Department of Immunology, Microbiology and Parasitology, University of the Basque Country UPV/EHU, Leioa, Spain
IKERBASQUE, Basque Foundation for Science, Bilbao, Spain
Correspondence: Vladimir R. Kaberdin, Department of Immunology, Microbiology and Parasitology, University of the Basque Country UPV/EHU, Barrio Sarriena S/N, 48940 Leioa, Spain. Tel.: +34 94601 8430; fax: +34 94601 3500; e-mail: firstname.lastname@example.org
The life and survival of the marine bacterium Vibrio harveyi during its adaptation in natural aquatic systems is highly influenced by the availability of nutrients and temperature. To learn about adaptation strategies evolved by this bacterium to cope with drastic temperature downshifts and nutrients depletion, we have studied the phenotypical and gene expression changes occurring in V. harveyi during its adaptation to cold seawater. We found that incubation in cold seawater up to 12 h did not cause any significant morphological changes in V. harveyi and had no effect on the number of viable and culturable cells. Microarray analysis revealed that the V. harveyi response to cold seawater leads to up- and downregulation of numerous genes controlling the central carbon metabolism, nucleotide and amino acid biosynthesis as well as DNA repair. In addition, expression of some genes controlling biosynthesis of lipids, molecular transport, and energy production was altered to likely affect the composition and properties of the V. harveyi cell envelope, thus implying the putative role of this compartment in adaptation to stress. Here, we discuss these results with regard to the putative adaptive responses likely triggered by V. harveyi to cope with environmental challenges in natural aquatic systems.
Marine bacteria including those of the genus Vibrio can act as primary or opportunistic pathogens affecting a large number of sea organisms that serve as a source of seafood worldwide. Recent studies have revealed that vibriosis (Egidius, 1987), a typical disease caused by Vibrio species in shrimps and in other industrially important sea organisms (lobster, oysters, etc.), is gradually becoming one of the major concerns during the ongoing climate changes (Vezzulli et al., 2010). Some types of vibriosis are caused by Vibrio harveyi (Martin et al., 2004; Haldar et al., 2010; Soto-Rodriguez et al., 2010; Zhou et al., 2012), a facultative anaerobic Gram-negative bacterium globally present in marine environments.
Many marine bacteria are highly adaptive and can survive and strive in natural aquatic systems under severe stress conditions including seasonal temperature downshifts and starvation. Moreover, regular trapping of marine microorganisms by ocean/sea currents (both cold and warm) as well as their subsequent release in new locations can also subject microorganisms to sudden environmental changes (including temperature shifts and metabolic stresses) specific for their new habitats. Adaptation of Gram-negative bacteria to low temperature and starvation has been studied previously using different model microorganisms including Vibrio species (Carroll et al., 2001; Yang et al., 2009; Wood & Arias, 2011). Although these studies revealed a number of strategies developed by phylogenetically distant bacteria to resist cold temperatures and limitation of nutrients, they also demonstrated that the putative adaptation mechanisms as well as the genes controlling these mechanisms only partly coincide and can often vary from species to species.
The lack of systematic data related to survival and growth of V. harveyi at low temperatures under limitation of nutrients and the potential value of such data for understanding the ecology of Vibrio species led to the present study aimed at investigating adaptation of V. harveyi to cold seawater. In the course of this study, we examined the dynamics of V. harveyi adaptation by monitoring the number of total and viable cells and their morphology using fluorescent and scanning electron microscopy as well as standard microbiological techniques. In addition, we compared V. harveyi cells at the transcriptome level before and after exposure to stress. Microarray analysis revealed a group of highly up- and downregulated genes controlling several major metabolic pathways and membrane functions. We discuss the putative roles of these genes in sustaining the V. harveyi capacity to survive and maintain cellular functions in natural aquatic systems.
