Lauren Davey, Department of Microbiology and Immunology, Dalhousie University, Halifax, NS, B3H 1X5 Canada Canadian Center for Vaccinology (CCfV), Dalhousie University and the Izaak Walton Killam (IWK) Health Centre, Halifax, NS, CanadaSearch for more papers by this authorAlejandro Cohen, Proteomics and Mass Spectrometry Core Facility, Life Sciences Research Institute, Dalhousie University, Halifax, NS, CanadaSearch for more papers by this authorJason LeBlanc, Department of Microbiology and Immunology, Dalhousie University, Halifax, NS, B3H 1X5 Canada Department of Pathology, Dalhousie University, Halifax, NS, Canada Department of Pathology and Laboratory Medicine, Nova Scotia Health Authority (NSHA), Halifax, NS, CanadaSearch for more papers by this authorScott A. Halperin, Department of Microbiology and Immunology, Dalhousie University, Halifax, NS, B3H 1X5 Canada Canadian Center for Vaccinology (CCfV), Dalhousie University and the Izaak Walton Killam (IWK) Health Centre, Halifax, NS, Canada Department of Pediatrics, Faculty of Medicine, Dalhousie University and the IWK Health Centre, Halifax, NS, CanadaSearch for more papers by this authorSong F. Lee, Corresponding Author Department of Microbiology and Immunology, Dalhousie University, Halifax, NS, B3H 1X5 Canada Canadian Center for Vaccinology (CCfV), Dalhousie University and the Izaak Walton Killam (IWK) Health Centre, Halifax, NS, Canada Department of Pediatrics, Faculty of Medicine, Dalhousie University and the IWK Health Centre, Halifax, NS, Canada Department of Applied Oral Sciences, Faculty of Dentistry, Dalhousie University, Halifax, NS, CanadaFor correspondence. E-mail Song.Lee@Dal.Ca; Tel. (902) 470-7522/494-8799; Fax 902-494-6621.Search for more papers by this author Lauren Davey, Department of Microbiology and Immunology, Dalhousie University, Halifax, NS, B3H 1X5 Canada Canadian Center for Vaccinology (CCfV), Dalhousie University and the Izaak Walton Killam (IWK) Health Centre, Halifax, NS, CanadaSearch for more papers by this authorAlejandro Cohen, Proteomics and Mass Spectrometry Core Facility, Life Sciences Research Institute, Dalhousie University, Halifax, NS, CanadaSearch for more papers by this authorJason LeBlanc, Department of Microbiology and Immunology, Dalhousie University, Halifax, NS, B3H 1X5 Canada Department of Pathology, Dalhousie University, Halifax, NS, Canada Department of Pathology and Laboratory Medicine, Nova Scotia Health Authority (NSHA), Halifax, NS, CanadaSearch for more papers by this authorScott A. Halperin, Department of Microbiology and Immunology, Dalhousie University, Halifax, NS, B3H 1X5 Canada Canadian Center for Vaccinology (CCfV), Dalhousie University and the Izaak Walton Killam (IWK) Health Centre, Halifax, NS, Canada Department of Pediatrics, Faculty of Medicine, Dalhousie University and the IWK Health Centre, Halifax, NS, CanadaSearch for more papers by this authorSong F. Lee, Corresponding Author Department of Microbiology and Immunology, Dalhousie University, Halifax, NS, B3H 1X5 Canada Canadian Center for Vaccinology (CCfV), Dalhousie University and the Izaak Walton Killam (IWK) Health Centre, Halifax, NS, Canada Department of Pediatrics, Faculty of Medicine, Dalhousie University and the IWK Health Centre, Halifax, NS, Canada Department of Applied Oral Sciences, Faculty of Dentistry, Dalhousie University, Halifax, NS, CanadaFor correspondence. E-mail Song.Lee@Dal.Ca; Tel. (902) 470-7522/494-8799; Fax 902-494-6621.Search for more papers by this author Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URLShare a linkShare onEmailFacebookTwitterLinked InRedditWechat Summary Recently, we identified a novel disulfide oxidoreductase, SdbA, in the oral bacterium Streptococcus gordonii. Disulfide oxidoreductases form disulfide bonds in nascent proteins using a CXXC catalytic motif. Typically, the N-terminal cysteine interacts with substrates, whereas the C-terminal cysteine is buried and only reacts with the first cysteine of the motif. In this study, we investigated the SdbA C86P87D88C89 catalytic motif. In vitro, SdbA single cysteine variants at the N or C-terminal position (SdbAC86P and SdbAC89A) were active but displayed different susceptibility to oxidation, and N-terminal cysteine was prone to sulfenylation. In S. gordonii, mutants with a single N-terminal cysteine were inactive and formed unstable disulfide adducts with other proteins. Activity was partially restored by inactivation of pyruvate oxidase, a hydrogen peroxide generator. Presence of the C-terminal cysteine alone (in the SdbAC86P variant) could complement the ΔsdbA mutant and restore disulfide bond formation in recombinant and natural protein substrates. These results provide evidence that certain disulfide oxidoreductases can catalyze disulfide bond formation using a single cysteine of the CXXC motif, including the buried C-terminal cysteine. Introduction Disulfide bonds between cysteine residues are important for the folding and function of secreted proteins, and proteins that are misfolded or lacking disulfide bonds are often unstable and prone to degradation. Disulfide bond formation is an oxidation reaction catalyzed by enzymes called thiol-disulfide oxidoreductases (Kadokura and Beckwith, 2010). These enzymes belong to the thioredoxin family, which are highly conserved and broadly distributed across all kingdoms of life. In addition to disulfide bond formation, this diverse class of enzymes carries out a range of redox reactions including disulfide reduction, isomerization, S-glutathionylation and reduction of peroxides (Atkinson and Babbitt, 2009; Collet and Messens, 2010). Some of the best characterized thioredoxin family members belong to the disulfide bond (Dsb) pathway of the Gram-negative bacterium Escherichia coli. In E. coli, disulfide bonds are formed by the periplasmic oxidoreductase DsbA. The active site of DsbA contains a disulfide bond that is transferred to substrate proteins. After each reaction, DsbA is left in a reduced state and the active site must be reoxidized to repeat another catalytic cycle. This function is performed by its redox partner, DsbB (Kishigami et al., 1995). Because DsbA is relatively indiscriminate, it can form disulfide bonds between any cysteine residues in a protein, sometimes leading to protein misfolding. The rearrangement of protein disulfides to achieve their proper conformation is catalyzed by the disulfide isomerase DsbC and its redox partner DsbD (Missiakas et al., 1995; Cho et al., 2007). Additional components of the system include the cytochrome c–reducing protein CcmG (Reid et al., 2001), and DsbG, which prevents oxidation of single cysteine residues (Depuydt et al., 2009). In the cytoplasm, proteins are maintained in a reduced state with the help of low molecular weight thiols and redox enzymes like thioredoxin and glutaredoxin (Cho and Collet, 2013). Despite their varying activities, all thioredoxin family enzymes share a characteristic thioredoxin fold, consisting of a four-stranded beta sheet between three alpha helices, and most have a Cys-X-X-Cys (CXXC) catalytic motif, where ‘X’ can be any amino acid (Atkinson and Babbitt, 2009; Shouldice et al., 2011). The two cysteines of the CXXC motif have distinct properties that have been well characterized and are generally shared by both oxidizing and reducing enzymes. The N-terminal cysteine is solvent exposed and usually has an acidic pKa, resulting in the formation of a reactive thiolate anion at physiological pH (Roos et al., 2013). This reactivity contributes to the N-terminal cysteine reacting with both substrate proteins and the redox partner (Shouldice et al., 2011). In contrast, the C-terminal cysteine is buried, and as a result substrates are sterically restricted to react with the N-terminal cysteine (Mössner et al., 2000). The C-terminal cysteine reacts solely with the first cysteine of the active site to form an intramolecular disulfide bond. Analyses of diverse disulfide oxidoreductases, ranging from E. coli DsbA to eukaryotic protein disulfide isomerase (PDI), have consistently demonstrated that both active site cysteines are required for activity in the cell (Bardwell et al., 1993; Yu et al., 1993; Walker et al., 1996; Tan et al., 2005). Unlike E. coli, little is known about disulfide bond formation in Gram-positive species, and there is considerable diversity in both the prevalence of disulfide bonded proteins and in the machinery used to generate them (Dutton et al., 2008; Daniels et al., 2010). DsbA homologs in Staphylococcus aureus and Bacillus subtilis have few substrates or phenotypes associated with them and appear to play a specialized role in genetic competence (Meima et al., 2002; Draskovic and Dubnau, 2005). In contrast, Mycobacterium tuberculosis encodes a DsbA homolog (Rv2969c) that is essential for optimal growth (Premkumar et al., 2013). In addition to DsbA-like enzymes, alternative types of disulfide catalysts have been identified, including two CcmG homologs in M. tuberculosis, DsbE and DsbF (Goulding et al., 2004; Chim et al., 2010), and a dimeric oxidoreductase predicted to be conserved among Corynebacterium spp. (Daniels et al., 2010). Although the oxidase activity of these enzymes has been confirmed in vitro, no natural substrates or phenotypes have been identified yet. Gram-positive facultative anaerobes like Streptococcus ssp. lack DsbA homologs entirely (Dutton et al., 2008). Nevertheless, pathogenic streptococci produce toxins that are stabilized by disulfide bonds, suggesting that these organisms possess disulfide catalysts that contribute to virulence (Marvaud et al., 1998; Baker et al., 2004). To better understand disulfide bond formation in Gram-positive bacteria, we have been investigating thiol-disulfide oxidoreductases in Streptococcus gordonii. We recently identified a novel disulfide oxidoreductase, SdbA, in S. gordonii, and found that that the ΔsdbA mutant has a pleiotropic phenotype (Davey et al., 2013). In this study, we evaluated the enzymatic activity of SdbA using cysteine point mutants in the CXXC catalytic motif. Unexpectedly, we found that SdbA was active in vitro using a single active site cysteine, at either the N-terminal or C-terminal position of the active site. In S. gordonii, mutants with a single N-terminal cysteine were inactive, whereas the single C-terminal cysteine alone catalyzed disulfide bond formation in both natural and recombinant protein substrates. The remarkable ability of SdbA to catalyze disulfide bond formation using a single cysteine demonstrates that certain disulfide oxidoreductases can carry out the same enzymatic function using different reaction pathways and suggests the potential existence of naturally occurring oxidoreductases that use a single, buried cysteine to oxidize their substrates. Virtually all antibodies contain highly conserved disulfide bonds that connect the β-sheets of the immunoglobulin domains in the variable heavy and light chains, and these structural disulfide bonds are critical for stable folding (Glockshuber et al., 1992; Proba et al., 1997). Because of their therapeutic potential, the production of recombinant antibodies has been well studied, and disulfide bond formation is a often limiting factor in their production (Jurado et al., 2002). Single chain variable fragment antibodies (scFv) are no exception, and aside from a few unusually stable variants, the proteins typically require two disulfide bonds (Glockshuber et al., 1992; Proba et al., 1997; Wörn and Plückthun, 2001; Jurado et al., 2002). This instability makes scFv antibodies a useful tool to study disulfide bond formation in S. gordonii. Previously, we found that S. gordonii SdbA was essential for the production of an scFv antibody (Davey et al., 2013). The protein contains four cysteines, which form two consecutive disulfide bonds, one in each variable fragment (Knight et al., 2008). The absence of scFv in the ΔsdbA mutant, and the fact that scFv production could be restored by sdbA complementation, indicated that SdbA catalyzed disulfide bond formation in this protein. Thus, to better understand how SdbA functions in S. gordonii, we started by investigating the role of SdbA in scFv production. To ensure that mutation of SdbA did not have unanticipated effects on the general sec pathway that could account for the absence of the scFv, we first tested production levels of a protein that does not contain disulfide bonds, the extracellular peptidyl-prolyl cis trans isomerase PrsA. Western blots showed that PrsA levels were similar between the parent and the ΔsdbA mutant, indicating that the sec pathway was functional in the ΔsdbA mutant (Fig. 1A). Next, we set out to find direct evidence that scFv from the ΔsdbA mutant lacked disulfide bonds. To analyze the oxidation state of scFv from the ΔsdbA mutant, degradation of the protein had to be prevented. To this end, we inactivated the extracellular serine protease and chaperone DegP in the ΔsdbA mutant. DegP has been identified as the sole protease responsible for elimination of unstable foreign proteins in related Gram-positive bacteria, making it a logical candidate for scFv degradation in S. gordonii (Poquet et al., 2000). Western blots showed that inactivation of degP in the ΔsdbA mutant fully restored scFv production, whereas complementation with a functional degP gene on the chromosome reversed the effect (Fig. 1B). Inactivation of degP alone did not alter the level of scFv production. To confirm that degradation required the protease activity of DegP, as opposed to its chaperone function, we introduced a point mutation in the active site serine required for protease activity (Krojer et al., 2010). The catalytically dead ΔsdbAdegPS235A mutant produced high levels of the scFv protein, confirming that the protease activity was responsible for degradation (Fig. 1B). DegP degrades scFv produced in the ΔsdbA mutant. A. Production of the 35 kDa lipoprotein PrsA was detected by Western blotting using anti-PrsA antisera as the probe (upper panel). The same samples were separated by SDS-PAGE and stained with Coomassie blue to show equal protein loading (lower panel). B. Western blots showing the production of the 33 kDa scFv protein. ScFv was detected using an anti-HA monoclonal antibody. Protein extracts were prepared from the parent, ΔsdbA, and ΔsdbAΔdegP mutants, as well as a ΔsdbAΔdegP mutant complemented with a functional degP gene on the chromosome (ΔsdbA DegP Compl), and a ΔsdbA mutant with a catalytically inactive DegP (ΔsdbA S235A DegP) (left panel). The ΔsdbA complemented mutant (SdbA Compl.) and single ΔdegP mutant are shown on the right panel. Since inactivation of degP stabilized scFv in the ΔsdbA mutant, we could use this strain to test the oxidation state of the protein. We analyzed disulfide bonding in scFv from the ΔsdbAΔdegP mutant using two approaches. First, scFv isolated from the parent and the ΔsdbAΔdegP mutant was alkylated with maleimide-PEG2-biotin to detect free thiols. Next, we used a differential thiol-trapping technique to detect disulfide bonds (Leichert and Jakob, 2004). This strategy uses iodoacetamide to block free thiols before reduction of disulfide bonds with dithiothreitol (DTT), and subsequent alkylation of the newly liberated thiols. The alkylation experiments showed that scFv isolated from the parent was fully oxidized, and free thiols were only detected when samples were reduced with DTT (Fig. 2A). In contrast, scFv from the ΔsdbAΔdegP mutant clearly contained free thiols indicative of unbonded cysteines. This reaction was blocked by iodoacetamide, which confirmed that the biotinylation was specific to free thiols (Fig. 2B). To ensure that the free thiols detected in the double mutant were from the loss of SdbA oxidoreductase activity, and not an accumulation of misfolded protein from the loss of DegP, we also tested the oxidation state of scFv isolated from a ΔdegP single mutant as a control. ScFv from the ΔdegP mutant was oxidized to a similar extent as the parent, confirming that reduced scFv obtained from the ΔsdbAΔdegP mutant was due to the loss of SdbA (Fig. 2A). Thus, SdbA is essential for disulfide bonding in the scFv, and DegP alone appears to be responsible for the rapid degradation of the misfolded protein. SdbA catalyzes disulfide bond formation. A. The oxidation state of scFv purified from the parent, ΔsdbAΔdegP, SdbAC86PΔdegP (C86PΔdegP), and ΔdegP mutant was determined by alkylation with maleimide-PEG2-biotin and detected using avidin alkaline phosphatase (Avidin-AP). The same protein samples were run in adjacent lanes and reacted with anti-HA to detect the total scFv protein as a loading control. The lower molecular weight bands recognized by avidin-AP correspond to degradation products that have lost the C-terminal HA-tag, but contain alkylated cysteines. Arrows indicate detection of free thiols. B. Differential thiol trapping to detect disulfide bonds. Free thiols in proteins were blocked with iodoacetamide (IAM) prior to reduction of disulfide bonds with DTT. The liberated cysteines were then detected by maleimide-PEG2-biotin and avidin-AP. Total protein loading was detected using anti-HA. C. Western blot to detect scFv produced in the SdbA CXXC active site mutants. Proteins were extracted from the parent, ΔsdbA, sdbAC86P (C86P), and the sdbAC89A (C89A) mutant, and from their corresponding ΔdegP mutants. Lower molecular weight bands are degradation products recognized by anti-HA. D. The oxidation state of the autolysin AtlS. Protein extracts from the parent and the sdbAC86P mutant (C86P) were alkylated with maleimide-PEG2-biotin and detected using avidin-AP. The same samples were run in adjacent lanes and reacted with anti-AtlS antisera to detect total AtlS as a loading control. The C-terminal cysteine of the CXXC motif alone is sufficient for SdbA activity in S. gordonii The conserved CXXC active site motif of thiol disulfide oxidoreductases is critical for their activity (Yu et al., 1993; Kishigami et al., 1995; Walker et al., 1996). To characterize the CXXC motif in SdbA, and confirm that the enzymatic activity of SdbA contributed to the pleiotropic phenotype observed in the ΔsdbA mutant, we used site-directed mutagenesis to replace one or both active site cysteines, generating sdbAC86P, sdbAC89A and sdbAC86P/C89A mutants. The constructs were introduced as single copies onto the chromosome of ΔsdbA mutants, and expressed under the native sdbA promoter. All mutations were confirmed by DNA sequencing. To assess the activity of the point mutants, we tested for scFv production. Similar to the ΔsdbA mutant, mutation of the SdbA active site C-terminal cysteine (C89A), or of both cysteines C86P/C89A, eliminated scFv production. Inactivation of degP rescued production in the mutants, demonstrating that the protein was produced, but was unstable and sensitive to degradation (Fig. 2C). Unexpectedly, mutation of the N-terminal cysteine (C86P) did not eliminate scFv production. The sdbAC86P mutant continued to produce scFv even in the presence of DegP, although inactivation of degP did increase the overall yield. The ability of this mutant to generate stable scFv protein suggested that SdbA remained functional using the single C-terminal cysteine (Fig. 2C). When constructing the mutants, the codon for proline was chosen for the sdbAC86P mutation because it generated a unique MscI restriction site that could be used to easily confirm the construct. Given the unexpected activity of the sdbAC86P mutant, we constructed a second set of SdbA cysteine mutants, in which the amino acid substitutions were reversed to ensure that the proline mutation did not affect activity. Analysis of scFv production by an sdbAC86A mutant showed that, like the sdbAC86P mutant, a low level of stably folded scFv protein was produced (Fig. S1). Similarly, mutation of the C-terminal cysteine, Cys89, to either an alanine or a proline abolished scFv production. Thus, the mutants behave similarly, regardless of the amino acid substitution introduced at the Cys86 position. As the C86P mutation could be easily confirmed by restriction analysis, we used this mutant in all subsequent experiments. To determine if scFv produced by the sdbAC86P mutant contained disulfide bonds, we tested the oxidation state by alkylation. To improve recovery, we used scFv purified from the sdbAC86PΔdegP mutant. As expected, the scFv consisted of a mixture of reduced and oxidized protein. Although free thiols were detected, likely corresponding to the portion of reduced scFv degraded by DegP (Fig. 2A), a portion of the protein was oxidized (Fig. 2B). Similar results were obtained when we tested the oxidation state the autolysin AtlS, a natural substrate of SdbA, which was oxidized to a similar extent in the sdbAC86P mutant and the parent (Fig. 2D). This indicated that SdbA could generate correctly folded disulfide bonded proteins using the catalytic C-terminal cysteine alone. Mutation of either active site cysteine in an oxidoreductase typically eliminates activity, and the same holds true for other members of the thioredoxin family (Yu et al., 1993; Zapun et al., 1994; Kishigami et al., 1995; Walker et al., 1996; Tan et al., 2005; Eser et al., 2009; Qin et al., 2011). Thus, the oxidase activity in the sdbAC86P mutant was surprising and warranted further investigation. Previously, we found that ΔsdbA mutants have a distinct phenotype that includes enhanced biofilm formation, deficiencies in autolytic activity and bacteriocin production, and clumping in liquid medium (Davey et al., 2013). To assess oxidase activity in S. gordonii, we tested the ability of the SdbA cysteine mutants to complement the ΔsdbA mutant phenotype. Analysis of biofilm formation by crystal violet staining showed that sdbAC86P fully complemented the ΔsdbA mutant phenotype and reduced biofilms levels to that of the parent (Fig. 3A). The sdbAC89A partially complemented the ΔsdbA mutant, while the sdbAC86P/C89A mutant failed to complement the ΔsdbA phenotype and produced significantly more biofilm (P   0.0001). The