Pleiotropic Defects Caused by Loss of the Proteasome-Interacting Factors Rad23 and Rpn10 of Saccharomyces cerevisiae

Corresponding author: Kiran Madura, Department of Biochemistry, Room 628, Robert Wood Johnson Medical School-UMDNJ, 675 Hoes Lane, Piscataway, NJ 08854., maduraki{at}umdnj.edu (E-mail)

Communicating editor: A. P. MITCHELL

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Rad23 is a member of a novel class of proteins that contain unprocessed ubiquitin-like (UbL) domains. We showed recently that a small fraction of Rad23 can form an interaction with the 26S proteasome. Similarly, a small fraction of Rpn10 is a component of the proteasome. Rpn10 can bind multiubiquitin chains in vitro, but genetic studies have not clarified its role in vivo. We report here that the loss of both Rad23 and Rpn10 results in pleiotropic defects that are not observed in either single mutant. rad23 rpn10 displays slow growth, cold sensitivity, and a pronounced G2/M phase delay, implicating overlapping roles for Rad23 and Rpn10. Although rad23 rpn10 displays similar sensitivity to DNA damage as a rad23 single mutant, deletion of RAD23 in rpn10 significantly increased sensitivity to canavanine, a phenotype associated with an rpn10 single mutant. A mutant Rad23 that is unable to bind the proteasome (^UbLrad23) does not suppress the canavanine or cold-sensitive defects of rad23 rpn10, demonstrating that Rad23/proteasome interaction is related to these effects. Finally, the accumulation of multiubiquitinated proteins and the stabilization of a specific proteolytic substrate in rad23 rpn10 suggest that proteasome function is altered.

THE ubiquitin/proteasome pathway has been implicated in a broad range of activities, primarily those involving protein degradation (HERSHKO 1991 ; HOCHSTRASSER 1996 ; PICKART 1997 ; VARSHAVSKY 1997 ). Ubiquitin is attached to proteolytic substrates by ubiquitin-conjugating (E2) enzymes and by substrate recognition (E3) factors that are also known as recognins/ubiquitin-protein ligases. Several lines of evidence indicate that the degradation of a substrate is preceded by the formation of a multiubiquitin (multi-Ub) chain (CHAU et al. 1989 ; PICKART 1997 ), a feature that may increase the affinity of a substrate for the proteasome (LAM et al. 1997 ; PIOTROWSKI et al. 1997 ). Ubiquitinated substrates are degraded by the 26S proteasome, a multicatalytic protease of well-defined composition and activity. The X-ray structure of the yeast 20S catalytic complex has been determined (GROLL et al. 1997 ), and the composition of the 19S regulatory particle was described recently (GLICKMAN et al. 1998B ).

In reconstituted experiments, long, unlinked and substrate-linked multi-Ub chains have an increased affinity for purified proteasomes (LAM et al. 1997 ; PIOTROWSKI et al. 1997 ). The search for a multi-Ub chain-binding protein in the proteasome led to the discovery of S5a in mammals (DEVERAUX et al. 1994 ), as well as its counterparts in Saccharomyces cerevisiae (Rpn10, also called Sun1 and Mcb1, VAN NOCKER et al. 1996B ) and Arabidopsis thaliana (Mbp1, DEVERAUX et al. 1995 ; VAN NOCKER et al. 1996A ). Surprisingly, deletion of RPN10 failed to cause a conspicuous defect, suggesting that the proteasome may contain other subunits that interact with multi-Ub chains (VAN NOCKER et al. 1996B ; FU et al. 1998 ). Furthermore, only a small portion of cellular Rpn10 is present in the proteasome at steady-state levels, suggesting that it may form transient interactions. Interestingly, GLICKMAN et al. 1998A reported that Rpn10 stabilizes a discrete subcomplex within the 19S regulatory particle, but the relevance of this activity to its ability to bind multi-Ub chains remains to be determined.

We reported recently that the nucleotide-excision repair protein Rad23 can interact with catalytically active 26S proteasome (SCHAUBER et al. 1998A ). A striking feature of Rad23 is the presence of an unusual N-terminal ubiquitin-like domain (UbL^R23) that mediates the interaction with the proteasome. This interaction may be important for efficient nucleotide-excision repair because removal of UbL^R23 causes sensitivity to DNA damage (WATKINS et al. 1993 ). As with Rpn10, only a small fraction of cellular Rad23 interacts with the proteasome. We therefore examined the possibility that Rpn10 and Rad23 might have overlapping functions. We report here that rad23 rpn10 (DLY140) has growth, cold-sensitive, and proteolytic defects that are not observed in either single mutant. DLY140 displays an apparent delay in the G2/M phase of the cell cycle, with cells containing 2N DNA and high levels of the mitotic cyclin Clb2. The canavanine sensitivity that was previously found in rpn10 was severely exacerbated in rad23 rpn10. In addition, we discovered increased levels of multiubiquitinated proteins in rad23 rpn10. Although our findings suggest an overlapping function for Rad23 and Rpn10, their activities are not redundant because the UV sensitivity of rad23 was not intensified in DLY140. Significantly, we found that the rad23 single mutant is sensitive to canavanine and is unable to degrade the proteasome substrate Ub-Pro-ß-galactosidase (Ub-Pro-ßgal). Taken together, these findings for the first time implicate a proteolytic function for Rad23. This conclusion is strengthened by the finding that ^UbLrad23, a mutant that is unable to interact with the proteasome, failed to complement the canavanine- and cold-sensitive phenotypes of rad23 rpn10.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Media, yeast strains, and plasmids:
Yeast transformations were performed using standard techniques (GUTHRIE and FINK 1991 ; GIETZ et al. 1992 ) to yield the strains described in Table 2. The RPN10 gene was amplified by polymerase chain reaction (PCR) with a 5' EcoRI and 3' KpnI restriction site using the following oligonucleotides: 5'-GCGAATTCATGGTATTGGAAGCTACAGTGTTAGTG-3' and 5'-GCGGTACCTATTTGTCTTGGTGTTGTTCAGGCTG-3'. The resulting DNA fragment and the plasmid pCS13 (Table 1) were digested with EcoRI and KpnI, ligated, and transformed into TOP10F' to yield pLC97. Similarly, the RAD23 gene was amplified by PCR with a 5' EcoRI and 3' KpnI restriction site using the following oligonucleotides: 5'-GCGGTACCTCAGTCGGCATGATCGCTGAA-3' and 5'-GCGAATTCATGACGAAGACCAAAGTAACAGAA-3'. The resulting DNA fragment and plasmid pKM1362 (MADURA and VARSHAVSKY 1994 ) were digested with KpnI and EcoRI, ligated, and transformed into TOP10F' to yield pDL120. pCS19 was constructed by in vivo recombination of pCEP10 (SCHAUBER et al. 1998A ) with SmaI-digested pUT11 (generously provided by F. Cross, Rockefeller University, New York) to generate a TRP1 derivative. DLY140 and DLY152 were constructed by in vivo recombination of EcoRI-digested pDG28 with MHY960 and DF5, respectively, and selected on synthetic medium lacking uracil.

