Role of the evolutionarily conserved cytochrome b tryptophan 142 in the ubiquinol oxidation catalyzed by the bc1 complex in the yeast Saccharomyces cerevisiae.

Trp-142 is a highly conserved residue of the cytochrome b subunit in the bc1 complexes. To study the importance of this residue in the quinol oxidation catalyzed by the bc1 complex, we characterized four yeast mutants with arginine, lysine, threonine, and serine at position 142. The mutant W142R was isolated previously as a respiration-deficient mutant unable to grow on nonfermentable carbon sources (Lemesle-Meunier, D., Brivet-Chevillotte, P., di Rago, J.-P, Slonimski, P. P., Bruel, C., Tron, T., and Forget, N.(1993) J. Biol. Chem. 268, 15626-15632). The mutants W142K, W142T, and W142S were obtained here as respiration-sufficient revertants from mutant W142R. Mutant W142R exhibited a decreased complex II turnover both in the presence and absence of antimycin A; this suggests that the structural effect of W142R in the bc1 complex probably interferes with the correct assembly of the succinate-ubiquinone reductase complex. The mutations resulted in a parallel decrease in turnover number and apparent Km, with the result that there was no significant change in the second-order rate constant for ubiquinol oxidation. Mutants W142K and W142T exhibited some resistance toward myxothiazol, whereas mutant W142R showed increased sensitivity. The cytochrome cc1 reduction kinetics were found to be severely affected in mutants W142R, W142K, and W142T. The respiratory activities and the amounts of reduced cytochrome b measured during steady state suggest that the W142S mutation also modified the quinol-cytochrome c1 electron transfer pathway. The cytochrome b reduction kinetics through center P were affected when Trp-142 was replaced with arginine or lysine, but not when it was replaced with threonine or serine. Of the four amino acids tested at position 142, only arginine resulted in a decrease in cytochrome b reduction through center N. These findings are discussed in terms of the structure and function of the quinol oxidation site and seem to indicate that Trp-142 is not critical to the kinetic interaction of ubiquinol with the reductase, but plays an important role in the electron transfer reactions that intervene between ubiquinol oxidation and cytochrome c1 reduction.

The ubiquinol-cytochrome c oxidoreductase (the cytochrome bc 1 complex; EC 1.10.2.2) is an integral multisubunit membrane protein that is involved in energy transduction in a wide range of organisms. All the bc 1 complexes contain a minimum of three redox-active polypeptide subunits carrying four prosthetic groups: a[2Fe-2S]iron-sulfur protein, a one-heme center cytochrome c 1 , and a two-heme center apocytochrome b, cytochrome b L , and cytochrome b H (1,2). This complex catalyzes the electron transfer from ubiquinol to cytochrome c coupled to a vectorial proton translocation through the membrane involving a mechanism known as the modified Q cycle (3)(4)(5)(6)(7). According to this mechanism, the bc 1 complex transfers two electrons from ubiquinol to two molecules of cytochrome c with a concomitant ejection of two protons on the positive side of the membrane, while two additional protons are translocated from the negative to the positive side of the membrane per pair of electrons transferred. This mechanism requires two quinone separate independent reaction domains: the ubiquinol oxidation domain (QP center) located on the positive side of the membrane, which is the target of three groups of specific inhibitors, especially the methoxyacrylate-type inhibitor exemplified by myxothiazol (for review, see Ref. 8), and the ubiquinone reduction domain (QN center) located on the negative side of the membrane, which is the target of other specific inhibitors exemplified by antimycin A (for review, see Ref. 8). The QP center is associated with two proteins, the [2Fe-2S]iron-sulfur protein and cytochrome b through heme b L . Electrons from ubiquinol diverge at center P: one is transferred to the ironsulfur protein, along with the production of an unstable semiquinone, and subsequently to cytochrome c 1 and cytochrome c; the second electron is transferred to heme b L and thereafter to heme b H and is recycled through a quinone molecule to form a stable semiquinone anion at the quinone reductase domain. The oxidation of a second molecule of ubiquinol at center P is required to reduce this stable semiquinone into ubiquinol at center N. Another essential feature of the Q cycle mechanism is that cytochrome b can be reduced through two pathways, the center P pathway, which is the more thermodynamically favorable, and the center N pathway when the thermodynamically favorable route through center P is blocked.
