Statement of Major Research Accomplishments and Interests, continued
2. Cytotoxic T lymphocyte selection, function, and aging
My other principal interest is centered on selection, recognition, function and homeostasis of cytotoxic T lymphocytes. We are studying these issues using a combination of structural, molecular, transgenic and functional approaches. Crystallographic modeling, TCR sequence and CDR3 length analysis, soluble and cell-bound TCR:peptide:MHC interaction measurements, transgenesis, site directed mutagenesis and functional immunological assays are combined to precisely dissect the CTL recognition and repertoire and to study CTL activation. Projects include: 2.1 CTL crossreactivity and positive selection; 2.2 The role of CTL repertoire, avidity and function in immune defense; 2.3 Optimal class I-restricted peptides as CTL vaccines; and 2.4 CTL dysregulation in immune senescence.
2.1 CTL crossreactivity and positive selection
Our model involves the class I molecule H-2K (Kb) and its natural gene conversion variant, H-2Kbm8 (Kbm8). These two molecules differ by four clustered substitutions at the floor of the peptide-binding site and exhibit marked differences in presentation of immunodominant ovalbumin (OVA-8) and Herpes Simplex virus glycoprotein B (HSV-8) peptides to cytotoxic T lymphocytes (CTLs), and in positive selection of CTLs specific for these antigens in the thymus in the absence of OVA-8 and HSV-8. Our efforts are aimed at understanding the molecular nature of these differences. This model offers several advantages:
i. H-2Kb was extensively functionally and structurally (X-ray crystallography) characterized and we are collaborating with Dr. Daved Fremont, (Washington University, St. Louis, MO) to extend these studies to the HSV-8:MHC complexes;
ii. H-2Kb has a number of useful natural mutants;
iii. many antigenic peptides restricted by this molecule were defined;
iv. we have obtained or produced several H-2Kb-restricted TCR Tg mouse lines specific for Kb-binding peptides, including two Tg lines specific for the HSV-8;
v. we can immunize CTLs using any MHC-binding peptide emulsified in an adjuvant;
vi. we use the Immunoscope method to quickly survey the CDR3 diversity of the CTL repertoire, and sequence the TCRs of interest; and
vii. we can directly follow virus-specific CTLs in vivo using the peptide:MHC tetramers.
We recently functionally addressed the nature of differences in peptide presentation, demonstrating that a subtle conformational change, dependent on the disturbance of a hydrogen bond network conserved in H-2Kb, can explain certain antigen-presenting properties of Kbm8. We further demonstrated that, compared to H-2Kb, H-2Kbm8 exhibits both loss (OVA-8)- and gain(HSV-8)-of-function for both peptide presentation and positive selection. Interestingly, Kbm8 selects crossreactive CTLs that recognize HSV-8 on both Kb and Kbm8, while Kb can select only mono-(HSV-8/Kb)-specific CTLs. This crossreactivity correlates to the utilization of defined TCR Vß and Va segments, providing us with an ideal model to dissect the molecular nature of TCR crossreactivity. To that effect, we derived a panel of HSV-8-specific CTL clones whose rearranged TCR genes have been sequenced. This molecular information was used to generate TCR single-chain transgenic mice, in which one half of the TCR is fixed. We are undertaking detailed mapping of the TCR:peptide:MHC contact, which, along with TCR mutagenesis and crystallography, should lead to a full molecular dissection of crossreactivity in this model. Subsequently, these studies will be extended to positive selection.
Our previous work pioneered the idea that, in order to be positively selected in the thymus, the TCR on a developing thymocyte must interact with a self peptide:MHC complex. This idea has been confirmed and extended by the work of other laboratories. But how many T cells in the repertoire are positively selected on specific self pep:MHC complexes, and how many need only the MHC? To answer these questions, we designated and engineered an H-2Kb variant, KbW9, that cannot bind peptides that bind to H-2Kb. While Kb and KbW9 bind nonoverlapping sets of peptides, their TCR contact area remains identical to the parental molecule, as shown by peptide elution and sequencing and TCR recognition. These studies elucidated many new details of the peptide:MHC contact. We also generated transgenic animals expressing Kb and KbW9 to analyze positive selection of a polyclonal and monoclonal (by interbreeding to TCR Tg mice) repertoire. This will allow us to test the limits of the plasticity of TcR repertoire selected by mutant MHC.
2.2 The role of CTL repertoire, avidity and function in immune defense
The above studies on antiviral CTL repertoire suggested the importance of the diverse CTL repertoire in controlling HSV infection in vivo. It is indeed critical to dissect the individual contributions of CTL repertoire diversity, in vivo CTL avidity for antigen and in vivo CTL peptide sensitivity and cytotoxic activity in mediating immune defense. We are currently investigating these problems in three relevant models: acute viral infection, subcutaneous lymphoma and skin transplant rejection models.
2.3 Optimal class I-restricted peptides as CTL vaccines
In a related set of studies, we are using our understanding of TCR:peptide:MHC interaction to produce efficacious anti-tumor CTL responses against poorly immunogenic peptides. The rationale of our strategy is to pick peptides that bind to the MHC poorly (so that they will not initiate a CTL response), but still above a certain threshold (so that they can target the tumor for destruction by activated CTLs), and engineer them into good class I binders. The animal is then immunized with such "heteroclitic" peptides and CTLs generated crossreact to the poor binder and destroy the tumor. This successful strategy is now being tested in a preclinical "humanized" mouse model of the prostate cancer.
2.4 CTL diversity and homeostasis in immune senescence
Aged mice and humans display limited T cell diversity, owing, in part, to the presence of age-related clonal T cell populations. Why and how are these cells generated, and what endows them with selective survival, growth and/or expansion properties is not known. Antigenic stimulation may be necessary for their onset and/or survival and expansion. Defects in clonal homeostasis, particularly in the mechanisms that ensure clonal constriction by apoptosis, could also contribute. Furthermore, clonal T cells may influence the immune system of the elderly by limiting the available T cell repertoire, and/or by disturbing cytokine balances necessary for a productive and flexible immune response. To that effect, we are evaluating the role of antigen stimulation in the onset of clonal expansion of CD8 T cells, the dependence of the clones on antigen:MHC stimulation for survival/expansion and the ability of the clones to undergo apoptosis. The hypothesis that T cell clones hamper the immune response by constricting the available T cell repertoire is supported by our current data.