Yam beans (Pachyrhizus Rich. ex DC., Fabaceae) are little-studied plants with edible tuberous roots native to South and Central America. The genus comprises five species, two wild (P. panamensis R. T. Clausen and P. ferrugineus (Piper) M. Sørensen) and three cultivated (P. ahipa (Wedd.) Parodi, P. erosus (L.) Urb., and P. tuberosus (Lam.) Spreng.). Yam beans are grown for their starchy root but are propagated exclusively through seeds. To stimulate root growth, farmers prune flower buds but leave either one pod on each plant or select a few plants dedicated to seed production. To set conservation strategies, it is necessary to understand how these different methods influence the crop's dynamics of genetic diversity, but this requires molecular tools that yield information on important parameters such as heterozygosity and allelic frequencies needed for the computation of most population genetic statistics. There are to date no available genetic markers for Pachyrhizus species. Socially and culturally important but economically marginalized, yam beans are “orphans” to crop science, and few resources have been invested in evaluating the current status of genetic diversity in these minor yet promising crops. The lack of molecular tools has probably stymied efforts to document these largely untapped genetic resources.
In this paper, we report the isolation and characterization of 17 polymorphic simple sequence repeat nuclear markers for P. ahipa and their successful cross-amplification in other Pachyrhizus species. Phylogenetic relationships among Pachyrhizus species remain largely unresolved. This new set of molecular markers will permit investigation of the phylogeography of the Pachyrhizus complex.
METHODS AND RESULTS
Total genomic DNA was extracted from herbarium specimens from 20 mg of lyophilized leaf tissue using NucleoSpin 96 Plant kits (Macherey-Nagel, Hoerdt, France) following the manufacturer's instructions. Purified DNA was eluted in a final volume of 200 µL, and final concentration was checked using a Nanodrop ND-1000 spectrophotometer (Labtech, Palaiseau, France). A sample of 3 µg total DNA at 60 ng/µL final concentration, representing a pool of 12 P. ahipa accessions spanning the whole distribution range of the species in Bolivia, was sent to Genoscreen (Lille, France) for production of enriched DNA libraries and 454 GS-FLX Titanium (Roche Applied Science, Meylan, France) pyrosequencing (Malausa et al., 2011). A total of 3454 sequences containing potential microsatellite motifs were produced. Following sequence cleaning and removal of duplicates, 252 primer pairs (only perfect repeats with at least five repeats) were designed using the QDD bioinformatics pipeline (Meglécz et al., 2010).
We selected a set of markers that would cover a wide range of amplification product sizes and could be used in multiplex reactions (i.e., that minimized differences in annealing temperatures and complementarity among primer pairs), targeting in priority loci with the longest di- and trinucleotide repeats (six repeats or more). A cost-efficient approach to selecting markers is to prescreen microsatellites for polymorphism using in silico DNA sequences (Hoffman and Nichols, 2011), but very little sequence information is available for the understudied genus Pachyrhizus. Blasting primer sequences against sequences available at GenBank for the closest Fabaceae species, we obtained the best results with the model crop Glycine max (L.) Merr. (subtribe Glycininae), with a mean query coverage (±SE) of 88% (±23) and 93% (±8) identity between G. max and P. ahipa homologous sequences. Targeting conserved flanking regions among distantly related species can also be a potent way to enhance cross-species utility of microsatellite markers (Dawson et al., 2010). Using microsatellite variability in G. max as a proxy to infer variability among putative microsatellites in Pachyrhizus spp., we targeted loci most likely to be polymorphic. Thirty-six primer pairs were tested in separate PCRs. Nine pairs failed to produce clear amplicons. A second test was carried out on the 27 primer pairs that amplified using a sample of 144 accessions (wild and cultivated) from herbarium specimens representing varietal, morphological, and potential genetic variation across the natural distribution area of the genus (Appendix 1). Multiplex PCR were carried out on an Eppendorf Mastercycler ep gradient thermocycler (Eppendorf, Hamburg, Germany) using phosphoramidite-labeled oligonucleotides (Applied Biosystems, Warrington, United Kingdom) in a final volume of 12.5 µL. Along with 1 µL of nondiluted DNA template, each well contained 6.25 µL of QIAGEN Type-it Master Mix (QIAGEN, Hilden, Germany), 1.25 µL of 10× primer mix (with primers at 2 µM), and 4 µL of RNase-free water. An initial activation step at 95°C for 30 s preceded 20 cycles of amplification, each starting with an annealing step of 90 s at 56°C and continuing with an extension at 72°C for 30 s. Amplification ended with a final extension at 60°C for 30 min. To ensure unambiguous peak assignment, primer pairs were pooled in two different sets (M1 and M2) as indicated in Table 1. Multiplex Manager 1.2 software (Holleley and Geerts, 2009) was used to optimize primer combinations.
Genotyping was performed on an ABI PRISM 3130 Genetic Analyzer (Perkin Elmer/Applied Biosystems, Foster City, California, USA). Each sample was prepared from 1 µL of PCR template to which 8.8 µL formamide and 0.2 µL GeneScan 500 LIZ Size Standard (Applied Biosystems) were added. Genotypes were extracted and analyzed using GeneMapper 4.0 software (Applied Biosystems). To reduce the risk of typing errors, allele peaks were checked by eye. Cross-species amplification tests succeeded for all loci across the genus. Six loci were strictly monomorphic across all species and were discarded. At the species level, 15 out of the 17 remaining loci were monomorphic in P. ahipa, six in the cultivated P. tuberosus, and four in the cultivated P. erosus (Table 2). Only two and three loci were monomorphic in the wild P. tuberosus and wild P. erosus, respectively. Number of alleles, observed and expected heterozygosities, and tests for deviation from Hardy–Weinberg equilibrium (HWE) were estimated using GenAlEx version 6.41 (Peakall and Smouse, 2006). Results for each locus and species are summarized in Table 2. The number of alleles ranged from three to 12, with a mean value of (±SE) 6.4 ± 3.0 alleles across loci and species. Expected heterozygosity ranged from 0.095 (AIP9) to 0.831 (AIP30). All loci showed significant deviation from HWE in the three cultivated species (P < 0.001). Linkage disequilibrium was checked using GENEPOP 4.1.4 (Rousset, 2008). Two pairs of loci showed significant linkage disequilibrium in the cultivated P. erosus after Bonferroni correction for multiple comparisons (P < 0.0004). Yam beans are predominantly self-pollinating species with outcrossing rates typically ranging between 2% and 4% (Sørensen, 1996), and physical linkage of loci cannot be distinguished from disequilibrium due to nonrandom mating.
Characteristics of the 17 microsatellite loci developed for Pachyrhizus spp.
Conservation of crop genetic resources hinges on the availability of efficient molecular tools to characterize population genetic structure and decipher the dynamics of crop genetic diversity. The case of Pachyrhizus illustrates the spillover benefits to be reaped from next-generation sequencing and research on model plants for the study of minor crops (Varshney et al., 2010). The markers we developed showed high levels of polymorphism and enough discriminant power for distinguishing among varietal groups within species. They will be available for a wide range of applications, from breeding to population genetic studies. Markers also revealed a surprisingly low level of genetic variability in the Bolivian root crop, P. ahipa. While the wild parent of the crop has yet to be identified, we will use the new markers to investigate the origin of P. ahipa. Results should shed new light on the evolutionary history of the Pachyrhizus genus.
Results of initial primer screening in Pachyrhizus ahipa, and P. tuberosus (wild and cultivated) for the 17 polymorphic loci. Cross-amplification tests were also carried in two wild species, P. ferrugineus and P. panamensis.