The Smallest Membrane Anchoring Subunit (QPs3) of Bovine Heart Mitochondrial Succinate-Ubiquinone Reductase

The cDNA encoding the smallest membrane-anchoring subunit (QPs3) of bovine heart mitochondrial succinate-ubiquinone reductase was cloned and sequenced. This cDNA is 1330 base pairs long with an open reading frame of 474 base pairs that encodes the 103 amino acid residues of mature QPs3 and a 55-amino acid residue presequence. The cDNA insert has an 820-base pair long 3′-untranslated region, including a poly(A) tail. The molecular mass of QPs3, deduced from the nucleotide sequence, is 10,989 Da. QPs3 is a very hydrophobic protein; the hydropathy plot of the amino acid sequence reveals three transmembrane helices. Previous photoaffinity labeling studies of succinate-ubiquinone reductase, using 3-azido-2-methyl-5-methoxy[3H]-6-decyl-1,4-benzoquinone ([3H]azido-Q), identified QPs3 as one of the putative Q-binding proteins in this reductase. An azido-Q-linked peptide with a retention time of 66 min is obtained by high performance liquid chromatography of the chymotrypsin digest of carboxymethylated and succinylated [3H]azido-Q-labeled QPs3 purified from labeled succinate-ubiquinone reductase by a procedure involving phenyl-Sepharose 4B column chromatography, preparative SDS-polyacrylamide gel electrophoresis, and acetone precipitation. The amino acid sequence of this peptide is NH2-L-N-P-C-S-A-M-D-Y-COOH, corresponding to residues 29–37. The structure of QPs3 in the inner mitochondrial membrane is proposed based on the hydropathy profile of the amino acid sequence, on the predicted tendencies to form α-helices and β-sheets, and on immunobinding of Fab′ fragmenthorseradish peroxidase conjugates prepared from antibodies against two synthetic peptides, corresponding to the NH2 terminus region and the loop connecting helices 2 and 3 of QPs3, in mitoplasts and submitochondrial particles. The ubiquinone-binding domain in the proposed model of QPs3 is probably located at the end of transmembrane helix 1 toward the C-side of the mitochondrial inner membrane.

QPs2 or QPs3. However, identification of the Q-binding domains in QPs2 or QPs3 requires knowledge of the amino acid sequences of these two subunits. Herein, we report cloning and nucleotide sequencing the cDNA encoding QPs3, the immunological determination of the topology of QPs3 in the inner mitochondrial membrane using monospecific polyclonal antibodies against two synthetic peptides corresponding to residues 1-14 and 55-66, isolating and sequencing an azido-Qlinked peptide from labeled QPs3, and the localization of the Q-binding domain in the proposed model of QPs3. The identity of QPs2 is also discussed.

EXPERIMENTAL PROCEDURES
Materials-Bovine serum albumin, dichlorophenolindophenol (DCPIP), Triton X-100, sodium cholate, TTFA, and phenyl-Sepharose CL-4B, were obtained from Sigma. Protein A-horseradish peroxidase conjugate, protein molecular weight standards, SDS, acrylamide, urea, and DEAE Affi-Gel blue were from Bio-Rad. Taq polymerase was from Promega. TA cloning kit was from Invitrogen. The bovine heart cDNA library constructed in the Uni-ZAP XR vector was from Stratagene. Insta-gel liquid scintillation mixture was from Packard Instrument Co. Oligonucleotides and peptides were synthesized by the DNA/Protein Core Facility at Oklahoma State University. Nitrocellulose membranes were from Schleicher & Schuell. All other chemicals were of the highest purity commercially available.
Enzyme Preparation and Assays-Intact bovine heart mitochondria (19), mitoplasts (20), submitochondrial particles (21), and succinate-Q reductase (22) were prepared and assayed as previously reported. Succinate-Q reductase was assayed, at room temperature, for its ability to catalyze TTFA-sensitive Q-stimulated DCPIP reduction by succinate using a Shimadzu UV-2101PC. The reaction mixture (1 ml) contains 40 mol of DCPIP, 100 mol of sodium potassium phosphate buffer, pH 7.4, 20 mol of succinate, 10 nmol of EDTA, 25 nmol of Q 0 C 10 , and 0.01% of Triton X-100. The reduction of DCPIP was followed by measuring the absorption decrease at 600 nm, using a millimolar extinction coefficient of 21 mmol Ϫ1 cm Ϫ1 . The concentration of TTFA used was 10 Ϫ4 M.
