Proceedings of The Physiological Society
University of Oxford (2011) Proc Physiol Soc 23, C63
Thermodynamic mutant cycles suggest no interface separation during channel gating at CFTR's non-canonical ATP binding site
A. Szollosi2, D. R. Muallem1, L. Csanády2, P. Vergani1
1. Neuroscience, Physiology and Pharmacology, University College London, London, United Kingdom. 2. Medical Biochemistry, Semmelweis University, Budapest, Hungary.
CFTR is an anion selective channel that plays an essential role in epithelial physiology. It is a member of the ATP Binding Cassette transporter superfamily. A conserved mechanism involving ATP binding and hydrolysis at cytosolic nucleotide binding domains (NBDs) drives both gating of the CFTR channel and transmembrane substrate translocation by transporters. NBDs can dimerize, and two composite ATP binding sites are formed at the interface. In CFTR and its closest relatives (medically relevant sulphonylurea receptors and multi drug resistance-associated proteins) only one of the two composite sites is catalytically active, while the other, site 1, is non-canonical, and binds ATP tightly but does not hydrolyze it. At site 2, ATP binding and hydrolysis have been linked to localized closing/dehydration and opening/hydration, respectively, of the NBD dimer interface. In turn these are coupled to CFTR channel opening and closing. However, little is known about movements around composite site 1 associated with gating. We used thermodynamic mutant cycles  to study the dynamics of the NBD dimer interface around composite site 1. T460 is a conserved residue in the NBD1 face of site 1. In homology models of CFTR, three residues (H1348, L1353 and H1375) on the NBD2 face of site 1 are close enough to interact with T460. In addition, the three positions yielded high correlations scores with T460, when multiple sequence alignments were analyzed to identify coevolving positions [2, 3]. We studied hydrolyitic and non-hydrolytic channel gating in single mutants (T460S, L1353M, H1348A, H1375A) and double mutants (T460S/L1353M, T460S/H1348A, T460S/H1375A). Mutation T460S caused a small acceleration of hydrolytic and non-hydrolytic closure. Mutation L1353M had very little effect. Mutations H1348A and H1375A dramatically slowed hydrolytic closing, H1348A also in a non-hydrolytic background. In the corresponding double mutants changes due to mutations at the NBD2 sites proved mostly additive with those caused by mutation T460S. Overall, changes in coupling energy (ΔΔGint) were either negligible or very small (ΔΔGint = 0.63±0.19 kT, n=8, p=0.007 in hydrolytic closing for T460S-L1353M pair; ΔΔGint= 0.43±0.14 kT, n=6, p= 0.01 in non-hydrolytic closing for T460-H1348 pair; mean±SEM, significance by t-test). It is likely (as suggested by homology models and coevolution analysis) that at least one of the residues tested here interacts with T460. If two residues interact, small or non-significant changes in energetic coupling throughout the gating cycle signify only limited relative movement. Therefore these results support a gating model in which ATP-bound composite site 1 remains closed throughout the gating cycle.
Where applicable, experiments conform with Society ethical requirements