The assumption that a highly conserved residue performs a vital functional or structural role in the protein is often used in site-directed mutagenesis strategies. Recent results have shown, however, that only a few of the evolutionarily conserved amino acids seem to be essential for bc 1 function (34). We report here on the mapping, sequencing, and biochemical study of four mutants: the inactive cytochrome b W142R mutant (26), the mutation of which affects an evolutionarily conserved residue, and three functional revertants isolated from this mutant. According to degli Esposti et al. (38), tryptophan at position 142 is conserved among the 800 cytochrome b sequences known to date, except in Paramecium aurelia and Bacillus PS3. Moreover, this position belongs to one of the most conserved regions of cytochrome b, spanning residues 131-148, which may interact with the QP center. The mutants/revertants were investigated regarding their effects on growth, various electron transfer activities, kinetic interaction between the substrate and the enzyme, cytochrome c 1 reduction, cytochrome b reduction through both centers P and N, and resistance to inhibitors. The results suggest that Trp-142 is important to the properties of the quinol oxidation site, especially to the electron transfer reactions that occur between quinol oxidation and cytochrome c 1 reduction.

EXPERIMENTAL PROCEDURES
Strains-The parental control diploid strain KM91 was obtained by crossing the haploid strain 777-3A (␣, ade1 Ϫ , op1 Ϫ , rho ϩ , mit ϩ ) with the haploid KL14-4A/60 (a, his1 Ϫ , trp2 Ϫ , OP1 ϩ , rho 0 ). To select for point mutations in mitochondrial DNA and at the same time to counter-select large deletions in this genome, the op1 Ϫ method was used (39). It is based on the following rationale. 1) the op1 Ϫ nuclear mutation (40) abolishes the activity of the main ADP/ATP translocator of the mitochondrial membrane (the AAC2 translocator). 2) In the presence of the op1 Ϫ mutation, the rho Ϫ /rho 0 mutations, which result from large deletions of mitochondrial DNA (41), lead to an immediate cell lethality even on glucose media (42) and therefore are counter-selected. 3) This cellular lethality is due to a complete loss of mitochondrial protein synthesis in rho Ϫ /rho 0 cells, which are always deficient in one or several elements of this machinery (rRNAs or tRNAs) and therefore cannot synthesize subunits 6 and 9 of the ATP synthase complex, while point mutations in cytochrome b or cytochrome oxidase genes do not abolish the synthesis of ATP synthase subunits and are viable on glucosecontaining media. 4) Therefore, mutagenesis of the op1 Ϫ rho ϩ , mit ϩ cells gives only op1 Ϫ rho ϩ survivors, with the op1 Ϫ rho Ϫ /rho 0 cells being dead. 5) To select among the survivors, the point mutations in the mitochondrial genome, a test cross is performed. The survivors (haploid) are crossed with the OP1 ϩ rho 0 haploids, and the diploids are selected on glucose media and tested for respiratory sufficiency on glycerol/ethanol media. The majority of the diploid clones are respiration-sufficient since the op1 Ϫ mutation is complemented by the wildtype allele in OP1 ϩ /op1 Ϫ heterozygous cells, and they do not carry any mitochondrial mutation, while only a minority carries a mitochondrial mit Ϫ mutation, which cannot be complemented by the OP1 ϩ rho 0 tester. Such haploid op1 Ϫ mit Ϫ clones are isolated (the first list of mutants is given in Ref. 39) and used for further studies. In this manner, the mutant 777-3A/V384 was isolated from the haploid strain 777-3A (␣, ade1 Ϫ , op1 Ϫ , rho ϩ , mit ϩ ) and then crossed with the strain KL14-4A/60 (a, his1 Ϫ , trp2 Ϫ , OP1 ϩ , rho 0 ), and the resulting diploid KM772 (␣/a, op1 Ϫ /OP1 ϩ , ade1 Ϫ /ADE1 ϩ , HIS1 ϩ /his1 Ϫ , TPR2 ϩ /trp2 Ϫ , rho ϩ , cob-V384) was constructed. It has the AAC2 ADP/ATP translocator, and its mitochondrial genome is homoplasmic for the W142R mutation in the cytochrome b gene.
