Vorhersage struktureller und biochemischer Eigenschaften von Cytochrom P450-Monooxygenasen und Laccasen

In the following update procedure a new version with a refined classification was realized and, besides integrating new sequences and structures, the CYPED was extended by biochemical properties, and by adding new functionalities. The amount of sequences and structures almost doubled. The CYPED entries could be linked to the entries of the CPK (Cytochrome P450 Knowledgebase, (Lisitsa et al., 2001)) and therefore data on biochemical properties, like substrates or inbibitors, could be provided. For this purpose a new algorithm was implemented, where the linkage of the entries could be realized by their high sequence similarity instead of a unique primary key. Thus, it was possible to link sequences which belong to the same protein entries although not being absolutely identical. Furthermore, information on mutations and CYP polymorphism was added. All data on allelic variants and their effects were extracted from the Home Page of Human Cytochrome P450 (CYP) Allele Nomenclature Committee (Oscarson und Ingelman-Sundberg, 2002) and listed on the respective protein entries. Since all CYPs share a common conserved fold, it was possible to derive a profile based on a structural alignment. With this profile all highly conserved secondary structure elements could be predicted and annotated among the CYPED entries, hence allowing a structural navigation in sequences without structural information. The structured organization of the sequence information on cytochrome P450 monooxygenases allows the systematic comparison of all sequences in order to identify conserved regions which are essential for structure and function and variable regions responsible for selectivity and interaction with redox partners. Unfortunately, structural information is limited to only 5 % of all CYPs. This makes it extremely difficult to draw conclusions on function and mechanism of the proteins. However, despite the profoundly low sequence similarity, the structures of all CYPs are preserved highly similar. Therefore, an alignment which was based on a superposition of all available CYP structures could be generated, which led to an HMM profile of conserved and variable regions. An automated script was developed which is able to map each CYPED sequence on the alignment and consequently predict secondary structures based on the provided structural information. Thus, it was possible to predict and annotate the conserved secondary structures among all CYPED sequences in an automated procedure. The results were published in the current online version of the database. Additionally, a web interface was implemented, which can be applied to perform such a prediction for any entered CYP sequences or similar sequences with the same fold. Furthermore, the structural prediction allowed the analysis of functionally relevant positions and sites in highly variable regions. Thus, a