The Ratio of Oxidative Phosphorylation Complexes I–V in Bovine Heart Mitochondria and the Composition of Respiratory Chain Supercomplexes

Abstract

The ratios of the oxidative phosphorylation complexes NADH:ubiquinone reductase (complex I), succinate:ubiquinone reductase (complex II), ubiquinol:cytochrome c reductase (complex III), cytochrome c oxidase (complex IV), and F1F0-ATP synthase (complex V) from bovine heart mitochondria were determined by applying three novel and independent approaches that gave consistent results: 1) a spectrophotometric-enzymatic assay making use of differential solubilization of complexes II and III and parallel assays of spectra and catalytic activities in the samples before and after ultracentrifugation were used for the determination of the ratios of complexes II, III, and IV; 2) an electrophoretic-densitometric approach using two-dimensional electrophoresis (blue native-polyacrylamide gel electrophoresis and SDS-polyacrylamide gel electrophoresis) and Coomassie blue-staining indices of subunits of complexes was used for determining the ratios of complexes I, III, IV, and V; and 3) two electrophoretic-densitometric approaches that are independent of the use of staining indices were used for determining the ratio of complexes I and III. For complexes I, II, III, IV, and V in bovine heart mitochondria, a ratio 1.1 ± 0.2:1.3 ± 0.1:3:6.7 ± 0.8:3.5 ± 0.2 was determined. The ratios of the oxidative phosphorylation complexes NADH:ubiquinone reductase (complex I), succinate:ubiquinone reductase (complex II), ubiquinol:cytochrome c reductase (complex III), cytochrome c oxidase (complex IV), and F1F0-ATP synthase (complex V) from bovine heart mitochondria were determined by applying three novel and independent approaches that gave consistent results: 1) a spectrophotometric-enzymatic assay making use of differential solubilization of complexes II and III and parallel assays of spectra and catalytic activities in the samples before and after ultracentrifugation were used for the determination of the ratios of complexes II, III, and IV; 2) an electrophoretic-densitometric approach using two-dimensional electrophoresis (blue native-polyacrylamide gel electrophoresis and SDS-polyacrylamide gel electrophoresis) and Coomassie blue-staining indices of subunits of complexes was used for determining the ratios of complexes I, III, IV, and V; and 3) two electrophoretic-densitometric approaches that are independent of the use of staining indices were used for determining the ratio of complexes I and III. For complexes I, II, III, IV, and V in bovine heart mitochondria, a ratio 1.1 ± 0.2:1.3 ± 0.1:3:6.7 ± 0.8:3.5 ± 0.2 was determined. oxidative phosphorylation blue-native polyacrylamide gel electrophoresis dodecylmaltoside NADH:ubiquinone reductase succinate:ubiquinone reductase ubiquinol:cytochrome c reductase cytochromec oxidase F1F0-ATP synthase The major components of the mammalian system of oxidative phosphorylation (OXPHOS)1 are the four complexes of the respiratory chain, NADH:ubiquinone reductase (complex I), succinate:ubiquinone reductase (complex II), ubiquinol:cytochrome c reductase (complex III), cytochromec oxidase (complex IV), and F1F0-ATP synthase (complex V). These complexes are functionally active when isolated as individual complexes. However, there is increasing evidence that complex V exists as a dimer in mitochondrial membranes and that the respiratory chain is a network of respiratory chain supercomplexes (1Arnold I. Pfeiffer K. Neupert W. Stuart R.A. Schägger H. EMBO J. 1998; 17: 7170-7178Crossref PubMed Scopus (371) Google Scholar, 2Schägger H. Pfeiffer K. EMBO J. 2000; 19: 1777-1783Crossref PubMed Scopus (1070) Google Scholar, 3Cruciat C.M. Brunner S. Baumann F. Neupert W. Stuart R.A. J. Biol. Chem. 2000; 275: 18093-18098Abstract Full Text Full Text PDF PubMed Scopus (219) Google Scholar). The respiratory chain supercomplexes from bovine heart mitochondria, which were isolated with high yield using blue native-polyacrylamide gel electrophoresis (BN-PAGE), contained complex I in monomeric form, complex III in dimeric form, and a variable copy number of complex IV. Associations of complexes II and V with other complexes of the system of oxidative phosphorylation could not be identified under the conditions of BN-PAGE so far. On the basis of a commonly used ratio for oxidative phosphorylation complexes I:II:III:IV:V of 1:2:3:6–7:3–5, which was published by Hatefi (4Hatefi Y. Annu. Rev. Biochem. 