A Brief Guide to the Immunology of Peripheral Tolerance

Steve Cobbold

Clonal Selection as the basis of Immune Specificity and Memory

Clonal selection

The whole basis of specificity within the immune system is that clones of lymphocytes are generated with random receptors for antigen. Within this random pool of receptors, some will have high affinity for self molecules, and these clones will normally be deleted at the site of their generation (ie. the thymus for T-lymphocytes). Other clones, by chance, will have high affinity for particular molecules (antigens) on, for example, an infecting flu virus, and this will allow the virus to be recognised by these clones. Recognition and activation then causes clonal expansion, so that a much higher proportion of T-lymphocytes will be available to recognize any further infection with the same flu virus, providing the basis of immunological memory. However, because the receptors are generated randomly, many will have a moderate affinity for both foreign proteins on infectious organisms, but possibly also a moderate affinity for self proteins not present at their site of generation (ie. non thymic antigens, in the case of T-lymphocytes). It is in considering how the immune system can distinguish between self and non-self in the case of these potentially cross-reactive clones that allows us to understand autoimmune diseases and how they might be treated.

T-lymphocyte recognition may look complicated (but it isn't really)

The heart of the immune response is based around T-lymphocytes that express a molecule called CD4 that is a receptor for the Major Histocompatibility Complex class II (MHC-II) molecule (that is one of the components of the Tissue Type). Protein molecules from the infectious organism or foreign graft (called antigens) are processed into peptides by specialized cells called Antigen presenting Cells (APC), and it is these peptides that bind to the MHC-II molecules that are recognised by the T-lymphocyte using its antigen receptor.

T cell and APC interaction molecules

A number of other adhesion molecules and growth factors are also used to send signals between the T- lymphocyte and the Antigen Presenting Cell, and only if all the signals are correct does the T-lymphocyte become activated and aggressive. An aggressive response results in the multiplication of that clone of T- lymphocytes, which also develop the ability to kill all further infected cells, using a similar recognition process to that shown above.

The difference between self and non-self

T-lymphocytes that recognize the majority of self molecules expressed throughout the body are inactivated or deleted centrally during development in the thymus. However, this sytem cannot deal with self molecules that are only found in specific organs or tissues outside the thymus (ie peripheral self antigens). It is the various molecules that allow communication between the antigen presenting cell and the T-lymphocyte that determine whether recognition, even of self molecules, leads to an aggressive response or not. An important part of determining which signalling molecules are present is the local environment within the architecture of the immune system. Infections tend to cause inflammation and the release of inflammatory cytokine molecules like interferon-gamma, and find their way into the immune system via the lymph system to the lymph-nodes. Therefore the lymph-node is specialized for presenting foreign antigens to generate immune responses.

Immune activation (non-self)

However, most of our body is not structured like a lymph-node, and if healthy should not have inflammation, and in this situation many of the important molecules for signalling aggression are not expressed. In this situation, although antigen can still be recognized by the T-lymphocyte, the response is not one of aggression, but rather of tolerance.

Immune tolerance (self)

This non-inflammatory environment can be reinforced by the presence of anti-inflammatory cytokines like IL-4 and IL-10, that can either be produced by healthy tissues, or by a population of T-lymphocytes that are protective and tend to yet further suppress any tendency to (self) aggression. We will return to the concept of suppressor T-lymphocytes below.

T-lymphocytes behave like hooligans

Because it is the whole local environment that determines whether a T-lymphocyte makes an aggressive response or not, there is a tendency for individual T-lymphocytes to behave like hooligans at a football match. Hooligans make most trouble when in gangs where a few ringleaders can make a whole crowd join in their aggressive behaviour.

The hooligan hypothesis

T-lymphocytes are similar in that unless sufficient numbers of them are helping each other become aggressive in the same local environment they instead default to the tolerant, perhaps even protective state. For normal body antigens, the vast majority of T-lymphocytes will already be in a state of self-tolerance, generated during development, and so ensuring that only very few, potentially auto-agggressive T-lymphocytes arise, and like hooligans isolated in a crowd they will be unable to cause much trouble on their own. Infectious organisms, however, will suddenly present a whole range of new foreign antigens to the immune system, able to be recognized by many different T-lymphocytes simultaneously, thereby providing the necessary incitement to aggression.

