Die weißbeinige Garnele, Litopenaeus vannamei, wurde erstmals als Penaeus vannamei beschrieben (Boone, 1931) und gehört zur Familie der Penaeidae. Ein weiterer Vertreter dieser Familie ist die Tigergarnele (Penaeus monodon). Das natürliche Habitat von L. vannamei ist der östliche Pazifik, speziell die Küstengewässer Mexikos, Mittelamerikas bis zu den peruanischen Küstengewässern im Süden. Hier findet man diese Garnelenart in der epipelagischen Zone der Pazifikküsten in Wassertiefen bis 72 Meter (Holthuis, 1980). Wassertemperatur, sowie Salinität unterliegen jahreszeitlichen, bzw. wetterbedingten Schwankungen, jedoch liegen in diesen tropischen, marinen Habitaten die Wassertemperaturen das ganze Jahr bei 20°C oder höher und die Salinität bei ca. 34‰ (US Department of Commerce). Die Jugendstadien dieser Garnelenart (Postlarven, Juvenile, Subadulte) kann man in den Küstengewässern (Lagunen, Mangroven, Mündungsgebiete) finden, wohingegen die adulten Tiere vor allem zur Fortpflanzung auch in tiefere Gewässer wandern (Food and Agriculture Organization of the United Nations, 2018). Aufgrund der Wanderungen während der verschiedenen Lebensphasen ist diese Garnelenart extremen Schwankungen in den Umweltbedingungen ausgesetzt. Schwankungen der Wassertemperaturen zwischen 20 und 36 °C stellen eine Herausforderung für diese Tiere dar. Durch extreme Wetterbedingungen (Regenzeit, Trockenzeit) sind diese Tiere in den Lagunen- und Mangrovengebieten Veränderungen in der Salinität von 2 bis 60 ‰ ausgesetzt. Diese extremen Bedingungen haben zur Folge, dass für juvenile Lebensstadien eine natürliche Mortalität von bis zu 90 % beschrieben ist (Edwards, 1977).
Der Lebenszyklus von L. vannamei beinhaltet verschiedene Stadien. Aus den Eiern schlüpfen nach ca. 12 bis 18 Stunden Nauplien die sich nach 2 Tagen zu Zoealarven, nach vier bis fünf Tagen zu Mysislarven und nach weiteren drei bis vier Tagen zu Postlarven weiterentwickeln. Das Postlarvenstadium dauert 10 bis 15 Tage und beinhaltet viele Unterstadien (PL 1 bis PL 30) bis sich die Tiere schließlich zur subadulten und nach mehreren Monaten zur adulten Garnele entwickeln (Food and Agriculture Organization of the United Nations, 2018). Mit einer Größe von maximal 23cm gehören sie zu den sog. Riesengarnelen und sind eine der am häufigsten
22
kultivierten und beliebtesten Garnelenarten weltweit. Sie werden in Mittel- und Südamerika (Brasilien, Ecuador, Mexiko, Venezuela, Honduras, Guatemala, Peru, u.a.), in Asien und Südostasien (China, Thailand, Indonesien, Vietnam, Malaysia, Taiwan, Indien, Philippinen u.a.) aber auch in den Vereinigten Staaten von Amerika gezüchtet (Food and Agriculture Organization of the United Nations, 2018).
Die Riesengarnele L. vannamei erfreut sich in Deutschland zunehmender Beliebtheit.
Laut FAO stieg die Produktion von L. vannamei in Deutschland von einer Tonne im Jahr 2010 auf acht Tonnen im Jahr 2017 (Food and Agriculture Organization of the United Nations, 2018). Die Entwicklungen der letzten Jahre lässt deutlich höhere Produktionsmengen für die Jahre 2018 und 2019 erwarten, da jährlich neue Kreislaufanlagen zur Produktion dieser Riesengarnelen hinzukommen. Somit stellen sie ideale Modelltiere für die Untersuchungen von Vibrionen in Kreislaufanlagen dar.
