The 2-oxoacid dehydrogenase multi-enzyme complex of the archaeon Thermoplasma acidophilum ) recombinant expression, assembly and characterization Caroline Heath 1 , Mareike G. Posner 1 , Hans C. Aass 1 , Abhishek Upadhyay 2 , David J. Scott 3 , David W. Hough 1 and Michael J. Danson 1 1 Centre for Extremophile Research, Department of Biology and Biochemistry, University of Bath, UK 2 Department of Biology and Biochemistry, University of Bath, UK 3 National Centre for Macromolecular Hydrodynamics, School of Biosciences, University of Nottingham, Sutton Bonington, UK In aerobic bacteria and eukaryotes, a family of 2-oxoacid dehydrogenase multi-enzyme complexes (OADHCs) functions in the pathways of central metabolism. The complexes are responsible for the oxi- dative decarboxylation of 2-oxoacids to their corre- sponding acyl-CoAs. Members of the family include the pyruvate dehydrogenase complex (PDHC), which catalyzes the conversion of pyruvate to acetyl-CoA and so links glycolysis and the citric acid cycle; the 2-oxoglutarate dehydrogenase complex (OGDHC), which catalyzes the conversion of 2-oxoglutarate to succinyl-CoA within the citric acid cycle; and the branched-chain 2-oxoacid dehydrogenase complex (BCOADHC), which oxidatively decarboxylates the branched-chain 2-oxoacids produced by the transami- nation of valine, leucine and isoleucine. The complexes comprise multiple copies of three component enzymes: 2-oxoacid decarboxylase (E1), dihydrolipoyl acyl-trans- ferase (E2) and dihydrolipoamide dehydrogenase (E3) [1–3]. E2 forms the structural core of the complex, with multiple polypeptide chains associating into octa- hedral (24-mer) or icosahedral (60-mer) configurations, depending on the particular complex and the source organism [2,4]. E1 and E3 bind noncovalently to the Keywords Archaea; metabolism; multi-enzyme complex; 2-oxoacid dehydrogenase; thermophile Correspondence M. J. Danson, Centre for Extremophile Research, Department of Biology and Biochemistry, University of Bath, Bath, BA2 7AY, UK Fax: +44 1225 386779 Tel: +44 1225 386509 E-mail: M.J.Danson@bath.ac.uk (Received 29 June 2007, revised 23 August 2007, accepted 24 August 2007) doi:10.1111/j.1742-4658.2007.06067.x The aerobic archaea possess four closely spaced, adjacent genes that encode proteins showing significant sequence identities with the bacterial and eukaryal components comprising the 2-oxoacid dehydrogenase multi- enzyme complexes. However, catalytic activities of such complexes have never been detected in the archaea, although 2-oxoacid ferredoxin oxidore- ductases that catalyze the equivalent metabolic reactions are present. In the current paper, we clone and express the four genes from the thermophilic archaeon, Thermoplasma acidophilum, and demonstrate that the recombi- nant enzymes are active and assemble into a large (M r ¼ 5 · 10 6 ) multi- enzyme complex. The post-translational incorporation of lipoic acid into the transacylase component of the complex is demonstrated, as is the assembly of this enzyme into a 24-mer core to which the other components bind to give the functional multi-enzyme system. This assembled complex is shown to catalyze the oxidative decarboxylation of branched-chain 2-oxoacids and pyruvate to their corresponding acyl-CoA derivatives. Our data constitute the first proof that the archaea possess a functional 2-oxo- acid dehydrogenase complex. Abbreviations BCOADHC, branched-chain 2-oxoacid dehydrogenase complex; CoASH, coenzyme-A; DLS, dynamic light scattering; E1, 2-oxoacid decarboxylase; E2, dihydrolipoyl acyl-transferase; E3, dihydrolipoamide dehydrogenase; FOR, ferredoxin oxidoreductase; IPTG, isopropyl thio-b- D-galactoside; M r , relative molecular mass; OADHC, 2-oxoacid dehydrogenase complex; OGDHC, 2-oxoglutarate dehydrogenase multienzyme complex; PDHC, pyruvate dehydrogenase complex; TPP, thiamine pyrophosphate. 5406 FEBS Journal 274 (2007) 5406–5415 ª 2007 The Authors Journal compilation ª 2007 FEBS E2 core. E1 may occur as a homodimer or as an a 2 b 2 hetero-tetramer, depending upon the source and the type of complex, although in all cases E3 is a dimer of identical subunits. E2 also forms the catalytic core of the complex; a lipoyl moiety, covalently attached to a lysine residue in the lipoyl domain of the E2, serves as a swinging arm, connecting the active sites of each enzyme and chan- nelling substrate through the complex [2,3]. Thus, E1 catalyzes the thiamine pyrophosphate (TPP)-dependent decarboxylation of the 2-oxoacid and the transfer of the resulting acyl group to the lipoic acid of E2. E2 then transfers the acyl-group to coenzyme-A (CoASH), after which E3 serves to reoxidize the dihydrolipoyl moiety. It does so by the reduction of the noncovalently bound cofactor FAD, in conjunction with a protein disulfide bond and an amino acid base, all of which are themselves then reoxidized by NAD + to form NADH. In Archaea [5] and anaerobic bacteria, the equiva- lent oxidation of 2-oxoacids is catalyzed by an un- related, and structurally more simple, family of 2-oxoacid ferredoxin oxidoreductases (FORs). This comprises the pyruvate FOR, the 2-oxoglutarate FOR and the 2-oxoisovalerate FOR, which catabolize the oxidative decarboxylation of pyruvate, 2-oxoglutarate and the branched-chain 2-oxoacids, respectively [6–8]. The pyruvate FOR from the halophilic archaeon Halo- bacterium halobium is an a 2 b 2 structure [9], whereas in Sulfolobus solfataricus and Aeropyrum pernix it is an ab dimer, and an octamer (a 2 b 2 c 2 d 2 )inPyrococcus furiosus, Methanothermobacter thermoautotrophicum and Archaeoglobus fulgidus [reviewed in 5,8]. The cata- lytic reaction of FORs does not involve a lipoic acid moiety or NAD + ; rather, the acyl-moiety formed on decarboxylation of the 2-oxoacid is handed on directly to CoASH, and the reducing equivalents to ferredoxin via the enzyme’s iron-sulfur centre [6–8]. No OADHC activity has ever been detected in any archaeon [5]. However, detection of E3 and lipoic acid