single C-terminal active site cysteine of SdbAC86P complements a ΔsdbA mutant. A. Crystal violet staining of biofilms grown in 24-well microtiter plates. Biofilms were grown with the parent, ΔsdbA mutant, sdbAC86P (C86P), sdbAC89A (C89A) and sdbAC86P/C89A (C86P/C89A) mutants. Bars represent the mean and standard deviation of triplicates. B. Autolysis of the parent, ΔsdbA mutant and the SdbA cysteine mutants. Each point represents the mean of triplicates. C. Zymogram showing AtlS activity in the parent, sdbAC86P (C86P) and sdbAC89A (C89A) mutants. The zymogram gel was stained with methylene blue to enhance contrast. An equal volume of the autolysin extracts was used for the zymogram (upper panel) and for the Western blot reacted with anti-AtlS antisera as a loading control (lower panel). D. Sth1 bacteriocin activity of the parent, ΔsdbA mutant and SdbA cysteine mutants. Bars show the growth of the indicator strain S. oralis 34. Error bars show mean and standard deviation of triplicates. E. Overnight cultures showing mutants that lack SdbA enzymatic activity form clumps and sediment to the bottom of the tube. Asterisks denote a significant difference from the parent (**P   0.01; ****P   0.0001; one-way ANOVA). Another important indicator of SdbA activity is autolysis, given that AtlS is a natural SdbA substrate (Davey et al., 2013). AtlS contains two cysteines that form a single disulfide bond. An autolysis assay demonstrated that, like the ΔsdbA mutant, the sdbAC89A and sdbAC86P/C89A mutants were highly resistant to lysis (Fig. 3B). The sdbAC86P mutant, however, restored autolysis and lysed at a rate that was comparable to the parent strain. To look at AtlS activity directly, we used zymography to test enzyme function. While AtlS from the sdbAC89A mutant generated a single, weak band of clearing in a zymogram, the sdbAC86P mutant produced two distinct bands, similar to the parent (Fig. 3C). This demonstrated that the sdbAC86P mutant generated enzymatically active AtlS, and agrees with the alkylation data showing that the autolysin had been oxidized in the sdbAC86P mutant (Fig. 2D). Next we tested for complementation of bacteriocin production. S. gordonii produces two bacteriocins, Sth1 and Sth2 (Heng et al., 2007), and ΔsdbA mutants are deficient in both. Bacteriocins were detected using an activity assay based on growth inhibition of the target strain S. oralis. Culture supernatants from the parent inhibited the growth of S. oralis, indicating the presence of bacteriocins. In contrast, no inhibitory activity was detected in the ΔsdbA and sdbAC86P/C89A mutants, and the sdbAC89A mutant displayed an intermediate phenotype, partially inhibiting the growth of S. oralis. The sdbAC86P mutant behaved similar to the parent and fully restored bacteriocin activity to the ΔsdbA mutant (Fig. 3D). Like the parent strain, the sdbAC86P mutant also prevented clumping and sedimentation observed in ΔsdbA strains grown in liquid culture (Fig. 3E). Taken together, the data support the notion that SdbAC86P was active in the cell with only the single C-terminal active site cysteine. In vitro enzyme assays were used to assess the activity of the SdbA cysteine variants. First, we carried out differential scanning fluorimetry to determine if the point mutations introduced major structural changes that might explain the inactivity of the sdbAC89A mutant. Compared with the melting temperature (Tm) of wild-type SdbA at 55.9°C, the inactive SdbAC89A variant was similar to the parent with a Tm of 56.8°C. Thus, the loss of enzyme activity in the sdbAC89A mutant did not appear to be due to misfolding as a result of the mutation. The active SdbAC86P variant had a Tm of 52.4°C, suggesting reduced stability and possible structural changes, but such changes did not affect activity as shown below. To assess the oxidase activity of the SdbA cysteine variants, we tested the ability of the purified enzymes to catalyze folding of reduced, denatured RNase A in the presence of glutathione. The wild-type enzyme efficiently catalyzed RNase A folding, while the cysteine-free SdbAC86P/C89A variant had lost all activity (Fig. 4A). In stark contrast to the activity of the enzymes in S. gordonii, where only the sdbAC86P mutant was active, both SdbA single cysteine variants exhibited comparable oxidase activity in vitro. Both of the SdbA single cysteine variants accelerated RNase A folding above background levels at approximately the same rate, indicating that either single cysteine of the CXXC active site was sufficient for activity in vitro (Fig. 4A). Neither mutant was active without prior treatment with GSSG (Fig. S3). This suggested that the SdbA single cysteine variants could use glutathione to catalyze disulfide bond formation. Both SdbA single cysteine variants are active in vitro. A. SdbA catalyzed refolding of reduced denatured RNase A in the presence of glutathione buffer. RNase A was incubated with the parent, SdbAC86P (C86P), SdbAC89A (C89A), SdbAC86P/C89A (C86P/C89A) or in buffer alone (RNase). Refolding was monitored by measuring the increase in absorbance at 296 nm as a result of RNase A catalyzed hydrolysis of cCMP. B. Alkylation with maleimide-PEG2-biotin to detect oxidation of RNase A catalyzed by SdbA single cysteine mutants. Maleimide-PEG2-biotin adds 0.5 kDa to each reduced cysteine, causing the alkylated RNase A (reduced) to migrate slower on a 15% SDS-PAGE gel stained with Coomassie blue. C. Reactivity of SdbA cysteines with DTNB. Reduced SdbA, the SdbA single cysteine variants, and SdbAC86P denatured with 6M guanidine HCl (denatured C86P) were reacted with excess DTNB and the absorbance was monitored at 412 nm. D. Mass spectrometry analysis of the oxidation state of SdbA cysteines following reaction with oxidized glutathione. The bars represent a semi-quantitative analysis of the peak intensities for various modifications on the active site cysteines (SH, thiol; GSSG, S-glutathionylation; SOH, sulfenic acid; SO2H, sulfinic acid; SO3H, sulfonic acid). The data are representative of two experiments. Oxidase activity was also assessed using a modified RNase A folding assay, in which we reacted SdbAC86P or SdbAC89A with reduced denatured RNase A and tested for oxidation by running the RNase A on a gel to visualize a shift in migration. Glutathione was omitted from the reaction buffer to minimize spontaneous refolding, and instead the SdbA single cysteine variants were oxidized with glutathione prior to the reaction. Following the refolding reaction, the RNase A was alkylated with maleimide-PEG2-biotin, which added 0.5 kDa per thiol to enhance the shift between oxidized and reduced proteins. The SDS-PAGE gel showed a clear shift between RNase A incubated alone versus samples incubated with either SdbAC86P or SdbAC89A, signifying that the protein had been oxidized (Fig. 4B). Kinetic analysis of enzyme activity was consistent with the results of the RNase A folding assay, showing that the efficiency of both SdbA single cysteine variants were lower than the parent but similar to each other (Fig. S2A). The kcat/Km values for the SdbAC89A and SdbAC86P variants with reduced RNase A were 68% and 78% of the wild type protein respectively. The CXXC motif of thioredoxin family proteins with oxidoreductase activity usually have an acidic, solvent-exposed N-terminal cysteine, and a buried C-terminal cysteine (Zapun et al., 1994). The unexpected activity of the single cysteine variants led us to test whether SdbA shares these characteristics, or if the activity could be attributed to unusual features of the active site. To determine if SdbA has an exposed N-terminal cysteine and buried C-terminal cysteine, solvent accessibility was tested using dithionitrobenzoic acid. Wild-type SdbA and SdbAC89A reacted rapidly, whereas the SdbAC86P variant only reacted after being denatured, confirming that the cysteine was indeed buried in the structure of the protein (Fig. 4C). The amount of thionitrobenzoic acid (TNB) released by the wild type was twice that of the single cysteine variants, which is indicative of a dithiol mechanism. During the reaction, the N-terminal cysteine, Cys86, reacts with DTNB to form a mixed disulfide with TNB and released one equivalent of free TNB. This is followed by a second reaction between Cys89 and the mixed disulfide to release a second unit of TNB. This suggests that despite being active with the single C-terminal cysteine, wild-type SdbA probably uses both cysteines. This is also consistent with the fact that both cysteines are conserved in the motif (Davey et al., 2013). It should also be noted that the N-terminal cysteine was determined to have an acidic pKa of 4.7, similar to other bacterial oxidoreductases (Fig. S2B) (Goulding et al., 2004). As such, the SdbA CXXC active site does share some typical characteristics of an oxidoreductase active site, despite the atypical activity of the C-terminal cysteine. There is evidence that low molecular weight thiols can contribute to oxidase activity by forming intermolecular disulfides, and a previous investigation of an E. coli DsbA single cysteine mutant found that it was active in vitro using glutathione (Wunderlich et al., 1995). To determine if the SdbA single cysteine variants could use low molecular weight thiols for activity, we tested for S-glutathionylation of the active site cysteines by mass spectrometry. Peptide fragments containing the catalytic site were analyzed for modifications on the cysteine residue and quantified by comparing the peak intensities. Each modification was expressed as a percentage of the total cysteine content. Surprisingly, both of the SdbA single cysteine variants were found to be almost completely S-glutathionylated, including the buried C-terminal cysteine that was inaccessible to DTNB (Fig. 4D and Fig. S4). This equivalent reactivity with glutathione likely explains why both enzymes were active in vitro, as the disulfide bond between SdbA and glutathione could be transferred to substrate proteins. These data suggest that SdbA can use either single cysteine of the active site to catalyze disulfide bond formation using low molecular weight thiols. As SdbA was functional using either cysteine of the active site in vitro, the results suggested that there was something specific to S. gordonii that inhibited activity of the N-terminal cysteine. When cultured under aerobic conditions, S. gordonii produces up to 1.6 mM of hydrogen peroxide (H2O2) as a metabolic byproduct of pyruvate oxidase (Liu et al., 2011), which slows the growth of competing bacteria and contributes to the success of S. gordonii in colonizing oral biofilms (Jakubovics et al., 2008). As the SdbA active site cysteines would be exposed to endogenous H2O2, we hypothesized that cysteine oxidation might contribute to the differences between SdbA activity in vitro and in the cell. H2O2 can oxidize the sulfhydryl group of thiols to sulfenic acid, a reversible modification that can contribute to disulfide bonding, while further oxidation to sulfinic and