Growth assays:
Yeast cultures were grown for ~15 hr in selective medium, and their densities were normalized to OD₆₀₀ ~1.0 to examine growth properties and OD₆₀₀ ~0.5 to estimate sensitivity to 1 µg/ml canavanine. Tenfold serial dilutions were prepared in sterile water, and 5 µl was plated on YPAD agar medium and incubated at 13°, 25°, and 30° to estimate growth properties. Aliquots were also spotted on synthetic complete (SC) medium and SC supplemented with 1 µg/ml canavanine (but lacking arginine) and incubated at 30°. The growth of the yeast strains was examined after 2–7 days on the basis of incubation temperature. The sensitivity to canavanine was determined after 5 days of growth at 30°.

Microscopy:
Yeast cultures were grown for ~15 hr at 30° in liquid culture or plated on YPAD medium for several days at 13°. Cells were collected and fixed in 20% ethanol for 20 min at 4°, sonicated for 5 min, and examined by microscopy. Cells were visualized with an Optiphot-2 light microscope (Nikon, Garden City, NY) using a x100 objective and photographed with a Nikon FX-35 camera. The fraction of budded and unbudded cells was determined with similar cultures that were not fixed in ethanol.

To examine cells by fluorescence-activated cell sorting (FACS), actively growing cells were fixed for 1 hr in 20% ethanol at 4°, washed in 50 mM Tris, pH 7.5, and treated with 1 mg/ml RNase for 2 hr at 37°. Cells were incubated at 55° for 1 hr in the presence of 0.04 mg/ml proteinase K, washed with PBS, and sonicated for 10 min. Cells were resuspended in 100 µg/ml propidium iodide, stored at 4°, and diluted fivefold prior to sorting.

Ultraviolet sensitivity assays:
Yeast cultures were grown overnight in synthetic medium and normalized to an OD₆₀₀ ~ 1.0. Appropriate dilutions were plated on YPAD plates and exposed to 254-nm UV light for 0, 10, 30, and 60 sec at 1.5 J/m²/sec, followed by incubation at 30° in the dark for 3 days. Colonies were counted and compared to an untreated control.

Protein degradation:
Bulk protein degradation assays were performed essentially as described by VAN NOCKER et al. 1996B . Yeast cultures were grown to exponential phase at 30° and incubated with ³⁵S-Translabel for 5 min, as described previously (BACHMAIR et al. 1986 ). The cells were pelleted, washed twice, and resuspended at 25° in chase buffer containing cycloheximide and excess unlabeled methionine and cysteine (BACHMAIR et al. 1986 ). Aliquots (5 µl) were withdrawn at 0, 10, 20, 30, 40, 50, and 60 min during the incubation in chase medium and were spotted in triplicate on filter disks presoaked in 50% trichloroacetic acid (TCA). Filters were dried, washed in cold 10% TCA for 10 min, and boiled in 10% TCA for 5 min. The filters were dried, transferred to scintillation vials, and quantitated. The radioactivity for each sample was normalized to the value obtained at 0 min.

Yeast strains expressing Ub-Pro-ßgal and Clb2-HA were propagated at 30° in raffinose-containing medium. The cells were transferred to medium containing 2% galactose to induce expression of Ub-Pro-ßgal and Clb2-HA. Actively growing cells were metabolically labeled for 5 min with ³⁵S-Translabel and chased in medium containing cycloheximide and excess unlabeled methionine and cysteine. Aliquots were removed at 0, 10, 30, and 60 min, combined with buffer A (50 mM HEPES, pH 7.5, 5 mM EDTA, 150 mM NaCl, 1% Triton X-100, BACHMAIR et al. 1986 ), and frozen immediately. The incorporation of ³⁵S into acid (TCA)-insoluble material was determined, and equal cpm were incubated with antibodies and examined as described previously (SCHAUBER et al. 1998B ). The stability of Ub-Pro-ßgal was determined by incubation of 2 x 10⁵ cpm of extract with anti-ßgal antibodies (Sigma, St. Louis), while Clb2-HA stability was determined by incubating 1 x 10⁶ cpm of extract with anti-HA antibodies (BabCo, Berkeley, CA). The immunoprecipitates were washed, resolved in an SDS-polyacrylamide, and exposed to X-ray film. Protein levels were quantitated by densitometry.

Immunoblotting:
For immunoblotting studies, cell extracts were prepared in lysis buffer (20 mM HEPES, pH 7.5, 100 mM potassium acetate, 5 mM Na · EDTA, 10% glycerol), and equal amounts of extract (1 mg total protein) were incubated with either 20 µl FLAG-agarose (Sigma) or 20 µl protein A-Sepharose (Repligen Co., Cambridge, MA) for a mock control. The immunoprecipitation reactions were adjusted to a final volume of 900 µl using buffer A and were incubated for 6–12 hr at 4°. The immunoprecipitates were washed three times with buffer A, resolved by SDS-polyacrylamide gel electrophoresis, transferred to 0.45 µm nitrocellulose (Bio-Rad, Richmond, CA), and blocked in 5% milk powder. The filter was incubated sequentially with antibodies against Rpt1, Rad23, and the FLAG epitope (Sigma). The immunoblots were developed with enhanced chemiluminescence (Renaissance ECL, New England Nuclear, Boston).

To visualize ubiquitin, 50 µg of total protein was resolved in a 12% SDS polyacrylamide gel and transferred to a 0.2-µm nitrocellulose filter (Bio-Rad). The filter was pretreated as described previously (SWERDLOW et al. 1986 ), blocked with 5% milk powder, and incubated with antibodies against ubiquitin (Sigma). The immunoblots were developed with enhanced chemiluminescence (Renaissance ECL, New England Nuclear).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Rad23 and Rpn10 are required for efficient growth:
The RAD23 gene was deleted by homologous recombination from an rpn10 strain (MHY960, acquired from M. Hochstrasser, University of Chicago) and its wild-type counterpart (DF5), resulting in strains DLY140 and DLY152, respectively (Table 2). Tenfold dilutions of exponential phase DLY140 and the congenic wild-type and single-mutant strains were plated on YPAD medium and incubated at 13°, 25°, 30°, and 37° for 2–7 days (Figure 1A). DLY140 showed extremely poor viability at 13°, and its growth was noticeably impaired at 25° and 30°. DLY140 displayed similar growth rates at 30° and 37°, demonstrating that it is not sensitive to high temperature (data not shown). In contrast, neither single mutant displayed a significant growth defect at any temperature, implicating an overlapping function for these proteasome-interacting factors.

