Davids Wagied, Zhang Zhaolei
Banting & Best Department of Medical Research (BBDMR), Donnelly Centre for Cellular & Biomolecular Research (CCBR), University of Toronto, 160 College Street, Toronto, ON M5S 3E1, Canada. wagied.davids@utoronto
BMC Evol Biol. 2008 Jan 24;8:23. doi: 10.1186/1471-2148-8-23.
Despite the prevalence of horizontal gene transfer (HGT) in bacteria, to this date there were few studies on HGT in the context of gene expression, operons and protein-protein interactions. Using the recently available data set on the E. coli protein-protein interaction network, we sought to explore the impact of HGT on genome structure and protein networks.
We classified the E. coli genes into three categories based on their evolutionary conservation: a set of 2158 Core genes that are shared by all E. coli strains, a set of 1044 Non-core genes that are strain-specific, and a set of 1053 genes that were putatively acquired by horizontal transfer. We observed a clear correlation between gene expressivity (measured by Codon Adaptation Index), evolutionary rates, and node connectivity between these categories of genes. Specifically, we found the Core genes are the most highly expressed and the most slowly evolving, while the HGT genes are expressed at the lowest level and evolve at the highest rate. Core genes are the most likely and HGT genes are the least likely to be member of the operons. In addition, we found the Core genes on average are more highly connected than Non-core and HGT genes in the protein interaction network, however the HGT genes displayed a significantly higher mean node degree than the Core and Non-core genes in the defence COG functional category. Interestingly, HGT genes are more likely to be connected to Core genes than expected by chance, which suggest a model of differential attachment in the expansion of cellular networks.
Results from our analysis shed light on the mode and mechanism of the integration of horizontally transferred genes into operons and protein interaction networks.
尽管水平基因转移（HGT）在细菌中普遍存在，但迄今为止，在基因表达、操纵子和蛋白质 - 蛋白质相互作用的背景下，关于HGT的研究较少。利用最近可获得的大肠杆菌蛋白质 - 蛋白质相互作用网络数据集，我们试图探索HGT对基因组结构和蛋白质网络的影响。
我们根据进化保守性将大肠杆菌基因分为三类：一组由所有大肠杆菌菌株共享的2158个核心基因，一组特定于菌株的1044个非核心基因，以及一组推测通过水平转移获得的1053个基因。我们观察到这些类别基因之间的基因表达能力（通过密码子适应指数衡量）、进化速率和节点连通性之间存在明显的相关性。具体而言，我们发现核心基因表达水平最高且进化最慢，而HGT基因表达水平最低且进化速率最高。核心基因最有可能且HGT基因最不可能成为操纵子的成员。此外，我们发现在蛋白质相互作用网络中，核心基因平均比非核心基因和HGT基因连接性更高，然而在防御COG功能类别中，HGT基因显示出比核心基因和非核心基因显著更高的平均节点度。有趣的是，HGT基因比随机预期更有可能与核心基因相连，这表明在细胞网络扩展中存在差异附着模型。
我们的分析结果揭示了水平转移基因整合到操纵子和蛋白质相互作用网络中的模式和机制。
