A Potent Quinone Reductase Encoded by ywqN (Qnr1) Protects Bacillus subtilis from Oxygen Radical Genotoxicity

TL;DR

A protein called YwqN in Bacillus subtilis helps protect the bacteria from harmful oxygen radicals and may be useful for cleaning up toxic dyes.

Contribution

Discovery of YwqN as a potent quinone reductase with protective roles against ROS and azo dyes in B. subtilis.

Findings

YwqN reduces menadione and naphthoquinone, showing potent quinone reductase activity.

YwqN and YhdA together protect B. subtilis from hydrogen peroxide-induced damage.

YwqN suppresses mutagenesis caused by oxidized guanine and superoxide-generating agents.

Abstract

ywqN encodes a protein with an unassigned function that shares partial 3D homology with B. subtilis YhdA, Pseudomonas putida ChrR, and Escherichia coli YieF, which are NADP(H)/FMN-dependent oxidoreductases that catalyze the reduction of diverse chemical pollutants, including Cr(VI). Here, we report that a recombinant His6-YwqN protein displays marginal chromate reductase activity but is capable of reducing synthetic azo dyes. Remarkably, His6-YwqN exhibits a potent quinone reductase activity, catalyzing the reduction of menadione (MD) and 1,4-naphthoquinone (NQ). The individual and combined roles of YwqN and YhdA in protecting B. subtilis from ROS-promoting agents were further tested. Sensitization to the oxidizing agent H2O2 required the simultaneous loss of both YwqN and YhdA. In contrast, strains deficient in ywqN, either alone or in combination with yhdA, exhibited similar but higher susceptibilities to the superoxide-generating agent MD compared with the WT strain. These results indicate that YwqN and YhdA contribute to protection against the deleterious effects of ROS in B. subtilis. Further results revealed that while YwqN, but not YhdA, prevented MD-induced mutagenesis, both proteins synergistically prevented RifR mutations induced by H2O2. Furthermore, overexpression of YwqN suppressed the hypermutagenesis phenotype of a B. subtilis strain deficient in the prevention/repair oxidized guanine (GO) system, which is prone to accumulate 8-oxoGs. In summary, YwqN counteracts the cytotoxic and genotoxic effects promoted by ROS in B. subtilis and represents a potential tool for the remediation of soils and effluents contaminated with carcinogenic azo dyes.

Figures

Full text

1. Introduction

NAD(P)H-dependent oxidoreductases catalyze the reduction or oxidation of distinct substrates coupled to the oxidation/reduction of a nicotinamide adenine dinucleotide cofactor (either NAD^+^ or NADP^+^) [1]. Some NAD(P)H-dependent oxidoreductases contain flavin mononucleotide (FMN) or, more frequently, flavin adenine dinucleotide (FAD) as a redox-active prosthetic group. NAD(P)H–flavoproteins are widespread in microorganisms, where they catalyze a diversity of reactions that play pivotal roles in central metabolic pathways [1,2]. Bacterial NAD(P)H/FMN oxidoreductases are of particular interest in bioremediation processes due to their ability to reductively cleave dyes containing an azo bond [1,3]. Large amounts of these synthetic dyes are used worldwide and released into the environment by textile industries, thus representing major sources of pollution [3,4]. Azo dyes are water-soluble and difficult to detect, and their degradation products, usually aromatic amines, are of human health concern due to their carcinogenic potential [4].

The three-dimensional structures of several bacterial azoreductases have been solved, and all display a common structural topology known as the flavodoxin-like fold [5,6]. This fold is composed of a central five-stranded parallel β-sheet flanked by five α-helices and is shared by some eukaryotic NAD(P)H oxidoreductases that possess quinone reductase activity [5,6]. In addition to azo compounds, azoreductases can reduce a broad range of substrates, including quinones, nitrofurans, and chromate [5,6,7]; therefore, NAD(P)H–quinone oxidoreductases and azoreductases can be considered part of a common enzyme family [3,6]. Notably, although proteins belonging to this family are highly divergent in their amino acid sequences, they share a conserved overall fold and similar enzymatic activities [6].

In addition to DNA repair proteins, peroxidases, and superoxide dismutases, FMN-dependent oxidoreductases constitute alternative cellular defenses against oxidative stress. Accordingly, in Pseudomonas putida, a biological role has been attributed to the FMN-Red protein ChrR, which confers resistance to H_2_O_2_ [8]. Similarly, the diaphorase NQO1 has been shown to protect mammals from the toxicity of ROS-generating carcinogenic compounds such as benzo-pyrene quinone, naphthoquinone, and benzoquinone [9]. Furthermore, the FMN-dependent oxidoreductase YhdA was shown to reduce hexavalent chromium and quinones, and to protect B. subtilis from the cytotoxic and genotoxic effects promoted by these ROS-generating agents [7].

