Resolution of the aerobic respiratory system of the thermoacidophilic archaeon, Sulfolobus sp. strain 7. III. The archaeal novel respiratory complex II (succinate:caldariellaquinone oxidoreductase complex) inherently lacks heme group.

An active respiratory complex II (succinate:quinone oxidoreductase) has been purified from tetraether lipid membranes of the thermoacidophilic archaeon, Sulfolobus sp. strain 7. It consists of four different subunits with apparent molecular masses of 66, 37, 33, and 12 kDa on sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The 66-kDa subunit contains a covalently bound flavin, the 37-kDa subunit is a possible iron-sulfur protein carrying three distinct types of EPR-visible FeS cluster, and the 33- and 12-kDa subunits are putative membrane-anchor subunits, respectively. While no heme group is detected in the purified complex II, it catalyzes succinate-dependent reduction of ubiquinone-1 and 2,6-dichlorophenolindophenol in the absence of phenazine methosulfate. The respiratory complex II of Sulfolobus sp. strain 7 appears to be novel in that it functions as a true succinate:caldariellaquinone oxidoreductase, although inherently lacking any heme group. This further indicates that the heme group of several respiratory complexes II may not be involved in the redox intermediates of the electron transfer from succinate to quinone.

The respiratory complex II (succinate:quinone oxidoreductase) is an iron-sulfur flavoprotein complex which serves as the sole membrane-bound component of the tricarboxylic acid cycle and one of the most important entry sites into the aerobic respiratory chain of a variety of aerobic organisms (1)(2)(3). The complex has been isolated from several sources and appears to be highly homologous in terms of the subunit composition and prosthetic groups (2,3). In general, it consists of three to four different subunits and contains one FAD, three distinct types of FeS cluster (a binuclear cluster called center S-1, one tetranuclear cluster called center S-2, and one trinuclear cluster called center S-3), and one or two protoheme IX as prosthetic groups. FAD binds covalently to the largest flavoprotein subunit (ϳ60 -75 kDa) via 8␣-N 3 -histidyl linkage, and all FeS clusters are probably located in the second largest iron-sulfur subunit (ϳ25-30 kDa). These two subunits constitute a peripheral portion of complex II, which can function as a water-soluble succinate:dye (such as phenazine methosulfate-DCIP) 1 oxidoreductase upon release from membranes (4).
For succinate:quinone oxidoreductase activity, however, two (or one in Bacillus subtilis enzyme (5)) smaller membranespanning subunits are required (2). These hydrophobic subunits contain low-potential b-type cytochrome(s). Aside from the structural importance, the functional role of protoheme has not been clarified due to the presence of a different number of protoheme per FAD and to the diverse redox behavior among complexes II upon reduction by succinate under steady-state conditions. Cytochrome b 560 of mitochondrial complex II is only slowly reducible by succinate (6), and that of highly active Paracoccus denitrificans complex II is not reducible at all (7); on the other hand, the protoheme center of the Escherichia coli enzyme is almost fully reduced by succinate and can be reoxidized by fumarate (8,9); and B. subtilis complex II contains two protoheme centers in a single membrane-anchor subunit (cytochrome b 558 ), of which only one is fully reducible by succinate (10,11). In addition, the primary structural comparison suggest that the membrane-anchor subunits are less homologous (3,12) compared to the cases of the flavoprotein and iron-sulfur subunits of mitochondrial and aerobic bacterial respiratory complexes II (2,3,(13)(14)(15)(16). The same observations have been reported for the fumarate reductase complexes, which are analogous in the structures and mediate the "reverse" reaction of respiratory complexes II (3,17,18).
