Role of DEAD-box RNA helicase genes in the growth of Yersinia pseudotuberculosis IP32953 under cold, pH, osmotic, ethanol and oxidative stresses

Yersinia pseudotuberculosis is an important foodborne pathogen threatening modern food safety due to its ability to survive and grow at low temperatures. DEAD-box RNA helicase CsdA has been shown to play an important role in the low-temperature growth of psychrotrophic Y. pseudotuberculosis. A total of five DEAD-box RNA helicase genes (rhlB, csdA, rhlE, dbpA, srmB) have been identified in Y. pseudotuberculosis IP32953. However, their role in various stress conditions used in food production is unclear. We studied the involvement of the DEAD-box RNA helicase-encoding genes in the cold tolerance of Y. pseudotuberculosis IP32953 using quantitative real-time reverse transcription (RT-qPCR) and mutational analysis. Quantitative RT-PCR revealed that mRNA transcriptional levels of csdA, rhlE, dbpA and srmB were significantly higher after cold shock at 3°C compared to non-shocked culture at 28°C, suggesting the involvement of these four genes in cold shock response at the transcriptional level. The deletion of csdA ceased growth, while the deletion of dbpA or srmB significantly impaired growth at 3°C, suggesting the requirement of these three genes in Y. pseudotuberculosis at low temperatures. Growth of each DEAD-box RNA helicase mutant was also investigated under pH, osmotic, ethanol and oxidative stress conditions. The five helicase-encoding genes did not play major roles in the growth of Y. pseudotuberculosis IP32953 under pH, osmotic, ethanol or oxidative stress.


Introduction
Yersinia pseudotuberculosis is a psychrotrophic foodborne pathogen that can cause yersiniosis, with symptoms including abdominal pain, fever and post-infectious systematic complications such as reactive arthritis [1]. It can grow at temperature as low as 0˚C [2], making it a threat to the safety of foods stored under chilled condition. a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 Bacteria use a variety of mechanisms to adapt to temperature downshifts. The increase of unsaturated or branched fatty acids in the membrane lipids of Escherichia coli [3,4], Bacillus subtilis [5] and Listeria monocytogenes [6,7] hinders phase separation of cell membrane phospholipids at low temperatures. Moreover, the cold-induced proteins in E. coli, including the cold shock protein family [8], DEAD-box RNA helicase CsdA [9], exoribonucleases PNPase and RNase R [10], facilitate efficient transcription and translation at low temperatures.
DEAD-box proteins are members in superfamily 2 of the helicase proteins [11,12]. These proteins are important in bacterial RNA metabolism due to their duplex RNA unwinding activity, coupled with ATP hydrolysis or ATP binding [13][14][15]. They participate in crucial biological processes including mRNA decay, ribosome assembly and translation initiation [15][16][17]. DEAD-box proteins are characterized by the presence of nine conserved motifs [11]. Despite highly conserved amino acid sequences, the number of DEAD-box protein-encoding genes in the genome sequence varies greatly in various species [18]. Y. pseudotuberculosis IP32953 is predicted to encode five DEAD-box RNA helicase genes, yptb0165 (rhlB), yptb0486 (csdA), yptb1214 (rhlE), yptb1652 (dbpA) and yptb2900 (srmB) [19]. The importance of csdA in the growth of Y. pseudotuberculosis at 3˚C has been shown by Palonen et al. [20] using insertion mutagenesis. Moreover, many DEAD-box proteins are also reportedly involved in the tolerance of other stressors such as heat, pH, oxidative and ethanol stress [21][22][23]. However, the role of various DEAD-box proteins in the cold tolerance of psychrotrophic Y. pseudotuberculosis is still unknown, as is whether DEAD-box RNA helicases are also relevant in the tolerance of other stressors. In the present study, we studied the role of five DEAD-box RNA helicaseencoding genes in Y. pseudotuberculosis under various conditions, including cold, acidic, alkalic, osmotic, ethanol and oxidative stresses.