sulfonic acid can cause irreversible damage (Cho and Collet, 2013; Roos et al., 2013). To test the effect of H2O2 on the SdbA cysteine variants, we used mass spectrometry to determine the oxidation state of the enzymes after exposure to H2O2 in vitro. SdbA was reacted with H2O2 in the presence of the sulfenic acid probe dimedone, and the proteins were subjected to trypsin digestion and analyzed by mass spectrometry. Peptide fragments containing the active site cysteines were assessed for modifications, and peak intensity integration was used as a semi-quantitative measure of the relative abundance of each modification. Despite their similar reactivity with glutathione, there was a striking difference between the SdbAC89A and SdbAC86P variants in their reactivity with H2O2. The cysteine in the SdbAC89A variant was detected almost entirely as sulfenic acid, representing 86.5% of the active site cysteines (Figs 5A, S5, S6). In contrast, the cysteine in the SdbAC86P mutant contained a mixture of sulfenic acid and S-glutathionylated species, corresponding to 37% and 62% of the total cysteine respectively. The glutathione modification apparently originated from E. coli during isolation of the protein. The N-terminal cysteine of the SdbAC89A mutant protein is sensitive to oxidation. A. Mass spectrometry analysis of the oxidation state of SdbA cysteines following reaction with hydrogen peroxide in the presence of the sulfenic acid probe dimedone. The bars represent a semi-quantitative analysis of the peak intensities for various modifications on the active site cysteines (SH, thiol; dimedone, sulfenic acid; SO2H, sulfinic acid; SO3H, sulfonic acid; GSSG, S-glutathionylation). The data are representative of two experiments. B. Detection of sulfenylated cysteines by Western blot. The SdbA variants were incubated with or without 0.1 mM hydrogen peroxide and detected with a biotin-tagged sulfenic acid probe, DCP-Bio1 and avidin alkaline phosphatase (upper panel). The same samples were separated by 15% SDS-PAGE and stained with Coomassie blue as a loading control (lower panel). C. Detection of scFv produced in the parent, sdbAC86P (C86P) and sdbAC89A (C89A) mutant, and their corresponding ΔspxB mutants. Membranes were probed with anti-HA to detect the scFv protein, and detection with anti-PrsA was used to standardize the protein concentration. D. The Western blot was analyzed by densitometry using ImageJ. Bars represent the ratio of scFv to PrsA. To confirm the mass spectrometry data, we tested for sulfenylation of the SdbA variants using a biotin-tagged sulfenic acid probe, DCP-Bio1. Western blots showed that the SdbAC89A variant produced a band detected by avidin alkaline phosphatase, indicating that the single N-terminal cysteine had been sulfenylated (Fig. 5B). Without H2O2, a weak band could still be detected in the SdbAC89A variant, likely due to oxidation by molecular oxygen during protein isolation. Unlike the N-terminal cysteine, the buried C-terminal cysteine was not highly susceptible to sulfenylation. It was surprising to find that sulfenylation had occurred primarily in the inactive SdbAC89A variant, given that sulfenic acid has been demonstrated to contribute to disulfide bond formation by proteins with single cysteine residues (Chng et al., 2012; Wang et al., 2014). Despite multiple attempts, we were unable to detect sulfenylation in SdbA immunoprecipitated from S. gordonii, possibly due to the low amount of total SdbA protein combined with the high reactivity of sulfenic acid. Therefore, as an alternative approach to investigate the effect of H2O2 on disulfide bond formation in the cell, we generated a H2O2 null mutant by inactivating the pyruvate oxidase gene spxB. The mutants were tested for scFv production as an indicator of SdbA activity, as described previously. Consistent with the notion that H2O2 and sulfenylation was detrimental to SdbA activity, inactivation of spxB partially restored scFv production to the sdbAC89A mutant (Fig. 5C and D). Thus, the SdbAC89A variant did have some activity in the cell, as was observed in the biofilm (Fig. 3A) and bacteriocin assays (Fig. 3D). Eliminating H2O2 also increased the scFv yield from the parent and sdbAC86P mutant, indicating that the reactive oxygen species generally inhibits SdbA, either directly or through oxidative damage of the scFv substrate. These results suggest that, in agreement with the in vitro data, SdbA has enzymatic activity using either single cysteine of the active site, but that the solvent exposed N-terminal cysteine in SdbAC89A was vulnerable to oxidation that inhibited activity. Oxidoreductases with mutations to the C-terminal cysteine of the CXXC motif, including DsbA (Kishigami et al., 1995), PDI (Walker et al., 1996) and thioredoxin (Balmer et al., 2003), can form disulfide linked complexes with their redox partners and substrates respectively. We hypothesized that sulfenic acid formation in the sdbAC89A mutant could contribute to complex formation which, if not resolved, would consume the enzyme rendering it inactive. First, we tested for intermolecular disulfide bond formation between SdbA and its natural substrate AtlS. AtlS was extracted by incubating cells in 4% SDS at room temperature and subsequently analyzed by non-reducing Western blots probed with anti-AtlS and anti-SdbA antisera. Extracts from the sdbAC89A mutant contained a ∼ 150-kDa band that was recognized by both antibodies, signifying a complex formed between SdbA (22 kDa) and AtlS (130 kDa) (Fig. 6A). The SdbA/AtlS complex was only weakly detected in the active sdbAC86P mutant and the parent. Given that AtlS was inactive in the sdbAC89A mutant, complex formation did not equate with oxidase activity and might inhibit function. SdbAC89A forms disulfide-linked complexes in the cell. A. Complex formation between SdbA and AtlS. Surface proteins extracted with 4% SDS from the parent, sdbAC86P (C86P) or sdbAC89A (C89A) mutant were analyzed under reducing and non-reducing conditions and probed with either anti-AtlS or anti-SdbA antisera. Arrows indicate the AtlS + SdbA complex detected by both anti-SdbA and anti-AtlS antisera, the intact AtlS monomer (AtlS1) and the processed AtlS monomer (AtlS2). The asterisk denotes a second SdbA complex with an unknown protein. B. Western blot of protein extracts from the parent, the SdbA single cysteine mutants and their corresponding ΔdegP mutants run under non-reducing and reducing conditions and probed with anti-SdbA antisera. Under reducing conditions, SdbA runs at 21 kDa. The same samples were separated on SDS-PAGE and stained with Coomassie blue to show the total protein (right). Finally, we tested the total disulfide-linked complexes in the cell using cells boiled in non-reducing sample buffer. The samples were run non-reducing Western blots and reacted with anti-SdbA antisera. Extracts from the sdbAC89A mutant produced a distinct banding pattern showing multiple complexes with other cysteine containing proteins (Fig. 6B). These complexes were absent in the parent and the active sdbAC86P mutant. When we compared complex formation between strains with and without a functional DegP protease, the sdbAC89A mutant consistently produced less total SdbA and fewer disulfide complexes, indicating that the complexes were sensitive to degradation. When we tested complex formation using recombinant SdbA and reduced RNase A in vitro, several high molecular weight bands were detected on the gel; however, the intensity and number of these bands was similar in both SdbA single cysteine variants, and the complexes were only a minor component of the total protein sample (Fig. S7). This suggested that complex formation was not an intrinsic characteristic of the SdbAC89A variant but was related to the environment in the cell. Taken together, the data show SdbA can catalyze disulfide bond formation using a single active site cysteine. Although the N-terminal cysteine in the SdbAC89A variant was active in vitro, it was inactive in the cell, apparently because of its propensity to form unstable complexes with other proteins. In contrast, the sdbAC86P mutant with a single, buried cysteine was active both in vitro and in the cell, and can complement the ΔsdbA mutant. In this study, we analyzed SdbA active site point mutants to confirm that the ΔsdbA mutant phenotype was directly related to its oxidoreductase activity. Our characterization of the SdbA active site showed that, unlike previously described thiol-disulfide oxidoreductases, SdbA is active with only a single cysteine of the CXXC motif. Although SdbA cysteine variants with a single cysteine in either the N-terminal or C-terminal position were active in vitro, only variants with a single C-terminal cysteine were active in S. gordonii under most conditions. This difference appears to be related to the environment in the cell, the structure of the protein and the difference between the individual catalytic cysteines in their susceptibility to oxidation. The single C-terminal cysteine of the SdbAC86P mutant produced disulfide-bonded proteins both in vitro and in S. gordonii, which was unexpected given that the C-terminal cysteine is buried and does not typically interact with substrates. In vitro activity required glutathione, which suggests that the mechanism of oxidation might involve a transfer of the disulfide bond from S-glutathionylated SdbA to its substrate. This type of mechanism is not unprecedented, and other thioredoxin family proteins, including DsbA mutants (Wunderlich et al., 1995) and glutaredoxin 3 (Eser et al., 2009), have been proposed to gain oxidase activity by forming a mixed disulfide between the N-terminal cysteine of the CXXC motif and glutathione. The disulfide bond is then transferred to substrate proteins, with glutathione as the final electron acceptor. We suggest that a similar scenario may occur in the sdbAC86P mutant (Fig. 7B). This is also in agreement with the mass spectrometry data, which showed that, despite being buried, the C-terminal cysteine forms a stable disulfide bond with glutathione. This mixed disulfide bond could be transferred to substrate proteins. Proposed model for SdbA activity in the cell. A. Disulfide linked complexes formed by sulfenylated SdbAC89A cannot be resolved and are subsequently degraded by DegP. B. The SdbAC86P mutant can use low molecular weight thiols to catalyze disulfide bond formation, adapted from Wunderlich et al. (1995) and Eser et al. (2009). C. Wild-type SdbA likely uses a dithiol mechanism but could potentially also use the same pathway as the SdbAC86P mutant. Although the sdbAC86P mutant used glutathione to catalyze disulfide formation in vitro, it is not clear how the enzyme functioned in S. gordonii. Initially, we hypothesized that sdbAC86P might function as a dimer; however, we found that purified SdbAC86P dimers were completely inactive in vitro, making it unlikely that they would be active in the cell (data not shown). Similarly, medium components did not contribute to oxidase activity, and the