View larger version (91K): In this window In a new window Download PPT slide   Figure 1. rad23 rpn10 displays slow growth and a cold-sensitive phenotype. (A) Yeast cultures were normalized to an optical density A600 ~1, and 10-fold dilutions were spotted on YPAD medium. Agar plates were incubated for 2–7 days at 30°, 25°, and 13°. The growth of rad23 and rpn10 was similar to that of the wild-type strain. In contrast, the growth of rad23 rpn10 was reduced at 30° and 25° and severely inhibited at 13°. (B) The growth of rad23 rpn10 was restored to wild-type levels after transformation with a plasmid encoding FLAG-Rpn10 expressed from PCUP1, or with a CEN-based plasmid expressing Rad23 from its own promoter. Significantly, a Rad23 mutant lacking the UbL domain (UbLrad23) did not suppress the cold-sensitive defect of rad23 rpn10 and only partially rescued the slow-growth phenotype, demonstrating that UbLR23/proteasome interaction has an important physiological effect. A vector control is shown in C. (The plates in B and C were incubated ~12 hr longer than those in A, accounting for the difference in growth.)

To confirm that the growth defects of DLY140 were caused by the loss of both Rad23 and Rpn10, we transformed the double mutant with plasmids encoding either FLAG-Rpn10 or single-copy Rad23 expressed from its own promoter. As expected, the unique phenotypes of DLY140 were completely rescued by plasmid-borne copies of either RPN10 or RAD23 (Figure 1B), but not by a vector control (Figure 1C). These studies demonstrate the requirement for either Rad23 or Rpn10 for suppressing the defects of DLY140.

The ubiquitin-like domain in Rad23 (UbL^R23) is required for its DNA repair activities (WATKINS et al. 1993 ). Our finding that UbL^R23 can form an interaction with the 26S proteasome (SCHAUBER et al. 1998A ) implicated a proteolytic function for Rad23 during DNA repair. We investigated whether Rad23/proteasome interaction was also required for the phenotypes associated with DLY140. rad23 rpn10 was transformed with a plasmid encoding ^UbLrad23, a mutant that lacks the UbL domain. Although ^UbLrad23 partially complemented the slow-growth phenotype of DLY140, it failed to suppress the cold sensitivity of DLY140 (Figure 1B), demonstrating that Rad23/proteasome interaction is important in the absence of Rpn10. These findings present a novel and important genetic link between Rad23 and the ubiquitin/proteasome pathway.

A role for Rad23 in cell-cycle progression:
To further characterize the growth defect of rad23 rpn10, we examined late exponential phase cells by microscopy. We found that a high fraction of DLY140 accumulated as large-budded cells in contrast to the wild-type or single-mutant strains (compare Figure 2D to Figure 2, A–C). The slow-growth and cold-sensitive phenotypes observed in DLY140 may be the result of the G2/M phase delay. Since the growth defect of DLY140 is intensified at 13°, we examined yeast cells that were incubated at 13° to determine if the cells arrested with a terminal defect. We observed profound morphological aberrations in rad23 rpn10 at 13° in addition to the accumulation of large-budded cells (Figure 2F). In contrast to the congenic wild-type strain (Figure 2E), DLY140 cells contained a large vacuole and numerous small vesicles. We also quantitated the number of cells that contained small and large buds in the four strains. These results (Figure 2G) corroborate the microscopic analysis (Figure 2, A–D). The ability of ^UbLrad23 to partially complement the slow-growth but not the cold-sensitive phenotype of DLY140 may reflect the complex growth defects of the double mutant. For instance, the large-budded phenotype that is observed at 30° may be distinct from the other morphological defects that are apparent only at low temperatures.

View larger version (63K): In this window In a new window Download PPT slide   Figure 2. rad23 rpn10 displays distinct morphological defects at 30° and 13°. Yeast cells were grown to late exponential phase at 30°, fixed in 20% ethanol, and visualized by microscopy. (A) Wild-type, (B) rad23, and (C) rpn10 cells displayed few large-budded cells, while (D) a high fraction of rad23 rpn10 contained large-budded cells. (E) Wild-type cells maintained normal morphology at 13°, but they were noticeably smaller than cells grown at 30°. (F) In contrast, the morphology of rad23 rpn10 cells was highly aberrant at 13°, and a large fraction was highly elongated and contained numerous small vesicles. (G) The percentage of unbudded cells and cells containing small and large buds in the strains described in A–D is shown.

To further characterize the G2/M phase delay, we stained yeast cells with propidium iodide and examined them by FACS. We found that ~70% of DLY140 and ~40% of wild-type and single-mutant cells contained 2N DNA (Figure 3A), consistent with the microscopic analysis (Figure 2, A–D). In agreement with the FACS analysis, we found a significant increase in the fraction of rad23 rpn10 cells that contained large buds, and a decrease in the proportion of unbudded cells (Figure 2G). We also examined the levels of Clb2-HA in rad23 rpn10, since the degradation of this mitotic cyclin occurs during the transition from mitosis to G1 (AMON et al. 1994 ). We expressed Clb2-HA in each of the four strains and measured stability in asynchronous cultures by pulse-chase methods. We found that the level of Clb2-HA was ~2.5-fold higher in DLY140 after 30 min in the chase buffer, relative to wild-type and single-deletion strains (Figure 3B). The accumulation of Clb2-HA may contribute to the delay in rad23 rpn10, although our data do not exclude the possibility that the stabilization of Clb2-HA is simply a consequence of the G2/M-phase delay in DLY140. Nonetheless, the accumulation of Clb2-HA in the double mutant provides a useful biochemical marker that reflects the delay in G2/M, which was predicted by the microscopic and FACS analyses. Further studies will be required to precisely define the point of delay during the cell cycle.

View larger version (31K): In this window In a new window Download PPT slide   Figure 3. An increased fraction of rad23 rpn10 cells contains 2N DNA and high levels of the mitotic cyclin Clb2-HA, consistent with a delay in G2/M phase. (A) Yeast cells were stained with propidium iodide and examined by FACS. Approximately 40% of wild-type and single-deletion cells contained 2N DNA, while ~70% of rad23 rpn10 accumulated with 2N DNA. (B) The stability of Clb2-HA was determined by pulse-chase analysis in asynchronously growing cultures. Clb2-HA accumulated to higher levels and was noticeably more stable in rad23 rpn10 compared to wild-type and single-deletion strains, consistent with the delayed progression through the G2/M phase.

Rad23 and Rpn10 form independent interactions with the 26S proteasome:
Plasmids encoding FLAG-Rad23 and FLAG-Rpn10 were transformed into rpn10 and rad23 cells, respectively. We prepared cell extracts from actively growing cells and incubated equal amounts of protein with FLAG-agarose beads. The beads were washed, and the bound proteins were separated in an SDS-polyacrylamide gel, transferred to nitrocellulose, and incubated with antibodies against Rpt1, a subunit in the 19S regulatory particle of the proteasome (GLICKMAN et al. 1998B ). We found that Rpt1 could be coprecipitated with both FLAG-Rad23 and FLAG-Rpn10, even in the absence of the other protein (Figure 4A), demonstrating that Rad23 and Rpn10 form independent interactions with the proteasome.