As opposed to the cases of mitochondrial and bacterial enzymes, little information is available for the respiratory complexes II from aerobic archaea. In earlier studies, it was reported that both an extreme halophile Halobacterium salinarium (f. halobium) (19) and an extreme thermoacidophile Sulfolobus acidocaldarius strain DSM 639 (20) contained the succinate dehydrogenase activities both in cytosol and membrane fractions. The recent work by Moll and Schä fer (21) has shown that S. acidocaldarius complex II consists of four different subunits, and that a 66-kDa flavoprotein subunit derived from the membrane-bound complex can alone function as a water-soluble succinate:dye oxidoreductase. While purified S. acidocaldarius complex II exhibited only a low caldariellaquinone reductase activity and did not contain any heme (21), their relationship remains to be clarified due to the presence of multiple membrane-bound b-type cytochromes of unknown function in S. acidocaldarius (22,23). Interestingly, a 18-kDa diheme cytochrome b of another thermoacidophilic archaeon, Thermoplasma acidophilum (24), recently appears to be a part of a putative complex II (25).
Sulfolobus sp. strain 7 (formerly S. acidocaldarius strain 7) is a typical thermoacidophilic archaeon that acquires energy by the aerobic respiration rather than simple fermentation (26 -28). The previous studies using the intact membranes suggested that succinate is the best respiratory substrate in terms of a specific oxygen uptake activity of the membranes at 50°C (26). In the preceding papers (29,30), we reported the detailed properties of the membrane-bound cytochromes of Sulfolobus sp. strain 7, and the purification, characterization, in vitro reconstitution, and the resolution analysis of the archaeal terminal oxidase supercomplex containing four distinct heme centers (b 562 , a 583 , and aa 3 ), one copper, and a Rieske-type FeS cluster. In this paper, we report the purification and characterization of the respiratory complex II (succinate:quinone oxidoreductase) from Sulfolobus sp. strain 7. This complex II appears to be novel in that it inherently lacks any heme group, while it can function as a succinate:caldariellaquinone oxidoreductase in vitro.

Materials
Ubiquinone-1 was a generous gift from Dr. Isao Tanaka, Eisai Co. (Tsukuba, Japan), and caldariellaquinone was isolated from the membranes of Sulfolobus sp. strain 7 (31) as described previously according to De Rosa et al. (32). Sodium cholate, Lubrol-PX, sodium Sarcosyl, and Triton X-100 were purchased from Sigma, and MEGA-9 from Dojin (Kumamoto, Japan). Hydroxylapatite (Bio-Gel HTP) was purchased from Bio-Rad, DEAE-Toyopearl from Tosoh Corp., and Superose-6 from Pharmacia Biotechnology Inc. 2-Heptyl-4-hydroxyquinoline-N-oxide and myxothiazol are the products of Sigma. Water was purified by the Milli-Q purification system (Millipore). Other chemicals mentioned in this study were of analytical grade.

Purification of the Sulfolobus Respiratory Complex II
Sulfolobus sp. strain 7 was cultivated at pH 2.5-3 and 75-80°C as described previously (26,33). For enrichment of cytochromes in the membranes, the cells were harvested in the early-middle exponential phase of growth and stored at Ϫ80°C until use.
Approximately 150 g (wet weight) of cells were disrupted by a French press (Otake Works, Tokyo), and the Sulfolobus membrane was prepared as described in Ref. 29 (suspended in 15 mM Tris-Cl, pH 7.5, containing 20% glycerol and 0.2 mM PMSF, at ϳ20 mg of protein/ml). It was immediately used for the following purification of the archaeal respiratory complex II, which was performed at room temperature unless otherwise specified.
Step 1: Lubrol-PX Extraction-To the membrane suspension was added a solution of 10% (w/v) Lubrol-PX and 2 mM PMSF to final concentrations of 4% and 1 mM, respectively. The mixture was gently stirred for 90 min at 4°C, and then was centrifuged at 130,000 ϫ g for 60 min at 4°C.
