Respiratory chain supercomplexes set the threshold for ......Upon cell fusion, respiration is recovered in hybrids cells, indicat-ing that mitochondria fuse and exchange genetic and

Respiratory chain supercomplexes set thethreshold for respiration defects in humanmtDNA mutant cybrids

Marilena D’Aurelio1, Carl D. Gajewski1, Giorgio Lenaz2 and Giovanni Manfredi1,*

1Department of Neurology and Neuroscience, Weill Medical College of Cornell University, New York, NY 10021, USA

and 2Dipartimento di Biochimica, G. Moruzzi, Universita di Bologna, Bologna 40126, Italy

Received April 26, 2006; Revised and Accepted May 24, 2006

Mitochondrial DNA (mtDNA) mutations cause heterogeneous disorders in humans. MtDNA exists in multiplecopies per cell, and mutations need to accumulate beyond a critical threshold to cause disease, becausecoexisting wild-type mtDNA can complement the genetic defect. A better understanding of the moleculardeterminants of functional complementation among mtDNA molecules could help us shedding some lighton the mechanisms modulating the phenotypic expression of mtDNA mutations in mitochondrial diseases.We studied mtDNA complementation in human cells by fusing two cell lines, one containing a homoplasmicmutation in a subunit of respiratory chain complex IV, COX I, and the other a distinct homoplasmic mutationin a subunit of complex III, cytochrome b. Upon cell fusion, respiration is recovered in hybrids cells, indicat-ing that mitochondria fuse and exchange genetic and protein materials. Mitochondrial functional comple-mentation occurs frequently, but with variable efficiency. We have investigated by native gelelectrophoresis the molecular organization of the mitochondrial respiratory chain in complementinghybrid cells. We show that the recovery of mitochondrial respiration correlates with the presence of supra-molecular structures (supercomplexes) containing complexes I, III and IV. We suggest that critical amountsof complexes III or IV are required in order for supercomplexes to form and provide mitochondrial functionalcomplementation. From these findings, supercomplex assembly emerges as a necessary step for respiration,and its defect sets the threshold for respiratory impairment in mtDNA mutant cells.

INTRODUCTION

The respiratory chain is the major source of the transmembraneelectrochemical gradient utilized in mitochondria to generateATP. In mammals, mitochondria contain multiple copiesof a circular DNA molecule (mtDNA), which encodes for13 protein components of the respiratory chain and a full comp-lement of tRNAs and rRNAs necessary for mitochondrialprotein translation (1).

The structural models explaining the supramolecular organ-ization of the respiratory chain have changed over the pastdecades. The initial one was the ‘solid-state’ model proposedby Chance and Williams (2). This model has later beenreplaced by the ‘liquid-state’ one, where functionally activeindividual complexes are free to diffuse laterally in the lipidbilayer of the inner membrane (3). In the ‘random collision’

model (4,5), mitochondrial electron transfer depends on therandom contacts between independent complexes and smalldiffusible molecules, such as coenzyme Q and cytochrome c.Only recently, the development of blue native gel electrophor-esis has allowed for the identification of supramolecular struc-tures containing multimers of respiratory chain complexes, inyeast and mammalian mitochondria, which are reminiscent ofthe original solid-state model (6–8). Furthermore, the archi-tecture of mammalian respiratory chain supercomplexes hasbeen determined by a combined electrophoretic and single-particle image-analysis approach (9). Flux control ratiostudies on bovine mitochondria demonstrated that the respirat-ory chain is functionally organized in partial supercomplexes(10). To date, however, it remains unclear to what extent theoxidative function of the respiratory chain is dependent uponthe formation of supercomplexes.

# The Author 2006. Published by Oxford University Press. All rights reserved.For Permissions, please email: [email protected]

*To whom correspondence should be addressed at: Department of Neurology and Neuroscience, Weill Medical College of Cornell University,525 E. 68th Street, A-505, New York, NY 10021, USA. Tel: +1 2127464605; Fax: +1 212 7468276; Email: [email protected]

Human Molecular Genetics, 2006, Vol. 15, No. 13 2157–2169doi:10.1093/hmg/ddl141Advance Access published on June 1, 2006

Dow

nloaded from https://academ

ic.oup.com/hm

g/article-abstract/15/13/2157/678241 by guest on 15 Novem

ber 2018

In humans, mutations in mitochondrial genes cause meta-bolic disorders biochemically characterized by respiratorychain defects (11). Pathogenic mtDNA mutations oftencoexist with wild-type (WT) mtDNA within the same cells,a condition called mtDNA heteroplasmy. In this case, thedegree of respiratory chain impairment depends on the relativeproportions of mutant and WT mtDNA, with a clear thresholdeffect, which varies among cell types, based on the metabolicrequirements for ATP. WT mtDNA can complement the lossof function of mutant mtDNA up to a certain threshold. Tocomplement the deleterious effects of the mutations, WTmtDNA, mtRNA and mtDNA-encoded proteins must be ableto diffuse within mitochondria. A more complex situationoccurs when different mtDNA species are segregated withindifferent mitochondria, because in this case, complementationrequires exchange of factors across the boundaries of mito-chondrial membranes. In mammalian cells, mitochondria areorganized in a dynamic network, which is subject to a constantreshaping process modulated by fusion and fission events(reviewed in 12). This reshaping process is likely to providethe mechanism whereby molecules are exchanged amongmitochondria, allowing for functional complementation.

Cell hybridization models have been used to confirm thehypothesis that mitochondria can fuse to provide complemen-tation. When cells containing homoplasmic levels of mutantmtDNA (i.e. 100% mutant mtDNA) and severely compro-mised respiratory function are fused with cells containingWT mtDNA, respiration is restored in the resulting hybridcells (13). Furthermore, the fusion of two cell lines, each con-taining a distinct homoplasmic mt-tRNA mutation, resulted inrestored respiration, suggesting that tRNAs were exchangedamong mitochondria (14).

To date, functional complementation between mitochondria-containing mutations in polypeptide–coding genes has notbeen explored. To address this issue, we have investigated com-plementation between two human cell lines, each containing ahomoplasmic mtDNA mutation in a gene encoding a subunitof distinct respiratory chain complexes. Both mutationsresult in a complete loss of their respective complexes. Wehave characterized the biochemical properties of protein–protein complementation in hybrids derived from fusion ofthese mutant cell lines. We have studied the direct effects ofprotein mutations on the assembly of individual respiratorychain complexes and on supermolecular complex formation,as well as the secondary effects on the assembly and functionof complexes that are not directly affected by the mutations.