Media, Growth Conditions, and Preparation of Mitochondria and "Crude bc 1 "-Media, growth conditions, and preparation of mitochondria were as described (41). Crude bc 1 complex was prepared as described (29).
Enzymatic Assays-Succinate oxidase, NADH oxidase, EtOH oxidase, and DBH 2 1 -cytochrome c oxidoreductase activities were meas-ured as described previously (43,44). The reduction of DB by NADH was measured by NADH oxidation monitored at 340 -425 nm; the reduction of DB by succinate was followed at 280 -325 nm in 10 mM phosphate buffer, pH 7.4, 2 mM KCN. Absorption coefficients of 6.2 and 16 mM Ϫ1 cm Ϫ1 for NADH and DB, respectively, were used for calculations. Reduction Kinetics of Cytochromes-Mitochondrial membranes were diluted to a concentration of 0.5 M cytochrome b in 0.65 M sorbitol, 10 mM KH 2 PO 4 , 2 mM EDTA, 0.1 mM MgCl 2 buffer, 0.3% bovine serum albumin, pH 6.5. Antimycin and myxothiazol were prepared in ethanol and added to final concentrations of 20 and 30 M, respectively. KCN was added to a final concentration of 2 mM. Succinate was added to a final concentration of 40 mM. Reduction kinetics were recorded on an SLM-AMINCO DW2 spectrophotometer equipped with a magnetic stirrer.
Inhibitor Titration-The myxothiazol I 50 values measured in the W142K, W142T, and W142S revertants were determined as described (22). I 50 represents the concentration of inhibitor that decreases the rate of DBH 2 -cytochrome c oxidoreductase by 50%. The action of myxothiazol on the mitochondria isolated from the W142R mutant was measured by inhibiting the cytochrome cc 1 reduction kinetics recorded at 551-540 nm. The initial slope of the reaction was plotted against the inhibitor concentration, and the concentration that led to a 50% decrease in kinetics was referred to as the I 50 value.
Measurement of the Complex II Cytochrome b Content-The amount of cytochrome b was determined spectrally. Cytochrome b from complex II has the property of being oxidized by fumarate even in the presence of dithionite (26,45). The dithionite-reduced minus (dithionite ϩ fumarate)-reduced absorbance spectra showed a peak centered at 562 nm, corresponding to the complex II cytochrome b. An absorption coefficient of 24 mM Ϫ1 cm Ϫ1 was used to calculate its concentration in each strain.
Codon 142 belongs to a very short exon (B2) of only 14 base pairs (46). Part of this exon, including codon 142, is involved in base pairings with intron bi2, which are crucial to the splicing of this intron (see Fig. 5 in Ref. 47). Not surprisingly, the mutants described above accumulated significant amounts of nonprocessed cytochrome b transcripts.
Two other amino acids, Gly-142 (GGA) and Met-142 (ATA), could have been obtained by monosubstitution at the Arg-142 codon (AGA). Given the fact that they have not been found among the high number of revertants analyzed, it seems likely that the corresponding nucleotide changes would result in a respiration-negative phenotype. However, it is difficult to ascertain whether Gly-142 and Met-142 are not compatible with the function of cytochrome b or whether the corresponding nucleotide changes impair splicing of cytochrome b pre-mRNA. Also, the serine codon AGC could have been derived by a single nucleotide change from the Arg-142 codon. Since the presence of serine at position 142 gives a functional cytochrome b, one could infer that this codon blocks the splicing of intron bi2. However, this codon is normally not used by yeast mitochondria for the synthesis of respiratory and phosphorylating complexes (48). It is therefore not surprising that it has not been found among the revertants we analyzed.