Step 2: DEAE-Toyopearl Column Chromatography-After diluting the Lubrol-PX extract with the same volume of 20 mM Tris-Cl buffer, pH 7.5, containing 2 mM PMSF, the extract was applied to a DEAE-Toyopearl column (3.0 ϫ 43 cm) equilibrated with 20 mM Tris-Cl buffer, pH 7.5, containing 1% (w/v) Lubrol-PX and 2 mM PMSF. The column was washed with 300 ml of the equilibration buffer, and the eluent was combined together with the flow-through fraction. The archaeal respiratory complex II was found in the combined fraction, while almost all of the a-and b-type cytochromes (29) as well as an NADH dehydrogenase activity (33) were still adsorbed on the column.
Step 3: Ammonium Sulfate Fractionation-To the combined flowthrough fraction (ϳ0.5 mg of protein/ml) was added 270 mg/ml of solid ammonium sulfate while stirring at 4°C. The floating pellet was collected and suspended in 20 mM potassium phosphate buffer, pH 6.0, containing 2% (w/v) Lubrol-PX and 2 mM PMSF. It was then dialyzed twice against a 50-fold volume of 10 mM potassium phosphate buffer, pH 6.8, containing 0.1% (w/v) Lubrol-PX and 2 mM PMSF for 12 h at 4°C.
Step 4: Hydroxylapatite Column Chromatography-The dialysate was applied to a hydroxylapatite column (2.0 ϫ 20 cm) equilibrated with 10 mM potassium phosphate buffer, pH 6.8, containing 1% (w/v) Lubrol-PX and 2 mM PMSF. The column was washed with 60 ml of the equilibration buffer, and eluted with 300 ml of the same buffer containing a linear gradient of 10 -200 mM potassium phosphate. The pooled complex II fractions were concentrated on an Amicon YM-30 ultrafil-tration membrane, and could be stored at Ϫ80°C in 0.5-ml droplets. As described below, the contaminants (two protein species) in the active fractions could be removed either by gel filtration column chromatography (Step 5a) or a preparative sucrose density gradient (Step 5b).
Step 5a: Gel Filtration-Further purification of the archaeal respiratory complex II was achieved by a Superose-6 column chromatography (Pharmacia Biotech Inc., 1.0 ϫ 30 cm) with the fast protein liquid chromatography system, which was equilibrated in 100 mM potassium phosphate buffer, pH 6.8, containing 1% (w/v) Lubrol-PX; flow rate was 0.2 ml/min.
Step 5b: Sucrose Density Gradient-The active complex II fractions obtained after Step 4 could be also purified by a preparative sucrose density gradient, instead of Step 5a. The pooled fractions were loaded onto a sucrose density gradient (5-15%) containing 100 mM potassium phosphate buffer, pH 6.8, and 0.2% (w/v) Lubrol-PX; the tubes of a Beckman SW 28Ti rotor were then spun at 28,000 rpm for 20 h at 10°C. The brown band was collected by a tubing, and was checked for purity by SDS-PAGE. The pooled fractions obtained either after Steps 5a or 5b, were used as a purified complex II of Sulfolobus, and were stored at Ϫ80°C until use.

Measurement of Enzymatic Activities
A succinate:ubiquinone-1 oxidoreductase activity was routinely measured by following the absorbance change on the reduction of ubiquinone-1 (50 M) at 278 nm using an extinction coefficient of 14.7 cm Ϫ1 mM Ϫ1 (9). A succinate-driven caldariellaquinone reductase activity was measured spectrophotometrically at 351 minus 341 nm using a differential absorption coefficient of 1.8 cm Ϫ1 mM Ϫ1 , according to Refs. 21 and 34. A succinate:DCIP oxidoreductase activity was monitored by following the absorbance change of DCIP at 600 nm (without addition of mediators). All these assays were performed at 50°C in the presence of 0.1% (w/v) Lubrol-PX and 5 mM KCN.