PCR Amplification-PCR amplification was performed in a minicycler from M. J. Research. The thermal cycle was set-up as follows: step 1, 95°C for 3 min; step 2, 94°C for 1 min for denaturation; step 3, 37°C for 2 min for annealing; and step 4, 70°C for 90 s for extension. A total of 30 cycles were performed with a final extension step of 7 min.
DNA Pure QPs3 was obtained from QPs by preparative SDS-PAGE essentially according to the previously reported method (10) except that the gels were pre-run for 10 h at 45 V with an anode buffer containing 0.1 M Tris-Cl, pH 8.9, 0.1 mM sodium thioglycolate and a cathode buffer containing 2% M Tris, 0.1 M Tricine, 0.1% SDS, and 0.1 mM sodium thioglycolate. The gel-eluted QPs3 protein was concentrated by membrane filtration, using centricon-10, to a protein concentration of 2 mg/ml and precipitated with cold acetone (Ϫ20°C). These precipitates were washed with 50% acetone, dried under argon, and subjected to NH 2 (10). The azido-Q-labeled reductase was precipitated by 43% ammonium sulfate saturation, separated by centrifugation at 48,000 ϫ g for 20 min, dissolved in 20 mM Tris-Cl, pH 7.8, and dialyzed against double-distilled water, overnight, with one change of water. [ 3 H]azido-Q-labeled QPs3 was purified to homogeneity from this dialyzed, labeled succinate-Q reductase as described above for isolation of pure QPs3.
Proteolytic Enzyme Digestion of QPs3-The reductive carboxymethylated and succinylated QPs3, 1 mg/ml, suspended in 50 mM ammonium bicarbonate buffer, containing 1 M urea was treated with chymotrypsin at 37°C for 2 h, using a chymotrypsin to QPs3 ratio of 1:50 (w/w). After the 2-h incubation, a second addition of chymotrypsin was made (1:100), and the digestion was continued at 37°C for 24 h.

Isolation of [ 3 H]Azido-Q-linked
Peptides-100 l aliquots of the chymotrypsin-digested, reductive carboxymethylated, succinylated, [ 3 H]azido-Q-labeled QPs3 were separated by reverse phase high performance liquid chromatography (HPLC) on a Synchropak RP-8 column (0.46 ϫ 25 cm) using a gradient formed from 90% acetonitrile in 0.1% trifluoroacetic acid and 0.1% trifluoroacetic acid with a flow rate of 0.8 ml/min. 0.8-ml fractions were collected and monitored for radioactivity and absorbance at 214 nm. Fractions with radioactivity were collected, dried, and sequenced. Those containing no radioactivity, with retention times of 18.96 and 31.15 were also collected, dried, and sequenced.
Production and Purification of Antibodies against QPs3 and Its NH 2terminal and Connecting Peptides-Pure QPs3 (see above) was used as antigen to raise antibodies in rabbits (15). Two polypeptides, one containing 14 amino acid residues (NH 2 -S-G-S-K-A-A-S-L-H-W-T-G-E-R-COOH) corresponding to residues 1-14 of QPs3 (the NH 2 -terminal peptide, see Fig. 4) and another containing 11 amino acid residues (NH 2 -T-D-Y-V-H-G-D-A-V-Q-K-COOH corresponding to residues 56 -66 (the connecting peptide, the loop between helices 2 and 3, see Fig. 4) were synthesized, purified, and used as antigens, after conjugation with ovalbumin, to raise antibodies in rabbits (24). Boosters were given weekly for 5 weeks, and sera were collected by cardiac puncture.
Purification of antibodies and preparation of the antibody FabЈ fragment-horseradish peroxidase conjugates were as previously reported (15). Horseradish peroxidase activity of the purified conjugate was assayed using a TMB peroxidase substrate kit (Bio-Rad) according to the manufacturer instructions.

