ELSEVIER Biochimica et Biophysica Acta 1317 (1996) 199-209

BB Biocht~Pic~a e t B i o p h y s i c a A c t a

Antimycin resistance and ubiquinol cytochrome c reductase instability associated with a human cytochrome b mutation

Mohamed Fouad Bouzidi a, Henri Carrier b, Catherine Godinot a,* Centre de G£n~tique MoHculaire et Cellulaire, UMR 5534, Centre National de la Recherche Scientifique, Universit~ Claude Bernard Lyon 1, F69622

Villeurbanne Cedex, France b Groupe de Recherche en pathologie neuromusclaire, Laboratoire de physiopathologie m£tabolique et r£nale, INSERM-CRI, 950201 Lyon, France

Received 31 July 1996; revised 11 September 1996; accepted 12 September 1996

Abstract

Progressive exercise intolerance was associated with a decreased maximal rate of ubiquinol cytochrome c reductase (complex III) activity in the muscle mitochondria of the studied patient and with a thirty five-fold increase in the 150 for antimycin A. In contrast, myxothiazol sensitivity was not altered. Complex III activity was stable at 37°C, but progressively decreased at 4°C. An heteroplasmic G to A mutation at position 15615 of the mitochondrial DNA, resulting in the replacement of the highly conserved Gly29 o in cytochrome b by Asp, was identified. Histochemical studies showed increased cytochrome oxidase and succinate dehydrogenase activities under the sarcolemma of type I fibres, After partial extraction of mitochondria from the muscle, the residual pellet contained a lower percentage of the mutation than did whole muscle, suggesting that the percentage of mutation is higher in the most readily extracted mitochondria, most probably present under the sarcolemma. In the current 8 transmembrane helix model of cytochrome b, Gly29 o lies at the end of the sixth transmembrane helix, facing the intermembrane space and close to the presumed sites of interaction between cytochrome b, the iron-sulfur protein and the 9.5 kDa protein. Since immunoblotting experiments showed a relative decrease in the proportions of these three subunits in the patient's mitochondria compared with the other complex III subunits, it is probable that the complex III instability and the relative decrease in these subunits are related to the mutation. The relationship between the decrease in the apparent affinity for antimycin A and the instability of complex III are discussed.

Keywords: Mitochondrial disease; Ubiquinol cytochrome c reductase; Cytochrome b; Human pathology; Antimycin A; Myxothiazol

1. Introduction

Complex III (bcl complex or ubiquinol cytochrome c reductase, EC 1.10.2.2) catalyzes proton translocation across the mitochondrial inner membrane during electron transfer between ubiquinol and cytochrome c. The enzyme purified from bovine heart mitochondria contains eleven distinct peptide subunits [1]: the largest subunits, core proteins I and II are important for the correct assembly of complex III; the next largest peptides, cytochromes b and cl and the 'Rieske' iron-sulfur protein, are the central catalytic units [1,2]. Among the six smallest subunits are the 13.4 kDa protein, the 9.5 kDa quinone-binding protein, the 9.2 kDa 'hinge' protein, essential for cytochrome c binding, the 8.0 kDa protein that has been shown to be the

* Corresponding author. Fax: + 33 72 440555. E-mail: godinot @cismsun.univ-lyon 1 .fr

mitochondrial targetting sequence of the iron-sulfur pro- tein, and the 7.2 and 6.4 kDa proteins. Cytochrome b, the only subunit encoded by mtDNA [3], contains two heme groups and is the major transmembrane component of Complex III [4]. MtDNA sequences evolve 6 -17 times faster than nuclear DNA sequences and 43 polymorphisms of the cytochrome b gene sequence have been described in human [5]. The majority either have no effect on the protein sequence or are commonly found in the normal population and have no known influence on cytochrome b function or assembly and on complex III activity. A de- cline in cytochrome b content during aging has been associated, at least in females, with a decrease in ubiquinol cytochrome c reductase activity [6]. However, the cause of the reduction in cytochrome b expression is still discussed. No specific mutation of human cytochrome b has been convincingly correlated to modifications of human cy- tochrome b properties, until now. Two mutations associ- ated to Leber Hereditary Optic Neuropathy (LHON) have

0925-4439/96/$15.00 Copyright © 1996 Elsevier Science B.V. All rights reserved. PH S0925-4439(96)00055-5

200 M. Fouad Bouzidi et al. / Biochimica et Biophysica Acta 1317 (1996) 199-209

been described in position 15257 or 15812 of the mtDNA sequence, both replacing a G with an A and converting Aspm to Asn and Val356 to Met, respectively [7]. Since ASP171 is highly conserved phylogenically and since all vertebrates have either a Val, Leu or Ile at position 356, these substitutions might have functional significance. However, no alteration in complex III activity has yet been demonstrated in the mitochondria of the blind patients bearing these mutations. Recently, a new G to A transition in position 15615 of the mtDNA sequence [8] was discov- ered in a patient presenting with progressive exercise intolerance linked to a decrease ubiquinol cytochrome c reductase activity in the muscle mitochondria [9]. This mutation replaces Gly290 of cytochrome b by Asp and, since GlY290 is conserved in cytochrome b from more than 900 species [10], this change is therefore a good candidate for explaining the Complex III deficiency.

This paper describes the first example in human of a mitochondrial complex III with modified properties related to a cytochrome b mutation.

mannitol, 0.5% bovine serum albumin, 1 mM EGTA, 10 mM Hepes, adjusted to pH 7.4 with KOH). It was then incubated for 5 min with 1 mg of subtilisin per g of muscle in 5 ml of buffer A, then diluted with 25 ml of buffer A and the mixture centrifuged at 15 000 X g for 5 min. The pellet was suspended in 30 ml of buffer A with a motor- driven Teflon-glass Potter homogenizer. The homogenate was centrifuged at 1000 X g for 5 min before filtering the supernatant through four layers of cheese-cloth to remove fat and fibrous tissue. The pellet was resuspended in 20 ml of buffer A, homogenized again, centrifuged and the su- pernatant filtered as above. The final pellet (C 1) was kept at - 80°C for DNA extraction, while the two filtrates were centrifuged at 15000 X g for 10 min. The mitochondria enriched pellets were suspended in 1.5 ml of buffer A, centrifuged at 1000 X g for 5 min and the supernatant was centrifuged at 15000 X g for 10 min. The final mitochon- drial fraction was suspended in a minimal volume of buffer A (approx. 20-50 mg protein/ml).