RESULTS

Fusion of cybrids containing two distinct mutantmtDNA results in hybrids with restored respiratorychain functions

Hybrid clones were obtained by fusion of two human osteosar-coma cybrid lines, one containing a homoplasmic stop-codonmutation in the cytochrome c oxidase subunit 1 gene(MTCO1), denoted as COX1-MT, and the other a frame-shiftmutation in the gene encoding cytochrome b (MTCyB),denoted as CyB-MT. Both parental cybrids completelylacked mitochondrial respiration (15,16). Each cybrid line

was engineered with a specific antibiotic resistance (i.e. hygro-mycin and puromycin, respectively), allowing for the double-selection of hybrid clones (nuclear complementation). Uponfusion, hybrids contained a mixture of variable proportionsof the two mutant mtDNAs. In addition, because mtDNArecombination occurs between the two mtDNA alleles, newmolecules of WT mtDNA are formed (17). Therefore, inmost clones, the sum of the two mutant alleles does notaccount for 100% of the mtDNA molecules (Figs 4–6).

To assess whether selective growth pressure has a signifi-cant impact on the frequency of hybrid clones, cybrid fusionwas performed under different growth conditions. The clonefrequency in non-selective medium, containing glucose, pyru-vate and uridine, reflects the efficiency of nuclear complemen-tation. The clone frequency in selective media lackingpyruvate and uridine (18), or containing galactose instead ofglucose as the major carbon source (19), reflects the frequencyof both nuclear and mitochondrial complementation, becausethese selective conditions only allow for survival of cloneswith restored mitochondrial respiratory chain activity. Thereis a small but statistically significant difference in the numberof hybrid clones obtained in non-selective and mitochondrialselective conditions (Fig. 1A). This indicates that the majorityof the hybrids that survived the nuclear selection can alsosurvive under mitochondrial selection and suggests thatcybrid fusion results both in nuclear and mitochondrial com-plementation. In all of the following experiments, we havestudied hybrid clones that are able to survive in selectivemedium. Because in the literature there are a few examples ofnuclear suppression of mtDNA defects (20,21), as a control,COX1-MT cybrids were also fused with cells devoid ofmtDNA (r0 cells) and resistant to neomycin. As expected,the resulting COX1-MT-r0 hybrids obtained by hygromycinand neomycin selection did not yield any viable clone undermitochondrial selective conditions (not shown), suggestingthat the hybridization procedure per se does not afford survivalunder mitochondrial selection.

Next, we investigated the growth of hybrid clones in galac-tose medium, as an indirect estimate of oxidative phosphoryl-ation function. On average, the number of hybrid cells countedafter 3 days of growth in galactose is ~60% lower than WTcells (Fig. 1B). The wide variability suggests that individualclones have varying levels of respiratory chain recovery. Asexpected, both parental cybrids fail to grow in galactosemedium. Oxygen consumption measurements confirm thatmitochondrial respiration is restored in hybrid cells.However, on average, coupled respiration in intact cells islower in hybrid cells than in WT controls, with a wide vari-ation (Fig. 1C). Cell respiration uncoupled with FCCP,which generates the maximal respiratory chain activity inintact cells, follows a similar trend (Fig. 1D). As expected,parental cybrids have no respiratory activity, either incoupled or in uncoupled conditions (Fig. 1C and D).

Correlation between mtDNA mutation levels, specificenzyme activities and respiration in hybrid cells

In Figure 2A, the relative abundance of each of the two mutantmtDNA species is correlated with cell respiration. Because theproportions of each heteroplasmic mutant mtDNA species

2158 Human Molecular Genetics, 2006, Vol. 15, No. 13

Dow


ic.oup.com/hm


ber 2018

varied in the same clone over time, the levels of bothmutations have been measured after each experiment. Fromthe tri-dimensional plot, it is apparent that a number ofclones have respiration similar to WT cells. These clonescontain variable proportions of the two mutant mtDNAs, butin each clone, these proportions were below 40 and 90% forCOX1-MT and CyB-MT, respectively. All the clones withproportions of mutant mtDNA above these levels had defec-tive respiration. The 40% threshold for COX1-MT was inagreement with the one that we had previously reported incybrids containing only COX1-MT mtDNA (16). Becausethe threshold for CyB-MT had not been established previously,we have looked at the correlation between the proportion ofCyB-WT and cell respiration. The plot in Figure 2B showshybrid clones with high proportions of COX1-WT, .60%(i.e. above the threshold for respiratory defect), and variableproportions of CyB-WT. The data points were interpolatedwith the equation Y ¼ Bmax � X/(Kd þ X ), where X is the per-centage of CyB-WT, Y the value of oxygen consumption, Bmax

the horizontal asymptote of the curve and Kd the percentage

of CyB-WT, which allows for 1/2 Bmax. On the basis of thisinterpolation, 5% of CyB-WT allows for 50% of the maximalrespiration, suggesting that the respiratory threshold forCyB-MT is very high, close to 90%. The hybrid clones withdefective respiration coincided with those with defectivegrowth in galactose (Fig. 1B). It is likely that hybrids withvery high proportions of mutant mtDNA, close to 100%, areunable to complement and do not survive the metabolic selec-tion, which may explain why a decrease in clone numbers isobserved under these conditions (Fig. 1A).

To study the correlation between mutation loads and activi-ties of respiratory chain complexes, we measured complex IV-and complex III-specific activities using exogenous substrates.There is no threshold for these specific activities, because theyboth show an inverse linear correlation with the proportions oftheir respective mutations (Fig. 2C and D). This suggests thatboth mutations affect the respective complex activities bydecreasing the total amount of assembled active enzymes,without decreasing their intrinsic kinetic parameters. It alsoexcludes the possibility that CyB-MT affects complex IV

Figure 1. Mitochondrial functional complementation. (A) Average number of surviving hybrid clones in three different growth conditions: non-selective mediumA (black bar), selective medium B (light gray bar) and selective medium C (dark gray bar), as described in Materials and Methods. Error bars represent SD.�indicates statistically significant differences (P, 0.05, estimated by unpaired two-tailed Student’s t-test) compared with non-selective medium. (B) Growthin galactose medium (medium B). 1 � 105 cells (indicated by the dashed line) plated in triplicate for each cell line were counted after 3 days of selection inmedium B. The value for hybrid clones is the average number of cells for 12 individual hybrid clones each counted in triplicate. Measurement of oxygen con-sumption in intact cells in the absence (C) or in the presence (D) of the uncoupling agent FCCP. The value for hybrid clones is the average rate of respiration in14 individual clones. Error bars indicate SD. WT, WT cybrid; COX1, 100% COX1-MT; CyB, 100% CyB-MT. Statistically significant differences between theWT average value and the average value in hybrid clones estimated by unpaired two-tailed Student’s t-test are indicated: �P , 0.05; ��P, 0.005.

Human Molecular Genetics, 2006, Vol. 15, No. 13 2159

Dow


ic.oup.com/hm


ber 2018

activity and vice versa, because in this case, the correlationswould not be linear.

