Mutants of Streptomyces clavuligerus with disruptions in different genes for clavulanic acid biosynthesis produce large amounts of holomycin: possible cross-regulation of two unrelated secondary metabolic pathways

A Streptomyces clavuligerus ccaR::aph strain, which has a disruption in the regulatory gene ccaR, does not produce cephamycin C or clavulanic acid, but does produce a bioactive compound that was identified as holomycin by high-performance liquid chromatography (HPLC) and infrared and mass spectrometry. S. clavuligerus strains with disruptions in different genes of the clavulanic acid pathway fall into three groups with respect to holomycin biosynthesis. (i) Mutants with mutations in the early steps of the pathway blocked in the gene ceaS (pyc) (encoding carboxyethylarginine synthase), bls (encoding a beta-lactam synthetase), or open reading frame 6 (ORF6; coding for an acetyltransferase of unknown function) are holomycin nonproducers. (ii) Mutants blocked in the regulatory gene ccaR or claR or blocked in the last gene of the pathway encoding clavulanic acid reductase (car) produce holomycin at higher levels than the wild-type strain. (iii) Mutants with disruption in cyp (coding for cytochrome P450), ORF12, and ORF15, genes that appear to be involved in the conversion of clavaminic acid into clavaldehyde or in secretion steps, produce up to 250-fold as much holomycin as the wild-type strain. An assay for holomycin synthetase was developed. This enzyme forms holomycin from holothin by using acetyl coenzyme A as an acetyl group donor. The holomycin synthase activities in the different clavulanic acid mutants correlate well with their production of holomycin.     

Induction of holomycin production and complex metabolic changes by the argR mutation in Streptomyces clavuligerus NP1

In bacteria, arginine biosynthesis is tightly regulated by a universally conserved regulator, ArgR, which regulates the expression of arginine biosynthetic genes, as well as other important genes. Disruption of argR in Streptomyces clavuligerus NP1 resulted in complex phenotypic changes in growth and antibiotic production levels. To understand the metabolic changes underlying the phenotypes, comparative proteomic studies were carried out between NP1 and its argR disruption mutant (designated CZR). In CZR, enzymes involved in holomycin biosynthesis were overexpressed; this is consistent with its holomycin overproduction phenotype. The effects on clavulanic acid (CA) biosynthesis are more complex. Several proteins from the CA cluster were moderately overexpressed, whereas several proteins from the 5S clavam biosynthetic cluster and from the paralog cluster of CA and 5S clavam biosynthesis were severely downregulated. Obvious changes were also detected in primary metabolism, which are mainly reflected in the altered expression levels of proteins involved in acetyl-coenzyme A (CoA) and cysteine biosynthesis. Since acetyl-CoA and cysteine are precursors for holomycin synthesis, overexpression of these proteins is consistent with the holomycin overproduction phenotype. The complex interplay between primary and secondary metabolism and between secondary metabolic pathways were revealed by these analyses, and the insights will guide further efforts to improve production levels of CA and holomycin in S. clavuligerus.     

Thioredoxin-thioredoxin reductase system of Streptomyces clavuligerus: sequences, expression, and organization of the genes.

The genes that encode thioredoxin and thioredoxin reductase of Streptomyces clavuligerus were cloned, and their DNA sequences were determined. Previously, we showed that S. clavuligerus possesses a disulfide reductase with broad substrate specificity that biochemically resembles the thioredoxin oxidoreductase system and may play a role in the biosynthesis of beta-lactam antibiotics. It consists consists of two components, a 70-kDa NADPH-dependent flavoprotein disulfide reductase with two identical subunits and a 12-kDa heat-stable protein general disulfide reductant. In this study, we found, by comparative analysis of their predicted amino acid sequences, that the 35-kDa protein is in fact thioredoxin reductase; it shares 48.7% amino acid sequence identity with Escherichia coli thioredoxin reductase, the 12-kDa protein is thioredoxin, and it shares 28 to 56% amino acid sequence identity with other thioredoxins. The streptomycete thioredoxin reductase has the identical cysteine redox-active region--Cys-Ala-Thr-Cys--and essentially the same flavin adenine dinucleotide- and NADPH dinucleotide-binding sites as E. coli thioredoxin reductase and is partially able to accept E. coli thioredoxin as a substrate. The streptomycete thioredoxin has the same cysteine redox-active segment--Trp-Cys-Gly-Pro-Cys--that is present in virtually all eucaryotic and procaryotic thioredoxins. However, in vivo it is unable to donate electrons to E. coli methionine sulfoxide reductase and does not serve as a substrate in vitro for E. coli thioredoxin reductase. The S. clavuligerus thioredoxin (trxA) and thioredoxin reductase (trxB) genes are organized in a cluster. They are transcribed in the same direction and separated by 33 nucleotides. In contrast, the trxA and trxB genes of E. coli, the only other organism in which both genes have been characterized, are physically widely separated.     

