iFace : A Bioinformatics Tool for the Analysis of Protein-Protein Interface

Received Jan 1 st , 2012 Revised Feb 5 th , 2012 Accepted Feb 8 th , 2012 Detailed knowledge of protein-protein interaction is essential to understand various biochemical and biological functions. In this paper, we present a bioinformatics tool to analyze the protein-protein interfaces using threedimensional structural information. iFace identifies protein-protein interaction sites and various interactions that contribute to the specificity and strength of the protein complex.


INTRODUCTION
Communication between proteins is vital for most biological process such as signaling pathways, respiration, enzyme regulation, receptor binding, the immune response etc. [1][2][3].Protein-protein interaction is one of the important mechanisms that establish protein-protein communication in cellular responses at various levels.Therefore, detailed knowledge of protein-protein interaction is essential for the better understanding of biological pathways and their impact on various diseases [4][5][6].
Although there have been many detailed studies of protein-protein interaction, it still remains as one of the most challenging problem in bioinformatics and molecular biology.Most of the available tools are restricted to the identification of interface residues and possible interactions at the interface.The detailed knowledge of interaction strength and the role of neighboring residues of each interface residue is important to understand the protein-protein mechanism.In this paper, we present a computational tool, iFace, which identifies interface residues, various interactions at the interface, estimated strength of the interaction and spatial neighbors of each interface residue.

Steps involved in iFace analysis
Given the coordinate set of 3D structure of a protein complex, iFace identifies protein-protein interaction sites, spatial neighbors of interface residues and computes various interactions such as hydrophobic interaction, hydrogen bonds, disulfide bonds and salt bridges.Various steps involved in iFace analysis are shown in Figure 1.

Identification of interface residues
For each protein complex, we identify interface residues and their spatial neighbors.Interface residues are defined as in Tsai et al. [25].Briefly, two residues are considered to be in contact across the interface if there is at least a pair of atoms, one from each residue, at a distance smaller than the sum of their vander Waals radii plus a threshold of 0.5 Angstrom

Identification of spatially neighboring residues
Spatially neighboring residues serve both structural and functional roles in proteins.Spatially neighboring residues were shown to have positive influence in identification of critical sites in proteins [26][27][28].iFace provides a list of spatially neighboring residues for each interface residue.The residues whose Cβ atoms are found within 5 Å distance from the Cβ of an interface residue are considered as spatial neighbors of the interface residue.

Identification of interactions at the interface
iFace program identifies the following interactions at the interface: hydrophobic interactions, hydrogen bonds, disulfide bonds and salt bridges.Interaction between hydrophobic side chains are identified using a distance cutoff of 5 Angstrom between apolar groups in the apolar side chains [18].The hydrogen bonds formed between subunits are identified using HBOND program which is a part of JOY suite [29].The hydrogen bonds are categorized into four classes: (i) main chain to main chain (MM); (ii) side chain to main chain amide (SN); IJCB Vol. 1, No. 1, 2012, 37 -42 http://www.ijcb.in(iii) side chain to main chain carbonyl (SO); (iv) side chain to side chain (SS).Disulphide bonds are recognized using the distance criteria employed originally in the MODIP program [30].

Strength of interaction
The strength of interaction is estimated as in Biro 2005 [31].Biro constructed amino acid interaction matrices to characterize the expected strength of interaction of two amino acids using three major physicochemical properties: size, charge and hydrophobicity [31].Each matrix contains 20 × 20 values for 20 amino acids and each value ranges from 1 to 20, where 1 is the lowest and 20 is the highest probability that two amino acids will interact with each other on the basis of a given physico-chemical property.We used hydropathy compatibility index (HCI) and charge compatibility index (CCI) to characterize the strength of interaction between two interface residues.HCI and CCI are calculated using the following formulas.where pI(A) and pI(B) are the isoelectric points of the amino acids A and B. This formula gives an index between 1 and 20.The lowest index indicates the lowest possible attraction between amino acids (Asp-Asp) while the highest index indicates the highest possible attraction between amino acids (Arg-Asp).

Input and output
Running iFace program is a straight forward procedure.It accepts input as a protein complex in PDB format.Prior to prediction, structures undergo a set of quality checks.If atoms present alternative locations or rotamers, only the first occurring rotamer is kept.If the PDB structure is NMR structure, the first model is considered for the prediction.However, iFace provides option to analyze the interface for all the models.iFace provides output in a convenient text format which can be parsed by simple scripts.For each complex, iFace provides the following information.a) List of interface residues b) List of spatially neighboring residues c) Residues that participate in the interface interaction d) Interaction type e) Estimated strength of interaction in terms of HCI and CCI

An example: Bovine seminal ribonuclease
Table 1 shows an analysis result for bovine seminal ribonuclease (PDB code 11ba) [32].This protein contains two identical subunits and each subunit has 124 amino acids.iFace identified 38 interface residues from chain A and 41 interface residues from chain B. As shown Table 1, the protein complex is stabilized by 8 hydrophobic interactions and 32 hydrogen bonds.In addition, iFace identified two disfulfide bonds at the interface.Hbond-MM: main chain to main chain (MM); hbond-SN: side chain to main chain amide; hbond-SO: side chain to main chain carbonyl; hbond-SS: side chain to side chain; HCI-hydropathy compatibility index; CCI -charge compatibility index