Protection of Bacillus pumilus Spores by Catalases

Abstract

Bacillus pumilus SAFR-032, isolated at spacecraft assembly facilities of the National Aeronautics and Space Administration Jet Propulsion Laboratory, is difficult to kill by the sterilization method of choice, which uses liquid or vapor hydrogen peroxide. We identified two manganese catalases, YjqC and BPUM_1305, in spore protein extracts of several B. pumilus strains by using PAGE and mass spectrometric analyses. While the BPUM_1305 catalase was present in six of the B. pumilus strains tested, YjqC was not detected in ATCC 7061 and BG-B79. Furthermore, both catalases were localized in the spore coat layer along with laccase and superoxide dismutase. Although the initial catalase activity in ATCC 7061 spores was higher, it was less stable over time than the SAFR-032 enzyme. We propose that synergistic activity of YjqC and BPUM_1305, along with other coat oxidoreductases, contributes to the enhanced resistance of B. pumilus spores to hydrogen peroxide. We observed that the product of the catalase reaction, gaseous oxygen, forms expanding vesicles on the spore surface, affecting the mechanical integrity of the coat layer, resulting in aggregation of the spores. The accumulation of oxygen gas and aggregations may play a crucial role in limiting further exposure of Bacilli spore surfaces to hydrogen peroxide or other toxic chemicals when water is present.

INTRODUCTION

The genus Bacillus is among a few bacterial genera that form endospores to survive adverse conditions. A bacterial endospore is a metabolically dormant form of life that is much more resistant to environmental challenges than a vegetative cell. These include heat, desiccation, lack of nutrients, and exposure to UV and gamma radiation, as well as organic chemicals and oxidizing agents (43). This exceptional resistance is attributed to the spore's structure and the biochemical properties of its components. This fact has been supported primarily by the investigation of Bacillus subtilis and Bacillus anthracis spores (19, 22, 24, 35). However, more information on the spore's physiology and biochemical composition is essential to understand their contribution to the survival strategies of different members of the Bacilli class.

Three compartments of spores are identified on the basis of their morphology, i.e., the core, cortex, and coat. The latter is the outermost layer, which directly interacts with the surrounding environment (18, 47). The coat actively limits the passage of harmful chemicals and is a gate that recognizes germinants (e.g., water and nutrients) (6, 17). The coat layer is composed largely of specific proteins that are not present in vegetative cells (23, 27, 35). For example, the coat protein laccase (CotA) is responsible for the formation and deposition of the brown protective pigment of B. subtilis spores that may protect them from hydrogen peroxide (25, 37), while CotE is involved in coat assembly (49). The CotJC protein has a manganese catalase domain, but catalase activity has not been reported in B. subtilis spores (44).

Bacillus pumilus SAFR-032 was repeatedly isolated from the spacecraft assembly facilities at the Jet Propulsion Laboratory (JPL), Pasadena, CA (36). Endospores of B. pumilus SAFR-032 are resistant to high concentrations of H₂O₂ (5% for 60 min) and simulated Martian UV irradiation (29, 38). The SAFR-032 genome was sequenced, and numerous protein-encoding genes were identified as putative candidates responsible for the strain's enhanced resistance to the various sterilization treatments tested (21). Understanding its hydrogen peroxide resistance is particularly important since H₂O₂ is used in current and newly developed sterilization methods (29, 46). So far, the major role in peroxide resistance has been attributed to small acid-soluble proteins associated with DNA (42). Recently, Bosak et al. (12) described a cyclic “sporulene” compound that is located in the inner membrane surrounding the core and limits hydrogen peroxide diffusion into the endospore. The coat was also found to protect the spore, as decoated spores have a lower ability to survive hydrogen peroxide treatment (40). From the genome data, Gioia et al. (21) reported that SAFR-032 has two germination catalase genes, katX1 and katX2, that are homologs of the B. subtilis germination catalase, the KatX protein. In addition, the SAFR-032 genome has two catalase genes, yjqC and bpum_1305, with two manganese ions replacing heme in the enzyme catalytic centers; hence, the proteins they encode are called pseudocatalases or manganese catalases (15). Manganese catalase (YjqC) was found in B. subtilis spore extract, but its biochemical function was not elucidated (33). B. subtilis also has the ydbD gene, which codes for a putative protein with a manganese domain, but there is no counterpart in the B. pumilus SAFR-032 genome (21, 32).

Although multiple manganese catalase genes are present in Bacillus species, manganese catalases are not widespread in nature and all of them have a lower affinity for hydrogen peroxide than heme catalases do (15). The presence of manganese catalases has been reported in Thermoleophilum album (2), Pyrobaculum calidifontis (3), Lactobacillus plantarum (31), and Thermus thermophilus (48). Manganese catalases from L. plantarum (9) and T. thermophilus (4) have been crystallized, and their molecular structures revealed that the active site contains dimanganese cations linked by two oxygen atoms. Both catalases are hexamers of about 30-kDa monomers with a four-helix bundle motif containing conserved glutamate, aspartate, and histidine residues, some of which are involved in manganese chelation (4, 9).

