Differential Involvement of the Five RNA Helicases in Adaptation of Bacillus cereus ATCC 14579 to Low Growth Temperatures

Abstract

Bacillus cereus ATCC 14579 possesses five RNA helicase-encoding genes overexpressed under cold growth conditions. Out of the five corresponding mutants, only the ΔcshA, ΔcshB, and ΔcshC strains were cold sensitive. Growth of the ΔcshA strain was also reduced at 30°C but not at 37°C. The cold phenotype was restored with the cshA gene for the ΔcshA strain and partially for the ΔcshB strain but not for the ΔcshC strain, suggesting different functions at low temperature.

Bacillus cereus is a human pathogenic sporulated bacterium which is associated with emetic and diarrheal types of food-borne illnesses (4). B. cereus is widespread in the environment and in a wide range of foods. The growth domains of B. cereus strains range from psychrotrophic to nearly thermophilic and correlate with several phylogenetic clusters (15), which presumably permit B. cereus to colonize many different habitats with different thermal regimes. Many foods are stored refrigerated before consumption, and in such cases, B. cereus has to adapt to low-temperature conditions.

B. cereus growth at low temperature takes place with a lag phase which may correspond to an adaptation phase (12). Cold is a stress which dramatically affects membrane fluidity, protein synthesis, and also the topology of nucleic acids (22). When exposed to low temperature, bacteria have to face a transient inhibition of protein synthesis mainly due to the presence of secondary structures in mRNA that are stabilized by cold conditions (16, 19). To overcome the translation interruption, cold-shocked cells synthesize cold-induced RNA helicases, which remove secondary structures from RNA duplexes in the presence of ATP, such as CsdA of Escherichia coli (19) or CshA of Bacillus subtilis (1). csdA and srmB deletion mutants of E. coli showed a cold-sensitive phenotype, and these RNA helicases have been described as involved in the biogenesis of the ribosomal 50S subunit at 20°C (10, 11). RNA helicases could also be involved in the degradation of mRNA by unwinding double-stranded mRNA, thereby allowing the action of RNase (8).

We have recently shown that the deregulation of the expression of one RNA helicase gene of B. cereus ATCC 14579 increased the lag phase of B. cereus at a low temperature (7). In this context, our aim was to investigate the role of the five putative RNA helicases present in the genome of B. cereus ATCC 14579 in its adaptation at low temperature, close to the growth limit.

Identification of putative RNA helicases in B. cereus ATCC 14579 and transcriptional analysis.

In silico analysis of the B. cereus ATCC 14579 genome revealed the presence of five genes predicted to encode RNA helicases. These five open reading frames are BC0259, BC4283, BC2103, BC5451, and BC2408, respectively renamed CshA to CshE, and they exhibit from 50 to 76% identity with the four B. subtilis RNA helicases. Nine motifs highly conserved in yeast eIF4A protein, representative of the DEXD-box family (13), were also found in each of the five B. cereus RNA helicases. These motifs were shown to be involved in the ATPase and helicase activity as well as RNA binding (5). The B. cereus CshA protein showed a high level of homology with CshA of B. subtilis (formerly named YdbR) and CsdA of E. coli. Those proteins all possess, in addition to the already described conserved motifs, a C-terminal glycine-arginine-rich domain thought to mediate the interaction of DEAD-box proteins with their specific partners or RNA binding (24) and necessary for E. coli CsdA function at 15°C (25). Reverse transcriptase-PCR (RT-PCR) experiments, performed as previously described (6) on RNA extracted from cells grown at 10°C, indicated that cshA transcripts were monocistronic, whereas cshB was cotranscribed with the downstream gene BC4282, encoding an endonuclease IV (data not shown). The presence of cotranscripts in the cshC locus, composed of six genes, was tested by pairs: each RT-PCR combination gave amplicons, indicating that cotranscriptions occur in this locus and that cshC is at least cotranscribed in the tested conditions with BC2102 and BC2104, encoding two small hypothetical proteins. The opposite orientation of cshD and cshE with their neighboring genes strongly suggested their monocistronic transcription. The presence of cold-regulating elements was previously described in the upstream region of cshA (7), but these cold-boxes are not present in the promoter region of the four other B. cereus RNA helicases. The cshA-to-cshE genes are present in five loci dispersed on the B. cereus ATCC 14579 chromosome and found in 9 out of the 10 B. cereus genomes completely sequenced (including B. weihenstephanensis KBAB4). The synteny of these loci is conserved in all strains. The B. cereus subsp. cytotoxicus 391-98 genome harbors only four RNA helicase genes, with the cshD gene missing.

Expression of RNA helicase genes at different temperatures.

