All reagents used were from Sigma–Aldrich unless specified otherwise. Cholesterol was dissolved in ethanol at 1 mg mL^-1and then further diluted into bacterial culture supernatants (Sups). Ham’s F-12 cell culture medium and Complete Serum-Free Medium (CSFM) were from Mediatech Inc., Manassas, VA, USA. Formulated Dulbecco’s Modified Eagle Medium (DMEM) came from Sigma–Aldrich. BSA was of >98% purity and essentially free from globulins and fatty acids. Amplex Red (AR) dye was from Invitrogen. The CytoTox-ONE Homogeneous Membrane Integrity Assay came from Promega, Madison, WI, USA. Succinic acid (SA) concentration was measured using SA assay kit from Megazyme. Protein concentration was estimated using the Bradford protein assay from Bio-Rad. Rabbit anti-streptolysin O antibody was from B-Bridge International, Santa Clara, CA, USA. Anti-rabbit IgG, HRP-linked antibody came from Cell Signaling Technology.

The listed kits were performed according to the manufacturers’ protocols unless otherwise indicated. Cell permeability was measured using the CytoTox-ONE Homogeneous Membrane Integrity Assay which indicated the amount of lactate dehydrogenase (LDH) secreted from cells with a damaged membrane. LDH was measured with a 10-minute coupled enzymatic assay that results in the conversion of resazurin into fluorescent resorufin detected at 530/590 nm. The extent of permeability was calculated as fluorescence relative to completely lysed cells.

For determining the presence of hydrogen peroxide and peroxynitrite the dye AR from Invitrogen was used with a modified protocol. AR (N-acetyl-3,7-dihydroxyphenoxazine) is a colorless derivative of dihydroresorufin that when oxidized produces a colorful fluorescent product resorufin which is detectable at 571 nm. AR and horseradish peroxidase (HRP) were added to DMEM without Phenol Red indicator supplemented with 1 g L^-1 of BSA at a concentration of 0.1 mM and 0.2 U mL^-1, respectively, prior to inoculation with bacteria as described below. Cultures were collected every hour, centrifuged at 10,000 × g for 5 min, and the Sups were removed. The absorbance of Sups was read at 571 nm via spectrophotometer.

The following procedure for bacterial propagation was used for every experiment unless stated otherwise. B. cereus strains 11778 and 14579 (designated by us as BC1 and BC2, respectively) were from the ATCC collection (Manassas, VA, USA). They were grown on agar plates containing Luria broth. Agar plates were re-streaked every seven days and were kept at 8°C until inoculation. Single colonies were used to inoculate into Luria broth and cultures were kept in an incubator shaker at 200 rpm, 37°C for 18 h. Overnight culture at 1:100 dilution was inoculated into either 10 mL of Complete Serum-Free Medium (CSFM), a nitrate rich medium containing 1 g L^-1 of BSA, or DMEM supplemented with 1 g L^-1of BSA. Cultures were grown either under microaerobic (stationary) conditions, in a 6-well plate in an incubator at 37°C, 5% CO₂, or under aerobic conditions, in loosely capped 50 mL tube shaken at 200 rpm at 37°C for 20 h unless specified otherwise. Under stationary conditions, the bacteria consume available oxygen and gradually become hypoxic, thus representing a microaerobic environment. Cultures were collected and the optical density (OD) of 200 μL of bacterial suspensions in a 96-well plate were measured in triplicates using a microplate reader at 600 nm.

Bacterial suspensions were centrifuged at 3000 × g for 15 min and Sups were removed from the bacterial pellet. Penicillin (100 μg mL^-1) and streptomycin (100 U mL^-1) were added to Sups to prevent any bacterial contamination. Fresh Sups were used immediately for challenge experiments.

Human small-airway epithelial cells (primary HSAECs) were from Cambrex Inc., Walkersville, MD, USA. Cells were cultured in Ham’s F-12 medium containing 10% fetal bovine serum, non-essential amino acids, L-glutamine, and pyruvate and grown in 37°C in a 5% CO₂ atmosphere. Cells were seeded into a 96-well plate at a density of 2.5 × 10⁴ per well and grown to confluence. For cytotoxicity assays, 200 μl of bacterial culture Sups were added per well and incubated for 20 min at 37°C, 5% CO₂ without shaking. After cell exposure, the plate was spun at 2000 × g for 5 min, Sups were removed, and 200 μL of 5% resazurin, a redox dye, dissolved in CSFM was added to each well. Resazurin measures cell survival because cellular metabolism breaks down the dye changing its color and fluorescence. Fluorescence was measured with an excitation at 530 nm and emission at 590 nm via a fluorescence reader to determine the differences in cell viability after 2-h incubation with resazurin. Viability was calculated as fluorescence relative to mock-treated cells. Overall, the assay generated consistent results; however, some variability in Sup toxicity was noticed between independently grown HSAEC cultures. This effect was not studied further.

