Construction of Fingerprint for Michelia Germplasm by Fluorescent SSR Markers

In this study, a fingerprinting system for 16 Michelia germplasms was established using high-throughput genotyping of fluorescent simple sequence repeat (SSR) markers. In total, 102 alleles were detected using 13 pairs of SSR markers from closely related species, and the number of alleles per locus ranged from 5 to 11, with an average of 7.8 alleles. Observed heterozygosity (Ho), expected heterozygosity (He), and polymorphic information content (PIC) ranged from 0.1250 to 0.5625 (mean 0.3650), 0.6703 to 0.9113 (mean 0.8099), and 0.5748 to 0.8714 (mean 0.7515), respectively. Among the four selected core primer pairs, the combinations LT106 and SGA5, LT106 and LT58, and SGA5 and MMA51 could unambiguously distinguish 16 Michelia germplasms. A cluster analysis showed the similarity coefficient of the 16 Michelia germplasms to range from 0.70 to 0.90, and different individuals from the same species clustered in the same branch. The fluorescence SSR genotyping system established in this study was efficient, rapid, and accurate; moreover, this approach provides a theoretical basis for identifying germplasms and for the protection of new varieties of the genus Michelia, and it provides a robust foundation for further breeding development of Michelia sp.

complicates the evaluation of genetic resources, variety selection, and popularization of Michelia varieties. At the same time, in view of the need for the legal protection of new plant varieties, a robust, efficient, and reliable molecular marker identification system is required, which will contribute to the protection of commercially registered varieties and can ensure further development of the Michelia planting industry in China.
The molecular marker technology is based on DNA and cannot be affected by the environment; moreover, this approach which is simple, efficient, and provides accurate and reliable results, has been widely used in various fields of genetics and genomics. Among molecular markers, simple sequence repeats (SSR) have been proven considerably useful as genetic markers owing to substantial genome coverage, hyper-variability, co-dominant inheritance, high interspecies transferability, and good reproducibility (Kalia et al., 2011). This approach has been successfully applied for phylogenetic analyses and germplasm identification of several woody species, e.g., Populus sp. Liu et al., 2016;Ou et al., 2017), Eucalyptus sp. , Pinus massoniana , Ginkgo biloba , Camellia oleifera , Amygdalus persica (Wang et al., 2016), Osmanthus fragrans (Duan et al., 2014), Tamarix chinensis (Ye et al., 2015), and Litchi chinensis (Madhou et al., 2013). However, the molecular biological study of Michelia in China commenced relatively late, and only one scientific report has been published so far (Wen et al., 2014). Therefore, EST-SSR primers for M. maudiae and Liriodendron tulipifera were chosen and screened for interspecies transferability and detection of polymorphisms, which were subsequently labeled using fluorescent probe and used to establish a fingerprinting system for 16 Michelia germplasms. The objective of our study was to establish an efficient, rapid, and accurate genotyping system for the genus Michelia, to provide a scientific theoretical basis for the identification of germplasm resources and for the legal protection of new varieties, and to provide reference for future breeding of Michelia.

Cross-species amplification
A set of 34 SSR primer pairs was selected to evaluate interspecific transferability in eight Michelia species: M. chapensis, M. maudiae, M. Jinxiu, M. wilsonii, M. foveolata, M. platypetala, M. compressa, and M. macclurei. Out of ten primer pairs for M. maudiae, only one primer pair failed to show amplification from M. platypetala samples, whereas the other nine primer pairs showed successful amplification from samples of all eight species, which indicated high interspecific transferability. Using the 24 L. tulipifera primer pairs, amplification efficiency ranged from 21% in M. wilsonii and M. macclurei to 38% in M. maudiae (Table 1). Therefore, nine primer pairs of M. maudiae and four primer pairs of L. tulipifera were selected and labeled with fluorescent probes for Michelia genotyping and fingerprinting.

Polymorphism evaluation
A total of 13 EST-SSRs were used to assess polymorphisms in 16 Michelia germplasms, which produced 102 alleles, and the number of alleles per locus ranged from 5 to 11 with an average of 4.8 alleles. The highest number of alleles (n=11) was amplified using SGA5, MMA72, and LT106, and the lowest number (n=5) was produced using Polymorphic information content (PIC) varied from 0.5748 to 0.8714, with an average of 0.7515. The highest PIC (0.8714) was produced using SGA5 and the lowest PIC (0.5748) was produced using LT186 (Table 2). Partial genotyping markers of Michelia germplasms using SGA5 is shown in Figure 1.

Fingerprinting of Michelia
Four core primer pairs were selected from 13 SSR primer pairs to establish a fingerprinting system for Michelia germplasms by further analyzing the allelic fragments of each locus (Table 3). Ten Michelia germplasms could be distinguished using the primer pair LT106, and eight Michelia germplasm could be distinguished using the primer pair SGA5. All 16 Michelia germplasms could be identified using a combination of the primers LT106 and SGA5 or a combination of the primers LT106 and LT58, or SGA5 and MMA51. Therefore, the four selected core primer pairs can be used to distinguish the germplasms examined here and also for reciprocal validation by using different combinations of primers.

