Characteristic of CgAhp
The ncg10877 gene of C. glutamicum is located at bp 969458 to 969940, encoding a putative alkyl hydroperoxide reductase (Ahp) of 160 amino acid residues with a molecular mass of 18.0 kDa. The NCgl0877 protein shared amino acid sequence identities of 51.3%, 53.9%, 52.9%, 50%, and 52.4% with the Cys-X-X-Cys active site motif-containing Ahp of Achromobacter xylosoxidans (ADP19073), Nitrosococcus watsonii (ADJ28176), Citrobacter freundii (AUV27934), Simplicispira suum (AVO42213), and Candidatus nitrosoglobus (BAW80394), respectively (Supplementary Fig. S1A). Ahp proteins were classified into two classes on the basis of functional characterization so far: disulfide oxidoreductases including 2-Cys AhpD and 1-Cys AhpF, and Prx of the thiol-based peroxidase family including AhpC and AhpE (Tartaglia et al.,1990; Wood et al., 2004). According to amino acid sequence, we found that NCgl0877 of C. glutamicum shared a highly conserved Cys-X-X-Cys (C-X-X-C) catalytic signature motif with 2-Cys AhpD, which was not the same as that of peroxide-degrading peroxidase with two active cysteines, such as AhpC (Supplementary Fig. S1B). Two-cysteine (2-Cys) AhpC metabolized peroxides via a conserved NH2-terminal cysteine residue (Cp), which undergoes oxidation. To complete the catalytic cycle, the Cys residue must be reduced via a C-terminal cysteine residue (CR), which was far from Cp, to result in the formation of an intermolecular disulfide bond in AhpC (Bryk et al., 2000; Zhang et al., 2019). 2-Cys AhpD restored oxidized peroxidases to their reduced form by the N- and C-terminal Cys of the Cys-X-X-Cys active site motif (Su et al., 2019). Therefore, we speculated that NCgl0877 of C. glutamicum may be act as oxidized peroxidase-reducing disulfide oxidoreductase but not peroxidase.
According to the catalytic CXXC motif, oxidized peroxidase-reducing disulfide oxidoreductases reported in C. glutamicum so far were classified into six different groups: Trx (Cys-Gly-Pro-Cys (C-G-P-C)), Mrx1(Cys-Pro-Tyr-Cys (C-P-Y-C)), AhpD (Cys-Gly-Thr-Cys (C-G-T-C) and Cys-Val-Tyr-Cys (C-V-Y-C)), DsbA-like Mrx1 (Cys-Pro-Phe-Cys (C-P-F-C)), NrdH (Cys-Val-Gln-Cys (C-V-Q-C)), and Mrx3 (Cys-Gly-Ser-Cys (C-G-S-C)) (Supplementary Fig. S1B). Strikingly, NCgl0877 of C. glutamicum formed a new group, preserving the Cys-Pro-Gly-Cys (C-P-G-C) active-site sequence motif. Moreover, the intervening residues between two cysteines in the catalytic CXXC motif of NCgl0877 were more similar to those of Mrx1 (C-P-Y-C) and DsbA-like Mrx1 (C-P-F-C), but was different from those of AhpD (C-G-T-C or C-V-Y-C) (Supplementary Fig. S1B). Previous studies showed that the difference of the intervening residues between two cysteines in the catalytic CXXC motif caused disulfide oxidoreductases to have different enzymatic rates and substrate preference properties (Rosado et al., 2017). Therefore, the discovery of new C-P-G-C active site motif prompted us to investigate the role of C. glutamicum NCgl0877.