In this study, we identified two manganese catalase proteins in the SAFR-032 spore coat that were annotated earlier as YjqC and BPUM_1305 and we detected catalase activity in spores of all of the B. pumilus strains investigated. The research presented here suggests that both proteins may contribute to the spores' H₂O₂ resistance. We observed that catalase activity and the product of the catalase reaction, gaseous oxygen, both contribute to hydrogen peroxide resistance. We hypothesize that the oxygen vesicles form a gaseous barrier on the spore surface that limits hydrogen peroxide diffusion into the spores' interior, further enhancing their resistance to hydrogen peroxide. As vesicles expand and break, they cause mechanical disruption of the coat layer, resulting in adjacent spore aggregation. The aggregate traps oxygen gas and allows spores to float on the liquid surface, further limiting their exposure to hydrogen peroxide.

MATERIALS AND METHODS

Bacterial strains and spore preparation.

The B. pumilus strains used in this work were SAFR-032 (JPL, Pasadena, CA), ATCC 7061 (American Type Culture Collection, Manassas, VA), BG-B79 (Department of Bacteriology, National Veterinary Institute, Uppsala, Sweden), 8A4, 8A6, and 14A1 (Bacillus Genetic Stock Center, The Ohio State University, Columbus, OH). Vegetative cells were maintained on tryptic soy agar (TSA; BD, Franklin Lakes, NJ) at 37°C.

Spores were prepared by inoculating Difco sporulation medium (DSM) plates (39) with overnight DSM broth cultures. The plates were allowed to grow for 48 h at 37°C and then placed at a suboptimal growth temperature of 25°C until >99% of the cells had sporulated. Sporulation and endospore purity were monitored with a phase-contrast microscope (Nikon Eclipse 55i, NIS-147 Elements Br. 3.0). Spores were purified to remove the remaining vegetative cells and cellular debris by using a modified method described earlier (39). Briefly, cells from agar plates were scraped into sterile deionized water and pelleted again by centrifugation after each washing step (30,000 × g, 10 min, 4°C). The pellets were washed with a solution containing 1 M KCl and 0.5 M NaCl, with 1 M NaCl, and then three times with deionized water. Finally, the endospore suspension was heat shocked at 80°C for 15 min and stored at 4°C while protected from light. Spore viability was estimated by serial dilutions, plating on TSA, and enumeration of CFU.

Preparation of spore protein extracts.

Total spore protein extracts of intact or decoated spores were prepared by the method described previously (10), with modification as follows. A spore suspension (450 μl, optical density at 580 nm [OD₅₈₀] of 20) in water was centrifuged (16,000 × g, 10 min, 25°C), and the pellet was suspended in 100 μl of 1× PAGE sample loading buffer (Bio-Rad, Hercules, CA). The suspension was boiled for 5 min, homogenized in a bead mill beater (Biospec Products Inc., Bartlesville, OK) with 0.1 g of 0.1 mm zirconia/silica (Biospec Products, Inc., Bartlesville, OK) for a total of 5.5 min (settings: 1 min followed by 9 × 30 s at 30-s intervals). The homogenate was boiled for 5 min and centrifuged at 16,000 × g for 1 min. Thirty microliters of supernatant was analyzed by 12% SDS-PAGE and visualized by Coomassie brilliant blue R-250 staining (41).

Decoating of spores and preparation of coat protein extracts.

Spore suspension (450 μl, OD₅₈₀ of 20) was centrifuged at 16,000 × g for 10 min to remove water. The spore coat was removed using detergent in the presence of salt and a reducing agent under alkaline conditions as described by Bagyan and Setlow (6), with the following modifications. The supernatant obtained from decoated spores was desalted using Sephadex G-25 PD-10 disposable columns (Amersham Biosciences, Uppsala, Sweden). All fractions containing proteins were combined and freeze-dried (Freeze Zone 6; Labconco Corporation, Kansas City, MO). The lyophilized material was suspended in 100 μl of 1× SDS sample loading buffer (Bio-Rad, Hercules, CA). A 30-μl sample of each extract was analyzed by 12% SDS-PAGE and visualized by Coomassie brilliant blue R-250 staining (41).

In-gel digestion.

For functional distribution of proteins and single-band identification, PAGE lanes were cut horizontally into slices 1.5 mm wide and then cut into cubes of ∼1 mm³. In-gel digestion was performed by a modified method described earlier (45). Briefly, each slice was destained overnight in 500 μl water containing 50% acetonitrile and 25 mM ammonium bicarbonate. The extract was discarded, and gel pieces were suspended in 200 μl of 100% acetonitrile to dehydrate until the pieces turned white. Gel pieces were rehydrated in a 25 mM solution of ammonium bicarbonate containing 10 mM dithiothreitol and incubated for 45 min at 50°C, followed by 30 min of incubation in a 25 mM ammonium bicarbonate solution containing 55 mM iodoacetamide (both from Sigma-Aldrich, St. Louis, MO) at room temperature in the dark. Gel pieces were washed with 25 mM ammonium bicarbonate and dehydrated by adding 200 μl of acetonitrile until the pieces turned white. Finally, the gel pieces were rehydrated in trypsin solution (12.5 ng/μl in 25 mM ammonium bicarbonate) and incubated at 37°C for 16 h. The peptides were eluted from the gel by 50% acetonitrile-water containing 0.1% formic acid, and the extract was concentrated with a SpeedVac. Just before liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis, dry samples were dissolved in 5% acetonitrile–0.1% formic acid and cleared by centrifugation at 16,000 × g for 10 min.

