Fish & Shellfish Immunology 43 (2015) 502e509

Contents lists available at ScienceDirect

Fish & Shellfish Immunology journal homepage: www.elsevier.com/locate/fsi

Full length article

Early steps in the European eel (Anguilla anguilla)eVibrio vulnificus interaction in the gills: Role of the RtxA13 toxin
s Callol a, b, c, David Pajuelo a, b, Lars Ebbesson d, Mariana Teles c, e, Agne Simon MacKenzie c, f, Carmen Amaro a, b, * a

ERI BIOTEDCMED, Universitat de Valencia, Spain Departament de Microbiologia i Ecologia, Universitat de Valencia, Spain
noma de Barcelona, Spain Institut de Biotecnologia i Biomedicina, Universitat Auto d Department of Biology, University of Bergen, Norway e
noma de Barcelona, Spain Department of Cell Biology, Physiology and Immunology, Universitat Auto f Institute of Aquaculture, University of Stirling, UK b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 4 December 2014 Received in revised form 7 January 2015 Accepted 9 January 2015 Available online 19 January 2015

Vibrio vulnificus is an aquatic gram-negative bacterium that causes a systemic disease in eels called warm-water vibriosis. Natural disease occurs via water born infection; bacteria attach to the gills (the main portal of entry) and spread to the internal organs through the bloodstream, provoking host death by haemorrhagic septicaemia. V. vulnificus produces a toxin called RtxA13 that hypothetically interferes with the eel immune system facilitating bacterial invasion and subsequent death by septic shock. The aim of this work was to study the early steps of warm-water vibriosis by analysing the expression of three marker mRNA transcripts related to pathogen recognition (tlr2 and tlr5) and inflammation (il-8) in the gills of eels infected by immersion with either the pathogen or a mutant deficient in rtxA13. Results indicate a differential response that is linked to the rtx toxin in the expression levels of the three measured mRNA transcripts. The results suggest that eels are able to distinguish innocuous from harmful microorganisms by the local action of their toxins rather than by surface antigens. Finally, the cells that express these transcripts in the gills are migratory cells primarily located in the second lamellae that relocate during infection suggesting the activation of a specific immune response to pathogen invasion in the gill. © 2015 Elsevier Ltd. All rights reserved.

Keywords: European eel Vibrio vulnificus Host-pathogen relationship Immune response rtxA13

1. Introduction Warm-water vibriosis is an acute haemorrhagic septicaemia caused by the gram-negative bacterium Vibrio vulnificus that mainly affects farmed European eel (Anguilla anguilla) [1e3]. Several important warm-water vibriosis outbreaks have been reported in eel farms in Europe since 1991 [4]. V. vulnificus is also a human pathogen that causes septicaemia after either raw seafood ingestion (primary septicaemia) or infections of wounds exposed to contaminated fishes or seawater (secondary septicaemia) with a high probability of death in immunocompromised patients [2]. The species is subdivided in biotypes and serovars from which only the

* Corresponding author. Dr. Moliner, 50. 46100-Burjassot (Valencia), Spain. Tel.: þ34 963543104. E-mail address: [email protected] (C. Amaro). http://dx.doi.org/10.1016/j.fsi.2015.01.009 1050-4648/© 2015 Elsevier Ltd. All rights reserved.

biotype 2-serovar E (Bt2-SerE) is able to infect both humans and eels, being recognized as a zoonotic agent [2]. Natural infection of eels occurs via waterborne contact with Bt2-SerE cells [5]. Bacteria attach to branchial lamellae and spread to the internal organs through the bloodstream [6]. Therefore, gills are the main portal of entry for this zoonotic pathogen into the eel. In teleosts, the gills together with the intestine form part of the mucosal immune system of fishes [7]. These tissues are covered by a mucus layer secreted by goblet cells, which contains multiple bactericidal compounds such as anti-microbial peptides, lysozyme and complement secreted by different types of immune cells associated to these surfaces [7,8]. Precisely, these cells are in charge of recognizing pathogens via specific receptors for highly conserved microbial pathogen associated molecular patterns (PAMPs). These receptors, called pattern recognition receptors (PRR), trigger the activation of innate defences which in turn stimulate adaptive immunity [9,10]. With regard to V. vulnificus recognition by the

