RNA was extracted from the fresh leaves of an individual of Chrysosplenium aureobracteatum plant sampled from Gwangdeoksan Mt. (Voucher No. KYI_2014358) (Table 1) using a RNeasy Plant Mini Kit (Qiagen, Valencia, CA, USA). The extracted RNA was used for a cDNA synthesis using a TruSeq cDNA Library Prep Kit, version 2 (Illumina, San Diego, CA, USA). The synthesized cDNA library was used for Illumina Hiseq 100 bp paired-end de novo transcriptome sequencing. Raw reads acquired by RNA sequencing were subjected to quality checking and trimming using Trimmomatic 0.32 (Bolger et al., 2014). EST-SSR motifs were obtained from the trimmed data using the Trinity (Hass et al., 2013) and the MIcroSAtellite (MISA) identification tools (Thiel et al., 2003). Primer sets were constructed to amplify the selected EST-SSR motifs using the Primer3 program (Rozen and Skaletsky, 1999). EST-SSR candidate loci (markers) were selected based on the number of SSR motif repetitions (5–8 times), the PCR product size (100–300 bp), the coverage depth (20–110), and the primer annealing temperature (55–60°C).

The candidate EST-SSR markers developed from C. aureobracteatum were applied to its most closely related species, C. barbatum, to confirm and obtain markers exhibiting polymorphism in both species. We examined 20 individuals of Chrysosplenium aureobracteatum and C. barbatum for the marker evaluation and cross-specific amplification test. The voucher specimens were deposited in the HHU (Herbarium of Hallym University) in Korea (Table 1).

PCR for the selected loci was conducted using a mixture with a total volume of 25 μL consisting of 10× Ex Taq buffer at 2.5 μL (TaKaRa Bio Inc., Otsu, Japan), 2.5 mM dNTP at 2 μL, 10 pmol forward and reverse primer labeled with fluorescent dye (FAM, HEX, NED) each at 0.5 μL, template DNA (5–10 ng) at 2 μL, Ex Taq DNA polymerase at 0.12 μL and SIGMA free water using a GeneAmp PCR System 9700 thermocycler (Applied Biosystems Inc., San Diego, CA, USA). The PCR process was performed under the following conditions: pre-denaturation (98°C, 5 min), denaturation (95°C, 1 min), annealing (59°C, 1d min), and extension (72°C, 1 min 30 sec). After going through the above reaction 35 times, the final extension was processed (72°C, 10 min). The PCR products were electrophoresed at 110 V in 1.2% agarose gel for 90 min, stained with ethidium bromide, and confirmed using a UV illuminator. Alleles were identified with an ABI 3730XL (automated sequencer) equipped with the GeneScan 500 LIZ Size Standard (Applied Biosystems). Genotyping was performed using GeneMapper 3.7 (Applied Biosystems), and peaks were scored manually by visual inspections. The PCR products with sizes deviating from the expected size ranges (as estimated by the Primer3 program) were eliminated from the genotyping process. Finally, the resulting genotypes were encoded to calculate the number of alleles (A), the observed heterozygosity (H_o), and the expected heterozygosity (H_e) using the GenAlEx 6.5 (Peakall and Smouse, 2012).

As a result of RNA sequencing, raw data of 8,486,210,570 bp were generated (GenBank BioProject number: PRJNA606349). The raw data were processed to produce 7,818,963,245 bp of trimmed data, with this data then utilized to produce a total of 70,753,963 bp of transcriptome data. A total of 15,547 SSR motifs were identified from the trimmed data. The identified SSR motifs consisted of 40% mono-nucleotides, 36% dinucleotides, 23% tri-nucleotides, and 1% tetra-, penta-, and hexa-nucleotides (Table 2).

From the identified SSR motifs, 96 candidate markers were selected in consideration of the number of repetitions, the coverage depth, the annealing temperature, and the PCR product size. Excluding nonspecific bands with sizes deviating from the expected sizes, a total of 17 EST-SSR loci (markers) showing polymorphisms for Chrysosplenium aureobracteatum and C. barbatum were finally developed (Table 3).

As a result of applying the 17 EST-SSR markers on a population (with 20 individuals) of C. aureobracteatum, polymorphisms were detected from 11 markers, and a total of 33 alleles were identified (2 to 7 alleles depending on the polymorphic markers). The observed heterozygosity (H_o) and the expected heterozygosity (H_e) ranged from 0.000 to 0.750 and 0.000 to 0.543, respectively.

Cross-specific amplification for a population (with 20 individuals) of C. barbatum confirmed that 13 markers were polymorphic out of 15 amplifiable markers, resulting in a total of 47 alleles (2 to 8 alleles depending on the polymorphic markers). The observed heterozygosity (H_o) and the expected heterozygosity (H_e) ranged 0.000 to 0.850 and 0.000 to 0.779, respectively (Table 4).

It is important to note that the six monomorphic markers in C. aureobracteatum (CA_005, CA_006, CA_010, CA_052, CA_066, CA_104) were highly polymorphic in C. barbatum. In contrast, the monomorphic markers for C. barbatum (CA_056, CA_073) exhibited polymorphism in C. aureobracteatum (Table 4). These results suggest that the EST-SSR markers developed for C. aureobracteatum are useful not only to examine the population genetic structure but also for species delimitation among related taxa.