Cold-induced expression of RNA helicase-encoding genes has been studied in bacteria and archaebacteria consequent to cold shock or low-temperature adaptation (9, 19, 20). We investigated the expression of the five B. cereus csh genes at 37°C and 10°C by real-time RT-PCR using the Quantifast SYBR green RT-PCR kit (Qiagen) according to the manufacturer's instructions. Two independent growths of B. cereus ATCC 14579 wild-type (WT) cells (Table 1) were performed in Luria-Bertani (LB) broth, under regulated batch conditions at 10°C and 37°C (pH was maintained at a constant 7.0, with stirring at 200 rpm and partial O₂ pressure [pO₂] at 100%) in a 2-liter bioreactor (Discovery 100MRU Inceltech). Total RNA was extracted as previously described (6) from cells sampled at three times of growth at 10°C and 37°C. Oligonucleotides listed in Table 1 with the “qRT” prefix were used for quantitative RT-PCR (qRT-PCR). The five genes were all overexpressed from 2.7- to 9.2-fold at 10°C compared to expression at 37°C (Table 2). The highest level of expression was observed for cshA when bacteria were grown at both 10°C and 37°C, suggesting an important role of this RNA helicase whatever the temperature. The csh genes were expressed at higher levels at the early growth phase (sample S1) than at other stages, except cshC, whose expression was constant along time. These results suggest a possible role of the five csh genes in B. cereus growth at 10°C.

TABLE 1.

Strains, plasmids, and primers

Strain or plasmid	Characteristicsb	Source or reference
Strains
B. cereus ATCC 14579		Laboratory collection
B. cereus ΔcshA	ATCC 14579 BC0259 Δ452-993	This work
B. cereus ΔcshB	ATCC 14579 BC4283 Δ398-897	This work
B. cereus ΔcshC	ATCC 14579 BC2103 Δ259-896	This work
B. cereus ΔcshD	ATCC 14579 BC5451 Δ410-1023	This work
B. cereus ΔcshE	ATCC 14579 BC2408 Δ386-1030	This work
E. coli ET12567	F−dam-13::Tn9 dcm-6 hsdM hsdR recF143 zjj-202::Tn10 galK2 galT22 ara14 pacY1 xyl-5 leuB6 thi-1	Laboratory collection
Plasmids
pMAD	Apr and Emr shuttle vectora	3
pMAD-ΔcshA	Recombinant pMAD harboring BC0259 Δ452-993	This work
pMAD-ΔcshB	Recombinant pMAD harboring BC4283 Δ398-897	This work
pMAD-ΔcshC	Recombinant pMAD harboring BC2103 Δ259-896	This work
pMAD-ΔcshD	Recombinant pMAD harboring BC5451 Δ410-1023	This work
pMAD-ΔcshE	Recombinant pMAD harboring BC2408 Δ386-1030	This work
pHT304	Apr and Emr cloning vector	2
pHT304ΩcshA	BC0259 (1,602 bp) and its promoter (364 bp) cloned in SacI and XbaI sites in pHT304	This work
pHT304ΩpcshA	BC0259 (1,602 bp) and a part of its promoter (197 bp) cloned in SalI and EcoRI sites in pHT304	This work
pHTXyl	Apr and Emr cloning vector	23
pHTXylΩcshA	Bc0259 (1,602 bp) cloned under the Pxyl promoter in XbaI and SacI in pHTxyl	This work
Primers
qRT-BC0259R/qRT-BC0259F	TAGAACTGCTGAATGTTTGG/TTTACCGATTTATGGTGGTC
qRT-BC2103F/qRT-BC2103R	TTCTTAAACGACCCATTCCG/GTACGTCCTGAGCGGTGAAT
qRT-BC2408F/qRT-BC2408R	GATGAAGCGGATCAAATGCT/AAAACGAAGTGCATCTGGCT
qRT-BC4283F/qRT-BC4283R	AGCACGCATGCCTAAAAACT/AGCAACTTGGTCTGCCATCT
qRT-BC5451F/qRT-BC5451R	TACTGCAAGAACGGGACGAG/TCCCGCCATTAAAGTACAGC
qRT-16SF/qRT-16SR	GGTAGTCCACGCCGTAAACG/GACAACCATGCACCACCTG

^aAp, ampicillin; Em, erythromycin.

^bΔ, deletion of the subsequent nucleotides.

TABLE 2.