2.4. Oxypolarographic measurements

2. Materials and methods

2.1. Chemicals

Chemicals were supplied either by Sigma Chemical Co. (St. Louis, MO, USA) or by Boehringer (Mannheim, Ger- many) except otherwise indicated. Taq DNA polymerase was from Appligene.

2.2. Collection of specimens

All experiments were performed with the informed consent of the patients. Controls were from patients com- plaining of muscle problems but that were subsequently shown to have no morphological, histochemical, biochemi- cal or genetic basis. The clinical condition of the patient with progressive exercise intolerance has been previously described [9].

A biopsy was taken from the vastus lateralis by open surgery, under local anesthesia. A sample was promptly snap-frozen in isopentane cooled by liquid nitrogen before making 8-/zm thick transverse sections. Two samples were kept in liquid nitrogen for molecular biology and enzy- matic analysis. Another sample was transported to the laboratory within half an hour in ice-cold transport buffer (150 mM KCI, 1 mM EDTA and 10 mM potassium phosphate buffer, pH 7.4), for mitochondria isolation.

2.3. Preparation of mitochondria

All the following operations were conducted at 0-4°C. The muscle sample was stripped of fat and connective tissues, weighed, finely chopped with scissors and washed in homogenization buffer A (70 mM sucrose, 220 mM

Substrate oxidation was measured by oxypolarography in an Hansatech oxygraph equipped with a Clark electrode using about 100 /zg of mitochondrial proteins in 0.5 ml of respiratory medium (225 mM sucrose, 1 mM EDTA, 5 mM MgC12, 10 mM potassium phosphate, 10 mM Tris- HC1, pH 7.4), in the presence of the indicated substrates. After 2 min of incubation at 30°C, 80/xM ADP was added and the rate of oxygen consumption was measured. Protein concentration was determined with the bicinchoninic acid reagent according to the manufacturer's instructions (Pierce). Those mitochondria not used for polarographic measurements were treated by the addition of a cocktail of protease inhibitors (4 mM phenylmethylsulfonyl fluoride, 1 mM EDTA and 2 mM e-aminocaproate, final concentra- tions), centrifuged at 17 000 X g for 20 min and the pellets stored at -80°C for enzymatic or immunological studies.

2.5. Measurement of ubiquinol cytochrome c reductase activity

Ubiquinol cytochrome c reductase activity was mea- sured either in freeze-thawed mitochondria suspended at about 0.2 mg of protein per ml of 10 mM potassium phosphate buffer, pH 7.5, containing the protease inhibitor cocktail, or in muscle homogenate prepared by homoge- nization (Teflon-glass Potter homogenizer) of 20-50 mg of frozen muscle in 0.4-1 ml of 10 mM potassium phos- phate buffer, pH 7.5, containing the protease inhibitor cocktail, followed by centrifugation at 600 X g for 5 min to remove nuclei and muscle fragments. Unless otherwise indicated, ubiquinol cytochrome c reductase activity was measured at 37°C from the quasi-linear initial rate of cytochrome c reduction at 550 nm in 1 ml of buffer containing 55 mM potassium phosphate, pH 7.5, 2.5 mg of

M. Fouad Bouzidi et al. / Biochimica et Biophysica Acta 1317 (1996) 199-209 201

fatty-acid free bovine serum albumin, 1 mM KCN, 0.1 mM horse cytochrome c, 20/xg of rotenone and 60 /zM of decylubiquinol prepared by dithionite reduction of 20 mM decylubiquinone, dissolved in DMSO (by heating for 10 min at 37°C), followed by elimination of excess dithionite by centrifugation at 17 000 × g for 15 min. After 1 min, 2.5 to 25 /xg of proteins (mitochondria or muscle ho- mogenate) were added and the quasi-linear initial rate of cytochrome c reduction measured. To test the inhibition of complex III by antimycin or myxothiazol, the mitochon- dria were preincubated for 1 min with the inhibitor before addition to the reaction cuvette. The data were calculated after subtraction of the rate observed either in the absence of protein or in the presence of 2 nmol of antimycin or myxothiazol. Myxothiazol and antimycin A were dissolved in absolute ethanol.

2.6. Immunological studies

Previously described procedures [9] were performed, using polyclonal rabbit anti-complex III antibodies, mono- clonal mouse anti-F1-ATPase /3 subunit antibody [11] or antibodies against subunits of complex IV (a monoclonal anti-subunit IV and a polyclonal anti-subunit II, generous gifts from B. Kadenbach and A. Lombes, respectively). Complex III proteins reacting with the antibodies were identified by comparing the migration pattern obtained using human mitochondria with that of complex III iso- lated from beef heart in the gel system described by Sch~igger et al. [12]. Control and patient mitochondrial proteins (30 /zg), separated by tricine-SDS-PAGE by the method of Sch~igger and von Jagow [13] on a 10% poly- acrylamide separating gel, were transferred to Immobilon P [14]. The proportion of each Complex III subunit present in the patient's mitochondria was estimated and compared with that from control mitochondria present on the same blot, using primary antibodies directed against the tested protein and peroxidase-labelled second anti-mouse or anti- rabbit immunoglobulin antibodies (Biosis) and the ECL immunodetection technique (Amersham). Preliminary ex- periments were performed with various amounts of mito- chondria in order to determine the optimal antibody dilu- tions and exposure times to avoid over-exposure of the X-ray film and to obtain signals representative of the protein amounts present on the nitrocellulose blot. The films were analysed by densitometry using a Sony camera coupled to a Macintosh IIFX and the image analysis software of Biolab.

