Hydrogen exchange kinetics in a membrane protein determined by nitrogen-15 NMR spectroscopy: use of the INEPT experiment to follow individual amides in detergent-solubilized M13 coat protein

Publisher Website
Google Scholar
Abstract
To coat protein of the filamentous coliphage M13 is a 50-residue polypeptide which spans the inner membrane of the Escherichia coli host upon infection. Amide hydrogen exchange kinetics have been used to probe the structure and dynamics of M13 coat protein which has been solubilized in sodium dodecyl sulfate (SDS) micelles. In a previous 1H nuclear magnetic resonance (NMR) study [O''Neil, J. D. J., and Sykes, B. D. (1988) Biochemistry 27, 2753-2762], multiple exponential analysis of the unresolved amide proton envelope revealed the existence of two slow "kinetic sets" containing a total of about 30 protons. The slower set (15-20 amides) originates from the hydrophobic membrane-spanning region and exchanges at least 105-fold slower than the unstructured, non-H-bonded model polypeptide poly(DL-alanine). Herein we use 15N NMR spectroscopy of biosynthetically labeled coat protein to follow individual, assigned, slowly exchanging amides in or near the hydrophobic segment. The INEPT (insensitive nucleus enhancement by polarization transfer) experiment [Morris, G. A., and Freeman, R. (1979) J. Am. Chem. Soc. 101, 760-762] can be used to transfer magnetization to the 15N nucleus from a coupled proton; when 15N-labeled protonated protein is dissolved in 2H2O, the INEPT signal disappears with time as the amid protons are replaced by solvent deuterons. Amide hydrogen exchanges is catalyzed by both H+ and OH- ions. Base catalysis is significantly more effective, resulting in a characteristic minimum rate in model peptides at pH .simeq. 3. Rate versus pH profiles have been obtained by using the INEPT experiment for the amides of leucine-14, leucine-41, tyrosine-21, tyrosine-24, and valines-29, -30, -31, and -33 in M13 coat protein. The valine residues exchange most slowly and at very similar rates, showing an apparent 106-fold retardation over poly(DL-alanine). A substantial basic shift in the pH of the minimum rate (up to 1.5 pH units) was also observed for some residues. Possible reasons for the shift include accumulation of catalytic H+ ions at the negatively charged micelle surface or destabilization of the negatively charged transition state of the base-catalyzed rection by either charge or hydrophobic effects within the micelle. The time-dependent exchange-out experiment is suitable for slow exchange rates (kex), i.e., less than (1-2) .times. 10-4 s-1. The INEPT experiment was also adapted to measure some of the more rapidly exchanging amides in the coat protein (leucine-14 at higher pH values and glycine-3) using either saturation transfer from water (kex .simeq. 1/T1,amide .simeq. 2 s-1) or exchange effects on the polarization transfer step itself (kex .simeq. JNH .simeq. 100 s-1, where JNH is the 15N-1H coupling constant). Retardations of individual rates with respect to poly(DL-alanine) are analyzed in terms of the local unfolding model for amide exchange. The results of all of these experiments are consistent with previous models of the coat protein in which a stable segment extends from the hydrophobic membrane-spanning region through to the C-terminus, whereas the N-terminal region is undergoing more extensive dynamic fluctuations.