The Cold-Induced Two-Component System CBO0366/CBO0365 Regulates Metabolic Pathways with Novel Roles in Group I Clostridium botulinum ATCC 3502 Cold Tolerance

Abstract

The two-component system CBO0366/CBO0365 was recently demonstrated to have a role in cold tolerance of group I Clostridium botulinum ATCC 3502. The mechanisms under its control, ultimately resulting in increased sensitivity to low temperature, are unknown. A transcriptomic analysis with DNA microarrays was performed to identify the differences in global gene expression patterns of the wild-type ATCC 3502 and a derivative mutant with insertionally inactivated cbo0365 at 37 and 15°C. Altogether, 150 or 141 chromosomal coding sequences (CDSs) were found to be differently expressed in the cbo0365 mutant at 37 or 15°C, respectively, and thus considered to be under the direct or indirect transcriptional control of the response regulator CBO0365. Of the differentially expressed CDSs, expression of 141 CDSs was similarly affected at both temperatures investigated, suggesting that the putative CBO0365 regulon was practically not affected by temperature. The regulon involved genes related to acetone-butanol-ethanol (ABE) fermentation, motility, arsenic resistance, and phosphate uptake and transport. Deteriorated growth at 17°C was observed for mutants with disrupted ABE fermentation pathway components (crt, bcd, bdh, and ctfA), arsenic detoxifying machinery components (arsC and arsR), or phosphate uptake mechanism components (phoT), suggesting roles for these mechanisms in cold tolerance of group I C. botulinum. Electrophoretic mobility shift assays showed recombinant CBO0365 to bind to the promoter regions of crt, arsR, and phoT, as well as to the promoter region of its own operon, suggesting direct DNA-binding transcriptional activation or repression as a means for CBO0365 in regulating these operons. The results provide insight to the mechanisms group I C. botulinum utilizes in coping with cold.

INTRODUCTION

Understanding the mechanisms by which food-borne pathogenic microorganisms cope with stress conditions they encounter in foods is of key importance in designing modern food safety measures. The ability of the anaerobic Gram-positive spore-forming Clostridium botulinum to survive, grow, and subsequently produce the extremely potent botulinum neurotoxin in foods (1) raises substantial concern over food safety (2, 3). Exposure of bacteria to sublethal stress can result in increased robustness and (cross-)protection toward harsher treatments, thus creating challenges in classical hurdle design in food processing (4). Identification of key mechanisms behind response and adaptation to the environmental hurdles C. botulinum encounters in the food chain could potentially allow development of targeted control measures.

Bacteria utilize two-component systems (TCSs) to sense environmental stimuli and activate adaptive mechanisms needed for survival and subsequent growth (5). The typical assembly of a TCS consists of a membrane-bound sensor histidine kinase, which senses a defined signal and undergoes a conformational change, phosphorylating its cognate response regulator protein (reviewed in reference 6). The activated response regulator subsequently induces transcription of genes required for survival in the situation at hand or represses genes that are unnecessary or even detrimental.

The role of TCSs in response and adaptation to low temperatures has been demonstrated in several prokaryotes. Most notably, the DesK/DesR TCS of Bacillus subtilis has been shown to respond to a temperature downshift and subsequently activate the transcription of an O₂-dependent membrane lipid desaturase-encoding des, resulting in increased membrane fluidity (7, 8). In addition, cold-related TCSs have been identified in Listeria monocytogenes (9), Haemophilus influenzae (10), the archaeon Methanolobus psychrophilus R15 (11), and Yersinia pseudotuberculosis (12).

As for C. botulinum, the reports regarding machineries related to sensing and adapting to low temperature are scarce. The cold shock protein CspB was shown to be important in cold tolerance of C. botulinum ATCC 3502 (13) and, recently, the TCS CBO2306/CBO2307 was shown to have an important role in cold adaptation in this organism (14). In addition, we have shown the TCS CBO0366/CBO0365 of C. botulinum ATCC 3502 to be induced upon temperature downshift (15); furthermore, insertional inactivation of either of the TCS genes resulted in a cold-sensitive phenotype (15). However, the mechanisms through which this TCS exerts cold tolerance in C. botulinum are unknown.

To further characterize the role of CBO0366/CBO0365 in cold shock response and cold tolerance of C. botulinum ATCC 3502, transcriptomic analysis of the differences in global gene expression between wild-type and cbo0365 mutant cultures at optimal and low temperatures was performed to identify genes putatively under transcriptional control of the CBO0365 response regulator. In addition, several differentially expressed genes, together with genes encoding components of related metabolic pathways, were mutated. The growth of these mutants at low temperatures was investigated to gain insight into the possible role of the mutated pathways under cold stress. Electrophoretic mobility shift assays (EMSAs) were carried out with recombinant CBO0365 protein to confirm direct regulation of several operons by the CBO0365 regulator.

MATERIALS AND METHODS

Construction of mutant strains.

C. botulinum ATCC 3502 (group I, type A) was used as a parent strain in the present study. An insertional knockout mutant for cbo0365 was constructed using the ClosTron technology (16) by Lindström et al. (15). Insertional inactivation of genes cbo0751, cbo0753, cbo1407, cbo2525, cbo2847, cbo3199, and cbo3202 was similarly accomplished with the ClosTron technology, with de novo synthesized intron targeting regions (DNA2.0, Inc., Menlo Park, CA) cloned into the ClosTron pMTL007C-E2 vector (17, 18). The intron insertion sites and orientations for the mutant strains are presented in Table 1, and the primers used for confirmation of the insertion site and orientation are described in Table 2. Single intron insertion was confirmed by Southern blotting with a probe targeted to the erm resistance marker within the inserted DNA fragment as described previously (20). A single intron insertion was similarly demonstrated for the previously constructed cbo0365 mutant.

TABLE 1.

Strains and plasmids used in this study

Strain or plasmid	Relevant properties	Source or reference
Bacterial strains
C. botulinum
ATCC 3502	Wild-type parental strain	ATCCa
ATCC 3502 cbo0365::intron 48|49S	Insertional disruption of cbo0365 at base 48 in sense orientation, erm	15
ATCC 3502 cbo0751::intron 254|255AS	Insertional disruption of cbo0751 at base 254 in antisense orientation, erm	This study
ATCC 3502 cbo0753::intron 196|197AS	Insertional disruption of cbo0753 at base 196 in antisense orientation, erm	This study
ATCC 3502 cbo1407::intron 428|429AS	Insertional disruption of cbo1407 at base 428 in antisense orientation, erm	This study
ATCC 3502 cbo2525::intron 122|123AS	Insertional disruption of cbo2525 at base 122 in antisense orientation, erm	This study
ATCC 3502 cbo2847::intron 509|510AS	Insertional disruption of cbo2847 at base 509 in antisense orientation, erm	This study
ATCC 3502 cbo3199::intron 467|468AS	Insertional disruption of cbo3199 at base 467 in antisense orientation, erm	This study
ATCC 3502 cbo3202::intron 167|168AS	Insertional disruption of cbo3202 at base 167 in antisense orientation, erm	This study
E. coli
CA434	Conjugation donor	University of Nottingham
Rosetta 2(DE3)/pLysS	Recombinant expression host, pRARE2 encoding seven rare tRNAs, catP	Novagen, Darmstadt, Germany
Plasmids
pMTL007C-E2	ClosTron plasmid, catP, L1.LtrB intron with ermB RAM, constitutive intron expression under fdx promoter	University of Nottingham (19)
pMTL007C-E2::cbo0751-254|255AS	pMTL007C-E2 with L1.LtrB retargeted to base 254 of cbo0751 in antisense orientation	This study
pMTL007C-E2::cbo0753-196|197AS	pMTL007C-E2 with L1.LtrB retargeted to base 196 of cbo0753 in antisense orientation	This study
pMTL007C-E2::cbo1407-428|429AS	pMTL007C-E2 with L1.LtrB retargeted to base 428 of cbo1407 in antisense orientation	This study
pMTL007C-E2::cbo2525-122|123AS	pMTL007C-E2 with L1.LtrB retargeted to base 122 of cbo2525 in antisense orientation	This study
pMTL007C-E2::cbo2847-509|510AS	pMTL007C-E2 with L1.LtrB retargeted to base 509 of cbo2847 in antisense orientation	This study
pMTL007C-E2::cbo3199-467|468AS	pMTL007C-E2 with L1.LtrB retargeted to base 467 of cbo3199 in antisense orientation	This study
pMTL007C-E2::cbo3202-167|168AS	pMTL007C-E2 with L1.LtrB retargeted to base 167 of cbo3202 in antisense orientation	This study
pET-28b(+)	Recombinant protein expression vector, kanR	Novagen
pET-28b(+)-cbo0365-HIS	pET-28b(+) harboring cbo0365 with N-terminal His6 tag	This study

^aATCC, American Type Culture Collection.