2.7. Mt DNA analysis

showed a G15615A mutation in the mtDNA, resulting in the generation of an Aat II restriction site in cytochrome b gene, and a T14766C mutation resulting in the loss of a MseI restriction site. Cytochrome b fragments were ampli- fied by polymerase chain reaction, using the 20-mer primers H15437 and L15777, for the first mutation, and H14640 and L14815, for the other, at concentrations of 0.5 /zM. The numbers used are those of the Cambridge mtDNA sequence [15] and the letter represents the heavy or light chain, respectively. The amplification was conducted in the presence of 0.2 mM of each dNTP, of 1 U of Taq DNA polymerase (Appligene) and of the Appligene buffer. Thirty cycles each of 1 min denaturation at 93°C, 1 min annealing (53°C, in the first case, or 55°C, in the second) and 1 min extension at 72°C were carried out. The ampli- fied DNA was digested with either Aat II at 37°C (muta- tion 15615) or Mse I (Tru 91) at 65°C (mutation 14766), as recommended by the enzyme manufacturer (Boehringer). After separation of the fragments by agarose electrophore- sis in TBE buffer (Tris-HCl 50 mM; boric acid 50 mM; EDTA 1 mM; pH 8.0) and staining with ethidium bromide, the relative amount of DNA, before and after restriction enzyme digestion was estimated by densitometric scanning of the gel, using a Sony camera coupled to a Macintosh IIFX and the image analysis software of Biolab. A stan- dard curve made with various dilutions of the amplified DNA was prepared to estimate the amount of digested DNA from the signal intensity.

2.8. Histochemistry

Histochemical methods included modified Gomori's trichrome [16], succinic dehydrogenase [17], myosin AT- Pase [18] and cytochrome oxidase reactions [19]. Quantita- tive data were obtained with a computer-aided semi-auto- matic analyser ASM 68K from Leica.

2.9. Human cytochrome b model

The human cytochrome b model was derived from hydropathic profile calculations based on the human se- quence [15] and comparison with other cytochrome b sequences [10]. The model, made by analogy with that proposed for the cytochrome b of several species, includes 8 transmembrane a-helices and one amphipathic c~-helix located between transmembrane helices III and IV and lying on the surface of the intermembrane face of the inner mitochondrial membrane. It is widely accepted [20-25]. The limits taken for the helices are those defined by Degli Esposti et al. [10].

Following proteinase K digestion, control and patient DNA was extracted either from C1 pellets partially de- pleted of mitochondria or from whole muscle, using the phenol-chloroform technique. Complete sequence analysis of the patient's cytochrome b, described elsewhere [8]

3. Results

Fig. 1 summarizes the data for the rates of oxidation of substrates entering the respiratory chain via complex I,

202 M. Fouad Bouzidi et al. / B iochimica et Biophysica Acta 1317 (1996) 199-209

8oo~ c

g~' ' i "0 o. 600-

, o o .

"6 6 200.

~" 2 o

Glutamate Succinate Aseorbate + Malate + Rotenone + TMPD

Fig. 1. Oxidation rate of substrates entering the mitochondrial respiratory chain via complex 1, II or IV. Substrate oxidation for the patient (right) and control (left) mitochondria were measured after addition of 80 /xM ADP, in the presence of 10 mM glutamate + 1 mM malate, 10 mM succinate + 0.6 /zM rotenone or 10 mM ascorbate + 50 /~M tetram- ethyl phenylene diamine, as described in Section 2. Control data are expressed as the mean of the rates measured for 16 control subjects _+ SEM.

1 O01 e- ra w 80 ' E ~o 60. c o 40 . u

20'

0 Core ~ Fe-S ,13.4 9.5 , n IV

I+11 Cyt. kDa subunits

Complex V Complex III Complex IV

Fig. 2. Immunological titration of various complex III, IV or V subunits in mitochondria from the patient and from controls. For each protein tested, the 100% level represents the mean signal intensity measured on the same blot for mitochondria from two controls. The bars indicate the variability between the two controls. Cytochrome b and subunit II of complex IV, both encoded by mitochondrial DNA, are indicated in gray and the other proteins, of nuclear origin in white.

complex II or complex IV. The oxidation rates of gluta- mate + malate (complex I) or succinate + rotenone (com- plex II) were markedly decreased while that for ascorbate + tetraphenylene diamine (complex IV) was essentially unchanged, suggesting that the patient was deficient in complex III activity.

A kinetic study of complex III activity showed that the K m for ubiquinol or cytochrome c were not significantly modified, while the maximal rate of ubiquinol cytochrome c reductase was reduced by approx. 80%, when compared to control mitochondria and expressed per mg of mito- chondrial protein (Table 1). It was not feasible to make a precise estimation of the amount of cytochrome b in the mitochondria by dual-wavelength spectrophotometric anal- ysis because of the high amount of sample required for this technique.

Table 1 Kinetic analysis of Complex III in patient and control mitochondria

Variable substrate Patient Control

Ubiquinol K m (/zM) 10 10+2 V m (/xmol • min- ]. mg of protein- i ) 0.3 1.5 + 0.3 Cytochrome c Km (/zM) 4 9+3 V m (/zmol.min- I -rag of protein -1) 0.3 1.5+0.3

Ubiquinol-cytochrome-c reductase activity was measured at 37°C, as described in the Section 2, in the presence of either 50/zM cytochrome c and various concentrations of ubiquinol or in the presence of 60 /zM ubiquinol and various concentrations of cytochrome c. Cytochrome ¢ reduction was followed by the difference in the spectra at 550 and 540 nm (extinction coefficient 19.0 mM- ]. cm- i according to [42]). K m and V m were determined from three different control mitochondrial prepara- tions. For each preparation, the best estimates of the K m and V m were calculated by least square analysis. Data are given as the mean of the three different estimations + the SEM.

However, the levels of complex III subunits have been compared by immunological assay, to those of subunits II and IV of cytochrome oxidase and to that of the subunit /3 of F1-ATPase in the mitochondria of the patient and of controls (Fig. 2). Although this technique is quantitatively much less precise than the spectrophotometric determina- tion of cytochrome b, it gives an idea of the relative amount of each protein present in patient 's mitochondria compared with those of controls. Amounts of subunits II and IV of cytochrome oxidase and F1-ATPase /3 subunit in the patient were normal while those of the complex III subunits were decreased. This decrease was barely signifi- cant for core I + core II subunits, cytochrome c~ or the 13.4 kDa protein while the levels of cytochrome b, the iron-sulfur protein and the 9.5 kDa protein were only about 20 to 30% of those seen in control mitochondria, which could explain the low maximal rate of ubiquinol cy- tochrome c reductase activity seen in the patient and might be related to the instability of assembled complex III (see below).

This lack of stability of the patient complex III activity was demonstrated by 4°C incubation of mitochondria pre- sent in an homogenate freshly prepared from muscle stored in liquid nitrogen. Fig. 3 shows that while the control complex III activity was as stable at 4°C as at 37°C, the pat ient 's complex III activity had decreased by 50%, after 7 h of incubation at 4°C, and, in contrast, remained stable at 37°C for at least 10 h.