in various archaea [10–12] led to the discovery of a putative OADHC operon in Haloferax volcanii [13] and our subsequent detection of similar putative ope- rons in the genome sequences of the aerobic archaea Thermoplasma acidophilum, S. solfataricus, Sulfolobus acidocaldarius, A. pernix, Pyrobaculum aerophilum, Halobacterium NRC1 and Ferroplasma acidophilum (M. G. Posner, unpublished data). We have previously expressed the E1a and E1b genes of the putative OADHC from T. acidophilum in Escherichia coli and shown the recombinant proteins to assemble into an a 2 b 2 enzyme that catalyzes the decarboxylation of the branched-chain 2-oxoacids and pyruvate [14]. In the current paper, we report the clon- ing and expression of the E2 and E3 genes of the same operon from T. acidophilum, and the in vitro assembly and characterization of an active 2-oxoacid dehydro- genase complex. This, then, is the first evidence that the putative OADHC operon in an archaeon encodes a 2-oxoacid dehydrogenase multi-enzyme complex that is func- tional in vitro and therefore may have physiological significance. Results Expression and purification of the E1, E2 and E3 components The E1 component The E1 a 2 b 2 recombinant enzyme was produced as described previously, and was shown to be catalyti- cally active with the branched-chain 2-oxoacids 4-methyl-2-oxopentanoate, 3-methyl-2-oxopentanoate and 3-methyl-2-oxobutanoate, and with pyruvate [14]. By dynamic light scattering (DLS), its M r was found to be 168 000 (± 6000), consistent with the value of 165 000 determined by gel filtration [14] and the expected value of 157 000 from the protein sequences. The E2 component The gene encoding the E2 component was PCR-ampli- fied from T. acidophilum genomic DNA and cloned into the pET28a expression vector, as described in Experimental procedures. The E2 protein was then expressed in two host systems: E. coli BL21(DE3) cells, in medium supplemented with lipoic acid but without isopropyl thio-b-d-galactoside (IPTG) induction, and E. coli BL21(DE3)pLysS cells without supplementation but with IPTG induction. Both methods yielded solu- ble E2 protein, although the level of expression was significantly greater in the pLysS cells. In each case, E2 was purified to > 95% homogeneity using His- Bind affinity chromatography followed by anion exchange chromatography. The E2 M r values predicted from the published gene sequence are 46 276 for unlipoylated protein and 46 464 for the polypeptide possessing a single lipoyl residue. Accordingly, mass spectrometric analysis revealed that the E2 expressed in E. coli grown in lipoic acid-supplemented medium comprised an approximately equimolar mixture of unlipoylated (M r ¼ 46 273) and lipoylated (M r ¼ 46 461) protein. Furthermore, from MS-analysis of tryptic fragments, C. Heath et al. 2-Oxoacid dehydrogenase complex from the Archaea FEBS Journal 274 (2007) 5406–5415 ª 2007 The Authors Journal compilation ª 2007 FEBS 5407 lipoylation was shown to have occurred at K42, which by sequence comparisons corresponds to the lipoylated lysine residue in bacterial E2 components (Fig. 1). However, the E2 protein produced by expression in induced cells, without lipoic acid supplement to the growth medium, was < 5% lipoylated. As reported below, consistent with these data is the observation that only the complex assembled using the lipoylated E2 component showed detectable catalytic activity in the overall complex assay. The E3 component The gene encoding the E3 component was PCR- amplified from T. acidophilum genomic DNA and cloned into the pET28a expression vector, as described in Experimental procedures. Expression of the E3 gene was carried out in E. coli BL21(DE3) cells. Small-scale expression trials with induced versus uninduced host cells revealed a higher E3 activity in the soluble cell extract from uninduced cells, and this was confirmed by SDS ⁄ PAGE. The soluble protein was purified to > 95% homogeneity by heat precipi- tation at 60 °C and His-Bind affinity chromatogra- phy. The absorption spectrum of the purified protein showed peaks at 375, 450 and 475 nm that are char- acteristic of the presence of FAD in the enzyme; using a molar absorption coefficient for FAD of 11 300 m )1 Æcm )1 at 455 nm [15], the flavin content was calculated to be 1.0 (± 0.1) FAD per polypep- tide (M r ¼ 49 867). Analytical gel filtration revealed an M r ¼ 100 000 for the purified enzyme, and DLS gave a similar value (117 000 ± 2000), suggesting a dimeric structure, as has been found for the bacterial and eukaryotic enzymes. The maximal specific activity of the enzyme was found to be 22 (± 1) lmolÆ min )1 Æmg )1 , a value that is considerably higher than the overall complex activity (see below). Assembly of OADHC from the recombinant components Complex assembly When E1, E2 and E3 were incubated at 55 °C for 10 min in 20 mm sodium phosphate buffer, pH 7.5, containing 2 mm MgCl 2 and 0.2 mm TPP, prior to assay, whole complex activity was subsequently detected with the branched-chain 2-oxoacids 4-methyl- 2-oxopentanoate, 3-methyl-2-oxopentanoate and 3-methyl-2-oxobutanoate, and with pyruvate. The 10-min incubation is necessary to allow E1-TPP bind- ing [16], but no increase in rate was seen when the incubation, after subunit mixing, was extended to 1 h at 4 °C, 25 °Cor55°C. With an E2 : E3 molar ratio fixed at 1 : 1 (E2 poly- peptide to E3 dimer), titration with E1 resulted in an increase in whole complex activity until a maximum was reached at a molar ratio (E1 ⁄ E2) of approximately 2–3 : 1 [E1 a 2 b 2 tetramer to E2 polypeptide] (Fig. 2). However, a reduction of the E3 ratio did not cause a significant change in whole complex activity, and by investigating the mixing of various amounts of the complex components it was found that maximum OADHC activity was achieved when the E1, E2 and E3 subunits were mixed in a molar stoichiometry of 3 : 1 : 0.1. However, as described below, this stoichio- metry does not equate to the amounts of the three enzymes in the assembled complex. B. subtilis PDHC B. subtilis BCOADHC Tp. acidophilum OADHC E. coli OGDHC E. coli PDHC (lipoyl domain 3) Fig. 1. Alignment of various E2 sequences around the lipoylated lysine residue. The T. acidophilum E2 protein sequence was aligned with those of the E2 components of the following OADHCs using the CLUSTALW multiple sequence alignment program: Bacillus subtilis PDHC; B. subtilis BCOADHC; E. coli OGDHC; E. coli PDHC (lipoyl domain 3). Residues 27–53 of the T. acidophilum sequence are shown, along with the aligned regions of the other E2 sequences, and the lysine residue that is lipoylated in each sequence is marked by a (*). 