Sem1p Is a Novel Subunit of the 26 S Proteasome from Saccharomyces cerevisiae*

The 26 S proteasome, which catalyzes degradation of polyubiquitinated proteins, is composed of the 20 S proteasome and the 19 S regulatory particle (RP). The RP is composed of the lid and base subcomplexes and regulates the catalytic activity of the 20 S proteasome. In this study, we carried out affinity purification of the lid and base subcomplexes from the tagged strains of Saccharomyces cerevisiae, and we found that the lid contains a small molecular mass protein, Sem1. The Sem1 protein binds with the 26 S proteasome isolated from a mutant with deletion of SEM1 but not with the 26 S proteasome from the wild type. The lid lacking Sem1 is unstable at a high salt concentration. The 19 S RP was immunoprecipitated together with Sem1 by immunoprecipitation using hemagglutinin epitope-tagged Sem1 as bait. Degradation of polyubiquitinated proteins in vivo or in vitro is impaired in the Sem1-deficient 26 S proteasome. In addition, genetic interaction between SEM1 and RPN10 was detected. The human Sem1 homologue hDSS1 was found to be a functional homologue of Sem1 and capable of interacting with the human 26 S proteasome. The results suggest that Sem1, possibly hDSS1, is a novel subunit of the 26 S proteasome and plays a role in ubiquitin-dependent proteolysis.

The 26 S proteasome, which catalyzes degradation of polyubiquitinated proteins, is composed of the 20 S proteasome and the 19 S regulatory particle (RP). The RP is composed of the lid and base subcomplexes and regulates the catalytic activity of the 20 S proteasome. In this study, we carried out affinity purification of the lid and base subcomplexes from the tagged strains of Saccharomyces cerevisiae, and we found that the lid contains a small molecular mass protein, Sem1. The Sem1 protein binds with the 26 S proteasome isolated from a mutant with deletion of SEM1 but not with the 26 S proteasome from the wild type. The lid lacking Sem1 is unstable at a high salt concentration. The 19 S RP was immunoprecipitated together with Sem1 by immunoprecipitation using hemagglutinin epitope-tagged Sem1 as bait. Degradation of polyubiquitinated proteins in vivo or in vitro is impaired in the Sem1-deficient 26 S proteasome. In addition, genetic interaction between SEM1 and RPN10 was detected. The human Sem1 homologue hDSS1 was found to be a functional homologue of Sem1 and capable of interacting with the human 26 S proteasome. The results suggest that Sem1, possibly hDSS1, is a novel subunit of the 26 S proteasome and plays a role in ubiquitindependent proteolysis.
In eukaryotic cells, the ubiquitin-proteasome system regulates various cellular processes (1)(2)(3). In this pathway, target proteins are polyubiquitinated by E1 1 /E2/E3 enzymes, and the thus formed polyubiquitin chain is recognized by the 26 S proteasome, and the protein portion is degraded in an ATP-dependent manner. The 26 S proteasome is composed of a core particle (CP, also known as the 20 S proteasome), which contains proteolytic active sites in its cavity, and the 19 S regulatory particle (RP), which regulates the catalytic activity of the CP (2)(3)(4). The CP, built of two copies each of seven distinct ␣ and seven distinct ␤ subunits as four stacked rings (␣ 1-7 ␤ 1-7 ␤ 1-7 ␣ 1-7 ), has three catalytic activities, namely chymotrypsin-like, trypsin-like, and peptidylglutamyl peptide hydrolyzing activities (2,4). The RP consists of at least 17 distinct subunits (3,5) and is composed of lid and base subcomplexes. The lid subcomplex consists of eight non-ATPase subunits (Rpn3, Rpn5 to Rpn9, Rpn11, and Rpn12), whereas the base subcomplex consists of six ATPase subunits (Rpt1 to Rpt6) and two non-ATPase subunits (Rpn1 and Rpn2) (6,7). Recently, additional proteasome-associated proteins have been identified and characterized (8,9). It has been reported that Rpn13 is a new subunit of the RP and that a mutant with deletion of RPN13 shows a defect in the ubiquitin fusion degradation (UFD) pathway (8).
With the exception of RPN9, RPN10, and RPN13 genes, all of the regulatory subunit genes are essential. Rpn10 binds the polyubiquitin chain both in its free form and when incorporated into the 26 S proteasome (10 -13). Rpn10 positions at the interface between the lid and base and strengthens the lid-base interaction (6). The base is thought to promote translocation of substrates into the CP (14). Binding of the RP to the CP is thought to open a narrow pore at the ends of the CP. Rpt2 functions as a gating device for the CP channel (15,16). Rpt5/ S6Ј have been reported to be additional intrinsic polyubiquitin chain-binding subunits (17). Rpn1 acts as a receptor for ubiquitin-like proteins such as Rad23 (12,13). Among the lid subunits, Rpn3, Rpn6, Rpn9, Rpn11, and Rpn12 have been characterized (18 -21): Rpn11 has been reported to be a novel metalloprotease that has a deubiquitinating enzyme activity, releasing polyubiquitin chains from substrates (22)(23)(24).
In this study, we purified the base and lid subcomplexes of the 26 S proteasome from the yeast Saccharomyces cerevisiae, and we found that the respective subcomplexes have additional small molecular mass subunits. Rpn13 was identified as an additional subunit of the base, whereas Sem1 was identified as an additional subunit of the lid. Rpn13 has been reported to be one of the RP subunits (8). Sem1 was originally identified as a suppressor of the sec15-1 temperature-sensitive mutation and a negative regulator of pseudohyphal differentiation (25). We have obtained several lines of evidence that indicate that Sem1 is a novel lid subunit of the 26 S proteasome and is necessary for enzymatic activity of the 26 S proteasome. First, the recombinant Sem1 is capable of binding with the 26 S proteasome purified from a mutant with deletion of SEM1 (⌬sem1) but not with that from the wild type in vitro. Second, the purified 26 S proteasome from the ⌬sem1 strain seems to be structurally equivalent to the wild type 26 S proteasome, but the 26 S proteasome and the lid from ⌬sem1 became unstable under high salt conditions. Third, the 19 S RP can be immunoprecipitated with Sem1 as bait. Fourth, the ⌬sem1 strain shows a temperature-sensitive phenotype, and polyubiquitinated proteins were accumulated in this mutant at the restrictive temperature. Fifth, there was little degradation of polyubiquiti-nated Sic1 by the 26 S proteasome from the ⌬sem1 strain in vitro. Finally, genetic interaction between RPN10 and SEM1 genes was observed. The human DSS1 (hDSS1) gene, a homologue of SEM1, is a candidate gene responsible for an autosomal dominant form of split hand/split foot malformation (SHFM) disorder (26), and the hDSS1 protein has been reported to interact with the breast cancer susceptibility antigen BRCA2 (27). We demonstrated that hDSS1 complements the phenotype of the S. cerevisiae ⌬sem1 mutant and that hDSS1 is capable of interacting with the human 26 S proteasome in mammalian cells, suggesting that hDSS1 is functionally conserved among eukaryotes. Thus, these biochemical and genetic findings suggest that Sem1p is a component of the 26 S proteasome and functions in ubiquitin-dependent proteolysis.