The sensitivity of the patient 's complex III to inhibitors known to bind to different loci on complex III was tested (Fig. 4). The inhibitory effect of myxothiazol on ubiquinol cytochrome c reductase was very similar on mitochondria from the patient and controls, while that of antimycin A was drastically reduced in the case of the patient 's mito- chondria, with the concentration required for 50% of inhi- bition of ubiquinol cytochrome c reductase activity being

M. Fouad Bouzidi et al . /Biochimica et Biophysica Acta 1317 (1996) 199-209 203

• ; i_~

x 4 x 40, o

8 8 i u . n . n . I . I . / / I

H

0 2 4 6 8 10 48

37°C

i I • i , i , l , ! , i

2 4 6 8 10 12

Incubation time (hours) Incubation time (hours)

Fig. 3. Influence of temperature on the stability of ubiqninol cytochrome c reductase. Ubiquinol cytochrome c reductase activity was measured on a freshly prepared muscle homogenate, as described in Section 2. The muscle homogenate was incubated either in an ice bucket at 0-4°C or in a water bath at 37°C. At the indicated times, samples were withdrawn from the muscle homogenate prepared from the patient (empty squares) or from a control (black diamonds) to measure the ubiquinol cytochrome c reductase activity.

about 0.03 and 1.1 nmol antimycin A per mg of protein in control and patient mitochondria, respectively, i.e. an in- crease of more than 35-fold in the patient.

In several species, cytochrome b mutations have been correlated with a loss of antimycin A sensitivity [25]; these data therefore suggested that such a mutation in this patient might be responsible for the observed decrease in antimycin A sensitivity. Sequencing of the patient's cy- tochrome b showed two differences from the Cambridge sequence [15], T to C and a G to A mutations at nu- cleotides 14766 and 15615 of the mtDNA sequence, re- spectively, replacing Ile 7 with Thr and Gly290 with Asp [8]. The first mutation results in the loss of an MseI restriction site in the mtDNA; MseI restriction of a cy- tochrome b fragment amplified by PCR and spanning nucleotides 14640 to 14815 should thus give rise to frag- ments of 127, 28, 17 and 4 bp in the control Cambridge sequence and to fragments of 144, 28 and 4 bp in the absence of the MseI site, as in mutation T14766C. Fig. 5 shows that a sample from a control, N1, gave one 144 bp

band and the expected band at 127 bp, the presence of the 144 bp band being due to incomplete cleavage, while samples from both the patient and another control, N2, gave a 144 bp fragment and bands smaller than 50 bp which were difficult to visualize. The absence of the 127 bp band, noted with samples from the patient and several control subjects such as N2, results from a polymorphism, which is apparently frequent in Caucasians [5] and there- fore unlikely to be related to the complex III deficiency. In fact, Fig. 4 shows that the specific activity of ubiquinol cytochrome c reductase and the sensitivity to antimycin A were similar in mitochondria prepared from control N1 (with mtDNA having the MseI site at position 14766) and from control N2 (mtDNA lacking this site), confirming that this mutation does not affect these properties of complex III.

Fig. 6 shows that the 341 bp fragment, spanning nu- cleotides 15437 to 15777, amplified from the DNA ex- tracted from the patient's muscle, could be partially cleaved by Aat II into two fragments of 182 and 159 bp, while no

100, > , ,

BO, .>

60,

~ 4 0 .

E 2 O . 0

O, 0

lOO

~ 8 o ._>

~ 4 0 .

~ 2 0

- - , . - - , ~ 0 1 2 3

Antimycin A (nmo l /mg prot)

L._ b .

. 0 . , - . - v

0 0.1 0.2 0.3

Myxot hlazol (nmol /mg prot)

Fig. 4. Inhibition of ubiquinol cytochrome c reductase by antimycin A (left) or myxothiazol (right) in patient (black squares) or control mitochondria. Controls with (triangles) or without (empty squares) the T 1 4 7 6 6 C mutation in mtDNA were used. Ubiquinol cytochrome c reductase activity was measured as described under Table 1, using 100 /,~M cytochrome c and 6 0 / z M decylubiquinol. Myxothiazol and antimycin A were dissolved in ethanol. 100% of activity corresponds to 1.3, 1.4 and 0.3 /zmol. rain- i .mg of protein- ] for controls with or without, the above mutation and for the patient, respectively.

204 M. Fouad Bouzidi et al. /Biochimica et Biophysics Acta 1317 (1996) 199-209

Mse I sites

@ 14640 14815

I II I

14787 14783

@

%v-- bp - 300- 176 - 150- - 144 -

127

so-

Fig. 5. Cytochrome b polymorphism at position 14766 revealed by

restriction with Mse I. The mtDNA of the patient P and of two controls,

Nl and N2, was amplified by PCR between positions 14640 and 14815 of

human mtDNA [15]. The fragments were digested by 5 units of Mse I for

3 h at 65°C under mineral oil, then separated by agarose gel electrophore-

sis (3% Multipurpose agarose). A: predicted restriction pattern of the

amplified fragment. The mutation at position 14766 deletes the Mse I

restriction site. B: agarose gel electrophoresis stained with ethidium

bromide. Left lane: DNA ladder; - indicates incubation without Mse I

and + indicates incubation with Mse I.

such cleavage was seen with control DNA (not shown).

Increasing the Aat II/DNA ratio resulted in no increase in the proportion of cleaved DNA: in lanes 3-5, from digests containing increasing amounts of muscle DNA and a fixed amount of Aat II, the amount of the 341 bp band and of the 182 + 159 bands increased simultaneously and, in all cases, about 85% cleavage was obtained. In lanes 1 and 2, the 341 bp fragment was amplified from DNA purified from the Cl pellet obtained after two successive mitochon- drial extractions of the muscle. In this case, the mitochon-

drial DNA, still bound to the nuclei and cell debris, was less extensively cleaved by Aat II (approximately 30%); a 3-fold increase in the DNA/Aat II ratio had no effect (compare lanes 1 and 2). Mse I digestion of the 176 bp fragment obtained after amplification of the Cl pellet DNA, (as described in Fig. 5) gave the same pattern as that

seen with muscle DNA (not shown). The incomplete cleavage obtained with Aat II therefore suggests hetero- plasmy of muscle mtDNA, with the more readily ex- tractable mtDNA containing a higher percentage of muta- tion. Since the mitochondria located under the sar- colemma, at the periphery of the fibre, are easier to extract than those inserted between the myofibrils [26], this sug- gested that the peripheral mitochondria may contain a higher percentage of mutation than mitochondria from other origins.