A. Callol et al. / Fish & Shellfish Immunology 43 (2015) 502e509

human immune system results obtained by using infected mice as an animal model propose that the Toll-like receptors, TLR2 and TLR5, would recognize the capsule and flagellin respectively leading to secretion of the proinflammatory cytokine IL-8 [11e14]. In fish however the mechanism by which the pathogen is perceived in particular by the eel is still unknown. In turn, pathogens react against the immune system by producing virulence factors that direct or indirectly cause lesions in the tissues that are responsible for an important part of the disease aetiology. Recently, it has been proposed that a toxin from the MARTX family (Multifunctional Autoprocessing Repeats-in-Toxin) is the main virulence factor involved in warm-water vibriosis in the eel [15]. These toxins have two external modules, the Ct and the Nt ends that contain the amino acid repetitions by which they associate to eukaryotic membranes and form a pore through which the internal module is liberated into the cytoplasm [16]. The internal module contains various domains with enzymatic activities that differ among different MARTX and are responsible for the toxic effect upon the eukaryotic cell [16]. V. vulnificus produces at least four types of MARTX among which the MARTX type III or RtxA13 is specific to Bt2-SerE [17]. Mutants deficient in RtxA13 are able to colonize gills and spread to the internal organs as efficiently as the wild-type strain without killing the eels [15]. The toxin has been associated to killing of fish immune cells that leads to a dysregulation in cytokine production and death [15]. This toxin has also been suggested to form a major bacterial defence against phagocytic activity [15,18,19]. In this study we assessed the expression levels of three nonspecific immune-related mRNA transcripts, tlr2, tlr5 and il-8 (a potent chemokine), as well as analysed the localization of the cells expressing these transcripts, during the first step of the warmwater vibriosis: gill colonization. Furthermore, we tested whether RtxA13 has some role in this early interaction by comparing expression levels in the gills from eels infected either with the pathogen (strain CECT4999) or with a derivative mutant deficient in RtxA13 (strain CT285). 2. Materials and methods

sampled and the mortality was only considered when the pathogen was recovered from internal organs in pure culture. 2.3. Warm-water vibriosis reproduction and sampling Warm-water vibriosis was reproduced in the laboratory by infecting eels by immersion in a bath containing the bacterium. Eels of 20 g (±5 g) were distributed into groups of 40 and were infected by 1 h immersion with either V. vulnificus R99 (wild type) or CT285 (DrtxA13) strains at the LD50 previously determined for the R99 strain. A control group was immersed in marine water. All animal groups were maintained in independent tanks (20 animals per tank; 60 L of water) under the above conditions for 1 week. Then, groups of five randomly selected eels were sampled at 0, 1, 6 and 12 h post-infection for blood extraction, gill dissection and liver sampling. The remaining eels, 20 per group, were maintained to control mortality. For blood extraction, eels were previously anaesthetised with a 250 mg/l of benzocaine and blood was extracted with heparinised syringes and was distributed into two aliquots: one for both bacterial and blood cell counts (erythrocytes and leukocytes) and the other for immunological parameter determination. Gills were dissected with sterilized surgical material. Left gill arches were used for gene expression studies where cartilage was removed and branchial lamellae were used for RNA extraction. In parallel, right gill arches were used for in situ hybridization (ISH). The liver was air exposed by dissection using sterile surgical material and was microbiologically sampled for bacterial counting and V. vulnificus recovery on TSA-plates. Bacterial counts from blood, gills or liver were performed by drop plating on TSA [21]. All the experiments described comply with the guidelines of the European Union Council (2010/63/EU) for the use of laboratory animals and have been approved by the department of environment (reference code 2014/pesca/RGP/028), as well as by the general directorate of agricultural and livestock production of the Generalitat de Valencia (reference number 2014/VSC/PEA/00094 tipo 2). 2.4. Non-specific immune blood parameters

2.1. Bacterial strains and growth conditions V. vulnificus Bt2SerE wild type (CECT4999, hereafter R99) and its derivative CT285, deficient in the gene rtxA13, were used in this study (Table 1). Both strains were grown on Tryptone Soy Agar (TSA) or Broth (TSB) supplemented with 0.5% (w/v) NaCl at 28  C for 24 h. Bacterial stocks were maintained at 80 in TSB plus 20% glycerol. 2.2. Fifty percent lethal dose determination (LD50) All the experiments were performed in tanks of 100 L containing 30 L of marine water maintained at 25  C with aeration. Prior to bacterial infections, the LD50 for the R99 strain was determined by immersion according to [5] and calculated with the procedure of Reed and Münch [20]. Moribund animals were microbiologically

Table 1 Characteristics of V. vulnificus biotype 2 serovar E strains used. Strain

Description

Colonization ability

Virulence

Reference

R99

V. vulnificus Bt2 SerE Wild type (CECT4999) CECT4999, DrtxA13

Yes

Yes

[29]

