Influence of Anaerobiosis and Low Temperature on Bacillus cereus Growth, Metabolism, and Membrane Properties

Abstract

The impact of simultaneous anaerobiosis and low temperature on growth parameters, metabolism, and membrane properties of Bacillus cereus ATCC 14579 was studied. No growth was observed under anaerobiosis at 12°C. In bioreactors, growth rates and biomass production were drastically reduced by simultaneous anaerobiosis and low temperature (15°C). The two conditions had a synergistic effect on biomass reduction. In anaerobic cultures, fermentative metabolism was modified by low temperature, with a marked reduction in ethanol production leading to a lower ability to produce NAD⁺. Anaerobiosis reduced unsaturated fatty acids at both low optimal temperatures. In addition, simultaneous anaerobiosis and low temperatures markedly reduced levels of branched-chain fatty acids compared to all other conditions (accounting for 33% of total fatty acids against more 71% for low-temperature aerobiosis, optimal-temperature aerobiosis, and optimal-temperature anaerobiosis). This corresponded to high-melting-temperature lipids and to low-fluidity membranes, as indicated by differential scanning calorimetry, 1,6-diphenyl-1,3,5-hexatriene (DPH) fluorescence anisotropy, and infrared spectroscopy. This is in contrast to requirements for cold adaptation. A link between modification in the synthesis of metabolites of fermentative metabolism and the reduction of branched-chain fatty acids at low temperature under anaerobiosis, through a modification of the oxidizing capacity, is assumed. This link may partly explain the impact of low temperature and anaerobiosis on membrane properties and growth performance.

INTRODUCTION

Consumer demand for refrigerated processed foods of extended durability (REPFEDs) has increased since the nineties (14), and 10⁸ to 10⁹ units were sold across Europe in 2004 (35). These products aim to improve nutritional and sensorial quality, leading to the development of mild heat processes which permit the survival of pathogenic spore-forming bacteria, such as Bacillus cereus. The Gram-positive spore-forming, facultative anaerobic bacteria B. cereus has the ability to grow at temperatures found in the chill chain, i.e., between 4 to 5 and 10°C depending on the strain (18). B. cereus can cause diarrheal and emetic food-borne poisonings (42). The risk for consumers depends on the ability of B. cereus to multiply during the shelf-life of REPFEDs. The food environment of REPFEDs in storage includes modified atmosphere or vacuum packaging, creating an anaerobic environment with refrigeration.

Cold adaption requires several changes in the bacterial cell (20, 34, 36, 37, 43). In particular, membrane modification plays a critical role. A low temperature rigidifies the cell membranes. To maintain exchanges with the environment, a bacterium has to modify its fatty acid composition to keep fluidity (19). Phospholipids are the main components of membranes, and their melting point varies with fatty acid composition. Certain modifications are known to decrease the melting point of fatty acids and to improve bacterial adaptation to low temperatures: chain length reduction, the synthesis of branched-chain fatty acids (BCFA), and, notably, anteiso- and unsaturated fatty acids (UFA) (24, 39). The aerobic growth of B. cereus was associated with a higher ratio of unsaturated membrane fatty acids at low temperature to that at optimal growth temperature (8, 19).

In the absence of an electron acceptor such as oxygen, the respiratory chain is unable to produce energy and stabilize the redox state (characterized by NADH/NAD⁺ balance), meaning the bacteria have to produce ATP and oxidize reducing equivalents via fermentative pathways (13). The anaerobic growth of B. cereus ATCC 14579 induces genes of mixed acid fermentation (46). The fermentative pathway of B. cereus is characterized by the production of acetate, ethanol, formate, lactate, and succinate (13, 33, 38).

Two recent studies described the growth of Bacillus weihenstephanensis, a psychrotrophic member of the B. cereus group, at low temperatures and with reduced oxygen concentrations (40, 44). In a meat-derived medium, B. weihenstephanensis growth at 8°C was slower under an atmosphere of nitrogen and carbon dioxide than under 2% oxygen with carbon dioxide. In a potato purée packaged under vacuum, growth at 7°C was delayed and reduced compared to growth in air. These studies indicated that removing oxygen reduced the growth of B. weihenstephanensis at low temperature, but they did not give clues to the possible impact of the absence of oxygen on cold adaptation mechanisms. However, the anaerobic growth of B. cereus permitted a higher resistance to heat and acid stress than aerobic growth (31).

For B. cereus, the effectiveness of fermentative metabolism at cold temperatures and the ability to adapt to cold temperatures in strict anaerobiosis remain unknown. The aim of this study was to investigate B. cereus adaptation to a combination of low temperature and the absence of oxygen, focusing on membrane fatty acids, membrane properties, and fermentative metabolism modifications.

MATERIALS AND METHODS

Strain and media.

The type strain Bacillus cereus ATCC 14579 was used (16). This strain belongs to the phylogenetic group IV defined by Guinebretière et al. (18), and its genome was sequenced in 2003 (22). All of the experiments were conducted in chemically defined medium (MOD; 17) supplemented with 1 g/liter of K₂HPO₄, 6 g/liter of (NH₄)₂SO₄, 0.04 g/liter of MgSO₄ (12) or in Luria broth (LB) adjusted to pH 7.2 with KOH. The pH of MOD was adjusted after autoclave sterilization to 7.2 using sterile prereduced KOH, filter sterilized, and deoxygenated, and 1 M glucose solution was added to reach a final concentration of 30 mM. MOD buffered at pH 7.4 with 0.08 M K₂HPO₄ and 0.02 M KH₂PO₄ was prepared for certain experiments. For uncontrolled batch cultures, oxygen was excluded from the media by heating under a flow of N₂ passed through a Hungate column to eliminate any trace of oxygen (21). Hungate tubes were filled under nitrogen flow.