View larger version (38K): In this window In a new window Download PPT slide   Figure 4. Rad23 and Rpn10 form independent interactions with the proteasome. (A) Yeast extracts were prepared from rad23 and rpn10 strains expressing FLAG-Rpn10 and FLAG-Rad23, respectively. Equal amounts of protein were incubated with FLAG-agarose (lanes 2, 4, 6, 8, and 10) or protein A-Sepharose as a mock control (lanes 1, 3, 5, 7, and 9), and the bound proteins were examined by immunoblotting with antibodies against Rpt1. Lanes 1 and 2 represent a wild-type strain that did not express a FLAG-tagged protein. The amount of Rpt1 that coprecipitated with FLAG-Rad23 from wild-type (lane 4) and rpn10 strains (lane 8) was essentially the same. Similarly, FLAG-Rpn10 coprecipitated an equivalent amount of Rpt1 from wild-type (lane 6) and rad23 (lane 10) cells. (B) Rad23 and Rpn10 do not compete for binding to the proteasome. Yeast cells expressing galactose-inducible Rad23 (PGAL1::RAD23) and FLAG-Rpn10 were grown in raffinose-containing medium. After the addition of 2% galactose, aliquots were withdrawn at intervals (0, 30, 60, and 120 min) and equal amounts of extract were incubated with FLAG-agarose (FLAG-IP). The immunoprecipitated proteins were analyzed sequentially in an immunoblot with antibodies against Rpt1 and Rad23. Despite the rapid increase of Rad23 levels, after the addition of galactose (Extract), there was no appreciable decrease in the amount of Rpt1 that coprecipitated with FLAG-Rpn10 (FLAG-IP). The abundance of FLAG-Rpn10 was relatively constant (Extract). Significantly, Rad23 was also detected in the FLAG-Rpn10 immunoprecipitates, demonstrating that both Rad23 and Rpn10 can bind the same proteasome. The asterisk represents a cross-reaction against the immunoglobulin heavy chain. FLAG-Rpn10 was expressed at similar levels in raffinose and galactose medium (data not shown).

To examine these interactions further, we expressed FLAG-Rpn10 and galactose-inducible Rad23 in DLY140. Actively growing cells were transferred from raffinose- to galactose-containing medium to induce expression of Rad23. Aliquots were withdrawn at the times indicated, and protein extracts were prepared. Equal amounts of protein were applied to FLAG-agarose, and Rpn10-interacting proteins were examined by SDS-PAGE and immunoblotting. We found that the interaction between Rpn10 and the proteasome was not affected by increasing levels of Rad23 (Figure 4B), because the level of Rpt1 that coprecipitated was essentially unchanged. Significantly, Rad23 was also detected in the FLAG-Rpn10 immunoprecipitate, demonstrating that Rpn10 and Rad23 can bind the same proteasome. The immunoprecipitation reactions were performed without the addition of exogenous ATP, which are conditions that promote the dissociation of the 19S and 20S particles of the proteasome. We speculate, therefore, that Rpn10 and Rad23 occupy different binding sites on the same 19S particle.

The UV sensitivity of rad23 is unaffected in rad23 rpn10:
Rad23 enables Rad14 and RNA polymerase II transcription factor IIH (TFIIH) to assemble into the nucleotide-excision repair complex in vitro (GUZDER et al. 1995A , GUZDER et al. 1995B ). Rad23 can also be purified in a complex with Rad4 in yeast and XP-C in humans, both of which are essential components of the nucleotide-excision repair complex. Mutations in XP-C cause xeroderma pigmentosum, which is characterized by a predisposition to skin cancer, ataxia, and neurological impairment (CLEAVER and KRAEMER 1994 ; FRIEDBERG et al. 1995 ). The biochemical activity of Rad23/Rad4 is unknown (MASUTANI et al. 1994 ; GUZDER et al. 1995A ) even though the dimer can preferentially bind damaged DNA (GUZDER et al. 1998 ). We reported previously that Rad4 can be copurified through several chromatographic steps with GST-Rad23 and subunits of the 26S proteasome (SCHAUBER et al. 1998A ). On the basis of this finding, we proposed that Rad23 might play a role in delivering proteolytic substrates to the proteasome. It is not known, however, if Rad4 constitutes one of these substrates.

We investigated whether the DNA repair defect of rad23 was further exacerbated in rad23 rpn10 because of the overlapping roles of Rad23 and Rpn10. We measured the survival of DLY140 after different doses of UV light and found that its sensitivity was similar to that of rad23 (Figure 5). This finding is significant because it demonstrates that the functions of Rad23 and Rpn10 are not redundant, but overlapping. Similarly, BIGGINS et al. 1996 reported that Rad23 and Dsk2 (another UbL-containing protein) have overlapping functions. Significantly, the UV sensitivity of rad23 was not further aggravated in rad23 rpn10, and, similar to rpn10, the dsk2 single mutant does not display any conspicuous growth defects. We have recently discovered that the UbL domain in Dsk2 (UbL^DSK) can interact with the proteasome (data not shown), suggesting that other UbL-containing proteins might also have proteolytic functions that intersect with Rad23 and Rpn10.

View larger version (15K): In this window In a new window Download PPT slide   Figure 5. Rpn10 is not required for the DNA repair function of Rad23. Yeast cells were exposed to 254 nm UV light at a fluence of 1.5 J/m2/sec, and survival was determined. Wild-type () and rpn10 () cells displayed similar sensitivity. In contrast, rad23 rpn10 () and rad23 () were sensitive to UV light.

Degradation of most short-lived proteins is unaffected in rad23 rpn10:
We determined whether the phenotypes of rad23 rpn10 were caused by an underlying defect in the ubiquitin/proteasome system. To test this idea, the degradation for short-lived proteins was measured in the wild-type and rad23 rpn10 strains, as described previously (VAN NOCKER et al. 1996B ). Proteins were labeled with ³⁵S-Translabel for 5 min, followed by incubation at 25° in chase buffer containing excess unlabeled methionine and cysteine. Aliquots were removed at intervals during the chase, and the amount of ³⁵S-labeled protein that was degraded was determined (see MATERIALS AND METHODS). Interestingly, DLY140 and wild-type cells showed similar levels of degradation of bulk short-lived proteins (Figure 6). Furthermore, the rates of degradation were not affected significantly in the presence of canavanine (data not shown) despite the sensitivity of rad23 rpn10 to this amino acid analog (see below).