Yes

No

[15]

CT285

503

Erythrocytes and leukocytes were observed at 40 under light microscope and total cells were counted using a dilution of 1:100 from whole blood in a Neubauer improved haemocytometer. In parallel, fresh blood was centrifuged at 800 g for 5 min, and the plasma obtained was used to determine IgM, prostaglandins and glucose levels. IgM levels in plasma were quantified using a commercial Fish IgM ELISA kit (Cusabio®) based on the competitive inhibition enzyme immunoassay. Prostaglandin levels were evaluated by a commercial EIA kit (Cayman chemical company) based on the measurement of PGE2 (prostaglandin E2) metabolite released into the plasma. Glucose levels in plasma were determined by enzymatic colourimetric analysis using commercial RTU kit rieux, France). (Biome 2.5. Primer and RNAprobe design Primers for the bacterial gene rtxA13 were taken from Lee et al. [15] (Table 2). Sequences for tlr2, tlr5 and il-8 were taken from our eel transcriptome shotgun assembly project database (Acc. number GBXM00000000, DDBJ/EMBL/GenBank) and were used for specific primer design. Primers for these eel mRNAs were designed and analyzed using bioinformatic programs and tools including Primer3 software [22] and OlygoAnalyzer (IDT SciTools) (Table 2). Conventional PCR were carried out and amplicons were separated on agarose gels stained with GelGreen™ and purified using

504

A. Callol et al. / Fish & Shellfish Immunology 43 (2015) 502e509

Table 2 Specific primers used for RT-qPCR and ISH riboprobe design. Organism

Gene

Primer

Sequence 50 -30

Length (bp)

Acc. Number

Reference

A. anguilla

tlr2

GBXM00005449

[37]

tlr5

321

GBXM00022707

This paper

A. anguilla

il-8

399

GBXM00014682

[37]

V. vulnificus

recA

60

e

[15]

V. vulnificus

rtxA13

ATGACCTGGGCTTGGCTTCA GGGCTCCAGCAGAATCAGGA GCGGTGGAAACGCACTGA GGTTGCTGAGCCTCAGACAC TAGGGGTGGATCTGCGGTGT GCTGCTTGTGTGTCTAACTTGTGC CGCCAAAGGCAGAAATCG ACGAGCTTGAAGACCCATGTG GAGTG ATGATGGGCGCTTTAC CAGCCGCGATGAGATGCT

360

A. anguilla

Fw Rv Fw Rv Fw Rv Fw Rv Fw Rv

60

e

[15]

Nucleospin® Extract II (MachereyeNagel) then cloned into a p-GEM easy vector (Promega) and transformed with JM109 competent cells (Promega) for sequencing and identity confirmation. Plasmids to be used for sequencing and as standards for absolute quantification were purified with Nucleospin® Plasmid (MachereyeNagel). Purified plasmids were also used to obtain DIG-labelled riboprobes for ISH. Transformed cells were stored at 80  C in TSB-1 plus 20% glycerol (v/v). 2.6. Gene transcription analysis After dissection, samples were directly treated with RNAlater (Qiagen), stored o/n at 4  C and kept at 80  C until use. Total RNA was isolated from tissues using 1 ml of TRI-reagent (Molecular Research Center, Inc.), following the manufacturer's instructions. RNA was quantified with a nanodrop 1000 spectrophotometer (Thermo scientific), and RNA integrity was verified using Agilent 2100 Bioanalyzer. Total RNA (400 ng) was used to synthesize cDNA with SuperScrip III Transcriptase (Invitrogen) and OligodT15 (Promega). cDNA was used as a template for absolute quantification in qPCR analysis. RT-qPCRs were carried out using SYBR® Green supermix (BioRad) in a final reaction volume of 20 ml. Standard curves were done using DNA plasmid purifications as a template for copy number determination. All qPCRs were performed using a MyiQ instrument (BioRad). Data were analyzed by one-way analysis of variance (ANOVA) followed by the post hoc multiple comparison by Bonferroni's method that was run for each gene to determine differences between groups (p < 0.05). All mRNA expression analysis with absolute quantifications had an R2 over 0.96 and primer efficiencies were between 99% and 115%. RT-qPCR of total bacterial RNA was performed in Power SYBR® Green PCR Master Mix (Applied Biosystems) with StepOne Plus RTPCR System (Applied Biosystems). The threshold cycle (CT) values were determined with StepOne Software V2.0 (Applied Biosystems) to establish the relative RNA levels of the tested genes. DNA polymerization was conducted from 60 to 95  C to obtain the melting curve for determining the PCR amplification specificity. The transcription level of rtxA13 vs recA, used as a housekeeping gene, was expressed as fold change (2DDCt) and calculated according to Livak's method [23]. 2.7. In situ hybridization (ISH) Gill samples were immediately immersed in neutral buffered formalin 4% (pH 7.4) for 1 h and treated for paraffin embedding and ISH was carried out to localize cells expressing mRNAs for tlr2, tlr5 and il-8. After DNA purification, RNA synthesis was performed with 10 DIG-labelling mix, RNA polymerase T7 and SP6 (Roche) to obtain sense (S) and anti-sense riboprobes (AS), which were stored

