Redesign of PCR Primers for Mitochondrial Cytochrome C Oxidase Subunit I for Marine Invertebrates and Application in All-Taxa Biotic Surveys
Total Page:16
File Type:pdf, Size:1020Kb
Molecular Ecology Resources (2013) doi: 10.1111/1755-0998.12138 Redesign of PCR primers for mitochondrial cytochrome c oxidase subunit I for marine invertebrates and application in all-taxa biotic surveys J. GELLER,* C. MEYER,† M. PARKER† and H. HAWK*1 *Moss Landing Marine Laboratories, 8272 Moss Landing Road, Moss Landing CA 95309, USA, †Department of Invertebrate Zoology, Smithsonian Institution, National Museum of Natural History, Washington DC 20013-7012, USA Abstract DNA barcoding is a powerful tool for species detection, identification and discovery. Metazoan DNA barcoding is primarily based upon a specific region of the cytochrome c oxidase subunit I gene that is PCR amplified by primers HCO2198 and LCO1490 (‘Folmer primers’) designed by Folmer et al. (Molecular Marine Biology and Biotechnology,3, 1994, 294). Analysis of sequences published since 1994 has revealed mismatches in the Folmer primers to many meta- zoans. These sequences also show that an extremely high level of degeneracy would be necessary in updated Folmer primers to maintain broad taxonomic utility. In primers jgHCO2198 and jgLCO1490, we replaced most fully degener- ated sites with inosine nucleotides that complement all four natural nucleotides and modified other sites to better match major marine invertebrate groups. The modified primers were used to amplify and sequence cytochrome c oxi- dase subunit I from 9105 specimens from Moorea, French Polynesia and San Francisco Bay, California, USA repre- senting 23 phyla, 42 classes and 121 orders. The new primers, jgHCO2198 and jgLCO1490, are well suited for routine DNA barcoding, all-taxon surveys and metazoan metagenomics. Keywords: biotic surveys, cytochrome c oxidase subunit I, DNA barcoding, Moorea, universal primers Received 20 March 2013; revision received 22 May 2013; accepted 5 June 2013 databases at the Web of Science on 21 February 2013 Introduction revealed 2967 citations of the Folmer et al. (1994) paper. The mitochondrial cytochrome c oxidase subunit I gene DNA barcoding, the use of diagnostic nucleotide (COI) has been used extensively for studies of popula- variation to identify species, is a further development tion genetics, phylogeography, speciation and systemat- that has used the COI gene as a primary tool (Savolai- ics. For many species and genera, genetic variation at nen et al. 2005; Stoeckle & Hebert 2008; Ward et al. this locus is sufficient to study processes that occur over 2009; Bucklin et al. 2011). Where interspecific variation relatively short and recent time intervals. Despite this does not overlap intraspecific variation, species can variation, some regions of the gene are sufficiently con- be reliably identified by DNA sequences. Studies served to design primers for the polymerase chain reac- have shown that this condition often, but not always, tion (PCR) that match a broad spectrum of organisms applies to COI. DNA barcoding, therefore, includes a (Hebert et al. 2003; Kress & Erickson 2012). Conse- degree of uncertainty (Meyer & Paulay 2005). The quently, for many studies, COI occupies a ‘sweet spot’ of method both relies upon and is improved by large variation allowing for meaningful population and inter- data sets, and formal protocols for DNA barcoding species studies while conserved enough for practicality. have been proposed to promote uniformity of data Two primers (LCO1490 and HCO2198), commonly quality (Ratnasingham & Hebert 2007; Kress & Erick- referred to as the ‘Folmer primers’, have been used son 2012). For animal taxa, the region of COI flanked extensively (Folmer et al. 1994). Query of citation by the Folmer primers has been designated as a DNA barcode region (barcodeoflife.org), and records with sufficient metadata are given the keyword BARCODE Correspondence: Jonathan Geller, Fax: 1-831-632-4403; E-mail: by GenBank. [email protected] Despite the popularity and success of the Folmer 1Present address: Departement de Biologie, Universite Laval, primers described above, they are not truly ‘universal’ 1045 Avenue de la Medecine, Quebec QC G1V 0A6, Canada in applicability. Our own experience and informal © 2013 John Wiley & Sons Ltd 2 J. GELLER ET AL. conversations with colleagues indicate that the Folmer programs exist that include algorithms for some of these primers often fail or perform poorly, producing faint factors [e.g. PRIMER3 (Rozen & Skaletsky 2000), PRIMER PRE- products despite attempts at optimization. Failures of MIER 6 (Premier Biosoft, Palo Alto, CA, USA), AMPLICON the Folmer primers are at least in part due to mis- (Jarman 2004)]. However, these computer programs usu- matches with the target annealing position for many ally are meant to design primers for