Immunologic tolerance is a state of immune unresponsiveness specific to a particular antigen or set of antigens induced by previous exposure to that antigen or set. Tolerance is generally accepted to be an active process and, in essence, a learning experience for T cells. It is said to occur mechanistically at 2 levels: centrally and peripherally. Induction of immunologic tolerance has been achieved and studied in numerous laboratory animal models, but it remains an elusive goal in clinical organ transplantation and in the management of autoimmune disease in humans.
Transplant tolerance is defined as a state of donor-specific unresponsiveness without a need for ongoing pharmacologic immunosuppression. Transplantation tolerance could eliminate many of the adverse events associated with immunosuppressive agents. A strategy to develop tolerance would also be of value in autoimmune diseases, such as juvenile-onset diabetes mellitus, rheumatoid arthritis, and multiple sclerosis.
Safe, reliable strategies for the induction of full tolerance have not yet been developed. However, during the study of achieving immune tolerance, ways in which to induce states of “partial tolerance” have been discovered, in which lower-than-conventional amounts of ongoing pharmacologic immunosuppression are needed. Nonetheless, immune tolerance remains the holy grail of transplantation immunology and clinical transplantation.
One of the first clinically relevant observations pertaining to transplantation was made by Dr. Earl Padgett in the 1920s, when skin grafts were noted to survive longer when skin from close relatives was used. 
Ten years later, a fellow American plastic surgeon, Dr. Barrett Brown, extended these observations when he transplanted a skin graft from the identical twin of the patient and achieved permanent survival of the skin graft.  Seemingly impossible without genetic identity, the field of transplantation was left silent for another 20 years.
In 1945, Dr. Ray Owen, who had a special interest in bovine blood groups, discovered that most dizygotic-twin cattle fetuses had placental anastomoses with each other and that they shared their blood supply in utero. These cattle maintained a stable mixture of each others’ red cells throughout their lives.
Dr. Owen observed that despite the twins’ having distinctly different blood groups, transfusions between them did not cause any transfusion reaction, which does occur between nontwin cattle of different blood groups. This mosaicism was an early example of immune tolerance. [3, 4]
In 1947, Rupert Billingham and Peter Medawar were studying the use of skin allografts in young cattle to distinguish identical and fraternal twins. This work was done to identify the sterile female twin of a male calf (ie, freemartin), a point of agricultural importance because the freemartin female bovine was totally useless to breeders and dairy farmers.
Billingham and Medawar found that the skin grafts were essentially accepted regardless of whether the cattle were identical or fraternal. However, skin grafts exchanged between unrelated cattle were always rejected. The study did not help in distinguishing the 2 kinds of twins.
In 1949, Billingham and Medawar came across Owen’s work in a monograph by Frank Macfarlane Burnet and Frank Fenner entitled “The Function of Antibodies.” This publication helped them make sense of their work. The authors postulated that the age of the animal at the time of its first encounter with a foreign body was the critical factor in determining its responsiveness and, hence, its recognition of nonself-antigens.
Billingham, Medawar, and Leslie Brent, [5, 6] Medawar’s postgraduate student, now envisioned the possibility of having adult animals accept tissue allografts by reproducing in the laboratory what occurred naturally in cattle. They aimed to prove that tolerance could be artificially induced.
If living cells from mouse strain CBA were injected into an adult mouse of strain A, some immunologic process destroyed the CBA cells, and the A-line mouse that received the CBA cells quickly destroyed any subsequent graft from the same donor strain. However, if the CBA cells were injected into an A-line fetus or newborn that was still immunologically immature, the CBA cells were accepted, and any future grafts from the strain A donor were accepted. 
Although these findings were not clinically applicable, they were considered to represent the first experimental breakthrough needed to initiate active research in the field of transplantation tolerance.  In 1960, Drs. Burnet and Medawar shared the Nobel Prize in Physiology or Medicine for their work in acquired immunologic tolerance to tissue grafts. 
New approaches in the field of transplantation immunology to induce allograft survival include the use of humanized mice—that is, immunodeficient mice engrafted with functional human immune systems. These currently provide a small animal model for the study of human immune responses. 
Go to Immunotherapeutic Targeting in Children for complete information on this topic.
Central (Intrathymic) Mechanisms of Tolerance
The chief mechanism of T-cell tolerance is the deletion of autoreactive T cells in the thymus so that the organism is rendered self-tolerant. Immature T cells migrate from the bone marrow to the thymus, where they encounter peptides derived from endogenous proteins bound to major histocompatibility complex (MHC) molecules on thymic epithelial cells.
