3D Structural Models of Transmembrane Proteins
Transmembrane proteins are macromolecules implicated in major biological processes and diseases. Because of their specific neighborhood, few transmembrane protein structures are currently available. The building of structural models of transmembrane prote
- PDF / 1,867,350 Bytes
- 15 Pages / 504 x 720 pts Page_size
- 84 Downloads / 181 Views
troduction Transmembrane proteins represent ~25% of proteins coded by genomes. They are composed of two major classes: all-a, e.g., rhodopsin and all-b, e.g., outer membrane proteins. They support essential biological functions as receptors, transporters, or channels. They are embedded in lipid membrane that constitutes a very specific neighborhood. As a result of this particularity, obtaining experimental 3D transmembrane structures is difficult. The total number of transmembrane proteins in the Protein DataBank (1) is limited, comprising ~1% of available structures (2). The design of structural models becomes an important axis of research. Indeed, more than two-thirds of the marketed drugs target a transmembrane protein and 50%, specifically a GPCR (3). Jean-Jacques Lacapère (ed.), Membrane Protein Structure Determination: Methods and Protocols, Methods in Molecular Biology, vol. 654, DOI 10.1007/978-1-60761-762-4_20, © Springer Science+Business Media, LLC 2010
387
388
de Brevern
Thus, most of the time it is not possible to work with an experimental structure, and so, the 3D structural model is one of the important research fields for understanding biological mechanisms and interactions (4). We present in this chapter the classical pipeline to build 3D structural models. The most common way to propose 3D structural models is on the basis of a comparative modeling process coupled with transmembrane segment predictions. Nonetheless, the principle goes far beyond the classical homology modeling as often the target structure is not directly related to the query sequence, i.e., it is not possible to simply align the sequence of the protein queries and targets. Thus, it is an iterative process mainly based – when possible – on multiple sequence alignments, bibliographic and web researches, molecular refinement, and helix–helix and helix–lipid interaction prediction tools. Two papers can be read to have pertinent examples of such an approach (5, 6). The chosen structural models must encompass most of the biochemical features and reflect the known experimental data. They may be used to analyze functional interaction properties.
2. Materials A recent computer with an internet connection is sufficient. Numerous software programs are available on-line, but some must be locally installed. Most of these software programs can be used with the Windows Operating System (OS) while a few work only with Linux OS. Tables 1–3 summarize some of the available tools that can be used, e.g., the secondary structure and interaction prediction methods. Updated and additional links can be found at http://www. dsimb.inserm.fr/~debrevern/TM/index.html.
3. Methods The main principle of the approach is described in Fig. 1. The first step consists in the true knowledge of the protein of interest (POI). Three complementary approaches are important: multiple sequence alignment, bibliographic and on-line research, and secondary structure prediction. Then, the alignment of the POI (query) sequence with a (target) sequence of a protein with an
Data Loading...