mutant produced similar amounts of scFv protein in both rich and minimal medium (Fig. S8). Streptococci do synthesize glutathione (Zheng et al., 2013), which could contribute to SdbAC86P activity, along with other low molecular weight thiols such as cysteine. Alternatively, SdbAC86P might interact with a redox partner that has not been identified yet. Interestingly, the C-terminal cysteine was inaccessible to DTNB, yet clearly reacted with glutathione and substrate proteins. A possible explanation for this discrepancy is that binding of a peptide substrate results in a conformation change that alters accessibility of the C-terminal cysteine. At the same time, this buried position might protect the cysteine from damaging oxidation by H2O2, allowing it to react preferentially with thiols instead, and maintain its enzymatic activity. Conformation changes induced by substrate binding have been demonstrated for the cytochrome c reductase ResA, which increases its catalytic activity in response to substrate binding (Lewin et al., 2006). Some peroxidases also undergo local unfolding around the active site during their catalytic cycle. In these enzymes, the peroxidactic cysteine reacts with substrates in a folded conformation and subsequently unfolds to allow access to the resolving cysteine (Hall et al., 2011). Intriguingly, the peroxidatic cysteine of peroxidases is in the equivalent position as the C-terminal cysteine in the SdbAC86P variant (Hall et al., 2011). In contrast to the C-terminal cysteine, the single N-terminal cysteine of the sdbAC89A mutant was solvent exposed and had an acidic pKa, two factors that would contribute to its reactivity with endogenous H2O2 (Hall et al., 2011), which appeared to inhibit enzyme activity in the cell. H2O2 probably inhibits all of the SdbA variants to some extent and could potentially cause oxidative damage of SdbA substrates as well; however, the effect was most pronounced in the SdbAC89A variant. The ability of the N-terminal cysteine to stabilize sulfenic acid was unexpected, and it is not clear why the cysteine is not equally susceptible to further oxidation to sulfinic or sulfonic acid (Fig. 5A). Our observation that sulfenylation was detrimental to the catalytic activity of the sdbAC89A mutant is counterintuitive, given that sulfenic acid species are highly reactive and can form disulfide bonds. Our data suggest that the sulfenylated N-terminal cysteine of the sdbAC89A mutant does indeed form disulfide bonds; however, the bonds formed via this pathway cannot be resolved and are degraded (Fig. 7A). Initially we thought that SdbA might function like the single cysteine rhodanese PspE, which can use sulfenic acid to catalyze disulfide bond formation in substrate proteins (Chng et al., 2012). Notably, however, PspE cooperates with the disulfide isomerase DsbC to complete its catalytic cycle and resolve intermolecular bonds between PspE and its substrates (Chng et al., 2012). Because S. gordonii might not have the requisite isomerases needed to resolve disulfide complexes using a similar mechanism, the SdbAC89A mutant ends up trapped with its substrates, resulting in a dead end for the enzyme. Typically, in thioredoxin family members with oxidase activity, the CXXC motif contains an intramolecular disulfide bond that is transferred to substrate proteins, thereby necessitating both cysteines for activity (Kadokura and Beckwith, 2010). As such, there are few reports of naturally occurring thioredoxin family proteins with a single C-terminal cysteine. Two eukaryotic PDI-like proteins with SXXC active sites that have been analyzed are human PDILT (Tokuhiro et al., 2012) and EtPDIL from the ampicomplexan Eimeria tenella (Han et al., 2014). However, neither of these proteins exhibited oxidoreductase activity in vitro, although human PDILT might function as a chaperone (Tokuhiro et al., 2012; Han et al., 2014). Generally, the most common thioredoxin family enzymes that use a single C-terminal cysteine are peroxiredoxins, whereas those that use a single N-terminal cysteine belong to the monothiol glutaredoxins, and there are no known single cysteine oxidoreductases (Atkinson and Babbitt, 2009). Nevertheless, in silico analyses have predicted the existence of DsbA-like thioredoxin family enzymes with single N or C-terminal cysteines, although these enzymes have not been characterized biochemically or biologically (Fomenko and Gladyshev, 2003; Atkinson and Babbitt, 2009). Our analysis of SdbA single cysteine variants supports these predictions by demonstrating that a single cysteine at either position of the CXXC motif has the potential for oxidase activity, even when the C-terminal cysteine is buried. Virtually all reports of single cysteine mutants with residual activity involve the exposed, N-terminal cysteine. For example, E. coli DsbA mutants were found to catalyze the oxidative folding of hirudin in vitro using a single N-terminal cysteine (Wunderlich et al., 1995). Similarly, a DsbA homolog in Francisella, FipB, had partial activity using a single N-terminal cysteine (Qin et al., 2011). PDI mutants also retain some isomerase activity with a single N-terminal active site cysteine (Walker et al., 1996). In each of these instances, the enzyme activity was limited to the reactive N-terminal cysteine of the CXXC motif, whereas variants with a single C-terminal cysteine were inactive. Thus, the activity of the S. gordonii single cysteine mutants, and particularly of the C-terminal mutant, was unexpected. One of the only examples of a mutated CXXC motif protein that did have a low level of activity using a single C-terminal cysteine did not have a typical CXXC active site. Mutants generated in the Bacillus subtilis cytochrome c reductase ResA with a single C-terminal cysteine showed a low level of activity in an in vitro assay, although they were inactive in the cell (Hodson et al., 2008). ResA, however, has an unusual CXXC motif, where both of the cysteines are sufficiently solvent exposed to react with DTNB, even when fully folded (Lewin et al., 2006). In contrast, our data indicate that SdbA has a typical oxidoreductase active site with an exposed, acidic N-terminal cysteine and buried C-terminal cysteine. A remaining question is whether wild-type SdbA uses the same mechanism as the SdbAC86P mutant. The data suggest that wild-type SdbA most likely uses a standard mechanism, using both cysteines, under most conditions (Fig. 7C). Although the single cysteine mutant complemented the ΔsdbA mutant phenotypes, production of the scFv protein was lower in the sdbAC86P mutant than in the parent. Similarly, kinetic analysis of the single cysteine variants showed that they were less efficient than the wild type. The reactivity of the wild-type enzyme with DTNB also indicates that SdbA uses both active site cysteines. Interestingly, the sdbAC86P mutant reached wild-type levels of scFv production in the H2O2 null ΔspxB mutant background, and it is conceivable that SdbA might use different pathways under different conditions (e.g. anaerobic biofilm), or that a single C-terminal cysteine may serve as a backup should the N-terminal cysteine become over oxidized to sulfonic acid. However, additional investigation is required to determine if the wild-type SdbA ever uses a monothiol mechanism in S. gordonii. If SdbA has a typical active site, a remaining question is why SdbA is functional with a single cysteine and what sets it apart from oxidoreductases that require two cysteines. Despite demonstrated oxidase activity, sequence analysis of SdbA reveals that it has a number of features that resemble a reductase. Like thioredoxin, SdbA has a tryptophan residue adjacent to the CXXC motif and an isoleucine located N-terminus to the conserved cis-proline, which has been reported to influence enzymatic activity (Ren et al., 2009). In addition, the SdbA active site, WCPDC, is identical to the conserved eukaryotic reductase TRP14 (Jeong et al., 2009). The M. tuberculosis disulfide oxidase DsbE is also homologous to a Gram-negative reductase, CcmG (Goulding et al., 2004), which suggests that protein homology might not be a good indicator of enzyme function in Gram-positive species. To better understand the catalytic mechanism of SdbA, analyses are underway to solve the enzyme structure. At this stage, we do not know if S. gordonii uses additional proteins to catalyze disulfide bond formation, and the identity of SdbA redox partner(s) have yet to be identified. Our data show that SdbA was essential to the production of the scFv antibody and a functional AtlS. This suggests that SdbA was the only oxidoreductase in S. gordonii capable of oxidizing these proteins, and therefore it is unlikely that a different enzyme could be forming disulfide bonds in the sdbAC86P mutant. Although stable scFv was not detected in the ΔsdbA mutant, our alkylation reactions indicated that a portion of the protein recovered from the ΔsdbAΔdegP mutant did contain oxidized cysteines. These could be from non-native disulfide bonds or other types of oxidation reactions, such as sulfenylation, that could give rise to background alkylation after reduction with DTT. Background oxidase activity is also known to occur E. coli DsbA mutants, although the mechanisms involved are unknown (Bardwell et al., 1991; Chng et al., 2012). We do not know if this background oxidation is spontaneous or enzyme catalyzed, or if SdbA is involved in isomerization of non-native bonds. However, we have not found evidence to suggest that SdbA has isomerase activity in vitro, and SdbA was inactive in a scrambled RNase A refolding assay, as well as an insulin reduction assay (data not shown). In summary, we have demonstrated that SdbA can sustain disulfide bond formation using a single cysteine of the CXXC active site motif. Point mutants with a single cysteine at either the N-terminal or C-terminal position of the CXXC motif exhibited oxidase activity in vitro, but only mutants with the C-terminal cysteine could maintain activity in S. gordonii. Remarkably, mutants with a single C-terminal cysteine complemented the phenotype of ΔsdbA mutants and generated disulfide bonds in both the native substrate AtlS and a recombinant disulfide bond containing protein. This study shows that certain disulfide oxidoreductases can function with the C-terminal buried cysteine of the CXXC motif. Experiments were carried out using S. gordonii SecCR1 as the parent strain. S. gordonii SecCR1 is a recombinant strain of S. gordonii Challis DL-1 that secretes a single chain variable fragment antibody against complement receptor 1 (CR1) (Knight et al., 2008). All other strains are described in Table 1. Unless otherwise noted, S. gordonii was grown in HTVG (per milliliter: 5 mg glucose, 35 mg tryptone, 100 mM HEPES, 0.04 μg p-aminobenzoic acid, 0.2 μg thiamine-HCl, 1 μg nicotinamide and 0.2 μg riboflavin, pH 7.6) (Burne et al., 1999) at 37°C, 5% CO2, without shaking. Cysteine free minimal medium (MM) was prepared as described previously (Kim and Senadheera, 