Activity-Guided Proteomic Profiling of Proteasomes Uncovers a Variety of Active (and Inactive) Proteasome Species

Abstract

Proteasomes are multisubunit, multicatalytic protein complexes present in eukaryotic cells that degrade misfolded, damaged, or unstructured proteins. In this study, we used an activity-guided proteomic methodology based on a fluorogenic peptide substrate to characterize the composition of proteasome complexes in WT yeast and the changes these complexes undergo upon the deletion of Pre9 (Δα3) or of Sem1 (ΔSem1). A comparison of whole-cell proteomic analysis to activity-guided proteasome profiling indicates that the amounts of proteasomal proteins and proteasome interacting proteins in the assembled active proteasomes differ significantly from their total amounts in the cell as a whole. Using this activity-guided profiling approach, we characterized the changes in the abundance of subunits of various active proteasome species in different strains, quantified the relative abundance of active proteasomes across these strains, and charted the overall distribution of different proteasome species within each strain. The distributions obtained by our mass spectrometry-based quantification were markedly higher for some proteasome species than those obtained by activity-based quantification alone, suggesting that the activity of some of these species is impaired. The impaired activity appeared mostly among 20S^Blm10 proteasome species which account for 20% of the active proteasomes in WT. To identify the factors behind this impaired activity, we mapped and quantified known proteasome-interacting proteins. Our results suggested that some of the reduced activity might be due to the association of the proteasome inhibitor Fub1. Additionally, we provide novel evidence for the presence of nonmature and therefore inactive proteasomal protease subunits β2 and β5 in the fully assembled proteasomes.

Graphical Abstract

Highlights

• •Activity-guided profiling of proteasomes from WT, Δα3, ΔSem1 yeast strains.
• •Proteasomes amounts found by guided profiling differ from whole cell proteomics.
• •Guided profiling versus peptidase assay shows proteasome distribution differences.
• •Reduced proteasome activity linked to Fub1 inhibition and immature β2/β5 subunits.

In Brief

Proteasomes, key protein complexes in eukaryotic cells, were analyzed in various yeast strains using an activity-guided proteomic approach. We discovered notable discrepancies in proteasome distributions between mass spectrometry and peptidase activity assays, indicating substantial portion of the proteasomes have reduced activity. This reduction might be linked to Fub1 inhibition. Moreover, immature β2 and β5 proteolytic subunits were identified in fully assembled proteasomes, potentially contributing to the decreased activity observed.

The proteasome is a key player in the ubiquitin proteasome system of all eukaryotic organisms. It is responsible for the controlled degradation of proteins in almost all types of cells. The proteasome is an energy-dependent self-compartmentalized protease complex present both in the cytoplasm and nucleus of the cell. It degrades ubiquitin-modified, damaged, unfolded, and misfolded proteins in a well-regulated manner. In eukaryotic cells, the proteasome population predominantly contains three types of proteasomes—30S, 26S, and 20S (Fig. 1A top). The 20S can also form a complex with the proteasome activator Blm10 (1)—termed here 20S^Blm10 (Fig. 1A bottom).

Fig. 1.

Activity-based proteomics to study active proteasome structures.A, the major types of proteasomes: 30S, 26S, 20S, and 20S^Blm10. The core particle 20S-CP is in red (alpha subunits are in shades of muted rose and the beta subunits in shades salmon pink), the regulatory particle—19S-RP is in blue, and Blm10—the proteasome activator is in green. B, the analysis workflow included side-by-side proteomic analyses of whole cells and activity-guided proteasome profiling for Δα3, ΔSem1, and WT strains. The activity-guided profiling utilized an in-gel activity assay, involving native protein extraction and analysis via a clear native gel with a fluorogenic peptide substrate. Proteasome peptidase activities were visualized under UV illumination. Concurrently, label-free proteomics compared the whole proteomes of the WT, ΔSem1, and Δα3 strains, following SDS-based protein extraction and SDS-PAGE. The slices of only active bands from the proteasome profiling and of the entire proteome lanes underwent in-gel tryptic digestion and were then analyzed by LC-MS to quantify the expression levels of various proteasome subunits. RP, regulatory particle; Sem1, suppressor of exocytosis mutation 1.

Structurally, all these active forms of the proteasome contain a common core particle (CP) called 20S. The 30S proteasome contains two regulatory particles—19S subcomplexes at both sides of its 20S core; the 26S proteasome contains one 19S at a single side of its 20S; and the 20S proteasome does not contain any regulatory particles. The 19S regulatory particle (RP) is mainly responsible for recognizing polyubiquitinated proteins, releasing ubiquitin chains, unfolding substrate proteins, and regulating the entry of the substrate into the 20S complex. The subunits of the 19S can be divided into subcomplexes termed ”base” and “lid” (2). The 20S core particle is the catalytic chamber that degrades the substrate into short peptides and amino acids with the help of the six catalytic subunits within it (3). The distribution of the different active proteasome species may vary under differing conditions (4). Attempts to determine this distribution using a variety of methods have led to different results. For example, SDS-polyacrylamide gel electrophoresis (PAGE) immunoblot analysis of the total amounts of proteasome subunits in Saccharomyces cerevisiae suggested that the majority of proteasomes are present as 26S or 30S (5), while a proteomics study based on affinity-tagging and purification of proteasomes from mammalian cells showed that around 50% of all proteasome complexes are 20S (6).

Mass spectrometry (MS) based structural and protein-protein interactions analyses of the proteasome holocomplex have been the focus of many research studies (7). Detailed structural characterization of the proteasome complex begins with its purification. Since the proteasome is an active protease complex, the challenge is to preserve its enzymatic activity for further functional studies while still obtaining a high purification yield. Given the high heterogeneity of proteasomes and the differences in stability and dynamics of proteasome interacting proteins (PIPs), the experimental conditions and the type of strategy used in the purification process highly influence the purity and composition of complexes obtained (30S, 26S, 20S proteasomes, and 20S with other activators), as well as the nature of the interactors purified (specific, stable, and/or dynamic).

Studies of native forms of yeast proteasome holocomplexes have been performed using 2D-gel electrophoresis by blue native PAGE followed by SDS-PAGE (8). This strategy, however, suffers from the reproducibility and complexity issues common to 2D-gel based methods. Affinity purification methods (such as immunoprecipitation, immunochromatography, and epitope tagging strategies), while costly and technically challenging, can be used as an alternative to generate a high yield of pure and functional proteasome complexes from a limited amount of starting material. Several affinity purification approaches have been developed over the past decade to purify proteasome complexes and these approaches differ in several ways: the bait protein used to catch the proteasome complexes (9), the use of an epitope-tagged proteasome subunit (10), or the use of antibodies against proteasome subunits (11). In addition to the above single-step purification approaches, tandem purification strategies have also been developed to ensure the quality of the purified proteasome complexes for downstream studies (12).