Peptide sequencing by LC-MS/MS.

Peptides were separated using reverse-phase LC on a nano-ACQUITY Ultra Performance Liquid Chromatograph (Waters Corporation, Milford, MA) and analyzed using a Q-TOF Premier tandem mass spectrometry system equipped with a nano-electrospray ionization (nano-ESI) source (7, 8). A 2-μl sample was loaded onto a Symmetry C₁₈ trap column (0.18 by 20 mm) and then separated on a BEH 130 C₁₈ analytical column (0.075 by 200 mm; Waters Corporation, Milford, MA). A 0.1% formic acid solution in H₂O was used as solvent A, and 0.1% formic acid in acetonitrile was used as solvent B. Peptides were trapped by 100% solvent A for 3 min at a 10-μl/min flow rate. Separation was performed at 35°C at 0.4 μl/min under the following conditions: (i) isocratic for 1 min, 95% solvent A and 5% solvent B; (ii) gradient for the next 44 min of separation, solvent A concentration gradually decreased from 95% to 50% and solvent B concentration increased from 5% to 50%; (iii) gradient for the next 5 min, solvent A concentration decreased to 10% and solvent B concentration increased to 90%; (iv) isocratic for 5 min, 10% solvent A and 90% solvent B; and (v) gradient for 5 min, solvent A concentration increased to 80% and solvent B concentration decreased to 20%. The nano-ESI source was set as follows: capillary voltage, 3.7 kV; cone voltage, 30 V; source temperature, 120°C; nebulizing gas pressure, 0.45 × 10⁵ Pa; collision energy, 5.0 V; detector voltage, 1,825 V. The data were acquired in MS survey mode. The following settings were used for the MS survey: a 300- to 2,000-Da mass range, a 1-s scan time, and a 0.1-s interscan delay. The threshold of MS/MS acquisition was set to 20 counts per second. The MS/MS acquisition settings were as followed: a mass range 50 to 2,000 Da, 3 ions selected from a single MS survey scan, a 2-s scan time, and a 0.05-s interscan delay. [Glu¹]-fibrinopeptide B (Sigma-Aldrich, St. Louis, MO) was used as a lockmass with a 30-s frequency.

Proteomic data analysis.

The raw data files were generated to peak lists and saved as pkl files by ProteinLynx Global Server (PLGS) 2.3 software (Waters, Milford, MA). For protein identification, Mascot software (Matrix Science, London, United Kingdom) was used and files were searched against the B. pumilus SAFR-032 protein sequence database. The database was created from the genome sequence information available at the National Center for Biotechnology Information website (ftp://ftp.ncbi.nih.gov/genomes/Bacteria/Bacillus_pumilus_SAFR_032_uid59017/NC_009848.faa). The parameters used by Mascot were as follows. (i) Trypsin was the specific enzyme, (ii) the peptide window tolerance was ±0.2, (iii) the fragment mass tolerance was ±0.1, (iv) the number of missed cleavage sites was 1, (v) an individual ion score of >18 was significant for peptide identification [P < 0.05; the individual ion score is −10 × log(P), where P is the probability that the observed match is the random match], and (vi) carbamidomethyl C was the only fixed amino acid modification set.

Enzyme assays.

Catalase activity was determined spectrophotometrically (Agilent/Hewlett Packard 8453 UV-Vis spectrophotometer; Agilent Technologies, Santa Clara, CA) by measuring the decomposition of H₂O₂ at 240 nm (ε₂₄₀ = 43.6 M⁻¹ cm⁻¹) (11). The spore suspension (OD₅₈₀ of ∼20) was centrifuged at 16,000 × g for 10 min, and water was discarded. The spores were suspended in 1 ml of 50 mM potassium phosphate buffer (pH 7.0) containing 15 mM (0.05%) H₂O₂ (EMD Chemicals, Merck KGaA, Darmstadt, Germany) so that the OD₅₈₀ was 1.0. Spore suspension aliquots of 1 ml were removed at T₀ and T_n at n time points and centrifuged at 16,000 × g for 5 min, and the concentration of hydrogen peroxide in the supernatant was measured.

Optimization of catalase activity.

The pH optimum for the catalase activity was determined at 25°C by measuring hydrogen peroxide decomposition by spores suspended in the following buffers: 50 mM glycine-HCl at pH 3.6, 50 mM potassium phosphate buffer at pH 5.8, 50 mM potassium phosphate buffer at pH 7.0, 50 mM potassium phosphate buffer at pH 8.0, and 50 mM glycine-NaOH at pH 10.0.

The optimum temperature for the catalase activity in spores was determined by using 15 mM (0.05%) H₂O₂ potassium phosphate buffer at pH 7.0 in a temperature range of 25 to 80°C.

Catalase inhibition assay.

One milliliter of spore suspension (OD₅₈₀ = 1.0) was centrifuged at 16,000 × g for 10 min, and the supernatant was discarded. Spores were suspended in 1 ml of prewarmed 15 mM (0.05%) H₂O₂ in 50 mM potassium phosphate buffer at pH 7.0. The inhibitors were added to final concentrations of 0.1 mM NH₂OH (freshly prepared before the measurement) and 5 mM NaN₃ (both from Sigma, St. Louis, MO) (2). After incubation at 37°C for 60 min with shaking (200 rpm), the spore suspension was centrifuged at 16,000 × g for 10 min. The H₂O₂ concentration in the supernatant was determined by measuring absorbance at 240 nm.