Uncontrolled batch cultures.

Uncontrolled batch cultures were conducted in Hungate tubes filled with 12 ml MOD or LB. The redox indicator resazurin was added at 0.2 mg liter⁻¹ to visualize the redox state of the media (38). The discoloration of the indicator indicates an oxidoreduction potential of less than −110 mV (23). Aerobic cultures were obtained using air-permeable caps. Tubes were incubated in a slanted position at 12, 15, 20, 30, and 37°C with shaking at 200 rpm. pH was measured at the initiation of the stationary phase (S417 pH meter; Hanna Instruments). Three independent cultures were performed for each condition. Inocula were made of overnight anaerobic cultures in MOD with 30 mM glucose at 37°C. The use of buffered MOD for uncontrolled batch cultures made it possible to keep the culture pH above 6.25 during the whole growth period under all conditions tested.

Batch cultures in pH- and gas-controlled environments.

Two bioreactors of 2 liters each were used for controlled batch cultures: an Inceltech Discovery 100 (Toulouse, France) at 37°C and an Adagio Pierre Guérin (Mauze-sur-le-Mignon, France) at 15°C. Temperature was controlled with a cryostat. Cultures were performed in 1.2 liters MOD, inoculated with cells harvested at the end of the exponential phase of uncontrolled aerobic cultures at the same temperature as that of the regulated cultures, and incubated overnight at 37°C and then for 4 days at 15°C. Cells from precultures were washed twice by the centrifugation and suspension of the pellets in saline water (9 g liter⁻¹ NaCl). Cultures were inoculated at an A₆₀₀ of 0.02. Bioreactors were kept under airflow to obtain aerobiosis or under nitrogen flow to obtain anaerobiosis. Gases were filtered at 0.2 μm. The pH was automatically adjusted to 7.2 with a sterile 3 M KOH or 6 M HCl solution. Shaking was kept at a minimum of 500 rpm and adjusted to maintain the partial pressure of oxygen (pO₂) at 100% (air). pO₂ was measured with a Mettler Toledo polarographic oxygen electrode calibrated on a medium flushed with air (100% value) or pure N₂ (0% value). Three independent cultures were performed for each condition.

Growth parameters.

Growth was monitored using a spectrophotometer (Helios Epsilon; Thermo Scientific, Rockford, IL). To avoid contaminating uncontrolled cultures with oxygen, the tubes were not opened before measuring A₆₀₀ values, meaning it was not possible to dilute certain cultures which grew marginally across the range of the linearity of the spectrophotometer. For batch-regulated cultures, samples were diluted with MOD to keep the A₆₀₀ below 0.3. The A₆₀₀ values were plotted over time and fitted with the modified Gompertz equation (51) to calculate the maximal specific growth rate (μ_max) and maximal absorbance. To measure the final biomass of the controlled batch cultures, 25 or 40 ml (depending on the growth conditions) was filtered with a membrane filter (pore size, 0.2 μm) and dried at 37°C until the weight remained constant.

Determination of glucose and metabolite concentrations.

In controlled cultures, a sample was taken at the beginning of the stationary phase and after the third (or fourth) stable A₆₀₀ measurement. The aliquots were centrifuged at 7,000 × g for 5 min at 4°C, and the supernatants were stored at −20°C until analysis. Glucose, acetate, and lactate concentrations were measured using Biosentec enzymatic kits (Toulouse, France), and ethanol, succinate, and formate were measured using a Biopharm enzymatic kit (Darmstadt, Germany) by following the manufacturers' instructions. Three technical replicates were done for each biological sample.

The moles of NADH and NAD⁺ formed by the central metabolism under fermentative growth conditions in controlled batch cultures were calculated according to the metabolites produced using the previous description of the central metabolism for B. cereus (50) and the following equation:

$${\mathrm{NAD}}^{+}\hspace{0.17em}\mathrm{formed}\mathrm{\hspace{0.17em}}\mathrm{moles}=\left(\mathrm{moles}\mathrm{\hspace{0.17em}}\mathrm{of}\mathrm{\hspace{0.17em}}\mathrm{glucose}\mathrm{\hspace{0.17em}}\mathrm{consumed}+\mathrm{moles}\mathrm{\hspace{0.17em}}\mathrm{of}\mathrm{\hspace{0.17em}}\mathrm{ethanol}\mathrm{\hspace{0.17em}}\mathrm{formed}+\mathrm{moles}\mathrm{\hspace{0.17em}}\mathrm{of}\mathrm{\hspace{0.17em}}\mathrm{acetate}\mathrm{\hspace{0.17em}}\mathrm{formed}-\mathrm{moles}\mathrm{\hspace{0.17em}}\mathrm{of}\mathrm{\hspace{0.17em}}\mathrm{formate}\mathrm{\hspace{0.17em}}\mathrm{formed}\right)/\left(2\times \mathrm{moles}\mathrm{\hspace{0.17em}}\mathrm{of}\mathrm{\hspace{0.17em}}\mathrm{ethanol}\mathrm{\hspace{0.17em}}\mathrm{formed}+2\times \mathrm{moles}\mathrm{\hspace{0.17em}}\mathrm{of}\mathrm{\hspace{0.17em}}\mathrm{succinate}\mathrm{\hspace{0.17em}}\mathrm{formed}+\mathrm{moles}\mathrm{\hspace{0.17em}}\mathrm{of}\mathrm{\hspace{0.17em}}\mathrm{lactate}\mathrm{\hspace{0.17em}}\mathrm{formed}\right)$$ (1)

Fatty acid composition.