EXPERIMENTAL PROCEDURES
Yeast Strains, Media, and Genetic Methods-The S. cerevisiae strains used in this study are listed in Table I. All strains are isogenic to W303. The Escherichia coli strain DH5␣ was used as a host for propagation and construction of plasmids. E. coli strain Rosetta (DE3) cells (Novagen) were used for expression of GST fusion proteins. Synthetic media (SD) were prepared as described previously (28). Synthetic media lacking appropriate nutrient(s) were used to select strains containing specific plasmids. Yeast nutrient-rich medium (YPD) consisted of 2% glucose, 2% polypeptone, 1% yeast extract (Difco), 400 g/ml adenine, and 20 g/ml uracil. Transformation of yeast was carried out by the method described by Schiestl and Gietz (29). For selection of G418 (Invitrogen) resistance, YPD plates containing 200 g/ml G418 were used. The permissive and restrictive temperatures for temperature-sensitive mutants were 25 and 37°C, respectively. The ⌬rpn10 strain (YYS80) was created by one-step disruption by PCR using a KanMX6 module by selection of geneticin resistance (30). To obtain ⌬sem1 mutants (YTS63), we first constructed the TSp361 plasmid as follows. An ϳ500-bp upstream fragment (EcoRI-XhoI) and a 500-bp downstream fragment (BglII-EcoRI) of the SEM1 ORF were amplified by PCR and cloned into the BamHI-XhoI gap of pRS306 to generate TSp361. TSp361 was linearized by digesting with EcoRI and used as a donor to transform W303-1A. The correct disruption was verified by PCR.
To obtain the tagged strains, the respective plasmids were constructed as follows. First, the sequences encoding the PreScission protease site, 3ϫ FLAG epitope (3FLAG) (Sigma), and the transcriptional terminator from the CYC1 gene were amplified and fused by sequential PCR (PFT cassette). Next, the 3Ј-segment of the ORF of the PRE1, RPT1, or RPN11 gene (without the stop codon) was fused in-frame to the PFT cassette by PCR and cloned into the YIp plasmid pRS303. To obtain the RPT6 5HA strain, the plasmid was constructed as follows. First, the sequences encoding five tandem repeats of the influenza virus hemagglutinin epitope (5HA) and the transcriptional terminator from the TDH3 gene were amplified and fused by sequential PCR (HT cassette). Next, the 3Ј-segment of the ORF of the RPT6 gene was fused in-frame to the HT cassette by PCR and cloned into the YIp plasmid pRS304. The resultant plasmid was then linearized by digestion with an appropriate restriction enzyme at a site within the coding sequence of the yeast gene and targeted into the genome.
Construction of Plasmids-The ORF of SEM1 containing the EcoRI and XhoI sites was amplified by PCR from genomic DNA of W303-1A and cloned into the pGEM-T plasmid (Promega). The ORFs of human DSS1 (hDSS1) and POH1 (hRPN11) containing the EcoRI and XhoI sites were amplified by PCR from a two-hybrid library of human HeLa cDNA (Clontech Laboratories) and cloned into the pGEM-T plasmid. All constructs were verified by DNA sequencing. To construct YCp-SEM1 3HA , the SEM1 gene with its promoter region, 518 bp upstream, was amplified by PCR and cloned into the pGEM-T plasmid, and the 786-bp fragment that had been produced from the above plasmid by digestion with PstI and XhoI was inserted into the PstI and XhoI sites of the pTS901CT plasmid (YCp, TRP1, 5ϫ HA epitope) (31).
For suppression analysis, the SEM1 gene with its own promoter region (upstream of 518 bp) and terminator region (downstream of 540 bp) was amplified by PCR and subcloned into the low copy plasmid pRS314. The hDSS1 gene was subcloned into the pKT10 plasmid, and the fragment containing the DSS1 gene flanked by 5Ј-and 3Ј-untranslated regions of the TDH3 gene was subcloned into pRS314.
For expressing GST fusion proteins, the SEM1 fragment that had been digested with EcoRI and XhoI was subcloned into the EcoRI and XhoI sites of the pGEX6P1 plasmid (Amersham Biosciences) and transformed into E. coli strain Rosetta (DE3) (Novagen). To generate the mammalian expression plasmids, the respective fragments of hDSS1 and POH1 were inserted into the EcoRI and SalI sites of the mammalian expression plasmid pCI-neo-3FLAG that had been generated by inserting the oligonucleotides encoding 3FLAG into the SalI and NotI sites of the pCI-neo plasmid (Promega).
Affinity Purification of the 26 S Proteasome and Its Subcomplexes-The 26 S proteasome was isolated by a modification of the method described by Verma et al. (8). The respective tagged and untagged strains of S. cerevisiae were grown to an absorbance of 600 nm (A 600 ) at 2.0 in YPD medium at 25°C. Cells were harvested and washed twice with ice-cold water and once with buffer A (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, and 10% glycerol). The pellet was stored at Ϫ80°C until use. Cells (4 g) were thawed, suspended in 4 ml of buffer B (buffer A containing 4 mM ATP, 10 mM MgCl 2 , and 2ϫ ATP regeneration system), and lysed by glass beads using a Bead-Beater (Biospec Products). An equal volume of buffer B was added to the lysate, and the mixture was centrifuged at 20,000 ϫ g for 30 min at 2°C. 80 l of anti-FLAG M2-agarose beads (Sigma) was added to the recovered supernatant, and the mixture was rotated for 1 h at 4°C. The beads were washed four times with 10 ml of buffer C (buffer A containing 2 mM ATP and 5 mM MgCl 2 ), twice with 10 ml of buffer C containing 0.2% Triton X-100, and again twice with 10 ml of buffer C. The beads were transferred to a 2-ml disposable column (Bio-Rad) and incubated with 100 l of 3ϫ FLAG peptide (100 g/ml) (Sigma) in buffer C for 30 min at 4°C to elute the 26 S proteasome. The purified 26 S proteasome was stored at Ϫ80°C. The base, lid, and CP were isolated using YYS98 (⌬rpn10 RPN1 3FLAG ), YYS99 (⌬rpn10 RPN113 FLAG ) and YYS37 (PRE13 FLAG ) strains, respec- tively, as follows. Anti-FLAG M2 beads, to which the 26 S proteasome from each of the above strains had been bound, were incubated with buffer D (50 mM Tris-HCl, pH 7.5, 500 mM NaCl, and 10% glycerol) for 1 h at 4°C to dissociate the 26 S proteasome to the lid, base, and CP. The beads were then washed four times with 10 ml of buffer D, incubated with buffer D containing 0.2% Triton X-100 for 5 min at 25°C, and washed twice with the same buffer and twice with buffer A. The base, lid, or CP, bound to the beads, was then eluted with 3ϫ FLAG peptide as described above. SDS-PAGE and Peptide Sequence Analysis-The isolated base and lid were resolved by 12.5% SDS-PAGE, and protein bands were excised from the SDS-polyacrylamide gel. The gels containing the respective protein bands were destained with 25 mM ammonium bicarbonate containing 50% acetonitrile and were completely dehydrated by evaporation using a SpeedVac vacuum centrifuge. After rehydration by adding a gel volume of 50 mM phosphate buffer, pH 7.8, containing 10 ng/l Staphylococcus aureus V8 protease (Wako Pure Chemicals) or 20 mM Tris-HCl, pH 9.0, containing 10 ng/l Achromobacter lyticus lysyl endopeptidase (Wako) at 4°C for 30 min, the gels were incubated at 37°C for 12 h to allow the above proteases to digest the respective proteins. The resultant peptides in the gels were extracted by adding a gel volume of a solution consisting of 50% acetonitrile and 5% trifluoroacetic acid. This extraction procedure was repeated twice. The extracted peptides were then combined, evaporated to about 20 l in a volume using a SpeedVac vacuum centrifuge, and acidified by addition of trifluoroacetic acid at the final concentration of about 0.1%. Peptide fragments were then separated by reverse-phase high performance liquid chromatography with a 1-80% acetonitrile gradient in 0.05% trifluoroacetic acid on a column (0.46 ϫ 10 cm) of TSK gel Super ODS (Tosoh, Japan). Peptides eluted as peaks were subjected to amino acid sequence determination on Procise 492 cLC (Applied Biosystems).