Analytical Methods
Absorption spectra were recorded either with a Hitachi U-3210 spectrophotometer equipped with a thermoelectric cell holder, or with a Shimadzu MPS-2000 spectrophotometer, except for the in vitro reconstitution experiments. EPR measurements were carried out using a JEOL JEX-RE1X spectrometer equipped with an Air Products model LTR-3 Heli-Tran cryostat system, in which temperature was monitored with a Scientific Instruments series 5500 temperature indicator/controller. Protein was measured by the BCA assay (Pierce Chemical Co.) with bovine serum albumin as a standard. The heme was determined by the pyridine hemochromogen method (35). The flavin content was determined as described by Singer (36) except the emission was measured at 513 nm (cf. Ref. 21) with a Shimadzu spectrofluorophotometer RF-540. Metal content analyses were carried out by inductively coupled plasma atomic emission spectrometry with a Seiko SPS 1500 VR instrument.
Polyacrylamide gel electrophoresis in the presence of SDS was carried out according to Laemmli (37), and proteins were visualized by Coomassie Brilliant Blue staining. Two-dimensional gel electrophoresis was performed as follows. The purified enzyme was treated with 4% Sarcosyl and 150 mM NaCl, and then subjected to 5% polyacrylamide gel electrophoresis in the absence of any detergent in the first dimension using the Laemmli discontinuous system; then, the lanes were cut out, soaked in 5% SDS and 10% 2-mercaptoethanol for 20 min at room temperature, and subjected to the 14% SDS-PAGE in the second dimension. Fluorescence gel photographs were taken with gels agitated in 25% methanol and 10% acetic acid for 60 min; gels were illuminated with a long-wavelength ultraviolet lamp, and occasionally a green Wratten filter (Kodak) was used on the camera lens (7). N-terminal amino acid sequence analysis was carried out by an Applied Biosystems gas-phase sequencer model 470A. Table I summarizes typical purification data of the respiratory complex II (succinate:quinone oxidoreductase) from the membranes of Sulfolobus sp. strain 7 by following a succinate:ubiquinone-1 oxidoreductase activity. Although the activity was not selectively solubilized when non-ionic Lubrol-PX was used, this could be partially overcome by choosing the "green membranes" prepared from the Sulfolobus cells harvested in the early to middle log phase, as a starting material for purification; such mem-branes are enriched with cytochromes (ϳ2-3-fold higher than those in the "brown membranes" prepared from the late-log to stationary phase cells) and show a higher succinate-driven oxygen uptake activity (data not shown). In addition, Lubrol-PX appeared suitable for purification of the archaeal respiratory complex II with respect to no significant change in the K m for succinate during solubilization and purification (K m ϭ ϳ280 M), as in the case reported for E. coli respiratory complex II (8,9). While a succinate:ubiquinone-1 oxidase activity became essentially free from the membrane-bound cytochromes at Step 2 and thereafter (see "Experimental Procedures"), the increase in specific activity is observed as the purification proceeds (Table I). The final preparation (obtained after the preparative sucrose density gradient step) reaches a 14.6-fold purification from the Lubrol-PX extract, with a specific activity of ϳ4.0 units/mg at 50°C and at pH 7.5 ( Table I).

Purification and Molecular Properties of the Respiratory Complex II from Sulfolobus sp. Strain 7-
The specific contents of covalently bound flavin and non-heme iron in the purified enzyme are ϳ5.6 and ϳ83 nmol/mg, respectively. It contained no detectable copper or heme iron.
The purified respiratory complex II of Sulfolobus sp. strain 7 consists of four different subunits with apparent molecular masses of 66, 37, 33, and 12 kDa, respectively (Fig. 1). Because of its thermostable nature, it exhibits an additional 90-kDa band on SDS-PAGE, when treated in the presence of 2% SDS and 2% 2-mercaptoethanol at 80 -90°C for 1 min or at room temperature ( Fig. 2A). When the 90-kDa band was cut out and electroeluted from SDS-polyacrylamide gels, and then subjected to SDS-PAGE again after complete denaturation, it appears to be composed of the 66-, 37-, and 12-kDa subunits (Fig.  2B). These data suggest the presence of very strong proteinprotein interactions among these three subunits, and of weaker interactions between the 33-kDa subunit and the rest of the subunits of the archaeal respiratory complex II.