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 01234567 Molar Ratio of E1:E2 Activity (AU/min) Fig. 2. Whole complex activity as a function of the E1 : E2 ratio. Lipoylated E2 and E3 were mixed in a molar ratio of 1 : 1 (E2 poly- peptide to E3 dimer), and to this mixture varying amounts of E1 were added. The mixture was then incubated at 55 °C for 10 min in the presence of 2 m M MgCl 2 and 0.2 mM TPP, following which samples were taken and assayed for whole complex activity (expressed as A (absorbance units).min )1 ) using the substrate 3-methyl-2-oxopentanoate. The molar ratio of E1 : E2 is expressed as E1 a 2 b 2 tetramer to E2 polypeptide. 2-Oxoacid dehydrogenase complex from the Archaea C. Heath et al. 5408 FEBS Journal 274 (2007) 5406–5415 ª 2007 The Authors Journal compilation ª 2007 FEBS OADHC activity The whole complex activity detected with 4-methyl-2- oxopentanoate, 3-methyl-2-oxopentanoate, 3-methyl-2- oxobutanoate, and pyruvate, but not with 2-oxogluta- rate, is entirely consistent with the substrate specificity observed for the recombinant E1 enzyme in the absence of the other complex components [14]. The relative activities are given in Table 1, along with those for the isolated E1 component and for the BCOADHC from bovine kidney. Using 3-methyl-2-oxopentanoate as substrate, the assembled T. acidophilum complex exhibited a hyperbolic dependence of velocity on the 2-oxoacid concentration, with K M ¼ 250 (± 40) lm and V max ¼ 4 (± 0.1) lmolÆmin )1 Æmg )1 (E2). This spe- cific activity is comparable to that of the Bacillus stearothermophilus PDHC [8–10 lmo1Æmin )1 Æmg )1 (E2) [17]. Relative molecular mass of the E2 core and the assembled complex Analytical centrifugation and DLS were used to deter- mine the M r of both the E2 core and the assembled complex. E2 core Sedimentation velocity analysis was carried out at 40 °C as described in Experimental procedures on recombinant E2 that was 50% lipoylated and had been purified by His BindÒ (Novagen, Merck Biosciences Ltd., Nottingham, UK) and then anion-exchange chro- matography. The sedimentation coefficient distribu- tions showed a major symmetrical peak constituting 95% of the total protein (Fig. 3). Analysis of the data by direct fitting using the finite element solution of the Lamm equation in sedfit [18], gave a sedimentation coefficient (corrected to water at 20°C and infinite dilution; s° 20,w ) ¼ 27 (± 1) S (Svedberg units; 1S = 10 )13 seconds) and an M r ¼ 1.1 (± 0.1) · 10 6 . From the E2 polypeptide M r of 46 276, the E2 protein there- fore comprises 23.8 polypeptides; that is, it assembles into a core of 24 subunits possessing octahedral sym- metry. DLS analysis at 55 °C, the optimum growth temper- ature of T. acidophilum, gave an E2 M r value ¼ 1.0 (± 0.1) · 10 6 , in good agreement with the value deter- mined by analytical centrifugation. This E2 sample was nonlipoylated, demonstrating that lipoylation is not a prerequisite for assembly of the core. Interest- ingly, the M r of the same E2 at 25 °C, with no prior heat treatment at 55 °C, was estimated by DLS to be 55 000, close to the sequence-predicted monomer value (46 276); this indicates that the assembly of E2 into a 24-mer core is temperature dependent. The assembled complex Sedimentation velocity analysis at 40 °C of assembled complex, created by mixing the components in a 3 : 1 : 0.1 (E1 a 2 b 2 : E2 polypeptide: E3 a 2 ) molar ratio, revealed three discrete protein peaks with s° 20,w values of 50S (24% of total protein), 19S (8%) and 6S Table 1. Substrate specificities of the T. acidophilum 2-oxoacid dehydrogenase complex. The T. acidophilum assembled complex was assayed as described in Experimental procedures. For 3-methyl-2-oxopentanoate, V max ¼ 4 lmolÆmin –1 Æmg –1 (E2), and K M ¼ 250 lM; other activities were determined at saturating sub- strate concentrations. 2-Oxoacid substrate Ratio of specific activities T. acidophilum Bovine kidney BCOADHC b Assembled complex E1 a 3-methyl-2-oxopentanoate 1.0 1.0 1.0 4-methyl-2-oxopentanoate 0.5 0.3 1.5 3-methyl-2-oxobutyrate 0.9 0.6 0.2 pyruvate 0.2 0.2 0.4 2-oxoglutarate 0 0 0 a Data from [14]. b Data from [25]. Fig. 3. Sedimentation coefficient distributions of the E2 protein. Sedimentation velocity of lipoylated E2 protein was carried out at 40 °C as described in Experimental procedures. The sedimentation velocity distribution was obtained by the c(s) method [34] using the program SEDFIT, and the data were then directly fitted using the finite element solution of the Lamm equation in SEDFIT [18] to give values of the sedimentation coefficient in the buffer of sedimenta- tion [20 m M Tris ⁄ HCl (pH 8.5), 0.4 M NaCl and 1 mM phenyl- methanesulfonyl fluoride]. C. Heath et al. 2-Oxoacid dehydrogenase complex from the Archaea FEBS Journal 274 (2007) 5406–5415 ª 2007 The Authors Journal compilation ª 2007 FEBS 5409 (68%), respectively. Unfortunately, attempts to fit these data using the finite element solution of the Lamm equation in sedfit [18] were unsuccessful, thus failing to give M r values for these proteins. Using the relationship that s µ (M r ) 2 ⁄ 3 for globular proteins, and incorporating the values of M r ¼ 6.1 · 10 6 and s° 20,w ¼ 60S for the E. coli PDHC [19], an approxi- mate value of M r ¼ 4.7 · 10 6 was calculated for what is assumed to be an assembled