YwqN from B. subtilis shares structural homology with the FMN/NADPH-dependent oxidoreductases YhdA, ChrR and YieF, which reduce Cr(VI) to Cr(III) while avoiding the generation of partially reduced species that promote oxidative stress [7,8,10,11,12]. The biochemical and physiological functions of YwqN, which have remained elusive to date, were here investigated. Our results reveal that a purified recombinant YwqN protein displays marginal chromate reductase activity but exhibits strong NADPH/FMN-dependent azoreductase and quinone reductase activities. Genetic analyses of B. subtilis strains deficient in YwqN and/or YhdA further demonstrate that both proteins play important roles in protecting B. subtilis from the deleterious and mutagenic effects of the oxygen radical generators H_2_O_2_ and MD. Remarkably, overexpression of YwqN counteracted the hypermutagenesis exhibited by a B. subtilis strain lacking the prevention/repair oxidized guanine (GO) system, thereby preventing the accumulation of the ROS-induced lesion 8-oxoG. Overall, these findings assign for the first time a functional role to YwqN in protecting B. subtilis from ROS-promoted DNA damage, and based on its biochemical properties, suggest its potential application in the remediation of soils and effluents polluted with carcinogenic azo dyes.

2. Results

2.1. YwqN Possesses Structural Homology with NADP(H)/FMN-Dependent Oxidoreductases

The biochemical and physiological roles of B. subtilis YwqN (BsuYwqN) have remained elusive to date [13]. Previous crystallographic studies reported that BsuYwqN adopts a similar fold to that reported for the tryptophan repressor protein [14]. As shown in this report, multiple-sequence analysis revealed that BsuYwqN shares 42% amino acid similarity with B. subtilis YhdA (BsuYhdA) and 34% similarity with E. coli YieF (EcoYieF), respectively (Figure 1). Previous biochemical evidence demonstrated that these oxidoreductases catalyze the FMN/NADPH-dependent reduction of hexavalent chromium and quinones, respectively [7,10]. Consistent with this, the primary structure of BsuYwqN contains the NAD(P)H-dh2-binding motif, LFVTPEYNXXXXXXLKNAIDXXS, characteristic of oxidoreductases from divergent origins [1,5,14] (Figure 1). Within this region, BsuYwqN contains the amino acid residues Pro_77_, Tyr_79_, Gly_85_, Leu_87_, Lys_88_, and Asp_92_, all of which are conserved in EcoYieF and BsuYhdA (Figure 1), supporting the dependence of these proteins on NAD(P)H as an electron donor.

A second motif required for FMN-binding was identified between the β1 sheet and α-helix 1 of BsuYwqN (Figure 1). This motif includes the invariant amino acids Gly_8_ and Arg_11_, the non-polar residues Gly_13_ and Gly_14_, as well as the polar amino acid Thr_16_ (Figure 1). These residues are likely involved in hydrogen bonding with the hydroxyl groups of the ribitol moiety of FMN, as described for YhdA [7] (Figure 2A). Notably, among the positively charged residues Arg and Lys involved in phosphate binding in YhdA and YieF, only Arg_11_ is conserved in YwqN; the corresponding lysine residue is replaced by Ser_12_, whose uncharged hydroxymethyl side chain may alter interactions with the FMN phosphate group (Figure 1) [15].

Further similarities and differences between BsuYwqN, BsuYhdA, and EcoYieF were established by 3D structural alignments (Figure 2A–C). All three proteins exhibited the typical folding of bacterial flavodoxin-like enzymes, consisting of a sandwich-like configuration with five α-helices and five parallel β-sheets. However, while EcoYieF and BsuYwqN shared a high level of structural similarity (Figure 2A), BsuYwqN exhibited notable differences, including a shorter N-terminal α-helix and a disordered C-terminal region, which correspond to longer α-helices in BsuYhdA and EcoYieF, respectively (Figure 2B,C).

Despite these structural differences, molecular docking analyses indicated that BsuYwqN, as well as EcoYieF and BsuYhdA, bind FMN within a conserved region that includes the GXXRXGXXT motif (Figure 1 and Figure 2A) [1,5,14]. Notably, the best-scored docking pose for FMN in BsuYwqN (i.e., −156.5) was comparable to the energies obtained from redocking of EcoYieF and BsuYhdA; namely, −166.5 and −167.3, respectively.

A detailed analysis of FMN–protein interactions, supported by the docking results, revealed that in EcoYieF and BsuYhdA, an arginine residue (R_15_ and R_11_, respectively) forms a stabilizing salt bridge with the FMN phosphate group. Additionally, a conserved glutamate residue (E_82_ and E_73_, respectively) establishes hydrogen bonds with the FMN isoalloxazine ring (Figure 2D,E). BsuYhdA and BsuYwqN also conserve a valine residue (V_104_ and V_120_, respectively), Thr_16_, and a glycine residue (G_106_ and G_155_, respectively), all of which participate in hydrogen bonding interactions (Figure 2E,F). In BsuYwqN, the glutamate residue is replaced by an aspartate, which retains the negative charge and can similarly interact with the FMN ring (Figure 2F). In addition, all three proteins contain a tyrosine residue (Y_83_ in EcoYieF; Y_74_ in BsuYhdA; and Y_79_ in BsuYwqN) that engages in multiple hydrophobic interactions with FMN (Figure 2D–F). Overall, FMN binding in these enzymes involves a combination of hydrogen bonding and hydrophobic interactions; however, the absence of a predicted arginine-mediated salt bridge in BsuYwqN may indicate a reduced FMN-binding affinity relative to the other proteins (Figure 2F).

Taken together, these findings indicate that YwqN shares key structural features with known chromate and quinone reductases and that its structural properties may allow it to catalyze FMN/NADPH-dependent reduction of environmental pollutants such as azo dyes, hexavalent chromium and quinones.