Double-positive (CD4+ and CD8+) thymocytes initially undergo random generation of different T-cell receptors (TCRs). Positive selection, also called thymic education, ensures that only clones with TCRs and moderate affinity for self-MHC are allowed to develop. Negative selection by means of apoptosis (programmed cell death) occurs when T cells do not produce functional TCRs, when TCR rearrangement fails, when T cells have low affinity for the MHC–self-peptide complex, or when T cells have extremely high affinity for such complexes.
Negative selection also results in the deletion of some thymocytes that have TCRs and reactivity and some thymocytes that interact with autoantigens presented by interdigitating cells and macrophages at the corticomedullary junction. The remaining cells lose either CD4 or CD8 and leave the thymus to function in the periphery as mature, functional CD4+ or CD8+ T cells.
Peripheral (Nonthymic) Mechanisms of Tolerance
Many potentially reactive T cells escape intrathymic deletion; this reflects the fact that many antigens are absent intrathymically or present at insufficient levels to induce tolerance in the thymus. Several peripheral (nonthymic) mechanisms prevent autoimmunity and are also capable of rendering peripheral T cell repertoires tolerant.
Sequestration of antigens into privileged sites
Some antigens are sequestered into privileged sites away from the immune system because of physical barriers, such as tight junctions, or immunologic barriers, such as expression of Fas ligand (FasL) or little expression of major histocompatibility complex (MHC) class I. Antigen-presenting cells (APCs), and subsequently T lymphocytes, may never encounter these self-antigens. Therefore, they remain ignorant of these antigens.
At some of these sites, proinflammatory lymphocytes are controlled by apoptosis resulting from the expression of FasL or the secretion of cytokines such as transforming growth factor-beta (TGF-beta) or interleukin (IL)–10.
When T cells enter these sites, their Fas interacts with the FasL of these sites, and they undergo apoptosis. Privileged sites include the brain, the testes, and the anterior chamber of the eye. Transplanted tissues are most likely to survive in these privileged sites because of the tight control of proinflammatory lymphocytes.
Apoptosis of T cells caused by persistent activation or neglect
Apoptosis, or programmed cell death, of lymphocytes is an important mechanism of immune control and homeostasis. Apoptosis contributes to the deletion of clones that are persistently activated and of activated lymphocytes when the immune response is no longer needed (eg, after an infection clears). Cells that are persistently stimulated undergo activation-induced cell death involving Fas-FasL signaling or tumor necrosis factor. Most T cells that remain after antigens are no longer present are deprived of the stimuli to survive and undergo passive cell death.
Apoptosis of donor-reactive lymphocytes is also known as the “deletional” method to induce tolerance and, in theory, represents the most fail-safe mechanism of tolerance. In the absence of donor-reactive lymphocytes, no matter what the immunological encounter, the response to donor antigens could not be induced.
T lymphocytes require 2 signals to become activated, to proliferate, and to differentiate. The first is the recognition of an appropriate MHC-antigen complex by the RT-cell receptors (TCRs) of the responsive lymphocyte. The second is delivered by costimulatory molecules also expressed by APCs; these are only able to engage once the first signal is activated. Lack of costimulation causes anergy; that is, T cells fail to respond to the MHC-peptide complex and remain unresponsive to subsequent challenges.
CD28 is the main costimulatory ligand expressed by naïve T cells encountering an antigen. Signaling by means of CD28 enhances T-cell proliferation by boosting IL-2 production by T cells, which promotes activation and proliferation. It also enhances expression of CD40 ligand, which interacts with CD40 on APCs and which induces upregulation of costimulatory molecules CD80 (B7-1) and CD86 (B7-2) to enhance costimulatory signaling to responding T cells.
Rigby et al have used blocking of the T-cell second signaling as an effective means of preventing autoimmunity and allograft rejection in many animal models. They studied the effects of anti-CD28 and CTLA4-Ig on diabetes development and the requirements to induce tolerance in nod.scid mice after the transfer of transgenic beta-cell reactive BDC2.5.NOD T-cells. 
Rigby and associates were successful in this set of experiments, and their work has helped to develop the understanding of natural regulatory mechanisms that may have a unique role in establishing targeted long-standing immune protection and peripheral tolerance. 
