Analysis Of The Genetic Diversity Of Scrophularia Scrophularis Based On Transcriptome SSR And SNP Markers And The Associated Sites Of Effective Components

Objective:The whole plant of Picria felterrae Lour.(Herba Picriae)has long been used in Guangxi,China,belonging in Zhuang medicine and listed as one of the ten traditional Chinese medicinal materials receiving major support from the Guangxi government during its 13th five-year period.Its wild resources are becoming fewer and fewer due to deterioration of its growth environment,and cultivated plants constitute the main source of market supply of that drug.Confounded germplasms resulting from non-breeding of cultivated varieties,together with degradation induced by long-term cultivation,has led to poor quality including the reduction in the content of bioactive ingredients,thus it is urgent to breed a new variety of Picria felterrae with high quality.With the rapid development of biotechnology and informatics,the theory and method of genomics have penetrated into many levels and aspects of germplasm resources research,enabling objective and accurate detection of differences between germplasms at genomic level.In this study,we sampled morphologically distinct germplasm resources of Picria felterrae from different areas and employed them to conduct genetic diversity and population structure analyses on a systematic basis using the composition and content of picfeltarraenins,the main bioactive ingredients,as the metric of good germplasms,based on transcriptomic SSR and SNP molecular markers;genes associated with picfeltarraenins were also investigated.We aimed to provide scientific basis and technical support for screening and identification of good germplasms and breed improvement of Picria felterrae.Methods:(1)Thirteen wild germplasms were sampled from Guangxi and Yunnan Provinces,and 50 germplasms were cultivated in Longzhou and Wuzhou,Guangxi.After multiplication in Nanning,Guangxi,samples were used for determination of picfeltarraenins IA and IB by reference to Standard Operating Procedures for High Performance Liquid Chromatography(HPLC)in "HPLC methods"(Appendix VID)of Pharmacopoeia of the People’s Republic of China(2015 Edition),and comparison of contents of picfeltarraenins IA and IB were done for samples from different areas and of different strains to preliminarily screen dominant populations and good germplasms with high levels of picfeltarraenins.(2)Sixty-three fresh leaf samples of Picria felterrae collected from Yunnan and Guangxi were subject to parallel sample transcriptome sequencing using the high-throughput transcriptome sequencing technology comprising RNA sample testing,library construction,and library quality control.Transcriptome genetic data of these samples were obtained after splicing and assembly.Unigene sequences were compared with NR,Swissprot,GO,COG,KOG,Pfam,and KEGG databases using the BLAST software to obtain Unigene annotation information.Encoded nucleotide and amino acid sequences were analyzed and predicted by database comparison and using the Getorf software.(3)The resulting transcriptome genetic information was compared with the reference genome sequences using STAR,a comparison software for transcriptome sequencing,and GATK was used to explore Single Nucleotide Polymorphism(SNP)loci.Based on SNP molecular markers,samples from 13 populations and 50 cultivated samples were subject to genetic structure analysis through phylogenetic,population structure,principal component,and genetic relation analyses to learn genetic and evolutionary relationships between samples.Moreover,the correlation between the bioactive ingredients and transcriptomic SNP and GEM loci were analyzed to screen the bioactive ingredients associated loci,and the function of candidate genes was predicted by comparing with a known database.(4)Simple sequence repeat(SSR)molecular markers were explored using MISA(a software for SSR identification)based on transcriptome sequences.EST-SSR primers were developed based on conserved sequences at both ends of SSR markers.4 samples were chosen randomly and used to test the polymorphism of the primers,and 20 pairs of polymorphic primers were selected and used for PCR amplification of test samples.The genetic structure of 18 populations and 69 individuals was analyzed according to the amplification results,and SSR molecular markers associated with the content of picfeltarraenins were screened using GLM simple linear regression and multiple stepwise regression in SPSS.Results:(1)Significant differences existed in the content of picfeltarraenin IA among 13 wild populations from Guangxi and Yunnan(p<0.05);very significant differences were found in picfeltarraenin IB and total glycosides between both provinces(p<0.01).By and large,populations from Yunnan were better than those from Guangxi,populations from Jinghong Yunnan ranking first in terms of dominance,followed by those from Menglang,Langcang,Yunnan.For the 50 cultivated samples,the levels of picfeltarraenin IA and total glycosides were not significantly different between Longzhou and Wuzhou,Guangxi(p>0.05);samples from Longzhou had significantly higher picfeltarraenin IB content than those from Wuzhou(p<0.05);samples from both cities shared almost the same germplasm level according to the content of total glycosides.Marked differences were present in picfeltarraenin contents among strains,individuals of strains HZP04,ZGBO,WZHZP21,ZGB08,and WZHJX03 being better as a whole.(2)323.55Gb clean data were obtained after 63 parallel samples were subject to transcriptome sequencing and sequencing quality was controlled,the Q30 base ratio higher than 86.28%.The preliminary Picria felterrae transcriptome database was created following splicing and assembly of sequencing data,519,730 Transcript sequences and 203,606 Unigene sequences were obtained,N50 of Transcript and Unigene being 3,365 and 751,respectively.Throuth comparison between Picria felterrae Unigene sequences and NR,Swissprot,GO,COG,KOG,Pfam,and KEGG databases,131183 Unigenes were annotated in 203606 transcripts,with an annotation ratio of 64.42%.After comparison of Unigenes to the known protein database,Unigenes with gene sequences same as or similar to know protein sequences were obtained.In total 153100 CDS nucleotide sequences and CDS amino acid sequences were obtained.For Unigenes not matched with known sequences,the Getorf software platform was used and 65536 CDS nucleotide sequences and 65536 CDS amino acid sequences were predicted.