Structural and Thermodynamic Approach to Peptide Immunogenicity

Carlos J. Camacho et al. study the relationship between peptide stabilities and their immunogenicity. They present a model system of several peptides corresponding to portions of murine HRS which are capable of inducing anti-protein antibodies of varying affinity and temporal persistence. On this system they show by molecular dynamics simulations, that sequences of the most immunogenic peptides correspond to highly ordered structural motifs in the parent protein. Competitive ELISAs provide direct evidence that these peptides share structural determinants with native protein. More interestingly they address the question of how can these less stable peptides induce the same immunogenic response. By Nir London

Carlos J. Camacho, Yasuhiro Katsumata, Dana P. Ascherman (PLOS Computational Biology

In the current paradigm of immune system recognition, B cell receptors recognize unprocessed protein structures. However there are evidence showing that short peptide immunization can trigger B cell responses targeting both the immunizing peptide and peptide like motifs contained within the intact protein. This despite the fact that the folding stability of such peptides is often quite low.

In the current study, the authors investigated this issue through detailed analysis of serologic profiles generated in mice immunized with 9 overlapping 18 amino acid peptides comprising the amino terminal portion of histidyl-tRNA synthetase (HRS), an autoantigen implicated in the pathogenesis of idiopathic inflammatory myopathy and the anti-synthetase syndrome.

The authors argue that complete peptide stability is not a necessary condition for effective binding of the antibody. They classify peptides according to the free energy of their protein-like motifs. (a) “stable” peptides (for which DeltaG<0 kcal/mol) that can form the same number of peptide-antibody complexes as stable protein despite a wide range of folding free energy values; (b) “weakly-stable” peptides with DeltaG>0 kcal/mol (but <8 kcal/mol) that have a drastic decrease in antibody binding events relative to the full protein; and, (c) “unstable” or “non-immunogenic” peptides with DeltaG>8 kcal/mol and resulting unfolding rates that preclude any effective binding.

For 6 of the 9 peptides, indeed mice immunized with the peptides generated antibody responses against the full protein. For two cases the temporal behavior of anti-protein and anti-peptide responses were similar over time (increasing antibody response) 3 peptides shows variable temporal patterns when compared and another two fail to generate anti-protein or anti-peptide antibodies at any time point.

Competition ELISAs complements these results by showing that for the two peptides that produced the strongest antibody response there is indeed competition for the binding of antibodies between the peptide and the full protein while for other peptides the effect is more variable.

To correlate these peptide immunization studies with the thermodynamically-defined categories of immunogenicity, the authors utilized MD simulations. MD simulations represent a method for probing the relative stability of different epitopes in their corresponding protein fold and for better defining the relevant “folded” state. Because recognition events occur within a nanosecond time scale, peptides were simulated over a 10 nanosecond period.

Although peptide conformations fluctuate to varying degrees, the composite profiles of the most structurally stable protein-like motifs provide a visual analogue showing relative stability and similarity to defined motifs found in murine HRS. Interestingly, each of the peptides with a cumulative RMSD value less than 4 Å triggers affinity maturation towards the full protein and/or peptide, whereas peptides with less stable backbone structures typically do not promote this temporal pattern of increasing antibody titer. Of note, MD simulations indicate that for those peptides capable of adopting higher order structure, the identified motifs can persist for several nanoseconds – a time period sufficient for antibody recognition.

These findings strongly suggest that the highly immunogenic peptides fall into the stable category where DeltaG values allow maximal peptide-antibody complex formation. Other peptides which are weakly stable, show that antibodies generated by their immunization generally bind protein, but not peptide, substrate antigens. The non-immunogenic peptides lack any form of structure resembling native HRS, and no new structural motifs are detected within the limited simulation time. The latter observation also reflects the fact that although MD simulations and resulting RMSD calculations based on backbone stability provide a framework for ranking the likelihood of forming high affinity peptide-antibody complexes, side chains remain a critical determinant influencing the specificity of this interaction. More specifically, the loop structure of one of the non-immunogenic peptides is flanked by highly unstable side chains blocking the relatively conserved backbone.

For those peptides generating strong antibody responses against the HRS protein, structural mapping indicates correspondence to well-defined domains that involve either Alpha-helices or linear motifs stabilized by a proline residue. To some extent, this result is expected because (in solution) such motifs should retain some of the stability present in native protein. The key question, however, is how peptides bearing only partial structural resemblance to native protein can bind antibodies with similar affinity to that of intact protein. Answering this question relies on the simple observation that although peptides should be destabilized when isolated from protein, this instability does not translate into an equivalent drop in affinity towards the repertoire of B cells receptors. Thermodynamic calculations reveal a relatively broad range of DeltaG values in which peptides are capable of triggering an immune response similar to full protein. Hence, as long as the peptide fold resembles that of the full protein, this class of peptides (defined as “stable”) should have antigenic properties similar to those of full protein. In the “weakly-stable” regime peptides are typically 10-100 times less likely than HRS protein to bind peptide-induced antibodies. In other words, the same antibodies that rarely bind isolated peptides can readily recognize the corresponding motif in the context of stable protein. Unlike their more stable counterparts, however, such weakly stable motifs typically do not promote affinity maturation against protein.

Carlos J. Camacho, Yasuhiro Katsumata, Dana P. Ascherman (2008). Structural and Thermodynamic Approach to Peptide Immunogenicity PLoS Computational Biology, 4 (11) DOI: 10.1371/journal.pcbi.1000231

Be Sociable, Share!

Enjoyed this Post ?

Subscribe by E-mail:

Subscribe in a reader. Follow us on twitter.

Powered by WordPress | Aeros Theme | WordPress Themes
© 2009 Rosetta Design Group LLC