Im Rahmen dieses Projektes wurde versucht diese ubiquitär vorkommenden und nützlichen, aber auch potentiell problematischen Bakterien sicher zu identifizieren, um herauszufinden, ob potentiell pathogene Spezies in RAS vorkommen (Publikation 1).
Außerdem wurde durch ein Screening der Isolate auf das Vorhandensein genetischer Loci, die für Pathogenitätsfaktoren kodieren, versucht die potentielle Gefahr für bakterielle Infektionen in RAS einzuschätzen (Publikation 2).
23
3 Publikation 1
Dies ist die publizierte Version des folgenden Artikels, der unter [https://doi.org/10.1111/jfd.12897] veröffentlicht wurde. Dieser Artikel darf für den nicht-kommerziellen Gebrauch, in Übereinstimmung mit der „Wiley Self-Archiving Policy“ [http://www.wileyauthors.com/self-archiving], verwendet werden.
This is the published version of the following article, which has been published at [https://doi.org/10.1111/jfd.12897]. This article may be used for non-commercial purposes in accordance with the Wiley Self-Archiving Policy [http://www.wileyauthors.com/self-archiving].
Bauer, J, Teitge, F, Jung, A, Adamek, M, Peppler, C, Steinhagen, D, Jung-Schroers, V, 2018. Recommendations for identifying pathogenic Vibrio spp. as part of disease surveillance programmes in Recirculating Aquaculture Systems for Pacific white shrimps (Litopenaeus vannamei), Journal of Fish diseases, DOI 10.1111/jfd.12897
24
Recommendations for identifying pathogenic Vibrio spp. as part of disease surveillance programmes in Recirculating Aquaculture Systems for Pacific white shrimps (Litopenaeus vannamei)
Julia Bauer1*, Felix Teitge1, Lisa Neffe1, Mikolaj Adamek1, Arne Jung2, Christina Peppler3, Dieter Steinhagen1, Verena Jung-Schroers1
1 Fish Disease Research Unit, University of Veterinary Medicine Hannover, Hannover, Germany
2 Clinic for Poultry, University of Veterinary Medicine Hannover, Hannover, Germany
3 Polyplan GmbH, Bremen
* Corresponding author. tel.: +49 511 9538578; fax: +49 511 9538587.
Postal address: Fish Disease Research Unit, Centre of Infectious Diseases, University of Veterinary Medicine Hannover, Bünteweg 17, D-30559 Hannover, Germany
E-mail address: julia.bauer@tiho-hannover.de
Abstract
Due to their pathogenic potential, identifying Vibrio species from recirculating aquaculture systems (RAS) for Pacific white shrimp (Litopenaeus vannamei) is of great importance to determine the risk for animal's as well as for the consumer's health. The present study compared identification results for a total of 93 Vibrio isolates, including type strains and isolates from shrimp aquaculture. Results from biochemical identifications, 16S rRNA sequencing, sequencing of the uridylate kinase encoding gene pyrH and analysis of the protein spectra assessed by MALDI‐TOF MS were compared. The results achieved by these different methods were highly divergent for many of the analysed isolates and for several Vibrio spp. difficulties in reliably identifying occurred. These difficulties mainly resulted from missing entries in digital databases, a low number of comparable isolates analysed so far, and high interspecific similarities of biochemical traits and nucleotide sequences between the closely related
25
Vibrio species. Due to the presented data, it can be concluded that for identifying Vibrio spp. from samples in routine diagnostics, it is recommended to use MALDI‐TOF MS analysis for a quick and reliable identification of pathogenic Vibrio sp. Nevertheless, editing the database, containing the main spectra of Vibrio is recommended to achieve reliable identification results.
Introduction
Production of aquatic animals for human consumption in recirculating aquaculture systems (RAS) is increasing in European countries, especially because of environmentally sustainable and ecological aspects. RAS are mostly operated as secondary mainstays using waste heat from biogas plants and have a planned aquaculture production quantity of between five and 30 tonnes per year.