TABLE 2.

Oligonucleotide primers used in this study

Primer	Use	Modification (sequence [5′–3′])a
cbo0366 RT	Gene-specific reverse transcription	CTTTTGATATTCCGCCCAAA
cbo0363-cbo0364 F	RT-PCR	AGTATGGTTTAATGTAGCGGTAGTAAG
cbo0363-cbo0364 R	RT-PCR	GCTTAGTGGCACAATATTTTCTTT
cbo0364 F	RT-PCR	TTTTCTGTTCCTTATATGGTTTGG
cbo0364 R	RT-PCR	TTTTCTGCTGACATTTCTTAATCA
cbo0365-cbo0366 F	RT-PCR	AAACAGTTTGGGGAGTTGGA
cbo0365-cbo0366 R	RT-PCR	TTTTTATCCATGCTCCAATGTCT
cbo0365 F	RT-PCR	TGATGCCTAAGATGGATGGT
cbo0365 R	RT-PCR	TCTTCATTTTCATTTCCATTTGA
cbo0366-cbo0366 F	RT-PCR	AAAATTAATTATGGCAGAGGATGAA
cbo0366-cbo0367 R	RT-PCR	AAATCTGCAGCACCTTTTACA
cbo0751 qPCR F	RT-qPCR	TCTGCGGGGACAGAAACTAAAC
cbo0751 qPCR R	RT-qPCR	CCCCAATCCTCTGTATGTTTTGC
cbo0753 qPCR F	RT-qPCR	TCCTGTGGAGAAAAGTGTGCTTG
cbo0753 qPCR R	RT-qPCR	TATGATGGGACAGTGTCGGTTG
cbo1407 qPCR F	RT-qPCR	ACTGGCTCAGAAATGGATGC
cbo1407 qPCR R	RT-qPCR	AAATTTAGGGGCCATGGAAG
cbo2226 qPCR F	RT-qPCR	TTAAGGCGGGGAGTAATGTG
cbo2226 qPCR R	RT-qPCR	CGGAACAATTGAAACACCTG
cbo2227 qPCR F	RT-qPCR	GGAGCTGTTGTTGGTTTGATTGG
cbo2227 qPCR R	RT-qPCR	TGCAACGAATGCTCCTGCTATC
cbo2847 qPCR F	RT-qPCR	GGCACTTGCAGCTGATTTAG
cbo2847 qPCR R	RT-qPCR	TCCTCTCTCCAAAAGAGTCTCC
cbo3199 qPCR F	RT-qPCR	GGAATACGGTGGAGCTGGTA
cbo3199 qPCR R	RT-qPCR	TGGAGCGCAACATAAAGATG
cbo3202 qPCR F	RT-qPCR	AGCCGACTATAGCAGCCGTA
cbo3202 qPCR R	RT-qPCR	TCCTCCGAATCCTGGAGTTA
16S rrn F	RT-qPCR, EMSA probe preparation	AGCGGTGAAATGCGTAGAGA
16S rrn F 5′ biotin	EMSA probe preparation	Biotin-AGCGGTGAAATGCGTAGAGA
16S rrn R	RT-qPCR, EMSA probe preparation	GGCACAGGGGGAGTTGATAC
cbo0751M 254|255a R	Intron insertion confirmation	GGTCCTCTAATCCCCAATCC
cbo0753M 196|197a R	Intron insertion confirmation	ATTTAATGAATAGTGACTCCATAATCC
cbo1407M 428|429a R	Intron insertion confirmation	AAATTTAGGGGCCATGGAAG
cbo2525M 122|123a R	Intron insertion confirmation	TCATCCTGACCACCAACATC
cbo2847M 509|510a R	Intron insertion confirmation	TCCTCTCTCCAAAAGAGTCTCC
cbo3199M 467|468a R	Intron insertion confirmation	ATCCTCTGTCTTGCCAATGC
cbo3202M 167|168a R	Intron insertion confirmation	CGGCTTTTCCATGGTTTCTA
EBS universal	Intron insertion confirmation	CGAAATTAGAAACTTGCGTTCAGTAAAC
cbo0365 F NheI	CDS cloning for protein overexpression	CCCGCTAGCATGTCAGCAGAAAAAATCCTTATTG
cbo0365 R XhoI	CDS cloning for protein overexpression	CGCCTCGAGCTATTTTTCAACCTTATATCCAACTCC
T7 promoter	Sequencing pET28b(+)-cbo0365-HIS	TAATACGACTCACTATAGGG
T7 terminator	Sequencing pET28b(+)-cbo0365-HIS	GCTAGTTATTGCTCAGCGG
Pcbo0364 150bp F	EMSA probe preparation	AACAGGGCAAATATAAGGAAAGTG
Pcbo0364 150bp F 5′ biotin	EMSA probe preparation	Biotin-AACAGGGCAAATATAAGGAAAGTG
Pcbo0364 150bp R	EMSA probe preparation	TCAATTTAACTTCCCCCATAACC
Pcbo0753 454bp F	EMSA probe preparation	TCTTCAGCAACTTTTCTTATTTTATCA
Pcbo0753 454bp F 5′ biotin	EMSA probe preparation	Biotin-TCTTCAGCAACTTTTCTTATTTTATCA
Pcbo0753 454bp R	EMSA probe preparation	TCTGATAGTGCTTTAATAATTTTTGC
Pcbo0753 180bp F	EMSA probe preparation	AGTTAATATTTTCATGTTCACATTT
Pcbo0753 180bp F 5′ biotin	EMSA probe preparation	Biotin-AGTTAATATTTTCATGTTCACATTT
Pcbo0753 180bp R	EMSA probe preparation	CAAGCTCATTAATCTCCCTCCT
Pcbo2525 223bp F	EMSA probe preparation	TTTTTGGTGCCTATAACAAAAGC
Pcbo2525 223bp F 5′ biotin	EMSA probe preparation	Biotin-TTTTTGGTGCCTATAACAAAAGC
Pcbo2525 223bp R	EMSA probe preparation	TTCATTAATAAAAACCTCCGTTCA
Pcbo3202 250bp F	EMSA probe preparation	CAAAGTTTTATTTGATGGCAGT
Pcbo3202 250bp F 5′ biotin	EMSA probe preparation	Biotin-CAAAGTTTTATTTGATGGCAGT
Pcbo3202 250bp R	EMSA probe preparation	TCCACTTAACCACCCCCTTA

^aRestriction enzyme sites in the primer sequences are underlined.

Culture conditions.

For RNA isolation for DNA microarray, quantitative real-time PCR and non-quantitative reverse transcription-PCR experiments, the bacterial cultures were anaerobically grown at 37°C in TPGY medium until the optical density at 600 nm (OD₆₀₀) reached 1.0, and the cultures were subjected to rapid temperature downshift to 15°C. Samples for total RNA isolation were withdrawn from the cultures immediately before the temperature downshift, and after 1 h anaerobic incubation at 15°C. The experiment was performed as three independent biological replicates.

Growth experiments to compare the growth of the ATCC 3502 wild-type and mutant strains were performed in an automatic turbidity reader (Bioscreen C Microbiology Reader; Growth Curves, Helsinki, Finland) placed in an anaerobic workstation with an internal atmosphere of 85% N₂, 10% CO₂, and 5% H₂ (MK III; Don Whitley Scientific, Ltd., Shipley, United Kingdom) as described previously (21). Strains were grown in three to five biological replicates in TPGY broth at 37°C for 24 h or at 17°C for 7 days or in TPGY broth supplemented with 0.1 mM sodium arsenite at 37°C for 2 days and at 20°C for 10 days. The OD₆₀₀ of the cultures was automatically monitored at regular intervals. Growth curves were constructed by plotting the OD₆₀₀ of each culture against time.