Peptide Mass Fingerprinting-Proteins on the SDS-PAGE gel were excised, destained as described above, and digested in gels with 10 g/ml modified trypsin (Promega) in 15 l of 20 mM ammonium bicarbonate. The resulting peptides were desalted using ZipTip-C 18 (Millipore) and mixed with 3 mg/ml ␣-cyano-4-hydroxycinnamic acid (Sigma) in 0.3% trifluoroacetic acid and 50% acetonitrile. The peptide mass fingerprints of proteins were obtained by a matrix-assisted laser desorption ionization/time of flight MS on Voyager DE-PRO (Applied Biosystems). A data base search was performed using MS-Fit.
Nondenaturing PAGE and In-gel Peptidase Assay-Nondenaturing PAGE on 4% polyacrylamide gel and the subsequent in-gel peptidase assay for the proteasome using a fluorogenic substrate, Suc-Leu-Leu-Val-Tyr-MCA (Peptide Institute, Japan), were carried out as described by Glickman et al. (5). Gels were run in a cold room until the dye (xylene cyanol) had run off. The gels were then incubated with the substrate for 10 min at 30°C in the presence of ATP and 0.05% SDS in buffer A for detection of the 26 S proteasome and the 20 S proteasome, respectively (5). The signals due to the proteasomes were visualized and photographed upon exposure to a UV transilluminator (360 nm).
Binding of GST-Sem1 with the 26 S Proteasome-GST or GST fusion protein was expressed and isolated from Rosetta (DE3) cells transformed with pGEX6P1 or pGEX6P1 containing the respective gene as described previously (13). GST-Sem1 or GST (1 g each) was diluted in 0.5 ml of binding buffer (buffer A containing 1 mM dithiothreitol, 0.2% Triton X-100, and 5% non-fat milk) and was then incubated with 10 l of glutathione-immobilized agarose beads (Amersham Biosciences) for 1 h at room temperature to allow GST-Sem1 or GST to bind to the beads. After washing with binding buffer, the beads were suspended in 200 l of binding buffer containing 2 mM ATP and 5 mM MgCl 2 . The affinity-purified 26 S proteasome from the wild type or ⌬sem1 strain (2 g each) was added to the beads, and the mixture was incubated for 30 min at room temperature. After the beads had been washed five times with 500 l of the same buffer, the proteins retained in the beads were eluted with SDS-loading buffer and subjected to SDS-PAGE, followed by Western blotting with anti-Rpt5, anti-Rpn9, and anti-20 S proteasome antibodies.
Immunoprecipitation with Anti-HA Antibody-immobilized Beads-The S. cerevisiae strains were grown to an A 600 at 1.0 in SD-Trp media at 30°C. Cells were pelleted by centrifugation and washed with buffer A. The pellet was suspended in 1 ml of buffer A and lysed by glass beads. An equal volume of buffer A was added to the lysate, and the suspension was centrifuged at 20,000 ϫ g for 30 min at 2°C. 25 l of anti-HA affinity matrix (Roche Applied Science) was added to the supernatant, and the mixture was rotated for 1 h at 4°C. The beads were washed four times with buffer A, and the bound proteins were then eluted with 40 l of 0.1 M glycine-HCl, pH 3.0, and immediately neutralized by addition of 1.5 l of 1.5 M Tris-HCl, pH 8.8.
Accumulation of Polyubiquitinated Proteins-The wild type and mutant strains of S. cerevisiae were cultured in YPD medium at 25°C to the logarithmic phase and then cultured at 37°C. Cells corresponding to an A 600 at 1.0 unit were periodically harvested and lysed by the mild alkali method (34). The extract was subjected to SDS-PAGE and then to Western blotting with an FK2 antibody that specially recognizes polyubiquitin chains (32) and also with an anti-actin antibody (Sigma) as a control.
Preparation of Polyubiquitinated Cdc34 and Gel Shift Assay for Its Binding with the 26 S Proteasome-Polyubiquitinated Cdc34 was produced using Cdc34/Ubc3 (E2) and Uba1 (E1). (The details of isolation procedures of Cdc34 and Uba1 will be published elsewhere.) A reaction mixture consisting of 25 mM Tris-HCl, pH 7.5, 50 mM NaCl, 10 mM dithiothreitol, 10 mM MgCl 2 , 10 mM ATP, 2.5 g/l ubiquitin (Sigma), 40 ng/l Uba1, and 200 ng/l Cdc34 was incubated for 12 h at 37°C. The level of polyubiquitination was estimated by Western blotting with an FK2 antibody. The resultant polyubiquitinated Cdc34 in the indicated amount was mixed with 3 pmol of the affinity-purified 26 S proteasomes in a final volume of 4 l, and the mixture was incubated for 15 min at room temperature and then subjected to nondenaturing PAGE, followed by in-gel peptidase assay for the proteasome.
Preparation of Polyubiquitinated Sic1 PY and in Vitro Degradation Assay-To prepare polyubiquitinated Sic1 PY , we developed a novel method using Rsp5 as an E3 ligase. (The details of this method will be published elsewhere.) In brief, the PY motif was introduced into the N-terminal region of Sic1, designated as Sic1 PY , to allow Rsp5 to bind to Sic1. The resultant Sic1 PY and the original Sic1 were expressed as fusion proteins with T7 tags at the N termini and HAT tags (Clontech) at the C termini, designated as T7-Sic1 PY -HAT and T7-Sic1-HAT, respectively, constructed by the pET system (Novagen). T7-Sic1 PY -HAT and T7-Sic1-HAT proteins were purified according to the method described by Verma et al. (23) and were then incubated with Uba1 (E1), Ubc4 (E2), and Rsp5 (E3). (The details of isolation procedures of Ubc4 and Rsp5 will be published elsewhere.) The resultant polyubiquitinated Sic1 PY (400 nM) was incubated with the affinity-purified 26 S proteasome (200 nM) from the wild type or ⌬sem1 strain at 30°C. The reaction was terminated by the addition of SDS-loading buffer, and the reaction mixture was then subjected to SDS-PAGE, followed by Western blotting with an anti-T7 antibody to detect the degree of degradation.
