【作者】 肖潇；

【导师】 宋亚军； 周冬生；

【作者基本信息】 中国人民解放军军事医学科学院 ， 军事预防医学， 2011， 硕士

【摘要】 副溶血性弧菌是革兰氏阴性的嗜盐弧菌,广泛分布于港口、海水沿岸、海河交界处等,通过污染的海产品引起人类急性胃肠炎及旅行者性腹泻。其主要毒力因子包括耐热性直接溶血素(TDH)、与TDH相关溶血素(TRH)、不耐热溶血素(TLH)和Ⅲ型分泌系统(T3SS)。这些毒力因子大多数具有细胞毒性和(或)肠毒性。根据菌体(O)抗原和荚膜(K)抗原可将副溶血性弧菌至少分为13个O血清群和71个K血清型。在1995年以前,与副溶血性弧菌相关的胃肠炎均由不同种血清型引起。然而,自从1996年2月在印度加尔各答首次爆发了由副溶血性弧菌引发的大规模食物中毒,这种血清型为O3:K6菌株(KP+,tdh+, trh-)迅速在世界各地广泛流行,尤其以沿海国家和地区为主。目前将这些1996年以后分离的血清型以O3:K6为主,包括O4:K68、O1:K25、O1:KUT,经脉冲场凝胶电泳(PFGE)、随机引物PCR(AP-PCR)及多位点序列分析(MLST)鉴定具有非常相似的指纹图谱(FPs)或序列型(STs)统称为―流行群‖。高纯度基因组DNA的提取脂多糖(LPS)是细菌外膜重要的组成部分,除此以外,某些革兰氏阴性菌还产生其它两种多糖:菌表多糖与荚膜多糖。例如,副溶血性弧菌通过表达大量的菌表多糖形成生物膜,使其有利于在环境中生存和传播。无论是哪种形式的多糖都会在一定程度上干扰基因组DNA的提取,同时也影响了DNA的特性。在本研究中,以黏液型细菌副溶血性弧菌RIMD2210633与肺炎克雷伯氏菌NTUH-K2044作为研究模型,以大肠埃希氏菌JM109作为对照,建立两种适用于提取黏性革兰氏阴性细菌基因组DNA的新方法。三种细菌基因组DNA的提取使用了三种方法,每种方法重复三次。方法Ⅰ为经典的饱和酚-氯仿抽提方法,方法Ⅱ与方法Ⅲ都在方法Ⅰ操作基础上进行了改进,分别增加了乙二醇甲醚或水饱和乙醚的步骤以去除多糖。三种方法得到的DNA产率各不相同,以方法Ⅰ最高,方法Ⅱ与方法Ⅲ由于增加了去除多糖的步骤明显降低了DNA产率。将各种方法提取的DNA经琼脂糖凝胶电泳,均有明亮的条带且无RNA污染与DNA降解。在本研究中,用A260/A280与A260/A230两个比值来评估三种方法提取的DNA纯度,三种方法在A260/A280均处于1.8～2.0,所以表明三种方法提取的DNA均无蛋白的污染。对于A260/A230≥2.0这个指标,只有方法II与方法III适合于所有实验菌株,而方法Ⅰ只适用于非黏性菌株JM109,对于副溶血性弧菌RIMD2210633与肺炎克雷伯氏菌NTUH-K2044其最高值也仅有1.4左右,因此存在严重的多糖污染。基于上述结果,本研究建立的方法II与方法III适用于黏性革兰氏阴性细菌基因组DNA的提取,这两种方法重复性高且花费少。当然,这两种方法也可以应用于革兰氏阳性细菌,只要在细胞裂解步骤(如对金黄色葡萄球菌加入溶葡萄球菌素)中加入溶菌酶即可。由于这两种方法所需成本较低,因此可大规模应用于黏性细菌基因组DNA的提取。基于差异分布的大片段基因簇(LVPC)的副溶血性弧菌基因分型本研究共使用251株副溶血性弧菌(S001-S251),其中包括193株临床分离株,58株环境分离株。基于之前的DNA芯片比较基因组学(M-GCH)研究,我们共筛选得到18个LVPC。用筛选得到的18个LVPC位点对251株副溶血弧菌进行PCR扩增,结果分别用“1”或“0”表示,其中“1”表示在该LVPC位点PCR扩增结果为阳性,“0”表示该LVPC位点PCR扩增结果为阴性。18个LVPC在一株菌中的存在或缺失可形成一个LVPC图谱,对所有菌株的LVPC图谱进行统计共有48个独特的LVPC图谱,即251株菌可分为48种LVPC型。根据获得的二元的LVPC数据通过非加权组平均法(UPGMA)对251株副溶血性弧菌进行聚类分析,得到最小扩展树(MS tree)。结合MS树与毒力分子标识的分布,将251株菌分成6个群,LVPC-C1～LVPC-C6。将基于LVPC与M-CGH数据构建的MS树进行比较,发现两者之间的拓扑结构非常相似。另外, LVPC所划分的5个群(LVPC-C1～LVPC-C5)与M-CGH生成的5个复合群(C1～C5)相对应。而对于LVPC-C6来说,该群只含有一株菌S093,经流行群特异性分子标识鉴定非流行群(GS-PCR+,tdh-;PGS+)。S093在基于LVPC所构建的MS树中为一个单独的单元型,存在于5个复合群之外,处于LVPC-C2与LVPC-C3之间,且与LVPC-C2遗传距离较近。其中,LVPC-C2所有菌株(共14株)均为1996年之前分离的旧O3:K6群。LVPC-C3中的所有菌株(63株)经流行群特异性分子标识鉴定均为流行群(GS-PCR+, tdh+, trh- ;PGS+)。由于S093处于LVPC-C2与LVPC-C3之间,即“流行群”与“旧O3:K6克隆群”之间,是一种过渡型可定义为“O3:K6的中间型”。这也再次印证了“旧O3:K6克隆群”、“O3:K6的中间型”与“流行群”之间的系统发育关系。总体来说,基于LVPC的分型方法相对于其它方法操作简单、费用低廉,只需进行普通的PCR扩增及琼脂糖凝胶电泳即可完成,可在短时间内对大量菌株进行分型。当然该方法的分辨率有限,因此可通过另外的分型方法弥补。基于可变数目串联重复序列(VNTR)的副溶血性弧菌基因分型本研究共选择了10个可变数目串联重复序列(VNTR)位点用于对251株副溶血性弧菌进行多态性分析,其中9个VNTR位点在251株实验菌株中呈现良好的多态性,1个位点由于在184株菌扩增结果为阴性,故舍去。在这些位点中位点VP2-7、VPTR1及VPTR6均呈现出较高的等位基因多态性,而位点VP1-10,VP1-11呈现出非常低的等位基因多态性。基于LVPC聚类分析结果,不同LVPC群的菌株由不同颜色标记,将这些有颜色标记的菌株的VNTR数据导入生物学软件BioNumerics 5.01,利用UPGMA方法对251株菌构建了MS树。将251株菌最终分成了222个基因型,基因型的差异主要是重复序列拷贝数不同造成的。只有来自LVPC-C2中大部分菌株与LVPC-C3在基于VNTR数据构建的MS树聚类到一起,而LVPC-C1、LVPC-C4及LVPC-C5在该MS树呈散在分布。上述结果说明基于VNTR的分型方法具有极高的分辨率,几乎可以将每株菌分开(251株菌分成222个VNTR型),包括那些同源性较高的流行群,这种区分能力是别的分型方法很难达到的。然而,大多数菌株在基于VNTR数据构建的MS树中呈散在分布,因而很难判断菌株之间的亲缘关系,其主要是由于VNTR的重复数在菌株中高度可变以至于无法对大量菌株进行整体的系统发育关系分析,但是由于该方法本身的优点如高分辨率与高通量,可用于对菌株的快速鉴定及亲缘关系较近菌株的区分。