MCP Papers in Press. Published on November 10, 2005 as Manuscript M500303-MCP200
An Integrated Mass Spectrometry-Based Proteomics Approach -QTAX to Decipher the 26S Proteasome Interacting Network
Cortnie Guerrero1, Christian Tagwerker2, 3, Peter Kaiser2, Lan Huang1*
1
Departments of Physiology & Biophysics and Developmental & Cell Biology, University of
California, Irvine, CA 92697 2
Department of Biological Chemistry, University of California, Irvine, CA 92697
3
Institute of Biochemistry, University of Innsbruck, Austria
*Correspondence should be addressed to Dr. Lan Huang (
[email protected]) Medical Science I, D233 Department of Physiology & Biophysics Department of Developmental & Cell Biology University of California, Irvine Irvine, CA92697-4560
Running title: QTAX for Deciphering the 26S Proteasome Interacting Network
1 Copyright 2005 by The American Society for Biochemistry and Molecular Biology, Inc.
SUMMARY We have developed an integrated proteomics approach to decipher in vivo protein-protein interactions and applied this strategy to globally map the 26S proteasome interaction network in yeast. We have termed this approach QTAX for quantitative analysis of tandem affinity purified in vivo cross-linked (X) protein complexes. For this work, in vivo formaldehyde cross-linking was employed to freeze both stable and transient interactions occurring in intact cells prior to lysis. To isolate cross-linked protein complexes with high purification efficiency under fully denaturing conditions, a new tandem affinity tag consisting of a hexahistidine sequence and an in vivo biotinylation signal was adopted for affinity-based purification. Tandem-affinity purification after in vivo cross-linking was combined with tandem mass spectrometry coupled with a quantitative SILAC strategy to carry out unambiguous protein identification and quantification of specific protein interactions. Using this method, we have captured and identified the full composition of yeast 26S proteasome complex as well as the two known ubiquitin receptors, Rad23 and Dsk2. Quantitative mass spectrometry analysis allowed us to distinguish specific proteasome interacting proteins (PIPs) from background proteins and led to the identification of a total of 64 potential PIPs, of which 42 are novel interactions. Among the 64 putative specific PIPs, there are ubiquitin pathway components, ubiquitinated substrates, chaperones, and transcription and translation regulators, demonstrating the efficacy of the developed approach in capturing in vivo protein interactions. The method offers an advanced technical approach to elucidate the proteasome’s dynamic protein interaction networks, and can find a wide range of applications in the studies of other macromolecular protein complex interaction networks.
2
ABBREVIATIONS PIP: proteasome interacting partner HB: histidine and biotin tag SILAC: stable isotope labeling of amino acids in cell culture OD: optical density LC MS/MS: liquid chromatography tandem mass spectrometry HRP: horse radish peroxidase CID: collision induced dissociation SCX: strong cation exchange m/z: mass/charge TAP: tandem affinity purification TEV: tobacco etch virus TEB: TEV elution buffer TOFMS: time of flight mass spectrometry Co-IP: co-immunoprecipitation wt: wild type, untagged strain L: light (12C6-Arg) H: heavy (13C6-Arg) L/H: relative abundance ratio of light (12C6-Arg) to heavy (13C6-Arg)UBL:Ubiquitin-like-domain UBA: Ubiquitin-associated domain QTAX: quantitative analysis of tandem affinity purified in vivo cross-linked (X) protein complexes
3
INTRODUCTION Protein ubiquitination and proteasome-mediated protein degradation are essential to the regulation of many important biological processes such as cell cycle progression, apoptosis, DNA repair, chromosome maintenance, transcriptional activation, metabolism, immune response, signal transduction, stress response, and cell differentiation, etc(1-3). Disruption of normal ubiquitin-proteasome degradation pathways has been implicated in a wide range of human disease. Therefore, the proteasome and ubiquitination components represent a class of “drugable” targets for pharmaceutical intervention. The 26S proteasome, the macromolecular degradation machine of the ubiquitin-proteasome pathway, consists of a self-compartmentalized 20S protease core that is capped at one or both ends by the 19S regulatory particle, or CAP (also known as PA700 in animal cells) (3-6). The 20S core particle, responsible for various proteolytic activities, is made up of two copies each of seven different α and seven different β subunits arranged into four stacked rings (α7β7β7α7). The two outer α rings are catalytically inactive, whereas three of the seven inner β subunits are catalytically active. Although the 20S core can degrade fully unfolded proteins in the absence of ATP and ubiquitin, protein degradation by 26S proteasomes is strictly ATP dependent and, in almost all cases, requires the presence of a ubiquitin chain on the substrate protein (7). The19S regulatory complex is composed of at least 18 different subunits, which are assembled into two main subcomplexes — a base that contains six ATPases plus two non-ATPase subunits and abuts the proteasome α-ring, and a lid subcomplex containing ten non-ATPase subunits that sits on top of the base (4).
Additional proteasome subunits continue to be identified owing to the
development of new protein purification and identification techniques (7-10), yet how they fit into the assembly of 19S has not been determined.
4
There are two major steps involved in the ubiquitin-proteasome-dependent degradation pathway: 1) enzymatic polyubiquitination of protein substrates; and 2) transportation to and recognition by the 26S proteasome prior to degradation. A cascade of enzymes, including the ubiquitin-activating enzyme (E1), ubiquitin-conjugating enzymes (E2s), and ubiquitin ligases (E3s), are responsible for ubiquitination of a target protein (11).
Once a protein is
polyubiquitinated, it is generally targeted to and degraded by the 26S proteasome. Little is known about how ubiquitinated substrates arrive at the proteasome.
Proteins with UBL
(ubiquitin-like) and UBA (ubiquitin-associated) domains, such as Rad23 and Dsk2, have been shown to function as polyubiquitin receptors, which recognize a group of ubiquitinated substrates and translocate them to the 26S proteasome for degradation (12-15). In addition, one of the 19S subunits, Rpn10, can bind to multiubiquitin chains (16) and function as a ubiquitin receptor (13,14). However, since Rpn10 is dispensable for the growth of yeast cells (17), the existence of other ubiquitin-binding proteins in the 19S complex is likely. Recently, in vitro cross-linking experiments on the purified 26S proteasome suggested that the non-ATPase subunits Rpn1 and Rpn2 can bind to proteins with ubiquitin-like domains (18), and Rpt5/S6, one of the six ATPase subunits, can specifically contact a proteasome-bound polyubiquitin chain (19). However, whether they function as Ub receptors in vivo is not clear. Since the known ubiquitin receptors are only responsible for a subgroup of substrates, additional ubiquitin receptors or pathways have been suggested for targeting various classes of ubiquitinated substrates (14,15), but their identities remain elusive. Mass spectrometry-based interactive proteomics has evolved as a powerful tool for mapping proteome-wide protein interaction networks (20-22). Combined with affinity purification, mass spectrometric analyses have identified a number of proteasome interacting proteins (PIPs)
5
including new proteasome subunits (7,8); however, many known interactions were not identified by this method including the known Ub receptors, Rad23 and Dsk2. This is most likely due to the fact that many in vivo interactions are transient and/or low affinity and generally dependent on the specific cellular environment in which they occur. Therefore, a key strategy of our new approach is to effectively capture and identify these weak interacting proteins in vivo. Chemical cross-linking stabilizes interactions through covalent bond formation, allowing the detection of protein-protein interactions in native cells or tissues that are weak and/or transient. The most commonly used cross-linking reagent is formaldehyde, a water soluble and cell membrane permeable molecule, which provides a fast and reversible cross-linking reaction (23). This method has been widely used for the study of protein-DNA and protein-protein interactions (2330). Recent reports demonstrated that formaldehyde cross-linking/mass spectrometry strategies are effective in capturing and identifying in vivo protein-protein interactions including low affinity and transient interactions (25,26,28). In this work, we describe an integrated approach, QTAX, for quantitative analysis of tandem affinity purified in vivo cross-linked (X) protein complexes and report the application of this strategy to decipher the 26S proteasome interaction network. MATERIALS AND METHODS Chemicals and antibodies All general chemicals for buffers and culture media were purchased from Fisher Scientific and/or VWR International. Arginine-13C6 was purchased from Cambridge Isotope Laboratories, Inc. (Andover, MA) and other amino acids were purchased from Sigma-Aldrich (St. Louis, MO). RGS-His antibody was obtained from Qiagen (Germantown, MD), and ImmunoPure streptavidin, HRP conjugated antibody, and Super Signal West Pico chemiluminescent substrate