When the isolated proteasome complexes by above methods are combined with MS-based proteomics they can provide additional insight in relation to the dynamic variations of the different proteasome complexes and their PIPs networks (13). Prior to MS based analysis the proteasome complexes can be chemically cross-linked before cell lysis followed by one-step or tandem affinity purifications (10). For quantitative analysis of the different proteasome complexes and the PIP either chemical labeling (14), stable isotope labeling by amino acids in cell culture labeling (12), or label-free (6) approaches were followed. Beyond enabling the identification of the proteasome complex interactors, these proteomic studies also enable the characterization of posttranslational modifications on proteasome complex subunits and potentially on their interacting partners. This quantitative proteomics approach could moreover allow the identification of transient and labile interactors, the measurement of proteasome complex dynamics in terms of composition and subcellular localization, and the determination of proteasome complex stoichiometry.

While the proteomic studies of proteasome complexes described above were successful, a lot of information may have been missed due to technical limitations. Many of these studies were based on affinity-based purification of the proteasomes, and as such their outcome was highly dependent on the specific proteasome subunit used for the purification (either by the addition of an affinity tag or by a specific antibody). Blue native PAGE is a simple method to separate and analyze protein complexes; however, it cannot provide information on proteasome activity. To overcome this gap in existing information, our primary aim was the development of a straightforward proteomic methodology that is based on common proteasome peptidase activity assays. This was used to study the changes in the yeast proteasome within two yeast strains generated by the deletion of two integral subunits of 26S proteasomes: suppressor of exocytosis mutation 1 (Sem1) and α3 (Pre9), hereafter, referred to as ΔSem1 and Δα3, respectively. These strains were selected due to their distinctive subunit composition and active proteasome complex distributions (15, 16). Sem1 is a multifunctional and intrinsically disordered small (8 kDa) acidic protein in S. cerevisiae that is conserved among eukaryotes. It is one of the 19S lid proteins (17) and is also crucial in its assembly (15). Sem1, while not an essential gene, regulates exocytosis and pseudohyphal differentiation in yeast (18). In S. cerevisiae, the α3 subunit gene is the only nonessential CP gene (16, 19). Yeast lacking α3 form an alternative proteasome by substituting an additional α4 subunit where α3 normally resides (16). Interestingly, similar alternative proteasomes that contain two copies of α4 without any α3 have also been observed in mammalian cells under certain stress conditions (20).

To examine changes in the proteasome among these strains relative to the WT we performed side-by-side proteomic characterizations of the whole cell and activity-guided proteasome profiling (Fig. 1B). Our analyses indicate that the amount of active proteasome complexes in the cells does not necessarily correspond to the cellular abundance of proteasome subunits. We determined the relative amount of each active proteasome species within each of the studied strains, investigated the distribution of various known PIPs, and assessed the maturation of the proteolytic subunits of the proteasome across different active proteasome complexes.

Experimental Procedures

Reagents

All reagents were ordered from Sigma-Aldrich unless specified otherwise.

Yeast Strains and Growth Conditions

All single yeast strains were purchased from Euroscarf:

WT: MATa BY4741; MATa; ura3Δ0; leu2Δ0; his3Δ1; met15Δ0 (Euroscarf # Y00000)

Δα3: MATa ura3Δ0; leu2Δ0; his3Δ1; met15Δ0, YGR135w::kanMX4 (Euroscarf # Y04765)

ΔSem1: MATa; ura3Δ0; leu2Δ0; his3Δ1; met15Δ0; YDR363w-a::kanMX4 (Euroscarf # Y04200)

Yeast strains were cultured at 30 °C in yeast extract peptone dextrose (YPD) or synthetic defined (SD) media according to the requirements of different experiments. YPD medium consisted of 1% yeast extract, 2% Bacto Peptone (Difco), and 2% dextrose (http://cshprotocols.cshlp.org/content/2015/9/pdb.rec085902). SD media consisted of 0.7% yeast nitrogen base supplemented with amino acids, adenine, and 2% dextrose (https://cshprotocols.cshlp.org/content/2015/9/pdb.rec085902).

Sample Preparation for Whole Proteome Analysis

Yeast cells sample preparation for whole-cell proteome analysis was performed as described before (21, 22). In brief, 5-mL yeast cultures of each strain were grown until A₆₀₀ reached 1.5. The cells were collected by centrifugation, washed twice with cold double-distilled water, and once in 500 μl 20% (v/v) trichloroacetic acid (TCA). Following 5 min centrifugation at 4000 rpm at room temperature, the cell pellet was resuspended in 100 μl 20% (v/v) TCA, glass beads were added, and the mixture was vortexed vigorously for 4 min. The supernatant was collected, and the beads were washed twice with 7% TCA to retrieve the remaining proteins. The supernatants from all steps were pooled and placed on ice for 45 min. Next, the samples were centrifuged for 10 min at 13,000 rpm (TCA precipitation) at 4 °C, and the supernatant was discarded. The pellets were washed twice with ice-cold 0.1% TCA and then resuspended in a Laemmli loading buffer. Equal protein amounts of each sample were separated by 12% SDS-PAGE and stained with Coomassie. The entire protein lane of each sample was cut into three horizontal gel pieces and processed by in-gel trypsin digestion procedure as described before (21) followed by peptides desalting using C18 StageTip (23).

Native Lysis Method

Cell pellets were collected by centrifugation of 20 ml culture of each strain in SD media at A₆₀₀ = 1.5. Pellets were washed twice in 1 ml of 25 mM Tris pH 7.4, 10 mM MgCl₂, 10% glycerol, 1 mM ATP (BDL), and 1 mM DTT (Buffer A). Next, each pellet was dissolved in 200 μl Buffer A, and 150 μl of glass beads were added. The samples were vortexed vigorously for 1 min and then placed on ice for 1 min. This cycle was repeated 8 times in total. After lysis, the supernatants were collected using a gel-loading tip and centrifuged at 14,000 rpm for 10 min at 4 °C. The clear supernatants of each sample were collected.

Native Gel, In-Gel Activity Assay

Native lysates (150 μgr) were run in 4% native PAGE in native running buffer (100 mM Tris base, 100 mM boric acid, 1 mM EDTA, 2.5 mM MgCl₂, 0.5 mM ATP (BDL), and 0.5 mM DTT) for 2 h at 130 V. For In-gel activity assay, the gel was incubated in 20 ml Buffer A supplemented with 20 μM Suc-LLVY-AMC (R&D Systems) at 30 °C for 10 min and imaged under UV light. To visualize 20S and 20S+Blm10 bands, the gel incubated in 20 ml Buffer A supplemented with 0.02% of SDS for an additional 10 min at 30 °C.