Growth Assay-The wild type and mutant strains of S. cerevisiae were cultured in YPD medium at 25°C to A 600 of 1.0 -2.0, and their densities were adjusted to A 600 of 0.5 by adding YPD medium. A 10-fold serial dilution was prepared in deionized water, and the serially diluted cells (2 l each) were spotted on YPD agar plates and incubated at 25 or 37°C for 2 days. For suppression analysis, low copy plasmids, pRS314-SEM1 and pRS314-DSS1, were transformed into the ⌬sem1 strain (YTS63). The transformants were spotted onto SD-Trp plates and incubated at 25 or 37°C for 2 days.
Mammalian Cell Culture, Transfection, and Immunoprecipitation-Human embryonic kidney (HEK) 293T cells were cultured in Dulbecco's modified Eagle's medium (Sigma) supplemented with 10% heat-inactivated fetal bovine serum (Invitrogen) at 37°C under a 5% CO 2 atmosphere. HEK293T cells were transfected with the pCI-neo-hDSS1 3FLAG or pCI-neo-POH1 3FLAG plasmid using Metafectene (Biontex) according to the manufacturer's protocol and cultured for 36 h at 37°C. The cells were washed with ice-cold phosphate-buffered saline, suspended with buffer A containing 0.2% Nonidet P-40 with or without 2 mM ATP and 5 mM MgCl 2 , and were then lysed using a 23-gauge needle. The lysate was centrifuged at 20,000 ϫ g for 30 min at 2°C, and the resulting supernatant was incubated with 20 l of anti-FLAG M2 beads for 1 h at 4°C. After the beads had been washed five times with the respective buffers, the materials bound to the beads were eluted with 50 l of 3ϫ FLAG peptide (100 g/ml) in buffer C and subjected to SDS-PAGE and to Western blotting. Alternatively, the eluates obtained from the above anti-FLAG M2 beads were subjected to an assay of proteasome activity with 0.1 mM Suc-Leu-Leu-Val-Tyr-MCA in buffer E (50 mM Tris-HCl, pH 7.5, 2 mM ATP, 10 mM MgCl 2 , and 1 mM dithiothreitol) for 20 min at 37°C. The reaction was stopped by the addition of 5% SDS, and fluorescence due to released 7-amino-4-methylcoumarin was measured with excitation at 380 nm and emission at 460 nm.

RESULTS
Affinity Purification of the 26 S Proteasome and Its Subcomplexes from S. cerevisiae-To isolate the 26 S proteasome and its subcomplexes from S. cerevisiae, Rpn1 (a base subunit), Rpn11 (a lid subunit), and Pre1 (a 20 S proteasome subunit) were individually tagged with 3ϫ FLAG at their C termini by chromosomal homologous recombination (Fig. 1A). To exclude the possibility that the epitope tag affects the 26 S proteasome assembly and its activity, the extracts from the above tagged strains were subjected to in-gel peptidase assays using Suc-Leu-Leu-Val-Tyr-MCA (Fig. 1B). It was found that the patterns of the signals due to the peptidase activities of the proteasomes from the above tagged and the wild type strains were essentially the same. In addition, the subunit composition patterns of the 26 S proteasomes affinity-purified from the tagged cells by using anti-FLAG antibody-immobilized beads were almost the same as the pattern observed for the 26 S proteasome purified by conventional chromatography (5), and peptide mass fingerprinting revealed that the 26 S proteasomes from the tagged strains contained all of the subunits and some proteasome-interacting proteins that had been reported previously (6 -9) (data not shown). Thus, we concluded that 3ϫ FLAG epitope tagging to the respective subunits should not affect the structure and enzymatic activity of the 26 S proteasome and that the respective tagged strains can be used to isolate the 26 S proteasome. Because the lid subcomplex has been reported to be easily dissociated from the 26 S proteasome prepared from a mutant with deletion of RPN10 (⌬rpn10) (6), the base and lid subcomplexes were isolated from the respective tagged ⌬rpn10 strains by affinity chromatography in the presence of 500 mM NaCl (Fig. 1C, base and lid). The polypeptide pattern of the 26 S proteasome was a close summation of those of the base, lid, and 20 S proteasome (CP) (Fig. 1C).
Small Molecular Mass Proteins Are Co-purified with the Lid and Base Subcomplexes-Closer inspection of Fig. 1C revealed that the affinity-purified lid and base subcomplexes contain two proteins with molecular masses of 19.5 (band 1 in Fig. 1C) and 17.5 kDa (band 2), respectively. Both bands were also detected in the 26 S proteasome but not in the 20 S proteasome. It is possible that these proteins are degradation products of 19 S RP itself, but the fact that these proteins were found to be co-fractionated with the respective subcomplexes by conventional purification procedures, including anion exchange chromatography and gel filtration (7), suggests that these small proteins are components of the 19 S RP or proteasome-interacting proteins.
To identify these proteins, the respective protein bands were excised from the gel and were in-gel digested with V8 protease or lysyl endopeptidase. The resultant peptides were extracted, separated by reverse-phase high performance liquid chromatography, and subjected to amino acid sequence determination. The identified peptides are listed in Table II. The sequence of each peptide completely matched the deduced sequence from the yeast ORF reported in the Saccharomyces Genome Data Base, allowing the assignment of each protein as the product of a specific chromosomal gene. Given that the S. cerevisiae genome is entirely known, the peptide sequence data are sufficient to assign each band as the product of a single gene. As a result, band 1 was identified as a small acidic protein, Sem1, whereas band 2 was identified as Rpn13, a recently identified novel subunit of the budding yeast 19 S RP (8). Thus, the latter result strongly suggests that Rpn13 is a base subunit.
