A conserved sequence motif in the Escherichia coli soluble FAD-containing pyridine nucleotide transhydrogenase is important for reaction efficiency

Soluble pyridine nucleotide transhydrogenases (STHs) are flavoenzymes involved in the redox homeostasis of the essential cofactors NAD(H) and NADP(H). They catalyze the reversible transfer of reducing equivalents between the two nicotinamide cofactors. The soluble transhydrogenase from Escherichia coli (SthA) has found wide use in both in vivo and in vitro applications to steer reducing equivalents toward NADPH-requiring reactions. However, mechanistic insight into SthA function is still lacking. In this work, we present a biochemical characterization of SthA, focusing for the first time on the reactivity of the flavoenzyme with molecular oxygen. We report on oxidase activity of SthA that takes place both during transhydrogenation and in the absence of an oxidized nicotinamide cofactor as an electron acceptor. We find that this reaction produces the reactive oxygen species hydrogen peroxide and superoxide anion. Furthermore, we explore the evolutionary significance of the well-conserved CXXXXT motif that distinguishes STHs from the related family of flavoprotein disulfide reductases in which a CXXXXC motif is conserved. Our mutational analysis revealed the cysteine and threonine combination in SthA leads to better coupling efficiency of transhydrogenation and reduced reactive oxygen species release compared to enzyme variants with mutated motifs. These results expand our mechanistic understanding of SthA by highlighting reactivity with molecular oxygen and the importance of the evolutionarily conserved sequence motif. Soluble pyridine nucleotide transhydrogenases (STHs) are flavoenzymes involved in the redox homeostasis of the essential cofactors NAD(H) and NADP(H). They catalyze the reversible transfer of reducing equivalents between the two nicotinamide cofactors. The soluble transhydrogenase from Escherichia coli (SthA) has found wide use in both in vivo and in vitro applications to steer reducing equivalents toward NADPH-requiring reactions. However, mechanistic insight into SthA function is still lacking. In this work, we present a biochemical characterization of SthA, focusing for the first time on the reactivity of the flavoenzyme with molecular oxygen. We report on oxidase activity of SthA that takes place both during transhydrogenation and in the absence of an oxidized nicotinamide cofactor as an electron acceptor. We find that this reaction produces the reactive oxygen species hydrogen peroxide and superoxide anion. Furthermore, we explore the evolutionary significance of the well-conserved CXXXXT motif that distinguishes STHs from the related family of flavoprotein disulfide reductases in which a CXXXXC motif is conserved. Our mutational analysis revealed the cysteine and threonine combination in SthA leads to better coupling efficiency of transhydrogenation and reduced reactive oxygen species release compared to enzyme variants with mutated motifs. These results expand our mechanistic understanding of SthA by highlighting reactivity with molecular oxygen and the importance of the evolutionarily conserved sequence motif. Pyridine nucleotide transhydrogenases catalyze the transfer of reducing equivalents between the two nicotinamide cofactors NAD(H) and NADP(H) and contribute to cellular redox homeostasis (1Hoek J.B. Rydstrom J. Physiological roles of nicotinamide nucleotide transhydrogenase.Biochem. J. 1988; 254: 1-10Crossref PubMed Scopus (300) Google Scholar, 2Blank L.M. Ebert B.E. Buehler K. Bühler B. Redox biocatalysis and metabolism: molecular mechanisms and metabolic network analysis.Antioxid. Redox Signal. 