Excision of Gel Bands of Native Gels

Active proteasome bands from 4% native gel were incubated with Suc-LLVY-AMC substrate and then excised under UV light and collected into clean tubes. For different strains, the blade was changed or cleaned thoroughly with 70% alcohol. Excised gel pieces were incubated with destaining solution at room temperature on a shaker overnight before in-gel trypsin digestion procedure as described before (21). The resulting peptides were desalted using C18 StageTip (23).

Comparison of Protein Extraction on Proteasomal Subunit Identification

WT strain cultures were grown in YPD media until an A₆₀₀ of 1.5 was reached. The cells were harvested by centrifugation, washed with ddH₂O, and each culture was then divided into two equal-volume aliquots. Proteins from one aliquot were extracted under denaturing conditions using TCA and SDS, as detailed in the “Sample Preparation for Whole Proteome Analysis” section. Proteins from the second aliquot were extracted under native conditions, as described in the “Native Lysis” section. The supernatant from each extraction was mixed with Laemmli loading buffer, and equal amounts of protein from each sample were briefly separated on a 12% SDS-PAGE, followed by Coomassie staining. Each lane of the gel was then subjected to in-gel tryptic digestion, followed by liquid chromatography with tandem mass spectrometry (LC-MS/MS) analysis.

LC-MS/MS Analysis

Whole proteome and activity-guided proteasome profiling samples were analyzed using a Q-Exactive HF mass spectrometer (Thermo Fisher Scientific) coupled to Easy nLC 1000 (Thermo Fisher Scientific). The peptides were resolved by reverse-phase chromatography on 0.075 × 180 mm fused silica capillaries (J&W) packed with Reprosil reversed-phase material (Dr Maisch; GmbH, Germany). Peptides were eluted with a linear gradient of 6 to 30% acetonitrile 0.1% formic acid for 60 min for active proteasomes profiling and 120 min for whole proteome analysis. In both cases these gradients were followed by a 15 min gradient of 28 to 95% acetonitrile 0.1% formic acid and a 15 min wash at 95% acetonitrile with 0.1% formic acid in water (at flow rates of 0.15–0.2 μl/min). The MS analysis was performed in positive mode using a range of m/z 300 to 1800, resolution of 60,000 for MS1 and 15,000 for MS2, using, repetitively, full MS scan followed by higher-energy collisonal dissociation of the 18 or 20 most dominant ions selected from the first MS scan with an isolation window of 1.3 m/z. Other settings used were as follows: a dynamic exclusion of 20 s, normalized collison enegry = 27, minimum automatic gain control target = 8 × 10³, and intensity threshold = 1.3 × 10⁵. Protein extraction comparison samples were analyzed using Exploris 480 mass spectrometer (Thermo Fisher Scientific) coupled to EvoSep One HPLC (Evosep). Samples were introduced onto the EvoTip, which was then washed twice with 20 μl of 0.1% formic acid. The washed peptides remained wet, maintained by applying 150 μl of 0.1% formic acid atop of the EvoTip until the MS analysis. The samples loaded on the EvoTips underwent chromatographic separation on a 15 cm × 150 μm analytical column, filled with 1.9-μm C18 beads (EV1106). Peptides were separated over an 88-min gradient according to the manufacturer standard method and with all other parameters consistent with the prior described procedure. The MS analysis of these samples were performed by Exploris 480 mass spectrometer (Thermo Fisher Scientific) in a positive mode (m/z 350–1200, resolution 120,000 for MS1 and 15,000 for MS2) using repetitively full MS scan followed by collision-induced dissociation (higher-energy collisonal dissociation, at normalized collison enegry = 27) of the 20 most dominant ions (charges +2 to +6) selected from the first MS scan with an isolation window of 1.3 m/z. A dynamic exclusion list was enabled with an exclusion duration of 30 s.

LC-MS/MS Data Analysis

Data analysis and label-free quantification (LFQ) of the whole proteome and proteasome active complexes were carried out using MaxQuant version 2.1.3.0 (https://maxquant.org) (24, 25). Proteasomal proteins N-terminal analysis was performed based on searches with MSFragger version 3.7 (26) via FragPipe version 19.1 (https://fragpipe.nesvilab.org/). In all searches the raw files were searched against the Uniprot S. cerevisiae searches of Jan 2020 (which include 6060 protein sequences). MaxQuant searches were performed using tryptic digestion mode with a minimal peptide length of seven amino acids. Search criteria included up to two missed cleavages, precursor and fragments tolerance of 20 ppm, oxidation of methionine, and protein N-terminal acetylation set as variable modifications. Default settings were used for all other parameters. Candidates were filtered to obtain a false discovery rate of 1% at the peptide and the protein levels. No filter was applied to the number of peptides per protein. For quantification, the match between runs module of MaxQuant was used with the LFQ normalization method enabled. For quantification, the match between runs modules of MaxQuant was used, and the LFQ normalization method was enabled (25) with default settings using LFQ quantification and match-between-runs options. MSFargger searches for proteasomal protein N-terminal cleavage were performed by using semi-tryptic digestion. Search criteria included precursor and fragments tolerance of 20 ppm with oxidation of methionine, and protein N-terminal acetylation set as variable modifications.

Experimental Design, Bioinformatics Data Analysis, and Statistics

The proteomic analyses were done using triplicate cell cultures (biological replicates). Biological replicates were individual cultures grown and prepared separately. Once these cultures reached appropriate A₆₀₀, they were treated as described.

Data analysis was carried using Perseus version 2.0.7 (https://maxquant.net/perseus/) (27). Perseus was used to generate principal component analysis plots, hierarchical clustering, and volcano plots. The multi sample ANOVA and two-sample T-tests were all performed with Perseus using the default settings (S0 = 0.1, false discovery rate = 0.05) unless otherwise specified. In the activity-guided proteasome profiling experiments the calculation of the distribution of active proteasome complexes in each strain was based on the sum of MaxLFQ intensities of all the 20S subunits in each sample. For example, in the WT proteasome distribution calculations the MaxLFQ intensities of all alpha and beta subunits were summed for each one of the four complexes (20S, 26S, 30S, and 20S^Blm10) in each one of the biological repeats (n = 3). The relative amount of each complex was calculated by dividing its sum of MaxLFQ intensities by the sum of MaxLFQ intensities obtained for all complexes. PIPs distribution calculations were conducted in a similar way but, in this case, the relative amount was based on the MaxLFQ intensity of the PIP itself. For Sem1 and Cuz1 the calculations were done using LFQ intensities and not MaxLFQ intensities.