Cytochrome Spectral Analysis-The absorbance spectra of mitochondria isolated from the parental and W142R, W142K, W142T, and W142S mutant strains are shown in Fig. 1. The ascorbate minus ferricyanide spectrum (bottom trace in each panel) shows the high potential c-type cytochrome (c ϩ c 1 ) reduction. The succinate minus ascorbate spectrum (middle traces in panels A and C-E) shows the reduction of the high potential heme b H . The succinate minus ferricyanide spectrum (middle trace in panel B) shows the reduction of heme b H ; indeed, in the presence of succinate, neither cytochrome cc 1 nor cytochrome oxidase is reduced in mutant W142R, which indicates that no electron transfer occurs between cytochrome b and cytochrome cc 1 in this mutant. Adding a few grains of dithionite to the sample reduced the low potential heme b L and the complex II cytochrome b, which is totally oxidized by a subsequent addition of fumarate (45). Thus, the dithionite plus fumarate minus succinate spectrum (top trace in each panel) shows the low potential heme b L reduction.
A decrease in the synthesis of all the cytochromes is usually observed in respiration-deficient strains, which use only ATP produced by fermentation for their growth. In some mutants, this decrease is in approximately the same proportion as that observed in their counterparts in the parental strain (22). In the results presented here, mutant W142R exhibits a greater decrease in the reduced heme b L than in the reduced heme b H or cytochrome cc 1 . All the revertants recovered a level of cytochrome synthesis similar to that in the parental strain (80 -100%).
Strain Growth, Electron Transfer Activities, and Inhibitor Titration-Mutant W142R did not grow on respiratory substrates and exhibited no quinol-cytochrome c reductase activity (Table I) and no oxygen consumption when ethanol, NADH, or succinate was used as substrate. Revertants W142K, W142T, and W142S recovered 12, 46, and 83% of the parental bc 1 complex activity, respectively (Table I); a similar recovery was observed in their various oxidase activities. All the mutant strains exhibited an NADH-DB reductase activity identical to that of the parental strain, whereas the succinate-DB oxidoreductase activity was severely affected in mutant W142R and only slightly modified (if at all) in the revertant strains. Revertant W142S grew on nonfermentable carbon sources with the same characteristics as the parental strain, whereas revertants W142K and W142T exhibited very low growth rates under these growth conditions: 1.2 and 4.5% of the parental counterpart, respectively. DBH 2 -cytochrome c reductase activity was titrated with myxothiazol and antimycin A in the parental and revertant strains. In the presence of myxothiazol, the I 50 values showed FIG. 1. Absorption difference spectra of mitochondria from the parental strain (KM91), mutant W142R, and revertants W142S, W142T, and W142K. Mitochondria were isolated from yeast cells grown on galactose medium and suspended to 2.5 mg/ml in 0.65 M sorbitol, 10 mM KH 2 PO 4 , 2 mM EDTA, 0.1 mM MgCl 2 buffer, 0.3% bovine serum albumin, pH 6.5. In each panel, the bottom trace shows the cytochrome c ϩ c 1 reduction induced by adding 40 mM ascorbate to the sample cuvette. The spectra were recorded after complete exhaustion of oxygen; the reference cuvette was maintained in the oxidized state by adding 10 M ferricyanide. In A and C-E, the middle trace shows the reduction of heme b H observed after adding 40 mM succinate to the sample cuvette and 40 mM ascorbate to the reference cuvette (succinate minus ascorbate spectrum). In B, the succinate minus ferricyanide spectrum shows the reduction of heme b H (see "Results" for explanations). The top traces show the reduction of heme b L observed after adding a few grains of dithionite and 40 mM fumarate to the sample cuvette and 40 mM succinate to the reference cuvette. that 10-and 5-fold increases in resistance toward this inhibitor occurred with revertants W142K and W142T, respectively, whereas no change in myxothiazol sensitivity was observed with revertant W142S. Mutant W142R did not exhibit any DBH 2 -cytochrome c reductase activity, but exhibited cytochrome c 1 reduction during a single turnover of the bc 1 complex (see "Reduction Kinetics of the bc 1 Complex Cytochomes"). This finding was used to assay the sensitivity of this mutant toward myxothiazol; a 4-fold higher sensitivity was observed in comparison with the parental strain. The sensitivity toward antimycin was not modified by mutation W142K, W142T, or W142S (data not shown), and it was previously observed that mutation W142R does not affect the antimycin sensitivity either (49).
