Protein expression and transcription profiles of three strains of Aeromonas salmonicida ssp. salmonicida under normal and iron-limited culture conditions

Background Aeromonas salmonicida is an important fish pathogen that produces a wide and varied array of virulence factors. Here we used iron deprivation by addition of the chelator 2’2-dipyridyl to induce the expression of several such virulence factors in three isolates of Aeromonas salmonicida (one avirulent and two virulent). By using SDS-PAGE followed by mass spectrometry, we identified proteins that appeared differentially expressed under these conditions. The differential transcription of the identified gene products were subsequently measured by reverse transcription quantitative real-time PCR (RT-qPCR). Results Our initial screening using SDS-PAGE identified five proteins that appeared differentially expressed in virulent and avirulent isolates or, within the same isolates, between bacteria cultivated under iron-rich or iron-deprived conditions. The transcription of the genes coding for these proteins were subsequently quantified by RT-qPCR. Results of this analysis demonstrated that the gene coding for alkyl hydroperoxide reductase (AhpC), a protein involved in oxidative stress response, was transcribed at a higher rate in the virulent strain as compared to the avirulent strain. Additionally, it was observed that addition of an iron chelator to the culture medium lead to a reduction of the transcription levels of the regulatory histone-like nucleoid structuring protein (H-NS). This was consistent in all three isolates. On the other hand, the transcription levels of the virulence array protein (VapA) and the protein ATP-synthetase F (ATPF) displayed only limited changes, despite being the dominant component of a protein fraction that displayed changes during the preliminary SDS-PAGE screening. This was true regardless of the culture conditions and of the isolates considered. Finally, transcription of the enzyme enolase was upregulated in the iron-deprived broths in all isolates. Conclusions We identified several genes differentially expressed under culture conditions known to lead to the overexpression of virulence factors. In addition, we identified alkyl hydroperoxide as being overexpressed in the virulent isolates compared to the avirulent isolates. The results from this study will contribute to enhance our understanding of the virulence of A. salmonicida and may suggest new directions for further research.