Materials and methods
Vibrio harveyi survival assays and viability tests
Vibrio harveyi strain ATCC 141126 was aerobically grown at 28 °C in marine broth (MB, Panreac). Stationary-phase V. harveyi populations were subjected to starvation and nonoptimal temperature by diluting (1 : 20) overnight V. harveyi cultures with cold (4 °C) natural seawater (from Port of Armintza in the north of Spain, 43°26′24″N and 2°54′24″W). The experiments were carried out in Erlenmeyer flasks cleaned with H2SO4 (97%, v/v) beforehand, rinsed with deionized water, and heated at 250 °C for 24 h to avoid any presence of residual organic substances. Large Erlenmeyer flasks containing 2 L filtered and subsequently autoclaved seawater were inoculated and incubated at 4 °C with shaking (90 rpm). Periodically, samples were collected in triplicate to determine the total number of cells, the numbers of viable and culturable bacteria. The total number of bacteria was determined by filtering bacterial cell populations through 0.22-μm-pore-size polycarbonate membrane filters (Millipore), followed by staining of the attached cells with acridine orange and direct counting individual cells using epifluorescence microscopy (Hobbie et al., 1977). Viable bacteria were estimated as bacteria with intact cytoplasmic membranes (MEMB+). These MEMB+ bacteria were counted with the aid of the Live/Dead BacLight™ kit (Invitrogen) as described by Joux & Lebaron (1997). The bacteria with intact cytoplasmic membranes (green fluorescence, MEMB+) and the permeabilized bacteria (red fluorescence) were enumerated separately. Culturability, expressed as colony-forming units, was evaluated by spreading aliquots on marine agar (MA, Oxoid) followed by incubation for 24 h at 28 °C.
The size measurements of unstressed and stressed cells were performed via image analysis of epifluorescence preparations (Massana et al., 1997) using an image analysis system, which included a video camera of high resolution (Hamamatsu 2400; Hamamatsu Photonics, Hamamatsu City, Japan). Digitized images of microscopic fields were analyzed by Scion Image 1.62ª software (for further details, see Supporting Information, Table S5).
Scanning electron microscopy
Samples of V. harveyi cells unstressed (overnight culture) and stressed (after 12 h incubation in cold seawater) were fixed with 2% glutaraldehyde in 0.1 M sodium phosphate buffer (pH 7.4). The suspensions of the fixed cells were filtered through 0.22-μm-pore-size membrane filters (GTTP filters, Millipore). The filters with the attached V. harveyi cells were further dehydrated by applying increasing series of ethanol (30%, 50%, 70%, 90%, and 100%). The filters with the attached dehydrated V. harveyi cells were overlaid with 1 mL of hexamethyldisilazane, incubated for 5 min, and air-dried. Finally, the samples were coated with gold, and imaging was carried out by analyzing samples in a Hitachi S4800 scanning electron microscope.
RNA isolation and microarray analysis
Vibrio harveyi ATCC 141126 cells were grown in triplicate in marine broth at 28 °C until the culture reached the stationary phase (12–16 h), and aliquots (20 mL) were diluted by sterile 4 °C seawater (1 : 20) and further incubated at 4 °C with shaking (90 rpm). Aliquots of stressed cell cultures withdrawn after 5 min and 12 h of incubation in triplicate were mixed with stop solution (5% phenol in ethanol) at ratio 8 : 1, incubated on ice for 15–20 min, and the cells were collected by centrifugation (15 min, 4 °C, 4400 g). The pelleted V. harveyi cells were used to isolate total RNA using TRIzol reagent and PureRNA mini Kit (Invitrogen) according to the vendor's instructions. Further analysis of RNA samples was carried out at the General Genomic Service (SGIker) of the University of the Basque Country as previously described (Rutherford et al., 2011). Briefly, after verifying RNA quality and integrity by Lab-chip technology on an Agilent 2100 Bioanalyzer with Agilent RNA 6000 Nano Chips, RNA has been retrotranscribed with SuperScript III reverse transcriptase (Invitrogen) and labeled using the SuperScript Indirect cDNA Labeling System (Invitrogen) to incorporate aminomodified nucleotides. After a purification step to remove unincorporated nucleotides, the aminomodified cDNA was coupled to fluorescent dyes (Cy5 or Cy3) and used for hybridization with custom microarray (Rutherford et al., 2011) produced by Agilent Technologies according to the design kindly provided by the laboratory of Prof. Bonnie Bassler (Princeton University). Raw data from Feature Extraction Software (FE processed signals) were subsequently processed on GeneSpring MultiOmic Analysis software 12.0 (Agilent Technologies) and were subjected to further statistical analysis.
