1329 Electrophoresis 2009, 30, 1329–1341 Emanuele Loro1 Elisabetta Gianazza2 Silvia Cazzola1 Adriana Malena1 Robin Wait3 Shajna Begum3 Carmen Brizio4 Federica Dabbeni-Sala5 Lodovica Vergani1 1 Dipartimento di Neuroscienze, Facoltà di Medicina, Università degli Studi, Padova, Italy 2 Gruppo di Studio per la Proteomica e la Struttura delle Proteine, Dipartimento di Scienze Farmacologiche, Università degli Studi, Milano, Italy 3 Kennedy Institute of Rheumatology Division, Faculty of Medicine, Imperial College London, Hammersmith, London, UK 4 Dipartimento di Biochimica e Biologia Molecolare, Università degli Studi, Bari, Italy 5 Dipartimento di Farmacologia, Facoltà di Medicina, Università degli Studi, Padova, Italy Received September 5, 2008 Revised October 24, 2008 Accepted October 30, 2008 Research Article Development and characterization of polyspecific anti-mitochondrion antibodies for proteomics studies on in toto tissue homogenates We describe the characterization of polyclonal antibodies directed against the whole mitochondrial subproteome, as obtained by hyperimmunization of rabbits with an organelle fraction purified from human skeletal muscle and lysed by sonication. After 2-DE separations with either blue native electrophoresis or IPG as first dimension and blotting, the polyspecific antibodies detect 113 proteins in human muscle mitochondria, representative of all major biochemical pathways and oxidative phosphorylation (OXPHOS) complexes, and cross-react with 28 proteins in rat heart mitochondria. Using as sample cryosections of human muscle biopsies lysed in urea/thiourea/CHAPS, the mitochondrial subproteome can be detected against the background of contractile proteins. When comparing with controls samples from mitochondrial encephalomyopathy with lactic acidosis and stroke-like episodes patients, immunoblotting shows in the latter a drastic reduction for the subunits of OXPHOS complex I as well as an increase of several enzymes, including ATP synthase. This finding is the first evidence at the proteomic level of massive up-regulation in a number of metabolic pathways by which the affected tissues try to compensate for the deficit in the OXPHOS machinery. Keywords: Antibodies / Mitochondria / Proteomics 1 Introduction Immunological detection in an unpurified sample is an expedient alternative to general protein detection after purification. Specific recognition of a single antigen with Correspondence: Dr. Elisabetta Gianazza, Gruppo di Studio per la Proteomica e la Struttura delle Proteine, Dipartimento di Scienze Farmacologiche, Università degli Studi, via G. Balzaretti 9, I-20133 Milano, Italy E-mail: elisabetta.gianazza@unimi.it Fax: +39-0250318284 Abbreviations: ACADV, acyl-CoA dehydrogenase, verylong-chain specific; ACON, aconitate hydratase; ADT, ADP/ATP translocase; AL4A1, D-1-pyrroline-5-carboxylate dehydrogenase; BN, blue native; CISY, citrate synthase; COX, cytochrome c oxidase; ETF, electron transfer flavoprotein; G3P, glyceraldehyde-3-phosphate dehydrogenase; IDHP, isocitrate dehydrogenase; KCRS, creatine kinase; MDHM, malate dehydrogenase; MELAS, mitochondrial encephalomyopathy with lactic acidosis and stroke-like episodes; NDUFA, NADH-ubiquinone oxidoreductase; OXPHOS, oxidative phosphorylation; SDH, succinate dehydrogenase (ubiquinone); VDAC, voltagedependent anion-selective channel protein & 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim DOI 10.1002/elps.200800576 monospecific antibodies is most often the focus not only in quantifications on unprocessed biological samples but also in immunoblotting procedures after electrophoretic separations. Nonetheless, the use of polyspecific antisera proves useful in some instances. In routine clinical laboratory methods as well as in experimental protocols aiming at recognizing disease markers (e.g. [1–3]), patient’s sera are analyzed as containing antibodies against either ‘‘self’’ (normal or pathological tissue) or ‘‘not-self’’ antigens (infectious agents, parasites, allergens). Polyspecific antisera raised through experimental hyperimmunization are used in diagnostic immunoelectrophoresis [4, 5] and in microbiology/parasitology methods (for isolate subtyping and for characterization of immunodominant antigens/protective antibodies [6–10]) but also in biochemical investigations (isolation of cDNA clones encoding proteins of complex structures [11]). With the use of commercial anti-rat serum antibodies we could detect in 2-DE blots of whole-brain tissue homogenates a sharp increase of serum proteins in the affected versus the unaffected hemisphere of post-stroke rats [12]. The first aim of the present investigation is then to develop and characterize polyspecific antibodies directed against the whole mitochondrion subproteome. The main www.electrophoresis-journal.com 1330 E. Loro et al. use of such an immunological reagent would be in the specific detection of at least the most abundant mitochondrial proteins in crude tissue extracts. A relevant application could be with biochemical investigations either on needle biopsies from muscles of patients reporting myopathy symptoms or on cell cultures, when only minute amounts of samples are available. A panel of antisera of various specificity against mitochondrial proteins and protein complexes is commercially available (see at http:// www.mitosciences.com/immunocytochemistry.html) but no product fits the requirements of polyspecificity inherent to the approach we mean to take. The occurrence in human patients of anti-mitochondrial autoantibodies [13] suggests human muscle mitochondrial proteins may act as highly immunogenic heteroantigens in rabbits. As a second step, the polyspecific antibodies are applied to the evaluation of the mitochondrial subproteome in muscle specimens from patients with mitochondrial encephalomyopathy with lactic acidosis and stroke-like episodes (MELAS), a rare type of myopathy secondary to an A-to-G transition mutation at nucleotide pair 3, 243 in the dihydrouridine loop of mitochondrial tRNA(Leu)(UUR) [14]. No such sample has ever been investigated before with proteomic procedures. Electrophoresis 2009, 30, 1329–1341 natants were pooled and centrifuged at 7000g for 10 min at 41C. The mitochondria in the final pellet were suspended in 3 mL of 5 mM MOPS, pH 7.4, containing 220 mM mannitol, 70 mM sucrose, 1 mM EDTA, 2 mM b-mercaptoethanol (MSEM) and centrifuged at 9000g for 10 min at 41C. The washed mitochondria were immediately frozen and stored in liquid nitrogen. 2.2 Antibody production Procedures involving animals and their care were conducted at Dipartimento di Neuroscienze, Università degli Studi di Padova, in conformity with the institution’s guidelines, which are in compliance with national and international rules and policies. Two milligrams of mitochondria from human skeletal muscle, suspended in 0.1 M NaCl at a final concentration of 0.5 mg/mL, were sonicated four times at 60 W for 5 s and added volume-to-volume with incomplete Freund’s adjuvant. This mixture was injected, in five immunizations at 2 wk intervals, in each of two male New Zealand white rabbits (Charles River, Calco, CO). Polyclonal polyspecific antisera were collected 5 months after the last injection. 