【作者】 杨慧盈；

【导师】 杨瑞馥； 杜宗敏；

【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， 2009， 博士

【摘要】 鼠疫是由鼠疫耶尔森氏菌引发的的一种烈性传染病,历史上曾经发生三次人间鼠疫的世界大流行。鼠疫菌的致病机制依赖于它和宿主之间的相互作用而实现。揭示鼠疫菌与人之间的蛋白相互作用是深入了解其致病机制的重要基础,同时为研究有效的预防和控制鼠疫的发生和传播的手段提供线索。本研究包括三部分:首先,采用高通量的酵母双杂交技术(Y2H),以推测的鼠疫菌毒力相关蛋白为诱饵筛选人脾脏cDNA文库,旨在获得鼠疫菌与人之间蛋白相互作用的初步网络,并且获得一批重要的蛋白相互作用(PPI),为深入揭示鼠疫菌的致病机制提供研究靶标;其次,应用Y2H阵列筛选技术筛选鼠疫菌III型分泌系统(T3SS)内部蛋白质间的相互作用,为进一步阐明T3SS的结构和功能奠定基础;第三,为进一步了解负调控因子LcrG在鼠疫菌T3SS基因表达中的作用,应用全基因组芯片进行了鼠疫菌野生株和lcrG突变株的比较转录谱学研究。第一部分:鼠疫菌毒力相关蛋白与人蛋白质相互作用的研究我们根据鼠疫菌和假结核菌的比较基因组学研究结果和本实验室开展的鼠疫菌在不同刺激条件下的表达谱、血清抗体谱、蛋白芯片筛选的试验结果,参考文献中鼠疫菌相关研究的最新进展最终确定可能与人蛋白相互作用的鼠疫菌蛋白152个(占鼠疫菌整个基因组的3.75%)。利用高通量Y2H技术筛选鼠疫菌与人之间的PPI。本研究采用基于Gateway重组技术的ProQuest? Two-Hybrid System,便于高通量平行操作。通过Gateway重组技术,成功构建了154个诱饵载体,对应于152个鼠疫菌蛋白的ORFs。其中存在自激活的诱饵有3个,分别是:YPCD1.26c,YPCD1.31c和YPMT1.61c。通过顺序转化法进行规模化Y2H筛选,共对151个无自激活活性的诱饵至少筛选一次人脾脏cDNA文库。筛选平板上长出的候选克隆分别通过生长实验鉴定HIS和ADE报告基因表型,X-Gal分析鉴定lacZ报告基因表型。对得到的阳性克隆进行测序,实验共得到成功测序的克隆1087个。用NCBI网站的BLAST功能进行数据库检索,得到符合ORFs读框的克隆833个。对应于非冗余猎物分子185个,鼠疫菌诱饵蛋白92个,及由此构成的359个相互作用对,组成我们的核心数据集。实验未筛选到文献报道的已知相互作用。在这359对相互作用对中,60对出现频率为两次,91对出现频率在三次或三次以上,分别占总数的16.7%和25.5%,说明我们的筛选结果可信度较高。随后采用实验验证和生物信息学分析的方法进一步评价相互作用的可信度。实验部分,进行了酵母回转实验和GST pull down实验。我们对读框正确的359对相互作用进行了回转实验,经过表型鉴定,208对相互作用表现为阳性,阳性率为57.9%。得到对应于67个诱饵、109个猎物蛋白的208对相互作用。挑选酵母回转实验阳性的鼠疫菌和人蛋白相互对21个,进行GST pull down实验验证,约占全部回转阳性的相互作用对的10%。21对相互作用中涉及6个猎物蛋白,由于其中1个猎物蛋白不能释放到上清中,导致7对相互作用暂时无法验证。在剩余的14对相互作用中,13对验证为阳性,阳性率为92.8%,证明通过Y2H筛选及回转验证实验得到的208对鼠疫菌和人蛋白之间的相互作用数据可靠性较高。利用双杂交筛选获得数据,参考文献及充分挖掘现有的公共数据库(BIND,BioGRID,DIP,GeneRIF,HPRD,IntAct,MINT和Reactome等),使用Cytoscape软件绘制并分析网络的结构和功能特点,评价网络的可靠性。从文献中的人PPI网络中提取与鼠疫菌蛋白发生相互作用的人的蛋白(HYT, human protein targeted by Y.pestis proteins)及这些蛋白间的相互作用,得到了HYT-HYT之间相互作用的网络。将此网络与鼠疫菌蛋白与人蛋白之间的相互作用网络融合得到了一个有176个节点,252条连线的网络。代表包括67个鼠疫菌蛋白和109个人蛋白的252对相互作用,其中含有109个HYT在已知的人PPI网络中拓展得到的44对相互作用。文献中报道的112个与EB病毒相互作用的人蛋白中有26个(26/112=23%)也与鼠疫菌相互作用,提示病毒与细菌可能在攻击宿主防御系统时采用了相似的策略,都是攻击人体PPI网络中的重要节点。这与目前已有的病原与宿主PPI的研究结果一致。在用GO biological process ID进行富集分析时发现,参与物种间相互作用的蛋白、多物种相互作用的蛋白、细胞黏附蛋白和一些细胞内的负调控蛋白得到显著性的富集;在用GO molecular function ID进行富集分析时发现,细胞骨架蛋白、转录因子结合蛋白、整合素蛋白、蛋白酶体蛋白等得到显著性的富集。这些蛋白很可能在鼠疫菌致病过程中发挥重要作用。说明我们筛到的PPI可能与鼠疫菌的感染过程密切相关。最后,我们根据深入文献调研,对筛选到的潜在相互作用进行了推测性的分析。鼠疫菌和宿主间的相互作用就是一系列鼠疫菌与人体蛋白的相互作用产成的协同作用,我们筛到的PPI在鼠疫菌感染人体的过程中具体发挥怎样的作用还有待于进一步深入的研究。本文以高通量的筛选方法研究鼠疫菌与人体蛋白之间的相互作用,获得了相互作用的初步网络,对网络进行了拓扑结构和功能分析,揭示了鼠疫菌与人之间PPI的主要方式,为深入研究鼠疫菌的致病机制提供了靶标。第二部分:鼠疫菌T3SS内PPI的研究pCD1质粒为3种对人致病的耶尔森氏菌所共有的,是耶尔森氏菌必须的毒力因子,它编码T3SS。该系统在鼠疫菌表面形成针状小体,能将至少6种效应蛋白(Yersinia outer membrane proteins,Yops)注射到宿主细胞中去。Yops一般具有蛋白酶的活性,通过对宿主蛋白的作用改变宿主细胞骨架结构、抵抗吞噬作用、干扰信号转导通路,从而抑制宿主正常的免疫响应。证实T3SS内部不同组分之间的PPI有助于预测和阐明Ysc机制的结构和功能。本实验选取鼠疫菌pCD1质粒编码基因中去除复制、分配相关基因、转座酶基因等后剩余57个基因,应用Y2H阵列筛选技术,进行T3SS内部相互作用的研究。实验共获得19对相互作用蛋白,其中的8对是以前的研究中曾经报道过的,剩余11对未见描述。相互作用主要分为三类:伴侣分子和分泌底物的相互作用,分泌调控复合体的相互作用与其他相互作用。有些相互作用发生在未知功能的假定蛋白之间;有些假定蛋白与已知的T3SS成分相互作用,提示他们可能是Ysc分泌机制新的成员。这些实验结果将有助于将来进一步阐明T3SS的结构和功能。确切的相互作用及意义有待于进一步研究。第三部分:负调控因子LcrG对T3SS基因表达影响的研究耶尔森氏菌T3SS由温度、钙离子、谷氨酸水平、PH值、营养获得和接触真核细胞等环境因素紧密调节,并与病原体的生命周期密切联系。T3SS装置又称作低钙反应刺激元(LCRS)。已知LcrG是Yops分泌的负调控因子,在37℃,体外Ca2+存在/缺失的条件下,lcrG突变株持续分泌Yops,并伴随生长抑制。为了更好的了解LcrG在调控鼠疫菌LCRS中的作用,我们构建了一个非极性的lcrG突变株,进行了比较转录组学分析。实验发现,lcrG突变株相比野生株有很多T3SS基因上调。免疫印记实验分析同样表明lcrG突变株分泌更多的YopM和LcrV蛋白。此外,通过体外(HeLa细胞和巨噬细胞)和体内(小鼠全身感染)实验分析lcrG突变株,均显示毒力显著减弱:在巨噬细胞中阻止TNF-α分泌的能力严重受损;BALB/c小鼠的LD50增加约600倍。尽管Yops在突变株中过表达,但转位到真核细胞中却明显受到阻碍,这一结论以前的报道一致。因此我们推断,LcrG不仅负调控Yops分泌,而且可能也负调控T3SS基因转录,使Yop合成和分泌保持在一个合适的水平。这一功能可能是通过间接机制显示的。更多还原