Thermoplasma complex of 50S. DLS is better able to deal with the polydispersity in the sample of assembled complex. Analysis of the auto- correlation curve using the volume distribution algo- rithm gives an M r of 5.0 (± 0.2) · 10 6 for the complex at 55 °C, in close agreement with that estimated by sed- imentation velocity experiments described above. To investigate the identity of the species in the assembled complex mixture, analytical gel filtration on Superdex TM 200 (GE Healthcare, Chalfont St Giles, UK) was carried out. Three protein species were observed (Fig. 4), one eluting in the column exclusion volume (where M r > 1.3 · 10 6 ), one of M r  160 000, and a third minor peak at M r  100 000. Whole com- plex activity was only detected in fractions at the exclusion volume, whereas E1 activity was detected in the second peak and E3 in the third. SDS ⁄ PAGE of protein from the exclusion volume peak revealed three protein bands, corresponding to E1a,E1b and E2 ⁄ E3 (the last two proteins run together due to their similar polypeptide size). The second peak contained predomi- nantly excess E1 protein, whilst the third minor peak comprised E1 and E3 protein that had not been com- pletely separated. In repeat experiments, uncomplexed E3 was not always detectable. Thus it is concluded that the species of M r ¼ 5 · 10 6 is assembled, catalytically active whole complex. Given that the molar ratio of E2 : E3 (E2 polypeptide: E3 a 2 ) is likely to be 1 : 0.1 at a maximum (that is, the ratio of mixed components), then an M r of 5.0 · 10 6 would correspond to an approximate stoichiometry (E1 a 2 b 2 : E2 polypeptide: E3 a 2 ) of 1 : 1 : 0.1. The other major peak in the sedimentation velocity analysis (6S) is pre- sumably excess E1. The minor 19S species remains unidentified, although it should be noted that a 19.8S species was also observed in the assembled E. coli PDHC, where it was concluded that it had the proper- ties of an incomplete aggregate of the component enzymes based on a trimer of the E2 chain [19]. Discussion As described in the introduction, the genomes of aero- bic archaea contain four genes whose translated pro- tein products show sequence identities to the E1a, E1b, E2 and E3 components of the 2-oxoacid dehydro- genase multi-enzyme complexes of aerobic bacteria and eukarya [reviewed in 5]. Furthermore, the genes appear to be arranged in an operon, transcriptional evidence for which has recently been gained in the hal- ophilic archaeon, H. volcanii [20]. The presence of 0 2 4 6 8 10121416 Elution volume (mL) 1 3 1 3 1 2 3 1 2 3 Enzyme activity AB Complex E1 E3 45.0 31.0 21.5 66.2 97.4 116.3 210.0 M 1 2 3 E2 and E3 E1α E1β A 280 Fig. 4. Analytical gel filtration and SDS ⁄ PAGE analysis of assembled OADHC. Assembled OADHC was created by mixing the recombinant protein components in a 3 : 1 : 0.1 (E1 a 2 b 2 : E2 polypeptide: E3 a 2 ) molar stoichiometry, followed by incubation at 55 °C for 10 min in the presence of 2 m M MgCl 2 and 0.2 mM TPP. (A) Gel filtration of the assembled complex was carried out at 25 °C as described in Experimental procedures, and the fractions were assayed for whole complex, E1 and E3 catalytic activity. Active fractions are indicated by bars above the elution profile. (B) Fractions were also analyzed by SDS ⁄ PAGE. M, marker proteins with their M r values given in kDa. Lanes 1–3 correspond to protein peaks 1–3 from the gel filtration; the identity of the component polypeptides is indicated alongside the gel. 2-Oxoacid dehydrogenase complex from the Archaea C. Heath et al. 5410 FEBS Journal 274 (2007) 5406–5415 ª 2007 The Authors Journal compilation ª 2007 FEBS these genes is unexpected, as all the archaea possess ferredoxin oxidoreductases that catalyze the equivalent oxidation of the same 2-oxoacids as those used by the bacterial and eukaryal dehydrogenase complexes; moreover, activity of the 2-oxoacid dehydrogenase complexes has never been detected in any archaeon. The obvious question raised by these observations is whether or not the archaeal ‘OADHC’ genes actually encode functional proteins that assemble into a multi- enzyme complex. The E1 component defines the sub- strate specificity of the whole complex, and we have previously reported heterologous expression of the a 2 b 2 E1 enzyme from T. acidophilum and shown it to be catalytically active with the branched-chain 2-oxo- acids and pyruvate [14]. In the current paper, the E2 and E3 genes have also been successfully cloned and expressed as soluble proteins in E. coli, and their cata- lytic activities have been demonstrated. Generation of an active E2 enzyme requires lipoyla- tion of a specific lysine residue. In E. coli the lipoyla- tion of its own E2 components can occur via two routes [21,22]. The endogenous pathway involves the covalent attachment of a C-8 intermediate of fatty acid biosynthesis to the target lysine on E2 by enzyme LipB; subsequently, LipA catalyzes the incorporation of sul- fur atoms to generate the lipoic acid moiety. However, if lipoic acid is supplied in the growth medium, E. coli preferentially uses its exogenous pathway, which employs lipoate protein ligase A; this enzyme catalyzes the adenylation of lipoic acid after uptake into the cell and its subsequent transfer to the E2 lysine residue. Knowing that the lipoylation process can take place across the species barrier, albeit with varying efficien- cies [2, and references therein], lipoylation of recombi- nantly expressed Thermoplasma E2 was tested and optimized. Up to 50% lipoylation was achieved when the rate and level of expression in E. coli was slowed down by decreasing the growth temperature and avoid- ing induction, whilst at the same time supplementing the growth medium with lipoic acid. Clearly, therefore, the E. coli machinery is able to rec- ognize the lipoyl-domain of the Thermoplasma E2 enzyme, the target lysine of which is flanked by D and V residues, as it is in the E2 protein of the E. coli OGDHC and of the Bacillus