Structural Analysis

The coordinates of the Protein Data Bank structures 5cz4 (WT yCP), 5fgi (yCPβ1-T1A-β2-T1A:carfilzomib), and 5cz7 (yCPβ5-T1A-K81R:bortezomib) (28) were superimposed as is with no further forced alignment done using ChimeraX version 1.3 (https://www.cgl.ucsf.edu/chimerax/) (29). For clarity, only the 20S core is shown. The remnants of prodomain β1 (residues −16 to 0) and β2 (residues −11 to 0) from structure 5fgi and the remnant of prodomain β5 (residues −4 to 0) from structure 5cz7 were highlighted in cyan (β1), green (β2), and yellow (β5). The rest of the 20S core is shown in red.

Results

Workflow

In-gel activity assay of the proteasome is commonly used for characterization of the levels of different active proteasomes in the studied sample. It is based on a clear native gel that is immersed in a buffer containing a fluorogenic peptide substrate. The peptidase activities are visualized in the gel under a UV illuminator (30). We took advantage of this assay's simple and quick proteasome separation and specifically focused on the detection of active proteasome complexes. We then proceeded to excise the active proteasome bands for in-gel digestion, followed by LC-MS/MS analysis (Fig. 1B).

The isolation of active proteasome species from a native gel has several advantages: it allows direct characterization of the different proteasomal species obtained under native conditions without the need to isolate them by other methods which may lead to biased results, such as purification of overexpressed and/or tagged proteasomal species; it allows concurrent characterization of different proteasomal species present in the same sample; the usage of quantitative proteomics enables an accurate determination of the relative abundance of the proteasomal species which does not depend on the efficiency of the activity assay, hence overcoming biases resulting from different efficiencies of the activity assay for different proteasomal species (30).

Whole Proteome Analysis Shows Elevation in Proteasomal Proteins in Mutant Strains Relative to the WT

As a reference for our activity-guided proteasome characterization we utilized label-free proteomics to compare the whole proteomes of WT, ΔSem1, and Δα3 strains and to determine the expression level of the different proteasome subunits.

Following common MS/MS data analysis steps, a total of 2597 proteins were identified and quantified (Supplemental Table S1). Principal component analysis of LC-MS/MS data of three repeats from each strain shows distinct clusters for each strain, indicating that the strains differ with respect to the whole proteome content (Supplemental Fig. S1).

Following multisample ANOVA statistical analysis 1228 proteins with significant abundance changes were identified. Hierarchical clustering of these 1228 proteins shows that they can be grouped into six distinct clusters with different expression profiles across the three strains (Fig. 2A). Gene ontology terms enrichment of these proteins (Supplemental Table S2) reveals several defined categories with significant changes. As shown in Figure 2A, one of these categories is proteasomal proteins, the levels of which are elevated in the mutants with respect to the WT. We performed a two-way t test to further compare each of the mutant strains with the WT (Fig. 2, B and C). Each of the mutant strains showed significant differences in the expression of around 600 proteins relative to WT. The amount of most proteasomal proteins is elevated in both mutants when compared against the WT (Fig. 2, B and C top panels). In order to test the overall changes in the 19S and 20S subcomplexes, we summed the LFQ intensities of the proteins that make up these subcomplexes in each of the mutants and compared them to the WT values (Fig. 2, B and C bottom panels). The results indicate that the increase in the expression level of the proteins that make up 19S in the Δα3 strain is higher relative to the expression of the proteins that make up the 20S (Fig. 2B bottom). In the case of ΔSem1, the change (relative to the WT) in the amount of the proteins that make up the 19S or the 20S is about the same (Fig. 2C bottom).

Fig. 2.

Whole cell proteomic data analysis of WT, ΔSem1, and Δα3.A, hierarchical clustering of the 1228 proteins that show significant abundance changes across the different strains. Lower cluster (marked) in light blue is of proteolysis-related proteins. B, top panel: t test results of the comparison of WT and Δα3 strain. CP subunits are shown in red and RP in blue. Known PIPs are shown in green. Bottom panel: average fold change of proteasome CP and RP subunits in Δα3 relative to WT. C, top panel: t test results of the comparison of WT and ΔSem1 strain. CP subunits are shown in red and RP in blue. Known PIPs are shown in green. Bottom panel: average fold change of proteasome CP and RP subunits in ΔSem1 relative to WT. CP, core particle; PIPs, proteasome interacting proteins; RP, regulatory particle; Sem1, suppressor of exocytosis mutation 1.

Proteomic Characterization of the Active-Proteasome Complexes

The abundance of the different proteasome subunits in the whole cell lysate does not necessarily represent their levels within the different active proteasome complexes. To characterize the proteomic changes in the active proteasome complexes in these strains we used native gel, excised the active bands of each strain, and submitted them to LC-MS analysis following in-gel digestion (as in Fig. 3A). If no active band was observed, we used the WT strain bands (analyzed on the same gel) as a reference and excised the corresponding regions from all lanes (Supplemental Table S3). Next, to characterize strain-related proteasomal content, we used LFQ and compared the abundance of all active proteasomes of each mutant relative to those of the WT (Fig. 3A). In the case of Δα3 strain, the relative abundance of all 20S subunits except Pre9 (α3 itself) and Pre6 (α4) is significantly lower than the WT (Fig. 3B). An additional copy of Pre6 (α4) compensates for the absence of α3 in this strain and forms a complete α-ring (16, 31). Therefore, its unique abundance change among the other 20S subunits is anticipated.

Fig. 3.

Activity-guided proteomic profiling of total active proteasome content.A, activity-guided analysis strategy for relative comparison of the total amount of proteasome complexes between strains. B, changes in proteasome subunits content of different proteasome complexes in Δα3 relative to WT. Top panel: t test results of the comparison of WT and Δα3 strain. CP subunits are shown in red and RP in blue. Known PIPs are shown in green. Bottom panel: average fold change of proteasome CP and RP subunits in Δα3 relative to WT. C, changes in proteasome subunits content of different proteasome complexes in ΔSem1 relative to WT. Top panel: t test results of the comparison of WT and ΔSem1 strain. CP subunits are shown in red and RP in blue. Known PIPs are shown in green. Bottom panel: average fold change of proteasome CP and RP subunits in ΔSem1 relative to WT. CP, core particle; PIPs, proteasome interacting proteins; RP, regulatory particle; Sem1, suppressor of exocytosis mutation 1.

It has been demonstrated that the absence of α3 alters the assembly, maturation, and stability of the 20S CP, reducing its quantity relative to the 19S RP. This, in turn, results in a surplus of assembled 19S particles, increasing the relative proportions of active 26S and 30S proteasomes in this strain (31). In the ΔSem1 strain the relative abundance of the 19S subunits is lower on average compared to the WT, while the abundance of the 20S subunits is higher (Fig. 3C).