2 Materials and methods 2.3 Antibody purification 2.1 Preparation of muscle homogenates and mitochondria Control samples of human skeletal muscle (quadriceps) were obtained by open biopsies from subjects undergoing orthopedic surgery, after informed consent of the patients. Pathological muscles were from two Italian male MELAS patients [15]. Patient 1, already described in [16], was 42 years old and had 70% mutant mtDNA in muscle; patient 2 was 20 years old and had 80% mutant mtDNA. Both presented MELAS symptoms and bilateral hearing loss, vomiting, diarrhea and abdominal pain [15]. All muscle samples were frozen in liquid nitrogen and stored at 801C until use. Muscle homogenates were prepared using a glass pestle in a glass potter, with a buffer containing 20 mM HEPES, pH 7.4, 100 mM KCl, 1 mM EDTA, 2 mM b-mercaptoethanol (buffer A) [17]. For some experiments, cryosections, 10 mm thick, prepared from vastus lateralis biopsies, were solubilized in urea/thiourea/CHAPS and protein concentration was evaluated with 2-D Quanta Kit (GE-Amersham). Human muscle mitochondria were isolated from fresh muscle specimens as described in [17]. Tissue was minced, washed free of blood cells and homogenized in cold buffer A. The homogenate was centrifuged at 800g for 10 min at 41C. The resulting supernatant was filtered through gauze to remove fat while the pellet was resuspended in buffer A, further homogenized and centrifuged. The filtered super& 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Antibodies were purified from rabbit sera as described in [18]. The antiserum was adjusted to pH 8.5 with 0.1 N NaOH. An aliquot of 2 mL, in continuous stirring, was added dropwise with 7 mL of 0.4% rivanol (2-ethoxy 6,9-diaminoacridine lactate) solution. The supernatant, containing the immunoglobulins, was decanted, centrifuged to remove sediments and decolorized with activated charcoal (approx. 100 mg, eventually removed by filtration trough a paper disc). IgG, precipitated by incubation overnight at 41C with an equal volume (5 mL) of saturated ammonium sulfate, were collected by centrifugation at 4000g for 20 min at 41C and finally resuspended in a volume of saline equivalent to half the volume of the starting antiserum (1 mL). Sulfate residues were removed by chromatography on a Sephadex G-25 column (PD10 from GE-Amersham). The purified immunoglobulin concentration was determined by Lowry’s assay [19]. 2.4 SDS-PAGE and electroblotting Total proteins from human muscle homogenates and purified mitochondria were separated by SDS-PAGE on a 7.5–17.5%T PAA gradient in the discontinuous buffer system of Laemmli [20]. The protein pattern was either revealed by CBB staining or electroblotted to nitrocellulose (NC) membrane [21]. www.electrophoresis-journal.com Electrophoresis 2009, 30, 1329–1341 2.5 Analysis of oxidative phosphorylation complexes by blue native-PAGE Oxidative phosphorylation (OXPHOS) complexes were extracted as described in [22, 23]. Aliquots of 2 mg of mitochondria were suspended in 2 mL of 400 mM sucrose, 1 mM EDTA, 0.2 mM PMSF and 0.2 M MOPS, pH 7.2, and centrifuged at 32 000g for 30 min at 41C. The pellet was resuspended in 200 mL of 1 M aminocaproic acid, 50 mM Bis-Tris/HCl pH 7.0 (buffer B) and added with 37.5 mL of freshly prepared 10% w/v lauryl maltoside. After 10 min incubation, the samples were centrifuged at 128 000g for 30 min at 41C. The supernatants, containing solubilized OXPHOS complexes, were collected and their protein content was determined by Bradford assay [24]. Samples were then frozen in liquid nitrogen and stored at 801C up to several months. Blue native (BN)-PAGE was performed using a minigel apparatus (Miniprotean Bio-Rad) as described in [17]. Different amounts of proteins (from 7.5 to 50 mg), in a loading buffer containing 7% v/v of a 5% w/v solution of Serva Blue G in 1 M aminohexanoic acid and 1% glycerol, were run on a 5–13%T gradient gel (Bio-Rad) and processed as described in [23]. Aliquots of 5 mg of rat heart mitochondrial complexes and of 7.5 mg of human skeletal muscle complexes were loaded for native separation followed by ingel activity assays (Section 2.6.) or native electroblotting (Section 2.7). Aliquots of 50 mg of proteins were run in first dimension for 2-DE separation BN-PAGE–SDS-PAGE (Section 2.8). Proteomics and 2-DE 1331 dilution (a gift of Prof. F. Dabbeni-Sala, Padova, Italy), and polyclonal antibodies against FAD covalently bound to the flavoprotein subunit of complex II, used at 1:500 dilution (a gift of Prof. R. Brandsch, Freiburg, Germany). Incubation lasted 4 h. Test polyspecific antibodies raised in rabbits in our laboratories (Section 2.3.) were used at a concentration of 0.6 mg/mL; incubation lasted overnight. The secondary antibodies (GE-Amersham), tagged with HRP, were used at a 1:2500 dilution; peroxidase zymography was visualized by enhanced chemiluminescence (ECL, GE-Pharmacia). 2.8 2-DE BN-PAGE–SDS-PAGE Single subunits of OXPHOS complexes, separated in 1-DBN-PAGE, were resolved by SDS-PAGE on a 7.5–17.5%T PAA gradient in the discontinuous buffer system of Laemmli [20] similar to the protocol in [23]. Equilibration between first and second dimensions was obtained by incubation for 30 min at 371C, in 2 mL of 10 mM Tris/ acetate, pH 8.2, containing 4% w/v SDS, 0.3% v/v b-mercaptoethanol, 20% v/v glycerol and 1% v/v Pyronin Red. Individual 1-DE lanes were overlaid to each SDS-PAGE slab; the 2-DE gels were run at 50 V until the tracking dye passed the stacking gel, then at 150 V. A pair of gels was always processed in parallel: one to be stained with CBB, the other to be blotted into NC membrane for immunodetection with rabbit polyspecific antibodies. Both 2-DE patterns were then scanned and analyzed with commercial software. All the CBB-stained spots positive to immunodetection were excised for identification through trypsinolysis and ESI-MS/MS. 2.6 In-gel activity assay Immediately after the electrophoretic run, enzymatic colorimetric reactions were performed on the BN-PAGE gel, according to the protocols described in [17]. Aliquots of 5 mg of rat heart mitochondrial complexes and of 7.5 mg of human skeletal muscle complexes were loaded for native separation. Single lanes were incubated for the zymograms for NADH-ubiquinone oxidoreductase (NDUFA) (EC 1.6.5.3); for ATPase (ATP synthase) (EC 3.6.3.14); for cytochrome c oxidase (COX); and for succinate dehydrogenase (ubiquinone) (SDH) (EC 1.3.5.1). Enzymatic reactions were visible after 1 h of incubation at room temperature with rat heart and after 2–3 h with human skeletal muscle samples. A reference lane was stained with CBB. 2.7 Immunodetection Reference antibodies were complex-specific: commercial monoclonal antibodies directed against complexes I and IV (Molecular Probes) used at 1:500 and 1:750 dilution, respectively; polyclonal antibodies against complex III, used at 1:200 dilution (a gift of Dr. R. Bisson, Padova, Italy); polyclonal antibodies against complex V, used at 1:10 000 & 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 2.9 2-DE IEF–SDS-PAGE First dimension was run on 4–10 NL IPG [25], second dimension on a 7.5–17.5%T PAA gradient in the discontinuous buffer system of Laemmli [20]. With IPG, two procedures were tried: (i) polymerization of an anodic IP plateau [26], reswelling in 8 M