Bacterial strains and growth conditions
Y. pseudotuberculosis IP32953 (gratefully received from Elisabeth Carniel, Institut Pateur, Paris, France) was used as the original strain for the construction of all mutant strains (S1 Table). E. coli DH5α was used for plasmid construction (S1 Table). Y. pseudotuberculosis IP32953 and plasmid-containing E. coli DH5α strains were grown in Luria-Bertani (LB) (Lennox, Sigma-Aldrich, St. Louis, MO) broth or on LB agar. Y. pseudotuberculosis IP32953 was incubated at 28˚C. E. coli DH5α was incubated with shaking at 30˚C when expressing temperature-sensitive plasmids pKD4 or pCP20, or at 37˚C when expressing pKD46 and pBluescript-tetR (hereafter called pBlue-tetR).

Cold shock, RNA isolation and RT-qPCR
Three replicate cultures were prepared by subculturing of individual colonies of Y. pseudotuberculosis IP32953 in LB broth at 28˚C overnight and diluted (1:100) in fresh LB broth. The diluted cultures were grown at 28˚C to the early logarithmic growth phase and the pre-coldshocked samples (T0) were collected (Fig 1A). One half of each culture was cold-shocked (temperature rapidly decreased from 28˚C to 3˚C) in an ice-water bath and incubated at 3˚C. The other half of each culture (non-shocked culture) was incubated at 28˚C. Samples were collected from the cold-shocked cultures at 30 min (T1) and 3 h (T2) after cold shock (Fig 1A). Ten milliliters of each sample was mixed with 2.5 ml of cold stop solution (900 μl of 99.6% ethanol and 100 μl of phenol [Sigma-Aldrich]) and incubated on ice for 30 min. Samples were centrifuged at 2˚C 5000 × g for 15 min. The cell pellets were stored at -70˚C until RNA isolation.
Total RNA from all samples was isolated using the GeneJET RNA purification kit (Thermo Fisher Scientific, Waltham, MA) followed by DNase digestion using an RNase-free DNA removal kit (Thermo Fisher Scientific) according to manufacturer's instructions. RNA concentration and purity were measured using the NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific). RNA integrity number was measured with an Agilent 2100 Bioanalyzer (Agilent Technologies Inc., Santa Clara, CA) and was at least 9.0 for all samples.  Table). Three independent colonies of IP32953 were grown at 28˚C, exposed to a rapid temperature downshift (cold shock [dash-dotted line]) at 3˚C when the optical density at 600 nm (OD 600 ) reached 0.5-0.7, and sampled for RT-qPCR analysis before cold shock (T0), and 30 min (T1) and 3 h (T2) after cold shock. Non-shocked cultures (solid line) served as controls. Error bars represent minimum and maximum values. (B) Relative mRNA expression levels of DEAD-box RNA helicase genes rhlB, csdA, dbpA, rhlE and srmB in three independent coldshocked (rapid decrease from 28˚C to 3˚C) cultures compared to pre-cold-shocked cultures at each time point (S3 Table). Significant differences (P < 0.05, Student's t-test) in relative expression levels are indicated with an asterisk. Error bars represent the minimum and maximum values of three biological replicates. A total of 900 ng of each RNA sample was used to synthesize cDNA using a DyNAmo cDNA Synthesis Kit (Thermo Scientific) for RT-qPCR according to manufacturer's instructions. The primers for RT-qPCR were designed with Primer-BLAST [24] (S3 Table). The total reaction volume of 20 μl contained 300 ng of random hexamers, 2 μl of M-MuLV RNase H + reverse transcriptase, 4 μl of template cDNA and a 0.5-μM concentration of each primer. The RT-qPCR was performed with a CFX96 Real-Time Detection System (Bio-Rad Laboratories Inc., Hercules, CA). The cycling protocol of Bio-Rad included initial denaturation at 95˚C for seven min, 40 cycles of denaturation at 95˚C for 10 s, annealing at 60˚C for 15 s, and extension at 72˚C for 20 s, with a final extension at 60˚C for one min. Fluorescence data were acquired after each extension step. A melt curve after a temperature upshift from 60˚C to 98˚C (0.5˚C / 5 s) was analysed after each run. The amplification reaction efficiencies for each primer pair were defined with a dilution series of pooled cDNA originating from the reverse transcription replicates of the biological replicates. Each dilution was assayed in triplicate. The reaction efficiency was calculated as 10 À 1 M ð Þ À 1, where M is the slope of the straight line from a semilogarithmic plot of the quantification cycle (Cq) as a function of the cDNA concentration using Biorad CFX Manager software (Bio-Rad Laboratories Inc.). Reaction efficiencies varied between 0.90 and 1.03.
The relative expression levels of the five helicase-encoding genes were calculated as described earlier by Palonen et al. (2011). Briefly, the expression ratio (R) for each target gene [25]. In this equation, E target is the amplification efficiency of the DEAD-box RNA helicase-encoding gene transcript, E 16Srrn is the amplification reaction efficiency of 16S rrn transcripts, ΔC q,target is the C q deviation between calibrator and sample for the DEAD-box RNA helicase-encoding gene transcript and ΔC q,16Srrn is the C q deviation between calibrator and sample for the 16S rrn transcripts. The cold-shocked samples collected at T1 and T2 were compared to the T0 sample.