subtilis BCOADHC (Fig. 1). How- ever, in addition to the identity of the neighbouring residues, it is the exact positioning of the lysine in the lipoyl domain that is fundamental to target lysine recog- nition [23], implying that the fold of the Thermoplasma enzyme has been conserved in this region. Whilst the Thermoplasma recombinant E2 has not been assayed for catalytic activity in isolation, assem- bly of the whole active complex from its individually expressed components shows that it is indeed a func- tional enzyme. This assembly process, studied by both analytical ultracentrifugation and DLS, has been dem- onstrated to involve the formation of a 24-mer E2 core, which binds E1 and E3 components to give a complex that has the same substrate specificity as that determined for the isolated E1 enzyme; namely, it is a branched-chain 2-oxoacid dehydrogenase complex that is also active with pyruvate. An E2 core that comprises an assembly of 24 polypeptide chains is consistent with other branched-chain 2-oxoacid multi-enzyme com- plexes from bacteria and eukarya [24,25], some of which also have activity with pyruvate. Additionally, the E1 subunit in those branched-chain complexes is also an a 2 b 2 oligomer [24,25]. What is particularly interesting is that the assembly of the E2 core in the Thermoplasma enzyme is temperature dependent, incu- bation to at least 40 °C (the temperature of the ultra- centrifugation) being required. The data in Fig. 2 show that overall complex activ- ity increased linearly with the ratio of E1 : E2 mixed together, as was found with the E. coli PDHC for example [26], until maximal activity was achieved at a mixing ratio of 2 : 1. However, both ultracentrifuga- tion and DLS estimate the M r of the assembled com- plex to be around 5 · 10 6 , closely fitting the tentative conclusion of an E1(a 2 b 2 ) ⁄ E2 ⁄ E3(a 2 ) stoichiometry of 1 : 1 : 0.1. These experiments indicate that not all the E1 molecules may be tightly bound to the E2 core and ⁄ or that assembly might be substrate enhanced. Whilst uncertainties remain over the exact subunit composition of the assembled complex, the important conclusion from our data is that the four OADHC ORFs in the archaeon T. acidophilum encode the com- ponents of a functional 2-oxoacid dehydrogenase multi-enzyme complex, the first to be identified in this domain of life. Thus, OADHCs were probably present in the common ancestor to the Bacteria and Archaea, and have been retained in aerobic members of each domain. The Thermoplasma enzyme possesses catalytic activity with branched-chain 2-oxo acids and pyruvate, but it remains to be established whether other archaeal OADHCs have the same or different substrate specific- ities. However, whatever specificity is found, the physi- ological role of these OADHCs in the archaea remains a mystery given the presence of active 2-oxoacid FORs that catalyze the equivalent chemical reactions. In studies of the OADHC from the halophilic archa- eon H. volcanii, we could find no growth substrate that would induce the expression of OADHC activity [20], nor was any physiological defect apparent in this organism when the E3 gene was inactivated by inser- tional mutagenesis [27]. Interestingly, Wanner & Soppa C. Heath et al. 2-Oxoacid dehydrogenase complex from the Archaea FEBS Journal 274 (2007) 5406–5415 ª 2007 The Authors Journal compilation ª 2007 FEBS 5411 [28] have found an additional gene cluster in H. volca- nii comprising three genes that would appear to code for OADHC E1a and E1b subunits, and an unat- tached lipoyl domain; however, no genes for a com- plete E2 or an E3 were present. Evidence for a function during nitrate-respirative growth on Casami- no acids was presented, although the metabolic sub- strate could not be identified. In conclusion, with respect to the archaeal four-gene OADHC cluster that we have studied in this paper, it is highly unlikely that the genes would have been retained in a highly sophisticated and functional state without a physiological role. We suggest that proteo- mic studies on this thermoacidophile need to be insti- tuted to reveal this role. Experimental procedures Materials Bacteriological media were purchased from Sigma-Aldrich (Poole, UK) or Fisher Scientific (Loughborough, UK). Expression vector pET28a, E. coli expression strain BL21(DE3), BugBuster Protein Extraction Reagent and BenzonaseÒ nuclease were purchased from Novagen-Merck (Nottingham, UK). E. coli JM109 cells, pGEM-T vector, restriction endonucleases, T4 DNA ligase and Taq polymer- ase were purchased from Promega (Southampton, UK). Vent DNA polymerase was from New England Biolabs (Hitchin, UK). Lipoic acid, phenylmethanesulfonyl fluoride and antibiotics were purchased from Sigma-Aldrich. SDS ⁄ PAGE molecular mass markers were from Bio-Rad (Hemel Hempstead, UK). Plasmids pET19b-E1a and pET28a-E1b, which, respec- tively, coexpress the a and b subunits of the T. acidophilum E1, have been described previously [14]. Bioinformatics The putative OADHC operon was identified in the T. aci- dophilum DSM1728 genome from the ENTREZ Nucleotide database (http://www.3.ncbi.nlm.nih.gov). E1a: Ta1438; E1b: Ta1437; E2: Ta1436; and E3: Ta1435. Recombinant DNA techniques E2 and E3 gene amplification Preparation of genomic DNA from T. acidophilum strain DSM 1728 has been described previously [29]. The E2 and E3 genes were PCR-amplified from this genomic DNA using primers that engineered restriction sites into the 5¢- and 3¢ ends of the gene products: NdeI and XhoI for the E2 gene, and NheI and EcoRI for E3. Oligonucleotides were as follows (restriction sequences are underlined): E2 forward: CGC CATATGTACGAATTCAAACTGC CAGACATAGG E2 reverse: CCG CTCGAGTCAGATCTCGTAGAT TATAGCGTTCGG E3 forward: CTACGAGA GCTAGCATGTACGATGC AATAATAATAGGTTC E3 reverse: TTTAAAAATG GAATTCAATGAGAT GGT. PCR amplification was carried out using Vent polymer- ase, and A-tails were added to the products with Taq poly- merase. Both genes were then separately cloned