MCB
Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental material
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Major, T.
Right arrow Articles by Voos, W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Major, T.
Right arrow Articles by Voos, W.

 Previous Article  |  Next Article 

Molecular and Cellular Biology, February 2006, p. 762-776, Vol. 26, No. 3
0270-7306/06/$08.00+0     doi:10.1128/MCB.26.3.762-776.2006
Copyright © 2006, American Society for Microbiology. All Rights Reserved.

Proteomic Analysis of Mitochondrial Protein Turnover: Identification of Novel Substrate Proteins of the Matrix Protease Pim1{dagger}

Tamara Major,1,2 Birgit von Janowsky,1,2 Thomas Ruppert,3 Axel Mogk,3 and Wolfgang Voos1*

Institut für Biochemie und Molekularbiologie, Universität Freiburg, Hermann Herder Str. 7, 79104 Freiburg,1 Fakultät für Biologie, Universität Freiburg, 79104 Freiburg,2 ZMBH, Im Neuenheimer Feld 282, 69120 Heidelberg, Germany3

Received 30 June 2005/ Returned for modification 8 August 2005/ Accepted 4 November 2005


   ABSTRACT
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
ATP-dependent oligomeric proteases are major components of cellular protein quality control systems. To investigate the role of proteolytic processes in the maintenance of mitochondrial functions, we analyzed the dynamic behavior of the mitochondrial proteome of Saccharomyces cerevisiae by two-dimensional (2D) polyacrylamide gel electrophoresis. By a characterization of the influence of temperature on protein turnover in isolated mitochondria, we were able to define four groups of proteins showing a differential susceptibility to proteolysis. The protein Pim1/LON has been shown to be the main protease in the mitochondrial matrix responsible for the removal of damaged or nonnative proteins. To assess the substrate range of Pim1 under in vivo conditions, we performed a quantitative comparison of the 2D protein spot patterns between wild-type and pim1{Delta} mitochondria. We were able to identify a novel subset of mitochondrial proteins that are putative endogenous substrates of Pim1. Using an in organello degradation assay, we confirmed the Pim1-specific, ATP-dependent proteolysis of the newly identified substrate proteins. We could demonstrate that the functional integrity of the Pim1 substrate proteins, in particular, the presence of intact prosthetic groups, had a major influence on the susceptibility to proteolysis.


   INTRODUCTION
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
In a cellular environment, especially under stress conditions, proteins are at risk for being inactivated by misfolding or aggregation (7). In order to maintain cellular protein function and integrity, a protein quality control system that preserves metabolic functions under all growth conditions is required. Protein quality control machineries in the cell consist of several types of chaperone proteins and proteases. The coordinated function of these components is required to stabilize damaged or denatured polypeptides and refold or remove them to avoid the destructive effects of protein aggregation (10, 11, 50). In a eukaryotic cell, mitochondria play crucial and diverse roles in many processes, for instance, ATP production, the metabolism of nutrients, biosynthesis of the Fe/S cluster, and signaling pathways leading to programmed cell death. In addition, the oxidative metabolism of mitochondria carries the risk of harmful protein modifications even under normal conditions (6). Increasing evidence that points to an important role of mitochondrial protein defects in the pathogenesis of human diseases such as neurological disorders, aging, cancer, and various neuromuscular syndromes is available (33, 43, 48). Hence, the protection of mitochondrial protein functions is an important aspect of cellular protein quality control.

Apart from the stabilizing action of molecular chaperones, the major defense mechanism against the accumulation of damaged polypeptides is their specific removal by proteolysis. In eukaryotic cells, the major cytosolic protease is the proteasome, a large hetero-oligomeric protein complex that processively degrades ubiquitinylated proteins in an ATP-dependent reaction (11, 25). While removal of damaged polypeptides of the endoplasmic reticulum is mainly achieved via the proteasome (15, 16), mitochondria contain their own proteolytic systems related to those found in bacterial cells (13, 27). The first ATP-dependent proteolytic activities in mitochondria had already been described more than 20 years ago (reviewed by Goldberg [12]). Indeed, the proteolytic enzymes of mitochondria belong to the superfamily of AAA+ proteins (ATPases associated with a variety of cellular activities) (23), which comprises many proteins with chaperone activity. An interesting aspect of the AAA-derived proteolytic systems in both bacteria and mitochondria is the interplay between protease and chaperone activities, supporting their functions as protein quality control components (39). The main protease responsible for the degradation of soluble matrix proteins is the protease Pim1 (for proteolysis in mitochondria), a homolog of the bacterial protease Lon (La) (40, 46, 49). Two other proteolytic systems, the m-AAA and the i-AAA proteases, are localized in the inner mitochondrial membrane, exposing their proteolytic sites to the matrix compartment or the intermembrane space, respectively (19). They have been implicated in the biogenesis of inner membrane protein complexes and in the degradation of nonassembled subunits (18). Pim1 is not essential for yeast survival, but it is needed for normal mitochondrial biogenesis. Yeast cells lacking the PIM1 gene lose mitochondrial DNA and are respiratory deficient (40, 46). As a member of the AAA+ protein family, Pim1 forms a large, ring-like heptameric complex. Unlike the proteasome or other ATP-dependent oligomeric proteases like the bacterial ClpP protease, ATPase and the proteolytic domain are located on the same polypeptide chain in Pim1 (38, 44). In this respect, Pim1 resembles a class of membrane-associated proteases found in bacteria and mitochondria, FtsH, Yme1, Yta10, and Yta12 (18, 42). Experiments using artificial reporter proteins indicate that Pim1 shows some preference for the degradation of misfolded, missorted, or nonassembled proteins in the mitochondrial matrix (29, 31, 32, 47). However, the role of Pim1 during mitochondrial protein quality control in vivo has not been established so far. In particular, the information about the extent of mitochondrial protein turnover and the identities of endogenous proteins that are prone to degradation is very limited. Since no covalent tagging mechanism like the ubiquitin system of eukaryotic cytosol has been identified for mitochondria, the molecular mechanisms that target proteins for proteolysis by Pim1 have not been established so far. In addition to its proteolytic activities, Pim1 protease has chaperone-like properties similar to those of membrane-bound AAA+ proteases (27). The chaperone-like interaction with nonnative proteins might be of relevance for the so-far-unknown molecular mechanism of substrate recognition. In mitochondria and in bacterial cells, chaperones of the mitochondrial Hsp70 and Hsp100 families have been found to cooperate with Pim1/LON-like proteases during degradation of misfolded polypeptides (2, 29, 35, 47).

A comprehensive understanding of the protein functions in eukaryotic cells with their compartmental organization requires an analysis of the protein composition and dynamics at the subcellular level (8, 17). Recently, the first proteomic approaches have been employed to establish the general protein content of purified mitochondria (22, 26, 37, 41). However, the mitochondrial proteome has to be considered as a nonstatic entity that shows characteristic changes according to the functional state of the cell. Therefore, proteomic techniques are ideally suited to monitor the dynamics of the organellar protein profile due to the adaptations to environmental conditions. In yeast, experimental studies have started to address this question by analyzing quantitative alterations in protein synthesis, protein modifications, and protein composition that occur in response to genetic or environmental changes at the cellular and mitochondrial levels (24).

In order to study the role of protein degradation in the mitochondrial protein quality control system, we utilized proteomic techniques to monitor changes in the mitochondrial proteome that are caused by proteolytic reactions. The protein composition was analyzed under different conditions by a quantitative comparison of the protein spot patterns obtained by two-dimensional (2D) polyacrylamide gel electrophoresis (PAGE). Potential degradation substrates were identified by mass spectroscopy. By this approach, a comprehensive overview of the protein turnover reactions in mitochondria was obtained under in vivo conditions. Endogenous mitochondrial substrate proteins of the matrix protease Pim1 were identified and validated by in organello degradation assays using imported radiolabeled substrate proteins.