For size exclusion chromatography experiments, bacteria were cultured in DMEM with or without BSA (1 g L^-1) in 6-well plates under microaerobic conditions at 37°C, 5% CO₂ for 20 h. Sups were collected and concentrated eightfold, via SpeedVac at 32°C. 1 mL of Sup concentrate was injected into a HiPrep 16/60 Sephacryl S-200 HR column (GE Healthcare Life Sciences) and run at a rate of 4 mL min^-1 in a 50 mM Tris running buffer, pH 7.5. Flow-through fractions were collected every 2 min between a run time of 13 and 135 min. The toxicity of fractions to HSAECs was assessed immediately using the 100 μL fractions which were added to a 96-well plate of confluent HSAECs containing 100 μL of DMEM per well. Cells were exposed for 1 h and resazurin was used as an indicator for cell survival. The fractions were stored at –20°C for mass spectrometry (MS) analysis.

To determine the protein composition of fractions, samples were prepared for liquid chromatography-tandem mass spectrometry (LC-MS/MS). Fractions were first concentrated 10-fold via SpeedVac at 32°C. Samples were resuspended and reduced in an 8 M urea/10 mM DTT mix for 30 min, alkylated by 50 mM iodoacetamide in the dark for 30 min, and then finally digested by trypsin (Thermo Fisher Scientific) at 37°C overnight. Peptides were purified via Zip-Tip (Millipore) and samples were analyzed by LC-MS/MS using a linear ion-trap mass spectrometer (LTQ, Orbitrap). After sample injection, the column was washed for 5 min with mobile phase A (0.4% acetic acid) and peptides eluted using a linear gradient of 0% mobile phase B (0.4% acetic acid, 80% acetonitrile) to 50% mobile phase B in 30 min at 250 nL/min, then to 100% mobile phase B for an additional 5 min. The LTQ mass spectrometer was operated in a data-dependent mode in which each full MS scan was followed by five MS/MS scans where the five most abundant molecular ions were dynamically selected for collision-induced dissociation using normalized collision energy of 35%. Tandem mass spectra were collected by Xcalibur 2.0.2 and searched against the NCBI mouse protein database using SEQUEST (Bioworks 3.3.1 software from Thermo Fisher) using tryptic cleavage constraints. Mass tolerance for precursor ions was 5 ppm and mass tolerance for fragment ions was 0.25 Da. SEQUEST filter criteria were: Xcorr vs. charge 1.9, 2.2, 3.5 for 1+, 2+, 3+ ions; maximum probability of randomized identification of peptide <0.01.

For native PAGE and SDS-PAGE, protein fractions were mixed with a DNA-loading buffer for native PAGE or with 2x Laemmli sample buffer supplemented with 50 mM DTT for SDS-PAGE. Fractions for SDS-PAGE were boiled with the buffer for 5 min. Samples were separated using 4–20% Tris-Glycine gels and then transferred to nitrocellulose membranes using iBlot Dry Blotting System (Life Technologies). Membranes were blocked with 5% BSA in PBS with 0.01% Tween 20 (PBST) for 1 h, and then incubated in a rabbit anti-streptolysin O antibody (1:1000) overnight. Membranes were washed in PBS with 0.05% Tween-20 and then incubated with an anti-rabbit IgG, HRP-linked antibody (1:5000) for 1 h at room temperature. Western blots were developed using SuperSignal West Dura Extended Duration Substrate (Thermo Fisher Scientific) and then imaged using ChemiDoc XRS+ System (Bio-Rad).

Each measurement was done in triplicate and experiments were performed at least twice for consistency. In figures, error bars indicate 95% confidence intervals of mean (two-tail t-test).

To determine bacterial culture supernatant (Sup) toxicity toward HSAECs we used two B. cereus non-emetic strains, ATCC 11778 and 14579 (designated BC1 and BC2, respectively). The BC2 was reported to express full toxinogenic potential including enterotoxins Hbl, Nhe, CytK, BceT, phospholipases, and the pore-forming CLO (Gohar et al., 2005). BC1 does not express enterotoxins except Nhe (Hansen and Hendriksen, 2001; Gohar et al., 2005), and our preliminary data indicated that it produced CLO (not shown). We compared the acidification and toxicity of Sups grown in aerobic and microaerobic conditions. The aerobic culture was shaken to allow gas diffusion throughout the volume of the bacterial culture. In the microaerobic conditions the amount of oxygen available to bacteria sedimented to the bottom of the wells is strongly limited by diffusion. The onset of anaerobic fermentation was detectable as a reduction of culture pH approximately after 4-h incubation. After 10-h inoculation the pH of microaerobic Sups was close to 5.3, whereas aerobic Sups showed only a small shift from 7.0 to 6.8. Sups demonstrated a detectable cytotoxicity as early as 2 h post-inoculation corresponding to early exponential phase (Figure 1). The release of toxic factors continued upon further incubation. As shown in Figure 2A, the Sups of both strains after 20 h had to be diluted with fresh medium in order to be tested in the dynamic range of the resazurin cell viability assay. BC2 showed the highest toxicity in aerobic cultures killing almost all HSAECs at 32-fold dilution when the exposure time was reduced to 20 min from 2 h (Figure 2). Less toxic BC1 Sups showed higher activity in microaerobic conditions compared to aeration.