By amino acid sequencing, Sem1 was identified as a protein bound to the affinity-purified lid. Sem1 is a small (89 amino acid residues) acidic protein with a theoretical molecular mass and pI of 10,386.15 Da and 4.16, respectively. The difference between the theoretical molecular mass (10.4 kDa) and apparent mobility in the SDS-polyacrylamide gel (19.5 kDa) of Sem1 may be caused by its acidity, because bacterially expressed Sem1 also migrated to a point about 19 kDa in SDS-PAGE (data not shown). The SEM1 (Suppressor of Exocytosis Mutations) gene is located at YDR363W-A and has been cloned as a multicopy suppressor of the temperature-sensitive mutation sec15-1 (25). The SEM1 gene is not essential, and its deletion strain shows a temperature-sensitive phenotype. Sem1 is highly conserved across evolution. Sem1 and the human homologue are 69% similar and 47% identical, whereas the mouse, rat, and human homologues are 100% identical (25). Thirteen amino acid residues are completely conserved throughout the species, suggesting a crucial role for these residues. The human homologue, DSS1 (GenBank TM accession number U41515), has been mapped to the locus affected in the autosomal dominant form of SHFM disorder (26). Because Sem1 or its homologue The crude extracts from the epitope-tagged yeast cells, YYS37 (PRE1 3FLAG ), YYS39 (RPN1 3FLAG ), and YYS40 (RPN11 3FLAG ), were subjected to nondenaturing PAGE followed by in-gel peptidase assay using Suc-Leu-Leu-Val-Tyr-MCA in the presence of ATP and 0.05% SDS. RP 2 CP, the double-capped proteasome; RP 1 CP, the single-capped proteasome. WT, wild type. C, SDS-PAGE analyses of the purified 26 S proteasome, base, lid, and 20 S proteasome (CP). The 26 S proteasome was purified from YYS39 (RPN1 3FLAG ). The lid, base, and CP were purified from YYS98 (⌬rpn10 RPN1 3FLAG ), YYS99 (⌬rpn10 RPN-11 3FLAG ), and YYS37 (PRE1 3FLAG ), respectively. The purified complexes were analyzed by SDS-PAGE and visualized by protein staining with Coomassie Brilliant Blue (CBB). The bands corresponding to tagged subunits are indicated by arrowheads. The no-tag lane refers to a control sample that had been affinity-purified from an untagged parental strain (W303-1A). Bands 1 and 2 were identified as Sem1 and Rpn13, respectively, as shown in Table II. M, molecular mass standards.
has not been reported to be a proteasome subunit in either yeast or other species, it should be tested whether Sem1 is a subunit of the 26 S proteasome.
GST-Sem1 Binds with Only the 26 S Proteasome from ⌬sem1 Strain-It is known that some lid subunit mutations result in a structural defect of the 26 S proteasome. For example, the 26 S proteasome purified from a ⌬rpn9 mutant lacked some subunits and was easily dissociated into subcomplexes (33), and the lid-less proteasome was affinity-purified from an rpn11 mpr1-1 mutant (23). To characterize the 26 S proteasome from the mutant with deletion of SEM1 (⌬sem1), the 26 S proteasomes were affinity-purified from YYS40 (RPN11 3FLAG ) and YTS67 (⌬sem1 RPN11 3FLAG ) and subjected to SDS-PAGE analysis ( Fig. 2A). The subunit composition of the 26 S proteasome from the ⌬sem1 strain was essentially the same as that from the wild type strain (SEM1) except for Sem1.
To determine whether the lack of Sem1 affects the structure and/or enzymatic activity of the 26 S proteasome, the affinitypurified 26 S proteasomes from the respective strains were separated by nondenaturing PAGE, followed by in-gel peptidase assay using Suc-Leu-Leu-Val-Tyr-MCA (Fig. 2B). The mobility and peptidase activity of the 26 S proteasome from the ⌬sem1 strain were found to be equivalent to those of the 26 S proteasome from the wild type SEM1 strain. The same result was obtained when peptidase activity was assayed under SDSfree conditions (data not shown). As judged from the data of SDS-PAGE and nondenaturing PAGE, the purified 26 S proteasome seems to be a double-capped form (RP 2 CP). Thus, these results imply that the global structure of the 26 S proteasome remains unchanged in the absence of Sem1. In addition, it should be noted that the band intensity of Sem1 in the SDS-polyacrylamide gel seems to be almost equivalent to that of Rpn13 (see Figs. 1C and 2A), implying that Sem1 is a stoichiometric component of the 26 S proteasome. To confirm this, we performed a GST pull-down assay using GST-Sem1 fusion protein. GST-Sem1 or GST, expressed and purified from E. coli, was bound to glutathione-immobilized agarose beads and incubated with the affinity-purified 26 S proteasome from the wild type SEM1 or ⌬sem1 strain. After washing, the beadbound proteins were eluted with SDS loading buffer and subjected to SDS-PAGE followed by Western blotting with anti-Rpt5 (base), anti-FLAG (Rpn11 3FLAG , lid), and anti-20 S proteasome antibodies (Fig. 2C). The 26 S proteasome from the ⌬sem1 strain was able to bind with GST-Sem1 (lane 9), whereas that from the wild type SEM1 strain did not bind with GST-Sem1 (lane 8). Taking account of the fact that several proteins, including Rad23 and Dsk2, known as extrinsic ubiquitin receptors for the 26 S proteasome, are able to bind with the affinity-purified 26 S proteasome from the wild type strain (12, 13) together with the above-stated finding that Sem1 is unable to bind with the intact 26 S proteasome, we propose that Sem1 is an intrinsic component of the 26 S proteasome rather than a proteasome-interacting protein.
The 26 S Proteasome and the Lid from ⌬sem1 Strain Are Unstable under High NaCl Conditions-To determine the structural role of Sem1 in the 26 S proteasome, we tested the stability of the 26 S proteasome and the lid subcomplex at high salt concentrations (Fig. 3A). The 26 S proteasomes were pulled down by using anti-FLAG antibody-immobilized beads from strains, including untagged control strain (Fig. 3A, I), RPN11 3FLAG strain (Fig. 3A, II), ⌬sem1 RPN11 3FLAG strain (Fig. 3A, III), and ⌬sem1 RPN11 3FLAG strain carrying YCp-SEM1 3HA plasmid (Fig. 3A, IV). The bead-bound 26 S proteasomes were treated with NaCl at various concentrations, and the bead-retained proteins were eluted with 3ϫ FLAG peptides and analyzed by SDS-PAGE and Western blotting with anti-FLAG (Rpn11 3FLAG , bait), anti-Rpn9 (lid), anti-Rpt5 (base), and anti-HA (Sem1 3HA ) antibodies. Because 3ϫ FLAG epitope had been tagged to Rpn11, the Rpn11 3FLAG -containing complexes should be pulled down and eluted with 3ϫ FLAG peptides in these experiments. Under control conditions with 0.1 M NaCl, equal amounts of the 26 S proteasomes from the respective strains were detected by both Western blotting and SDS-PAGE analyses (Fig. 3A, lanes 6 -8), and under conditions in which the bead-bound 26 S proteasome had been treated pre-FIG. 2. GST-Sem1 binds with the 26 S proteasome isolated from ⌬sem1 strain. A, subunit compositions of the 26 S proteasomes purified from the wild type SEM1 strain, YYS40 (RPN11 3FLAG ), and the ⌬sem1 strain, YTS67 (⌬sem1 RPN11 3FLAG ). The proteasomes were resolved by 12% SDS-PAGE, and protein bands were visualized by Coomassie Brilliant Blue (CBB) staining. The no-tag lane refers to a control sample that had been affinity-purified from an