Diversity of Vibrios associated with reared clams in Galicia (NW Spain)

1 1 2 Diversity of Vibrios associated with reared clams in Galicia (NW3 Spain).4 5 Roxana Beaz Hidalgo1, Ilse Cleenwerck2, Sabela Balboa1, Marjan De Wachter2,6 Fabiano L. Thompson3, Jean Swings2, Paul De Vos2 & Jesús L. Romalde1*7 8 1 Departamento de Microbiología y Parasitología. Facultad de Biología e Instituto de9 Acuicultura. Universidad de Santiago de Compostela. 15782. Santiago de Compostela.10 2 BCCM/LMG Bacteria Collection, Laboratory of Microbiology, Ghent University,11 Ghent, Belgium.12 3Department of Genetics, Institute of Biology, Federal University of Rio de Janeiro13 (UFRJ), Rio de Janeiro, Brazil.14 15 16 17 18 19 20 21 22 Accepted in: Systematic and Applied Microbiology, March, 200823 24 25 26 27 * Corresponding author:28 Phone: +34 981563100 # 1325329 Fax: +34 98159690430 E-mail: [email protected]31 32 33 2 ABSTRACT34 The aim of the present study was to characterize and identify vibrios isolated from35 cultured clams in Galicia (NW Spain). A total of 759 isolates were obtained,36 phenotypically characterized, grouped and assigned to the genus Vibrio. Subsequently,37 the genomic diversity of 145 representative strains was analyzed by means of amplified38 fragment length polymorphism (AFLP) and revealed a high genetic diversity amongst39 those isolates. Only 57 out of 145 strains could be identified to the species level, which40 were distributed in 13 AFLP clusters. V. cyclitrophicus, V. splendidus and V.41 alginolyticus were the most abundantly represented species. Eighty eight isolates42 remained unidentified, 59 were distributed over 16 clusters, while 29 were unclustered.43 Sequencing of the 16S rRNA and two house-keeping (rpoA and recA) genes of44 representative strains belonging to the 8 unidentified clusters with higher number of45 isolates confirmed their assignation to the Vibrionaceae family, some of them probably46 representing new species within the genus. The present study confirms that the47 phenotypic characterization of vibrios is not enough to identify them at species level. A48 wide diversity of vibrios was found in cultured clams of all four geographic locations49 analyzed. In total, more than 12 Vibrio species and at least three potential new species50 in this genus were identified.51 52 Keywords: Vibrios, diversity, clam, AFLP, sequencing.53 3 INTRODUCTION54 The culture of clams represents an important marine resource and is of great economic55 importance for the coast of Galicia (NW Spain). Natural clam beds of authoctonous56 carpet-shell clam (Ruditapes decussatus) have been harvested beyond their maximum57 sustainable yield, which has led to the introduction of the foreing species Manila clam58 (Ruditapes philippinarum) to cover the consumer demands, with the risk of introducing59 new pathogenic strains [20]. Periodically, mortalities occur in these clam populations60 which, until now, could not be deffinetively associated to any infectious agent or61 environmental factor.62 The family Vibrionaceae is autochthonous to aquatic environments including estuarine,63 coastal waters and sediments world wide, and some species are well known pathogens64 of marine organisms including fish, corals and molluscs [8, 21, 24, 32]. Vibriosis65 mainly affects nursery cultures of juvenile bivalves. Vibrios associated with infection in66 bivalve molluscs belong to the species such as Vibrio tapetis, Vibrio crassostreae,67 Vibrio alginolyticus and Vibrio pectenicida [15, 17, 19]. In Spain, mass mortalities have68 been reported in adult clams due to V. tapetis infections [4], and in seed and juveniles69 due to Vibrio splendidus II, V. neptunius and