Index | Search | Home | Table of Contents
Kapteyn, J. and J.E. Simon. 2002. The use of RAPDs for assessment of identity, diversity, and quality of Echinacea. p. 509513. In: J. Janick and A. Whipkey (eds.), Trends in new crops and new uses. ASHS Press, Alexandria, VA.
J. Kapteyn and J.E. Simon
The genus Echinacea, Asteraceae, is comprised of nine species, which are perennial herbs indigenous to North America and which have been traditionally used as medicinal plants for centuries. Three Echinacea species, E. angustifolia DC var. angustifolia, E. purpurea (L.) Moench, and E. pallida (Nutt.) Nutt., are currently being traded internationally in the natural products market. Echinacea products constitute a significant portion of this growing, multi-billion dollar industry. Echinacea is considered of value as a nonspecific immune stimulant; the pharmacological activity of Echinacea extracts has been widely studied with over 350 published studies to date (Briskin 2000) together providing strong evidence for an immune modulating activity of these extracts.
The increasing popularity of Echinacea products has led to expansion in wildcrafting and commercial cultivation to meet the growing demand for plant material. Commercial supplies of E. purpurea are obtained from cultivated sources; E. angustifolia and E. pallida have until recently been supplied largely from indigenous habitats in the United States (Foster 1993). The threat to the genetic diversity present in wild populations due to indiscriminant overharvesting and the need for preservation of these genetic resources creates an incentive for the determination of the genetic variability present within these three species. The quality of botanical materials used in the manufacture of natural products has also been of concern with Echinacea products. Parthenium integrifolium is a common prairie perennial with a root morphology similar to that of E. angustifolia; roots of this plant have been a frequent adulterant of dried roots of E. angustifolia.
DNA fingerprinting techniques such as random amplified polymorphic DNA (RAPD) (Williams et al. 1990) permit the identification of taxa and the determination of phylogenetic relationships and intraspecific diversity at a molecular genetic level. The use of such techniques for germplasm characterization facilitates the conservation and utilization of plant genetic resources, permitting the identification of unique accessions or sources of genetically diverse germplasm. The ability of this method to distinguish between taxa also has useful implications in botanical quality analysis.
This study used RAPD markers to determine the genetic relationships of the three Echinacea species of commercial interest, to evaluate the level of diversity present within germplasm of each of the three species, and to compare accessions of each species available from different sources including the USDA National Plant Germplasm System and commercial sources.
RAPD markers were also identified that are capable of distinguishing the presence of Parthenium integrifolium L., an adulterant of E. angustifolia, in DNA samples extracted from combined tissue of the two species. The 17 species-specific markers generated for the three Echinacea species in this study may also be useful in the identification of Echinacea species included in samples of botanical material or finished products.
A total of 19 accessions of Echinacea were used in this study, including one outgroup E. atrorubens Nutt. accession. A complete list of the accessions and their sources is available (Kapteyn et al. 2002). Seed from each accession was germinated after stratification with (2-chloroethyl) phosphonic acid (Sari et al. 1999), and plants of each accession were grown in a greenhouse at Purdue University, West Lafayette, Indiana. DNA was isolated separately from four plants of each accession for RAPD analysis. P. integrifolium plants were obtained from Prairie Nursery, Inc. and grown in a greenhouse at Rutgers University, New Brunswick, New Jersey.
Total DNA was isolated from 1 g young leaf tissue using a modified method of Doyle and Doyle (1987). RAPD reactions were performed in 25 ml volumes containing 25 ng of template DNA, PCR buffer (50 mM KCl; 10 mM Tris-HCl pH 8.8; 0.1% Triton X-100), 3.0 mM MgCl2, 0.25 mM each dNTP, 0.2 mM primer, and 1.25 units Taq DNA polymerase. Amplification reactions were performed in a Perkin-Elmer 9600 Thermal Cycler programed for 40 cycles of 94°C for 30 s, 45°C for 60 s, and 72°C for 60 s, with a 72°C hold for 10 min after the completion of 40 cycles.
PCR products were separated on 1.6% agarose/0.5X TBE gels. Gels were visualized and photographed after staining with ethidium bromide.
RAPD products were scored for presence or absence of each amplicon evaluated. Only those bands that could be unequivocally scored across all samples were included in the analysis. Pairwise similarity matrices were generated using Jaccards coefficient of similarity (Jaccard 1908). Principal coordinate analysis (Gower 1966) was also performed to display the relationships in three dimensions. All procedures were performed using NTSYS-pc (Rohlf 1998).
