Introduction

Goldenseal (Hydrastis canadensis L.), belongs to Ranunculaceae family and is native to deciduous woodlands of eastern North America (Foster and Duke 2000). It is an economically important medical herb, which in 1999 was ranked fifth among the top thirteen herbal products sold in the USA with sales exceeding $44 million (Blumenthal 1999). In 2004 and 2005, combined cultivated and wild-harvested goldenseal roots that entered the market as dried material was 42,000 and 41,000 kg respectively, an almost 50% increases from 21,000 kg in 2003 (AHPA 2007). Goldenseal preparations have been used traditionally for the treatment of eye infections, bronchitis, pharyngitis, and other digestive disorders (Foster and Duke 2000). Its medicinal properties are attributed to a group of secondary metabolites called alkaloids—mainly berberine and hydrastine. These alkaloids have antibiotic activities and therefore are effective against many bacterial infections (Seazzocchio et al. 2001). Very little is known about the genetics of secondary metabolism, and currently no model species exist for studying medicinal plants or plants with complex secondary metabolite production (Cole et al. 2007), therefore research in medicinal plants especially in areas of gene regulation and function in secondary metabolism has progressed slowly. It is quite evident that research and biotechnology in agriculture and horticulture fields have benefited tremendously from availability of vast amount of information from fully mapped genomes of plant species such as rice (Oryza sativa) and Arabidopsis (Arabidopsis thaliana). Current plant model species however, have limited applications in research of medicinal plants with complex secondary metabolism. The search for a suitable model species for medicinal plants is ongoing, and it is anticipated that identifying and fully sequencing the genome of such a model species will greatly facilitate research into gene regulation and function, and biotechnology in medicinal plants.

Most of the goldenseal supplied to the market is harvested from the wild (AHPA 2007), and because of increasing demand; wild populations are overharvested and could potentially lead to loss of valuable germplasm. On the other hand, deforestation and habitat fragmentation as a result of land development activities greatly disturbs natural habitats of goldenseal and restricts gene flow among its populations. In 1997 goldenseal was listed as an Appendix II species of the Convention for International Trade and Endangered Species (CITES) of Wild Fauna and Flora (CITES 1997), and is designated as threatened in the US and Canada and therefore granted a high priority for protection (Bannerman 1997). Restoration of goldenseal into the wild is a major concern for plant conservationists, and commercial cultivation is necessary to supply the growing demand and reduce harvest pressure on wild populations. Conventional propagation of goldenseal through seed is slow given high mortality of seedling in early stages, and harvesting of the rhizome for vegetative propagation seems not feasible because the species is endangered and there is always a possibility of losing mother plants during this process (He et al. 2007). Micropropagation has been suggested as a potential technique to rapidly produce goldenseal planting stock for conservation and commercial cultivation, and thus far a number of optimal in vitro propagation protocols have been developed (Hall and Camper 2002; Bedir et al. 2004; Liu et al. 2004; He et al. 2007). However, with in vitro propagation often genetic disturbances do occur and this leads to somaclonal variations (Sliwinska and Thiem 2007). The resulting genetic variability can be consequential; therefore it is imperative to check for genome stability of in vitro produced plants especially if they are to be used as a source of planting stock to restore unique locally threatened wild population germplasm or to produce secondary metabolites with pharmacological activities.

Currently, very little is known about the genetics of secondary metabolism in goldenseal and its genome size is not even known. In this study, we sought to estimate the nuclear DNA content of goldenseal and assess genome size stability of its in vitro regenerated plantlets compared to wild plants using Flow cytometry (FCM). FCM is a fast, relatively inexpensive, and accurate technique for estimating nuclear DNA content, and has been successfully used to assess genome size stability of in vitro regenerated medicinal plants (Sliwinska and Thiem 2007).