【Abstract】 Plague is a severe epidemic caused by Y. pestis and has caused three pandimes in human history. The pathogenic mechanism of Y. pestis depends on its interactions with the hosts. Revealing the protein-protein interactions (PPI) between Y. pestis and its host will not only lay a significant foundation for thoroughly understanding of its pathogenic mechanism, but also provide clues to develope efficient means for plague preventing and controlling. There are three parts in this research. First, by applying a high throughput Yeast Two-Hybrid (Y2H) System, the predicted virulence-related proteins of Y. pestis were used as baits to screen human spleen cDNA library. This study aims to obtain a primary PPI network between Y. pestis and huamn and a set of important interactions, so as to provide targets for deeply revealing the pathogenic mechanism of Y. pestis. Second, we systemically analyze the protein interactions within Y. pestis T3SS using yeast two hybrid system. Our results will help us to elucidate the structure and function of T3SS. Third, to better understand the role of negative regulator LcrG on T3SS, we compared the transcriptional profiles from wild-type strain and ?lcrG mutant by using whole-genome microarrays.PART 1 Investigation of interactions between Y.pestis virulence related proteins and human proteomeWe identified 152 Y. pestis proteins (about 3.75% of the Y. pestis full genome) that might interact with human proteins according to the results of previous studies carried in our lab, including the comparative genomics of Y. pestis and Y. pseudotuberculosis, the gene expression profile under different stimuli, serum antibody profile of infected hosts, and the current progress of Y. pestis research from references. Our research adopted the ProQuest Two-Hybrid System (Invitrogen) based on the Gateway Recombination Technology, which facilitate the high throughput parallel operation. Through the gateway recombination technology, 154 bait vectors corresponding to 152 ORFs of Y. pestis proteins (two of them were fragmented) were constructed. Three of them (YPCD1.26c, YPCD1.31c and YPMT1.61c) were self-activating baits.High-throughput Y2H screening were carried out using sequential transformation method. 151 non-self-activating baits were screened against the human spleen cDNA library for at least once. The candidate clones grown on the selective plates were selected further by analyzing the auxotrophic phenotypes of reporter genes his and ade and the X-Gal assay to identify the phenotype of reporter gene lacZ, respectively. The prey-vectors form positive clones were isolated and sequenced. Total of 1087 preys were successfully sequenced. The BLAST program of the NCBI website was used to search the database, 833 clones containing the inframe inserts of full or partial ORFs were identified, which were corresponding to 185 non-redundant prey molecules. These 185 prey formed 359 interactions with 92 proteins of Y. pestis, which constituted our core data set. No known interactions reported previously were screened out in our experiment. Among these 359 interaction pairs, 60 occurred twice and 91 occurred three times or more, which represented 16.7% and 25.5% of the total PPIs number, respectively, suggesting that our screening results should be credible.Subsequently, experimental validation and bioinformatics analysis were adopted to determine the reliability of those interactions. In the experimental part, we evaluated the technical false positive by plasmid retransformation in yeast and GST pull-down