urea, on-gel anodic sample application and run ‘‘upside up’’ in a Multiphor chamber [27] of 8 cm-long, 7 mm-wide strips; (ii) polymerization of both anodic and cathodic, 2 cm long, IP plateaus [26], in-gel rehydration with 7 M urea, 2 M thiourea and 4% w/v CHAPS and run ‘‘upside down’’ in a Protean IEF cell [28] of 12 cm-long and 3.5 mm-wide strips. Equilibration between first and second dimensions was according to standard protocols [29]. Second dimension was run in a MiniVE chamber (GE-Amersham), with slab size 85  70  1.5 mm3. Typical sample loads were 50–100 mg of mitochondrial proteins and 50–400 mg of either cryosection lysate or muscle homogenate. The latter contains a very high concentration of BSA from the extraction buffer (Section 2.1.); hence the total protein load in the experiments was proportionally higher. All detection procedures after 2-DE were as in Section 2.8. www.electrophoresis-journal.com 1332 Electrophoresis 2009, 30, 1329–1341 E. Loro et al. 2.10 MS In-gel digestion with trypsin was performed according to published methods [30–32] modified for use with a robotic digestion system (Genomic Solutions, Huntington, UK) [33]. Cysteine residues were reduced with DTT and derivatized by treatment with iodoacetamide. The gel pieces were dehydrated with acetonitrile and dried at 501C, prior to addition of modified trypsin (Promega, Madison, WI; 10 mL at 6.5 ng/mL in 25 mM ammonium hydrogen carbonate). After incubation at 371C for 8 h, peptides were sequentially extracted with 25 mM ammonium hydrogen carbonate, 5% formic acid and acetonitrile. Tandem electrospray mass spectra were recorded using a Q-Tof hybrid quadrupole/orthogonal acceleration time of flight spectrometer (Micromass, Manchester, UK) interfaced to a Micromass CapLC capillary chromatograph. Samples were dissolved in 0.1% aqueous formic acid, injected onto a Pepmap C18 column (300 mm  0.5 cm; LC Packings, Amsterdam, NL), and eluted into the mass spectrometer with an acetonitrile/0.1% formic acid gradient (5–70% acetonitrile over 20 min). The capillary voltage was set to 3500 V. A survey scan over the m/z range 400–1300 was used to identify protonated peptides with charge states of 2, 3 or 4, which were automatically selected for data-dependent MS/ MS analysis, and fragmented by collision with argon. The resulting product ion spectra were transformed onto a singly charged m/z axis using a maximum entropy method (MaxEnt3, Micromass) and proteins were identified by correlation of uninterpreted spectra to entries in Swiss-Prot/ TrEMBL, using ProteinLynx Global Server (Version 1.1, Micromass). The database was created by merging the FASTA format files of Swiss-Prot, TrEMBL and their associated splice variants (1 768 175 entries at the time of writing). No taxonomic, or protein mass and pI constraints were applied. One missed cleavage per peptide was allowed, and the fragment ion mass tolerance window was set to 100 ppm. For confirmation of identifications spectra were also searched against the NCBI nr database using MASCOT. Hits that rested on a single matching peptide were confirmed by manual interpretation of sequence specific fragment ions using the MassLynx program PepSeq (Micromass). 3 Results To evaluate the diverse immunological responses induced by mitochondrial proteins in the two hyperimmunized rabbits, proteins from whole human skeletal muscle homogenate and from purified skeletal muscle mitochondria were resolved in gradient SDS-PAGE, electroblotted and probed with immunoglobulins purified from the two polyspecific antisera. ECL development of HRP zymography had identical duration for all NC membranes. Figure 1A shows how the antibodies produced by the two rabbits (left and middle panels) recognize different specific mitochondrial proteins with different affinity, both when in a purified sample and in whole muscle homogenate. On the basis of Figure 1. Immunodetection of mitochondrial proteins with polyspecific antibodies raised in rabbits after 1-DE or 2-DE and electroblotting. A 5 1-DE of varying amounts, as marked, of mitochondria purified from human skeletal muscle and of whole human skeletal muscle homogenates. Primary antibody from rabbit-1 in the blot to the left, from rabbit2 in the middle, from both rabbits to the right. Rightmost, Mr listing of protein standards. B 5 2-DE of mitochondria, left (same as in Fig. 3), and of muscle homogenate, right (same as in Fig. 6). The extra spots detected in muscle homogenate versus purified mitochondria are marked with their names. & 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com Electrophoresis 2009, 30, 1329–1341 Proteomics and 2-DE 1333 whole muscle homogenate and not in mitochondria (Fig. 1A) and due to cross-reactivity with serum albumin (see Tables S1 and S2 in Supporting Information). When comparing the immunostaining of the two types of samples after 2-D (IEF–SDS-PAGE) (Fig. 1B) two more peculiar spots were present in whole muscle homogenate (right) but not in purified mitochondria (left): (i) one at high molecular weight, due to cross-reactivity with myosin and (ii) the other at pI 7–29 kDa, identified as carbonic anhydrase, a protein that accounts for 8% of total muscle proteins (see Tables S1 and S2 in Supporting Information). BN-PAGE, developed to investigate membrane-bound protein complexes and designed for the mitochondrial OXPHOS system [22], was used to evaluate which complexes, and which of their constituent subunits, were recognized by the rabbit polyspecific antibodies. Figure 2A illustrates the 1-DE run on a 5–13%T PAA gradient gel of 7.5 mg of human skeletal muscle mitochondrial complexes isolated by a one-step protocol [23]. The separation of the individual complexes, on the basis of their size, is visualized both by CBB staining and by in-gel activity assays specific for the different complexes: complex II or SDH (EC1.3.5.1), complex IV or cytochrome c oxidase (EC 3.6.3.14) (COX), complex V or ATPase (EC 3.6.3.14) and complex I or NADH dehydrogenase (EC1.6.5.3) (NDUFA). We also analyzed the relative presence of the different complexes by native electroblotting and immunodetection with all available monospecific antibodies, lane ‘‘blot (mono-)’’ (a composite picture from five independent experiments). These signals are to be compared with the banding obtained with the rabbit polyspecific antibodies, lane ‘‘blot (poly-)’’. The polyspecific antibodies recognize all the complexes in a specific way, but with different affinity. In fact complexes I, V, III and IV develop bands of similar intensity to blot (mono-), indicating a high antigenic activity of these four complexes; on the contrary complex II gives 3 Figure 2. Detection of mitochondrial complexes purified from this evidence, in all of the following experiments a balanced pool of the immunoglobulins from the two antisera was used for immunodetection (right panel). The pattern of 1-D immunodetection presented a single discrepancy between purified mitochondria and whole cell extract – a band, about 66 kDa in size, detected only in & 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim human muscle mitochondria: (A) 5 after 1-DE (BN-PAGE); (B and C) 5 after 2-DE (BN-PAGE–SDS-PAGE). (A) The strips, bottom up, correspond to CBB staining; to immunodetection with