Respiratory chain complex assembly in human cybrid cells

To understand the effects of the mtDNA mutations on thecomposition of the respiratory chain, we have investigatedthe assembly of its complexes by blue native gel electro-phoresis. We first determined the appropriate conditions tosolubilize mitochondrial membranes while preserving intactrespiratory chain complexes. We tested two detergents at

various concentrations in WT cybrid cells. First, increasinglauryl maltoside from 0.01 to 0.4% results in an increase ofthe amounts of individual complexes detectable by westernblot (Fig. 3A, left lanes). At the same time, it results in adecrease of the bands migrating higher than complex I, pre-sumably corresponding to supercomplexes. In addition,increasing the concentration of lauryl maltoside causes afaster migration of complex IV, possibly resulting fromremoval of ancillary proteins or conformational changes.Lauryl maltoside concentrations higher than 1% result indisassembly of complex I (data not shown). Secondly, digito-

Figure 2. Mitochondrial threshold for respiration and specific enzymatic activities. (A) Rates of coupled respiration as a function of the proportions of CyB-MTand COX1-MT in 14 hybrid clones, parental cybrids 100% CyB-MT and 100% COX1-MT, and WT cybrids. Note that in the tridimensional graph, respiratoryrates are represented by the height of each individual pin. (B) Plot of the rate of coupled respiration as a function of the relative proportions of CyB-WT in 13hybrid clones and the parental 100% CyB-MT. The data are fitted by a single rectangular hyperbole, two parameters, with equation Y ¼ Bmax � X/(Kd þ X ); Y isthe rate of respiration, Bmax ¼ 17.08, the horizontal asymptote, the maximum of respiration, X the percentage of CyB-WT and Kd ¼ 5.01 the percentageCyB-WT which corresponds to 1/2 Bmax. Respiratory chain complex III (III), complex IV (IV) and the mitochondrial matrix citrate synthase (CS) activitiesmeasured spectrophotometrically using specific substrates and inhibitors on isolated mitochondria from 16 individual hybrids clones. Values are averages oftwo independent mitochondria preparations each measured in duplicate. (C) Plot of III/CS activities ratio as a function of the relative proportion ofCyB-MT. (D) Plot of IV/CS activities ratio as a function of the relative proportion of COX1-MT. In both (C) and (D), data are fitted by linear regressioncurves as shown by the relative equations and R2 values.


Dow


ic.oup.com/hm


ber 2018

Figure 3. Assembly and structural organization of the respiratory chain complexes. (A) Respiratory chain complexes and supercomplexes from WT cybridresolved by blue native gel electrophoresis followed by western blot and detected with a mixture of monoclonal antibodies: anti-39 KDa subunit forcomplex I, anti-70 KDa subunit for complex II, anti-core 2 for complex III and anti-COX1 for complex IV. Different percentages of lauryl maltoside and digi-tonin were used for mitochondrial solubilization. A mixture of high molecular weight purified proteins was used as a molecular mass marker: ferritin dimer,880 KDa; thyroglobulin, 669 KDa; ferritin monomer, 440 KDa; catalase 232 KDa; lactate dehydrogenase, 140 KDa; albumin, 66 KDa. (B) Mitochondrial com-plexes and supercomplexes from WT cybrids, solubilized with 0.33% digitonin, resolved in five individual lanes by blue native gel electrophoresis, blotted andseparately probed with antibodies: anti-39 KDa subunit for complex I, anti-core 2 for complex III, anti-COX 1 for complex IV, anti-70 KDa subunit for complexII and anti-b subunit for complex V. (C) Effect of decreasing digitonin concentrations on complex II in WT and parental mutant cybrids. Solubilization ofcomplex II, whose assembly is unaffected by the mtDNA mutations, is similar in WT and mutant cells. I, II, III, IV and V indicate the positions of complexesI–V. III2, dimeric complex III; I þ III2, complex I associated with dimeric complex III; III2 þ IV, dimeric complex III associated with monomeric complex IV;I þ III2 þ IV(1 – 4), supercomplex containing complex I, dimeric complex III and monomeric (1), dimeric (2), trimeric (3) and tetrameric (4) complex IV. ‘Insol.’indicates non-solubilized mitochondrial proteins.


Dow


ic.oup.com/hm


ber 2018

nin concentrations between 0.33 and 5.0% allow for the detec-tion of abundant high migrating bands, whereas the amountsof individual complexes are lower with lauryl maltoside. Nodifference is apparent among samples treated with variousconcentrations of digitonin (Fig. 3A). On the basis of theseresults, we chose to use 0.4% lauryl maltoside for the detec-tion of individual respiratory chain complexes and 0.3% digi-tonin for the detection of supercomplexes.

In lauryl maltoside-treated samples, the apparent molecularweight of the major band detected with anti-complex III core2antibody is �460 KDa, suggesting that complex III is in ahomodimeric form (III2). Anti-subunit I of complex IV(COX1) antibody detects a major band of �200 KDa, indicat-ing that complex IV is mostly resolved in its monomeric form.A higher migrating band, reacting with both core2 and COX1antibodies, is detected at �600 KDa. Second dimension, dena-turing, electrophoresis shows that this band corresponds to asupercomplex composed of complex III and complex IV(III2 þ IV, not shown). Such a III2 þ IV supercomplex hasbeen previously demonstrated in mammalian cells (22). Therelative proportion of this III2 þ IV complex appears toincrease with progressive solubilization with lauryl maltoside(Fig. 3A), suggesting that the interactions between the com-ponents of complex III and complex IV are strong and notcompletely disrupted by the solubilization.

To confirm that the high migrating bands observed in thedigitonin-treated samples correspond to supercomplexes con-taining multimers of the various respiratory chain enzymes,the same WT cybrid sample was loaded side-by-side in fouradjacent wells. After electrophoresis and blotting, each lanewas cut and probed with an antibody against subunits of differ-ent complexes. As expected, using antibodies against com-plexes I, III and IV, at least one band migrating at a commonposition can be detected (band 1 in Fig. 3B), whereascomplex V and complex II migrate at different positionsbecause they are not part of the same supercomplex. In additionto the common supercomplex band, COX1 antibodies alsoreveal the 200 KDa monomeric complex IV plus three highmigrating bands (bands 2–4 in Fig. 3B), which presumably cor-respond to supercomplexes containing variable numbers ofcomplex IV, as previously reported by Schagger (7). Thesethree higher migrating bands are better detected with theCOX1 antibody than with antibodies against the other com-plexes, possibly because complex IV subunits are the mostabundant or because epitopes for the other complexes may bepartially masked in these high molecular weight structures.

