Bacillus pumilus KatX2 confers enhanced hydrogen peroxide resistance to a Bacillus subtilis PkatA::katX2 mutant strain

Background Bacillus pumilus cells exhibit a significantly higher resistance to hydrogen peroxide compared to closely related Bacilli like Bacillus subtilis. Results In this study we analyzed features of the catalase KatX2 of B. pumilus as one of the most important parts of the cellular response to hydrogen peroxide. KatX2, the vegetative catalase expressed in B. pumilus, was compared to the vegetative catalase KatA of B. subtilis. Data of our study demonstrate that B. pumilus can degrade toxic concentrations of hydrogen peroxide faster than B. subtilis. By replacing B. subtilis katA gene by katX2 we could significantly enhance its resistance to H2O2 and its potential to eliminate this toxic compound. Mutant cells showed a 1.5- to 2-fold higher survival to toxic concentrations of hydrogen peroxide compared to wild type cells. Furthermore, we found reversible but also irreversible oxidations of the KatX2 protein which, in contrast to KatA, contains several cysteine residues. Conclusions Our study indicates that the catalase KatX2 plays a major role in the increased resistance of B. pumilus to oxidative stress caused by hydrogen peroxide. Resistance to hydrogen peroxide of other Bacilli can be enhanced by exchanging the native catalase in the cells with katX2. Electronic supplementary material The online version of this article (doi:10.1186/s12934-017-0684-y) contains supplementary material, which is available to authorized users.


Background
In their natural habitat (soil) as well as in industrial fermentation processes cells are frequently exposed to different stress conditions like heat, osmotic or oxidative stress [1][2][3][4][5][6][7]. Oxidative stress can be caused by a variety of reactive oxygen species (ROS) like superoxide (O 2 ·− ), hydrogen peroxide (H 2 O 2 ) and hydroxyl radical (OH·). Such conditions may interfere with the fermentation processes by hampering the cell growth, killing the cells or by impairing the quality of the fermentation product.
Bacillus pumilus strains show a significantly increased resistance to oxidative stress caused by hydrogen peroxide compared to its relatives belonging to the Bacillus subtilis species complex [8][9][10]. A mechanistic understanding of this phenomenon could help to improve the oxidative stress resistance of other production strains which are already in use.
Catalases are key enzymes for the degradation of hydrogen peroxide. In many well-known Bacillus species KatA is the primary catalase expressed in vegetative cells [11,12]. The genome of B. pumilus does not encode any gene homolog of KatA [13]. Instead, it encodes two genes annotated as catalases KatX1 and KatX2. KatX1 shows a homology of 83% to B. subtilis KatX, the spore catalase of this organism [14]. KatX2 shows only about 50% homology to both the KatX and the KatA amino acid sequence of B. subtilis. Like KatA, KatX2 is an iron dependent catalase.
In this study and also in previous studies using B. pumilus Jo2 we never observed an expression of katX1, neither on the proteome nor on the transcriptome level. In

Open Access
Microbial Cell Factories *Correspondence: birgit.voigt@filkfreiberg.de 4 Present Address: Research Institute for Leather and Plastic Sheeting, Meißner-Ring 1-5, 09599 Freiberg, Germany Full list of author information is available at the end of the article contrast, KatX2 showed high induction rates on both levels following H 2 O 2 treatment in B. pumilus Jo2 [8]. This indicates that KatX2 may take on the role as the major vegetative catalase in B. pumilus.
In this study we compared the abundance and the activity of the catalase from B. pumilus with those of related organisms like B. subtilis. We replaced B. subtilis katA gene by the B. pumilus katX2 gene and studied the resistance of this mutant to hydrogen peroxide. Furthermore, since KatX2 contains several cysteine residues, we analyzed the oxidation of this protein under control conditions and during hydrogen peroxide provoked oxidative stress.