single sequences or taxa. Relatively few full-length COI gene sequences for groups of similar sequences. Design of primers for were available in 1994, when the primers were pub- alignments of highly divergent sequences is a more lished. Present-day analysis of full-length COI difficult task. sequences often reveals mismatches with the primers. In this study, we describe changes to the standard For this reason, a common strategy has been to obtain Folmer primers that were meant to correct mismatches a few sequences using Folmer primers under nonstrin- for many marine invertebrate species. In alignments of gent annealing temperatures, with additives, or with COI, variation in 3rd codon positions is extensive, and reamplification of weak products, then to design new no primer of 20–30 bp can be suggested that is 100% primers specific for the particular study (e.g. lepidopt- conserved across animal phyla. However, degeneracy in erans: Hebert et al. 2004; fishes: Ward et al. 2005; bry- the PCR primer can accommodate this variation in the ozoans: Mackie et al. 2006). This primer customization priming region. Degeneracy is created during primer is tolerable for focal taxon studies, but workflows for synthesis by mixing nucleotides at the variable sites, major biodiversity surveys, routine DNA barcoding or thereby creating a pool of primers containing all variants. identification of unknowns preclude frequent primer This has the downside of diluting the effective primers: redesign. only a small proportion of the primer mix will be an More universal COI primers would be useful for exact match to any template, and many primers in the biodiversity and barcoding studies. The existing body of pool will poorly match the target sequence. Further, data and the literature for the fragment flanked by sequence variation in the primer pool makes it difficult the Folmer primers places a positional constraint on pri- to predict interprimer interactions or the potential for mer design. A few alternative primers have been sug- mispriming. A different approach is to use inosine gested. Meyer (2003) redesigned the Folmer primers, nucleotides (dITP) at variable positions. Inosine nucleo- using sequences in the Folmer et al. (1994) paper to make tides form pair bonds with all natural nucleotides, them degenerate in the 3′ region (dgLCO1490 and thus increasing potential target sequences (PTS) without dgHCO2198). Meusnier et al. (2008) proposed ‘mini increasing degeneracy. barcode’ primers that amplify an internal fragment of In this study, we suggest a more universal version of the standard barcode region from a broader swath of the Folmer primers using degenerate positions and inter- taxa at the cost of sequence length. A 21 February 2013 nal inosines and show its applicability to biotic surveys survey of the BOLDSystems public primer database, the in Moorea, French Polynesia and San Francisco Bay, CA. repository for DNA barcoding primers maintained by We conclude that the redesigned primers are broadly the Barcode of Life Database, revealed minimally 418 dif- applicable and complement the standard Folmer primers ferent primers targeting the COI gene for various taxa. in DNA barcoding applications. An alternative to the Folmer primers is thus highly desirable. Methods Fortunately, mitochondrial genomics has flourished in recent years, and over 3000 complete or nearly com- Cytochrome c oxidase subunit I sequences for represen- plete mitochondrial genomes were known in 2011 from tative marine invertebrate taxa were acquired from Gen- 28 phyla (http://mi.caspur.it/mitozoa) (Lupi et al. 2010). Bank using available complete mitochondrial genomes This allows more comprehensive alignments of the COI in 2009. Sequences were aligned in Geneious (Biomatters, gene and identification of mismatches in the region tar- New Zealand), and the consensus of nucleotides at the geted by the Folmer primers. These alignments are the positions that correspond to LCO1490 and HCO2198 was basis for the design of new or improved primers for the determined. Positions with fourfold degeneracy were COI region. The molecular interactions involved in replaced with dITP. Positions with twofold degeneracy primer annealing are complex and involve nucleotide were synthesized with mixed nucleotides to create a pri- complementarity to the target, primer homoduplex and mer pool. Resulting primers were named jgLCO1490 and heteroduplex formation, potential for secondary struc- jgHCO2198 to make the relationship to the original tures in incompletely denatured template, ionic strength Folmer primers explicit. of the PCR buffer and nucleotide-neighbour effects in the All morphologically discernable invertebrate species target region. Ideal primer design would include an were collected from settling plates placed in San accurate model for these effects, and several computer Francisco