Background
First described in the 19th century, Aeromonas salmonicida is considered today as one of the oldest known fish pathogens. Much has since been discovered with regard to its virulence including identification of extracellular protein.
For example, the P1 and P2 proteases, caseinase and metallo-caseinase [1] and glycerophospholipid cholesterol acyltransferase (GCAT) [2]. However, the importance of such extra-cellular products was shown to be limited prompting some investigators to suggest that bacterial virulence may predominantly be attributed to bacterial growth [3]. A number of adhesion factors such as the A-layer, type I and type IV pili have also been identified [4][5][6]. More recently, the presence of a type III secretory apparatus [7] was also described. Unlike other virulence factors, the T3SS of A. salmonicida has been reported to be playing a major role in bacterial virulence [8][9][10]. The T3SS system is remarkable because of its thermo-sensitivity, which is lost when the organism is cultured at temperatures above 25°C [11,12]. Such thermo-sensitivity is of particular interest because the bacterium can be cultured at temperatures as high as 34.5°C, with a recommended temperature for the culture of Aeromonads as high as 28°C [13]. Therefore, it is likely that several clinical isolates have lost the type III secretion system they once carried [12] and have become avirulent. This loss may have occurred for the type strain of A. salmonicida ssp. salmonicida ATCC 33658 T, which is avirulent and does not possess a type III secretion system [8,14].
In most bacterial pathogens, virulence factors are tightly regulated [15]. One such regulator is the ferric uptake regulator (Fur) which is sensitive to the oxidative-stress response [16]. Fur controls bacterial response to oxidative molecules, including those generated by free radicals from iron, as well as the acquisition of iron [17]. In addition, Fur and related metalloregulatory proteins are also involved in the regulation of a number of virulence genes [16] in a variety of organisms, including E. coli [18], Streptococcus [19], Pseudomonas [20], Shigella [21], Edwardsiella [22] and Vibrio [23,24]. Not surprisingly, such regulatory mechanisms of bacterial virulence are present in A. salmonicida. For example, molecules associated with the quorum sensing system have been described [25,26] and shown to play a role in regulating the production of proteases [26] including Asap1 [27]. Similarly, a correlation was demonstrated between iron depletion and the production of virulence factors such as superoxide-dismutase [28], hemin receptor HutE, and the expression levels of genes encoding a putative DapD and Flp/Fap-pilin proteins [29]. Finally, three iron-regulated outer membrane proteins (IROMPs) have been described in A. salmonicida [30], all were TonB dependent: The first IROMP was a ferric siderophore receptor of 85 kDa in size while the second was identified as a colicin receptor homologue (FstC) of 73 kDa and likely to also act as a ferric receptor homologue. Finally, the last protein was 76 kDa in size and homologous to the heme receptor protein HutA from Vibrio cholera.
In A. salmonicida, it has been shown that the bacterial growth was slowed in iron-deprived cultures [30]. This is especially important during the infectious process because free iron is kept at a minimal level inside the host [31]. In fact, it has been shown that activation of human monocytes by the interferon-γ leads to reduction in their uptake of transferrin, therefore reducing the availability of intracellular iron [31]. For this reason, therapeutants aiming at further reducing the iron availability and acquisition for the bacterial pathogen have received scrutiny as alternative to antibiotherapy: Adjunction of the iron chelator deferasirox was shown to complement the efficiency of vancomycin against methicillin-resistant Staphylococcus aureus [32]. On the contrary, deferoxamine, another chelator, had a negative effect on the outcome of infection by Yersinia but only a limited effect on Klebsiella infection and no noticeable effect on infections with Aeromonas hydrophila [33]. Furthermore, inhibitors of siderophore production such as acyl-sulmamoyl adenosines have also been tested and proved efficacious against Mycobacterium tuberculosis and Yersinia pestis [34]. Finally, the monoterpenoid phenol carvacrol was shown to suppress the virulence of Salmonella enterica [35]. Unfortunately, despite these promising findings, this approach has yet, to the best of our knowledge, to be applied against fish pathogens. It is worth noting, however, that carvacrol is present in essential oils and might contribute to the protective effect of these when administrated as food feed additive [36,37].
The importance of iron availability in virulence and its regulatory mechanisms is therefore well established. Indeed, it was discovered that, at least as far as the major IROMPS were concerned, culture under iron-limited conditions up-regulated the same proteins than growth in vivo during the infectious process [30]. Consequently, the comparison of the transcription profile of an isolate between iron-rich and iron-deprived conditions as well as between virulent and avirulent isolate is likely to facilitate the identification of factors uniquely associated with a virulent phenotype. Here, we applied a preliminary SDS-PAGE screening, followed by quantitative RT-PCR to compare the protein transcription and expression profiles between three different isolates: the avirulent type strain ATCC 33658 T as well as two clinical isolates A-14390 and A-15233 under normal and iron-deprived culture conditions. Results from this study will enhance our understanding of the virulence in aquatic pathogens such as in A. salmonicida.

Preliminary SDS-PAGE screening
When examined by SDS-PAGE, the various samples produced similar pattern of bands (cf. Figure 1). However, five protein bands ranging from 50 kD to 15 kD in size appeared differentially expressed. Those five bands were excised, analysed by mass spectrometry and identified based on their molecular weight and the analysis of their mass spectra by Mascot.
A band of approximately 50 KD (WCP1) was observed in the whole cell crude extract and appeared more strongly visible under iron-limited conditions in the ATCC 33658 T and A-15233 isolates. Among the proteins identified in this fraction (Table 1) Among the outer-membrane proteins only one band of 50 kD (OMP2) was over-expressed (cf. Additional file 1: Figure S1) in isolate A-15233 under iron-limited conditions. This band mostly comprised the proteins VapA (score 3930) and Enolase (score 232).

