Hamamelidaceae comprises 31 genera and more than 140 species (Li et al., 1999) and includes several ornamental genera within Corylopsis Siebold & Zucc., Fothergilla L., Hamamelis L., Loropetalum R. Br., and Parrotia C. A. Mey. Fothergilla has been used in ornamental plantings for over two centuries and there are fewer than 15 cultivars, whereas Loropetalum has 19 cultivars and Corylopsis and Parrotia have five or fewer cultivars (Dirr, 1998). Hamamelis species are used as an astringent and are more widely recognized, with more than 75 cultivars (Marquard et al., 1997). Many cultivars from these genera are commercially available, but the pedigrees are not well known as they are often selected from spontaneous mutations, wild-grown seedlings, or open-pollinated crosses, as is the case with Hamamelis (Marquard et al., 1997).

The phylogeny of Hamamelidaceae has been examined and, based on nrDNA ITS sequences, a well-supported phylogeny with three clades was resolved (Li et al., 1999). Fothergilla, Hamamelis, and Parrotia were in one clade, whereas Corylopsis and Loropetalum were in a different clade. Corylopsis, Loropetalum, and Parrotia are native to Asia. Loropetalum chinense (R. Br.) Oliv. is found in Japan and southeastern China, whereas L. lanceum Hand.-Mazz. is more widely distributed throughout Japan, China, and northeastern India (Zhang et al., 2003). Only four populations of L. subcordatum (Benth.) Oliv. remain, making it one of the most endangered angiosperm species in China (Gong et al., 2010). Parrotia persica (DC.) C. A. Mey. is a deciduous tree endemic to northern Iran; it is the only extant species in the genus and could become a conservation concern if habitat destruction continues (Sefidi et al., 2011). Species from the genus Hamamelis are found on both the North American and Asian continents (Zhang et al., 2003). Fothergilla is the only genus exclusively limited to North America. The genus is found in the southeastern United States and includes only two species, F. major Lodd. and F. gardenia L., as well as the hybrid Fothergilla ×intermedia. Both species are of conservation concern, F. major in Tennessee and F. gardenii in both Florida and Georgia (USDA, 2012).

Molecular markers can be used to determine diversity in wild populations and assist in breeding and conservation studies. The purpose of this study was to develop a microsatellite-enriched library from Fothergilla ×intermedia to establish loci capable of distinguishing species and cultivars, to assess genetic diversity for use by ornamental breeders, and to test these loci for crossover to other members of Hamamelidaceae, such as those in the related genera Corylopsis, Hamamelis, Loropetalum, and Parrotia.

METHODS AND RESULTS

Leaf tissue or unopened flower buds were obtained from the J. C. Raulston Arboretum at North Carolina State University (Raleigh, North Carolina, USA), Spring Grove Cemetery and Arboretum (Cincinnati, Ohio, USA), Arnold Arboretum of Harvard University (Boston, Massachusetts, USA), and University of Tennessee Gardens (Knoxville, Tennessee, USA). Approximately 100 mg of tissue was homogenized in 2.0-mL microcentrifuge tubes (Fisher Scientific, Pittsburgh, Pennsylvania, USA) containing silica beads (2.3 mm; BioSpec Products, Bartlesville, Oklahoma, USA) and frozen in liquid nitrogen for 5 min followed by agitation in a Bio101 FastPrep Homogenization System FP120 (Thermo Savant, Waltham, Massachusetts, USA) for 30 s at the 5.0 speed setting and freezing and agitation were repeated once. DNA was isolated using the QIAGEN DNeasy Plant Mini Kit (QIAGEN, Valencia, California, USA) with the following modifications: 2% (w/v) insoluble polyvinylpyrrolidone (PVP) and 6 µL of RNase were added to 600 µL of AP1 buffer and cell lysis incubation was 20 min. DNA was quantified using a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, Delaware, USA). An enriched microsatellite library was created from accession number AA#182-96^*A (Appendix 1) following the procedures of Wang et al. (2007). Microsatellite-containing sequences were identified using Imperfect SSR Finder (Stieneke and Eujayl, 2007). Primer3 (Rozen and Skaletsky, 1999) was used to design 50 primer pairs, which were screened for amplification against a subset of four Fothergilla samples. A single 10-µL PCR reaction contained 10 ng DNA, 2.5 mM MgCl₂, 1× Gene Amp PCR Buffer II (Applied Biosystems, Carlsbad, California, USA), 0.2 mM dNTPs, 0.25 µM primers (forward and reverse), 5% dimethyl sulfide (DMSO; Fisher Scientific), 0.4 unit AmpliTaq Gold DNA Polymerase (Applied Biosystems), and sterile water. The reactions were amplified using the following conditions: 94°C for 3 min; 35 cycles of 94°C for 40 s, 55°C for 40 s, and 72°C for 30 s; and a final extension at 72°C for 4 min. The amplicons were separated by electrophoresis through a 2% agarose gel stained with ethidium bromide. Twelve loci were polymorphic, whereas the remaining loci did not amplify, produced a smear pattern, or were monomorphic. The polymorphic loci were used to characterize a larger sample size.