RNA extraction.

Samples from three replicate C. botulinum ATCC 3502 wild-type or cbo0365 mutant cultures were collected at the time points described above for total RNA isolation. The samples were collected into sterile plastic tubes containing ice-cold ethanol-phenol (9:1) stop solution (Sigma-Aldrich, St. Louis, MO) in a ratio of sample to stop solution of 5 to 1, mixed thoroughly, and incubated on ice for 30 min. Cells were harvested by centrifugation (4°C, 8,000 × g) for 5 min. The cell pellets were immediately frozen to −70°C until RNA extraction.

The cell pellets were thawed on ice for 5 min and used for RNA extraction with the RNeasy minikit or RNeasy Midi kit (Qiagen GmbH, Hilden, Germany) according to the manufacturer's instructions. The cells were lysed with a solution containing 25 mg lysozyme (Sigma-Aldrich, St. Louis, MO)/ml and 250 IU of mutanolysin (Sigma-Aldrich)/ml in Tris-EDTA buffer (pH 8.0; Fluka BioChemica, Buchs, Switzerland) under agitation at 37°C for 30 min. To ensure efficient removal of all genomic DNA, an additional DNase treatment was carried out using the DNA-free kit (Ambion, Austin, TX) according to the manufacturer's instructions.

The RNA yield and purity (A₂₆₀/A₂₈₀) were checked using the NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific, Inc., Waltham, MA). The A₂₆₀/A₂₈₀ ratio was >2.0 for all samples. Integrity of RNA was confirmed with miniaturized gel electrophoresis with the Agilent Bioanalyzer (Agilent Technologies, Inc., Santa Clara, CA). The RNA integrity number was >9.2 for all RNA samples.

cDNA synthesis.

For DNA microarray analysis, a total of 2 μg of each RNA sample was reverse transcribed into cDNA and simultaneously labeled with fluorescent dyes. In brief, each 30-μl labeling reaction mixture contained 0.2 μg of random hexamers (Invitrogen)/μl, 0.01 M dithiothreitol (DTT; Invitrogen), 1.3 U of RNase inhibitor (Invitrogen)/μl, 0.5 μM dATP, dTTP, and dGTP, 0.2 μM dCTP, 1.7 nmol of Cy-3 (two replicates of wild type, one replicate of cbo0365 mutant) or Cy-5 (two replicates of cbo0365 mutant, one replicate of wild type)-labeled dCTP (GE Healthcare, Pittsburgh, PA), 13 U of SuperScript III reverse transcriptase (Invitrogen)/μl, and appropriate buffer (1× First-Strand Buffer; Invitrogen) and then incubated at 46°C for 3 h. RNA hydrolysis and reaction inactivation were performed by addition of 15 μl of 0.1 M NaOH and 0.5 mM EDTA and incubation at 70°C for 15 min. The reactions were subsequently neutralized by addition of 15 μl of 0.1 M HCl. The cDNA was purified with QIAquick PCR purification kit (Qiagen), with final elution volume of 40 μl. The cDNA concentration of each sample was measured with NanoDrop.

For RT-qPCR, a total of 500 ng of each 15°C RNA sample was used for cDNA synthesis with the DyNAmo cDNA synthesis kit (Thermo Fisher Scientific) as instructed by the manufacturer. Each 20-μl reaction, which contained 15 ng of random hexamers/μl, 10 IU of Moloney murine leukemia virus RNase H⁺ reverse transcriptase solution (Thermo Fisher Scientific), and appropriate buffer containing deoxynucleoside triphosphates (dNTPs) and MgCl₂ in a final concentration of 5 mM (1×; Thermo Fisher Scientific), was incubated at 25°C for 10 min and at 37°C for 30 min, inactivated at 85°C for 5 min, and finally chilled to 4°C. Two replicate RT reactions were made for each RNA sample.

For reverse transcription-PCR (RT-PCR), a 20-μl reaction mixture containing 1 μg of C. botulinum ATCC 3502 RNA from a 37°C sample and 2 μM primer cbo0366-RT was incubated at 65°C for 5 min for denaturation and on ice for 1 min for annealing. RT was performed in a reaction mixture containing 0.5 mM concentrations of each dNTP (Thermo Fisher Scientific), 5 mM DTT, 40 U of RNase inhibitor (RNaseOUT), First-Strand Buffer (50 mM Tris-HCl, 75 mM KCl, and 3 mM MgCl₂), and 200 U of SuperScript III reverse transcriptase (Invitrogen) or 2 μl of water for the no-RT control. The reaction mixtures were incubated at 55°C for 60 min. RNA hydrolysis, reaction inactivation, and neutralization were performed as described above for the cDNA synthesis procedure for microarray analysis.

Transcriptomic analysis with DNA microarrays.

The in situ-synthesized DNA microarrays (8x15K; Agilent Technologies) were custom designed to cover 3,641 chromosomal (out of the total of 3,648) and all 19 plasmid-borne open reading frames (ORFs) in the ATCC 3502 genome (22). Depending on the length of ORF, a total of 3 to 14 60-mer oligonucleotide probes were designed for each ORF.

For array hybridizations, biological replicate samples of the wild type and mutant were labeled with either Cy5 or Cy3 as described above. A total of 300 ng of Cy3-labeled cDNA and 300 ng of Cy5-labeled cDNA (wild type versus mutant from the same condition) were mixed into a final volume of 18 μl, and 0.1 mg of salmon sperm DNA (Invitrogen)/ml was added. The DNA was denatured at 95°C for 2 min, chilled on ice, and finally mixed with a blocking agent (Hi-RPM GE hybridization kit; Agilent Technologies) and hybridization buffer (Hi-RPM GE hybridization kit) as instructed by the manufacturer. A volume of 50 μl of the mix was pipetted onto the DNA microarrays. The arrays were hybridized at 65°C overnight and washed as instructed (Gene Expression wash buffer kit; Agilent Technologies).

The slides were scanned (Axon GenePix Autoloader 4200 AL; Westburg, Leusden, Netherlands) at 532 and 635 nm using a 5-μm resolution. Image processing was performed with GenePix Pro 6.0 software (Axon Instruments), and data analysis was performed with the R limma package (23). The foreground and local background intensities of each spot were characterized by the mean and median pixel values of the spot, respectively. Local background was subtracted from the foreground signal using the “normexp” method, with an offset value of 50 (24). For comparison between different arrays, the signal intensities measured in the Cy5 and Cy3 channels were converted into a logarithmic (log₂) scale and normalized using the loess method (25). Statistical analysis was performed to find differentially expressed genes between the wild type and the cbo0365 mutant. The analysis was done separately for each probe to control for variation within a coding sequence. A moderated t test with empirical Bayes variance shrinkage (“eBayes” function) was applied to each probe on the array, and the resulting P values were converted into false discovery rate (FDR) values by a Benjamini-Hochberg adjustment (“topTable” function) (26). For each ORF, the probe with a median unmodified P value for the expression difference was chosen to represent the ORF. Of the representative ORFs, those with a FDR of <0.05 were subsequently considered to have a significant difference in expression. Furthermore, of these significantly differently expressed ORFs, ones with a log₂ fold change of <−2.0 or >2.0 in either or both of the sampling points were considered to be included in the CBO0365 regulon.

RT-qPCR.