1985; 54: 1015-1069Crossref PubMed Google Scholar) in 1985, and a recent finding that complex I is almost quantitatively associated with complex III (2Schägger H. Pfeiffer K. EMBO J. 2000; 19: 1777-1783Crossref PubMed Scopus (1070) Google Scholar), a model for a network of mammalian respiratory chain complexes has been presented. According to this model, the respiratory chain exists as a mixture of two supercomplexes in a 2:1 ratio: a large supercomplex comprises complexes I, III, and IV; and a smaller supercomplex comprises complexes III and IV as well but is lacking complex I. However, the recovery of complex I after BN- and SDS-PAGE in a second dimension has not been tested in the previous work (2Schägger H. Pfeiffer K. EMBO J. 2000; 19: 1777-1783Crossref PubMed Scopus (1070) Google Scholar), and the ratio of OXPHOS complexes is not exactly known. An often-cited review of the ratio of OXPHOS complexes by Capaldiet al. (5Capaldi R.A. Gonzalez D.H. Zhang Y.Z. Yanamura W. J. Bioenerg. Biomembr. 1988; 20: 291-311Crossref PubMed Scopus (52) Google Scholar) gives concentration ranges of complexes per milligram of protein. Markedly differing complex I:II:III:IV:V ratios can be calculated from these data depending on the technique used for complex I determinations. A ratio of 1:1.5:1.9–4.1:4.6–7.7:4–4.2 is calculated when complex I is quantified using FMN contents. These compositional data were rounded off to a 1:2:3:6:x ratio. However, a 1:3:3.8–8.2:9.2–15.2:8–8.4 ratio is calculated on the basis of original data sets using immunologic techniques to quantify complex I (5Capaldi R.A. Gonzalez D.H. Zhang Y.Z. Yanamura W. J. Bioenerg. Biomembr. 1988; 20: 291-311Crossref PubMed Scopus (52) Google Scholar, 6Smith S. Cottingham I.R. Ragan C.I. FEBS Lett. 1980; 110: 279-282Crossref PubMed Scopus (16) Google Scholar). Different materials were used in the cited studies, such as bovine and rat mitochondria, Keilin-Hartree particles, and sonic particles, which may be a major reason for the large variations in the contents and ratios of the complexes. The commonly used ratio of OXPHOS complexes in bovine sonic particles, which was published in a review article by Hatefi (4Hatefi Y. Annu. Rev. Biochem. 1985; 54: 1015-1069Crossref PubMed Google Scholar), was based on the content of FMN (complex I), covalently bound FAD (complex II), cytochrome c1 (complex III), cytochromeaa3 (complex IV), and F1 (complex V) and adjusted to the nearest integer relative to complex I. Unfortunately, no S.D.s or ranges are given, except for complexes IV and V. The use of the FMN content in sonic particles as a single measure to quantify complex I seems critical. Complex I might have been overestimated if the sample contained FMN from another flavoprotein, and all ratios would need correction by a factor of 2 if complex I contained a second FMN, as postulated by Albracht and Hedderich (7Albracht S.P.J. Hedderich R. FEBS Lett. 2000; 24275: 1-6Crossref Scopus (78) Google Scholar). Vinogradov and King (8Vinogradov A.D. King T.E. Methods Enzymol. 1979; 55: 118-127Crossref PubMed Scopus (41) Google Scholar) used Keilin-Hartree preparations to determine cytochromes, acid-extractable FMN and FAD, and acid-nonextractable flavin. Assuming that all acid-extractable FMN is bound to complex I (one FMN per complex I), and all acid-nonextractable flavin is part of complex II (one FAD per complex II), a ratio of complexes I:II:III:IV of 1:1.3:<2.7:5.6 can be calculated from their data. The value for complex III is uncertain, because no correction for interfering cytochrome b of complex II was made. Assuming that, in contrast to sonic particles, 20% of total FMN of Keilin-Hartree particles is not part of complex I gives a complex I:II:III:IV ratio of 1:1.6:<3.4:7, which is close to the commonly used ratio. Again the critical FMN content is the only measure to quantify complex I, and novel approaches not relying on the assumptions concerning FMN and FAD contents seem to be required to obtain reliable ratios of OXPHOS