Table 1.

Characteristics of 12 microsatellite loci isolated from Fothergilla ×intermedia for three Corylopsis, 15 Fothergilla, 14 Hamamelis, two Loropetalum, and two Parrotia accessions.^a

The 36 accessions of Hamamelidaceae were genotyped and analyzed in the same manner as primer screening (Appendix 1). PCR products were separated using the QIAxcel Capillary Electrophoresis System (QIAGEN) and sized with a 25–300-bp marker. Raw allele data for each individual were binned into allelic classes using FLEXIBIN (Amos et al., 2007). We used a conservative 2-bp allelic category size determination standard error range for reproducibility and the 2-bp resolution of the QIAxcel Capillary Electrophoresis System. Ploidy level varies from 4x to 6x in Fothergilla (Ranney et al., 2007) and 2x to 6x in Corylopsis as compared to the diploid Parrotia and Hamamelis species (Zhang et al., 2003). Due to ploidy variation, each allele was scored as either 0 (absent) or 1 (present), and nonamplified loci were scored as missing data. The data were analyzed using GenAlEx version 6.5 (Peakall and Smouse, 2012). For cluster analysis, genetic similarity indices were calculated for all pairwise comparisons using Dice's similarity coefficient and then clustered using the unweighted pair group method with arithmetic mean (UPGMA) using NTSYS-pc 2.20q (Rohlf, 2008). The cophenetic coefficient between the Dice similarity matrix and the UPGMA dendrogram was calculated (Rohlf, 2008).

Table 2.

Unique alleles detected at 12 microsatellite loci for five genera within Hamamelidaceae.^a

Fig. 1.

Cluster analysis of three Corylopsis, 15 Fothergilla, 14 Hamamelis, two Loropetalum, and two Parrotia accessions (see Appendix 1) after unweighted pair group method with arithmetic mean (UPGMA) clustering using 12 microsatellite loci coded as dominant (presence/absence) markers. The similarity of the accessions was calculated using the Dice coefficient. The cophenetic correlation coefficient value (r = 0.91), suggested a strong fit between the Dice similarity matrix and the UPGMA dendrogram using the parameters of Sneath and Sokal (1973).

Twelve primer pairs were polymorphic and used to genotype 36 accessions of Hamamelidaceae (Table 1). The number of alleles ranged from four (Foth029) to 15 (Foth16), and Shannon's information index ranged from 0.07 to 0.14. Five loci amplified across all genera: Foth001, Foth002, Foth004, Foth029, and Foth032. In total, 128 alleles were identified and 90 were unique to individual genera (Table 2). Foth045 amplified only Fothergilla accessions and 63 unique alleles were detected. Ten loci amplified in Corylopsis, with five loci detecting a total of seven unique alleles. Ten loci amplified in Hamamelis, with six loci detecting nine unique alleles. Nine loci amplified the Parrotia accessions and detected three unique alleles. For Loropetalum, five loci amplified and detected eight unique alleles. Cluster analysis grouped the accessions into five groups, which were separated by genus (Fig. 1). The cophenetic correlation coefficient value (r = 0.91) suggested a strong fit between the Dice similarity matrix and the UPGMA dendrogram using the parameters of Sneath and Sokal (1973).

CONCLUSIONS

We have developed the first set of microsatellites from Fothergilla and have demonstrated cross transfer to other species within Hamamelidaceae. These loci provide a set of markers to evaluate genetic diversity of natural and cultivated collections and assist ornamental plant breeders for genetic studies of five popular genera of woody ornamental plants.