untagged parental strain, W303-1A. The position of Sem1 is indicated by an asterisk. M, molecular mass standards. B, nondenaturing PAGE and in-gel peptidase assay of the proteasomes purified from the wild type SEM1 and ⌬sem1 strains. The proteasomes were visualized by in-gel peptidase assay as in Fig. 1B. The 26 S proteasomes from both strains were detected at the same positions corresponding to the double-capped proteasome (RP 2 CP) and had almost equal peptidase activities. C, GST pull-down assay using GST-Sem1 as bait. One g each of GST or GST-Sem1 was bound to glutathione-immobilized agarose beads and was then incubated with 2 g each of the 26 S proteasome purified from the wild type SEM1 strain (b), the 26 S proteasome purified from ⌬sem1 strain (c), or control sample purified from no-tag strain (a) in the same volume. The materials bound to the beads were eluted with SDSloading buffer and analyzed by Western blotting with anti-Rpt5 (base), anti-FLAG (Rpn11 3FLAG , lid), and anti-20 S proteasome antibodies. viously with 0.5 M NaCl for 5 min at 25°C (lanes 9 -11) or with 1 M NaCl for 30 min at 25°C (lanes 12-14), dissociation of the 26 S proteasome was detected in all cases, especially in the case of the ⌬sem1 strain (Fig. 3A, lanes 10 and 13). These results imply that the 26 S proteasome from the ⌬sem1 strain is more sensitive to NaCl treatment, which causes dissociation of the 26 S proteasome into the base and the 20 S proteasome in comparison with the case of the wild type SEM1 strain (Fig. 3A, compare lane 9 with 10 and lane 12 with 13). Densitometric analysis revealed that all lid subunits were dissociated in the ⌬sem1 strain under the most severe conditions (Fig. 3B, lane  13), although equal amounts of Rpn11 3FLAG remained bound to the beads in both cases of the wild type and ⌬sem1 strains (Fig.  3B, lanes 12 and 13). In addition, it should be noted that instability of the lid in the ⌬sem1 strain was fully suppressed by expressing Sem1 3HA (lanes 11 and 14 in Fig. 3A and lane 14 in Fig. 3B). In this case, Sem1 3HA , as well as Sem1, was found to bind to the lid, as detected by protein staining with SYPRO Orange (Fig. 3A, bottom panel). These results clearly indicate that Sem1 is a component of the lid and is required for assembly of the integral structure of the lid and the 26 S proteasome. The 19 S RP Is Able to be Co-immunoprecipitated with Sem1-To eliminate the possibility that Sem1 is contained in other complexes, e.g. an exocytic complex, we carried out immunoprecipitation with Sem1 3HA as bait. Cell extracts prepared from two tagged transformants, YTS63 cells carrying YCp-SEM1 3HA and YYS143 (RPT6 5HA ) cells (control), were subjected to immunoprecipitation with anti-HA antibody-immobilized beads, and the bead-bound proteins were eluted at low pH. The respective eluates were resolved by SDS-PAGE and analyzed by Western blotting with anti-Rpn9 (lid), anti-Rpt5 (base), and anti-HA (Rpt6 5HA and Sem1 3HA ) antibodies (Fig. 4A) and by protein staining (Fig. 4B). Equivalent amounts of almost all 19 S RP subunits were recovered in either case of immunoprecipitation with Rpt6 5HA or Sem1 3HA as bait (Fig.  4B, lanes b and c). These results indicate that Sem1 is a stoichiometric component of the 19 S RP and that the 19 S RP is the only major Sem1-containing complex.
Polyubiquitinated Proteins Are Accumulated in ⌬sem1 Strain at a Restrictive Temperature-The above biochemical results unambiguously indicate that Sem1 is a component of the lid, but the 26 S proteasome lacking Sem1 seems to have normal peptidase activity (see Fig. 2B). On the other hand, it has been reported that accumulation of polyubiquitinated proteins was detected at a restrictive temperature in several lid subunit mutants such as ⌬rpn9 and rpn12-1 strains (33). In our background strain W303, the ⌬sem1 strain showed a temperature-sensitive phenotype as described previously (27) (see Fig.  9A). Therefore, we next carried out an experiment to determine whether polyubiquitinated proteins are accumulated in the ⌬sem1 strain at a restrictive temperature. The wild type W303-1A and ⌬sem1 strains were grown at 25°C to mid-logarithmic phase and then grown at 37°C. The cells were harvested at the indicated times, and the total cell lysate was prepared and analyzed by Western blotting with an FK2 antibody recognizing polyubiquitin chains. As shown in Fig. 5, large amounts of high molecular mass polyubiquitinated proteins accumulated at the restrictive temperature in the ⌬sem1 strain in comparison with the case of the wild type, indicating that the function of the 26 S proteasome in the degradation of polyubiquitinated proteins is defective in the ⌬sem1 strain at the restrictive temperature. Thus, it can be concluded that Sem1 plays a role in ubiquitin-dependent proteolysis.
The 26 S Proteasome Lacking Sem1 Is Unable to Degrade a Polyubiquitinated Protein in Vitro-To obtain definitive in vitro evidence that Sem1 is required for degradation of polyubiquitinated proteins, we next carried out two in vitro experiments including a polyubiquitin chain binding assay and a polyubiquitinated protein degradation assay. A gel shift assay has recently been developed to evaluate polyubiquitin chain binding activity of the 26 S proteasome (12). According to this method, polyubiquitinated Cdc34 was prepared in the presence of ubiquitin, Uba1 (E1), and ATP (Fig. 6A) and was incubated  5-8). The bead-bound 26 S proteasome was treated with 0.5 M NaCl for 5 min at 25°C (lanes 9 -11) or with 1 M NaCl for 30 min at 25°C (lanes 12-14) and then washed with NaCl at the respective concentrations. The proteins, bound to the beads, were then eluted with FLAG peptide and analyzed by Western blotting with anti-FLAG (Rpn11 3FLAG , bait), anti-Rpn9 (lid), anti-Rpt5 (base), and anti-HA (Sem1 3HA ) antibodies (upper panel). Input refers to 6% of the whole cell lysate used for immunoprecipitation (lanes 1-4). The purified proteins were also analyzed by SDS-PAGE followed by protein staining with SYPRO Orange (bottom panel). The positions of Sem1 and Sem1 3HA are indicated by an asterisk and an arrowhead, respectively. M, molecular mass standards. B, densitometric analysis of the 26 S proteasomes pretreated with 1 M NaCl. The protein bands detected in the SDSpolyacrylamide gel in A (lanes 12-14) were analyzed by ImageGauge software. The lid subunits corresponding to the respective peaks are indicated above, and Rpn11 3FLAG is indicated by an arrowhead.
with the 26 S proteasome purified from the wild type SEM1 or ⌬sem1 strain. The resultant reaction mixture containing the 26 S proteasome and polyubiquitinated Cdc34 was subjected to nondenaturing PAGE and to in-gel peptidase assay (Fig. 6B). In both cases of the SEM1 and ⌬sem1 strains, the 26 S protea-somes were gel-shifted to similar points in the presence of polyubiquitinated Cdc34, suggesting that the 26 S proteasome retains the ability to recognize/bind polyubiquitin chains even in the absence of Sem1. In a control experiment using original Cdc34, i.e. when ubiquitin or E1 enzyme was omitted, gel shift of the 26 S proteasome was not detected (data not shown).
