Microsatellite markers for the critically endangered elm species Ulmus gaussenii ( Ulmaceae )

Qi-Fang Geng, Jie Yang, Jia He, Dan-Bi Wang, En Shi, Wei-Xiang Xu, Nasreen Jeelani, Zhong-Sheng Wang* and Hong Liu School of Life Sciences, Nanjing University, 22 Hankou Road, Nanjing 210093, Jiangsu, China State Key Laboratory of Pharmaceutical Biotechnology, School of Life Sciences, Nanjing University, 22 Hankou Road, Nanjing 210093, Jiangsu, China International Center for Tropical Botany, Department of Earth and Environment, Florida International University, 11200 S. W. 8 Street, Miami, FL 33199, USA College of Forestry, Guangxi University, 100 Daxue East Road, Nanning 530004, Guangxi, China


INTRODUCTION
Ulmus gaussenii W. C. Cheng (Ulmaceae) was named by Prof. Wanjun Zheng in 1955, and its natural range is restricted to the Inebriate Pavilion of the limestone Langya Mountain in Anhui Province, eastern China (Zheng et al., 1999).U. gaussenii (Anhui elm) is now one of the rarest and most endangered elm species, with approximately 26 mature trees known to survive in nearly 10 hectares of wild area (Sun, 2006).U. gaussenii is a medium deciduous tree and listed as a critically endangered species by the International Union for Conservation of Nature (Fu and Jin, 1992).To provide an effective conservation management plan for this critically endangered species, there is an urgent need to evaluate its genetic variation.Microsatellites, or simple sequence repeats (SSRs), are highly polymorphic, codominant markers that have been widely used in studies of population genetics (Zane et al., 2002;Hedrick, 2004;Abdul-Muneer, 2014).

Plant materials and DNA extraction
We collected fresh leaves of 48 individuals of U. gaussenii from one population in Langya Mountain, Chuzhou, Anhui Province, China (32°17.1317′N,118°17.0101′E).Samples were dried in silica gel and stored at room temperature until use.Total genomic DNA was extracted from the dried leaves using a modified cetyltrimethyl ammonium bromide method (Zhou et al., 1999).
Fragments flanked by an SSR at one end were amplified from the AfaI, AluI, EcoRV and HaeIII DNA libraries using one of the compound SSR primers (AC) 6 (TC) 5 , (TC) 6 (AC) 5 , (AC) 6 (AG) 5 or (TC) 6 (TG) 5 and an adaptor primer (5′-CTATAGGGCACGCGTGGT-3′).The amplified smeared fragments were cloned into the Trans pEASY-T1 Cloning vector system in accordance with the manufacturer's instructions (Beijing Transgen Biotech, Beijing, China).Recombinant clones were identified using blue/white screening on LB agar plates containing ampicillin, X-gal and IPTG.Insert-positive clones were amplified using the M13 forward and reverse primers.After electrophoresis on a 1.5% agarose gel, amplified fragments longer than 350 bp were sequenced using the M13 primer (Beijing Genomics Institute, Shenzhen, China).For each fragment containing the (AC) 6 (TC) n , (TC) 6 (AC) n , (AC) 6 (AG) n or (TC) 6 (TG) n compound SSR sequences at one end, a specific primer was designed using Primer3 software (Rozen and Skaletsky, 2000).The specific primer and corresponding compound SSR primer were used as a compound SSR marker.
To characterize the developed compound SSR markers, we performed a polymerase chain reaction (PCR) amplification in a 10-μl reaction mixture containing about 10 ng of template DNA, 1× AmpliTaq Gold 360 Master Mix (Applied Biosystems), 0.4 μl of 360 GC Enhancer (Applied Biosystems), and 0.5 μM of each specific primer and a fluorescent dye-labeled (6-FAM, VIC, NED or PET; Applied Biosystems) compound SSR primer.
PCR was performed with a Veriti Thermal Cycler (Applied Biosystems) under the following conditions: 10 min at 95 °C; 38 cycles of 30 s at 95 °C, 30 s at the annealing temperature of each designed specific primer, and 1 min at 72 °C; and a 7-min extension at 72 °C after the final cycle.The degree of polymorphism of SSR loci was surveyed by automated scanning detection using an ABI 3730 genetic analyzer (Applied Biosystems).Allele sizes were determined using GeneMapper analysis software version 4.0 (Applied Biosystems) with a LIZ-500 DNA size standard (Applied Biosystems).
Data analysis Null allele frequencies were determined with MICRO-CHECKER version 2.2.3 with the Oosterhout algorithm method (Van Oosterhout et al., 2004).The number of alleles per locus (N A ), observed heterozygosity (H O ) and expected heterozygosity (H E ) were calculated using GENALEX 6.5 (Peakall and Smouse, 2006).Deviation from Hardy-Weinberg equilibrium (HWE) and linkage disequilibrium (LD) were tested using Genepop version 4.2 (Raymond and Rousset, 1995;Rousset, 2008).The P values for HWE and LD were corrected for multiple comparisons by applying a sequential Bonferroni correction (Rice, 1989).Inbreeding coefficients (F IS ) were calculated using the FSTAT program version 2.9.3 (Goudet, 1995).The deviation from zero was tested in this population by 1000 per mutation tests with a sequential Bonferroni correction.
To evaluate whether the population had undergone recent bottlenecks, we adopted Wilcoxon's signed-rank test with 1,000 iterations using BOTTLENECK software version 1.2.02 (Piry et al., 1999).We ran BOTTLENECK using three possible mutation models: Infinite Allele Model (IAM), Stepwise Mutation Model (SMM) and Two Phase Model (TPM; 30% IAM and 70% SMM).These models are based on the principle that in a recently bottlenecked population, the observed heterozygosity is higher than the expected equilibrium heterozygosity estimated from the observed number of alleles (Cornuet and Luikart, 1996).