V. alginolyticus [8, 20]. Other molluscs are70 also susceptible to infection by vibrios including scallops Pecten maximus [17], oysters71 Crassostrea gigas [34] and abalone, Haliotis diversicolor supertexta [16]. However,72 scarce information exists on the occurrence of vibrios or any other bacteria that are73 pathogenic to cultured clams [3,19].74 The main aim of this study is to analyze the Vibrio populations associated with cultured75 clams in Galicia, using, a polyphasic approach which include phenotypic76 characterization, amplified fragment length polymorphism (AFLP) analysis and77 sequencing of the 16S rRNA and two house-keeping genes, in order to determine the78 4 normal microbiota and also to be able to identify potential new pathogens for these79 bivalve molluscs.80 81 MATERIALS AND METHODS82 Sampling, bacterial strains and growth conditions83 Cultured clams were monthly sampled from 4 geographic locations (A to D) in the84 Atlantic Northwest of Spain (Fig. 1) from March 2004 to September 2005. Organisms85 were transported alive on ice to the laboratory and analyzed within 3 hours. A total of86 15-20 clams, depending on the size, were aseptically dissected and samples of87 hepatopancreas, mantle, gonad, gills and extrapalial fluid were collected. For seed or88 young clams, the whole organism of 15-20 individuals was collected and treated as the89 dissected organs. Homogenates were prepared by adding a volume of saline solution90 (0.85%) to the weighed organ samples (1:1). Samples were serially diluted in saline91 solution (0.85%), plated on marine agar (MA) (Pronadisa, Madrid, Spain) and92 thiosulphate citrate bile sucrose agar (TCBS, Oxoid Ltd., Basingstoke, UK), and93 incubated at 23 ± 1ºC for 10 days (MA) or 48h (TCBS). Pure cultures of the different94 colony morphologies were recovered on MA. Isolates were routinely grown on Marine95 Agar (Pronadisa) at 23 ± 1ºC for 24 hours. Stock cultures were maintained frozen at96 –80oC in Marine Broth (Oxoid) supplemented with 15% glycerol (v/v).97 Phenotypic analysis98 The bacteria isolated (n = 759) were subjected to a set of phenotypic tests: cell99 morphology and motility, Gram stain, oxidase, growth on TCBS, susceptibility to the100 vibriostatic agent 0/129, production of arginine dihydrolase, lysine and ornitine101 decarboxylase, Glucose fermentation, indole, hydrolysis of gelatin, starch, esculin and102 Tween 80, reduction of nitrate to nitrite, production of gas from glucose, Voges103 5 Proskauer, Methyl red, growth at different temperatures (4ºC, 37ºC, 44ºC) and at104 different salinities (0%, 0.5%, 3%, 6%, 8%). [1, 9, 24].105 Genotypic analyses106 A total of 145 representative strains of all phenotypic groups were selected for further107 genotypic analysis. At least one strain was selected for each phenotypic group. The 145108 selected strains for fingerprinting by AFLP included 33 isolated from site A, 36 from109 site B, 29 from site C and 47 from site D.110 Genomic DNA extraction111 A loop of bacterial cells was harvested from fresh cultures for genomic DNA extraction112 using Easy DNA (Invitrogen, Barcelona, Spain) kit. Concentration and purity of DNA113 were estimated measuring optical densities at 260 and 280 nm using a114 spectrophotometer Spectra Max Plus 384 (MDS Inc., Sunnyvale, CA). DNA integrity115 was verified by loading DNA samples on a 1% agarose gel in 1X TAE buffer (40Mm116 Tris/Acetate, 1Mm EDTA, pH 8.0). Extracted DNA was maintained at –20ºC until117 further use.118 AFLP analysis119 AFLP is based on the amplification of subsets of genomic restriction fragments by using120 PCR [11, 33]. DNA stock samples were diluted to obtain an initial concentration of 1 µg121 of