2.2. Elucidation of the Biochemical Function of BsuYwqN

As indicated above, the biochemical function of BsuYwqN has remained elusive; therefore, we sought to express its encoding gene to generate a recombinant protein in E. coli. To this end, the ORF of ywqN lacking the first and last codons was PCR-amplified and cloned into the expression vector pQE30 to generate a His_6_-tagged BsuYwqN protein, as described in the Materials and Methods. Based on its amino acid composition, a molecular mass of 20.47 kDa was predicted for BsuYwqN [15]. In agreement with this prediction, purification of His_6_-YwqN by metal affinity chromatography yielded a ~20 kDa protein with a high level of purity, as revealed by SDS-PAGE (Table S1 and Figure 3 inset).

We next investigated the ability of purified His_6_-BsuYwqN to catalyze the reduction of distinct substrates, beginning with the azo compound methyl red (MR). These assays were initially performed in the absence of FMN, but in the presence of an excess of NADPH. Under these conditions, the protein was unable to reduce the azo dye MR. As noted above, analysis of the structural properties of YwqN revealed the absence of key amino acid residues involved in FMN binding (Figure 1 and Figure 2F); therefore, we inquired whether BsuHis_6_-YwqN contained an associated flavin cofactor. Ultraviolet–visible absorbance spectroscopy revealed the absence of a characteristic peak centered at approximately 450 nm, which would be indicative of a bound oxidized flavin in BsuHis_6_-YwqN (Figure 3) [16]. This result suggested that FMN was either dissociated from the recombinant purified protein or was not efficiently incorporated during its production in E. coli. We, therefore, supplemented FMN to the buffers employed during the purification of BsuHis_6_-YwqN, as described in the Materials and Methods. This procedure yielded a purified protein displaying a UV–visible spectrum with two distinct peaks at 375 and 455 nm, consistent with the presence of a flavin cofactor [16] (Figure 3). Furthermore, this FMN-reconstituted form of BsuHis_6_-YwqN was able to oxidize the cofactors NADH and NADPH, with similar kinetics, as determined by monitoring absorbance at 340 nm (Figure S1). As expected, no oxidation of these cofactors was detected in the control reaction lacking the FMN-reconstituted enzyme (Figure S1).

Having determined the cofactor requirements of BsuHis_6_-YwqN, we next tested its capacity to reduce three distinct azo dyes, namely, MR, methyl orange (MO) and disperse orange (DO). The results revealed that BsuYwqN catalyzed the decolorization of all three compounds, exhibiting, over time, a higher efficiency toward MO than toward the other two azo dyes (Figure S1 and Table 1). We further explored the enzymatic properties of BsuYwqN by experimentally determining its catalytic efficiency during the reduction of the three azo dyes, as described in the Materials and Methods. The enzyme exhibited similar catalytic efficiency values of 577, 667, and 630 M^−1^ s^−1^ during reduction of MR, MO, and DO, respectively (Table 1). Altogether, these results support the notion that BsuYwqN possesses structural and functional properties of azoreductases, thereby revealing its potential for bioremediation applications.

We next determined the catalytic efficiencies of BsuYwqN during reduction of the quinones MD and NQ. The enzyme was able to reduce both compounds with essentially similar catalytic efficiencies, namely, 12,500 M^−1^ s^−1^ for MD and 12,600 M^−1^ s^−1^ for NQ, respectively (Table 1). Notably, BsuYwqN reduced both quinones with catalytic efficiencies ~20-fold higher than those determined for the azo compounds (Table 1). Taken together, these results indicate that YwqN possesses structural and functional properties for the efficient reduction of both quinone compounds and azo dyes but displays better properties for catalyzing quinone reduction.

We finally investigated whether BsuYwqN exhibited chromate reductase activity. To this end, the FMN-reconstituted enzyme was incubated with increasing concentrations of hexavalent chromium in a range of 20–120 ppm for 30 min. Under these conditions, the enzyme displayed specific activities corresponding to approximately 10% and 0.1% of those calculated for the reduction of azo dyes and quinone compounds, respectively (Table S2). These results support the conclusion that BsuYwqN is considerably more efficient at reducing azo dyes and quinones than hexavalent chromium.

2.3. Determination of the Physiological Role of BsuYwqN

As demonstrated above, BsuYwqN possesses quinone reductase activity; therefore, we explored a possible role for this protein in conferring protection to B. subtilis against the cytotoxic and mutagenic effects promoted by oxygen radicals. To this end, we first challenged a B. subtilis strain deficient in ywqN with increasing doses of the ROS generators H_2_O_2_ and MD. Compared with the WT strain, the single loss of YwqN did not sensitize B. subtilis cells to the noxious effects of H_2_O_2_ (Figure 4A). To investigate whether this result could be explained by a suppressive effect mediated by the oxidoreductase YhdA [5,7], we disrupted the gene encoding this protein in both the WT strain and in a YwqN-deficient background. Notably, in comparison with the WT strain, loss of YhdA alone did not significantly affect the H_2_O_2_ susceptibility of B. subtilis; however, simultaneous loss of both YwqN and YhdA generated B. subtilis cells that were approximately 2.7-fold more sensitive to the noxious effects of this oxidizing agent than the WT strain (Figure 4A). Meanwhile, sensitivity to MD increased approximately 1.4- and 1.6-fold in strains deficient in ywqN or yhdA, respectively, relative to the WT strain, and this effect did not significantly increase in the double mutant lacking both ywqN and yhdA (Figure 4B). Altogether, these results suggest that while YwqN and YhdA complement each other’s functions to counteract the noxious effects of hydrogen peroxide, they seem to work in a common pathway to prevent MD-promoted cytotoxicity (Figure 4A,B).