To reveal the contribution of CLO to the toxicity of BC1 and BC2 the Sups of microaerobic and aerobic cultures were incubated with 10 μg mL^-1of cholesterol for 1 h prior to HSAEC exposure. CLO is a thiol-activated PFT that is known to be inhibited by the addition of cholesterol (Tweten, 2005). Results from Figure 2B demonstrate cholesterol caused a substantial reduction in the toxicity of microaerobic Sups from both strains. The incubation of cholesterol with Sups from aerobic cultures had little effect (data not shown) consistent with downregulation of CLO expression or/and its inactivation by oxygen (Tweten, 2005). The partial abrogation of Sup toxicity indicated contribution of other virulence factors in addition to CLO.

To evaluate the effect of B. cereus PFTs on host cell membrane the amount of cell permeability in the Sup-treated HSAECs was measured using the Cytotox Homogenous Membrane Integrity Assay (Promega) which is based on the amount of released LDH. Both strains demonstrated a concentration-dependent increase in cell permeability, although the BC2 Sup required much lower concentration (higher dilution) to reach the effect comparable with BC1 Sup (Figure 3). The permeabilizing activity of BC1 Sup was almost completely inhibited in the presence of cholesterol (Figure 3A) while the BC2 Sup demonstrated only a partial cholesterol dependence (Figure 3B). The major drop in viability of exposed HSAECs for both strains took place at rather low levels of permeability when only a fraction of killed cells demonstrated LDH release. These results may be explained by a relatively high contribution of non-permeabilizing toxic factors. It is also likely that the membrane pores produced by the Sups were too small for the efficient translocation of LDH (which is a rather big tetrameric enzyme) but sufficient for the toxic effect of bacterial metabolic products (Walev et al., 2001).

In order to further characterize virulence factors responsible for toxicity, microaerobic BC1 and BC2 cultures were grown, and the Sups from these cultures were concentrated and then fractionated using column chromatography. Sup fractions were added to HSAECs and the toxicity of each was measured using the redox-sensing resazurin dye. For both strains the main portion of toxicity was found in a peak corresponding to the proteins of approximate molecular mass of 50–60 kDa overlapping with the tail of the BSA peak (data not shown). LC-MS/MS analysis of the toxic fractions revealed a list of candidate toxic proteins (Table 2) along with the corresponding numbers of spectral counts (identified tryptic peptides) which can serve as a semi-quantitative measure of the protein abundance. Among the known pathogenic factors the fractions contained large amounts of collagenase, all three Hbl subunits, CLO and a hemolytic sphingomyelinase. No Nhe binding subunit, NheC, necessary for the enterotoxin activity was identified in either toxic fraction. It is plausible that the Nhe migrated through the column as dissociated individual subunits and therefore the NheC was separated from NheAB.

Previous research reported that supplementation of culture medium with BSA enhances the toxicity of the Bacillus species (Elleboudy et al., 2011; Popova et al., 2011; St John et al., 2013). To investigate this effect, cultures of BC1 and BC2 were grown in microaerobic conditions in DMEM supplemented with or without 1 g L^-1 of BSA. Figure 4 demonstrates BSA significantly increased the toxicity of bacterial Sups from both strains. The OD of these cultures indicated minor differences in growth that were not significant (0.72 OD⁶⁰⁰ in DMEM versus 0.80 OD⁶⁰⁰ in DMEM with BSA). Sups from BC1 cultures grown without BSA retained toxicity up to a fourfold dilution compared to the 32-fold dilution for the Sups from cultures grown in the presence of BSA. Reduction of the Sup pH from 6.9 to 5.4 was also noticed in the cultures grown in the presence BSA indicative of the stimulation of acidic anaerobic fermentation (Popova et al., 2011).

