The Structure of the 16S Ribosomal RNA

This project, supported by NIH NLM-05652 and the Culpeper Foundation, is part of the Helix Group at Stanford School of Medicine. Please address inquiries to russ.altman@stanford.edu.

1. Summary of Project Goals

Understanding macromolecular structure provides insight into the structural basis for function. It may also suggest methods for perturbing or abolishing function. It is therefore a subject of intense interest in molecular biology. Methods for determining and refining medium-sized protein and nucleic acid molecules (containing on the order of 500 amino acids or 100 bases) are mature and well understood. The size and flexibility of large macromolecular complexes makes them difficult to study with the primary methods for determining structure (x-ray crystallography and nuclear magnetic resonance). For now, the structures of large complexes must be probed with experimental methods that provide structural information that is relatively sparse and relatively noisy. Such methods include determination of molecular shape and volume, estimating proximity relationships with cross-linking, labeling, protection experiments, and many others. In addition, computer-based prediction methods may provide some information about substructures (for example, secondary structure or class of fold). In using these data to construct structural models, it is important to acknowledge the levels of uncertainty in these data sources, and to understand that they will most likely be compatible with a family of structures, instead of a single structure.

 We have previously described a system, PROTEAN-1, which is able to use multiple sources of uncertain data to produce estimates of structure. PROTEAN-1 is based on a constraint-satisfaction paradigm and is summarized below. PROTEAN is able to model the structure of a large ribonucleoprotein complex (the central domain of the 30S ribosomal subunit) from a combination of constraints derived from sequence analysis and experiments. The ease with which the data analysis protocols can be changed, and the consequent effects on the structural model can be gauged, makes PROTEAN a good program for exploring the structural implications of sparse and noisy data. The underlying constraint satisfaction engine can guarantee that the resulting structures satisfy all the constraints provided, and that they provide an upper bound on the set of structures that are compatible with the data.

 The ribosome is the site of translation of messenger RNA into protein. It is composed of two subunits. In prokaryotes, the large subunit is called 50S and the small subunit is called 30S. The 30S subunit consists of a single strand of RNA (the 16S rRNA, 1542 bases), and 21 proteins ranging in molecular weight from 9 kD to 61 kD. The 30S subunit can exist independently from the 50S subunit and be reconstituted from its components in vitro. The 30S subunit is the site of translation initiation. An understanding of its structure is useful for understanding the mechanisms of protein translation, and perhaps for designing drugs to modify these mechanisms.
 

2. Project Personnel

3. Shared Software and Data from Project

  1. Model of 16S rRNA from hydroxyl footprinting data.


    2. We have created a model of the 16S rRNA structure using hydroxyl data and a computational method based on constraint satisfaction to define the allowable conformational space for each secondary structural element in the context of protein positions. A description of the work and coordinates of the model are available here .
     

  2. The proteanD display program


    3. We published a paper in J. Mol. Graphics describing the program, proteand which is designed to display various representations of structural uncertainty for macromolecules, including overlapping stick drawings, ellipsoids of uncertainty, and secondary structure accessible volumes. These representations are closely related to our methodology for computing the structure of ribosomal RNA. The program runs on Silicon Graphics (SGI) machines and sample input files and binary executable code are available in ftp://ftp-smi.stanford.edu/pub/altman/tar.proteand.
     

  3. Model of the 16S Central Domain


    4. The initial model of the central domain of the 16S ribosomal RNA structure is contained as a sample input to the display program, proteanD.
     

  4. Reusing problem solving techniques: Ribosomes as an example.


    5. With John Gennari and Mark Musen of the Knowledge Modeling Group at Stanford Section on Medical Informatics, we have investigated the use of reusable constraint satisfaction methods in the context of ribosome modeling. The propose-and-revise methodology can be extended to the ribosome problem, although at considerable loss of problem solving efficiency. See references below.
     

References

   PubMed link to papers related to this project.
Last update March 10, 2004.
help@helix.stanford.edu