Volumes of 25 ml (or 100 ml for cultures at 15°C under nitrogen) were collected at the beginning of the stationary phase. Samples were washed with saline. Lipid fatty acids obtained from 40 mg of cells (fresh weight) were transesterified via the ester link method (41). The reaction consisted of alkaline methanolysis breaking the ester link in the lipid and producing fatty acid methyl esters by reaction with 5 ml of 0.2 M KOH in methanol at 37°C for 1 h. One ml of 1 M acetic acid then was added to lower the pH to <7.0, and we checked the pH with pH test strips. Fatty acid methyl esters then were extracted by adding 3 ml of hexane. Supernatant (apolar phase) was transferred into clean tubes and concentrated by evaporation at room temperature under a continuous flow of nitrogen to obtain approximately 20 μl of extract. Extracts were injected into a gas chromatography-mass spectrometry (GC-MS) system (Shimadzu QP 2010-01) and processed by GC-MS real-time analysis. A 100-μl volume of a solution of a methylated C19 fatty acid (Sigma) at 0.1 g liter⁻¹ was mixed into samples before treatment and used as a standard. The equivalent chain length (ECL) method and/or derivatization method to detect the position of unsaturation (4,4-dimethyloxazoline [DMOX]) and the position of ramification (picolinyl) were used to identify fatty acids as described by Brillard et al. (8). The composition of the fatty acids from uncontrolled cultures in buffered MOD was similar to that of the controlled cultures grown with the same temperature and atmosphere conditions (data not shown).

DSC.

Lipids used for differential scanning calorimetry (DSC) were obtained by chloroform-methanol (1:2) extraction (7) from controlled cultures in MOD sampled at the beginning of the stationary phase. Samples represented 25 ml of culture (100 ml for cultures at 15°C under nitrogen). After centrifugation at 7,000 × g for 10 min at 10°C, the supernatants were discarded and the pellets were suspended in 3.75 ml chloroform-methanol (1:2), 1.25 ml chloroform, 1.25 ml distilled water and vigorously vortexed after each solvent addition. After centrifugation at 2,000 × g for 5 min at 10°C, the organic phase was harvested and evaporated under a continuous flow of nitrogen. Before storage at −20°C, tubes were filled with nitrogen to prevent oxidation. Five mg ml⁻¹ of the dried lipids was resuspended in 1 mM EDTA, 10 mM HEPES, 50 mM KCl at pH 7.4. Melting points were estimated with a DSC III Setaram microcalorimeter (Caluire, France). Data were collected in the range of −3 to 25°C at a scan rate of 0.5°C min⁻¹. The endothermic peak observed during heating was used to estimate the melting point.

FT-IR.

Fourier transform infrared spectroscopy (FT-IR) was used to evaluate the membrane properties of B. cereus cells grown under the different test conditions by measuring the symmetric stretching energy of CH₂ (2, 3, 27). Cells were collected from uncontrolled cultures in buffered MOD at the start of the stationary phase. Infrared spectra were obtained with a Tensor 27 spectrophotometer (Bruker Optics, Wissenbourg, France) equipped with an attenuated total reflectance (ATR) ZnSe crystal and connected to OPUS software (San Francisco, CA). Spectra were obtained from wavenumbers in the range of 800 to 4,000 cm⁻¹ with 32 scans, a resolution of 4 cm⁻¹, and an interval of 2 cm⁻¹. Cells obtained from samples (volumes between 12 and 36 ml depending on the A₆₀₀ values) were rinsed twice and suspended in 0.1 M phosphate buffer at pH 6.25 and then deposited on the ZnSe crystal. Spectra were read at 30 min postdeposition at room temperature (23°C). This time and temperature allowed a continuous film to form at the crystal surface. Spectra were processed with OPUS software by smoothing and derivatization (second derivative using the Savitsky-Golay algorithm).

Anisotropy fluorescence.

A previously described method using 1,6-diphenyl-1,3,5-hexatriene (DPH) as a probe was used to characterize cell membrane fluidity (5, 10, 45). The hydrophobic properties of DPH allow the molecule to be inserted into the lipid membrane of fresh cells. DPH anisotropy is associated with the membrane fluidity of the cells. Fluorescence was measured on an ISS Koala 2 (Champaign, IL) spectrofluorometer equipped with a Xenon lamp/monochromator excitation source and monochromator/photon-counting electronics/PMT (photon multiplicator tube) for emission detection. The excitation wavelength was set at 350 nm and the emission wavelength at 450 nm. Polarizers on excitation and emission sides were used to measure anisotropy according to the following equation:

$$r=\left(I\parallel -GI\perp \right)/\left(I\parallel +2GI\perp \right)$$ (2)

where I‖ and I⊥ are fluorescence intensity parallel and perpendicular, respectively, to the vertically polarized excitation beam. The G factor corrects for the polarization-dependent properties of the detection system and was calculated by the following equation:

$$G=\mathrm{IHH}/\mathrm{IHV}$$ (3)

where IHH and IHV are the fluorescence intensities measured under horizontal excitation and horizontal emission and vertical emission, respectively. Anisotropy (r) was calculated automatically by ISS Vinci software. All samples were placed in a temperature-regulated chamber and stirred at constant speed during measurements. At the start of the stationary phase in buffered MOD uncontrolled batches, cells were harvested and rinsed twice with morpholineethanesulfonic acid (MES) buffer at pH 6.2, and 1 ml of the suspension (adjusted to an A₆₀₀ of 1.2) was mixed with the probe solution at a final DPH concentration of 6 μM. Anisotropy was measured on fresh cells held at 5, 18, and 39°C.