   MATERIALS AND METHODS
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Yeast strains and isolation of mitochondria. The following Saccharomyces cerevisiae strains were used in this study: WVY38 (wild type [WT]) (his3-{Delta}200 leu2-{Delta}1 ura3-52 trp1-{Delta}63), KRY01 (WT rho–) (MAT{alpha} ade2-101 his3-{Delta}200 leu2-{Delta}1 ura3-52 trp1-{Delta}63 lys2-801 rho–), and KRY02 (pim1{Delta}) (MAT{alpha} ade2-101 his3-{Delta}200 leu2-{Delta}1 ura3-52 trp1-{Delta}63 lys2-801 pim1{Delta}::HIS3) (29). The following yeast strains containing the Fe/S cluster assembly mutations were obtained from the Euroscarf strain collection: BY4741 (WT) (MATa his3{Delta}1 leu2{Delta}0 met15{Delta}0 ura3D), Y05278 (ssq1{Delta}) (MATa his3{Delta}1 leu2{Delta}0 met15{Delta}0 ura3{Delta} YLR369w::kanMX4), and Y04889 (nfu1{Delta}) (MATa his3{Delta}1 leu2{Delta}0 met15{Delta}0 ura3{Delta}0 YKL040c::kanMX4). Mitochondria were isolated from these strains according to previously published procedures (30) after growth at 30°C in YP medium (1% yeast extract, 2% Bacto peptone) containing 3% glycerol (WVY38) or 2% glucose (KRY01, KRY02, BY4741, Y05278, and Y04889).

Two-dimensional electrophoresis. Mitochondria (250 µg) were resuspended in resuspension buffer (250 mM sucrose, 10 mM MOPS [morpholinepropanesulfonic acid]-KOH, pH 7.2, 80 mM KCl, 5 mM MgCl2) supplied with 4 mM NADH and an ATP-regenerating system (3 mM ATP, 20 mM creatine phosphate, 200 µg/ml creatine kinase) and incubated for 1 h or 4 h at 30°C or 39°C. After reisolation, pelleted mitochondria were lysed by vigorous shaking in 100 µl of lysis buffer {6 M urea, 2 M thiourea, 4% (wt/vol) 3-[(3-cholamidopropyl)- dimethylammonio]-1-propanesulfonate (CHAPS), 1% (wt/vol) dithiothreitol (DTT), 40 mM Tris, protease inhibitors} at 4°C for 5 min. Insoluble proteins were removed by subsequent centrifugation at 13,000 x g for 10 min, and the supernatant was used for two-dimensional electrophoresis. For isoelectric focusing, a solution containing 7 M urea, 2 M thiourea, 2% (wt/vol) CHAPS, 0.4% (wt/vol) DTT, 1.25% (vol/vol) ampholites (pH 3 to 10), and 0.001% (wt/vol) bromophenol blue was added to the soluble mitochondrial protein extract. Aliquots of 250 µl were applied to dry, 13-cm, nonlinear pH 3 to 10 immobiline strips (Amersham Biosciences) for overnight in-gel rehydration at 20°C in an IPGphor isoelectric focusing system (Amersham Biosciences). Reswelling was followed by protein focusing for 1 h at 200 V, 1 h at 500 V, and then 1 h at 1,000 V. Finally, the voltage was increased to 8,000 V for 6 h. After an equilibration step in buffer (50 mM Tris-HCl, pH 8.8, 6 M urea, 2% [wt/vol] sodium dodecyl sulfate [SDS], 30% [vol/vol] glycerol, 0.001% [wt/vol] bromophenol blue), proteins were separated by 10 to 14% linear gradient SDS-PAGE at 25 mA per gel for 4 h. Protein detection was performed by Coomassie brilliant blue G-250 staining. To account for experimental variations, three independent samples were assayed for each condition and analyzed together. After normalization to the total spot volume per gel, protein spots were matched and quantified using ImageMaster 2D Platinum software (Amersham Biosciences). To express differences in the protein amounts of the samples, a normalized change value was determined. First, the average spot volumes were established for the individual experimental conditions. The relative differences in the amounts were obtained by calculating the ratio (change) (x) between the spot volumes of the different experimental conditions. A normalized change value was calculated as x – 1 for ratios greater than 1 and as 1 – 1/x for ratios smaller than 1. Unchanged protein spots result in a normalized change value of 0, while positive values represent an increased protein amount and negative values represent a decreased protein amount. In contrast to a representation by ratio or percentage, the normalization procedure results in an equal representation of the relative increase and reduction in the amounts. To avoid a misinterpretation of small quantitative changes in spot volumes due to the intrinsic variations of the 2D-PAGE procedure, we assigned a confidence threshold of 0.2 for the normalized change value, below which changes in protein amounts were regarded as not significant.

Mass spectrometry and identification of proteins. Individual spots were excised after 2D-PAGE, reduced with DTT, alkylated with iodoacetamide, and digested with trypsin (5, 36) using a Digest Pro mass spectrometry (MS) liquid-handling system (Intavis AG). Following digestion, tryptic peptides were extracted from the gel pieces with 50% acetonitrile-0.1% trifluoroacetic acid, concentrated, and analyzed by MS. Tandem mass spectrometry (MS/MS) was performed on a quadrupole time-of-flight (TOF) (Q-TOF) mass spectrometer. For matrix-assisted laser desorption ionization-TOF MS (Ultraflex, Bruker), the sample was desalted using a ZipTip (Millipore) and spotted onto a steel target using cyano-4-hydroxy-cinnamic acid as a matrix. The peptide mass fingerprint (PMF) was acquired after external calibration (peptide calibration standard II; Bruker). Liquid chromatography-electrospray ionization (LC-ESI) Q-TOF MS was done using a nanoHPLC system (Ultimate [equipped with a Famos autosampler]; Dionex, The Netherlands) coupled to an ESI Q-TOF hybrid mass spectrometer (Qstar pulsar; Applied Biosystems). The column (Inertsil 3-µm C18 material, 75 µm by 150 mm; LC Packings) was connected with a nano-ESI emitter (New Objectives). A total of 2,000 V was applied via liquid junction. The Q-TOF operated in positive-ion mode. One MS survey scan (0.7 s) was followed by one information-dependent product ion scan (3 s).

For protein identification by PMF, peptide masses were labeled manually in comparison to a control sample taken from a spot of an empty area of the same gel. The PMF was searched against the NCBI-nr database using Mascot (Matrix Science). The algorithm was set to use Saccharomyces cerevisiae as taxonomy and trypsin as the enzyme, allowing for one missed cleavage site at maximum and assuming either carbamidomethyl or oxidation as a fixed modification of cysteine and methionine, respectively. Mass tolerance was set to 100 ppm. Protein hits were taken as identified if the mascot score exceeded the significance level. From LC-ESI Q-TOF MS analysis, the noninterpreted MS/MS spectra were searched against the NCBI database using Mascot software with the same settings as described above except that the mass tolerance was set to 1.1 Da and 0.1 Da for MS and MS/MS, respectively. Proteins identified by a single peptide are listed in the tables if the following criteria were fulfilled: (i) the scoring value exceeded the Mascot homology threshold, (ii) manual interpretation of the fragment spectrum resulted in a consecutive stretch of at least four amino acids, and (iii) a database search with this stretch of amino acids using that Mascot query returned the same protein. Sequence coverage, Mowse score, and number of matching peptides are given in the supplemental material (see Table S1 in the supplemental material). Proteins not listed were identified by manual comparison with previously published reference 2D gels of yeast mitochondria according to their reproducible running behavior (24, 37). All spot positions of identified protein sequences were checked for plausibility by a comparison with the published molecular weight and isoelectric point.