Construction of mutant strains
All five DEAD-box RNA helicase-encoding gene deletion mutants (S1 Table) were constructed as described by Datsenko and Wanner with slight modification [26]. Briefly, a competent Y. pseudotuberculosis IP32953 wild-type strain was transformed with the λ Red expression plasmid pKD46 (S1 Table). The transformants were grown in ampicillin-containing LB broth, induced with 10% of arabinose and made competent for electroporation as described previously [27]. Linear PCR products were made on the template of kanamycin-resistance cassette flanked by FLP recognition target (FRT) sequences from pKD4 plasmid. The primers (S2 Table) were designed to contain 59 nucleotides from the flanking region of the sequences to be deleted from the IP32953 strain. A 40-˚l aliquot of pKD46-containing competent cells was mixed with 400 ng of the linear PCR product in an ice-cold 0.2-cm cuvette (Sigma-Aldrich). Electroporation was performed at 2.5 kV with 25 mF and 200 O, immediately followed by addition of 1 ml super optimal broth with catabolite repression medium (Sigma-Aldrich) and incubated at 28˚C. The culture was evenly spread on kanamycin-containing LB plates after three hours of incubation. Kanamycin-resistant recombinants were screened by colony PCR with primers k1, k2 and up-/down-stream primers for target genes (S4 Table). The confirmed recombinants were transformed with FLP expression plasmid pCP20 to excise the kanamycin resistance marker (S1 Table) and grown on an LB medium at 37˚C to promote curing of the pkD46 and pCP20 plasmids. The final deletion was confirmed by colony PCR with up-/downstream primers for target genes (S4 Table).

Complementation of mutant strains
The pBluescript-tetR plasmid [28] was used as a vector. The promoter region was predicted by analysing the 500-bp sequence upstream of each DEAD-box RNA helicase target gene with BPROM software (Softberry Inc., Mount Kisco, NY). The coding sequences and native promoters of dbpA and srmB were amplified by PCR using the primers com yptb1652 NotI, com yptb1652 XhoI, com yptb2900 NotI and com yptb2900 XhoI, respectively (S4 Table). The resulting PCR products and pBlue-tetR were digested with NotI and XhoI (Thermo Fisher Scientific) and ligated with T4 ligase (New England Biolabs, Ipswich, MA), producing pBlue-tetR-dbpA and pBlue-tetR-srmB. The complementary plasmid pBlue-tetR-csdA was constructed by digesting plasmid pBluescript-csdA [20] with NotI and XhoI, and ligated to pBlue-tetR.