Increased catalytic performance of the 2-oxoacid dehydrogenase complexes in the presence of thioredoxin, a thiol–disulfide oxidoreductase

Bacterial and mammalian pyruvate and 2-oxoglutarate dehydrogenase complexes undergo an irreversible inactivation upon accumulation of the dihydrolipoate intermediate. The first component of the complexes, 2-oxoacid dehydrogenase, is affected. Addition of thioredoxin protects from this inactivation, increasing catalytic rates and limiting degrees of the substrate transformation to products, acyl-CoA and NADH. Although the redox active cysteines of thioredoxin are essential for its interplay with the complexes, the effects are observed with both dithiol and disulfide forms of the protein. This indicates that thioredoxin affects an SH/S–S component of the system, which is present in the two redox states. The complex-bound lipoate is concluded to be the thioredoxin target, since (i) both dithiol and disulfide forms of the residue are available during the catalytic cycle and (ii) the thioredoxin reaction with the essential SH/S–S group of the terminal component of the complex, dihydrolipoyl dehydrogenase, is excluded. Thus, the thioredoxin disulfide interacts with the dihydrolipoate intermediate, while the thioredoxin dithiol reacts with the lipoate disulfide. Kinetic consequences of such interplay are consistent with the observed thioredoxin effects. Owing to the essential reactivity of the SH/S–S couple in thioredoxin, the thiol–disulfide exchange between thioredoxin and the lipoate residue is easy reversible, providing both protection (by the mixed disulfide formation) and catalysis (by the appropriate lipoate release). In contrast, non-protein SH/S–S compounds prevent the inactivatory action of dihydrolipoate intermediate only at a high excess over the complex-bound lipoate. This interferes with the catalysis-required release of the residue from its mixed disulfide. Therefore, only thioredoxin is capable to ‘buffer' the steady-state concentration of the reactive dithiol. Such action represents a new thioredoxin function, which may be exploited to protect other enzymes with exposed redox-active thiol intermediates.     

Thioredoxins in<Emphasis Type="Italic">Arabidopsis</Emphasis> and other plants

Regulation of disulfide dithiol exchange has become increasingly important in our knowledge of plant life. Initially discovered as regulators of light-dependent malate biosynthesis in the chloroplast, plant thioredoxins are now implicated in a large panel of reactions related to metabolism, defense and development. In this review we describe the numerous thioredoxin types encoded by the Arabidopsis genome, and provide evidence that they are present in all higher plants. Some results suggest cross-talk between thioredoxins and glutaredoxins, the second family of disulfide dithiol reductase. The development of proteomics in plants revealed an unexpectedly large number of putative target proteins for thioredoxins and glutaredoxins. Nevertheless, we are far from a clear understanding of the actual function of each thioredoxin in planta. Although hampered by functional redundancies between genes, genetic approaches are probably unavoidable to define which thioredoxin interacts with which target protein and evaluate the physiological consequences.     

Thioredoxins and glutaredoxins as facilitators of protein folding

Thiol-disulfide oxidoreductase systems of bacterial cytoplasm and eukaryotic cytosol favor reducing conditions and protein thiol groups, while bacterial periplasm and eukaryotic endoplasmatic reticulum provide oxidizing conditions and a machinery for disulfide bond formation in the secretory pathway. Oxidoreductases of the thioredoxin fold superfamily catalyze steps in oxidative protein folding via protein-protein interactions and covalent catalysis to act as chaperones and isomerases of disulfides to generate a native fold. The active site dithiol/disulfide of thioredoxin fold proteins is CXXC where variations of the residues inside the disulfide ring are known to increase the redox potential like in protein disulfide isomerases. In the catalytic mechanism thioredoxin fold proteins bind to target proteins through conserved backbone-backbone hydrogen bonds and induce conformational changes of the target disulfide followed by nucleophilic attack by the N-terminally located low pK(a) Cys residue. This generates a mixed disulfide covalent bond which subsequently is resolved by attack from the C-terminally located Cys residue. This review will focus on two members of the thioredoxin superfamily of proteins known to be crucial for maintaining a reduced intracellular redox state, thioredoxin and glutaredoxin, and their potential functions as facilitators and regulators of protein folding and chaperone activity.     

