Isolation and characterization of twelve polymorphic microsatellite markers in the endangered Hopea hainanensis (Dipterocarpaceae)

Abstract Microsatellite markers were isolated and characterized for Hopea hainanensis Merrill & Chun, an endangered tree species with scattered distribution in Hainan Island and northern Vietnam. Twenty‐six microsatellite markers were developed based on next‐generation sequencing data and were genotyped by capillary electrophoresis on an ABI 3730xl DNA Analyzer. Twelve markers were found to be polymorphic in H. hainanensis. GENODIVE analyses indicated that the number of alleles ranged from 2 to 6 per locus, and the observed and expected heterozygosity varied from 0 to 0.755 and from 0.259 to 0.779, respectively. Primer transferability was tested with Hopea chinensis Hand.‐Mazz. and Hopea reticulata Tardieu, in which 3 and 7 microsatellite markers were found to be polymorphic, separately. The results showed that H. reticulata and H. hainanensis had similar levels of genetic diversity. A neighbor joining dendrogram clustered all individuals into two major groups, one of which was exclusively constituted by H. hainanensis, while the other consisted of two subgroups, corresponding to H. reticulata and H. chinensis, respectively. The 12 polymorphic microsatellite markers could be applied to study genetic diversity, population differentiation, mating system, and fine‐scale spatial genetic structures of H. hainanensis as well as its close relatives, facilitating the conservation and restoration of these endangered but valuable Hopea species.

endangered according to the IUCN Red List of Threatened Species (Ly et al., 2018). In addition to the highly valued wood, H. hainanensis is rich in bioactive compounds. The extracts from stems and barks were reported to have potent antioxidant activities, which could be used as candidates for pharmaceutical products or food additives (Ge et al., 2009).
Trees of this family dominate Southeast Asia's tropical forests, accounting for 20%-50% of forest basal area and often well over 50% of canopy trees (Ashton, 1988;Ghazoul, 2016). Many species of this family constitute important timber resources and thus have been heavily exploited by local countries in tropical Asia. The unsustainable exploitation for timber and deforestation for agriculture render many dipterocarp species now being classified as endangered (Ghazoul, 2016). Understanding the genetic diversity, population structure and mating system of these endangered species is crucial and of priority for the effective management and conservation (Frankham, 1995). Population genetic studies focused on dipterocarp species have been carried out for the purpose of conservation and restoration (Finger et al., 2012;Ismail et al., 2014). Microsatellite markers are widely used to estimate genetic diversity, fine-scale spatial genetic structure, gene flow, and mating system for endangered species in Dipterocarpaceae (Finger et al., 2012;Lee et al., 2013; de Morais et al., 2015). However, the development of informative microsatellite markers is first step in population genetic studies.
In this study, we sequenced the genome of H. hainanensis using next-generation sequencing technology. Based on the assembled contigs, 26 novel microsatellite markers were developed and characterized using 50 individuals of this species, 12 of which were found to be polymorphic. The marker transferability was tested upon two additional Hopea species, H. chinensis Hand.-Mazz. and H. reticulata Tardieu. These newly developed microsatellite markers could be used as a universal tool in population genetic studies of H. hainanensis as well as its close relatives. and low-quality bases with default parameters (Bolger et al., 2014).

| MATERIAL S AND ME THODS
Clean reads were extended and merged by overlapping paired-end reads using FLASH with minimum and maximum overlaps of 20 and 100 bp, respectively (Magoc & Salzberg, 2011). The extended reads were clustered by CD-HIT with the minimum identity of 98% (Fu et al., 2012). Microsatellite motifs were screened by MISA (Thiel et al., 2003) with search parameters set as follows: at least six repeats for dinucleotide motifs, five repeats for tri-and tetranucleotide motifs, and four repeats for penta-and hexanucleotide motifs.
Two adjacent microsatellite motifs with base pairs less than 100 between each other were recognized as a compound microsatellite and discarded. Microsatellites with sufficiently long flanking regions were retained, based on which primers were designed and examined The PCR products were separately combined with a GeneScan 500 LIZ Size Standard (Life Technologies) and resolved by capillary electrophoresis on an ABI 3730xl DNA Analyzer (Applied Biosystems) at the TIANYI Biotechnology Company. Capillary electrophoresis is the preferred method for SSR genotyping because of its high resolving power and good repeatability (Mason, 2015 Normal University, who initiated the whole genome sequencing of the two Hopea species), allelic dosage was analyzed based on the ratios between peak intensities following the MAC-PR method (Esselink et al., 2004). GENODIVE version 3 was adopted to estimate genetic diversity and test deviation from Hardy-Weinberg equilibrium (HWE), as this software can take account of missing dosage information for partial heterozygotes of autopolyploid (Meirmans, 2020). Another challenge posed by autopolyploidy is polysomic inheritance, under which double-reduction may occur and bias the results of standard population genetic analyses (Huang et al., 2019). However, genotypic ambiguities caused by unknown allelic dosage in autopolyploid could not be fully resolved with the MAC-PR method (Esselink et al., 2004). Huang et al. (2020) developed a new software package named POLYGENE for estimating population genetic statistics directly from allelic phenotypes (electrophoresis band types). For a microsatellite locus, POLYGENE could infer the possible genotypes and their posterior probabilities based on the allelic phenotype, and then, it estimates the allele frequencies through an iterative algorithm designed by Kalinowski and Taper (2006). Therefore, population genetic analyses were further performed using POLYGENE which take into account both double-reduction and genotypic ambiguities faced by microsatellite studies on polyploids (Huang et al., 2020). Hopea chinensis is a diploid species; thus, it was analyzed under the diploid model with GENODIVE and POLYGENE (Trang & Triest, 2016). Analysis of molecular variance (AMOVA, Excoffier et al., 1992)    to 0.779, respectively. Comparable results were obtained through POLYGENE analyses ( Table 2). The observed and expected heterozygosities ranged from 0 to 0.755 and from 0.255 to 0.757, respectively. The polymorphism information content (PIC) of the 12 loci ranged from 0.222 to 0.719. Deviation from HWE was detected in a large number of loci, and the estimated inbreeding coefficients (F IS ) were apparently different from zero, indicating a nonrandom mating in natural populations of H. hainanensis. The census population size of this species is extremely small (Ly et al., 2018). Small populations are expected to experience severe inbreeding and genetic drift, resulting in departure of HWE. Another possible contribution to departure from HWE is double reduction, which could take place during meiosis in autopolyploid (Huang et al., 2019). The negative value of F IS observed at a few loci (HHA01 and HHA11) suggested an excess of heterozygotes, which might be caused by the stochastic nature of mutation across SSR loci (Putman & Carbone, 2014).

| RE SULTS AND D ISCUSS I ON
An analysis of molecular variance (AMOVA) for H. hainanensis revealed that 80.0% of total genetic variation was partitioned within populations (Table 3). High proportion of variation was generally found to be maintained within populations of dipterocarp species, which is mainly attributed to outcrossing and woody nature of these species (Ghazoul, 2016 These co-dominant markers can be applied to assess the genetic diversity, population structure and mating system of H. hainanensis, which lays foundation for efficient conservation and management of this endangered species. In addition, the successful cross-amplification of seven and three polymorphic microsatellite markers in H. reticulata and H. chinensis, respectively demonstrates the potential application of these markers in population genetic researches of other Hopea species. Council.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

F I G U R E 1
The neighbor joining tree based on the chord genetic distance constructed for all individuals of the three Hopea species.