This result can be correlated with histochemical results

from the patient. Fig. 7 shows a striking accumulation of

Aat II site

0 A 15437 15777

I I I

Cl Muscle

11 213 4 51

0 B 1000~ bp

%r- 300 = - 341

150- = 182 159

Fig. 6. Percentage heteroplasmy of mutation 15615 in muscle or Cl pellet

from the patient. The mtDNAs from the patient’s whole muscle and the

Cl pellet prepared as described in Section 2 were amplified by PCR

between position 15437 and 15777 of human mtDNA [15]. The fragments

were digested by 3 units of Aat II for 4 h at 37”C, then separated by

agarose gel electrophoresis (2% Multipurpose agarose). A: predicted

restriction pattern of the amplified fragment. The mutation at position

15615 creates an Aat II restriction site. B: agarose gel electrophoresis

stained with ethidium bromide. Left lane: DNA ladder; 1 and 2: 5 and 10

~1 of the PCR reaction from the Cl pellet DNA; 3-5: 2, 5 and 10 ~1 of

the PCR reaction from the whole muscle DNA, after digestion with Aat

II.

subsarcolemmal mitochondria, using the modified Gomori’s trichrome. In serial cross sections of the vastus lateralis muscle, an intense reaction was seen at the periph- ery of all type I fibres (negative for myofibrillar ATPase activity at pH 9.4, as shown in Fig. 8A), with the succinic dehydrogenase (Fig. 8B) and the cytochrome oxidase stains (Fig. IX). In dd’t’ a 1 ion, the percentage of type I fibres, determined by counting 400 fibres, was of 63% instead of

Fig. 7. Mitochondrial aggregates in a muscle cross section. A normal fibre (left) is surrounded by others harbouring red-stained ragged subsar-

colemmal mitochondrial aggregates with moditied Gomori’s trichrome

stain. Scale bar = 50 Km.

M. Fouad Bouzidi et al. / Biochimica et Biophysica Acta 1317 (1996) 199-209 205

Fig. 8. Histochemistry on serial muscle cross-sections. The clear type 1 fibres are more abundant than the dark type 2 ones and exhibit increased activity of succinic dehydrogenase and cytochrome oxidase within subsarcolemmal mitochondrial aggregates. A: myofibrillar ATPase, B: succinic dehydrogenase, C: cytochrome oxidase. Scale bar = 100 /zm. Numbers indicates some type 2 fibres that can be followed by their shape in the serial sections.

the normal 36% seen for the same muscle (vastus lateralis) in sedentary adult males, giving a type 1/type 2 ratio of 1.7 for the patient and 0.6 for controls.

4. Discussion

A complex III deficiency in the muscle mitochondria of the reported patient was first suggested by the low oxygen consumption, observed with either glutamate + malate or with succinate + rotenone, and the normal respiration with ascorbate + tetramethyl phenylene diamine. Taylor et al. [27] have shown that complex III has a much greater effect on electron flux via succinate cytochrome c reductase in muscle than in other tissues such as liver. It is therefore not surprising that a decrease in the maximal rate of ubiquinol cytochrome c reductase activity induced such a dramatic decrease in the rate of substrate oxidation.

Histochemical analysis of the patient's muscle biopsy demonstrated an increase in the relative amount of type I fibres, when compared to the same muscle of healthy non-athletes control males, and a striking accumulation of mitochondria under the sarcolemma of all such fibres, that is a common finding in other human respiratory chain defects. The increase in the proportion of oxidative type I fibers means that, in the deficient muscle, a compensation has taken place in which the production of these fibers have been induced by an as yet unknown mechanism. The fact that it is those mitochondria, which are most readily extractable from muscle, which contain a higher proportion of mutated DNA suggests that the defective mitochondria accumulate under the sarcolemma. Indeed, the percentage of DNA containing the G15615A mutation in whole muscle (about 85%) is higher than in the C1 pellet (about 30%) obtained after two successive mitochondrial extractions of muscle. In animal muscle, it has been shown that the mitochondria located under the sarcolemma are much eas- ier to extract than those inserted between the myofibrils

[26], suggesting that, in the muscle fiber, intact mitochon- dria remain near the myofibrils where they will be most effective in muscle contraction. This observation might help to explain the threshold effect often reported in mitochondrial pathology in which 80 to 90% of the mtDNA must be mutated before any serious clinical signs are seen [28].

The complex III deficiency detected in this patient was due to a decrease in the maximal rate of ubiquinol cy- tochrome c reductase activity, thus blocking the electron flux through the respiratory chain. The Krn of the complex for ubiquinol or cytochrome c was not significantly differ- ent from that of control mitochondria suggesting that the decreased activity might be linked to a defect in biogene- sis, assembly or stability of the complex. Immunological studies using a polyclonal antiserum directed against vari- ous complex III subunits demonstrated that, while several other subunits (the /3 subunit of ATP synthase or subunit II or IV of cytochrome oxidase) were present in normal quantities and the amounts of core protein I and II, cy- tochrome cl and the 13.4 kDa proteins were not signifi- cantly different from those in control mitochondria, the amounts of cytochrome b, the iron-sulfur protein and the 9.5 kDa protein were all markedly decreased. It cannot be ascertained that the immunological titration of cytochrome b present in the mitochondria is as precise as would be a spectrophotometric titration (which cannot be performed with the limited amount of mitochondria obtained in a human biopsy). However, the immunological titration indi- cates that cytochrome b, the Fe-S protein and the 9.5 kDa protein are present in an amount corresponding to about 20 to 30% of the controls. Similarly, there is about 20% of residual ubiquinol cytochrome c reductase activity. It is likely that this activity corresponds to the amount of correctly assembled complex and that the turnover of the assembled complex, if it could be expressed as activity per mole of assembled cytochrome b, would not be affected by the mutation. The mutation induces an assembly and

206 M. Fouad Bouzidi et al. / Biochimica et Biophysica Acta 1317 (1996) 199-209

stability defect with a simultaneous decrease of cy- tochrome b, the Fe-S protein and the 9.5 kDa protein. When a complex is assembled and the Fe-S and 9.5 kDa subunit inserted, the complex therefore seems as active as the native complex, although the structure may be dis- torted, if a mutation is present. Similarly, in yeast mutants, mutations on the 11 kDa subunit (homologous to the human 9.5 kDa subunit) or on the 14 kDa subunit were associated to a coordinated decrease of the 11 kDa subunit, of the Fe-S protein and of cytochrome b [29,30]. However, in the patient mitochondria, the 14 kDa subunit was not lowered, a phenomenon that has not been observed in yeast.