Validation of microarray data by quantitative real-time PCR
Total V. harveyi RNA was isolated from three independent cell cultures according to the procedure described in the previous section. Further analysis of RNA samples has been carried out at the General Genomic Service (SGIker) of the University of the Basque Country, Spain. Briefly, after verifying RNA quality and integrity by Lab-chip technology on an Agilent 2100 Bioanalyzer with Agilent RNA 6000 Nano Chips, RNA was used for cDNA synthesis using AffinityScript Multiple Temperature cDNA Synthesis Kit (Agilent Technologies). Aliquots (1 μg each) of each RNA sample were individually reverse-transcribed according to the vendor's instructions. The reaction mixtures containing cDNAs were diluted and used for quantitative PCR amplification in 7900HT Fast Real-Time PCR System (Applied Biosystems) to determine the relative changes in the level of 18 transcripts and the reference transcript (16S rRNA gene) used for normalization. The gene-specific primers used for RT-qPCR are listed in Table S3. The software to control the amplification process and obtain the quantitative PCR data was SDS 2.4 (Applied Biosystems). The Ct values obtained in this step were further used to calculate the relative expression (for details, see Table S4).
The morphology and viability of V. harveyi is not affected during 12 h incubation in cold seawater
Incubation in cold seawater can have potentially adverse effects on morphology and survival of bacteria. We first compared the morphological appearance of unstressed and stressed V. harveyi cells using fluorescent and scanning electron microscopy (Fig. 1). Aliquots of V. harveyi overnight cultures were diluted (1 : 20) by sterile cold seawater from the Bay of Biscay, and the diluted cultures were further incubated for 12 h at 4 °C. Samples of V. harveyi cells taken before and after incubation in seawater were stained using acridine orange (Hobbie et al., 1977) and were further examined by epifluorescent microscopy. The results of this examination (Fig. 1a) demonstrated that during their incubation in cold seawater up to 12 h, V. harveyi cells did not show any significant changes in their morphology and size, when compared to the cells from the original overnight culture. Moreover, the results of three independent experiments obtained by analyzing the size of 150–210 cells (in each experiment) corroborate the initial observation that 12 h of incubation in cold seawater do not change the morphology and size of V. harveyi cells (see Table S5). The lack of phenotypical changes was also supported by images obtained using scanning electron microscopy. The shape and the integrity of unstressed (control overnight culture) and stressed (incubated for 12 h in cold seawater) cells were very similar (Fig. 1b).
In addition, enumeration of the cells present in aliquots taken after 0.5, 1, 2, 3, 4, 5, 6, and 12 h of V. harveyi incubation in cold seawater by fluorescent microscopy revealed that the total number of V. harveyi cells remained essentially the same (Fig. 2). Likewise, the number of cells, which maintained the integrity of cytoplasmic membrane and retained capacity to form colonies on marine agar, remained unchanged.
Gene expression analysis
To learn more about specific adaptation mechanisms triggered by V. harveyi to survive during seasonal changes in temperature and nutrient availability, we compared V. harveyi transcriptome profiles after short and long exposure to cold seawater. V. harveyi cells were grown in marine broth at 28 °C overnight, subsequently diluted with sterile cold seawater (1 : 20), and further incubated at 4 °C. Total RNAs for microarray analysis were isolated after 5-min (control cells) and 12-h (stressed cells) incubation in cold seawater. Comparison of these two gene expression points by hybridizing isolated total V. harveyi RNAs with a custom microarray revealed profound changes that occur at the transcriptome level. A large number of V. harveyi transcripts have showed twofold (or higher) changes in abundance [see Table S1 (downregulated genes) and Table S2 (upregulated genes)].