To validate the expression values obtained from the microarray experiments, quantitative real-time reverse transcription-PCR (RT-qPCR) was performed for selected genes, representing genes markedly less, more, or not differently expressed in the cbo0365 mutant in the microarray experiment. The DyNAmo Flash SYBR green qPCR kit (Thermo Fisher Scientific) was used according to the manufacturer's instructions to set up the RT-qPCRs. Each reaction consisted of 1× DyNAmo Flash SYBR green Master Mix (Thermo Fisher Scientific), 0.5 μM forward and reverse primers (Table 2), and 4 μl of cDNA in a total volume of 20 μl. The reactions were performed in the Rotor-Gene RG3000 thermal cycler (Qiagen) with initial heating at 95°C for 15 min to activate the DNA polymerase, followed by either 40 cycles of denaturation at 95°C for 10 s and then annealing and extension at 60°C for 20 s (primers cbo0365, cbo0366, cbo0753, cbo2227, cbo2525, cbo3199, cbo3202, and 16S rrn) or 40 cycles of denaturation at 95°C for 10 s, annealing at 55°C for 15 s, and extension at 72°C for 15 s (primers cbo0751, cbo1407, cbo2226, and cbo2847). Data acquisition was performed during the extension step. The amplification reaction efficiencies and the appropriate sample dilution factor for each primer pair were defined with dilution series of pooled cDNA. Rotor-Gene 6 software (Qiagen) was used to set the threshold fluorescence levels for each primer pair and to calculate the reaction efficiencies. The reaction efficiencies of the primers varied between 0.91 and 1.05. A sample dilution factor of 1:20 was used with primer pairs for cbo0365, cbo0366, cbo3199, and cbo3202, of 1:10³ for cbo0751, cbo0753, cbo1407, cbo2226, cbo2227, cbo2525, and cbo2847, and of 1:10⁵ for 16S rrn. The 1:10⁵-diluted no-RT controls were analyzed with the 16S rrn primers and reaction conditions described above, and no-template controls were included in each run. A melting-curve analysis was included in each run to confirm primer specificity.

The relative quantification of target gene expression in the cbo0365 mutant, normalized to reference gene (16S rrn) and calibrated to the samples taken from the wild-type strain, were determined by calculating the expression ratios of target genes with the Pfaffl method (27), which incorporates the reaction efficiency of each primer pair into the ratio calculations. The quantification cycle (Cq) values measured for 16S rrn remained stable throughout the experiment, supporting its use as a reliable normalization control. 16S rrn is the only previously reported reference gene for C. botulinum (28, 29), and no other suitable reference genes have been reported.

RT-PCR.

To identify the transcript structure of the cbo0365 locus, RT-PCR was performed on cDNA synthesized from wild-type ATCC 3502 RNA. Each reaction consisted of 1 μl of C. botulinum ATCC 3502 first-strand cDNA template, a 200 μM dNTP mixture (Thermo Fisher Scientific), 0.5 μM concentrations of each forward and reverse primer (primers targeted to the intergenic region between cbo0363 and cbo0364 [cbo0363-cbo0364f + cbo0363-cbo0364r], cbo0364 alone [cbo0364f + cbo0364r], and cbo0365 alone [cbo0365M-f + cbo0365-r1] or to the intergenic region between cbo0365 and cbo0366 [cbo0365-cbo0366-f + cbo0365-cbo0366-r]; Table 2), and 2 U of DyNAzyme II DNA polymerase (Thermo Fisher Scientific) in a 1× reaction buffer (Thermo Fisher Scientific). The PCR run consisted of 30 cycles of denaturation at 94°C for 30 s, annealing at 55°C for 30 s, and extension at 72°C for 1 min 20 s. The PCR products were visualized under UV light by ethidium bromide staining after agarose gel electrophoresis.

EMSAs.

To overexpress and purify the CBO0365 protein, a 699-bp DNA fragment containing the coding sequence (CDS) of cbo0365 was PCR amplified using the Phusion DNA polymerase (Thermo Fisher Scientific) with the primers cbo0365 F NheI and cbo0365 R XhoI (Table 2), incorporating an NheI site at the 5′ end and an XhoI site at the 3′ end of the CDS. The fragment was purified with the GeneJET PCR purification kit, digested with NheI and XhoI (New England BioLabs), and cloned into the pET-28b(+) expression vector (Novagen, Darmstadt, Germany) incorporating an N-terminal His₆ tag into the CDS. The plasmid was propagated in Escherichia coli TOP10 electrocompetent cells (Invitrogen) in LB broth containing 30 μg of kanamycin/ml. Plasmids were isolated from cultures with a GeneJET plasmid miniprep kit (Thermo Fisher Scientific) and verified by sequencing. Recombinant CBO0365 protein was expressed in a culture of E. coli Rosetta 2(DE3)/pLysS (Novagen) grown at 37°C in 100 ml of Luria-Bertani (LB) broth containing 30 μg of kanamycin/ml and 34 μg of chloramphenicol/ml. The cultures were grown to an OD₆₀₀ of 0.8, the protein expression was induced with 1 mM IPTG (isopropyl-β-d-thiogalactopyranoside), and the bacteria were further cultured for an additional 5 h. Cells were harvested by centrifugation at 10,000 × g for 15 min at 4°C and frozen overnight at −20°C. The cells were resuspended in 10 ml of binding buffer (0.5 M NaCl, 20 mM Tris-HCl, 16 mM imidazole [Sigma-Aldrich]; pH 7.9) and sonicated. Cell debris was separated from the soluble fraction by centrifugation (10,000 × g for 15 min at 4°C), and the supernatant was passed through a 0.45-μm-pore-size syringe-end filter. The filtrate was applied to 1 ml of His-Bind Ni²⁺-NTA resin (Novagen) and incubated at 4°C for 1 h under gentle agitation. The suspension was applied to a gravity flow chromatography column (Novagen), washed with 10 ml of binding buffer and 10 ml of wash buffer (0.5 M NaCl, 20 mM Tris-HCl, 60 mM imidazole [Sigma-Aldrich]; pH 7.9), and eluted in three 1-ml fractions with elution buffer (0.5 M NaCl, 20 mM Tris-HCl, 1 M imidazole; pH 7.9). The purity of the eluted fractions was assessed with SDS-PAGE with Coomassie blue staining, and the protein concentrations were approximated with a Bradford assay using bovine serum albumin (BSA) as a standard. The eluted protein fractions were pooled into a final concentration of ∼1 mg/ml. The pooled protein was dialyzed into storage buffer (50 mM HEPES [pH 7.5; Sigma-Aldrich], 100 mM NaCl, 25 mM MgCl, 1 mM EDTA, 10% glycerol) with Novagen D-tube dialyzers (molecular mass cutoff, 6 to 8 kDa) and stored at 4°C for a maximum of 2 weeks.

The ability of recombinant CBO0365 protein to bind in vitro to putative promoter sequences of the operons of interest was investigated with nonradioactive EMSAs. To produce 5′-biotin-labeled DNA probes for the EMSA, DNA fragments containing the putative promoter sequences of cbo0364, cbo2525, and cbo3202 and a fragment within the coding sequence of 16S rrn were PCR amplified with Phusion DNA polymerase (Thermo Fisher Scientific) using the 5′-biotin-labeled forward primers and unlabeled reverse primers described in Table 2. In addition, two fragments of the noncoding region upstream of cbo0753 were similarly constructed: a 454-bp fragment containing the entire noncoding region between cbo0752 and cbo0753 and a fragment containing the 180-bp region directly upstream of cbo0753. Cold probes were similarly produced using only unlabeled primers. The PCRs for labeled probes were electrophoresed in nondenaturing polyacrylamide gels for 90 to 120 min and visualized with ethidium bromide staining. Fragments of the desired size were excised from the gels with a sterile scalpel blade. The gel slices were destained in nuclease-free water twice for 20 min each time and then crushed, and DNA was eluted overnight at 37°C into 3 volumes of gel extraction buffer (300 mM sodium acetate, 1 mM EDTA). The eluted DNA buffer solution was sterile filtered, DNA was ethanol precipitated and dissolved in nuclease-free water. Unlabeled probes were purified with the GeneJET PCR purification kit (Thermo Fisher Scientific). The concentration and purity of the probes was determined with the NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific).

