Introduction

The Atlantic cod (Gadus morhua L.) is a well-known teleost species inhabiting the Atlantic Ocean from North Carolina to Greenland in the west and from the Bay of Biscay to the Arctic Ocean in the east. The species has a long history as an important fish for consumption, but in recent decades stocks have declined dramatically particularly in the eastern Atlantic (Christensen et al. 2003). The species is presently an attractive candidate for aquaculture with commercial farming already underway in Norway, Canada and Scotland. Selection programmes have been established in Norway, Canada and Iceland to develop stock for the aquaculture industry that performs well on commercial traits such as growth rate and disease resistance.

Genetic markers are imperative for proper management of cod, for cod aquaculture, as well as for ensuring a sustainable coexistence of wild and farmed stocks. There is a need for additional genetic markers for refining the stock structure and assigning fish to populations as well as for improvements in aquaculture production both for pedigree tracing and quantitative trait loci (QTL) identification (Nielsen et al. 2006; Westgaard and Fevolden 2007). Recently, we have published a number of microsatellite markers derived from both genomic library enriched for repetitive sequences (Wesmajervi et al. 2007; Westgaard et al. 2007b) and expressed sequence tags (ESTs) (Stenvik et al. 2006; Westgaard et al. 2007a). In contrast to genomic microsatellites, EST-derived microsatellite markers have the potential to be functional markers (Varshney et al. 2005). Here we report the isolation and characterisation of 20 novel microsatellite loci containing di- and tri- nucleotide repeats, which add to those previously reported (Brooker et al. 1994; Miller et al. 2000; O’Reilly et al. 2000; Jakobsdottir et al. 2006; Wesmajervi et al. 2007; Westgaard et al. 2007a, b; Stenvik et al. 2006).

In an ongoing genomics program we are generating ESTs from cDNA libraries of different developmental stages of Atlantic cod (unpublished results). Approximately 1700 ESTs from cod embryos were investigated for the presence of microsatellite motifs using Tandem Repeats Finder software (Benson 1999) after setting the alignment score threshold to 30. A set of 54 flanking primer pairs were designed for EST sequences containing di- and tri- nucleotide motifs using Primerselect (DNASTAR). For fluorescent detection the forward primers were dye labelled (PET, NED, VIC or 6-FAM) (Applied Biosystems; AB). A PIG-tail sequence 5′-GTTTCTT (Brownstein et al. 1996) was included in the reverse primers (Table 1). Genomic DNA from 96 unrelated individuals, originated from randomly selected wild-caught Norwegian coastal- and North-East Arctic cod, was analysed to evaluate the efficiency of amplification and polymorphism of the markers by polymerase chain reaction (PCR) in a GeneAmp PCR system 2720 (AB). In order to utilise a single-step co-amplification of microsatellites, four markers (labelled with different dyes) were combined in a single reaction using Qiagen Multiplex PCR kit (Qiagen, Norway). The amplification was carried out in 10 μl reaction volume containing 10–50 ng template, 1× Qiagen Multiplex PCR Master Mix (3 mM MgCl2) 10 nM of each forward and reverse primer. Thermal cycling conditions were: 15 min at 95°C, followed by 32 amplification cycles at 94°C for 30 s, 56°C for 90 s and 72°C for 60 s and then a final extension at 72°C for 30 min. Genotyping was performed using a 3130xl Genetic Analyzer (AB). Data were collected automatically and sized with GeneMapper software (AB) using the GeneScan-500-LIZ size standard (AB). Out of 54 primer pairs 20 were readily scorable and polymorphic (Table 1). Statistical calculations were performed using the program Genepop version 3.4 (Raymond and Rousset 1995). The number of alleles per locus (A), observed heterozygosity (H O), expected heterozygosity (H E) and inbreeding coefficient (F IS) were calculated. Possible genotyping errors were examined with MICRO-CHECKER version 2.2.3 (Van Oosterhout et al. 2004). The number of alleles varied from 2 to 10 per locus and the observed and expected heterozygosity ranged from 0.01 to 0.67 and 0.03 to 0.70, respectively (Table 1). The loci were further tested for both Hardy–Weinberg equilibrium (HWE) and linkage disequilibrium and significance levels were determined after 100 batches of 5,000 iterations each. Three of the loci (Gmo-C213, Gmo-C246 and Gmo-C247) deviated from HWE after sequential Bonferroni correction (Rice 1989), probably due to the small sample size, genetic drift and the presence of null alleles. Four of 253 pairwise exact tests rejected genotypic equilibrium between loci (Gmo-C213 and Gmo-C222, Gmo-C233 and Gmo-C229, C223 and Gmo-C236 and C229 and Gmo-C236) after sequential Bonferroni correction. BLAST analysis gave highly significant hits for 10 loci (Table 2). These represent well-characterised genes thus defining the associated microsatellites as Type I markers, all located to the 3′UT region of the genes. In conclusion, these 20 polymorphic microsatellite loci provide useful tools for studies of population genetics, reproductive ecology and for constructing linkage maps of Atlantic cod.

Table 1 Characteristics of 20 microsatellite loci isolated from Atlantic cod (Gadus morhua L.)
Full size table
Table 2 Annotation of gene-associated microsatellites by BLAST analysis
Full size table