In order to ensure that digitonin solubilizes mitochondrialmembranes similarly in WT and mutant cells, a digitonin titra-tion was performed in control and in 100% mutant COX1 orCyB parental cybrids. We looked at complex II by bluenative western blot, because this complex is not affecteddirectly or indirectly by either mutation. As expected, decreas-ing digitonin below 0.3% had similar effect on the solubility ofcomplex II in all cell lines (Fig. 3C).

The assembly of the respiratory chain complexes isdefective in hybrid cells

The amounts of assembled respiratory chain complexes in hybridclones were estimated using blue native gel electrophoresis of

mitochondrial proteins solubilized with 0.4% lauryl maltoside.Compared with WT cybrids, mutant hybrids show loss of com-plexes III and IV, which worsens with increasing CyB-MT andCOX1-MT loads, respectively (Fig. 4A). In addition, the super-complex III2 þ IV is reduced when compared with WT cellsin hybrid clones containing levels of mutant mtDNA abovethe threshold for defective respiration (40 and 90%, forCOX1-MT and CyB-MT, respectively). Interestingly, in someclones, there is a loss of complex I, which is not geneticallyaffected by the two mutations. Complex I defect is most pro-nounced in the clones with the highest proportions of eitherCOX1-MT (51%) or CyB-MT (96%). Complex II is unaffectedby the mtDNA mutations and, therefore, was used to normalizethe levels of the other complexes in Figure 4. We also studiedthe levels of assembled complexes in the parental homoplasmicmutant cybrids (Fig. 4B). In COX1-MT cybrids, there is a com-plete lack of complex IV and supercomplex III2 þ IV, whereascomplex I is decreased by ~30%. In CyB-MT cybrids, thereare no complexes III and III2 þ IV, and complex I is severelydecreased by ~90%. Complex III is not decreased inCOX1-MT cybrids and complex IV is not decreased inCyB-MT cybrids, suggesting that the assembly of complexesIII and IV are independent from each other. Note that, inCyB-MT and COX1-MT homoplasmic cybrids, the levels ofcomplexes IV and III, respectively, appear higher than in WTcells. This is because, unlike WT cells, the supercomplexIII2 þ IV cannot form in mutant cybrids, and each complex isresolved in an individual band. In-gel complex I activity isreduced in the hybrid clones and in the parental mutantcybrids when compared with WT cells, consistent with thedecrease of assembled complex I (Fig. 4C and D). To confirmthat cells containing high levels of COX1-MT mtDNA havereduced complex I, we also studied heteroplasmic cybridCOX1-MT cells. Clones with mutation levels .90% have aclear reduction of complex I, whereas a clone with 33%mutant mtDNA shows normal complex I content (Fig. 4E).

Recovery of supercomplex assembly correlateswith recovery of mitochondrial respiratory functionin hybrid cells

The proportion of mutant mtDNA correlates with the levelsof their respective enzymatic activities (Fig. 2C and D).However, they are not linearly correlated with mitochondrialrespiration (Fig. 2A). Therefore, we have investigatedwhether respiration is dependent on the presence of respiratorychain supercomplexes. Mitochondrial membranes are solubil-ized with 0.33% digitonin to preserve supercomplex integrity.The high molecular weight bands, corresponding to a multi-meric structure composed of complexes I þ III2 þ IV(1 – 4),are clearly detectable in the representative clones c and dshown in Figure 5A, which have very low amounts of individ-ual III2 and IV. In these clones, the proportions of bothmutations are below the threshold for respiratory defect(Fig. 2A). Conversely, hybrid clones a and b, in whichCyB-MT mtDNA is above the threshold (i.e. .90%,Fig. 2A and B), have very reduced levels of supercomplexesI þ III2 þ IV(1 – 4). As a comparison, WT cybrids show the pre-sence of supercomplexes as well as individual complexes III2,IV and III2 þ IV (Fig. 5A). Homoplasmic CyB-MT cybrids


Dow


ic.oup.com/hm


ber 2018

with no complex III2 have a complete lack of supercomplexes,but contain individual complex IV. In contrast, hybrid clone e,which contains COX1-MT mtDNA above the respiratorydefect threshold (i.e. .40%), has no individual complex IV,

but a large amount of III2 (Fig. 5A), suggesting that all theavailable complex IV is part of a supercomplex, whereasmost of complex III is in its dimeric form. In this clone, thereduced amount of complex IV available is sufficient for the

Figure 4. Defective respiratory chain complex assembly in hybrid clones and parental mtDNA mutant cybrids. Mitochondrial respiratory chain complexes from(A) hybrids (B) and WT cybrid and from 100% COX1-MT and CyB-MT cybrids solubilized with 0.4% lauryl maltoside, resolved by blue native gel electro-phoresis, blotted and sequentially probed with antibodies: anti-39 KDa subunit for complex I, anti-core 2 for complex III, anti-COX1 for complex IV andanti-70 KDa subunit for complex II. Relative proportions of CyB-MT and COX1-MT are indicated at the bottom for each hybrid clone. Note that the proportionsof the two mutant mtDNA species do not add up to 100%, because hybrid clones contain WT molecules originating from mtDNA recombination (17). Theamounts of respiratory complexes normalized by the amount of complex II (II) and expressed as percentage of WT cells as estimated by densitometry ofthe western blot bands are indicated below each lane. (C) Complex I in-gel activity in the same samples, as in (A). (D) Complex I in-gel activity in thesame samples, as in (B). After resolving complex I by blue native gel electrophoresis, a specific NADH-dependent colorimetric assay was performed and itsintensity estimated by densitometry. Complex I activity is indicated below each lane as a percentage of WT cybrid cells. (E) Blue native western blot solubilizedand detected as described in (A). Complex I content is reduced in cybrids containing �90% COX1-MT mtDNA.


Dow


ic.oup.com/hm


ber 2018

formation of I þ III þ IV with stochiometry 1:2:1 or 1:2:2, butnot of supercomplexes with complex IV stochiometry higherthan 2. This is not readily apparent in the blot in Figure 5A(long exposure), but it is better demonstrated in Figure 5B,where gel running conditions are optimized to resolve thehigh molecular weight structures, and clone e shows amarked reduction in supercomplex content when comparedwith WT.