cgahp null mutant was sensitive to organic peroxide stress
Recently, Si et al. (2020) found that the ncgl0877 geneof C. glutamicum (designated cgahp) was one of the main targets of OasR, which was strongly linked to the oxidative stress response in C. glutamicum. Therefore, we speculated that CgAhp may also play a role in oxidative stress response. To gain the physiologically functional insights into the oxidative stress resistance of CgAhp, we investigated the phenotype of a cgahp null mutant in C. glutamicum RES167 strain, obtained by homologous recombination-based gene knock-out, with regards to ROS resistance by an agar-based disc diffusion assay. As shown in Supplementary Fig. S2, although CgAhp was viewed as non-essential in C. glutamicum RES167 under normal growth conditions, the Δcgahp(pXMJ19) strain (the mutant lacking cgahp with the empty plasmid pXMJ19) resulted in decreased tolerance to CHP and t-BHP as compared with WT(pXMJ19) strain (the wild-type C. glutamicum strain with the empty plasmid pXMJ19), giving a significantly larger inhibition zone than WT(pXMJ19) strain. To confirm that the sensitivity to reagents was owing to the absence of cgahp gene, the complementary strain Δcgahp(pXMJ19-cgahp) was constructed by the introduction of plasmid pXMJ19 in trans containing the wild-type C. glutamicum cgahp gene into Δcgahp null mutant and complementation experiments were carried out. As shown in Supplementary Fig. S2B, the complementary strain Δcgahp(pXMJ19-cgahp) produced significantly smaller inhibition zone under CHP and t-BHP again, which were the equivalent of those of the WT(pXMJ19) strains, indicated that resistant phenotypes were almost fully restored in Δcgahp(pXMJ19-cgahp) strains. However, there was no significant difference between the WT(pXMJ19), Δcgahp(pXMJ19), Δcgahp(pXMJ19-cgahp) strains upon H2O2, HClO, diamide, CDNB, IAM, STR, CIP, CdCl2, and NiSO4 challenge. A recent report showed that the lack of oasR gene only resulted in increased resistance to organic peroxides (OPs) and the oasR expression was induced by OPs but not by other oxidants (Si et al., 2020). Moreover, disulfide oxidoreductases Mrx1, Trx and NCgl0018 in C. glutamicum was involved in inorganic oxides, alkylation agents, heavy metal resistance (Li et al., 2021; Che et al., 2020; Chen et al., 2021). These results, combined with up-regulated expression of cgahp in ΔoasR (Si et al., 2020), indicated that OasR specifically responded to OPs to release the inhibition of cgahp, thereby leading to OPs resistance. This phenomenon enabled cells to initiate specific detoxification pathways and respond quickly to environmental pressures. Therefore, CgAhp was involved in OPs stress resistance.
Formation of an intramolecular disulfide bond Cys42-Cys45 under oxidative stress
CgAhp contains a conserved catalytic motif at position 42-45 consisting of C42-P-G-C45 (Supplementary Fig. S1). Sequence alignment indicates that Cys42 might be the nucleophilic cysteine residue, while Cys45 might be the resolving Cys residue (Supplementary Fig. S1). Therefore, we speculated that Cys42 and Cys45 might participate in the formation of disulfide bonds. To confirm this speculation, we mutated the first and the second cysteine of the CXXC motif to serine to gain two variants of CgAhp, namely, CgAhp:C42S and CgAhp:C45S. CgAhp WT and these two variants of CgAhp with and without CHP treatment were used to perform DTNB analysis and NBD-Cl modification. As shown in Fig. 1A, the DTT-treated CgAhp WT contained 1.73 ± 0.45 thiol groups per monomer, but the thiol content decreased to 0.21 ± 0.12 when CgAhp WT was treated with CHP. The difference of 1.51 thiol groups between the two preparations was linked to the fully oxidation of CgAhp WT after CHP treatment.
NBD-Cl can specifically react with free sulfhydryl groups of cysteines and cysteine sulfenic acid, but not with cysteines that are present as sulfinic acid or sulfonic acid. The covalent attachment of NBD-Cl generated an absorption peak at ∼420 nm upon reaction with thiol groups, whereas it peaked at ∼347 nm upon reaction with sulfenic acids (Baker and Poole 2003). Following the reaction with NBD-Cl, the absorption spectra of CgAhp:C42S variants were unchanged before and after exposure to CHP or both CHP and MSH, exhibiting only the 420 nm peak (Fig. 1B). The CgAhp:C42S variant showed one thiol per monomer before and after CHP or both CHP and MSH treatment, indicating that Cys45 was still in thiol form under exposure to CHP (Fig. 1A). However, CgAhp:C45S under CHP treatment lost one thiol group, compared to the thiol content of DTT-treated state, indicating that Cys42 did not exist as a thiol in CHP-treated CgAhp:C45S variant (Fig. 1B). The redox state of thiol in CgAhp:C45S was further examined using NDB-Cl modification. Non-CHP-treated CgAhp:C45S modified with NBD-Cl produced a new covalently attached spectral species with a λmax of 420 nm, consistent with previously characterized thiol adducts with NBD-Cl (Cys-S-NBD). CHP-treated and NBD-labeled CgAhp:C45S had an absorbance maximum (λmax) of 347 nm, representing the NBD-modified product Cys-S(O)-NBD (Ellis and Poole 1997), which clearly signified the detection and trapping of approximately stoichiometric amounts of SOH at Cys42, the only Cys in this variant (Fig. 1B). However, no NBD-Cl labelling peaks occurred in CHP- and MSH-treated CgAhp:C45S proteins. Previous studies showed that MSH reacted with sulfenic acid (Cys-SOH) to form Cys-SSM (Chen et al., 2021; Li et al., 2021). Cys-SSM did not react with NBD-Cl. Therefore, combined with our result that Cys42 was oxidized to form a sulfenic acid (Cys42-SOH), we speculated that S-mycothiolation occurred on Cys42 of CgAhp in the presence of MSH and CHP.
