|
Our laboratory studies many
aspects of T cells. T cells use receptors made up of a and b chains to recognize
invading organisms, usually in the form of peptide fragments derived from the
invader bound to specialized proteins called major histocompatibility complex
proteins (MHC) bourne on cell surfaces. The laboratory uses many techniques to
study these cells ranging from Xray crystallography to measurement of gene
expression using gene arrays and assessment of the fate of T cells in normal and
immunized animals.
T cell positive and negative
selection.
T cells are produced in the thymus and, after they have matured migrate to
other lymphoid organs such as the spleen and lymph nodes, where they drive
specific immune responses. In the thymus T cell precursors assemble the genes
which will code for their aß receptors from a number of choices. The choices are
such that each of the trillion T cells in each human being, or each of the
hundred million T cells in a mouse, bears a different aß receptor. Consequently
T cells differ in their ability to recognize different invaders. Before T cells
leave the thymus, however, they must pass 2 tests, called positive and negative
selection. They must have a low but perceptible ability to react with MHC
proteins bound to peptides of their host. On the other hand, they must not react
too avidly with MHC + self peptides. If they react with too low or too high an
avidity, they die.
Although positive and
negative selection have been studied for years by many groups, much still
remains unknown about the processes. Our recent experiments have focussed on the
role of self peptides in the phenomena. To study this, we have created mice
which express a single class II MHC protein, bound to a single peptide. All the
class II restricted, CD4+ T cells which appear in such animals appear to have
been positively and negatively selected on this single MHC/peptide combination.
Most of these cells react with the same MHC molecule, bound to one of the many
mouse peptides with which it is engaged in normal mice. In normal mice these
cells must die in the thymus, demonstrating that negative selection deletes a
very large proportion of all the cells which are positively selected.
Our experiments now
concentrate on the difference between the TCRs on cells selected in thymuses
expressing 2 different peptides bound to the same MHC protein. What will be the
effects of the different peptides? To immunize these mice and compare their aß
receptor repertoires we prepare dendritic cells expressing class II bound to a
single foreign peptide. Studies of the T cells generated by immunizing with the
dendritic cells are underway.
Effects of pre-T alpha.
Others have shown that after developing thymocytes have rearranged their
TCR ß chains, and before they have rearranged TCR a, they express their ß chains
on their surfaces, combined with a protein called pre Ta. This combination
delivers signals to the thymocyte which drive it to the next developmental
stage. There are many mysteries about the pre Ta/TCRß combination. First, it
seems to be missing a domain, that equivalent to Va on mature cells. Secondly,
the means by which this complex signals the thymocyte is unknown and there is
always the possibility that it has a ligand. We have made a monoclonal antibody
to mouse pre Ta. We are using this antibody and other methods to study the
structure and function of pre Ta.
T cell activation in animals.
Activation induced T cell
death.
Introduction of antigen into animals causes T cells specific for that
antigen to divide rapidly, then die. The death is caused by several processes of
which the most well known is engagement of Fas on the surfaces of the activated
T cells by Fas ligand on other cells, and consequent activation of caspases
within the cells. Experiments in our laboratory indicate that activated T cells
also die because they contain increased amounts of reactive oxygen species and
these molecules, directly or indirectly, lead to the deaths of the cells. Recent
experiments suggest that a third process may also be involved in this death. We
have used Affymetrix gene arrays to compare gene expression in T cells which are
destined to live or die and have thus identified several sets of proteins which
may be involved in activation induced T cell death. Such proteins include the
usual suspects, for example members of the Bcl-2 and NF-kB families, and also
more unusual candidates, such as the glycolytic enzymes and uncoupling proteins.
At the moment we are
particularly interested in the role of the proapoptotic protein, BAD, in T cell
death. Our experiments on this subject include studies of its structure and its
interactions with anti-apoptotic proteins such as Bcl-xl using conventional
protein chemistry and X ray crystallography. Experiments on its distribution in
living and dying T cells are also planned.
