Genetic differentiation has been observed in marine species even when no obvious barriers to gene flow exist, and understanding such differentiation is essential for effective fisheries management. Highly differentiated outlier loci can provide information on how genetic variation might contribute to local adaptation but may also be affected by historical demographic events. A locus which aligned to a predicted zona pellucida sperm-binding protein 3 gene (ZP3) in Atlantic cod (Gadus morhua) was previously identified as the highest outlier based on F_ST in a RADseq study of Pacific cod (Gadus macrocephalus) across the West Coast of North America. However, because of the limited length of the RAD sequence and restricted geographic area of sampling, no conclusion on the functional significance of the observed variation was possible. In other marine species, ZP3 is involved in reproductive isolation, local adaptation, and has neofunctionalized as an antifreeze gene, and so it may provide important insights in functional population structure of Pacific cod. Here, we sequenced a 544 bp region of ZP3 in 230 Pacific cod collected from throughout their geographic range. We observed striking patterns of spatial structuring of ZP3 haplotypes, with a sharp break near Kodiak, Alaska, USA where populations within ~200 km of each other are nearly fixed for different haplotypes, contrasting a pattern of isolation by distance at other genetic markers in this region (F_ST = 0.003). Phylogenetic analysis of ZP3 haplotypes revealed that the more southern haplotypes appear to be ancestral, with the northern haplotype evolving more recently, potentially in response to a novel selective pressure as Pacific cod recolonized northern latitudes after glaciation. The sharp break in haplotype frequencies suggests strong selective pressures are operating on small spatial scales and illustrates that selection can create high divergence even in marine species with ample opportunities for gene flow.

PCRs for the variable region of the putative zona pellucida subunit 3 gene (primer ZP_GM, F: 3' gcaatctgagggtaggacca 5', R: 3' aacgcagtgatccacaaaga 5') were carried out in a 25 µl volume, with Phusion 5X buffer (New England Biolabs, Ipswich, MA), 10mM dNTPs, 10µM forward and reverse primers, 0.2µL Phusion Taq polymerase and approximately 200ng DNA. Thermal cycling conditions were 98°C for 30 sec, followed by 5 cycles of 98°C for 10 sec, 63-59°C touchdown for 30 seconds (-1°C each cycle for 5 cycles) and 72°C for 30 seconds, and then by 30 cycles of 98°C for 10 seconds, 58°C for 30 seconds and 72°C for 30 seconds, and concluded at 72°C for 5 minutes.

A total of 230 Pacific cod were sequenced to screen 544 bp of the variable region of ZP3 using primers ZP_GM (Table 1). Sanger sequencing was performed bidirectionally using forward and reverse primers at MCLAB (320 Harbor Way, South San Francisco, CA). Contigs were aligned using Sequencher v. 5.0 (Gene Codes Corporation, Ann Arbor, MI) and scores below 80% quality were discarded. Sequence calls were double checked and confirmed by two readers when ambiguities were present. Consensus sequences were aligned in BioEdit v. 7.2 (Hall, 1999). Sanger DNA sequence data was transformed from unphased sequences to fasta files with two haplotypes for all individuals. Haplotypes for ambiguous (heterozygous) nucleotides at ZP3 segregating sites were inferred using Bayesian methodology with a priori expectations based on coalescent theory in PHASE v2.1.1 and SeqPHASE (Stephens, Smith, & Donnelly, 2001; Flot, 2010). The most likely pair of haplotypes was selected for each individual based on the highest posterior probability.

Flot, F.J. (2010). SeqPHASE: a web tool for interconverting PHASE input/output files and FASTA sequence alignments, Molecular Ecology Resources, 10 (1), 162–166.

Stephens, M., Smith, N.J., & Donnelly, P. (2001). A new statistical method for haplotype reconstruction from population data. The American Journal of Human Genetics, 68(4), 978–989.