6
were from Pierce Biotechnology (Rockford, IL). Trypsin was purchased from Promega Corp. (Madison, WI) and endoproteinase Lys-C from WAKO chemicals (Osaka, Japan). Yeast strains and culture conditions Standard yeast growth media and conditions were used (31). All strains used in this study, except the commercial TAP-tagged strains (Open Biosystems, AL), are isogenic to 15Daub∆, a bar1∆ ura3∆ns, a derivative of BF264-15D (32). Rpn11 and Rpt5 were tagged with the HB-tag (Tagwerker et al., manuscript in preparation) at their C-terminus at the chromosomal locus by a single-step PCR strategy (33). These strains were used for all experiments performed in YEPD media. For the SILAC experiments, arginine auxotroph strains were constructed by deleting ARG4 with a hygromycine resistant marker following a PCR-based strategy (34). For the initial experiments, the Rpn11-HB strain was grown in 100-500 ml YEPD media at 200rpm, 30ºC to an OD600∼ 1.5. The strains used in the SILAC experiments were grown in 500-1000 ml synthetic complete media supplemented with 20mg/l of either heavy arginine (wild type, untagged strain) or light arginine (Rpn11-HB strain) to a final OD600 ∼ 0.9 before cross-linking experiments and subsequent purifications. In vivo Formaldehyde Cross-linking To identify the optimal cross-linking conditions, different concentrations of formaldehyde (13%) were directly added to the yeast cell culture for various incubation times (10 to 30 min.) and cells were incubated at 30ºC. The cross-linking reaction was quenched for 10 min at 30oC by addition of 2.5 M glycine with a final concentration of 0.125 M. After cross-linking, cells were collected, washed with ice cold water, and frozen at -80ºC prior to lysis. It was determined that
7
1% formaldehyde for 10 min at 30oC was the optimal cross-linking condition, which was used in all subsequent experiments. Tandem Affinity Purification of Cross-Linked Proteasome Interacting Proteins (PIPs) Frozen cells were lysed by bead-beating in lysis buffer (8 M urea, 300 mM NaCl, 50 mM NaH2PO4, 0.5% NP-40, 1 mM PMSF, pH 8; 1 ml lysis buffer per 100 ml cultured cells) and cellular debris was removed by centrifugation. Clarified lysate was initially incubated with NiSepharose 6 Fast Flow beads (Amersham Biosciences). After binding, the beads were washed twice with 20 bed volumes of lysis buffer pH 8, twice with 20 bed volumes of lysis buffer, pH 6.3, and once with 20 bed volumes of lysis buffer, pH 6.3 + 10 mM imidazole to remove nonspecifically bound proteins. The proteins were eluted from the Ni-Sepharose beads using 10 bed volumes of lysis buffer, pH 4.3. The eluate was brought to pH 8 with 1 M TRIS, pH 8 and applied directly to immobilized streptavidin beads (Pierce, Immunopure Immobilized Streptavidin).
To remove any remaining contaminants, streptavidin beads were washed
stringently using 10 bed volumes each of buffer 2 (8 M Urea, 0.2 M NaCl, 0.2% SDS, 100 mM Tris, pH 7.5), and buffer 3 (buffer 2 containing 2% SDS). Residual urea and SDS were removed using buffer 4 (0.2 M NaCl, 100 mM Tris, pH 7.5). Both binding reactions were allowed to proceed overnight at room temperature. Gel Electrophoresis and Immunoblotting Cell lysates, wash fractions, and elution fractions were separated by SDS-PAGE (7% or 10% gel). Proteins were transferred to a PVDF membrane and analyzed by immunoblotting. The RGS6H antibody was used at a 1:2000 dilution (Qiagen, Germantown, MD). Biotinylated proteins were detected using a streptavidin-HRP conjugate (1:5000).
8
Quantitative Analysis Using SILAC Arginine auxotroph strains were grown in synthetic complete media supplemented with heavy arginine (wild type, untagged strain) or light arginine (Rpn11-HB strain). Each strain was initially grown to stationary phase overnight (5-20ml culture), then transferred to a larger culture (100-200ml) until reaching early stationary phase. The culture was then transferred to a final volume of either 500ml or 1L and incubated until reaching OD600 ∼ 0.9. Finally, the cells were in vivo cross-linked, collected, lysed, clarified, and purified as described above. Equal amounts of lysate from the heavy and light labeled cultures were combined prior to tandem affinity purification. Liquid Chromatography –Tandem Mass Spectrometry (LC MS/MS) At the last step of the tandem affinity purification, the proteins were bound on streptavidin beads and, due to extremely high affinity binding, could not be eluted. Therefore, direct on-bead digestion was performed. For direct trypsin digestion, the urea buffer was replaced with 50 mM NH4HCO3 and the proteins were digested overnight at 37oC. For Lys-C/ trypsin digestion, LysC was directly added to the 8 M urea buffer and streptavidin beads for 4 hours at 37oC and then the urea concentration was reduced to < 2 M for trypsin digestion overnight. After digestion, the tryptic peptides were extracted from the streptavidin beads with 25% acetonitrile/0.1% formic acid three times. The extracts were pooled, concentrated using speedvac, and acidified by 0.1% formic acid prior to mass spectrometric analysis. For 1-dimentional (1-D) LC MS/MS analysis, the tryptic digests were directly injected onto the column, whereas the Lys-C/trypsin digests were desalted first with C18 Ziptips (Millipore, MA) before analysis.
1-D LC MS/MS was carried out by nanoflow reverse phase liquid
9
chromatography (RPLC) (Ultimate LC packings, Dionex) coupled on-line to a quadrupoleorthogonal-time-of-flight tandem mass spectrometer (QSTAR XL, Applied Biosystems/MDS Sciex). RPLC was performed using a PepMap column (75 µm ID x 150 mm long, LCpackingsDionex) and the peptides were eluted using a linear gradient of 0% B to 35% B in 100 min at a flow of 200 nL/min. Solvent A contained 98% H2O/2% acetonitrile/0.1% formic acid, whereas solvent B was composed of 98% acetonitrile/2% H2O /0.1% formic acid. The QSTAR MS was operated in an information-dependent mode in which each full MS scan was followed by three MS/MS scans where the three most abundant peptide molecular ions were dynamically selected for collision induced dissociation (CID), thus generating tandem mass spectra. In general, the ions selected for CID were the most abundant in the MS spectrum, except that singly charged ions were excluded and dynamic exclusion was employed to prevent repetitive selection of the same ions within a preset time. Collision energies were programmed to be adjusted automatically according to the charge state and mass value of the precursor ions. To increase the number of MSMS spectra acquired from any given sample and improve the dynamic range of mass spectrometric analysis, multiple LC MS/MS runs were performed on the same sample with the exclusion lists (i.e. the m/z list of the ions being sequenced from the previous runs) generated from the previous LC MSMS runs using Mascot script within the Analyst program. For the 2-D LC MS/MS, the digests were first separated by strong cation exchange (SCX) chromatography, which was performed using an AKTA system (Amersham Biosciences). Solvent A (25% acetonitrile, pH 3 adjusted with formic acid) and solvent B (solvent A with 400 mM NH4Cl) were used to develop a salt gradient. The digests were separated using a 1 mm x 15 cm polysulfoethyl A column (Poly LC, Columbia, MD) at a flow rate of 90 µl/min. Peptide elution was monitored by UV detection at 215 and 280 nm. A typical separation employed 0% B
10
from 0–10 min to allow for sample loading and removal of non-peptide species, followed by a gradient of 0–100% B from 10–30 min. Fractions were manually collected based on UV absorbance. All of the SCX fractions were desalted off-line using C18 Ziptips (Millipore, MA) prior to LC MS/MS. Database searching and abundance ratio calculation The monoisotopic masses (m/z) of both parent ions and their corresponding fragment ions, parent ion charge states (z) and ion intensities from the tandem mass spectra (MS/MS) acquired were automatically extracted using the script in the Analyst software and directly submitted for automated database searching for protein identification using two different search engines, Protein Prospector (UC, San Francisco) and Mascot (Matrix Science), to improve the confidence level of the protein identifications in the large data sets. The LC-Batchtag program within the developmental version of Protein Prospector was used for database searching.
