Molecular modeling of the hypoxia-inducible factor-1 (HIF-1)

Citation
G. Michel et al., Molecular modeling of the hypoxia-inducible factor-1 (HIF-1), THEOR CH AC, 101(1-3), 1999, pp. 51-56
Citations number
27
Categorie Soggetti
Physical Chemistry/Chemical Physics
Journal title
THEORETICAL CHEMISTRY ACCOUNTS
ISSN journal
1432881X → ACNP
Volume
101
Issue
1-3
Year of publication
1999
Pages
51 - 56
Database
ISI
SICI code
1432-881X(199902)101:1-3<51:MMOTHF>2.0.ZU;2-E
Abstract
Hypoxia-inducible factor-1 (HIF-1) is a heterodimeric transcription factor activated by hypoxia. It is composed of two different subunits, HTF-1 alpha and ARNT (aryl receptor nuclear translocator). When activated, HIF-1 media tes the differential expression of genes such as erythropoietin and vascula r endothelial growth factor (VEGF). This work is aimed at defining the 3D s tructure of HIF-1. In this work we use the powerful modeling approach in or der to obtain a first model of HIF-1. The two subunits HIF-l alpha and ARNT belong to the bHLH (basic helix-loop-helix) PAS (per, ahr/arnt, sim) famil y. We focused on the bHLH domain since the protein/DNA interactions are med iated by the basic domains and the dimerization by hydrophobic interactions between the two helices. Firstly, we aligned the sequence of each subunit along with sequences of other bHLH factors, using the program CLUSTALW. Suc h alignments could not detect the sequence corresponding to the bHLH domain . We therefore used the program GIBES, which detects common motifs between sequences of different proteins. We indeed obtained common residues which c ould constitute the bHLH motif. In order to validate the target sequences, we submitted both sequences to a secondary structure prediction algorithm, PHD. Secondly, we made sequence/structure alignments in order to find a tem plate using the BLAST program. We were then able to obtain a structural mod el of the heterodimer by means of HOMOLOGY. Once the four-helix bundle was constructed, we generated both loops using the program SCWRL for adding sid echains to the protein backbone based on the backbone-dependent library. Th irdly, we minimized the model of the heterodimer to avoid steric clashes in its structure, using the program DISCOVER. A combination of tools, such as Ramachandran plots, physicochemical properties, and energetic profiles, en abled us to validate the 3D model. The next step will be to dock the struct ure onto the DNA recognition site. This first study already allowed us to o btain a structural view of HIF-1. It has to be confirmed by X-ray structure analysis of the protein and we will then be able to understand the protein DNA interaction of this transcription factor and its role played in the ce llular response to hypoxia.