Pacific white shrimps (Litopenaeus vannamei) are one of the most frequently cultivated shrimp species worldwide. Their native habitats are regions of the Eastern Pacific from Central to South America where water temperatures are constantly above 20°C. The main producing countries of L. vannamei are located in North‐ and South America and Asia (FOOD AND AGRICULTURE ORGANIZATION OF THE UNITED NATIONS, 2017). Although in Germany the production of penaeid shrimps, like L. vannamei, in RAS has also increased over the last few years, most shrimps consumed in Germany are still reared in third countries and approximately 40,000 tonnes of shrimps and related species are imported annually. Overall, the import of frozen shrimps into the European Union ranks second in value after imported salmon (EUMOFA—
EUROPEAN MARKET OBSERVATORY FOR FISHERIES AND AQUACULTURE PRODUCTS, 2016). These imports often exceed maximum residue levels of pharmaceuticals set by German law because of different regulations regarding treatments of food animals in the exporting countries.
Hence, notifications in the German Rapid Alert System for food and feed show multiple reports on imported shrimps in 2016 and 2017, such as drug residues (tetracycline, nitrofuran, furazolidone, chloramphenicol, malachite green), bacterial contaminations (Salmonella sp., V. alginolyticus, V. cholera, V. vulnificus) or the preservative
26
maximum levels (sulphide) being exceeded (FEDERAL OFFICE OF CONSUMER PROTECTION AND FOOD SAFETY, 2016, 2017). In many shrimp‐producing countries, prophylactic treatments against bacterial diseases are conducted on a routine basis, also using medications that are prohibited in many importing countries for use in food animals, such as malachite green (Wijegoonawardena & Siriwardena, 2000). For prophylactic treatments, also Watch List antibiotics are used (Holmström et al., 2003) and resistances in Vibrio spp. against reserve group antibiotics (Jeyasanta, Lilly, & Patterson, 2017) indicate a frequent use of these agents (World Health Organisation, W, 2017).
Considering the amount of annually imported shrimps into the EU and the mentioned problems in shrimp‐producing tropical countries, a regional production of L. vannamei would be a logical step towards achieving an ecological production. The intention to achieve ecological goals in aquaculture, such as reducing wastewater discharge from aquaculture plants and a reduction in the use of medication, and particular antibiotics, leads to a balancing act between saving resources and safeguarding animal health. A production of shrimps in RAS would allow a reduction of wastewater discharge but entails a more unstable biological balance than natural aquatic habitats, because of high stocking densities, low water exchange and high feeding rates. Therefore, the total bacterial count in the recirculating water may increase very quickly and even non‐
pathogenic bacteria can cause infections in shrimps. As part of the microflora of sea water, especially Vibrio spp. can often be found in high amounts in marine RAS (Urakawa & Rivera, 2006). When replicating very fast, Vibrio spp. are often the cause of bacterial infections in shrimps (Moriarty, 1997).
Vibrio spp. are prevailing organisms within the physiological microflora in marine environments and brackish waters. Around 80 species of Vibrio are described (Buller, 2014). Currently, 58 species of the genus Vibrio spp. have been divided into 14 clades according to results from a multi‐locus sequence analysis (Thompson et al., 2005;
Sawabe et al., 2013) to analyse and understand the evolution of these bacterial species. The amount of Vibrio spp. within the physiological microflora in marine and brackish habitats can reach up to 40% (Urakawa & Rivera, 2006).
27
In L. vannamei, Vibrio spp. can be found on the skin, in the digestive tract, in the hepatopancreas and in the haemolymph of healthy animals. Usually, Vibrio spp. are introduced to RAS via the stocking of shrimps (Vandenberghe et al., 1999). Most of these Vibrio species are harmless for the animals, and a symbiotic relationship between these bacteria and the host is of great importance (Urakawa & Rivera, 2006).