Cybrids containing very high proportions of COX1-MT(.90%) and no CyB-MT mtDNA (16) lack complex IV andhave no detectable supercomplex I þ III2 þ IV(1 – 4), whereasindividual complex III2 and a partially formed supercomplex(I þ III2) are detectable (Fig. 5A, right lanes). A second, dena-turing, dimension western blot of COX1-MT cybrids confirmsthe composition of this partial supercomplex, because it showsthe presence of subunits of complex III (core2) and I

Figure 5. Respiratory chain supercomplexes in hybrid clones. (A) Respiratory chain complexes and supercomplexes from WT cybrids, hybrid clones, 100%CyB-MT cybrids and 91, 99 and 100% COX1-MT cybrids were solubilized with 0.33% digitonin, resolved by blue native gel electrophoresis (5–13% polyacryl-amide gradient gel) and detected with a mixture of antibodies: anti-39 KDa subunit for complex I, anti-70 KDa subunit for complex II, anti-core 2 for complex IIIand anti-COX1 for complex IV. Relative proportions of CyB-MT and COX1-MT are indicated for each cell line. (B) Mitocondrial supercomplexes from WTcybrids, hybrid clones, 100% CyB-MT cybrids and 100% COX1-MT cybrids solubilized with 0.33% digitonin, resolved by blue native gel electrophoresis (5–8% polyacrylamide gradient gel) and detected with the antibody anti-COX 1. (C) Mitochondrial complexes and supercomplexes subunits from COX1-MTcybrids (left) and WT cybrids (right) solubilized with digitonin and resolved by first dimension blue native gel electrophoresis (as in A), followed by separationby second dimension, denaturating, gel electrophoresis of gel strips placed side by side and detection with a mixture of antibodies: anti-30 KDa subunit ofcomplex I and anti-core 2 subunit of complex III. Note that in this experiment, complex IV is detected using anti-COX IV instead of COX1 antibodies toavoid overlap with the subunits of complexes I and III.


Dow


ic.oup.com/hm


ber 2018

(30 KDa), but not of a subunit of complex IV (COX IV;Fig. 5C). In this second dimension, WT cells show broadbands detected by antibodies against complexes I, III andIV, in correspondence to the migration points of the multipleforms of the supercomplexes in the first dimension, confirmingthe presence of all three complexes in these structures.

To further investigate how respiration correlates with theamounts of individual respiratory chain complexes and withthe assembly of supercomplexes, we assayed in parallel thelevels of each individual complex, the presence of supercom-plexes and respiration in hybrid clones containing increasingproportions of CyB-MT mtDNA. The relative amount ofcomplex III (i.e. III2 and III2 þ IV) in samples solubilizedwith lauryl maltoside is shown in Figure 6A (top figure).We find a linear inverse correlation between the proportionof CyB-MT mtDNA and complex III content (Fig. 6B). Thisresult was expected on the basis of the linear inverse corre-lation between the proportion of CyB-MT mtDNA andcomplex III activity shown in Figure 2C. However, the corre-lation between complex III content and respiration is not linear(Fig. 6C). Instead, we observe a hyperbolic curve similar tothat describing the correlation between the proportion ofCyB-MT mtDNA and respiration (Fig. 2B). This type of cor-relation curve suggests that a respiration defect occurs onlywhen complex III content falls below �10% of normal. Inthe clones tested, a decrease in respiration appears to beassociated with a marked decrease of supercomplex content(Fig. 6A bottom figure). Clones with CyB-MT mtDNA,90% have normal respiration and abundant supercomplex(right lanes), whereas clones .90% have reduced respirationand very low or undetectable supercomplex (left lanes).

DISCUSSION

Functional complementation between mitochondria-containingpathogenic tRNA mutations has been demonstrated previously(13,14,23), although one report suggests that it may be a rareevent (24). Here, we demonstrate complementation resultingfrom exchange of mtDNA-encoded polypeptides among mito-chondria. We show that complementation of respiratory chainfunction occurs in hybrid clones containing a mixture of twopathogenic mutations in polypeptide-coding genes, COX1and CyB. Metabolic selection in conditions that only allowfor survival of cells with respiratory chain function (i.e.uridine and pyruvate deprivation or galactose medium)decreases the number of hybrid clones only by 25%(Fig. 1A). This indicates that complementation of mtDNA pro-teins is a frequent event following cell hybridization and thatdiffusion of molecules within fused organelles is very efficient.Rapid and efficient exchange of proteins in the matrix of fusedmitochondria has been clearly demonstrated, and membranepotential is necessary for this process (25). The parentalcybrids used in this study have no residual respiratory chainfunction. However, it is likely that they can maintain a mito-chondrial membrane potential sufficient for fusion by intra-mitochondrial hydrolysis of glycolytic ATP (26).

Although all selected hybrid clones show restored respirat-ory functions, the degree of complementation is very variable(Fig. 1B–D). The level of the recovery of respiration depends

on the proportions of WT COX1 and CyB genes present ineach hybrid clone. The correlation between mutation loadand respiration shows a clear threshold effect, which differsfor the two mutations: hybrid cells are able to tolerate amuch higher proportion of mutant CyB than mutant COX1before showing a decrease in respiration [Fig. 2A and (16)].When looking at specific enzymatic activities of individualcomplexes, measured with exogenous electron donors andacceptors (Vmax), activities of complexes III and IV show aninverse linear correlation with the proportions of the respect-ive mutations (Fig. 2C and D). As expected, the content ofassembled individual complexes inversely correlates with theproportion of mutant mtDNA (Fig. 6B). On the basis ofthese findings, we conclude that in human cells, complexesIII and IV can be assembled independently from each other,as previously demonstrated in yeast (6), and that their specificenzymatic activities are correlated to the amounts of individ-ual assembled complexes. However, this correlation does notfully explain the threshold effect observed for respiration(Figs 2B and 6C), suggesting that the amounts of individualcomplexes III are not rate-limiting for respiration, until theydecrease below a threshold level.

The difference in the threshold for respiratory defectbetween CyB-MT and COX1-MT mtDNA [Fig. 2B and(16)] may be explained by a faster turnover of complex IVwhen compared with complex III. In addition, it is possiblethat the stoichiometry of the two complexes in the supermole-cular structures is different, because of single complex IIIdimers, but multiple complex IV units appear to be containedin the supercomplex bands detected in Figure 3B.

We show a decrease in the levels of complex I in the hybridclones containing high levels of CyB-MT or COX1-MTmtDNA (Fig. 4A and C). Complex I stability has beenshown to depend on the levels of complex III in various organ-isms, including bacteria (27) and mammalian cells (8,28,29).Here, we confirm this finding, but we also show for the firsttime that complex I stability can also be affected bycomplex IV, in human cells, although high COX1 mutationlevels are needed to observe destabilization of complex I(Fig. 4). In addition, the mutation level necessary to observethis phenomenon could vary in different cell types, becausemuscle cells from patients with nuclear-encoded complex IVdefects, which contain only 10% of normally assembledcomplex IV, do not show a marked reduction of complex I(8,30). These variations could be associated with the differentrates of turnover of mitochondrial respiratory chain complexesin different cell types.

