Final Technical Report Food Standard Agency Project Q01099 Extending the Fish Species Lab-on-a-Chip Capillary Electrophoresis PCR-RFLP Database Steve Garrett, John Dooley, Marie-Anne Clarke & Helen M. Brown. Department of Chemistry & Biochemistry Campden BRI Gloucestershire, GL55 6LD 1 Executive Summary Food fraud is a significant problem which affects both the industry and consumer. Manufacturers have to ensure the authenticity of food ingredients to enable compliance with food safety and labelling regulations. If they are found to be supplying incorrectly labelled products they run the risk of prosecution, damage to brand image and incur substantial financial penalties. As part of a programme to develop analytical procedures suitable for use by local authority food enforcement laboratories, the Food Standards Agency has supported the development of a number of DNA based food authenticity tests which use the Agilent 2100 Bioanalyzer, a lab-on-a-chip capillary electrophoresis (CE) end-point detection system. It is a relatively low-cost, user-friendly system, which is suitable for uptake by laboratories that do not have extensive facilities and expertise in DNA analysis. A lab-on-a-chip CE DNA profiling approach for fish species identification was originally developed and evaluated in FSA project Q010697. The method used Polymerase Chain Reaction (PCR) amplification of a 464bp region of mitochondrial target in the cytochrome b (cyt b) gene and three restriction enzyme digests to produce species specific DNA profiles. The main objective of this follow up project Q01099 was to extend the range of species covered in the cyt b 464 bp PCR-RFLP (Restriction Fragment Length Polymorphism) profiles database. This was initially carried out by obtaining authentic fish species and generating cyt b DNA sequence data experimentally. However, during the course of the project, sequence data was also obtained from the ‘Fish Trace’ project. This collaborative study funded by the European Commission generated cyt b gene sequence information on thousands of authentic fish samples from different catch locations in Europe, covering approximately 200 different species. Following retrieval of the Fish Trace cyt b sequences, a large database of theoretical PCR-RFLP profiles was constructed using in-house bioinformatics approaches. The database contained the PCR-RFLP profiles for the three restriction endonucleases, Dde I, Hae III and Nla III, identified in project Q01069 as the most appropriate for the discrimination of fish species. Approximately 30 of the 200 Fish Trace species exhibited intra-species profile variability. In these species an altered DNA sequence either led to a further enzyme cutting site or loss of a cutting site. For these species, only one of the three enzyme profiles was affected. Some of the variations in profile appeared to be linked to the catch location of the fish samples but the extensive study of the fish populations and profiles differences were not the focus of this work. In some instances closely related species gave the same profiles for the three enzymes. For example, Yellowfin tuna and Bigeye tuna gave the same profile. Therefore the use of other restriction enzymes in addition to Dde I, Hae III, Nla III would be required to enable species identification in such instances. The Fish Trace database was then challenged with experimental data derived from 35 authentic and commercial fish samples. The lab-on-a-chip CE system produced DNA fragments that were consistent with the theoretical profiles, although there was an overestimation of the sizes with the Series II DNA 1000 lab chip kit. The original DNA 500 lab chip kit used in Q01069 produced PCR-RFLP 2 DNA fragment sizes that were closer to the theoretical sizes, however these kits have been replaced by the Series II DNA 1000 lab chip kit in 2006. A further objective for this project was the development of alternative PCR-RFLP profile approaches for canned salmon and tuna. The 464 cyt b target was too large to be amplified from canned products due to difficulties in amplification from highly degraded DNA, so alternative smaller cyt b targets were chosen. For salmon, a 168 bp target was identified by comparison of a number of salmon species cyt b sequences. Two restriction enzymes, Bfa I and Dde I, allowed discrimination of Coho, Chum, Pink, Red and Chinook salmon species. These five species are found in the Pacific Ocean and are associated with commercial products. A PCR-RFLP assay developed was successfully applied to canned materials. For canned tuna, an existing published PCR-RFLP assay15 was evaluated for use on the lab-on-a-chip CE system. The method used a 176 bp cyt b amplicon and three enzymes, MboI, Bsi YI and MnlI, to discriminate Albacore, Yellowfin tuna, Bigeye, Bluefin and Skipjack tuna species. In a blind study, all the unknown canned tuna samples were identified correctly by the analyst. Some work was carried out to determine the feasibility of applying a published method18 for discriminating between King and Queen scallops to the lab-on-a-chip CE system. These are the two most common scallop species available to the UK consumer, with the King being the higher quality and more expensive of the two. The method used amplification of an ITS2 region (an internal transcribed spacer region which flanks the 5.8S ribosomal RNA gene) followed by digestion with the enzyme Nla III. Although the method was successfully applied for lab-on-a-chip analysis, there was a lack of available authentic samples to fully validate the method. The final objective involved testing the feasibility of using the cytochrome oxidase I (COI) gene as an alternative target for PCR-RFLP identification of fish species. This is the target sequence used by an international collaboration in molecular taxonomy to enable the 'barcoding' of all living organisms (http://www.dnabarcodes.org/). The 'barcoding ' PCR primers were applied to DNA from over 30 fish species and the resulting PCR products were sequenced. Unlike the 464 cyt b target, different COI primer combinations had to be used in order to achieve amplification for all species. Also, the size of the COI target sequence was approximately 700bp, making it considerably larger than the 464bp cyt b target and less amplifiable from processed fish products. Theoretical COI PCR-RFLP profiles were produced for all the species again using the enzymes DdeI, Hae III and Nla III. Unique profiles were produced for 29 species. The COI PCR-RFLP would make an appropriate alternative target if the amplicon size could be reduced and a single set of primers could be developed. 3 GLOSSARY ................................................................................................................................... 7 1 INTRODUCTION ......................................................................................................................... 9 1.1 FISH SPECIES IDENTIFICATION USING CYTOCHROME B PCR-RFLP ................................................................ 9 1.2 BENEFITS OF LAB-ON-A-CHIP CE ........................................................................................................... 9 1.3 PCR-RFLP AND SEQUENCE DATABASES................................................................................................ 10 1.4 SMALL AMPLICONS FOR CANNED FISH SPECIES IDENTIFICATION. ................................................................ 10 1.5 OTHER SEAFOOD IDENTIFICATION ........................................................................................................ 11 2 MATERIALS AND METHODS ..................................................................................................... 12 2.1 SEAFOOD MATERIALS ........................................................................................................................ 12 2.1.1 Authentic fish samples .......................................................................................................... 12 2.1.2 Additional commercial fish samples ..................................................................................... 14 2.2 EXTRACTION OF DNA FROM FISH MATERIALS ........................................................................................ 14 2.2.1 CTAB DNA extraction method ............................................................................................... 15 2.2.2 Promega’s Maxwell™ 16 DNA extraction robot ................................................................... 15 2.3 DNA ANALYSIS ................................................................................................................................ 15 2.3.1 DNA amplification ................................................................................................................. 15 2.3.1.1 464bp cyt b amplicon (general fish) ............................................................................................................... 15 2.3.1.2 168bp cyt b amplicon (canned salmon) ......................................................................................................... 15 2.3.1.3 176bp cyt b amplicon (canned tuna) ............................................................................................................. 16 2.3.1.4 ITS-2 amplicon (scallop ) ................................................................................................................................. 16 2.3.1.5 700bp cytochrome oxidase I (COI) amplicon .................................................................................................