Shabalina Svetlana A, Ogurtsov Aleksey Y, Rogozin Igor B, Koonin Eugene V, Lipman David J
National Center for Biotechnology Information, National Library of Medicine, National Institutes of Health, Bethesda, MD 20894, USA.
Nucleic Acids Res. 2004 Mar 18;32(5):1774-82. doi: 10.1093/nar/gkh313. Print 2004.
Sequencing of multiple, nearly complete eukaryotic genomes creates opportunities for detecting previously unnoticed, subtle functional signals in non-coding regions. A genome-wide comparative analysis of orthologous sets of mammalian and yeast mRNAs revealed distinct patterns of evolutionary conservation at the boundaries of the untranslated regions (UTRs) and the coding region (CDS). Elevated sequence conservation was detected in approximately 30 nt regions around the start codon. There seems to be a complementary relationship between sequence conservation in the approximately 30 nt regions of the 5'-UTR immediately upstream of the start codon and that in the synonymous positions of the 5'-terminal 30 nt of the CDS: in mammalian mRNAs, the 5'-UTR shows a greater conservation than the CDS, whereas the opposite trend holds for yeast mRNAs. Unexpectedly, a approximately 30 nt region downstream of the stop codon shows a substantially lower level of sequence conservation than the downstream portions of the 3'-UTRs. However, the sequence in this poorly conserved 30 nt portion of the 3'-UTR is non-random in that it has a higher GC content than the rest of the UTR. It is hypothesized that the elevated sequence conservation in the region immediately upstream of the start codon is related to the requirement for initiation factor binding during pre-initiation ribosomal scanning. In contrast, the poorly conserved region downstream of the stop codon could be involved in the post- termination scanning and dissociation of the ribosomes from the mRNA, which requires only the mRNA-ribosome interaction. Additionally, it was found that the choice of the stop codon in mammals, but not in yeasts, and the context in the immediate vicinity of the stop codons in both mammals and yeasts are subject to strong selection. Thus, genome-wide analysis of orthologous gene sets allows detection of previously unrecognized patterns of sequence conservation, which are likely to reflect hidden functional signals, such as ribosomal filters that could regulate translation by modulating the interaction between the mRNA and ribosomes.
对多个近乎完整的真核生物基因组进行测序，为在非编码区检测先前未被注意到的微妙功能信号创造了机会。对哺乳动物和酵母mRNA直系同源集进行全基因组比较分析，揭示了非翻译区（UTR）和编码区（CDS）边界处不同的进化保守模式。在起始密码子周围约30 nt的区域检测到序列保守性升高。起始密码子上游5'-UTR约30 nt区域的序列保守性与CDS 5'-末端30 nt同义位置的序列保守性之间似乎存在互补关系：在哺乳动物mRNA中，5'-UTR的保守性高于CDS，而酵母mRNA则呈现相反趋势。出乎意料的是，终止密码子下游约30 nt的区域显示出比3'-UTR下游部分低得多的序列保守水平。然而，3'-UTR这个保守性较差的30 nt部分中的序列并非随机，因为它的GC含量高于UTR的其余部分。据推测，起始密码子上游区域序列保守性升高与起始前核糖体扫描过程中起始因子结合的需求有关。相反，终止密码子下游保守性较差的区域可能参与核糖体从mRNA上的终止后扫描和解离，这仅需要mRNA-核糖体相互作用。此外，研究发现，哺乳动物（而非酵母）中终止密码子的选择以及哺乳动物和酵母中终止密码子紧邻区域的上下文都受到强烈选择。因此，对直系同源基因集进行全基因组分析能够检测到先前未被识别的序列保守模式，这些模式可能反映了隐藏的功能信号，例如可能通过调节mRNA与核糖体之间的相互作用来调控翻译的核糖体筛选机制。
