James H. Doudna Cate
In order for life to exist, inert genes must be translated into action, action that occurs when proteins are made by the ribosome. A large, complex machine, the ribosome is poorly understood despite intensive experimental efforts for decades. Key questions about the fundamental nature of protein synthesis remain unanswered. For example, how does the ribosome read the genetic code? And how do certain antibiotics, so useful in reducing infections, cripple ribosomes? In mammals, how are ribosomes regulated, and how do certain viruses and cancers circumvent this regulation? Probing the structure of prokaryotic ribosomes provides the best approach for understanding the highly-conserved functions of the ribosome in all kingdoms of life. With ribosome structures to 7-10 = or higher resolution, we will be able to compare the known biochemical, genetic, and phylogenetic data to the medium-resolution electron density map, revolutionizing the kinds of questions that can be asked about the function of the ribosome during the translation cycle. High-resolution ribosome structures (3=) will build the foundation for insight into the "business end" of the genetic code
Crystallization of 70S ribosome complexes: To investigate translational fidelity, my structural studies will focus on 70S ribosomal complexes. Transfer RNA is thought to bind to the ribosome in at least seven different modes involving both classical and hybrid states . The classical states and some of the hybrid states may form stable ribosome complexes and produce well-ordered crystals. Although the present crystals of 70S ribosomes are from a prokaryotic thermophile, previous experiments on E. coli ribosomes will guide our approaches to forming stable complexes for crystallization. Other complexes that may provide better crystals include those with elongation factor G (EFG), EF-Tu/tRNA ternary complex bound to the ribosome, or with antibiotics that target the ribosome.
In collaboration with Drs. Marat Yusupov, Gulnara Yusupova, and Harry Noller, we have obtained crystals of a 70S ribosome complex that diffract x-rays to a nominal resolution of 5 =. From these crystals, we have solved the structure of the ribosome to 10 = and are now working to obtain phases for the structure factors to about 7 =. This is a comparable phasing resolution to the initial nucleosome structure .
What will we be able to see at 10 = or better resolution? In the nucleosome core particle solved at 7 = resolution, the DNA was clearly distinguishable from proteins, as were the DNA major and minor grooves. In the 10 = electron density maps of the ribosome, we are beginning to see the major and minor grooves of RNA double helices. We have also unambiguously identified the positions of tRNA in the 3 of its binding states. Based on the extensive biochemical data collected on similar prokaryotic ribosomes, we have also begun to model the RNA and protein positions in the 30S and 50S subunits. While high-resolution structures of the ribosome will be necessary to build complete models, our preliminary modeling points to numerous biochemical experiments to test ribosome function.
While the 70S electron density maps are still in the process of refinement, we have taken the phases from one of the heavy-atom derivatives to use in a control experiment as a gauge of the quality of the phases. We have collected 7-9 = resolution data sets from two different 70S ribosome complexes containing either an intact tRNA or partial tRNA in the ribosomal P site. When using the x-ray phases in combination with difference amplitudes from these complexes ( |F|full tRNA -|F|partial tRNA), at 10 = resolution we obtain an electron density map in which the location of the tRNA in the P site is clearly seen. The major and minor grooves in the RNA helices are clearly visible, and the position of the 3'-CCA of the tRNA is prominent . From this control experiment, we know that the heavy-atom phasing is working very well so far.
The importance of cations in ribosome folding and function: All of the fundamental activities of the ribosome depend on ionic conditions. For example, as the concentration of Mg2+ drops, 70S ribosomal particles dissociate into 50S and 30S subunits. Further lowering the Mg2+ concentration causes the subunits to disassemble. In the other direction, ribosome reconstitution has specific Mg2+ and polyamine requirements . Most importantly, translation fidelity in vitro depends on a narrow range of ionic conditions, above and below which ribosomes lose decoding accuracy. These results, along with my discovery of a specific magnesium ion core in the P4-P6 domain from the T. thermophila group I intron , suggest there are specific binding sites for Mg2+ in the ribosome that play key roles in ribosome structure and function.
We will use the 70S ribosome crystals to locate and analyze these sites in a manner similar to the approach used on the P4-P6 group I intron domain . However, the ribosome poses additional problems related to its size and the resolution to which its structure is known. To overcome the limits of resolution in a 7-10 = map, I will have to rely on markers like osmium (III) hexammine and pentammine and lanthanides to identify potential magnesium sites in the electron density . In combination with known biochemical data on important phosphate oxygens and careful modeling of the RNA into the electron density map, we will develop hypotheses as to which parts of the ribosomal RNA bind metal ions.
The RNA we identify crystallographically in possible magnesium binding sites will be tested for its functional roles via phosphorothioate substitution interference . Functional tests such as folding compactness or subunit association will be used to separate active from inactive RNAs, which can then be analyzed by iodine-EtOH cleavage and reverse-transcription. Phosphorothioates which interfere with Mg2+ binding should be suppressed to some extent by manganese ion binding. Combining the crystallographic data with tests of ribosome activity may finally show the specific roles metal ions play in translation.