Prior to the DNA-protein binding reactions, 0 to 4 μM His₆-tagged CBO0365 protein was phosphorylated in binding buffer containing 20 mM HEPES (pH 7.9; Sigma-Aldrich), 60 mM KCl, 5 mM MgCl₂, 1 mM EDTA, 1 mM DTT, 0.3 mg of BSA/ml, 5% glycerol, and 50 mM lithium potassium acetyl phosphate (Sigma-Aldrich) for 1 h at 25°C. To test the binding capacity of recombinant CBO0365 to the promoter fragments produced above, 20 fmol of 5′-biotin-labeled DNA probe was added to the reactions, and binding was allowed to proceed for 30 min at 25°C. The reactions were loaded into precast 5% native polyacrylamide gels (Bio-Rad, Inc., Hercules, CA) and run at 115 V for 90 to 160 min in prechilled 0.5× Tris-borate-EDTA (TBE) buffer at 4°C. The DNA fragments were transferred to a positively charged nylon membrane (Roche Applied Science, Indianapolis, IN) with a Bio-Rad Trans-Blot electrotransfer apparatus (Bio-Rad) at 100 V for 30 min in 0.5× TBE buffer at 4°C. The membrane detection was performed with Pierce chemiluminescent nucleic acid detection module (Thermo Fisher Scientific) according to the manufacturer's instructions and imaged by exposure to chemiluminescent film (Amersham HyperFilm ECL; GE Healthcare). Control reactions without acetyl phosphate were similarly performed, and additional reactions to determine the specificity of each DNA-protein interaction were performed with a 200-fold molar excess of unlabeled DNA probe added to reactions with 4 μM protein.

Microarray accession number.

The microarray data have been deposited in the NCBI Gene Expression Omnibus (GEO) under accession number GSE26587.

RESULTS

Identification of genes in the putative CBO0365 regulon.

To identify the CDSs under putative regulation of the CBO0365 response regulator, we compared the transcriptomes of the wild-type ATCC 3502 and cbo0365 mutant strains during growth at 37°C and 1 h after a temperature downshift to 15°C using DNA microarrays based on the ATCC 3502 genome. Genes expressed significantly less or more (median log₂ ratios of <−2.0 or >2.0, an FDR of <0.05) in the cbo0365 mutant than in the wild-type strain were considered to be directly or indirectly positively or negatively, respectively, regulated by CBO0365.

At 37°C, a total of 150 chromosomal CDSs showed significantly different expression between the cbo0365 mutant and the wild-type strain. One hour after temperature downshift to 15°C, the number of differentially expressed chromosomal CDSs was 141. All chromosomal CDSs markedly (log₂ ratio of <−2.0 or >2.0) less or more expressed in the cbo0365 mutant than in the wild type at only one temperature were similarly less or more expressed at the other (FDR < 0.05), albeit with a log₂ expression difference falling outside the defined cutoff values. Hence, the putative CBO0365 regulon was practically not affected by temperature (Table 3).

TABLE 3.

Genes of C. botulinum ATCC 3502 with significant differences in expression between cbo0365 insertional mutant and wild-type strains in early logarithmic growth at 37°C and 1 h after a temperature downshift to 15°C^a

Functional classificationb	CDS	Predicted protein product(s)/genomic context	Log2 fold changec
			37°C	15°C
Chemotaxis and mobility	cbo0719	Methyl-accepting chemotaxis protein	2.5	2.3
	cbo2226-cbo2227	Flagellar motor protein	3.3 to 3.7	3.2 to 3.4
Detoxification	cbo0753-cbo0757	Arsenical resistance	2.3 to 3.0	2.4 to 3.0
Drug/analog sensitivity	cbo0072	Putative Na+-driven multidrug efflux pump	2.2	2.1
Transport and binding	cbo2521-cbo2525	Phosphate binding and transport	3.4 to 5.5	3.6 to 4.7
Adaptation	cbo0550	Carbon starvation protein A	2.1	1.8
Carbohydrate metabolism	cbo0880-cbo0883	Glucoside uptake	2.2 to 3.7	2.5 to 3.6
	cbo3169	ABC transporter, periplasmic binding protein	2.8	3.1
Sporulation	cbo0069	Spore maturation protein A	2.6	2.5
Degradation of proteins	cbo0344	Probable metallopeptidase	–1.2	–2.0
Biosynthesis of cofactors	cbo0422-cbo0425	Biosynthesis of pantothenate	3.0 to 4.4	3.1 to 4.1
Electron transport	cbo3197-cbo3199	ABE fermentation	–2.9 to −3.3	–3.2 to −3.8
	cbo3200-cbo3202	ABE fermentation	–3.0 to −3.7	–3.2 to −4.0
Fatty acid biosynthesis	cbo0502	3-Oxoacyl-[acyl-carrier protein] reductase	2.1	2.5
Cell envelope	cbo0715	Membrane protein	2.4	2.3
	cbo2101	Putative UDP-glucose epimerase	1.6	2.5
	cbo2309	Membrane protein	2.0	1.7
	cbo2804	Putative exported protein	2.1	1.9
	cbo2929	Putative exported protein	1.8	2.0
	cbo2937	Membrane protein	–2.7	–2.3
	cbo2972	Membrane protein	2.7	2.4
	cbo3016	N-Acetylmuramoyl-l-alanine amidase	2.3	1.7
	cbo3222	Putative secreted protein	1.8	2.6
	cbo3352	Membrane protein	2.4	2.0
Regulation	cbo0043	Putative RNA polymerase sigma factor	–2.5	–2.7
	cbo0430	TetR family transcriptional regulator	2.1	1.9
	cbo1924	Two-component response regulator	2.7	2.8
	cbo2034	GntR family transcriptional regulator	3.1	2.7
	cbo2164	AraC family transcriptional regulator	2.2	2.1
	cbo2527	Two-component response regulator	1.7	2.0
Extrachromosomal elements	cbo0800	Putative transposase	2.3	3.2
	cbo1681	Prophage 1 protein	2.3	2.3
	cbo1724, cbo1727-cbo1728, cbo1730-cbo1735, cbo1737, cbo1739-cbo1741, cbo1743, cbo1745, cbo1750	Prophage 1	–2.1 to −2.7	–1.4 to −3.0
	cbo2325-cbo2337, cbo2339-cbo2356, cbo2358-cbo2360, cbo2367-cbo2372, cbo2375-cbo2378, cbo2383-cbo2384, cbo2389	Prophage 2	–1.8 to −5.7	–1.9 to −6.2
	cbo3019-cbo3040, cbo3042-cbo3044	Putative genomic island	2.1 to 3.4	1.7 to 3.4
Hypothetical proteins	cbo0043A	Conserved hypothetical protein	–4.1	–5.2
	cbo0257	Putative isochorismatase	2.1	2.3
	cbo0427	Hypothetical protein	3.2	3.5
	cbo0641-cbo0642	Conserved hypothetical protein	1.6 to 2.3	2.3 to 2.8
	cbo0716	Conserved hypothetical protein	1.9	2.1
	cbo1247	Hypothetical protein	2.3	2.0
	cbo2310	Conserved hypothetical protein	2.7	3.4
	cbo2528	Conserved hypothetical protein	1.7	2.2
	cbo2693	Hypothetical protein	2.2	1.8
	cbo2936	Conserved hypothetical protein	–3.3	–3.1
	cbo3211	Conserved hypothetical protein	3.1	2.3
Not classified	cbo2183	Putative phenazine biosynthesis-like protein	2.3	2.0
	cbo2850-cbo2852	Dehydrogenases	5.1 to 5.6	4.2 to 5.3
	cbo3210	Putative phosphoesterase	2.9	2.7
	cbo3571	Nitroreductase (pseudogene)	2.0	2.1

^aGenes of C. botulinum ATCC 3502 with significant (FDR < 0.05) difference in expression (>2.0 or <−2.0 log₂ fold) between cbo0365 insertional mutant and wild-type strain in early logarithmic growth at 37°C and 1 h after temperature downshift to 15°C. FDR, false discovery rate.

^bAccording to Riley (55).

^ccbo0365 mutant versus ATCC 3502 wild type.

The chromosomal CDSs less expressed in the cbo0365 mutant than in the wild type both at 37 and 15°C were arranged in six loci and included cbo0043 encoding a putative RNA polymerase sigma factor, cbo0344 encoding a metallopeptidase, cbo3197-cbo3199 and cbo3200-cbo3202 related to acetone-butanol-ethanol (ABE) fermentation, 16 CDSs related to prophage 1 (the entire prophage is cbo1679 to cbo1755 [22]), and 49 CDSs related to prophage 2 (the entire prophage is cbo2313 to cbo2394 [22]) (Table 3).