The immunological data show that the cytochrome b mutation does not modify the amount of core protein I and II, cytochrome cl and the 13.4 kDa protein present in the mitochondria but does seem to impair the assembly of the cytochrome b, the Fe-S protein and the 9.5 kDa protein into a functional complex. These results are consistent with the model for the structure and assembly of bovine com- plex III proposed by Gonzalez-Halphen et al. [31]. Indeed, in this model, the 9.5 kDa protein and the iron-sulfur

protein interact with cytochrome b on the intermembrane side of the mitochondrial membrane. If the structure of human cytochrome b predicted by analogy with yeast or bovine cytochrome b is correct (Fig. 9), the mutation at position 15615 which replaces the highly conserved Gly290 [10] with Asp introduces a negative charge in the trans- membrane helix VI of the cytochrome b model, close to the outer surface of the inner mitochondrial membrane. Since Asp, in addition to being acidic, is also a more bulky residue than Gly, the normal interactions between cy- tochrome b, the Fe-S protein and the 9.5 kDa protein could be disrupted and might well explain the specific decrease in the amounts of iron-sulfur and 9.5 kDa pro- teins. This suggestion is also consistent with the model of assembly of complex III proposed by Hemrika et al. [30]: the assembly of the sub-complexes of the various subunits occurs before the binding of the 9.5 kDa protein (11 kDa in yeast) and of the Fe-S protein (the last to be added), after association of the preassembled sub-complexes to cytochrome b. The preassembled sub-complexes are resis- tant to protease digestion, which explains why they are found in normal amounts in the patient mitochondria. The

Centre P M y x o t h i a z o / T T N SL ,P YGTL VQ wl

G v ~ L A W F S S I A I Elso A H W T I F14o

S $6o "~o R M A D Q

D P s V [ ] I Y II N III W

H ~y Lp A M . W '~ I y V 6130

• 6LF " R~,Y ~,~ A A I TT Ai~l N M AT ? i LQ AG, s LA T - - L F'" L 20 / f ~ ' ~ c'° IF c' ,L I f~S LL ,L Lml~IL F IN 'G

~ G R 14o2 T A (3 S ~ S F L y'~lO I R K ~ N p LlO N M S M K p P T M-N~ L T N' Antimycin

P H

'°(DFs Centre N

intermembrane space

T IV T L

A

G GY s []Y

V D

R m P S

D P 6

V k2so k

D , FS P

L2F L L

L GL ~ D A

@Y p H Q

T

L G21o I TS HSDKI

N P L A 2so L N

T T

KI H PP

p270 L E w ~ Y FAYT 1~80

L R

~ VI PV S VI I P v S Y P FT I VI I I

o , W I Q G340 Q V

G S A • I LT L I L~L~r~

i S L L r r A D L T FYT

I , , L I ' ' w ,L' |P 'MA Isis: pMLT V M H L I Q S I p L I E N

Salo R K M K F L K

Q Q S M M W HOoc-A3Bo

mitochondrial matrix Fig. 9. Folding model of human cytochrome b. The model has been derived by analogy with models proposed for yeast or beef cytochrome b. The limits of the helices correspond to those predicted as most probable by Degli Espsoti et al. [10] from the hydropathy profiles and alignment of sequences from different species. The four histidine residues thought to be ligands to the two heine groups are boxed in gray [25]. Amino acid residues known to be the loci of mutations involved in yeast resistance to antimycin A or to myxothiazol [23-25,35,36] are surrounded by circles or by squares, respectively. Amino acids thought to interfere both with cytochrome b assembly and antimycin resistance [35] are underlined, stained in gray and put in bold face. The amino acids thought to be involved in the ubiquinone hydrophobic binding pocket are underlined and stained in gray [40]. The vertical arrows indicate the orientation of the helices in the mitochondrial inner membrane. The mutations I7T and GEa0D deduced from the patient cytochrome b gene sequence are indicated in bold and surrounded by gray squares. The amino acids D171 and V356 underlined correspond to the mutations 15257 and 15812, respectively. These mutations are polymorphisms which have been proposed to interfere with the Leber Hereditary Optic Neuropathy [7].

M. Fouad Bouzidi et al. /Biochimica et Biophysica Acta 1317 (1996) 199-209 207

4°C instability of the complex III from the patient' s sample suggests that the binding of the 9.5 kDa and (or) the Fe-S protein are important in stabilizing complex III. In addi- tion, the stability of complex III seen at 37°C in contrast to the decrease at 4°C suggests the involvement of membrane lipids. As an hypothesis, it could be suggested that lipids might be involved in the association between cytochrome b, the 9.5 kDa protein and the Fe-S protein. At 37°C, phospholipids have a much higher flexibity than at 4°C, where they are in a gel structure. Alternatively, some lipids might be essential parts of the complex III structure, as shown recently by X-ray cristallography in the case of the complex IV structure [32]. Freezing the lipid structure at 4°C might induce a weaker binding of the 9.5 kDa and of the Fe-S protein than at 37°C.

To our knowledge, this is the first report of a human cytochrome b mutation inducing a specific antimycin A resistance. Myxothiazol resistance has been previously re- ported for the platelet mitochondria of a patient with intermittent ataxia and lactic acidosis [33] and it was suggested that the illness resulted from a mutation in the quinol oxidase domain of mitochondrial cytochrome b; however, the cytochrome b mutation was not identified. Studies performed using antimycin A-resistant yeast mu- tants have generally proposed that antimycin A resistance is often associated with cytochrome b mutations located in the vicinity of center N while myxothiazol resistance is associated to mutations located in the vicinity of center P (Fig. 9). Sequencing cytochrome b gene from the patient discussed in this report [8] has revealed two differences when compared to the sequence described by Anderson et al. [15]. The first, a T14766 to C change, is a common cytochrome b polymorphism [5] which replaces Ile 7 with Thr, near the center N of the cytochrome b model, to which a number of yeast mutants resistant to antimycin A have been mapped (Fig. 9). Our data show that antimycin A sensitivity was not impaired in the mitochondria from a normal individual bearing this polymorphism and thus, the polymorphism itself cannot be responsible for the an- timycin A resistance, seen in this patient. In addition, this TL4766C mutation was homoplasmic, present in the healthy mother and sisters of the patient, while the mutation G156~sA was heteroplasmic and only found in the complex III-deficient patient's muscle [8].