Construction of mutant strains
A linear DNA fragment carrying a 600 kb upstream-fragment with the promoter region including the ribosome binding site of B. subtilis katA, the B. pumilus katX2 gene from the ATG start-codon to the stop-codon, the spectinomycin resistance gene from pUS19 [16] and a 600 kb downstream fragment beginning right behind the B. subtilis katA stop codon was created with a two-step fusion PCR (Additional file 1: Figure S1) [17]. Firstly, single DNA fragments of the upstream-and downstream-region of B. subtilis katA, the B. pumilus katX2 gene and the spectinomycin resistance gene were created using chromosomal DNA of B. subtilis 168, B. pumilus SAFR-032 or the pUS19 plasmid and primers extended by several nucleotides homologous to the connecting upstream/downstreamfragment (Table 1). PCR-fragments were separated from template and primer DNA by electrophoresis in a 0.8% agarose gel, cut out in the absence of UV irradiation and purified using the Qiaquick Gel Extraction Kit (Qiagen, Germany). The upstream fragment and B. pumilus katX2 gene as well as the spectinomycin resistance gene and the downstream fragment were fused through their complementary ends and fusion products were purified as described above. In a second fusion step the upstream-katX2 fragment and the spectinomycin resistance-downstream fragment were fused to create the complete linear DNA fragment. It was purified as described above. The purified linear fragment was used for transformation of naturally competent B. subtilis cells [18,19].
Transformation was carried out by incubating of 1 µg purified linear DNA with 1 ml competent cells for 1 h in a 100 ml shake flask at 37 °C and 70-100 rpm. 5 ml of LB was added and cells were incubated for 1.5 h at 37 °C and 200 rpm.
Mutants were selected on LB agar plates containing 200 µg/ml spectinomycin. For the verification of the mutation, chromosomal DNA was amplified and sequenced by Eurofins (http://www.eurofinsgenomics. eu/de/home.aspx).

Sample preparation
Cells were harvested or stressed at an OD 500 of 0.6 with various concentrations of H 2 O 2 . Samples used for 2D-PAGE analyses, fluorescence thiol modification Samples used for the catalase activity assay were harvested at an OD 500 of 0.6 or stressed using one-tenth of these concentrations for 10 min and then harvested. Cells were harvested by centrifugation (20,000×g, 4 °C, 10 min) followed by two washing steps with 100 mM Tris-HCl buffer, pH 7.5. Cell disruption was done by sonication after resuspension in TE buffer (10 mM Tris, pH 7.5, 10 mM EDTA) containing 1.4 mM PMSF. For absolute protein quantification an in-solution digestion of proteins with TE buffer without PMSF was used. Protein concentration was determined with RotiNanoquant (Roth).

Survival assay
The survival of B. subtilis 168 and the PkatA::katX2 mutant was analyzed by adding 500 µM H 2 O 2 to exponentially growing cells at an OD 500 of 0.6. Cells were diluted using 0.9% NaCl to appropriate concentrations before addition of hydrogen peroxide, 3 and 15 min after hydrogen peroxide addition and plated on LB agar plates (Invitrogen). Colony forming units were counted after overnight incubation at 37 °C.

2D-PAGE, gel imaging, relative quantification and protein identification
200 µg protein were adjusted to 306 µl with 2 M thiourea/8 M urea, mixed with 34 µl CHAPS solution (20 mM DTT, 1% w/v CHAPS, 0.5% v/v Pharmalyte, pH 4-7 or 3-10) and loaded onto commercially available IPG strips (SERVA Electrophoresis) in the pH-range of 4-7. IEF was performed according to Büttner et al. [20]. Equilibration of the strips containing the focused proteins was performed in solutions containing DTT and iodoacetamide, respectively, as described by Görg et al. [21]. Gels of 12.5% acrylamide and 2.6% bisacrylamide were used for separation in the second dimension. Gels were stained with Flamingo Fluorescent Gel Stain (Bio-Rad Laboratories) according to the manufacturer`s instructions. 2D-PAGE was done with three independent biological replicates.
Analysis of the gel images and spot quantification was performed as described by Wolf et al. using the Delta2D software version 4.4 (Decodon) [22]. Protein spots were excised from the gels (Ettan Spot Picker, GE Healthcare), digested and spotted onto MALDI targets (Ettan Spot Handling Workstation, GE Healthcare). MS-analysis of the targets was performed by MALDI-TOF-MS/ MS using the Proteome Analyzer 4800 (Applied Biosystems) and peak lists were searched with MASCOT search engine version 2.1.0.4 (Matrix Science) and search parameters as described by Wolf et al. [22].