Hydrogen peroxide resistance.

The assay of resistance to 5% hydrogen peroxide was performed for all Bacillus strain spores as described previously by using ∼10⁸ spores ml⁻¹ (29). Also, the same experiment was performed with the SAFR-032 strain with hydroxylamine as a manganese catalase inhibitor (30). To assess the contribution of manganese catalase to spore survival, 10 μl of 0.1 M NH₂OH at pH 6.0 was added to the 823-μl spore suspension, the suspension was incubated at room temperature (25°C) for 10 min, and 167 μl of 30% H₂O₂ was added. The suspension was incubated at 25°C (with shaking at 150 rpm) for 60 min. At the end of the incubation period, a 100-μl sample was diluted 1:10 with bovine catalase (100 μg/ml in phosphate-buffered saline [PBS; Sigma-Aldrich, St. Louis, MO], sterilized with a 0.2-μm-pore-size polyethersulfone filter; VWR) and then further diluted. The bovine catalase was used to decompose any remaining hydrogen peroxide that would potentially inhibit spore germination. Dilutions of 1:100 and 1:1,000 contained 100 μl of bovine catalase solution (100 μg/ml in PBS) because hydroxylamine also inhibited bovine catalase activity.

Scanning electron microscopy.

A 50-μl volume of B. pumilus SAFR-032 spore suspension (OD₅₈₀ = 20) was added to 450 μl of 2% albumin containing 15 mM (0.05%) H₂O₂. After 5 min of incubation on ice, the aliquots were subjected to flash freezing in liquid nitrogen, followed by a solvent replacement procedure (26). The samples frozen in 2-ml microtubes were overlaid with 1.5 ml of a −80°C chilled solution of acetone containing 5% glutaraldehyde (Electron Microscopy Sciences, Hatfield, PA). After 3 days, the solution was decanted and fresh chilled 5% glutaraldehyde in acetone was added to the samples. The next day, the procedure was repeated and samples were moved from −80°C to −20°C. In the final step, the glutaraldehyde solution was replaced with anhydrous acetone and the samples were kept for 2 days at 5°C. The samples were dried at room temperature for 2 days. Both samples were carbon coated before examination with an Amray 1830 scanning electron microscope (SEMTech Solutions, Inc., North Billerica, MA).

RESULTS

Proteomic analysis of total spore protein extracts.

B. pumilus SAFR-032 and ATCC 7061 spore protein extracts were analyzed by SDS-PAGE. The protein profiles obtained from the gel slice extracts of both strains' spores were similar, with few significant differences in the intensities of protein bands. We identified 374 and 350 proteins for SAFR-032 and ATCC 7061 (data not shown). The most numerous group in both strains were proteins dedicated to translation, including proteins of both ribosomal subunits (19%), but there were only a few proteins dedicated to amino acid metabolism, suggesting that germinating spores use protein stored in spores as an amino acid source. We found several oxidative stress response proteins that may contribute to the higher reactive oxygen species (ROS) resistance of SAFR-032 spores (Table 1). However, manganese catalase (BPUM_1305), manganese superoxide dismutase (SodA), laccase (CotA), catalase KatX1, peroxiredoxin (YkuU), and thiol peroxidase (Tpx) were detected in both strains. We did not detected manganese catalase (YjqC), superoxide dismutase (SodF), glutathione peroxidase (Bsa), or thioredoxin (TrxA) in our proteomic analysis of ATCC 7061 gel slices. YjqC was very abundant in SAFR-032 spore protein extract but was not present in that of ATCC 7061 (Fig. 1). The PAGE-predicted molecular mass of YjqC corresponded to the calculated mass (∼31 kDa). The BPUM_1305 protein, with a predicted molecular mass of 33 kDa, was detected in both strains, although a band corresponding to BPUM_1305 was difficult to distinguish because of the similarity of the molecular masses of other proteins. However, the band of this protein was possible to visualize when the coat protein fractions were extracted and analyzed by PAGE separately. The two manganese catalases showed 27.6% identity (data not shown), and the peptide amino acid sequences detected by LC-MS/MS were unique to each of the proteins. In addition to the main band, YjqC was detected in multiple PAGE slices in the SAFR-032 extract, which did not corresponded to its molecular mass, so the PAGE and mass spectrometry results confirmed the existence of cross-linking between coat proteins.

Table 1.