Real-time qPCR Comparison between virulent and avirulent strains
RT-qPCR was performed on genes selected based on the results from the SDS-PAGE screening and its results are summarised in Figure 2 and Table 2. Only one gene, ahpC, was transcribed at a significantly higher rate in the virulent isolates compared to the avirulent ATCC 33658 T strain (mean fold change: 5.8E +01 ± 1.86E +00 , as calculated with the 2 -ΔΔC T method). This higher rate of transcription was true under both iron rich and iron-deprived culture conditions.
Comparison between iron-rich and iron deprived conditions WCP1 and OMP2 (50 kD), both contained the same two proteins of interest: VapA and enolase. RT-qPCR failed to detect any significant difference in the level of transcription of VapA (mean fold change: 3.64E + 01, with a confidence interval ranging from 2.64E + 01 to 5.14E + 01). Enolase, on the other hand was more clearly upregulated (mean fold change: 6.10E + 02, ranging from 1.68E + 02 to 2.23E + 03).
The 22 kD band contained AhpC, a protein that was found to be upregulated in the virulent isolates. Additionally, it was found that deprivation of iron also led to a slightly elevated transcription of this gene in all strains (mean fold change: 3.14E +01 ).
The 19 kD band contained AtpF and, when examined by SDS-PAGE, it appeared fainter in the ATCC 33658 T isolate when cultivated under iron-deprived conditions. This suggested that the protein might be less expressed under these conditions. However, analysis by qPCR indicated a minor up-regulation at the level of transcription (2.50E + 01, ranging from 1.32E + 01 and 4.73E + 01).
The 15 kD band contained the regulator H-NS. Results of the RT-qPCR revealed that it was significantly overexpressed under iron-replete conditions (average transcription ratios of 1.89E + 03, ranging from 1.09E + 03 to 3.29E + 03).