complexes. Knowing the accurate ratios of respiratory chain complexes is essential for the correct interpretation of kinetic and spectroscopic data obtained with native mitochondrial membranes, because the type and number of supercomplexes and the fraction of free individual complexes will define the overall throughput rates of the respiratory chain. Knowing the accurate ratios and the structural organization of the building blocks of the respiratory chain is also essential to evaluate the importance of substrate channeling versus pool function of quinone and cytochrome c (see “Discussion”). Therefore, we reassessed the stoichiometry of the respiratory chain complexes, in particular the complex I:III ratio, by using three novel and independent approaches. Imidazole was obtained from Fluka; dodecylmaltoside was from Biomol; and decylquinazolinamine was a kind gift from Aventis Crop Science (Frankfurt, Germany). Inhibitors and all other chemicals were purchased from Sigma. Bovine heart mitochondria were prepared according to the method of Smith (9Smith A.L. Methods Enzymol. 1967; 10: 81-86Crossref Scopus (478) Google Scholar). Complex I, which contains one FMN per monomer, was prepared as described by Okun et al. (10Okun J.G. Lümmen P. Brandt U. J. Biol. Chem. 1999; 274: 2625-2630Abstract Full Text Full Text PDF PubMed Scopus (301) Google Scholar). The FMN content was 0.99 nmol of FMN/mg of protein. Complex III, which contains two heme b per monomer or one cytochrome c1 per monomer, was isolated according to the method of Schäggeret al. (11Schägger H. Link Th.A. Engel D. von Jagow G. Methods Enzymol. 1986; 126: 224-237Crossref PubMed Scopus (229) Google Scholar). The cytochrome b content of the preparation used was 3.5 nmol/mg. Complex IV was prepared according to the method of Brandt et al. (12Brandt U. Schägger H. von Jagow G. Eur. J. Biochem. 1989; 182: 705-711Crossref PubMed Scopus (58) Google Scholar). The preparation used had a cytochrome a content of 9.8 nmol/mg of protein. Complex V was prepared as described elsewhere. 2H. Schägger, unpublished procedures. Difference spectra of dithionite-reduced minus ferricyanide-oxidized samples were recorded in the range of 650–500 nm using a Shimadzu UV 300 spectrophotometer. The following absorption coefficients were used: cytochrome b of complex II, ε561–575 nm = 26 mm−1cm−1; cytochromesbL and bH of complex III, ε562–575 nm = 2 × 28.5 mm−1cm−1; and cytochromea + a3 of complex IV, ε603–630 nm = 24 mm−1cm−1. NADH:ubiquinol reductase activity of complex I was measured at pH 8, 30 °C, as decylquinazolinamine-sensitive oxidation of NADH (200 μm,340–400 nm; ε = 3.4 mm−1cm−1) using 75 μm decylbenzoquinone as electron acceptor. Succinate:dichloroindophenol reductase activity of complex II was measured at pH 7, 30 °C, as thenoyltrifluoroacetone-sensitive reduction of dichloroindophenol (0.002%, 610–750 nm; ε = 20.5 mm−1cm−1) using 50 mmsuccinate as a substrate. Ubiquinol:cytochrome c reductase activity of complex III and cytochrome c oxidase activity of complex IV were measured at 25 °C, pH 7.2, by the antimycin-sensitive reduction and cyanide-sensitive oxidation of cytochrome c, respectively (70 μm cytochromec, 550–540 nm; ε = 19.1 mm−1cm−1). Decylbenzoquinol (75 μm) was used as substrate for cytochrome creductase. Oligomycin-sensitive ATP hydrolysis of complex V was measured at 25 °C using an assay coupled to the oxidation of NADH. Shortly before the test, 0.25 mm NADH, 1 mmphosphoenolpyruvate, 2.5 units/ml lactate dehydrogenase, and 2 units/ml pyruvate kinase were added to the test buffer (250 mmsucrose, 50 mm KCl, 5 mm MgCl2, 2 mm NaCN, and 20 mm Tris/HCl, pH 7.5). Tricine-SDS-PAGE, BN-PAGE, and two-dimensional PAGE (BN-PAGE/SDS-PAGE; Ref. 13Schägger H. Cramer W.A. von Jagow G. Anal. Biochem. 1994; 217: 220-230Crossref PubMed Scopus (1051) Google Scholar), three-dimensional PAGE (BN-PAGE/modified BN-PAGE/SDS-PAGE; ref. 2Schägger H. Pfeiffer K. EMBO J. 2000; 19: 1777-1783Crossref PubMed Scopus (1070) Google Scholar), Coomassie blue staining, and densitometric protein quantification (14Schägger H. Electrophoresis. 