Redox potential of human thioredoxin 1 and identification of a second dithiol/disulfide motif

Thioredoxin (Trx1) is a redox-active protein containing two active site cysteines (Cys-32 and Cys-35) that cycle between the dithiol and disulfide forms as Trx1 reduces target proteins. Examination of the redox characteristics of this active site dithiol/disulfide couple is complicated by the presence of three additional non-active site cysteines. Using the redox Western blot technique and matrix assisted laser desorption ionization time-of-flight mass spectrometry mass spectrometry, we determined the midpoint potential (E0) of the Trx1 active site (-230 mV) and identified a second redox-active dithiol/disulfide (Cys-62 and Cys-69) in an alpha helix proximal to the active site, which formed under oxidizing conditions. This non-active site disulfide was not a substrate for reduction by thioredoxin reductase and delayed the reduction of the active site disulfide by thioredoxin reductase. Within actively growing THP1 cells, most of the active site of Trx1 was in the dithiol form, whereas the non-active site was totally in the dithiol form. The addition of increasing concentrations of diamide to these cells resulted in oxidation of the active site at fairly low concentrations and oxidation of the non-active site at higher concentrations. Taken together these results suggest that the Cys-62-Cys-69 disulfide could provide a means to transiently inhibit Trx1 activity under conditions of redox signaling or oxidative stress, allowing more time for the sensing and transmission of oxidative signals.     

Sites of interaction of thioredoxin with sorghum NADP-malate dehydrogenase

The activation pathway of the chloroplastic NADP-dependent malate dehydrogenase (MDH) by reduced thioredoxin has been examined using a method based on the mechanism of thiol/disulfide interchanges, i.e. the transient formation of a mixed disulfide between the target and the reductant. This disulfide can be stabilized when each of the partners is mutated in the less reactive cysteine of the disulfide/dithiol pair. As NADP-MDH has two regulatory disulfides per monomer, four different single cysteine mutants were examined, two for the C-terminal bridge and two for the N-terminal bridge. The results clearly show that the nucleophilic attack of thioredoxin on the C-terminal bridge proceeds through the formation of a disulfide with the most external Cys377. The results are less clear-cut for the N-terminal cysteines and suggest that the Cys24-Cys207 disulfide bridge previously proposed to be an intermediary step in MDH activation can form only when the C-terminal disulfide is reduced.     

Gain of function in an ERV/ALR sulfhydryl oxidase by molecular engineering of the shuttle disulfide

The ERV/ALR sulfhydryl oxidase domain is a versatile module adapted for catalysis of disulfide bond formation in various organelles and biological settings. Its four-helix bundle structure juxtaposes a Cys-X-X-Cys dithiol/disulfide motif with a bound flavin adenine dinucleotide (FAD) cofactor, enabling transfer of electrons from thiol substrates to non-thiol electron acceptors. ERV/ALR family members contain an additional di-cysteine motif outside the four-helix-bundle core. Although the location and context of this "shuttle" disulfide differs among family members, it is proposed to perform the same basic function of mediating electron transfer from substrate to the enzyme active site. We have determined by X-ray crystallography the structure of AtErv1, an ERV/ALR enzyme that contains a Cys-X4-Cys shuttle disulfide and oxidizes thioredoxin in vitro, and compared it to ScErv2, which has a Cys-X-Cys shuttle and does not oxidize thioredoxin at an appreciable rate. The AtErv1 shuttle disulfide is in a region of the structure that is disordered and thus apparently mobile and exposed. This feature may facilitate access of protein substrates to the shuttle disulfide. To test whether the shuttle disulfide region is modular and can confer on other enzymes oxidase activity toward new substrates, we generated chimeric enzyme variants combining shuttle disulfide and core elements from AtErv1 and ScErv2 and monitored oxidation of thioredoxin by the chimeras. We found that the AtErv1 shuttle disulfide region could indeed confer thioredoxin oxidase activity on the ScErv2 core. Remarkably, various chimeras containing the ScErv2 Cys-X-Cys shuttle disulfide were found to function efficiently as well. Since neither the ScErv2 core nor the Cys-X-Cys motif is therefore incapable of participating in oxidation of thioredoxin, we conclude that wild-type ScErv2 has evolved to repress activity on substrates of this type, perhaps in favor of a different, as yet unknown, substrate.     