Adjuvants and the survival of
activated T cells.
Immunologists were surprised when they discovered that activated T cells
die in animals because it has been assumed that these cells would survive to
form the pool of memory cells which would protect the animal against a second
infection with the same antigen. In retrospect, of course, death was not a
surprising fate for most of the activated cells since, were they all to survive,
the animal would rapidly be overwhelmed by its accumulated immunological memory
cells. The circumstances under which T cells are activated does, however, affect
the T cells to some extent. Activation in the presence of natural adjuvants such
as bacterial lipopolysaccharide or virus infection does tend to save some of the
T cells from death. Perhaps this distinction between activation in the absence
or presence of adjuvants is one of the methods by which the immune system
distinguishes between self and invading organisms.
Adjuvants do many different
things. Our interests focus on their ability to prevent activated T cell death.
Inflammatory cytokines play a role in the phenomenon but other molecules,
especially certain members of the TNF receptor family such as OX40 are probably
also important. Again we have used analyses on gene arrays to pick additional
candidates.
Memory T cells.
In the course of experiments on T cells in older animals (see below) we
noticed that memory T cells in mice are proliferating in the apparent absence of
antigen. Their division rate is slow. They go through one round of division
approximately every 12 days. Later we found that, for CD8+ T cells, this
division was driven by IL-15 and perhaps IL-7. Unexpectedly we also found that
IL-2 prevents accumulation of the dividing memory cells, by killing the cells as
they proliferate. In the absence of IL-15, the numbers of CD8+ memory T cells in
animals fall. In the absence of IL-2, their numbers increase dramatically. Now
we would like to find out if we can use these observations to improve
immunizations or, conversely, to facilitate treatment of autoimmune diseases.
T cell homeostasis and the
MHC and TCR structure and function.
We have long been interested in the structure and function of MHC class II
proteins. When we solved the structure of the class II protein, IEk bound to a
peptide from mouse hemaglobin we were surprised to find a cluster of negatively
charged amino acids at the base of one on the peptide binding pockets of the
protein. This cluster is preserved in human DR. We are now studying the effects
of this cluster on the properties of the protein. Unexpectedly, the cluster
appears to stabilize the protein, perhaps by partcipating in an extended
hydrogen bonding network.
H-2M (DM) is known to
participate in replacing the invariant chain peptide, CLIP, with more strongly
binding peptides in class II. We are using crystallography and site specific
mutation to study this process.
In some people T cells
react with antigens which are not peptides but which might bind to and modify
the structure of peptides. One example of such an antigen is the nickel ion. We
have expressed T cell receptors which can react with nickel bound to DR and are
studying the interaction between these molecules.
Although several Xray
crystallographically solved structures of TCR and MHC have been published, much
remains to be known about the interactions between these two important proteins.
For example, others have predicted that the geometry with which TCRs react with
MHC and peptide is different if the MHC involved is class II versus class I. We
are approaching this problem in two ways. First, we are studying the propeties
of the reaction between TCRs and MHC/peptide combinations which react with each
other with greatly different kinetics. The idea is to investigate whether T cell
responses are controlled by the off rate of the TCR/MHC/peptide interaction, as
has been predicted by others, or by their overall affinity for each other.
Secondly we are preparing TCR/MHC/peptide combinations which have very high
affinity for each other in the hope that this will improve their ability to
co-crystallize.
Effects of TCR affinity on T
cell priming and function.
Others have shown that polymers of MHC and peptide can be used to identify
antigen specific T cells. Later we showed that this method can also be used to
measure the affinity of the TCRs on T cells for their MHC/peptide ligands.
Repeated priming with antigen increases the average affinity of T cells for the
MHC/antigen complex. We are now investigating how the affinity maturation of the
T cell population occurs, and what its consequences may be for the immune
response of the animal.
John Kappler and Philippa
Marrack are Howard Hughes Medical Institute Scholars.
View the HHMI faculty page for J Kappler?
View the HHMI faculty page for P Marrack?
| 
