Docking of transmembrane helices into four helix bundles in the high affinity IgE receptor

The high affinity IgE receptor (Fc-epsilon-RI) consists of one alpha, one beta and two identical gamma subunits and it contains 7 transmembrane (TM) helices. Since a direct experimental structure determination of this intact receptor would be complex, a combination of computational chemistry with experimental work can be used to elucidate the 3D structure. The basic structural building block in plasma membrane proteins of both prokaryotic and eukaryotic cells is the apolar, often slightly amphipathic, transmembrane alpha helix. A major problem in membrane protein structure is that the seven transmembrane helices can adopt many different spatial arrangements. Helix-helix packing plays an important role in the formation of TM helix bundles, as well as in the folding of the whole receptor. The most suitable helix-helix packing arrangements were usually found using rules derived from soluble helical proteins and by rotating the helices such that the most hydrophobic sides would face the lipids. ((1) under bar) Molecular modelling and mutagenesis results ((2) under bar), ((3) under bar), ((4) under bar) have suggested that the amino acid sequence of TM helices contains in-formation. which directs specific helix-helix interactions. In this work, spatial arrangement of the transmembrane bundle were studied, as well as, helix - helix interactions in the high affinity IgE receptor, using a nonsubjective procedure, namely a low resolution docking procedure. This procedure created 35 possible four helix bundles, that were divided into three categories: "clockwise", "anticlockwise" and "crossed-loop" arrangements. "Crossed-loop" bundles were discarded due to distance constraints imposed by connecting loops. Those arrangements are also statistically less favourable. Two four-helix bundles were chosen based upon the following criteria: agreement with pro-nosed topology, correct TM helix arrangement, high interaction energy between TM helices, and appropriate orientation of the hydrophobic moments towards lipid bilayer. Our results are in good agreement with lipid binding sites which were predicted using molecular mechanics. ((6) under bar), ((7) under bar) As in any purely modelling study, the result is not a replacement for an experimental high resolution three-dimensional structure; however, the insights obtained suggest a number of avenues for further study and form a framework for rational strategies for site-directed mutagenesis, drug binding and second messenger events.