Discussion
In order to investigate the virulence factors of the fish pathogen A. salmonicida, three bacterial isolates, one known to be avirulent and two virulent ones were cultivated under iron rich and iron deprived conditions. The bacteria in these cultures were lysed and the cell lysates were analysed on SDS page. Five fractions (of size ranging from 50 to 15 kD) that appeared differentially expressed between the virulent and avirulent isolates or between iron-rich and iron-deprived culture conditions were excised and analysed by mass-spectrometry and identified by Mascot.
Fraction WCP1 contained proteins of~50 kD in size. These were found to be upregulated under iron-deprived conditions upon examination of the SDS-PAGE. Based on  the PMSS method the most represented of these was the virulence array protein and enolase (the same proteins present in OMP2).
Of the two, only enolase appeared markedly upregulated under iron-deprived conditions when analysed by qPCR (mean fold change of 6.10E + 02, as calculated by the 2 -ΔΔC T method). VapA is responsible for the paracrystalline A-layer, an important virulence factor of A. salmonicida [4]. Enolase is a virulence factor known to play a role in the plasminogen binding of a number of bacteria, including Aeromonas hydrophila [38]. It is also known that the expression of the enolase of A. hydrophila is increased during the infection process [39].
No differences were found in the transcription levels of either VapA or enolase between the virulent and avirulent isolates. This finding was consistent with that of Vanden Bergh et al. [40] that determined that the secretion of both of these proteins was not limited in a T3SS deficient mutant.
Band WCP2 contained several proteins of approximately 22kD in size. The most interesting among these was YP_001142399.1, an alkyl hydroperoxide reductase C (AHPC). On the SDS-PAGE, this protein appeared to be more highly expressed in the two virulent isolates under iron-deprived conditions. This was consistent with the results from the RT-qPCR analysis (mean fold change: 5.8E + 01 between virulent and avirulent isolates). To a lesser extent, the transcription of this protein was also upregulated under iron-deprived conditions (3.14E + 01, ranging from 2.23E + 01 to 4.64E + 01). AHPC binds heme molecules and plays a role in its metabolism [41]. However, AHPC has mostly been studied in relation with its role in the resistance against oxidative stress including resistance against macrophages [42]. It has already been demonstrated that APHC is over-expressed in other bacterial pathogens, such as Vibrio salmonicida during the infection process [43]. Furthermore, this regulation is sometimes achieved through ferric metabolism [44][45][46], as seems to be the case for A. salmonicida [47].
A fraction of 19 kD in size (WCP3) contained AtpF, a protein known to play a role in the resistance of microorganisms against acidic conditions for example, in Helicobacter pylori [48]. SDS-PAGE screening displayed a difference in the intensity of this 19 kD band in the avirulent isolate. This change could be due to one of the proteins within this fraction being down-regulated under irondeprived conditions. However, when RT-qPCR analysis was conducted to confirm this observation, it failed to detect such downregulation of the transcription levels when the bacterium was cultivated under iron-deprived conditions. Rather, the results indicated that the transcription of the atpf gene was slightly upregulated (transcription ratio of 2.50E + 01, ranging from 1.32E + 01 to 4.73E + 01). This suggested that another protein of within this band might have been responsible for the change observed in the band intensity or that this protein's expression is regulated post-transcriptionally.
Perhaps more interesting was the finding of the H-NS regulator (YP_001142759.1) within the 15 kD band (WCP4). This protein appeared down-regulated under iron-deprived conditions (1.89E + 03, with a confidence interval ranging from 1.09E + 03 to 3.29E + 03). hns is part of a multifactor regulatory system in Shigella flexneri where it represses transcription of the virulence protein IcsA [49]. hns is similarly negatively regulated in the presence of Fur in Salmonella enterica serovar typhimurium [50] and down-regulation of hns in turn allows for the expression of hil, resulting in the eventual activation of the Salmonella pathogenicity island I (SPI-1). Interestingly, no direct homolog of hilA appears present within the genomes of aeromonads. Therefore, if H-NS plays the role of a virulence activator in A. salmonicida like it does in the case of Salmonella consistent with our findings, it does so through different means. Contrary to our expectations, only one band from the outer-membrane preparation was found to be overexpressed under iron-deprived conditions. The contents of this band were similar to WCP1, containing both VapA and the enzyme enolase. This contrasted with the results reported by Ebanks et al. [30] who, while focussing on outer-membrane proteins, identified three iron-regulated outer membrane proteins (IROMP) on the surface of A. salmonicida: a ferric siderophore receptor, a colicin receptor homologue (FstC) and a heme receptor protein.
The reason for this discrepancy is not known but several factors might explain it. For example, Ebanks et al. have previously suggested that the number of IROMPs might vary depending of the isolate studied [30]. Alternatively, this discrepancy may be related to the method of our preliminary screening. Because we only targeted proteins that appeared differentially expressed on the SDS-PAGE, it is likely that a number of products escaped our detection: Either the difference in their expression levels was too small to be noticeable on the SDS-PAGE, or they were part of a larger fraction and their difference in expression was masked by other proteins in this faction whose expression was constant. This is particularly true for higher molecular weight bands that have poor resolution on 1-D gels. In contrast, Ebanks et al. used 2Dgel electrophoresis to isolate the various proteins. Addition of this second dimension yields better resolution of the various proteins thus offering improved sensitivity than the one employed in the present study and these authors were therefore likely to identify proteins that would remain masked using SDS-PAGE, especially higher molecular weight molecules such as the ones, ranging from 73 to 85 kDa, described by these authors [30].

Bacterial strains and culture conditions
Three different isolates of Aeromonas salmonicida spp. salmonicida were used: strain ATCC 33658 T known to be avirulent [8] and to lack the plasmid carrying the type III secretion system [14,51] and two virulent isolates obtained from disease outbreaks in trout farms in Austria (isolates A-14390 and A-15233). In previous challenges, the strain ATCC 33658 T was shown to induce 70% mortalities when 10 8 CFU were injected intra-peritoneally, while 10 7 CFU failed to induced any mortality. In contrast injection of 5x10 3 CFU induced 90% mortalities when using the isolate A-14390 and 70% mortality with the isolate A-15233). All isolates have been conserved at −80°C and have been used in several fish trials at our institution where they displayed high virulence.
All three isolates were cultured in the same fashion: beads were removed from the −80°C freezer and used to inoculate blood agar plates. After 2 days of cultivation at 15°C, single colonies were removed from these plates and used to inoculate 5 ml of BHI broth with or without iron (iron depletion attained by the addition of 100 μM of the chelator 2'2-dipyridyl to the BHI broth [28]). Cultures were incubated at 15°C for 48 hours before being processed for whole cell crude extract or outermembrane protein extraction.