into the intermediate vector pGEM-T using T4 DNA ligase, and the clones were amplified in E. coli strain JM109 grown in Luria–Bertani LB media [1% (w ⁄ v) tryptone, 1% (w ⁄ v) NaCl, 0.5% (w ⁄ v) yeast extract] supplemented with carben- icillin (50 lgÆmL )1 ). Plasmids were extracted using the BD Biosciences (Palo Alto, CA) NucleoSpin Plasmid kit, and the genes were sequenced for fidelity. The E2 and E3 genes were then excised from pGEM-T, using the appropriate restriction endonucleases, purified by electrophoresis in a 0.8% (w ⁄ v) agarose gel, and extracted from the gel using the Qiagen (Hilden, Germany) Qiaex II Gel Extraction kit. Construction of expression vectors pET28a-E2 and pET28a-E3 Expression vector pET28a was prepared for recombinant ligation by NdeI ⁄ XhoI restriction endonuclease digestion for E2, and NheI ⁄ EcoRI digestion for E3, and purified by gel electrophoresis and gel extraction as already described. E2 and E3 genes were separately ligated into the vectors using T4 DNA ligase, generating the recombinant expression vec- tors pET28a-E2 and pET28a-E3. This pET28a vector thus introduced a 20- and a 23-amino acid sequence at the N-ter- mini of the E2 and E3 recombinant protein products, respec- tively, with each containing a 6-histidine tag sequence. Expression and purification of OADHC components E2 expression and purification Two different methods of expression were used. Several col- onies of E. coli strain BL21(DE3), freshly transformed with plasmid pET28a-E2, were picked from LB agar plates and used to inoculate 2 L of LB medium supplemented with kanamycin (30 l gÆmL )1 ) and 0.2 mmdl-lipoic acid. Incu- bation was at 30 °C for 20 h in darkness, with no induction by IPTG, after which cells were harvested by centrifugation at 6000 g. Alternatively, an overnight culture (20 mL, A 600 0.6) of freshly transformed E. coli BL21(DE3)pLysS cells was used to inoculate 1 L of LB medium supple- mented with kanamycin (30 lgÆmL )1 ). After induction with IPTG (at A 600 0.6) and subsequent overnight incubation at 37 °C, cells were harvested as described before. The former 2-Oxoacid dehydrogenase complex from the Archaea C. Heath et al. 5412 FEBS Journal 274 (2007) 5406–5415 ª 2007 The Authors Journal compilation ª 2007 FEBS method allows production of recombinant E2 of which up to 50% is lipoylated, whereas in the latter method only 5% of the E2 is lipoylated. Purification of the recombinant E2 was carried out at 25 °C, unless otherwise stated. Samples were analyzed by SDS ⁄ PAGE on a 10% (w ⁄ v) polyacrylamide gel at each step of the purification, and protein concentrations were determined from A 280 values. Frozen cells were disrupted by resuspension in BugBuster (5 mLÆg )1 wet cells) supple- mented with Benzonase nuclease (1 lLÆmL )1 ), incubated on ice with gentle agitation for 30 min, and centrifuged at 16 000 g for 20 min at 4 °C to pellet cell debris. The solu- ble cell extract was subjected to His BindÒ Resin chroma- tography, and fractions containing E2 protein were pooled and dialyzed overnight into 20 mm Tris ⁄ HCl buffer, pH 9.0, 10% (v ⁄ v) glycerol. The protein was then subjected to anion exchange chromatography on an Amersham Bio- sciences (Chalfont St Giles, UK) A ¨ kta FPLC system, using a 5 mL Q-Sepharose Hi-Trap column equilibrated with 50 mm Tris ⁄ HCl buffer, pH 8.5. Protein was eluted over a 0–0.8 m gradient of NaCl in the same buffer, at a flow rate of 1 mLÆmin )1 over 60 min. Fractions containing E2 were stored at 4 °C in the elution buffer supplemented with 1mm phenylmethanesulfonyl fluoride. E3 expression and purification For expression of the E3 enzyme, a 20 mL overnight cul- ture of transformed BL21(DE3) (A 600 0.6) was used to inoculate 1 L of LB medium supplemented with kanamycin (30 lgÆmL )1 ). Cells were incubated at 37 °C until 5 h after the A 600 had reached 0.6 (with no induction by IPTG), and were then collected by centrifugation. Cell disruption was carried out as described for the purification of recombinant E2. The soluble cell extract was subjected to heat precipita- tion at 65 °C for 5 min, and precipitated material removed by centrifugation at 16 000 g for 20 min at 4 °C. E3 in the remaining soluble fraction was purified by His-Bind Resin chromatography, dialyzed overnight into 20 mm Tris ⁄ HCl buffer, pH 8.4, and then stored at 4 °C. Assembly of the OADHC multi-enzyme complex OADHC was assembled in vitro by mixing together recom- binant E1a,E1b, E2 and E3 proteins at 55 °C for 0–1 h. Molar ratios to enzyme E2 varied from 0.5 to 6.0 (E1a 2 b 2 ) and 0.01–1.0 (E3a 2 ). Each assembled complex was assayed for overall complex activity as described below. SDS-PAGE Analysis of protein purity and determination of polypeptide M r values were carried out by SDS ⁄ PAGE in a resolving gel containing 10% (w ⁄ v) acrylamide [30]. Enzyme assays E1 enzymic activity was assayed spectrophotometrically by following the 2-oxoacid-dependent reduction of 2,6-dichlo- rophenolindophenol (DCPIP) at 595 nm [31]. Assays were carried out at 55 °Cin20mm potassium phosphate (pH 7.0), 2 mm MgCl 2 and 0.2 mm TPP. Buffer and recom- binant E1a 2 b 2 enzyme were pre-incubated at 55 °C for 10 min; 50 lm DCPIP was then added and the assay started by the addition of the 2-oxoacid substrate (pyru- vate, 2-oxoglutarate, 4-methyl-2-oxopentanoate, 3-methyl-2- oxopentanoate or 3-methyl-2-oxobutanoate). E3 was assayed at 55 °Cin50mm EPPS buffer (pH 8.0) containing 0.4 mm dihydrolipoamide and 1 mm NAD + . The reaction, in a final volume of 1 mL, was started by the addition of enzyme, and activity was monitored by measur- ing the production of NADH at 340 nm. Overall complex activity was assayed at 55 °Cin50mm potassium phosphate buffer (pH 7.0) containing 2.5 mm NAD + ,1mm MgCl 2 , 0.2 mm TPP, 0.13 mm CoASH and 2.6 mm cysteine-HCl [32]. Buffer and assembled enzyme complex were pre-incubated at 55 °C for 10 min to allow binding of TPP to E1. The assay, in a final volume of 1 mL, was started by the