Next, spontaneous Rif^R^ mutation frequencies, as well as those promoted by H_2_O_2_ or MD, were determined in a B. subtilis strain deficient in ywqN. The results revealed that, relative to the wild-type parental strain, loss of YwqN caused a mild but significant increase in both the spontaneous and H_2_O_2_-induced Rif^R^ mutation frequencies (Figure 5). Of note, while a pDG148 construct, carrying an IPTG-inducible Phs-ywqN cassette, restored the mutation frequencies of the ywqN mutant to those exhibited by the WT strain, such an effect was not observed with the empty vector control (Figure S3).

A previous report revealed that the oxidoreductase YhdA from B. subtilis (BsuYhdA) counteracts the mutagenic effects derived from the accumulation of ROS-induced genetic lesions [7]. Consistent with this, we found that loss of YhdA had a higher impact on both the spontaneous and H_2_O_2_-promoted mutagenesis than loss of ywqN alone (Figure 5). Notably, simultaneous loss of YwqN and YhdA caused a significant increase in spontaneous and H_2_O_2_-promoted Rif^R^ mutation frequencies relative to the WT strain and to strains carrying single deficiencies in yhdA or ywqN (Figure 5).

We further explored the mutagenic phenotypes associated with both oxidoreductases by challenging strains carrying single or double deficiencies in ywqN and/or yhdA, after being challenged with the superoxide radical generator, MD [17]. The results showed that relative to the WT strain, the loss of YwqN, but not of YhdA, increased the menadione-induced mutation frequency by approximately three-fold, and such an effect was not further enhanced in the strain deficient in both ywqN and yhdA (Figure 6). Of note, IPTG-induced expression of ywqN from the pDG148-Phs-ywqN plasmid, but not from the empty vector, restored MD-induced mutation frequencies to wild-type levels, thereby confirming the role of YwqN in protecting B. subtilis from MD-induced mutagenesis (Figure S3).

As shown in this study, YwqN confers protection to B. subtilis from the noxious effects of MD and H_2_O_2_. The ROS generated by these compounds can target DNA, producing strand breaks and oxidized bases, including the highly mutagenic lesion 8-oxoG [7]. Antioxidant enzymes such as KatA and SodA, as well as the prevention/repair GO system, counteract these effects [7]. Thus, to better understand the mechanism by which YwqN confers protection to B. subtilis against ROS-induced mutagenesis, an extrachromosomal plasmid designed to overexpress ywqN from an IPTG-inducible promoter was introduced into B. subtilis strains deficient in either the prevention/repair GO system or the major vegetative catalases KatA and KatB (Figure S4). Our results revealed a significant decrease in both spontaneous and H_2_O_2_-induced mutagenesis in the ΔGO strain carrying the plasmid that overexpresses ywqN (pDG148-ywqN) but not in the GO-deficient strain harboring the empty pDG148 vector (Figure 7A). Notably, overexpression of ywqN also reduced spontaneous H_2_O_2_- and MD-induced mutagenesis significantly in the strain deficient in KatA and KatB; however, this protective effect was not observed in the strains carrying the empty vector pDG148 (Figure 7B,C). Taken together, these results strongly suggest that the quinone reductase activity of YwqN enables B. subtilis to counteract the cytotoxic and mutagenic effects promoted by hydroxyl and superoxide radicals.

3. Discussion

ywqN from B. subtilis encodes a protein whose function has remained undetermined [13]. In this study, we first employed an in silico approach to demonstrate that YwqN shares structural features with bacterial NAD(P)H/FMN-dependent oxidoreductases that catalyze the reduction of hexavalent chromium, industrial dyes, and quinone compounds. We then determined the substrate preference, kinetic parameters, and biochemical properties of a recombinant His_6_-YwqN protein, which allowed us to assign YwqN a primary function as a quinone reductase. Furthermore, our results demonstrated that YwqN plays an important role in counteracting the cytotoxic and genotoxic effects promoted by superoxide and hydroxyl radicals in B. subtilis.

A previous crystallographic study suggested that YwqN adopts a similar fold to that of the tryptophan repressor [14]. Here, we determined that YwqN exhibits secondary and tertiary structural similarities with oxidoreductases that utilize NAD(P)H and FMN to catalyze the reduction of a broad range of substrates.

Primary sequence analysis of YwqN revealed the presence of independent domains predicted to bind the cofactors NAD(P)H and FMN. Within the NAD(P)H-binding region, YwqN contains only six of the thirteen amino acids reported to be absolutely conserved among diverse bacterial oxidoreductases [1,5,14]. A recent study that classified NAD(P)H-binding motifs of oxidoreductases of diverse origins based on their amino acid sequences defined six distinct 1d motifs to discriminate between NAD- and NADP-binding proteins [18]. Here, we found that NADH and NADPH can be efficiently reduced by BsuYwqN (Figure S1). Therefore, the YwqN residues Pro_72_, Tyr_74_, Gly_80_, Leu_82_, Lys_83_, and Asp_87_, which are located in the NAD(P)H motif and are conserved in EcoYieF and BsuYhdA (Figure 1), can be proposed as components of an additional, functional NAD(P)H-binding motif in bacterial oxidoreductases.