Oxidative stress response proteins identified in B. pumilus SAFR-032 spore extract

Identified proteina	NCBI accession no.	Expected/calculated molecular masses (Da)	% sequence coveragea	Mascot scorea,b	Peptide(s)a
Catalase KatX1	YP_001488916.1	60,947/67,350	4	55	K.FYTEEGNWDLVGNNLK.I, R.TFSYSDTQR.Y
			5	53	R.DALKFPDLVHAFKPDPVTNR.Q, K.NLVNTLATCQK.D
Glutathione peroxidase (BsaA)	YP_001487155.1	18,227/17,470	7	19	K.ELESSIIALLDK
Iron superoxide dismutase (SodF)	YP_001487090.1	32,422/27,390	6	55	R.YDALEPFISK.E, R.LEILQAER.H
Manganese catalase	YP_001486548.1	33,407/37,770	21	60	R.LLIDLPRPDHPDANGAAAVQELLGGK.F, K.ALEVATGVDVGK.M
			12	93	K.GSHPIDGKPLTVIEGTPQGAPVPDYR.E
Manganese superoxide dismutase (SodA)	YP_001487459.1	22,397/22,590	14	117	K.LEITSTPNQDSPLTEGK.T, K.FKSDFAAAAAGR.F, K.HHNTYVTNLNK.A,
			12	80	R.FGSGWAWLVVNNGK.L
Multicopper oxidase/laccase (CotA)	YP_001485796.1	58,882/40,280	4	117	K.VWPYLEVEPR.K
			2	65	R.LGSVEVWSIVNPTR.G
Peroxiredoxin (YkuU)	YP_001486562.1	20,578/22,590	26	112	K.YPLAADTNHTVSR.E, R.EYGVLIEEEGIALR.G
			31	179	R.VLQALQTGGLCPANWKPGQK.T, R.GLFIINPEGELQYQTVFHNNIGR.D
Sporulation-related manganese catalase (YjqC)	YP_001487572.1	31,115/31,150	38	438	K.ELQYEAKPSKPDPLYAK.K, K.YIIASGNLMADFR.A, R.ANLNAESQGR.L, R.GVKDMLSFLIAR.D, K.DMLSFLIAR.D, R.DTYHQNMWIAAIK.E, R.DTYHQNMWIAAIKELEER.E, K.ELEEREGDVVVPTTFPR.S
Thiol peroxidase (Tpx)	YP_001487800.1	17,954/17,470	21	60	R.WCGANGIENVETLSDHR.D, K.VVYTEYVSEATNHPNYEK.A
			43	199	K.VGEQAPDFTVLTNSLGEVSLSDLTGK.V, K.VTIISVIPSIDTGVCDAQTR.R, R.FNEEAAGLGPVNIYTISADLPFAQAR.W
Thioredoxin (TrxA)	YP_001487729.1	11,445/8,610	32	72	K.IDVDDNQETAGK.Y, K.YGVMSIPTLLVLK.D, K.EALAELVNK.H

^aBold font indicates results common to both SAFR-032 and ATCC 7061, while lightface font and italics indicate results for SAFR-032 and ATCC 7061, respectively.

^bMascot search probability-based MOWSE score. The ion score is −10 × log(P), where P is the probability that the observed match is a random event. Individual ion scores of >18 indicate identity (P < 0.05). Protein scores are derived from ion scores as a nonprobabilistic basis for ranking of proteins hits.

Fig 1.

SDS-PAGE analysis of proteins extracted from intact and decoated B. pumilus spores. Lanes: ST, molecular mass standards (sizes in kDa are on the left); 1, SAFR-032; 2, ATCC 7061; 1A and 2A, intact spore protein extracts; 1B and 2B, decoated spore protein extracts; 1C and 2C, coat protein extracts. Lanes 1A and 1B were cut into ∼1.5-mm slices from the bottom to the top for protein identification by LC-MS/MS. A similar amount of protein (∼10 μg) was loaded into each well. The black arrowheads indicate a sporulation-related manganese catalase (YjqC). The white arrowheads indicate the second manganese catalase (BPUM_1305). The numbers indicate the excised PAGE bands.

We also screened other B. pumilus strains for the presence of the manganese catalases. BPUM_1305 was detected in all of the strains analyzed, and ATCC 7061 had the highest number of peptides detected, Mascot score, and sequence coverage (Fig. 2 and Table 2). YjqC was absent from ATCC 7061 and BG-B79 spores.

Fig 2.

SDS-PAGE analysis of B. pumilus spore protein extracts. Lanes: ST, molecular mass standards (sizes in kDa are on the left); 1, SAFR-032; 2, ATCC 7061; 3, BG-B79; 4, 8A4; 5, 8A6; 6, 14A1. A similar amount of protein was loaded into each 12% SDS-PAGE well. The rectangles (∼1 cm) were excised, digested in the gel with trypsin, and sequenced. The arrowheads indicate the sporulation-related manganese catalase (YjqC). BPUM_1305 was identified in all of the gel extracts tested.

Table 2.

Manganese catalases identified by SDS-PAGE and LC-MS/MS in B. pumilus spores

Strain	YjqC			BPUM_1305
	Mascot scorea	% coverage	No. of peptides	Mascot scorea	% coverage	No. of peptides
SAFR-032	461	34	9	119	16	3
ATCC 7061	—b	—	—	199	23	5
BG-B79	—	—	—	154	15	4
8A4	142	20	5	43	4	1
8A6	248	25	6	28	4	1
14A1	269	20	5	15	8	1

^aMascot search probability-based MOWSE score. The ion score is −10 × log(P), where P is the probability that the observed match is a random event. Individual ion scores of >18 indicate identity (P < 0.05). Protein scores are derived from ion scores as a nonprobabilistic basis for ranking of proteins hits.

^b—, not detected.

Localization of oxidative stress response proteins.