polyspecific antibodies raised in rabbits (blot poly-); to a composite picture of the immunodetections of individual complexes (banding as marked) with already available monospecific antibodies (blot mono-); to the zymogram for NDUFA (EC 1.6.5.3); to the zymogram for ATPase (ATP synthase F0) (EC 3.6.3.14); to the zymogram for COX; to the zymogram for SDH (EC 1.3.5.1). Protein load 5 7.5 mg. (B) The maps correspond to CBB staining (bottom) and to immunodetection with polyspecific antibodies raised in rabbits after electroblotting (top). Protein load as marked. The spots detected by the polyclonal antisera were excised from the CBB gel and processed for protein identification through trypsinolysis and ESI-MS/MS. Each of these spots is marked with a number, the same as associated to the matching entry in the alphabetical list of identified proteins in Table S1. (C) The identified spots are labeled by function, in order to highlight supramolecular assemblies. I–V 5 OXPHOS complexes I–V; downwards triangles 5 enzymes of carbohydrate and amino acid metabolism; upwards triangles 5 enzymes of lipid metabolism; squares 5 PDH subunit; pentagon 5 prohibitin; circle 5 transport. www.electrophoresis-journal.com 1334 E. Loro et al. a weaker signal when probed with the polyspecific than with the monospecific antibodies. To further dissect the properties of the rabbit polyspecific antibodies raised in our laboratories, the single subunits of the native complexes were separated in second dimension by SDS-PAGE, as also shown in Fig. 2. One gel was stained with CBB (Fig. 2B, bottom); another blotted and probed with the polyspecific antibodies (Fig. 2B, top). All the spots that were positive both to immunodetection and to CBB staining were excised, trypsin-digested and analyzed by ESI-MS/MS. The identified proteins are listed, in alphabetical order, in Supporting Information Table 1S, each entry being identified by a number that also marks the corresponding spot in Fig. 2B (CBB stain panel). Table 2S in Supporting Information provides the full list of matching peptide sequences for each spot. In panel 2C the proteins are labeled on the basis of their biochemical function or of the metabolic pathway they take part into. In several instances we observe the comigration of subunits of different OXPHOS complexes, and/or the comigration of OXPHOS subunits with various matrix or membrane proteins. This finding supports the concept of supramolecular interactions among enzymes resulting in the organization of a ‘‘metabolon’’, defined as a functional complex of enzymes sequential in a metabolic pathway and bound to a cell structural element. A similar procedure was performed in another set of experiments, in which 50 mg of human skeletal muscle mitochondrial proteins were resolved by IEF on IPG, followed by SDS-PAGE in the second dimension. Figure 3 compares the protein patterns after immunoblotting and after CBB staining. The MS-identified proteins are listed in Supporting Information Table 1S, while matching sequences are listed in Table 2S of Supporting Information. Cross-immunoreactivity towards related antigens, namely mitochondrial complexes from a different species, the rat, and a different type of muscular tissue, the heart, was then tested, as illustrated in Fig. 4. Resolution of rat heart OXPHOS by 2-DE separation BN-PAGE–SDS-PAGE was performed as described in Fig. 2B. As shown in the list of Table 1S, many of the rat proteins are detected by the polyspecific antibodies raised against the corresponding human proteins. To test whether the use of the polyspecific antibodies could detect differences between control and pathological samples, we thoroughly characterized mitochondria from MELAS patients. As shown in Fig. 5A by the densitometric evaluation of the zymograms developed after BNPAGE, enzymatic activity of ATPase and of SDH are unaffected by the pathology. Conversely, the activity of COX is strongly reduced and the activity of NDUFA is virtually abolished. Figure 5B is the 2-DE (BN-PAGE–SDS-PAGE) of the mitochondrial complexes from a MELAS sample as immunodetected after electroblotting; the arrowheads mark changes in spot intensity versus a matched control. A decrease in the intensity of complex I subunits is evident together with an increase of ADP/ATP translocase (ADT) and of ATP synthase subunits (B, D, O, g and f subunits). & 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Electrophoresis 2009, 30, 1329–1341 Figure 3. Detection of mitochondrial proteins purified from human skeletal muscle after 2-DE (IEF–SDS-PAGE). CBB staining (bottom) and immunodetection with polyspecific antibodies raised in rabbits after electroblotting (top). Protein load as marked. Spots analysis and map labeling were as in Fig. 2. Finally, the applicability of rabbit polyspecific antisera was tested towards a high background of unrelated antigens such as in total skeletal muscle. In a first step (Fig. 6A, top) the patterns of MELAS muscle lysates (right; N 5 2) were compared to control samples after SDS-PAGE and immunoblotting; Fig. 6B plots the ratio between the integrated optical densities of the stained bands, MELAS versus controls, when normalized for the b-actin signal in each lane (Fig. 6A, bottom). Band intensity is higher in pathological samples at Mr ca. 80 and 40 kDa, much higher at Mr ca. 30 kDa. Then, as shown in Fig. 6C, 300 mg of whole proteins from normal and MELAS skeletal muscle were resolved by 2-DE through IEF–SDS-PAGE. Even at this high concentration, the most prominent spots detected by Ponceau staining are only slightly positive to immunodetection. In this 2-DE procedure alternative protocols for sample preparation and IPG run had been compared (data not shown). Direct solubilization of muscle cryosections in urea/thiourea/CHAPS and anodic ongel loading afforded both highest sample loading and least streaking. As a further advantage, the use of cryosections requires the lowest total amount of material with virtually no waste and is thus compatible with biopsy processing. www.electrophoresis-journal.com Proteomics and 2-DE Electrophoresis 2009, 30, 1329–1341 1335 Figure 5. Characterization of mitochondrial proteome in MELAS. (A) Ratios between the densitometric data for enzyme activities after BN-PAGE in control and MELAS samples; abbreviations for protein names are as in Fig. 2A. (B) Immunoblotting after 2-DE (BN-PAGE–SDS-PAGE) on mitochondrial complexes. The upwards and downwards arrowheads stand for increase or decrease of spot intensity versus control. Figure 4. Detection of mitochondrial complexes purified from rat heart mitochondria after 2-DE (BN-PAGE–SDS-PAGE). CBB staining (middle) and immunodetection with polyspecific antibodies raised in rabbits after electroblotting (top). Protein load as marked. Spots analysis and map labeling were as in Fig. 2. Bottom panel: same as in Fig. 2C. Figure 6C shows a clearly increased abundance of a number of immunoreactive components in the total MELAS muscle lysate in comparison with a matched control sample. This difference may be observed in the middle panels (blotting and detection with polyspecific antibodies). Specifically the areas of increased immunological reactivity correspond to the regions where, in Fig. 3, the following proteins are found to migrate: ACADV (acyl-CoA dehydrogenase, very-long-chain specific), AL4A1 (D-1-pyrroline-5carboxylate dehydrogenase), aconitate hydratase (ACON), ATP synthase a and b (ATPA and ATPB), citrate synthase (CISY), glyceraldehyde-3-phosphate dehydrogenase (G3P), isocitrate & 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim dehydrogenase (IDHP), KCRS (creatine kinase), MDHM (malate dehydrogenase) and VDAC1 (voltage-dependent anion-selective channel protein 1; synonym: porin). 