Relative expression of cshA, cshB, cshC, cshD, and cshE genes in Bacillus cereus ATCC 14579 at 10°C and 37°C

Gene	Relative mRNA level (n-fold)	Relative expression in samplec
		S1	S2	S3
cshA	at 10°C vs 37°Ca	+6.6	+4.0	+5.7
	at 10°Cb	1	−1.5	−2.1
cshB	at 10°C vs 37°C	+4.0	+2.7	+3.3
	at 10°C	1	−3.2	−1.3
cshC	at 10°C vs 37°C	+4.9	+3.8	+4.4
	at 10°C	1	+1.1	1
cshD	at 10°C vs 37°C	+5.4	+2.7	+2.8
	at 10°C	1	−1.4	−1.5
cshE	at 10°C vs 37°C	+9.2	+7.8	+7.6
	at 10°C	1	1	−1.5

^aThe mRNA level for each gene was normalized to the RNA level of the ssu gene encoding 16S RNA and quantified by the 2^−ΔΔCT method as previously described (21). Only ratios of ≤0.5 and ≥2 were considered to be significant (i.e., P ≤ 0.05) according to the accuracy of the method.

^bmRNA level was determined by the following formula: 2^−ΔΔCT with ΔΔC_T = ΔC_{T 10°C for Sx} − ΔC_{T 10°C for S1.}

^cThe data are the means of results for two replicates on mRNA samples extracted from two independent cultures at 37°C and at 10°C (with a coefficient of variation of <12 %, except for cshA and cshB in S2 at 10°C, results for which were, respectively, 13.4 and 22.4%). + and − indicate, respectively, up- and downregulated genes. At 37°C, sample S1 corresponds to cells harvested at an OD₆₀₀ of 0.3, S2 at an OD₆₀₀ of 0.9, and S3 at an OD₆₀₀ of 2.0; at 10°C, S1 corresponds to an OD₆₀₀ of 0.2, S2 to an OD₆₀₀ of 0.5, and S3 to an OD₆₀₀ of 1.2.

Impact of RNA helicase gene deletions on B. cereus growth at low temperature.

To study the role of the RNA helicase genes in B. cereus cold adaptation, mutant strains were constructed for each csh gene by allelic exchange, as previously described (3), without insertion of any antibiotic resistance cassette to avoid polar effect on surrounding genes (Table 1). Growth at 30°C in LB broth of the ΔcshB, ΔcshC, ΔcshD, and ΔcshE strains was similar to that of the WT, whereas growth of the ΔcshA strain was slightly impaired (Fig. 1 A). Bacterial counts were in accordance with optical density at 600 nm (OD₆₀₀) values (Fig. 1B). Growth of all five mutants, including the ΔcshA strain, was similar to that of the WT at 37°C and 42°C (data not shown). At 10°C, compared to growth of the WT, growth of the cshD and cshE isogenic mutants was clearly not impaired (data not shown), whereas the other mutants showed impaired (ΔcshB) or an absence of (ΔcshA and ΔcshC) cold growth (Fig. 1C). These phenotypes contrast with those described for B. subtilis ΔcshA or ΔcshB strains which were not cold sensitive (17). The minimum growth temperature was approximately 12°C for the ΔcshB strain and was about 15°C for the ΔcshA and ΔcshC strains (data not shown). CFU counts revealed that the number of ΔcshB cells remained constant at 10°C (Fig. 1D), suggesting that the increase in OD₆₀₀ values was due to an increase in mutant cell length. CFU numbers of the ΔcshA and ΔcshC strains decreased slightly at 10°C, whereas the OD₆₀₀ remained constant. This could be due to a slight loss of viability at 10°C or to the formation of aggregates. However, when the temperature was shifted to 37°C, growth of mutant cells was observed, suggesting that the absence of growth at low temperature of the ΔcshA, ΔcshB, and ΔcshC strains was mainly due to their altered ability to adapt to cold conditions.

FIG. 1.

Growth at 30°C (A and B) and 10°C (C and D) of B. cereus ATCC 14579 (▪) and the ΔcshA (□), ΔcshB (▵), ΔcshC (×), ΔcshD (○), and ΔcshE (⋄) mutants in LB broth. Changes in OD₆₀₀ (A and C) and numeration of bacterial cells in log CFU ml⁻¹ (B and D) are presented. Independent cultures were performed in triplicate; error bars indicate standard deviations.

