Determining valine side-chain rotamer conformations in proteins from methyl 13C chemical shifts: application to the 360 kDa half-proteasome.

A method is presented for determining Val side-chain χ(1) rotamer distributions in proteins based exclusively on measured (13)C(γ1) and (13)C(γ2) chemical shifts. The approach selects an ensemble of 20 χ(1) values, calculates average methyl (13)C(γ1,γ2) chemical shifts via theoretical quantum chemical calculations and maximizes the agreement with the experimentally measured shifts using a genetic algorithm. The methodology is validated with an application involving six proteins for which (13)C(γ) chemical shifts and three-bond methyl-backbone scalar couplings are available. The utility of the methodology is demonstrated with an application to the 360 kDa 'half-proteasome' where the χ(1) rotameric distributions of Val residues are calculated on the basis of chemical shifts. For the most part the χ(1) profiles so obtained compare very well with those generated from the high-resolution (2.3 Å) X-ray structure of the proteasome. Both NMR and X-ray distributions are cross-validated by comparing calculated (1)H-(13)C methyl residual dipolar couplings with measured values, and the level of agreement is at least as good for the NMR derived χ(1) values. Notably, as the resolution of the X-ray data improves (rotamer distributions from 3.4 and 2.3 Å X-ray structures are compared with the NMR data), the agreement with the NMR gets significantly better. This emphasizes the importance of NMR approaches for the study of high molecular weight complexes that can be recalcitrant to high resolution X-ray analysis.