Degradation assay. To monitor the degradation of reporter proteins in the mitochondrial matrix, substrate proteins were synthesized by in vitro transcription and translation in the presence of [35S]methionine using rabbit reticulocyte lysate (Amersham). Radiolabeled preproteins were imported into isolated mitochondria (representing 300 µg of total mitochondrial protein) for 20 min at 25°C (29). After the import reaction, nonimported precursor proteins were digested by incubation with proteinase K (100 µg/ml) for 10 min on ice. Treatment with proteinase K was stopped by the addition of 1 mM phenylmethylsulfonyl fluoride. Mitochondria were reisolated by centrifugation for 10 min at 10,000 x g and washed with SEM buffer (250 mM sucrose, 1 mM EDTA, 10 mM MOPS-KOH, pH 7.2) containing 1 mM phenylmethylsulfonyl fluoride. Mitochondria were then reisolated and resuspended in 800 µl of resuspension buffer (250 mM sucrose, 10 mM MOPS-KOH, pH 7.2, 80 mM KCl, 5 mM MgCl2) containing 4 mM NADH and an ATP-regenerating system (3 mM ATP, 20 mM creatine phosphate, 200 µg/ml creatine kinase) or 10 units/ml apyrase and 0.1 mM oligomycin. Samples were further incubated for various time periods (0, 30, 60, 120, 180, and 240 min) at 30°C. At indicated time points, aliquots of 100 µl were taken, and after the reisolation step, the samples were analyzed by 12.5% SDS-PAGE. The stability of the protease substrates was determined by autoradiography using ImageQuant software (Amersham Biosciences) and Excel 2001 (Microsoft).


   RESULTS
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Analysis of the yeast mitochondrial proteome by 2D-PAGE. We analyzed the changes in the mitochondrial protein composition that were generated by the activity of mitochondrial proteases in intact mitochondria isolated from wild-type yeast cultures grown on a nonfermentable carbon source. The mitochondria were incubated for 4 h under native buffer conditions in the presence of an ATP-regenerating system. The protein composition was analyzed before and after the incubation period by 2D-PAGE to monitor changes in protein abundance. Separation of mitochondrial proteins by 2D-PAGE resulted in the detection of about 270 protein spots that were separated and matched reproducibly in all gels, which is well within the range of those found in similar studies (1, 37). Protein spots were excised and subjected to identification by matrix-assisted laser desorption ionization-TOF mass spectroscopy. The sequences obtained corresponded to 52 different mitochondrial proteins (see Table S1 in the supplemental material). Due to their unambiguous and reproducible separation by 2D-PAGE, the spots corresponding to the proteins Cpr3, Kgd1, Kgd2, Leu4, Mef1, Mmf1, Ssc1, and Yil157c were identified by comparison with previously published standard gels (24, 37).

In total, 60 mitochondrial proteins were analyzed, which would represent a coverage of about 11% of all mitochondrial proteins from yeasts listed in the MitoP2 database. To assess the representation of different mitochondrial protein types in our analysis, we determined the submitochondrial localization of the identified proteins and compared them with data from the MitoP2 database. A profile of our protein group revealed that 72% of all identified proteins are localized in the matrix space, 16% are localized in the inner membrane, and 3.5% are localized in intermembrane space and outer membrane, while for 5%, the localization is not known (Fig. 1A). Except for a lower relative abundance of inner membrane proteins and a higher proportion of matrix proteins, we observed a distribution that was similar to the general localization pattern of mitochondrial proteins according to the MitoP2 database (Fig. 1B). The relative lower abundance of membrane proteins is a typical phenomenon of the proteome analysis by 2D-PAGE and can be attributed to the standard solubilization procedure used.


Figure 1
View larger version (28K):
[in this window]
[in a new window]
  		FIG. 1. Submitochondrial localization and functional classification of the proteins identified from the 2D-PAGE analysis of the mitochondrial proteome. (A and B) Submitochondrial localization of proteins identified in this study (A) and according to the MitoP2 database (B). MA, mitochondrial matrix; IM, inner membrane; IMS, intermembrane space; OM, outer membrane; ND, not defined. (C and D) Functional classes of mitochondrial proteins identified in this study (C) and those included in the MitoP2 database (D). The percentage of total proteins detected within each category is indicated.

 
We have also categorized the identified proteins according to their function (Fig. 1C) and compared them with the functional distribution of all mitochondrial proteins included in the MitoP2 database (Fig. 1D). In good correlation with the sublocalization results, we found a relative enrichment of proteins with metabolic functions (52% versus 25%) that are predominantly soluble proteins of the mitochondrial matrix. Due to their localization in the inner mitochondrial membrane, we hardly detected proteins of the carrier and transporter subclasses (2% versus 8% in the database). In addition, the majority of proteins involved in protein synthesis, mainly ribosomal subunits, were not sufficiently resolved by the 2D-PAGE procedure used. From the representation of mitochondrial proteins concerning subcellular localization and functional classes, we concluded that the experimental method was particular suited for the analysis of soluble matrix proteins, being the main proteolytic targets of the protease Pim1.

Dynamic behavior of the mitochondrial proteome. To monitor the quantitative changes in the mitochondrial proteome, the spot intensities belonging to the same protein were added up, and the protein amounts before and after the incubation period were compared. Two different temperature conditions were chosen. Incubation at 30°C corresponded to normal growth conditions, while 39°C represented mild heat shock conditions. From the ratio of the protein amounts before incubation to those after incubation, the normalized change values were calculated, representing the relative differences in protein abundance before and after the incubation period. Proteins that showed a decreased amount after the incubation, presumably due to proteolysis, are depicted with negative change values (a value of –1 represents 50% reduction). All proteins identified and analyzed are listed in Table 1. The quantitative analysis revealed four classes of proteins with a differential turnover behavior. Selected proteins from these groups are shown in Fig. 2A. A small group of proteins (Ilv1, Kgd1, Mef1, Lsc1, Lys4, and Zeo1) was strongly degraded at 30°C but less affected at higher temperatures. Another group of proteins (Alt1, Atp2, Dld1, Ilv2, Ilv6, Mae1, Pil1, Rim1, and Yjl200c) was degraded at both 30 and 39°C, usually with a slightly higher rate at the elevated temperature. A third group (Atp11, Hsp60, Kgd2, Mcr1, and Phb1) showed a temperature-sensitive behavior, being only degraded at elevated temperatures. The majority of the detected proteins were stable at both temperatures. All functional classes of proteins like biogenesis components, protective enzymes, and some membrane transporters were represented in this group (Mas1, Ilv5, Ssc1, Sod2, Por1, etc.). Also, the amounts of abundant mitochondrial proteins implicated in metabolic functions like the tricarboxylic acid (TCA) cycle (Aco1, Idh2, and Mdh1) were not significantly changed under the conditions analyzed. In these cases, we often observed a small positive change value that was generally lower than the significance threshold of 0.2. Since biosynthesis of the analyzed proteins cannot occur under the chosen conditions, the positive values may arise from the normalization procedure or may be correlated with the intrinsic difficulties in quantifying highly abundant proteins or membrane proteins.


View this table:
[in this window]
[in a new window]
  		TABLE 1. List of identified mitochondrial proteins

 

Figure 2
View larger version (24K):
[in this window]
[in a new window]
  		FIG. 2. Temperature-dependent changes in the mitochondrial proteome. (A) Relative changes in protein abundance during a time course of 4 h. Purified mitochondria were incubated at the indicated temperatures under native buffer conditions in the presence of an ATP-regenerating system. The reduction of protein amounts of selected proteins during the incubation period at 30°C and 39°C is represented as normalized (Norm.) change values. As an example, a reduction of 50% would result in a negative value of 1. Error bars indicate the relative error of the change values. (B) Absolute changes in protein abundance. Shown are the differences in spot volumes before and after the incubation period in arbitrary units (AU) obtained from the Imagemaster 2D quantification of the respective protein spots. Error bars indicate the relative errors of the changes in protein amounts.

