Book of Abstracts: Albany 2007
June 19-23 2007
In silico Prediction and Analysis of Substrate Specificity for Acyl Adenylate and Glycosyltransferase Superfamily of Enzymes
Acyl CoA synthetase (ACS) and glycosyltransferase (GTrs) superfamily of enzymes are involved in the biosynthesis of various biologically and industrially important natural products. The enzymes of these two superfamilies activate a wide variety of substrates and play a major role in increasing the structural as well as functional diversity of various secondary metabolites in microbes and plants (1). Based on their substrate preference they can be further classified into various subsfamilies. A large number of sequences belonging to these superfamilies have been deposited in public databases, without any functional annotation. The identification of function at the level of substrate specificity (subfamily classification) is essential for identifying novel biochemical/metabolic pathways in genomes (2). In this work, we have used an in silico approach to investigate the relationship between sequence and substrate specificity for these two groups of enzymes.
ACS or AMP-forming superfamily catalyze transfer of a wide variety of acyl, aryl, and aminoacyl moieties from their corresponding acids to the phosphopantetheine group of CoA or other carrier proteins through formation of a thioester bond. Due to low overall sequence similarity within the superfamily, it is difficult to ascertain their substrate preference. Recently, several examples have been reported where the enzyme shows high sequence similarity to one subfamily whereas experimental studies indicate that they utilize substrate of a different subfamily (3). We have used a knowledge-based approach which involves compilation of substrate specificity information for various experimentally characterized ACS and derivation of profile HMMs for each subfamily. Benchmarking on a test set of ACS sequences indicate that these HMM profiles can accurately differentiate probable substrates from non-substrates (sensitivity = 0.91 to 1, specificity = 0.96 to 1). Using homologous crystal structures, we also identified a limited number of contact residues crucial for substrate specificity determination, i.e., specificity determining residues (SDRs). Patterns of SDRs from different subfamilies have been used to derive predictive rules for correlating them to substrate preference (sensitivity = 0.83 to 1, specificity = 0.96 to 1). The changes in SDRs can be directly correlated to changes in substrate specificity; thus, the substrate preference of the uncharacterized enzymes can be predicted by identifying putative SDRs. The power of the SDR approach was demonstrated by correct prediction of substrate in the above mentioned ambiguous cases. The prediction protocol when used on NR database helped to identify new members of the superfamily and also assigned substrate preference to uncharacterized sequences. The prediction tool is available at http://linux1.nii.res.in/~pankaj/poss.html as a web server. Furthermore, molecular modeling of the substrate in the active site has been carried out to understand the structural basis of substrate preference. Such information about SDRs can potentially help in rational design of novel proteins with altered specificities by site directed mutagenesis.
GTrs are involved in transferring various types of sugar groups from an activated donor to proteins, nucleic acids, carbohydrates, lipids and also secondary metabolites. As for ACS, comprehensive analysis of sequence/structural features was carried out for antibiotic GTrs (4). Using a knowledge-based approach similar to that used for ACS analysis, a computational protocol SEARCHGTr was developed for correlating sequences of GTrs to the chemical structures of their corresponding substrates. This software (http://www.nii.res.in/searchgtr.html) (4) predicts the donor/acceptor specificity and also identifies putative substrate binding residues. Benchmarking on a set of GTrs of known substrate specificity indicate that the program can predict the acceptor specificity of antibiotic GTrs with an accuracy of 77%.
References and Footnotes
Pankaj Kamra Khurana*
National Institute of Immunology