Label-free quantification (LC-IMS E )
In-solution digestion of protein extracts with trypsin was done according to the method described previously [23]. Desalting of peptides prior to mass spectrometry analysis using stage tips was achieved using a standard protocol [24]. For absolute quantification the peptide mix was spiked with a tryptic digest of yeast alcohol dehydrogenase (Waters) at a final concentration of 50 fmol/µl.
The nanoACQUITY ™ UPLC ™ system (Waters) was used to separate the peptide mixture and to introduce the samples into the mass spectrometer. The peptide mixture was directly loaded on an analytical column (nano-ACQUITY ™ UPLC ™ column, BEH300 C18, 1.7 mm, 75 mm_200 mm, Waters). Separation of peptides for IMS E (MS E with ion mobility separation) was done with a 90 min gradient from 5% buffer B to 40% buffer B. All MS E analyses were performed as previously described [23]. The only modification was, that the collision energy was alternated between 4 eV in the precursor ion trace and a ramp 25-45 eV for fragment ion trace. Wave velocity was ramped from 1000 to 400 m/s, wave height was set to 40 V.
LC-IMS E data were processed using PLGS v3.0.1. Processing parameters were set as follows: Chromatographic peak width and MS TOF resolution were set to automatic, lock mass charge 2 set to 785.8426 Da/e with a lock mass window of 0.25 Da, low energy threshold 200.0 counts, elevated energy threshold 20.0 counts, intensity threshold 750 counts. The data were searched against a randomized B. subtilis 168 database (NCBI, version August 2014) with added amino acid sequence of B. pumilus SAFR-032 KatX2 protein, laboratory contaminants and yeast ADH1 sequence (8438 entries). For positive protein identification the following criteria had to be met: 1 fragment ion matched per peptide, 5 fragment ions matched per protein, 1 peptide matched per protein; 2 missed cleavages allowed, primary digest reagent: trypsin, fixed modification: carbamidomethylation C (+57.0215), variable modifications: deamidation N, Q (+0.9840), oxidation M (+15.9949), pyrrolidonecarboxylacid N-TERM (−27.9949). The protein false discovery rate (FDR) was set to 5%. For the final analysis only 2 peptide identifications were considered. A protein had to be identified in at least two out of 3 technical replicates per time point; this took the FDR on protein level to less than 3%. Three biological replicates for each time point were analyzed.
Data generated by the IMS E mode were corrected for detector saturation effects by implementing a correction factor based on the ion accounting output files that were created for each sample by the PLGS software. The correction factor (cf ) was calculated using the following equation.
where ΣIpeptide and ΣIproduct are the matched peptide/product intensity sums, m is the median of the ratios ΣIpeptide ΣIproduct calculated for every protein quantified in a sample.

Catalase activity assay
Cells were grown to an OD 500nm of 0.6 and cytosolic cell extracts were prepared as described above. Right before starting the assay, a working solution of 25 µg/ml lactoperoxidase and 0.5 M dicarboxidine dihydrochloride (both Sigma-Aldrich) was prepared. To determine the degradation of H 2 O 2 by cytosolic protein extracts, 5 µg of protein extract were filled up with catalase assay buffer (Catalase Assay Kit, BioVision) to 200 µl. H 2 O 2 was added to a final concentration of 2 mM and the extracts were incubated at 30 °C. To measure the H 2 O 2 concentration remaining in the culture at certain time points, 25 µl of the cell extract was mixed with 500 µl working solution and absorbance at 450 nm was measured as described [25]. The experiments were done in triplicates.

Fluorescence thiol modification assay and analysis of protein modifications
Proteins with reversibly oxidized cysteines were visualized using a protocol described by Hochgräfe et al. [26]. Protein extracts were purified and pre-stained with BODIPY (Thermo Fisher Scientific) as described and loaded onto IPG-strips in the pH-range 4-7 (SERVA Electrophoresis). 2D-PAGE was performed as described above in the dark. Following fluorescence scanning of reversibly oxidized proteins (BODIPY staining), the gels were stained with Flamingo Fluorescent Gel Stain (Bio-Rad Laboratories). Spot quantification and MS-analyses were performed as described above. Oxidation was determined by calculating and comparing the volume of Flamingo stained spots (protein amount) to the volume of BODIPY stained spots (expressed as % volume representing the portion of the spot volume of an individual spot of the entire spot volume detected on the gel). Three biological replicates were performed.
For the analysis of possible modifications, protein spots were excised from the gels as described above, destained (0.2 M NH 4 HCO 3 , 30% acetonitrile) and double digested with trypsin and chymotrypsin (both Promega). Peptide extraction was performed by covering the gel pieces with ultra-pure water (prepared with a Sartorius Stedim unit) and 15 min incubation in an ultrasonic water bath. Peptides were detected by LC-MS/MS using an Orbitrap Elite (Thermo Fisher Scientific). Database searches

