Combined structure and dynamics study of a WW domain; WW3* domain from human Nedd4‐1

Abstract

Understanding the molecular basis of protein‐protein interactions is a pre‐requisite to the a priori design of therapeutics that combat the majority of human diseases. The focus of this thesis is the interaction between the third WW domain of human neuronal precursor cellexpressed developmentally down‐regulated protein 4 (Nedd4‐1) and the cognate C‐terminal PY motif from the α‐subunit of the human epithelial sodium channel (ENaC). Nedd4‐1 regulates ENaC on the cell surface through ubiquitination and subsequent proteasomal degradation. Mutations in the PY motif leads to impaired ENaC regulation, which leads to sodium imbalance giving rise to an acute form of hypertension known as Liddle’s syndrome. Besides Liddle’s syndrome, impaired WW domain – PY motif interactions are implicated in several other human diseases, including the majority of cancers, Huntington’s disease, Alzheimer’s disease and muscular dystrophy. Determination of the solution structure of the apo‐WW3* domain revealed that it contains a three stranded anti‐parallel beta sheet structure, a characteristic of WW domains. The apo‐ WW3 domain shows an identical fold when compared with the solution structure of the WW3 domain–αENaC peptide complex. Despite the identical fold, differences in side chain orientations were observed for residues in the vicinity of and at the peptide binding interface. Rotameric averaging of the side chains in the apo‐WW3* domain, changes to fixed staggered rotamers for the Cα‐Cβ bond in the peptide‐bound state of the WW3* domain, as indicated by 3J couplings that measure the χ1 torsion angle; thus highlighting the likelihood that conformational fluctuations drive peptide binding. The role of conformational fluctuation is further highlighted in the analysis of dynamics between free‐ and the αENaC peptide bound‐state of the WW3* domain. Both, the free‐ and bound‐state of the WW3* domain showed similar ps‐ns timescale dynamics, as determined by model‐free analysis of laboratory frame 15N relaxation experiments measured at three magnetic fields (14.1, 18.8 and 21.2 T). Here, the values of the square of the generalized order parameter (S2) per residue were very similar between the apo‐ and the peptide‐bound states. The major finding of this analysis was the observation of chemical exchange contributions (Rex) the majority of the core residues in the apo‐WW3* domain. These Rex values were found to be quenched upon binding of the αENaC peptide, with only three residues in the third β‐strand showing minor Rex values. A difference in the internal motions of the amide bond vectors between apo‐ and αENaC peptide‐bound WW3* domain also points towards the role of dynamics in peptide recognition. Molecular dynamics simulations carried out on the apo‐WW3* domain highlight the conformational fluctuations at the ns timescale. Several regions of the WW3* domain sample a wide range of conformations, some of which resemble the conformations of the peptide‐bound WW3* domain. In the apo‐WW3* simulation, the χ1 angle and Cα chemical shift distributions indicate that the side chain of one of the key peptide binding residues, T447, samples conformations observed in structures of both the apo‐ and αENaC peptide‐bound WW3* domain. The length of the loop between the first and the second β‐strands was observed to fluctuate between four and seven residues in the simulation of the apo‐WW3* domain, due to fluctuations of the backbone torsion angles and hydrogen bond network. Such conformational fluctuations were either absent or present to a minor extent in the simulations of the αENaC peptide‐bound WW3* domain. The results of this study indicate that the interaction of the WW3* domain and the αENaC peptide is driven by a conformational selection mechanism. Thus, this study provides an understanding of how dynamics along with structural characteristics are key factors contributing to the high affinity binding of a protein‐peptide complex. Since WW domain–PY motif interactions are involved in the pathogenesis of several human diseases, characterisation of such protein‐peptide interactions should help in the design of more effective diseasecombating drugs.