Antimycin A binding has been studied in many yeast mutants. In 1983, Chevillotte-Brivet et al. [34] described several antimycin-resistant mutants carrying a missense mutation in three exon loci (box1, box4 and box6), which resulted in a total loss of both electron transfer activity and the high affinity binding site for antimycin. They sug- gested that all three portions of the polypeptide chain participate in the creation of the antimycin binding site, and that the antimycin-binding component was not cy- tochrome b itself but that an alteration in cytochrome b structure caused an effect on the antimycin-binding site. However, at that time, the exact location of the various

mutations was not known. More recently, of the six novel missense mitochondrial cytochrome b yeast mutants char- acterized by Lemesle-Meunier et al. [35], two, G340E and L282F (corresponding to G339 and L28 l, in man) have been shown to result in severely impaired stability and assembly of the bcl complex and loss of high affinity antimycin binding site. In the 8 transmembrane a helices cy- tochrome b folding model, yeast L282 (man Lz8 l) is located in the loop connecting helices V and VI and yeast G340 (man G339) near the end of helix VII, both L282 (L28 l) and G340 (G339) , facing the intermembrane space. All other cytochrome b mutations involving antimycin A resistance are generally located on the matrix side of the mitochondrial membrane (Fig. 9). Although, this is simply a model and not a structure defined by X-ray crystallogra- phy, it is consistent with the results obtained by many laboratories using different approaches (see reviews [10,23-25,36]). In this model, L282 and G340 are located in the same area of the cytochrome b structure as G290. It is shown that the G340E and L282 F mutations in yeast not only induce antimycin A resistance but also impair assem- bly of the bcl complex [35]. We here show that, in a patient beating a G290D mutation in cytochrome b, the stability of complex III is decreased at 4°C and the amount of cytochrome b, the iron-sulfur protein and the 9.5 kDa protein are decreased. This suggests that the G290D cy- tochrome b mutation impairs the normal assembly or stability of the complex, resulting in a decrease in enzyme activity and also in a defective antimycin A binding by a mechanism similar to that proposed by Lemesle-Meunier et al. [34] for the yeast G340 E and L282F mutants. In addition, the observation of van Keulen and Berden [37] showing that the 9.5 kDa subunit (yeast 1 1 kDa subunit) might be directly involved in the binding of antimycin A is also consistent with our conclusions.

Antimycin A has also been previously reported to cause a conformational change in the cytochrome b c l complex resulting in weaker binding of the Rieske iron-sulfur pro- tein [38-40]. It is therefore possible that the G290D muta- tion which loosened the interaction between cytochrome b, the iron-sulfur protein and the 9.5 kDa protein prevents antimycin A induced inhibition in this way.

According to the folding model of yeast cytochrome b described by Link et al. [23], on top of cytochrome b are two extramembranar domains, the loop between helix I and II and the helical domain between helices III and IV. Between these is a cleft large enough to accommodate both the loop between helices V and VI and parts of the iron-sulfur protein. Helix VI containing G290 appears to be close to yeast N256 (man N255), in the loop linking helices V and VI. In this model, the top of helix VI would be close to this cleft. Changing a Gly to an Asp could therefore easily hinder the normal interaction between helix VI and the iron-sulfur protein. Geier et al. [41] have shown that, in the yeast mutant N256Y, the association between the iron-sulfur protein and cytochrome b was

208 M. Fouad Bouzidi et al . / Biochimica et Biophysica Acta 1317 (1996) 199-209

decreased to such an extent that no iron-sulfur was found in the isolated bcl complex; in contrast, its level was normal in the original membrane, suggesting that the muta- tion N256Y of yeast cytochrome b weakened its interaction with the iron-sulfur protein which rendered the complex unstable during isolation. In the same study, data obtained with various mutants suggested that yeast residues V292 (equivalent to V29 ~ in human) and yeast M295 (1294) in helix VI might contribute to the generation of an hy- drophobic cleft in the binding pocket of ubiquinone, with F129 (F128) and I125 (MI24) in helix III; the latter is thought to be in close proximity to helix VI. If this model also holds for human cytochrome b, the replacement of human G290 (G291 in yeast) by a D which is a bulkier and charged group, might perturb the structure of this pocket. However, in the G290 D human mutation, no change in the apparent ubiquinone affinity was seen. In addition, these yeast mutants were myxothiazol-resistant rather than antimycin- resistant. The most likely explanation is therefore that the antimycin A resistance observed in the G290 D human mutant is not due to a direct interference between an- timycin A and G290 but to an indirect decrease in the stability of binding between cytochrome b and the iron- sulfur protein.

In conclusion, although a defect in the interaction be- tween cytochrome b and the iron-sulfur protein or the 9.5 kDa protein is an attractive hypothesis for explaining the changes seen in the properties of complex III in this patient, it will remain a hypothesis until the determination of the crystallographic structure of complex III can show if such interaction between the cytochrome b G290 and the iron-sulfur protein and/or the 9.5 kDa protein is possible. In addition, it would be interesting to create, in a species in which it is possible to carry out site directed mutagenesis, a cytochrome b mutant homologous to the G290 D human mutant to study its effects on the modifications of complex III properties and assembly.

Acknowledgements

This work was supported by grants from the 'Associa- tion Fran~aise contre les Myopathies', the 'R6gion Rh6ne-Alpes', the CNRS and the 'Universit6 Claude Bernard de LYON I'. The authors would like to thank F. Flocard (MD) for referring the patient. M.F.B. was sup- ported by the Moroccan government. Dr Sch~igger is grate- fully acknowledged for performing the identifications of the subunits reacting with our polyclonal anti-complex III antibody.

References

[1] Sch~igger, H. (t994) in A practical guide to membrane protein purification, (yon Jagow, G. and Sch~gger, H., eds.), pp. 107-124, Academic Press, New York.