Results and discussion
Growth and survival of the B. subtilis PkatA::katX2 mutant We exchanged B. subtilis 168 katA by the katX2 gene from B. pumilus. Only the coding region starting with the ATG start codon of the B. pumilus KatX2 gene was fused to the katA-promoter region of B. subtilis. There was no difference in the growth behavior of the B. subtilis PkatA::katX2 mutant compared to the wild type strain under control conditions. After the treatment of the wild type strain and the B. subtilis PkatA::katX2 mutant with 50 µM H 2 O 2 also only a small impact on the growth of both strains could be detected (Fig. 1a). In contrast, the mutant strain showed a significantly lower impact on the growth rate than the wild type when the cells were treated with 200 µM H 2 O 2 (Fig. 1b). Mutant cells continued growth up to an optical density of about 1 whereas the wild type reached a final OD of about 0.75. Increasing the hydrogen peroxide concentration up to 2 mM, a concentration which B. pumilus can withstand [8], resulted in a nearly complete cessation of growth in both B. subtilis strains (Fig. 1c).
The survival of the B. subtilis 168 strain and the PkatA::katX2 mutant was analyzed using 500 µM H 2 O 2 . 23% of the wild type cells survived 3 min after peroxide treatment (Fig. 2). At the same time point about 50% of the PkatA::katX2 mutant cells were still alive. 15 min after addition of H 2 O 2 32% of the katX2-expressing cells were alive whereas only 18% of the wild type cells survived at this time point. These data indicate a 1.5-to 2-fold higher survival of the katX2-expressing cells under this oxidative stress conditions. To gain information on the absolute concentration of the catalases in the cytoplasm, the gel-and label-free quantification approach LC-IMS E was conducted. The results of this experiment revealed a concentration of KatA of 0.0165 fmol/ng protein extract in exponentially growing B. subtilis cells (Fig. 3). KatA accumulation increased up to 0.1153 fmol per ng protein extract in hydrogen peroxide stressed cells. In B. pumilus we measured an amount of 0.0385 fmol KatX2 per ng protein extract. 20 min after addition of H 2 O 2 , catalase accumulation increased up to 0.1076 fmol per ng cell extract in B. pumilus cells. Based on the lower basal accumulation the induction rate following H 2 O 2 treatment was significantly higher in B. subtilis cells.

Quantification of the catalase protein spots
Exponentially growing B. subtilis PkatA::katX2 mutant cells contained 0.0259 fmol KatX2 catalase per ng protein extract (Fig. 3). Following hydrogen peroxide treatment the KatX2 accumulation increased up to 0.1431 fmol per ng cell extract and therefore it was higher than the KatX2 accumulation in H 2 O 2 stressed B. pumilus cells.

Catalase activity
The sequence differences between the KatX2 and the KatA-group may lead to a higher enzymatic activity and/ or to a higher stability of the KatX2 enzyme, which could explain the enhanced hydrogen peroxide tolerance of B. pumilus compared to related organisms.
We used cell extracts containing equal amounts of protein to analyze the degradation of H 2 O 2 by the catalases. For this we prepared extracts from exponentially growing cells as well as from cells previously stressed by low amounts of H 2 O 2 (5 µM for B. subtilis, 200 µM for B. pumilus) (Fig. 4) [8,11,12]. As expected, the added H 2 O 2 concentration was reduced faster in pre-stressed extracts. In both cases, the B. pumilus cell extract degraded the hydrogen peroxide faster than the extracts from B. subtilis. B. subtilis protein extracts degraded only about 50% of the added H 2 O 2 within 10 min, whereas 80-90% of it was degraded by the B. pumilus extracts. The faster degradation of hydrogen peroxide by control cell extracts of B. pumilus could be explained by the higher catalase amount present in theses extracts (Fig. 3). However, in stressed cells the amount of the catalase is comparable between B. subtilis and B. pumilus. Therefore, the faster degradation of hydrogen peroxide by extracts from stressed B. pumilus cells cannot be due to a higher amount of the catalase. Using the absolute protein quantities determined with the LC-IMS E method, we caculated specific enzyme activities for the catalases. For KatA a specific activity of 6.212 × 10 7 units (µg/min) per The hydrogen peroxide degradation of extracts from exponentially growing unstressed B. subtilis PkatA::katX2 mutant cells was comparable to those shown by the wild type. The specific activity was 8.628 × 10 7 units per mg catalase. Pre-stressed extracts of the mutant cells degraded the added H 2 O 2 much faster than the corresponding extracts from the other strains and organisms, even faster than the pre-stressed B. pumilus cell extracts (Fig. 4). This may be explained by the higher induction rate of the recombinant catalase KatX2 in B. subtilis following hydrogen peroxide treatment compared to the induction rate observed in stressed B. pumilus, which resulted in a slightly higher amount of the catalase in the cells ( Fig. 3; Table 2) [8,11].