Statistical analysis.

Effects of temperature and gaseous atmosphere on A_600max, μ_max, pH, fluorescence anisotropy, and fatty acid composition were tested by analysis of variance (ANOVA). Means were compared with the Tukey's honest significant difference (HSD) test (Systat 9.0; SPSS, Chicago, IL). Growth parameters, end product metabolites, and the melting point of the controlled culture were tested by a Student's t test (Systat 9.0; SPSS, Chicago, IL). For all tests, the null hypothesis was rejected at a P value of <0.05. The NADH/NAD⁺ ratio between anaerobic cultures grown at 15 and 37°C and its 95% confidence interval were calculated by linear regression in Microsoft Excel 2010.

RESULTS

Interaction of low temperature and anaerobiosis in uncontrolled growth conditions: growth parameters.

B. cereus was grown at 37, 30, 20, 15, and 12°C in MOD and LB under aerobic or anaerobic growth conditions. No growth was observed in anaerobiosis at 12°C in either medium, whereas growth was observed at all temperatures under air (Fig. 1). At all temperatures permitting growth, maximum A₆₀₀ (A_600max) values were at least 4-fold higher under aerobiosis than anaerobiosis (Fig. 1). In anaerobiosis in both MOD and LB, the A_600max decreased significantly (4-fold) from that at 30°C (in MOD, P < 0.01; in LB, P < 0.05) to that at 15°C, which was the lowest temperature at which growth was observed.

Fig 1.

Maximum population (A_600max) of B. cereus grown at different temperatures under air (white bars) and under nitrogen (gray bars) (N₂) in uncontrolled batch cultures. (A) Growth in LB medium. (B) Growth in MOD medium. Value are the means from three biological replicates, and bars represent standard errors. Different letters show mean values that are significantly different according to Tukey's test (P < 0.05) for A_600max. NG, no growth observed during 527 h.

Interaction of low temperature and anaerobiosis in controlled growth conditions. (i) Growth parameters.

Controlled batch cultures of B. cereus ATCC 14579 were performed in MOD at pH 7.2 and at temperatures of 15 and 37°C. The temperature 15°C was chosen as the cold condition because it was the lowest temperature that enabled significant growth in uncontrolled cultures under both air and nitrogen. MOD was used because it is a chemically defined medium and gave globally the same results as LB in uncontrolled batch cultures.

Temperature had a strong impact on μ_max (Table 1), leading to a 15-fold drop in μ_max between 15 and 37°C for both aerobic and anaerobic cultures (P < 0.005). Growth rate was doubled in aerobiosis compared to that in anaerobiosis at both temperatures. There was no synergistic effect of low temperature and anaerobiosis on growth rate. Glucose consumption was complete under all conditions except at 15°C under nitrogen (89% of glucose consumed). Under air, the biomass and A_600max obtained at the two tested temperatures were not significantly different (P > 0.05): biomass at 37 and 15°C was 3.26 and 3.91 g liter⁻¹, respectively (Table 1). At both 37 and 15°C, the biomasses obtained under nitrogen were markedly lower than that under air (P < 0.01). Differences between the amount of biomass in nitrogen and air cultures were amplified at 15°C compared to those at 37°C. In air, a low incubation temperature had no impact on biomass production. On the contrary, in anaerobiosis a low incubation temperature reduced biomass production (biomass differences were considered significant at P < 0.001). This indicates a synergistic effect of anaerobiosis and growth at 15°C on biomass production (Fig. 2 and Table 1).

Table 1.

Growth parameters of B. cereus grown in controlled batches in MOD under different conditions^a

Temp and atmosphere	Growth parameter
	Biomass (g liter−1)	μmax (h−1)	Glucose consumption (%)	Biomass yield (g mmol−1 of glucose)
15°C
N2	0.48 ± 0.027	0.06 ± 0.007	89 ± 13.9	0.018 ± 0.002
Air	3.91 ± 0.47	0.13 ± 0.004	100	0.137 ± 0.004
37°C
N2	2.09 ± 0.05	1 ± 0.094	100	0.036 ± 0.001
Air	3.26 ± 0.16	1.96 ± 0.157	100	0.111 ± 0.01

^aEach value is the mean of results obtained from three independent cultures ± standard deviations. μ_max values were estimated using a modified Gompertz model (51).

Fig 2.

Growth of B. cereus in controlled batch culture at 37°C (A) and 15°C (B), under air (▲ ■ ●) and under nitrogen (△ □ ○), in MOD at pH 7.2. Three independent growth curves are presented for each condition.

(ii) End product metabolites.