Nevertheless, there are different species of Vibrio spp. that are known to have a pathogenic potential for shrimps and can cause mass mortalities, resulting in massive economic losses.
In many incidents of disease or mortality, a high number of particular species, such as V. harveyi or V. alginolyticus, can be detected (Vandenberghe et al., 1999). Especially, V. harveyi seems to be associated with the infection of larvae as well as adult shrimps and can cause massive losses in Penaeus spp. (Karunasagar, Pai, Malathi, &
Karunasagar, 1994) and specifically in L. vannamei (Zhou et al., 2012). Also, V. alginolyticus seems to cause bacterial infections in L. vannamei (Liu, Cheng, Hsu,
& Chen, 2004).
Several of the shrimp pathogenic Vibrio species have a zoonotic potential as well and are described as causative agents for clinical infections in humans. Human infections may either arise from a contact with contaminated water or due to the consumption of contaminated seafood (Buller, 2014). Three major Vibrio spp. responsible for human gastrointestinal infections were detected in seafood products in France (Robert‐Pillot, Copin, Himber, Gay, & Quilici, 2014). V. parahaemolyticus was detected in 31.1%, V. vulnificus in 12.6% and V. cholerae in 0.6% of the tested products, respectively, and 25% of the V. parahaemolyticus isolates were positive for specific virulence genes.
These facts point out the risk of food‐borne infections due to the consumption of insufficiently heated seafood products. Another Vibrio sp. with a high zoonotic potential is V. fluvialis. Infections can present with severe enteritis (Allton, Forgione, & Gros, 2006), including diarrhoea (Chowdhury et al., 2012), but also cases of cellulitis and cerebritis are described (Huang & Hsu, 2005). This makes V. fluvialis an emerging pathogen and a severe concern to public health (Igbinosa & Okoh, 2010; Ramamurthy, Chowdhury, Pazhani, & Shinoda, 2014).
28
The health risks for shrimps, kept in RAS, as well as the zoonotic potential arising from specific Vibrio species, especially V. alginolyticus, V. fluvialis, V. harveyi, V. parahaemolyticus and V. vulnificus, underline the necessity for a rapid and accurate identification of these bacteria associated with shrimp aquaculture in order to prevent disease outbreaks and to maintain high product quality. In this study, we compared different diagnostic methods for identifying Vibrio spp. in general and especially for aforementioned species from shrimp aquaculture recirculation systems. Our aim was to identify a rapid diagnostic method to discriminate pathogenic Vibrio species from closely related environmental species in farmed shrimps.
Materials and Methods
Bacterial isolates
In total, 93 isolates of Vibrio spp. were analysed, whereof 82 isolates originated from different aquaculture facilities stocked with Pacific white shrimps (Litopenaeus vannamei) of different age groups and were collected during bacteriological examinations of the systems. Additionally, 11 type strains from the German Collection of Microorganisms and Cell Cultures (DSZM) were included in this study.
The isolates from RAS originated from water samples, swabs from the biofilm of RAS surfaces and swabs from the biofilm of shrimp surfaces. All samples were cultivated on Columbia sheep blood (CSB) agar, on Columbia sheep blood agar with the addition of 1.5% NaCl, and on Thiosulphate Citrate Bile Salts Sucrose (TCBS) agar (Oxoid Deutschland GmbH, Germany) at 25°C for 48 hr. Growing colonies were sub‐cultured on CSB agar at 25°C for a further 24 hr. Subcultures were morphologically described, and all colonies were separately dissolved in Veal Infusion Broth (Bacto Veal Infusion Broth, Becton, Dickinson and Company, New Jersey, USA) and stored at −80°C until further use. The type strains of different Vibrio spp. were processed in the same manner as the field isolates.