View larger version (16K): In this window In a new window Download PPT slide   Figure 6. The degradation of bulk short-lived proteins is not affected in rad23 rpn10. Yeast cells were grown to exponential phase at 30°, labeled with 35S-Translabel for 5 min, washed, and incubated at 25° in medium containing excess cold methionine and cysteine. Aliquots (5 µl) were withdrawn at 10-min intervals and treated as described in MATERIALS AND METHODS. The turnover of generally short-lived proteins was equivalent in the wild type () and rad23 rpn10 (). The values reflect measurements that were obtained in triplicate, and they are representative of two independent experiments.

The proteolytic defects of rpn10 are intensified by loss of Rad23:
The growth of many mutants in the ubiquitin/proteasome pathway is inhibited by canavanine (SEUFERT and JENTSCH 1990 ; VAN NOCKER et al. 1996B ; RAMOS et al. 1998 ), an amino acid analogue whose incorporation into proteins causes misfolding. Loss of Rpn10 was previously shown to cause moderate sensitivity to canavanine (VAN NOCKER et al. 1996B ). We show here that the loss of Rad23 also causes the cells to become sensitive to canavanine. However, loss of both Rad23 and Rpn10 resulted in extreme sensitivity to canavanine, and plating assays showed >100-fold reduced survival (Figure 7A). Significantly, the canavanine sensitivity of DLY140 was suppressed by either FLAG-Rpn10 or Rad23, but not by ^UbLrad23 (Figure 7B). Similarly, the cold sensitivity was not suppressed by ^UbLrad23 (Figure 1B), demonstrating the importance of Rad23/proteasome interaction. These findings suggest that the proteolytic defect in rad23 rpn10 may occur at the level of the proteasome.

View larger version (92K): In this window In a new window Download PPT slide   Figure 7. The canavanine sensitivity of rpn10 is exacerbated by the loss of RAD23. (A) Late exponential phase cultures were normalized to an optical density of A600 ~0.5. Tenfold dilutions were plated on synthetic medium lacking arginine and supplemented with 1 µg/ml canavanine, and they were incubated at 30° for 5 days. rad23 rpn10 displayed extremely poor growth on medium containing canavanine. In contrast, rad23 and rpn10 displayed moderate sensitivity to canavanine. The growth of these strains on medium lacking canavanine was similar to that observed in Figure 1A, although the longer incubation of these plates at 30° permitted rad23 rpn10 to form a patch of growth. (B) The sensitivity of rad23 rpn10 to canavanine was alleviated by expression of FLAG-Rpn10 or Rad23, but not by UbLrad23, confirming the requirement for Rad23/proteasome interaction. A vector control is shown in C.

To investigate the potential proteolytic role of Rad23, we examined the levels of multiubiquitinated proteins in wild-type, rad23, rpn10, and rad23 rpn10 strains. We resolved 50 µg of total protein from each strain by SDS-PAGE and incubated an immunoblot with antiubiquitin antibodies. We detected higher levels of multiubiquitinated proteins in DLY140, relative to the wild-type and single-deletion strains, consistent with the idea that Rad23 and Rpn10 provide overlapping proteolytic functions (Figure 8A).

View larger version (65K): In this window In a new window Download PPT slide   Figure 8. Accumulation of multiubiquitinated proteins and stabilization of Ub-Pro-ßgal in rad23 rpn10. (A) The levels of multiubiquitinated proteins were examined in extracts prepared from wild-type (lanes 1 and 2), rad23 rpn10 (lanes 3 and 4), rpn10 (lanes 5 and 6), and rad23 (lanes 7 and 8). Multiubiquitinated proteins were present at higher levels in rad23 rpn10 than in wild-type or single-mutant strains. The levels of multiubiquitinated proteins increased slightly in rad23 rpn10 in the presence of 3 µg/ml canavanine (even-numbered lanes). (B) The stability of Ub-Pro-ßgal was measured by pulse-chase analysis in wild type, rad23, rpn10, and rad23 rpn10. Actively growing yeast cells were incubated with 35S-Translabel, and ß-gal was immunoprecipitated from equal cpm of extract. Ub-Pro-ßgal was noticeably stabilized in the double deletion relative to rpn10 and wild-type strains although there was also significant stabilization in rad23. The asterisk represents a 90-kD degradation product. Significantly, the 90-kD fragment was detected neither in rad23 rpn10 nor rad23.

The loss of Rpn10 caused stabilization of Ub-Pro-ßgal (VAN NOCKER et al. 1996B ), a well-characterized substrate of the ubiquitin/proteasome pathway (JOHNSON et al. 1992 , JOHNSON et al. 1995 ). Ub-Pro-ßgal was also stablilized in the rad23 single mutant, providing compelling evidence of a proteolytic role for Rad23. Curiously, however, despite the stabilization of Ub-Pro-ßgal, its levels were significantly lower in rad23 than in the other strains. As expected, Ub-Pro-ßgal was completely stabilized in DLY140 and accumulated to high levels (Figure 8B). We also detected a distinct 90-kD degradation product in rpn10 and the wild-type strain. The release of this proteolytic fragment, which is indicative of the degradation of Ub-Pro-ßgal (BACHMAIR et al. 1986 ), was detected in neither rad23 nor rad23 rpn10.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The N-terminal, ubiquitin-like domain in Rad23 is required for efficient nucleotide-excision repair, and a regulatory function has been proposed for this motif (WATKINS et al. 1993 ). We discovered that UbL^R23 forms a stable interaction with catalytically active proteasomes (SCHAUBER et al. 1998A ), and recent studies indicate that this interaction occurs with the 19S regulatory particle (data not shown). These findings have led to the hypothesis that Rad23 might perform a role at the proteasome that is relevant to its DNA repair functions.

Rpn10 is a component of the 19S regulatory particle of the 26S proteasome (VAN NOCKER et al. 1996B ; GLICKMAN et al. 1998B ). Rpn10 and its mammalian counterpart (S5a) were isolated by their ability to interact with multiubiquitin chains, suggesting that Rpn10/S5a could play a central role in proteasome-mediated degradation. Surprisingly, Rpn10 is not required for cellular viability since a deletion of RPN10 caused only subtle defects, including stabilization of an engineered substrate, moderate sensitivity to an amino acid analog, and instability of the 19S particle under high-salt conditions (GLICKMAN et al. 1998A ). Similar to Rad23, only ~5% of Rpn10 is associated with the proteasome at steady-state levels, suggesting that a similar mechanism might regulate their interactions with the proteasome.

We report the discovery of a novel genetic interaction between the proteasome-interacting factors Rad23 and Rpn10. The simultaneous deletion of both genes caused pleiotropic defects, including slow growth, cold sensitivity, G2/M phase delay, increased sensitivity to the amino acid analog canavanine, and the accumulation of multiubiquitinated proteins. The suppression of the sensitivity of rad23 rpn10 to canavanine and cold temperature required UbL^R23, indicating a need for Rad23/proteasome interaction. The sensitivity of rad23 rpn10 to canavanine and the accumulation of multiubiquitinated proteins in this mutant indicate that the defects are caused, at least in part, by impairment of the function of the ubiquitin/proteasome pathway.