Catalytic Properties of the Reductase-The bc 1 complex steady-state electron transfer mechanism can be described as a ping-pong two-site scheme in which ubiquinol and cytochrome c interact with two catalytically independent sites (50,51). The experimental V max /K m ϭ k min value (with V max expressed as the turnover number) gives the apparent second-order rate constant characteristic of the arrival and binding of the substrate to its site. This is valid as long as these steps are not rate-limiting, which is the case in the strains studied here since they all exhibited Michaelis-type kinetic behavior. The results given in Table II indicate that all the revertants exhibit similar close k min values with DBH 2 (variations inferior to an order of magnitude) compared with those of the parental strain, indicating that the kinetic interaction of ubiquinol with the reductase is slightly or not affected by the mutations at position 142.
Reduction Kinetics of the bc 1 Complex Cytochromes-The effects of the various mutations on the cytochrome cc 1 and cytochrome b electron transfer pathways were determined by studying the kinetics of the substrate-induced reduction at wavelength pairs of 551-540 and 562-575 nm, respectively. These kinetics give a global value including 1) the rate of arrival and binding of the substrate to its site and 2) the rate of the various electron transfer steps occurring between this binding site and cytochrome cc 1 or b. In the mutant strain, as in the revertants, the kinetics, especially those of cytochrome b reduction through center N, were found to be slower when DBH 2 was used as substrate than with succinate or NADH. This probably indicates that the natural Q 6 -type quinone, which is involved when succinate or NADH is used, was a more efficient electron donor than DBH 2 . In all our mutants/revertants, the NADHquinone reductase activity was not found to differ from that of the parental strain, whereas the complex II activity was found to be severely affected in the W142R mutant (Table I). The kinetics were recorded using NADH (data not shown) and succinate (see below) as substrates; even with the W142R mutant, the same relative results were obtained with these two substrates when the mutant/revertant strains were compared with the parental strain. These results indicate that even the low W142R complex II activity was not rate-limiting in the various bc 1 complex electron pathway kinetic reactions we measured.
Reduction of cytochrome cc 1 by succinate in the presence of KCN was carried out in mitochondrial membranes isolated from the mutant, revertant, and parental strains (Fig. 2). The results reveal that mutation W142R leads to a strongly negative effect on the cytochrome cc 1 reduction kinetics, which is partially corrected by the W142K reversion and almost or completely corrected by the W142T and W142S reversions. Control tests with crude bc 1 complexes from W142R and W142K and the parental strains were carried out to determine cytochrome c 1 reduction using DBH 2 as substrate; in comparison with the parental strain, the cytochrome c 1 reduction rates exhibited the same relative decrease as that observed in mitochondrial membranes (data not shown).
Cytochrome b was reduced through center P in the presence of antimycin and through center N in the presence of myxothiazol in the parental, mutant, and revertant strains (Fig. 3). Mutations W142R and W142K both affected cytochrome b reduction through center P, but the effect of mutation W142R was stronger than that of W142K; W142T and W142S had no detectable effects on cytochrome b reduction through center P (Fig. 3a). Although Trp-142 is located at center P, mutation W142R also had an effect on the kinetics of cytochrome b reduction through center N (Fig. 3b). W142K, W142T, and W142S had no detectable effect on cytochrome b reduction through center N.
The cytochrome b reduction kinetics were also recorded in the absence of inhibitor in the mutant, revertant, and parental strains. Adding succinate to the parental strain mitochondrial membranes led, at steady state, to a 18% reduction of the anaerobiosis-reduced cytochrome b. In the W142S, W142T, and

TABLE II Catalytic properties of the ubiquinol-cytochrome c reductase
The DBH 2 -cytochrome c reductase activity was measured in mitochondria isolated from the parental wild-type strain (KM91), mutant strain W142R, and revertants W142K, W142T, and W142S. The V max values are expressed in turnover numbers (moles of cytochrome c reduced per mole of cytochrome b/second). The experimental k min (V max / K m ) value corresponds to the apparent kinetic second-order constant of the arrival and binding of the substrate to its site. W142K mutant strains, 23, 33, and 58% of the anaerobiosisreduced cytochrome b were reduced at steady state, respectively, after the addition of succinate; in the W142R mutant, cytochrome b was fully reduced under these conditions. Mutant W142R exhibited slow cytochrome b reduction kinetics, similar to those recorded in the presence of myxothiazol (data not shown).