4 Discussion Given the importance of mitochondria in human diseases [15], protocols that permit the analysis of the mitochondrial subproteome from small amounts of biopsy tissue and even cell culture material would have great diagnostic utility. In this report we describe the possibility to develop polyspecific antibodies directed against the whole human mitochondrial subproteome by rabbit hyperimmunization with purified skeletal muscle mitochondria, and the application of these antibodies in protocols using small amounts of human muscle. Such a strategy overcomes the limitations of the protocols www.electrophoresis-journal.com 1336 E. Loro et al. Electrophoresis 2009, 30, 1329–1341 Figure 6. Detection of mitochondrial proteins in whole human skeletal muscle from control and MELAS patients. (A) Comparison of the SDS-PAGE pattern for five controls (left) and two MELAS (right) muscles after immunoblotting using polyclonal anti-mitochondrion antibodies (top) and after reprobing the NC membrane with commercial anti-actin antibodies (bottom). (B) Results from densitometric evaluations of the SDS-PAGE patterns in panel A, plotted as ratios between MELAS and control samples. (C) 2-DE (IEF–SDSPAGE) and electroblotting of lysed cryosections from control and MELAS muscle: Ponceau Red staining (left), immunodetection with polyspecific antibodies raised in rabbits (right). Experimental: muscle cryosections directly lysed in urea/ thiourea/CHAPS, loaded on gel near the anode. The upwards and downwards arrowheads stand for increase or decrease of spot intensity versus control. described thus far to obtain purified mitochondria from gram amounts of muscle as starting material – a requirement hardly met with human specimens. 4.1 Characterization of the polyspecific anti-mitochondrion antibodies Specificity of antibodies rests primarily on the purity of the antigen. The mitochondria used to produce the polyclonal antibodies belong to the same batch already used to define the reference 2-DE map of human muscle mitochondria [28]. In this preparation MS procedures identified only a negligible number of non-mitochondrial proteins [28], which has been confirmed by the new MS data listed in Tables 1S and 2S of Supporting Information. & 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim The characterization of polyspecific antibodies specificity was performed against human muscle mitochondria and rat heart mitochondria. About 200 gel features from 2-DE, either BN-PAGE–SDS-PAGE or IEF–SDS-PAGE, were recognized, and analyzed with ESI-MS/MS. The antibodies detected (Table 1S) 20 homologous proteins both in human and in rat samples, 85 proteins only in human muscles and 8 proteins only in rat heart. This high degree of cross-reactivity may be explained by the high sequence identity between mammalian mitochondrial proteomes as it had already been reported for humans, mice and cows (Mitomap [34, 35]). On the basis of Entrez Gene [36], the Human Protein Reference Database [37] and MitoP [38], 113 proteins were functionally categorized; no function could be assigned to 21 polypeptides listed in TrEMBL [39]. The distribution of the functionally classified proteins is shown in Fig. 7 and www.electrophoresis-journal.com Electrophoresis 2009, 30, 1329–1341 Figure 7. Distribution of the 113 functionally classified proteins detected by polyspecific antibodies. Assignments to functional classes were made on the basis of (i) classifications provided in the MITOP database and (ii) information provided at National Center for Biotechnology Information’s Entrez Gene website (http://www.ncbi.nih.gov/entrez/query.fcgi?db 5 gene) and Human Protein Reference Database (http://www.hprd.org). their localization in enzymatic pathways is visualized in Fig. 8. Depending on the electrophoretic procedure in first dimension, the proteins recognized by the polyspecific antibodies in the 2-DE maps are mostly proteins embedded in, or connected to, the mitochondrial membranes when resolved by BN-PAGE (Figs. 2 and 4) and mostly proteins from the mitochondrial matrix space when resolved by IEF (Fig. 3). As an example, the ADT1 is a very hydrophobic and basic inner-membrane carrier, and one of the principal proteins in mitochondrial preparations [40]. ADT1 was not observed in IEF–SDS-PAGE but was found in BN-PAGE–SDS-PAGE (Figs. 2B and 4). 4.1.1 OXPHOS Subunits of the five complexes that constitute the OXPHOS machinery represent the main group, amounting to 48% of the functionally classified mitochondrial proteins recognized by the antibodies (Fig. 7). This includes 14/45 complex I subunits, 2/4 complex II subunits (plus 2 complex II-related subunits), 7/11 complex III subunits, 4/13 complex IV subunits and 11/16 complex V subunits. 4.1.2 Metabolism The enzymes of the TCA cycle (plus pyruvate dehydrogenase) are represented by 11 distinct proteins (enzymes or enzyme subunits) whereas lipid metabolism is represented by 16 proteins – an indication of the primary role of fatty acid oxydation as energy source in skeletal muscle. 4.1.3 Enzyme association The protocol BN-PAGE–SDS-PAGE allows the isolation of active OXPHOS enzymes without loss of subunits, as shown & 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Proteomics and 2-DE 1337 by in-gel activities in Fig. 2A. As apparent in Figs. 2B and 4, and confirmed by MS analysis, the second dimension electrophoresis resolves from the position where each OXPHOS complex has migrated in first dimension not only the specific complex subunits but also significant amounts of other mitochondrial membrane proteins together with unexpected proteins known to be localized in the matrix (bottom panels in Figs. 2 and 4). Some of these proteins are involved in metabolic processes (TCA cycle, amino acid and lipid metabolism), some take part in protein folding (prohibitin) or are transport proteins (VDAC and ADT1). These impurities might be due to simple contamination; however it appears that the concentration of detergent we used (1.5% lauryl maltoside, chosen to minimally disrupt complexes and to maintain their activity according to Zerbetto et al. [17]) may be too low to weaken native contacts with OXPHOS complexes. This hypothesis is confirmed by several reports that, using different technical approaches, described physical interactions between OXPHOS complexes and proteins of b-oxidation and TCA cycle, resulting in enzyme assemblies of high functional efficiency [41–45]. The in-gel activity assay for complex I (Fig. 2A) recognizes a sharp band at 1 MDa and a lighter band at 200 kDa. Consistently we observe the presence of two complex I subunits (NADH B22 subunit, NADH 39 kDa subunit) intermixed with complex IV subunits after second dimension migration (Fig. 2B). Comigration of mixed elements of