The pKa of the cysteine residues
Previous research has shown the pKa of the nucleophilic cysteine involved in the reaction was a determining factor for the rates of the thiol-disulfide exchange reactions (Jensen et al., 2014). The nucleophilic cysteine in the CXXC motif of oxidoreductases was often in the local electrostatic environment due to the influence of nearby residues (Hansen et al., 2005), leading to the phenomenon that the pKa value of the N-terminal cysteine in the CXXC motif was lower than that of cysteine (8.6) (Lillig et al., 2008). As the low pKa value of the nucleophilic cysteine, the N-terminal cysteine could perform a nucleophilic attack on the substrate disulphide (Lillig et al., 2008). Thus, the pKa of active site residues in CgAhp was determined by recording the absorption at 240 nm during a pH titration (Roos et al., 2013), since the thiolate ion has a higher absorption at this wavelength than the thiol group. As shown in Supplementary Fig. S3, the pKa values of the nucleophilic Cys42 and the resolving Cys45 were less than 6 and 8.39, respectively. The result indicated that the low pKa value made Cys42 function as the nucleophilic Cys. Moreover, the pKa of the Cys45 (8.39) was already lower than the pKa of the MSH sulfur (8.76) (Sharma et al. 2016), which made Cys45 more attack Cys42-SSM mixed disulfide, leading to the formation of a Cys42-Cys45 disulfide. Together, the above results indicated that Cys45 resolved the mixed disulfide Cys42-SSM or Cys42-SOH, leading to the formation of a Cys42-Cys45 disulfide. The result was consistent with the result of Chen et al. reported for C. glutamicum NCgl0018 (2021).
Oxidized CgAhp (CgAhp-S2) was preferentially reduced with electrons from the Lpd/SucB/NADH pathway
Since C. glutamicum used three ubiquitous electron transfer pathways, i.e. the MSH/Mtr/NADPH system, TrxR/NADPH system, and Lpd/SucB/NADH system, to reduce disulphide bonds between the active site cysteines in oxidoreductase, we identified possible electron donor pathways coupled to oxidized CgAhp reduction. To do so, CgAhp was first oxidized with a 10-fold molar excess of diamide to obtain CgAhp-S2 with a single disulfide bond between its active site cysteines (CgAhpox).
Next, CgAhp-S2 was added as a substrate for the electron transfer pathways mentioned above. By monitoring the decrease in the absorption at 340 nm due to NADPH or NADH consumption, we found that oxidized CgAhp-S2 obviously obtained electrons transferred by the Lpd/SucB/NADH pathway, while slight electron transfer was observed when the MSH/Mtr/NADPH or TrxR/NADPH electron transfer pathway was used (Supplementary Figs. S4 A-C). Further, we determined steady-state kinetics. As shown in Supplementary Figs. S4 D-F, the Km value, kcat value, and catalytic coefficient of CgAhp-S2 for the MSH/Mtr/NADPH, TrxR/NADPH, or Lpd/SucB/NADH electron donor pathway were calculated to be 12.51±2.37 μM, 0.03±0.002 s-1, and 2.39±0.08×103 M-1 s-1; 4.85±0.89 μM, 0.11±0.01s-1, and 2.27±0.13×104 M-1 s-1, or 1.21±0.13 μM, 19.61±0.39 s-1, and 1.63±0.04×107 M-1 s-1, respectively. It is worth noting that although CgAhp-S2 could be reduced by the three electron pathways, the catalytic coefficient of CgAhp-S2 with the Lpd/SucB/NADH pathway was several orders of magnitude higher than those with the TrxR/NADPH and MSH/Mtr/NADPH pathways, indicating CgAhp-S2 preferred the Lpd/SucB/NADH pathway. So, CgAhp-S2 seemed to be mainly reduced by the Lpd/SucB/NADH pathway in C. glutamicum but not the TrxR/NADPH and MSH/Mtr/NADPH reducing systems, in line with the result of Su et al. reported for C. glutamicum AhpDs (2019).