The mass
accuracy for parent ions and fragment ions were set as ±100 ppm and 300 ppm, respectively. An in-house Mascot program was also used for database searching and the mass accuracy for parent ions was set as ±100 ppm and 0.3 Da was used for the fragment ion mass tolerance. Both SwissProt and NCBInr public databases were queried to identify the purified proteins since each database contains unique protein entries. In addition, the Search Compare program within the developmental version of Protein Prospector (35) was used to make a list of proteins that differed between samples. For the SILAC experiments, the Search Compare program was also used to calculate the relative abundance ratios of Arg-containing peptides based on either ion intensities of monoisotopic peaks or their areas observed in the LC MS spectra at the time when the peptides were sequenced and subsequently identified during database searching. The proteins
11
identified by one or two peptides were confirmed by manual inspection of the MSMS spectra. The relative abundance ratios were also validated by checking the raw spectra. Validation of the selected specific PIPs Several newly identified putative proteasome interacting partners were validated by coimmunoprecipitation experiments. Yeast strains expressing endogenous levels of TAP-tagged versions of the proteins were purchased (Open Biosystems, AL). Each TAP-tagged strain was grown to an OD600 of 1.5 and lysed in a buffer containing 25mM Tris, pH 7.5, 200mM NaCl, 0.2% NP-40, 2mM DTT, phosphatase inhibitors (50mM NaF, 0.1mM Na3VO4, 10mM Na4P2O7, 5mM EDTA, 5mM EGTA) and protease inhibitor complete (Roche Diagnostics, Germany). Approximately 5mg protein lysate was added to 15µl antigen affinity gel rabbit IgG (MP Biomedicals, Ohio) and incubated at 4°C for 1.75 hours. Beads were washed 3x1ml with wash buffer (25mM Tris, pH 7.5, 150mM NaCl, and 0.2% triton), and 1 x 1ml with TEB buffer (50mM TRIS pH7.5, 1mM EDTA, 1mM DTT). After washing, 30µl TEB buffer and 10 units of TEV protease were added and cleavage was allowed to proceed overnight at 4°C. Proteins that eluted from the beads were analyzed by immunoblotting with antibodies directed against the 19S proteasome subunit Rpt6 (GeneTex, TX). A yeast strain expressing a TAP-tagged version of the 20S subunit Pre1 was used to purify the 26S proteasome as a positive control. To prevent dissociation of the 19S complex from the TAP-tagged of the 20S subunit Pre1 purification was carried out in the presence of an ATP regenerating system (7). The TEV elution from this strain was diluted 1:25 before separation on SDS-PAGE to reduce the signal obtained by immunoblot analysis. A wild type, untagged strain was used as a negative control. RESULTS
12
A new tandem affinity tag for purification of the 26S proteasome complex and its interacting proteins The QTAX strategy for purifying and identifying the 26S proteasome interacting network is illustrated in Fig. 1. In order to use a tandem affinity purification compatible with in vivo crosslinking under fully denaturing conditions, we adapted a new tandem-affinity tag (HB-tag) that has been developed to specifically purify covalently linked proteins such as ubiquitinated or sumoylated proteins (Tagwerker and Kaiser, manuscript in preparation). The HB-tag consists of a hexahistidine sequence combined with a signal peptide that serves as a biotinylation signal in vivo (36).
HB-tagged proteins are efficiently biotinylated in vivo in yeast as well as in
mammalian cells (>95%) at a specific lysine residue present in the tag. To increase the affinity of the original HB-tag to Ni2+-chelate resin, a second hexahistidine sequence was added. HBtagged proteins can be purified by Ni2+-chelate chromatography followed by binding to streptavidin resin under fully denaturing conditions. To purify the 26S proteasome and its interacting proteins, two regulatory proteasome subunits, the non-ATPase subunit Rpn11 or the ATPase subunit Rpt5 were selected for HB tagging since these subunits have been successfully used with other tags for purification purposes (7,8). Rpn11 and Rpt5 were tagged at their Ctermini at the chromosomal locus so that endogenous levels of the tagged proteins were expressed. The HB-tagged forms of both of the essential subunits Rpn11 and Rpt5 appeared to be fully functional since the growth rate and cell morphology of each tagged strain was indistinguishable from that of an untagged strain (data not shown). As shown in Fig. 1B, prior to cell lysis, in vivo formaldehyde cross-linking was used to freeze the 26S proteasome interaction networks in yeast cells expressing a HB-tagged proteasome subunit.
The cross-linked
proteasome complexes were isolated using sequential purification on Ni-sepharose and
13
immobilized streptavidin under fully denaturing conditions.
The purified complexes were
subsequently digested with trypsin and their composition was identified by peptide sequencing using LC MS/MS analysis and automated database searching. To evaluate the purification efficiency of the cross-linked proteasome complexes, we monitored each purification step by immunoblotting. As shown in Fig. 2A, all of the crosslinked products containing an Rpn11-HB proteasome subunit were bound efficiently on Nisepharose since there was no visible amount left in the flow through after the binding. Proteins eluted from the Ni-sepharose were subsequently incubated with immobilized streptavidin for the second purification step. As shown in lane 5 (flow through after streptavidin binding) in Fig. 2A, the Rpn11-HB containing complexes were efficiently captured on streptavidin beads. These results demonstrated that HB tagged Rpn11 is compatible with formaldehyde cross-linking and that cross-linking did not interfere with protein purification. Similar results were observed from cells expressing Rpt5-HB (data not shown). Optimization of in vivo cross-linking of HB tagged proteasome interacting proteins using formaldehyde To optimize the cross-linking condition, different concentrations of formaldehyde were tested at various incubation times in yeast cells expressing the Rpn11-HB. As shown in Fig. 2B, immunoblot analysis indicated that protein-protein cross-linking efficiency did not improve with formaldehyde concentration above 1%. Similarly, cross-linking efficiency appeared to plateau after 10 min of incubation at 30˚C with 1% formaldehyde (data not shown). Mass spectrometric analysis was also used to determine the optimal formaldehyde concentration for in vivo crosslinking. Since the optimal cross-linking condition is expected to provide the highest yield of specific cross-linking products but the lowest non-specific background, we have compared a list
14
of proteins identified from samples cross-linked with different formaldehyde concentrations. The number of proteasome subunits identified was comparable in the samples that were crosslinked with 1% or 2% formaldehyde for 10 minutes at 30˚C, but the number of peptides matched to each subunit was higher in the 1% formaldehyde sample. Cross-linking with a final concentration of 3% formaldehyde resulted in a significantly lower amount of peptides. This is probably due to the fact that higher concentrations of formaldehyde (>2%) tend to cause more proteins to cross-link with DNA, leading to a reduced amount of total soluble protein in the lysate. Identification of the 26S proteasome complex after in vivo formaldehyde cross-linking and tandem affinity purification After optimizing the cross-linking and affinity purification conditions, we isolated the crosslinked 26S proteasome complexes from cells expressing either Rpn11-HB or Rpt5-HB. Using these strains, we were able to purify and identify the complete composition of the 26S proteasome complex, as summarized in Table I. In addition to all essential 26S proteasome subunits, the newly assigned proteasome subunits including Rpn13 (7), Sem1 (9), Ubp6 and Ecm29 (8) were captured. A wild type, untagged strain was processed in parallel and did not result in the identification of any proteasome subunits. Given that the purification was carried out under fully denaturing conditions, these results indicate that the identified proteasome subunits were specifically present in the in vivo cross-linked protein complexes containing the HB-tagged baits Rpn11 or Rpt5. Quantitative analysis of proteasome interacting proteins (PIPs) using SILAC and tandem mass spectrometry
15
Although the tandem affinity purification technique using Ni-NTA/Streptavidin resins has improved purification efficiency significantly compared to one-step Ni-NTA affinity purification, there are still some background proteins present after the final purification step. To distinguish specific PIPs from background proteins, we employed a quantitative approach using the SILAC strategy (37,38), which is illustrated in Fig. 3. Briefly, one population of cells are grown in media containing the natural form of an essential amino acid,
12
C6-Arginine (Arg),
while another population (e.g. untagged wild type cells) are grown in media containing a stable, isotope-labeled analogue, 13C6-Arg. Equal amounts of protein from the two different populations of cells are mixed and affinity purified together. The bound proteins on streptavidin beads are digested and analyzed by LC MS/MS. If a protein is present in both samples, the resulting arginine containing peptides will be observed as pairs (12C6-Arg vs.