DISCUSSION
In this study, four new cytochrome b structures were characterized and used to analyze the functional and structural role of the evolutionarily invariant tryptophan 142 in the QP region of the bc 1 complex. This was done by isolating three functional revertants from a nonfunctional cytochrome b mutant; the initial mutation, W142R, is located in the extramembranous loop between transmembrane ␣-helices 3 and 4 on the positive side of the membrane (Fig. 4) (26). This loop, together with the C-terminal region of helix 3, might be involved in the center P catalytic subdomain (1,36). The secondary mutations, which occur at the level of the original mutation, change Arg-142 into lysine, threonine, and serine.
The cytochrome b W142R mutation, which leads to a respiration-deficient phenotype, has been shown to abolish the bc 1 complex electron transfer activity. One rather unexpected consequence of this mutation is the low succinate-quinone oxidoreductase activity observed in the corresponding mutant strain (Table I). When normalized either to the amount of protein (milligrams) or to the amount of complex II cytochrome b, this activity is severely affected, which indicates that electron transfer through complex II is modified as a result of mutation W142R. This is the first time, as far as we know, that a point mutation in the bc 1 complex cytochrome b, which does not impair the subunit composition of the complex (52), has been found to impair the complex II specific activity. One possible explanation for our results might be that the structural effect induced by mutation W142R in the bc 1 complex interfered with the correct assembly of the succinate-ubiquinone reductase complex. It is noteworthy that other cytochrome b mutated residues similarly affect both complex III and complex II specific activities. 2 Replacing tryptophan at position 142 with lysine, threonine, or serine does not affect the kinetic interaction of ubiquinol with the reductase (Table II). Tryptophan is a large, aromatic, and polar but neutral residue. Depending on the environment, it can either contribute to the hydrophobicity of a protein region or act as a hydrogen bond donor through its nitrogen free doublet. In the reaction center, His-215 (in Rhodopseudomonas viridis) and Thr-222 (in Rhodopseudomonas spheroides) are hydrogen-bonded to one carbonyl of the QA molecule (53,54). As noted by degli Esposti et al. (38), there exists a relevant homology between the conserved yeast cytochrome b peptide W 142 GATV and the conserved peptide H 215 GATV of the chloroplast D2 subunit or its homologous bacterial reaction center M subunit. By analogy with what occurs in the reaction center, degli Esposti et al. (38) suggested that in yeast cytochrome b, either Trp-142 or Thr-145 might be hydrogen-bonded to ubiquinol. The polar characteristic of tryptophan would then be important to the binding of quinol. Our results show that the Trp-142 substitutions observed in the revertant strains introduced polar residues that were liable to exhibit a hydrogen bond and did not affect the kinetic interaction of ubiquinol with the reductase. Two theoretically possible reversions from codon AGA (which codes for arginine in the mutant) that may code for the two nonpolar residues, methionine and glycine, were never selected. Whether this is due to the incompatibility of those codons with the pre-mRNA splicing or to the incompatibility of a nonpolar residue at position 142 with a functional bc 1 complex is unknown. Trp-142 is conserved among all the 800 species sequenced to date, except in Bacillus PS3, where it is replaced with the nonpolar aromatic residue phenylalanine, and in P. aurelia, where it is replaced with isoleucine. It will be necessary to obtain mutants W142G, W142I, W142M, and/or W142F before we can draw any conclusion about the interference of the polar character of residue 142 with the kinetic interaction of ubiquinol with the reductase. This can be envisaged by isolating mutants from the intronless cytochrome b gene strain (30).