CgAhp functioned as a weak peroxidase but not oxidase
As CgAhp played an important role in the resistance to OPs stresses, we therefore examined the possible role of CgAhp as a peroxidase. H2O2 and CHP were added to the Lpd/SucB/NADH pathway in the presence and absence of CgAhp, and the ability of CgAhp to reduce H2O2 and CHP was investigated by following the absorption decrease of NADH at 340 nm (Figs. 2A and B). A very weak peroxidase activity of CgAhp for CHP was observed, as the addition of CgAhp to the reaction mixture containing CHP resulted in a slight increase of NADH consumption at 340 nm. No peroxidase activity of CgAhp for H2O2 was detected. The peroxidase activity of CgAhp was further corroborated by monitoring the consumption of H2O2 and CHP in a Fox assay (Figs. 2C and D), indicating that CgAhp by itself showed minimal peroxidatic activity by linking to the Lpd/SucB/NADH pathway, as reported (Bryk et al., 2002). Thus, the result showed that CgAhp was not thiol-dependent alkyl peroxidases.
To investigate its putative DsbA-oxidoreductase activity, we used E. coli RNase I as a substrate. RNase I was active with its four disulfide bonds correctly formed, making it an ideal model enzyme for oxidative protein folding evaluation (Messens et al., 2007). We used methylene blue intercalated RNA as a substrate to check the oxidase activity at 659 nm after the incubation of reduced unfolded RNase I with CgAhp (Greiner-Stoeffele et al., 1996). CgAhp did not have a capacity of catalyzing the oxidative refolding of RNase I (Fig. 2E). Reduced RNase I (unfolded) demonstrated 14.6% of activity relative to folded RNase I (100%). In contrast, in the presence of E. coli DsbA, which has been proven to be an oxidase, 67.7% of activity was recovered (Rosado et al., 2017). Thus, CgAhp did not act as an oxidase.
CgAhp behaved like C. glutamicum AhpD in regenerating thiol-dependent peroxidase coupled to the Lpd/SucB/NADH electron pathway
To determine whether CgAhp was able to regenerate thiol-dependent peroxidases, the catalytic constants of peroxidase with the CgAhp/Lpd/SucB/NADH system as the recycling reductants were determined under steady-state conditions at saturating concentrations of peroxides and different concentrations of the recycling reductant CgAhp (0 to 500 µM). As shown in Table 1, the kcat and Km values of MPx for CHP with the CgAhp/Lpd/SucB/NADH system were 4.11±0.53 s-1 and 28.92±0.31 µM, respectively. This corresponded to catalytic efficiencies of 14.27×104 M-1 s-1, in accordance with data obtained on AhpD from C. glutamicum (AhpD2, around 34.7×104 M-1 s-1) (Su et al., 2019). Similar results were also observed in Prx, Ohr, and OsmC with the CgAhp/Lpd/SucB/NADH system as the terminal electron acceptor for CHP elimination. The catalytic efficiencies of Prx, Ohr, and OsmC for CHP with the CgAhp/Lpd/SucB/NADH system were 9.3×104 M-1 s-1, 106.9×104 M-1 s-1, and 186.3×104 M-1 s-1, respectively. Although the CgAhp/Lpd/SucB/NADH system supported the peroxidase activities of MPx and Prx when H2O2 was used as substrate, it showed comparably low activity. Of note, the CgAhp/Lpd/SucB/NADH reducing system facilitated Ohr and OsmC activity very poorly when H2O2 was used as substrate, in line with the results of Si et al. (2015a; 2019) reported for only and mainly organic peroxide-detoxifying Ohr and OsmC in C. glutamicum, respectively. Previous studies showed that Ohr and OsmC could employ Trx regeneration systems in reducing CHP substrate (the catalytic efficiencies of Ohr and OsmC were 10×104 and 21.2×104, respectively), they had higher affinity for CgAhp than Trx in vitro. These data indicated that the CgAhp/Lpd/SucB/NADH system was more efficient in supporting the peroxidase