The end products of mixed acid fermentation (acetate, ethanol, formate, lactate, and succinate) and glucose consumption (38) were assayed as indicators of the metabolic pathways used by B. cereus ATCC 14579 during growth at low temperature and/or anaerobiosis. Acetate was the major metabolite produced under air at 37°C (0.8 mol mol⁻¹ of glucose consumed), followed by succinate and formate (0.06 and 0.01 mol mol⁻¹ of glucose consumed, respectively). Under air, less acetate was produced at 15°C (0.03 mol mol⁻¹ of glucose consumed) than at 37°C. At 37°C under nitrogen, the main products, in decreasing molar concentrations per mole of consumed glucose, were lactate, formate, acetate, ethanol, and succinate (Table 2), which is consistent with previous observations (26, 38). Anaerobic cultures of B. cereus at 15°C did not produce significantly less lactate (P > 0.05) than at 37°C but did produce significantly less ethanol (P < 0.01). Acetate was the only metabolite produced in larger amounts at 15°C under nitrogen than at 37°C, but the difference was not significant (P > 0.1) (Table. 2). The NADH/NAD⁺ ratio was dependent on the fermentative pathway (Fig. 3) and was 0.98 under the 15°C N₂ condition and 0.54 under the 37°C N₂ condition (significant difference at P < 0.05; Table 2). This suggests an imbalance in redox state at low temperature. The production under nitrogen of metabolites from the acetyl-CoA node (ethanol plus acetate) was the same at both 37 and 15°C (no significant difference in mol mol⁻¹ of glucose consumed; P > 0.75; Table 2). Production in terms of moles of ethanol per mole of glucose consumed represented 35% of acetyl-coenzyme A (CoA) at 37°C versus 6% at 15°C. As the production of 1 mol ethanol from acetyl-CoA oxidizes 2 mol NADH to NAD⁺ (Fig. 3), this means that the level of NADH oxidation during anaerobic growth at 15°C was low. In addition, under nitrogen, the combined acetate and ethanol concentration was higher than the formate concentration at 15°C (P < 0.05) but was lower than that at 37°C. More pyruvate was available at 15°C than at 37°C for use by the pyruvate dehydrogenase enzyme, which consumes more NAD⁺ than the pyruvate formate lyase (Fig. 3 and Table 2).

Table 2.

Production of fermentative end product metabolites for B. cereus grown in controlled batches in MOD under different conditions^a

Temp and atmosphere	Metabolite production (mol mol−1 of glucose consumed)						NADH/NAD+ ratio
	Acetate	Ethanol	Formate	Lactate	Succinate	Acetate, ethanol, and acetyl-CoA
15°C
N2	0.522 ± 0.085	0.033 ± 0.007	0.388 ± 0.068	1.114 ± 0.144	0.007 ± 0.002	0.556	0.98
Air	0.032 ± 0.060	0.000	0.058 ± 0.032	0.000	0.026 ± 0.004	0.032	ND
37°C
N2	0.350 ± 0.029	0.188 ± 0.022	0.605 ± 0.051	1.321 ± 0.063	0.021 ± 0.003	0.538	0.54
Air	0.792 ± 0.119	0.000	0.014 ± 0.012	0.000	0.061 ± 0.017	0.792	ND

^aValues are the mean results obtained from three independent cultures ± standard deviations. The NADH/NAD⁺ ratio was calculated with equation 1.

Fig 3.

Fermentative metabolism of B. cereus; data are from references 30 and 50.

(iii) Fatty acid composition and lipid melting points.

Under air at both temperatures and under nitrogen at 37°C only, branched-chain fatty acids (BCFA) were the main B. cereus membrane fatty acids. In particular, iso-15 and iso-17 prevailed in nitrogen, and iso-15, iso-17, and iso-17:1 (5) prevailed in air (Tables 3 and 4). In contrast, at 15°C under nitrogen, iso-15 and iso-17 or iso-17:1 (5) were nearly absent.

Table 3.

Fatty acid composition in percentages of the total amount of fatty acids extracted from B. cereus grown under different conditions^a

Temp and atmosphere	Fatty acid composition (%)
	i12	n12	i13	a13	i14	n14	i15	a15	n15	i16	i16:1(5)	i16:1(10)	n16	C16:1(6)	C16:1(10)	i17	a17	i17:1(5)	i17:1(10)	n17	a17:1(10)	n18	C18:1(9)	C18:2
15°C
N2	1.50 ± 0.196 a	0.40 ± 0.227 ab	12.57 ± 2.306 a	6.87 ± 1.547 a	2.40 ± 0.635 bc	3.33 ± 1.052 a	4.07 ± 0.487 c	2.65 ± 0.533 b	2.14 ± 0.493 a	0.34 ± 0.035 c	0.07 ± 0.034 c	0.05 ± 0.013 c	49.39 ± 6.371 a	2.56 ± 1.416 a	1.83 ± 1.059 b	1.54 ± 0.301 c	0.22 ± 0.015 c	0.92 ± 0.527 c	0.04 ± 0.010 c	0.45 ± 0.088 a	0.07 ± 0.027 b	4.85 ± 0.425 a	0.73 ± 0.343 a	0.12 ± 0.030 a
Air	0.12 ± 0.011 c	0.08 ± 0.006 ab	6.53 ± 0.268 b	0.60 ± 0.148 b	4.53 ± 0.168 a	1.65 ± 0.023 b	21.61 ± 0.081 b	5.06 ± 1.416 a	0.40 ± 0.017 c	2.71 ± 0.069 b	2.18 ± 0.187 a	0.94 ± 0.087 b	4.50 ± 0.825 c	1.89 ± 0.234 ab	7.57 ± 0.057 a	7.16 ± 0.455 b	0.59 ± 0.292 c	24.54 ± 2.573 a	1.89 ± 0.183 b	0.18 ± 0.096 b	0.29 ± 0.044 b	1.34 ± 0.513 b	3.49 ± 2.440 a	0.04 ± 0.017 b
37°C
N2	0.50 ± 0.117 b	0.35 ± 0.084 ab	10.41 ± 0.656 a	1.65 ± 0.367 b	2.58 ± 0.298 b	2.55 ± 0.208 ab	28.90 ± 3.710 a	5.77 ± 0.278 a	1.26 ± 0.301 b	3.66 ± 0.300 a	0.04 ± 0.029 c	0.10 ± 0.057 c	20.51 ± 2.988 b	0.06 ± 0.029 b	2.04 ± 0.400 b	14.93 ± 1.527 a	1.96 ± 0.227 b	0.27 ± 0.055 c	0.19 ± 0.032 c	0.09 ± 0.002 b	0.17 ± 0.179 b	1.16 ± 0.323 b	0.32 ± 0.066 a	0.08 ± 0.010 ab
Air	0.14 ± 0.003 c	0.45 ± 0.029 a	4.04 ± 0.218 b	0.70 ± 0.045 b	1.54 ± 0.070 c	2.45 ± 0.218 ab	28.88 ± 1.153 a	6.56 ± 0.115 a	0.34 ± 0.067 c	4.06 ± 0.467 a	0.29 ± 0.004 b	1.66 ± 0.235 a	4.16 ± 0.445 c	0.20 ± 0.080 b	8.31 ± 0.155 a	14.04 ± 0.676 a	3.02 ± 0.371 a	8.08 ± 0.681 b	6.17 ± 1.234 a	0.12 ± 0.018 b	3.59 ± 0.595 a	0.57 ± 0.109 b	0.55 ± 0.114 a	0.09 ± 0.018 ab

