Microsatellite markers for Juglans cinerea L. and their utility in other Juglandaceae species

Abstract

Ten polymorphic microsatellite markers were found to amplify in butternut (Juglans cinerea; Juglandaceae). These microsatellite loci were found to amplify across most of nine other species and five hybrids examined. Loci were highly polymorphic, with 18 to 32 alleles per locus across species. These nuclear microsatellite markers will be useful in examining genetic diversity within and among populations of butternut, and in distinguishing butternut from interspecific hybrids.

Introduction

Butternut, Juglans cinerea L., is a short-lived tree species being extirpated throughout most of its native Eastern North American range by butternut canker (Schultz 2003; Nielsen et al. 2003), a disease caused by the exotic fungus Sirococcus clavigignenti-juglandacearum (Renlund 1971). The species was listed as endangered by the Committee on the Status of Endangered Wildlife in Canada in 2003, and in the USA it is currently considered a Regional Forester Sensitive Species.

Previous research based on allozyme loci reported low levels of genetic diversity across butternut populations in the northern portion of the species range (Morin et al. 2000). Here, we identify a suite of ten nuclear microsatellite markers that amplify in J. cinerea and examine the utility of these markers in other Juglandaceae species. Four of these markers have previously been published for J. nigra (Woeste et al. 2002; Victory et al. 2006) and J. regia (Dangl et al. 2005).

Methods

Primers for this research were derived by further sequencing of a black walnut (Juglans nigra) microsatellite library described by Woeste et al. (2002). Leaf samples of the following species and hybrids were obtained: Carya illinoensis and Juglans microcarpa from L.J. Grauke, USDA ARS National Germplasm Repository for Pecans and Hickories, Somerville, TX; Juglans ailanthifolia, J. cathayensis, J. hindsii, J. mandschurica, J. regia, and J. x paradox from Malli Aradhya, USDA ARS National Clonal Germplasm Repository, Davis, CA; J. nigra from Mark Coggeshall, University of Missouri Agroforestry Research Center; and J. x quadrangulata. J. x royal, and J. major from the germplasm collection of the USDA Forest Service Hardwood Tree Improvement and Regeneration Center, West Lafayette, IN. Samples of J. cinerea were collected by the authors with the assistance of numerous collaborators.

DNA Extraction, quantification, amplification and genotyping

Genomic DNA was extracted from leaves as described by Victory et al. (2006). All samples were quantified using a ND-1000 Spectrophotometer (NanoDrop Technologies, Inc.). PCR conditions and genotyping was performed as described in Victory et al. (2006) except products were separated using an ABI 3730 sequencer. Two positive and one negative control were run with each PCR to ensure accurate scoring. Failed reactions were repeated for accuracy.

Analysis

GDA (Lewis and Zaykin 2001) software was used to determine mean sample size over all loci (n), mean number of alleles per locus (A), expected (H _e ) and observed (H _o ) heterozygosity, linkage disequilibrium, and fixation index (f). Linkage disequilibrium was only tested for J. cinerea, the only species for which there was a large sample size (n = 422). SAS (v. 9.1; Cary, NC) Proc Princomp and Proc Candisc were used to perform principle components analysis (PCA) and canonical discriminant analysis (CDA), respectively. For both SAS procedures, missing data resulting from non-amplification or from unscorable, multiple peaks were replaced with the grand mean for each locus. Each allele/species combination was considered an independent unit of analysis (i.e., each bi-allelic genotype at each locus generated two data points). To make the dataset more balanced, eight J. cinerea genotypes were chosen arbitrarily and included in the PCA and CDA.

Results and discussion

Loci amplified across the majority of species examined (Table 1, 2) and were highly polymorphic, showing 18–32 alleles per locus across species; however, polymorphism within taxa was generally low, ranging from 1.3 for J. x quadrangulata to 13.0 for J. cinerea. This was likely due to limited sample sizes. Linkage disequilibrium was detected in 60% of the pairwise comparisons between loci. Four of the ten loci (WGA 004, WGA 204, WGA 221, and WGA 256) amplified across all species. The loci did not show a high degree of ascertainment bias (Table 2), as the allele numbers for J. ailanthifolia and J. major were as high or higher as those for J. nigra, the species from which the primers were originally derived. Transfer of WGA 004, 204, 221 and 256 to C. illinoensis is unusual and may point to either strong conservation of these loci or a relatively recent divergence of the genera (Hale et al. 2005).

Table 1 Primer sequences and label information for microsatellite loci that amplify in Juglans cinerea

Table 2 Allele size range (size) and number observed (N_a) for each locus and overall allelic richness (A) for each species

The usefulness of these microsatellite loci across species is reflected in the number of alleles per locus (Table 2). Both J. regia and J. mandschurica showed elevated fixation levels compared to other groups (Table 3). In the case of J. regia, this may reflect a domestication bottleneck. The hybrids J. x intermedia and J. x paradox are complex; they are intersectional hybrids that may contain the genomes of more than two species, and they may be the result of backcrosses or intercrosses (Potter et al. 2002). As a consequence, these taxa may contain genomic incompatibilities that limit recombination and affect inbreeding. Analysis using principle components and plotting of the first two principle component scores, which explained 41% of the variance, showed a clear clustering of the section Cardiocaryon alleles and a discrete position for the alleles of J. major and J. regia (Fig. 1). J. cinerea alleles were located much nearer to J. nigra and other members of section Rhysocaryon than to the Asian members of section Cardiocaryon, with which butternut is sometimes lumped (Fig. 1). The taxa could all be significantly separated from one another using canonical discriminant analysis, with the exception of J. mandschurica and J. ailanthifolia (data not shown).

Table 3 Mean sample size over all loci (n), mean number of alleles per locus (A), expected heterozygosity (H_E), observed heterozygosity (H_O), and an estimate of the fixation index (f). Species are arranged as in Table 2

Fig. 1

Scatterplot of individual species/allele units on first two axes of PCA, which account for 41% of the total variance

The interspecific butternut hybrid J. x bixbyi is vigorous, difficult to distinguish from butternut, produces large numbers of fruit and may be more resistant to butternut canker. Cross-species amplification of these loci may prove useful in distinguishing butternut from hybrids. For example, there was a 13 base pair difference in allele size ranges for WGA 090 between J. ailanthifolia and J. cinerea (the two parent species of J. x bixbyi). Indeed, all J. x bixbyi individuals examined in this study contained an allele of the J. cinerea size and an allele of the J. ailanthifolia size, as expected. However, before a distinct range of allele sizes can be confirmed for a species, additional individuals must be genotyped. Results of PCA and CDA presented here should be considered heuristic and not an indication of phylogeny since allelic states were considered random deviations from a grand mean and not according to any biological model such as stepwise mutation, and the PCR products were not sequenced to verify the relationships among the length variants. The role of homoplasy in determining identity of state is likely in some cases.