二级基因分型法本研究目的是将两种分型方法相结合以梯阶式分析将亲缘关系远近的菌株区分开,简单的说即用一种分型方法将菌株大致分为几个群,随后采用分辨率较高的分型方法对同一群中的菌株分为不同的型。这样不仅提高方法的分辨能力,而且能将一种分型方法的不足由另外一种分型方法弥补。本研究建立了一种新的基因分型方法用于鉴定菌株以及判断菌株之间的亲缘关系。首先通过操作简便快速的基于18个LVPC位点的分型方法和5个毒力分子标识(GS-PCR, tdh, T3SS2α, trh和T3SS2β)先将251株菌按其致病性进行初步分类,得到6个LVPC群,随后再利用基于VNTR的二级分型方法将每个LVPC群进行更精细的区分。将6个LVPC群的菌株基于VNTR数据分别构建了MS树,LVPC-C1分成了34个等位基因型,LVPC-C2为13,LVPC-C3为55,LVPC-C4为70,LVPC-C5为49。总之,基于LVPC与VNTR的二级基因分型方法对于区分副溶血性弧菌是一种非常实用的方法。该方法得到的分型结果易于判读、具有很好的重复性以及在一定程度上实现了高通量,可对大量全球收集的菌株建立遗传指纹图谱样数据库,从而有利于副溶血性弧菌的鉴定、分型、溯源及风险评估。更多还原

【Abstract】 Vibrio parahaemolyticus is a Gram-negative halophilic bacterium widely found in estuarine and marine environments, which causes gastroenteritis and traveler‘s diarrhea through consumption contaminated seafoods. The main virulence factors of V. parahaemolyticus includes thermostable direct haemolysin (TDH), TDH-related toxin (TRH), thermolabile haemolysin (TLH), two type III secretion systems, which play distinct roles in cytotoxicity and enterotoxicity, respectively.Serotyping of V. parahaemolyticus isolates has identified more than 13 O antigen groups and 71 K antigen types. Before 1995, V. parahaemolyticus associated gastroenteritis was caused by many different serogroups. Since 1996, so-called“pandemic clones,”the majority of which belong to serotype O3:K6, have caused worldwide outbreaks of gastroenteritis. Previous the pulsed field gel electrophoresis (PFGE), arbitrarily primed–PCR (AP-PCR) and multilocus sequence typing (MLST) studies revealed that the new O3:K6 and its derivates (O4:K68,O1:K25 and O1:KUT ) gave very similar fingerprint patterns or sequence types (STs), suggesting that they constitute a clonal complex. These strains are collectively called the ’pandemic group’ that is thought to be responsible for the pandemic outbreaks. Extraction of high-purity genomic DNALipopolysaccharide is the major component of the outer membrane in Gram-negative bacteria, and some of these bacteria produce two other kinds of polysaccharide as well, i.e., exopolysaccharide and capsular polysaccharide. V. parahaemolyticus expresses abundant exopolysaccharides, which is essential for its environmental survival and transmission by providing access to nutrients and protection from predators and adverse natural environmental conditions. Polysaccharides are problematic for DNA extraction from mucoid bacteria. In the present study, mucoid V. parahaemolyticus RIMD2210633 and K. pneumonia NTUH-K2044 were used as the model organisms, and Escherichia coli JM109 as control. Two methods were established for extracting genomic DNA from mucoid Gram-negative bacteria.In this study, genomic DNA was extracted with the three methods. Method I is a classical phenol/chloroform extraction approach. Methods II or III was modified from the method I by introducing methoxyethanol or ethyl ether for removal of polysaccharides, respectively. Experiments were performed in triplicates. The yields of DNA obtained from the three methods