This is expected, given that Sem1 is a stoichiometric component of the 19S and hence is essential for efficient lid assembly (32). This correlates with the in-gel peptidase activity assay, which shows a decrease in the amount of active 26S and 30S in ΔSem1 relative to WT (Fig. 3A and Supplemental Fig. S2). Sem1 is notably absent from these graphs, since it was identified by a single unique peptide due to its small size and typical tryptic digestion pattern, leading to its LFQ not being calculated. This single and unique Sem1 peptide was identified only in Δα3 and WT (in all three biological repeats).

As for known PIPs, the levels of Blm10 (33) and Ecm29 (34) in ΔSem1 were similar to the WT (Fig. 3C). This is in contrast to the relative abundance changes of Ecm29 in the whole cell proteomic comparison (Fig. 2C), which shows that it is significantly enriched in ΔSem1 compared to WT (2.1 ± 0.2 fold, p-value = 9.1E-5) despite the Blm10 abundance not changing significantly (Fig. 2C). Our results indicate that Ecm29 levels are significantly higher in Δα3 proteasomes compared to WT (3.4 ± 0.45 fold, p-value = 7.9E-4) while Blm10 levels are similar (Fig. 3B). In this case, the levels in the proteasome complexes correlate well with the whole cell levels that show similar trends (Fig. 2B). Ecm29 tethers the 20S to the 19S and preserves their interaction when ATP is absent (34, 35); Ecm29 is known to recognize faulty proteasomes (36, 37); and Ecm29 plays a role in proteasome changes under oxidative stress (38). The assembly issue of the 20S proteasome in Δα3, resulting in a high ratio of fully assembled 19S to 20S, is evident from in-gel peptidase activity results. These results demonstrate that the proportion of 30S and 26S proteasome complexes relative to 20S is higher in Δα3 compared to that found in WT cells. (Supplemental Fig. S2). To ensure that the Δα3 strain retains the appropriate levels of functional 30S and 26S proteasomes, it is likely that Ecm29 has been upregulated compared to WT and is now associated with these proteasome complexes. Other PIPs like the adaptor protein Cic1 (39) and the proteasome inhibitor Fub1 (40, 41) were identified and quantified but did not show significant changes between WT and Δα3 or ΔSem1. It is worth noting that Fub1 was only identified and quantified in the activity-guided proteomic profiling of the proteasomes (Fig. 3, B and C) and was not quantified in the whole cell proteome analysis. In contrast, Cuz1, the Cdc48-associated UBL/Zn-finger protein known to deliver ubiquitinated substrates to the proteasome (42, 43), exhibited higher expression in Δα3 and ΔSem1 relative to WT in the whole proteome analysis (7.1 ± 2.1 fold, p-value = 0.015 and 5.3 ± 2.1 fold, p-value = 0.01; Fig. 2, B and C, respectively). However, its identification by the activity-guided proteasome profiling was much weaker, preventing quantification under the analysis settings that were used (Supplemental Table S3). These results collectively suggest that the expression levels of proteasomal proteins and PIPs in cells significantly differ from their amounts in the assembled active proteasomes. The observed variations may be attributed to the different extraction methods used: denaturing for total proteome analysis and native for activity-guided proteasome profiling. To investigate this further we conducted an additional experiment to examine the quantities of proteasome subunits extracted from the same WT yeast cultures using both denaturing and native methods (Supplemental Fig. S3A). This comparison revealed that native extraction enriches 20S subunits while reducing the amount of all 19S subunits (Supplemental Fig. S3B and Supplemental Table S4). This discrepancy could stem from the presence of inactive or damaged 26S and 30S proteasomes in insoluble protein deposits (44, 45) excluded during native extraction, while the assembled and active 20S complexes remain soluble and are included in the analyzed samples. Conversely, the LFQ quantitative comparison (Supplemental Fig. S3C and Supplemental Table S7) of WT proteasome subunits in our whole proteome studies (Fig. 2, Supplemental Table S1) compared to the WT subunits in our activity-guided proteasome profiling (Fig. 3, Supplemental Table S4) revealed a markedly different scenario. This comparison showed relatively higher amounts of both 20S and 19S subunits in the activity-guided profiling (Supplemental Fig. S3C and Supplemental Table S5). Together, these results indicate that the differences in proteasome proteins and PIPs abundances observed between the activity-guided profiling and whole proteome studies do not stem from the protein extraction methods used in each experiment.

Relative Amount of Active Proteasome Species in Each Strain

We performed a pairwise LFQ comparison of the different active proteasome complexes relative to the 26S species (Fig. 4A and Supplemental Table S6) to characterize the changes in proteasome subunits and PIPs in each active proteasome species within a strain. The comparison of WT 26S to 30S proteasomes showed that the amount of CP subunits is higher in the 26S species than in the 30S species (Fig. 4B), with an average fold change of ∼2. The intensities of the RP subunits were very similar in both species, and since the 30S species contains two RP subcomplexes, the results suggest that there is twice as much 26S than 30S in WT cells. In the comparisons of WT 26S to 20S (Fig. 4C) and 26S to 20S^Blm10 (Fig. 4D) the expected absence of the 19S subunits from the 20S samples is reflected. In the 20S sample, there are more 20S subunits (1.33 fold) than in the 26S sample, while in the 20S^Blm10 sample there are less (0.77 fold). Abundance changes were also observed for Ecm29 (0.0011 fold, p-value = 9.5E-5) but its lower level in the 26S relative to the 30S is anticipated given its roles in the assembly of the 20S and 19S species and in proteasome quality control. The higher amount of Blm10 in the 26S (∼15 fold, p-value = 0.007) sample suggests that some of the 26S proteasomes are hybrid proteasomes that have Blm10 on one side of the 20S and 19S on the other as reported previously (1). This indicates that these two species (26S, and 26S^Blm10) could not be resolved in our activity gel.

Fig. 4.

Distribution of proteasome complexes within each strain.A, analysis strategy of proteasome complexes within each strain based on the relative comparison to the 26S band of each strain. B, comparison of proteins identified in WT 26S band to those that were identified in 30S. CP subunits are shown in red and RP in blue. Known PIPs are shown in green. C, comparison of proteins identified in WT 26S band to those that were identified in 20S. CP subunits are shown in red and RP in blue. Known PIPs are shown in green. D, comparison of proteins identified in WT 26S band to those that were identified in 20S+Blm10 band. CP subunits are shown in red and RP in blue. Known PIPs are shown in green. E, comparison of proteins identified in ΔSem1 26S band to those that were identified in ΔSem1 26S-Lidless band. CP subunits are shown in red and RP in blue. Rpn subunits that show 2 to 3 fold change are labeled cyan and those that show >4 fold change are marked in purple. Known PIPs are shown in green. F, the relative distribution of active proteasome complexes in the different yeast strains. For each strain the relative distribution of proteasome content calculated from LFQ analysis of active proteasome bands is shown on the left (marked as “MS”) and the distribution calculated based on active proteasomes band intensity in native gels is shown on the right (marked “Activity”). CP, core particle; LFQ, label-free quantification; PIPs, proteasome interacting proteins; RP, regulatory particle; Sem1, suppressor of exocytosis mutation 1.

