pocketZebra: a web-server for automated selection and classification of subfamily-specific binding sites by bioinformatic analysis of diverse protein families

Abstract

The new web-server pocketZebra implements the power of bioinformatics and geometry-based structural approaches to identify and rank subfamily-specific binding sites in proteins by functional significance, and select particular positions in the structure that determine selective accommodation of ligands. A new scoring function has been developed to annotate binding sites by the presence of the subfamily-specific positions in diverse protein families. pocketZebra web-server has multiple input modes to meet the needs of users with different experience in bioinformatics. The server provides on-site visualization of the results as well as off-line version of the output in annotated text format and as PyMol sessions ready for structural analysis. pocketZebra can be used to study structure–function relationship and regulation in large protein superfamilies, classify functionally important binding sites and annotate proteins with unknown function. The server can be used to engineer ligand-binding sites and allosteric regulation of enzymes, or implemented in a drug discovery process to search for potential molecular targets and novel selective inhibitors/effectors. The server, documentation and examples are freely available at http://biokinet.belozersky.msu.ru/pocketzebra and there are no login requirements.

Ligand‐based Modeling for the Prediction of Pharmacophore Features for Multi‐targeted Inhibition of the Arachidonic Acid Cascade

“…A functional classification of a protein family is performed by the Zebra application based on the SSPs . This is followed by the classification of subfamily‐specific binding sites by the algorithm Pocketzebra . The latter algorithm utilised the multiple sequence alignment of MAPEG family (Supplementary Figure 1) and the crystal structure of mPGES‐1 to classify this family into two subfamilies: MGST1/FLAP/mPGES‐1 and LTC4S/MGST2/MGST3.…”

“…[38] This is followed by the classification of subfamily-specific binding sites by the algorithm Pocketzebra. [12] The latter algorithm utilised the multiple sequence alignment of MAPEG family ( Supplementary Figure 1) and the crystal structure of mPGES-1 to classify this family into two subfamilies: MGST1/ FLAP/mPGES-1 and LTC4S/MGST2/MGST3. The most significant pocket (POC3) along with the SSPs is displayed for the representative enzymes of subfamilies 1 and 2, namely mPGES-1 and LTC4S, in Figure 3 and all the subfamilyspecific pockets are listed in decreasing order of significance in Supplementary Figure 2.…”

Study of Functional and Allosteric Sites in Protein Superfamilies

“…Not all positions in protein structures are equally susceptible to variation in the course of evolution, reflecting differing selection pressure on amino acids residues with different functional roles. That makes it possible to apply a bioinformatic analysis of protein superfamilies to the study of the evolutionary relationship of amino acid residues in functional and regulatory binding sites [ 39 ] ( Table ). Totally conserved positions play a key role in a function common to all proteins within a superfamily; e.g., they are involved in the enzyme’s catalytic mechanism.…”

“…In this regard, the subfamily- specific positions (SSPs) – conserved within functional subfamilies, but different between them – attract special attention [ 48 , 49 ]. SSPs are observed in both catalytic and allosteric sites, and their presence can be a very powerful factor for the identification of functional and regulatory centers in protein structures [ 39 ]. Identification of statistically significant subfamily- specific positions can help understand the difference in the organization of binding sites within evolutionarily related proteins.…”

Robust enzyme design: Bioinformatic tools for improved protein stability

“…However, because the amount of experimental data is limited, similarity-based clustering strategies have been proposed to automatically predict functional subfamilies in large datasets [67]. SSPs were shown to be responsible for functional discrimination among homologous proteins; they are preferentially associated with catalytic and regulatory sites in enzymes [69]. However exploitation of the SSPs as hotspots for protein engineering has, so far, been very limited.…”