^aEach value is the means with standard errors of percentages obtained from three independent cultures. For each fatty acid, values with the same letter were not significantly different (P > 0.05) according to Tukey's HSD test. The nomenclature of fatty acids is the following: C_x, straight saturated chain where x is the number of carbons; C_x:y(n), unsaturated straight chain, where x is the number of carbons, y is the number of unsaturation (number of double bonds) in the chain, and n is the place of the unsaturation in the chain (starting from the carboxylic function); ax and ix, branched-chain fatty acids, respectively anteiso- and iso-fatty acid with x, which is the number of carbons (including the ramification). Branched-chain fatty acids can be saturated or unsaturated and use the same nomenclature as that for straight-chain fatty acids.

Table 4.

Groups and length of fatty acids as percentages of the total amount extracted from B. cereus anaerobic and aerobic controlled cultures grown at 15 and 37°C^a

Temp and atmosphere	Fatty acid length (%)										Anteiso/iso ratio
	C11	C12	C13	C14	C15	C16	C17	C18	Unsaturated chain	Branched chain
15°C
N2	1.50 ± 0.20	19.84 ± 3.63	2.40 ± 0.63	10.05 ± 0.70	2.61 ± 0.42	56.57 ± 3.24	0.09 ± 0.00	5.70 ± 0.73	6.39 ± 2.62	33.31 ± 6.33	0.42 ± 0.01
Air	0.14 ± 0.03	6.92 ± 0.45	5.00 ± 0.91	27.94 ± 1.93	7.12 ± 1.46	48.96 ± 1.41	0.09 ± 0.01	3.60 ± 2.67	43.06 ± 0.36	79.66 ± 3.18	0.09 ± 0.02
37°C
N2	0.50 ± 0.12	12.41 ± 1.10	2.58 ± 0.30	37.22 ± 3.27	5.06 ± 0.34	40.13 ± 1.40	0.11 ± 0.02	1.56 ± 0.30	3.26 ± 0.12	71.10 ± 4.37	0.16 ± 0.02
Air	0.14 ± 0.00	5.22 ± 0.18	1.54 ± 0.06	37.92 ± 1.05	6.34 ± 0.29	47.52 ± 1.13	0.12 ± 0.02	1.22 ± 0.23	28.92 ± 2.82	82.81 ± 0.38	0.20 ± 0.01

^aEach value is the mean of results obtained from three independent cultures ± standard deviations.

The lower incubation temperature of aerobic culture significantly increased the proportion of unsaturated fatty acids (UFA) from 29% at 37°C to 43% at 15°C (P < 0.001) (Table 4), as demonstrated by the increase of iso-17:1 (5) (P < 0.001) (Table 3). In the absence of oxygen, the proportion of UFA was not significantly influenced by the lower temperature (P > 0.25) and was low compared to the proportion of UFA in aerobic culture at both temperatures, i.e., 3% at 37°C and 6% at 15°C under nitrogen versus 29% at 37°C and 43% at 15°C under air (P < 0.001) (Table 4). Both low temperature and anaerobiosis increased the proportions of short-chain fatty acids (P < 0.001). For instance, C12 represented 12% at 37°C and 20% at 15°C under nitrogen and 5% (37°C) and 7% (15°C) under air (Table 4). Under aerobic culture, the proportion of BCFA did not differ between 15 and 37°C (P > 0.75). Compared to aerobic cultures, anaerobic cultures at 37°C contained slightly less BCFA (P < 0.05): 71 and 83%, respectively. In contrast, at 15°C, anaerobic cultures contained less than half the BCFA of aerobic cultures (P < 0.001): 33 and 80%, respectively (Table 4).

The properties of the B. cereus lipids were assessed by DSC. This method can evaluate the melting point of a mixture of lipids extracted from cell membranes. Results show that the lipids synthesized by B. cereus under air had lower melting points than the lipids synthesized under nitrogen at both growth temperatures (P < 0.01). In particular, lipids produced at 15°C under nitrogen had the highest melting point and therefore were the least fluid (Table 5).

Table 5.

Melting point (T_m) of lipids extracted from B. cereus in controlled batches and measured by DSC^a

Atmosphere	Tm (°C) by growth temp and batch no.
	15°C		37°C
	1	2	1	2
N2	3.68	4.07	2.12	0.74
Air	1.23	−0.57	−0.35	−2.1

^aFor each condition, results obtained with cells from two independent cultures are presented.

Membrane properties. (i) FT-IR spectroscopy.

