Martin Che L, Sukarna Tika Y, Akther Saymon, Ramrattan Girish, Pagan Pedro, Di Lia, Mongodin Emmanuel F, Fraser Claire M, Schutzer Steven E, Luft Benjamin J, Casjens Sherwood R, Qiu Wei-Gang
Department of Biology, The Graduate Center, City University of New York, New York, USA.
Department of Biological Sciences and Center for Translational and Basic Research, Hunter College, City University of New York, New York, USA.
mBio. 2015 Apr 14;6(2):e00011-15. doi: 10.1128/mBio.00011-15.
Phylogenomic footprinting is an approach for ab initio identification of genome-wide regulatory elements in bacterial species based on sequence conservation. The statistical power of the phylogenomic approach depends on the degree of sequence conservation, the length of regulatory elements, and the level of phylogenetic divergence among genomes. Building on an earlier model, we propose a binomial model that uses synonymous tree lengths as neutral expectations for determining the statistical significance of conserved intergenic spacer (IGS) sequences. Simulations show that the binomial model is robust to variations in the value of evolutionary parameters, including base frequencies and the transition-to-transversion ratio. We used the model to search for regulatory sequences in the Lyme disease species group (Borrelia burgdorferi sensu lato) using 23 genomes. The model indicates that the currently available set of Borrelia genomes would not yield regulatory sequences shorter than five bases, suggesting that genome sequences of additional B. burgdorferi sensu lato species are needed. Nevertheless, we show that previously known regulatory elements are indeed strongly conserved in sequence or structure across these Borrelia species. Further, we predict with sufficient confidence two new RpoS binding sites, 39 promoters, 19 transcription terminators, 28 noncoding RNAs, and four sets of coregulated genes. These putative cis- and trans-regulatory elements suggest novel, Borrelia-specific mechanisms regulating the transition between the tick and host environments, a key adaptation and virulence mechanism of B. burgdorferi. Alignments of IGS sequences are available on BorreliaBase.org, an online database of orthologous open reading frame (ORF) and IGS sequences in Borrelia.
While bacterial genomes contain mostly protein-coding genes, they also house DNA sequences regulating the expression of these genes. Gene regulatory sequences tend to be conserved during evolution. By sequencing and comparing related genomes, one can therefore identify regulatory sequences in bacteria based on sequence conservation. Here, we describe a statistical framework by which one may determine how many genomes need to be sequenced and at what level of evolutionary relatedness in order to achieve a high level of statistical significance. We applied the framework to Borrelia burgdorferi, the Lyme disease agent, and identified a large number of candidate regulatory sequences, many of which are known to be involved in regulating the phase transition between the tick vector and mammalian hosts.
系统发育足迹法是一种基于序列保守性从头识别细菌物种全基因组调控元件的方法。系统发育方法的统计效力取决于序列保守程度、调控元件长度以及基因组间的系统发育分歧水平。在早期模型的基础上，我们提出了一种二项式模型，该模型使用同义树长度作为中性预期，以确定保守基因间隔序列（IGS）的统计显著性。模拟结果表明，二项式模型对于进化参数值的变化具有鲁棒性，这些参数包括碱基频率和转换与颠换比率。我们使用该模型，利用23个基因组在莱姆病物种组（广义伯氏疏螺旋体）中搜索调控序列。该模型表明，目前可用的伯氏疏螺旋体基因组集合不会产生短于5个碱基的调控序列，这表明需要更多广义伯氏疏螺旋体物种的基因组序列。尽管如此，我们表明，先前已知的调控元件在这些伯氏疏螺旋体物种的序列或结构中确实高度保守。此外，我们有足够把握预测出两个新的RpoS结合位点、39个启动子、19个转录终止子、28个非编码RNA以及四组共调控基因。这些假定的顺式和反式调控元件提示了调控蜱与宿主环境之间转变的新的、伯氏疏螺旋体特异性机制，这是伯氏疏螺旋体的一种关键适应性和毒力机制。IGS序列比对可在BorreliaBase.org上获取，该网站是伯氏疏螺旋体直系同源开放阅读框（ORF）和IGS序列的在线数据库。
虽然细菌基因组大多包含蛋白质编码基因，但它们也包含调控这些基因表达的DNA序列。基因调控序列在进化过程中往往是保守的。因此，通过对相关基因组进行测序和比较，人们可以基于序列保守性识别细菌中的调控序列。在这里，我们描述了一个统计框架，通过该框架可以确定需要测序多少个基因组以及在何种进化相关性水平上，以达到较高的统计显著性。我们将该框架应用于莱姆病病原体伯氏疏螺旋体，并鉴定出大量候选调控序列，其中许多已知参与调控蜱传播媒介与哺乳动物宿主之间的相变。