complex I, three subunits of complex III (core subunits 1 and 2, subunit 14 kDa) and complex IV isolated in active state (Fig. 2A) seems to indicate an interaction preserved by a mild solubilization protocol and made evident by polyspecific antibodies. These findings are in line with previous evidences showing that complexes I, III and IV assemble into so-called ‘‘supercomplexes’’ leading to the formation of supramolecular structures called ‘‘respirasomes’’ (first described in [46] and reviewed in [47, 48]), which seem to be of great functional importance [47, 48]. Recent structural analyses revealed the association of ATP synthase with the phosphate carrier and ADT1 into the functional metabolon called ‘‘ATP synthasome’’ [49, 50]. In the human skeletal sample, ADT runs in association with complex I, V–III, IV–II (Fig. 2C) and with VDAC, consistent with the described associations in ATP synthasome [49, 50] and in VDAC–ADT interaction complexes [51]. Subunits of complex V, running in association with subunits of complexes I, III and IV (subunits 2 and 4), may indicate that, under our experimental conditions, a small amount of complex V was retained in supercomplexes of OXPHOS machinery, as previously observed in experiments of immunocapture of functionally active human heart complex V [42] and complex I [43]. Subunits of complex I of muscle mitochondria run also in association with prohibitin 2 and mitofilin (Fig. 2C and Supporting Information Table 1S). Mitofilin is a mitochondrial inner membrane protein controlling mitochonwww.electrophoresis-journal.com 1338 E. Loro et al. Electrophoresis 2009, 30, 1329–1341 Figure 8. Mapping of the proteins detected by the polyspecific anti-mitochondrion antibodies on a simplified chart of the main metabolic pathways. Names of the relevant enzymes are boxed, according to the following graphical code: light grey filling for all identified enzymes; dark grey filling for the enzymes (or enzyme subunits) up-regulated in MELAS patients; black filling for the enzymes (or enzyme subunits) decreased in MELAS patients; framing with both a continuous and a dashed thick line for enzymes (or enzyme subunits) recognized in both human skeletal muscle and rat heart samples; framing with a thick dashed line for enzymes recognized only in rat heart samples; framing with a thin continuous line and white filling for enzymes not identified in the present investigation. Abbreviations were as in the caption to Fig. 1 of [28], except: AATC: aspartate aminotransferase, cytoplasmic; AATM: aspartate aminotransferase, mitochondrial; ACSM: medium-chain acyl-Coa synthetase; ADT1: ADP/ATP translocase 1; AL4A1:D-1-pyrroline-5carboxylate dehydrogenase; ALDH2: aldehyde dehydrogenase; CMC1: calcium-binding mitochondrial carrier protein Aralar1 (mitochondrial aspartate glutamate carrier 1); DECR: 2,4-dienoyl-CoA reductase; IDHP: isocitrate dehydrogenase (NADP); KCRS: creatine kinase, sarcomeric; M2OM: mitochondrial 2-oxoglutarate/malate carrier protein; MDHC: malate dehydrogenase, cytoplasmic; MDHM: malate dehydrogenase, mitochondrial; MPC: mitochondrial pyruvate carrier (unidentified); NNTM: NAD(P) transhydrogenase; PRDX3: thioredoxin-dependent peroxide reductase; SODM: superoxide dismutase (Mn); THIOM: thioredoxin; TRXR2: thioredoxin reductase 2; VDAC1: voltage-dependent anion-selective channel protein 1; VDAC2: voltage-dependent anion-selective channel protein 2. drial cristae morphology that assembles into large multimeric complexes, for which a close cosedimentation with complex I has been previously demonstrated [45]. Prohibitins are a family of proteins with low molecular weight, which assemble into large structures in the mitochondrial inner membrane and act as membrane-bound chaperones for the stabilization of newly synthesized mitochondrial translation products [52]. In particular, a role in the assembly or degradation of mitochondrial complex I has been suggested [53]. We found that prohibitins comigrate with complex I at 1 MDa (Fig. 2C), in agreement with previous findings [54]. In rat heart, SDH flavoprotein subunits comigrate with electron transfer flavoprotein (ETF) a subunit, IDHP and ADT (Fig. 4, bottom). Similarly, in human muscle, SDH flavoprotein comigrates with medium chain acyl-CoA & 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim dehydrogenase, CISY, IDHP and ADT, whereas NADH dehydrogenase subunit 75 kDa comigrates with the a subunit of trifunctional enzyme (Fig. 2, bottom). Existence of functional assemblies between enzymes of the b-oxidation pathway and OXPHOS (specifically, complexes I and II) had been previously reported [55, 56] with the suggestion of a functional channeling of NADH and CoA/ acetylCoA by a b-oxidation metabolon [55]. In a similar way we interpret the presence of different matrix soluble enzymes such as CISY-IDHP and transenoylCoA isomerase (Fig. 2B) at a position (200 kDa) corresponding to activity and presence of complex IV in first dimension as supportive evidence of interactions between inner membrane and matrix. Previous reports [57, 58] had already shown interactions among TCA cycle enzymes as well as interactions between these enzymes and the cis www.electrophoresis-journal.com Electrophoresis 2009, 30, 1329–1341 surface of the inner mitochondrial membrane. It was proposed that these multiple interactions could result in a supramolecular protein–membrane complex defined as the TCA cycle metabolon [58]. 4.1.4 Non-mitochondrial proteins Four distinct cytoplasmic proteins – the glycolytic enzyme G3P, carbonic anhydrase 3, and the structural proteins actin and myosin – have been recognized by the polyspecific antimitochondrion antibodies. These may represent either minor contaminants, proteins that have physical associations with the mitochondria or proteins/protein isoforms that are also present inside the mitochondria. It is noteworthy that the recognized isoform of G3P exhibits basic pI value and may electrostatically associate with the mitochondrial outer membrane, as it has been previously reported [40, 59]. Actin and myosin may represent artifacts of the isolation procedure, but may also reflect the cytoskeletal architecture with which mitochondria are intimately associated. In a recent report 2-DE (BN–SDSPAGE) of OXPHOS from five rat organs resulted in the identification of 92 non-redundant soluble and membraneembedded non-OXPHOS proteins including 36 proteins known or presumed to be localized to non-mitochondrial compartments [60]. 4.1.5 Undetected proteins Since the antibodies identify only about 10% of the known mitochondrial proteins, it is likely that the list is underestimated. We did not observe antibodies against proteases, such as mitochondrial processing peptidase, against proteins involved in cell death, such as Smac and AIF, in protein targeting, such as components of the protein translocation machinery, and in RNA–DNA–protein synthesis, such as mitochondrial ribosomal proteins or DNA polymerase. Although we have assembled a list of physico-chemically and functionally diverse proteins recognized by our polyspecific antibodies, our choice of methodology may not be optimal for certain proteins. For example, in the case of OXPHOS complexes our antibodies did not identify very hydrophobic proteins (for example ATPase subunit C, the proteolipid anchor of complex V). Similarly the absence of other proteins may reflect either a low abundance or physical properties/ chemistries unfavorable to detection by our analytical strategies (BN-PAGE, IEF). Therefore, additional investigations and strategies, such as immunoprecipitation, may further extend the list of the mitochondrial proteins identified by these polyspecific antibodies. 