The other characteristic of tryptophan, namely its aromatic structure, does not seem to be critical to the kinetic interaction of quinol with the reductase since none of the mutants studied have an aromatic residue at position 142 and all exhibit k min values that are similar to those of the parental strain. Thus, neither the size nor the shape of residue 142 seems to affect the kinetic interaction of quinol with the reductase, contrary to what occurs in Rhodopseudomonas capsulatus with Gly-143 (yeast numbering) (35). Indeed, replacing Trp-142 with amino acids of various sizes (arginine, lysine, threonine, and serine) did not affect the k min value. On the whole, Trp-142 does not seem to be involved in the kinetic interaction of quinol with the reductase. Thus, it is conceivable that this amino acid is not critical to the binding of ubiquinol, although it is highly conserved.
The overall data obtained 1) by characterizing mutants selected for their center P inhibitor resistance from both mitochondrial (15,55) and bacterial (36) systems and 2) by performing inhibitor binding studies on cytochrome b-deficient mutants and their revertants (22,56,57) and on mutants obtained by carrying out site-directed mutagenesis (32,35) indicate that two topologically different conserved regions of the protein are involved in the QP inhibitor pocket. The first region includes the C-terminal region of helix 3 and the adjacent part of loops 3 and 4; the second region is formed by loops 5 and 6 and the N-terminal region of helix 6. Myxothiazol  (17), 129 (17,15), 132 (58), 133 (56), 137 (15,17,22), and 143 (35). In the present study, Trp-142 emerges as a new residue involved in the interaction of cytochrome b with myxothiazol. Moreover, depending on which amino acid occupies position 142, the strain exhibits either resistance or greater sensitivity toward this inhibitor, suggesting that the various amino acids induce different conformational states of the QP pocket, leading to a lesser or greater binding of myxothiazol. Studies on cytochrome b inhibitor resistance mutants of various origins have provided a great deal of information about the residues involved in the binding of the various inhibitors acting at centers P and N. Two types of amino acid linked to the inhibitor-binding pockets cannot, however, be found by selection of inhibitor resistance mutants: those with which a mutation would lead to a partial or total loss of bc 1 complex function and those with which the amino acid change would lead to a greater sensitivity toward inhibitors. The results presented here and elsewhere (22,26,57) show that functional analysis of cytochrome b-deficient mutants and their revertants can serve to pinpoint these amino acids.
The cytochrome cc 1 reduction kinetics were found to be severely to slightly affected by mutations W142R, W142K, and W142T (Fig. 2); on the other hand, the same kinetics were observed in both the mutant W142S and the parental strain. Measurement of the reduction kinetics of cytochromes cc 1 and b is limited by the lag time of the apparatus, which is of the same order of magnitude as the reduction rate obtained with the parental strain. We cannot therefore rule out the possibility that mutation W142S may have an effect on the cytochrome cc 1 reduction kinetics that would not be measurable under our experimental conditions. The fact that the cytochrome cc 1 reduction kinetics are modified by at least three different residues at position 142 (arginine, lysine, and threonine) shows that Trp-142 plays an important role in the electron transfer reactions that intervene between the ubiquinol oxidation site and cytochrome c 1 .
Electron transfer through center N was recorded in the presence of myxothiazol and was found to be the same with all the strains, except for the deficient mutant W142R (Fig. 3a). Among the four mutations we studied, only arginine causes this phenotype, indicating that a long distance effect is possible, but probably requires a strong structural modification. It should be noted, however, that mutant W142R has kept all the bc 1 complex subunits (52).
Arginine or lysine at position 142 affects the kinetics of cytochrome b reduction through center P, whereas threonine or serine does not interfere with this electron transfer pathway (within the limits of our experimental conditions) (Fig. 3b). This seems to indicate that tryptophan 142 present in the wild-type cytochrome b sequence is not involved in this cytochrome b reduction pathway. Mutation W142R impairs the cytochrome b reduction kinetics through both centers P and N, but the effect on center P is stronger. This result is consistent with the center P location of amino acid 142.