Next, we tested whether the 26 S proteasome from the ⌬sem1 strain retains the activity to degrade polyubiquitinated proteins in vitro. Polyubiquitinated Sic1 is known as a good substrate for the 26 S proteasome (8, 23), but its preparation using an E3 ligase complex SCF Cdc4 in the present study was technically difficult. We therefore prepared polyubiquitinated Sic1 PY by our original method using Rsp5 as an E3 ligase. To allow Rsp5 to bind to Sic1, the PY motif (35), a sequence required for Rsp5 binding, was introduced into the N-terminal region of Sic1, designated as Sic1 PY . T7 epitope-and HAT epitope-tagged Sic1 PY , T7-Sic1 PY -HAT, was highly polyubiquitinated by Uba1, Ubc4, and Rsp5 in a PY motif-dependent manner, and the resultant polyubiquitinated-Sic1 PY was efficiently degraded by the purified 26 S proteasome (Fig. 7A). By using this polyubiquitinated-Sic1 PY as a substrate, the activities of the 26 S proteasomes purified from the wild type SEM1 and ⌬sem1 strains were measured. As shown in Fig. 7, B and C, ϳ80% of polyubiquitinated Sic1 PY (Ubn-Sic1 PY ) was quickly degraded by the 26 S proteasome from the wild type strain within 3 min under the conditions used, whereas the degradation by the 26 S proteasome lacking Sem1 proceeded very slowly and ϳ50% of polyubiquitinated-Sic1 PY remained undigested after 10 min. Thus, the 26 S proteasome lacking Sem1 has a defect in the degradation of the polyubiquitinated protein, strongly supporting the above idea that Sem1 plays a role in ubiquitin-dependent proteolysis.
The ⌬rpn10 ⌬sem1 Double-deletion Mutant Displays a Synthetic Phenotype for Vegetative Growth-Because Sem1 was found to be a component of the lid, we next analyzed genetic interactions among SEM1 and other proteasome subunit genes. Because the mutant with deletion of RPN10 that encodes a proteasomal intrinsic ubiquitin receptor has been well characterized and shows a mild phenotype (10,11), a ⌬sem1 ⌬rpn10 double-deletion mutant was constructed to analyze the genetic interaction between SEM1 and RPN10. It was found that this double-deletion mutant is viable but shows slower growth than that of the wild type strain or those of single deletion mutants (Fig. 8). The wild type strain and ⌬rpn10  (FK2). B, gel shift assay using polyubiquitinated Cdc34. Polyubiquitinated Cdc34 (Ubn-Cdc34) in the indicated amount was incubated with 3 pmol of the 26 S proteasome purified from the wild type SEM1 or ⌬sem1 strain as in Fig. 2A. The mixture was subjected to nondenaturing PAGE and in-gel peptidase assay as in Fig.  1B. Note that the Ubn-Cdc34-bound 26 S proteasomes in both cases were gel-shifted to almost the same positions in comparison with the original ones (RP 2 CP). CBB, Coomassie Brilliant Blue. mutant did not show any growth defect; the ⌬sem1 mutant showed slight temperature sensitivity at 35°C, and the ⌬sem1 ⌬rpn10 strain showed clear high temperature-sensitive and cold-sensitive phenotypes. Thus, the synthetic effect between SEM1 and RPN10 is consistent with the above idea that Sem1 is a component of the 26 S proteasome.
The Human Sem1 Homologue, hDSS1, Is Capable of Binding with the Human 26 S Proteasome-It has been reported that the morphology and growth defect at 35°C in the mutant with deletion of DSS1 (⌬dss1) in yeast Schizosaccharomyces pombe are significantly rescued by expressing human DSS1 (hDSS1) (27). To determine whether hDSS1 complements the temperature-sensitive phenotype of the ⌬sem1 strain of S. cerevisiae, the hDSS1 gene was cloned and expressed under the TDH3 promoter in the ⌬sem1 strain. As shown in Fig. 9A, the expression of hDSS1 rescued the growth defect of the ⌬sem1 strain at the restrictive temperature. This result suggests that hDSS1 has the same function as that of S. cerevisiae Sem1 or a function overlapping that of S. cerevisiae Sem1.
The above-stated finding that hDSS1 is a functional homologue of Sem1 led us to speculate that hDSS1 is also a compo-nent of the human 26 S proteasome. To determine this, FLAG epitope-tagged hDSS1 or POH1 (a human homologue of Rpn11) (control) was transiently overexpressed in HEK293T cells, and the cell extracts were subjected to immunoprecipitation with anti-FLAG antibody-immobilized agarose beads. The resultant immunoprecipitates were analyzed by Western blotting with antibodies against the human 26 S proteasome, i.e. anti-MSS1 (human Rpt1), anti-p45 (human Rpt6), and anti-(human ␣5 subunit of the 20 S proteasome) antibodies (Fig. 9B). Both the CP subunit and the RP subunits MSS1 and p45 were coimmunoprecipitated with hDSS1 as well as with the 19 S RP subunit POH1. To determine whether the complexes immunoprecipitated using hDSS1 3FLAG and POH1 3FLAG as baits have the abilities to hydrolyze Suc-Leu-Leu-Val-Tyr-MCA, a peptidase assay for the proteasome was carried out in the presence of ATP (Fig. 9C). Approximately equal levels of peptidase activities were detected in both the cases of hDSS1 3FLAG and POH1 3FLAG . When immunoprecipitation was carried out in the absence of ATP, the level of peptidase activity decreased to one-third the level in the presence of ATP in either case (data not shown). These results strongly suggest that hDSS1 binds with or is incorporated into the human 26 S proteasome. DISCUSSION In this study, we found that Sem1 is a novel lid component of the 26 S proteasome of S. cerevisiae. We carried out affinity purification of the 26 S proteasome and lid and base subcomplexes from tagged strains of S. cerevisiae, and we found that the highly purified lid has an additional small molecular mass protein (Fig. 1C). Comparison of the intensities of protein bands on SDS-polyacrylamide gel showed that this protein seems stoichiometric to other RP subunits, and we found to our surprise that this protein is Sem1, as identified by amino acid sequencing. Subsequently, we carefully carried out several experiments to determine whether Sem1 is a component of the 26 S proteasome. First, it was found by GST pull-down assays that GST-Sem1 binds with the 26 S proteasome from the ⌬sem1 strain but not with the 26 S proteasome from the wild type strain (Fig. 2C). Second, analysis of the stability of the 26 S proteasome revealed instability of the lid lacking Sem1 under high salt conditions (Fig. 3). Third, immunoprecipitation using HA epitope-tagged Sem1 as bait showed that almost all 19 S RP subunits are able to be immunoprecipitated together FIG. 7. The 26 S proteasome lacking Sem1 has a defect in degradation of polyubiquitinated Sic1 PY in vitro. A, in vitro polyubiquitination of Sic1 PY , the PY motif-introduced Sic1, and its degradation by the 26 S proteasome. Recombinant T7-Sic1 PY -HAT and T7-Sic1-HAT (control) proteins were incubated with ubiquitinating enzymes, Uba1, Ubc4, and Rsp5. The resultant polyubiquitinated Sic1 PY was incubated with the affinity-purified 26 S proteasome (26 S). The levels of ubiquitination and degradation of Sic1 PY were detected by Western blotting with anti-T7 antibody. Note that the wild type Sic1 was neither ubiquitinated nor degraded by the 26 S