Low-copy nuclear gene regions offer increased phylogenetic utility for species- and population-level studies of plants as compared to chloroplast and nuclear ribosomal markers (Zimmer and Wen, 2012), yet sampling these regions remains challenging due to the dearth of universal primers and barriers to sequencing whole or partial nuclear genomes from multiple individuals. Consequently, assessing the phylogenetic limits of custom-designed target sequences or primers for low-copy nuclear gene regions is critical to fully realizing their broader impacts for advancing plant systematics. We report the results of a cross-amplification study incorporating primers for 91 low-copy nuclear gene loci created by Gostel et al. (2015) for species-level phylogenetics of Malagasy Commiphora Jacq. (Burseraceae). Primers for these markers were developed using genomic resources from two rosid orders by mapping sequence data from a transcriptome of Bursera simaruba (L.) Sarg. (Burseraceae; Sapindales) (Matasci et al., 2014) to 950 putative low- or single-copy nuclear gene loci of Arabidopsis thaliana (L.) Heynh. (Brassicaceae; Brassicales) (Duarte et al., 2010). Gostel et al. (2015) further optimized the primers for microfluidic PCR-based target enrichment, a method that allows simultaneous and cost-effective amplification of multiple loci (Blow, 2009; Uribe-Convers et al., 2016).
We tested cross-amplification of these primers using 10 taxa that have varying phylogenetic distances from B. simaruba within Sapindales and included A. thaliana as the outermost limit of the survey. Sapindales is a widespread group that includes ca. 6700 species within nine families (Angiosperm Phylogeny Group, 2016) (Fig. 1). Molecular phylogenies of this order often lack sufficient phylogenetic support along their backbone as well as at the species level (e.g., Fine et al., 2014; Grudinski et al., 2014), thus our understanding of Sapindalean systematics could benefit from an expanded phylogenetic toolkit such as that provided by the Gostel et al. (2015) primers.
METHODS AND RESULTS
Taxonomic sampling and molecular methods—Appendix 1 contains accession information for the 11 taxa sampled; Fig. 1 displays their phylogenetic relationships. Bursera simaruba (Bursera Jacq. ex L. subgenus Bursera) and C. grandifolia Engl. were included as positive controls; prior work has shown that all or most of the custom-designed primers amplify PCR product in these two species (Gostel et al., 2015). For experimental taxa, we included B. tonkinensis Guillaumin, which is sister to Commiphora (Weeks and Simpson, 2007), as well as Aucoumea Pierre, the monotypic genus sister to Bursera and Commiphora (Weeks et al., 2014). One species from each of Boswellia Roxb. ex Colebr., Canarium L., and Protium Burm. f. were included, as well as Beiselia Forman, the monotypic genus sister to all other Burseraceae (Weeks et al., 2014). We included one species of Anacardiaceae, the family that is sister to Burseraceae (Weeks et al., 2014), and one species of Rutaceae, which represents the Sapindalean clade sister to Burseraceae–Anacardiaceae– Kirkiaceae (Muellner-Riehl et al., 2016). Arabidopsis thaliana (Brassicales) was included because its genomic resources were used in primer design and can test the applicability of these primers to other closely related rosid lineages (Wang et al., 2009).
Whole genomic DNA was extracted from taxa using the FastPrep FastDNA Spin Kit (Bio101 Systems, La Jolla, California, USA) or the cetyltrimethylammonium bromide (CTAB) method (Weeks et al., 2005). Primer development for the 91 markers is detailed by Gostel et al. (2015); primer sequences are listed in Table 1. Markers were amplified via PCR in 15-µL reactions including: 0.15 µL of forward and reverse primers (50 µM), 0.75 µL spermidine (4 mM), 7.5 µL GoTaq Green Master Mix (Promega Corporation, Madison, Wisconsin, USA), 5.6 µL nuclease-free water, and 1 µL genomic DNA (0.1–25.8 ng/µL). Markers that failed to amplify for B. simaruba and C. grandifolia were then trialed using reaction chemistry based on that recommended for microfluidic PCR-based target enrichment including: 0.15 µL of forward and reverse primers (50 µM); FastStart High Fidelity PCR System reagents (Roche Diagnostics, Mannheim, Germany), composed of 1.5 µL FastStart High Fidelity Reaction Buffer without MgCl2 (10× concentration), 2.7 µL MgCl2 (25 mM), 0.75 µL DMSO, 1.2 µL Nucleotide Mix (10 mM), 0.15 µL FastStart High Fidelity Enzyme Blend (5 U/µL); 0.75 µL Loading Reagent (Fluidigm Corporation, San Francisco, California, USA); 6.8 µL nuclease-free water; and 1 µL genomic DNA.