There are various levels of tolerance and it is also reversible

An aggressive response after a naive T-lymphocyte has recognized antigen leads to activation, then proliferation. Also, in order to provide immunological memory, many of the adhesion molecules and growth factor receptors are upregulated, making repeated activation more likely and rapid. Finally, T- lymphocytes can terminally differentiate into effector cells with a variety of functions aimed at either directly killing target cells or directing other parts of the immune system to do so. If, however, the T- lymphocyte responds in a non-aggressive manner, it can go through analagous stages giving different levels of tolerance - first, various interaction molecules (eg CD4, T-cell receptor, CD28 and growth factors like IL-2) can be down-regulated, leading to, in effect, a memory for non-aggressive responses. Eventually the T-lymphocyte can become almost completely anergic, unable to make any response, and may eventually die by apoptosis (commit suicide).

Reversibility of tolerance/activation

The idea of hysteresis, where the degree of response (positive aggression or negative tolerance on the Y axis) resulting from the degree of stimulus (on the X axis) can be used to demonstrate how any state between the two extremes of terminal differentiation and apoptosis can be reversed, but that there is a large resistance to such change. In other words, tolerant T-lymphocytes require a major extra inflammatory stimulus, such as a life threatening viral infection throughout the body, in order to reverse tolerance and hopefullly provide some extra (cross-reactive) immunity in such a dire situation. If, after the infection clears up, these reactivated T-lymphocytes are NOT returned to tolerance, then this is one way in which an autoimmune disease could result. Primed or previously activated T-lymphocytes can conversely be made tolerant if we can efficiently block their reactivation signals. This we can achieve with monoclonal antibodies against these signally molecules (eg CD4 and CD8), is the basis of many of our tolerance model systems in vivo, and holds out some hope to a clinical treatment of autoimmunity and graft rejection.

T-lymphocytes with suppressive properties

It now seems that tolerant T-cells are not always completely non-responseive (anergic) but can actually have protective properties by acting to suppress aggresive responses occuring close by. Suppressor T-lymphocytes have been the subject of much controversy over the last twenty years or so, and it used to be thought that there were separate lineages of T-lymphocytes, specialized align = center to see different components of the foreign antigen, and even able to recognize the internal image of another T- lymphocytes antigen receptor (anti-idiotype).

Lineage model of suppression

More recently, it has become apparent that suppressor T-lymphocytes may simply represent an alternative path of differentiation than that to effector cells, and may even be identical to those T-lymphocytes that have resulted from a non-aggresive response by becoming tolerant. This alternative differentiation causes the T- lymphocyte to make a different pattern of growth factors or cytokines, that can either directly suppress aggressive T-lymphocytes or may act to make the local environment less inflammatory.

Choices model of suppression

These two alternative pathways of differentiation are somtimes termed Th1 for the aggressive pathway making inflammatory cytokines like interferon-gamma and tumour necrosis factor, and Th2 (or the newly recognized Tr1 subset) for the protective pathway where IL-4, IL-10 and TGF-beta are made to inhibit inflammation.

Tolerance is infectious

Very recently, we have found that once the immune system is tolerant of, for example, a skin graft, you cannot break tolerance by infusing even large numbers of naive or even aggressive T-lymphocytes. If you follow the fate of these infused cells, you find that over a period of about two weeks, they are themselves guided towards tolerance by the tolerant immune system, and these secondary tolerant T-lymphocytes can even act to suppress any further infusions of potentially aggressive cells. So it looks almost as if tolerance is infectious, being passed on from one set of tolerant T-lymphocytes to another.

Infectious tolerance - Th1/Th2 hypothesis

One hypothesis that could explain infectious tolerance is that Th2-type T-lymphocytes make up the tolerant population, and make cytokines like IL-4 that are known to influence naive T- lymphocytes to follow the same pathway of differentiation while at the same time suppressing the generation of aggressive (Th1) responses. Therefore, Th2 cells beget Th2 cells beget more Th2 cells etc.... This is generally known as "Immune Deviation". A more recent idea is that there is a special T cell subset that regulates Th1 responses, termed Tr1 that is induced by IL-10 and acts to down regulate antigen presentation by making cytokines such as TGF-beta. This is perhaps a more attractive hypothesis because we and others have recently shown that Th2 cells are not necessarily protective. For example, both Th1 and Th2 T cell clones directed against the H-Y antigen can reject male skin grafts after transfer in back in vivo.