29 Molecular biological species identification Analysis of the 16S rRNA gene
Two different PCRs for the purpose of examining the nucleotide sequence of the 16S rRNA gene were performed. With one PCR (Jiang, Gao, Xu, Ye, & Zhou, 2011), that should result in a secure species identification, almost the whole 16S rRNA gene sequence with a length of 1465 base pairs (bp) was analysed for all 93 Vibrio spp.
included in this study. The resulting sequence was trimmed to a length of 1420 bp (variable regions V1‐V8). The result of this 16S rRNA sequencing was used as a standard identification method, and additional methods applied were tested against the result achieved with this method (Clarridge, 2004).
DNA was extracted from single colonies of each bacterial isolate using a commercially available DNA extraction kit in accordance with the manufacturer's instructions (Qiagen, Hilden, Germany). The DNA concentration was measured using spectrophotometry (NanoDrop ND‐1000 Lab, Peqlab Biotechnologie GmbH, Erlangen, Germany), adjusted to a concentration of 10 ng/μL with PCR grade water (Thermo Fisher Scientific) and kept at −20°C until further use. The PCR mixture consisted of 5x KAPA2G Buffer A (5 μl), 10 mM dNTPs Mix (0.2 mM each, 0.5 μl), forward and reverse primers (Jiang et al., 2011) (Table 1), each 10 mM (0.5 μl), 10 ng/μl of template DNA (5 μl) and 0.25 U/μl KAPA2G Robust HotStart Polymerase (0.05 μl) (VWR International GmbH, Erlangen, Germany). Nuclease‐free water was added to a final volume of 25 μl.
An endpoint PCR was performed using a thermocycler (SensoQuest, Göttingen, Germany) with the 27 f and 1492 r primers and PCR profile as listed in Table 1.
30
Table 1Primers and PCR protocols used in this study 1 = initial denaturating;2 - 4 = cycles (denaturation, annealing, extension); 5 = final extension
31
All PCR products were applied to a 1% agarose gel with the addition of 4 μl Gel Red Nucleic Acid DNA marker (Biotium, Inc., Freemont, USA) and 1x TBE (Tris‐boracic acid–EDTA) buffer. Afterwards, DNA amplicons were separated in an electrical field.
A DNA Ladder (100 base pairs (bp), Carl Roth GmbH, Karlsruhe) was used to determine the product size. Resulting bands were visualized under UV light at 302 nm.
PCR products that showed bands of the anticipated size were sent for sequencing to a commercial company and were sequenced from both sides (LCG genomics, Berlin, Germany). Sequences obtained from the samples were aligned using the multiple sequence alignment tool Kalign (Lassmann & Sonnhammer, 2005) and then subjected to phylogenetic analysis using the clustalW2 program and a neighbour joining algorithm available at https://www.ebi.ac.uk/Tools/phylogeny/simple_phylogeny/
(Larkin et al., 2007; Saitou & Nei, 1987; Kimura, 1980; McWilliam et al., 2013).
Resulting trees were visualized using FigTree v1.4.3.
In addition, a PCR was performed that amplifies a fragment of the 16S rRNA sequence of 720 base pairs. This PCR appeared more suitable for use in routine diagnostic, because it results in a smaller amplicon product which would require sequencing from only one side of the nucleotide. DNA extraction and PCR mixture were as described above, and the PCR was performed using the primer pair and PCR protocol reported by Wilson et al. (1990) as listed in Table 1. Sequences obtained with this PCR were trimmed to a length of 570 bp. Thus, a significantly shorter 16S rRNA sequence was obtained, which included only the variable regions V5‐V8 of the gene.
To identify Vibrio species, all obtained 16S rRNA nucleotide sequences were aligned to sequences from the EzBioCloud database (ChunLab, Inc., Seoul, Korea) as well as to sequences from the standard nucleotide BLAST (National Center for Biotechnology Information, U.S. National Library of Medicine, Bethesda, USA) database. All 16S rRNA sequences obtained during the current study are available at GenBank, as listed in Supporting Information Data S1.
