【作者】 赵静；

【导师】 王起山；

【作者基本信息】 上海交通大学 ， 畜牧学， 2016， 硕士

【摘要】 奶牛乳房炎是奶牛养殖业最常见、最复杂的疾病之一,也是制约奶业发展的重要因素。奶牛乳房炎作为乳腺的一种炎症,是由遗传因素、环境因素、病原体、管理因素等综合因素作用的结果,然而到目前为止,奶牛乳房炎的相关机理,特别是其遗传抗性机制还不是特别清楚。随着统计基因组学、芯片技术和高通量测序技术的发展,使我们从基因组和表达组层面全面研究乳房炎抗性机制成为可能。在基因组层面,虽然已经大量开展数量性状基因座（quantitative trait locus,QTL）定位的研究,但因QTL区间跨度太大,其研究成果仍不能很好地运用到育种实践中,因此对QTL的深入挖掘十分必要。据此,本研究分别从基因组和表达组两个层面系统挖掘奶牛乳房炎相关的候选基因集,并基于真实的奶牛乳房炎表型数据和简化测序数据初步核实验证相关基因集合,具体开展工作如下:一、基因组层面挖掘奶牛乳房炎相关的候选基因集:本研究先从当前QTL数据库中收集整理了关于奶牛乳房炎的QTL共76个,通过数据库的对比,发现QTL内共包含1336个基因。我们进一步从GO、KEGG、IPA三个方面对这些基因进行分析,分别得到42、182、91个基因可能与奶牛乳房炎相关,去冗余后三个分析总共发现284个基因组成奶牛乳房炎候选基因集。对比分析结果显示,其中三个分析结果中有1个共同基因即FABP4基因;GO、KEGG分析结果中有7个共同基因即LIAS、HPSE、COQ3、ENPP3、ENPP1、APOA1和IDO1基因;GO、IPA分析结果中有2个共同基因即BMPR1B和FAM103A1;KEGG、IPA分析结果中有20个共同基因即PPT1、GCNT1、ST3GAL1、ATP6V1F、ATP6V1H、FUT9、FUT4、ATP6V1B1、ATP6V1B2、ATP6V1C1、ST3GAL3、ATP6V0B、PTS、DGKD、AASS、MTHFD1L、LAP3、SCD5、MYC和UCHL1基因。二、表达组层面挖掘奶牛乳房炎相关的候选基因集:本研究先从NCBI的GEO数据库中收集了五套已发表的Affymetrix平台的与奶牛乳房炎的发生发展相关的全基因组表达谱芯片数据,分别为GSE24560、GSE24217、GSE32186、GSE25413、GSE15025,实验设计分别是基于大肠杆菌感染。然后利用基因集富集分析法（Gene Set Enrichment Analysis,GSEA）进行分析,分别得到显著上调和下调通路。最后根据感染时间分别进行整合分析,其中1、6、24 h分别得到共同的显著下调通路为15、30、17个,整合所有的结果发现5个共同通路,包括趋化因子信号通路（Chemokine signaling pathway）、Toll样受体信号通路（Toll-like receptor signaling pathway）、查加斯病通路(Chagas disease（American trypanosomiasis）,弓形虫病通路（Toxoplasmosis）和MAPK信号通路（MAPK signaling pathway）。三、奶牛乳房炎候选基因集初步核实验证:通过全基因组关联分析（Genome-wide association study,GWAS）方法对奶牛乳房炎的遗传机制进行研究,我们首先利用本实验室开发的GGRS测序平台对278头奶牛进行简化基因组测序,共获得108,936个SNP。然后利用ssGWAS方法进行GWAS,发现716个显著SNP位点(显著性阈值为0.05/108936（28）4.59?10^-7),占所有SNP数的0.66%。我们进一步对上述GO分析,KEGG分析和IPA在线分析得到284个基因与GWAS结果进行核实验证。被检测出的108,936个SNP中落在284个基因内的SNP共497个,其中显著阈值以上的SNP有48个,占奶牛候选基因内总SNP数的比例为9.66%。对两个显著SNP比率进行卡方检验,结果显示两者之间存在显著差异,这在一定程度上说明我们发现的奶牛乳房炎候选基因集的准确性较好。更多还原

【Abstract】 Mastitis is one of the most common and complex diseases affecting dairy industry.As a kind of mammary gland inflammation,mastitis affect by genetic,environmental,pathogens and many other factors,however,the mechanism of mastitis in dairy cows is not clear.As the development of the statistic genomics,chip technology and high-throughput sequencing technologies,it make us study mastitis resistance mechanism from the genome and expression level possible.At the genome level,though studies have been carried out and large numbers of quantitative trait loci（quantitative trait loci,QTL）have been found,but the results can’t be well applied to breeding practice because the QTL is too big,so it is necessary to dig deep for QTL.At the expressing level,quite a few researchers have use gene chip to study cow mastitis,but repeatability is not high and each study results have some differences,so we preprocessing microarray data in the same method and implement enrichment of gene analysis,then integrate the results of the analysis.So this study from two aspect of genome and expressing to mining related set of candidate genes,then using real phenotypic and sequencing data to verified the result gene set,details and results are described below:Genome level mining mastitis related set of candidate genes:first,we collected 76 cow mastitis QTLs from QTL database,by comparing the database,we found 1336 genes within QTL,then these genes were analyzed from the GO,KEGG,IPA three aspects,42,182,91 genes were found associated with mastitis respectively,after redundancy a total of 248 genes were identified.There was one common gene in three analysis results that is FABP4 gene,7 common gene in GO and KEGG analysis results that is LIAS,HPSE,COQ3,ENPP3,ENPP1,APOA1 and IDO1 gene,2 common gene in GO and IPA analysis results that is BMPR1B and FAM103A1,20 common gene in KEGG and IPA analysis results that is gene PPT1,GCNT1,ST3GAL1,ATP6V1F,ATP6V1H,FUT9,FUT4,ATP6V1B1,ATP6V1B2,ATP6V1C1,ST3GAL3,ATP6V0B,PTS,DGKD,AASS,MTHFD1L,LAP3,SCD5,MYC and UCHL1.Expression level mining mastitis related set of candidate:we searched NCBI GEO（Gene Expression Omnibus）database and selected five recently published commercial Affymetrix microarray datasets,GSE24560,GSE24217,GSE32186,GSE25413 and GSE15025 which were all about cow mastitis.Then applied gene set enrichment analysis,and up-regulated and down-regulated pathways were identified.There were 15,30,17 common up-regulated pathways in 1,6,24 h result respectively.Five common pathways in all result,including Chemokine signaling pathway,Toll-like receptor signaling pathway,Chagas disease（American trypanosomiasis）,Toxoplasmosis and MAPK signaling pathway.Mastitis candidate gene set verification:we first use GGRS sequencing platform for 278cows simplified genome sequencing,a total of 108936 SNPs were found.Then using GWAS ssGWAS method,found that 716 significant SNPs(threshold is 0.05/108936=4.59*10^-7),which is0.66%of all the SNPs.Among 108936 SNPs there were 497 SNPs within the 248 genes which were identified by GO,KEGG and IPA analysis,and among 497 SNPs there were 48 SNPs significant,and 9.66%of all SNPs within gene set.Chi-square test was carried on two significant SNPs ratio,result showed that they were significant differences,this shows that the accuracy of cow mastitis gene sets that we found is high.更多还原