 
The analysis of the relative differences in the protein amounts, indicated by the change values (Fig. 2A), provided the identities of the proteins that were most affected by mitochondrial proteolysis. However, since small absolute changes in the amount of less abundant proteins lead to high change values, these proteins might not represent the most prominent targets of mitochondrial proteases. On the other hand, a degradation of a high number of polypeptides of an abundant protein may result only in a low change value. In order to identify the major protease substrates in mitochondria, representing the bulk of degraded polypeptides, we also monitored the changes in the absolute protein amounts after the incubation period (Fig. 2B). Most of the proteins that showed large negative change values (Alt1, Ilv1, Ilv6, Mcr1, Lys4, Rim1, and Yjl200c) were of relatively low abundance. Although a substantial portion of these proteins was degraded, the total mass of degraded polypeptides was relatively low, resulting in small absolute change values. The proteins Lsc1, Dld1, and Ilv2, having an intermediate abundance, showed both high absolute and relative differences before and after incubation, identifying them as important substrate proteins with a high degree of proteolysis. We observed that a substantial amount of polypeptide material of two highly abundant proteins, Atp2 and Hsp60, was degraded. However, due to their high abundance, the effects on the total protein amount were small, indicated by small change values. Nevertheless, the degradation of Atp2 and Hsp60 seems to be specific, since other highly abundant proteins like Ssc1, Ald4, Aco1, Ilv5, and Por1 were not affected. The absolute protein amounts are given in the supplemental material (see Table S2 in the supplemental material). In summary, we conclude that the major mitochondrial enzymes are generally maintained at a stable level. However, the high turnover rates of a subset of mitochondrial proteins indicated that these proteins have a high susceptibility to degradation by mitochondrial proteases already at normal conditions, in many cases. Hence, mitochondrial proteolysis seemed to be highly selective for certain protein species and was not correlated with protein abundance.

Identification of Pim1 substrate proteins. Since the alterations of protein amounts in the time course experiments reflect the combined activity of all mitochondrial proteases, we set out to identify the proteins that were specifically degraded by the matrix protease Pim1. We reasoned that Pim1-specific degradation substrates should accumulate in pim1{Delta} mitochondria due to the absence of the protease. By a quantitative comparison of the protein spot patterns of wild-type and pim1{Delta} mitochondria, the substrate proteins could be identified. Since pim1{Delta} yeast strains have a rho– phenotype, we prepared pim1{Delta} mitochondria and the corresponding rho– wild type from yeast strains grown on fermentable carbon sources and analyzed the mitochondrial protein spot patterns by 2D-PAGE (Fig. 3). To enhance the differences of the protein composition due to proteolysis, we additionally incubated the isolated mitochondria for 1 h at 30°C. The protein spots that showed a significantly increased amount in pim1{Delta} mitochondria were excised and sequenced by mass spectroscopy. We were able to identify five proteins, Ilv1, Ilv2, Lsc1, Lys4, and Yjl200c, as putative substrates of Pim1. The accumulated proteins had widely different molecular weights but had isoelectric points mainly in the neutral pH range. The quantitative analysis of the differences in the spot patterns of the 2D gels (Fig. 4A) revealed that these proteins had positive normalized change values between 0.4 and 0.7 (40 to 70% higher spot volumes) in pim1{Delta} mitochondria. Proteins with change values lower than 0.2 were not considered significant to avoid experimental artifacts of the 2D-PAGE procedure.


Figure 3
View larger version (74K):
[in this window]
[in a new window]
  		FIG. 3. 2D-PAGE analysis of the proteome of wild-type and pim1{Delta} mitochondria. Intact isolated mitochondria from wild-type and pim1{Delta} cells were incubated for 1 h at 30°C and analyzed by 2D-PAGE. Protein spots were visualized by Coomassie staining. A typical master 2D map of mitochondrial proteins (wild type) indicating the location of the enlarged sections is shown. Gel sections of WT and pim1{Delta} mitochondria indicate the localization of protein spots with an increased abundance in pim1{Delta} and control proteins (Aco1, Mdh1) (arrows). MW, molecular mass.

 

Figure 4
View larger version (29K):
[in this window]
[in a new window]
  		FIG. 4. Quantification of the differences in mitochondrial protein composition between wild-type and pim1{Delta} mitochondria. (A) Normalized (Norm.) change values for selected proteins of wild-type and pim1{Delta} mutant mitochondria from the experiment in Fig. 3 are shown with gray bars. The proteins are sorted according to degree of change. The relative errors of the change values are represented by error bars. (B) Time course of degradation of selected proteins in wild-type mitochondria. The amount of each protein (spot volume) present at 0 h, after 1 h, and after 4 h of incubation of isolated mitochondria at 30°C was determined. Mitochondrial protein composition was analyzed by 2D-PAGE as described in the text. The protein amount present in the sample at 0 h was set as 100%. The average of three independent experiments is shown. Error bars show the relative errors of the spot volumes.

 
The identified proteins are metabolic enzymes. Lsc1, the {alpha}-subunit of succinyl coenzyme A (CoA) ligase, is involved in the TCA cycle. For the protein Yjl200c, an uncharacterized open reading frame with postulated aconitate hydratase activity, an involvement in TCA cycle reactions seems plausible. Three candidate proteins are involved in different amino acid biosynthetic pathways, like homoaconitase (Lys4), involved in lysine biosynthesis; Ilv1, the threonine deaminase needed for isoleucine biosynthesis; and Ilv2, the acetolacetate synthase with functions in valine and isoleucine biosynthesis. The analysis of the differences in protein abundance between wild-type and pim1{Delta} mitochondria also showed that most metabolic enzymes and proteins involved in biosynthetic reactions (like chaperones) were not changed significantly. Again, only a small subset of the mitochondrial proteins seemed to be sensitive to proteolysis by Pim1 under normal temperature conditions. An increase in the protein amount in the pim1{Delta} mutant mitochondria does not necessarily reflect an inhibited degradation reaction but could also result from an increased protein expression. To directly assess the degradation rates of the identified proteins, we selected some of the identified proteins and performed time course turnover experiments in isolated wild-type mitochondria at 30°C using 2D-PAGE (Fig. 4B). All selected proteins, Ilv1, Ilv2, Lsc1, Lys4, and Yjl200c, indeed showed a strong, time-dependent reduction in the amounts of the endogenous proteins in intact mitochondria. In comparison, the control proteins malate dehydrogenase (Mdh1) and aconitase (Aco1) remained stable under the same conditions.

In organello degradation assays verify Pim1-dependent degradation of endogenous substrate proteins. In order to characterize the direct involvement of Pim1 in the degradation of the putative substrate proteins, we utilized an in organello proteolysis assay based on the in vitro import of radiolabeled substrate proteins into isolated mitochondria (29, 47). The proteins Lys4, Yjl200c, Ilv1, Ilv2, and Lsc1 were efficiently degraded after import into wild-type mitochondria (Fig. 5A to E). The degradation rates of the different proteins varied slightly, most probably due to their individual structural properties. The homoaconitase Lys4 showed the strongest degradation after import into mitochondria. A Pim1-dependent degradation behavior of the five most prominent substrate proteins identified in the 2D-PAGE analysis (Fig. 4A) was therefore confirmed by the in organello proteolysis assays by their significantly lower degradation rates in pim1{Delta} mitochondria. In summary, the proteins Lys4, Yjl200c, Ilv1, Ilv2, and Lsc1 represent a novel class of endogenous substrate proteins of the mitochondrial protease Pim1 that are degraded under normal temperature conditions. As a control protein for the specificity of the in organello degradation assays, we imported the TCA cycle enzyme malate dehydrogenase, Mdh1, which was not degraded (Fig. 6A). Interestingly, another enzyme of the TCA cycle, aconitase (Aco1), showed a different behavior (Fig. 6B). While the endogenous aconitase was stable in the 2D-PAGE analysis, the newly imported protein was degraded, albeit at a comparably slow rate.


Figure 5
View larger version (25K):
[in this window]
[in a new window]
  		FIG. 5. Degradation kinetics of selected proteins after import into wild-type and pim1{Delta} mitochondria. The proteolytic breakdown of radiolabeled mitochondrial proteins was monitored at 30°C after the import in wild-type and pim1{Delta} mitochondria. To obtain comparable conditions, the [rho0] wild-type strain KRY01 was used. Aliquots were taken and analyzed by SDS-PAGE at various time points (0, 30, 60, 120, 180, and 240 min). The remaining protein amounts in each sample were quantified by autoradiography. The amount of protein present at 0 min was set to 100% for wild-type and pim1{Delta} mitochondria. The diagrams show the means and the standard errors of the means of at least three independent experiments.