To determine whether accessions of the same species differed from one another, Arlequin (Schneider et al. 2000) was used to analyze the population genetic structure of the RAPD data within each of the three species. Prior to AMOVA analysis using Arlequin, the haplotypic data for each species was pruned, removing those bands whose observed frequency was greater than or equal to 1-(3/N) to ensure that unbiased estimates of population-genetic parameters could be achieved (Lynch and Milligan 1994). Subsequent AMOVA analysis proceeded with 67 markers for E. purpurea, 72 for E. angustifolia, and 70 for E. pallida. For this analysis, it was assumed that each accession represented a separate population in Hardy-Weinberg equilibrium, with the four individuals of each accession representing the population sample.
A matrix of Euclidian square distances was computed using the pairwise difference method. This matrix was used for the analysis of genetic structure including partitioning of variation among and within populations and the calculation of pairwise population Fst values, which were subsequently tested for significance.
Several primers were screened in amplifications of DNA samples of E. angustifolia, E. pallida, E. purpurea, and P. integrifolium to identify molecular markers specific to P. integrifolium. A 420 bp RAPD marker was generated using primer O-13 that was specific to and found in all P. integrifolium samples screened and absent from all Echinacea individuals. As E. angustifolia is most often subject to adulteration by P. integrifolium, quality analysis using the P. integrifolium-specific RAPD marker proceeded with these two species. Coarsely processed bulked tissue of each species was combined in ratios of 1:1, 1:3, 9:1, and 99:1 E. angustifolia:P. integrifolium prior to grinding and DNA extraction. DNA samples were subsequently prepared for RAPD analysis and DNA from the combined tissue samples and from individuals of both species were amplified using O-13 to determine the potential utility of this RAPD marker in discerning the presence of P. integrifolium in a mixed botanical sample. RAPD products from the quality analysis were separated on 5.0% acrylamide/1X TBE gels, which were handled as previously described.
Initially, 57 primers were tested for their ability to generate amplification products. Of the 43 primers that produced amplification products, 22 were chosen for their ability to generate unambiguously scoreable RAPD bands. A total of 101 bands were scored, with an average of 5 bands scored per primer. The number of scoreable bands generated by a single primer ranged from as few as 1 to as many as 8. Product sizes ranged from 400 to 2100 bp. Many other visibly polymorphic fragments were generated, however they were not considered in the analysis due to their weak or non-reproducible amplification, or due to an inability to resolve closely migrating fragments. Portions of gels showing typical amplification products are shown in Fig. 1.
Fig. 1. Portions of two typical agarose gels showing RAPD profiles generated with primers OPC-2 (A) and OPH-13 (B). Lane M=pGEM DNA marker, lanes 14=E. angustifolia, lanes 58=E. pallida, lanes 912= E. purpurea.
This study identified 17 diagnostic markers suitable for discrimination of the three commercially relevant species plus E. atrorubens. A marker was considered diagnostic if it was present at a frequency of 0.95 for all individuals of a given taxon and present at a frequency of 0.05 for all individuals of each taxon being discriminated against. The reciprocal case was also considered diagnostic.
Pairwise genetic similarities (data not shown) generated using Jaccards coefficient of similarity (Jaccard 1908) ranged from as low as 0.185 between pur4-2 and ang16-2 (E. purpurea and E. angustifolia, respectively), to as high as 0.978 between pur1-1 and pur4-3 (both E. purpurea). Genetic similarity was highest among E. atrorubens at 0.873, followed by E. pallida at 0.816, E. purpurea at 0.802 and E. angustifolia at 0.790. Only the mean similarity of E. atrorubens was significantly different from the other taxa (P = 0.05).
Cluster analysis of the genetic similarity values was performed to generate a dendrogram illustrating the overall genetic relationships between the species studied and the accessions and individuals within those species. The dendrogram was constructed using UPGMA clustering (Fig. 2). Four distinct clusters comprised of each of the four Echinacea species were formed. The E. angustifolia and E. pallida clusters were most closely linked, joining at a similarity of 0.45. The E. atrorubens cluster joins this cluster at a similarity of 0.37, and E. purpurea joins the cluster comprised of the previous three species at a similarity of 0.29.