It was shown that the synthesis of complex I subunits andtheir assembly occurs normally in complex III mutant cells,but in the absence of complex III, complex I is rapidlydegraded (28). This indicates that complex III may act as achaperone that stabilizes complex I, and on the basis of ourdata, complex IV may further stabilize the supercomplex.Because I þ IV intermediates are never detected, the inter-actions between complexes I and IV are probably indirectthrough complex III. A supercomplex I þ III2 can form inthe absence of complex IV (Fig. 5A), and a stable supercom-plex III2 þ IV is observed after solubilization with laurylmaltoside (Fig. 4A). These are likely to be supercomplexassembly intermediates, perhaps forming assembly cores for


Dow


ic.oup.com/hm


ber 2018

Figure 6. Supercomplex levels are restored in hybrid clones with full respiratory complementation. (A) Top figure: hybrid clones containing variable proportionsof mutant mtDNA solubilized in lauryl maltoside and analyzed by blue native gel western blot as in Figure 4A. The relative amount of each complex normalizedby complex II and expressed as a percentage of WT is indicated below each lane. Bottom figure: the same samples are solubilized with digitonin to detect super-complexes. Respiration rates are indicated below each lane and expressed as a percentage of WT cells. The relative proportions of each mutant mtDNA species inthe clones analyzed are shown at the bottom. (B) Complex III (i.e. the sum of III and III þ IV) content is inversely correlated with the proportion of CyB-MTmtDNA. Solid diamonds represent clones studied in Figure 4A, whereas open diamonds represent clones studied in Figure 6A. (C) Correlation plot betweencomplex III content and respiration expressed as percentages of WT cells in clones studied in Figure 6A. The data are fitted by a single rectangular hyperbole,as in Figure 2B, with Kd ¼ 1.8 and Bmax ¼ 111.7.564.


Dow


ic.oup.com/hm


ber 2018

the addition of complex I and complex IV. However, itappears that fully assembled complex I is not necessary forthe assembly and stability of complexes III and IV (8),although a partial defect of complex III assembly has beenreported in a subset of individuals with mutations of nuclear-encoded complex I subunits (22). Nuclear-encoded assemblyfactors have been implicated in the formation and stabilizationof individual respiratory chain complex IV (31,32); thus, it ispossible that similar factors regulate the assembly of super-complexes, but this possibility still remains to be explored.

In WT cells, individual complexes III2 and IV are detectedunder mild solubilization conditions with digitonin, suggestingthat they are not part of a supercomplex. These individualcomplexes may have a function in respiration per se, andthey may also act as a functional reserve. When the amountsof individual complexes decrease, due to mtDNA mutationsaffecting their subunits, the reserves become depleted. Thishas virtually no effect on respiration as long as enough super-complex can form. Any further decrease in the amounts ofcomplexes III or IV results in a depletion of the supercom-plexes and a significant reduction in respiration. These con-clusions are based on the correlation observed between thepresence of supercomplexes in native western blots andrespiratory function. To obtain further confirmation, it wouldbe useful to disrupt the supercomplexes without affecting theirindividual components and test the effects on mitochondrial

respiration. Unfortunately, this could only be achieved byknocking down a hypothetical ‘supercomplex assemblyfactor’, whose existence is purely speculative at this stage.

Figure 7 shows a schematic model of supercomplex assem-bly and stabilization in WT and mtDNA mutant cells. Wepropose a semi-solid-state model of the respiratory chain,where respiring supercomplexes exist in a dynamic equili-brium with randomly organized, enzymatically active, isolatedcomplexes. Furthermore, we do not exclude that in the nativestate (i.e. without any detergent manipulation), other proteinsmay be associated with the supercomplex, forming structures,with molecular weight even higher than those detected byblue-native electrophoresis.

There are several potential advantages to a supramolecularorganization of the respiratory chain. For example, it is poss-ible that such structures allow for a more efficient utilization ofavailable substrates and cofactors than in a random collisionmodel. It is also possible that they decrease the probabilityof electron escape to generate reactive oxygen species (8).In hybrid cells, complex I, the principal source of free radicalsthrough semi-quinone (33), is always associated with com-plexes III and IV (Fig. 5), possibly to prevent excessive freeradical production.

In conclusion, this study of COX1-MT and CyB-MT hybridhuman cells provides novel clues for understanding the func-tional structures underlying mitochondrial respiration and the

Figure 7. Model of respiratory chain complexes structure and assembly in WT and mtDNA mutant cells. (A) In WT mammalian cells, respiratory chain com-plexes are organized in supramolecular structure (supercomplexes) composed of monomeric complex I, dimeric complex III and mono-, di-, tri- or tetramericcomplex IV. The supercomplexes coexist with a pool of partially assembled supercomplexes (III2 þ IV) and individual complexes (dimeric complex III andmonomeric complex IV), indicating that a functional ‘solid-state’ model, like the supercomplexes, can exist in equilibrium with randomly organized, enzima-tically active, isolated complexes. (B) In 100% CyB-MT cybrids, the complete loss of complex III results in disassembly and degradation of complex I; onlycomplex IV, mostly organized in a monomeric form, remains in the mitochondrial inner membrane. (C) In 100% COX1-MT cybrids, the total loss of complex IVresults in a partial decrease of complex I. The residual complex I is completely assembled with complex III; complex III dimers are unaffected by the COX1mutation. From this model, it is clear that complex III constitutes the structural core to which complexes I and complex IV bind to form a stable supramolecularstructure. Complex I, because of its instability, cannot exist as an individual complex.


Dow


ic.oup.com/hm


ber 2018

mechanisms that modulate the biochemical and clinical phe-notypes in mitochondrial disorders. We suggest that the pre-sence of supercomplexes sets the threshold for functionalcomplementation. Therefore, the ability to form sufficientlevels of supercomplexes emerges as one of the critical stepsmodulating the expression of mitochondrial diseases associ-ated with mutations in the respiratory chain subunits.

MATERIALS AND METHODS

Cell culture

Parental COX1-MT and CyB-MT cybrid lines were culturedin Dulbecco’s Modified Eagle’s medium (DMEM, Invitrogen)containing 4.5 g/l glucose, 110 mg/l pyruvate and 50 mg/mluridine (medium A) supplemented with 10% fetal bovineserum (FBS, Cellgro).

Cell fusions and nuclear selection of hybrid cells were per-formed using 50% (w/v) polyethylene glycol (ATCC) and acombination of puromycin (Sigma-Aldrich) and hygromycin(Invitrogen) as described (17) in medium A.

For metabolic selection, 14 days after fusion, hybrid cloneswere grown in triplicate plates with mitochondrial selectionmedia. Selection medium B contained 4.5 g/l galactose,110 mg/l pyruvate and 50 mg/ml uridine (all fromSigma-Aldrich). Selection medium C contained 4.5 g/lglucose, no pyruvate and no uridine. Both media B and Cwere supplemented with 10% dialyzed FBS. One week later(21 days after fusion), the number of surviving hybrid cloneswas counted in triplicate plates for each of the selection con-ditions. In addition, a portion of the hybrid clones selected inmedium C were collected by the cylinder method (18), culturedin the same medium and used for further studies.