Materials and methods

Plant materials

Pea [Pisum sativum cv. Ctirad (2C = 9.09 pg, Doležel et al. 1998)] was used as internal reference standard. Pea seedlings were grown in a greenhouse in 150 × 150 mm round pots (Dillen Products, Middlefield, OH, USA) using regular Sunshine Mix 1 (SUNGRO Horticulture, Seba Beach, Canada), under natural light conditions, temperature of 21 ± 3°C, and plants were hand watered daily. Goldenseal plants used in this study were collected from two wild populations in West Virginia (Monongalia County, WV, USA) and from in vitro propagated plantlets initiated from leaf explants of wild plants (Fig. 1a, b).

Fig. 1
figure 1

Wild growing plants (a) and In vitro plantlets (b) of H. canadensis, and histograms of relative fluorescence intensity (c and d) obtained after analysis of their nuclei isolated simultaneously with that of Pisum sativum cv. Ctirad (2C = 9.06 pg, as an internal reference standard). Inset each histogram; the peak number, mean peak value, and coefficient of variation percentage (CV, %) value of each peak is given

Full size image

In vitro plantlet regeneration

In vitro propagated plants were initiated from wild plants through callus phase using a tissue culture protocol for goldenseal as described by Bedir et al. (2004). The cultures were transferred every 4 weeks to fresh medium and regenerated shoots were used for subsequent sub-culturing for shoot multiplication and were maintained in culture for more than 3 years prior to this study.

Nuclei preparation

Nuclei were obtained from young leaves of goldenseal (wild plants or in vitro plantlets) and pea by chopping them with new scalpel blades (No. 10 blade; PhytoTechnology Laboratories, Shawnee Mission, KS, USA) in AE nuclei isolation buffer (Arumuganathan and Earle 1991). The nuclei isolation buffer was two staged, herein referred as buffer A and B. Buffer A comprised of AE buffer supplemented with 50 μg/ml of RNase A (Product Number P4642; Sigma–Aldrich Chemical Company, St. Louis, MO, USA). RNase was heated to 90°C for 15 min to inactivate DNases. Buffer B comprised of AE buffer supplemented with 50 μg/ml propidium iodide (PI) (Product Number P4170; Sigma–Aldrich Chemical Company, St. Louis, MO, USA). All buffers were freshly prepared and stored on ice and in the dark at all times. Nuclei from both sample and reference standard were either separately or simultaneously isolated following a two step protocol; first, 40 mg of fresh young leaves were chopped on ice in a plastic Petri dish containing 1 ml of buffer A, and the resulting slurry was filtered through a 40 μm nylon filter (Product Number. 9086473; BD Falcon, Franklin Lakes, NJ, USA) and centrifuged (Marathon 16 km, Fisher Scientific) at 14,000 rpm for 20 s. The supernatant was discarded and the pellet re-suspended in 400 μl of buffer B. The stained nuclei were allowed to stand at 25 ± 1°C (room temperature) in the dark for 30 min before they were analyzed in a flow cytometer.

Flow cytometry

Stained nuclei were analyzed using a FACS Calibur flow cytometer equipped with a 15 mW 488 nm air-cooled argon-ion laser and Cell Quest Pro software (BD Bioscience, San Jose, CA, USA). Prior to running samples, linearity of the instrument was assessed using the BD DNA QC Particles Kit (Cat. No. 349523; BD Bioscience, San Jose, CA, USA). Doublets, disintegrated nuclei, and other cell debris were eliminated from analysis by gating on the forward and side scatter profiles of the samples. The gate was uniformly maintained across all samples in each run. For each sample 7,000–10,000 events (nuclei) were collected and the resulting histograms were analyzed using FCS Express Version 3 De NOVO software (FCS Research, Los Angeles, CA, USA).