Optical and EPR Properties- Fig. 3 shows the optical spectra of the purified respiratory complex II. The air-oxidized spectrum is characteristic of an iron-sulfur flavoprotein, and is partially reducible by succinate. The purified Sulfolobus enzyme contained no succinate-reducible protoheme center typically observed in the mitochondrial and several bacterial respiratory complexes II (2, 3); the absence of any heme group was further confirmed by the pyridine ferrohemochrome method (data not shown).
The inset in Fig. 3 shows the visible spectra of the 90-, 66-, and 33-kDa bands obtained by electroelution from SDS-polyacrylamide gels (see Fig. 2). The absorption spectra of the 90and 66-kDa bands showed maxima at 445 and around 350 nm, similar to the case of FAD (maxima at 450 and 375 nm). The blue-shift of near UV peak (from 375 to 350 nm) in the flavincontaining bands is characteristic of 8␣-substituted flavins (38,39), but this was not confirmed in this study. Fig. 4 shows the EPR spectra of the purified respiratory complex II at 10 -25 K. The air-oxidized enzyme of Sulfolobus (as isolated) elicits the EPR signal at g ϭ 2.02 at 10 K (Fig. 4A), which is characteristic of a [3Fe-4S] 1ϩ(1ϩ,0) cluster, probably corresponding to the oxidized S-3 center of mitochondrial respiratory complex II (2,3). This EPR signal was also detectable in the archaeal membranes (data not shown). Upon addition of 5 mM succinate, the S-3 center in the archaeal respiratory complex II was reduced, and the resulting EPR spectra consisted of at least two overlapping S ϭ 1/2 species (Fig. 4B); of these, the rhombic resonances at g z,y,x ϭ 2.03, 1.94, and 1.90 are readily detectable at 25 K (data not shown), and are attributable to a reduced plant-ferredoxin-type [2Fe-2S] 1ϩ(2ϩ,1ϩ) cluster, probably corresponding to the S-1 center of mitochondrial respiratory complex II (2, 3). The succinate-reduced semiquinone radical feature at g ϭ 2.005 (Fig. 4B) is further reducible by dithionite, being diamagnetic, and the resulting EPR spectrum is attributable to the reduced S-1 center at 25 K (Fig. 4C). Thus, the complexity of the EPR spectrum of the succinate-reduced enzyme shown in Fig. 4B suggests a spinspin interaction between the archaeal S-1 center and the radical center (possibly flavin semiquinone radical), which further  indicates their close proximity in the archaeal respiratory complex II, as in the case of the mitochondrial complex (2). Besides the EPR signal attributed to the reduced S-1 center, the EPR spectrum of the dithionite-reduced enzyme elicits an another weak S ϭ 1/2 resonances at g ϭ ϳ2.08 and ϳ1.88, spanning a wide range of the magnetic field, which is detectable at 10 K ( Fig. 4D) but not at 25 K or higher (Fig. 4C). These weak and broad "wing" EPR signals are reminiscent of the reduced S-2 center of the mitochondrial complex II (2, 3), and most likely attributable to a [4Fe-4S] 1ϩ(2ϩ,1ϩ) cluster.
In the succinate-reduced intact membranes of Sulfolobus, the EPR signals of the reduced S-1 center could be also detected in the presence of cyanide, but not in the absence of cyanide (data not shown; cf. Ref. 29). Thus, of three different FeS clusters present in the archaeal respiratory complex II, at least centers S-1 and S-3 are directly involved in the membranebound aerobic respiratory chain from succinate to molecular oxygen (29,30), while the functional participation of the S-2 center remains to be investigated (cf. Ref. 40).