Analysis of the primary and tertiary structures of BsuYwqN revealed a divergence between its predicted FMN-binding motif and those of the oxidoreductases EcoYieF and BsuYhdA (Figure 1 and Figure 2). We hypothesized that such structural differences could negatively impact the affinity of BsuYwqN for the FMN cofactor. In support of this hypothesis, recombinant His_6_-YwqN purified by metal affinity chromatography lacked the characteristic absorbance peaks of flavin-containing proteins. Furthermore, under conditions of NADPH saturation, the FMN-deficient enzyme was unable to catalyze the reduction of the azo dye MR. It must be pointed out that flavin-dependent oxidoreductases exhibiting a low FMN affinity, as found here for YwqN, have been previously reported in E. coli, Pseudomonas aeruginosa, and Deinococcus radiodurans [19,20,21]. Previous studies reported the reconstitution of apo forms of bacterial flavodoxins with FMN concentrations ranging from 1 to 100 mM [19,20,22]. Here, we found that supplementation with 20 mM FMN was enough to reconstitute BsuYwqN activity, as demonstrated by its ability to efficiently catalyze the reduction of NADH, NADPH, three azo dyes, and the quinones MD and NQ (Figure S1). Furthermore, using MR as a substrate, His_6_-YwqN exhibited maximal dye reduction at pH 7.5 and at temperatures between 30 and 40 °C (Figure S5).

Our results aimed at determining the range of substrates that can be processed by BsuYwqN revealed its capacity to catalyze the reduction of distinct azo dyes and quinone compounds. However, in contrast with ChrR from P. putida, EcoYieF and BsuYhdA, BsuYwqN was found to be inefficient at reducing hexavalent chromium [7,10,11,12]. We speculate that the low affinity of BsuYwqN for FMN and/or the structural differences it displays relative to EcoYieF and BsuYhdA (Figure 1 and Figure 2) negatively impact the recognition and reduction of hexavalent chromium [23,24].

Although BsuYwqN exhibited stronger activity toward quinones than toward azo dyes, determination of initial reaction rates revealed that reduction of these compounds did not follow classical Michaelis–Menten kinetics. Therefore, we determined reduction rates at low substrate concentrations, where the reaction velocity was proportional to the substrate concentration (Figure S2) and employed a simplified form of the Michaelis–Menten equation [25,26] to calculate the catalytic efficiencies of YwqN over the aforementioned substrates. Of note, BsuYwqN reduced the dyes MR, MO and DO with similar efficiencies; however, the enzyme was ~ 20-fold more efficient at reducing MD and NQ than the azo dyes (Table 1). Overall, in terms of cofactor requirements, catalytic behavior and ability to reduce quinones and azo dyes, BsuYwqN parallels bacterial flavodoxins such as WbrA from E. coli and Archaeoglobus fulgidus, AzoA from Enterococcus faecalis, BsuYhdA, AcpD from E. coli, AzoR from Rhodobacter sphaeroides, and Azo1 from Staphylococcus aureus [27,28,29,30,31,32].

In natural environments, B. subtilis can be exposed to soil pollutants such as hexavalent chromium and quinones; these compounds can promote the intracellular generation of ROS by acting as substrates of oxidoreductases [33,34]. Reduced semiquinone anions stimulate the production of superoxide and indirectly hydroxyl radicals [35], which can impact DNA, leading to mutagenesis and cell death [36].

Here, in agreement with our biochemical evidence revealing that ywqN encodes a potent FMN/NAD(P)H quinone reductase, we found that its encoded product confers protection to B. subtilis from the mutagenic effects of menadione and H_2_O_2_. Our genetic analyses of B. subtilis strains deficient in YwqN and/or YhdA indicated that YwqN plays a more important role in counteracting the mutagenic effects of menadione, whereas YhdA contributes better to preventing hydrogen peroxide-promoted mutagenesis. Notably, the combined loss of YwqN and YhdA generated a more pronounced mutagenic phenotype, indicating a synergistic role in protecting B. subtilis against hydrogen peroxide-induced DNA damage. Therefore, BsuYhdA not only shares the ability to reduce hexavalent chromium with PpuChrR but also confers protection against the noxious effects of H_2_O_2_ [8]. Meanwhile, YwqN seems to be more involved in protecting B. subtilis from electrophilic stress by catalyzing the two-electron reduction of ubiquinone to ubiquinol, thereby preventing the formation of mutagenic superoxide radicals. Similar protective roles have been reported for the mammalian flavoprotein reductases NQO1 and NQO2, as well as for the bacterial oxidoreductases AZR from Rhodobacter sphaeroides, MdaB from Helicobacter pylori and BsuYhdA [7,9,37,38,39,40]. Consistent with this model, overexpression of ywqN suppressed the mutagenic phenotypes of B. subtilis strains deficient in the GO prevention/repair system or in the major catalases KatA and KatB. These results further support a role for YwqN in counteracting the mutagenic effects of ROS and 8-OxoG and reinforce the notion that YwqN catalyzes a two-electron transfer reduction reaction that avoids the generation of electrophilic stress.

