Type 1 Diabetes: Cellular, Molecular & Clinical Immunology
Theoretical Essay D - Why Doesn't Everyone Develop Autoimmune Diabetes?
David W. Talmage and Richard J. Sanderson
The islet cells of the pancreas are particularly vulnerable
to autoimmune attack because they are more easily destroyed by nonspecific oxidants
and inflammation than other tissues.1 For this reason the autoimmune process
in the pancreas does not require T cells or antibody specific for surface membrane
antigens. A cellular inflammation (insulitis) in response to any tissue antigen
(eg, cytoplasmic protein) will apparently do the job. What prevents the immune
system from responding to these self proteins?
We know that a key organ in the development of tolerance to self is the thymus,
where both positive and negative selection of T cells occur.2 The most important
negative selection is the deletion of T cells with the ability to respond to
self major histocompatibility complex (MHC) antigens. The MHC antigens responsible
for the deletion are on the surface of antigen-presenting cells (APCs), probably
macrophages and dendritic cells.
Positive selection in the thymus greatly increases the frequency of T cells
with low affinity for MHC antigens. However, this low affinity would have to
be below the threshold required to activate T cells in order to escape the process
of negative selection described above. Positive selection explains why T cell
responses are more easily generated to peptides complexed to self MHC antigens
than to not-self MHC antigens. The quilt of low-level affinity to self MHC antigens
requires only a few tufts of high affinity to specific peptides to exceed the
threshold required for activation.3 Thus, T cells with affinity for islet cell
peptides not found in the thymus should be particularly available for activation,
because they would receive positive but not negative selection in the thymus.
So what prevents an immune response to these islet-specific proteins?
The answer to the last question is that there must be an extra-thymic process
or processes to suppress the response to self peptides not present in the thymus.
And we have evidence that such an extra-thymic suppression does take place.
Cultured islet or thyroid grafts, which survive for long periods in allogeneic
hosts, induce a resistance to rejection in their hosts, called tolerance.4 This
is an active process that takes weeks to months to develop, and although difficult
to transfer, the tolerance has been successfully transferred through parabiosis.5
This tolerance suppresses the rejection of new grafts as well as old and also
prevents the rejection of grafts after the transfer of activated immune cells
specific for the graft.
If tolerance can be acquired to foreign antigens, then it must also be possible
to acquire tolerance to self antigens. We do not believe that this acquired
tolerance to self antigens is due to blocking antibody. Although antibody to
foreign MHC antigens has been detected in graft recipients, antibody to minor
antigens, which are usually tissue-specific peptides, has not been detected.
And the acquired tolerance to cultured grafts develops to minor antigens as
well as major.4
Our proposal is based on the fact that in addition to their specificity, T cells
can be divided into several classes based on the lymphokines they produce.6,7
One class of T cells, called Th1 cells, produces both interleukin-2 (IL-2) and
interferon gamma, two lymphokines that are necessary for the cellular infiltration
and rejection processes of delayed-type hypersensitivity (DTH). Other T cells
make other lymphokines and are collectively called Th2 cells, although they
probably do not constitute homogeneous group. Unfortunately, there are currently
no surface markers capable of identifying the lymphokine-producing class of
T cells and investigators have had to rely on the production of a specific lymphokine
to identify T cell class in this way.
Our proposal is that Th1 and Th2 cells represent mutually suppressing classes
of T cells.8 Thus, if either class becomes dominant in the T cell response to
a particular peptide, if suppresses the response of the other class. According
to this proposal, tolerance is not do so much to the absence of a T cell response
but to the dominance Th2 cells in that response. A number of different findings
seem to point in this direction. It has long been known that antibody production
and delayed hypersensitivity appear to be mutually exclusive. Recently, Sanderson
has shown that IL-4, a product of Th2 cells that is required for most antibody
responses, inhibits the clustering of lymphocytes that occurs in response to
PMA and calcium ionophore.9 LaRosa and coworkers have shown that the infiltration
that occurs in NOD mice made resistant to the development of diabetes by the
injection of complete Freund's adjuvant (CFA) has a significantly higher content
of IL-4-producing cells than the infiltrate of uninjected mice.10 All of these
findings suggest that IL-4 suppresses the production of interferon and IL-2
by Th1 cells.
Our proposal that IL-4 inhibits the production of interferon should be easily
testable. If this hypothesis should be confirmed, than the following model of
tolerance to tissue-specific self-peptides would seem likely.
The Proposed Model
Th1 and Th2 cells are mutually suppressing. Thus, one or the
other cell population tends to become dominant in the T cell response to a particular
peptide. The outcome depends on the dose of antigen and the manner in which
it is presented. In a normal individual not susceptible to diabetes, the response
to self peptides is dominated by Th2 cells. This is probably also true with
viral and bacterial antigens. The production of IL-4 by Th2 cells enables the
animal to make a quick antibody response that is essential to resistance to
the microbial antigens. At the same time damaging autoimmune DTH to self peptides
is avoided; there is also relatively little autoantibody production to self
peptides, because most of these are too small to induce such a response. Antibody
responses to the larger protein MHC antigens are avoided because T cells specific
for these antigens are deleted by the negative selection in the thymus.
In an individual susceptible to diabetes, the specific clones of Th2 cells that
are needed to induce tolerance to certain self peptides (eg, islet cell peptides)
are deleted in the thymus or blocked from leaving because of a cross-reaction
with a particular MHC class II antigen. In the absence of these specific Th2
cells, the specific clone of Th1 cells is free to expand and produce interferon,
which activates the inflammatory macrophages that kill the islet cells by some
nonspecific process.
A number of predictions of the model or easily testable. Already mentioned is
inhibition of IL-2 and interferon by IL-4. IL-4 injected along with islet transplants
and diabetic NOD mice should inhibit the recurrence of diabetes. Anti-II-4 given
to NOD mice made diabetes resistant by the injection of CFA should develop diabetes
anyway. The lymphocytic infiltration around cultured graphs in tolerant animals
should be relatively rich in Th2 cells.
The model also has the advantage of explaining a puzzling phenomenon that has
been known for some time. This is the apparent mutual exclusion of antibody
production and delayed hypersensitivity. If IL-4 is suppresses for Th1 cells,
than this means that the helper factor required for antibody production is paradoxically
suppressing the development of delayed hypersensitivity. The only explanation
for such a strange behavior is that this action has some benefit to the organism,
such as preventing the development of autoimmunity. The recent demonstration
that anti-IL-4 can blocked the development of tolerance supports this proposal.11
Reference List - links to PubMed available in Reference List.
For comments, corrections or to contribute teaching slides, please contact Dr. Eisenbarth at: george.eisenbarth@ucdenver.edu