2010; 13: 349-394Crossref PubMed Scopus (83) Google Scholar). Besides the well-studied membrane transhydrogenases (EC 1.6.1.2) that couple proton transfer across the cell membrane to cofactor transhydrogenation (3Jackson J.B. Proton translocation by transhydrogenase.FEBS Lett. 2003; 545: 18-24Crossref PubMed Scopus (73) Google Scholar), also soluble transhydrogenases (STHs) exist (4Argyrou A. Blanchard J.S. Flavoprotein disulfide reductases: advances in chemistry and function.Prog. Nucl. Acid Res. Mol. Biol. 2004; 78: 89-142Crossref PubMed Scopus (176) Google Scholar). STHs are energy-independent FAD-containing enzymes catalyzing the reversible reaction:NAD++ NADPH ⇌NADH +NADP+(1) STHs (EC 1.6.1.1) are evolutionary related to the flavoprotein disulfide reductase (FDR) family (4Argyrou A. Blanchard J.S. Flavoprotein disulfide reductases: advances in chemistry and function.Prog. Nucl. Acid Res. Mol. Biol. 2004; 78: 89-142Crossref PubMed Scopus (176) Google Scholar). The two families share conserved sequence motifs including the GXGXXG motif involved in binding the nucleotide cofactors NAD(P)+ and FAD (5Kleiger G. Eisenberg D. GXXXG and GXXXA motifs stabilize FAD and NAD(P)-binding rossmann folds through Cα-H···O hydrogen bonds and van der waals interactions.J. Mol. Biol. 2002; 323: 69-76Crossref PubMed Scopus (157) Google Scholar) but also show notable differences including a strongly conserved threonine (CXXXXT) in STHs instead of the redox-active disulfide motif CXXXXC in FDRs and a tyrosine replacing histidine in the His-Glu pair (YXXXXE in STHs) in the terminal motif. The main function of STHs in vivo is to oxidize excess of NADPH (6Sauer U. Canonaco F. Heri S. Perrenoud A. Fischer E. The soluble and membrane-bound transhydrogenases UdhA and PntAB have divergent functions in NADPH metabolism of Escherichia coli.J. Biol. Chem. 2004; 279: 6613-6619Abstract Full Text Full Text PDF PubMed Scopus (457) Google Scholar), forming NADH which is thus made available to supply electrons to the respiratory chain (7Zhao H. Wang P. Huang E. Ge Y. Zhu G. Physiologic roles of soluble pyridine nucleotide transhydrogenase in Escherichia coli as determined by homologous recombination.Ann. Microbiol. 2008; 58: 275-280Crossref Scopus (10) Google Scholar). As crossroads of the intracellular redox status, STHs from Escherichia coli and Pseudomonas fluorescences have found wide use in improving the yield of value-added chemicals produced in cell factories. Indeed, by transferring electrons between nicotinamide carriers, transhydrogenases replenish the NAD(P)H pool required by the metabolic pathway of the target compound. However, most metabolic engineering reports employing STHs benefit from the NADPH generation at the expense of NADH oxidation (8Sánchez A.M. Andrews J. Hussein I. Bennett G.N. San K.Y. Effect of overexpression of a soluble pyridine nucleotide transhydrogenase (UdhA) on the production of poly(3-hydroxybutyrate) in Escherichia coli.Biotechnol. Prog. 2006; 22: 420-425Crossref PubMed Scopus (94) Google Scholar, 9Wang Z. Gao C. Wang Q. Liang Q. Qi Q. Production of pyruvate in Saccharomyces cerevisiae through adaptive evolution and rational cofactor metabolic engineering.Biochem. Eng. J. 2012; 67: 126-131Crossref Scopus (34) Google Scholar, 10Jan J. Martinez I. Wang Y. Bennett G.N. San K.Y. Metabolic engineering and transhydrogenase effects on NADPH availability in Escherichia coli.Biotechnol. Prog. 2013; 29: 1124-1130Crossref PubMed Scopus (33) Google Scholar, 11Xu W. Yao J. Liu L. Ma X. Li W. Sun X. et al.Improving squalene production by enhancing the NADPH/NADP+ ratio, modifying the isoprenoid-feeding module and blocking the menaquinone pathway in Escherichia coli.Biotechnol. Biofuels. 