RESULTS AND DISCUSSION
A total of 172 clones with positive inserts were chosen and sequenced.One hundred and thirty-nine sequences were found to contain compound SSR motifs, of which 31 were discarded because the repeat was too close to one end of the insert to design a suitable primer or because the sequence was identical to that of another insert(s).Therefore, only 108 sequences [(AC) 6 (TC) n (3), (TC) 6 (AC) n ( 27), (AC) 6 (AG) n ( 43) and (TC) 6 (TG) n (35)] were suitable for designing specific primers.Using these primer pairs, 49 loci were successfully amplified and yielded a single band of the right size.
For further characterization, the 48 individuals from Langya Mountain were genotyped using the procedure described above.Twelve of the 49 loci were polymorphic.Allelic diversity per locus ranged from two to nine with an average of 4.83 (Table 1).The observed (H O ) and expected (H E ) heterozygosities ranged from 0.021 to 0.750 and 0.225 to 0.744, respectively.Inbreeding coefficient (F IS ) values ranged from -0.157 to 0.960 with an average of 0.340 (Table 1).Significant deviation from LD was detected in two pairs of loci (Ulmg30 and Ulmg42, Ulmg100 and Ulmg102) following Bonferroni correction (P < 0.05).Significant deviations from HWE were found at nine loci after the Bonferroni corrections (P < 0.05, Table 1).The null allele frequencies as predicted from MICRO-CHECKER software ranged from -0.085 to 0.403 per locus (Table 1).The BOTTLENECK test indicated a recent population bottleneck in the U. gaussenii population under IAM, SMM and TPM (Wilcoxon test, P < 0.05), which may account for the significant LD and HWE deviations at these studied loci.
The significant deviation from LD may be due to demographic effects, such as a recent population bottleneck (as noted above) or genetic drift, rather than to a linkage relationship between these loci.The deviation from HWE may be due to several factors such as the presence of null alleles, a high level of inbreeding, a bottleneck effect, restricted population size, a founder effect, the Wahlund effect or genetic drift.In our case, for six loci (Ulmg27, Ulmg30, Ulmg40, Ulmg45, Ulmg61 and Ulmg102), the deviation from HWE as well as their sig-nificant F IS might be attributed to null alleles, or to a high level of inbreeding.While the probable presence of null alleles at these six loci was inferred from MICRO-CHECKER, a high level of inbreeding in U. gaussenii could be the most important contributing factor for the deviation from HWE and high F IS in these six loci (Islam et al., 2014).The restricted population size and paucity of reproductive individuals may result in a high level of inbreeding in the U. gaussenii population.
The remaining three loci (Ulmg42, Ulmg44 and Ulmg65) exhibiting deviation from HWE were not identified as possibly having null alleles and, in these cases, other factors should be considered to explain the deviation.Among the causes of HWE enumerated above, we consider inbreeding to be unlikely, as the F IS values for these three loci were small.One of the remaining factors such as a bottleneck effect, restricted population size, a founder effect, the Wahlund effect or genetic drift is more likely to cause the observed HWE deviation.
The 12 polymorphic microsatellite markers described here will be available for population genetic studies of U. gaussenii, which in turn will contribute to the conservation and protection of this endangered species.

Table 1 .
(Van Oosterhout et al., 2004)tellite loci isolated from Ulmus gaussenii IS : inbreeding coefficient, with F IS in bold indicating significant deviation from zero with a sequential Bonferroni correction; *: significant deviation from Hardy-Weinberg equilibrium following Bonferroni correction, with P < 0.05; r: null allele frequencies calculated with the Oosterhout algorithm(Van Oosterhout et al., 2004), with presence (yes) and absence (no) indicated below.
a : annealing temperature of the primer pair; H o : observed heterozygosity; H E : expected heterozygosity; F