The HLA–B27 peptidome: Building on the cornerstone

Abstract

The pathogenetic role of HLA–B27 in the spondylarthritides (SpA) is an elusive problem that remains unsolved 37 years after the discovery of the extraordinary association of this antigen with ankylosing spondylitis (AS) (1,2) and almost 20 years after the establishment of a transgenic animal model (3) in which the direct involvement of B27 in the pathogenesis of these diseases was demonstrated. In the absence of definitive evidence for the mechanism of this association, 3 major hypotheses are currently the focus of intense debate. First is the arthritogenic peptide hypothesis (4), which is based on the canonical antigen-presenting properties of B27. This hypothesis assumes that there is an autoimmune mechanism involving specific T cell recognition and molecular mimicry between microbial and selfantigens presented by this molecule. Second is the misfolding hypothesis (5), which proposes an autoinflammatory mechanism (independent of antigen presentation) following activation of the unfolded protein response upon accumulation of misfolded B27 heavy chain in the endoplasmic reticulum. Third is the surface homodimer hypothesis (6), which is based on the expression of covalent B27 heavy-chain homodimers at the cell surface that can be recognized by leukocyte receptors, providing the basis for an immunomodulatory role of noncanonical B27 forms with potentially pathologic consequences (7). Recently, we proposed that the pathogenesis of SpA might be better explained not on the basis of just a particular feature of B27, but from a comprehensive consideration of its whole biology, which is determined by the interdependency of multiple molecular properties (8). In adopting this viewpoint, peptide binding is recognized as an essential feature of HLA–B27, largely determining its immunologic specificity, as well as its folding and stability, which are critical to the formation of heavy-chain homodimers. Thus, peptides emerge as the cornerstone of HLA–B27 biology and pathophysiology (8). This idea might be supported by the identification of endoplasmic reticulum aminopeptidase 1 (ERAP-1), which is involved in the processing of HLA class I ligands, as a susceptibility factor for AS (9). Therefore, the characterization of the B27 peptidome becomes a critical issue in our search for an understanding of the pathogenesis of AS. What can be expected from a knowledge of the B27-bound peptide repertoire that may be relevant to the pathogenesis of SpA? It would be naive to assume that a purely proteomic approach may lead to a straightforward identification of putative arthritogenic peptides because, among other reasons, we ignore the features these peptides might have or the proteins from which they arise. However, other basic issues that are critical to a proper assessment of the very concept of arthritogenic peptide and to its potential features can be addressed. Some of these issues are as follows: 1) the accurate establishment of the B27 binding motif and its variation as a function of peptide size, which would improve the prediction of B27 epitopes and help in their identification as naturally processed antigens; 2) the influence of the binding motif, and deviations therefrom, on the molecular stability of B27; 3) the characterization of peptide features that influence the binding mode and presentation to T cells, as shown for Arg5 residues among B*2705 ligands (10); 4) the identification of the parental proteins of B27-bound peptides, including tissue-specific ones, that might be targets of site-specific autoimmune attack; 5) the identification of self-derived B27 ligands with similarity to microbial epitopes, which may provide a basis for molecular/antigenic mimicry; 6) an estimation of the frequency of such peptides, which would inform us of the likelihood that molecular mimSupported by grant SAF2008/00461 from the Ministry of Science and Innovation and by an institutional grant from the Fundacion Ramon Areces to the Centro de Biologia Molecular Severo Ochoa. Jose A. Lopez de Castro, PhD: Centro de Biologia Molecular Severo Ochoa, Consejo Superior de Investigaciones Cientificas, and Universidad Autonoma de Madrid, Madrid, Spain. Address correspondence and reprint requests to Jose A. Lopez de Castro, PhD, Centro de Biologia Molecular Severo Ochoa, Calle Nicolas Cabrera No. 1, Universidad Autonoma, 28049 Madrid, Spain. E-mail: aldecastro@cbm.uam.es. Submitted for publication August 25, 2009; accepted in revised form October 2, 2009.