The chromosomal CDSs significantly more highly expressed in the cbo0365 mutant than in the wild type at 37 and 15°C were arranged in 32 loci and included those with predicted functions in sporulation (cbo0069), multidrug resistance (cbo0072), pantothenate biosynthesis (cbo0422 to cbo0425), fatty acid biosynthesis (cbo0502), arsenic resistance (cbo0753 to cbo0756), glucoside uptake (cbo0880 to cbo0883), flagellar rotation (cbo2226 and cbo2227), phosphate binding and transport (cbo2521 to cbo2525), dehydrogenation (cbo2850 to cbo2852), and a putative genomic island (cbo3016 to cbo3044) (Table 3).

Validation of the DNA microarray results.

To validate the DNA microarray expression data, we performed a relative gene expression analysis using RT-qPCR of the ABE fermentation pathway genes cbo1407, cbo2847, cbo3199, and cbo3202, the arsenical resistance operon genes cbo0751 and cbo0753, putative flagellar protein-encoding cbo2226 and cbo2227, and the phosphate ABC transporter substrate binding protein-encoding cbo2525, in the ATCC 3502 wild type and cbo0365 mutant 1 h after temperature downshift to 15°C. An R² correlation value of 0.98 was observed between the log₂ expression ratios of the microarray and RT-qPCR experiments (Fig. 1), although the microarray experiment appeared to slightly underestimate the largest expression differences. The higher expression differences suggested by the RT-qPCR experiments than by the DNA microarray analysis are probably attributed to the different normalization procedures used in the two approaches. In addition, the validation results supported the use of the cutoff log₂ ratios of <−2.0 and >2.0 in the microarray experiment, since the most prominent discrepancies in the expression ratios between DNA microarray and RT-qPCR experiments were observed in genes with expression changes under these cutoff values (cbo0751, cbo1407, and cbo2847 [Fig. 1]).

FIG 1.

Confirmation of DNA microarray results with quantitative reverse transcription-PCR (RT-qPCR). Correlation of log₂ fold changes in expression of cbo0751, cbo0753, cbo1407, cbo2226, cbo2227, cbo2525, cbo2847, cbo3199, and cbo3202 between C. botulinum ATCC 3502 cbo0365 mutant and the wild-type strain 1 h after a temperature downshift from 37 to 15°C observed in the DNA microarray (x axis) and RT-qPCR (y axis) experiments.

Inactivation of putative CBO0365 regulon genes results in cold-sensitive phenotypes.

To test whether the metabolic pathways with components under putative transcriptional control of CBO0365 have roles in cold tolerance of C. botulinum ATCC 3502, insertional inactivation of several genes related to these metabolic events was performed. We inactivated the ABE fermentation pathway-related cbo1407 (bdh), cbo2847 (ctfA), cbo3199 (bcd), and cbo3202 (crt) encoding NADH-dependent butanol dehydrogenase, butyrate-acetoacetate coenzyme A (CoA)-transferase subunit A, butyryl-CoA dehydrogenase, and 3-hydroxybutyryl-CoA dehydratase, respectively, the arsenic resistance-related cbo0751 (arsC) and cbo0753 (arsR), and cbo2525 (phoT) encoding a phosphate ABC transporter substrate-binding protein and tested the growth of these mutants at low temperature. Inactivation of cbo3202 resulted in complete inability to grow at 17°C (Fig. 2A). A slight decrease in the growth rate of the cbo3199 mutant at 17°C was also observed (Fig. 2A). The cbo1407 and cbo2847 mutants both had a markedly lower growth rate at 17°C than the wild-type strain (Fig. 2B). The putative ABE fermentation pathway components of C. botulinum ATCC 3502 and the effect of inactivation of related genes on cold tolerance are summarized in Fig. 3.

FIG 2.

Mutants of several genes under putative regulation of CBO0365 and within related metabolic pathways show impaired growth at low temperature. (A to D) Average growth of C. botulinum ATCC 3502 wild type (WT) and mutants with insertionally inactivated cbo3199 (bcd) and cbo3202 (crt) encoding two central enzymes of the ABE fermentation pathway (A), cbo1407 (bdh) and cbo2847 (ctfA) encoding components of the ABE pathway (B), arsenical resistance operon components cbo0751 (arsC) and cbo0753 (arsR) (C), and cbo2525 (phoT) encoding a phosphate ABC transporter (D) at 17°C. Error bars denote the minimum and maximum values of five biological replicates.

FIG 3.

Putative ABE fermentation pathway of C. botulinum ATCC 3502 and effect of inactivation of related components on cold tolerance. The genes putatively encoding the pathway components are inferred from homology to C. acetobutylicum ATCC 824. No homologues for C. acetobutylicum acetoacetate decarboxylase adc were discovered in the ATCC 3502 genome, suggesting inability of C. botulinum ATCC 3502 to produce acetone. Acidogenic reactions are presented with dashed, and solventogenic reactions with solid arrows. Bold arrows represent the central reactions of the pathway. The existence of reactions presented with red arrows is uncertain in C. botulinum ATCC 3502 due to missing acetone production pathway genes. Insertional inactivation of genes crt, bcd, ctfA, and bdh (blue background) resulted in deteriorated growth at 17°C.

Markedly impaired growth was observed for the cbo0751 mutant at 17°C, and inactivation of cbo0753 completely abolished growth at this temperature (Fig. 2C), suggesting an important role for these arsenic resistance protein-encoding genes in cold tolerance of C. botulinum ATCC 3502. Similarly, inactivation of cbo2525 resulted in a phenotype almost completely unable to grow at 17°C (Fig. 2D). No differences in growth at optimal conditions (37°C) were observed between the wild-type strain and any of the mutant strains investigated here (data not shown).

cbo0365 locus.

The structure of the cbo0365 locus suggests that cbo0365, encoding a putative response regulator, and cbo0366, encoding a putative sensor histidine kinase, are cotranscribed together with cbo0364. The 3′ end of the cbo0364 coding sequence overlaps with 58 bases of the 5′ end of the coding sequence of cbo0365, and the 3′ end of the cbo0365 coding sequence overlaps with 11 bases of the 5′ end of the coding sequence of cbo0366. Transcription of the three genes into the same mRNA, independent of cbo0363, was confirmed by RT-PCR analysis of RNA extracted from the ATCC 3502 wild-type cells. Extracted RNA was reverse transcribed with a gene-specific primer targeted to the 3′ end of cbo0366. PCR was performed using primer pairs (Table 2) targeted to the intergenic region between cbo0363 and cbo0364, the CDS of cbo0364, the CDS of cbo0365, or to the intergenic region between cbo0365 and cbo0366 (Fig. 4). No products were obtained with primers targeted to the intergenic region between cbo0366 and cbo0367 from cDNA synthesized with random hexamers, confirming that cbo0367 was not included in the same transcript (data not shown). In addition, no-RT controls failed to yield amplification products, indicating that there was no DNA contamination in the RNA samples (Fig. 4).

FIG 4.

RT-PCR was performed to show the transcriptional arrangement of the cbo0365 locus. PCR primer pairs were targeted to cbo0363-cbo0364 (A), cbo0364 alone (B), cbo0365 alone (C), and cbo0365-cbo0366 (D). Templates included cDNA from ATCC 3502 wild-type culture RNA synthesized with a primer targeted to the 3′ end of cbo0366 (lanes 1), no-RT control of RNA from ATCC 3502 wild-type culture (lanes 2), positive control; genomic DNA of ATCC 3502 wild type (lanes 3), and water (lanes 4). M, DNA molecular weight marker VIII (Roche Applied Science).

A BLAST search against sequenced C. botulinum genomes other than ATCC 3502 suggested cbo0366/cbo0365 to be highly conserved among group I C. botulinum but also to share 43 to 44% (cbo0366) to 66% (cbo0365) amino acid similarity with a TCS present in three group II C. botulinum strains. The CDS of cbo0364 encodes a protein predicted to harbor six transmembrane helices and a type 2 phosphatidic acid phosphatase (PAP2) superfamily domain (NCBI Conserved domain database [http://www.ncbi.nlm.nih.gov/cdd]). However, although a protein with 41 to 43% amino acid identity could be identified in B. cereus, the function of CBO0364 is unknown. A possible function as an auxiliary phosphatase fine-tuning the CBO0366/CBO0365 phosphorelay can be speculated for CBO0364, based on the predicted functional domain and its transcriptional association with cbo0365 and cbo0366.