T lymphocytes also express CD152 (CTLA-4) after CD28 binds to B7-1 and B7-2 on APCs. The interaction of CTLA-4 and B7 molecules decreases opportunities for B7-CD28 binding and down-regulates T-cell activities (eg, production of IL-2) to reduce T-cell proliferation. CD28 interacts with B7 molecules, first leading to T-cell activation. However, after this effect peaks, upregulation of CTLA-4 with its relatively high affinity for B7 molecules limits the degree of activation.
Verbinnen et al recently studied the involvement of regulatory T cells (Tregs) and deletion of alloreactive cells in the induction and maintenance of tolerance after costimulation blockade (CTLA-4) in a mouse model of graft-versus-host reaction.  The study showed that clonal deletion of host-reactive T cells was a major mechanism responsible for tolerance. 
Regulatory T cells
Tregs, also called suppressor T cells, suppress the activation of clone-specific T-cell activity. Tregs account for 10-15% of CD4+ T cells and express a transmembrane protein called CD25, an alpha chain of the receptor for IL-2. These CD4+ CD25+ Tregs are anergic to TCR-mediated activation but potently suppress the activation of other T cells.
Not all CD25+ T cells are regulators. Some naive T cells upregulate CD25 in response to antigen, a change that represents an active rather than suppressive immune response. The thymus produces anergic but suppressive CD4+ CD25+ T cells (which are also identified by the expression of FoxP3, the transcription factor responsible for their development). These T cells suppress activation and expansion of autoreactive CD4+ CD25– populations.
Studies in mice have shown that Tregs are antigen specific and that they regulate peripheral tolerance by producing suppressive cytokines, such as IL-10 and TGF-beta. They depend on continuous antigen exposure to stay active. Removal of the antigen reduces the quantity of cells.
In allograft rejection, direct stimulation of T cells in response to donor antigen presented by donor APCs had been the focus of transplantation for many years. However, indirect antigen presentation, in which self-APCs present donor peptide in an MHC-restricted fashion, is responsible for the induction of antigen-specific Tregs that can directly and indirectly suppress other alloreactive T cells. 
Positive costimulation with CD28 appears to be necessary for the development of intrathymically derived Tregs, but costimulation blockade with CTLA-4 is needed for peripherally acquired suppressor Tregs to develop.
Tregs are also responsible for maintaining tolerance by broadening suppression through what is termed linked suppression to additional antigens expressed in the tolerated tissue and to further cohorts of naïve T-cells as they develop. Immunologist Herman Waldmann described this phenomenon as a process of infectious tolerance.
Tregs from tolerant animals can be transferred to naïve animals, in which they subsequently confer antigen-specific tolerance, including tolerance to skin and organ allografts. This is referred to as adoptive tolerance and has been recognized, though not completely understood, since the 1990s, when Dr. Metcalfe at Cambridge published a landmark paper describing this tolerant phenomenon. 
Tolerance induction via expansion and transfer of donor Tregs to an allograft recipient or via ex vivo development of Tregs from recipient T cells is an intriguing but yet-untested strategy in humans. Currently, in heart transplantation, analysis of whether the FOXP3 gene expression in the peripheral blood reflects antidonor immune responses is under way; this is an exciting approach to broadening the translational approach of tolerance induction by way of Tregs. 
Oral administration of protein antigen is used to induce specific immunologic unresponsiveness and may be a potential future therapy for some autoimmune disorders. [15, 16, 17] The specific mechanisms of how oral antigen induces tolerance are still unclear, and its importance in clinical transplantation is unlikely to be relevant.
Studies have demonstrated that feeding antigen in high dosages results in the deletion of reactive T cells, but it does not promote the generation of Tregs. Small doses of antigen, administered orally or intranasally, induce antigen-specific antibodies, increase levels of the inhibitory cytokines IL-4 and IL-10, and induce Tregs in gut-associated lymphoid tissue (GALT).
IL-4 is a factor involved in the differentiation of TGF-beta–secreting T cells from naïve splenic T cells, also known as T helper 3 (Th3) cells.  Th3 cells have suppressive properties for T helper 1 (Th1) and T helper 2 (Th2) cells.
A phenomenon known as bystander suppression is observed with oral tolerance because oral antigen–induced Tregs also secrete antigen-nonspecific cytokines after fed antigen triggers them. Nonspecific cytokines can suppress inflammation in the microenvironment where the fed antigen is localized. Therefore, knowledge of the specific antigen being targeted might not be needed as long as the fed antigen can induce Tregs to get to the target and suppress inflammation.