13
C6-Arg) due to the
incorporation of different isotope labeled arginine during cell culture. The relative abundance of proteins present in two different samples can thus be quantified based on the ratios of the peak intensities of arginine peptide pairs determined in LC MS experiments. In the LC MS spectra, peptides observed as singlets indicate that they are either arginine containing peptides derived from the proteins only present in one sample, or peptides without arginine (e.g. lysine containing tryptic peptides or C-terminal peptides). This quantitative method is a powerful approach to identify specific interacting proteins of interest (38). We first evaluated the conditions for the SILAC experiment in yeast cells. Wild type Saccharomyces cerevisiae can synthesize all amino acids, which might lead to incomplete incorporation of isotopically labeled amino acids during in vivo labeling of cells. To circumvent this potential problem, we generated yeast strains that are auxotrophic for arginine by deletion of the ARG4 gene to prevent endogenous arginine from competing with the labeled 13C6-Arg in the
16
medium. In order to test the incorporation of labeled arginine, cells were grown either in the presence of 13C6-Arg or of 12C6-Arg for 8 hours. Proteins were extracted from both samples and separated by 1-D gel electrophoresis. Gel bands corresponding to the same proteins were excised from the two samples, digested with trypsin, and analyzed by LC MS/MS. The results demonstrated that incorporation of 13C6-Arg was close to completion (>95%), which was further improved with overnight incorporation; comparable to what has been reported by Jensen and coworkers (39). Interestingly, in our yeast strain carrying the arg4∆ deletion, no 13C6-Arg was converted to
13
C5-Proline. This is advantageous since conversion of
13
C6-Arg to
13
C5-Proline
was previously observed in mammalian cells (40) and requires correction for the quantitation of proline-containing peptides. Analysis and quantitation of several other randomly chosen proteins confirmed the excellent labeling efficiency (data not shown), indicating that an arg4∆ yeast strain is suitable for SILAC experiments using 13C6-Arg labeling. To identify the specific PIPs after in vivo cross-linking, we cultured an Rpn11-HB expressing strain in 12C6-Arg and a wild type strain in 13C6-Arg containing medium (Fig. 3). After in vivo formaldehyde cross-linking, equal amounts of cell lysate from the two different populations of cells were mixed for tandem affinity purification and subsequent trypsin digestion. The digests were analyzed by both 1-D and 2-D LC MS/MS to improve the detection sensitivity and dynamic range of the analysis. The list of proteins identified were summarized, compared, and validated, and their relative peptide abundance ratios of Arg-containing peptide pairs were calculated using the Search Compare program within Protein Prospector (35). Three different groups of proteins were classified based on their relative peptide abundance ratios as follows: A). The 26S proteasome subunits. The first group of proteins are the subunits of the 26S proteasome complex.
We expected that the Rpn11-HB cross-linked complexes would be
17
enriched significantly during the tandem affinity purification from the cells expressing Rpn11HB, but not from the wild type cells. Since Rpn11-HB tagged cells were grown in the medium containing light Arg (L), whereas wild type cells were cultured in heavy Arg (H) medium, the relative abundance ratios (L/H) of the proteasome subunits were expected to be much larger than 1. For all of the identified proteasome subunits including 19S and 20S core subunits, we observed the same pattern of the Arg-containing peptides (light vs. heavy forms) in the LC MS experiments. This is exemplified on two Arg-containing peptides from two different proteasome subunits Rpt6 (an ATPase subunit) and Scl1 (α1 subunit of 20S) (Fig. 4A and 4B). Their sequences were determined by MS/MS. Since only one arginine is present in each sequence, the mass differences between the assumed peptide pairs should be 6 Da if both light and heavy forms were present.
Based on the defined mass differences, no corresponding
13
C6-Arg labeled
peptides were observed for any identified 12C6-Arg labeled proteasome peptides, resulting in the observation of Arg-containing peptides as singlets, not pairs. This suggests that the proteasome subunits were selectively enriched and purified from the mixed lysates (Rpn11-HB and wild type cells) so, therefore, their relative abundance ratios are high. B). Putative specific proteasome interacting proteins (PIPs). The second group of proteins identified is classified as the putative specific PIPs with the relative abundance ratios > 1.5. We have identified a total of 64 putative specific PIPs that are summarized in Table II. To our knowledge, 42 of these putative PIPs are newly identified and have not been reported before. Among these 64 proteins are polyubiquitins, ubiquitin receptors, ubiquitinated substrates, chaperons, transcription factors, translational factors, and karyopherin, etc. Since polyubiquitin and the ubiquitin receptors (Rad23 and Dsk2) are known to be directly involved in ubiquitinproteasome degradation pathways (13,14), their interactions with the proteasome are believed to
18
be highly specific and their ratios are also high. As shown in the TOF MS spectra (Fig. 5A and B), only the light forms of a ubiquitin peptide was clearly observed as either triply charged (MH33+ 487.6) or doubly charged (MH22+ 730.9) ions, and no 13C6-Arg labeled forms of the same peptide were detected. The peptide sequence was determined as LIFAG(48K)*QLEDGR by a series of y ions (y1~y9) obtained in the MS/MS spectrum (MH33+ 487.6) shown in Fig. 5C, resulting from the polyubiquitin chain linked to lysine at position 48. After trypsin digestion, the remnant of one ubiquitin molecule on the 48Lys of the other ubiquitin has two glycine residues (i.e. GG) (41). It is noteworthy that only Lys48-linked polyubiquitin chains were detected to be associated with proteasome in this experiment and no other polyubiquitin-chain linkage types were observed. This is consistent with the notion that Lys48-linked chains are the predominant degradation signal for ubiquitinated substrates (1,3,17).
Similarly, the
12
C6-Arg-containing
peptides of the known ubiquitin receptors Rad23 and Dsk2 were also identified as singlets due to the absence of the 13C labeled forms of these peptides. As an example, the TOF MS spectrum of a tryptic peptide (MH22+ 752.85) is shown in the inset of Fig.6 and its sequence was determined as QLNDMGFFDEDR, which matched unambiguously to Dsk2. As shown in Fig. 6B, the relative abundance ratio of this particular peptide should be >8 even if the intensity of the 13C6Arg labeled peptide (MH22+ 755.85) was the same as the noise level (~7 counts). Similar results were obtained for Rad23, indicating that Ub receptors and the components directly involved in the Ub-proteasome degradation pathways can be specifically enriched. In addition to polyubiquitin, Rad23, and Dsk2, several other PIPs, i.e. Sse1/Sse2, Rps24A/B, Rpl13A/B, and Rps2, also have high ratios as listed in Table II. In comparison to the PIPs with high ratios, the Arg-containing peptides from other putative PIPs have been observed as pairs, not as singlets, and their relative abundance ratios are ranging from 1.6 to 6. As an example, the
19
selected Arg-peptide pairs from Ssa1 and Yef3 are given in Fig. 4C and D, and their relative abundance ratios of L/H were determined as 5.1 and 3.2, respectively.
These results are
reproducible from multiple MS analysis and different preparations. C). Background proteins. The last groups of proteins with relative abundance ratios ranging from 0.8 to 1.5, i.e. close to 1, are considered background proteins due to non-specific interactions. The representative TOFMS spectra of the Arg-containing peptide pairs from two unchanged proteins are shown in Fig. 4 E and F. The first peptide sequence was determined as EGDDVADAFQR and matched to pyruvate carboxylase Pyc1 with a ratio of 1.12 (Fig. 4E) and the second was identified as YAQDGAGIER and matched to the 60S ribosomal subunit Rpl3 with a ratio of 1.16 (Fig. 4F). Our results show that the ratios of different peptides resulting from the same proteins, and the ratios from different proteins in this group are reproducible within experimental error. Therefore, based on the relative abundance ratio (~1), it is easy to identify these background proteins. It is worth mentioning that three of previously reported putative PIPs, Acc1, Ilv6, and Shm2 (7) had relative abundance ratios close to one, indicating that under our experimental conditions they are likely nonspecific interacting proteins. Validation of the Selected Identified PIPs. We have selected 8 proteins with different abundance ratios to validate their interactions with the proteasome by an independent method using a different epitope tag. The proteins chosen in order of highest to lowest ratios were: Ssa1, Ybr025c, Yef3, Eft1, Tef4, Pfk1, Gus1, and Acc1. We used yeast strains expressing endogenous levels of TAP-tagged versions of these proteins and purified protein complexes based on the affinity of the TAP-tag to IgG-sepharose. Protein complexes were proteolytically released from the IgG-sepharose by TEV protease cleavage at the TEV-site present in the TAP tag. Protein complexes were analyzed for the presence of
20
proteasomes by immunoblotting using an antibody directed against the proteasome subunit Rpt6. The immunoblotting results are illustrated in Fig. 7. A yeast strain expressing the TAP-tagged 20S proteasome subunit Pre1 was used as a positive control (Fig. 7, lanes 1 and 9), and an untagged yeast strain was used as a negative control (Fig. 7, lanes 2 and 7). We found that all proteins tested with a ratio of 1.6 or higher (Ssa1, Ybr025c, Yef3, Eft1, and Tef4) interacted specifically with the proteasome (Fig. 7, lanes 3-6 and 10), whereas the proteins with ratios less than 1.6 (Pfk1, Gus1, Acc1) did not show positive interactions under these conditions (Fig. 7, lanes 8, 11 and 12). These results suggest that the putative PIPs with ratios >1.5 are most likely to be specific proteasome interacting proteins. However, it is important to emphasize that these experiments did not involve cross-linking and purification was carried out under native conditions. Therefore, it is possible that proteins that tested negative in this assay but had abundance ratios >1 are interacting with the proteasome in vivo, but that their interactions are very unstable or transient and only able to be captured in combination with in vivo cross-linking. DISCUSSION We have presented a novel integrated proteomics approach, QTAX, for the analysis of protein complexes including transient and weak interaction partners, which has been successfully applied to decipher the 26S proteasome interaction networks from budding yeast in vivo. This presents an improvement over other strategies using affinity purification and mass spectrometric analysis either with or without cross-linking. Combined with the SILAC strategy, a total of 64 putative PIPs were identified and they are involved in a variety of biological processes, including proteolysis, RNA processing, translation, metabolism, cell cycle, protein folding, etc (Table II). In addition to some known interactions, 42 putative new PIPs were identified, suggesting that many proteins may be involved in the ubiquitin-proteasome pathway.