RESULTS AND DISCUSSION
Isolation of the cDNA Encoding QPs3 from the Beef Heart cDNA Library in ZAP by PCR Cloning-Since the amino acid sequences of the membrane-anchoring subunits in succinate-Q reductase from different species show little conservation (25), we did not use the homology probing strategy to obtain the cDNA for QPs3. The availability of anti-QPs3 antibodies in our laboratory together with our previous success in immunological screening of a beef heart cDNA expression library in gt11 to obtain cDNAs for the Rieske iron sulfur protein (26), the QPc 9.5 kDa (27) of ubiquinol-cytochrome c reductase, and the QPs1 (15) of succinate-Q reductase encouraged us to use the immunological screening method to isolate the cDNA for QPs3. However, no positive result was obtained. This failure to obtain cDNA for QPs3 by the immunological screening method is probably due to the low titer of antibodies against QPs3 rather than the lack of QPs3 cDNA in the cDNA library used because we also failed to obtain a positive clone from other beef heart cDNA libraries, such as ZAP (from Stratagene). In an effort to obtain high titer anti-QPs3 antibodies, we added several more booster injections to rabbits and tried to raise antibodies in chickens. However, both attempts failed.
It was reported (28) that specific cDNA inserts are obtained from the cDNA libraries constructed in gt11 by PCR amplification using synthetic guessmers. The design of synthetic guessmers requires knowledge of a partial amino acid sequence of the target protein. When purified QPs3 (Fig. 1A, lane 6) was subjected to protein sequencing, 43 residues from the NH 2 terminus were obtained, NH 2 SGSKAASLHWTG-ERVVSVLLLGLIPAAYLNPCSAMDYSLAATL-. This enabled us to use the PCR cloning method to isolate the QPs3 cDNA from a beef heart cDNA library in ZAP. A 110-bp cDNA fragment was amplified from a beef heart cDNA library in ZAP (4 ϫ 10 6 plaque-forming unit) by PCR using two synthetic guessmers, 5Ј-GCTGCCTCCCTGCACTGGAC-3Ј (the sense primer) and 5Ј-GCAGCCAGGGAGTAGTCCAT-3Ј (the antisense primer). The sense primer represents the guessed sequence for residues 5-11 (A-A-S-L-H-W-T) with the degenerate third base of the codon of T 11 being omitted. The antisense primer represents the guessed sequence for residues 41-35 (A-A-L-S-Y-D-M). Since W and M have no degeneracy in their genetic codes, the presence of W 10 and M 35 in the QPs3 partial sequence enables us to design these two PCR guessmers with specificity at the 3Ј end, five specific bases in the 3Ј end of the sense guessmer, and three in the antisense guessmer.
The PCR reaction consists of 20 mM Tris-Cl, pH 8.3, 1.5 mM MgCl 2 , 30 mM KCl, 0.05% Tween 20, 100 g of autoclaved gelatin, 100 M dNTP, 2 units of Taq polymerase, and 200 pmol of guessmers. The resulting 110-bp PCR product was cloned into a PCR vector (TA cloning kit) and sequenced. The DNA sequence of this 110-bp PCR product translated to match the amino acid residues between Trp 10 and Met 35 of the chemically determined partial NH 2 -terminal sequence of QPs3.
Based on the nucleotide sequence for residues 10 -35 of QPs3, two gene specific primers, 5Ј-TTGCTCCTGGGCCTAAT-TCC-3Ј, corresponding to residues 21-27 (sense primer), and 5Ј-AGGAGCAAAACACTGAACAAC-3Ј, corresponding to residues 22-17 (antisense primer), were synthesized and used with the primers of vector ZAP, T7 and T3, respectively, in the subsequent PCR reactions to yield 3Ј-and 5Ј-RACE products. These two PCR products were cloned into the PCR II vector and sequenced. The 3Ј-RACE product is confirmed by matching the deduced amino acid residues 21-42 of QPs3 with chemically determined partial NH 2terminal amino acid sequence, and residues 59-66 and 44-49 with chymotryptic peptides of QPs3 with retention times of 18.96 and 31.15 min, respectively, on an HPLC chromatogram. Although the 5Ј-and 3Ј-RACE products can be joined together by PCR using a marathon cDNA amplification method (29), no effort was made to obtain a combined RACE product in this investigation. However, for future structure-function studies of QPs3, we have obtained a 331-bp BamHI-EcoRI cDNA fragment encoding mature QPs3 protein by PCR amplification from the beef heart cDNA library in ZAP using two primers, GGATCCTCTGGTTCCAAG (sense primer) and GAATTCTAAAAGGTCAGAGC (antisense), and cloned into a PCR vector.