Possible Assignments of Fp and Ip Subunits-The two-dimensional PAGE analysis of the archaeal respiratory complex II and the visible spectrum of its 66-kDa subunit (see Figs. 1-3) suggest that the 66-kDa subunit is the flavoprotein subunit containing a covalently bound flavin. This was further verified by the N-terminal amino acid sequence analysis of the subunit (up to 21 residues), which contained a consensus FAD-binding -GXGXAG-motif and showed high homology to other flavoprotein subunits of several bacterial respiratory complexes II and fumarate reductase complexes (3, 14, 16, 41) (Fig. 5).
The close spatial proximity of the three FeS centers and the strict conservation of at least 10 cysteinyl residues in the ϳ30-kDa iron-sulfur subunits of the mitochondrial and several bacterial respiratory complexes II and fumarate reductase complexes suggest that all three FeS centers are probably located in their Ip subunits (2,3). Since most non-metalloproteins from extreme thermophiles do not contain any (or few) cysteinyl residue(s), the chemical modification of cysteine residues with iodoacetamide at pH 8.0 was used to identify the iron-sulfur subunit of the Sulfolobus respiratory complex II which is also expected to contain ϳ10 cysteinyl residues (cf. Refs. 3 and 25). The purified enzyme was denatured at 90°C for 10 min in 200 mM Tris-Cl buffer, pH 8.0, containing 2% SDS and 20 mM dithiothreitol, and subsequently incubated at room temperature for 35 min in the presence of 100 mM iodoacetamide, pH 8.0. Fig. 6 clearly shows that alkylation of the purified archaeal respiratory complex II with iodoacetamide prior to SDS-PAGE affected the mobility of the 37-kDa subunit with an increment of ϳ0.5 kDa, but not those of other subunits. This indicates that the 37-kDa subunit is probably the iron-sulfur subunit of the archaeal respiratory complex II. Unfortunately, further confirmation either by N-terminal amino acid sequencing or by preparing a soluble succinate dehydrogenase dimer consisting solely of the flavoprotein and iron-sulfur subunits (cf. Ref The 90-kDa subcomplex was recovered from polyacrylamide gels by electroelution, then subjected to 13% SDS-PAGE after complete denaturation in the presence of 2% SDS and 5% 2-mercaptoethanol for 90 min at 50°C. Std, molecular size standards (Bio-Rad, low-range marker proteins); proteins were visualized by Coomassie Brilliant Blue staining.
FIG. 3. The room temperature optical spectra of purified respiratory complex II from Sulfolobus sp. strain 7, and its constituents recovered by electroelution from SDS-polyacrylamide gels (inset). The air-oxidized (solid trace) and succinate-reduced (dotted trace) absolute spectra of purified respiratory complex II of Sulfolobus sp. strain 7 at room temperature. The sample (0.65 mg/ml) was dissolved in 100 mM potassium phosphate buffer, pH 6.8, containing 0.2% (w/v) Lubrol-PX. No heme group was detected in the preparation. Inset, absolute spectra at pH 8.5 of the air-oxidized forms of electroeluted 90-kDa subcomplex (p90, solid trace), 66-kDa subunit (p66, dotdash), and 33-kDa subunit (p33, dotted trace), indicating that the former two contain a covalently bound flavin.
from the purified complex, has so far appeared unsuccessful (data not shown).
