Class II Major Histocompatibility Complex (MHC) molecules are membrane glycoproteins that bind peptide fragments derived from exogenous protein sources, including viral and bacterial pathogens, and transport them to the cell surface for recognition by helper T cells (Cresswell, 1994). Unlike class I MHC molecules, class II MHC molecules are found only on a limited number of cell types (Germain, 1994). These specialized antigen-presenting cells express up to three forms, or isotypes, of class II MHC molecules (Korman et al., 1985; Mengle-Gaw & McDevitt, 1985). Each isotype is encoded at a unique genetic locus. At present, structural information on class II MHC molecules is limited to members of one isotype: HLA-DR in humans, and its murine homologue, I-E. However, the strongest genetic associations of class II MHC molecules with autoimmunity have been established for the other class II isotypes: HLA-DQ and its mouse homologue I-A, and an isotype exclusive to humans called HLA-DP (Lafuse, 1991). In particular, the I-A isotype has been implicated in a number of mouse models of autoimmunity, including diabetes (Wicker et al., 1995), rheumatoid arthritis (Chiocchia et al., 1993), and experimental allergic encephalomyelitis (Martin et al., 1992). It is important, therefore, to investigate whether any unique structural features of I-A molecules can be correlated with autoimmune diseases.
Class II MHC molecules are heterodimeric proteins consisting of noncovalently associated alpha and beta chains. The extracellular portion of each chain consists of one-half of a peptide binding site (the alpha1 or beta1 domain) and one Ig-like domain (the alpha2 or beta2 domain). Three putative asparagine-linked glycosylation sites, located on the alpha1 (alpha78), alpha2 (alpha119), and beta1 (beta19) domains, are present in all class II protein sequences. To date, three different structures of class II MHC molecules have been reported for HLA-DR1 (Brown et al., 1993; Jardetzky et al., 1994; Stern et al., 1994), HLA-DR3 (Ghosh et al., 1995), and I-Edk (Fremont et al., 1996). In these human and mouse class II structures, the amino terminal alpha1 and beta1 domains form the characteristic peptide binding groove that was initially described for human (Bjorkman et al., 1992) MHC class I molecules. The peptide-binding site architecture is formed from two long anti-parallel alpha-helical segments that sit on top of, and traverse, an eight-stranded anti-parallel beta sheet. The peptide-binding domain is supported by the membrane proximal Ig-like domains, and is positioned distal from the membrane surface for interaction with its ligand, the T cell receptor. For most class I and class II MHC molecules, specific positions (or anchors) in the bound peptide are conserved (Rammensee et al., 1995), often corresponding to specific pockets in the peptide binding groove of the MHC molecule that accommodate peptide side chains of a certain size and charge (Saper et al., 1991; Matsumura et al., 1992a). In general, class II MHC molecules will bind peptides that have certain residues located within a nine residue "core" motif. Peptides bound to a class II MHC molecule assume a remarkably regular secondary structure conformation that is similar to a polyproline type II ribbon-like helix (Stern etal., 1994; Fremont et al., 1996; Jardetzky et al., 1996; reviewed in Stern and Wiley, 1994; Wilson, 1996). Importantly, a number of conserved side chains, primarily from the two alpha-helices, interact with main-chain atoms of the bound peptide and help orient and fix the peptide in the groove (Stern & Wiley, 1994; Jardetzky et al., 1996), imparting a degree of sequence-independent peptide binding capability to the MHC molecule.
X-ray structural studies of I-A molecules require protein in sufficient quantity and quality for crystallization experiments. Soluble I-Ad for crystallization purposes was produced using a leucine zipper to aid in correct heterodimeric pairing of the alpha- and beta-chains (Scott et al., 1996). These I-Ad molecules could then be loaded with specific peptides. However, purification of these I-Ad-peptide complexes by anion exchange chromatography and hydrophobic interaction chromatography did not yield crystals of X-ray diffraction quality. We report here that tethering of a high affinity ovalbumin peptide to the beta-chain of the MHC molecule, in conjunction with preparative IEF purification and leucine zipper removal, lead to successful crystallization of an I-Ad-peptide complex that diffracts to 2.6 Å. |
peer agreement (net): +2