aus dem Spring, Christian;Schomburg, Dietmar - Sequence families of classified enzymes

ECCB 2002 Poster sorted by: Author | Number

Next | Previous poster (in order of the view you have selected)

Title: Sequence families of classified enzymes - Correlation with the function	P11
aus dem Spring, Christian; Schomburg, Dietmar Christian.aus-dem-Spring@uni-koeln.de, d.schomburg@uni-koeln.de Institute of Biochemistry, University of Cologne, Germany

It is widely accepted that enzymes are adapted from pre-existing enzymes via gene modification rather than invented de novo [1]. This is routinely used in structure and function prediction to transfer information by homology. The underlying hypothesis is that function and structure can be transferred between similar sequences because they have been conserved over long periods of time. This assumption can be confirmed in the case of protein structures but in the case of protein function it is far less justified [2,3]. Previous studies have shown that groups of divergently related enzymes exist whose members seem to catalyze different overall reactions using the functional classification defined by the International Enzyme Commission (EC) as reference [4]. Nevertheless there is considerable evidence that nature has used common binding sites and common mechanistic features to catalyze the analogous reactions with different substrates, and, further, has used common mechanistic features to catalyze different reactions [5]. Therefore the relationship between sequence and reaction was investigated.
We analyzed all 32,945 enzyme sequences labeled with any EC class in recent releases of SWISS-PROT and TrEMBL, evaluating all 1,703 occuring EC classes. An all-vs-all comparison was performed using the standard pairwise alignment method of BLAST. The determined sequence similarities were then used to group the sequences into clusters of related enzymes using the complete linkage clustering method. This method is known to produce very tight clusters because sequences are only added to an existing cluster when they show similarity to all sequences already part of the cluster. This avoids miss-linkage through multi-domain proteins. The resulting classification splits the enzyme space into well defined groups of enzymes, which are closely related to natural groups of biological families. Different indices of validity were applied to access the quality of our classification. A comparison to the classification of the EC systems shows that about 20% of all cluster contain enzymes which are assigned to different EC classes. Examples were analyzed with the result that those divergently related enzymes whose members catalyze different reactions still share a common partial reaction, intermediate, or transition state.

[1] Jensen, R.A. (1976), Enzyme recruitment in evolution of new function, Annu Rev Microbiol. 30:409-25.
[2] Chothia, C., Lesk, A. (1986), The relation between the divergence of sequence and structure in proteins, EMBO J. 5(4):823-6.
[3] Karp, P.D. (1998), What we do not know about sequence analysis and sequence databases,Bioinformatics. 14(9):753-4
[4] Shah, I., Hunter, L. (1997), Predicting enzyme function from sequence: a systematic appraisal, Proc Int Conf Intell Syst Mol Biol.5:276-83.
[5] Babbitt, P.C., Gerlt, J.A. (1997), Understanding enzyme superfamilies. Chemistry As the fundamental determinant in the evolution of new catalytic activities. J Biol Chem. 5;272(49):30591-4.