DNA in 26 µl of sterile Milli Q water. DNA was digested with TaqI (5´TCGA3´) and122 HindIII (5´AAGCTT3´) (Amersham Pharmacia Biotech, Sweden) and double-stranded123 adaptors were ligated to the ends of restriction fragments with T4 ligase (Amersham124 Pharmacia Biotech) to generate template DNA for PCR amplification. Two subsequent125 PCR amplifications, pre-selective PCR and selective PCR, were performed using126 primers and conditions previously described [25] in a GeneAmp PCR System 9600127 thermocycler (Applied Biosystems, USA). Separation of the selective PCR products128 6 was performed on a 36 cm denaturing polyacrilamide gel (4.25% Acrylamide, 6 M Urea129 in 1 x TBE/89 mM Tris + 89 mM Boric acid + 2mM EDTA, pH 8.3) on an ABI Prism130 377 DNA sequencer (Applied Biosystem). The level of reproducibility was controlled131 by generating the AFLP pattern of the strain V. alginolyticus LMG 4409T from the132 BCCM/LMG Bacteria collection (Belgian Co-ordinate Collections of Micro-133 organisms/Laboratory for Microbiology of the Ghent University, Belgium) in each134 AFLP assay performed. Tracking and normalization of the lanes were performed using135 the Gene Scan 3.1 software (Applera Co., Norwalk, CT), and the subsequent numerical136 was carried out by the BioNumerics 4.5 software (Applied Maths, Sint-Martens-Latem,137 Belgium). Similarity values were calculated using the Dice coefficient (tolerance value138 of 0.3%) and a dendrogram was constructed using the UPGMA algorithm. The AFLP139 patterns were compared with the profiles of the Vibrionaceae database at the140 BCCM/LMG Bacteria Collection, containing 544 reference profiles representing most141 validly described species generated mainly by Thompson et al. [25]. The cut-off level of142 similarity for defining the AFLP clusters was 63%. In addition, strains showing143 similarities higher than 88% were considered as clones following the criteria of144 Thompson et al. [25].145 Sequencing of the 16S rRNA, rpoA and recA genes146 Clusters that could not be identified by AFLP were further analyzed by sequencing the147 16S rRNA gene of representative strains as previously described [20]. 16S rRNA genes148 were amplified by PCR with universal primers pA and pH [10]. Primers [5]149 corresponding to internal conserved regions of this gene were used for the sequencing150 reactions in Tgradient machine(Biometra, Germany), using a GenomeLab DTCS-Quick151 Start Kit (Beckman Coulter, USA). Sequencing of the house-keeping genes rpoA and152 recA was performed according to Thompson et al. [23, 32].153 7 Sequencing products were analysed using an ABI Prism 373A DNA sequencer (Applied154 Biosystems). Sequence data analysis was performed with DNAstar Seqman program155 (Lasergene, USA). Gene sequences of the novel strains were subjected to a BLAST156 search against the latest release of the GenBank and related sequences were obtained.157 Phylogenetic trees were constructed by neighbor-joining (NJ), distance matrices were158 calculated using Kimura´s two parameter correction and stability of groupings and159 bootstrap analysis (1000 replicates) was performed with the program Mega version 3.1.160 161 RESULTS AND DISCUSSION162 A total of 759 strains were obtained from cultured clams, assigned to the Genus Vibrio163 on the basis of 25 morphological, physiological and biochemical characteristics, and164 grouped into 29 distinct clusters (Table 1). V. splendidus turned out to be represented by165 428 isolates i.e. 56% of the total number of isolates. The next most numerous groups166 were V. alginolyticus (48 isolates, 6.3%), V. diazotrophicus (41 isolates, 5.4%), V.167 aestuarianus (32 isolates, 4.2%), V. pelagius I / V. superstes (31 isolates, 4.1%) and V.168 fluvialis (28 isolates, 3.7%). The rest of the groups contained less than 20 isolates. Some169 groups were quite homogeneous, consisting of isolates having almost identical170 properties; others were somewhat heterogeneous including isolates with variable171 phenotypic traits. Recently, Noguerola and Blanch [18] have proposed a set of172 dichotomous keys for rapid identification of Vibrio isolates. With minor exceptions, our173 results were in agreement with the identification schemes porposed in their work.174 Discrepancies were observed for some traits in few species, such as growth at differeent175 temperatures and salinities or Voges-Proskauer test in V. diazotrophicus, V. scophthalmi176 and V. natriengens. These discrepancies were probably due to the fact that those authors177 8 studied only the type strains of the different species instead of a large number of178 environmental isolates as included in our work.179 The AFLP analysis was performed on 145 representatives of the different phenotypical180 groups and included also type and reference strains of practically all known species of181 the genus Vibrio. The AFLP patterns contained from 65 to 156 bands between 50 to 536182 bp in size. No clonality was observed in any of the clusters, being the AFLP similarity183 values among strains always under 88%. The similarity among patterns obtained for V.184 alginolyticus LMG 4409T employed as positive control in every AFLP run was always185 higher than 88%, supporting the meaning of this value as cut-off for clonality.186 A total of 94 clusters and 81 unclustered strains were obtained (Dendrogram is available187 as Supplementary File). A total of 57 Galician isolates were identified to species level188 and distributed in 13 clusters. Species identified were V. cyclitrophicus (17 strains), V.189 splendidus (16 strains in two clusters), V. alginolyticus (5 strains), V. diabolicus (4190 strains), V. crassostreae (4 strains), V. chagasii (3 strains), V. mediterranei (2 strains),191 V. ichthyoenteri (1 strain), V. parahaemolyticus (1 strain), V. pectenicida (1 strain) and192 V. lentus (1 strain).193 Most clam strains (n = 88) remained unidentified by AFLP, since they did not clustered194 with any type strain. Fifty nine of them were distributed over 16 clusters, while 29 were195 unclustered. The clusters obtained in the present work were similar to those obtained by196 Thompson et al. [25] using the DICE coefficient and the WARD algorithm, although an197 important improvement of the method for identification purposes was achieved, since198 no groups harboring multiple type strains were observed. On the other hand, some of199 the species analyzed including V. splendidus, V. harveyi or V. diazotrophicus, grouped200 in more than one cluster indicating the existence of intraspecific diversity within these201 9 Vibrio species. Similar results have been previously observed by other authors using202 this technique for typing of different Vibrio species [13, 28].203 A wide geographical distribution was observed for most phenotypic and AFLP groups204 established, being not possible any association among groups and a specific site. The205 highest diversity was observed in site D, but this fact is probably related to the higher206 number of strains (33%) obtained from that site.207 A comparison of the phenotypic and the AFLP results showed that, for isolates that208 could be identified at species level by both procedures, identification was coincident209 only in 29.82% of the cases. Major discrepancies were observed in the heterogeneous210 species, such as V. splendidus [8]. Thus, some isolates considered as