Structural Immunoinformatics - Immunoinformatics - page 50

Chapter 3
Structural Immunoinformatics
Shoba Ranganathan, 1,3 Joo Chuan Tong, 2 and Tin Wee Tan 3
1 Department of Chemistry and Biomolecular Sciences & Biotechnology Research Institute,
Macquarie University, Sydney, NSW 2109, Australia, [e-mail]
2 Institute for Infocomm Research, 21 Heng Mui Keng Terrace, Singapore 119613,
[e-mail]
3 Department of Biochemistry, Yong Loo Lin School of Medicine, National University of
Singapore, 8 Medical Drive, MD7 #02-03, Singapore 117597, [e-mail]
Abstract. Normal adaptive immune responses operate under major histocompatibility
complex (MHC) restriction by binding to specific short antigenic peptides. Sequence-
structure-function information is critical in facilitating the understanding of principles
governing MHC-specific peptide recognition and binding. Three-dimensional structures of
bound peptide ligands to MHC receptors are today characterized in great number using X-ray
crystallography, offering a rich source of information for structural analysis. By utilizing
information derived from available experimental structures, it is possible to predict binders for
alleles that have not been studied extensively and offers an alternative to sequence-based
approaches that require a large dataset for training. This chapter will introduce the use of
structural descriptors, as well as comparative modeling and docking techniques for predicting
whether a peptide sequence can bind to a specific MHC allele.
3.1 Introduction
The binding of peptide ligands to MHC molecules plays a key role in the activation
of normal adaptive immune responses and an intricate theoretical problem that
remains unsolved. For an MHC molecule to recognize antigenic peptides, geometric
and electrostatic complementarity between the receptor and ligand is essential for the
formation of chemical bonds between their functional groups, which in turn
determines the stability of the complex. In this context, the introduction of structural
information can greatly facilitate our understanding of how well a peptide ligand can
associate with an MHC molecule.
In recent years, an increasing number of protein structures have been
experimentally determined and deposited in the Protein Data Bank (PDB; Bernstein,
Koetzle, Williams, Meyer, Brice, Rodgers, Kennard, Shimanouchi, and Tasumi
1977), providing a wealth of information for structural analysis and prediction.
Together with the development of new structural modeling and docking techniques,
the use of structure-based approaches to predict potential T-cell epitopes is
increasingly successful, often producing modeled structures accurate to within 2.00Å