While our results strongly suggest that YwqN plays an antimutagenic role in actively growing B. subtilis cells, future studies will explore its contribution to stationary phase-associated mutagenesis, a process in which oxidative stress has been shown to promote genetic diversity in nutritionally stressed bacteria [36].

Finally, in addition to its physiological role in protecting B. subtilis from ROS-induced DNA damage, the broad substrate range of YwqN highlights its potential applicability in bioremediation of soils and effluents contaminated with carcinogenic azo dyes.

4. Materials and Methods

4.1. Bacterial Strains, Culture Conditions, and Reagents

All B. subtilis and E. coli strains and the plasmids used in this work are listed in Table 2. The growth medium used routinely was lysogeny broth (LB; Lennox formulation, Sigma-Aldrich; St. Louis, MO, USA). When required, ampicillin (Amp; 100 μg mL^−1^), kanamycin (Kan; 10 μg mL^−1^ for B. subtilis and 50 μg mL^−1^ for E. coli), tetracycline (Tet; 10 μg·mL^−1^), spectinomycin (Sp; 100 μg·mL^−1^), chloramphenicol (Cm; 5 μg·mL^−1^), kanamycin (Kan; 25 μg·mL^−1^), erythromycin (Ery; 5 μg·mL^−1^), or rifampin (Rif; 10 μg·mL^−1^) was added to the medium. Liquid cultures were incubated with vigorous aeration (shaking at 250 rpm) at 37 °C. Cultures on solid media were grown at 37 °C. The optical density (OD) of liquid cultures was monitored with a Pharmacia Ultrospec 2000 spectrophotometer (Peapack, NJ, USA) set at 600 nm.

The azo dyes and quinones used in this study were methyl red (MR), methyl orange (MO), disperse orange (DO), menadione (MD), and 1,4-naphtoquinone (NQ). All reagents were purchased from Sigma-Aldrich. The wavelengths of maximum absorption of the azo compounds used were 430 nm for MR, 465 nm for MO, 445 for DO, and 340 nm for NQ, MD, NADH and NADPH.

4.2. Genetic and Molecular Biology Techniques

Preparations of competent E. coli or B. subtilis cells and their transformation with DNA were performed as previously described [42,43]. Chromosomal DNA from B. subtilis was purified according to a previously described protocol [44]. Small-scale preparation of plasmid DNA from E. coli cells, enzymatic manipulations, and agarose gel electrophoresis were performed by standard techniques [43]. Medium-scale preparation and purification of plasmid DNA were accomplished by using commercial ion-exchange columns according to the instructions of the supplier (Sigma-Aldrich).

4.3. Generation of a Construct to Overexpress ywqN in B. subtilis

A construct to overexpress ywqN in WT and distinct repair/prevention- and antioxidant-deficient B. subtilis strains was constructed as follows. A 721 bp fragment encompassing the open reading frame (ORF) of ywqN and extending 24 bp upstream of ORF and 154 bp downstream of the stop codon was amplified by PCR using chromosomal DNA from strain B. subtilis 168 and the synthetic oligonucleotide primers 5′-GCGTCGACCGTAAACAAAGGAGCAGATGC-3′ (forward, containing a SalI site, underlined) and 5′-GCGCATGCTGCTTGACTTGCAGCGTGGTT-3′ (reverse, containing a SphI site, underlined). DNA amplification was carried out using high-fidelity Vent DNA polymerase according to the manufacturer’s recommendations (New England BioLabs, Ipswich, MA, USA). The PCR product was digested with SalI and SphI and ligated into the SalI/SphI-digested expression vector pDG148. The resulting construct (pPERM1877) was first amplified in E. coli XL-10 Gold and then introduced by transformation into competent cells of B. subtilis ΔkatB::Cm^r^ ΔkatA::Ery^r^ (PERM1275) and B. subtilis ΔytkD:Cm^r^ ΔmutM:Tet^r^ ΔmutY:Sp^r^ (ΔGO; PERM1699) to generate the ywqN-overexpressing strains B. subtilis PERM1867 and PERM1876, respectively (Table 2). Additionally, a B. subtilis ΔkatB::Cm^r^ ΔkatA::Ery^r^ (PERM 1275) strain carrying the empty vector pDG148 was generated and designated PERM1868 (Table 2). A B. subtilis ΔGO strain, carrying the empty vector pDG148 (PERM1709), was also employed in this study (Table 2). Experiments employing these strains were amended with 1 mM IPTG.

4.4. Generation of a B. subtilis Strain Deficient in ywqN and yhdA

To generate a B. subtilis strain deficient in ywqN and yhdA, chromosomal DNA isolated from B. subtilis PERM1817 (ΔywqN::Ery) was used to transform competent cells of B. subtilis PERM1818 (yhdA::Kan) to generate B. subtilis strain PERM1905 (ΔywqN::Ery^r^ ΔyhdA::Kan^r^) (Table 2).