Kinetic Properties-The purified respiratory complex II from Sulfolobus sp. strain 7 showed a succinate:ubiquinone-1 oxidoreductase activity in the presence of 0.1% (w/v) Lubrol-PX at 50°C and at pH 6.8; the K m for succinate ϭ 280 M, and the K m for ubiquinone-1 ϭ 20 M with the V max ϭ 14 mol/min/mg. These kinetic parameters are not significantly altered from the membrane-bound state at 50°C and at pH 6.8 (the K m for succinate ϭ ϳ280 M, and the K m for ubiquinone-1 ϭ 23 M). The purified enzyme can use DCIP as an alternative good electron acceptor in the absence of any electron mediator: the K m for DCIP ϭ 89 M, and the V max ϭ 13.6 mol/min/mg. When either ubiquinone-1 or DCIP were used as an electron acceptor, the pH optimum of the reaction was at pH 6.5-6.8 (data not shown). Since the optimal pH for growth of Sulfolobus is pH 2.5-3 (26,42), this suggests that the catalytic site of the archaeal respiratory complex II is exposed to the cytoplasmic side. The catalytic reaction of the enzyme was temperature-dependent as reported for S. acidocaldarius enzyme by Moll and Schä fer (21), and a difficulty was found in measuring the enzymatic activities at room temperature (data not shown). The purified Sulfolobus respiratory complex II can also use caldariellaquinone, which is a benzo-[b]-thiophen derivative of quinone characteristic of Sulfolobales (32) and is the probable physiological electron acceptor of the enzyme; the K m for caldariellaquinone ϭ ϳ60 M with the V max ϭ ϳ1.8 mol/min/mg. Since the membrane-bound NADH dehydrogenase purified from the same species could not use this compound as an electron acceptor (33), the low activity of the purified complex II toward caldariellaquinone most likely reflects not only the difficulty of handling this very hydrophobic compound in vitro, but also the assay conditions far below the optimal temperature for growth of the archaeon (50°C versus ϳ80°C; see Ref. 26). Thus, these data suggest that Sulfolobus complex II can function as a succinate:caldariellaquinone oxidoreductase in vitro, in spite of the absence of any heme group typically observed for mitochondrial and bacterial respiratory complexes II (2,3). This is further supported by the in vitro reconstitution experiments, proving that it functions as a primary dehydrogenase of the cyanide-sensitive active succinate-oxidizing respiratory chain in the presence of the cognate terminal oxidase supercomplex, caldariellaquinone, and succinate (29).
Because of the molecular identity of cytochrome b 562 in the purified terminal oxidase supercomplex and the succinate-reducible cytochrome b 562 in the archaeal intact membranes (29), the active respiratory complex II of Sulfolobus sp. strain 7 appears to lack any heme group inherently, as in the case of E. coli fumarate reductase complex which catalyzes a "reverse" menaquinol:fumarate oxidoreductase reaction (17). DISCUSSION Active Respiratory Complex II of Sulfolobus sp. Strain 7-The respiratory complex II of Sulfolobus sp. strain 7 purified in this study consists of four different subunits. The largest 66-kDa subunit contains a covalently bound flavin, the second largest 37-kDa subunit most likely contains three distinct types of FeS cluster (center S-1, a [2Fe-2S] 2ϩ,1ϩ cluster; center S-2, a [4Fe-4S] 2ϩ,1ϩ cluster; and center S-3, a [3Fe-4S] 1ϩ,0 cluster) whose EPR properties are comparable to those of the corresponding FeS clusters in mitochondrial and several bacterial respiratory complexes II and fumarate reductase complexes (2,3). In the membrane-bound state, at least centers S-1 and S-3 of the archaeal complex II seem to be directly involved in the aerobic respiratory chain, while the participation of center S-2 remains to be studied further. These two subunits probably form the catalytic core for oxidation of succinate, and the additional two smaller subunits (the 33-and 12-kDa subunits) may serve as the putative membrane anchor for reduction of caldariellaquinone. The 12-kDa subunit shows strong interaction with the 66-and 37-kDa subunits in vitro, thus most likely associating directly to these subunits in vivo. Interestingly, its apparent molecular mass is comparable to the quinone-binding subunits of mitochondrial respiratory complex II and of E. coli fumarate reductase complex (2,3,12,43,44), further implying that it might serve for caldariellaquinone binding. Thus, the overall architecture of the Sulfolobus respiratory complex II is very similar to those of mitochondrial and other prokaryotic respiratory complexes II and fumarate reductase complexes in terms of the subunit composition and the prosthetic groups other than heme (2,3).