Preparation of the whole cells and outer-membrane proteins
The whole cell crude protein extract was obtained by centrifugation of the bacterial cultures at 3,220 g for 20 min using an Eppendorf 5810R tabletop centrifuge (Eppendorf; Vienna; Austria). Pellets were rinsed 3 times in 0.9 M PBS before being resuspended in 400 μl lysis buffer (50 mM Tris-HCl; 150 mM sodium chloride; 20 mM ethylene diamine tetraacetic acid; 1% sodium deoxycholate; 1% Triton X-100; all ingredients from Sigma; Steinheim am Albuch; Germany). Pellets were sonicated twice on ice for 30 seconds using a UW1070 Sonopuls (Bamdelin; Berlin; Germany). The samples were then centrifuged at 3,220 g at 4°C for 10 minutes and the supernatants were removed and mixed with an equal quantity of Laemmli sample buffer (Bio-Rad; Vienna, Austria).
Outer-membrane proteins (OMPs) were purified based on the protocol by Wilcox [52]. Briefly, the bacterial cultures were pelleted by centrifugation at 3,220 g for 20 minutes and the pellets were rinsed 3 times in 0.9 M PBS before resuspension in 5 ml of 50 mM Tris-HCl and sonication on ice. After centrifugation, supernatants were transferred to a second tube and 500 μl of 20% sodium N-lauroylsarkosynate (Fluka Analytical; Sigma Aldrich; Steinheim am Albuch; Germany) was added. These preparations were then incubated for 30 minutes at room temperature before being ultra-centrifuged at 50,000 g for 1 hour at 4°C in a Sorvall Ultra Pro 80 centrifuge (Thermo-Sorvall; Waltham; USA). Supernatants were set aside and the pellets were resuspended in 15 μl 50 mM Tris-HCl and mixed with 150 μl doublestrength Laemmli sample buffer.

Initial screening and mass spectrophotometry
All preparations were stored at −20°C for no more than 24 hrs. The concentrations within these preparations were then measured by absorbance using a Biophotometer (Eppendorf; Vienna; Austria). Subsequently, the preparations were diluted to the same concentration and an equal amount of total protein were resolved by SDS-PAGE using a 4% polyacrylamide stacking gel and a 16% resolving gel [53].
Gels were run using the Mini-Protein Tetra Cell (Bio-Rad laboratories GmbH) at 120 V for 70 min in 1× running buffer (25 mM Tris, 250 mM glycine, pH 8.3, 0.1% SDS). Proteins were visualized by staining with Coomassie Brilliant Blue R-250 (Sigma-Aldrich) and the Precision plus protein all blue standard (Bio-Rad, Vienna, Austria) was used to estimate the molecular weight of the separated proteins.
Bands that appeared differentially expressed, either between the virulent isolates and the avirulent strain or under iron rich and iron-deprived culture conditions, were excised. These bands were subsequently digested in-gel using trypsin (Trypsin Gold, Mass Spectrometry Grade, Promega, Madison, WI, USA). The extracted peptides were then analyzed in a nano-LC MS/MS using an Ultimate 3000 RSLC HPLC system (Dionex, Amsterdam, Netherlands) followed by mass spectrometric analysis in a HCT Esquire mass spectrometer (Bruker Daltonics, Bremen, Germany). Processed spectra were identified using Mascot in the NCBI database (Table 1). In order to identify the proteins that were likely to be the most important contributor to the detected variations, we employed the peptide-matching score summation (PMSS) technique which scores the various proteins depending on their relative abundance.