While analyzing these data, we selected a group of genes that are highly up- or downregulated (see Table 1) and therefore likely play the major role in V. harveyi responses to cold seawater. Apart from several stress-related genes (e.g. genes involved in stringent response) as well as numerous genes coding for the components of the protein-synthesizing (e.g. genes encoding ribosomal proteins and RNA-modifying enzymes), many other selected genes encode transporters, various enzymes involved in biosynthesis and transport of amino acids, nucleotides, lipids, and other essential biomolecules as well as enzymes that function in the central carbon metabolism (e.g. glycolysis, TCA cycle). Further analysis of their possible contribution to V. harveyi capacity to survive in and adapt to cold seawater (see 'Vibrio harveyi survival assays and viability tests'), suggests that the observed changes in the V. harveyi transcriptome likely affect all major cellular functions including metabolism, transport, energy production, and gene expression processes.
Table 1. Vibrio harveyi genes highly up- or downregulated during adaptation to cold seawater
Specific biological pathway
Highly up- and downregulated genes
Locus tag (gene name)
Annotation of the genes and their product is according to the classification used in the kegg database (http://www.genome.jp/). Up- or downregulation of genes is indicated with upward (↑) or downward (↓) arrows, respectively.
BirA family transcriptional regulator, biotin operon repressor/biotin-[acetyl-CoA-carboxylase] ligase
Transcriptional regulator LuxR
Bicyclomycin/multidrug efflux system protein
Multiple antibiotic transporter
Control of Na+/H+ balance
Antisense control by sRNAs
CsrB/RsmB RNA family
Spot 42 RNA
Validation of microarray data by quantitative real-time PCR
To validate the microarray data, the alterations in the level of transcripts (18 in total), selected from the group of highly up- and downregulated genes that control the key biological pathways (Table 1), have also been assessed by quantitative real-time PCR (Table S4). In pilot experiments, we have determined that the level of 16S rRNA gene after 5-min (control cells) and 12-h (stressed cells) incubation in cold seawater remained unaltered (data not shown), and therefore, we used this rRNA as a reference to normalize qRT-PCR data. The analysis has been performed with RNA corresponding to three biological replicates using gene-specific primers (Table S3). As seen in Table 2, the result of the qRT-PCR analysis confirmed up- and downregulation of the above genes, thus corroborating microarray data.
Table 2. Validation of microarray data by quantitative real-time PCR (qRT-PCR)
Locus tag (gene name)
Regulation (12 h vs. 5 min)
Total RNA was isolated in triplicate from V. harveyi cells after 5-min and 12-h incubation in cold seawater, and the differences in the level of individual transcripts determined by qRT-PCR and microarray analysis are presented for highly up- and downregulated transcripts.
Outer membrane protein assembly factor YaeT/Omp85/BamA
Periplasmic protein TonB
Iron transport protein
Although sea microorganisms possess a tremendous potential to adapt and adequately respond to environmental changes, which they face in natural aquatic systems, the specific mechanisms involved in their adaptation are still poorly understood. This is particularly true for some Vibrio pathogens including V. harveyi, an emerging pathogen infecting a large number of sea organisms (e.g. shrimps) important for seafood industry. The survival of Vibrio species in natural aquatic systems can be challenged by numerous environmental factors (e.g. temperature, composition of seawater as well as the presence of symbiotic or predator organisms) potentially limiting their growth and survival (Johnson, 2013).
Here, we investigated adaptation of V. harveyi in cold seawater by monitoring phenotypical and gene expression changes. The experiments were performed on stationary-phase bacteria to study responses of nondividing bacteria to temperature and nutritional stress. We found that the overall shape, size, and integrity of V. harveyi cells have not been affected upon 12 h of incubation in sterile seawater. Moreover, the number of viable and culturable cells remained the same during the time course, which implies that V. harveyi likely elicits specific adaptation mechanisms maintaining its culturability under these stress conditions.
