New pulse schemes for recording intermolecular NOEs in a molecular complex consisting of 15N,13C labeled and unlabeled components are presented. The pulse sequences select for magnetization transferred from protons on the unlabeled component to proximal protons of the labeled molecule. Filtering (suppression of signal from 13C labeled molecules) is accomplished using adiabatic 13C inversion pulses which are swept at a rate which is tuned according to the one-bond 1H−13C scalar coupling vs carbon chemical shift profile of the labeled molecule in the complex. Significantly improved spectra are obtained relative to data recorded with other purging schemes. Improvements are demonstrated in experiments where intermolecular NOEs between labeled RNA-unlabeled peptide and labeled protein−unlabeled peptide are recorded. A discussion of structural information obtained for a complex of the amino-terminal arginine rich domain of the N protein from bacteriophage λ and boxB RNA using the new methodology is presented.

The interaction between the amino-terminal transactivation domain (TAD) of p53 and TFIIH is directly correlated with the ability of p53 to activate both transcription initiation and elongation. We have identified a region within the p53 TAD that specifically interacts with the pleckstrin homology (PH) domain of the p62 and Tfb1 subunits of human and yeast TFIIH. We have solved the 3D structure of a complex between the p53 TAD and the PH domain of Tfb1 by NMR spectroscopy. Our structure reveals that p53 forms a nine residue amphipathic alpha helix (residues 47-55) upon binding to Tfb1. In addition, we demonstrate that diphosphorylation of p53 at Ser46 and Thr55 leads to a significant enhancement in p53 binding to p62 and Tfb1. These results indicate that a phosphorylation cascade involving Ser46 and Thr55 of p53 could play an important role in the regulation of select p53 target genes.

The structure of the complex formed by the arginine-rich motif of the transcriptional antitermination protein N of phage lambda and boxB RNA was determined by heteronuclear magnetic resonance spectroscopy. A bent alpha helix in N recognizes primarily the shape and negatively charged surface of the boxB hairpin through multiple hydrophobic and ionic interactions. The GAAGA boxB loop forms a GNRA fold, previously described for tetraloops, which is essential for N binding. The fourth nucleotide of the loop extrudes from the GNRA fold to enable the E. coli elongation factor NusA to recognize the N protein/RNA complex. This structure reveals a new mode of RNA-protein recognition and shows how a small RNA element can facilitate a protein-protein interaction and thereby nucleate formation of a large ribonucleoprotein complex.

A procedure is described for the efficient preparation of isotopically enriched RNAs of defined sequence. Uniformly labelled nucleotide 5'triphosphates (NTPs) were prepared from E.coli grown on 13C and/or 15N isotopically enriched media. These procedures routinely yield 180 mumoles of labelled NTPs per gram of 13C enriched glucose. The labelled NTPs were then used to synthesize RNA oligomers by in vitro transcription. Several 13C and/or 15N labelled RNAs have been synthesized for the sequence r(GGCGCUUGCGUC). Under conditions of high salt or low salt, this RNA forms either a symmetrical duplex with two U.U base pairs or a hairpin containing a CUUG loop respectively. These procedures were used to synthesize uniformly labelled RNAs and a RNA labelled only on the G and C residues. The ability to generate milligram quantities of isotopically labelled RNAs allows application of multi-dimensional heteronuclear magnetic resonance experiments that enormously simplify the resonance assignment and solution structure determination of RNAs. Examples of several such heteronuclear NMR experiments are shown.