The PCR thermocycler protocol followed that of Gostel et al. (2015) and included three alternating standard and Cot cycles (Mathieu-Daude et al., 1996), beginning with 2 min at 50°C, 20 min at 70°C, and 10 min at 95°C. The first set of 10 standard cycles included a denaturation step at 95°C for 15 s, annealing at 60°C for 30 s, and extension at 72°C for 1 min. Two Cot cycles followed, including four steps consisting of 95°C for 15 s, 80°C for 30 s, 60°C for 30 s, and 72°C for 1 min. Standard and Cot cycles alternated two more times with eight, two, eight, and five cycles, respectively. After 35 cycles, samples were held at 4°C prior to being visually verified via agarose gel electrophoresis (1% agarose; 94 V for 40 min). Low DNA mass ladder (Invitrogen, Carlsbad, California, USA) was included in the first and last wells of each gel to guide length estimation of PCR products.
Marker amplification results—Table 1 contains amplification results for the low-copy nuclear loci, including the range of amplicon lengths for all taxa and GenBank numbers for markers sequenced by Gostel et al. (2015) for B. simaruba and C. grandifolia that had ≥15 sequence reads mapped. Table 2 summarizes marker amplification success for each taxon. Ninety primer pairs amplified product in B. simaruba and, on average, 54 primer pairs worked for other Burseraceae taxa. The low number of markers amplified in Aucoumea (16) was unexpected given its close relationship to Bursera. This result may have been caused by primer mismatch due to increased genetic change within this monotypic genus, as evidenced by its long branch within Burseraceae phylogeny (Weeks et al., 2014). In total, nine primer pairs worked for every Burseraceae taxon tested, and if Aucoumea is excluded as an outlier, the panel of family-universal primer pairs increases to 26. Thirty-four and 26 primer pairs generated product in Anacardiaceae and Rutaceae, respectively, while only two primer pairs worked in Arabidopsis. Comparing the Burseraceae panel to that of Anacardiaceae and Rutaceae reveals 16 and 12 successfully amplified regions in common, respectively, with eight shared among the three families. PCR chemistry may have suppressed amplification of markers, as high-fidelity PCR reagents were not used due to their high cost. Among the positive controls, high fidelity as compared to standard PCR reagents increased amplification success by 8% (Bursera, 83 to 90 primer pairs) and 85% (Commiphora, 39 to 72 primer pairs). Thus, our experimental results report a conservative baseline for the cross-amplification success of these primer pairs.
Our study demonstrates that 90 of 91 primer pairs for novel low-copy nuclear loci developed by Gostel et al. (2015) for B. simaruba successfully amplify product in a broad range of Sapindalean taxa and effectively expand the phylogenomic toolkit for this order. Twenty-six markers amplify all Burseraceae taxa (excluding Aucoumea) and eight amplify all Sapindalean groups tested. Our results present a new source for universal targets or primers for phylogenetic reconstruction of taxa within Sapindales. Future efforts will include sequencing amplicons to determine the number of phylogenetically informative characters for each locus.
Primer pair sequences and validation results by taxon.
Number of primer pairs amplified of the 91 primer pairs tested for each of the 11 taxa.
The authors thank Cíntia Silva-Luz for providing leaf material of Schinus and Beiselia. Research was supported in part by the National Science Foundation (grant no. 1403150 to M.R.G. and A.W.) and the Provost/COS/ ESP Institutional Graduate Fellowship Research Award to E.S.C. from George Mason University. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation. Publication of this article was funded in part by the George Mason University Libraries Open Access Publishing Fund.