To learn more about the putative nature of these mechanisms, we employed microarray analysis and assessed gene-specific variations in the level of individual transcripts. We found that incubation in cold seawater induces global changes in the V. harveyi gene expression program affecting a large number of transcripts (Table S1 and S2). The observed changes were further analyzed with a major focus on highly up- and downregulated genes (Table 1) and in the context of the action of two major stress factors (i.e. (1) limitation of nutrients/microelements; and (2) low temperature) affecting V. harveyi survival.
In our experimental system, the first factor is partly associated with adaptation to stress caused by a decrease in the concentration of amino acids (partly eliminated when V. harveyi cells were transferred from marine broth to cold seawater) and by limited availability of alternative carbon sources. We found that limitation of nutrients leads to significant downregulation of genes controlling the central carbon metabolism, biosynthesis of lipids, amino acids, and nucleotides (Table 1).
Several downregulated genes that control the central carbon metabolism encode the key enzymes of glycolysis, tricarboxylic acid cycle, and glyoxylate/dicarboxylate metabolism. Moreover, the fine tuning of these metabolic pathways appears to be attained through the action of small regulatory RNAs known for their essential roles in post-transcriptional control of gene expression (Kaberdin & Bläsi, 2006). Among several sRNAs that have been upregulated in V. harveyi (namely CsrB/RsmB; M1 RNA, an RNA component of RNase P, 6S/SsrS RNA, tmRNA, and Spot 42 RNA, see Table S2), two sRNAs (CsrB/RsmB and Spot 42) are specifically known for their role in sugar metabolism. Previous studies of sugar metabolism in bacteria revealed that the first one (CsrB/RsmB) is involved in antagonizing the action of the global regulator CsrA able to positively (or negatively) control various biological pathways including quorum sensing, glycolysis, acetate metabolism, motility, gluconeogenesis, biofilm formation, and others (Babitzke & Romeo, 2007; Timmermans & Van Melderen, 2010). Moreover, the role of CsrB in regulation of quorum sensing in V. cholerae has been demonstrated recently (Lenz et al., 2005). Thus, an increase in the abundance of this small RNA in V. harveyi is likely linked to a regulatory mechanism(s) serving to adjust one (or several) biological pathways affected by the stress conditions used. Even higher degree of upregulation (more than 40-fold) has been observed for Spot 42. This sRNA is highly conserved in bacteria including the Vibrionaceae family (Hansen et al., 2012). Although the precise function of this sRNA in adaptation of V. harveyi to cold seawater remains to be determined, previous characterization of its mode of action in Escherichia coli (Beisel & Storz, 2011) and the nature of the recently predicted targets of this sRNA in other bacteria such as Aliivibrio salmonicida (Hansen et al., 2012) suggest that V. harveyi Spot 42 likely plays a critical role in regulating the central carbon metabolism.
Besides transcriptome changes affecting the central carbon metabolism, there were also significant alterations in the expression level of several genes known for their role in lipid biogenesis (Table 1). A considerable decrease in the level of the fab mRNA encoding the second enzyme of fatty acid biosynthesis (i.e. 3-oxoacyl-ACP synthase) suggests a reduction in de novo synthesis of lipids.
With few exceptions, significant downregulation was also detected for many genes involved in biosynthesis of amino acids. However, in contrast to genes involved in biosynthesis of other amino acids, expression of genes encoding several enzymes of histidine biosynthesis including those involved in purine metabolism was increased dramatically (Table 1). The anticipated reduction in protein synthesis caused by amino acid starvation and temporal arrest of V. harveyi growth is likely linked to co-regulation of other genes known to control protein functionality under normal conditions. Indeed, our microarray data (Table 1) point to a significant decrease in the level of transcripts encoding the main players of chaperone-mediated protein folding (IbpA, GroEL, GroES, and DnaK). This response is reminiscent of a cold-induced reduction in chaperone expression previously observed in V. parahaemoliticus (Yang et al., 2009).