2019; 12: 1-9Crossref PubMed Scopus (29) Google Scholar), a reaction that corresponds to the opposite direction proposed as native function of STHs (Equation 1). Similarly to their use in metabolic engineering, the emerging field of cell-free synthetic biology (12Shi T. Han P. You C. Zhang Y.H.P.J. An in vitro synthetic biology platform for emerging industrial biomanufacturing: bottom-up pathway design.Synth. Syst. Biotechnol. 2018; 3: 186-195Crossref PubMed Scopus (39) Google Scholar) (sometimes referred to as “synthetic biochemistry” (13Bowie J.U. Sherkhanov S. Korman T.P. Valliere M.A. Opgenorth P.H. Liu H. Synthetic biochemistry: the bio-inspired cell-free approach to commodity chemical production.Trends Biotechnol. 2020; 38: 766-778Abstract Full Text Full Text PDF PubMed Scopus (81) Google Scholar)) has also shown the value of STHs to sustain the synthesis of diverse classes of biomolecules, from opioids as hydromorphone (14Boonstra B. Rathbone D.A. French C.E. Walker E.H. Bruce N.C. Cofactor regeneration by a soluble pyridine nucleotide transhydrogenase for biological production of hydromorphone.Appl. Environ. Microbiol. 2000; 66: 5161-5166Crossref PubMed Scopus (63) Google Scholar) to fatty acids derivatives as in the case of p-nitrophenoxydecanoic acid (15Mouri T. Shimizu T. Kamiya N. Goto M. Ichinose H. Design of a cytochrome P450BM3 reaction system linked by two-step cofactor regeneration catalyzed by a soluble transhydrogenase and glycerol dehydrogenase.Biotechnol. Prog. 2009; 25: 1372-1378Crossref PubMed Scopus (25) Google Scholar), up to the formation of antioxidant species such as reduced glutathione within phospholipidic compartments (16Partipilo M. Ewins E.J. Frallicciardi J. Robinson T. Poolman B. Slotboom D.J. Minimal pathway for the regeneration of redox cofactors.JACS Au. 2021; 1: 2280-2293Crossref PubMed Scopus (11) Google Scholar). These last two examples exploit the NADPH-generating action of the soluble transhydrogenase from E. coli, catalyzing the supply of demanded reactants. The above highlights the reversible nature of the enzymatic transhydrogenation rendering transhydrogenases attractive and flexible biocatalysts. Despite a partial biochemical characterization (17Cao Z. Song P. Xu Q. Su R. Zhu G. Overexpression and biochemical characterization of soluble pyridine nucleotide transhydrogenase from Escherichia coli.FEMS Microbiol. Lett. 2011; 320: 9-14Crossref PubMed Scopus (18) Google Scholar) of the soluble transhydrogenase SthA (also known as UdhA) from E. coli, numerous mechanistic aspects remain to be elucidated. Not only is the mechanism by which NADPH is formed still obscure but also possible unwanted side reactions of the flavoprotein with molecular oxygen have not been studied (18Chaiyen P. Fraaije M.W. Mattevi A. The enigmatic reaction of flavins with oxygen.Trends Biochem. Sci. 2012; 37: 373-380Abstract Full Text Full Text PDF PubMed Scopus (182) Google Scholar, 19Gran-Scheuch A. Parra L. Fraaije M.W. Systematic assessment of uncoupling in flavoprotein oxidases and monooxygenases.ACS Sustain. Chem. Eng. 2021; https://doi.org/10.1021/acssuschemeng.1c02012Crossref Scopus (5) Google Scholar, 20Mattevi A. To be or not to be an oxidase: challenging the oxygen reactivity of flavoenzymes.Trends Biochem. Sci. 2006; 31: 276-283Abstract Full Text Full Text PDF PubMed Scopus (236) Google Scholar). Insight into these properties of SthA would facilitate its broader use in metabolic engineering and industrial biomanufacturing. In this work, we present the study of the purified soluble transhydrogenase (SthA) from E. coli. We initially characterize the reaction in which (thio)NADPH is produced from NADH. Then, we describe a hitherto unreported reactivity of SthA with