32
Analysis of the uridylate kinase encoding gene pyrH
A 469‐bp fragment of the nucleotide sequence of the uridylate kinase encoding gene pyrH was amplified. PyrH was chosen because most Vibrio spp. could be securely differentiated from each other by means of the pyrH sequence (Thompson et al., 2005). DNA extraction and PCR reaction mixture were as described above. The PCR was performed using forward and reverse primers and the PCR protocol described by Thompson et al. (2005) as listed in Table 1. Obtained DNA sequences were aligned to known pyrH sequences using the standard nucleotide BLAST database (National Centre for Biotechnology Information, U.S. National Library of Medicine, Bethesda, USA, https://blast.ncbi.nlm.nih.gov/Blast.cgi). A phylogenetic tree was generated as described above. PyrH sequences obtained during this study are available at BankIt as detailed in Supporting Information Data S1.
Identification using mass spectrometry MALDI‐TOF
With MALDI‐TOF MS, bacteria can be identified according to their protein spectrum by mass spectrometry. For this analysis, all bacteria were grown as pure culture on CBS for 24 hr at 25°C. From each isolate, a single colony was applied in duplicate to a stainless steel target plate (MSP 96 target polished steel BC, microScout target, Bruker Corporation, Billerica, USA), air‐dried, and subsequently covered with 1 μl of the matrix solution (IVD matrix HCCA‐portioned, Bruker Corporation, Billerica, USA). As a reference, a bacterial test standard (IVD Bacterial Test Standard, Bruker Corporation, Billerica, USA) was applied on the target plate as well. Following air‐drying, the samples were analysed immediately using the Bruker microflex MALDI‐TOF system.
Mass spectra were acquired using the MALDI‐TOF MS spectrometer in a linear positive mode. The resulting protein spectra were compared to the Bruker Biotyper database using the MBT Compass Library (BTyp2.0‐Sec.Library 1.0) and the programs MBT compass and flexControl. Outputs of results were evaluated by Score Values: 2.3‐3.0: highly probable species identification, 2.0‐2.299: secure genus identification, 1.7‐1.99: probable genus identification, 0‐1.699: not reliable identification.
33 Biochemical Identification
The isolates were analysed by biochemical methods, using the manual microorganism identification systems API 20E (BioMerieux, Nuertingen, Germany) following the manufacturer's instructions except for incubation temperature, which was 25°C for all systems and the addition of 2% NaCl (w/v) in all test media. This identification system was applied because it is frequently used in laboratories when performing routine bacteriological diagnostics. The API strips were read after 24 hr and finally evaluated after 48‐h incubation time. Additionally, all isolates were also tested for β‐haemolysis, swarming of the bacteria on CSB agar, methyl red reaction, hydrolysis of aesculin, lactose, maltose, mannose, trehalose, xylose and salicin fermentation as well as growth on TCBS agar. All test media were supplemented with 2% NaCl (Table 3).
Bacterial characteristics were evaluated after an incubation of 24 hr at 25°C, and biochemical reactions were analysed after 48 hr of incubation at 25°C. Unclear or negative reactions were re‐evaluated after 72 and 96 hr of cultivation, respectively.
Species identification was performed by comparing the results with different databases, and a sample was assigned to the species with biochemical characteristics matching best to the test results. By using the apiweb database (BioMérieux sa, Lyon, France), an identification based only on the results of the Api 20E profiles was made.
This database contains biochemical characteristics of six Vibrio species (V. alginolyticus, V. cholerae, V. fluvialis, V. mimicus, V. parahaemolyticus and V. vulnificus). Additionally, all observed biochemical characteristics were compared to the identification tables provided by Buller that includes 78 different Vibrio species (Buller, 2014). Furthermore, the isolates were identified by using an online database (http://www.microrao.com/vibrio_ident.htm) which provides a tool for identifying Vibrio sp., considering up to 44 biochemical traits.