In addition, similar to regulation in other Gram-negative bacteria (Jin et al., 2012), amino acid starvation appeared to trigger the so-called stringent response. We anticipate the occurrence of this response due to the observed increase in the level of mRNA encoding the putative ppGpp synthase (see Table 1). The regulator molecule ppGpp and other factors such as CgtA and DksA have been previously shown to play important roles in V. cholerae adaptation to amino acid starvation and to other stress conditions (Pal et al., 2011) and therefore might also have similar functions in V. harveyi. Moreover, we also observed upregulation of the sspA gene encoding the stringent starvation protein A (SspA). Previous studies in E. coli revealed that this protein is important in E. coli stress responses during stationary phase and when cells face limitation of nutrients (Williams et al., 1994). The E. coli SspA was found to function as transcription factor, and at least in part, it can functionally be substituted by SspA orthologs of other bacteria including the V. cholerae counterpart (Hansen et al., 2005).
The anticipated reduction in metabolic activities and efficiency of gene expression implies that V. harveyi starvation in seawater should inevitably lead to a reduced consumption of energy. Consistently, we observed a considerable decrease in the abundance of transcripts encoding cytochrome c oxidase subunit II and inorganic pyrophosphatase, that is, enzymes involved in energy production (see Tables 1 and 2). In agreement with the downregulation of these genes, our microarray data also revealed a concomitant decrease in the expression level of transcripts encoding the enzymes (namely coproporphyrinogen III oxidase and uroporphyrinogen decarboxylase) involved in the biosynthesis of heme, an important cofactor of energy-generating enzymes during aerobic growth.
Reactions of the central carbon metabolism, amino acid and fatty acid biosynthesis as well as biosynthetic pathways of modified tetrapyrroles (e.g. heme and cobalamin) involve carbon dioxide (CO2) and bicarbonate ion ( ) endogenously produced in bacteria and required for cell growth under normal conditions. Downregulation of several genes (e.g. hemE) encoding CO2-generating enzymes such as uroporphyrinogen, oxaloacetate, and orotidine 5′-phosphate decarboxylases (see Table 1; Tables S1 and S2) likely leads to a decrease in carbon dioxide production. The anticipated reduction in CO2 production is, in turn, consistent with upregulation of birA expression known to negatively control the biosynthesis of biotin, a cofactor of decarboxylases. Moreover, it is also correlated with a strong upregulation of the gene encoding carbon anhydrase, an enzyme facilitating rapid interconversion of carbon dioxide and water to bicarbonate, thus suggesting an important role for this enzyme in maintaining the proper equilibrium in V. harveyi under stress (Merlin et al., 2003).
Apart from numerous genes encoding metabolic enzymes and regulatory proteins whose expression was altered in response to starvation, significant up- and downregulation of many other genes is apparently linked to the function of V. harveyi cell envelope representing the front line of defense against environmental threats in bacteria. Some regulatory mechanisms apparently serve to adjust the composition and physical properties of the cellular envelope to preserve the maximal functionality of the outer and inner membranes in response to the drastic temperature downshift (cold shock). One of the major challenges that bacteria face under cold-shock conditions is a partial loss of the membrane fluidity, in turn impairing important membrane functions such as energy production, cell division, and transport. In this regard, our finding that temperature downshift leads to the upregulation of genes involved in lipoprotein biogenesis [lolC, int (Table 1) and to a lesser degree of lolA and lolB (Table S1)] suggests that efficient production and incorporation of lipoproteins into the outer membrane may play a critical role in V. harveyi adaptation to low temperatures.
A somewhat different role can be envisaged for another ancillary protein, BamA (YaeT), encoded by the V. harveyi yeaT gene. An increase in expression of this gene (see Table 1) is likely critical to provide a high level of YeaT production. This protein is required for efficient insertion of several beta-barrel proteins (into the outer membrane of bacteria (Dautin & Bernstein, 2007) and therefore could be critical for V. harveyi cell envelope biogenesis under stress.
