Cypripedium tibeticum King ex Rolfe (Orchidaceae), known as Chinese lady slipper, is a unique orchid species with high ornamental and medicinal value. It is endemic to southwestern China, including Tibet, Gansu, Yunnan, Guizhou, and Sichuan provinces. Cypripedium tibeticum mainly grows in alpine meadows, scrub forests, and forest margins at high altitudes (2800–4200 m a.s.l.). The natural population size of this orchid is usually limited by habitat fragmentation and overharvesting (Chen and Tsi, 1998; Fay and Chase, 2009; Swarts and Dixon, 2009; Bronstein et al., 2014). Due to habitat destruction and human activities, the survival of C. tibeticum in its native range is threatened, and it is currently listed as a national protected plant (Luo et al., 2003; Qian et al., 2014).
Microsatellites (simple sequence repeats [SSRs]) are widely used molecular markers that are applied in population genetics to investigate genetic diversity, population structure, and evolutionary history (Hodel et al., 2016; Vieira et al., 2016). To date, microsatellite markers have been developed for several Cypripedium L. species; however, scarcely any markers are applicable to C. tibeticum (Fay and Cowan, 2001; Pandey and Sharma, 2013; Yamashita et al., 2016). In this study, we developed and characterized 20 polymorphic SSR markers in C. tibeticum and tested their applicability in two related species: C. flavum P. F. Hunt & Summerh. and C. bardolphianum W. W. Sm. & Farrer. These microsatellites will be valuable tools for ecological, phylogeographic, and conservation studies of C. tibeticum and other related Cypripedium species.
METHODS AND RESULTS
Plant sample collection and DNA extraction—Leaf samples of C. tibeticum were collected from three locations in Sichuan Province, including Huanglong, Kangding, and Xiaojinxian in China (population codes: HL, KD, and XJX, respectively [Appendix 1]). Genomic DNA was extracted from fresh leaf tissue of each individual using the DNeasy Plant Mini Kit according to manufacturer instructions (QIAGEN, Valencia, California, USA).
Development and screening of microsatellite markers—Total genomic DNA was digested with the restriction enzyme HaeIII, and products were 5′ phosphorylated and a single A was added at the 3′ end. DNA was ligated to a Solexa adapter and sequenced on a single HiSeq 2500 flow cell (Illumina, San Diego, California, USA). A total of 7,210,538 reads longer than 100 bp were obtained, and raw reads with quality scores less than 25 and lengths shorter than 25 bp after stripping the adapters were filtered using SeqPrep ( https://github.com/jstjohn/seqprep). De novo assembly was performed using CLC Genomics Workbench (QIAGEN) and produced 1,920,476 contigs, where parameters were set as: average coverage value of 25× and minimum contig length of 299 bp. MSATCOMMANDER 1.0.8 software was used to identify contigs carrying di-, tri-, and tetranucleotide repeats with a minimum of five repeats and a minimum tract length of 100 bp (other parameters were set to default settings) (Faircloth, 2008). Ninety-four primer pairs were designed for microsatellite loci candidates using Primer Premier 5.0 software (Premier Biosoft International, Palo Alto, California, USA) and tested with 15 C. tibeticum samples (five individuals from each population) for amplification efficiency and polymorphism. Each 15-µL PCR reaction contained ∼15 ng of genomic DNA, 1.5 µL of 10× PCR buffer, 0.8 µL of fluorescently labeled TP-M13 (5 mM), 1.0 µL of each primer (10 mM), 1.5 µL of dNTP (10 mM), and 0.1 µL of Platinum Taq DNA polymerase (Invitrogen, Carlsbad, California, USA). All candidate primer pairs were tested by a touchdown PCR protocol as follows: 94°C for 2 min; five cycles of 94°C for 30 s, 60–56°C (Δ1°C touchdown per cycle) for 30 s, 72°C for 30 s; followed by 35 cycles of 94°C for 30 s, 55°C for 30 s, 72°C for 30 s; and a final extension at 60°C for 30 min.
Table 1.
Characteristics of 20 microsatellite loci developed in Cypripedium tibeticum.

Results from testing primer pairs showed that 26 pairs produced bands within the expected size range on agarose gels in all the samples tested. The size of amplified fragments from these loci was further analyzed on an ABI PRISM 3130×l Genetic Analyzer (Applied Biosystems, Waltham, Massachusetts, USA), and allele sizes were scored by GeneMapper version 3.2 software (Applied Biosystems). Twenty loci were polymorphic and showed clear peaks. The authenticity of these amplified loci was confirmed by Sanger sequencing of representative PCR products, and sequences have been deposited in GenBank (Table 1). Sequence library data of this study were deposited to the Sequence Read Archive of the National Center for Biotechnology Information (NCBI; BioProject ID: PRJNA393499).
Data analysis and results—These 20 microsatellite loci were PCRamplified in an additional 74 individuals from three C. tibeticum populations collected in Sichuan Province. For each locus, the observed number of alleles, effective number of alleles, observed heterozygosity, and expected heterozygosity were calculated using PopGene32 version 1.32 (Yeh et al., 1999). Tests for Hardy–Weinberg equilibrium and linkage disequilibrium were performed by GENEPOP Web version 4.2 (Rousset, 2008). The total number of alleles per locus ranged from two to 21 (mean ± SD: 6.350 ± 4.320). The observed heterozygosity and expected heterozygosity ranged from 0.261 to 0.967 (0.664 ± 0.143) and from 0.441 to 0.960 (0.745 ± 0.119), respectively (Table 2). Of the 20 polymorphic loci, nine loci in the Huanglong population, six loci in the Kangding population, and five loci in the Xiaojinxian population deviated significantly from Hardy–Weinberg equilibrium, respectively (P < 0.05; Table 2). Linkage disequilibrium was not detected at any locus.
The utility of these 20 microsatellite loci developed for C. tibeticum was also detected in two other Cypripedium species: C. flavum and C. bardolphianum. Eighteen loci were successfully PCR-amplified in C. flavum, and 17 loci were amplified in C. bardolphianum (locus M294 did not amplify in C. bardolphianum, M370 and M886 did not amplify in either species). Genotyping results showed locus M372 was monomorphic in both species, and M136 was polymorphic in C. bardolphianum but monomorphic in C. flavum. All the other markers were polymorphic in both species (Table 3).
Table 2.
Genetic variation of the 20 polymorphic microsatellite loci in three populations of Cypripedium tibeticum.a

Table 3.
Characteristics and polymorphism of 20 microsatellite loci developed for Cypripedium tibeticum in C. flavum and C. bardolphianum.a

CONCLUSIONS
In this study, we developed and validated 20 polymorphic microsatellite markers for the orchid C. tibeticum, most of which showed applicability in two related Cypripedium species: C. flavum and C. bardolphianum. These markers will be useful for population genetic investigation and species conservation in natural habitats of C. tibeticum and other closely related species.
ACKNOWLEDGMENTS
The authors thank Dejun An for help with sample collection and Dr. Brad S. Coates (Research Geneticist, U.S. Department of Agriculture–Agricultural Research Service) for assistance in editing this manuscript. This study was supported by the National Natural Science Foundation of China (NSFC 31300366), the Natural Science Foundation of Shaanxi Province (2015JQ3091), the Natural Science Foundation of Xi'an Science and Technology Bureau (CXY1531WL28), and the Scientific Research Foundation of the Education Department of Shaanxi (15JK2146).