21
In this work, in vivo cross-linking using formaldehyde was carried out to fix protein interactions prior to cell lysis, which can stabilize weak and transient interactions during purification processes.
To eliminate the non-specific purification background (i.e. proteins
bound to affinity resins non-specifically) and prevent the formation of non-covalent interactions after cell lysis, affinity purification of the cross-linked complexes under fully denaturing conditions was preferred although immunoaffinity purification under native condition has been applied to the study of M-Ras interacting proteins (28). Previously, the only tag available that can be used for affinity purification under fully denaturing conditions is the His tag, which has been successfully employed for the purification of ubiquitin and ubiquitin-like modified proteins using Ni-NTA chromatography (41,42), however, single step purification using the His tag alone is not very specific (42,43). To improve the purification specificity, a new tandem affinity tag, the HB tag was adopted in this work, which has proven to be compatible for both in vivo crosslinking and subsequent tandem affinity purification of the cross-linked protein complexes using Ni-NTA chromatography and affinity binding to streptavidin resins under fully denaturing conditions (e.g. 8M urea). The extremely high affinity binding of biotin to the streptavidin resin (Kd=10-15) allows stringent wash steps to effectively remove non-specific interactions, thus achieving higher purification specificity in comparison to one-step Ni-NTA purification. This was further confirmed by mass spectrometric analysis of the background proteins from one-step and two-step purification. The number of the background proteins (~40) obtained with two-step Ni-NTA/Streptavidin purification is about 16% of that (~240) obtained with only one-step NiNTA purification from untagged yeast cells after formaldehyde cross-linking. Furthermore, cross-linking did not seem to increase nonspecific purification background since the number of background proteins are comparable from cross-linked and non-cross-linked cells (data not
22
shown).
In comparison with previous in vivo cross-linking/affinity purification strategies
(25,26,28), quantitative purification of cross-linked proteasome complexes using the HB tag was carried out using two-step processes under fully denaturing conditions and the purified crosslinked proteasome complexes were analyzed directly by mass spectrometry without separation by SDS-PAGE, which may be advantageous for reserving physiologically formed protein interaction profiles and improving the sensitivity of protein identification. The identification of the full composition of 26S proteasome complex for the first time in one single experiment suggests that the proteasome assembly is dynamic and some of the loosely bound subunits can be lost even during one-step affinity purification. Although many peptides (>10) of the 19S and 20S alpha subunits were obtained during mass spectrometric analysis (Table I), most of the 20S beta subunits were identified by much smaller number of peptides (1~3), possibly due to the fact that the beta subunits were embedded in the center of the stacked ring structure of 20S proteasome complex, and cross-linking may block some potential trypsin digestion sites since lysines are the major sites for cross-linking reaction. Therefore, crosslinking reversal may be useful to expose more lysine residues for trypsin digestion, leading to the detection of additional peptides for improved sequence coverage. Comparing the results from Rpn11-HB and Rpt5-HB tagged cells, the majority of the proteins identified are very similar. As shown in Table I, some proteins seem to be found preferentially in one type of the tagged complex. Rad23 and Dsk2 were found only in Rpn11-HB tagged cells, whereas Nas6 was only identified in Rpt5-HB-tagged cells. The human homolog of Nas6, Gankyrin, was found to interact directly with Rpt 5 using a yeast two-hybrid system (44), which agrees well with our results. There are some surprising differences as the number of peptides identified for some proteins changed quite dramatically. For example, 47 of Ecm29 peptides from Rpt5-HB tagged
23
cells were identified, which was much higher than the 5 Ecm29 peptides from Rpn11-HB tagged cells. This suggests that subcomplexes of the proteasome may exist in vivo or that some PIPs can be differentially enriched after cross-linking due to their physical locations. Therefore, using several different HB-tagged proteasome subunits will be beneficial to capture all proteasome interacting proteins. The SILAC strategy was successfully employed to quantitatively distinguish the proteasome complex and its interacting proteins from background proteins. Based on the validation results using co-IP analysis, we have assigned the proteins with abundance ratios >1.5 as putative specific PIPs (Table II).
Interestingly, most of the identified putative PIPs have relative
abundance ratios ranging from 1.6 ~ 6, smaller than those of the known components involved in ubiquitin-proteasome pathways. The differences in the relative abundance ratios among the PIPs may be due to inherent differences in their interactions with the proteasome complex and their role in the ubiquitin-proteasome degradation pathways. Among the identified PIPs are a number of abundant cellular proteins such as heat shock proteins, elongation factors, ribosomal proteins, etc. Due to the large quantity of these proteins present in the cells, it is difficult to completely prevent their non-specific purification, which is thus leading to a relatively small abundance ratio. Nevertheless, these proteins are likely bona fide proteasome interactors because their relative abundance ratios were consistently above 1.5. Among the identified chaperone proteins, different classes were detected including Ssa, Ssb, and Kar2 members of the Hsp70 family, Sse1/Sse2, Hsc82 and Hsp82 (the budding yeast homologs of Hsp90), and all have relative abundance ratios >2. This suggests that their interactions with the proteasome are specific. The co-IP experiment has further confirmed that the Hsp70 family member Ssa1 in budding yeast associates with the proteasome specifically. In higher eukaryotes,
24
it has been shown that Hsc70/Hsp70 members facilitate in the delivering of aggregation-prone substrates for degradation by interacting with the proteasome through an adaptor protein (45,46). In addition, Hsp 90 family members have been suggested to play a role in the proteasome structural integrity and assembly through their interactions with the 26S proteasome (47). Therefore, different chaperon members may play distinct roles in modulating protein degradation by the proteasome. Previously, Tef1/2 (i.e. EF-1 α) has been identified as a putative PIP by mass spectrometry (7). In this work, Tef1/2 interaction with the proteasome has also been detected and determined to be specific since it has a ratio of 1.6. This agrees well with recent reports that the elongation factor 1 complex (αβγ) interacts with the 26S proteasome (48-50), and Tef1 possesses a chaperone-like activity and can interact with both ubiquitinated substrates and the proteasome subunit Rpt1 to promote co-translationally damaged protein degradation (49,50). In addition to Tef1/2, several other translational elongation factors including Yef3/Hef3, Eft1/2, and Tef4 were also identified as specific PIPs by the SILAC ratios (>1.6), which was further confirmed by coIP analysis. Although these interactions have not been detected before, we suspect that these translational elongation factors may also be involved in the degradation of co-translationally damaged proteins, and they may serve as the molecular linkage between the pathways of protein synthesis and degradation. In summary, QTAX is a powerful integrated proteomics approach to capture and identify stable and transient protein-protein interactions occurring in the cellular environment.
In
comparison to the existing methods, we have made significant technical improvements in two major aspects for the study of protein interactions using in vivo cross-linking. First, the new tandem affinity tag (HB tag) has demonstrated to be effective for purification of the cross-linked
25
proteasome complexes and their interacting proteins under fully denaturing conditions, resulting in substantial reduction of nonspecific binding during the purification in comparison to one-step Ni-NTA purification. Second, quantitative mass spectrometry using the SILAC strategy in yeast cells has been employed to study proteasome interactions, resulting in the effective differentiation of specific interactions from nonspecific interactions.
The QTAX strategy
described here has shown that in vivo cross-linking using formaldehyde effectively captures proteasome complexes and their interacting partners including some that have not been identified previously. Further genetic and biochemical studies of these interactors may help to reveal how the 26S proteasome is integrated into various biological pathways and how different PIPs play a role in substrate recognition and translocation to the 26S proteasome for degradation. ACKNOWLEDGMENTS We wish to thank Prof. Ralph Bradshaw for critically reading of the manuscript and Prof. A.L. Burlingame for allowing us to use the development version of Protein Prospector software package. We would also like to thank Karen LeBlanc for the purified proteasome complex for the co-IP experiments and members of the Huang and Kaiser laboratories for their help during this study. This work was supported by National Institutes of Health grants (GM-74830 to L.H. and GM-66164 to P.K.), the Dept. of the Army (PC-041126 to L.H.) and the California Breast Cancer Research Program (11NB-0177 to P.K.).