4.5. Design of an E. coli Strain to Overproduce a YwqN Protein Containing an N-Terminal His6-Tag

The ORF of ywqN lacking the first and stop codon was amplified by PCR, using chromosomal DNA from B. subtilis 168 and the oligonucleotide primer set 5′- GCGGATCCAAAATTGCGGTTATTAACG-3′ (containing a BamHI site, underlined) and 5′- GCGTCGACTATCGCATCGCTTCTTTTC-3′ (containing a SalI site, underlined). The PCR fragment (540 bp) was first cloned into pJET1.2/blunt vector (Thermo Fisher Scientific, Waltham, MA, USA) and replicated in E. coli DH5α. This plasmid was digested with BamHI/SalI, and the 519-bp ywqN fragment was cloned into the BamHI/SalI site of pQE30 (Qiagen; Germantown, MD, USA). The resulting construct was introduced by transformation into competent cells of E. coli XL-10 Gold to generate the strain E. coli PERM1892 (Table 2).

4.6. Purification of His6-YwqN

E. coli strain PERM1892 was grown at 37 °C in 100 mL of LB medium to an optical density at 600 nm (OD_600_) of 0.5. At this point, the culture was supplemented with IPTG to a final concentration of 1.0 mM, and the expression of ywqN was induced for 4 h. Cells were collected by centrifugation and washed twice with 10 mL of 100 mM Tris-HCl at pH 7.5 (buffer A). Cells were resuspended and incubated in 10 mL of buffer A containing 250 mM NaCl, 1 mM phenyl-methyl-sulfonyl-fluoride (PMSF), 1 mM DTT, 20 μM FMN and lysozyme (2 mg mL^−1^) for 30 min at 4 °C and then disrupted by sonication. The cell lysate was subjected to centrifugation (27,200× g) to eliminate undisrupted cells and cell debris, and the supernatant was applied to 5 mL of a nickel–nitrilotriacetic acid (Ni-NTA)–agarose (Qiagen; Valencia, CA, USA) column equilibrated with 100 mM Tris-HCl (pH 7.5)–250 mM NaCl (buffer B). The column was washed with 150 mL of buffer A containing 10% glycerol and 20 μM FMN (buffer C) containing 25 mM imidazole, and then the protein bound to the resin was eluted with 6 mL of buffer C containing 300 mM imidazole. Aliquots (15 μL) were analyzed by SDS-PAGE, as previously described [45].

4.7. Homology and Structural Analyses

The amino acid sequences of B. subtilis YwqN and YhdA were obtained from the SubtiWiki database [13]. The E. coli YieF amino acid sequence was downloaded from the National Center for Biotechnology Information (NCBI). Multiple sequence alignments were performed using the T-coffee WEB server [46]. Secondary structure information was obtained from the Protein Data Bank (PDB) for BsuYwqN (PDB 1rli), BsuYhdA (PDB 1nni) and EcoYieF (PDB 3svl). Secondary structure alignment and visualization were carried out using the PyMOL Molecular Graphics System (Schrödinger LLC, v.3.1.6.1).

While reported 3D structures of BsuYhdA and EcoYieF were co-crystallized with FMN, the available structure of BsuYwqN lacks a bound FMN molecule. Therefore, we employed molecular docking analyses to predict the FMN binding mode in BsuYwqN. To this end, firstly, FMN was redocked into BsuYhdA and BsuYieF using Molegro Virtual Docker [47] to identify the optimal search algorithm and scoring function. The best-performing parameters were MolDock Optimizer (algorithm) and MolDock Score [GRID] (scoring function). These parameters were then used to predict the FMN binding mode in BsuYwqN. All dockings and redocking procedures were run with 10,000 maximum iterations and an initial population of 50 poses, using a docking region of 10 Å and a grid size of 0.2 Å. For FMN redocking, poses were evaluated based on their RMSD relative to the binding mode observed in crystallized complexes BsuYhdA (PDB 1nni) and EcoYieF (PDB 3svl). For FMN docking in BsuYwqN, the FMN structure from PDB 1nni was used as a ligand, and RMSD values were calculated relative to this reference. The top-scoring pose was selected as the representative model in Figure 2.

Protein–ligand interaction analyses with FMN were performed using the Protein–Ligand Interaction Profiler (PLIP) server from the University of Dresden [48]. All structural images of docking and protein–ligand interactions were generated using the PyMOL Molecular Graphics System (Schrödinger LLC, v.3.1.6.1).