The elimination of canavanyl proteins requires the activities of the ubiquitin/proteasome pathway (SEUFERT and JENTSCH 1990 ). Nonetheless, some mutants of the ubiquitin pathway are sensitive to canavanine but do not display an obvious biochemical defect in proteolysis. We note that although rad23 rpn10 is highly sensitive to growth on canavanine-containing medium, its ability to degrade bulk short-lived proteins is similar to that of the wild-type counterpart. It is therefore possible that the sensitivity of rad23 rpn10 to canavanine is not related to impaired proteolysis. However, the significant accumulation of multiubiquitinated proteins in rad23 rpn10 is consistent with a proteolytic defect. We suggest, therefore, that the cellular sensitivity to canavanyl proteins may be a much more acute sensor of a proteolytic defect than the measurement of bulk protein turnover.

The sensitivity to canavanine was less severe in rpn10 (VAN NOCKER et al. 1996B ) than in rad23 (Figure 7A). Significantly, the double mutant was ~100-fold more sensitive to canavanine than either single mutant. Taken together, the defects of rad23 rpn10, which include canavanine sensitivity, accumulation of multiubiquitinated proteins, and stabilization of Ub-Pro-ßgal are consistent with altered proteolytic function. We note that the stabilization of Ub-Pro-ßgal in rad23 and the sensitivity of this mutant to canavanine have revealed an Rpn10-independent proteolytic function for Rad23. The accumulation of Clb2-HA in rad23 rpn10 provides a plausible explanation for the delay in the G2/M phase of the cell cycle (Figure 3B). However, our results do not resolve whether Clb2 stabilization is caused directly by the loss of both Rad23 and Rpn10, or by an indirect effect resulting from the delayed progression through G2/M phase.

To determine if the functional overlap between Rad23 and Rpn10 was of a specific nature, we examined the effect of deleting RAD23 from a strain lacking another proteasome subunit, Son1/Ufd5. FUJIMURO et al. 1998 showed that Son1 can be isolated in association with the proteasome. A proteolytic function for Son1 was also revealed in studies by JOHNSON et al. 1995 , who found that it is a component of the ubiquitin fusion degradation (UFD)-targeting pathway. Unlike rad23 rpn10, we found that rad23 son1 failed to display any synthetic defects. Similarly, deletion of RAD23 from ufd2 (rad23 ufd2) had no adverse effects. Ufd2 was recently reported to encode a novel factor that can bind and extend short multi-Ub chains (KOEGL et al. 1999 ). Ufd2 is also required for the UFD-targeting system (JOHNSON et al. 1995 ). These findings suggest that the genetic link between Rad23 and Rpn10 is specific, and is not a general consequence of mutating two unrelated proteolytic factors.

Several proteasome mutants, including cim3-1, cim5-1 (GHISLAIN et al. 1993 ), and nin1-1 (KOMINAMI et al. 1995 ), display a G2/M phase cell-cycle defect. Significantly, RPN10 was previously isolated as a high-copy suppressor of both nin1-1 and nin1, implicating Rpn10 in progression through mitosis (KOMINAMI et al. 1997 ). BIGGINS et al. 1996 showed that loss of both Rad23 and Dsk2 caused cells to arrest growth at 37° with 2N DNA content because of a failure to duplicate the spindle pole body. The previous findings and our studies described here collectively suggest that Rad23 may also play an important role in cell-cycle progression, although its activity is expected to be redundant with other factors, including Rpn10 and Dsk2.

On the basis of our findings, we propose that Rad23 could augment either of the two known activities of Rpn10: the recognition of multiubiquitinated substrates by the proteasome (VAN NOCKER et al. 1996B ) or the stabilization of the 19S regulatory particle of the proteasome (GLICKMAN et al. 1998A ). Alternatively, the possibility that Rad23 might mediate a novel mechanism for translocating substrates to the proteasome is suggested by the copurification of the DNA repair protein Rad4 with Rad23 and the proteasome (SCHAUBER et al. 1998A ). It is not known, however, if Rad4 constitutes one of the Rad23-specific substrates. An underlying assumption of this idea is that proteolytic substrates could be targeted to the proteasome by either Rad23 or Rpn10. The translocation of these substrates or their recognition by the proteasome would fail in strains lacking both Rad23 and Rpn10, leading to the pleiotropic defects observed in rad23 rpn10. The stabilization of Ub-Pro-ßgal (Figure 8B) in rad23 rpn10 is consistent with a requirement for both Rad23 and Rpn10 in the degradation of specific proteasomal substrates. However, because the degradation of bulk short-lived proteins is unaffected in rad23 rpn10 (Figure 6), we believe that only a subset of cellular substrates is targeted for degradation by Rad23 and Rpn10.

Recent studies have shown that several proteins contain significant sequence and structural similarity to ubiquitin (HOCHSTRASSER 1998 ). Some of these UbL sequences are processed and post-translationally ligated to other cellular proteins in a reaction that is identical to ubiquitin conjugation (JOHNSON and BLOBEL 1997 ; JOHNSON et al. 1997 ). However, the significance of these modifications is not known. In contrast, the UbLs present in a distinct set of proteins that includes Rad23 and Dsk2 are not processed and ligated to other proteins (WATKINS et al. 1993 ; BIGGINS et al. 1996 ). It was shown previously that the rad23 dsk2 double mutant has a temperature-sensitive growth defect that is caused by a failure to duplicate the spindle pole body (BIGGINS et al. 1996 ). This result implicates an overlapping function for Rad23 and Dsk2 that is similar to what we have observed in rad23 rpn10. We discovered that the UbL domain in Dsk2, as well as a UbL present in an uncharacterized open reading frame (YOL111c), can interact with the proteasome (data not shown). On the basis of these findings, we speculate that Rpn10 and UbL-containing proteins might promote the degradation of distinct subsets of proteins through regulated interaction with the proteasome. The degradation of a proteolytic target in trans has been described previously with an engineered substrate (JOHNSON et al. 1990 ). A similar role is performed by antizyme, which facilitates the translocation of ornithine decarboxylase to the proteasome for degradation (MURAKAMI et al. 1992A , MURAKAMI et al. 1992B ).