 

Figure 6
View larger version (15K):
[in this window]
[in a new window]
  		FIG. 6. Stability of control proteins in wild-type and pim1{Delta} mitochondria. Radiolabeled preproteins, Mdh1 (A) and Aco1 (B), were imported into wild-type (WT) and pim1{Delta} mitochondria, and their proteolysis was monitored for 240 min at 30°C. The amount of reporter proteins was quantified as described in the legend to Fig. 5. The means and the standard errors of the means of at least three independent experiments are shown in both diagrams.

 
Since Pim1 belongs to the AAA+ protein family of ATP-dependent proteases, we tested whether the newly identified native substrate proteins were degraded in an ATP-dependent fashion. The radiolabeled substrate proteins were imported into isolated wild-type mitochondria that have been depleted from ATP after completion of the import reaction (Fig. 7A to E). In agreement with a Pim1-dependent proteolysis, degradation of all identified substrate proteins was strongly dependent on a high matrix ATP level.


Figure 7
View larger version (22K):
[in this window]
[in a new window]
  		FIG. 7. ATP-dependent degradation of putative Pim1 substrates. After import of the indicated radiolabeled preproteins, Lys4 (A), Yjl200c (B), Ilv1 (C), Lsc1 (D), and Ilv2 (E), in wild-type mitochondria, the samples were incubated at 30°C for 240 min either with an ATP-regenerating system (+ ATP) or with apyrase and oligomycin (– ATP). The degradation of the imported proteins was analyzed by SDS-PAGE and autoradiography as described in the legend to Fig. 5.

 
In the 2D-PAGE time course analysis of intact mitochondria (Fig. 1A), the proteins Ilv1, Lsc1, and Lys4 showed a less efficient degradation at the elevated temperature of 39°C than that under normal conditions (30°C). To confirm this temperature-dependent loss of protease susceptibility, we reproduced the temperature effects in the in organello proteolysis assay. Indeed, after import into isolated wild-type mitochondria, the degradation rates of the proteins Ilv1, Lsc1, and Lys4 were strongly reduced at 39°C compared to those under the standard conditions at 30°C (Fig. 8A to C). As a control protein, we analyzed Alt1, a putative alanine transaminase. Similar to its behavior in the in vivo time course, Alt1 was more strongly degraded at elevated temperatures after import in organello, exhibiting the opposite behavior as Ilv1, Lsc1, and Lys4 (Fig. 8D). The reduced degradation rates at high temperatures are most likely due to a higher tendency to form insoluble aggregates (B. von Janowsky and W. Voos, unpublished data).


Figure 8
View larger version (28K):
[in this window]
[in a new window]
  		FIG. 8. Temperature-dependent degradation of Pim1 substrates. The proteins Ilv1 (A), Lsc1 (B), Lys4 (C), and, as a control, Alt1 (D) were imported as radiolabeled precursors for 20 min at 30°C into the matrix of wild-type mitochondria. After completion of import, their stability was assayed for 240 min at 30°C or alternatively at 39°C as described in the legend to Fig. 5.

 
The absence of an intact Fe/S cluster increases susceptibility to proteolysis. We observed that most of the identified Pim1 substrates exhibit a complex tertiary structure that includes different types of prosthetic groups like the Fe/S cluster (Aco1, Lys4, and Yjl200c), pyridoxalphosphate (Ilv1), and thiamine pyrophosphate (Ilv2). We reasoned that defects in the biosynthesis or assembly of such prosthetic groups would perturb the structure of the respective enzymes, leading to the recognition and degradation by Pim1. As an example, we chose to analyze the influence of Fe/S cluster assembly on the rate of protein degradation. To directly demonstrate the correlation between proteolysis and a damaged or absent Fe/S cluster, we analyzed the degradation of the substrate proteins in mitochondria derived from the mutant yeast strains ssq1{Delta} and nfu1{Delta}. The proteins Ssq1 and Nfu1 have been shown to be involved in the mitochondrial synthesis of the Fe/S cluster (9, 34). When we imported the Yjl200c precursor into ssq1{Delta} or nfu1{Delta} mutant mitochondria, we indeed observed an increased degradation rate (Fig. 9A). A similar observation was obtained for aconitase (Aco1) that was strongly degraded in the Fe/S cluster assembly mutants (Fig. 9B). Surprisingly, the protein Lys4 did not show a difference in degradation between wild-type and ssq1{Delta} or nfu1{Delta} mutant mitochondria (Fig. 9C). However, since Lys4 already showed a high degradation rate in wild-type mitochondria, it is likely that even defects in its Fe/S cluster might not be able to enhance the rate of proteolysis under the chosen conditions. As a control, we tested the effect of the Fe/S biogenesis mutants on the degradation of the {alpha}-subunit of the succinate-CoA ligase (Lsc1) that does not contain an Fe/S cluster as a cofactor. This protein did not exhibit any difference in the degradation rates between wild-type and mutant mitochondria (Fig. 9D), excluding a nonspecific effect on mitochondrial proteolysis. At least for the two Fe/S cofactor-containing proteins analyzed, a correlation between the degradation behavior and the impairment of Fe/S cluster synthesis was observed. We conclude that the integrity of the prosthetic group, and its effects on the overall tertiary structure of the proteins, is one of the major determinants for the susceptibility to Pim1-dependent degradation.


Figure 9
View larger version (32K):
[in this window]
[in a new window]
  		FIG. 9. Proteolysis of proteins after import into Fe/S cluster assembly mutants. The degradation of the imported and processed radiolabeled proteins Yjl200c (A), Aco1 (B), Lys4 (C), and control protein Lsc1 (D) was analyzed in wild-type, ssq1{Delta} and nfu1{Delta} mitochondria. The stability of the proteins was examined and quantified as described in the legend to Fig. 5.

 

   DISCUSSION
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
In this study, we investigated mitochondrial protein composition in the yeast Saccharomyces cerevisiae using proteomic techniques in order to get information about the extent of organellar protein turnover and the identities of mitochondrial protease substrates. Although a fairly comprehensive picture of the general protein composition of mitochondria has been established by various proteome studies (17, 24, 26, 37, 41), not much is known about the dynamic behavior of the mitochondrial proteome. By analyzing purified mitochondria using 2D-PAGE, we monitored the quantitative changes in protein composition that were due to organellar proteolysis. The 2D-PAGE technique provided a straightforward and reliable method for the quantitative analysis of a protein set with a limited number of polypeptides, as is represented by the mitochondrial proteome. We were able to reproducibly quantify 60 different mitochondrial proteins. Although the total number of different protein species that were analyzed by the 2D-PAGE approach was limited compared to database listings (MitoP2), most of the proteins with intermediate to high abundance were included in our analysis, representing the bulk of mitochondrial protein mass with a bias for the detection of soluble matrix proteins. Since most of these proteins fall into the functional class of metabolic enzymes, which play the major role in the adaptation to environmental changes and are mainly affected by stress conditions, the procedure used allowed, in particular, the investigation of proteases of the matrix compartment responsible for the turnover of soluble proteins. In the model organism used, the yeast Saccharomyces cerevisiae, the major mitochondrial protease of the matrix is the protein Pim1, a homolog of the bacterial Lon protease.

To date, information about the endogenous target proteins of mitochondrial proteases has been very limited (19, 27, 39, 45). Our proteomic analysis of protease substrates represents the first step in defining the role of proteolytic enzymes in the quality control reaction of mitochondria under in vivo conditions. The 2D-PAGE analysis of protein turnover in intact isolated mitochondria showed a relatively high general stability of the mitochondrial proteome. Only about 5 to 10% of all detected protein spots showed a significantly reduced amount after 4 h of incubation under normal conditions (data not shown). Although the low turnover rate was surprising, it correlates well with a previous study that analyzed peptides exported from mitochondria that were generated by proteolysis (1). Other results indicate that the mitochondrial protein composition does not change dramatically even under conditions where a high degree of metabolic adaptation seems to be required, like the switch from fermentative to nonfermentative conditions (24). In our study, we observed some differences in the amounts of individual mitochondrial proteins between cells grown on glucose and those grown on glycerol (data not shown). Although repression of gene expression by glucose may have an influence on mitochondrial protein composition, the overall protein spot pattern showed only small differences.

