Plants of species within Lespedeza Michx. sect. Macrolespedeza Maxim. (Fabaceae) are deciduous shrubs mainly distributed in the temperate forests of East Asian countries such as Korea, China, and Japan (Ohashi and Nemoto, 2014). Based on morphological and molecular studies (Lee, 1965; Akiyama, 1988; Han et al., 2010; Xu et al., 2012), hybridization is frequent among these species. Therefore, their taxonomic entities are often unclear.

One representative, L. maritima Nakai, is a Korean endemic that primarily occurs at low elevations on mountains along the southern coast of the Korean Peninsula. This species is distinguished from others within the genus by its unique morphological traits, e.g., leaflets with leathery surfaces and revolute margins, as well as abundant sprouting from the basal portion of the plant. Despite these distinctive characteristics, the taxonomic entity of L. maritima has been controversial. First described by Nakai (1923), this species has since then been regarded either as a hybrid between L. cyrtobotrya Miq. and L. maximowiczii C. K. Schneid. (Lee, 1965) or as a synonym of L. thunbergii (DC.) Nakai subsp. thunbergii (= L. formosa (Vogel) Koehne subsp. velutina (Nakai) S. Akiyama & H. Ohba) (Akiyama, 1988), thus making its boundary of taxonomic delimitation more ambiguous. Previous phylogenetic analysis (Xu et al., 2012) indicated low resolution for chloroplast DNA (cpDNA) and nuclear ribosomal internal transcribed spacer (nrITS) sequences among species. In plants such as those of Quercus L., codominant polymorphic microsatellite markers have been used to confirm that hybridization is frequent among those species (see, e.g., Lee et al., 2014). Advances in sequencing techniques, i.e., next-generation sequencing (NGS), have made it possible for researchers to obtain microsatellite libraries rapidly. Thus, microsatellite markers could be applied to resolve the above-mentioned debate about L. maritima and other Lespedeza species if they were available for that genus. Here, we used NGS to develop 28 microsatellite markers to clarify the taxonomic entity of L. maritima.