V. splendidus211 based on phenotypic results, were identified as V. cyclitrophicus, V. crassostreae, V.212 chagasii or were unidentified by AFLP.213 Sequencing of the 16S rRNA gene of representative strains belonging to the major 8214 unidentified clusters, confirmed that these isolates belonged to the genus Vibrio. Cluster215 5 (6 strains) could be assigned to V. tasmaniensis (99.74% similarity). Clusters 37 (7216 strains), 48 (3 strains) and 86 (4 strains) may constitute three potential new species217 taken into account their 16S rRNA, rpoA and recA similarities (Table 2), as well as218 their relative positions to known vibrios in the phylogenetic concatenated tree (Fig. 2).219 These three potential new species are pending to further analysis to satisfy current220 requirements for new species descriptions [22]. Cluster 37 was close to the V.221 halioticoli-like group showing sequence similarities in the 16S rRNA gene of 99.4%222 with Vibrio comitans, 99.15% with Vibrio rarus and 99.0% with Vibrio inusitatus223 (Table 2). However, rpoA and recA similarities were lower than 97 and 93%224 respectively (Table 2), and DNA-DNA hybridization values obtained among these225 species did not reach 36% (data not shown). Cluster 48 was close to the V. splendidus-226 16 395 [33] P. Vos, R. Hogers, M. Bleeker, M. Reijans, T. Van de Lee, M. Hornes, A. Frijters,396 J. Pot, J. Peleman, M. Kuiper, M. Zabeau. AFLP: a new technique for DNA397 fingerprinting. Nucleic Acids Res. 23 (1995) 4407-4414.398 399 [34] M. Waechter, F. Le Roux, J.L. Nicolas, E. Marissal, F. Berthe. Characterization of400 pathogenic bacteria of the cupped oyster Crassostrea gigas. C. R. Biol. 325 (2002)401 231-238.402 17 Table 1. Data of preliminary phenotypic data and AFLP identification clusters. NT, not tested. RP= Ruditapes philippinarum, RD= Ruditapes decussatus Phenotypic group AFLP clusters Phenotypic identification AFLP identification Geographic sites P1 5, 6, 7, 9, 13, 19, 48, 68, 77, 26 V. splendidus I V. splendidus, V. chagasii, V. cyclitrophicus, unidentified A, B, C, DRP, DRD P2 5, 7, 8, 9, 13, 16, 19, 48 V. splendidus II V. splendius, V. crassostreae, V. chagasii, V. cyclitrophicus, unidentified A, B, C, DRP, DRD P3 28, 29 V. alginolyticus V. alginolyticus, V. diabolicus A, B, C, DRP, DRD P4 7, 8, 19, 68, 70 V. diazotrophicus V. splendidus, V. crassostreae, V. cyclitrophicus, unidentified A, B, C, DRP, DRD P5 5, 14, 19, 29 V. aestuarianus V. lentus, V. cyclitrophicus, V. diabolicus, unidentified A, B, C, DRP P6 36, 37, 53, 67, 70 V. pelagius I / V. superstes Unidentified A, B, C, DRP, DRD P7 6, 7, 9, 16, 26 V. fluvialis V. splendidus, unidentified A, B, C, DRP, DRD P8 5, 37 V. natriengens Unidentified A, B, C, DRP P9 28, 53 Vibrio sp. V. alginolyticus, unidentified A, B, C, DRP, DRD P10 9, 19, 26, 70, 82 V. pacinii V. cyclitrophicus, unidentified A, B, C, DRP, DRD P11 13, 16, 19 V. lentus V. chagasii, unidentified A, B, C, DRP, DRD P12 22, 48, 70 V. tasmaniensis / V. fortis V. tapetis, unidentified A, B, C, DRD P13 28, 58 V. fischeri / V. logei V. alginolyticus, unidentified. A, B, C, DRD P14 70, 82 V. scophthalmi Unidentified A, B, C, DRP, DRD P15 8, 19 V. cyclitrophicus V. cyclitrophicus, V. crassostreae A, B, C, DRP, DRD P16 52 V. ordalii V. ichthyoenteri B, C P17 8, 29, 47, 55 V. harveyi / V. vulnificus V. crassostreae, V. diabolicus, V. parahaemolyticus, V. mediterranei A, B, C P18 19, 24 V. nereis V. cyclitrophicus, V. pectenicida A, B, C, DRP P19 19, 77 V. mytili V. cyclitrophicus, unidentified B, DRD P20 22 V. tapetis V. tapetis A, DRP P21 19, 69, 82 V. pelagius II V. cyclitrophicus, unidentified B, DRD P22 37 V. cicinnantiensis Unidentified B, DRP, DRD P23 37 V. pectenicida