We performed spore decoating by using detergent in the presence of salt and a reducing agent under alkaline conditions to extract the coat proteins of SAFR-032 and ATCC 7061 spores. SDS-PAGE analysis of decoated spores and coat protein extracts revealed that some of the proteins are present only in the coat (Fig. 1). Bands from the coat fraction were excised from the PAGE gel for ESI-MS/MS sequencing and Mascot identification. The distinctive abundant band in native spores of SAFR-032 was identified as a sporulation-related manganese catalase (YjqC). YjqC was not detected in decoated spore protein extracts, but it was abundant in the coat fraction, confirming the localization of this protein in the spore coat. BPUM_1305 was identified in the coat fractions of both strains. Additional oxidoreductases were identified in the coat bands as well. Table 3 summarizes the results for the oxidative stress response enzymes detected in the SAFR-032 coat fraction.

Table 3.

Oxidative stress response proteins identified in B. pumilus SAFR-032 coat fraction

Identified protein/band no.a	NCBI accession no.	Expected molecular mass (Da)	Sequence coverage (%)c	Mascot scoreb,c	Peptidesa
Iron superoxide dismutase/4, 9	YP_001487090.1	32,422	9	195	R.YDALEPFISK.E, K.HHQSYVDGLNQAELALK.R, R.QIEQSFESYDAFK.A, R.LEILQAER.H, K.AAYVDR.W
Outer spore coat protein	YP_001485796.1	58,882	7	138	K.FVDELPIPEVAEPVK.K, R.DFEATGPFFER.E, K.VWPYLEVEPR.K
(CotA)/multicopper oxidase (laccase)/3, 5, 8; 3, 7, 9			6	65	R.ILNASNTR.T, R.TLTLTGTQDK.Y, R.LGSVEVWSIVNPTR.G
Manganese catalase/6, 8; 5, 6	YP_001486548 0.1	33,407	36	329	R.VYEMTDHPTAR.E, R.EMIGYLLVR.G, K.ALEVATGVDVGK.M, K.LYTFSDTDYKDINK.I, R.GGVHVVAYAK.A
			24	215	K.YFDSAR.K, R.KFEDQNIHTK.L, K.LYTFSDTDYK.D, K.GSHPIDGKPLTVIEGTPQGAPVPDYR.E, R.ELPEEFAPGISK.E, R.LLIDLPRPDHPDANGAAAVQELLGGK.F
Sporulation-related manganese catalase (YjqC)/1, 2, 3, 4, 5, 6, 7	YP_001487572.1	31,115	49	611	MFYHIK.E, K.ELQYEAKPSKPDPLYAK.K R.LLDKAPVK.E, K.YIIASGNLMADFR.A, R.ANLNAESQGR.L, R.LYEMTDDR.G K.DMLSFLIAR.D, R.DTYHQNMWIAAIK.E, R.EGDVVVPTTFPR.S, K.QQVSYDLFNFSR.G R.SMDGKGEFR.Y, R.YISAPVVFGSAPK.L, K.ELFNTPK.K

^aResults common to both SAFR-032 and ATCC 7061 are in bold, while results for SAFR-032 and ATCC 7061 are in lightface and italics, respectively.

^bMascot search probability-based MOWSE score. The ion score is −10 × log(P), where P is the probability that the observed match is a random event. Individual ions score >18 indicate identity (P < 0.05). Protein scores are derived from ions scores as a nonprobabilistic basis for ranking proteins hits.

^cNormal font and italics indicate results for SAFR-032 and ATCC 7061, respectively.

Catalase enzymatic properties.

Catalase activity optimization was performed at various pHs and temperatures for SAFR-032 and ATCC 7061 spore suspensions (data not shown). Catalase activity was measured over a pH range of 3.6 to 10.0, and the highest activity was detected in 50 mM potassium phosphate buffer at pH 7.0. Catalase activity was detected from 25 to 80°C, with optimal temperatures of 70 and 37°C for spores of SAFR-032 and ATCC 7061, respectively (Table 4). Moreover, catalase activity was detected in the spores of all of the strains at all of the temperatures tested. The ATCC 7061 strain had the highest activity at 37°C, while the spores of the other strains showed similar activity at 70°C, which suggests that this enzyme is thermostable.

Table 4.

Catalase specific activities detected in B. pumilus spore suspensions

Conditionsa	Mean catalase activity (nmol min−1 ml−1 OD580−1) ± SD
	SAFR-032	ATCC 7061
37°C	112 ± 1	224 ± 3
70°C	180 ± 3	208 ± 8
37°C, 60-min H2O2 treatment	53 ± 1	25 ± 5
37°C, 0.1 mM NH2OH	24 ± 3	35 ± 4
37°C, 5 mM NaN3	87 ± 7	186 ± 10

^aEach in 15 mM H₂O₂.

Hydroxylamine is a good inhibitor of manganese catalase at a concentration of 0.1 mM (30). The catalase activities were inhibited by hydroxylamine in all spores of B. pumilus strains. Sodium azide was used as an inhibitor of heme catalase at a concentration of 5 mM (28). SAFR-032 catalase activity was inhibited by 75% and that of ATCC 7061 was inhibited by 83% by incubation with hydroxylamine, while the activities of both strains decreased by only 12% in the presence of sodium azide (Table 4).

Spore resistance to hydrogen peroxide.

