elF4E, the mRNA cap binding protein, is a master switch that controls eukaryotic translation. To be active, it must bind elF4G and form the elF4F complex, which also contains elF4A. Translation is downregulated by association of elF4E with 4E-BP, which occupies the elF4G binding site. Signalling events acting on 4E-BP cause it to dissociate from elF4E, and elF4E is then free to bind elF4G to form the active elF4F complex. We have solved the structure of the yeast elF4E/m 7 Gpp complex in a CHAPS micelle. We determined the position of the second nucleotide in a complex with m 7 GpppA, and identified the 4E-BP binding site. elF4E has a curved eight-stranded antiparallel β-sheet, decorated with three helices on the convex face and three smaller helices inserted in connecting loops. The m 7 G of the cap is intercalated into a stack of tryptophans in the concave face. The 4E-BP binding site is located in a region encompassing one edge of the β-sheet, the adjacent helix a2 and several regions of non-regular secondary structure. It is adjacent to, but does not overlap the cap-binding site.

Careful pulse sequence design and optimization is critical to the success of a given NMR experiment. Over the past several decades the level of sophistication of NMR pulse sequences has increased tremendously, leading to large spectral sensitivity and resolution improvements, to data sets with far fewer artifacts, and to much more rapid acquisition times, opening up a wide range of applications. Here I briefly highlight how pulse sequence ‘engineering’ has evolved, focusing on liquid state NMR, and, in particular, on the HNCA-class of triple-resonance experiment. In many respects, the evolution of triple-resonance NMR mirrors the evolution of solution state NMR experiments in general, with ‘tricks’ that first appeared in triple-resonance pulse sequences or that were motivated by them now incorporated into a broad range of experiments.

Previous NMR studies on the ternary complex of human dihydrofolate reductase (hDHFR) with methotrexate (MTX) and NADPH detected six long-lived bound water molecules. Two of the water molecules, WatA and WatB, stabilize the structure of the protein while the other four, WatC, WatD, WatE and WatF, are involved in substrate binding and specificity. WatE may also act as a proton shuttle during catalysis. Here, the contributions of individual residues to the binding of these water molecules are investigated by performing NMR experiments on ternary complexes of mutant enzymes, W24F, E30A and E30Q. W24 and E30 are conserved residues that form hydrogen bonds with WatE in crystal structures of DHFR. Nuclear Overhauser effects (NOEs) are detected between WatE and the protein in all the mutant complexes, hence WatE still has a long lifetime bound to the complex when one of its hydrogen-bonding partners is deleted or altered by mutagenesis. The NOEs for WatE are much weaker, however, in the mutants than in wild-type. The NOEs for the other water molecules in and near the active site, WatA, WatC, WatD and WatF, also tend to be weaker in the mutant complexes. Little or no change is apparent in the NOEs for WatB, which is located outside the active site, farthest from the mutated residues. The decreased NOE intensities for the bound water molecules could be caused by changes in the positions and/or lifetimes of the water molecules. Chemical shift and NOE data indicate that the mutants have structures very similar to that of wild-type hDHFR, with possible conformational changes occurring only near the mutated residues. Based on the lack of structural change in the protein and evidence for increased structural fluctuations in the active sites of the mutant enzymes, it is likely that the NOE changes are caused, at least in part, by decreases in the lifetimes of the bound water molecules.