Modification of the KatX2 protein after peroxide stress
Unlike the KatA protein, KatX2 contains three cysteines. Under oxidative stress conditions cysteine residues can be oxidized [26,27]. This could irreversibly damage the protein, e.g. when oxidized to cysteine sulfinic and sulfonic acid as shown for the GapA protein of Staphylococcus aureus [28]. To prevent this, Bacillus cells protect cysteine residues in proteins by reversibly oxidizing them using different low-molecular-weight thiol compounds such as bacillithiol [5,29]. Using the fluorescence thiol modification assay described by Hochgräfe et al. [26] we analyzed reversible thiol-modifications in the B. pumilus KatX2 protein. This procedure uses two different staining methods, one for protein accumulation and one for reversible thiol oxidations. Quantification of proteins was done using relative spot volumes (volume of a spot compared to the volumes of all spots visible on the 2D-gel). The ratio between the spot volumes of a protein spot in the thiol modification staining and the protein accumulation staining is an indicator for the amount of reversible oxidations of the cysteine residues in a protein. In exponentially growing cells, KatX2 cysteine residues were nearly completely reduced (Fig. 5a). Hydrogen  peroxide treatment caused a significant increase of reversible cysteine oxidation (Fig. 5b). The ratio of thiol modification to protein accumulation increased from about 0.5 to 1.36. When the oxidative stress is too harsh or continues too long, the capacity of the cells to protect proteins by reversible thiol oxidations might get exhausted leading to irreversible oxidation of cysteine residues to cysteine sulfinic and/or sulfonic acid. To test this hypothesis, LC-MS/MS analysis was performed to search for such irreversible oxidation of cysteine residues in the different catalase spots excised from the 2D-gels. Modifications were verified by analysing fragment ions of cysteine containing peptides. We found evidence for the oxidation of the thiol group of cysteine 461 in one of the catalase 2D-gel spots. Mass shifts of +32 and +48 were detected in KatX2 expressed by the B. subtilis PkatA::katX2 mutant representing sulfinic (+32) and sulfonic (+48) acid formation (Fig. 6a). In B. pumilus we only detected formation of sulfonic acid in the KatX2 cysteine 461 (Fig. 6b).

Concluding remarks
By exchanging the B. subtilis KatA catalase with B. pumilus KatX2, we could show that this catalase enables a higher resistance of B. subtilis to oxidative stress. The B. subtilis PkatA::katX2 mutant showed an increased growth compared to the B. subtilis wild type when treated with hydrogen peroxide. However, H 2 O 2 concentrations comparable to those used in B. pumilus studies nearly completely ceased growth of the B. subtilis PkatA::katX2 mutant. This fact points to further features of B. pumilus like a better protection of its proteins or a more effective machinery for repair of damaged compounds which are responsible for the enhanced resistance to hydrogen peroxide beside the high catalase activity of KatX2. Previous studies have demonstrated that oxidative stress indeed can lead to oxidation of secreted proteins which are often produced with microbial hosts [12]. B. pumilus katX2 catalase might be an interesting marker gene for the engineering of new production strains which confer an enhanced resistance against hydrogen peroxide. Improving resistance of production strains to upcoming oxidative stress during fermentations has already been tried by mutating a catalase to increase catalytic efficiency [30]. This could ensure an improved quality of overproduced target proteins in bacterial production hosts.