(3)SNP molecular markers of the 63 Picria felterrae parallel samples spanned from 57355 to 107932,where the quantity of SNP markers of wild samples T01-T13 was 61430-81517 and the number of SNP markers of cultivated individuals T14-T63 were 57355-107932,the span of cultivated varieties greater than that of wild type.The number of homozygotic SNP markers was 41852-74564,whereas that of heterozygous SNP markers was 4094-47084.According to the genetic structure analysis,13 populations could be divided into 2 groups,namely 8 populations from Yunnan listed as one group and 2 ones from Yunnan and 3 ones from Guangxi as the other group;the 50 individual samples could be divided into two groups,namely 4 individuals from Wuzhou listed as one group and the rest as the other group.For the correlation between SNP and GEM loci and the content of picfeltarraenins,it was found based on SNP analysis that there were 5 picfeltarraenin IA related loci and 4 picfeltarraenin IB related loci when the threshold p<1.00E-06;based on GEM analysis,there were 184 picfeltarraenin IA related loci,2421 picfeltarraenin IB related loci,and 45 loci associated with both ingredients when the threshold p<1.00E-06;based on both SNP and GEM analysis,there were 5 picfeltarraenin IA related loci and 6 picfeltarraenin IB related loci when the threshold p<1.00E-03;functional genes at the loci mainly involved in signal transcription,methylation,and gene modification.(4)The evaluation of Unigene sequences with a total base number of 47,941,629bp yielded 13768 SSR molecular markers,and 6 different types of SSR were identified.48 pairs of polymorphic primers were obtained from 100 pairs of primers which were randomly selected from primers developed based on the sequences at both ends.71 allelic genes were amplified from 18 population samples using 20 pairs of polymorphic primers randomly selected(3.55 alleles per pair on average).For different primers,the P value varied from 0%to 40.7%;the PIC varied from 0 to 0.7941,and the I varied from 0 to 1.8143.The Obs_Het varied from 0 to 0.4423;the Exp_He varied from 0 to 0.8269;the Fis value varied from 0.0953 to 0.6639;the Fit value for different groups varied from 0.0626 to 0.8587;the genetic differentiation coefficient(Fst)varied from 0 to 0.6866.The gene flow(Nm)varied from 0.1144 to 0.7594.For different populations,the Nei varied from 0 to 0.4016;the I varied from 0 to 0.6209,the population similarity coefficient varied from 0.3814 to 0.9686,and the genetic distance varied from 0.0319 to 0.9638.18 populations were divided into 4 groups at a genetic distance of 0.3213,samples from Yunnan falling into 3 groups and 3 populations from Guangxi"classified as one group.A total of 76 allelic genes were amplified from 69 individuals using 20 pairs of SSR polymorphic primers,3.8 allels per locus on average.The allele polymorphism rate varied from 0 and 59%.For different loci,the polymorphism information content(PIC)varied from 0 to 0.6211;the Shannon polymorphism information index varied from 0 to 1.2401,the Nei’s gene diversity index(Nei)varied from 0 to 0.6823;the mean observed heterozygosity was 0.3824;the mean genetic differentiation coefficient(Fst)was 0.3659;the mean gene flow(Nm)was 0.4332.The inter-sample similarity coefficient varied from 0.5856 to 0.9506,and genetic distance varied from 0.0506 to 0.5325;69 individuals fell into 13 groups at a genetic distance of 0.24.The simple linear regression and multiple stepwise regression showed that there were 5 loci associated with picfeltarraenins IA and IB,respectively,among which only one locus was associated with both ingredients.Conclusion:The appraisal of 13 wild population samples collected from Yunnan and Guangxi based on the quality evaluation standard set forth in Pharmacopoeia of the People’s Republic of China demonstrates that wild germplasm resources from Yunnan are obviously better than those from Guangxi and have a bigger potential for screening good Picria felterrae germplasm resources.Although Guangxi is the main producing area of Herba Picriae,cultivated Picria felterrae individuals are of varying quality,those from Wuzhou may even show degeneration.We created a preliminary public Picria felterrae transcriptome database using the transcriptome sequencing technology,which fills the gap of genetic information of Picria felterra.We developed Picria felterrae SSR and SNP molecular markers using transcriptome data and analyzed the genetic evolutionary relation of Picria felterrae at different levels.Different degrees of differentiation were found in different habitats,and wild samples from Guangxi and Yunnan display marked differentiation.The analysis results are basically consistent with the geographical distribution difference of Picria felterrae test germplasms,but there is still interregional genetic cross.This provides a foundation for research into Picria felterrae dominant populations.As a whole,cultivated varieties from Guangxi show a lower level of genetic differentiation among themselves,but there is still distinct differentiation between individuals,which can be used for screening of good strains.Adopting the content of bioactive ingredients in Picria felterrae as the target trait and using different molecular markers and association strategies,we singled out a large number of molecular marker loci associated with picfeltarraenins;the function of genes on these loci mainly involves signal transcription,methylation,and gene modification.Our results may provide the most direct genetic information for seed breeding of Picria felterrae to facilitate the selection of good germplasms of this species.