Test for inhibitors

Some plants contain compounds in their cytosol that inhibit PI intercalation/fluorescence therefore compromising the reliability of estimating nuclear DNA contents particularly if detection of small differences in DNA content is desired (Price et al. 2000). One of the key recommendations of the 2003 Plant Genome Size Workshop held at the Royal Botanical Gardens, Kew (London, UK) was that all plant species whose nuclear DNA content is determined by FCM should be tested for the presence PI inhibitors. Presence of inhibitor is confirmed if mean peak position of reference standard is lower in the presence of target sample (Price et al. 2000). In this study we first tested for the presence of such inhibitors in the cytosol of goldenseal, by comparing the peak positions of pea nuclei that were separately processed and those that were simultaneously processed (co-chopped as one sample, with leaves layered on top of each other) with goldenseal.

Data collection and analysis

A total of 12 goldenseal plants were analyzed, 8 were wild grown and 4 were in vitro propagated plantlets. Each plant sample was analyzed in triplicates with each replicate measurement performed on different days. Nuclear DNA content of goldenseal was estimated using the linear relationship according to the following equation (Doležel et al. 2007):

$$ {\text{Sample 2C value}}\,\left( {\text{pg}} \right) = {\text{Reference 2C value}} \times {\frac{\text{Sample\,2C\,mean\,peak\,position}}{\text{Reference\,2C\,mean\,peak\,position}}} $$

Conversion from picograms (pg) to base-pair numbers (bp) was done as follows: 1 pg DNA = 0.978 × 109 bp (Doležel et al. 2003). The results were analyzed using one-way analysis of variance in SAS (SAS Institute, Inc. 2003).

Results and discussion

From FCM analysis, the difference between mean peak position of pea nuclei prepared separately and those prepared by co-chopping with goldenseal was not significant (Fig. 2a, b). This implies that either addition of 0.25% DTT into the isolation buffer was able to minimize the cytosolic effects of goldenseal, or that the secondary compounds (alkaloids) present in goldenseal do not inhibit PI staining of pea DNA. The 2C nuclear DNA content of goldenseal was estimated to be 2.62 ± 0.020 pg. There was no significant difference (P ≤ 0.05) in nuclear DNA content between wild plants and in vitro plantlets (Fig. 1c, d; Table 1). This implies that the genome size of goldenseal plantlets produced in vitro remained stable even after long-term culture (>3 years) with repeated sub-culturing (every 4 weeks). Variation in genome size has been observed in medicinal plantlets produced through tissue culture (Sliwinska and Thiem 2007). From this study it is inferred that goldenseal plantlets produced in vitro using the protocol developed by Bedir et al. (2004) maintains the same genome size as wild plants, and presents a potential for use in mass production of goldenseal for restoration of threatened wild populations or commercial cultivation.

Fig. 2
figure 2

Histograms of relative fluorescence intensity of Pisum sativum cv. Ctirad (internal reference standard) nuclei separately processed (a) and simultaneously processed (co-chopped) with Hydrastis canadensis (b)

Full size image
Table 1 Nuclear DNA content and genome size (mean ± SD) of Hydrastis canadensis estimated with flow cytometry using P. sativum cv. Ctirad (2C = 9.09 pg) as an internal reference standard
Full size table

Based on records from Plant DNA C-values Database (Bennett and Leitch 2005), goldenseal genome size falls in the lower end of C-value distribution of both its family (155 records, ranging from 1.10–50.02 pg/2C) and among angiosperms (total records 4427, ranging from 0.02–254.80 pg/2C). With our estimation, goldenseal genome size is approximately 8.2 times bigger than Arabidopsis (Arabidopsis thaliana, 0.32 pg/2C; Bennett et al. 2003) and 2.6 times that of rice (Oryza sativa, 1.0 pg/2C; Bennett and Smith 1991). Determining the genome size is among the initial steps in deciding on species and strategy for genome sequencing (Gregory 2005). This report is the first to estimate nuclear DNA content of goldenseal and therefore provides important information needed to facilitate further genomic and molecular studies of this economically important medicinal species, and for use in identifying a suitable candidate for a model species to study medicinal plants.