[2] Hatefi, Y. (1985) Ann. Rev. Biochem. 54, 1015-1069. [3] Weiss, H. and Ziganke, B. (1974) Eur. J. Biochem. 41, 63-71. [4] Katan, M.B., Pool, L. and Groot, S.P. (1976) Eur. J. Biochem. 65,

95-105. [5] Wallace, D.C., Lott, M.T. and Brown, M.D. (1995) The report of the

committee on human mitochondrial genome 1995. In Human gene mapping 1995: a compendium, (Cutichia A.J., ed.), pp. 910-954, John Hopkins University Press, Baltimore, (also available at http://www.gen.emory.edu/mitomap.html).

[6] Boffoli, D., Scacco, S.C., Vergari, R., Persio, M.T., Solarino, G., Laforgia, R. and Papa, S. (1996) Biochim. Biophys. Acta 1315, 66-72.

[7] Brown, M.D., Voljavec, A.S., Lott, M.T., Torroni, A., Yang, C.C. and Wallace, D.C. (1992) Genetics 130, 163-173.

[8] Dumoulin, R., Saignol, I., Ferlin, T., Hulin, S., Stepien, G. and Mousson, B. (1995) J. Mol. Med. B36.

[9] Bouzidi, M.F., Sch~igger, H., Collombet, J.M., Cartier, H., Flocard, F., Quard, S., Mousson, B. and Godinot, C. (1993) Neuromnsc. Disord. 3, 599-604.

[10] Degli Esposti, M., De Vries, S., Crimi, M., Ghelli, A., Patarnello, T. and Meyer, A. (1993) Biochim. Biophys. Acta 1143, 243-271.

[11] Moradi-Ameli, M. and Godinot, C. (1983) Proc. Natl. Acad. Sci. USA 80, 616-617.

[12] Sch~igger, H., Link, T.A., Engel, W.D. and yon Jagow, G. (1986) Methods Enzymol. 126, 224-237.

[13] Sch~igger, H. and yon Jagow, G. (1987) Anal. Biochem. 166, 368-379.

[14] Matsudaira, P. (1987) J. Biol. Chem. 262, 1035-1038. [15] Anderson, S., Banquier, A.T., Barrel, B.G., de Bruijn, M.H.L.,

Coulson, A.R., Drouin, J., Eperon, I.C., Nierlich, D.P., Rose, B.A., Sanger, F., Schreier, P.H., Smith, A.J.H., Staden, R. and Young, I.G. (1981) Nature 290, 457-465.

[16] Engel, W.K. and Cuningham, G.C. (1963) Neurology 13, 919-923. [17] Watterburg, L.W. and Leong, J.L. (1960) J. Histochem. Cytochem.

8, 296-300. [18] Padykula, H.A. and Herman, E. (1955) J. Histochem. Cytochem. 3,

170-183. [19] Seligman, A.M., Kamowsky, M.J., Wasserkrug, H.L. and Hanker,

J.S. (1968) J. Cell. Biol. 38, 1-14. [20] Rao J.K.M. and Argos P. (1986) Biochim. Biophys. Acta 869,

197-214. [21] Brasseur, R. (1988) J. Biol. Chem. 263, 12571-12575. [22] di Rago J.P., Macadre, C., Lazowska, J. and Slonimski, P.P. (1993)

FEBS Lett. 328, 153-158. [23] Link, T.A., Wallmeier, H. and von Jagow, G. (1994) Biochem. Soc.

Trans. 22, 197-203. [24] Beattie, D.S., Jenkins, H.C. and Howton, M.M. (1994) Arch.

Biochem. Biophys. 312, 292-300. [25] Trumpower, B.L. and Gennis, R.B. (1994) Annu. Rev. Biochem. 63,

675 -716. [26] Herpin, P. and BarrY, H. (1989) Comp. Biochem. Physiol. 92,

59-65. [27] Taylor, R.W., Birch-Machin, M.A., Bartlett, K., Lowerson, S.A. and

Tumbull, D.M. (1994) J. Biol. Chem. 269, 3523-3528. [28] Ballinger, S.W., Shoffner, J.M. and Wallace D.C. (1994) Curr.

Topics Bioenerget. 17, 59-98. [29] Hemrika, W., Lobo-Hadju, G., Berden, J.A. and Grivell, L.A. (1994)

FEBS Lett. 344, 15-19. [30] Hemrika, W., De Jong, M., Berden, J.A. and Grivell, L.A. (1994)

Eur. J. Biochem. 220, 569-576. [31] Gonzalez-Halphen, D., Lindorfer, M.A. and Capaldi, R.A. (1988)

Biochemistry 27, 7021-7031. [32] Iwata, S, Ostermeier, C., Ludwig, B. and Michel, H. (1995) Nature

376, 660-669. [33] Parker, W.D., Freeman, F., Haas R. and Parks, J.K. (1988) Biochim.

Biophys. Acta 936, 133-138.

M. Fouad Bouzidi e t al. / Biochimica et Biophysica Acta 1317 (1996) 199-209 209

[34] Chevillotte-Brivet, P., Lemesle-Meunier, D., Forget, N. and Pajot, P. (1983) Eur. J. Biochem. 129, 653-661.

[35] Lemesle-Meunier, D., Chevillotte-Brivet, P., di Rago, J.P., Slonim- ski, P.P., Bruel, C., Tron, T. and Forget, N. (1993) J. Biol. Chem. 268, 15626-15632.

[36] Colson, A.M. (1993)J. Bioenerg. Biomemb. 25, 211-220. [37] van Keulen, M.A. and Berden, J.A. (1984) Biochim. Biophys. Acta

808, 32-38 [38] Rieske J.S., Baum, H., Stoner, C.D. and Lipton, S.H.J. (1967) J.

Biol. Chem. 242, 4854-4866.

[39] Rieske, J.S. (1976) Biochim. Biophys. Acta 639, 129-155. [40] Graham, L.A., Brandt, U., Sargent, J.S. and Trurnpower, B.L.

(1992) J. Bioenerg. Biomemb. 25, 245-257. [41] Geier, B.M., Haase, U. and von Jagow, G. (1994) Biochem. Soc.

Trans. 22, 203-209. [42] Copperstein, S.J. and Lazarow, A. (1951) J. Biol. Chem. 189,

665 -669.