at 80  C until use. Paraffin blocks were cut in 5 mm sections with a microtome. Sections were rehydrated and treated with proteinaseK for 3:30 min to increase probe penetration, after fixation with 4% paraformaldehyde immersed in acetic anhydride to decrease the background and dehydrated again. Hybridizations were done with 350 ng of RNAprobe per slide o/n at 65  C, slides were treated with RNase A and incubated o/n with 1:3000 anti-DIG at RT in a humidified chamber. Visualization was carried out by NBT/BCIP method until signal appeared. Pictures were taken with LEICA DM LB microscope. 3. Results 3.1. Bacterial infection and rtxA13 expression Warm-water vibriosis was reproduced by infecting eels with the R99 strain and early colonization and invasion was monitored for 6 h by counting culturable bacteria recovered from each organ on TSA plates (Fig. 1A). The same experiment was also performed infecting eels with the mutant CT285 (defective in toxin production). Similarly to that previously reported [15], both strains were able to colonize gills from time 0 and spread to blood and liver without significant differences in terms of early colonization and invasion ability (Fig. 1A). Regarding rtxA13 transcription, the mRNA was expressed by the R99 strain in the gills from time 0 until the end of the experiment, without significant differences between sample time. RtxA13 was identifiable in blood and internal tissues from 1 h post-infection displaying a significant increase over time similar to that previously described [15] in both blood and liver (Fig. 1B). 3.2. Non-specific immune blood parameters Blood parameters including erythrocyte and leucocyte counts as well as, immune response parameters in plasma were evaluated throughout 12 h post-infection. The number of erythrocytes present after challenge did not change significantly either along the sampling time or between strains (Fig. 2A). However, leucocyte counts significantly increased at 6 h and returned to basal levels at 12 h post-infection but only when the eels were bath-infected with the wild-type strain (Fig. 2B). Furthermore significant differences in leucocyte numbers in blood between eels infected with the wild-type strain and those infected with the mutant were observed at 1 h and 6 h post-infection where higher counts were observed in the eels infected with the wild-type strain (Fig. 2B). In parallel, immunological parameters such as IgM, prostaglandins and glucose levels were evaluated. These three parameters did not significantly vary in relation to the strain used for the challenge or the sampling time (Fig. 2C-E).

A. Callol et al. / Fish & Shellfish Immunology 43 (2015) 502e509

505

Fig. 1. Eel colonization and invasion by Bt2SerE. Eels were bath-infected either with R99 (closed bars) or CT285 (open bars) and gills (i), blood (ii) and liver (iii) were sampled at 0, 1, 3 and 6 h post-infection for bacterial counting by drop plate methodology (A) and rtxA13 mRNA transcription quantification by qRT-PCR (B). Bacterial counts are expressed as log of colony forming units (CFU) on TSA-1 plates and the transcription level of rtxA13 vs recA (housekeeping gene) as fold change (2DDCt) calculated according to Livak's method [23]. Ttest statistical analysis was performed and significant differences are indicated as * (P < 0.05), ** (P < 0.01) or *** (P < 0.001).

3.3. mRNA expression analysis mRNA abundance levels for tlr2, tlr5 and il-8 in the gills of eels infected with V. vulnificus wild type and the mutant strains are shown in Fig. 3. tlr2 and tlr5 were upregulated 2-fold (p < 0.05) and 5.5-fold (p < 0.01) respectively after 1 h of bath-infection with the wild type strain and returned to basal levels at 6 h post-infection. In contrast, the mRNA abundance of both genes remained practically constant and similar to those of control samples in the gills of eels infected with CT285. Although transcription levels at 6 h were not statistically different to that of the controls, a clear tendency was observed towards downregulation as differences in transcription levels at 6 h post-infection between the wild type and the mutant strains were statistically significant at a p value