The kinetic properties of the archaeal complex II are very similar to other prokaryotic respiratory complexes II. It catalyzes electron transfer from succinate to DCIP in the absence of any mediators, and also to ubiquinone-1. The K m for succinate was 0.28 mM (not altered from the membrane-bound state), which is higher than those of the protoheme-containing respiratory complexes II from beef heart mitochondria (0.02 mM) (45) and E. coli (0.071 mM) (9), but lower than those of the water-soluble succinate dehydrogenase dimers derived from complexes II (4,19,46). Such "intermediate" K m values for succinate have been reported also for other archaeal complexes II (e.g. from H. salinarium strain R1 (ϳ0.7 mM) (19) and S. acidocaldarius strain DSM 639 (1.42 mM) (21)), indicating that it may be characteristic of the archaeal complexes II.
The most notable feature of the purified complex II of Sulfolobus sp. strain 7 is the absence of any heme group (Fig. 3), as in the cases reported for S. acidocaldarius complex II (21) and certain bacterial fumarate reductase complexes (3,17,47). Although not reacting rapidly in the presence of Lubrol-PX, the Sulfolobus respiratory complex II functions as an active succinate:caldariellaquinone oxidoreductase in the in vitro succinate-oxidizing reconstitution system in cyanide-and 2-heptyl-4-hydroxyquinoline-N-oxide-sensitive manners (29), indicating that it inherently lacks any heme group. The in vivo function of the complex is most likely the same, since the archaeon grows aerobically and contains a 2-oxoacid:ferredoxin oxidoreductase, another key enzyme of the archaeal "oxidative" tricarboxylic acid cycle (28, 48 -50).
Mechanical and Evolutionary Implications-The lipophilic part of respiratory complex II usually carries one (e.g. mitochondrial (44,45,51), E. coli (8,9), and P. denitrificans complexes II (7)) or two protoheme groups (e.g. B. subtilis complex II (11)). The absence of any heme group in Sulfolobus complex II indicates that the low-potential heme b in other respiratory complex II may not be involved in the redox intermediates of the electron transfer from succinate to quinone, especially if one assumes that all respiratory complex II possess essentially similar intramolecular electron transfer mechanisms. We assume that at least one of the putative lipophilic subunits of Sulfolobus complex II participates in binding and reducing caldariellaquinone, and that low-potential heme b centers in complex II play only a structural role rather than functional. The presence of two succinate-reducible FeS centers, S-1 and S-3, and a stable radical feature in the archaeal complex II is in line with the "linear-sequence model" involving at least center S-3 and stable semiquinone for intramolecular electron flow through respiratory complexes II (3) rather than the "dual pathway model" involving both the heme b/center S-2 pair and the center S-1/center S-3 pair (52). On the other hand, the role of an another low-potential center S-2, which is also present in Sulfolobus respiratory complex II and is not reducible by succinate, remains to be clarified, but the spatial proximity of three FeS clusters in complexes II indicates that it may largely affect the redox equilibrium of the intramolecular electron transfer from the flavin site to the quinone/quinol couple (cf. Refs. 2, 40, and 53).
Quite recently, a part of a putative succinate dehydrogenase operon of the thermoacidophilic archaeon, T. acidophilum, has been cloned and sequenced (25), and appears to encode the 18-kDa diheme cytochrome b (24). The absence of heme groups in the Sulfolobus complex II suggests that the heme b center in respiratory complexes II might have lost its redox function after divergence of the thermoacidophilic archaea (54,55), representing an evolutionary relic at the present time (cf. Ref. 2 for review), and that the putative loss of heme groups in the early respiratory complex II of Sulfolobus sp. strain 7 may represent a unique evolutionary event.