The hypothesis that it is the balance of Th2 (or Tr1) over Th1 responses that leads to tolerance, and explains its infectious nature is very topical, but it is almost certainly only one part of the mechanism of suppression and immunoregulation. For example, we do not yet know how suppression is able to operate between the two major subsets of T-lymphocytes that express either CD4 or CD8 and only recognize antigen presenting cells or targets that have the appropraite Major Histocompatability Complex molecules (MHC Class II and MHC- Class I, respectively).

Models of linked suppression

It is now possible to use populations of T-lymphocytes all expressing the same (transgenic) antigen receptor, for example against the male antigen on MHC Class-II cells (H-Y) or a defined MHC Class-I antigen (Kb), and look at the requirements for them to regulate each other. It is this type of experiment that has demonstrated that T-lymphocytes tolerant to one antigen can suppress another set of T-lymphocytes specific for a completely different antigen, if the two antigens are linked on the same antigen presenting cell, presumably bringing them into very close proximity.

Molecular mechanisms of tolerance and immune privilege

In the late 1990s, Andrew Mellor and David Dunn discovered that an enzyme called IDO, which breaks down (or catabolises) the essential amino acid tryptophan, is important to maintain immune tolerance to the fathers "foreign" antigens expressed by the foetus during pregnancy. By collaborating with this group, we have shown that regulatory Tr1 cells, though signals transmitted via CTLA4, induce the enzyme IDO in dendritic cells. This IDO breaks down tryptophan and induces what we have termed a local "immune privileged" microenvironment deprived of this essential amino acid. This is sensed in T lymphocytes by the signalling molecule GCN2, which can stop CD8 T cells from proliferating and acting as cytotoxic T cells, so that the nearby tissue is protected from immune attack. We beleive this is part of the molecular explanation for infectious tolerance.

Linked suppression via IDO and tryptophan catabolism

Over the last few years it has become clear that there are more functional subsets of helper T-lymphocytes than just the original Th1, Th2 and Tr1. In particular, we now think that in the presence of transforming growth factor beta (TGFbeta), an alternative pathway of differentiation can generate a whole new set of T lymphocytes including regulatory T cells (Tregs), Th17 cells and Th9 cells. The Treg cells, which can be recognised by their expression of the master regulator gene foxp3, are now known to be important in immune tolerance both to self antigens (to control autoimmune disease) and to foreign antigens (to control allergy and maintain transplantation tolerance). Like the Tr1 cells in the infectious tolerance experiments above, Tregs express high levels of CTLA4 and also produce anti-inflammatory cytokines such as TGFbeta and IL-10. We have shown that all these can act to induce enzymes in dendritic cells and other cells within the tissues that consume essential amino acids in a manner similar to IDO catabolising tryptophan. This depletion of essential amino acids is then sensed by both GCN2 (as above) as well as by another molecule called mTOR, and this not only stops the T lymphocytes from proliferating and differentiating into harmful effector T cells, but actually pushes them to express foxp3 and become a new cohort of Tregs.

Infectious tolerance via amino acid consumption and inhibition of mTOR signalling

We believe this is an important molecular mechanism which explains how infectious tolerance works. It is also interesting to note that the immunosuppressive drug rapamycin is an mTOR inhibitor, and this may explain why it is one of the few conventional immunosuppressive drugs thought to be tolerance permissive.


In summary, the immune system has developed many mechanisms of distinguishing between self tissues and foreign infections. We are able to exploit and expand some of these natural mechanisms using, for example, monoclonal antibodies that block critical signalling molecules on the T-lymphocytes, such as CD4. This should allow us to develop treatments to re-establish self tolerance in autoimmune diseases, or to generate specific tolerance to foreign organ grafts.

I hope this has been of some use to you, and that I have not managed to put you off immunology for life! If you are confused but still interested why not go back and read the simple version again?


Return to TIG Home Page

Steve Cobbold - updated 29th June 2009.