The requirement for the UbL of Rad23 for suppressing the phenotypes of rad23 rpn10 raises several interesting points. It was shown previously that a multi-Ub-chain-binding domain that is present in Rpn10 is dispensible for activity, although it is required for forming a high-affinity interaction with multi-Ub chains in vitro (FU et al. 1998 ). Since the proteolytic role of UbL^R23 was revealed only in a strain lacking Rpn10, it will be of interest to determine if the multi-Ub chain-binding domain in Rpn10 plays any role in suppressing the phenotypes of rad23 rpn10. Additionally, Rad23, Dsk2, and Rpn10 have been linked functionally using biochemical and genetic strategies. It will therefore be important to determine if Dsk2/Rpn10 also displays a genetic relationship that is analogous to Rad23/Rpn10 and Rad23/Dsk2 (BIGGINS et al. 1996 ). A detailed study of the biochemical functions of these proteins, and their role during the cell cycle, can provide insights into the significance of proteolysis during the cell cycle in yeast and mammalian cells.

	ACKNOWLEDGMENTS

We thank T. Kinzy (UMDNJ), M. Hochstrasser (University of Chicago), F. Cross (Rockefeller University), and W. Seufert (University of Heidelberg, Germany) for plasmids, strains, and reagents. We thank T. Dekas (Rutgers University/UMDNJ) for assistance with FACS analysis and members of the laboratory for their critical review of this manuscript. The studies described here were supported by a grant to K.M. from the National Institutes of Health (GM52058). D.L. was supported by a predoctoral fellowship from the American Heart Association (9810001T).

Manuscript received February 8, 1999; Accepted for publication June 1, 1999.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

AMON, A., S. IRNIGER, and K. NASMYTH, 1994  Closing the cell cycle circle in yeast: G2 cyclin proteolysis initiated at mitosis persists until the activation of G1 cyclins in the next cycle. Cell 77:1037-1050[Medline].

BACHMAIR, A., D. FINLEY, and A. VARSHAVSKY, 1986  In vivo half-life of a protein is a function of its amino-terminal residue. Science 234:179-186[Abstract/Free Full Text].

BARTEL, B., I. WUNNING, and A. VARSHAVSKY, 1990  The recognition component of the N-end rule pathway. EMBO J. 9:3179-3189[Medline].

BIGGINS, S., I. IVANOVSKA, and R. D. ROSE, 1996  Yeast ubiquitin-like genes are involved in duplication of the microtubule organizing center. J. Cell Biol. 133:1331-1346[Abstract/Free Full Text].

CHAU, V., J. W. TOBIAS, A. BACHMAIR, D. MARRIOTT, and D. ECKER et al., 1989  A multi-ubiquitin chain is confined to specific lysine in a targeted short-lived protein. Science 243:1576-1583[Abstract/Free Full Text].

CLEAVER, J. E., and K. H. KRAEMER, 1994 Xeroderma pigmentosum, pp. 2949–2971 in The Metabolic Basis of Inherited Disease, edited by C. R. SCRIVER, A. L. BEAUDET, W. S. SLY and D. VALLE. McGraw-Hill, New York.

DEVERAUX, Q., V. USTRELL, C. PICKART, and M. RECHSTEINER, 1994  A 26S protease subunit that binds ubiquitin conjugates. J. Biol. Chem. 269:7059-7061[Abstract/Free Full Text].

DEVERAUX, Q., S. VAN NOCKER, D. MAHAFFEY, R. VIERSTRA, and M. RECHSTEINER, 1995  Inhibition of ubiquitin-mediated proteolysis by the Arabidopsis 26S protease subunit S5a. J. Biol. Chem. 270:29660-29663[Abstract/Free Full Text].

FRIEDBERG, E. C., G. C. WAKJER and W. SIEDE, 1995 DNA Repair and Mutagenesis. American Society for Microbiology, Washington, DC.

FU, H., S. SADIS, D. M. RUBIN, M. GLICKMAN, and S. VAN NOCKER et al., 1998  Multiubiquitin chain binding and protein degradation are mediated by distinct domains within the 26S proteasome subunit Mcb1. J. Biol. Chem. 273:1970-1981[Abstract/Free Full Text].

FUJIMURO, M., K. TANAKA, H. YOKOSAWA, and A. TOH-E, 1998  Son1p is a component of the 26S proteasome of the yeast Saccharomyces cerevisiae.. FEBS Lett. 423:149-154[Medline].

GHISLAIN, M., A. UDVARDY, and C. MANN, 1993  S. cerevisiae 26S protease mutants arrest cell division in G2/metaphase. Nature 366:358-361[Medline].

GIETZ, R. D. and A. SUGINO, 1988  New yeast-Escherichia coli shuttle vectors constructed with in vitro mutagenized yeast genes lacking six-base pair restriction sites. Gene 74:527-534[Medline].

GIETZ, D., A. ST. JOHN, R. A. WOODS, and R. H. SCHIESTL, 1992  Improved method for high efficiency transformation of intact yeast cells. Nucleic Acids Res. 20:1425-1434[Free Full Text].

GLICKMAN, M. H., D. M. RUBIN, O. COUX, I. WEFES, and G. PFEIFER et al., 1998a  A subcomplex of the proteasome regulatory particle required for ubiquitin-conjugate degradation and related to the COP9-signalosome and eIF3. Cell 94:615-624[Medline].

GLICKMAN, M. H., D. M. RUBIN, V. A. FRIED, and D. FINLEY, 1998b  The regulatory particle of the Saccharomyces cerevisiae proteasome. Mol. Cell. Biol. 18:3149-3162[Abstract/Free Full Text].

GROLL, M., L. DITZEL, J. LOWE, D. STOCK, and M. BOCHTLER et al., 1997  Structure of 20S proteasome from yeast at 2.4 A resolution. Nature 386:463-471[Medline].

GUTHRIE, C., and G. R. FINK, 1991 Guide to Yeast Genetics and Molecular Biology. Academy Press, New York.

GUZDER, S. M., V. BAILLY, P. SUNG, L. PRAKASH, and S. PRAKASH, 1995a  Yeast DNA repair protein RAD23 promotes complex formation between transcription factor TFIIH and DNA damage recognition factor RAD14.. J. Biol. Chem. 270:8385-8388[Abstract/Free Full Text].

GUZDER, S. N., Y. HABRAKEN, P. SUNG, L. PRAKASH, and S. PRAKASH, 1995b  Reconstitution of yeast nucleotide excision repair with purified Rad proteins, replication protein A, and transcription factor TFIIH. J. Biol. Chem. 270:12973-12976[Abstract/Free Full Text].

GUZDER, S. N., P. SUNG, L. PRAKASH, and S. PRAKASH, 1998  Affinity of yeast nucleotide excision repair factor 2, consisting of the Rad4 and Rad23 proteins, for ultraviolet damaged DNA. J. Biol. Chem. 273:31541-31546[Abstract/Free Full Text].

HERSHKO, A., 1991  The ubiquitin pathway for protein degradation. Trends Biochem. Sci. 16:265-268[Medline].

HOCHSTRASSER, M., 1996  Ubiquitin-dependent protein degradation. Annu. Rev. Genet. 30:405-439[Medline].