Observations under phase-contrast microscope showed that after growth at low temperatures (i.e., close to their minimum growth temperature), the morphologies of the three cold-sensitive mutants were differently affected compared to that of WT cells (Fig. 2 A): ΔcshA cells were long, formed large aggregates, as previously described for E. coli ΔcsdA (19), and were curved at the pole (Fig. 2B). ΔcshB cells formed very long filaments (Fig. 2C), while ΔcshC cells were short and stocky (Fig. 2D). Staining of ΔcshB cells with the membrane stain FM 4-64 revealed red fluorescent areas, confirming the presence of cell membrane septa within filamentous cells (data not shown). When observed under transmission electron microscopy (TEM), the cell structure of the three csh mutants was similar to that of the WT at 30°C (data not shown). After growth at low temperature, cell structures of the ΔcshA, ΔcshB, and ΔcshC strains were strongly modified compared to that of WT cells (Fig. 3 A): ΔcshA cells grown at 15°C were long and incompletely divided with thickened membranes (Fig. 3B), which were also seen in ΔcshB cells (Fig. 3C). ΔcshC cells also exhibited a thicker membrane than WT cells on TEM images but looked like ghosts with a scattered cytoplasm (Fig. 3D). Viability of mutant cells grown at 12°C or 15°C was tested using the LIVE/DEAD bacterial viability test and compared to that of the WT strain (Fig. 2E to H): approximately 52% and 44.5% of cells stained red, corresponding to bacteria which lost their membrane integrity for WT and ΔcshB strains, respectively, grown at 12°C. At 15°C, 38% of red cells were counted for the WT strain and 45% for the ΔcshA and ΔcshC strains. These results showed that cell viability at a permissive low temperature was not affected in mutants compared to WT strains, despite the marked alterations in their morphologies.

FIG. 2.

Morphology of cells grown at low temperatures of B. cereus WT (A and E) and the ΔcshA (B and F), ΔcshB (C and G), and ΔcshC (D and H) strains. Cells grown at 12°C (WT, ΔcshB, and ΔcshC) or 15°C (ΔcshA) and harvested at an OD₆₀₀ of 1.0 were observed under phase-contrast microscopy at a magnification of ×1,000 (left-hand panel) or after staining with a LIVE/DEAD bacterial viability test (Invitrogen) (right-hand panel). Red and green cells were counted in 10 random microscope fields.

FIG. 3.

Transmission electron micrographs of B. cereus WT (A), ΔcshA (B), ΔcshB (C), and ΔcshC (D) cells grown at low temperatures (×18,000 magnification). Cells were grown until late exponential phase and prepared as described previously (26). Bars indicate 500 nm. Cells in the presented pictures were representative of the majority of the bacteria observed.

Complementation of cold-sensitive phenotypes.

In vivo complementation of ΔcshA cells with a plasmidic copy of cshA (which was under the control of the entire promoter region; see pHT304ΩcshA in Table 1) fully restored the WT phenotype of growth at 12°C (Fig. 4), showing the specific involvement of CshA in the cold-sensitive phenotype of the ΔcshA strain. As shown in Fig. 4, complementation of the cshA mutation was not obtained if the trans cshA copy was placed under the control of a truncated promoter region (pHT304ΩpcshA lacking the distal cold-box in Table 1) or an inducing xylose promoter (pHTXylΩcshA in Table 1), suggesting that the presence in its 5′ noncoding region of three conserved motifs similar to cold-box elements, described in many bacteria as modulators of cold-induced mRNA stability, was essential (7, 14).

FIG. 4.

Complementation assays with the cshA gene at 12°C of B. cereus ΔcshA strain. Horizontal lines represent the promoter region of cshA (solid lines) or the Pxyl promoter (dashed line); boxes are putative cold-boxes. Colonies on LB agar plates were grown for 7 days at 12°C from 10-fold serially diluted bacterial suspensions.

Introduction of a copy of cshA in ΔcshB cells complemented in part its cold phenotype as revealed by growth in LB broth at 12°C (data not shown). Morphology of cells of the ΔcshB strain carrying the pHT304ΩcshA plasmid was closer to that of the WT than to that of the ΔcshB cells. Such an additional copy of cshA did not alter the growth at 30°C. This strongly suggests that CshA and CshB could have similar functions in B. cereus cold adaptation or that CshA could have multiple overlapping functions as described for E. coli CsdA (18). In contrast, the cold phenotype due to the cshC deletion was not complemented at 12°C on LB plates or in LB medium by the introduction in trans of cshA. A possible polar effect of cshC mutation of the downstream genes BC2104 and BC2105 was unlikely because the corresponding mRNAs were still detected in ΔcshC cells by RT-PCR (not shown).

This study demonstrated that the five RNA helicase-encoding genes present in the B. cereus ATCC 14579 genome were upregulated in response to low-temperature conditions but that only three were involved during growth at low temperature. Their deletion raised the minimum growth temperature threshold of B. cereus and, at low temperatures permissive for growth, caused strong modifications of cell morphology. The CshA RNA helicase may have a stronger role in B. cereus temperature adaptation, as the deletion of cshA already altered growth at 30°C, unlike that of cshB and cshC. These three RNA helicases may have different functions, as their deletion caused different modifications, with possible overlapping roles between CshA and CshB at low temperature. The exact function of each RNA helicase in RNA metabolism in the cold adaptation of B. cereus needs further investigation.