Growth experiment and growth curve analysis
Five replicate cultures were prepared by subculturing of individual colonies of Y. pseudotuberculosis IP32953 wild-type strain, deletion mutant strains and the mutants with complementary plasmids or vectors into 5 ml of LB broth containing 100 μg/ml of ampicillin when appropriate and grown overnight at 28˚C. To study the growth at 3˚C and 28˚C, cultures were diluted into fresh LB broth containing ampicillin when appropriate to OD 600 of around 0.01.
A volume of 300˚l of each dilution was pipetted into wells of microtiter plates. The microtiter plates were tightly sealed to prevent evaporation during the growth. The strains were grown in the Bioscreen C Microbiology Reader (Growth Curves, Helsinki, Finland). The OD 600 levels of the cultures were monitored at 15-min intervals at 28˚C for 72 h, or at 1-h intervals at 3˚C for 720 h. The OD 600 values were ensured to correspond with the number of viable cells by performing bacterial colony counting using Plate Count Agar plates (Oxoid, Thermo Fisher Scientific) from the wild-type strain and from all the mutants grown at 3˚C.
To analyse the defective growth of the deletion mutants, the mean maximum growth rate of each strain at 28˚C and 3˚C were obtained with DMFit (https://www.combase.cc/index.php/ en/), using the model of Baranyi and Roberts [29]. The statistical significances of differences between the fitted maximal growth rates of the mutant strains and the wild-type strain were tested using the Student's t test [30].

Relative expression levels of DEAD-box RNA helicase-encoding genes
In the cold-shocked cultures, the relative expression levels of csdA and rhlE were 3.7-and 3.9-fold upregulated (P < 0.05) at 30 min after cold shock compared to T0 cultures, respectively (Fig 1B). The transcripts of csdA were also 2.2-fold upregulated at 3 h after cold shock compared to T0 cultures.

Growth of DEAD-box RNA helicase gene deletion mutants at 3˚C and 28˚C
Growth of the wild-type Y. pseudotuberculosis IP32953 and DEAD-box RNA helicase gene deletion mutants ΔrhlB, ΔcsdA, ΔrhlE, ΔdbpA and ΔsrmB was monitored at 3˚C and 28˚C (Fig  2). At 3˚C, no observable growth was detected for ΔcsdA until 600 h of incubation. Mutants ΔdbpA and ΔsrmB showed defective growth compared to the wild-type strain. The maximum growth rate of ΔdbpA and ΔsrmB decreased by at least 50% (P < 0.001) compared to the wildtype strain ( Table 1). The defective growth of ΔcsdA, ΔdbpA and ΔsrmB at low temperatures was confirmed with viable cell counting. The viable cell concentration of ΔcsdA was below the detection limit (200 CFU/ml) after five days of incubation until day 18 at 3˚C (Table 2). Three replicates of ΔcsdA grown at 3˚C were inoculated into fresh LB broth at day 18 and incubated at 28˚C for 96 h. The viable cell concentration of these cultures reached an average of 7.28 log10 (CFU/ml). Mutants ΔdbpA and ΔsrmB showed significantly lower (P < 0.001) colony counts compared to the wild-type strain at day five at 3˚C and resumed growth in subsequent time points (Table 2). Mutant ΔrhlE displayed a slightly lower growth rate and maximal OD 600 value compared to the wild type at 3˚C (Fig 2B; Table 1). However, the viable cell concentration of ΔrhlE did not significantly differ from that of the wild type (Table 2).  Table) or at 1-h intervals at 3˚C (S6 Table) At 28˚C, the growth rates of ΔrhlB, ΔrhlE, ΔdbpA and ΔsrmB were similar with the wildtype strain, and slightly lower maximum optical density compared to the wild-type strain was observed in ΔcsdA (Fig 2A). The viable cell concentration of these mutants was also similar with that of wild type during growth at 28˚C, except a minor difference between ΔcsdA and wild type after 11 hours incubation (Table 3).
To verify that the cold-sensitive phenotypes of ΔcsdA, ΔdbpA and ΔsrmB at 3˚C were specifically due to the absence of these genes, the mutant strains were complemented with a wildtype copy of each deleted gene. The complementation restored the growth of ΔcsdA, ΔdbpA and ΔsrmB (S6 Table;

Growth under acidic, alkalic, osmotic, ethanol and oxidative stresses
The wild type Y. pseudotuberculosis IP32953 and DEAD-box RNA helicase gene deletion mutants ΔrhlB, ΔcsdA, ΔrhlE, ΔdbpA and ΔsrmB were also incubated and monitored at pH 4.6, pH 4.8 or pH 9.0, in 4% NaCl, in 3% ethanol and in 10 mM H 2 O 2 (Fig 3). These five mutants showed minor decreases in their growth rates under ethanol stress compared to the wild type (Fig 3; Table 4). None of the five mutants displayed impaired growth rate under pH, salt, and oxidative stresses compared to that of wild type.