Design of the oligonucleotide primers and preparation of the standard curves
Based on the initial screening and mass spectrometry results, five genes of particular interest were selected for investigation by quantitative RT-qPCR: alkyl hydroperoxide reductase C (ahpC), ATP-synthetase F (ATP-F), enolase, histone-like nucleoid structuring protein (hns) and the virulence array protein (vapa). Several proteins identified in the gels were known to be housekeeping genes or genes unlikely to be involved in virulence and were therefore not included further in the study. Primers were designed using the software Primer 3 to target short sequences (ranging from 112 bp to 180 bp in size) within these genes as well as within the 16S RNA sequence (Table 3). The optimum annealing temperature for each primer was determined using a gradient PCR assay and standard curves were generated for its gene using tenfold dilutions (from 10 −1 to 10 −10 ).

RNA isolation and cDNA synthesis
A. salmonicida spp. salmonicida were grown in triplicate at 15°C under the conditions previously described. Subsequently, bacterial cells were lysed in lysis buffer and homogenised using a QIAshredder kit (Qiagen, Hilden, Germany). The RNA was isolated from these lysates using a Qiagen RNeasy kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. The samples were treated on-column with DNAse (Qiagen, Hilden, Germany) to exclude DNA contamination and eluted in 20 μl RNAse free water. cDNA was synthesised using the iScript cDNA synthesis kit (Bio-Rad, Vienna, Austria) following the manufacturer's instructions.

Quantitative real-time PCR assay
Real-Time PCR (qPCR) was carried out in 25 μl endvolumes in technical duplicates using a CFX96 Touch real-time PCR detection system (Bio-Rad, Vienna, Austria). The reaction mixtures comprised of 5 μl cDNA template, 12.5 μl SYBR Green Supermix dye (Bio-Rad, Vienna, Austria) including a 7 pmol final concentration of each primer. The qPCR program started with 3 minutes at 95°C, followed by 39 cycles of 94°C, 59°C and 72°C, at 30 seconds each, with the plate reading step at 59°C. The thermal profile for the melt-curve analysis was determined by incubation at 59°C for 1 minute followed by gradual increase of temperature to 95°C during which change in fluorescence was monitored. No template controls (NTCs) were included in duplicate to control for the presence of primer-dimers. The results obtained were analysed using the Bio-Rad CFX Manager 2.1 software (Bio-Rad, Vienna, Austria). The ΔCT values obtained were normalized based on the 16S-RNA and the normalized transcription ratios were calculated based on the 2 -ΔΔC method [54] in order to compare the gene transcription levels between the virulent and avirulent strains and also within strains between the two culture conditions.

Conclusions
In the present study, we used SDS-PAGE, mass spectrometry and RT-qPCR to analyze the protein expression profiles of three strains of A. salmonicida ssp. salmonicida under normal and iron deprived culture conditions. We determined that the protein AHPC which is involved in resistance against oxidative stress was over-expressed in the virulent isolates compared to the avirulent isolates including marginal over-expression under iron poor conditions in all isolates. Similarly, the enzyme enolase was overexpressed when the bacteria were cultivated in broth enriched with 2'2 dipyridyl. In contrast, the regulatory protein H-NS was under-expressed under such iron-limited condition which suggests the existence of a regulatory pathway coordinated through this molecule. This putative pathway however, does not act through hilA as is the case in the genus Salmonella because no homologue of the molecule could be identified in the aeromonads' genome. Thus we believe that these findings enhance our understanding of the virulence mechanisms and regulation in A. salmonicida. However, additional research will be required to ascertain the precise mechanisms by which the availability of iron regulates the expression of the genes identified in this study.

Additional file
Additional file 1: Figure S1. SDS-PAGE of the outer-membrane preparation displaying OMP2. Well A contains Isolate A-15233 (virulent) cultivated under normal conditions while well B contains Isolate A-15233 (virulent) cultivated under iron-deprived.