were different and method I got the highest yields. Polysaccharide removal procedures in methods II and III significantly reduced the yields of DNA. Electrophoresis assay of DNA samples indicated that DNA bands were clear and clean, and no signs of RNA smears and of degraded DNA were observed. In this study, the OD260/OD280 and OD260/OD230 ratios were calculated for each DNA sample to evaluate the three DNA extraction methods. All the three methods gave the OD260/OD280 values between 1.8 and 2.0 for all the three bacteria, indicating the relevant DNA samples were free of protein contamination. Both methods II and III gave the OD260/OD230 values≥2.0 for all the three strains, while the method I gave the values≥2.0 for only non-mucoid JM109. For the DNA samples isolated from RIMD2210633 and NTUH-K2044 using the method I, the OD260/OD230 values were less than 1.4 indicating severe of polysaccharides in them.In the present work, we established two reproducible and cost-effective methods for routine genomic DNA extractions from mucoid Gram-negative bacteria. And the two methods could also be applied to mucoid Gram-positive bacteria if lysozyme (e.g. lysostaphin specific for Staphylococcus) was used in the cell lysis step. Large variably-presented gene cluster(LVPC) -based genotypingTwo hundred and fifty one V. parahaemolyticus isolates were tested in this study, including193 clinical isolates and 58 non-clinical ones. From our previous M-GCH study, we indentified 18 LVPC. Single gene was chosen from each LVPC, and PCR assays were set up to screen for the presence (1) or absence (0) of each LVPC within the 251 strains tested. A total of 48 different LVPC profiles were identified.According to the binary LVPC data, we constructed an unweighted pair-group method with arithmetic means(UPGMA)-based Minimum Spanning Tree (MS tree) that depicted the phylogentic structure of the 251 strains. Based on the MS tree structure in conjunction with the frequency profiles of virulence markers, these 251 strains could be assigned into six complexes, LVPC-C1 to LVPC-C6. Notably, the MS trees of LVPC and M-CGH shared a similar topologic structure. In addition, LVPC-C1 to LVPC-C5 generally corresponded to the five M-CGH complexes (C1 to C5), respectively. LVPC-C6 contained only a single strain S093 (GS-PCR+, tdh-, T3SS2α-, trh- and T3SS2β-), which was not assigned to any complex in the M-CGH study.All the 14 LVPC-C2 strains were the pre-1996 ?old‘O3:K6 ones (GS-PCR-, tdh-, T3SS2α-, trh+ and T3SS2β+); and12 of them were characterized previously by MLST to constitute a clonal complex namely the old-O3:K6 clone. All the 63 pandemic strains (GS-PCR+, tdh+, T3SS2α+, trh- and T3SS2β-) tested were assigned into the LVPC-C3 complex . S093 was most likely a phylogenetic intermediate between the pandemic and old-O3:K6 clones. We proposed it as the ?intermediate-O3:K6 clade‘. In addition, the genetic relationship in the LVPC MS tree herein supported the