degradation signal. EMBO J. 11, 497–505 Shabek, N., Iwai, K., and Ciechanover, A. (2007) Ubiquitin is degraded by the ubiquitin system as a monomer and as part of its conjugated target. Biochem. Biophys. Res. Commun. 363, 425– 431 Hanson, P. I., and Whiteheart, S. W. (2005) AAA⫹ proteins: have engine, will work. Nat. Rev. Mol. Cell. Biol. 6, 519 –529

MINIMAL LENGTH OF PROTEASOME SUBSTRATES

31. 32.

33.

34.

35.

36. 37. 38. 39. 40. 41.

42.

43.

44.

Navon, A., and Goldberg, A. L. (2001) Proteins are unfolded on the surface of the ATPase ring before transport into the proteasome. Mol. Cell 8, 1339 –1349 Smith, D. M., Kafri, G., Cheng, Y., Ng, D., Walz, T., and Goldberg, A. L. (2005) ATP binding to PAN or the 26S ATPases causes association with the 20S proteasome, gate opening, and translocation of unfolded proteins. Mol. Cell 20, 687– 698 Prakash, S., Tian, L., Ratliff, K. S., Lehotzky, R. E., and Matouschek, A. (2004) An unstructured initiation site is required for efficient proteasome-mediated degradation. Nat. Struc. Mol. Biol. 11, 830 – 837 Lee, C., Schwartz, M. P., Prakash, S., Iwakura, M., and Matouschek, A. (2001) ATP-dependent proteases degrade their substrates by processively unraveling them from the degradation signal. Mol. Cell 7, 627– 637 Reits, E., Griekspoor, A., Neijssen, J., Groothuis, T., Jalink, K., van Veelen, P., Janssen, H., Calafat, J., Drijfhout, J. W., and Neefjes, J. (2003) Peptide diffusion, protection, and degradation in nuclear and cytoplasmic compartments before antigen presentation by MHC class I. Immunity 18, 97–108 Thrower, J. S., Hoffman, L., Rechsteiner, M., and Pickart, C. M. (2000) Recognition of the polyubiquitin proteolytic signal. EMBO J. 19, 94 –102 Takeuchi, J., Chen, H., and Coffino, P. (2007) Proteasome substrate degradation requires association plus extended peptide. EMBO J. 26, 123–131 Zhang, M., Pickart, C. M., and Coffino, P. (2003) Determinants of proteasome recognition of ornithine decarboxylase, a ubiquitin-independent substrate. EMBO J. 22, 1488 –1496 Buchberger, A. (2002) From UBA to UBX: new words in the ubiquitin vocabulary. Trends Cell Biol. 12, 216 –221 Lambertson, D., Chen, L., and Madura, K. (2003) Investigating the importance of proteasome-interaction for Rad23 function. Curr. Genet. 42, 199 –208 Finney, N., Walther, F., Mantel, P. Y., Stauffer, D., Rovelli, G., and Dev, K. K. (2003) The cellular protein level of Parkin is regulated by its ubiquitin-like domain. J. Biol. Chem. 278, 16054 – 16058 Hanna, J., Hathaway, N. A., Tone, Y., Crosas, B., Elsasser, S., Kirkpatrick, D. S., Leggett, D. S., Gygi, S. P., King, R. W., and Finley, D. (2006) Deubiquitinating enzyme Ubp6 functions noncatalytically to delay proteasomal degradation. Cell 127, 99 –111 Raasi, S., and Pickart, C. M. (2003) Rad23 ubiquitin-associated domains (UBA) inhibit 26 S proteasome-catalyzed proteolysis by sequestering lysine 48-linked polyubiquitin chains. J. Biol. Chem. 278, 8951– 8959 Biggins, S., Ivanovska, I., and Rose, M. D. (1996) Yeast ubiquitinlike genes are involved in duplication of the microtubule organizing center. J. Cell Biol. 133, 1331–1346 Received for publication June 11, 2008. Accepted for publication August 21, 2008.

133