Multiple multilocus DNA barcodes from the plastid genome discriminate plant species equally well.
Fazekas AJ, Burgess KS, Kesanakurti PR, Graham SW, Newmaster SG, Husband BC, Percy DM, Hajibabaei M, Barrett SC. · 2008
Read Multiple multilocus DNA barcodes from the plastid genome discriminate plant species equally well. by Fazekas AJ, Burgess KS, Kesanakurti PR, Graham SW, Newmaster SG, Husband BC, Percy DM, Hajibabaei M, Barrett SC. (2008) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc2475660.txt.
Introduction
Fazekas and co-authors tested candidate DNA barcodes for land plants in 2008. Animal barcoding works well with one standard gene, but plants lacked an agreed region. The team compared eight regions from the plastid genome, the small DNA set inside plant chloroplasts, plus one from the mitochondria, across 251 samples.
The samples covered 92 species in 32 diverse plant genera, including several taxonomically tricky groups. Single regions resolved between 7% and 59% of species as clean separate groups, with a spacer region called trnH-psbA the best single marker. Getting the sequences at all varied: one gene, matK, needed ten primer pairs to reach 88% recovery. Combining four to seven regions lifted resolution only to 69% to 71%, barely above good two-region mixes. The authors said practical ease of recovery deserves as much weight as raw variability.
Catalog record
| CREATOR | Fazekas AJ, Burgess KS, Kesanakurti PR, Graham SW, Newmaster SG, Husband BC, Percy DM, Hajibabaei M, Barrett SC. |
|---|---|
| YEAR | 2008 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 63 KB |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Multiple Multilocus DNA Barcodes from the Plastid Genome Discriminate Plant Species Equally Well By Aron J. Fazekas, Kevin S. Burgess, Prasad R. Kesanakurti, Sean W. Graham, Steven G. Newmaster, Brian C. Husband, Diana M. Percy, Mehrdad Hajibabaei, Spencer C. H. Barrett, Robert DeSalle Published in: PLoS ONE DOI: 10.1371/journal.pone.0002802 ABSTRACT A universal barcode system for land plants would be a valuable resource, with potential utility in fields as diverse as ecology, floristics, law enforcement and industry. However, the application of plant barcoding has been constrained by a lack of consensus regarding the most variable and technically practical DNA region(s). We compared eight candidate plant barcoding regions from the plastome and one from the mitochondrial genome for how well they discriminated the monophyly of 92 species in 32 diverse genera of land plants (N = 251 samples). The plastid markers comprise portions of five coding (rpoB, rpoC1, rbcL, matK and 23S rDNA) and three non-coding (trnH-psbA, atpF–atpH, and psbK–psbI) loci. Our survey included several taxonomically complex groups, and in all cases we examined multiple populations and species. The regions differed in their ability to discriminate species, and in ease of retrieval, in terms of amplification and sequencing success. Single locus resolution ranged from 7% (23S rDNA) to 59% (trnH-psbA) of species with well-supported monophyly. Sequence recovery rates were related primarily to amplification success (85–100% for plastid loci), with matK requiring the greatest effort to achieve reasonable recovery (88% using 10 primer pairs). Several loci (matK, psbK–psbI, trnH-psbA) were problematic for generating fully bidirectional sequences. Setting aside technical issues related to amplification and sequencing, combining the more variable plastid markers provided clear benefits for resolving species, although with diminishing returns, as all combinations assessed using four to seven regions had only marginally different success rates (69–71%; values that were approached by several two- and three-region combinations). This performance plateau may indicate fundamental upper limits on the precision of species discrimination that is possible with DNA barcoding systems that include moderate numbers of plastid markers. Resolution to the contentious debate on plant barcoding should therefore involve increased attention to practical issues related to the ease of sequence recovery, global alignability, and marker redundancy in multilocus plant DNA barcoding systems. INTRODUCTION The development of DNA barcoding markers for discriminating among the >300,000 species of land plants is an unresolved problem, in contrast to some other groups of organisms where effective barcoding markers are now available. A portion of the mitochondrial cytochrome c oxidase 1 (COI or cox1) gene sequence is currently being used as a universal barcode in certain animal groups [1], [2], but mitochondrial genes are generally thought to lack promise for plants [3]–[6], primarily because of their low nucleotide substitution rates (e.g., [7]). It is generally agreed that a multilocus approach based on plastid (‘chloroplast’) data is currently the most effective strategy for species identification