Recombinant CBO0365 protein binds to putative promoter regions of cbo0364, cbo3202 (crt), cbo0753 (arsR), and cbo2525 (phoT).

To confirm direct CBO0365-mediated regulation of genes and/or operons with markedly affected expression in the cbo0365 mutant, we purified recombinant His₆-tagged CBO0365 protein and tested its ability to bind to the putative promoter regions of these operons. In vitro phosphorylated CBO0365 was shown to bind to a 150-bp fragment derived from the noncoding DNA region directly upstream of cbo0364, thus suggesting CBO0365 to directly control the transcription of its own operon (Fig. 5A), a phenomenon commonly observed with TCSs (30). Furthermore, binding to the promoter regions of cbo3202 (Fig. 5B), to cbo2525 (Fig. 5C), and to the full-length noncoding region between cbo0752 and cbo0753 (Fig. 5D) was observed. No binding was observed to a 180-bp fragment derived from sequence directly upstream of cbo0753 (Fig. 5E), suggesting that the binding site for CBO0365 resides in the 274-bp noncoding region directly downstream of cbo0752. In vitro phosphorylation of CBO0365 for 60 min by acetyl phosphate was essential for any DNA-binding activity, and no binding was observed in reactions with nonphosphorylated protein. Specificity of the protein-DNA interactions between CBO0365 and the promoter fragments were confirmed by prevention of signal shifts in the presence of a 200-fold molar excess of unlabeled competitor probe. Furthermore, no binding to a control DNA fragment derived from the coding sequence of 16S rrn was observed (Fig. 5F). These data, along with the markedly higher or lower expression of cbo0753, cbo2525, and cbo3202 in the cbo0365 mutant than in the wild-type strain, suggest direct regulation of related operons by CBO0365 and the ability for the CBO0365 protein to function either as an activator or repressor of transcription.

FIG 5.

Phosphorylated CBO0365 binds in vitro to promoter regions of cbo0364, cbo3202 (crt), cbo0753 (arsR), and cbo2525 (phoT). (A to D) EMSA results obtained with 0 to 4 μM phosphorylated CBO0365 protein showing binding to double-stranded biotin-labeled DNA probes of putative promoter regions of cbo0364 (A), cbo3202 (B), cbo0753 (C), and cbo2525 (D). (E and F) No binding was observed to a short fragment directly upstream of cbo0753 (E) or to a negative-control fragment from the coding region of 16S rrn (F). No DNA-binding activity was observed for nonphosphorylated CBO0365 (A to E). Specificity was confirmed with addition of 200-fold molar excess of nonlabeled competitor probe.

Mutants of cbo0365, cbo0751 (arsC), and cbo0753 (arsR) show impaired resistance to sodium arsenite.

Since disruption of the regulation of the arsenical resistance operon cbo0753-cbo0756 (arsRDAB) in the cold-sensitive cbo0365 mutant was observed, the growth characteristics of the cbo0365, cbo0751 (arsC), and cbo0753 (arsR) mutants in the presence of 0.1 mM sodium arsenite at 37 and 20°C were investigated to gain further information of the functionality of this pathway and its possible relation to cold tolerance. At 37°C, inactivation of cbo0751 encoding the arsenate reductase ArsC, a central detoxifying enzyme, expectedly resulted in a phenotype with sensitivity to sodium arsenite markedly increased from that of the wild-type strain (Fig. 6A). Inactivation of cbo0365 also resulted in significantly deteriorated growth in the presence of 0.1 mM sodium arsenite (Fig. 6A) despite higher expression of the arsenical resistance operon arsRDAB in the mutant. Moreover, inactivation of cbo0753 (arsR), the putative arsenical resistance operon repressor, almost completely abolished growth in the presence of 0.1 mM sodium arsenite (Fig. 6A). At 20°C, addition of 0.1 mM sodium arsenite to the growth medium resulted in complete lack of growth for the cbo0365 mutant (Fig. 6B), whereas the wild-type strain was still able to grow under these conditions (Fig. 6B).

FIG 6.

Mutants of cbo0365, cbo0751 (arsC), and cbo0753 (arsR) show impaired resistance to sodium arsenite. (A and B) Average growth of C. botulinum ATCC 3502 wild type (WT) and mutants with insertionally inactivated cbo0365 or the arsenical resistance operon components cbo0751 (arsC) and cbo0753 (arsR) at 37°C in the presence of 0.1 mM sodium arsenite (A) and of C. botulinum ATCC 3502 wild type (WT) and a mutant with insertionally inactivated cbo0365 at 20°C in the presence of 0.1 mM sodium arsenite (B). Error bars denote the minimum and maximum values of three (A) or five (B) biological replicates.

DISCUSSION

The two-component signal transduction system CBO0366/CBO0365 plays an important role in the cold shock tolerance and growth of group I C. botulinum strain ATCC 3502 at low temperature (15). In the present study, we sought to gain insight into the mechanisms regulated by the CBO0365 response regulator, through which the cold-sensitive phenotypes of the CBO0366/CBO0365 TCS mutants (15) could be mediated. Since the growth of the cbo0365 mutant at low temperatures was extremely poor (15), the effects of cbo0365 mutation on the transcriptome were investigated in optimal growth conditions and after a temperature downshift. By comparison of the RT-qPCR-confirmed transcriptomic profiles of the cbo0365 mutant and the ATCC 3502 wild type at 37 and 15°C, several genes and operons with markedly affected expression in the mutant were discovered. These genes and operons were considered to be directly or indirectly, positively or negatively, regulated by CBO0365. Mutational analysis of the identified genes and components of related metabolic pathways suggested novel roles for acetone-butanol-ethanol (ABE) fermentation, resistance to arsenic, and phosphate uptake-related genes in cold tolerance of C. botulinum ATCC 3502. The presence of direct regulatory links was confirmed by showing recombinant CBO0365 to bind in vitro to several putative target gene promoters.

The genome of C. botulinum ATCC 3502 harbors genes highly similar to the components of the ABE fermentation machinery in the solventogenic Clostridium acetobutylicum ATCC 824, suggesting a solvent-producing capability for group I C. botulinum, with the exception for acetone production (22). Among the CDSs significantly less expressed in the cbo0365 mutant than in the wild-type strain were six genes involved in the formation of butyryl-CoA from acetyl-CoA, thus encoding the central enzymes of the ABE fermentation pathway (31). The locus structure in ATCC 3502 suggests these genes to be arranged in two operons, the first one including cbo3202 (crt) encoding 3-hydroxybutyryl-CoA dehydratase, cbo3201 (hbd) encoding 3-hydroxybutyryl-CoA dehydrogenase, and cbo3200 (thl) encoding acetyl-CoA acetyltransferase. The second putative operon includes cbo3199 (bcd) encoding short-chain specific acyl-CoA dehydrogenase and cbo3198 and cbo3197 encoding electron transfer flavoprotein alpha- and beta subunits (etfAB). An EMSA analysis showed recombinant CBO0365 to bind to the putative promoter of cbo3202, suggesting direct regulation of the cbo3202-cbo3200 operon by the CBO0366/CBO0365 TCS. The disruption of regulation of the ABE fermentation pathway as an explanation for the cold sensitivity of the CBO0366/CBO0365 TCS mutants was further supported by the inability of the cbo3202 mutant to grow at 17°C. Moreover, disruption of cbo3199, putatively encoding an enzyme catalyzing the reaction following the one catalyzed by the product of cbo3202, resulted in a moderate effect on cold tolerance of C. botulinum ATCC 3502.

A common way for bacteria to counter cold-induced decrease in membrane fluidity is to increase membrane fatty acid unsaturation (32, 33). Unsaturated fatty acid (UFA) synthesis has been thoroughly characterized in Escherichia coli (32, 34); however, the mechanisms for UFA synthesis and particularly the enzymes catalyzing the isomerization reaction responsible for diverting the unsaturated carbon bond into the growing acyl chain remain unidentified in clostridia (35). Group II C. botulinum has been shown to increase the unsaturated fatty acid content of its lipid membrane at low temperature (36), but the mechanisms for such adjustments are unclear. The cbo3202-encoded enzyme belongs to the crotonase superfamily, which harbors enzymes with diverse functions related to acyl-acyl carrier protein and acyl-CoA modifications—the central steps in lipid biosynthesis (37). Thus, a possible explanation for the markedly cold-sensitive phenotype exhibited by the cbo3202 mutant is that the putatively cbo3202-encoded 3-hydroxybutyryl-CoA dehydratase possesses an alternative function in fatty acid synthesis.