activity of Ohr and OsmC when CHP was used as substrate but not H2O2. Moreover, CgAhp preferentially supported the peroxidase activity of Ohr and OsmC. When CHP was used as substrate, the catalytic efficiencies of MPx and Prx with the CgAhp/Lpd/SucB/NADH system was significantly lower than data obtained on C. glutamicum MPx and Prx with the Trx system (MPx, 58.5×104 M−1 s−1; Prx, 264.1×104 M−1 s−1) (Si et al., 2015b, 2017), and about 8-20 times lower than those of the C. glutamicum Ohr and OsmC-catalyzed reaction with the CgAhp/Lpd/SucB/NADH system. In addition, the catalytic coefficients of MPx and Prx with the CgAhp/MSH/Mtr/NADPH and CgAhp/TrxR/NADPH system as the recycling reductants were detected. As shown in Table 1, although CgAhp could support the peroxidase activity of MPx and Prx linked to MSH/Mtr/NADPH or TrxR/NADPH systems, their catalytic coefficients were only at a low rate of 3-70 M-1 s-1, several orders of magnitude lower than those with the Lpd/SucB/NADH pathway, indicating that CgAhp linked to the CgAhp/MSH/Mtr/NADPH or CgAhp/TrxR/NADPH system was highly unlikely in vivo.
All together, these results indicated that (i) CgAhp preferably produced a robust, NADH-dependent, peroxidase activity of Ohr and OsmC; (ii) CgAhp was an important cytoplasmic alkyl hydroperoxide oxidoreductase involved in regeneration of oxidized peroxidase; (iii) CgAhp was linked to the Lpd/SucB/NADH pathway.
Disulfide bond reduction by CgAhp using electrons from the Lpd/SucB/NADH pathway
Peroxidase MPx and Prx metabolizing peroxides in vivo could form two different states of intramolecular disulfide bond (S-S) and the protein-MSH mixed disulfide (Si et al., 2015b, 2017). CgAhp, supporting the peroxidase activity of MPx and Prx, had C-P-X-C motif that was similar to that of Mrx1 and DsbA-like Mrx1 (Supplementary Fig. S1). Therefore, we investigated the possible role of CgAhp in reducing the intramolecular disulfide (S-S) and mixed disulfide of peroxidase in a coupled Lpd/SucB/NADH pathway. Because once determined that Cys64 of MPx and Cys84 of Prx had no direct function in the catalytic mechanism and that Cys79 of MPx and Cys97 of Prx formed a disulfide with Cys36 of MPx and Cys63 of Prx at the end of the catalytic cycle, respectively, we decided to use MPx:C64S and Prx:C84S to study the follow experiments (Si et al., 2015b, 2017). We employed in vitro assay system (see Materials and Methods) by using pre-oxidized disulfide bonded MPx:C64S-S2 and Prx:C84S-S2 as substrates together with C. glutamicum Lpd, SucB and NADH. From the Michaelis-Menten kinetic plot, we obtained kcat of 0.06 ± 0.01 s−1 and 0.05± 0.01 s−1, and Km of 2.07 ± 0.17 μM and 2.33 ± 0.22 μM, which result in a specificity constant (kcat/Km) of 2.84× 104 M−1s −1 and 2.32× 104 M−1s −1, respectively (Figs. 3A and B). Compared with MPx with C. glutamicum Trx (8.43× 104 M −1 s−1) (Perde et al., 2015), the catalytic efficiency of MPx disulfide reduction with CgAhp is slightly lower. To further assess whether CgAhp possessed general thiol-disulfide redox activity as Trx, we further tested the capacity of recombinant CgAhp to reduce disulfide compound Insulin. Insulin, containing disulfides, was typically used for determining the ability of protein disulfide reduction (Holmgren et al., 1979). We therefore detected the ability of CgAhp/Lpd/SucB/NADH system to reduce insulin. As shown in Table 2, CgAhp was shown to reduce insulin disulfides in the presence of the Lpd/SucB/NADH pathway. Its rate of precipitation was higher than those of Mrx1 (22.67 ± 0.02) and DsbA-like Mrx1 Rv2466c (13.03 ± 0.02) from M. tuberculosis (Rosado et al., 2017). These results demonstrated that CgAhp was almost as effective as Trx in reducing disulfide bonds.
