PhD Lecture by Sune Pletscher-Frankild, CBS

Degenerate Peptide Specificity of MHC Molecules and T cell Receptors in Immunodominance and Tolerance

Friday August 21, 2009 at 12:00
CBS, DTU, Lyngby, Building 208, Auditorium 062
Assessment Committee:    Professor Anders Gorm Pedersen, DTU (Chairman)
Professor Hanne Frøkiær, KU
Associate professor José A. M. Borghans, University Medical Center Utrecht
Chair of defense:Associate professor Thomas Blicher, DTU
Supervisor:Professor Ole Lund, DTU

MHC molecules have evolved highly degenerate peptide-binding motifs to allow broad coverage of microbial antigens. Somatically recombined T cell receptors (TCR) are also very cross-reactive to maximize the overall probability that at least one epitope is recognized in any virulent microorganism. However, the width of MHC and TCR binding speci?city cannot be stretched beyond certain limits as it increases the rate of negative selection which in turn increases the elimination of immature thymocytes. Host survival therefore depends on a ?ne balance between antigen coverage and T cell diversity. Here, we study the effects of degenerate speci?city of TCRs and MHC on tolerance and immunodominace, respectively. Chapter 1 provides a detailed introduction to T cell biology and maturation ending with a perspective on how our results from paper IV complement the understanding of T cell function. Paper I is a technical paper primarily dealing with the unbiased evaluation of competing MHC-binding predictors and paper III is a follow up. Paper II describes the initial development of a database of known pathogens. Although the technical results in paper I, II and III are important for vaccine optimization, the focus of the discussion (chapter 3) is on the results with biological implications presented in paper IV and a new manuscript included as chapter 2. In paper IV we investigate whether T cell recognition is predictable using a primitive model of TCR cross-reactivty based on peptide similarity. We test if peptides that stimulate the same T cell clone are similar in sequence. Data from various studies were then analyzed and we could conclude that cross-reactivity is indeed predictable. We then predicted a set of HIV-peptides which would cross-react with self-epitopes and found that none of these were immunogenic. This is presumably due to central tolerance which limits the antigenic coverage of T cells demonstrating that tolerance is also predictable. Chapter 2 describes how we developed a method to estimate the binding frequency of MHC molecules. The method was tested on experimental data on non-natural peptides showing a very high predictive performance. We could then show that HLA molecules roughly bind 1:20 to 1:200 or 0.5-5% of all natural peptides. These estimates are larger than current estimates and in detail show that HLA-A molecules bind about three times more peptides than HLA-B. It is currently not well understood why some epitopes are immunodominat and others are not, but we do know that the vast majority of immunodominant epitopes are HLA-B restricted. The ?nding that HLA-B molecules are the most speci?c is therefore interesting to the search for the causes behind the phenomenon of immunodominance.

Everybody is welcome. Registration is not necessary.