Virtual models of the HLA class I antigen processing pathway Nikolai Petrovsky a,b, * , Vladimir Brusic b,c a Autoimmunity Research Unit, The Canberra Hospital, ACT 2606, Australia b Medical Informatics Centre, University of Canberra, ACT 2601, Australia c Institute for Infocomm Research, Singapore 119613, Singapore Accepted 21 June 2004 Abstract Antigen recognition by cytotoxic CD8 T cells is dependent upon a number of critical steps in MHC class I antigen processing including proteosomal cleavage, TAP transport into the endoplasmic reticulum, and MHC class I binding. Based on extensive exper- imental data relating to each of these steps there is now the capacity to model individual antigen processing steps with a high degree of accuracy. This paper demonstrates the potential to bring together models of individual antigen processing steps, for example pro- teosome cleavage, TAP transport, and MHC binding, to build highly informative models of functional pathways. In particular, we demonstrate how an artificial neural network model of TAP transport was used to mine a HLA-binding database so as to identify HLA-binding peptides transported by TAP. This integrated model of antigen processing provided the unique insight that HLA class I alleles apparently constitute two separate classes: those that are TAP-efficient for peptide loading (HLA-B27, -A3, and -A24) and those that are TAP-inefficient (HLA-A2, -B7, and -B8). Hence, using this integrated model we were able to generate novel hypoth- eses regarding antigen processing, and these hypotheses are now capable of being tested experimentally. This model confirms the feasibility of constructing a virtual immune system, whereby each additional step in antigen processing is incorporated into a single modular model. Accurate models of antigen processing have implications for the study of basic immunology as well as for the design of peptide-based vaccines and other immunotherapies. Ó 2004 Elsevier Inc. All rights reserved. Keywords: Transporter associated with antigen processing; Major histocompatibility complex; Human leukocyte antigen; Artificial neural networks 1. Introduction Cytotoxic T lymphocytes (T cells) respond to and de- stroy infected (e.g., viruses, bacteria, parasites, or fungi), mutated (e.g cancerous), or foreign (e.g., transplanted) tissue through the highly specific recognition of major histocompatibility complex (MHC)-presented peptides on the surface of host cells. Intracellular antigen pro- cessing pathways determine which peptides are available for binding to MHC class I molecules and are thereby important in shaping immune responses to particular proteins. MHC class I antigen presentation involves multiple steps—proteasome cleavage, transport into the endoplasmic reticulum (ER), binding to MHC mol- ecules, and transport of peptide/MHC complexes to the cell surface for recognition by T cells (Fig. 1). The first three steps involve protein cleavage and peptide binding and are consequently highly peptide sequence specific. Proteasomes regulate a range of important cellular processes through their ability to selectively degrade cytosolic proteins [1]. The proteasomal pathway is a ma- jor source of peptides presented by the MHC class I route as shown by the fact that impairment of proteos- omal cleavage leads to greatly reduced MHC class I pre- sentation [2]. The core of the proteasome complex, human 20S proteasome, consists of 28 subunits and has considerable structural heterogeneity [3]. The 20S proteasome can degrade proteins alone, or in associa- 1046-2023/$ - see front matter Ó 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.ymeth.2004.06.005 * Corresponding author. Fax: +61-2-62603372. E-mail address: nikolai.petrovsky@anu.edu.au (N. Petrovsky). www.elsevier.com/locate/ymeth Methods 34 (2004) 429–435