HOCHSTRASSER, M., 1998  There's the rub: a novel ubiquitin-like modification linked to cell cycle regulation. Genes Dev. 12:901-907[Free Full Text].

JOHNSON, E. S. and G. BLOBEL, 1997  Ubc9p is the conjugating enzyme for the ubiquitin-like protein Smt3p. J. Biol. Chem. 272:26799-26802[Abstract/Free Full Text].

JOHNSON, E. S., D. K. GONDA, and A. VARSHAVSKY, 1990  cis-trans recognition and subunit-specific degradation of short-lived proteins. Nature 346:287-291[Medline].

JOHNSON, E. S., B. BARTEL, W. SEUFERT, and A. VARSHAVSKY, 1992  Ubiquitin as a degradation signal. EMBO J. 11:497-505[Medline].

JOHNSON, E. S., P. C. M. MA, I. M. OTA, and A. VARSHAVSKY, 1995  A proteolytic pathway that recognizes ubiquitin as a degradation signal. J. Biol. Chem. 270:17442-17456[Abstract/Free Full Text].

JOHNSON, E. S., I. SCHWIENHORST, R. J. DOHMEN, and G. BLOBEL, 1997  The ubiquitin-like protein Smt3p is activated for conjugation to other proteins by an Aos1p/Uba2p heterodimer. EMBO J. 16:5509-5519[Medline].

KOEGL, M., T. HOPPE, S. SCHLENKER, H. D. ULRICH, and T. U. MAYER et al., 1999  A novel ubiquitination factor, E4, is involved in multiubiquitin chain assembly. Cell 96:635-644[Medline].

KOMINAMI, K., G. N. DEMARTINO, C. R. MOOMAW, C. A. SLAUGHTER, and N. SHIMBARA et al., 1995  Nin1p, a regulatory subunit of the 26S proteasome, is necessary for activation of Cdc28p kinase of Saccharomyces cerevisiae.. EMBO J. 14:3105-3115[Medline].

KOMINAMI, K., N. OKURA, M. KAWAMURA, G. N. DEMARTINO, and C. A. SLAUGHTER et al., 1997  Yeast counterparts of subunits S5a and p58 (S3) of the human 26S proteasome are encoded by two multicopy suppressors of nin1-1.. Mol. Biol. Cell 8:171-187[Abstract].

LAM, Y. A., W. XU, G. N. DEMARTINO, and R. E. COHEN, 1997  Editing of ubiquitin conjugates by an isopeptidase in the 26S proteasome. Nature 385:737-740[Medline].

MADURA, K. and A. VARSHAVSKY, 1994  Degradation of G by the N-end rule pathway. Science 265:1454-1458[Abstract/Free Full Text].

MASUTANI, C., K. SUGASAWA, J. YANAGISAWA, T. SONOYAMA, and M. UI et al., 1994  Purification and cloning of a nucleotide excision repair complex involving the xeroderma pigmentosum group C protein and a human homologue of yeast RAD23. EMBO J. 13:1831-1843[Medline].

MURAKAMI, Y., D. TANAKA, Y. MATSUFUJI, and S. HAYASHI, 1992a  Antizyme, a protein induced by polyamines, accelerates the degradation of ornithine decarboxylase in Chinese-hamster ovary-cell extracts. Biochem. J. 283:661-664.

MURAKAMI, Y., S. MATSUFUJI, T. KAMEJI, S. HAYASHI, and K. IGARASHI et al., 1992b  Ornithine decarboxylase is degraded by the 26S proteasome without ubiquitination. Nature 360:597-599[Medline].

PICKART, C. M., 1997  Targeting of substrates to the 26S proteasome. FASEB J. 11:1055-1066[Abstract].

PIOTROWSKI, J., R. BEAL, L. HOFFMAN, D. D. WILKINSON, and R. E. COHEN et al., 1997  Inhibition of the 26S proteasome by polyubiquitin chains synthesized to have defined lengths. J. Biol. Chem. 272:23712-23721[Abstract/Free Full Text].

RAMOS, P. C., J. HOCKENDORFF, E. S. JOHNSON, A. VARSHAVSKY, and R. J. DOHMEN, 1998  Ump1p is required for proper maturation of the 20S proteasome and becomes its substrate upon completion of the assembly. Cell 92:489-499[Medline].

SCHAUBER, C., L. CHEN, P. TONGAONKAR, I. VEGA, and D. LAMBERTSON et al., 1998a  Rad23 links DNA repair to the ubiquitin/proteasome pathway. Nature 391:715-718[Medline].

SCHAUBER, C., L. CHEN, P. TONGAONKAR, I. VEGA, and K. MADURA, 1998b  Sequence elements that contribute to the degradation of yeast G. Genes Cells 3:307-319[Abstract].

SEUFERT, W. and S. JENTSCH, 1990  Ubiquitin-conjugating enzymes UBC4 and UBC5 mediate selective degradation of short-lived and abnormal proteins. EMBO J. 9:543-550[Medline].

SEUFERT, W., B. FUTCHER, and S. JENTSCH, 1995  Role of a ubiquitin-conjugating enzyme in degradation of S- and M-phase cyclins. Nature 373:78-81[Medline].

SWERDLOW, P. S., D. FINLEY, and A. VARSHAVSKY, 1986  Enhancement of immunoblot sensitivity by heating of hydrated filters. Anal. Biochem. 156:147-153[Medline].

VAN NOCKER, S., Q. DEVERAUX, M. RECHSTEINER, and R. D. VIERSTRA, 1996a  Arabidopsis MBP1 gene encodes a conserved ubiquitin recognition component of the 26S proteasome. Proc. Natl. Acad. Sci. USA 93:856-860[Abstract/Free Full Text].

VAN NOCKER, S., S. SADIS, D. M. RUBIN, M. GLICKMAN, and H. FU et al., 1996b  The multiubiquitin-chain-binding protein Mcb1 is a component of the 26S proteasome in Saccharomyces cerevisiae and plays a nonessential, substrate-specific role in protein turnover. Mol. Cell. Biol. 16:6020-6028[Abstract].

VARSHAVSKY, A., 1997  The ubiquitin system. Trends Biochem. Sci. 22:383-387[Medline].

WATKINS, J. F., P. SUNG, L. PRAKASH, and S. PRAKASH, 1993  The Saccharomyces cerevisiae DNA repair gene RAD23 encodes a nuclear protein containing a ubiquitin-like domain required for biological function. Mol. Cell. Biol. 13:7757-7765[Abstract/Free Full Text].

This article has been cited by other articles:

G. Gabriely, R. Kama, R. Gelin-Licht, and J. E. Gerst Different Domains of the UBL-UBA Ubiquitin Receptor, Ddi1/Vsm1, Are Involved in Its Multiple Cellular Roles Mol. Biol. Cell, September 1, 2008; 19(9): 3625 - 3637.