Similar comparisons were also conducted for the active proteasomes of Δα3 (Supplemental Fig. S4 and Supplemental Table S7) and ΔSem1 (Supplemental Fig. S5 and Supplemental Table S8). In the case of the ΔSem1 strain, we also included a comparison of the lidless 26S proteasome, which is unique to this strain (Supplemental Fig. S6A). This particular proteasome species displayed distinct ratios for various proteasomal subcomplexes and subunits as shown in Figure 4E and Supplemental Fig. S6B. The 20S subunits and the 19S base Rpt1-6, Rpn1, Rpn13, and Ubp6 were approximately 1.4 times more abundant in the 26S compared to the lidless 26S. As expected, most of the lid subunits Rpn5, Rpn6, Rpn8, and Rpn9 showed significantly higher abundance in the 26S relative to the lidless 26S (about 2–3 times higher). The same trend of higher relative abundance in the 26S was observed for the base subunits Rpn2 and Rpn10, possibly reflecting Rpn2's position and its contacts with the lid subunits (46) and Rpn10's role as the hinge between the base and lid (46, 47). The most significant changes in abundance were found for Rpn7, Rpn12, and Rpn3, which were all significantly higher in the 26S band. These results match Sem1's position in the proteasome structure (Supplemental Fig. S6C) and supports its role as a chaperone for lid assembly, which involves the binding of Rpn7 and Rpn13 before they associate with the rest of the lid followed by the binding of Rpn12 to complete the lid assembly (32, 48).

In the next step, we conducted proteomic analyses of the different active proteasome complexes to determine the relative amount of each type of proteasome within each strain. For this purpose we used the LFQ intensities of all the 20S subunits as a reference since they are common to all active proteasome complexes. We calculated the relative proportion of each active proteasome species out of the total active proteasomes in each strain (Fig. 4F). The analysis revealed that the majority of proteasomes in the cell are 20S in the WT strain, (Fig. 4F left). The 20′S share of the total active proteasomes (39%) is equivalent to the amounts of the 26S and 30S proteasomes combined. Furthermore, the relative amount of the WT 26S proteasomes (27%) is approximately twice that of the WT 30S proteasomes (14%). Interestingly, WT 20S^Blm10 comprises about one-fifth of the total proteasomes in the cell, and its amount is higher than that of the WT 30S proteasomes. The MS-based distribution of active proteasomes differs significantly from the proportions based on the in-gel peptidase activity of the WT sample (Fig. 4F and Supplemental Table S9). According to the activity assay, most of the WT proteasomes are 26S (32%), while the 30S and 20S account for 27% of the total each, and the smallest portion consists of 20S^Blm10, which makes up only 14% of the total.

The MS-based quantification of Δα3 proteasomes (Fig. 4F middle) reveals a very high proportion of 30S (40%) and a very small portion of 20S proteasomes (7%). The rest of active proteasomes are split almost equally between 26S (28%) and 20S^Blm10 (25%) proteasomes. By peptidase-activity-based quantification the relative share of 30S and 26S is much larger and reaches over 90% of the proteasomes while 20S^Blm10 portion is only 5% and the 20S is only 3%. These quantifications once again highlight the 20S assembly issues due to the lack of α3. The reduction in fully assembled 20S in this strain may lead to a large excess of assembled 19S that readily reacts with any available 20S; thus, most of the active Δα3 proteasomes contain RP.

For ΔSem1, the MS-based quantification (Fig. 4F right) indicates that more than half of the proteasomes in this strain are 20S (53%) and almost none are 30S proteasomes (5%). The remainder is split between 20^Blm10 (19%), 26S (13%), and lidless-26S (10%). This confirms Sem1's role in the lid and RP assembly, which causes a shortage in assembled 19S and affects the association of the CP and RP. By peptidase-activity-based quantification the relative share of 30S and 26S is much larger and reaches over 50% of the proteasomes while the 20S and 20S^Blm10 portions are reduced to 26% and 15% respectively.

Relative Amount of PIPs in the Different Proteasome Complexes

Reduction and enhancement of proteasomal activity can be caused by the presence of additional proteins that can bind specifically to the proteasome and alter its activity (1, 36, 40, 49). Therefore, we used the same type of quantitative analysis described above (Fig. 4F) to monitor the distribution of the known PIPs. In our dataset, we identified and quantified 19 PIPs known to play essential roles in proteasome assembly, storage, or activity modulation among a multitude of ubiquitin-proteasome system related proteins. (Supplemental Table S10). Of these, we focused on PIPs that are known to affect proteasome activity (Fig. 5) and which may explain the results described above (Fig. 4F). Based on all the LFQ intensities that were obtained for each particular PIP in each proteasome complex we calculated its distribution across those complexes as described in the Experimental Procedures section.

Fig. 5.

Distribution of known proteasome-interacting-proteins. The relative amount of known PIPs in each of the active proteasome complexes was calculated from the LFQ analysis of active proteasome bands. The distribution of complexes is indicated by the 20S (right). The distribution of the PIPs among the various proteasome complexes is shown in purple for WT, in green for Δα3 and in gold for ΔSem1. The 20S and Ubp6 distributions are shown as references for the distribution of the active proteasomes and RP, respectively. LFQ, label-free quantification; PIPs, proteasome interacting proteins; RP, regulatory particle; Sem1, suppressor of exocytosis mutation 1.

Our analysis revealed that only one protein, Cic1, was exclusively associated with the 30S proteasome in all three strains. The specific binding of Cic1 to the 30S proteasome was previously demonstrated using a tagged version of Cic1 (39). However, this binding was not linked to Cic1's suggested function as a proteasomal adaptor for specific types of substrates nor to its localization within the nucleus (39).