Sequence Analysis of QPs3- Fig. 2 shows the nucleotide sequence and the deduced amino acid sequence of QPs3. The QPs3 cDNA is 1330 base pairs long with an open reading frame of 474 base pairs that encodes 158 amino acid residues, of which 103, starting with serine, belong to mature QPs3 and 55, starting with methionine, constitute an NH 2 -terminal presequence. In addition, the cDNA has 820 nucleotides of 3Ј noncoding sequence and contains a poly(A) tail.
The presequence of QPs3 is rich in the basic amino acid arginine and contains the hydroxy amino acid serine. This is characteristic of the cleavable amino acid terminal presequences that are essential for the import of mitochondrial proteins encoded by nuclear DNA (30). Since the QPs3 presequence lacks Arg at the Ϫ2 position relative to the mature amino terminus, QPs3 may be matured by a two-step cleavage process (31). Mitochondrial proteins with leader peptides containing the R-X-(F)-X-X-(S) motif, where r ϭ arginine at the Ϫ10 position, X ϭ other amino acid at Ϫ9; (F) ϭ hydrophobic residues at Ϫ8; and (S) ϭ serine, threonine, or glycine at Ϫ5, are thought to cleave first by a matrix processing protease between residues at Ϫ9 and Ϫ8 since arginine at position Ϫ10 is at position Ϫ2 relative to the cleaved bond. The remaining octapeptide is subsequently removed by an intermediate-specific protease. Although the QPs3 presequence lacks the arginine Ϫ10 in the common motif of this two-step cleavage, the maturation of QPs3 may still follow the same process because mutation of arginine Ϫ10 , in the human ornithine transcarbamylase precursor or rat malate dehydrogenase precursor, to alanine (32, 33) does not alter the two-step maturation of these two proteins.
The molecular mass of mature QPs3, determined from the deduced amino acid sequence, is 10,989 daltons, which is fairly close to the 9 kDa estimated from SDS-PAGE. Fig. 3 compares the amino acid sequence of bovine QPs3 with those of Saccharomyces cerevisiae Sdh4 (34), adult Ascaris suum cytbS (35), and Escherichia coli SdhD (36) and FrdD (37). Bovine QPs3 shares little sequence homology with gene products of S. cerevisiae sdh4, E. coli sdhD, and frdD but has 50% sequence identity with the A. suum (adult) cytochrome b small subunit of fumarate reductase (35). It is noteworthy that the similarity between bovine QPs3 and A. suum cybS in the region from Leu-23 to His-60 of QPs3 (68%) is higher than that of the entire peptide. This region contains a conserved histidine residue that matches His-71 of E. coli SdhD, which provides the heme b ligand (38). His-81 of E. coli FrdD that was reported to be involved in Q-binding is also in this region (39). It should be mentioned that the QPs3 cDNA was found to have 85% sequence identity to the DNA sequence of a Homo Sapiens cDNA clone (R91018) according to a BLAST search of the Expressed Sequence Tag (EST) data base.
The Proposed Structure of QPs3 in the Mitochondrial Inner Membrane- Fig. 4 shows the proposed structure of QPs3 in the inner mitochondrial membrane. This structural model, constructed from hydropathy plots of the amino acid sequence of QPs3 (40), predicted tendencies to form ␣-helices and ␤-sheets and the binding of FabЈ fragment-horseradish peroxidase con- jugates, prepared from antibodies against synthetic peptides corresponding to residues 1-14 and 55-66 of QPs3, in mitoplasts, submitochondrial particles (SMP), and alkali-treated submitochondrial particles. In this model, QPs3 has three transmembrane helices corresponding to residues 15-34 (helix I), 37-56 (helix II), and 67-89 (helix III). The NH 2 -terminal region, residues 1-14, and the loop connecting helices II and III, residues 57-66, are extruded from the M-side of the inner mitochondrial membrane. The loop connecting helices I and II, residues 36 -37, and the COOH-terminal region, residues 90 -103, are on the C-side of the membrane.