However, discrepancies between the transcriptome (representing the gene expression pattern) and the proteome (representing the protein amounts) indicate that different turnover rates for individual proteins do exist. Despite a relatively stable proteome in general, our analysis revealed that certain endogenous proteins are subjected to a specific proteolysis. Based on the behavior at different incubation temperatures, four classes of different protein stabilities were identified. As expected, the first class, comprising the majority of stable proteins, didn't show significant proteolysis at either normal (30°C) or elevated (39°C) temperatures. The second class of proteins, representing normal protease substrates, was degraded at both normal and elevated temperatures. A third group of proteins exhibited a temperature-sensitive proteolysis, being degraded only at 39°C. These proteins obviously have heat-labile conformations that make them prone to heat damage at elevated temperatures, thereby increasing protease susceptibility. Interestingly, a fourth group showed a significant turnover only at normal temperatures but remained stable at 39°C. This behavior could also be reproduced by in organello proteolysis assays, which analyze the degradation of individual proteins after import into isolated mitochondria. In previous experiments, we observed that destabilized, aggregation-prone model proteins were completely resistant against Pim1-dependent proteolysis at elevated temperatures (29). Indeed, many of the tested Pim1 substrates have a high tendency to form insoluble aggregates already at slightly elevated temperatures (B. von Janowsky and W. Voos, unpublished). A possible reason for this behavior is that the respective proteins have an unstable tertiary structure that makes them susceptible to proteolysis at normal temperatures but at the same time prone to aggregation at elevated temperatures.

The amount of protease molecules is relatively limited under normal conditions, restricting the amount of proteolysis that can occur in mitochondria. Some proteins showed a high degree of relative change, with a majority of the polypeptides degraded during the incubation period. These proteins represent either intrinsically instable polypeptides or proteins that have to be removed due to regulative processes. However, those proteins do not necessarily reflect the bulk of degraded polypeptides in mitochondria. Interestingly, the proteins with the highest absolute reduction were the F1Fo ATPase subunit Atp2 (F1ß), the chaperone Hsp60, and the inner membrane protein D-lactate dehydrogenase (Dld1). Since Hsp60 was subjected to proteolysis predominantly at high temperatures, it could be speculated that some of the large oligomeric Hsp60 protein complexes become destabilized under these conditions, resulting in the degradation of nonassembled subunits. The observation that a heat shock protein like Hsp60 was degraded at higher temperatures seems to be counterintuitive. However, despite its prominent role in the de novo folding of proteins, there is no evidence that Hsp60 proteins play a prominent role in cellular thermoprotection, at least if the more important "holding" chaperones like Hsp70 are functional (21). The reason for the high sensitivity of Dld1 at all temperatures is unknown but could be related to its complicated biogenesis pathway (28), possibly resulting in a high number of nonassembled polypeptides.

Since the protein Pim1 is a major component of the mitochondrial protein quality control system in the matrix compartment, we were especially interested in native substrate proteins of this protease. Most studies that addressed proteolysis by Pim1 used artificially constructed reporter proteins that were introduced into mitochondria (29, 47). Those studies concluded that Pim1 preferentially degrades misfolded or damaged protein substrates. This selectivity was supported by experiments where mutated forms of endogenous mitochondrial proteins were found to be degraded by Pim1. For example, in Saccharomyces cerevisiae mitochondria, a mutant form of the enzyme acetohydroxy acid reductoisomerase (Ilv5) that was prone to aggregation became a substrate of Pim1 (2). In addition, two proteins were described to be Pim1 substrates in vivo, the F1ß-subunit of the F1Fo ATPase (Atp2) and the ß-subunit of the matrix-processing peptidase (Mas1). However, these proteins were only degraded when the yeast cells were grown under heat stress conditions (39°C) (40).

By a comparison of the steady-state protein compositions of wild-type and pim1{Delta} mitochondria, we were able to directly identify novel protein substrates of Pim1. We found five proteins (Ilv2, Ilv1, Lsc1, Lys4, and Yjl200c) that occurred in significantly higher amounts in pim1{Delta} mitochondria. Although the analysis of pim1{Delta} mitochondria required a switch to fermentable growth conditions, the overall similarities of the protein spot pattern of mitochondria grown on different carbon sources made it unlikely that a major protein was missed in the identification of Pim1-specific substrate proteins. The newly identified substrate proteins also showed a time-dependent turnover in wild-type mitochondria, indicating that their accumulation was indeed due to absent proteolysis. Surprisingly, most of the previously proposed Pim1 substrates were not found in our analysis. Although we observed some degradation of Atp2 in wild-type mitochondria, it did not seem to be a substrate of Pim1 due to its absent accumulation in pim1{Delta} mitochondria, at least at normal temperatures. Either Pim1 is not responsible for the degradation of Atp2 or the significantly lower steady-state protein amount of Atp2 in the rho– cells grown under fermentable conditions has an effect on its degradation efficiency. In contrast, Mas1 did not show any proteolysis under the conditions analyzed. However, since the degradation of Atp2 and Mas1 was originally observed only under stress conditions in intact cells, indirect effects on protein import or folding, leading to the observed instability, cannot be ruled out. We also did not find an accumulation of the proteins Dld1 and Mcr1 in pim1{Delta} mitochondria. Although they showed a significant turnover in the time course experiment, it is highly unlikely that they were degraded by Pim1. As membrane proteins or proteins of the intermembrane space, respectively, they are not exposed to the matrix space and therefore are not accessible for the protease Pim1. Similar to these proteins, Alt1, a putative aminotransferase of the matrix, showed no Pim1 dependence (data not shown) and is possibly degraded by other mitochondrial proteases. A certain substrate overlap between the soluble Pim1 protease and the m-AAA protease of the inner membrane that exposes its active site to the matrix cannot be excluded (32).

To validate the Pim1-dependent degradation, we subjected the set of newly identified substrate proteins to an in organello degradation assay. This assay is based on the import of radiolabeled precursor proteins that were synthesized by in vitro transcription/translation. By this import reaction, isolated mitochondria are loaded with protease substrate proteins in a short, pulse-like reaction. After the removal of nonimported polypeptides, the degradation reaction can be followed quantitatively in a subsequent incubation, representing a chase period. Since mitochondria remain intact throughout the incubation, proteolysis occurs essentially under in vivo conditions. After import of the proteins Ilv2, Ilv1, Lsc1, Lys4, and Yjl200c into isolated wild-type mitochondria, we observed specific, Pim1-dependent degradation in all cases. Degradation was also dependent on a high matrix ATP level, very similar to the behavior of artificial reporter proteins that have been used extensively as Pim1 substrates (29, 32, 47). The slow degradation of some substrate proteins, especially Lys4, in pim1{Delta} mitochondria most likely indicates that other mitochondrial proteases may partially take over the function of Pim1. In general, the in organello degradation assays essentially reflected the protease susceptibility of the analyzed proteins very well. However, it should be considered that the degradation susceptibility of newly imported precursor proteins could be different from the behavior of the same protein under steady-state conditions. In case of missing assembly partners or cofactors required for full enzymatic activity, folding to a native conformation might not be achieved, resulting in premature degradation. A case like this is represented by the TCA cycle enzyme aconitase (Aco1) that was stable under steady-state conditions but was slowly degraded after import into isolated mitochondria (see below).

What might be the possible reasons for the increased susceptibility of the newly identified proteins to proteolysis? Interestingly, we observed that three of the Pim1 substrate proteins (Ilv1, Lsc1, and Lys4) belonged to the group of substrate proteins that exhibited a high turnover rate at normal conditions but that were stable at higher temperatures. This is an indication that the overall conformational stability of the polypeptide might play an important role in recognition by Pim1. In addition, two of the new Pim1 substrates are members of an enzyme family containing an Fe/S cluster as a prosthetic group: Lys4, the homoaconitase involved in lysine biosynthesis; and Yjl200c, an uncharacterized open reading frame that has significant homology to aconitase. It has been shown previously that Fe/S clusters are generally very sensitive to oxidative damage (3). Since our experiments have not been performed under oxygen-limiting conditions, a certain degree of oxidative damage could be expected even under normal conditions. In case of a damaged or lost Fe/S cluster, the respective enzymes are likely to lose their structural integrity. The resulting nonnative conformation of the proteins would then serve as a signal for their removal by the protease Pim1. A similar mechanism might also apply to the proteins threonine deaminase (Ilv1) and acetolacetate synthase (Ilv2), which catalyze steps in the biosynthesis of amino acids. Both require a cofactor for their function, pyridoxal phosphate in the case of Ilv1 and thiamine pyrophosphate in the case of Ilv2.