1995; 16: 763-770Crossref PubMed Scopus (74) Google Scholar) were performed as described. The spectrophotometric-enzymatic assay described here for determining the ratio of complexes II, III, and IV use of a differential solubilization of complexes II and III and parallel assays of spectra and catalytic activities in the samples before and after Bovine heart mitochondria were in mm and 20 mm pH 7, and by ratios from to protein of sample was as a The second was for 30 at × Succinate:dichloroindophenol reductase activity of complex II was quantitatively in all but c reductase activity of complex III was only using a ratio of 1 and using of and of the total cytochrome b absorption of bovine heart mitochondria were in the A and Knowing the of complexes II and III in the and the coefficients for the of isolated complexes II and III, the of complex II to the total heme b absorption in bovine heart mitochondria was calculated as and from these two ± from and the ratio of complex was determined as ± Complex III and IV were calculated from minus spectra of the total heme b absorption to complex II, a complex ratio of ± 0.2 = was The overall ratio of complexes determined by this approach was ± ± to complex III = These ratios to monomeric complexes independent of their A electrophoretic-densitometric approach to measure the of OXPHOS complexes use of a solubilization of all OXPHOS complexes from bovine heart mitochondria by by of the complexes by BN-PAGE, of the subunits by SDS-PAGE in a second and densitometric quantification of subunits after Coomassie blue solubilization and recovery of the complexes after BN-PAGE was a for this and the staining indices of individual subunits of complexes had to be determined. Bovine heart mitochondria were under conditions used for BN-PAGE but on a mitochondria of were in 2 of 50 mm and 50 mm pH 7, and by of dodecylmaltoside ratio, was as a The second was 30 at × minus spectra and catalytic activities were measured using the sample before and the after The recovery of all complexes was and Bovine heart mitochondria were for BN-PAGE as described using a ratio of An of the was used for BN-PAGE (70 of total protein to a sample BN-PAGE, the two of complexes I and V were as in and on of an gel of the after ultracentrifugation was with and the gel as a and Coomassie blue staining, the staining of the of isolated complex I and of the of isolated complex V were with the of the The from the gel were ± for complex I and ± for complex V = The was identified as the of a not of this by of seem to be because this would the recovery from and the measured recovery would not be The but not the was used to quantify the recovery of complex because the recovery using the was of complexes III and IV have been by using a the sample was to BN-PAGE, complexes III and IV were quantified BN-PAGE, the two were and by SDS-PAGE using of complexes III and IV as in the The yield of complex III was with complex III and with complex IV, which gives to BN-PAGE (14Schägger H. Electrophoresis. 1995; 16: 763-770Crossref PubMed Scopus (74) Google Scholar). of isolated complexes in the were to and the subunits were quantified after Coomassie blue staining using a of an gel used for determination of staining indices is in The staining of the subunits in I were to 1 and 1 and to the I and II subunits of complex III, which were used as in Coomassie blue-staining indices of subunits of OXPHOS = ± + ± + ± protein ± protein ± ± ± ± ± ± ± ± ± blue-staining of individual subunits were to 1 and 1 of the the staining indices were to the staining of I and II of complex III, which were used as in in a Coomassie blue-staining of individual subunits were to 1 and 1 of the the staining indices were to the staining of I and II of complex III, which were used as in of the two-dimensional used for determination of the ratios of complexes in bovine heart mitochondria is in A complex ratio to complex III = 3) of ± ± ± = was determined this the recovery of complexes IV and V after BN-PAGE, which is with complexes I and III, the complex ratio is approach using staining indices for individual subunits was also to bovine heart The ratio of complexes was to be almost to the ratio determined for mitochondria not A second electrophoretic-densitometric approach with the solubilization described but was used of dodecylmaltoside ratio, to respiratory chain The of complexes in the after