proposed phylogenetic relationship between the old-O3:K6, pandemic, and intermediate-O3:K6 groups. Variable number of tandem repeat(VNTR)-based genetypingA total of ten VNTR loci were tested in our strain collection. VPTR7 gave negative PCR amplification for 184 of the 251 strains, and was thus discarded from further VNTR assay.The rest nine VNTR loci were polymorphic among the 251 strains. VP2-07 and VP1-10 gave the highest and lowest allelic diversity, respectively.The categorical VNTR dataset of the 251 strains was directly analyzed to build a MS tree. The 251 strains were subdivided into 222 distinct genotypes. Only the strains from LVPC-C2 or LVPC-C3 were grouped together in the VNTR MS tree of the 251 strains, while those from LVPC-C1, LVPC-C4, and LVPC-C5 showed scattered distribution in the MS tree.Overall, the VNTR-based genotyping method had a much higher resolution than the LVPC-based one in discriminating the V. parahaemolyticus strains. Although certain VNTR loci seem to mutate too rapidly to be useful in determining global phylogenetic relationships, they are useful for strain identification and may identify rapidly evolving polymorphisms that are useful for discriminating very closely related strains, such as V. parahaemolyticus serotype O3:K6 strains. Two-step genotype approachAn optimal typing method should provide the ability to distinguish among both distantly and closely related strains. Thus, multiple methods are often required to adequately discriminate among closely related isolates. In fact, each of the methods utilizes different genetic markers to differentiate different strains, and each method has its weaknesses that can be complemented by another. Given the disadvantages and advantages of both LVPC- and VNTR-based genotyping methods, we proposed the combined used of LVPC and VNTR. For this purpose, PCR-screening of 18 LVPCs and five virulence markers (GS-PCR, tdh, T3SS2α, trh and T3SS2β) should be carried out for the test strains, which will primarily group the strains into various complexes with different pathogenicity characteristics. The strains from each LVPC-based complex would be further analyzed with the nine VNTRs loci, as the VNTR-based genotyping worked well to give a much more detailed discrimination of the strains.The VNTR-based method was applied to 251 V. parahaemolyticus isolates that were genotyped previously by LVPC-based method. The isolates of V. parahaemolyticus belonging to the same LVPC complex were distinguishable by the size variations in the 9 informative VNTRs loci. The LVPC-C1 gave 34 VNTR allelic profiles, the LVPC-C2 gave 13, the LVPC-C3 gave 55, the LVPC-C4 gave 70, and the LVPC-C5 gave 49 profiles respectivly.This combined scheme could be used to construct a genetic fingerprint-like database based on a large collection of global strains, which would be very useful for identification, genotyping, source-tracking, and risk evaluation of V. parahaemolyticus.更多还原