4.2 Analysis of lysates from whole muscle of MELAS patients No proteomic data on specimens from MELAS patients are currently available in the literature. After a transcriptomic & 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Proteomics and 2-DE 1339 investigation, in 1999 Heddie et al. [61] reported on the ‘‘coordinate induction of energy gene expression in tissues of mitochondrial disease patients’’, including a MELAS subject. Our data, based on the results of electrophoretic separations plus immunoblotting using the polysepcific antimitochondrion antibodies, fully confirm their data and extend the list of the affected proteins specific to the mitochondrial compartment. BN-PAGE-based procedures focus information on proteins of mitochondrial membranes. The lack of complex I activity when MELAS samples are assayed in-gel after 1-DE separation (Fig. 5A) is in agreement with the finding of a strongly reduced presence of both CBB stained (data not shown) and immunoreactive material migrating at the position typical for NDUFA subunits after 2-DE separation (BN-PAGE–SDS-PAGE) (Fig. 5B). The reduction of activity for complex IV (Fig. 5A) has no obvious counterpart in a quantitative alteration in 2-DE pattern (either CBB- or immunostained, Fig. 5B). However, a qualitative alteration of the pattern is observed, featuring a minor shift in BNPAGE mobility, which could be caused by ROS-mediated oxidation, as it has been previously suggested [62]. The normal ATPase activity is consistent with previous observations of milder bioenergetic defect in cells harboring 3243 MELAS mutation in comparison with other mtDNA point mutations [63, 64] Conversely, after 1-DE with SDS-PAGE (Fig. 6A-B), the use of polyclonal antibodies on MELAS whole muscle samples shows an overall increase in the amount of immunoreactive material in comparison with whole control muscle. This finding is likely connected with the distinctive proliferation of mitochondria in the tissues of mitochondrial disease patients. Indeed, a hallmark of MELAS muscles is the histological evidence of an increased number of subsarcolemmar mitochondria made evident by the trichromic staining [15]. However, the percent increase of protein content in MELAS versus control is different in different molecular size ranges, and is maximal for lower molecular weight species. With a more detailed investigation through 2-DE protocols (IEF–SDS-PAGE), specific energy pathways are found to be affected: ADT1 is found increased in 2-DE with BN-PAGE as first dimension, ACADV, AL4A1, ACON, CISY, G3P, IDHP, KCRS, MDHM and VDAC in 2-DE with IEF as first dimension, ATP synthase subunits with both types of 2-DE experiments. The increase in the number of mitochondria as well as in the amount of specific enzymes is most likely a compensatory phenomenon in response to energy deprivation. Our data are thus consistent with the transcriptomics findings reported by Heddi et al. [61] who observed increased levels of transcripts for mtDNA as well as for nuclear OXPHOS genes. Specifically, among the 16 genes tested for, they reported up-regulated transcription for ATPB, ADT and KCRS, for which we find an increase in protein levels. Their observations and ours thus suggest that muscle tissue attempts to counteract OXPHOS defects associated with mtDNA mutations by stimulating mitochondrial biogenesis and by boosting the metabolic pathwww.electrophoresis-journal.com 1340 Electrophoresis 2009, 30, 1329–1341 E. Loro et al. ways that may circumvent the absence of OXPHOS complex I as well as the reduction in activity of complex IV. In the metabolic map of Fig. 8 most of the affected enzymes are seen to cluster as closely linked to one another. Four enzymes sequential in the Krebs cycle – MDHM, CISY, ACON and IDHP – are involved. Complex II directly acts on Krebs cycle intermediates by converting succinate to fumarate and complex III accepts reducing equivalents from FADH2 through ETF and UQ. FADH2 is mainly produced by fatty acid catabolism, and ACADV, another of the upregulated enzymes in MELAS, catalyzes one of the ratelimiting steps of this process. The practical relevance of the up-regulation of the Krebs cycle may be highlighted by the report of a case of successful treatment with succinate of an uncontrolled convulsive MELAS patient [65]. Increasing evidence supports the notion that G3P is a housekeeping gene with multiple functions linked to various intracellular localizations [66]. The role of G3P in mitochondria and its up-regulation in MELAS are worth notice on the basis of G3Ps novel connection with cellular survival [67] and deserve further investigation. Similarly, the up-regulation in MELAS of AL4A1 might be the result of a stress response [68]. AL4A1 catalyzes a necessary step in the proline-NO homeostasis. Proline oxidase, a mitochondrial enzyme that oxidizes proline to form pyrroline-5-carboxylate, the substrate of AL4A1, is inhibited by lactate [69], and lactic acidosis is one of the most typical features of MELAS. Reduction in mitochondrial production of pyrroline-5-carboxylate secondary to lactate inhibition was hypothesized as the most likely cause of hypocitrullinemia, recorded in MELAS patients, and consequent altered NO homeostasis [70]. Hypocitrullinemia may be exacerbated by the up-regulation of AL4A1, further unbalancing these metabolic pathways. This work was supported by grants from AFM (Grant 11643, Call 2005), MIUR (FIRB, Grant RBNE03B8KK, Call 2003) and from Università degli Studi di Milano (FIRST 2006) to Dr. Elisabetta Gianazza; from AFM (Grant 11032, Call 2005) to Lodovica Vergani. C.B. was supported by a postdoctoral research fellowship (Giovani Ricercatori) financed by FIRB (Grant RBNE03B8KK, Call 2003). The authors have declared no conflict of interest. 6 References [1] Bernatsky, S., Ramsey-Goldman, R., Clarke, A., Curr. Opin. Rheumatol. 2006, 18, 129–134. [2] Fernandez Madrid, F., Cancer lett. 2005, 230, 187–198. [3] Saif, M. W., Zalonis, A., Syrigos, K., Expert Opin. Biol. Ther. 2007, 7, 493–507. [4] Grabar, P., Williams, C. A., Biochim. Biophys. Acta 1953, 10, 193–194. [5] Clarke, H. G., Freeman, T., Clin. Sci. 1968, 35, 403–413. [6] Needham, G. R., Jaworski, D. C., Simmen, F. A., Sherif, N., Muller, M. T., Exp. Appl. Acarol. 1989, 7, 21–32. [7] Petersen, C., Gut, J., Leech, J. H., Nelson, R. G., Infect. Immun. 1992, 60, 2343–2348. [8] Sambri, V., Armati, S., Cevenini, R., FEMS Immunol. Med. Microbiol. 1993, 7, 67–71. [9] Fallon, P. G., Doenhoff, M. J., Parasite Immunol. 1995, 17, 261–268. [10] Dumke, R., Catrein, I., Pirkil, E., Herrmann, R., Jacobs, E., Int. J. Med. Microbiol. 2003, 292, 513–525. [11] Birkett, C. R., Parma, A. E., Gerke-Bonet, R., Woodward, R., Gull, K., Gene 1992, 110, 65–70. [12] Sironi, L., Guerrini, U., Tremoli, E., Miller, I. , Gelosa, P., Lascialfari, A., Zucca, I. et al., J. Neurosci. Res. 2004, 78, 115–122. 