4.8. Enzyme Assays

YwqN activity was measured by spectrophotometry in a JENWAY Genova Plus equipment (Cole Parmer; Vernon Hill, IL, USA) set at room temperature. Oxidation of nicotinamide cofactors was assayed in a reaction mixture containing 100 mM of NADH or NADPH and 0.5 μM of YwqN in 50 mM Tris-HCl at pH 7.5 and monitored for 10 min at 340 nm. To determine oxidation of MD and NQ, both compounds to a final concentration of 50 mM were mixed with 0.5 μM of His_6_ –YwqN and 300 mM of NADPH at pH 7.5 at room temperature and recorded at 340 nm for 10 min. Azoreductase activity was determined by dissolving the corresponding azo dye at a final concentration of 60 μM in 50 mM Tris-HCl at pH 7.5. An amount of 0.5 μM YwqN was added to the reaction mixture, and the reaction started by the addition of 1 mM NADPH was followed for 40 min. Initial rates for azo dyes were determined by monitoring the absorbance decrease at a suitable wavelength (MR: 430 nm, MO: 465 nm, and DO: 445 nm). To determine quinone reductase activity, assays were performed in a reaction mixture consisting of 300 μM NADPH and 0.5 μM of enzyme in 50 mM Tris-HCl at pH 7.5. The reaction was started by the addition of different concentrations of 2-methyl-1,4-naphtoquinone (menadione; MD) or 1,4-naphthoquinone (NQ). The quinone reductase activity was followed by NADPH consumption at 340 nm [22,47]. The degradation rate of the dyes was obtained as the concentration (μM) of each degraded dye/min. The kinetic analysis of YwqN activity was performed by plotting the degradation rate as a function of the concentration of each dye. Determination of initial reaction rates revealed that reduction of the compounds analyzed did not fit a Michaelis–Menten behavior, i.e., a hyperbolic curve showing an increase in the enzyme reaction rate as a function of the substrate concentration, which reaches a maximum velocity upon saturation of the enzyme with the substrate [25,26]. Therefore, reduction kinetics curves at low concentrations of the substrates, where the velocity was proportional to the substrate concentration (Figure S2), were employed together with the simplified Michaelis–Menten equation (k_cat_/K_M_ = vo/([E_t_].[S]) to calculate the catalytic efficiencies of YwqN’s azoreductase and quinone reductase [25,26].

4.9. Determination of Optimal pH and Temperature for His6-YwqN-Dependent Methyl Red Reduction

To determine the optimum pH, 0.5 μM of His_6_-YwqN was incubated at 30 °C with 1 mM NADPH and 100 μM methyl red in either 50 mM MES (pH 6.5) or 50 mM Tris/HCl (pHs 7.0, 7.5, 8.0, 8.5 and 9.5). The optimum temperature was determined in reactions set at 10, 20, 30, 40, 50 or 60 °C containing 50 mM Tris/HCl (pH 7.5), 0.5 μM of His_6_-YwqN, 1 mM NADPH and 100 μM methyl red. The reactions were incubated for 30 min, and the fraction of methyl red remaining was determined as described above. These experiments were repeated two times, and values were plotted as averages of duplicate determinations.

4.10. Determination of B. subtilis Cells’ Sensitivity to Oxidizing Agents

The susceptibility to hydrogen peroxide (H_2_O_2_) and menadione was determined from dose–response curves of B. subtilis strains exposed or not to increasing amounts of the oxidizing agents. To this end, the wild-type and mutant strains B. subtilis ΔywqN (PERM1817), ΔyhdA (PERM1818), and ΔywqN ΔyhdA (PERM1905) (Table 2) were cultured at 37 °C in LB medium to an OD_600_ of 0.7. At this point, cell samples of the strains of interest were treated with different concentrations of either H_2_O_2_ (0–80 mM) or menadione (0–120 μM) and incubated for 1 h at 37 °C with shaking. Bacterial viability was estimated by counting the colony-forming units (CFUs) on LB–agar plates. To this end, samples of the bacterial suspensions were collected and serially diluted in 1X phosphate-buffered saline (PBS) (137 mM NaCl; 2.7 mM KCl; 10 mM Na_2_HPO_4_ and 1.8 mM KH_2_PO_4_), and aliquots were plated on LB–agar plates. Colonies were counted after overnight incubation at 37 °C. Data were reported as 90% lethal dose (LD_90_) values, namely, the concentration of H_2_O_2_ or menadione that killed 90% of the bacterial population.

4.11. Determination of Spontaneous and H2O2- or MD-Induced Mutation Frequencies

The frequencies of B. subtilis Rif^R^ mutants in the presence or absence of hydrogen peroxide or menadione were determined as follows. B. subtilis strains WT, ΔywqN, ΔyhdA, ΔywqN/ΔyhdA, and ΔkatA ΔkatB or ΔGO bearing the pDG148-empty vector (PERM1868 or PERM1709; Table 2) or the pDG148-ywqN construct (PERM1867 or PERM1876; Table 2) were propagated in A3 (Antibiotic 3 medium; Difco, Detroit, MI, USA) to an OD_600_ of 1.0; then, each culture was split into two subcultures. One of the subcultures of each strain was left untreated, and the other was amended with an LD_25_ of H_2_O_2_ or menadione, respectively. The untreated and treated cultures were shaken at 37 °C for 16 h. Mutation frequencies were determined from three independent cultures and plating aliquots of each culture, amended with menadione or H_2_O_2_ or unamended, onto six LB plates containing 10 mg mL^−1^ rifampin (Rif) and 1 mM IPTG. The same procedure was repeated using LB plates lacking rifampin. The total number of colonies and Rif^R^ colonies was counted after 24 h of incubation at 37 °C to determine mutation frequencies.

4.12. Statistical Analyses

Statistical analyses to compare the results of mutation frequencies were performed using R software (Version R-4.5.2). A Shapiro–Wilk normality test was performed for all data, as well as a homoscedasticity of variances test. Variables that met the assumptions of normality and homoscedasticity of the data were analyzed using a one-factor analysis of variance (ANOVA), followed by a Tukey’s Honestly Significant Difference (HSD) test for multiple comparisons. For data that did not meet normality assumptions, a Mann–Whitney-Wilcoxon U-test was applied to assess differences between groups. The statistical significance used in all tests was p = 0.05.