molecular oxygen in the absence of oxidized cofactors as electron acceptors. This reaction involves full reduction of FAD and it leads to the formation of superoxide anion and hydrogen peroxide (H2O2). The reactivity of the flavoprotein with dioxygen and production of reactive oxygen species (ROS) also takes place during transhydrogenation in the presence of the oxidized cofactor (thio)NADP+ as electron acceptor, although only as minor side-reaction (2%) compared to the hydride transfer between nicotinamide cofactors. Finally, we explore the characteristic CXXXXT motif in the FAD-binding domain of SthA, both by replacing cysteine (C45A) and by restoring the redox-active disulfide center typical of FDRs via threonine mutagenesis (T50C). We determined the ability of C45A and T50C to bind FAD and to perform transhydrogenation, as well as to produce ROS. These SthA mutants reveal that the conservation of the CXXXXT motif in STHs is necessary for efficient enzymatic transhydrogenation. SthA from E. coli K-12 with a 10x-histidine tag at the C-terminus was overproduced in E. coli MC 1061 (Fig. 1A). The yellow protein was purified from the soluble fraction by affinity chromatography followed by size-exclusion chromatography (SEC profile shown in Fig. 1B). The absorbance spectrum of the purified protein indicated the presence of bound oxidized FAD with absorbance maxima at 370 nm and 450 nm (Fig. 1C). For accurate determination of enzyme concentration (21Aliverti A. Curti B. Vanoni M.A. Identifying and quantitating FAD and FMN in simple and in iron-sulfur-containing flavoproteins.Met. Mol. Biol. 1999; 131: 9-23PubMed Google Scholar), the extinction coefficient at 450 nm was determined: ε450 = 12.1 mM−1 cm−1. The protein retained its intense yellow color during the purification protocol suggesting a tightly bound FAD cofactor. This was confirmed by measuring the ratio between the wavelengths at 280 and 450 nm (A280/450): a value of around 6 indicated that most protein is in its holo form, although we could detect some small loss of FAD upon SEC (Fig. S1). The protein had a tendency to aggregate upon storage at -80 °C, which was prevented by the addition of glycerol (5–10%). SthA protomers have a calculated molar mass of around 54 kDa as confirmed by SDS-PAGE analysis (Fig. 1A). Yet, the hydrodynamic radius of 8.2 ± 0.2 nm measured by dynamic light scattering (DLS) indicated a molar mass of 467.3 ± 23.0 kDa (Fig. 1D), suggesting an octameric native state. This result was confirmed by interpolating the elution volume upon SEC using standards of known molecular weight (Fig. S2), obtaining a molar mass of 437.7 ± 13.9 kDa. Electron microscopy of the negatively stained–purified SthA (Fig. S3) showed particles with a size matching the hydrodynamic radius obtained by DLS. Although there are several studies demonstrating the SthA capacity to mediate the hydride transfer from NADH to NADP+ (8Sánchez A.M. Andrews J. Hussein I. Bennett G.N. San K.Y. Effect of overexpression of a soluble pyridine nucleotide transhydrogenase (UdhA) on the production of poly(3-hydroxybutyrate) in Escherichia coli.Biotechnol. Prog. 2006; 22: 420-425Crossref PubMed Scopus (94) Google Scholar, 9Wang Z. Gao C. Wang Q. Liang Q. Qi Q. Production of pyruvate in Saccharomyces cerevisiae through adaptive evolution and rational cofactor metabolic engineering.Biochem. Eng. J. 2012; 67: 126-131Crossref Scopus (34) Google Scholar, 10Jan J. Martinez I. Wang Y. Bennett G.N. San K.Y. Metabolic engineering and transhydrogenase effects on NADPH availability in Escherichia coli.Biotechnol. Prog. 2013; 29: 1124-1130Crossref PubMed Scopus (33) Google Scholar, 11Xu W. Yao J. Liu L. Ma X. Li W. Sun X. et al.Improving squalene production by enhancing the NADPH/NADP+ ratio, modifying the isoprenoid-feeding module and blocking the menaquinone pathway in Escherichia coli.Biotechnol. Biofuels. 