2012). Streptococcus oralis 34 was grown in Brain Heart Infusion medium (BHI, Difco). E. coli XL-1 Blue (Stratagene) was grown in Luria–Bertani medium (LB) at 37°C with shaking. Antibiotics were used at the following concentrations: for S. gordonii, erythromycin 10 μg ml−1, tetracycline 10 μg ml−1, spectinomycin 250 μg ml−1 and kanamycin 250 μg ml−1; for E. coli, ampicillin 100 μg ml−1 and tetracycline 10 μg ml−1. sdbA complemented on chromosome and expressed from the native sdbA promoter, TetR, SpecR, KanR Davey et al. (2013) sdbA coding for the double cysteine mutation located on the chromosome and expressed from the native sdbA promoter, TetR, SpecR, KanR Davey et al. (submitted) sdbA coding for the cysteine 86 to proline mutation located on the chromosome and expressed from the native sdbA promoter, TetR, SpecR, KanR This study sdbA coding for the cysteine 89 to alanine point mutation located on the chromosome and expressed from the native sdbA promoter, TetR, SpecR, KanR This study sdbA coding for the cysteine 86 to alanine point mutation located on the chromosome and expressed from the native sdbA promoter, TetR, SpecR, KanR This study sdbA coding for the cysteine 89 to proline point mutation located on the chromosome and expressed from the native sdbA promoter, TetR, SpecR, KanR This study ΔsdbAΔdegP, degP complemented on chromosome and expressed from the native degP promoter, TetR, SpecR, KanR, ErmR This study ΔsdbA, degP coding for the serine to alanine mutation located on the chromosome and expressed from the native degP promoter, TetR, SpecR, KanR, ErmR This study DegP-deficient mutants were constructed in the S. gordonii SecCR1 parent and ΔsdbA mutant using the primers listed in the supplementary Table S1. A single ΔdegP mutant was constructed by insertional inactivation with an erythromycin resistance cassette (ermAM) (Claverys et al., 1995), and a ΔsdbAΔdegP double mutant was constructed using a kanamycin resistance cassette (aphA3) amplified from pDL276 (Dunny et al., 1991). Inactivation of degP in the SdbA cysteine point mutants was achieved by insertion of the ermAM gene. PCR products were digested with restriction enzymes as indicated in Table S1 and ligated together with T4 DNA ligase (New England Biolabs, Whitby, ON, Canada). The ligation products were amplified using the outside primers and the resulting constructs were used to transform S. gordonii SecCR1 and S. gordonii ΔsdbA mutants as described previously (Knight et al., 2008). Transformants were selected on BHI containing the appropriate antibiotics, and insertion of the resistance cassettes was confirmed by PCR. The same strategy was used to mutate spxB with the ermAM cassette. Construction of a degP complemented mutant was achieved by introducing a functional degP gene back onto the chromosome. The entire degP reading frame and a portion of the upstream intergenic region was amplified with the primer pair SL752/SL993. This fragment was digested with BamHI and ligated to aphA3. The ligation product was amplified by PCR and digested with KpnI. This construct was then ligated to a 531 bp segment of the downstream portion of degP amplified with the primers SL822/SL755. The resulting construct was used to transform the ΔdegP mutant, replacing the ermAM cassette with a functional degP gene and a kanamycin resistance marker by double cross-over homologous recombination. Complementation of the degP gene was confirmed by PCR analysis and with Western blots using anti-HtrA (DegP) antisera (1:500 dilution, a gift from Dr. Jeffrey Weiser, University of Pennsylvania). All point mutations were generated on the chromosome and expressed from their native promoters. A serine (ACT) to alanine (ACG) mutation was introduced in the DegP active site at amino acid position 235. The mutation was constructed by overlapping PCR using Phusion high-fidelity DNA polymerase (New England Biolabs) and the primers SL752 and SL991 (upstream), and SL992 and SL993 (downstream) (Heckman and Pease, 2007). The mutated degP gene was digested with KpnI and ligated to aphA3 followed by a 531 bp downstream portion of degP, as described above. The construct was amplified with the primer pair SL752/SL755 and the PCR product was used to transform the ΔdegP mutant by homologous recombination, replacing ermAM with the S235A degP gene. Transformants were selected on BHI with kanamycin, and replica plating was used to identify kanamycin-resistant, erythromycin-sensitive colonies. A similar approach was used to construct SdbA cysteine point mutants. Overlapping PCR with the primer pairs SL756 and SL1039 (upstream), and SL1038 and SL759 (downstream) was used to construct a cysteine (TGT) to alanine (GCT) mutation at position 89. The two fragments were combined as template for PCR and amplified with the outside primers SL756/SL803. The resulting construct was then cut with BamHI and ligated to fragment containing aphA3 followed by a 548 bp portion of the region downstream of sdbA. The ligated DNA was amplified using the outside primers SL756/SL759. This PCR product was used to transform the S. gordonii ΔsdbA mutant by homologous double cross-over recombination, replacing the ermAM gene with the SdbA C89A construct. sdbA was expressed from its own promoter on the chromosome. Transformants were selected on BHI with kanamycin, and replica plating was used to identify kanamycin-resistant, erythromycin-sensitive colonies. The same approach was used to generate SdbA C86A and C89P mutations using the primer pairs SL1102/SL1103 and SL1104/SL1105 respectively. These fragments were ligated to aphA3 and used to transform the ΔsdbA mutant exactly as described above. The sdbAC86P mutant was constructed by a slightly different strategy. The upstream portion of sdbA amplified by the primer pairs SL756/SL975 and the downstream portion of sdbA amplified by the primer pairs SL974/SL803 were cloned into pBluescript. Primers SL974 and SL975 contained the cysteine (TGT) to proline (CAA) mutation. The resulting pBluescriptsdbAC86P was digested with BamHI and ligated to aphA3 and a downstream portion of sdbA. The ligated DNA was amplified using primers SL762/SL759 and transformed into the ΔsdbA mutant. This strategy was used because it created a unique MscI site (TGG TGT → TGG CCA) that could be used as a second approach to quickly confirm the mutation. In addition to PCR and restriction analysis, all point mutations were confirmed by DNA sequencing. The point mutated sdbA variants were amplified from S. gordonii chromosomal DNA with Phusion polymerase using the prime pair SL764/SL763. The PCR product was cleaned using a QIAquick PCR purification kit (Qiagen) and sequenced in both directions using the primers SL764 and SL763. The point mutated degP gene was prepared using the same approach, except that the gene was amplified with the primers SL754/SL755 and sequenced using SL752 and SL993. Plasmids expressing point mutated SdbA variants on the pQE30 expression vector (described below) were also confirmed by sequencing. Plasmids were isolated from E. coli using a QIAprep Spin Miniprep kit (Qiagen) and sequenced using the primers SL764/SL763. Sequencing was carried out at the McGill University and Génome Québec Innovation Centre. Analysis of biofilm formation, eDNA release, autolysis, zymography and bacteriocin production performed as described previously (Davey et al., 2013). Production of scFv and PrsA was tested by Western blotting. Cells from 3 ml of overnight culture grown in HTVG were pelleted by centrifugation (14 000 × g, 3 min, 4°C) and suspended in 50 μl SDS-PAGE sample buffer. Samples were boiled for 5 min and separated by SDS-PAGE, followed by Coomassie blue staining to ensure equal protein loading. The same volume of protein was loaded for Western blotting. ScFv was detected with an anti-HA antibody (Sigma-Aldrich, Oakville, ON, Canada, 1:30 000 dilution), and PrsA was detected using a 1:1000 dilution of mouse anti-PrsA antisera (Davis et al., 2011). To assess the effect of endogenous hydrogen peroxide on scFv production, overnight cultures of spxB+ and ΔspxB strains were diluted 1:20 into fresh HTVG and grown to an optical density at 600 nm of ∼ 0.600. Cells were collected by centrifugation and boiled in SDS-PAGE sample buffer as described above. ScFv was detected with anti-HA, and duplicate samples were probed with anti-PrsA as a loading control. ScFv antibody was purified from the parent and the ΔdegPΔsdbA and sdbAC86P mutants by affinity chromatography. Cultures were grown overnight in 500 ml HTVG. Cells were pelleted by centrifugation and suspended in 50 ml of column wash buffer with urea (50 mM NaH2PO4, 300 mM NaCl, 5 mM imidiazole, 8 M urea) to extract and denature scFv. Denaturing conditions were required to make the hexahistidine tag on scFv accessible for affinity purification. Following overnight incubation at 4°C, the cells were removed by centrifugation at 20 000 × g, 15 min. ScFv was purified from the supernatant with a NiCAM affinity column (Sigma-Aldrich) using standard techniques. Purified scFv was precipitated with 9% (v/v) trichloroacetic acid (TCA) and 0.18% (v/v) deoxycholate on ice for 30 min, followed by centrifugation at 15 000 × g, 10 min, 4°C and washed twice with acetone. Pellets were then suspended in 5 mM EZ-link maleimide-PEG2-biotin (Thermo Scientific) in 100 mM Tris (pH 7.0) and 1% (v/v) SDS, and incubated for 30 min at room temperature, followed by 10 min at 37°C. Excess maleimide-PEG2-biotin was removed by TCA precipitation. The resulting pellets were solubilized in 100 mM Tris (pH 7.0), 1% (v/v) SDS, 8 M urea. To prepare positive controls, the extracts were reduced with 100 mM dithiothreitol (DTT) in 10 mM Tris (pH 8.1) for 30 min at room temperature and TCA precipitated prior to the addition of maleimide-PEG2-biotin. To test the oxidation state of AtlS, 200 ml cultures of the parent and the sdbAC86P mutant were grown to an OD600 = 0.9, and the cells were collected by centrifugation at 10 000 × g, 10 min, 4°C. Cell pellets were suspended in 2 ml of 5 M LiCl and incubated at 4°C with slow rotation to extract surface proteins. Cells were removed by centrifugation at 15 000 × g, 10 min, 4°C, and supernatants containing extracted AtlS were alkylated with maleimide-PEG2-biotin as described above. To specifically detect disulfide-bonded cysteines, iodoacetamide was used to block the free thiol groups. Purified scFv was TCA precipitated and washed twice with ice-cold acetone. The resulting pellets were suspended in 200 mM Tris (pH 8.1), 100 mM iodoacetamide and incubated for 20 min on ice in the dark, followed by TCA precipitation to remove excess iodoacetamide. Disulfide bonds were then reduced with DTT and alkylated with maleimide-PEG2-biotin. To detect biotinlylated proteins, the samples were boiled in sample buffer and run in duplicate on a 12.5% SDS-PAGE. Proteins were transferred to a nitrocellulose membrane and reacted with avidin alkaline phosphatase (1:60 000; Sigma-Aldrich). Total scFv concentration in the samples was used as a