using assays were ± for complex I, for complex II, ± for complex III, ± for complex IV, and ± for complex which well with the solubilization by BN-PAGE was used to the supercomplexes and individual OXPHOS complexes, and a SDS-PAGE to the subunits of the Coomassie subunits were for densitometric the + and + of complex I and the 1 and 2 of complex III. The of these was by protein The total staining of of these subunits in supercomplexes and free complexes was and with the in a two-dimensional gel from a sample The of complexes with of complexes were ± for complex I, ± for complex II, ± for complex III, for free complex IV (complex IV bound to supercomplexes could not be and ± for complex V = also that the recovery of complexes was and was to determine the ratio of complexes I and III in a that was independent of the staining indices of individual two-dimensional of mitochondrial complexes, we that ± of total complex I and ± of total complex III = were bound to ± and ± as free individual complexes I and III. supercomplexes a complex I monomer, which to a complex III and a variable complex IV copy as in that of total complex I and of total complex III are in a complex I monomer would of total complex III and in free Therefore, the complex I:III ratio determined this is A electrophoretic-densitometric approach to determine the complex I:III ratio with the solubilization using and dodecylmaltoside as described but after BN-PAGE, a second BN-PAGE using in the The this approach was to of the because the protein I of complex III and the + of complex I were close in the two-dimensional gel two-dimensional native which are native in (BN-PAGE/modified all complexes that as individual complexes in BN-PAGE are on a in the is in in which dodecylmaltoside was used for solubilization and of isolated complexes. However, solubilization by to supercomplexes in the BN-PAGE, which the individual complexes in the BN-PAGE and were this of individual complexes, complexes I and III, which were and on of and complexes I and III from samples were to the and were used as for this was to use of a complex for densitometric quantification except that were and such as the protein of complex III. were subunits that were not from the + of complex I, which was close to the of complex V mitochondria and the three-dimensional we measured that ± of total complex I and ± of total complex III = were in ± and ± were in free complex I monomer would of total complex III and in free The complex I:III ratio determined this was ratios determined in this work are in of published ratios of OXPHOS complexes in bovine heart (4Hatefi Y. Annu. Rev. Biochem. 1985; 54: 1015-1069Crossref PubMed Google work ratios were to complex III = in a The ratios were to complex III = The of this work was to determine the accurate complex I:III ratio in bovine heart mitochondria by novel and independent approaches to obtain a basis for a recent structural model of the mammalian (2Schägger H. Pfeiffer K. EMBO J. 2000; 19: 1777-1783Crossref PubMed Scopus (1070) Google Scholar) and A ratio of complex which is has been determined by using after of FMN and heme from sonic particles (4Hatefi Y. Annu. Rev. Biochem. 1985; 54: 1015-1069Crossref PubMed Google Scholar). However, recent data that complex I might have been overestimated because of FMN, because a ratio of complex to a ratio of complexes I:III was determined by techniques (10Okun J.G. Lümmen P. Brandt U. J. Biol. Chem. 1999; 274: 2625-2630Abstract Full Text Full Text PDF PubMed Scopus (301) Google Scholar). assays were based on the that complex I one FMN as isolated complex I (10Okun J.G. Lümmen P. Brandt U. J. Biol. Chem. 1999; 274: 2625-2630Abstract Full Text Full Text PDF PubMed Scopus (301) Google Scholar). contrast to this a second FMN, which can be from complex I was postulated to spectra of complex I (7Albracht S.P.J. Hedderich R. FEBS Lett. 2000; 24275: 1-6Crossref Scopus (78) Google Scholar). could Smith et al. S. Cottingham I.R. Ragan C.I. FEBS Lett. 1980; 110: 279-282Crossref PubMed Scopus (16) Google Scholar) determined an FMN content of nmol/mg of bovine sonic particles, assays a complex I content of the of two of FMN per complex I were almost all published data on mitochondrial complex I contents would need correction by a factor of used three novel electrophoretic-densitometric approaches to determine the complex I:III ratio. complex I one FMN was used for determining the Coomassie blue-staining ratios of subunits according to method and were independent of the FMN content of complex I. The an complex I:III ratio of 1.1 ± is close to the ratio published by Hatefi (4Hatefi Y. Annu. Rev. Biochem. 