2019; 12: 1-9Crossref PubMed Scopus (29) Google Scholar, 15Mouri T. Shimizu T. Kamiya N. Goto M. Ichinose H. Design of a cytochrome P450BM3 reaction system linked by two-step cofactor regeneration catalyzed by a soluble transhydrogenase and glycerol dehydrogenase.Biotechnol. Prog. 2009; 25: 1372-1378Crossref PubMed Scopus (25) Google Scholar, 16Partipilo M. Ewins E.J. Frallicciardi J. Robinson T. Poolman B. Slotboom D.J. Minimal pathway for the regeneration of redox cofactors.JACS Au. 2021; 1: 2280-2293Crossref PubMed Scopus (11) Google Scholar), a thorough kinetic study of SthA is still lacking. First, we investigated which buffer and pH conditions are optimal for the hydride transfer from NADH to thioNADP+ (Fig. 2A) by following the reduction of the latter. The replacement of the oxygen with a sulfur in the amide side-chain of the nicotinamide moiety of thioNADP+ allows to monitor the cofactor reduction at 400 nm, discriminating it from NAD(P)H which instead absorbs at 340 nm (22RydströM J. Hoek J.B. Ernster L. 2 nicotinamide nucleotide transhydrogenases.Enzyme. 1976; 13: 51-88Crossref Scopus (61) Google Scholar). The optimal buffer for (thio)NADPH generation was 100 mM Tris, tested in the pH range of 7.5 to 9.0, showing the highest SthA catalytic performance at pH 8.0. In potassium phosphate (KPi) buffer, the activity of SthA was strongly reduced. Very low activity (∼2% when compared with Tris, pH 8.0) was detected in KPi at pH 6.0. At pH values of 7.5 and 8.0, the activity in KPi buffer was over 6-fold lower than in Tris buffer of the same pH, indicative of inhibition by phosphate (Fig. 2B). By increasing the KPi concentration, we observed that the SthA-transhydrogenation activity leveled off to a basal activity of 15% at 50 mM or higher. When potassium was replaced with sodium phosphate, we observed the same inhibitory trend but with slightly higher basal activity values (20–25%) than those of KPi. The low activity was ascribed exclusively to the presence of phosphate since we did not detect any effect on the SthA-mediated transhydrogenation at high ionic strength with sodium or potassium chloride (Fig. S4). Then, we determined the kinetic parameters (Fig. 2C and Table 1) of the NADH-consuming reaction catalyzed by SthA. Fixing the amount of thioNADP+ at 0.5 mM, we calculated an apparent affinity constant (KM) for NADH (Fig. 2C, on the left) of 2.6 ± 0.4 mM, with a turnover number (kCAT) of 9.2 ± 0.5 s−1. At high NADH concentrations (above 7 mM), substrate inhibition was observed, with an estimated inhibition constant (KI) of 12.1 ± 2.0 mM. In the case of thioNADP+ (Fig. 2C, on the right), we estimated a significantly lower KM (121 ± 40 μM) than that of NADH (kept constant at 10.0 mM during the kinetics), while the kCAT value was similar (15.3 ± 0.5 s−1) to the one calculated for the reduced cofactor. Substrate inhibition was again evident at high concentrations of thioNADP+, resulting in a KI value of 585 ± 257 μM.Table 1Kinetic parameters of SthA for the transhydrogenase activity (upper panel) and oxidase activity (lower panel)Transhydrogenase activity: NADH + thioNADP+ → NAD+ + thioNADPHSubstrateKM (mM)KI (mM)VMAX (μmolˑ min−1ˑmg1)kCAT (s−1)kCAT/KM (s−1ˑmM−1)CosubstrateNADH2.6 ± 0.412.1 ± 2.010.2 ± 0.59.2 ± 0.53.60.5 mM thioNADP+thioNADP+0.12 ± 0.040.59 ± 0.2617.0 ± 3.515.3 ± 3.1127.510.0 mM NADHOxidase activity: NAD(P)H + O2 → NAD(P)+ + H2O2 + O2•-SubstrateApparent KM (μM)KI (mM)Apparent VMAX (nmolˑ min−1ˑmg1)Apparent kCAT (s−1)kCAT/KM (s−1ˑmM−1)CosubstrateNADH49.1 ± 10.4--137.1 ± 8.80.12 ± 0.012.40.2 mM O2 (53Reynafarje B. Costa L.E. Lehninger A.L. O2 solubility in aqueous media determined by a kinetic method.Anal. Biochem. 