Fig. 2. Dendrogram showing the genetic relationships of 76 Echinacea individuals. The individuals are labeled with a species and an accession identifier, plus an individual designator.
Application of the frequency parameters of Lynch and Milligan (1994) for obtaining unbiased estimates of population-genetic parameters to the data sets of each of the three Echinacea species separately resulted in the elimination of 34 markers for E. purpurea, 31 markers for E. pallida, and 29 markers for E. angustifolia. The remaining markers that satisfied the specified criteria were used in the subsequent AMOVA analysis of each species.
AMOVA analysis using the Arlequin program enabled a partitioning of the overall RAPD variation between the within accession and among accession covariance components (Table 1). AMOVA analysis did not reveal any significant differences between the E. purpurea accessions; all of the diversity (98.0%) was attributable to variation within the accessions. Variation was similarly partitioned for E. pallida and E. angustifolia, with most of the variation again being found within the accessions (82.6% and 78.2%, respectively). Partitioning of variation between accessions was significant for both of these species, however.
Table 1. Summary of AMOVA analysis. Statistics include: degrees of freedom (df), sum of squares (SSD), variance-component estimates (CV), and percentages of the total variance (% Total) contributed by each component.
|Analysis||Source of variation||df||SSD||CV||% Total|
|E. purpurea||Among populations||3||11.5||0.07||2.0 NS|
|E. pallida||Among populations||6||32.4||0.62||17.5|
|E. angustifolia||Among populations||6||36.1||0.79||21.8|
NS = not significant (significance tests after 1023 permutations)
Pairwise population comparisons were performed as the calculation of Fst values (data not shown) and a test of significance of these values using a non-parametric permutation approach (Excoffier et al. 1992). Table 2 shows a matrix of the results of the tests of significance of Fst values for E. angustifolia. Eleven of 21 pairwise population comparisons found accessions to be different for E. pallida, and 10 of 21 pairwise comparisons were found to be significant for E. angustifolia.
Table 2. Matrix of significant Fst P values for pairwise E. angustifolia accession comparisons. A + indicates a significant difference between accessions.
Of the several primers screened, two generated markers that were unique to P. integrifolium in amplifications of template DNA from this species, E. angustifolia, E. pallida, and E. purpurea. Subsequent quality analysis proceeded with the P. integrifolium-specific 420 bp marker generated with primer O-13 and adulterated samples of E. angustifolia only. This marker could be detected among amplification products of RAPD reactions containing template DNA extracted from combined bulked tissues of P. integrifolium and E. angustifolia (Fig. 3) at levels down to 10% P. integrifolium.
Fig. 3. Polyacrylamide gel of RAPD profiles generated with primer OPO-13 for P. integrifolium, E. angustifolia, and template DNA from combined samples. Lane M= pGEM DNA marker, Lane 12=P. integrifolium individuals, lane 3=1:1 E. angustifolia:P. integrifolium, lane 4=3:1 E. angustifolia:P. integrifolium, lane 5=9:1 E. angustifolia:P. integrifolium, lane 6=99:1 E. angustifolia:P. integrifolium, lane 78=E. angustifolia individuals.
The RAPD technique has been successfully used in a variety of taxonomic and genetic diversity studies (Jain et al. 1994; Li et al. 1999; McGrath et al. 1999; Nebauer et al. 1999; Rodriguez et al. 1999), and was found by Wolf et al. (1999) and by us to be suitable for use with Echinacea species in its ability to reproducibly generate polymorphic markers. Wolf et al. (1999) has published a report demonstrating the utility of the RAPD technique for the discrimination of E. purpurea, E. angustifolia, and E. pallida using two different primers. Our study reports an additional 17 RAPD markers capable of distinguishing among the commercially relevant Echinacea species and E. atrorubens, and extends the application of those markers to the identification of the genetic relationships between those species and the diversity and structure present within those species (Kapteyn et al. 2002). RAPD data was used to partition genetic variation between the within accession and among accession levels for the three commercial species and to identify E. pallida and E. angustifolia accessions which may serve as potential sources of unique genetic material. The sensitivity of the RAPD method also permitted the detection of P. integrifolium adulteration of E. angustifolia at levels down to 10%. The results presented here demonstrate the utility of using RAPD markers to characterize interspecific relationships, to evaluate germplasm diversity in Echinacea species, to identify potential sources of unique germplasm material, and to identify the presence of specific adulterants in botanical samples.