CgAhp reduced intramolecular disulfide bond via a dithiol mechanism
To check whether CgAhp used one active site cysteine or two active site cysteines in the reaction process, we mutated the first, the second, and both cysteines in the catalytic CXXC active site motif of CgAhp to serine, respectively, and we expressed and purified CgAhp:C42S, CgAhp:C45S, and CgAhp:C42SC45S mutants to homogeneity. The functionalities of the CgAhp:C42S, CgAhp:C45S, and CgAhp:C42SC45S to reducing oxidized MPx:C64S-S2 and Prx:C84S-S2 were tested in progress curves by following the oxidation of NADH in the presence of Lpd and SucB. As shown in Figs. 4A and B, electron transfer was almost the same as back ground levels when CgAhp:C42S, CgAhp:C45S, or CgAhp:C42SC45S was present, indicating that both cysteines in the CXXC motif of CgAhp were essential for transferring electron for peroxidase. Only the sample with oxidized MPx:C64S-S2 and Prx:C84S-S2 coupled to the CgAhp/Lpd/SucB/NADH electron transfer pathway showed consumption of NADH. Mutants could not replace CgAhp WT. As such, CgAhp was functioning as a dithiol reductase with essential N-terminal cysteine and C-terminal cysteine.
OasR negatively regulated CgAhp expression in C. glutamicum
The expression of C. glutamicum AhpD was induced by oxidative stress (Su et al., 2019). Moreover, cgahp mutants exhibited sensitivity to organic peroxide. Therefore, to test whether cgahp expression responded to OPs, qRT-PCR at the transcriptional level was performed. Transcriptional and translational lacZY genes fused to the probable promoter region of cgahp were constructed as described in Materials and Methods. The β-Galactosidase activity of Pcgahp::lacZY chromosomal promoter fusion reporter was determined in bacterial cells treated with different concentrations of CHP and t-BHP (Fig. 5A). The results showed that the increased levels of β-Galactosidase activity were attributed to cgahp promoter in the CHP- and t-BHP-induced WT(pXMJ19)(Pcgahp::lacZY) reporter strains. The enhanced β-Galactosidase activity observed under induction (Fig. 5A) was consistent with the increased mRNA levels in the WT(pXMJ19) strains induced by CHP- and t-BHP by qRT-PCR analysis (Fig. 5B). This indicated that CHP- and t-BHP induced the expression of cgahp in this pathway, thereby increasing the resistance of C. glutamicum to OPs conditions.
Recently, Si et al. (2020) found CgAhp was one of the main targets of OasR by microarray analysis, which was strongly linked to the oxidative stress response in C. glutamicum. Therefore, we detected OasR's regulatory capacity for CgAhp. As shown in Fig. 5A, the β-Galactosidase activity were approximately four times higher in strain ΔoasR(pXMJ19) (strain lacking oasR gene contained empty pXMJ19) than in the WT(pXMJ19) strain, indicating that the cgahp promoter was negatively regulated by OasR. The negative regulation of cgahp by OasR was also confirmed by qRT-PCR, with the observation that the mRNA levels of cgahp increased by 4-fold in the mutant ΔoasR(pXMJ19) mutant and restored to the wild-type level in the complemented strain ∆oasR(pXMJ19-oasR) (Fig. 5B). Moreover, transcription of cgahp was not increased in the
ΔoasR(pXMJ19) mutant under the CHP- and t-BHP-induced conditions. However, cgahp transcription increased WT(pXMJ19) strain (Figs. 5A and B). To further determine whether OasR regulated CgAhp expression directly, we examined the interaction between OasR and the CgAhp promoter using EMSA. Incubation of a 220-bp DNA element containing the cgahp promoter (Pcgahp) sequence (-220 to -1 relative to the ATG start codon of the first ORF of the cgahp gene) with His6-OasR led to the formation of DNA-protein complexes, and the abundance of such complexes depended on the amount of OasR (Fig. 5C left panel). However, both a 220-bp control DNA fragment amplified from the cgahp coding open reading frame region and BSA instead of His6-OasR showed no detectable binding (Fig. 5C, lane 6 and 7). Thus, OasR directly repressed the expression of cgahp.