26
FIGURE LEGENDS Figure 1. The QTAX strategy for the purification of proteasome interacting proteins (A) Schematic representation of a proteasome subunit fused to the HB-tag at its C-terminus. The HB-tag shown here consists of a bacterially derived peptide that induces biotinylation in vivo flanked by two hexahistidine tags. One of the hexahistidine-tags contains an RGS6xHis epitope that can be used for immunoblot analysis. (B) Outline of the QTAX strategy for quantitative analysis of tandem affinity-purified in vivo cross-linked protein complexes. Formaldehyde is first added to the cells to cross-link proteins in vivo. The cells are then lysed and proteins are first purified by Ni-chelate chromatography. The eluent is subsequently applied to immobilized streptavidin beads for the second purification step. Cell lysis and both purification steps are carried out under fully denaturing conditions to prevent any non-covalent interactions. Proteins bound to streptavidin beads are directly digested with trypsin to release peptides that are analyzed by LC MS/MS and proteins are identified using automated database searching. Figure 2. In vivo cross-linking and tandem affinity purification (A) Tandem affinity purification of in vivo cross-linked Rpn11-HB containing proteasomes using Ni-sepharose and streptavidin beads. The Rpn11-HB cross-linked complexes were efficiently isolated using two-step affinity purification as shown by immunoblotting using anti-RGS6xHis antibodies. Lane 1: cell lysate; lane 2: unbound fraction after incubation with Ni2+-sepharose.; lane 3: wash with a buffer (pH 6.3) containing 8 M urea and 10mM immidazole; lane 4: Elution from Ni2+-sepharose (i.e. load to streptavidin); lane 5: unbound fraction after binding to streptavidin beads. The samples were separated on a 7% SDS-PAGE gel and Rpn11-HB was detected by immunoblotting using antiRGS6xHis antibodies. (B) Test of cross-linking efficiency using different concentrations of
27
formaldehyde as indicated for 10 min. The samples were separated on a 10% SDS-PAGE gel and detected as in (A).
Figure 3. Schematic diagram of a quantitative strategy using SILAC coupled with in vivo crosslinking and tandem affinity purification to identify specific proteasome interacting proteins. Cells expressing Rpn11-HB were grown in light medium (12C6-Arg), whereas wild type cells (untagged) were grown in heavy medium (13C6-Arg).
Figure 4. TOF MS Spectra of six Arg-containing peptide pairs from the selected proteins including two unchanged proteins, Pyc1 (L/H 1.12) (A) and 60S Rpl3 (L/H 1.16) (B); two proteasome subunits, Rpt6 (C) and Scl1 (D); and two putative PIPs, Ssa1 (L/H 5.1) (E) and Yef3 (L/H 3.2) (F). L--light (12C6-Arg), H--heavy (13C6-Arg). The relative abundance ratio of the Arg-containing peptides was calculated as L/H based on either monoisotopic peak intensity or area. All the peptides were observed as either doubly or triply charged ions and identified as peptides containing one arginine; therefore, the mass differences between each peptide pair is 6 Da. The ratios for proteasome subunits could not be calculated since the corresponding
13
C-
labeled peptides were not observed.
Figure 5. TOF MS Spectra of an Arg-containing peptide from polyubiquitin observed as (A) triply charged ion (MH33+ 487.6); (B) doubly charged ion (MH22+ 730.9). No detectable
13
C6-
labeled peptides were observed in these spectra; the same pattern of Arg-containing peptides (L vs. H) as in Figure 4(C) and (D). (C) MS/MS spectrum of the triply charged ion, MH33+ 487.6, and the sequence was determined as LIFAG[48K(GG)]QLEDGR, a ubiquitin-modified peptide.
28
Figure 6. (A) MS/MS spectrum of a tryptic peptide (MH22+ 752.86) and its sequence, determined as QLNDMGFFDFDR, matched to the Ub receptor, Dsk2. (B) TOFMS spectrum of this peptide and no 13C labeled peptide was observed.
Figure 7. Validation of selected PIPs identified with various abundance ratios by coimmunoprecipitation using TAP tagged yeast strains. Lanes 1 and 9 are positive controls using yeast strains expressing the TAP-tagged 20S subunit Pre1. For a better comparison, 25 times less of this sample was loaded as compared to the other samples. Lanes 2 and 7 are negative controls using a wild type yeast strain. Lane 3 (Ssa1-TAP, L/H 5.1), lane 4 (Ybr025c-TAP, L/H 3.3), lane 5 (Yef3-TAP, L/H 3.2), lane 6 (Eft1-TAP, L/H 2.7), lane 8 (Acc1-TAP, L/H 1.0), lane 10 (Tef4-TAP, L/H 1.6), lane 11 (Pfk1-TAP, L/H 1.5) and lane 12 (Gus1-TAP, L/H 1.3). Yeast strains expressing the TAP-tagged proteins indicated were lysed, protein complexes were purified under native conditions based on the affinity of the TAP-tag to IgG-sepharose and eluted by TEV protease cleavage. Eluted protein complexes were analyzed for the presence of proteasomes by immunoblotting with antibodies directed against the 19S subunit Rpt6. All tested proteins with abundance ratio >1.5 showed co-purification of the proteasome subunit Rpt6. The purification of 26S proteasome from the TAP-tagged Pre1 strain was carried out in the presence of an ATP regenerating system.
29
REFERENCES: 1. 2. 3. 4. 5. 6. 7.
8. 9. 10. 11. 12. 13. 14. 15. 16. 17.
Baumeister, W., Walz, J., Zuhl, F., and Seemuller, E. (1998) The proteasome: paradigm of a self-compartmentalizing protease. Cell 92, 367-380 Pickart, C. M. (2001) Mechanisms underlying ubiquitination. Annu Rev Biochem. 70, 503-533 Pickart, C. M., and Cohen, R. E. (2004) Proteasomes and their kin: proteases in the machine age. Nat Rev Mol Cell Biol. 5, 177-187 Voges, D., Zwickl, P., and Baumeister, W. (1999) The 26S proteasome: a molecular machine designed for controlled proteolysis. Annu Rev Biochem. 68, 1015-1068 Smalle, J., and Vierstra, R. D. (2004) The ubiquitin 26s proteasome proteolytic pathway. Annu Rev Plant Biol. 55, 555-590 Kloetzel, P. M. (2001) Antigen processing by the proteasome. Nat Rev Mol Cell Biol. 2, 179-187 Verma, R., Chen, S., Feldman, R., Schieltz, D., Yates, J., Dohmen, J., and Deshaies, R. J. (2000) Proteasomal proteomics: identification of nucleotide-sensitive proteasomeinteracting proteins by mass spectrometric analysis of affinity-purified proteasomes. Mol Biol Cell. 11, 3425-3439 Leggett, D. S., Hanna, J., Borodovsky, A., Crosas, B., Schmidt, M., Baker, R. T., Walz, T., Ploegh, H., and Finley, D. (2002) Multiple associated proteins regulate proteasome structure and function. Mol Cell. 10, 495-507 Sone, T., Saeki, Y., Toh-e, A., and Yokosawa, H. (2004) Sem1p is a novel subunit of the 26 S proteasome from Saccharomyces cerevisiae. J Biol Chem. 279, 28807-28816 Schmidt, M., Haas, W., Crosas, B., Santamaria, P. G., Gygi, S. P., Walz, T., and Finley, D. (2005) The HEAT repeat protein Blm10 regulates the yeast proteasome by capping the core particle. Nat Struct Mol Biol. 12, 294-303 Hershko, A., and Ciechanover, A. (1998) The ubiquitin system. Annu Rev Biochem. 67, 425-479 Saeki, Y., Saitoh, A., Toh-e, A., and Yokosawa, H. (2002) Ubiquitin-like proteins and Rpn10 play cooperative roles in ubiquitin-dependent proteolysis. Biochem Biophys Res Commun. 293, 986-992 Elsasser, S., Chandler-Militello, D., Muller, B., Hanna, J., and Finley, D. (2004) Rad23 and Rpn10 serve as alternative ubiquitin receptors for the proteasome. J Biol Chem. 279, 26817-26822 Verma, R., Oania, R., Graumann, J., and Deshaies, R. J. (2004) Multiubiquitin chain receptors define a layer of substrate selectivity in the ubiquitin-proteasome system. Cell 118, 99-110 Elsasser, S., and Finley, D. (2005) Delivery of ubiquitinated substrates to proteinunfolding machines. Nat Cell Biol 7, 742-749 Deveraux, Q., Ustrell, V., Pickart, C., and Rechsteiner, M. (1994) A 26 S protease subunit that binds ubiquitin conjugates. J Biol Chem. 269, 7058-7061 van Nocker, S., Sadis, S., Rubin, D. M., Glickman, M., Fu, H., Coux, O., Wefes, I., Finley, D., and Vierstra, R. D. (1996) 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
30
18. 19. 20. 21. 22.