addition of the 2-oxoacid substrate as for the assay of E1, and OADHC activity was monitored by measuring the production of NADH at 340 nm. Kinetic parameters were determined by the direct linear method of Eisenthal & Cornish-Bowden [33]. Gel filtration Analytical gel filtration was carried out at 25 °C on the Amersham Biosciences A ¨ kta FPLC system, using a Superdex 200 10 ⁄ 300 GL column. Protein standards were: b-amylase (M r ¼ 200 000), alcohol dehydrogenase (150 000), BSA (66 000), carbonic anhydrase (29 000) and cytochrome c (12 400). For analysis of E3, the column was equilibrated with 20 mm sodium phosphate (pH 7.0), 0.1 m NaCl and 10% (v ⁄ v) glycerol; peak fractions were assayed for E3 enzymic activity. Analysis of assembled complex was carried out in 20 mm sodium phosphate (pH 7.0), 2 mm MgCl 2 , 0.1 m NaCl and 10% (v ⁄ v) glycerol. Peak fractions were assayed for E1, E3 and OADHC activity. Analytical ultracentrifugation All analytical ultracentrifugation experiments were carried out on a Beckman XL-A analytical ultracentrifuge (Beck- man-Coulter, CA). Sedimentation velocity experiments were carried out at 15 000 r.p.m., and cells were scanned every 5 min at 280 nm. For sedimentation of E2, the buffer was 20 mm Tris ⁄ HCl (pH 8.5), 0.4 m NaCl and 1 mm phen- ylmethanesulfonyl fluoride; for whole complex, the buffer was the same as that used in the analytical gel filtration C. Heath et al. 2-Oxoacid dehydrogenase complex from the Archaea FEBS Journal 274 (2007) 5406–5415 ª 2007 The Authors Journal compilation ª 2007 FEBS 5413 analysis, 20 mm sodium phosphate (pH 7.0), 2 mm MgCl 2 , 100 mm NaCl and 10% (v ⁄ v) glycerol. The set temperature on the centrifuge was 40 °C, and the solution densities were directly measured at this temperature using an Anton-Paar DMA 5000 high-precision density-meter. Sedimentation velocity distributions were obtained by the c(s) method [34] using the program sedfit. Data were then directly fitted using the finite element solution of the Lamm equation in sedfit [18] to give values of the sedimentation coefficient and of M r . Dynamic light scattering All DLS measurements were performed using a Zetasizer Nano S from Malvern Instruments Ltd. (Malvern, UK). Prior to DLS measurements, protein solutions (1 mgÆmL )1 )in20mm sodium phosphate buffer, pH 7.5, 10% glycerol (v ⁄ v) and 0.1 m NaCl were filtered through a 0.02 lm membrane filter (WhatmanÒ Anotop 10, Fisher Scientific, Loughborough, UK) to remove dust particles. However, it was found necessary to filter enzyme E3 through a 0.22 lm membrane filter (Millipore, Watford, UK). DLS measurements were carried out at 25 °Cor 55 °C. M r values were derived from the measured hydro- dynamic radii using the Protein Utilities feature of the dispersion technology software, version 4.10, supplied with the instrument. Mass spectrometry Determination of the E2 polypeptide mass A 100 pmol sample of recombinant E2 was injected on to a MassPrep on-line desalting cartridge (2.1 · 10 mm) (Waters, Milford, MA), eluted with an increasing acetoni- trile concentration [2 vol acetonitrile + 98 vol aqueous for- mic acid (1%, v ⁄ v) to 98 vol acetonitrile + 2 vol aqueous formic acid (1%, v ⁄ v)] and delivered to an electrospray ion- ization mass spectrometer (LCT, Micromass, Manchester, UK) that had previously been calibrated using myoglobin. An envelope of multiply charged signals was obtained and deconvoluted using maxent1 software to give the molecular mass of the protein. Mapping the lipoylation of E2 Recombinant E2 (50 pmol) was dialyzed against 50 mm ammonium bicarbonate on a VS membrane disc (Millipore) for 30 min. Sequencing grade, modified porcine trypsin (Promega) (60 ng) was added and the sample incubated at 37 °C for 16 h. A portion of the sample was diluted in 5% (v ⁄ v) formic acid and the peptides separated using an Ulti- Mate nanoLC (LC Packings, Amsterdam, the Netherlands) equipped with a PepMap C18 trap and column. The eluant was sprayed into a Q-Star Pulsar XL tandem mass spec- trometer (Applied Biosystems, Foster City, CA) and ana- lyzed in Information Dependent Acquisition mode. The MS ⁄ MS data generated were analyzed using the Mascot search engine (Matrix Science, London, UK), with lipoyla- tion selected as a possible lysine modification. The MS ⁄ MS spectrum corresponding to the modified peptide was also interpreted ‘manually’ using BioAnalyst (Applied Biosys- tems) tools. Acknowledgements MJD and DWH thank the US Air Force Office of Sci- entific Research (Arlington, VA, USA) for generous financial support. CH and MGP gratefully acknowl- edge the receipt of Postgraduate Studentships from the UK Biotechnology and Biological Sciences Research Council and from the University of Bath, respectively. We thank Dr Jean van den Elsen (University of Bath, UK) for allowing us to use the DLS Zetasizer Nano S, and Dr Catherine Botting, BMS Mass Spectrometry and Proteomics Facility, University of St Andrews, UK, for carrying out the MS analyses. References 1 Perham RN (1991) Domains, motifs, and linkers in 2-oxo acid dehydrogenase multienzyme complexes – a paradigm in the design of a multifunctional protein. Biochemistry 30, 8501–8512. 2 Perham RN (2000) Swinging arms and swinging domains in multifunctional enzymes: catalytic machines for multistep reactions. Ann Rev Biochem 69, 961–1004. 3 Perham RN, Jones DD, Chauhan HJ & Howard MJ (2002) Substrate channelling in 2-oxo acid dehydrogenase multienzyme complexes. Biochem Soc Trans 30, 47–51. 4 Izard T, Ævarsson A, Allen MD, Westphal AH, Per- ham RN, de Kok A & Hol WGJ (1999) Principles of quasi-equivalence and Euclidean geometry govern the assembly of cubic and dodecahedral cores of pyruvate dehydrogenase complexes. Proc Natl Acad Sci USA 96, 1240–1245. 5 Danson MJ, Lamble HJ & Hough DW (2007) Central metabolism. In Archaea: Molecular and Cell Biology (Cavicchioli R, ed.). Chapter 12, pp. 260–287. ASM Press, Washington, DC. 6 Kerscher L & Oesterhelt D (1982) Pyruvate – ferredoxin oxidoreductase: new findings on an ancient enzyme. Trends Biochem Sci 7, 371–374. 