Stern S, Weiser B, Noller H F
Thimann Laboratories, University of California, Santa Cruz 95064.
J Mol Biol. 1988 Nov 20;204(2):447-81. doi: 10.1016/0022-2836(88)90588-8.
We have derived a model for the three-dimensional folding of Escherichia coli 16 S ribosomal RNA, using interactive computer graphic methods. It is based on (1) the secondary structure derived from comparative sequence analysis, (2) the three-dimensional co-ordinates for the centers of mass of the 30 S subunit proteins, and (3) the locations of sites in 16 S rRNA that interact with specific ribosomal proteins, from footprinting and crosslinking studies. We present a detailed description of the derivation of the model. About 75% of the RNA chain is sufficiently constrained to provide a useful model. This contains most of the universally conserved core of the molecule. In all but a few instances, protected and crosslinked sites can be placed within or very close to their cognate proteins, while obeying stereochemical rules. The overall shape of the model and locations of specific regions of the RNA correspond well to data derived from electron micrographs of 30 S subunits, although such data were not used to construct the model. Phylogenetic variations in the structure are readily accommodated; as an example, we have modeled the 950-nucleotide mammalian mitochondrial 12 S rRNA by superimposing it on the E. coli structure. The three major RNA domains, as defined by secondary structure, appear to exist as autonomous structural units in three dimensions, for the most part. There is an extensive interface between the 5' and central domains, whereas the 3' major domain has relatively little apparent contact with the rest of the structure. The 5', central and 3' major domains form structures that resemble the body, platform and head, respectively, seen in electron micrographs of 30 S subunits. We discuss possible roles for the ribosomal proteins in stabilizing specific structural features of the RNA during ribosome assembly. The decoding site, as deduced from footprinting and crosslinking studies involving the tRNA anticodon stem-loop, is well-localized. Bases protected from chemical probing by the anticodon stem-loop line the cleft of the subunit. The conserved loop at position 530, which contains some of the bases protected by A site-bound tRNA, is remote (approx. 80 A) from the decoding site. Protection of these bases by the anticodon stem-loop is thus unlikely to be due to direct contact.
我们使用交互式计算机图形方法，推导了大肠杆菌16S核糖体RNA的三维折叠模型。该模型基于：（1）通过比较序列分析得出的二级结构；（2）30S亚基蛋白质质心的三维坐标；（3）足迹法和交联研究中确定的16S rRNA与特定核糖体蛋白相互作用位点的位置。我们详细描述了该模型的推导过程。约75%的RNA链受到充分限制，从而提供了一个有用的模型。这包含了该分子大部分普遍保守的核心区域。在绝大多数情况下，受保护和交联的位点能够放置在与其对应的蛋白质内部或非常接近的位置，同时遵循立体化学规则。尽管构建该模型时未使用30S亚基电子显微镜数据，但模型的整体形状和RNA特定区域的位置与这些数据非常吻合。该结构的系统发育变异很容易得到解释；例如，我们通过将950个核苷酸的哺乳动物线粒体12S rRNA叠加在大肠杆菌结构上，对其进行了建模。由二级结构定义的三个主要RNA结构域在三维空间中大多似乎作为自主结构单元存在。5'和中央结构域之间有广泛的界面，而3'主要结构域与结构的其余部分的明显接触相对较少。5'、中央和3'主要结构域形成的结构分别类似于30S亚基电子显微镜图像中看到的主体、平台和头部。我们讨论了核糖体蛋白在核糖体组装过程中稳定RNA特定结构特征的可能作用。通过涉及tRNA反密码子茎环的足迹法和交联研究推断出的解码位点定位良好。反密码子茎环保护免受化学探针作用的碱基排列在亚基的裂隙处。位于530位的保守环，包含一些被A位点结合的tRNA保护的碱基，距离解码位点较远（约80埃）。因此，反密码子茎环对这些碱基的保护不太可能是由于直接接触。