loading control and was detected with anti-HA monoclonal antibodies. Experiments were repeated at least three times to ensure reproducibility. Recombinant plasmids to produce SdbAC86P, SdbAC89A and SdbAC86P/C89A were constructed by cloning in frame fragments of the genes amplified with the primer pair SL763/764 into the expression vector pQE-30 (Qiagen) behind an N-terminal His6-tag. Expression and purification of the recombinant proteins was carried out as described previously (Davey et al., 2013). Oxidation of SdbA single cysteine mutants was carried out using 0.5 mg ml−1 protein in 100 mM Tris (pH 8.8), 200 mM KCl, 1 mM EDTA and incubated with 100 mM oxidized glutathione (GSSG, Sigma-Aldrich) for 1 h at room temperature. Excess glutathione was then removed by dialysis against 100 mM sodium phosphate buffer (pH 7), and aliquots were stored at −80°C. Production of reduced, denatured RNase A (Sigma-Aldrich) and analysis of oxidase activity by RNase A catalyzed cCMP hydrolysis were carried out as described previously (Daniels et al., 2010; Davey et al., 2013). Briefly, 10 μM of purified SdbA, or the single cysteine variants, was prepared in a redox buffer containing 0.2 mM GSSG, 0.1 mM GSH, 2 mM EDTA, in 100 mM Tris acetate (pH 8.0). Reduced, denatured RNase A (10 μM) was added and incubated for 2 min prior to the addition of 4.5 mM cCMP substrate. RNase A catalyzed cCMP hydrolysis was detected as the increase in absorbance at 296 nm. The kinetics of RNase A refolding was calculated using a previously described method (Lyles and Gilbert, 1991; Rupp et al., 1994). The reactions were run as described above, using 5–30 μM of reduced denatured RNase A and monitored over a 30 min period. Data analysis was carried out using GraphPad Prism 5 (GraphPad Software). RNaseA folding was also tested by alkylation with maleimide-PEG2-biotin. SdbAC86P or SdbAC89A (20 μM) was combined with reduced, denatured RNaseA (10 μM) in 100 mM Tris buffer pH 8.0, 1 mM EDTA. Proteins were incubated at ambient temperature for 20 min, and the reaction was quenched by the addition of 9% (v/v) TCA. Following TCA precipitation, protein pellets were suspended in column wash buffer and applied to a NiCAM column to trap SdbA via its N-terminal His6-tag. RNaseA remaining in the unbound and wash fractions was pooled and reacted with maleimide-PEG2-biotin as described above. Alkylated proteins were separated on 15% SDS-PAGE gels and stained with Coomassie blue. To test for SdbA-RNaseA mixed disulfides, SdbA was eluted from the column, and the eluate was run on a 15% SDS-PAGE gel under non-reducing conditions. The solvent accessibility of the individual SdbA cysteines was tested by reaction with dithionitrobenzoic acid (DTNB) (Wunderlich et al., 1995). A 15 μM solution of purified reduced SdbA, SdbAC86P or SdbAC89A was prepared in 0.1 M sodium phosphate buffer (pH 8.0). The reaction was started by adding 200 μM of DTNB from a 10 mM stock solution prepared in 0.1 M sodium phosphate buffer (pH 8.0), and the absorbance at 412 nm was recorded. The pH-dependent ionization of the SdbA C86 thiol was followed by the absorbance of the thiolate anion at 240 nm as described previously (Goulding et al., 2004). The pH dependence of the thiolate-specific absorbance signal s = (A240/A280)red/(A240/A280)ox was fitted according to the Henderson-Hasselbalch equation. The pH-dependent absorbance of the cysteine-free SdbAC86P/C89A variant was used as a control. Sulfenylation of SdbA cysteines was detected using the sulfenic acid probe DCP-Bio1 (Millipore), which specifically biotinylates sulfenic acid. Purified His6-SdbA variants (25 μM) were reacted with 1 mM DCP-Bio1 and 100 μM hydrogen peroxide in phosphate-buffered saline (PBS) for 30 min at room temperature. Excess DCP-Bio1 was removed by TCA precipitation. Samples were analyzed by SDS-PAGE and Western blot by probing with avidin alkaline phosphatase. Modification of the SdbA active site cysteines following reaction with glutathione or hydrogen peroxide were tested separately. SdbA was reacted with glutathione as described above. To test for oxidation of the SdbA active site cysteines following exposure to hydrogen peroxide, 25 μM of SdbAC89A or SdbAC86P was reacted with 100 μM hydrogen peroxide and 10 mM dimedone in PBS for 30 min at room temperature. Dimedone stabilizes sulfenic acid, whereas sulfinic and sulfonic acid species can be detected without derivatization. Following the reaction, 5 μg of SdbA protein was electrophoresed on an SDS-PAGE gel under non-reducing conditions. The 18 kDa bands corresponding to SdbA were excised from the gel and prepared for mass spectrometry by trypsin digestion according to Shevchenko et al. with some modifications (Shevchenko et al., 2006). Briefly, reduction with DTT and alkylation with iodoacetamide was omitted in the procedure to allow the detection of cysteine residues. Gel bands were digested with trypsin (Promega, Madison, WI) for 12 h at 37°C. Peptides were extracted from the gel bands by with 100 μl of a 50% acetonitrile-5% formic acid solution. The extract was dried by vacuum centrifugation (SPD SpeedVac Thermo Electron Corp. Waltham, MA); the tryptic peptides were resuspended in 20 μl of a 3% acetonitrile, 0.5% formic acid solution. Liquid chromatography-tandem mass spectrometry (LC-MS/MS) was performed using a nano flow liquid chromatography system (Ultimate3000, ThermoScientific) interfaced to a hybrid ion trap-orbitrap high resolution tandem mass spectrometer (VelosPro, ThermoScientific) operated in data-dependent acquisition (DDA) mode. Briefly, 1 μl of each sample was injected onto a capillary column (C18 Onyx Monolithic, 0.10 × 150 mm Phenomenex) at a flow rate of 300 nl min−1. Samples electro-sprayed at 1.2 kV using a dynamic nanospray probe with fused silica non-coated emitters (20 μM ID with 10 μM ID tip PicoTip Emitter from New Objective). Chromatographic separation was carried out using 90 min linear gradients (Mobile Phase A: 0.1% formic acid in MS-grade water, mobile phase B: 0.1% formic acid in MS-grade acetonitrile,) from 3% B to 35% B over 60 min, then increasing to 95%B over 5 min. MS/MS spectra were acquired using both collision induced dissociation (CID) and higher-energy collisional dissociation (HCD) for the top 15 peaks in the survey 30 000 resolution MS scan. The raw files were acquired (Xcalibur, ThermoFisher) and exported to Proteome Discoverer 2.0 (ThermoFisher) software for peptide and protein identification using SequestHT search algorithm (full trypsin digestion with two maximum missed cleavages, 10 ppm precursor mass tolerance and 0.8 Da fragment mass tolerance). Database searching was done using the UniprotKB E. coli or S. gordonii databases, with the appended SdbA/thioredoxin C86P and C89A mutations, accordingly. Sulfenic acid with dimedone (+138 Da increase in mass), sulfinylation (+32 Da), sulfonylation (+48 Da), and S-glutathionlyation (+305 Da) and oxidized methionines were selected as dynamic (variable) modifications (Table S2). Semi-quantitative analysis was done using Xcalibur QualBrowser by integrating the extracted ion chromatograms (XIC) of each modified and unmodified peptide. XICs were obtained using a 0.02 Da mass window for each peptide precursor. Comparison across the modified and unmodified peptides was always performed on equally charged precursors, (typically 2 + or 3+) whenever the peptides exhibited multiple charged precursors. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium (Vizcaíno et al., 2014) via the PRIDE partner repository with the dataset identifier PXD002827. A 25 μl solution of purified SdbA (1 μM) was prepared in 20 mM phosphate buffer (pH 7.5), 150 mM NaCl, 10 mM EDTA and 5× SYPRO orange (Invitrogen). Melting curve analysis was performed using a LightCycler 2.0 real-time PCR instrument (Roche Diagnostics) by increasing the temperature from 37°C to 95°C with a ramp rate of 0.05°C s−1. During thermal denaturation, continuous fluorescence was captured in all channels. Data were analyzed with melting temperature (Tm) analysis software provided by the manufacturer (version 4.05). The raw data captured from the 610 nm channel were exported and reanalyzed using the Multicode RTx analysis software version 1.6.4.1 (Eragen BioSciences). Results were analyzed by one-way analysis of variance with Bonferroni post tests using GraphPad Prism 5. Funding for this study was provided by the Natural Sciences and Engineering Research Council of Canada (NSERC). LD is a recipient of a NSERC post-graduate scholarship, an IWK graduate studentship and a Nova Scotia Health Research Foundation Scotia Scholars award. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article. Atkinson, H.J., and Babbitt, P.C. (2009) An atlas of the thioredoxin fold class reveals the complexity of function enabling adaptations. PLoS Comput Biol 5: e1000541. Baker, M.D., Gendlina, I., Collins, C.M., and Acharya, K.R. (2004) Crystal structure of a dimeric form of streptococcal pyrogenic exotoxin A ( SpeA1). Protein Sci 1: 2285– 2290. Wiley Online Library Balmer, Y., Koller, A., del Val, G., Manieri, W., Schürmann, P., and Buchanan, B.B. (2003) Proteomics gives insight into the regulatory function of chloroplast thioredoxins. Proc Natl Acad Sci USA 100: 370– 375. Bardwell, J.C., McGovern, K., and Beckwith, J. (1991) Identification of a protein required for disulfide bond formation in vivo. Cell 67: 581– 589. Bardwell, J.C., Lee, J.O., Jander, G., Martin, N., Belin, D., and Beckwith, J. (1993) A pathway for disulfide bond formation in vivo. Proc Natl Acad Sci USA 90: 1038– 1042. Burne, R.A., Wen, Z.T., Chen, Y.M., and Penders, J.E.C. (1999) Regulation of expression of the fructan hydrolase gene of Streptococcus mutans GS-5 by induction and carbon catabolite repression. J Bacteriol 181: 2863– 2871. Chim, N., Riley, R., The, J., Im, S., Segelke, B., Lekin, T., et al. (2010) An extracellular disulfide bond forming protein (DsbF) from Mycobacterium tuberculosis: structural, biochemical, and gene expression analysis. J Mol Biol 396: 1211– 1226. Chng, S.S., Dutton, R.J., Denoncin, K., Vertommen, D., Collet, J.F., Kadokura, H., and Beckwith, J. (2012) Overexpression of the rhodanese PspE, a single cysteine containing protein, restores disulphide bond formation to an Escherichia coli strain lacking DsbA. Mol Microbiol 85: 996– 1006. Wiley Online Library Cho, S.H., and Collet, J.F. (2013) Many roles of the bacterial envelope reducing pathways. Antioxid Redox Signal 18: 1690– 1698. Cho, S.H., Porat, A., Ye, J., and Beckwith, J. (2007) Redox active cysteines of a membrane electron transporter DsbD show dual compartment accessibility. EMBO J 26: 3509– 3520. Wiley Online Library Claverys, J.P., Dintilhac, A., Pestova, E., V, Martin, B., and Morrison, D.A. (1995) Construction and evaluation of new drug resistance cassettes for gene disruption mutagenesis in Streptococcus pneumoniae, using an ami test platform. Gene 164: 123– 128. Collet, J.F., and Messens, J. (2010) Structure, function, and mechanism of thioredoxin proteins. Antioxid