Disulfide bond formation in the Escherichia coli cytoplasm: an in vivo role reversal for the thioredoxins.

Cytoplasmic proteins do not generally contain structural disulfide bonds, although certain cytoplasmic enzymes form such bonds as part of their catalytic cycles. The disulfide bonds in these latter enzymes are reduced in Escherichia coli by two systems; the thioredoxin pathway and the glutathione/glutaredoxin pathway. However, structural disulfide bonds can form in proteins in the cytoplasm when the gene (trxB) for the enzyme thioredoxin reductase is inactivated by mutation. This disulfide bond formation can be detected by assessing the state of the normally periplasmic enzyme alkaline phosphatase (AP) when it is localized to the cytoplasm. Here we show that the formation of disulfide bonds in cytoplasmic AP in the trxB mutant is dependent on the presence of two thioredoxins in the cell, thioredoxins 1 and 2, the products of the genes trxA and trxC, respectively. Our evidence supports a model in which the oxidized forms of these thioredoxins directly catalyze disulfide bond formation in cytoplasmic AP, a reversal of their normal role. In addition, we show that the recently discovered thioredoxin 2 can perform many of the roles of thioredoxin 1 in vivo, and thus is able to reduce certain essential cytoplasmic enzymes. Our results suggest that the three most effective cytoplasmic disulfide-reducing proteins are thioredoxin 1, thioredoxin 2 and glutaredoxin 1; expression of any one of these is sufficient to support aerobic growth. Our results help to explain how the reducing environment in the cytoplasm is maintained so that disulfide bonds do not normally occur.     

Three-dimensional solution structure of the reduced form of Escherichia coli thioredoxin determined by nuclear magnetic resonance spectroscopy

The three-dimensional solution structure of reduced (dithiol) thioredoxin from Escherichia coli has been determined with distance and dihedral angle constraints obtained from 1H NMR spectroscopy. Reduced thioredoxin has a well-defined global fold consisting of a central five-strand beta-sheet and three long helices. The beta-strands are packed in the sheet in the order beta 1 beta 3 beta 2 beta 4 beta 5, with beta 1, beta 3, and beta 2 parallel and beta 2, beta 4, and beta 5 arranged in an antiparallel fashion. Two of the helices connect strands of the beta-sheet: alpha 1 between beta 1 and beta 2 and alpha 2 between beta 2 and beta 3. Strands beta 4 and beta 5 are connected by a short loop that contains a beta-bulge. Strands beta 3 and beta 4 are connected by a long loop that contains a series of turn-like or 3(10) helical structures. The active site Cys-Gly-Pro-Cys sequence forms a protruding loop between strand beta 2 and helix alpha 2. The structure is very similar overall to that of oxidized (disulfide) thioredoxin obtained from X-ray crystal structure analysis but differs in the local conformation of the active site loop. The distance between the sulfurs of Cys 32 and Cys 35 increases from 2.05 A in the disulfide bridge to 6.8 +/- 0.6 A in the dithiol of reduced thioredoxin, as a result of a rotation of the side chain of Cys 35 and a significant change in the position of Pro 34. This conformational change has important implications for the mechanism of thioredoxin as a protein disulfide oxidoreductase.     

DsbD-catalyzed transport of electrons across the membrane of Escherichia coli

Dsb proteins catalyze folding and oxidation of polypeptides in the periplasm of Escherichia coli. DsbC reduces wrongly paired disulfides by transferring electrons from its catalytic dithiol motif (98)CGYC. Genetic evidence suggests that recycling of this motif requires at least three proteins, the cytoplasmic thioredoxin reductase (TrxB) and thioredoxin (TrxA) as well as the DsbD membrane protein. We demonstrate here that electrons are transferred directly from thioredoxin to DsbD and from DsbD to DsbC. Three cysteine pairs within DsbD undergo reversible disulfide rearrangements. Our results suggest a novel mechanism for electron transport across membranes whereby electrons are transferred sequentially from cysteine pairs arranged in a thioredoxin-like motif (CXXC) to a cognate reactive disulfide.     