5 Concluding remarks [13] Bogdanos, D. P., Baum, H., Vergani, D., Clin. Liver Dis. 2003, 7, 759–777, vi. In conclusion in the present report we described that polyspecific anti-mitochondrion antibodies may be used to detect informative features on the main components of mitochondrial subproteome from small amount of human muscle of patients, processed through 1-D and 2-D electrophoresis. The same may be expected to apply also to cell cultures. For the latter, it may then be possible to test in parallel the effects of a number of treatments either aiming at a more detailed definition of the underlying biochemical defects or mimicking a possible therapeutic intervention while using a limited amount of material and avoiding lengthy fractionation procedures. In addition to the qualitative and quantitative information on individual proteins, such as post-translational modifications and/or mass reduction, we deem that this new tool could help shed further light on the supramolecular organization of mitochondrial enzymes and on their association with nonmitochondrial proteins. [14] Goto, Y., Nonaka, I., Horai, S., Nature 1990, 348, 651–653. & 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim [15] McFarland, R., Taylor, R. W., Turnbull, D. M., Curr. Top. Dev. Biol. 2007, 77, 113–155. [16] Vergani, L., Malena, A., Sabatelli, P., Loro, E., Cavallini, L., Magalhaes, P., Valente, L. et al., Brain 2007, 130, 2715–2724. [17] Zerbetto, E., Vergani, L., Dabbeni-Sala, F., Electrophoresis 1997, 18, 2059–2064. [18] Hurn, B. A. L., Chantler, S. M., Methods Enzymol. 1980, 70, 104–141. [19] Lowry, O. H., Rosebrough, N. J., Farr, A. L., Randall, R. J., J. Biol. Chem. 1951, 193, 265–275. [20] Laemmli, U. K., Nature 1970, 227, 680–685. [21] Towbin, H., Staehelin, T., Gordon, J., Proc. Natl. Acad. Sci. USA 1979, 76, 4350–4354. [22] Schägger, H., von Jagow, G., Anal. Biochem. 1991, 199, 223–231. www.electrophoresis-journal.com Electrophoresis 2009, 30, 1329–1341 [23] Schägger, H., Cramer, W. A., von Jagow, G., Anal. Biochem. 1994, 217, 220–230. [24] Bradford, M. M., Anal. Biochem. 1976, 72, 248–254. [25] Gianazza, E., Giacon, P., Sahlin, B., Righetti, P. G., Electrophoresis 1985, 6, 53–56. [26] Gianazza, E., Guerini Rocco, A., Marchetto, A., Vergani, L., Electrophoresis 2007, 28, 2953–2956. [27] Fratelli, M., Demol, H., Puype, M., Casagrande, S., Eberini, I., Salmona, M., Bonetto, V. et al., Proc. Natl. Acad. Sci. USA. 2002, 99, 3505–3510. [28] Gianazza, E., Vergani, L., Wait, R., Brizio, C., Brambilla, D., Begum, S., Giancaspero, T. A. et al., Electrophoresis 2006, 27, 1182–1198. [29] Görg, A., Postel, W., Weser, J., Günther, S., Strahler, S. R., Hanash, S. M., Somerlot, L., Electrophoresis 1987, 8, 122–124. [30] Jeno, P., Mini, T., Moes, S., Hintermann, E., Horst, M., Anal. Biochem. 1995, 224, 75–82. [31] Wilm, M., Shevchenko, A., Houthaeve, T., Breit, S., Schweigerer, L., Fotsis, T., Mann, M., Nature 1996, 379, 466–469. [32] Shevchenko, A., Wilm, M., Vorm, O., Mann, M., Anal. Chem. 1996, 68, 850–858. [33] Wait, R., Gianazza, E., Eberini, I., Sironi, L., Dunn, M. J., Gemeiner, M., Miller, I., Electrophoresis 2001, 22, 3043–3052. [34] Kogelnik, A. M., Lott, M. T., Brown, M. D., Navathe, S. B., Wallace, D. C., Nucleic Acids Res. 1996, 24, 177–179. [35] Schilling, B., Murray, J., Yoo, C. B., Row, R. H., Cusack, M. P., Capaldi, R. A., Gibson, B. W., Biochim. Biophys. Acta 2006, 1762, 213–222. [36] Maglott, D., Ostell, J., Pruitt, K. D., Tatusova, T., Nucleic Acids Res. 2007, 35, D26–D31. [37] Peri, S., Navarro, J. D., Amanchy, R., Kristiansen, T. Z., Jonnalagadda, C. K., Surendranath, V., Niranjan, V. et al., Genome Res. 2003, 13, 2363–2371. [38] Cotter, D., Guda, P., Fahy, E., Subramaniam, S., Nucleic Acids Res. 2004, 32, D463–D467. [39] Bairoch, A., Apweiler, R., Nucleic Acids Res. 1996, 24, 21–25. [40] Hartmann, C. M., Gehring, H., Christen, P., Eur. J. Biochem. 1993, 218, 905–910. [41] Sumegi, B., Srere, P. A., J. Biol. Chem. 1984, 259, 15040–15045. [42] Aggeler, R., Coons, J., Taylor, S. W., Ghosh, S. S., Garcia, J. J., Capaldi, R. A., Marusich, M. F. J. Biol. Chem. 2002, 277, 33906–33912. [43] Murray, J., Zhang, B., Taylor, S. W., Oglesbee, D., Fahy, E., Marusich, M. F., Ghosh, S. S. et al., J. Biol. Chem. 2003, 278, 13619–13622. Proteomics and 2-DE 1341 [46] Schagger, H., Pfeiffer, K., EMBO J. 2000, 19, 1777–1783. [47] Boekema, E. J., Braun, H. P., J. Biol. Chem. 2007, 282, 1–4. [48] Lenaz, G., Genova, M. L., Am. J. Physiol. Cell Physiol. 2007, 292, C1221–C1239. [49] Ko, Y. H., Delannoy, M., Hullihen, J., Chiu, W., Pedersen, P. L., J. Biol. Chem. 2003, 278, 12305–12309. [50] Chen, C., Ko, Y., Delannoy, M., Ludtke, S. J., Chiu, W., Pedersen, P. L., J. Biol. Chem. 2004, 279, 31761–31768. [51] Vyssokikh, M. Y., Brdiczka, D., Acta Biochim. Pol. 2003, 50, 389–404. [52] Nijtmans, L. G., de Jong, L., Artal Sanz, M., Coates, P. J., Berden, J. A., Back, J. W., Muijsers, A. O. et al. EMBO J. 2000, 19, 2444–2451. [53] Bourges, I., Ramus, C., Mousson de Camaret, B., Beugnot, R., Remacle, C., Cardol, P., Hofhaus, G. et al., Biochem. J. 2004, 383, 491–499. [54] Nijtmans, L. G., Artal, S. M., Grivell, L. A., Coates, P. J., Cell. Mol. Life Sci. 2002, 59, 143–155. [55] Eaton, S., Middleton, B., Sherratt, H. S., Pourfarzam, M., Quant, P. A., Bartlett, K., Adv. Exp. Med. Biol. 1999, 466, 145–154. [56] Parker, A., Engel, P. C., Biochem. J. 2000, 345, 429–435. [57] Porpaczy, Z., Sumegi, B., Alkonyi, I., Biochim. Biophys. Acta 1983, 749, 172–179. [58] Robinson, J. B., Jr., Srere, P. A., J. Biol. Chem. 1985, 260, 10800–10805. [59] Taylor, S. W., Fahy, E., Zhang, B., Glenn, G. M., Warnock, D. E., Wiley, S., Murphy, A. N. et al., Nat. Biotechnol. 2003, 21, 281–286. [60] Reifschneider, N. H., Goto, S., Nakamoto, H., Takahashi, R., Sugawa, M., Dencher, N. A., Krause, F., J. Proteome Res. 2006, 5, 1117–1132. [61] Heddi, A., Stepien, G., Benke, P. J., Wallace, D. C., J. Biol. Chem. 1999, 274, 22968–22976. [62] Lenaz, G., Baracca, A., Carelli, V., D’Aurelio, M., Sgarbi, G., Solaini, G., Biochim. Biophys. Acta 2004, 1658, 89–94. [63] James, A. M., Wei, Y. H., Pang, C. Y., Murphy, M. P., Biochem. J. 1996, 318, 401–407. [64] Pallotti, F., Baracca, A., Hernandez-Rosa, E., Walker, W. F., Solaini, G., Lenaz, G., Melzi D’Er.il, G. V. et al., Biochem. J. 2004, 384, 287–293. [65] Oguro, H., Iijima, K., Takahashi, K., Nagai, A., Bokura, H., Yamaguchi, S., Kobayashi, S., Intern. Med. 2004, 43, 427–431. [66] Sirover, M. A., J. Cell. Biochem. 2005, 95, 45–52. [67] Colell, A., Ricci, J. E., Tait, S., Milasta, S. , Maurer, U., Bouchier-Hayes, L., Fitzgerald, P. et al., Cell 2007, 129, 983–997. [68] Yoon, K. A., Nakamura, Y., Arakawa, H., J. Hum. Genet. 2004, 49, 134–140. [44] Schägger, H., de Coo, R., Bauer, M. F., Hofmann, S., Godinot, C., Brandt, U., J. Biol. Chem. 2004, 279, 36349–36353. [69] Dillon, E. L., Knabe, D. A., Wu, G., Am. J. Physiol. 1999, 276, G1079–G1086. [45] John, G. B., Shang, Y., Li, L., Renken, C., Mannella, C. A., Selker, J. M., Rangell, L. et al., Mol. Biol. Cell 2005, 16, 1543–1554. [70] Naini, A., Kaufmann, P., Shanske, S., Engelstad, K., De Vivo, D. C., Schon, E. A., J. Neurol. Sci. 2005, 229– 230, 187–193. & 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com