The important role of cofactors is corroborated by the degradation behavior of aconitase (Aco1), which, in contrast to other members of the Fe/S cluster enzyme family, was not changed in our turnover experiments in intact mitochondria. It was observed that aconitase was specifically degraded in mammalian mitochondria only under oxidative stress conditions (4). A similar situation might be present after import in mitochondria. A delayed assembly of the Fe/S cluster would result in the accumulation of nonfunctional apoenzymes that are degraded by Pim1. In the case of the endogenous, fully functional aconitase, the overall structural stability of the assembled enzyme seemed to be sufficient to prevent proteolysis. The role of a damaged or missing Fe/S cluster as a cause for proteolytic removal could be confirmed by the analysis of protein degradation in mutant mitochondria where components of the Fe/S cluster assembly machinery had been deleted. Indeed, the degradation rates of newly imported Yjl200c and Aco1 were significantly increased in ssq1{Delta} and nfu1{Delta} mitochondria, apparently by the inability of these mitochondria to synthesize a functional Fe/S cluster (20). One of the identified Pim1 substrates, Lsc1, does not contain a cofactor or prosthetic group. It is part of a heterodimeric enzyme complex in vivo, being the {alpha}-subunit of the succinyl-CoA ligase of the TCA cycle. Since Lsc1 belonged to the group of aggregation-prone protease substrates, its conformational stability is probably so low that it is recognized by Pim1 already at normal temperatures.

In summary, we propose that changes in a complex tertiary structure, particularly defects in the function or assembly of prosthetic groups, make proteins susceptible to degradation by the mitochondrial matrix protease Pim1. The altered, nonnative conformation most probably serves as a signal for recognition and subsequent removal by Pim1 in the absence of a covalent modification like the ubiquitin tag. A similar substrate selection behavior has been observed for the bacterial membrane-bound protease FtsH, a homolog of the AAA+ proteases of the mitochondrial inner membrane (14). The biochemical determination of the structural requirements for the mechanism of substrate recognition and proteolysis by Pim1 will be an important topic of future investigations.


   ACKNOWLEDGMENTS
 
We thank N. Zufall for expert technical assistance, D. Hofmann for help with the 2D-PAGE technique, and T. Langer and S. Rospert for discussion and critical reading of the manuscript.

This work was supported by the Deutsche Forschungsgemeinschaft (Schwerpunktprogramm 1132, grant VO 657/4-2, to W.V.).


   FOOTNOTES
 
* Corresponding author. Mailing address: Institut für Biochemie und Molekularbiologie, Universität Freiburg, Hermann Herder Str. 7, 79104 Freiburg, Germany. Phone: 49-761-203-5280. Fax: 49-761-203-5261. E-mail: wolfgang.voos{at}biochemie.uni-freiburg.de. Back

{dagger} Supplemental material for this article may be found at http://mcb.asm.org/. Back


   REFERENCES
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 
1. Augustin, S., M. Nolden, S. Müller, O. Hardt, I. Arnold, and T. Langer. 2005. Characterization of peptides released from mitochondria: evidence for constant proteolysis and peptide efflux. J. Biol. Chem. 280:2691-2699.[Abstract/Free Full Text]

2. Bateman, J. M., M. Iacovino, P. S. Perlman, and R. A. Butow. 2002. Mitochondrial DNA instability mutants of the bifunctional protein Ilv5p have altered organization in mitochondria and are targeted for degradation by Hsp78 and the Pim1p protease. J. Biol. Chem. 277:47946-47953.[Abstract/Free Full Text]

3. Beinert, H., R. H. Holm, and E. Munck. 1997. Iron-sulfur clusters: nature's modular, multipurpose structures. Science 277:653-659.[CrossRef][Medline]

4. Bota, D. A., and K. J. Davies. 2002. Lon protease preferentially degrades oxidized mitochondrial aconitase by an ATP-stimulated mechanism. Nat. Cell Biol. 4:674-680.[CrossRef][Medline]

5. Catrein, I., R. Herrmann, A. Bosserhoff, and T. Ruppert. 2005. Experimental proof for a signal peptidase I like activity in Mycoplasma pneumoniae, but absence of a gene encoding a conserved bacterial type I SPase. FEBS J. 272:2892-2900.[CrossRef][Medline]

6. Davies, K. J. 1993. Protein modification by oxidants and the role of proteolytic enzymes. Biochem. Soc. Trans. 21:346-353.[Medline]

7. Dobson, C. M. 2003. Protein folding and misfolding. Nature 426:884-890.[CrossRef][Medline]

8. Dreger, M. 2003. Proteome analysis at the level of subcellular structures. Eur. J. Biochem. 270:589-599.[Medline]

9. Dutkiewicz, R., B. Schilke, H. Knieszner, W. Walter, E. A. Craig, and J. Marszalek. 2003. Ssq1, a mitochondrial Hsp70 involved in iron-sulfur (Fe/S) center biogenesis. Similarities to and differences from its bacterial counterpart. J. Biol. Chem. 278:29719-29727.[Abstract/Free Full Text]

10. Esser, C., S. Alberti, and J. Höhfeld. 2004. Cooperation of molecular chaperones with the ubiquitin/proteasome system. Biochim. Biophys. Acta 1695:171-188.[Medline]

11. Goldberg, A. L. 2003. Protein degradation and protection against misfolded or damaged proteins. Nature 426:895-899.[CrossRef][Medline]

12. Goldberg, A. L. 1992. The mechanism and functions of ATP-dependent proteases in bacterial and animal cells. Eur. J. Biochem. 203:9-23.[Medline]

13. Goldberg, A. L., R. P. Moerschell, C. H. Chung, and M. R. Maurizi. 1994. ATP-dependent protease La (lon) from Escherichia coli. Methods Enzymol. 244:350-375.[Medline]

14. Herman, C., S. Prakash, C. Z. Lu, A. Matouschek, and C. A. Gross. 2003. Lack of a robust unfoldase activity confers a unique level of substrate specificity to the universal AAA protease FtsH. Mol. Cell 11:659-669.[CrossRef][Medline]

15. Kleizen, B., and I. Braakman. 2004. Protein folding and quality control in the endoplasmic reticulum. Curr. Opin. Cell Biol. 16:343-349.[CrossRef][Medline]

16. Kostova, Z., and D. H. Wolf. 2003. For whom the bell tolls: protein quality control of the endoplasmic reticulum and the ubiquitin-proteasome connection. EMBO J. 22:2309-2317.[CrossRef][Medline]

17. Kumar, A., S. Agarwal, J. A. Heyman, S. Matson, M. Heidtman, S. Piccirillo, L. Umansky, A. Drawid, R. Jansen, Y. Liu, K. H. Cheung, P. Miller, M. Gerstein, G. S. Roeder, and M. Snyder. 2002. Subcellular localization of the yeast proteome. Genes Dev. 16:707-719.[Abstract/Free Full Text]

18. Langer, T. 2000. AAA proteases: cellular machines for degrading membrane proteins. Trends Biochem. Sci. 25:247-251.[CrossRef][Medline]

19. Langer, T., and W. Neupert. 1996. Regulated protein degradation in mitochondria. Experientia 52:1069-1076.[CrossRef][Medline]

20. Lill, R., and U. Mühlenhoff. 2005. Iron-sulfur-protein biogenesis in eukaryotes. Trends Biochem. Sci. 30:133-141.[CrossRef][Medline]

21. Mogk, A., T. Tomoyasu, P. Goloubinoff, S. Rüdiger, D. Roder, H. Langen, and B. Bukau. 1999. Identification of thermolabile Escherichia coli proteins: prevention and reversion of aggregation by DnaK and ClpB. EMBO J. 18:6934-6949.[CrossRef][Medline]