The amount of cytochrome b reduced by succinate during steady state in the absence of inhibitor, together with the results obtained in the cytochrome cc 1 and cytochrome b reduction kinetic experiments (Figs. 2 and 3), provides some helpful information about Trp-142 and the mutant phenotypes. Since mutation W142R abolishes respiratory activity, it is surprising that none of the electron transfer pathways are abolished in this mutant. At steady state, W142R cytochrome b is entirely reduced by succinate, which means that the bc 1 complex can oxidize ubiquinol at least once, but that under multiple turnovers, the electrons may be blocked. In the catalytic switch model (59), this would mean that the mutant bc 1 complex switches to the b state and is locked in this conformation; the switch cannot be triggered back to the FeS state. W142K reversion improves the cytochrome cc 1 and cytochrome b reduction kinetics as compared with W142R; the bc 1 complex recovers 12% activity, and cytochrome b is 58% reduced during steady state. W142T reversion brings the cytochrome b reduction kinetics back to the parental value and improves cytochrome c 1 reduction as compared with W142K. Even if only the cytochrome cc 1 reduction step is affected by the W142T mutation, the amount of reduced cytochrome b present during steady state is 15% higher than in the parental strain. Finally, based on the cytochrome cc 1 and cytochrome b reduction kinetics, the W142S reversion brings both the cytochrome cc 1 and cytochrome b reduction kinetics back to the parental value with, however, a 5% more reduced cytochrome b during steady state and 20% less bc 1 activity observed than in the parental strain. Mutation W142S therefore probably impairs the cytochrome cc 1 reduction kinetics (to an extent that was not detectable under our experimental conditions). Reversions W142K, Shown is the predicted diagram of the eight-membrane-spanning ␣-helix cytochrome b (10, 11,30). The four conserved histidines, thought to be the ligands to the two heme groups, are numbered and boxed. Hemes b L and b H are predicted to be coordinated to histidines 82 and 183 and to histidines 96 and 197, respectively. The shaded region spanning residues 131-148 is the longest fragment, exhibiting an average 70% identity to the other species. In this region, amino acid residues that are loci of mutations conferring resistance toward myxothiazol and/or modifying the bc 1 complex activity are indicated with bold circles. Tryptophan at position 142 is located on the positive side of the inner mitochondrial membrane and is indicated with a black circle.
W142T, and W142S gradually correct the effect of arginine on the bc 1 complex activity; they may induce conformational state dynamics at center P, which may be intermediate between the mutant b state-locked conformation and the parental b state/ FeS state fast switching conformation.
The quinol-cytochrome c oxidoreductase complex activity depends on the quinol affinity for its site and on the kinetic parameters of the intra-and intermolecular electron transfer. The protein space located between two oxidoreduction centers is generally referred to as the intervening space. Most of the mutations that confer resistance to inhibitors and any mutation that modifies the bc 1 electron transfer activity will affect an amino acid located at the quinol-binding site or in the intervening spaces or both. The quinol-binding site, the inhibitor-binding sites, and the intervening spaces are different functional areas that may have common topological domains.
Glycine at position 143 (yeast numbering) has been found to be involved in the interaction of the quinol/quinone couple with center P and in myxothiazol binding (35). Cysteine 133 and glycine 137 are critical to the interaction of quinol with center P and are involved in the intervening spaces between the quinol oxidation site and cytochrome c 1 (56). 3 Glycine 137 and, to a lesser extent, cysteine 133 are also involved in myxothiazol binding (22,56). Four respiration-deficient yeast mutants with impairments at positions 131, 133, 137, and 142 (26) of the cytochrome b sequence have been isolated, and a mutation affecting the cytochrome b from R. capsulatus at position 158 (homologous to position 143 in yeast) led to a nonphotosynthetic phenotype (35). All these residues belong to the longest region of cytochrome b (spanning amino acids 131-148), with an average 70% identity to the other known cytochrome b sequences (calculated from the sequence alignment presented in Ref. 38) (Fig. 4). Four residues also located in this region (at positions 132, 133, 137, and 143) are thought to interfere with the center P inhibitors. This region is therefore likely to have an important role in the structure and function of center P.
Our results suggest that tryptophan 142 is involved in the myxothiazol-binding site, but probably not in the interaction of quinol with center P. It would belong to the intervening space between the quinol-binding site and cytochrome c 1 since its mutation to arginine, lysine, threonine, and probably serine leads to a decrease in the electron transfer rate between ubiquinol and cytochrome c 1 . Trp-142 could either be an essential component of the structural architecture of one of the center P intervening spaces or be directly involved in one of the electron transfer reactions that occur at center P.