Catherine Etchebest

A New Approach to the Rapid Determination of Protein Side Chain Conformations

Two efficient algorithms have been developed which allow amino acid side chain conformations to be optimized rapidly for a given peptide backbone conformation. Both these approaches are based on the assumption that each side chain can be represented by a small number of rotameric states. These states have been obtained by a dynamic cluster analysis of a large data base of known crystallographic structures. Successful applications of these algorithms to the prediction of known protein conformations are presented.

A critical comparison of search algorithms applied to the optimization of protein side-chain conformations

Three different optimization algorithms are applied to solving the problem of finding the best side‐chain conformations with a test set of 14 globular proteins having known crystallographic conformations. It is shown that simulated annealing, simple and modified genetic algorithms, and a heuristic combinatorial approach achieve similar optimal solutions, with the exception of simulated annealing applied to the largest proteins. The efficiency of the different algorithms, however, shows wide variations. General conclusions are drawn concerning the optimal approach to such problems.

The gramicidin A channel

The inclusion of the presence and flexibility of the CH2CH2OH end chain in the computation of the energy profile for single occupancy by Na+ of the gramicidin A channel modifies substantially the profile obtained without that chain. The binding site (deepest minimum) in the profile is situated at 10.5 Å from the center of the channel, in satisfactory agreement with the conclusions based on 13C‐NMR studies. The existence of an external minimum at the mouth is confirmed.

The gramicidin A channel: the energy profile for single and double occupancy in a head-to-head β6.33,3-helical dimer backbone

The energy profile for Na+ in the channel formed by the gramicidin A β‐helical dimer backbone was computed introducing all the terms in the theory of intermolecular interactions. The effect of allowing the ion to reach its successive optimal positions shows the presence of a series of energy minima associated with different carbonyls. The presence of a second ion lowers the central barrier for the first one and facilitates its progression and exit. The energy profile for double occupancy indicates the presence of two symmetrical minima at about 13 Å from the center.

The gramicidin A channel: comparison of the energy profiles of Na+, K+ and Cs+: Influence of the flexibility of the ethanolamine end chain on the profiles

Energy profile Gramicidin A Ethanolamine end Caesium Potassium Sodium Energy barrier Theoretical computation

Prediction of the positioning of the seven transmembrane α-helices of bacteriorhodopsin: A molecular simulation study

We have applied a search strategy for determining the optimal packing of protein secondary structure elements to the rotational positioning of the seven transmembrane helices of bacteriorhodopsin. The search is based on the assumption that the relative orientations of the helices within the bundle are conditioned principally by inter-helix side-chain interactions and that the extra-helical parts of the protein have only a minor influence on the bundle conformation. Our approach performs conformational energy optimization using a predetermined set of side-chain rotamers and appropriate methods for sampling the conformational space of peptide fragments with fixed backbone geometries. The final solution obtained for bacteriorhodopsin places each of the seven helices to a precision of a few degrees in rotation around the helical axis and to a few tenths of an ångström in translation along the helical axis with respect to the best experimental structure obtained by electron diffraction, except for helix D, where our results support the suggestion that this helix should be displaced along its axis toward its N terminus. The perspectives of such an approach for the determination of the structures of other transmembrane helical bundles are discussed.

Structure and dynamics of bacteriorhodopsin Comparison of simulation and experiment

Global features of the structure and dynamics of bacteriorhodopsin are investigated using molecular modelling, dynamical simulations and neutron scattering experiments. The simulations are performed on a model system consisting of one protein molecule plus intrinsic water molecules. The simulation‐derived structure is compared with neutron diffraction data on the location of water and with the available electron microscopy structure of highest resolution. The simulated water geometry is in good accord with the neutron data. The protein structure deviates slightly but significantly from the experiment. The low‐frequency vibrational frequency distribution of a low‐hydration purple membrane is derived from inelastic neutron scattering data and compared with the corresponding simulation‐derived quantity.

Rotational orientation of transmembrane α-helices in bacteriorhodopsin: A neutron diffraction study

The rotational orientation of the seven transmembrane alpha-helices (A-G) in bacteriorhodopsin has been investigated by neutron diffraction. The current model of bacteriorhodopsin is based on an electron density map obtained by high-resolution electron microscopy (EM). Assigning helix rotational positions in the EM model depended on fitting large side-chains, mainly aromatic residues, into bulges in the electron density map. For helix D, which contains no aromatic residues, the EM map is more difficult to interpret. For helices A and B, whose position and orientation had been determined previously by neutron diffraction, the positions defined by EM agree within experimental error with these earlier conclusions. The orientation of all seven helices has been examined by using neutron diffraction on bacteriorhodopsin samples with specifically deuterated valine, leucine and tryptophan residues. Experimental peak intensities were compared to those predicted for an extensive set of structural models. The models were generated by (1) rotating all helices around their axis; (2) moving deuterated residues in the extramembrane loops about their probable positions and changing the weight of their contribution to the neutron diffraction pattern; (3) allowing deuterated side-chains to change their conformation. The analysis confirmed exactly the positions previously determined for helices A and B. For an optimal fit to the data to be obtained, the other five helices, including helix D, must lie either at or within 20 degrees of their position in the current EM model. The complementarity of medium-resolution EM, neutron diffraction and model building for the structural study of integral membrane proteins is discussed.

The gramicidin A channel: The energy profile calculated for Na+ in the presence of water with inclusion of the flexibility of the ethanolamine tail

The effect of the conformational freedom of the ethanolamine tail of gramicidin A on the energy profile for the transfer of Na+, computed in the presence of water, shows an appreciable lowering of the minimum at 10.5 Å, and a splitting of the entrance barrier. The deep energy region at the channel mouth remains, however, the deepest one and contains a site of strong interaction of the ion with the Trp 11 carbonyl, at about 12 Å from the center.

Energy Profile of Cs+ in Gramicidin A in the Presence of Water. Problem of the Ion Selectivity of the Channel

The effect of water present at the mouth and inside the channel of Gramicidin A on the energy profile calculated for a caesium ion is determined. The total optimal interaction energy computed for the system GA-Cs+-(22 waters) leads to an energy profile characterized by a deep minimum at 11 A followed by an entrance energy barrier of 7 Kcal/mol expanding until 9 A from the center. After this point, a second minimum less deep than the previous one is observed, itself followed by a central barrier. The shape of the profile at the entrance is governed by the balance between the progressive desolvation process of the ion and the increase of favorable hydrogen bond interactions implying both the water molecules and GA. The comparison of this energy profile with that obtained in vacuo shows that the presence of water molecules does not modify the pathway of the ion which, owing to its size, is constrained essentially to remain on the channel axis. The comparison Na+ versus Cs+ indicates that although the phenomena involved are globally the same, differences between the two profiles appear due firstly to the difference in the affinity of the two ions for water and secondly to their respective size. This last difference implies that the number of water molecules present in the interior of the channel during the cation progression is reduced roughly by one in the case of caesium. The desolvation barrier computed for Cs+ is half the corresponding value for Na+, a result in agreement with the observed selectivity.