The absorption profiles on FT-IR can show the phase transition of membranes as a function of temperature (27, 28) and thus give an indication of the fluidity-related membrane properties. FT-IR was used to compare the properties of B. cereus membranes as a function of growth conditions (2). We used the second derivative peak of the spectra at around 2,850 cm⁻¹ to visualize the impact of growth conditions on the vibrational energy of CH₂ symmetric stretching (Fig. 4). At 37°C, the aerobic and anaerobic fermentative cultures had second derivative peaks at the same wavenumber. Low-temperature adaptation under aerobic conditions yielded a higher wavenumber, indicating increased membrane fluidity. In contrast, under anaerobic conditions at 15°C, the B. cereus cultures had the lowest wavenumber of second derivative peaks, indicating less-fluid membranes (Fig. 4). This indicated a strong impact of anaerobiosis on membrane properties during growth at low temperature, leading to less-fluid membranes.

Fig 4.

Second derivative of spectra from bacteria analyzed by infrared spectroscopy and calculated with the Savitsky-Golay algorithm (OPUS software). Symbols: □, cells grow at 15°C under N₂; ○, cells grown at 15°C under air; ■, cells grown at 37°C under N₂; ●, cells grown at 37°C under air. The spectra were focused around 2,850 cm⁻¹ to visualize the impact of growth conditions on the stretching mode of CH₂. Curves are the means from three biological replicates, and peaks were normalized at the same intensity for the three replicates. No differences were observed in wavenumber values between replicates. Spectra were acquired at 23°C.

(ii) Fluorescence anisotropy.

Fluorescence anisotropy with 1,6-diphenyl-1,3,5-hexatriene (DPH), which is used to represent global membrane fluidity (5, 10, 45), was measured at test temperatures of 5, 18, and 39°C on cells from aerobic and anaerobic cultures grown at 15 and 37°C (Fig. 5). At 39°C, anisotropy values were not significantly different between cultures (P > 0.1). As expected, the membrane fluidity of all cell cultures decreased (increasing anisotropy) with decreasing test temperature. However, cells grown at 15°C under air kept the highest fluidity. At a test temperature of 18°C, membrane fluidity was similar between aerobic and anaerobic cells cultivated at 37°C (P > 0.85) but markedly different between aerobic and anaerobic cells cultivated at 15°C (P < 0.05). Cell membrane fluidity was higher in aerobic cultures grown at 15°C than in aerobic cultures grown at 37°C, which reflects a growth adaptation at low temperature. Cell membrane fluidity was not higher in anaerobic cultures grown at 15°C than in cultures grown at 37°C.

Fig 5.

Fluorescence anisotropy of 1,6-diphenyl-1,3,5-hexatriene (DPH) measured at 5, 18, and 39°C. Symbols: □, cells grown at 15°C under N₂; ○, cells grown at 15°C under air; ■, cells grown at 37°C under N₂; ●, cells grown at 37°C under air. Point are the means from at least 2 independent cultures. Error bars represent extreme values.

DISCUSSION

This work confirms that the combination of anaerobiosis and cold temperature considerably limits the growth of B. cereus sensu lato (40, 44). The absence of oxygen impaired the growth of B. cereus ATCC 14579 at 12°C (Fig. 1), whereas it is able to grow at temperatures as low as 10°C under air (9).

Reducing the temperature from 37 to 15°C caused a 15-fold drop in μ_max under both air and anaerobiosis. The absence of oxygen markedly reduced the production of biomass at all temperatures tested in both controlled and uncontrolled cultures (Fig. 1 and 2). This points to a synergistic effect of anaerobiosis and low temperature on biomass production, which suggests a possible interaction between anaerobiosis and low temperature on B. cereus growth.

Under anaerobiosis, metabolism shifted toward the fermentative pathway, which is less favorable than the respiratory chain in terms of oxidizing capacity, ATP production (30), and biomass (26, 38). We detected a significant modification in fermentative metabolism at low temperature occurring at the acetyl-CoA node. At the optimal temperature, enough ethanol was produced to allow the oxidation of NADH into NAD⁺ to maintain the redox balance, which is consistent with previous studies showing that at 37°C the expression of adhA and adhE was considerably increased in anaerobic conditions compared to that of aerobiosis (13). In contrast, ethanol was almost absent at 15°C, to the profit of acetate, which could favor energy production at the cost of the ability of the fermentative pathway to oxidize NADH. At 15°C in anaerobiosis, acetyl-CoA production was higher than formate production, whereas both compounds were produced in similar amounts at 37°C. This suggests a conversion of pyruvate to acetyl-CoA at 15°C in anaerobiosis by the pyruvate dehydrogenase (Pdh) enzyme rather than by the pyruvate formate lyase (Pfl). The latter produces both acetyl-CoA and formate from pyruvate (Fig. 3 and Table 2). A similar Pdh-driven conversion of pyruvate was observed during oxic and anoxic respirations (11, 50). The loss of balanced oxidizing capacity (equation 1) could be responsible for the incomplete consumption of glucose during growth at 15°C under anaerobiosis, since the synthesis of glyceraldehyde 3-phosphate (3P) in glycolysis requires NAD⁺ (Fig. 3).

UFA synthesis plays a role in microbial cold adaptation (19). Two UFA synthesis pathways have been described to date (24): (i) an aerobic mechanism conserved in most living organisms (1) where UFA are synthesized by the desaturation of existing fatty acids attached to membranes with the action of a fatty acyl-desaturase (29), and (ii) an anaerobic mechanism where UFA are synthesized by a dehydratase and an isomerase, such as FabA and FabB enzymes in Escherichia coli, FabM in Enterococcus faecalis (47), or FabZ in Clostridium acetobutylicum (49). This second mechanism is not active in all organisms and is not the predominant UFA synthesis pathway in bacilli (24). The homolog of fabA in B. cereus ATCC 14579 is repressed in anaerobiosis but not aerobiosis at the optimal temperature (46). Therefore, the absence of oxygen should prevent or strongly reduce UFA synthesis by B. cereus and thus could limit cold adaptation.

