Two-dimensional 1H nuclear magnetic resonance study of pike pI 5.0 parvalbumin ( Esox lucius) : Sequential resonance assignments and folding of the polypeptide chain

Abstract

The structure of α pike 5.0 parvalbumin under its Ca-loaded form (or PaCa 2) is studied in solution by two-dimensional 1H nuclear magnetic resonance (n.m.r.) at 360 MHz using a conventional strategy of sequential assignments, which involved correlated spectroscopy, relayed coherence transfer spectroscopy and nuclear Overhauser enhancement spectroscopy. In order to overcome the problem of spectral overlapping due to the presence of 108 residues in the protein, experiments were performed at different pH and temperature values, either in 1H 2O or in 2H 2O solutions. The amino acid sequence of pike 5.0 parvalbumin is thus fully characterized by nearly the totality of its NH, C αH and C βH resonances originating from the different residues (421 protons assigned among 429 in total). When associated with the remaining side resonances, these sequence-specific assignments provide a basis for establishing the secondary organization and tertiary folding of the polypeptide chain. Pike 5.0 parvalbumin was selected as a characteristic representative of the α phylogenic series, for which no crystalline structure is presently available, in contrast with the β series for which two crystalline structures have been determined. A parvalbumin molecule with a single polypeptide chain of 108 amino acids represents one of the highest molecular weights analyzed so far by two-dimensional n.m.r. spectroscopy. The use of a moderate magnetic field strength, with 1H nuclei resonating at 360 MHz, is justified by the fact that ring current effects are operating favorably in this globular protein with a high phenylalanine content. A three-dimensional structure has been generated by the “distance geometry” or DISGEO computational procedure on the basis of about 450 interproton nuclear Overhauser enhancement connectivities (short, medium and long-range) in conjunction with a selection of phi and chi dihedral angle constraints. The coherence of the calculated structure, which displays all the features of the typical folding of a parvalbumin protein, provides a good test of reliability of the n.m.r. data collected so far. Although similar to a β parvalbumin in the folding of its polypeptide chain, the α parvalbumin studied here differs markedly from a β parvalbumin in the length of its C-terminal F-helix domain, which includes 11 residues instead of ten in the latter. As suggested by the comparison of parvalbumin primary structures, the lengthening of the F-helix is likely to be a general feature of α parvalbumins, in relation to their enhanced conformational stability.