23. 24. 25.
26.
27. 28. 29. 30. 31. 32. 33.
Saeki, Y., Sone, T., Toh-e, A., and Yokosawa, H. (2002) Identification of ubiquitin-like protein-binding subunits of the 26S proteasome. Biochem Biophys Res Commun. 296, 813-819 Lam, Y. A., Lawson, T. G., Velayutham, M., Zweier, J. L., and Pickart, C. M. (2002) A proteasomal ATPase subunit recognizes the polyubiquitin degradation signal. Nature 416, 763-767 Aebersold, R., and Mann, M. (2003) Mass spectrometry-based proteomics. Nature 422, 198-207 Ho, Y., Gruhler, A., Heilbut, A., and et.al. (2002) Systematic identification of protein complexes in Saccharomyces cerevisiae by mass spectrometry. Nature 415, 180-183 Gavin, A., Bosche, M., Krause, R., and Grandi P, M. M., Bauer A, Schultz J, Rick JM, Michon AM, Cruciat CM, Remor M, Hofert C, Schelder M, Brajenovic M, Ruffner H, Merino A, Klein K, Hudak M, Dickson D, Rudi T, Gnau V, Bauch A, Bastuck S, Huhse B, Leutwein C, Heurtier MA, Copley RR, Edelmann A, Querfurth E, Rybin V, Drewes G, Raida M, Bouwmeester T, Bork P, Seraphin B, Kuster B, Neubauer G, Superti-Furga G. (2002) Functional organization of the yeast proteome by systematic analysis of protein complexes. Nature 415, 141-147 Jackson, V. (1999) Formaldehyde cross-linking for studying nucleosomal dynamics. Methods 17, 125-139 Kuo, M. H., and Allis, C. D. (1999) In vivo cross-linking and immunoprecipitation for studying dynamic Protein:DNA associations in a chromatin environment. Methods. 19, 425-433 Schmitt-Ulms, G., Legname, G., Baldwin, M. A., Ball, H. L., Bradon, N., Bosque, P. J., Crossin, K. L., Edelman, G. M., DeArmond, S. J., Cohen, E., and Prusiner, S. B. (2001) Binding of neural cell adhesion molecules (N-CAMs) to the cellular prion protein. J Mol Biol. 314, 1209-1225 Schmitt-Ulms, G., Hansen, K., Liu, J., Cowdrey, C., Yang, J., DeArmond, S. J., Cohen, F. E., Prusiner, S. B., and Baldwin, M. A. (2004) Time-controlled transcardiac perfusion cross-linking for the study of protein interactions in complex tissues. Nat Biotechnol. 22, 724-731 Orlando, V., Strutt, H., and Paro, R. (1997) Analysis of chromatin structure by in vivo formaldehyde cross-linking. Methods. 11, 205-214 Vasilescu, J., Guo, X., and Kast, J. (2004) Identification of protein-protein interactions using in vivo cross-linking and mass spectrometry. Proteomics 4, Otsu, M., Omura, F., Yoshimori, T., and ikuchi, M. (1994) Protein disulfide isomerase associates with misfolded human lysozyme in vivo. J Biol Chem. 269, 6874-6877 Hall, D. B., and Struhl, K. (2002) The VP16 activation domain interacts with multiple transcriptional components as determined by protein-protein cross-linking in vivo. J. Biol. Chem. 277, 46043-46050 Guthrie, C., and Fink, G. R. (1991) Guide to Yeast Genetics and Molecular Biology, Academic Press, Inc., San Diego. Reed, S. I., Hadwiger, J. A., and Lorincz, A. T. (1985) Protein kinase activity associated with the product of the yeast cell division cycle gene CDC28. Proc Natl Acad Sci U S A 82, 4055-4059 Petracek, M. E., and Longtine, M. S. (2002) PCR-based engineering of yeast genome. Methods Enzymol. 350, 445-469
31
34. 35.
36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48.
Manivasakam, P., Weber, S. C., McElver, J., and Schiestl, R. H. (1995) Micro-homology mediated PCR targeting in Saccharomyces cerevisiae. Nucleic Acids Res. 23, 2799-2800 Chalkley, R. J., Baker, P. R., Huang, L., Hansen, K. C., Allen, N. P., Rexach, M., and Burlingame, A. L. (2005) Comprehensive analysis of a multidimensional liquid chromatography mass spectrometry dataset acquired on a QqTOF mass spectrometer: 2. New developments in protein prospector allow for reliable and comprehensive automatic analysis of large datasets. Mol Cell Proteomics Epub ahead of print, Cronan, J. E. J. (1990) Biotination of proteins in vivo. A post-translational modification to label, purify, and study proteins. J Biol Chem. 265, 10327-10333 Ong, S. E., lagoev, B., Kratchmarova, I., Kristensen, D. B., Steen, H., Pandey, A., and M., M. (2002) Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics. Mol Cell Proteomics. 1, 376-386 Blagoev, B., Kratchmarova, I., Ong, S. E., Nielsen, M., Foster, L. J., and Mann, M. (2003) A proteomics strategy to elucidate functional protein-protein interactions applied to EGF signaling. Nat Biotechnol. 21, 315-318 Gruhler, A., Olsen, J. V., Mohammed, S., Mortensen, P., Faergeman, N. F., Mann, M., and Jensen, O. N. (2005) Quantitative phosphoproteomics applied to the yeast pheromone signaling pathway. Mol Cell Proteomics. Mol Cell Proteomics 4, 310-327 Ong, S. E., Kratchmarova, I., and Mann, M. (2003) Properties of 13C-substituted arginine in stable isotope labeling by amino acids in cell culture (SILAC). J Proteome Res. 2, 173-181 Peng, J., Schwartz, D., Elias, J. E., Thoreen, C. C., Cheng, D., Marsischky, G., Roelofs, J., Finley, D., and Gygi, S. P. (2003) A proteomics approach to understanding protein ubiquitination. Nat Biotechnol. 21, 921-926 Wohlschlegel, J. A., Johnson, E. S., Reed, S. I., and Yates, J. R., 3rd. (2004) Global analysis of protein sumoylation in Saccharomyces cerevisiae. J Biol Chem 279, 4566245668 Denison, C., Rudner, A. D., Gerber, S. A., Bakalarski, C. E., Moazed, D., and Gygi, S. P. (2005) A proteomic strategy for gaining insights into protein sumoylation in yeast. Mol Cell Proteomics 4, 246-254 Rezvani, K., Mee, M., Dawson, S., McIlhinney, J., Fujita, J., and Mayer, R. J. (2003) Proteasomal interactors control activities as diverse as the cell cycle and glutaminergic neurotransmission. Biochem Soc Trans. 31, 470-473 Luders, J., Demand, J., and Hohfeld, J. (2000) The ubiquitin-related BAG-1 provides a link between the molecular chaperones Hsc70/Hsp70 and the proteasome. J Biol Chem. 275, 4613-4617 Wiederkehr, T., Bukau, B., and Buchberger, A. (2002) Protein turnover: a CHIP programmed for proteolysis. Curr Biol 12, R26-28 Imai, J., Maruya, M., Yashiroda, H., Yahara, I., and Tanaka, K. (2003) The molecular chaperone Hsp90 plays a role in the assembly and maintenance of the 26S proteasome. Embo J 22, 3557-3567 Tokumoto, T., Kondo, A., Miwa, J., Horiguchi, R., Tokumoto, M., Nagahama, Y., Okida, N., and Ishikawa, K. (2003) Regulated interaction between polypeptide chain elongation factor-1 complex with the 26S proteasome during Xenopus oocyte maturation. BMC Biochem 4, 6
32
49. 50.
Chuang, S. M., Chen, L., Lambertson, D., Anand, M., Kinzy, T. G., and Madura, K. (2005) Proteasome-mediated degradation of cotranslationally damaged proteins involves translation elongation factor 1A. Mol Cell Biol 25, 403-413 Chuang, S. M., and Madura, K. (2005) Saccharomyces cerevisiae Ub-conjugating enzyme Ubc4 binds the proteasome in the presence of translationally-damaged proteins. Genetics
33
Table I. Summary of the 26S proteasome subunits identified using in vivo cross-linking, tandem affinity purification and mass spectrometry Acc# Gene Name # of Peptides # of Peptides identified w/Tagged identified w/Tagged Rpn11 Rpt5 19S regulatory subunits P38764 P32565 P40016 Q12250 Q12377 Q06103 Q08723 Q04062 P38886 P43588 P32496 6323453* O94742 P33299 P40327 P33298 P53549 P33297 Q01939
RPN1 RPN2 RPN3 RPN5 RPN6 RPN7 RPN8 RPN9 RPN10 RPN11 RPN12 RPN13 SEM1 RPT1 RPT2 RPT3 RPT4 RPT5 RPT6
57 68 31 34 27 25 24 30 14 34 17 5 2 36 37 36 36 43 35
53 55 26 35 20 22 19 27 11 24 11 5 1 35 34 30 38 42 33
Others P43943 P50086 P38737
UBP6 NAS6 ECM29
8 0 4
7 2 47
20S Core subunits P21243 P23639 P23638 P40303 P32379 P40302 P21242 P38624 P25043 P25451 P22141 P30656 P23724 P30657
SCL1 PRE8 PRE9 PRE6 PUP2 PRE5 PRE10 PRE3 PUP1 PUP3 PRE1 PRE2 PRE3 PRE4
15 18 12 9 10 12 8 5 2 1 2 3 2 6
12 10 8 8 10 14 10 9 0 1 1 3 2 5
* This protein entry is not present in swissprot database, only in NCBInr.