Unidentified B, DRP P24 6, 47 V. mediterranei V. mediterranei, V. splendidus A, C, DRP P25 NT V. furnisii NT B P26 NT V. gazogenes NT C P27 Unclustered V. gallicus Unidentified DRD P28 Unclustered V. agarivorans Unidentified A P29 9 V. wodanis Unidentified B 18 Table 2. Sequence similarities (%) of the genes 16S rRNA, recA and rpoA of three potential new Vibrio species and closely related species. Species R33677 (Cluster 37) 16S recA rpoA R33624 (Cluster 48) 16S recA rpoA R33755 (Cluster 86) 16S recA rpoA V. comitans V. rarus V. inusitatus V.splendidus V. gigantis V. pomeroyi A. wodanis A. fischeri A. logei A. salmonicida 99.4 93.0 99.8 99.1 84.0 94.0 99.0 91.0 96.0 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 97.6 86.5 97.8 97.3 86.0 97.0 97.1 86.0 97.8 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 98.1 82.0 98.1 97.1 89.0 97.2 97.8 82.0 97.4 96.0 81.0 97.5 19 Table 3. Accession numbers for sequences of the genes 16S, recA and rpoA determined in the laboratory. Strain Cluster 16S recA rpoA R33660 R33635 R33634 R33656 R33677 R33624 R33727 R33755 V. gigantis DSM 18531T V. crassostreae LMG 22240T V. comitans LMG23416T V. rarus LMG 23674T V. inusitatus LMG 23434T 5 9 16 26 37 48 70 86 - - - - - EF599163 EF500162 EU541607 EU541606 EF599161 EU541605 EF599164 EU541604 - - - - - EU541589 EU541590 EU541592 EU541591 EU541585 EU541587 RU541588 EU541586 EU541593 EU541594 EU541597 EU541599 EY541600 EU541569 EU541570 EU541572 EU541571 EU541565 EU541566 EU541568 EU541567 EU541573 EU541574 EU541577 EU541578 EU541579 20 Figure legends Fig. 1.- Sampling sites in the Galician coast (NW Spain). Fig. 2.- NJ Phylogenetic tree based on concatenated sequences of three genes recA, rpoA, and the 16S rRNA gene. Horizontal branch lengths are proportional to evolutionary divergence. Bootstrap percentages from 1000 replicates appear next to the corresponding branch. Similar results were obtained using the Maximum-Parsimony method (data not shown). 21 Fig. 1.- Beaz-Hidalgo et al. D A C B 22 Fig. 2. BeazHidalgo et al. V. furnisii ATCC 35016T V. fluvialis ATCC 33809T V. proteolyticus ATCC 15338T V. cholerae ATCC 14035T V. cincinnatiensis ATCC 35912T V. gazogenes ATCC 29988T V. penaeicida LMG19663T V. tubiashii ATCC 19109T V. nereis ATCC 25917T V. natriegens ATCC14048T V. alginolyticus ATCC 17749T V. campbelli ATCC 25920T V. harveyi ATCC 14126T (X74706.1) V. aginolyticus group V. pectenicida A365 (Y13830.1) V. scophthalmi CECT 4638T (VSU46579) V. ichthyoenteri LMG 19664T (AJ437192.1) V. aestuarianus ATCC 35048T (X74689.1) V. ordalii ATCC 33509T (X74718.1) V. mediterranei CIP 103203T (X74710.1) V. pelagius ATCC 25916T (X74722.1) V. chagasii LMG 13237 (AJ490157.1) R33624 (Cluster 48) V. cyclitrophicus LMG 21359T (AM162656.1) R33634 (Cluster 16) V. lentus CECT 5110T (AJ278881.1) R33660 (Cluster 5) V. tasmaniensis LMG 21574T (AJ514912.1) V.splendidus ATCC 33125T (X74724.1) V. kanaloae LMG 20539T (AM162657.1) R33635 (Cluster 9) R33727 (Cluster 70) V. crassostreae CAIM 1405T (EF094887.1) R33656 (Cluster 26) V. gigantis CAIM 25T (EF094888.1) V. pomeroyi LMG 20537T (AJ491290.1) V.comitans LMG 23416T (DQ922915) V.inusitatus LMG 23434T (DQ922920) R33677 (Cluster 37) V. halioticoli IAM14596T (AB000390) V. neonatus LMG 19972T (AY426979) V. ezurae LMG 19970T (AY426980) V. rumoiensis LMG 20038T (AB013297.1) R33755 (Cluster 86) A. fisheri ATCC 7744T (X74702.1) A. wodanis NVI 88/441T (AJ132227.1) A. salmonicida NCMB 2262T (X70643) A. logei NCIMB 2252T (AJ437616.1) 100 100 100 100 94 100 100 100 100 94 100 100 100 100 63 43 59 93 65 97 60 38 52 25 28 36 63 25 22 100 100 93 84 92 46 39 73 61 100 44 37 41 57 30 0.01 A. fischeri group V. halioticoli group V. splendidus group