Redox Signal 13: 1205– 1216. Daniels, R., Mellroth, P., Bernsel, A., Neiers, F., Normark, S., von Heijne, G., and Henriques-Normark, B. (2010) Disulfide bond formation and cysteine exclusion in Gram positive bacteria. J Biol Chem 285: 3300– 3309. Davey, L., Ng, C.K.W., Halperin, S.A., and Lee, S.F. (2013) Functional analysis of paralogous thiol disulfide oxidoreductases in Streptococcus gordonii. J Biol Chem 288: 16416– 16429. Davis, E., Kennedy, D., Halperin, S.A., and Lee, S.F. (2011) Role of the cell wall microenvironment in expression of a heterologous SpaP-S1 fusion protein by Streptococcus gordonii. Appl Environ Microbiol 77: 1660– 1666. Depuydt, M., Leonard, S.E., Vertommen, D., Denoncin, K., Morsomme, P., Wahni, K., et al. (2009) A periplasmic reducing system protects single cysteine residues from oxidation. Science 326: 1109– 1111. Draskovic, I., and Dubnau, D. (2005) Biogenesis of a putative channel protein, ComEC, required for DNA uptake: membrane topology, oligomerization and formation of disulphide bonds. Mol Microbiol 55: 881– 896. Wiley Online Library Dunny, G.M., Lee, L.N., and LeBlanc, D.J. (1991) Improved electroporation and cloning vector system for Gram positive bacteria. Appl Environ Microbiol 57: 1194– 1201. Dutton, R.J., Boyd, D., Berkmen, M., and Beckwith, J. (2008) Bacterial species exhibit diversity in their mechanisms and capacity for protein disulfide bond formation. Proc Natl Acad Sci USA 105: 11933– 11938. Eser, M., Masip, L., Kadokura, H., Georgiou, G., and Beckwith, J. (2009) Disulfide bond formation by exported glutaredoxin indicates glutathione\'s presence in the E. coli periplasm. Proc Natl Acad Sci USA 106: 1572– 1577. Fomenko, D.E., and Gladyshev, V.N. (2003) Identity and functions of CxxC-derived motifs. Biochemistry 42: 11214– 11225. Glockshuber, R., Schmidt, T., and Plueckthun, A. (1992) The disulfide bonds in antibody variable domains: effects on stability, folding in vitro, and functional expression in Escherichia coli. Biochemistry 31: 1270– 1279. Goulding, C.W., Apostol, M.I., Gleiter, S., Parseghian, A., Bardwell, J., Gennaro, M., and Eisenberg, D. (2004) Gram positive DsbE proteins function differently from Gram negative DsbE homologs. A structure to function analysis of DsbE from Mycobacterium tuberculosis. J Biol Chem 279: 3516– 3524. Hall, A., Nelson, K., Poole, L.B., and Karplus, P.A. (2011) Structure-based insights into the catalytic power and conformational dexterity of peroxiredoxins. Antioxid Redox Signal 15: 795– 815. Han, H., Dong, H., Zhu, S., Zhao, Q., Jiang, L., Wang, Y., et al. (2014) Molecular characterization and analysis of a novel protein disulfide isomerase like protein of Eimeria tenella. PLoS ONE 9: e99914. Heckman, K.L., and Pease, L.R. (2007) Gene splicing and mutagenesis by PCR-driven overlap extension. Nat Protoc 2: 924– 932. Heng, N.C.K., Tagg, J.R., and Tompkins, G.R. (2007) Competence dependent bacteriocin production by Streptococcus gordonii DL1 (Challis). J Bacteriol 189: 1468– 1472. Hodson, C.T.C., Lewin, A., Hederstedt, L., and Le Brun, N.E. (2008) The active site cysteinyls and hydrophobic cavity residues of ResA are important for cytochrome c maturation in Bacillus subtilis. J Bacteriol 190: 4697– 4705. Jakubovics, N.S., Gill, S.R., Vickerman, M.M., and Kolenbrander, P.E. (2008) Role of hydrogen peroxide in competition and cooperation between Streptococcus gordonii and Actinomyces naeslundii. FEMS Microbiol Ecol 66: 637– 644. Wiley Online Library Jeong, W., Jung, Y., Kim, H., Park, S.J., and Rhee, S.G. (2009) Thioredoxin related protein 14, a new member of the thioredoxin family with disulfide reductase activity: implication in the redox regulation of TNF alpha signaling. Free Radic Biol Med 47: 1294– 1303. Jurado, P., Ritz, D., Beckwith, J., de Lorenzo, V., and Fernández, L.A. (2002) Production of functional single chain Fv antibodies in the cytoplasm of Escherichia coli. J Mol Biol 320: 1– 10. Kadokura, H., and Beckwith, J. (2010) Mechanisms of oxidative protein folding in the bacterial cell envelope. Antioxid Redox Signal 13: 1231– 1246. Kim, J., and Senadheera, D. (2012) TcyR regulates L-cystine uptake via the TcyABC transporter in Streptococcus mutans. FEMS Microbiol Lett 328: 114– 121. Wiley Online Library Kishigami, S., Kanaya, E., Kikuchi, M., and Ito, K. (1995) DsbA-DsbB interaction through their active site cysteines. J Biol Chem 270: 17072– 17074. Knight, J.B., Halperin, S.A., West, K.A., and Lee, S.F. (2008) Expression of a functional single chain variable fragment antibody against complement receptor 1 in Streptococcus gordonii. Clin Vaccine Immunol 15: 925– 931. Krojer, T., Sawa, J., Huber, R., and Clausen, T. (2010) HtrA proteases have a conserved activation mechanism that can be triggered by distinct molecular cues. Nat Struct Mol Biol 17: 844– 852. Leichert, L.I., and Jakob, U. (2004) Protein thiol modifications visualized in vivo. PLoS Biol 2: e333. Lewin, A., Crow, A., Oubrie, A., and Le Brun, N.E. (2006) Molecular basis for specificity of the extracytoplasmic thioredoxin ResA. J Biol Chem 281: 35467– 35477. Liu, X., Ramsey, M.M., Chen, X., Koley, D., Whiteley, M., and Bard, A.J. (2011) Real-time mapping of a hydrogen peroxide concentration profile across a polymicrobial bacterial biofilm using scanning electrochemical microscopy. Proc Natl Acad Sci USA 108: 2668– 2673. Lyles, M.M., and Gilbert, H.F. (1991) Catalysis of the oxidative folding of ribonuclease A by protein disulfide isomerase: pre-steady-state kinetics and the utilization of the oxidizing equivalents of the isomerase. Biochemistry 30: 619– 625. Marvaud, J.C., Eisel, U., Binz, T., Niemann, H., and Popoff, M.R. (1998) TetR is a positive regulator of the tetanus toxin gene in Clostridium tetani and is homologous to botR. Infect Immun 66: 5698– 5702. Meima, R., Eschevins, C., Fillinger, S., Bolhuis, A., Hamoen, L.W., Dorenbos, R., et al. (2002) The bdbDC operon of Bacillus subtilis encodes thiol-disulfide oxidoreductases required for competence development. J Biol Chem 277: 6994– 7001. Missiakas, D., Schwager, F., and Raina, S. (1995) Identification and characterization of a new disulfide isomerase like protein (DsbD) in Escherichia coli. EMBO J 14: 3415– 3424. Wiley Online Library Mössner, E., Iwai, H., and Glockshuber, R. (2000) Influence of the pK(a) value of the buried, active site cysteine on the redox properties of thioredoxin like oxidoreductases. FEBS Lett 477: 21– 26. Wiley Online Library Poquet, I., Saint, V., Seznec, E., Simoes, N., Bolotin, A., and Gruss, A. (2000) HtrA is the unique surface housekeeping protease in Lactococcus lactis and is required for natural protein processing. Mol Microbiol 35: 1042– 1051. Wiley Online Library Premkumar, L., Heras, B., Duprez, W., Walden, P., Halili, M., Kurth, F., et al. (2013) Rv2969c, essential for optimal growth in Mycobacterium tuberculosis, is a DsbA like enzyme that interacts with VKOR derived peptides and has atypical features of DsbA like disulfide oxidases. Acta Crystallogr D Biol Crystallogr 69: 1981– 1994. Wiley Online Library Proba, K., Honegger, A., and Plückthun, A. (1997) A natural antibody missing a cysteine in VH: consequences for thermodynamic stability and folding. J Mol Biol 265: 161– 172. Qin, A., Scott, D.W., Rabideau, M.M., Moore, E.A., and Mann, B.J. (2011) Requirement of the CXXC motif of novel Francisella infectivity potentiator protein B FipB, and FipA in virulence of F. tularensis subsp. tularensis. PLoS ONE 6: e24611. Reid, E., Cole, J., and Eaves, D. (2001) The Escherichia coli CcmG protein fulfils a specific role in cytochrome c assembly. Biochem J 58: 51– 58. Ren, G., Stephan, D., Xu, Z., Zheng, Y., Tang, D., Harrison, R.S., et al. (2009) Properties of the thioredoxin fold superfamily are modulated by a single amino acid residue. J Biol Chem 284: 10150– 10159. Roos, G., Foloppe, N., and Messens, J. (2013) Understanding the pK(a) of redox cysteines: the key role of hydrogen bonding. Antioxid Redox Signal 18: 94– 127. Rupp, K., Birnbach, U., Lundström, J., Van, P.N., and Söling, H.D. (1994) Effects of CaBP2, the rat analog of ERp72, and of CaBP1 on the refolding of denatured reduced proteins. Comparison with protein disulfide isomerase. J Biol Chem 269: 2501– 2507. Shevchenko, A., Tomas, H., Havlis, J., Olsen, J., V, and Mann, M. (2006) In gel digestion for mass spectrometric characterization of proteins and proteomes. Nat Protoc 1: 2856– 2860. Shouldice, S.R., Heras, B., Walden, P.M., Totsika, M., Schembri, M.A., and Martin, J.L. (2011) Structure and function of DsbA, a key bacterial oxidative folding catalyst. Antioxid Redox Signal 14: 1729– 1760. Tan, J., Lu, Y., and Bardwell, J.C.A. (2005) Mutational analysis of the disulfide catalysts DsbA and DsbB. J Bacteriol 187: 1504– 1510. Tokuhiro, K., Ikawa, M., Benham, A.M., and Okabe, M. (2012) Protein disulfide isomerase homolog PDILT is required for quality control of sperm membrane protein ADAM3 and male fertility. Proc Natl Acad Sci USA 109: 3850– 3855. Vizcaíno, J.A., Deutsch, E.W., Wang, R., Csordas, A., Reisinger, F., Ríos, D., et al. (2014) ProteomeXchange provides globally coordinated proteomics data submission and dissemination. Nat Biotechnol 32: 223– 226. Walker, K.W., Lyles, M.M., and Gilbert, H.F. (1996) Catalysis of oxidative protein folding by mutants of protein disulfide isomerase with a single active-site cysteine. Biochemistry 35: 1972– 1980. Wang, L., Zhang, L., Niu, Y., Sitia, R., and Wang, C.C. (2014) Glutathione peroxidase 7 utilizes hydrogen peroxide generated by Ero1α to promote oxidative protein folding. Antioxid Redox Signal 20: 545– 556. Wörn, A., and Plückthun, A. (2001) Stability engineering of antibody single chain Fv fragments. J Mol Biol 305: 989– 1010. Wunderlich, M., Otto, A., Maskos, K., Mücke, M., Seckler, R., and Glockshuber, R. (1995) Efficient catalysis of disulfide formation during protein folding with a single active site cysteine. J Mol Biol 247: 28– 33. Yu, J., McLaughlin, S., Freedman, R.B., and Hirst, T.R. (1993) Cloning and active site mutagenesis of Vibrio cholerae DsbA, a periplasmic enzyme that catalyzes disulfide bond formation. J Biol Chem 268: 4326– 4330. Zapun, A., Cooper, L., and Creighton, T.E. (1994) Replacement of the active site cysteine residues of DsbA, a protein required for disulfide bond formation in vivo. Biochemistry 33: 1907– 1914. Zheng, X., Zhang, K., Zhou, X., Liu, C., Li, M., Li, Y., et al. (2013) Involvement of gshAB in the interspecies competition within oral biofilm. J Dent Res 92: 819– 824. The full text of this article hosted at iucr.org is unavailable due to technical difficulties. Please check your email for instructions on resetting your password. If you do not receive an email within 10 minutes, your email address may not be registered, and you may need to create a new Wiley Online Library account. Can\'t sign in? Forgot your username? Enter your email address below and we will send you your username If the address matches an existing account you will receive an email with instructions to retrieve your username