22. Mootha, V. K., J. Bunkenborg, J. V. Olsen, M. Hjerrild, J. R. Wisniewski, E. Stahl, M. S. Bolouri, H. N. Ray, S. Sihag, M. Kamal, N. Patterson, E. S. Lander, and M. Mann. 2003. Integrated analysis of protein composition, tissue diversity, and gene regulation in mouse mitochondria. Cell 115:629-640.[CrossRef][Medline]

23. Neuwald, A. F., L. Aravind, J. L. Spouge, and E. V. Koonin. 1999. AAA+: a class of chaperone-like ATPases associated with the assembly, operation, and disassembly of protein complexes. Genome Res. 9:27-43.[Abstract/Free Full Text]

24. Ohlmeier, S., A. J. Kastaniotis, J. K. Hiltunen, and U. Bergmann. 2004. The yeast mitochondrial proteome, a study of fermentative and respiratory growth. J. Biol. Chem. 279:3956-3979.[Abstract/Free Full Text]

25. Pickart, C. M., and R. E. Cohen. 2004. Proteasomes and their kin: proteases in the machine age. Nat. Rev. Mol. Cell Biol. 5:177-187.[CrossRef][Medline]

26. Prokisch, H., C. Scharfe, D. G. Camp, W. Xiao, L. David, C. Andreoli, M. E. Monroe, R. J. Moore, M. A. Gritsenko, C. Kozany, K. K. Hixson, H. M. Mottaz, H. Zischka, M. Ueffing, Z. S. Herman, R. W. Davis, T. Meitinger, P. J. Oefner, R. D. Smith, and L. M. Steinmetz. 2004. Integrative analysis of the mitochondrial proteome in yeast. PLoS Biol. 2:795-804.

27. Rep, M., and L. A. Grivell. 1996. The role of protein degradation in mitochondrial function and biogenesis. Curr. Genet. 30:367-380.[CrossRef][Medline]

28. Rojo, E. E., B. Guiard, W. Neupert, and R. A. Stuart. 1998. Sorting of D-lactate dehydrogenase to the inner membrane of mitochondria. Analysis of topogenic signal and energetic requirements. J. Biol. Chem. 273:8040-8047.[Abstract/Free Full Text]

29. Röttgers, K., N. Zufall, B. Guiard, and W. Voos. 2002. The ClpB homolog Hsp78 is required for the efficient degradation of proteins in the mitochondrial matrix. J. Biol. Chem. 277:45829-45837.[Abstract/Free Full Text]

30. Ryan, M. T., W. Voos, and N. Pfanner. 2001. Assaying protein import into mitochondria. Methods Cell Biol. 65:189-215.[Medline]

31. Savel'ev, A. S., I. E. Kovaleva, L. A. Novikova, L. V. Isaeva, and V. N. Luzikov. 1999. Can foreign proteins imported into yeast mitochondria interfere with PIM1p protease and/or chaperone function? Arch. Biochem. Biophys. 363:373-376.[CrossRef][Medline]

32. Savel'ev, A. S., L. A. Novikova, I. E. Kovaleva, V. N. Luzikov, W. Neupert, and T. Langer. 1998. ATP-dependent proteolysis in mitochondria. m-AAA protease and PIM1 protease exert overlapping substrate specificities and cooperate with the mtHsp70 system. J. Biol. Chem. 273:20596-20602.[Abstract/Free Full Text]

33. Schapira, A. H. 1999. Mitochondrial involvement in Parkinson's disease, Huntington's disease, hereditary spastic paraplegia and Friedreich's ataxia. Biochim. Biophys. Acta 1410:159-170.[Medline]

34. Schilke, B., C. Voisine, H. Beinert, and E. Craig. 1999. Evidence for a conserved system for iron metabolism in the mitochondria of Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 96:10206-10211.[Abstract/Free Full Text]

35. Sherman, M., and A. L. Goldberg. 1992. Involvement of the chaperonin DnaK in the rapid degradation of a mutant protein in Escherichia coli. EMBO J. 11:71-77.[Medline]

36. Shevchenko, A., M. Wilm, O. Vorm, and M. Mann. 1996. Mass spectrometric sequencing of proteins silver-stained polyacrylamide gels. Anal. Biochem. 68:850-858.

37. Sickmann, A., J. Reinders, Y. Wagner, C. Joppich, R. Zahedi, H. E. Meyer, B. Schönfisch, I. Perschil, A. Chacinska, B. Guiard, P. Rehling, N. Pfanner, and C. Meisinger. 2003. The proteome of Saccharomyces cerevisiae mitochondria. Proc. Natl. Acad. Sci. USA 100:13207-13212.[Abstract/Free Full Text]

38. Stahlberg, H., E. Kutejova, K. Suda, B. Wolpensinger, A. Lustig, G. Schatz, A. Engel, and C. K. Suzuki. 1999. Mitochondrial Lon of Saccharomyces cerevisiae is a ring-shaped protease with seven flexible subunits. Proc. Natl. Acad. Sci. USA 96:6787-6790.[Abstract/Free Full Text]

39. Suzuki, C. K., M. Rep, J. M. van Dijl, K. Suda, L. A. Grivell, and G. Schatz. 1997. ATP-dependent proteases that also chaperone protein biogenesis. Trends Biochem. Sci. 22:118-123.[CrossRef][Medline]

40. Suzuki, C. K., K. Suda, N. Wang, and G. Schatz. 1994. Requirement for the yeast gene LON in intramitochondrial proteolysis and maintenance of respiration. Science 264:273-276.[Abstract/Free Full Text]

41. Taylor, S. W., E. Fahy, B. Zhang, G. M. Glenn, D. E. Warnock, S. Wiley, A. N. Murphy, S. P. Gaucher, R. A. Capaldi, B. W. Gibson, and S. S. Ghosh. 2003. Characterization of the human heart mitochondrial proteome. Nat. Biotechnol. 21:281-286.[CrossRef][Medline]

42. Tomoyasu, T., T. Yuki, S. Morimura, H. Mori, K. Yamanaka, H. Niki, S. Hiraga, and T. Ogura. 1993. The Escherichia coli FtsH protein is a prokaryotic member of a protein family of putative ATPases involved in membrane functions, cell cycle control, and gene expression. J. Bacteriol. 175:1344-1351.[Abstract/Free Full Text]

43. Turner, C., and A. H. Schapira. 2001. Mitochondrial dysfunction in neurodegenerative disorders and ageing. Adv. Exp. Med. Biol. 487:229-251.[Medline]

44. van Dijl, J. M., E. Kutejova, K. Suda, D. Perecko, G. Schatz, and C. K. Suzuki. 1998. The ATPase and protease domains of yeast mitochondrial Lon: roles in proteolysis and respiration-dependent growth. Proc. Natl. Acad. Sci. USA 95:10584-10589.[Abstract/Free Full Text]

45. Van Dyck, L., and T. Langer. 1999. ATP-dependent proteases controlling mitochondrial function in the yeast Saccharomyces cerevisiae. Cell. Mol. Life Sci. 56:825-842.[CrossRef][Medline]

46. Van Dyck, L., D. A. Pearce, and F. Sherman. 1994. PIM1 encodes a mitochondrial ATP-dependent protease that is required for mitochondrial function in the yeast Saccharomyces cerevisiae. J. Biol. Chem. 269:238-242.[Abstract/Free Full Text]

47. Wagner, I., H. Arlt, L. van Dyck, T. Langer, and W. Neupert. 1994. Molecular chaperones cooperate with PIM1 protease in the degradation of misfolded proteins in mitochondria. EMBO J. 13:5135-5145.[Medline]

48. Wallace, D. C. 1999. Mitochondrial diseases in man and mouse. Science 283:1482-1488.[Abstract/Free Full Text]

49. Wang, N., S. Gottesman, M. C. Willingham, M. M. Gottesman, and M. R. Maurizi. 1993. A human mitochondrial ATP-dependent protease that is highly homologous to bacterial Lon protease. Proc. Natl. Acad. Sci. USA