1985; 145: 406-418Crossref PubMed Scopus (198) Google Scholar)NADPH91.8 ± 30.5--221.0 ± 28.30.20 ± 0.032.20.2 mM O2The measurements were carried out in 50 mM Tris, pH 7.5 at 30 °C in biological triplicate (n = 3, the errors indicate the s.e.m.), using for each of them single technical replicates. Open table in a new tab The measurements were carried out in 50 mM Tris, pH 7.5 at 30 °C in biological triplicate (n = 3, the errors indicate the s.e.m.), using for each of them single technical replicates. Since the nucleotide adenine cofactors AMP, ADP, and ATP are reported as activators of SthA for the reaction in the opposite direction (production of NADH) (17Cao Z. Song P. Xu Q. Su R. Zhu G. Overexpression and biochemical characterization of soluble pyridine nucleotide transhydrogenase from Escherichia coli.FEMS Microbiol. Lett. 2011; 320: 9-14Crossref PubMed Scopus (18) Google Scholar), we investigated whether they also affected the reaction generating (thio)NADPH (Fig. 2D). Indeed, at the concentration of 5.0 mM, all three adenine nucleotides increased the rate of thioNADPH formation (reported as percentages in Table S1). ADP was the best reaction activator, followed by AMP and ATP, respectively. In recent years, a growing number of reports in the literature have shown how various flavoenzymes display side reactivity with molecular oxygen, forming ROS (19Gran-Scheuch A. Parra L. Fraaije M.W. Systematic assessment of uncoupling in flavoprotein oxidases and monooxygenases.ACS Sustain. Chem. Eng. 2021; https://doi.org/10.1021/acssuschemeng.1c02012Crossref Scopus (5) Google Scholar, 23Messner K.R. Imlay J.A. Mechanism of superoxide and hydrogen peroxide formation by fumarate reductase, succinate dehydrogenase, and aspartate oxidase.J. Biol. Chem. 2002; 277: 42563-42571Abstract Full Text Full Text PDF PubMed Scopus (230) Google Scholar). We decided to test if SthA can also catalyze the formation of ROS. We hypothesized that this uncoupling activity (Fig. 3A) would start with the reduction of FAD into FADH2 by NAD(P)H and then be followed by the reoxidation of the reduced flavoenzyme by dioxygen. Such transfer of reducing equivalents to the final donor (O2) would lead to the formation of the ROS species superoxide anion and/or H2O2. We tested this hypothesis by the addition of NADH to a reaction mixture containing SthA and devoid of any oxidized cofactor NAD(P)+ with or without oxygen (Fig. 3, B and C). By monitoring the absorbance at 340 and 450 nm with a stopped-flow setup (full absorbance spectra available on Fig. S5), we followed at the same time the redox status of both the nicotinamide and the flavin-embedded cofactors. Using the stopped-flow instrument, we could not only monitor rapid kinetics but it also allowed to control the oxygen concentration in the reactions (24van Berkel W.J. Benen J.A. Eppink M.H. Fraaije M.W. Flavoprotein kinetics.Flavoprotein Protoc. 1999; 131: 61-85Crossref Google Scholar, 25Valentino H. Sobrado P. Performing anaerobic stopped-flow spectrophotometry inside of an anaerobic chamber.Met. Enzymol. 2019; 620 (Elsevier Inc, 620, 51–88): 51-88Crossref PubMed Scopus (7) Google Scholar). In aerobic conditions (Fig. 3B, left panel), a rapid reduction of the prosthetic flavin (≤30 ms) was initially observed—as absorbance decrease at the wavelength of 450 nm (red line)—with the concomitant oxidation of a fraction of the NADH pool visible at 340 nm (black line). Following a short stationary phase, the formed FADH2 and the remaining NADH were fully oxidized in the range of tens of seconds. Under anaerobic conditions (Fig. 3B, right panel), the first half-reaction was