Identifying Fishes through DNA Barcodes and Microarrays

Marc Kochzius,Aglaia Antoniou,Serge Planes,Manfred Nölte,Janet Hauschild,Alessia Cariani,Cemal Turan,Caroline Hervet,Christian Seidel,Daniel Campo,Dietmar Blohm,Monica Landi,Fausto Tinti,H Weber ,K Van Kappel ,Eva García Vázquez ,V Marteinsson ,Guðmundur Ó Hreggviðsson ,Sigríður Hjörleifsdóttir ,Sandeep K Botla ,Antoniοs Magoulas ,Moleyur M.n Venugopal

doi:10.1371/journal.pone.0012620

Abstract

BackgroundInternational fish trade reached an import value of 62.8 billion Euro in 2006, of which 44.6% are covered by the European Union. Species identification is a key problem throughout the life cycle of fishes: from eggs and larvae to adults in fisheries research and control, as well as processed fish products in consumer protection.Methodology/Principal FindingsThis study aims to evaluate the applicability of the three mitochondrial genes 16S rRNA (16S), cytochrome b (cyt b), and cytochrome oxidase subunit I (COI) for the identification of 50 European marine fish species by combining techniques of “DNA barcoding” and microarrays. In a DNA barcoding approach, neighbour Joining (NJ) phylogenetic trees of 369 16S, 212 cyt b, and 447 COI sequences indicated that cyt b and COI are suitable for unambiguous identification, whereas 16S failed to discriminate closely related flatfish and gurnard species. In course of probe design for DNA microarray development, each of the markers yielded a high number of potentially species-specific probes in silico, although many of them were rejected based on microarray hybridisation experiments. None of the markers provided probes to discriminate the sibling flatfish and gurnard species. However, since 16S-probes were less negatively influenced by the “position of label” effect and showed the lowest rejection rate and the highest mean signal intensity, 16S is more suitable for DNA microarray probe design than cty b and COI. The large portion of rejected COI-probes after hybridisation experiments (>90%) renders the DNA barcoding marker as rather unsuitable for this high-throughput technology.Conclusions/SignificanceBased on these data, a DNA microarray containing 64 functional oligonucleotide probes for the identification of 30 out of the 50 fish species investigated was developed. It represents the next step towards an automated and easy-to-handle method to identify fish, ichthyoplankton, and fish products.

Highlights

World fishery production reached 143.6 million tons in 2006, 77% of which were used for complete cyt b and rhodopsin gene sequences of European marine fish species [21], as well as partial 16S rRNA gene (16S), cyt b, and cytochrome oxidase subunit I (COI) sequences of anchovies [11] to enable a sequence-based identification of specimens.in course of developing a unifying identification system for animal species an universal marker has been proposed to serve as a so-called ‘‘DNA barcode’’ [22,23]
FN688905–FN689348) sequences of 50 fish species from European seas was obtained and these sequences are available at the EMBL sequence data base
Insertions, and deletions were observed in the cyt b and COI sequences, indicating that they represent

Summary

Introduction

In course of developing a unifying identification system for animal species an universal marker has been proposed to serve as a so-called ‘‘DNA barcode’’ [22,23]. This DNA barcode is the sequence of the ‘‘Folmer fragment’’ [24], a polymorphic part of the mitochondrial cytochrome oxidase subunit I gene (COI), which can be used to identify closely related species as well as higher taxa in many animal phyla. The applicability of COI for species identification in fish [25] triggered the international initiative for barcoding all fishes (FISHBOL; www.fishbol.org) [3,26]. Species identification is a key problem throughout the life cycle of fishes: from eggs and larvae to adults in fisheries research and control, as well as processed fish products in consumer protection

Objectives

Methods

Results

Conclusion