7 Schut GJ, Menon AL & Adams MWW (2001) 2-Keto acid oxidoreductases from Pyrococcus furiosus and Thermococcus litoralis. Methods Enzymol 331, 144–158. 8 Ragsdale SW (2003) Pyruvate ferredoxin oxidoreductase and its radical intermediate. Chem Rev 103, 2333–2346. 2-Oxoacid dehydrogenase complex from the Archaea C. Heath et al. 5414 FEBS Journal 274 (2007) 5406–5415 ª 2007 The Authors Journal compilation ª 2007 FEBS 9 Plaga W, Lottspeich F & Oesterhelt D (1992) Improved purification, crystallization and primary structure of pyruvate: ferredoxin oxidoreductase from Halobacterium halobium. Eur J Biochem 205, 391–397. 10 Danson MJ, Eisenthal R, Hall S, Kessell SR & Williams DL (1984) Dihydrolipoamide dehydrogenase from halophilic archaebacteria. Biochem J 218, 811–818. 11 Smith LD, Bungard SJ, Danson MJ & Hough DW (1987) Dihydrolipoamide dehydrogenase from the ther- moacidophilic archaebacterium Thermoplasma acidophi- lum. Biochem Soc Trans 15, 1097–1097. 12 Pratt KJ, Carles C, Carne TJ, Danson MJ & Stevenson KJ (1989) Detection of bacterial lipoic acid: a modified gas chromatographic – mass spectrometric procedure. Biochem J 258, 749–754. 13 Jolley KA, Maddocks DG, Gyles SL, Mullan Z, Tang S-L, Dyall-Smith ML, Hough DW & Danson MJ (2000) 2-Oxoacid dehydrogenase multienzyme complexes in the halophilic Archaea? Gene sequences and protein structural predictions. Microbiology 146, 1061–1069. 14 Heath C, Jeffries AC, Hough DW & Danson MJ (2004) Discovery of the catalytic function of a putative 2-oxo- acid dehydrogenase multienzyme complex in the ther- mophilic archaeon Thermoplasma acidophilum. FEBS Lett 577, 523–527. 15 Massey V (1960) Identity of diaphorase and lipoyl dehy- drogenase. Biochim Biophys Acta 37, 314–322. 16 Mann S, Melero CP, Hawksley D & Leeper FJ (2004) Inhibition of thiamin diphosphate dependent enzymes by 3-deazathiamin diphosphate. Org Biomol Chem 2, 1732–1741. 17 Chauhan HJ, Domingo GJ, Jung HI & Perham RN (2000) Sites of limited proteolysis in the pyruvate decar- boxylase component of the pyruvate dehydrogenase multienzyme complex of Bacillus stearothermophilus and their role in catalysis. Eur J Biochem 267, 7158–7169. 18 Schuck P (1998) Sedimentation analysis of non-interact- ing and self-associating solutes using numerical solu- tions to the Lamm equation. Biophys J 75, 1503–1512. 19 Danson MJ, Hale G, Johnson P, Perham RN, Smith J & Spragg P (1979) Molecular weight and symmetry of the pyruvate dehydrogenase multienzyme complex of Escherichia coli. J Mol Biol 129, 603–617. 20 Al-Mailem DM, Hough DW & Danson MJ (2007) The 2-oxoacid dehydrogenase complex from Haloferax vol- canii. Extremophiles, doi: 10.1007/s00792-007-0091-0. 21 Miller JR, Busby RW, Jordan SW, Cheek J, Henshaw TF, Ashley GW, Broderick JB, Cronan JE & Marletta MA (2000) Escherichia coli LipA is a lipoyl synthase: in vitro biosynthesis of lipoylated pyruvate dehydrogenase complex from octanoyl-acyl carrier protein. Biochemistry 39, 15166–15178. 22 Morris T, Reed K & Cronan J Jr (1995) Lipoic acid metabolism in Escherichia coli: the lplA and lipB genes define redundant pathways for ligation of lipoyl groups to apoprotein. J Bacteriol 177 , 1–10. 23 Wallis NG & Perham RN (1994) Structural dependence of post-translational modification and reductive acetyla- tion of the lipoyl domain of the pyruvate dehydrogenase multienzyme complex. J Mol Biol 236, 209–216. 24 Ævarsson A, Seger K, Turley S, Sokatch JR & Hol WGJ (1999) Crystal structure of 2-oxoisovalerate dehydroge- nase and the architecture of 2-oxo acid dehydrogenase multienzyme complexes. Nature Struct Biol 6, 785–792. 25 Pettit FH, Yeaman SJ & Reed LJ (1978) Purification and characterisation of branched-chain a-ketoacid dehy- drogenase complex of bovine kidney. Proc Natl Acad Sci USA 75, 4881–4885. 26 Bates DL, Danson MJ, Hale G, Hooper EA & Perham RN (1977) Self-assembly and catalytic activity of the pyruvate dehydrogenase multienzyme complex of Escherichia coli. Nature 268, 313–316. 27 Jolley KA, Rapaport E, Hough DW, Danson MJ, Woods WG, Dyal I & Smith ML (1996) Dihydrolipoa- mide dehydrogenase from the halophilic archaeon Halo- ferax volcanii: homologous overexpression of the cloned gene. J Bacteriol 178, 3044–3048. 28 Wanner C & Soppa J (2002) Functional role for a 2-oxo acid dehydrogenase in the halophilic archaeon Haloferax volcanii. J Bacteriol 184, 3114–3121. 29 Sambrook J & Russell DW (2001) Molecular Cloning: A Laboratory Manual, 3rd edn. Cold Spring. Harbour Laboratory Press, Cold Spring Harbour, NY. 30 Laemmli UK (1970) Cleavage of structural proteins dur- ing the assembly of the head of bacteriophage T4. Nat- ure 227, 680–685. 31 Lessard IAD & Perham RN (1994) Expression in Esc- herichia coli of genes encoding the E1a and E1b subun- its of the pyruvate-dehydrogenase complex of Bacillus stearothermophilus and assembly of a functional E1 Component (a 2 b 2 ) in-vitro. J Biol Chem 269, 10378– 10383. 32 Domingo GJ, Chauhan HJ, Lessard IAD, Fuller C & Perham RN (1999) Self-assembly and catalytic activity of the pyruvate dehydrogenase multienzyme complex from Bacillus stearothermophilus. Eur J Biochem 266, 1136–1146. 33 Eisenthal R & Cornish-Bowden A (1974) The direct linear plot. A new graphical procedure for estimating enzyme kinetic parameters. Biochem J 139, 715–720. 34 Schuck P (2000) Size distribution analysis of macromole- cules by sedimentation velocity ultracentrifugation and Lamm equation modeling. Biophys J 78, 1606–1619. C. Heath et al. 2-Oxoacid dehydrogenase complex from the Archaea FEBS Journal 274 (2007) 5406–5415 ª 2007 The Authors Journal compilation ª 2007 FEBS 5415 . The 2-oxoacid dehydrogenase multi-enzyme complex of the archaeon Thermoplasma acidophilum ) recombinant expression, assembly and characterization Caroline. In the current paper, we report the clon- ing and expression of the E2 and E3 genes of the same operon from T. acidophilum, and the in vitro assembly and