identical to the aerobic condition. However, the second half-reaction did not take place and FADH2 remained in the reduced state, since no O2 was available to accept the electrons taken over by the flavin. Consequently, also NADH could not be further oxidized to NAD+. We then determined the kinetic parameters for the oxidative activity using either NADPH or NADH as electron donor in the absence of any NAD(P)+. This revealed that SthA aerobically oxidizes both NADH and NADPH (Table 1 and Fig. S6 for kinetic plots), essentially acting as an NAD(P)H oxidase with a kCAT of 0.1 to 0.2 s−1. For this uncoupling reaction, it shows a higher affinity for NADH than NADPH but a higher turnover number when NADPH was electron donor. Using NADH, SthA showed a 75-fold higher kCAT value for the transhydrogenation than the oxidase activity, although the catalytic efficiency (kCAT/KM) for the hydride transfer between cofactors was only 1.5 times higher than the oxygen reduction. The millimolar range of the KM for NADH during transhydrogenation lies behind such a low catalytic efficiency. Next, we focused on the formation of the ROS species. We used ferricytochrome c (26Butler J. Jayson G.G. Swallow A.J. The reaction between the superoxide anion radical and cytochrome c.Biochim. Biophys. Acta. 1975; 408: 215-222Crossref Scopus (129) Google Scholar) as the reporter for production of superoxide anion. Time course experiments revealed that SthA produces superoxide anion in the order of a few micromolar (Fig. 3C), in the presence of an excess of 1.0 mM NAD(P)H as substrate. After 10 min of reaction, NADH led to the accumulation of almost 7.5 μM of superoxide anion (dark blue line), three times higher than the superoxide reached from an equimolar amount of NADPH (light blue line). Intriguingly, we also detected superoxide formation during the transhydrogenation between 1.0 mM NADH and 0.15 mM thioNADP+ (ocher line), which generated similar levels of radical species to those observed in the absence of oxidized nicotinamide cofactors as electron acceptors. Then, we tested both the uncoupling and the transhydrogenase activities under the same experimental conditions as in Figure 3C but this time by measuring the production of H2O2 (Fig. 3D). H2O2 was produced in all cases, at higher levels than the superoxide anion (30–40 μM H2O2 in 10 min of reaction), showing that H2O2 is the main product of the SthA uncoupling activity, which was confirmed by comparing the ROS formation reaction rates (Fig. 3E). Indeed, we estimated the generation of H2O2 (green bars) to be more than 5 times and 30 times faster than the superoxide anion (blue bars), respectively from the oxidation of NADH and NADPH. As expected, the reaction rates for the ROS release fit well with the turnover numbers calculated for the NAD(P)H consumption (Table 1). Unlike the sequence-related FDR family, STHs characteristically contain a threonine replacing the C-terminal cysteine in the CXXXXT motif equivalent to the redox-active disulfide center in FDRs. Due to its evolutionary conservation among STHs (Figs. 4A, and S7 for the phylogenetic tree), the presence of the threonine in position 50 suggests a role for this residue to operate the enzymatic transhydrogenation. On the other hand, the cysteine in position 45 emerges as the link between STH and FDR families, as it is mechanistically crucial for the reduction of thiol substrates in the latter protein group (4Argyrou A. Blanchard J.S. Flavoprotein disulfide reductases: advances in chemistry and function.Prog. Nucl. Acid Res. Mol. Biol. 2004; 78: 89-142Crossref PubMed Scopus (176) Google Scholar). Therefore, we characterized the single-mutants C45A and T50C to establish the importance of Cys45 and