Respiratory chain analyses

Growth of hybrid clones in galactose (medium B) was deter-mined by plating 100 � 103 cells in 9 � 35 mm2 plasticPetri dishes. Cells were counted in triplicate daily, for threeconsecutive days.

Oxygen consumption was measured on intact cells, usingpyruvate as substrate, with or without the protonophore carbo-nyl cyanide p-trifluoromethoxyphenylhydrazone (FCCP,2 mM, Sigma-Aldrich), in a thermostatic oxygraph chamberequipped with a Clark-type electrode (Hansatech) as described(16). The CyB-MT mtDNA threshold for respiration defectwas obtained with a non-linear regression fitting curve,using the application program SigmaPlot (Sigmaplot ScientificGraphing Software, Version 9.01).

Enzymatic assays of respiratory chain complex activitieswere performed on isolated mitochondria obtained from 10semi-confluent 150 mm dishes (�100–150 � 106 cells) asdescribed (34). The activities of complex III, IV and citratesynthase were measured by spectrophotometric assays asdescribed (34,35).

For complex I in-gel activity assay, samples were separatedby blue native gel electrophoresis and incubated in 2 mM

Tris–HCl, pH 7.4, 0.1 mg/ml NADH (Sigma-Aldrich),and 2.5 mg/ml nitrotetrazolium blue (NTB, Sigma-Aldrich).After overnight incubation at room temperature, the stained

gel was washed in distilled water and imaged with a digitalscanner.

MtDNA analyses

The relative proportions of COX1-MT and CyB-MT mtDNAspecies were determined in each hybrid clone by PCR-RFLP(restriction fragment length polymorphism) analysis ofdigested radiolabeled products, as described (17).

Blue native electrophoresis

Mitochondrial membranes were isolated from 2.5 � 106 cellsas described (36). Various concentrations of two different deter-gents were tested to optimize the solubilization of mitochon-drial proteins: lauryl maltoside (Sigma-Aldrich) from 0.01–0.4% (w/v), and digitonin (Sigma-Aldrich) from 0.3–5%(w/v). Ten microliters of sample was loaded on a 5–13% gradi-ent polyacrylamide gel. Electrophoresis was performed asdescribed (36). Transfer of proteins onto a PVDF membrane(BioRad) was carried out overnight at 30 V at 48C. For immu-nodetection of protein complexes, monoclonal antibodies(Invitrogen) against the following subunits were used:39 KDa of complex I, 70 KDa of complex II, core 2 ofcomplex III, subunit I of complex IV and subunit b ofcomplex V. Native high molecular weight markers were fromAmersham Biosciences.

For second dimension, gel electrophoresis, a lane excisedfrom the first dimension native gel was first treated withdenaturing buffer containing 15 mM b-mercaptoethanol and1% SDS for 30 min and then washed in 1% SDS for 1 h.The gel strip was then electrophoresed on a tricine–SDS–polyacrylamide gel as described (37). For immunodetection ofproteins, monoclonal antibodies (Invitrogen) against the follow-ing subunits were used: 30 KDa of complex I, core 2 of complexIII and subunit IV of complex IV. Quantification of respiratorychain complexes was performed by densitometric analyses ofband intensities on digital images of the western blots usingthe Scion2 image software.

ACKNOWLEDGEMENTS

We thank Dr Anatoly A. Starkov for critical reading of themanuscript. This work was supported by grants from theUnited Mitochondrial Disease Foundation, Muscular Dystro-phy Association (GM), NIH/NINDS K02 NS047306 (GM)and Telethon Italia Fondazione ONLUS (MD).

Conflict of Interest statement. None declared.

REFERENCES

1. Attardi, G. and Schatz, G. (1988) Biogenesis of mitochondria. Annu. Rev.Cell Biol., 4, 289–333.

2. Chance, B. and Williams, G.R. (1955) A method for the localization ofsites for oxidative phosphorylation. Nature, 176, 250–254.

3. Hatefi, Y., Haavik, A.G., Fowler, L.R. and Griffiths, D.E. (1962) Studieson the electron transfer system. XLII. Reconstitution of the electrontransfer system. J. Biol. Chem., 237, 2661–2669.


Dow


ic.oup.com/hm


ber 2018

4. Hochli, M. and Hackenbrock, C.R. (1976) Fluidity in mitochondrialmembranes: thermotropic lateral translational motion of intramembraneparticles. Proc. Natl Acad. Sci. USA, 73, 1636–1640.

5. Hackenbrock, C.R., Chazotte, B. and Gupte, S.S. (1986) The randomcollision model and a critical assessment of diffusion and collision inmitochondrial electron transport. J. Bioenerg. Biomembr., 18, 331–368.

6. Schagger, H. and Pfeiffer, K. (2000) Supercomplexes in the respiratorychains of yeast and mammalian mitochondria. Embo J., 19, 1777–1783.

7. Schagger, H. (2002) Respiratory chain supercomplexes of mitochondriaand bacteria. Biochim. Biophys. Acta, 1555, 154–159.

8. Schagger, H., de Coo, R., Bauer, M.F., Hofmann, S., Godinot, C. andBrandt, U. (2004) Significance of respirasomes for the assembly/stabilityof human respiratory chain complex I. J. Biol. Chem., 279, 36349–36353.

9. Schafer, E., Seelert, H., Reifschneider, N.H., Krause, F., Dencher, N.A.and Vonck, J. (2006) Architecture of active mammalian respiratory chainsupercomplexes. J. Biol. Chem., 281, 15370–15375.

10. Bianchi, C., Genova, M.L., Parenti Castelli, G. and Lenaz, G. (2004) Themitochondrial respiratory chain is partially organized in a supercomplexassembly: kinetic evidence using flux control analysis. J. Biol. Chem.,279, 36562–36569.

11. DiMauro, S. (2004) Mitochondrial diseases. Biochim. Biophys. Acta,1658, 80–88.

12. Chen, H. and Chan, D.C. (2005) Emerging functions of mammalianmitochondrial fusion and fission. Hum. Mol. Genet., 14 (Spec no. 2),R283–R289.

13. Ono, T., Isobe, K., Nakada, K. and Hayashi, J.I. (2001) Human cells areprotected from mitochondrial dysfunction by complementation of DNAproducts in fused mitochondria. Nat. Genet., 28, 272–275.

14. Takai, D., Isobe, K. and Hayashi, J. (1999) Transcomplementationbetween different types of respiration-deficient mitochondria withdifferent pathogenic mutant mitochondrial DNAs. J. Biol. Chem.,274, 11199–11202.