One of the important functions of the bacterial envelope is linked to the transport of solutes and biomolecules. Control of nutrient/metal ions transporter expression plays a central role in regulation of bacterial growth and adaptation to changing environmental conditions. We found that several V. harveyi transporter-encoding genes were up- and downregulated by exposure to cold seawater. Microarray data suggest that, at least in part, V. harveyi attempts to compensate the reduced expression of biosynthetic genes by upregulation of transporter genes controlling the uptake of amino acids (e.g. amino acid ABC transporter ATP-binding protein) and nucleotides (e.g. putative xanthine/uracil permease).
In addition to the lolC gene (discussed above), some highly upregulated genes are involved in the control of polar amino acids and iron uptake. The latter encode the energy-dependent transport TonB-ExbB-ExbD complex, bacterioferritin-associated ferredoxin, iron transport protein, and co-chaperone HscB. The upregulation of iron acquisition and downregulation of the iron storage genes known to be involved in iron homeostasis in other Vibrio species (e.g. V. cholerae; Mey et al. (2005)) suggest that V. harveyi adjusts its metabolism to strive in the seawater under iron-limiting conditions.
Finally, a somewhat unexpected group of transcripts with a significant change in expression was found to be related to DNA stability and genetic information processing. Exposure of V. harveyi to cold seawater appears to affect the integrity of DNA and subsequently change the expression level of several DNA repair genes (e.g. lexA) listed in Table 1. We also found that, similar to genes regulated in the human pathogen V. vulnificus by cold shock (Wood & Arias, 2011), a large group of genes controlling protein synthesis, in particular several genes involved in ribosome and tRNA biogenesis (e.g. rpsJ), were likewise upregulated in V. harveyi (Table 1). Their upregulation appears to be redundant and, unlike downregulation of other genes controlling ribosome and tRNA biogenesis (Table S1), contrasts with the reduced capacity of V. harveyi to carry out protein synthesis under starvation conditions. A possible explanation for this paradox is likely rooted in the partial loss of autoregulation usually exerted by the products of these genes (i.e. ribosomal proteins) by means of their binding to the 5′ UTRs of their cognate polycistronic RNAs to inhibit ribosome binding and subsequent translation of these transcripts (Babitzke et al., 2009). Thus, it seems likely that low temperature (4 °C) prevents the correct folding of the corresponding 5′ UTR, thereby preventing their autoregulation and subsequently leading to the excessive production (upregulation) of these transcripts under cold-shock conditions. Based on these observations, we propose that, in addition to the increased propensity of RNA to form structures under cold-shock conditions that potentially prevent the correct translation/transcription of bacterial transcripts (Phadtare et al., 2000), bacterial adaptation to low temperature can also be impaired by misfolding of RNA structures (e.g. translational operators or riboswitches) that have essential regulatory functions in vivo.
In summary, although microarray analysis revealed several group of genes that were affected by starvation and cold shock in the same manner as they are regulated in some other Vibrio species (Carroll et al., 2001; Yang et al., 2009; Wood & Arias, 2011), it also disclosed a considerable number of new stress-related genes (e.g. genes encoding guanylate kinase and putative xanthine/uracil permease that are involved in nucleotide salvage pathways as well as periplasmic alpha-amylase, an enzyme involved in utilization of polysaccharides) most likely important for adaptation of V. harveyi and other marine bacteria in natural aquatic systems.
The custom microarray design for gene expression analysis of V. harveyi was kindly provided by the lab of Prof. Bonnie Bassler (Princeton University). All microarray analysis and real-time PCR analysis procedures were performed at the Gene Expression Unit of the Genomics Core Facility – SGIker of the University of the Basque Country UPV/EHU. The work was supported by the Spanish Ministry of Science (CGL2011-26252 and BFU2011-25455) and the Basque Government (research project IT376-10, grants BFI-2011-85 and BFI09.103) and by Ikerbasque (Basque Foundation for Science).