assay. Plasmid retransformation assay were carried out in 375 pairs PPIs and 208 pairs shown to be positive through phenotypic identification, with a positive rate of 57.9% that is comparable with the published YTH work . We obtained 208 PPIs between 67 baits and 109 preys. Among the 208 PPIs that shown to be positive in the plasmid rescue from yeast experiment, 21 were selected for GST pull-down assay, representing 10% of the total retransformation-positive protein pairs (208 pairs). Of these 21 interaction pairs, six prey proteins were involved, and 7 interaction pairs were unable to be validated because of the inability of a prey protein to release into the supernatant of cell lysate. Of the rest 14 interaction pairs, 13 were validated positive and one negative by GST pull-down assay, resulting in a positive rate of 92.8%, indicating that the data set of 208 pairs of protein-protein interactions between Y. pestis and human protein had comparatively high reliability. With the thoroughly mining the current public database (BIND, BioGRID, DIP, GeneRIF, HPRD, IntAct, MINT and Reactome etc.), we construct a network using the data set consisting of 208 PPIs, and analyzed the characteristics of its structure and function. Network between HYT-HYT (HYT, human protein targeted by Y.pestis proteins) were also obtained by extracting HYT and their interactions from human interaction networks form existing references. These two networks were fused with each other to form a network with 176 nodes and 252 lines, representing 252 pairs of interactions involving 67 Y. pestis proteins and 109 human proteins, which including 44 pairs of interactions between 109 HYT, which was extracted form the known human protein interaction networks. The relatively high reliability of those interactions was validated by topological analysis and functional GO analysis. Among the 112 human proteins that were found to interact with Epstein-Barr virus by YTH, 26 also interacted with Y. pestis (26/112=23%), which suggested that the virus and the bacteria might use similar strategy to attack the important nodes of human protein interaction network and the host’s defending system. This result was consisted with the current understanding of pathogen-host protein interaction. When processing GO biological process ID enrichment analysis, proteins involved in interspecies interactions and multi-species interactions, cell adhesion proteins, some intracellular negative regulatory proteins were significantly enriched. Processing GO molecular function ID enrichment analysis, cytoskeleton proteins, transcription factor binding proteins, integrins and proteasome proteins were significantly enriched. These proteins may play important roles in the pathogenesis of Y. pestis. Finally, we conducted predictive analysis of those interactions based on deeply investigation of related references. Interactions between Y.pestis and its host were coordinated effects produced by a series of Y. pestis and human protein interactions, and further studies are needed to define the roles of those PPIs in the process of Y. pestis infection.This study investigated the PPIs between Y. pestis and its host by high throughput screening method. We obtained a primary interaction network between Y. pestis and human proteins, which had been subjected to topological analysis and functional analysis further. The results revealed the main pattern of protein-protein interactions between Y. pestis and human and provided