and species recognition in plants [3]–[6], [8]–[10] (but see [11]). However, both the identity and number of the most appropriate regions for plant barcoding remain contentious (e.g., the competing proposals reviewed in [12]). Plant researchers have proposed several different barcode regions. These focus on coding and non-coding regions located primarily in the plastid genome. Kress et al. [4] suggested that two non-coding regions (the nuclear ITS region and the plastid trnH-psbA intergenic spacer) may have potential as universal plant barcodes, but subsequently proposed the combination rbcL and trnH-psbA [8]. Other combinations involving three plastid regions have also been proposed by a working group that includes the Royal Botanical Gardens, Kew, UK (http://www.rbgkew.org.uk/barcoding). These comprise the trnH-psbA region and portions of two coding regions (matK and rpoC1), or three coding regions combined (matK, rpoB and rpoC1) [9]. Other regions, such as a portion of the plastid 23S rDNA locus (referred to as the “universal plastid amplicon” or UPA [13]), and the plastid trnL–trnF intergenic spacer [14] have been proposed. Additionally, it has been suggested that the mitochondrial cox1 locus not be discounted as a plant barcode marker without sufficient evidence (P.D.N. Hebert, University of Guelph, pers. comm.). Based on a study focused primarily on Orchidaceae, Lahaye et al. [11], recently proposed that matK be adopted “as a universal DNA barcode for flowering plants” (either alone or in combination with trnH-psbA), leaving its suitability for other plants largely unresolved. Finally, at a recent conference (Second International Barcode of Life Conference), the Plant Working Group of the Consortium for the Barcode of Life (PWG-CBOL) proposed additional combinations of non-coding and coding plastid regions (i.e., matK+atpF–atpH+trnH-psbA; matK+atpF–atpH+psbK–psbI; see [12]). Because of the size of the task in establishing a reference barcoding database for all plants, it is widely accepted that a single (“universal”) set of barcode regions should be adopted. To date there has been no comprehensive empirical test of the utility of all gene regions and multilocus combinations currently under consideration. The assessment of different solutions is complicated by the multiplicity of ideas on what the most relevant criteria are for choosing among different single and multilocus barcoding strategies, and by lack of clarity about the weightings that ought to be placed on their relative importance. The amount of variation obtained per sequencing read is obviously a key parameter (the more the better, within limits). However, other technical considerations are likely to be at least as important, such as the ease of sequence recovery, the frequency of DNA sequencing artifacts, and the question of whether DNA sequence alignment is useful for identifying species using barcodes. The Barcode of Life Data System (BOLD, www.boldsystems.org [15]) currently relies on aligned sequences, but other alignment-independent algorithms have been proposed or are in development (e.g., [16], [17]). Here, we compare the ability of nine different gene regions to discriminate among species within clades of closely related organisms — portions of five coding plastid regions (rbcL, matK, rpoC1, rpoB and 23S rDNA), three non-coding plastid intergenic spacer regions (trnH-psbA, atpF–atpH and psbK–psbI), and the mitochondrial cox1 gene (the animal barcode region). We tested these regions individually and in various combinations (including the most commonly proposed ones) for a diverse set of genera from major clades of land plants from temperate N. America. The surveyed genera include bryophytes, a lycophyte, and several monilophytes and gymnosperms, in addition to numerous angiosperms. Each species is represented by accessions from multiple populations. We apply a distinct approach to evaluating species resolution here, by assessing whether clusters of populations that correspond to monophyletic species in a floristic sample are well supported (see [11] for a somewhat similar approach). A high degree of correlation should exist between support for species monophyly and the ability of DNA barcoding marker systems to discriminate species, since the robustness of the branch that subtends a single species (representing a monophyletic cluster of populations) determines whether that species is distinct from closely related species, at least within the limits of sampling error. Our survey necessarily focuses on gene-tree monophyly, because this is what is quantifiable using markers belonging to a single linkage group (e.g., the plastid genome). However, in a few cases we find evidence of gene-tree paraphyly that also raises questions about the monophyly of some sampled species. It is worth emphasizing that species assessed to have well-supported monophyly in a local context may not