The Secrets of the Class II MHC Peptidome Start To Be Revealed

The paper by Rudensky et al. (1) from Charles Janeway’s laboratory, selected as a Pillars article, represents the first attempt to examine the sequences of natural peptides extracted from class II MHC (MHC II) molecules. This study was carried out in 1991, prior to Donald Hunt’s introduction of sophisticated mass spectrometry approaches to isolate and sequence large numbers of MHC-bound peptides (2, 3). Rudensky et al. examined a series of peptides extracted from a B cell line that expressed both H2^b and H2^k MHC alleles. They isolated the MHC II molecules, released the peptides, fractionated the extract by reverse phase HPLC, and then performed sequence analysis of the various fractions. Although only a small number of peptides were identified, nevertheless important information was obtained. Their findings were confirmed and amplified as the MHC II peptidomes were examined subsequently in various other laboratories, but this time by mass spectrometry technology.

What was the state of the field of Ag presentation in 1991? New information had been forthcoming on the nature of T cell recognition of Ags. We already knew that Ag processing was required for generation of peptides that bound to MHC proteins, and that the peptide–MHC complex was the substrate of T cell recognition. Moreover, initial information was already available on features of peptides bound to class I MHC (MHC I) and MHC II. Townsend et al. (4) had published their seminal data on the peptides derived from the influenza nucleoprotein, which were extended later by others to show that MHC I peptides were of limited size, between 8 and 10 residues (5–7). Of note is the pioneering work from Rammensee’s group, in which they extracted MHC I peptides from cell lines, did bulk sequence analysis, and obtained information on preferred residues at certain amino acid positions in the short peptides (8–10). Concerning MHC II peptides, with Paul Allen, we had obtained proof that a dominant hen egg-white lysozyme (HEL) peptide preferentially bound to I-A^k (11). In contrast to the data that subsequently emerged from MHC I studies, the HEL peptide sequence was much longer, 16 aa in length (11). As we examined this peptide by binding studies, it became apparent that there was a “core” of nine residues in which MHC contact and TCR contact residues could be defined (12). Moreover, we had also shown that MHC II proteins could also bind self lysozyme with the same affinity as HEL; that is, by binding parameters, the MHC was not making the self–non-self discrimination. Later, the actual sequence of the naturally processed peptide was identified as a 9-mer core with variable flank residues such that the predominant peptide encompassed 14 residues (13). In an important analysis, Howard Gray’s laboratory, which had also found binding of peptides to purified MHC II, went on to show that a number of self peptides could be isolated from I-A^d and I-E^d molecules. These self peptides competed with the foreign peptides for MHC binding. However, in their paper, the peptides were not chemcally characterized (14). In a different study, Gray’s group went on to isolate a 10-residue peptide from HEL extracted from I-E^d molecules of HEL-pulsed APCs (15). In brief, this was the state of the growing field of Ag presentation at the time that Rudensky et al. carried out their important analysis in the 1991 Nature report. Analysis of the peptidome was the logical next step to follow, which they took with the techniques at hand at that moment.

The Rudensky et al. paper isolated only a few peptides, but these gave them a reasonable perspective on the nature of the MHC II peptidome. There were several highlights in their study. All the peptides were between 13 and 17 residues, confirming the analysis that had been done with synthetic peptides. The peptides bound to MHC II were larger in size than those binding to MHC I. They correctly speculated that the ends of the peptide-binding site of MHC II were likely more open than in MHC I. This point was later confirmed when Wiley’s group for the first time reported the structure of HLA-DR1 and showed that peptides lay in an extended conformation within an “open ended” binding groove in this MHC II molecule (16). This mode of peptide binding to MHC II molecules has held up as more structures of peptide–MHC II complexes were resolved (17). Another important issue in the Rudensky et al. paper is the source of the peptides: they derived from proteins of the vesicular system as well as from exogenous proteins, that is, bovine albumin in the culture media. That membrane proteins contributed substantially to the peptidome was not appreciated until this study. Also of interest, they identified a sequence from murine leukemia virus.

The finding that self peptides occupied the binding site raised the important issue of their role. Subsequent analyses by Hunt et al. (3) and others calculated that at the very least, ∼2000 different autologous peptides were bound to MHC II, some at just a few copies but others, being dominant, were much more highly represented, all related to their sequence. We now know that many MHC II peptides are selected as families consisting of the 9-mer core and flanking residues of variable lengths. The flanking residues are important and can affect the topology of the TCR contact residues (18). Many of the self peptides bind weakly and can exchange with peptides during their transport to the plasma membrane. These self peptides could compete with foreign peptides, as first proven in one of our own studies (19).

The autologous MHC II peptidome can also affect the response of the T cells during their differentiation. This is an issue that Rudensky et al. emphasized and extended in a second paper published in the same issue (20), in which they describe a peptide derived from the I-Eα-chain, detected by the Y-Ae Ab. As discussed in the featured Pillars article (1) and expanded in their second report, self peptides could have a role in modulating the response by affecting the T cell repertoire in the thymus. I find their speculation very appealing, although the biochemical features of the thymic peptidome among various alleles still need to be examined. Perhaps in addition to the role of self peptides suggested by Rudensky et al. (1, 20), these peptides could select nonconventional self-reactive T cells that could participate in autoimmunity (21).

The Rudensky et al. paper was an important contribution to the field. It was the beginning of a number of analyses of the MHC peptidome that were considerably expanded following the studies by Hunt and Engelhard. These efforts now continue to generate more information, using the powerful tools of tandem mass spectrometry, with which hundreds of peptides can now be analyzed and sequenced (3).