The Nature of The “Groovy” Peptide
The MHC groove, which binds a single peptide, imposes some well‐defined restrictions on the nature and length of the peptide that it can accommodate. However, at the majority of positions in the peptide ligand, a surprising degree of redundancy is permitted and this relates in part to residues interacting with the T‐cell receptor rather than the MHC. Thus, each MHC molecule has the potential to bind hundreds or even thousands of different peptide sequences so long as at certain amino acid positions the peptides share characteristic conserved anchor residues for that particular MHC allele. Different MHC alleles will bind a different range of peptides thanks to the difference in sequence of the binding groove of the different MHC variants.
Binding to MHC class I
X‐ray crystallographic analysis reveals the peptides to be tightly mounted along the length of the groove in an extended configuration with no breathing space for α‐helical structures (Figure 5.19). The molecular forces involved in peptide binding to MHC and in TCR binding to peptide MHC are similar to those seen between antibody and antigen (i.e., noncovalent).
The naturally occurring peptides can be extracted from purified MHC class I and sequenced. They are predominantly 8–10 residues long; because the MHC class I peptide‐binding groove is closed at both ends, any peptides that are slightly longer than this have to bulge upwards out of the cleft. Analysis of the peptide pool sequences indicates amino acids with defined characteristics at certain key positions (Table 5.1). These are called anchor positions and represent the amino acid side‐chains required to fit into allele‐specific pockets in the MHC groove (Figure 5.20a). There are usually two, sometimes three, such major anchor positions for class I‐binding peptides, frequently at peptide positions 2 (P2) and 9 (P9) but sometimes at other positions. For example, the highly prevalent HLA‐A*0201 has a pocket that will accept leucine, methionine, or isoleucine at peptide position P2 and a pocket that will accept leucine, valine, isoleucine, or methionine at P9 (Table 5.1). In some HLA alleles, instead of a major anchor pocket there are two or three more weakly binding pockets. Even with the constraints of two or three anchor motifs, each MHC class I allele can accommodate hundreds or even thousands of different peptides. Thus, so long as the criteria for the anchor positions are met, the other amino acids in the sequence can vary. Allele‐independent hydrogen‐bonding to conserved residues at either end of the MHC class I groove occurs at the N‐ and C‐termini of the peptide.
Except in the case of infection, the natural class I ligands will be self peptides derived from proteins endogenously synthesized by the cell, histones, heat‐shock proteins, enzymes, leader signal sequences, and so on. About 75% of these peptides originate in the cytosol and most of them will be in low abundance, say 100–400 copies per cell. Thus proteins expressed with unusual abundance, such as oncofetal proteins in tumors and viral antigens in infected cells, should be readily detected by resting T‐cells.
Binding to MHC class II
Unlike class I, the class II groove is open at both ends and therefore can bind longer peptides, typically about 15–20 amino acids long. However, just as for class I, it is a stretch of about 9 amino acids that are directly involved in the interaction and this portion is referred to as the peptide binding register. The other amino acids can extend from each end of the groove, quite unlike the strait‐jacket of the class I ligand site (Figure 5.19 and Figure 5.20), and are susceptible to proteolytic trimming. With respect to class II allele‐specific binding pockets for peptide side‐chains, the motifs are based on three or four major anchor residues, typically but not invariably at P1, P4, P6, and P9 (Figure 5.20b).
Unfortunately, it is difficult to establish these preferences for the individual residues within a given peptide. This is because although the length of the class II groove is similar to that of class I, the open nature of the groove in class II places no constraint on the length of the ligand. Thus, each class II molecule binds a collection of peptides of varying length, and analysis of such a naturally occurring pool isolated from the MHC would not establish which amino acid side‐chains were binding preferentially to the nine available sites within the groove. One approach to get around this problem is to study the binding of soluble class II molecules to very large libraries of random‐sequence nonapeptides expressed on the surface of bacteriophages.
Peptide binding leads to a transition from a more open conformation to one with a more compact structure extending throughout the peptide‐binding groove. The range of concentrations of the different peptide complexes that result will engender a hierarchy of dominance of epitopes with respect to their ability to interact with T‐cells.