A small percentage of the spores of all of the B. pumilus strains (∼10⁸ ml⁻¹), except ATCC 7061, survived 1 h of incubation in 5% H₂O₂. We then used ∼5.63 × 10⁸ ml⁻¹ SAFR-032 spores for the enumeration test with hydroxylamine. The survival of spores treated with 5% H₂O₂ was 0.12%; however, in the presence of 0.1 mM hydroxylamine, spore survival decreased to 0.0046%. In the control containing 0.1 mM hydroxylamine, approximately 62% of the SAFR-032 spores survived, suggesting some toxicity of hydroxylamine for B. pumilus spores (Fig. 3).

Fig 3.

Viability of SAFR-032 spores (5.63 × 10⁸) after 1 h of exposure to 5% hydrogen peroxide. The bars represent means of three replicates, and the error bars represent the standard deviations. Bars: 1, untreated control (standard deviation [SD], ±1.2 × 10⁸); 2, treatment with 5% H₂O₂ (SD, ±5.6 × 10⁵); 3, control treated with 0.1 mM NH₂OH (SD, ±1.3 × 10⁸); 4, treatment with 0.1 mM NH₂OH for 10 min before exposure to 5% H₂O₂ (SD, ±7 × 10³).

We also found that the catalase activity of SAFR-032 decreased by 53% and that of ATCC 7061 decreased by 89% after a 60-min exposure to 5% hydrogen peroxide (Table 4). We know from previous studies (29) and work performed in our laboratory that ATCC 7061 spores do not survive a 60-min exposure to 5% H₂O₂.

The addition of albumin, followed by rapid freezing and solvent replacement, allowed the visualization and preservation of oxygen vesicles deforming the spore coat as they emerge from the spore surface in the presence of hydrogen peroxide. During the gradual temperature increase, the acetone solubilized the ice in the sample and glutaraldehyde-albumin copolymer fixed frozen oxygen vesicles on the B. pumilus spore surface (Fig. 4B and D). Control spore samples (Fig. 4A and C) without hydrogen peroxide showed some deformations of the spore surface, but the difference from H₂O₂-treated samples is profound.

Fig 4.

Spores of B. pumilus SAFR-032 visualized by scanning electron microscopy after treatment with H₂O₂. In order to preserve oxygen vesicles, the reaction was performed in 2% albumin and a freezer substitution solvent replacement procedure (acetone containing 5% glutaraldehyde) was used after the rapid freezing of samples in liquid nitrogen. Shown are control spores (magnifications: A, ×20,000; C, ×50,000) and spores treated with H₂O₂ (magnifications: B, ×20,000; D, ×50,000; E, ×60,000; F, ×110,000). The white arrows indicate oxygen gas vesicles fixed on the spore surface during hydrogen peroxide treatment. The spores in sample E were suspended in water, treated with H₂O₂, and freeze-dried; no albumin or glutaraldehyde was used. Panel E provides evidence that treatment with H₂O₂ causes part of the spore surface to be dislodged and that the resulting membrane fragments bind spores together.

DISCUSSION

The research presented here suggests that catalase activity detected in the B. pumilus spore coat contributes to the spore's resistance to hydrogen peroxide exposure. Genes for sporulation-related manganese catalase (YjqC) and other oxidative stress response proteins were previously identified in the genome of B. pumilus SAFR-032 (21).

To verify the genomic information, we investigated SAFR-032 and ATCC 7061 spore protein extracts. The latter strain served as a control since it is more sensitive to the 5% hydrogen peroxide treatment (29). Proteomic analysis of these two strains revealed that manganese catalase (BPUM_1305) was present in both strains, while the sporulation-related manganese catalase protein (YjqC) was found only in SAFR-032. Generally, our goal was to compare the model organism (SAFR-032) with the control (ATCC 7061), as another research group did (29). We included other B. pumilus strains for comparison of hydrogen peroxide resistance. Catalase activity was detected in other strains, and those strains survived treatment with 5% hydrogen peroxide. Further investigation revealed the presence of BPUM_1305 in other B. pumilus strains tested. In addition to ATCC 7061 spores, YjqC was also absent from B. pumilus BG-B79, a strain isolated from a biogas plant in Sweden (5). Interestingly, the BG-B79 strain has 100% 16S rRNA sequence similarity to SAFR-032 on the basis of the sequences deposited in the NCBI database. This result indicates that proteomic analysis can distinguish differences between strains that genomic analysis cannot. The two manganese catalases have features in common with enzymes characterized earlier. The YjqC catalase has extensive homology to crystallized L. plantarum catalase (46.8%) (9), while the BPUM_1305 amino acid sequence is more similar to that of T. thermophilus catalase (54.5%) (4). The high optimum temperature of the enzyme may suggest a horizontal gene transfer from a thermophile. The molecular masses of YjqC and BPUM_1305 are 31 and 33 kDa, respectively. Our results indicated that BPUM_1305 may be common in B. pumilus spores, while YjqC may be more strain specific.