Discussion
The ceased growth of ΔcsdA and impaired growth of ΔdbpA and ΔsrmB at 3˚C indicate that the DEAD-box RNA helicase-encoding genes csdA, dbpA and srmB are required for optimal growth of Y. pseudotuberculosis at low temperatures. Gene rhlE only played a minor role in low-temperature growth. The growth-ceased csdA mutant continued to replicate after moving from 3˚C to 28˚C indicating that the bacterial cells were still viable. In addition, slight growth was observed for ΔcsdA after 600 h incubation at 3˚C. It is possible that the bacterium is very slowly adapting to low temperature by replacing the deficiency of the csdA by the enhancement of other cold tolerance mechanisms. Previous studies have shown that protein CsdA participates in RNA degradation and ribosome assembly in E. coli after a temperature downshift [9,31,32]. However, the mechanism of CsdA in bacterial cold tolerance is still unclear. It seems that CsdA is the most important DEAD-box RNA helicase compared to the rest four in the cold tolerance. The  Table). The points represent the median OD 600 values of five independent cultures. Error bars represent minimum and maximum values.
distinctive importance of CsdA might stem from its homodimeric form, which is less frequently observed in DEAD-box helicases compared to monomeric form [33], since the dimerization domain of protein CsdA is responsible for its enzymatic function and structural stability [14]. Genes dbpA and srmB were less important than csdA in the cold tolerance of Y. pseudotuberculosis according to the mutational experiment and transcriptional analysis. In E. coli, SrmB was required for optimal growth and assembly of the 50S ribosomal subunit at low temperature (22˚C) [34,35]. However, deletion of DbpA in E. coli did not cause a cold sensitive phenotype or inappropriately assembled ribosome at low temperatures (< 30˚C) [36,37]. The significance of DbpA may only be observed in psychrotrophic gram-negative strains at refrigerated temperatures. Understanding the mechanisms of the ability of Y. pseudotuberculosis to grow at low temperatures may provide new tools for controlling this foodborne pathogen.
In our study, gene rhlE did not have a notable impact on the cold growth of Y. pseudotuberculosis. The significant transcriptional response under cold shock stress implies that rhlE may function for only a short time after the cell's exposure to cold stress. Protein RhlE may be a backup helicase for CsdA, as overexpression of RhlE could suppress the cold-sensitive ΔcsdA in E. coli [38,39]. However, overexpressed RhlE exacerbates the cold-sensitive phenotype of ΔsrmB [39]. The helicase RhlE may indirectly participate in cold tolerance by adjusting the activities of CsdA and SrmB.
Gene rhlB was not related to the cold tolerance of Y. pseudotuberculosis according to our mutational or transcriptional analysis. Though RhlB, together with RNase E, PNPase and enolase, was identified as part of the degradosome in Y. pseudotuberculosis [40], loss of its encoding gene rhlB did not harm growth in any stress conditions nor the optimal condition in our study. Further studies are required to confirm whether RhlB is needed for Y. pseudotuberculosis growth under other untested stresses.
The five DEAD-box RNA helicase-encoding genes are not essential for Y. pseudotuberculosis growth under acidic, alkalic, osmotic, ethanol or oxidative stresses. All five genes had only minor influence on growth when exposed to ethanol stress. Although DEAD-box proteins have been found to potentially be involved in tolerance to pH, osmotic, ethanol or oxidative stresses in gram-positive B. cereus and L. monocytogenes [22,23], they do not appear to play a major role in these stresses in Y. pseudotuberculosis.

Conclusion
The helicases CsdA, DbpA and SrmB are important for cold growth of Y. pseudotuberculosis. CsdA had the strongest impact, with the deletion mutant showing ceased growth at low