This study found significant modifications in fatty acid profiles between different growth conditions. The fatty acid composition of aerobic cultures was consistent with previous observations, with a high proportion of iso-15, iso-17, and n16 (Table 3) (8, 19). Under air, the decrease in temperature led to an increase in the proportion of UFA. The modification of the anteiso-/iso-fatty acid ratio, described as a way for B. subtilis (6) and B. cereus (19) to adapt to cold temperature, was not observed in this work, which is in agreement with a previous one (8). Differences in anteiso-/iso-fatty acid ratios between studies are probably due to differences in media, growth culture conditions, or strains selected. The absence of oxygen considerably decreased UFA synthesis at both 15 and 37°C. In contrast, there was a marked decrease in BCFA in the absence of oxygen, but only at low temperatures. The latter observation represents the most striking feature of fatty acids from B. cereus grown at low temperature without oxygen (Table 4). Kaneda (25) noted that fatty acid biosynthesis requires NAD⁺ to eliminate the amine radical of the precursor amino acid and add the CoA group. The lower production of NAD⁺ by fermentative metabolism at low temperature (Table 2) therefore may reduce BCFA synthesis, consequently affecting membrane properties.

B. cereus cultured anaerobically at low temperature produced less UFA, which is not consistent with an adaptation to cold temperature. In addition, B. cereus cells lost most of the BCFA, the main fatty acids of its membrane, which are known to be specific to bacilli, to influence the phase transition temperature of their membranes and to be necessary for their growth (25, 48). Cold adaptation is usually characterized by membrane lipids with a lowered melting point (32). At 15°C under nitrogen, B. cereus synthesized lipids that had the highest melting point, even higher than that of the lipids produced at 37°C (Table 5), which runs totally counter to cold adaptation theory. Bacteria could use other mechanisms to adapt membrane properties. For instance, proteins are able to modulate membrane fluidity (4, 15). Here, we characterized the membrane properties using in situ measurements of lipid properties by FT-IR at around 2,850 cm⁻¹. This wavenumber corresponds to the vibrational energy of CH₂ symmetric stretching, which is dependent on the conformation of the carbohydrate chain. For a given carbohydrate composition, melting corresponds to an increase in vibrational energy accompanied by a 1.5- to 3-cm⁻¹ increase in wavenumber (27, 28). The peak of absorbance by FT-IR is used to detect a global modification in membrane properties induced by growth conditions, with an increase in wavenumber indicating a more fluid membrane (2, 3).

Here, we combined FT-IR with fluorescence anisotropy, which is widely used to measure membrane fluidity. A higher anisotropy equates to a lower fluidity (5, 10, 45). FT-IR did not find any differences between B. cereus cells grown at 37°C in aerobiosis and anaerobiosis (Fig. 4). These results are consistent with the fluorescence anisotropy measurements, which found that cells grown at 37°C under air and nitrogen had the same anisotropy when measured at 39 and 18°C.

Cells grown at 15°C had different IR spectra, with opposite response patterns after growth under anaerobiosis versus those under aerobiosis. The vibration energy for CH₂-symmetric stretching was higher for cells grown under air than under nitrogen (Fig. 4). This could correspond to a different conformation of the carbohydrate chain, resulting in a more fluid state in aerobiosis than in anaerobiosis. Anaerobic cultures at 15°C gave a clearly different FT-IR response, with presumably the least-fluid lipids, compared to those of cultures grown in the other conditions, which is consistent with the highest melting point of their lipids and their markedly different fatty acid composition. Fluorescence anisotropy confirmed that bacteria produced in aerobiosis at 15°C presented the most fluid membranes between 5 and 18°C, which denotes a cold adaptation and could be explained by adapted fatty acid biosynthesis. It also confirmed that membrane fluidity was much lower for anaerobic cultures at 15°C than for cultures in air (Fig. 5). According to fluorescence anisotropy, membranes of anaerobic cultures at 15°C were not less fluid than those of cultures at 37°C, in contrast to findings with FT-IR (Fig. 4 and Fig. 5). Fluorescence anisotropy measures the mobility of a probe inserted in the membrane and, in contrast to FT-IR, integrates the impact of more membrane components on fluidity than strictly fatty acids. In any case, the membrane fluidity of anaerobic cultures at 15°C is similar to or lower than that of cultures at 37°C, and this reveals a failure to adapt to low temperatures. Results from all methods showed that anaerobiosis had little impact on membrane properties for cells grown at 37°C but a strong impact for cells grown at 15°C. Anaerobiosis not only hindered the modification of fatty acids for cold adaptation (i.e., unsaturation) but also, combined with cold, caused a profound alteration of fatty composition (i.e., BCFA) and consequently of membrane properties.

In conclusion, our work underlined the difficulties for B. cereus to adapt to cold temperatures in the absence of oxygen, notably due to a significant limitation of maximal growth. Cold modifies the fermentative metabolism of B. cereus toward a lower production of oxidized cofactors, which could reduce the production of BCFA. This, combined with the reduction of UFA in the absence of oxygen, presumably leads to membranes with low fluidity at cold temperatures. This may reduce exchanges between the cell and its environment, which would explain the observed growth limitation.