The mechanism of high Mr thioredoxin reductase from Drosophila melanogaster

Drosophila melanogaster thioredoxin reductase-1 (DmTrxR-1) is a key flavoenzyme in dipteran insects, where it substitutes for glutathione reductase. DmTrxR-1 belongs to the family of dimeric, high Mr thioredoxin reductases, which catalyze reduction of thioredoxin by NADPH. Thioredoxin reductase has an N-terminal redox-active disulfide (Cys57-Cys62) adjacent to the flavin and a redox-active C-terminal cysteine pair (Cys489'-Cys490' in the other subunit) that transfer electrons from Cys57-Cys62 to the substrate thioredoxin. Cys489'-Cys490' functions similarly to Cys495-Sec496 (Sec = selenocysteine) and Cys535-XXXX-Cys540 in human and parasite Plasmodium falciparum enzymes, but a catalytic redox center formed by adjacent Cys residues, as observed in DmTrxR-1, is unprecedented. Our data show, for the first time in a high Mr TrxR, that DmTrxR-1 oscillates between the 2-electron reduced state, EH2, and the 4-electron state, EH4, in catalysis, after the initial priming reduction of the oxidized enzyme (Eox) to EH2. The reductive half-reaction consumes 2 eq of NADPH in two observable steps to produce EH4. The first equivalent yields a FADH--NADP+ charge-transfer complex that reduces the adjacent disulfide to form a thiolate-flavin charge-transfer complex. EH4 reacts with thioredoxin rapidly to produce EH2. In contrast, Eox formation is slow and incomplete; thus, EH2 of wild-type cannot reduce thioredoxin at catalytically competent rates. Mutants lacking the C-terminal redox center, C489S, C490S, and C489S/C490S, are incapable of reducing thioredoxin and can only be reduced to EH2 forms. Additional data suggest that Cys57 attacks Cys490' in the interchange reaction between the N-terminal dithiol and the C-terminal disulfide.     

Thioredoxin-dependent Redox Regulation of Chloroplastic Phosphoglycerate Kinase from Chlamydomonas reinhardtii

In photosynthetic organisms, thioredoxin-dependent redox regulation is a well established mechanism involved in the control of a large number of cellular processes, including the Calvin-Benson cycle. Indeed, 4 of 11 enzymes of this cycle are activated in the light through dithiol/disulfide interchanges controlled by chloroplastic thioredoxin. Recently, several proteomics-based approaches suggested that not only four but all enzymes of the Calvin-Benson cycle may withstand redox regulation. Here, we characterized the redox features of the Calvin-Benson enzyme phosphoglycerate kinase (PGK1) from the eukaryotic green alga Chlamydomonas reinhardtii, and we show that C. reinhardtii PGK1 (CrPGK1) activity is inhibited by the formation of a single regulatory disulfide bond with a low midpoint redox potential (−335 mV at pH 7.9). CrPGK1 oxidation was found to affect the turnover number without altering the affinity for substrates, whereas the enzyme activation appeared to be specifically controlled by f-type thioredoxin. Using a combination of site-directed mutagenesis, thiol titration, mass spectrometry analyses, and three-dimensional modeling, the regulatory disulfide bond was shown to involve the not strictly conserved Cys²²⁷ and Cys³⁶¹. Based on molecular mechanics calculation, the formation of the disulfide is proposed to impose structural constraints in the C-terminal domain of the enzyme that may lower its catalytic efficiency. It is therefore concluded that CrPGK1 might constitute an additional light-modulated Calvin-Benson cycle enzyme with a low activity in the dark and a TRX-dependent activation in the light. These results are also discussed from an evolutionary point of view.     

Protein disulfide-isomerase is a substrate for thioredoxin reductase and has thioredoxin-like activity

We have demonstrated that calf liver protein disulfide-isomerase (Mr 57,000) is a substrate for calf thymus thioredoxin reductase and catalyzes NADPH-dependent insulin disulfide reduction. This reaction can be used as a simple assay for protein disulfide-isomerase during purification in place of the classical method of reactivation of incorrectly oxidized ribonuclease A. Protein disulfide-isomerase contains two redox-active disulfides/molecule which were reduced by NADPH and calf thioredoxin reductase (Km approximately 35 microM). The isomerase was a poor substrate for NADPH and Escherichia coli thioredoxin reductase, but the addition of E. coli thioredoxin resulted in rapid reduction of two disulfides/molecule. Tryptophan fluorescence spectra were shown to monitor the redox state of protein disulfide-isomerase. Fluorescence measurements demonstrated that thioredoxin--(SH)2 reduced the disulfides of the isomerase and allowed the kinetics of the reaction to be followed; the reaction was also catalyzed by calf thioredoxin reductase. Equilibrium measurements showed that the apparent redox potential of the active site disulfide/dithiols of the thioredoxin domains of protein disulfide-isomerase was about 30 mV higher than the disulfide/dithiol of E. coli thioredoxin. Consistent with this, experiments using dithiothreitol or NADPH and thioredoxin reductase-dependent reduction and precipitation of insulin demonstrated differences between protein disulfide-isomerase and thioredoxin, thioredoxin being a better disulfide reductase but less efficient isomerase. Protein disulfide-isomerase is thus a high molecular weight member of the thioredoxin system, able to interact with both mammalian NADPH-thioredoxin reductase and reduced thioredoxin. This may be important for nascent protein disulfide formation and other thiol-dependent redox reactions in cells.     