proteasome. B, degradation of polyubiquitinated Sic1 PY by the 26 S proteasome from the wild type SEM1 or ⌬sem1 strain. The 26 S proteasome (200 nM) from each strain or control sample from no-tag strain was incubated with polyubiquitinated Sic1 PY (Ubn-Sic1 PY ) (400 nM) at 30°C. The reaction was terminated by the addition of SDS-loading buffer at the indicated times, and the reaction mixture was subjected to Western blotting with anti-T7 antibody. C, time dependence of the 26 S proteasome-mediated degradation of polyubiquitinated Sic1 PY . The signals on the gels in B were quantified by ImageGauge software (Fuji) and plotted as a function of incubation time. Note that the half-life (t1 ⁄2 ) in the case of the 26 S proteasome from SEM1 strain is ϳ1 min, whereas that from ⌬sem1 strain is more than 10 min. with Sem1 (Fig. 4). Fourth, it was found by proteasomal degradation assays that the 26 S proteasome lacking Sem1 has a defect in degradation of polyubiquitinated proteins both in vivo and in vitro (Figs. 5 and 7). Finally, genetic analysis revealed genetic interaction between SEM1 and RPN10 (Fig. 8). Based on the results described above, we conclude that Sem1 is a novel subunit of the 26 S proteasome, and we therefore propose that Sem1 be renamed Rpn15 to strengthen its identity as a component of the lid subcomplex. As the 26 S proteasome lacking Sem1 has the ATP-dependent peptidase activity (Fig. 2) and retains the ability to bind polyubiquitinated proteins (Fig.  6), it can be inferred that this Sem1-deficient 26 S proteasome has a defect in the release of polyubiquitin chains from substrates or the translocation of substrates into the CP. Little accumulation of the original protein Sic1 PY in assays of degradation of polyubiquitinated Sic1 PY (Fig. 7) prefers the former possibility, i.e. impairment of deubiquitination in the Sem1deficient 26 S proteasome. In addition, we showed that human DSS1 is a functional homologue of Sem1 (Fig. 9A) and that it interacts with the 26 S proteasome in mammalian HEK293T cells, as revealed by immunoprecipitation (Fig. 9B), suggesting that hDSS1 is also a subunit of the human 26 S proteasome.
It is well known that most of the proteasomal genes are contained in the same cluster that consists of functionally related genes with similar gene expression patterns (36). It is also known that Rpn4 functions as a transcriptional activator of genes encoding proteasomal subunits through its binding to an upstream activating sequence, 5Ј-GGTGGCAAA-3Ј, a proteasome-associated control element (PACE) (37). Recently, it has been reported that the sequence 5Ј-AGTGGCAA-3Ј is an alternative PACE and that both PACEs exist in the promoters of all 32 proteasomal genes and some proteasome-associated protein genes (9). Our search of the Saccharomyces Genome Data Base revealed that the latter PACE sequence is present upstream of the SEM1 gene (42-35 bp upstream), suggesting that expression of the SEM1 gene is regulated by Rpn4. The presence of the PACE in SEM1 provides support for our proposal that Sem1 is a component of the 26 S proteasome.
The SEM1/DSS1 genes are highly conserved from yeast to mammals. S. cerevisiae SEM1 was originally identified as a multicopy suppressor of the sec15-1 temperature-sensitive mu-tation (25). SEM1 is not an essential gene, and deletion of the SEM1 gene results in suppression of some exocyst mutations and triggers pseudohyphal growth in diploid cells (25). The fact that the 19 S RP is the only major Sem1-containing complex (Fig. 4) raises the possibility that the 26 S proteasome is involved in exocytosis. It is possible that a putative protein functioning in exocytosis is polyubiquitinated and degraded by the 26 S proteasome. The S. pombe DSS1 gene is also a nonessential gene and shows an elongated phenotype, caused by cell cycle delay at a restrictive temperature (27). It is well known that the 26 S proteasome play key roles in cell cycle progression (1,3). The hDSS1 gene was identified as one of three candidate genes at 7q21.3-q22.1 involved in an autosomal dominant form of SHFM disorder, a heterogeneous limb developmental disorder, and is expressed during limb development (26). Thus, it can be inferred that Sem1/DSS1 is involved in regulation of exocytosis, cell cycle, and/or development. In addition, the hDSS1 protein is capable of interacting with the product of the breast cancer susceptibility gene BRCA2 (27). The crystal structure of a complex composed of the C-terminal domain (ϳ800 residues) of BRCA2 and hDSS1 has been determined, and the binding site between BRCA2 and hDSS1 is highly conserved across evolution (38). In mammalian cells, BRCA2 binds to RAD51 and plays a role in DNA double-strand break repair (39). Recently, Ustilago maydis DSS1 has been reported to play an essential role in DNA repair, recombination, and genome maintenance (40). The phenotypes of U. maydis mutants with deletion of DSS1 reflect those of U. maydis mutants deficient in Brh2, the U. maydis BRCA2 homologue, and U. maydis RAD51 (40 -42). Thus, it is possible that DSS1 is involved in DNA repair and recombination through BRCA2.
Our findings that yeast Sem1 is a subunit of the 26 S proteasome and that hDSS1 is capable of binding with the human 26 S proteasome in mammalian cells, together with the fact that hDDS1 interacts with BRCA2, led us to assume that BRCA2 could interact with hDSS1 bound to the 26 S proteasome. Recently, a complex termed BRCC containing BRCA1, BRCA2, and RAD51 was identified as a ubiquitin E3 ligase complex that enhances cellular survival following DNA damage (43). It is likely that hDSS1 functions as an adaptor between the 26 S proteasome and the E3 ligase complex BRCC through FIG. 9. Human Sem1 homologue, hDSS1, interacts with the human 26 S proteasome. A, human DSS1 complements the temperature sensitivity of yeast ⌬sem1 cells. Low copy plasmids (pRS314) containing the SEM1 gene with its own promoter and the hDSS1 gene with the TDH3 promoter were transformed into YTS63 (⌬sem1) cells. The resultant transformant was cultured, and 10-fold diluted cells were spotted on SD-Trp plates and incubated at 25 or 37°C for 2 days. WT, wild type. B, co-immunoprecipitation of the endogenous human 26 S proteasome with overexpressed hDSS1. HEK293T cells were transfected with the pCI-neo (mock) (a), pCI-neo-POH1 (hRPN11) 3FLAG (b), or pCI-neo-hDSS1 3FLAG (c) plasmid, and immunoprecipitation (IP) was performed using anti-FLAG antibody-immobilized beads. The resultant immunoprecipitates were subjected to Western blotting with anti-MSS1 (hRpt1), anti-p45 (hRpt6), and anti-(h␣5) antibodies. C, peptidase assay for the proteasome. The immunoprecipitates in B were subjected to peptidase assay using Suc-Leu-Leu-Val-Tyr-MCA as a substrate.
its direct interaction with BRCA2. Further studies on interaction between the human 26 S proteasome and BRCA2, as well as structural elucidation of the human 26 S proteasome possibly containing hDSS1, are necessary to determine the role of hDSS1 in DNA repair and ubiquitin-dependent proteolysis.