It was previously discovered that the effect of BSA may be relevant to stabilization of secreted CLO. Under certain conditions, CLO monomers spontaneously undergo pre-oligomerization which renders it unable to bind to cholesterol located on the cell membrane, resulting in reduced membrane permeabilization (Cowell et al., 1978; Gilbert, 2005). Additionally, CLO was reported to have two conformations depending on the oxidation or reduction of its cysteine residue. The active form contains a free sulfhydryl group and, if oxidized, results in protein inactivation. To determine the contribution of the above mechanisms to the CLO stabilization by BSA, BC1 cultures were grown in the presence or absence of BSA and concentrated by membrane filtration. To test if the concentration of Sups caused CLO inactivation the retentates were diluted with corresponding fresh media to the original volume and their toxicities were assessed after incubation in the presence or absence of cholesterol. After concentration, CLO only remained functional in Sups grown in the presence of BSA (Figure 5A).

The retentates were run on PAGE gels in native and denaturing conditions to analyze the protein mobility and pattern. Standard western blotting conditions were followed using antibody against highly homologous streptolysin O (Heffernan et al., 2007). On SDS-PAGE gel in the presence of DTT the protein demonstrated a single band corresponding to dimer with approximately 120 kDa in size (left panel in Figure 5B). The presence or absence of BSA appeared to have no effect on dimer formation in Sups. Right panel in Figure 5B demonstrates a native PAGE gel containing samples run without a reducing buffer. Sups from cultures grown in the presence of BSA had two bands of CLO, indicating different charged forms of the toxin. Previous research has identified the oxidized form of CLO migrates faster than its reduced form (Cowell et al., 1976). These results suggested BSA did not prevent CLO dimerization but instead protected an active conformation of CLO from being oxidized. BC2 strain demonstrated similar behavior; however, the amount of produced CLO was substantially reduced (data not shown).

Bacillus species can undergo anaerobic fermentation that results in the generation and secretion of acidic metabolic products. Sups from BC1 grown in microaerobic conditions were titrated to pH 5.4, supplemented with SA and incubated with HSAECs. The pH value of 5.4 in Sups indicate SA which has pK_as of 4.2 and 5.6 could be a potential contributor in its partially protonated state. The concentration of SA in Sups was found to be 1.4 ± 0.1 mM (mean ± SD) regardless of BSA supplementation. To determine if succinate could potentiate B. cereus virulence, cells were briefly exposed to Sups which were then removed and replaced with medium containing different concentrations of SA. The supplementation with SA significantly enhanced the cytotoxicity of Sups (Figure 6).

Previous research demonstrated that microaerobic cultures of B. anthracis generate NO and its derivatives such as NO₂ and peroxynitrite. These substances enhance bacterial virulence through induction of mitochondrial stress in the exposed host cells (St John et al., 2013). The release of reactive species is detectable as a capacity of bacterial cultures to oxidize the dye AR as a test substrate (St John et al., 2013). Since B. cereus was also reported to produce peroxynitrite in microaerobic cultures and conditions of mild acid stress (Mols et al., 2010; Mols and Abee, 2011) we assessed the level of B. cereus-generated reactive substances using the AR test.

AR is a colorless and non-fluorescent derivative of dihydroresorufin that generates a red derivative of resorufin when oxidized by hydrogen peroxide or peroxynitrite in the reaction catalyzed by HRP. Aliquots of bacterial cultures grown in the presence of AR and HRP were collected every hour and the color intensities of Sups were measured. A medium containing both AR and HRP reagents served as a control. Figure 7 illustrates BC1 and BC2 microaerobic cultures generated significant amounts of oxidized AR. Upon a prolonged incubation with BC2 the intensity of released red color declined. This effect is typically associated with consumption of available oxygen (Towne et al., 2004). In addition, resorufin can be converted into a colorless product by peroxidase.

To further elucidate the nature of the oxidizing substances, HSAECs were incubated with Sups supplemented with ascorbic acid (AA), a commonly used biologically relevant antioxidant (Bendich et al., 1986). For example, AA is effective in providing complete scavenging of NO-derived radicals from solution (Balavoine and Geletii, 1999). On the other hand, intracellular AA was reported to work as a pro-oxidant enhancing susceptibility of cells to the toxic effect of peroxynitrite in mitochondria (Guidarelli et al., 2001). We added AA to microaerobic Sups for 2 h at 0, 0.5, 1, and 2 mM concentrations and the viability of HSAECs was tested after 20-min exposure. Sups supplemented with ascorbate were considerably more toxic, and the BC1 Sup demonstrated the highest effect (Figures 8A,B). Control cells exposed to the AA-supplemented culture medium were non-toxic in spite of some acidification of the Sups by AA (Figure 8 legend). The BC1 toxicity-enhancing effect of AA took place only in the presence of BSA (Figure 8C) in agreement with the previous studies implicating BSA in the capture and stabilization of reactive substances in Sups (St John et al., 2013).