34
Table II. Summary of the Putative Proteasome Interacting Proteins (PIPs) Identified by QTAX with the Relative Abundance Ratios >1.5
Gene Name
ORF Name
L/H ratio (average)
*RPS2 UBI4 DSK2 RAD23 RPL13A/B *SSE1/SSE2 *RPS24A/B SSA3 SSA4 SSA2 SSA1 *KAR2 SSB2 *SEC53 *YBRO25C *SSB1 *YEF3 *HEF3 *SAM1/2 *EFT1/2 *RPS11A/B *FPR1 HSC82 ENO2 ENO1 *KAP123 *HSP82 *LEU1 *PMA1/2 *LYS21 PAB1 *LYS20
YGL123W YLL039C YMR276W YEL037C YDL082W YPL106C YER074W YBL075C YER103W YLL024C YAL005C YJL034W YNL209W YFL045C YBR025C YDL229W YLR249W YNL014W YLR180W YOR133W YDR025W YNL135C YMR186W YHR174W YGR254W YER110C YPL240C YGL009C YGL008C YDL131W YER165W YDL182W
no C 13 no C 13 no C 13 no C 13 no C 13 no C 13 no C 5.9 5.7 5.2 5.1 5.1 3.5 3.4 3.3 3.2 3.2 3.1 3.1 2.7 2.7 2.7 2.6 2.5 2.5 2.5 2.4 2.1 2.1 2.1 2.1 2
13
Std. Dev.
Gene Name
ORF Name
L/H ratio (average)
Std. Dev.
N/A N/A N/A N/A N/A N/A N/A 1.0 1.2 1.1 0.9 -0.6 0.7 0.5 0.5 0.7 0.5 1.1 0.1 --0.5 0.6 0.7 -0.6 0.3 0.2 0.3 0.0 0.3
PGK1 *CYC1/7 TIF2 *TIF1 *FBA1 *ERG20 *GND1/2 BMH1/2 *ACT1 *TSA1 *CDC19 *TAL1 *CDC60 ADH1 PDC1 *PDC5/6 *ASN1 *RPS0A/B *YHB1 *ALD6 *SER1 *THS1 *TEF4 TDH1/2/3 TEF1/2 KRS1 *DPS1 *VAS1 *RPS27A/B *GLN1 RPL5 *SAH1
YCR012W YJR048W YJL138C YKR059W YKL060C YJL167W YHR183W YER177W YFL039C YML028W YAL038W YLR354C YPL160W YOL086C YLR044C YLR134W YPR145W YGR214W YGR234W YPL061W YOR184W YIL078W YKL081W YJL052W YPR080W YDR037W YLL018C YGR094W YKL156W YPR035W YPL131W YER043C
2 2 1.9 1.9 1.9 1.9 1.9 1.8 1.8 1.8 1.8 1.8 1.8 1.7 1.7 1.7 1.7 1.7 1.7 1.7 1.7 1.7 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6
0.6 0.2 0.5 0.5 0.3 0.1 0.3 0.3 0.1 0.3 0.4 0.2 -0.6 0.3 0.3 0.2 0.2 0.3 0.1 0.3 0.2 0.1 0.2 0.3 0.2 0.2 0.1 0.1 0.1 0.2 0.2
* PIPs captured that have not been previously reported
-- Only 1 pair was used for the ratio calculation.
35
Proteasome subunit
B.
6xhis
Rpn11
A.
HB tag biotin
6xhis
Substrate Recognition
?
R
? ?
HB
HB
Rad23
19S
? 20S
?
1.Freeze interaction (in vivo cross-linking)
2.Affinity Purification I.Ni2+-NTA II.Streptavidin
26S
MS/MS 4. 1-D or 2-D LC MS/MS 5. Database Searching
3.On-beads Digestion
Poly ubiquitin(Ub) chain cross-linkage Ubiquitinated substrates
P1:LSPLAQELR P2:QQLAPYSDDLR P3:GGGSLAEYHAK
Protein Identification
R
known Ub receptor
?
Unknown Ub receptor
P(1, 2,..): Protein 1, 2, …
36
Figure 1
A.. Lane
1
2
3
4
5
Anti-RGSHis Blot
B. Final Conc. of FA (%)
0
1
2
3
Figure 2 37
HB-Rpn11 12C
6-Arg
Wild Type 13C
6-Arg
In vivo cross-linking Lysis mixing Tandem affinity purification (Ni2+-NTA/Streptavidin resins) Trypsin Digestion LC MS/MS Protein Identification and Quantitation ∆=n x 6Da
m/z
38
Figure 3
1000
No 13C
500
B
100
L
80 60
670.80 671.30
L
VEGSGGGDSEVQR (Proteasome subunit Rpt6)
Ion Counts
1500
745.37 745.69
mANLSQIYTQ (Proteasome subunit α1)
40
No 13C
20
0
0 746
748
750
669
670
671
672
L
600
ATAGDTHLGGEDFDNR (Ssa1)
∆ 561.23
400
H
200
D
Ion Counts
800
559.24
Ion Counts
C
L/H=5.1
1200 1000 800
L
LSVATADNR (Yef3)
∆
600
H
400
L/H=3.2
0 558
559
560
561
562
563
564
473
474
475
476
477
478
1000
L/H=1.12
500
F 3000 2500
L
540.26
3500
EGDDVADAFQR (Pyc1)
1500
H ∆
YAQDGAGIER (60S Rpl3)
2000 1500
L/H=1.16
1000 500
0
0 610
612
614
479
m/z Ion Counts
∆
H
614.75
L
2000
611.73
m/z Ion Counts
675
200
0
E
674
m/z 473.73
m/z
673
476.75
744
543.27
Ion Counts
A
616
618
540
m/z
542
544
546
m/z
Note: ∆ = 6 Da
Figure 4 39
1000
No 13C
500
350
L
300 250 200
730.90 731.41
Ion Counts
1500
L
B
MH33+
Ion Counts
2000
487.60 487.93
A
MH22+
150 100
No 13C
50 0
0 488
a2
491
492
729
I/L
730
731
m/z
732
733
734
735
m/z
86.10
700
490
199.18
C
489
GG
487
L I F A G K* Q L E D G R
E
y2
G G
L
232.14
400
Q
K* G
589.30
y6 2+
(y10)
y7 y8 959.54
1016.55
476.22
y3
617.81
100
175.12
y1
y4
347.17
200
227.17
F
A
y5
b2
717.35
300
120.08
Ion Counts
500
D
0 200
400
600
800
1000
m/z 40
Figure 5
y9 1087.50
G
R
GG
600
10
200 400
y7
600
y5 y6
y4
800
41
1000
20
7 5 5 .8 5
L
y9
5
1200
7 5 2 .8 6 7 5 3 .3 5
80
752
y10
1 3 7 6 .4 9
25 60
1 2 6 3 .5 8
30
Io n C o u n ts
QLNDMGFFDFDR
1 0 3 5 .5 0
y3
9 0 3 .5 2
B
1 1 4 9 .5 3
15 8 4 6 .4 1
a2 4 3 7 .2 4
y1
6 9 9 .3 4
1 7 5 .1 3
20
5 5 2 .2 6
2 1 4 .1 6
Io n C o u n ts
A 35
753
MH22+
40
No 13C
0 754 755
1400
m/z
Figure 6
756 757
m/z
y8
y11
0 1600
Lane
1
2
3
4
5
6
TAP strain
Pre1
-control
Ssa1
Ybr025c
Yef3
Eft1
Ratio (L/H)
High
*
5.1
3.3
3.2
2.7
Lane
TAP strain Ratio (L/H)
7
8
-control Acc1 *
1.0
9
10
11
12
Pre1
Tef4
Pfk1
Gus1
high
1.6
1.5
1.3
Figure 7
42