Though it is accepted that the coat protects spores from hydrogen peroxide, the exact mechanism has not been determined (40). CotA present in B. subtilis 168 spores was identified as a laccase (multicopper oxidase) and found to protect spores from hydrogen peroxide (25, 37). We detected CotA in the whole spore extracts and the coat fractions of SAFR-032 and ATCC 7061. However, B. pumilus spores are white and lack the brown pigment produced by laccase, which is a characteristic trait of B. subtilis spores. The proteomic analyses of SAFR-032 coat proteins confirmed both catalase proteins' localization in the coat. Furthermore, the genomic analysis (21) and our proteomic data revealed other oxidoreductase proteins in coat fraction as potential contributors to resistance to other ROS (Table 3). Our results strongly indicate that the SAFR-032 spore coat is filled with oxidative stress response proteins. The YjqC protein was detected previously in B. subtilis spore protein extracts by Kuwana et al. (33), and Abhyankar et al. (1) found that YjqC is located in the outer spore coat of this species. In B. anthracis, superoxide dismutases are located in the spore exosporium, which is a characteristic spore structure present in most pathogenic members of the class Bacilli (16). The presence of superoxide dismutases in B. anthracis spores facilitates their virulence-enhancing survival in the macrophage vacuole. Analogously, the presence of large amounts of manganese catalase enzymes in the outer layers in SAFR-032 spores might help this strain to survive the multiple hydrogen peroxide treatments used in clean rooms at the National Aeronautics and Space Administration (NASA) JPL. Our observations suggest that although BPUM_1305 has higher activity, it is less stable over reaction time than YjqC. Therefore, both catalases' activities may contribute to the higher resistance of SAFR-032 spores to hydrogen peroxide. Similar conclusions about the additive effect of enzyme activities were reached when superoxide dismutases of B. anthracis spores were investigated (16). Knocking out two out of four superoxide dismutases surprisingly led to an increase in superoxide dismutase enzyme activity, presumably to compensate for the loss of the deleted dismutases. There was a total loss of activity when all four superoxide dismutases were deleted. Therefore, the increased catalase activity of ATCC 7061 spores that we observed may compensate for the absence of YjqC, although spores of this strain are not as resistant because of the enzymatic instability of BPUM_1305.

We were not able to obtain yjqC or bpum_1305 knockout mutants by using a method described previously for B. pumilus strains that involves the use of high-voltage electroporation transformation (13). As B. pumilus SAFR-032 lacks comS, a competence gene that is essential for competence regulation in B. subtilis (20), it was not possible to transform the bacterium through natural competence. Therefore, we took a biochemical approach and used hydroxylamine, a manganese catalase-specific inhibitor. Hydrogen peroxide disproportionation by B. pumilus spore catalase was strongly inhibited by hydroxylamine, which was selected as the strongest manganese catalase inhibitor in a previous study (30). Sodium azide weakly inhibited the activity further, confirming that the heme catalases may contribute to spore resistance as well. Also, the ability of B. pumilus SAFR-032 spores to survive was weaker in the presence of hydroxylamine, indicating the protective role of manganese catalases against hydrogen peroxide.

We conclude that YjqC and BPUM_1305 take part in the defense of B. pumilus against hydrogen peroxide. The localization of these two enzymes in the spore coat contributes to the dynamic control of hydrogen peroxide influx into the spore's interior. In addition, evolving oxygen as a final product of the catalase reaction may further protect spores. Our hypothesis is supported by the scanning electron microscope observation of SAFR-032 spores treated with hydrogen peroxide (Fig. 4). The solvent replacement method allowed us to preserve the ice-trapped oxygen vesicles that emerge from the spore's surface in a solution containing hydrogen peroxide. We conclude that oxygen-filled vesicles accumulating within the spore coat form a gas barrier that limits the diffusion of peroxide to the spores' interior so that not only do catalases enzymatically destroy hydrogen peroxide but the accumulated reaction product, oxygen, is retained on the spore surface, providing some additional protection. Our findings also suggest that trapped oxygen provides buoyancy that helps the spores to escape full submersion in H₂O₂. This is especially evident when pelleted spores float to the surface of water containing more than 1% hydrogen peroxide. Previously, other researchers observed a “doughnut-like” deformation of spores of Bacillus sp. strain 34hs1 (34) and an undulated surface of spores of B. pumilus FO-036b after a 60-min exposure to 5% hydrogen peroxide (29), confirming our observations that exposure to H₂O₂ mechanically disturbs the spore coat (Fig. 4E). The sterilization method developed recently by our group uses supercritical fluid carbon dioxide containing 3.3% water and 0.1% hydrogen peroxide (vol/vol/vol) to achieve a 4- to 8-log reduction of the viability of various microbial species, including SAFR-032 spores, and kills dry spores of SAFR-032 with greater efficiency (14, 46) than wet spores. We believe that dry spores lack the gaseous oxygen barrier formation that protects wet spores in an aqueous environment. Also, the sterilization with hydrogen peroxide vapor used by the NASA JPL kills spores through the formation of a gaseous H₂O₂ plasma cloud in an electromagnetic field (29), thus preventing spores from forming protective oxygen vesicles.

We observed that SAFR-032 is one of the most efficient spore formers we have used in our laboratory and it is able to form a very rigid biofilm that floats on the surface of liquid medium (14). There are still some questions to be answered. What is the physiological role of oxygen retention on the spore surface? Are catalases involved in the change in buoyancy of SAFR-032 vegetative cells before they float, form a biofilm, and sporulate? We believe that SAFR-032 has emerged as an important microbial model for further studies of how members of the class Bacilli respond to toxic environmental conditions.