Another means for decreasing the membrane lipid melting point is to increase the proportion of branched-chain fatty acids (BCFA), especially the anteiso-BCFA (32, 33). The synthesis of BCFA is described to be initiated with branched-chain acyl-CoA primers, such as 2-methylbutyryl-CoA, produced from the branched-chain α-keto acids by the branched-chain α-keto acid dehydrogenase Bkd in several Gram-positive bacteria (38). The utilization of branched-chain α-keto acids and the presence of BCFA in C. botulinum and in the closely related Clostridium sporogenes have been previously reported, although some controversy on the results exists (39–42). No homologues for genes encoding the Bkd enzymes can be detected in the genomes of either of these clostridial species, suggesting alternate, as-yet-uncharacterized mechanisms for BCFA primer production in clostridia. Hypothetically, conversion of the central ABE fermentation acyl-CoA intermediates into structurally closely related BCFA primers could serve as means to initiate BCFA synthesis in clostridia. Thus, the cold-sensitive phenotype observed in the cbo3202 and cbo3199 mutants could be attributed to the lack of putative BCFA synthesis precursors. In conclusion, the central ABE fermentation intermediates and the enzymes catalyzing their synthesis appear to have an important role in cold tolerance of C. botulinum and should be further investigated in attempts to elucidate the mechanisms of fatty acid biosynthesis in clostridia.

To further characterize the role of the ABE fermentation pathway components in cold tolerance of C. botulinum ATCC 3502, mutants of cbo1407 and cbo2847 putatively encoding butanol dehydrogenase and CoA-transferase A, respectively, were constructed. The putatively cbo1407-encoded butanol dehydrogenase catalyzes aldehyde-alcohol transformation, ultimately resulting in solvent (butanol and/or ethanol) production (31). Mutation of cbo1407 resulted in slower growth at 17°C, suggesting the importance for solvent formation at low temperature. One of the important toxic effects of butanol accumulation in solventogenic clostridia is the direct fluidization of the lipid membranes (43). Since the strictly anaerobic C. botulinum lacks the lipid desaturase system increasing cold-induced O₂-dependent membrane fluidity in B. subtilis (44, 45), other strategies to rapidly counter membrane solubilization at temperature downshift are possibly present. The importance of intact solventogenic mechanisms at a low temperature could be attributed to the lipid-solubilizing effect of solvents and subsequently to a reduced efficiency of rapid membrane adaptation in mutants with impaired solventogenesis.

The cbo2847 encoding CoA-transferase A is related to acid reassimilation after a switch to solventogenesis in C. acetobutylicum and coupled to acetone production with acetoacetate decarboxylase (46). However, since the acetoacetate decarboxylase-encoding adc is missing from the genome of ATCC 3502, the functionality of the CoA-transferase A remains to be confirmed in C. botulinum. Nevertheless, inactivation of cbo2847 resulted in a cold-sensitive phenotype similar to the cbo1407 mutant, suggesting a role for this enzyme in tolerance to low temperature in C. botulinum ATCC 3502.

To the authors' knowledge, the ABE fermentation pathway of C. botulinum has not been characterized, probably because the extreme toxicity of most laboratory strains and the unavailability of nontoxigenic surrogates restricts the usage of available research equipment and facilities. Thus, the ultimate effects of mutations in the ABE fermentation pathway on the metabolism of C. botulinum and the impact of altered solvent production in adaptation to low temperature remain to be characterized.

The expression levels of cbo0753-cbo0756 encoding components of the putative arsenical resistance operon were found to be significantly higher in the cbo0365 mutant strain than in the wild-type strain at both 37 and 15°C. Furthermore, recombinant CBO0365 was shown to bind to the promoter region of cbo0753 (arsR), suggesting direct transcriptional control the of the ars operon by CBO0365. In our previous studies, an association between arsenic resistance and robustness toward low temperature was observed in C. botulinum (47, 48). The Nordic (group I) C. botulinum type B strains form two distinct clusters BI and BII (49, 50), which differ in their genomic content and, in particular, in their arsenical resistance operon structure (47). The lack of arsenic resistance-encoding genes in the cluster BI strains was consistent with significantly decreased robustness to the presence of sodium arsenite (47) or low temperature (48). In contrast, the cluster BII strains harboring a full complement of the arsenic resistance machinery showed tolerance to sodium arsenite (47) and to the lower end of the growth temperature range (48), with a 3.3°C lower minimum growth temperature than that of the arsenic-sensitive cluster BI strains. These findings were supported by the impaired or completely abolished growth of the cbo0751 (arsC) and cbo0753 (arsR) mutants, respectively, at 17°C compared to the wild-type strain. These data suggest a novel important role for an intact arsenical resistance operon exerting cold tolerance of C. botulinum. A possibility for interconnections between mechanisms for arsenic tolerance and oxidative stress response could be speculated as the factor causing the cold-sensitive phenotype of mutants in which the metal-sensing arsenic resistance mechanisms are disturbed. Strict regulation of metal homeostasis is crucial in countering oxidative stress, which has been demonstrated to arise second from cold stress (4).

ArsR is a trans-acting repressor that, together with ArsD, forms a balanced regulatory circuit that fine-tunes the transcription levels of the ars operon (51). Such a delicate regulation and ultimately arsenic resistance are prone to be imbalanced by disruption of any of the related regulators. Indeed, the inactivation of arsC, encoding the central detoxifying enzyme arsenate reductase, resulted in impaired tolerance to sodium arsenite. Moreover, the cbo0365 and arsR mutants were both arsenic sensitive, the latter presenting almost completely abolished growth in the presence of 0.1 mM sodium arsenite, whereas the wild-type strain was expectedly able to grow under these conditions (47). The growth defect was even more pronounced when the arsenic stress was combined with cold stress: at 20°C, the presence of sodium arsenite completely abolished the growth of the cbo0365 mutant, whereas the mild cold stress alone was previously observed to allow growth, albeit with a significantly deteriorated rate (15). These data support a role for an intact arsenical resistance operon and its undisturbed regulation in robustness to low temperature in group I C. botulinum.

Arsenate uptake into cells has been shown to be facilitated by the pho operon-encoded phosphate transport system (52). Among the operons significantly more expressed in the cbo0365 mutant than in the wild-type strain was cbo2521-cbo2525 (pho) putatively encoding a phosphate uptake system. To characterize the potential role of this cellular process in cold tolerance, cbo2525 (pstS) putatively encoding a phosphate uptake system protein was inactivated. A role for this mechanism in cold tolerance of C. botulinum was demonstrated by the deteriorated growth of the cbo2525 mutant at 17°C. The similar expression differences observed for both pho and ars operons suggest a regulatory link between the phosphate uptake and arsenic detoxifying mechanisms. This hypothesis is further supported by the EMSA result showing binding of CBO0365 to putative promoters of both operons and thus direct control of expression by the CBO0366/CBO0365 TCS. How phosphate uptake and arsenic transport and detoxification ultimately are related to clostridial cold tolerance remains unknown.

Inactivation of the cold adaptation-related cbo0365 response regulator markedly affected the global gene expression pattern of C. botulinum ATCC 3502 both at 37 and at 15°C. Characterization of these transcriptional differences suggests previously uncharacterized roles for the ABE fermentation, arsenic resistance, and phosphate uptake mechanisms in cold tolerance of C. botulinum ATCC 3502. Moreover, the results provide an explanation for the cold-sensitive phenotype exhibited by the cbo0366 and cbo0365 mutants (15). Several of the identified mechanisms were shown to be under the direct transcriptional control of the CBO0366/CBO0365 TCS. The genetic responses of C. botulinum to environmental stress are poorly understood and, apart from those with demonstrated roles in cold tolerance (14, 15) or toxigenesis (20, 53, 54), the roles of TCSs in this pathogen are largely unknown. Identification of the key mechanisms behind growth at low temperature provides novel approaches to control the food safety hazards this notorious pathogen causes in food processing.