15. Rana, M., de Coo, I., Diaz, F., Smeets, H. and Moraes, C.T. (2000)An out-of-frame cytochrome b gene deletion from a patient withparkinsonism is associated with impaired complex III assembly and anincrease in free radical production. Ann. Neurol., 48, 774–781.

16. D’Aurelio, M., Pallotti, F., Barrientos, A., Gajewski, C.D., Kwong, J.Q.,Bruno, C., Beal, M.F. and Manfredi, G. (2001) In vivo regulation ofoxidative phosphorylation in cells harboring a stop-codon mutation inmitochondrial DNA-encoded cytochrome c oxidase subunit I. J. Biol.Chem., 276, 46925–46932.

17. D’Aurelio, M., Gajewski, C.D., Lin, M.T., Mauck, W.M., Shao, L.Z.,Lenaz, G., Moraes, C.T. and Manfredi, G. (2004) Heterologousmitochondrial DNA recombination in human cells. Hum. Mol. Genet.,13, 3171–3179.

18. King, M.P. and Attardi, G. (1989) Human cells lacking mtDNA:repopulation with exogenous mitochondria by complementation. Science,246, 500–503.

19. Mattiazzi, M., Vijayvergiya, C., Gajewski, C.D., DeVivo, D.C., Lenaz, G.,Wiedmann, M. and Manfredi, G. (2004) The mtDNA T8993G (NARP)mutation results in an impairment of oxidative phosphorylation that can beimproved by antioxidants. Hum. Mol. Genet., 13, 869–879.

20. Hao, H., Morrison, L.E. and Moraes, C.T. (1999) Suppression of amitochondrial tRNA gene mutation phenotype associated with changes inthe nuclear background. Hum. Mol. Genet., 8, 1117–1124.

21. Deng, J.H., Li, Y., Park, J.S., Wu, J., Hu, P., Lechleiter, J. and Bai, Y.(2006) Nuclear suppression of mitochondrial defects in cells without theND6 subunit. Mol. Cell. Biol., 26, 1077–1086.

22. Ugalde, C., Janssen, R.J., van den Heuvel, L.P., Smeitink, J.A. andNijtmans, L.G. (2004) Differences in assembly or stability of complex I

and other mitochondrial OXPHOS complexes in inherited complex I

deficiency. Hum. Mol. Genet., 13, 659–667.

23. Nakada, K., Inoue, K., Ono, T., Isobe, K., Ogura, A., Goto, Y.I., Nonaka, I.and Hayashi, J.I. (2001) Inter-mitochondrial complementation:

mitochondria-specific system preventing mice from expression of disease

phenotypes by mutant mtDNA. Nat. Med., 7, 934–940.

24. Enriquez, J.A., Cabezas-Herrera, J., Bayona-Bafaluy, M.P. and Attardi, G.

(2000) Very rare complementation between mitochondria carrying

different mitochondrial DNA mutations points to intrinsic geneticautonomy of the organelles in cultured human cells. J. Biol. Chem.,

275, 11207–11215.

25. Legros, F., Lombes, A., Frachon, P. and Rojo, M. (2002) Mitochondrialfusion in human cells is efficient, requires the inner membrane potential,

and is mediated by mitofusins. Mol. Biol. Cell., 13, 4343–4354.

26. Buchet, K. and Godinot, C. (1998) Functional F1-ATPase essential inmaintaining growth and membrane potential of human mitochondrial

DNA-depleted rho degrees cells. J. Biol. Chem., 273, 22983–22989.

27. Stroh, A., Anderka, O., Pfeiffer, K., Yagi, T., Finel, M., Ludwig, B. andSchagger, H. (2004) Assembly of respiratory complexes I, III, and IV into

NADH oxidase supercomplex stabilizes complex I in Paracoccusdenitrificans. J. Biol. Chem., 279, 5000–5007.

28. Acin-Perez, R., Bayona-Bafaluy, M.P., Fernandez-Silva, P.,

Moreno-Loshuertos, R., Perez-Martos, A., Bruno, C., Moraes, C.T. andEnriquez, J.A. (2004) Respiratory complex III is required to maintain

complex I in mammalian mitochondria. Mol. Cell, 13, 805–815.

29. Blakely, E.L., Mitchell, A.L., Fisher, N., Meunier, B., Nijtmans, L.G.,Schaefer, A.M., Jackson, M.J., Turnbull, D.M. and Taylor, R.W. (2005)

A mitochondrial cytochrome b mutation causing severe respiratory chainenzyme deficiency in humans and yeast. FEBS J., 272, 3583–3592.

30. Diaz, F., Thomas, C.K., Garcia, S., Hernandez, D. and Moraes, C.T.

(2005) Mice lacking COX10 in skeletal muscle recapitulate the phenotypeof progressive mitochondrial myopathies associated with cytochrome c

oxidase deficiency. Hum. Mol. Genet., 14, 2737–2748.

31. Shoubridge, E.A. (2001) Cytochrome c oxidase deficiency. Am. J. Med.

Genet., 106, 46–52.

32. Herrmann, J.M. and Funes, S. (2005) Biogenesis of cytochrome

oxidase-sophisticated assembly lines in the mitochondrial innermembrane. Gene, 354, 43–52.

33. Ohnishi, S.T., Ohnishi, T., Muranaka, S., Fujita, H., Kimura, H.,

Uemura, K., Yoshida, K. and Utsumi, K. (2005) A possible site ofsuperoxide generation in the complex I segment of rat heart mitochondria.

J. Bioenerg. Biomembr., 37, 1–15.

34. Trounce, I.A., Kim, Y.L., Jun, A.S. and Wallace, D.C. (1996) Assessment

of mitochondrial oxidative phosphorylation in patient muscle biopsies,lymphoblasts, and transmitochondrial cell lines. Meth. Enzymol.,

264, 484–509.

35. Fato, R., Cavazzoni, M., Castelluccio, C., Parenti Castelli, G., Palmer, G.,Degli Esposti, M. and Lenaz, G. (1993) Steady-state kinetics of

ubiquinol–cytochrome c reductase in bovine heart submitochondrialparticles: diffusional effects. Biochem. J., 290, 225–236.

36. Nijtmans, L.G., Henderson, N.S. and Holt, I.J. (2002) Blue native

electrophoresis to study mitochondrial and other protein complexes.Methods, 26, 327–334.

37. Brookes, P.S., Pinner, A., Ramachandran, A., Coward, L., Barnes, S.,

Kim, H. and Darley-Usmar, V.M. (2002) High throughputtwo-dimensional blue-native electrophoresis: a tool for functional

proteomics of mitochondria and signaling complexes. Proteomics,2, 969–977.


Dow


ic.oup.com/hm


ber 2018

Documents

Respiratory chain supercomplexes set the threshold for ......Upon cell fusion, respiration is recovered in hybrids cells, indicat-ing that mitochondria fuse and exchange genetic and