Thr50 for the catalytic properties of SthA. Compared to the WT enzyme, both the mutants revealed a lower flavin content after the protein purification (Table 2 and Fig. S8 for the SEC profiles), indicating that Cys45 and Thr50 might promote a more stable embedding of the prosthetic FAD cofactor in SthA. Indeed, in silico docking of the FAD molecule to the Alphafold-predicted structure of SthA (Fig. 4B) revealed the close proximity of Cys45 and Thr50 to the ribityl moiety and the N5 and N10 atoms of the isoalloxazine ring, respectively. The lower capacity to retain the prosthetic group in the SthA mutants even matched the amount of unbound FAD eluted during the purification procedure (Figs. S8 and S1). When we verified their capacity to transfer reducing equivalents from NADH to thioNADP+, both C45A and T50C displayed a comparable one-order-of-magnitude loss of the activity found in the WT protein (Fig. 4C, yellow bars).Table 2Flavin content of the SthA-variants designed in our studyVariantRatio 280/450ε450 nm (mM-1 cm-1)WT6.0 ± 0.312.1C45A7.4 ± 0.213.0T50C8.8 ± 0.212.6The ratio 280/450 corresponds to the ratio between the wavelengths at 280 and 450 nm measured during the elution of the purified protein sample on SEC. The reported values correspond to the average of five different purification batches (n = 5, s.e.m. as errors). The extinction coefficients at 450 nm (ε450 nm) for each protein variant were determined according to established protocols for flavoproteins (21Aliverti A. Curti B. Vanoni M.A. Identifying and quantitating FAD and FMN in simple and in iron-sulfur-containing flavoproteins.Met. Mol. Biol. 1999; 131: 9-23PubMed Google Scholar). Open table in a new tab The ratio 280/450 corresponds to the ratio between the wavelengths at 280 and 450 nm measured during the elution of the purified protein sample on SEC. The reported values correspond to the average of five different purification batches (n = 5, s.e.m. as errors). The extinction coefficients at 450 nm (ε450 nm) for each protein variant were determined according to established protocols for flavoproteins (21Aliverti A. Curti B. Vanoni M.A. Identifying and quantitating FAD and FMN in simple and in iron-sulfur-containing flavoproteins.Met. Mol. Biol. 1999; 131: 9-23PubMed Google Scholar). We then studied the reductive half-reaction by determining the apparent dissociation constant (Kd) for NADH and reduction rate constant (kred) of the protein-embedded flavin by stopped-flow kinetic measurements (Fig. 4D and Table 3, representative spectra available in Fig. S9). WT SthA was found to be efficiently and fully reduced by NADH with a kred of 303 s−1 and Kd of 88 μM. Mutations in the CXXXXT motif increased the Kd by 3 to 4 fold compared to WT SthA. The higher Kd values suggest that the combination of Cys45 and Thr50 is needed for optimal binding of the nicotinamide substrate. On the other side, both mutant SthAs displayed significantly higher kred values. The C45A mutation doubled the kred, while restoration of the disulfide center through the T50C mutation led to a 7-fold increase of the reduction rate.Table 3Pre–steady-state kinetic parameters for the reductive half-reaction of SthA mutantsVariantSubstratekred (s-1)Kd (μM)WTNADH303 ± 5088 ± 27C45ANADH710 ± 160300 ± 100T50CNADH2200 ± 400227 ± 61The values of reduction rate constant of the SthA variants (kred) and apparent dissociation of NADH constant (Kd) come from biological triplicates (n = 3, s.e.m. as errors) measured by stopped-flow spectrophotometry in anaerobic conditions. Open table in a new tab The values of reduction rate constant of the SthA variants (kred) and apparent dissociation of NADH constant (Kd) come from biological triplicates (n = 3, s.e.m. as erro