The sidedness of the membrane in this model was determined immunologically with FabЈ-horseradish peroxidase conjugates prepared from anti-QPs3, anti-NH 2 -terminal peptide (residues 1-14), and anti-connecting peptide (residues 57-66) antibodies in bovine heart mitoplasts (digitonin-treated intact mitochondria), SMP (reverse orientation), and alkaline-treated SMP (SMP devoid of succinate dehydrogenase). The peroxidase activity assays of these three particles are shown in Fig. 5. Since the peroxidase activity observed with preimmune FabЈhorseradish peroxidase-treated preparations is assumed to be due to nonspecific binding, it is subtracted from that of the anti-QPs3, anti-NH 2 -terminal peptide, or anti-connecting peptide FabЈ-horseradish peroxidase-treated preparations. The intactness of mitoplasts and submitochondrial particle preparations was established by the absence and presence of rotenonesensitive NADH-Q reductase activity.
When mitoplasts and SMP preparations were treated with anti-QPs3 FabЈ fragment-peroxidase conjugates, peroxidase activity was detected in both preparations. The slightly higher activity in treated SMP suggests that QPs3 is a transmembranous protein with slightly more mass exposed on the matrix side of the membrane. When an alkali-treated SMP preparation is treated with anti-QPs3 FabЈ fragment-horseradish per- oxidase conjugate, only a slight increase in peroxidase activity is observed, indicating that few epitopes on the matrix side of QPs3 are covered by succinate dehydrogenase.
When mitoplasts and SMP preparations are treated with FabЈ fragment-horseradish peroxidase conjugates prepared from anti-N-terminal and connecting peptides antibodies, peroxidase activity is observed only on SMP, indicating that the NH 2 -terminal end and the loop connecting helices II and III are exposed on the M-side of the mitochondrial inner membrane.
It should be mentioned that during the course of immunological studies of QPs3 we observed that antibodies against QPs3, the NH 2 -terminal, and the connecting peptides crossreact with QPs2 (Fig. 1, B-D, lanes 3 and 5). They do not react with QPs1 or proteins in ubiquinol-cytochrome c reductase (see Fig. 1). Antibodies against QPs2 also react with QPs3 (data not shown). This immunocross-reaction of QPs2 and QPs3 is expected because the QPs3 sequence is contained in QPs2. This is evident from the following observations. (i) When an electro-phoretically pure QPs2 preparation (Fig. 1A, lane 5) is subjected to protein sequencing, a major peptide with a partial NH 2 -terminal amino acid sequence of Ser-Pro-Ser-His-His-Ser-Gly-Ser-Lys-Ala-is obtained. This sequence contains five amino acid residues from the COOH terminus of the presequence and five amino acid residues from the NH 2 terminus of mature QPs3, suggesting that QPs2 is incompletely processed QPs3. (ii) When chymotrypsin-digested QPs2 and QPs3 are subjected to HPLC separation, identical chromatograms are obtained (data not shown). (iii) When peptides with identical retention times, from the respective QPs2 and QPs3 HPLC chromatograms, are sequenced, identical sequences are obtained. (iv) The azido-Q-labeled peptide from QPs2 is identical to that from QPs3 (described in the next section). At present, we do not know whether QPs2 is the same or a different gene product than QPs3. This requires further investigation.
Isolation and Characterization of the Ubiquinone-binding Peptides of QPs3-When succinate-Q reductase is photoaffinity labeled with [ 3 H]azido-Q derivative, about 50% of bound azido-Q is in QPs1, 22% in QPs2, and 25% in QPs3 (10). The smaller azido-Q uptake by QPs2 or QPs3, compared with QPs1, questions the Q-binding role of these two subunits. One way to establish a Q binding role for QPs3 is to isolate an [ 3 H]azido-Q-linked peptide from [ 3 H]azido-Q-labeled QPs3 obtained from [ 3 H]azido-Q-labeled succinate-Q reductase.
[ 3 H]azido-Q-labeled QPs3 is isolated from [ 3 H]azido-Q labeled succinate-Q reductase by a procedure involving phenyl-Sepharose CL-4B column chromatography, preparative SDS-PAGE, electrophoretic elution of proteins from gel slices, and acetone precipitation. The use of a hydrophobic column, phenyl-Sepharose CL-4B, and elution with different detergents results in the isolation of [ 3 H]azido-Q-labeled QPs from [ 3 H]azido-Q-labeled succinate-Q reductase. Since the FP and IP subunits of succinate dehydrogenase are less hydrophobic than those of QPs subunits, they are eluted with detergents having less hydrophobicity than that used for eluting QPs. This column chromatographic step also removes most of the non-protein bound [ 3 H]azido-Q from QPs. Pure [ 3 H]azido-Q-labeled QPs3 is isolated from the labeled QPs by preparative SDS-PAGE using high resolution gel system in the presence of 8 M urea. The use of preparative SDS-PAGE not only separates QPs3 from other QPs subunits, it also further removes non-protein bound azido-Q adducts from QPs3. QPs3 in gel slices is eluted by electrophoretic eluting. The SDS present in the eluted protein solution was removed by 50% acetone precipitation.