Ecm29 specifically interacts with the 30S proteasome in both WT and ΔSem1. This interaction corresponds to the reported binding of Ecm29 to both the 20S and 19S proteasomes which links them together (34). Despite this, previous research using 2D native-PAGE showed that Ecm29 can also interact with the 26S proteasome (35), but this discrepancy may be resultant from the different methodologies used and also since the reported interaction of Ecm29 with 26S proteasomes was shown for only purified proteasomes. Despite the relatively low abundance of 30S proteasomes in ΔSem1 (Fig. 4F), Ecm29 still specifically interacted with these proteasomes. This may be caused by the lack of Sem1 generating incomplete proteasomes (such as Lidless 26S) that cannot properly bind Ecm29 despite the higher abundance of Ecm29 found in ΔSem1 cells (Fig. 2C). Surprisingly, Ecm29 was detected in all proteasome species in Δα3 except for the 20S proteasome. This could be explained by the unique nature of Δα3 and the reported ability of Ecm29 to function as a quality control factor by binding to aberrant proteasomes (such as Δα3) and ensuring their inhibition by gate closure (36).

Blm10 is known to bind to the proteasome CP in both 20S and 26S (50). In all strains tested, Blm10 was mostly or exclusively detected in the 20S^Blm10 proteasome. Unlike a previous report suggesting that the majority of WT Blm10 is bound to the 26S proteasome (1), our findings indicate that only 25% of WT Blm10 is bound to the 26S and most Blm10 is found in 20S^Blm10. Notably, Blm10 was only detected in 20S^Blm10 in the ΔSem1 strain, likely due to the assembly defect of the RP that reduces the 26S proportion compared to WT. In contrast to this, the proportion of Blm10 bound to 26S in the Δα3 strain was higher than in WT, possibly due to the altered 20S assembly that leads to much less 20S in this strain than in WT.

Although Blm10 is typically described as a proteasome activator (1, 50) it has been reported to bind to both sides of the 20S and generate an inactive form of proteasome called Blm10₂-CP (50). The formation of this Blm10₂-CP was found to be enhanced in open-gate proteasome mutants such as α3 or α7 that lack their N-terminal domain (50). In Δα3 we observed a significant difference between the proportion of 20S^Blm10 obtained based on chymotrypsin peptidase activity (5%) and that determined by MS (25%) (Fig. 4F). It is therefore possible that the gel region we assigned and excised as Δα3 20S^Blm10 may have contained a mix of 20S^Blm10 and the Blm10₂-CP.

The proteasome inhibitor Fub1 has been less studied compared to other PIPs. Fub1 has been shown to physically interact with the CP subunits (41). When comparing the relative quantities of proteasomes and Fub1 in different strains (as depicted in Supplemental Fig. S7A and B, respectively) it is evident that Fub1 is more abundant in the Δα3 strain, which corroborates recent findings (40). In WT, Fub1 is exclusively associated with the 20S proteasome, while in Δα3 Fub1 is present in both the standalone CP and CP-RP complexes with a preference for 20S and 20S^Blm10. This is consistent with previous findings for this strain (40). In the ΔSem1 strain, Fub1 is exclusively associated with 20S and 20S^Blm10 in proportions that mirror the overall distribution of these proteasomes in this strain. Hence, the presence of Fub1 could also account for some of the observed divergence between the MS-based and activity-based quantifications detailed above.

Evaluation of Proteasome Proteolytic Subunits Activation

Most proteases, including the proteasome proteolytic subunits (β1, β2, and β5), are initially translated as inactive zymogens and must undergo proteolytic processing to become active. This self-catalytic process occurs during the final steps of CP assembly after the two halves of the CP have been dimerized, and results in the exposure of the active site threonine of the proteolytic subunits (51). We wanted to test if the underestimation of some proteasome species based on their activity (Fig. 4F) might be explained by certain alterations in the maturation process of β1, β2, and β5. However, typical bottom-up proteomics approaches, which rely on conservative peptide searches that assume fully specific tryptic digestion, are not able to detect endogenous proteolytic processing such as the maturation of beta subunits. To identify proteasome proteolytic subunit activation events, we analyzed all active proteasome samples using a peptide search based on semitryptic cleavage specificity (Supplemental Table S11). Through this method, we identified several signature peptides that span the cleaved propeptide domain (Fig. 6A), providing insight into the activation status of the proteases in each complex. This type of analysis only looks at a small and specific group of peptides so it's likely that no signature peptide will be identified if the level of the studied complex is low. The MS/MS identifications of proteolytic beta subunits signature peptides had high scores, and high peptide sequence coverage therefore provided conclusive evidence for their presence in the different samples (Supplemental Figs. S8 and S9). We also identified semitryptic signature peptides that indicate the maturation of β6 and β7 subunits.

Fig. 6.

Activation of proteasome proteolytic subunits.A, tryptic and semi-tryptic peptides that span the propeptide domain of the proteasome proteolytic subunits. The identification of these signatures indicates the activation status of the proteolytic subunits: β1, β2, and β5. The propeptide domain is shown in yellow and the active protease chain is shown in gray. The catalytic threonine is shown in red. B, the number of identifications of the different proteasome protease prodomain peptides among the various proteasome complexes. WT in purple, Δα3 in green, and ΔSem1 in gold. The size of the dot represents the number of biological repeats in which the peptide was identified. Peptides that indicate activation of the beta subunits are in lines above the partition line and peptides that indicate the presence of the uncleaved prodomain are below the line. C, yeast proteasome structures that include proteolytic subunits propeptide domains. The structures are from (28) that were aligned and overlaid using ChimeraX (29). For clarity, only the 20S core (in red) is shown from different viewpoints. The remnants of prodomain β1 (residues Gly3 to Gly19, in cyan) and β2 (residues Asn18 to Gly29, in green) are from structure 5fgi and the remnant of prodomain β5 (residues Ile71 to Gly75, in yellow) from structure 5cz7. Sem1, suppressor of exocytosis mutation 1.

As expected, active β1, β2, and β5 subunits were identified in all WT samples, except for β5 in 20S^Blm10 samples (Fig. 6B, top lines). A similar trend was observed for ΔSem1, in which we identified at least two different active proteolytic subunits in all proteasome complexes. We found less active proteases in the Δα3 complexes. This was most noticeable in the Δα3 20S sample, where there was no indication of any active β subunit. Conversely, the signature peptides which indicate the presence of nonmature and inactive β subunits (Fig. 6B bottom lines) were identified in all proteasome complexes and strains. Nonmature β2 (lacking the trypsin-like activity) was identified in all WT, ΔSem1, and Δα3 strains while nonmature β5 (lacking the chymotryptic-like activity) was identified in WT 20^Blm10, ΔSem1 20S, and in all Δα3 complexes aside from the 26S. Of note, the presence of nonmatured β2 and β5 was particularly prevalent in Δα3 20^Blm10 where all four possible nonmature signature peptides appeared in all the biological repeats.