Cluster analysis
Genetic similarity coefficients were calculated and clustered based on UPGMA using the 13 fluorescently amplified SSR markers (  SSR markers are widely distributed in eukaryotic and prokaryote genomes. Compared with genomic SSR, EST-SSR markers located in a transcription region are typically less polymorphic than genomic SSR (He, 2010), therefore they tend to be more conserved and thus are more compatible between species (Xu et al., 2006;Yadav et al., 2011), genera (Yadav et al., 2011;Li and Sun, 2012), and families (Decroocq et al., 2003). A fingerprinting system for 16 Michelia germplasms was successfully established using EST-SSR markers of M. maudiae and L. tulipifera. In this study, the transferability of EST-SSRs of Liriodendron and Michelia was 21-38%, which was lower than that observed by 24-71%) and lower than that of Liriodendron and Magnolia (54%). However, the transferability of EST-SSRs is generally affected by the proportion of conserved SSR-flanking sequences (Cordeiro et al., 2001), the position of the primer sequences in the EST , genome size, life span, and the type of mating system (Ueno et al., 2009).
Unlike traditionally used marker such as RAPD, ISSR, and AFLP, SSRs are co-dominant, which means that homozygous or heterozygous loci can be distinguished and the complete genetic information can be extracted (Song et al, 2012). Moreover, owing to hyper-variability, closely related species and even hybrid progeny can be accurately identified. For example, in species related to Osmanthus fragrans (Duan et al., 2014), Populus , Ginkgo biloba , or Camellia sinensis (Chen et al., 2017), and in hybrid progeny of Juglans regia (Pollegioni et al., 2009), Ziziphus jujuba , Myrica rubra  and other woody plants, fingerprinting systems were successfully established using SSR markers. Compared with traditional high-resolution agarose gel electrophoresis and polyacrylamide gel electrophoresis, the technique of fluorescence SSR capillary electrophoresis genotyping based on high-throughput sequencing are easy to use, less time consuming, highly efficient, cost-effective, highly accurate and reproducible (Liesebach et al., 2010;Fariaet al., 2011). To test large numbers of individuals, multiple PCRs can be employed to save costs and increase throughput (Drašnarová et al., 2014;Wu et al., 2016). The fingerprinting system established in this study may help distinguish eight closely related species by using different primer combinations and can also be used to identify different individuals of M. chapensis and M. maudiaespecies, emphasizing the high efficiency and accuracy of SSR markers.
The classification of genera and species of the Magnoliaceae has traditionally been an issue of disagreement, including the phylogenetic classification of Michelia (Zhang, 2007). The phylogenetic history and divergence of the genus Michelia has been widely discussed, and at present, there are two taxonomic systems, one suggested by Liu et al. (1996) and the other by Chen and Nooteboom (1993). In the current study, eight species of Michelia were assigned to five groups. In line with Li (2008; (Zhang, 2007), and a different number or type of markers and different sample numbers may also lead to discrepancies in taxonomic classifications. Therefore, in order to accurately classify the genus Michelia, multiple criteria such as morphology, cytology, and molecular biology should be applied.  Table 4). From each individual, young leaves were collected, immediately frozen using liquid nitrogen, and stored at -70°C. A commercially available plant DNA extraction kit (DP305; Tiangen Biochemistry Technology Co. Ltd, Beijing) was used to isolate total DNA, according to the manufacturer's instructions. DNA integrity was assessed by 0.8% agarose gel electrophoresis. Purity and DNA concentration were determined using a UV spectrophotometer, and the samples were stored at -20°C.

Primer screening and transferability detection
Out of 49 SSR markers derived from M. maudiae (Sun et al., 2010) and L. tulipifera (Yang et al., 2012), 34 markers with high transferability were selected for cross-species amplification. Genomic DNA of individuals 1, 6, 11, 12, 13, 14, 15, and 16 was used as templates for preliminary primer screening after which 13 SSR markers were selected for fingerprinting (Table 5). All markers were labeled with fluorescent groups at the 5' end as mentioned above and were synthesized by Shanghai Lingen Biotechnology Co., Ltd.
PCR reactions of 10 µL contained 5 µL PCR mix, 3.4 µL ddH2O, 0.3 µL forward primer, 0.3 µL reverse primer, and about 1 µL DNA template. The reaction was performed using the following thermocycling program: 94 °C for 5 min, followed by 35 cycles of 94 °C for 50 s, 56 °C for 50 s, and 72 °C for 90 s, and a final extension step at 72 °C for 7 min. PCR products were visualized by electrophoresis using 1.5% agarose gels.

Genotyping system establishment
The PCR reaction volume for genotyping was 15 µL ( Table 6). The reaction was performed under the following thermocycling conditions: 94 °C for 3 min, followed by 35 cycles of 94 °C for 15 s, 55 °C for 15 s, and 72 °C for 30 s, and a final extension step at 72 °C for 3 min. The PCR products were visualized by agarose gel electrophoresis; after this, the products were diluted 10-fold and mixed with the internal standard of ROX 500. The denaturation reactions of a volume of 10 µL contained 8.25 µL HiDi, 0.25 µL ROX 500, and 1.5 µL diluted PCR product. After 5 min of denaturation at 95 °C, the reaction was immediately placed on ice for 3 min and then sequenced using an ABI 3730xl DNA Analyzer (Applied Biosystems, Foster City, CA), according to the manufacturer's instructions.

Data analyses
The number of alleles (Na), He, Ho, and PIC were calculated for each EST-SSR locus using MSA software.  Alleles were transformed to a 0/1 matrix according to the fragments, and SM similarity coefficients were calculated. The software NTSYS-pcversion 2.1 was used to cluster the 16 individuals of Michelia using a UPGMA method.