Although the isolated [ 3 H]azido-Q-labeled QPs3 is pure and free of free azido-Q, it is highly aggregated and resistant to proteolytic enzyme digestion. Inclusion of 0.1% SDS and 2 M urea in the digestion mixture does not increase proteolysis. Modification of isolated azido-Q labeled QPs3 by reductive carboxymethylation followed by succinylation renders the protein susceptible to chymotrypsin digestion. Reductive carboxymethylated and succinylated QPs3 is not completely soluble in aqueous solution; the solution becomes clear only after chymotrypsin digestion. A similar situation was observed with azido-Q labeled-cytochrome b (23). Fig. 6 shows the radioactivity distribution among the chymotryptic peptides of QPs3 separated by HPLC. Most of the radioactivity was found in fraction 66 (P-66). It should be mentioned that the HPLC chromatograms and radioactivity distribution of the chymotryptic peptides of QPs2 are identical to those of QPs3 (data not shown).
When P-66 from QPs3 was sequenced, a partial NH 2 -terminal sequence of Leu-Asn-Pro-Cys-Ser-Ala-Met-Asp-Tyr, corresponding to residues 29-37 in QPs3, is obtained. An identical sequence is obtained for the radioactivity containing fraction from QPs2. Thus the Q-binding domain in the proposed structure of FIG. 5. Binding of Fab fragment-horseradish peroxidase conjugates prepared from antibodies against QPs3, the NH 2 -terminal and connecting peptide of QPs3 with mitoplasts, submitochondrial particles, and alkali-treated submitochondrial particles. The indicated amounts of mitoplasts (z), submitochondrial particles ( ), and alkaline-treated submitochondrial particles (p) were mixed with 10 milliunits of FabЈ-horseradish peroxidase conjugates prepared from anti-QPs3 antibodies (A), anti-NH 2 -terminal peptide antibodies (B), and anti-connecting peptide antibodies (C) in 50 mM sodium phosphate, pH 7.4, containing 0.25 M sucrose and incubated at 4°C for 3 h. The mixtures were centrifuged at 30,000 ϫ g for 15 min, and the precipitate was suspended in 50 mM sodium phosphate, pH 7.4, containing 0.25 M sucrose. This procedure was repeated three more times before aliquots of the suspension were taken and horseradish peroxidase activity was assayed. Preimmune FabЈ-horseradish peroxidase conjugate treated with mitoplasts, submitochondrial particles and alkali-treated SMP in the identical manner was used as control. The activities indicated are after subtracting the control activity.
QPs3, is probably located at the end of transmembrane helix 1, near the C-side of the membrane. Recall that the Q-binding domain of QPs1 is on the M-side of the mitochondrial inner membrane.
The finding that the Q-binding domains in QPs3 and QPs1 of bovine succinate-Q reductase are on opposite sides of the membrane is in line with a two-Q binding site hypothesis formulated from inhibitor studies of this enzyme complex (41). The presence of two quinone-binding sites in E. coli fumarate reductase is suggested by mutational studies (39) and by inhibitor kinetics analysis of putative Q-binding site mutants (41). When these two quinone binding sites in E. coli fumarate reductase are incorporated into a proposed mechanism of Q reduction in photoreaction centers (42,43), Glu-29, Ala-32, His-82, Trp-86 of FrdC and His-80 of FrdD are considered participants in a Q B -type site, and FrdD Phe-57, Glu-59, and Ser-60 in an apolar Q A -type site (39). According to the proposed structure of E. coli FrdC and FrdD, the Q B -type site is located at the cytoplasmic side and the Q A -type site at periplasmic side. If this reasoning is applied to beef heart mitochondrial succinate-Q reductase, the Q-binding domain identified in QPs1 (10) would be the Q B -type site and the domain in QPs3 is the Q A -type site. More detailed information on Q-binding must await determination of the three-dimensional structure of succinate-Q reductase.