In the CP assembly process the autocatalytic cleavage of the β subunits prodomain is accompanied by the degradation of the proteasome-assembly factor UMP1 (51, 52). We were able to identify UMP1 peptides only in some of the Δα3 proteasome samples (Supplemental Table S9). This fits well with the very low activity observed for this species (Supplemental Fig. S2) and its relative high abundance determined by MS (>25%, Fig. 4F). Evidence for nonmatured 20^Blm10 was found for all three strains indicating that this may be a common character of this complex and may also contribute to its altered activity. Interestingly, structural modeling indicates the presence of the prodomains of the beta subunits that does not alter the overall structure of the 20S proteasome. Furthermore, these additional chains can be easily accommodated within the 20S proteasome's cavity (Fig. 6C).

Discussion

The proteasome is the primary contributor to the proteolysis of most cellular proteins in eukaryotes. Consequently, any fluctuations in the levels of cellular proteasome complexes can disrupt protein homeostasis. To investigate such changes in active proteasome ratios within cells we present a straightforward yet potent method for activity-guided MS-based proteomic profiling of proteasome complexes. This method relies on peptide substrates, preserves the native and active forms of proteasome, and separates these various active complexes using clear native PAGE.

This concept is not new; in fact, it was one of the first methods used for the initial characterization of the proteasome as an ATP-dependent proteolytic complex (53). Over the years, the PAGE separation process and the activity measurements have been streamlined and standardized (54); so, it is easy to implement this method in any biochemical or cell biology laboratory. Overall, this approach is simple, cost-effective, quick, and can be broadly applied to cultured cells or tissue samples as sources of proteasomes.

In this study, we used this method to determine the composition and relative abundance of each type of active proteasome in different yeast strains. In the WT strain we found that the 20S proteasome was the most abundant, with a relatively large quantity of 20S^Blm10 species also present. This relatively higher abundance of yeast 20S compared to 26S and 30S in the WT is in accordance with MS-based studies of the abundance of constitutive proteasomes in mammalian cells (6). Aberrations in CP or RP assembly, caused by Δα3 and ΔSem1, respectively, strongly affected the distribution of the different proteasomes.

Interestingly, in all of the studied strains we observed a discrepancy between the proteasome distribution determined by MS as opposed to that determined by monitoring in-gel peptidase activity (Fig. 4F). This discrepancy was mostly noticed regarding the relative quantification of the 20S and 20S^Blm10 species. The discrepancy between MS-based and activity-based quantifications is partly related to the limitations of the peptidase activity assay. While activity-based quantifications are useful in detecting active proteasomes, they do not necessarily provide accurate data on their relative amounts. The substrate access to the CP catalytic cavity is controlled by the α subunits ring which surrounds the entry channel (55). The RP can open the entry channel allowing the fluorogenic substrate to enter more easily through the CP gate in 26S and 30S proteasomes. In contrast, 20S and 20S^Blm10 proteasomes typically require the addition of small amounts of detergent, like SDS, to promote gate opening (54). This is likely leading to altered quantification and overestimation of the 30S and 26S complexes. Furthermore, the in-gel peptidase activity assay is qualitative by nature and is read at a single time point, usually after a relatively long incubation time, which can lead to saturation and therefore inaccurate quantification.

To investigate other factors that might contribute to the observed reduction in proteasome activity of the 20S and 20S^Blm10 we characterized and quantified the PIPs associated with each proteasomal species. This analysis provided clear and novel information regarding the distribution of known PIPs among the different active proteasome complexes. Our PIP mappings suggest that changes in proteasomal activity, particularly of 20S and 20S^Blm10, may be related to proteasome inhibition by Fub1. Recent research has revealed the mechanism by which Fub1 inhibits proteasomes in Δα3 (40), but the question of how it enters the CP cavity has been left unanswered. As Δα3 proteasome is known to have an open CP gate, it was suggested that Fub1 recognizes proteasomes with an aberrant CP gate and enters through them into the matured CP while simultaneously blocking all proteolytic subunits (40). Unlike Δα3, WT and ΔSem1 proteasomes have a typical CP entry gate which is not continuously open, emphasizing the enigma of Fub1's distribution. One possible answer, suggested by Rawson et al (40), is that Fub1 might be part of a quality control mechanism that inhibits a subset of CPs with abnormal entry gates. Another option is the existence of an alternative entry mechanism for CPs with a closed entry gate.

We also identified signature peptides for the maturation of the proteolytic subunits of the proteasome, which surprisingly revealed the presence of noncleaved and therefore inactive β2 and β5 subunits in the assembled proteasomes. These uncleaved subunits may also account for the reduced activity observed in the proteasomes. The autocatalytic cleavage of the proteolytic proteasomal β subunits is a critical event in the assembly of the CP (28, 51, 56, 57). This process takes place in the final step of assembly and requires the dimerization of two half-proteasome particles, each composed of a full α ring and a full β ring (51). This sequence of events ensures that the CP chamber is sealed off from the cellular environment before its proteolytic activity is initiated.

The presence of nonmature signature peptides indicates the presence of full or partial β2 and β5 propeptide domains. It has been shown before that yeast strains carrying a single mutation at the catalytic threonine of β1, β2, or both are still viable, and that their proteasomes can accommodate the remnants of the propeptide domain (28, 58). However, outright deletion of the β5 propeptide domain or the disruption of its cleavage process leads to significant phenotypic variations (28). Here, for the first time, we show that proteasomes with disrupted or incomplete processing of β2 and β5 propeptide domains are present in WT yeast. Additional studies are required to determine the mechanism that allows proteasome assembly despite the incomplete propeptide removal.

In conclusion, we have demonstrated that our activity-guided proteomic method for studying the proteasome is a valuable addition to the proteasome research toolbox. It is simple, cost-effective, quick, and provides new and detailed information that can be obtained by the current commonly-used proteasome characterization methods.

Data Availability

The data that support the findings of this study have been deposited to the ProteomeXchange Consortium (http://proteomecentral.proteomexchange.org) via the PRIDE partner repository with the dataset identifiers: whole cell label-free analysis PXD045047; activity-guided profiling of proteasomes PXD045051; comparison of yeast protein extraction under native and denaturation conditions PXD048380. The annotated mass spectra from the MaxQuant output can be viewed via MS-Viewer (59). The search key for the whole cell label-free analysis is:0xd802n87v and the link is: https://msviewer.ucsf.edu/prospector/cgi-bin/mssearch.cgi?report_title=MS-Viewer&search_key=0xd802n87v&search_name=msviewer. The search key for the Activity-Guided In-gel: is oyqbcider4 and the link is https://msviewer.ucsf.edu/prospector/cgi-bin/mssearch.cgi?report_title=MS-Viewer&search_key=oyqbcider4&search_name=msviewer.

Supplemental data

This article contains supplemental data (29, 54, 60).

Conflict of interest

The authors declare no competing interests.