A study on efficient screening of the primers for selecting polymorphic SSR markers based on the re-sequencing data in Pyrus

Abstract: 【Objective】SSR markers are widely used for cultivar identification and QTL mapping. Theprimers were firstly synthesized and then screened by amplified fragment length in PCR in differentsamples in the past. The length of reads in the second generation sequencing of Illumina hiseq has been150 bp resently. The repeat length of the core sequence of SSR is usually about 50 bp. Some reads areable to completely cover repeats, which make it possible to evaluate SSR polymorphism based on thereads. With the popularity and low cost of the next generation sequencing, a rapid method for screening SSR marker polymorphism among samples based on the re-sequencing data was developed in this study.【Methods】Six Pyrus accessions were re-sequenced including two cultivars of P. pyrifolia:'Zaosheng-xinshui''Qiushui'; three wild species: P. calleryana, P. pashia, P. betulaefolia and one cultivar of P.communis'early red comice' (BIG Data Center, http://bigd.big.ac.cn/gsa: CRR019693-CRR019698) . Atotal of 446 pairs of SSR primers reported in previous studies and 16 newly developed pairs of primerswere evaluated for the test of the polymorphism in Pyrus species. The left and right flanking regions (120 bp) and corresponding SSR sequences of 462 loci were isolated as a database. Firstly, the re-se-quencing data of each sample was mapped by Magicblast to the database mentioned above, and themapped reads to each SSR locus were isolated. Secondly, two 11 bp short sequences around the SSRsite were matched to the mapped reads by an in-house Bioperl script, and the length between two shortsequences was calculated. The sequence lengtsh of all SSR loci in each sample were obtained. Thirdly, the fluorescence PCR were performed to validate the prediction resulst. An economic method for fluo-rescent labeling of PCR fragments was adopted to test the polymorphism. A tail (M13 universal se-quence, TGTAAAACGACGGCCAGT) was added to the 5ʹ end of each of the ten reverse primers. Thetail primers were labeled with the following four dyes: FAM, HEX, ROX, and TRAMA. PCR productswere electrophoresed on 3% agarose gels to check their quality. Two independent PCRs were performedto confirm stable amplification. The PCR products with different fluorescent tails were diluted, thenmixed with the internal size standard LIZ500, and loaded onto an ABI 3700 XL Genetic Analyzer.【Results】The average number of reads mapped to individual SSR loci was more than 16. More than 220 loci were detected in each Pyrus accession. The number of detected loci in P. betulaefolia and P. communis'early red comice'was lower than those of the other accessions, which might be related to their dis-tant relationship with'Dangshansuli' (the reference Pyrus genome in the mapping process) .The poly-morphic fragments could be found in 380 of the 446 developed pairs of primers among different Pyrusaccessions. The rest of 66 pairs of primers could not generate available mapped reads in six of the Py-rus accessions, therefore the length of the repeat sequence in these loci could not be obtained. In someloci, such as NAUpy52 s (316 bp) , NAUpy19 m (342 bp) , NAUpy35 e (374 bp) , and NAUpy31 i (386 bp) , the sequence length was more than 150 bp, making it difficult to find an available matched reads.Some loci with three or more bands indicated that they were multiple copy sites and should be deleted.The loci with one and two bands were reserved. Twenty-nine SSR loci in the selected region of pearchromosome 3 were accurately analyzed in'Zaoshengxinshui'and'Qiushui', of which ten locishowed polymorphism in the re-sequencing data. Nine pairs of primers showed polymorphism in thefluorescence PCR, suggesting that the method used in this study was reliable. Only NAU57 d in all prim-ers failed to show polymorphism. The results of fluorescence PCR in SAAS301, SAAS303, SAAS306, SAAS307, and SAAS309 were consistent with the prediction result of the re-sequencing data.NAUpy64 d and SAAS 313 had two bands in'Zaoshengxinshui'in the PCR amplification, but oneband was detected in the prediction by re-sequencing data, suggesting that the other band might not bedetected by re-sequencing data. This could be improved by increasing the sequencing depth in the fu-ture. The primer of SAAS308 had two bands in'Zaoshengxinshui'in the prediction, but only one bandwas shown by fluorescent PCR, which indicated that the other band might not be amplified effectivelythrough PCR. More efficient PCR primers should be designed. The difference of band length in PCRand predicted length in re-sequencing data of SAAS311 primer is 1-2 bp, which might be caused by theerror in fluorescence detection. Although the results of some primers in the re-sequencing data were notentirely consistent with the results of fluorescence PCR, the pre-screened primers showed good poly-morphism in'Zaoshengxinshui'and'Qiushui', thus this method is very suitable for pre-screening, which would significantly improve the efficiency of primer screening.【Conclusions】In this study, alarge number of pairs of SSR primers could be screened at one time, and the polymorphism of differentpairs of primers could be obtained before the experiment of traditional PCR, which would improve theefficiency of primer screening. This method not only could be applied to pear re-sequencing data, but al-so to similar studies on other plants and animals.