Characterization of Escherichia coli-Anabaena sp. hybrid thioredoxins

Thioredoxin is a small redox protein with an active-site disulfide/dithiol. The protein from Escherichia coli has been well characterized. The genes encoding thioredoxin in E. coli and in the filamentous cyanobacterium Anabaena PCC 7119 have been cloned and sequenced. Anabaena thioredoxin exhibits 50% amino acid identity with the E. coli protein and interacts with E. coli enzymes. The genes encoding Anabaena and E. coli thioredoxin were fused via a common restriction site in the nucleotide sequence coding for the active site of the proteins to generate hybrid genes, coding for two chimeric thioredoxins. These proteins are designated Anabaena-E. coli (A-E) thioredoxin for the construct with the Anabaena sequence from the N-terminus to the middle of the active site and the E. coli sequence to the C-terminus, and E. coli-Anabaena (E-A) for the opposite construct. The gene encoding the A-E thioredoxin complements all phenotypes of an E. coli thioredoxin-deficient strain, whereas the gene encoding E-A thioredoxin is only partially effective. Purified E-A thioredoxin exhibits a much lower catalytic efficiency with E. coli thioredoxin reductase and ribonucleotide reductase than either E. coli or Anabaena thioredoxin. In contrast, the A-E thioredoxin has a higher catalytic efficiency in these reactions than either parental protein. Reaction with antibodies to E. coli and Anabaena thioredoxins shows that the antigenic determinants for thioredoxin are located in the C-terminal part of the molecule and retain the native conformation in the hybrid proteins.(ABSTRACT TRUNCATED AT 250 WORDS)     

Biosynthesis and molecular genetics of cephamycins

Cephamycin C is produced in a nine steps pathway by the actinomycetes S. clavuligerus and N. lactamdurans. The genes encoding the biosynthesis enzymes are clustered in both microorganisms as well as in the cephabacin producer Lysobacter lactamgenus, a Gram negative bacterium. The clusters of genes include genes encoding enzymes common to the biosynthesis of penicillin and cephalosporin C by the eukaryotic producers Penicillium chrysogenum and Cephalosporiun acremonium and genes for steps specific for the formation of the precursor -aminoadipic acid as well as for the enzymes involved in the late modification of the cephalosporin intermediates of the pathway. Present are also genes for proteins involved in the export and/or resistance to cephamycin C. In S. clavuligerus a gene encoding a regulatory protein controlling the formation of cephamycin C and clavulanic acid is also present in the cluster.     

Transfer of electrons across the cytoplasmic membrane by DsbD, a membrane protein involved in thiol-disulphide exchange and protein folding in the bacterial periplasm

Reduction of non-native protein disulphides in the periplasm of Escherichia coli is catalysed by three enzymes, DsbC, DsbG and DsbE, each of which harbours a catalytic Cys-X-X-Cys dithiol motif. This dithiol motif requires continuous reduction for activity. Genetic evidence suggests that the source of periplasmic reducing power resides within the cytoplasm, provided by thioredoxin (trxA) and thioredoxin reductase (trxB). Cytoplasmic electrons donated by thioredoxin are thought to be transferred into the periplasm via the DsbD membrane protein. To understand the molecular nature of electron transfer, we have analysed the membrane topology of DsbD. DsbD is exported by an N-terminal signal peptide. The N- and C-terminal domains are positioned in the periplasmic space, connected by eight transmembrane segments. Electron transfer was shown to require five cysteine sulphydryl of DsbD. Trans complementation of mutant DsbD molecules revealed intermolecular electron transfer. We discuss a model whereby the membrane-embedded disulphides of DsbD accept electrons from cytoplasmic thioredoxin and transfer them to the C-terminal periplasmic dithiol motif of DsbD.