lots of targets for elucidating the pathogenic mechanism of Y. pestis.PART 2 Protein-Protein interactions within Type III Secretion System of Y. pestispCD1 is a common plasmid owned by three pathogenic Yersinia species. It is the essential virulence determinant of Y. pestis and encodes T3SS that can form injectisome on the surface of the bacteria and inject at least six Yops(Yersinia outer membrane proteins)into the host cell. Yop effectors usually have various protease activities, that render the pathogens’ability to inhibit the normal immune response of the host by alter the cytoskeletal structure of the hots cell, resist phagocytosis and interfere with signal transduction pathways.Identifying the interaction of among the components of T3SS is helpful for predicting and elucidating the structure and function of the Ysc injectisome. This study selected 57 genes from the pCD1 plasmid encoded-genes by excluding those related with plasmid replication and partition, and the putative transposase genes. YTH matrix technique was used to study the interactions within T3SS. Totally 19 pairs of interaction proteins were obtained in the research. Eight pairs of which were reported previously, and the rest 11 pairs hadn’t been reported yet. The interactions were classified into three major types: i) interaction between the chaperon and secretory substrate; ii) interaction between secretory regulation complexes; iii) other interactions. Several putative proteins with unknown function were involved in interactions with the known T3SS proteins or another unknown function protein, indicating that they might be new members of T3SS. These results are helpful for elucidating the T3SS structure and function in the future. Further studies are needed to define the precise interaction and significance.PART 3 Study on the regulatory role of LcrG on the T3SS transcriptionYersinia T3SS is tightly regulated by environmental factors that are closely related to the life cycle of the pathogens, such as temperature, calcium and glutamine levels, pH, nutrient availability, and contact with eukaryotic cells. It is known that LcrG is a negative regulator for the secretion of Yops. At mammalian body temperature, i.e. 37℃, ?lcrG mutant consistently secretes Yops in the presence or absence of Ca2+ in vitro, accompanied by growth restriction.In this work, to better understand the roles of LcrG in the regulation of LCRS in Y. pestis, we compared the transcriptional profiles from wild-type strain and ?lcrG mutant by using whole-genome microarrays. It was found that most T3SS genes were up-regulated in the ?lcrG mutant when compared with the wild-type strain, and immunoblotting analysis also revealed that more YopM and LcrV proteins were secreted by the ?lcrG mutant. Furthermore, the virulence of the ?lcrG mutant was dramatically attenuated in both in vitro (HeLa cells and macrophage cells) and in vivo (mice systemic infection) analysis. The capability to inhibit TNF-αsecretion in macrophages was severely impaired and the LD50 in BALB/c mice was increased by approximately 600 folds. Although Yops were overexpressed in the ?lcrG mutant, the translocation of them into the eukaryotic cells was severely hampered, which was consistent with the previous report.We therefore speculate that LcrG could play a negative regulatory role not only in the Yop secretion but also in the transcription of T3SS genes, and this function is possibly exerted through indirect mechanisms.更多还原