Scripps Research Logo


Modulation of T cell signaling strength by Themis, and its role in thymocyte differentiation

We recently identified and characterized a novel gene and protein called Themis (Fu et al., 2009. PMCID: PMC2782925). It is only expressed in the T cell lineage, and is predominantly found in CD4+8+ (DP) thymocytes. It is highly conserved in all vertebrates and is a member of a completely new structural class of proteins. It has no classical domains other than a polyproline region. In collaboration with Oreste Acuto (University of Oxford, UK), we found that the polyproline region binds to an SH3 domain in Grb2 and this function is required for Themis to function in thymocyte development.

Themis-knockout mice have a defect in development of mature CD4+ thymocytes and T cells. This defect in development occurs at positive selection, the checkpoint where the TCR is selected to be able to recognize self MHCp enough to get a small signal, leading to its differentiation into a mature thymocyte. Our recent studies on Themis indicate that it is involved in setting the threshold for TCR signaling by reducing signaling strength in response to relatively low affinity ligands, such as those that stimulate the differentiation of thymocytes into mature T cells (positive selection). However, it allows strong responses to high affinity ligands such as those that cause the thymic deletion of potentially autoreactive T cells.


Visualization of intermolecular interactions in live cells

We developed Foerster resonance energy transfer (FRET) microscopic methods to enable us to analyze intermolecular interactions in living cells. We were the first group to use live-cell FRET microscopy to look at protein-protein interactions in the immunological synapse (Zal et al., 2002). In doing this we made an important methodological advance that allows quantitative molecular interaction information (FRET efficiency) to be obtained from the FRET method most compatible with live-cell imaging (ratio of fluorescence emission between acceptor and donor) (Zal and Gascoigne, 2004). It also allows FRET data to be obtained from time-lapse and 3-D imaging where photobleaching is a major problem. Neither was previously possible. Earlier FRET studies relied on labeled antibodies, and so could not look at molecular interactions during cell-cell interactions. We showed that the TCR-CD4 interaction is induced at the immunologic synapse by antigen recognition, which was important because it showed that the co-receptor is not strongly pre-associated with the TCR, as suggested from other studies (Zal et al., 2002). Antagonists did not induce the close TCR-CD4 interaction and in fact blocked the interaction induced by agonist. We also used FRET to analyze TCR-co-receptor interactions induced by positive and negative-selecting peptides for which we had previously measured the TCR:MHCp affinity (Yachi et al, 2006). Although there was a good relationship between TCR affinity and T cell activation, there were exceptions. These were explained by the kinetics with which TCR interacts with CD8: negative-selecting ligands induced a fast, brief, FRET signal, whereas positive selecting ligands induced a slower FRET response. This work is being continued as described in the next paragraph.


Early events in TCR signaling 

Initial stimulation through the TCR induces the binding of CD8 to MHC molecules, enhancing the activation in response to agonist ligands. The mechanism by which CD8 is initially brought into close proximity of the TCR may be through intra-cytoplasmic interactions between the CD3 subunits of TCR and the Lck molecule that is associated with CD8, rather than through the binding of CD8 to MHC. We have developed a TIRF microscopic FRET method for investigating these interactions. We intend to identify the molecular mechanisms for these interactions between particular CD3 molecules and the different domains of Lck. We collaborate in these studies with the lab of Cheng Zhu (Georgia Inst. of Technology, USA) who developed a highly time-resolved biophysical method for measuring TCR/CD8/MHCp interactions as well as 2-dimensional affinity. We will also investigate the role of mechanical stimulation of the TCR on these early interactions.


Endogenous non-stimulatory peptides as co-agonist for TCR stimulation

An unexpected finding from the TCR-CD8 interaction studies was that endogenous peptides on an antigen-presenting cell can increase T cell sensitivity to antigen, even though the T cell does not react to the endogenous peptides on their own. This is particularly relevant when the antigenic peptides are few, within a sea of endogenous peptides, as is usually the case. It is likely that viruses (e.g. CMV) or tumors that downregulate MHC expression on the infected cell are making use of this co-agonist phenomenon to avoid detection by T cells. The pathogen is unlikely to be able to remove all of the antigenic peptides (and T cells can respond to a very small number of antigenic peptides), but if the bulk of the non-stimulatory MHCp’s were missing, a specific T cell would not be activated by a few antigenic MHCp.

We demonstrated that for at least some MHC class I-restricted T cells, unlike for MHC class II-restricted cells, the specificity of the peptide is unimportant, probably because of the efficiency with which the CD8-MHC class I interaction recruits the kinase Lck to the immunological synapse (Yachi et al., 2005). We, in collaboration with Dr Arup Chakraborty (MIT) and Keith Gould (Imperial College London), have resolved the differences between the MHC class I- and class II-restricted T cells for their requirements for particular kinds of endogenous peptides in order to show this “co-agonist” activity. We now know that when the affinity of the co-receptor is high, the TCR can make use of many endogenous MHCp complexes, and so does not show peptide-specificity for the endogenous co-agonist peptides. In contrast, when the co-receptor affinity for MHC is weak (as in the CD4-MHC class II interaction), the TCR requires a stronger interaction with the co-agonist in order to get activation, so that peptide-specificity for the non-stimulatory ligands is evident.


The role of PKCη in T cell development and activation

We found that the protein kinase Cη isoform is upregulated after TCR ligation in developing thymocytes, and likewise in natural positive selection. Of the PKC isoforms, only PKCθ has been shown to have a special role in T cells, where it is recruited to the immunological synapse during antigen-recognition. Because PKCθ-deficient mice have very little defect in thymic selection, this suggested that PKCη could be replacing PKCθ in the developing thymocytes, and this proved to be the case (Fu et al, 2011. PMCID:PMC3242502). PKCθ knockout mice have no defect in thymocyte development, but PKCh/PKCθ double KOs have a significant defect. Despite redundant or overlapping roles in many aspects of T cell activation,  PKCh KO mice show a unique defect in homeostatic proliferation, that is not shared with PKCθ KO mice.