METHODS AND RESULTS

We sampled 47 individuals of L. maritima from Bogil Island (n = 32; 34°08′16″N, 126°31′49″E) and Namhae Island (n = 15; 34°44′33″N, 128°01′20″E), located in the South Sea of Korea. To avoid repeated collections from the same plant, we selected individuals that grew at least 5 m apart. The leaves were preserved with silica gel, and voucher specimens were deposited in the Herbarium of Inha University (IUI), Incheon, Korea (Appendix 1). Genomic DNA, used to acquire the microsatellite library, was extracted from the silica gel–dried leaves of one individual (voucher no. DP149122) with a DNeasy Plant Mini Kit (QIAGEN, Seoul, Korea). The extracted DNA (300 ng) was fragmented to 500 bp using a Covaris S220 (Covaris, Woburn, Massachusetts, USA). A TruSeq Nano DNA Library Preparation Kit (Illumina, San Diego, California, USA) was used for library preparation after sequencing on the Illumina MiSeq platform at Life Is Art of Science (LAS; Daejeon, Korea;  http://lascience.co.kr/). In all, 5,379,227 pairs for 10,758,454 reads (300 × 300) were generated. Through SSR_pipeline version 0.951 (Miller et al., 2013), we selected the microsatellite library containing di- or trinucleotide repeats for which at least 10 or five, respectively, were obtained from total reads. Of the obtained microsatellite library, we initially discarded those that had flanking regions of less than 100 bp and more than 20 repeats. Then we attempted reference mapping of total paired reads to each remaining sequence using Geneious R 7.1.8 (Biomatters, Auckland, New Zealand). For this process, putative multicopy loci with exceptionally high coverage were removed, while certain other loci were selected based on the following criteria: (1) loci showing unique patterns that had two separate alleles and few variations at the site to which a primer was attached, and (2) no additional single nucleotide polymorphisms (SNPs) in the flanking region. From those, we designed 65 primer pairs with Primer3 (Rozen and Skaletsky, 1999) and added the M13(−21) sequence (5′-TGTAAAACGACGGCCAGT-3′) to the 5′ end of all forward primers intended for labeling with a fluorochrome. To test for amplification and polymorphism of each primer pair, we conducted PCR for 47 individuals of the two populations with a Gene-Amp PCR System 2700 Thermal Cycler (Applied Biosystems, Foster City, California, USA). Each reaction mixture (10 µL total volume) contained 5 ng of DNA plus 5 µL of 2× Plus Mix (Dongsheng Biotech, Guangdong, China), which comprised 0.4 mM dNTPs, 2× PCR buffer with 4 mM MgSO₄, and 0.4 U·µL^-1 of Taq DNA polymerase. The mixtures also contained the appropriate 0.08 µM forward M13 (−21)-tagged primer, a 0.3 µM reverse primer, and a 0.3 µM M13(−21) labeled fluorescent marker (NED, PET, VIC, or 6-FAM). Conditions included initial denaturation at 94°C for 3 min; then 30 cycles at 94°C for 30 s, 53°C for 45 s, and 72°C for 45 s; with a final extension at 72°C for 7 min. Afterward, the PCR products were visualized on 2% agarose gels and resolved to genotype on an ABI 3730XL sequencer with GeneScan 500 LIZ Size Standard (Applied Biosystems). The sizes of the alleles were determined with GeneMapper 3.7 (Applied Biosystems). Using GenAlEx 6.5 (Peakall and Smouse, 2006), we estimated the total number of alleles (A_T) and number of alleles (A), plus values for observed heterozygosity (H_o), expected heterozygosity (H_e), and deviation from Hardy–Weinberg equilibrium (HWE) for each population. Null allele frequencies were calculated by MICRO-CHECKER version 2.2 (van Oosterhout et al., 2004). Additionally, we tested cross-amplification with four related species (L. bicolor Turcz., L. cyrtobotrya, L. thunbergii subsp. thunbergii, and L. maximowiczii) in successfully amplified primers, using two individuals per species. Among those species, the first three were sampled at Baegun Mountain (37°29′32″N, 126°30′58″E) while L. maximowiczii was collected at Biseul Mountain (35°42′16″N, 128°31′58″E). Sequences of the raw paired-end reads, successfully amplified and scored, are available in the SRA database of the National Center for Biotechnology Information (NCBI) (Bioproject no. PRJNA290063).

Table 1.

Characteristics of 28 microsatellite loci developed in Lespedeza maritima.

Table 2.

Genetic diversity values for 47 individuals of Lespedeza maritima across 28 microsatellite loci.

Table 3.

Results of cross-amplification for the 28 microsatellite markers for Lespedeza maritima in four related species.

Of the 65 initially designed primer pairs, 45 were successfully amplified for 47 individuals from our two populations. Among those 45 primer pairs, 17 showed scoring errors. Finally, we selected 28 pairs, all of which were polymorphic (Table 1). The values for genetic diversity for the 47 individuals of L. maritima were as follows (Table 2): A_T varied from two to 13 ; H_o and H_e per locus ranged from 0.067 to 0.867 and from 0.064 to 0.848, respectively; eight loci (LMS5, LMS8, LMS11, LMS18, LMS24, LMS45, LMS52, and LMS62) showed significant deviation from HWE on Bogil Island, while six loci (LMS11, LMS17, LMS24, LMS33, LMS45, and LMS49) showed significant deviation from HWE on Namhae Island. Evidence of null alleles detected at most loci (except for LMS17) indicated significant deviations from HWE. Therefore, this lack of equilibrium could be explained by the presence of those null alleles. Our results from the cross-amplification showed that all loci, except LMS5, LMS8, LMS33, LMS34, LMS39, and LMS45, were well amplified for the four related species (Table 3).

CONCLUSIONS

We have developed a set of 28 polymorphic microsatellite markers for L. maritima. Most are well amplified in four related species within section Macrolespedeza: L. bicolor, L. cyrtobotrya, L. thunbergii subsp. thunbergii, and L. maximowiczii. Future studies will use these new markers as tools to reveal the taxonomic entity of L. maritima.