1985; 54: 1015-1069Crossref PubMed Google Scholar), but is a by using independent the obtained in this work is close to the ratio after FMN from sonic particles, we that there is no evidence for a second FMN of complex I, as by Albracht and Hedderich (7Albracht S.P.J. Hedderich R. FEBS Lett. 2000; 24275: 1-6Crossref Scopus (78) Google Scholar) and Smith et S. Cottingham I.R. Ragan C.I. FEBS Lett. 1980; 110: 279-282Crossref PubMed Scopus (16) Google Scholar). FMN seems not to be a with the of sonic that complex I contains one FMN, isolated complex I, and that FMN is not bound in to a mitochondrial protein other complex I. Knowing the accurate stoichiometry of oxidative phosphorylation complexes is to structural of the mammalian with Complex are and only of total complex I was in free in the of seems that all complex I is bound to complex III in the of but of total complex III is not bound to complex I at the measured 1.1 ± ratio of complexes ratio is with a model of the of large and supercomplexes in a 2:1 ratio 2Schägger H. Pfeiffer K. EMBO J. 2000; 19: 1777-1783Crossref PubMed Scopus (1070) Google and of complex I to complex III the for quinone substrate which would a independent of the of a substrate A quinone pool function has been in mammalian mitochondria Eur. J. Biochem. PubMed Scopus Google Scholar), but the of the quinone pool function for the electron complexes I and III would Assuming that of complex I to complex III not the to the quinone complex II and other such and would have to all complex III, also the fraction of complex III that is the large However, if of complex I to complex III from the quinone complex II and other would only have to the smaller NADH and of complexes III and IV. this as a we have on the importance of quinone pool function versus substrate channeling for NADH and oxidation A second major of this work was to obtain accurate data for complex Hatefi (4Hatefi Y. Annu. Rev. Biochem. 1985; 54: 1015-1069Crossref PubMed Google Scholar) measured acid-extractable FMN and acid-nonextractable FAD to determine the complex ratio and a complex ratio of for bovine sonic used an and a spectrophotometric-enzymatic The ratio of complexes is to determine by because the two complexes with However, we use of the solubilization of with bovine mitochondrial can complexes II and V using because complex are under these and the supercomplexes seem to be to an network of supercomplexes (2Schägger H. Pfeiffer K. EMBO J. 2000; 19: 1777-1783Crossref PubMed Scopus (1070) Google Scholar). solubilization of complex II, with and and the for the determination of the complex ratio ± The is not with the commonly used ratio by Hatefi (4Hatefi Y. Annu. Rev. Biochem. 1985; 54: 1015-1069Crossref PubMed Google Scholar). Complex II is to complex III, but the ratio of monomeric complex II to dimeric complex III is close to that there might a of the two complexes, such an could not be by The ratio of FAD and cytochrome c1 in isolated cytochrome c reductase is is two of complex II are bound to one complex III dimer Methods Enzymol. 1967; 10: Scopus Google Scholar). determined a complex ratio of using a approach and an independent electrophoretic-densitometric previous (4Hatefi Y. Annu. Rev. Biochem. 1985; 54: 1015-1069Crossref PubMed Google Scholar, R. Methods Enzymol. 1979; 55: Scholar). et al. PubMed Scopus Google Scholar) the of to complexes III and V in rat heart and The of to heart complex III was almost to the of and a ratio of complexes III and which the ratio that we using an independent electrophoretic-densitometric have also been obtained using techniques for determining the F1 content in sonic particles F. G. Biochem. 1986; PubMed Scopus Google Scholar, Hatefi Y. 1988; PubMed Scopus Google Scholar). were in the range from which is to a complex ratio of and Brandt for the kind gift of bovine complex I.