Oral administration of antigens has suppressed or reversed autoimmune disorders in several animal models; thus, they can induce self-tolerance. However, to date, attempts at oral tolerance have not been successful in any human autoimmune diseases, including diabetes, rheumatoid arthritis, and multiple sclerosis.
Other major antigens encountered during transplantation include ABO blood-group antigens, present on many tissues and on red blood cells (RBCs). ABO compatibility depends on the presence or absence of specific antibodies in the recipient’s serum against specific antigens on donor RBCs.
Adults typically have preformed antibodies to donor blood-group antigen. Therefore, transplantation of solid organs from ABO-incompatible donors is contraindicated because of the risk of hyperacute rejection mediated by the preformed antibodies to the donor RBCs. This contraindication does not apply to newborns and infants, because their ABO isohemagglutinins are not yet formed or well developed. 
Dr. Lori West and colleagues, in a study of 10 infants and children aged 4 hours to 14 months who received ABO-incompatible heart transplants between 1996 and 2000, reported an 80% survival rate with 2 early deaths that were not attributed to ABO incompatibility; no morbidity was attributed to ABO incompatibility.  Plasma exchange was done during cardiopulmonary bypass to remove antibodies, and immunosuppression therapy was given afterward. Rejection was monitored by means of endomyocardial biopsy. Follow-up ranged from 11 months to 4.6 years.
The findings reported by West et al mimic findings from animal models of neonatal tolerance in which immunologically immature recipients could accept incompatible grafts and develop tolerance. Recent follow-up with these recipients reveals that this strategy to accept ABO-incompatible donor hearts for infant transplantation significantly improves the likelihood of transplantation and reduces waiting list mortality without adversely altering outcomes after transplantation. 
ABO-incompatible heart transplantation during infancy results in the development of B-cell tolerance to donor blood-group A and B antigens. This tolerance occurs by eliminating donor-reactive B lymphocytes, which may depend on persistence to some degree of antigen expression.
In adults, transplantation of ABO-incompatible livers has been performed in emergency situations with reasonable success, though reported cases have not involved tolerance. Transplantation of ABO-incompatible organs in adults by using the antibody-depleting strategies of plasmapheresis or splenectomy or use of anti-CD20 antibody, along with intensive immunosuppression, has been successfully undertaken in a few centers.
Of interest, long-term plasmapheresis is not necessary, and ongoing graft function with immunosuppression can be achieved despite the return of antibodies against donor blood groups.
Induction of Tolerance in Transplant Patients
Clinical research surrounding clinical allograft transplantation has been conducted to induce full or partial tolerance in transplant patients. These strategies are not yet ready for general clinical use and will not be so until further evidence-based studies are available.
The holy grail of organ transplantation is full immunologic tolerance, a state of indefinite survival of a well-functioning allograft, without the need for maintenance immunosuppression. In addition, the host must retain a normal immune response and thus not suffer from immunosuppression-related infections, neoplasia, or other drug-related adverse effects. Rare cases of operational tolerance after transplantation, with complete cessation of immunosuppressive therapy (usually because of noncompliance), have been reported.
Most studies of the intentional induction of immunologic tolerance have involved patients with hematologic malignancies. Full tolerance was achieved with myeloablative therapy before organ transplantation in combination with induced donor chimerism by means of bone marrow transplantation and excellent human leukocyte antigen (HLA) matching.  Mixed chimerism retains a graft-versus-host T-cell effect that allows for transplant acceptance despite subsequent disappearance of the donor chimerism.
Myeloablative therapy includes total-body irradiation and lymphoablative methods, such as total lymphoid irradiation and use of azathioprine and corticosteroids. However, the complications of full tolerance and the untimeliness of donor organs with regard to preparation time for myeloablative therapy before transplantation preclude routine application of these therapies.
Cosimi and Sachs, in a study of mixed chimerism involving a small number of patients, found that patients had transient chimerism for several weeks, and graft survival was approximately 70% in the long term.  Nonmyeloablative conditioning was used, such as peritransplantation low-dose total-body irradiation or thymic irradiation plus antithymocyte globulin therapy combined with splenectomy. Donor-specific marrow infusion was given at the time of transplantation. Cyclosporine was given for about a month after transplantation and then stopped.
Bühler et al adopted this protocol for patients with multiple myeloma and end-stage renal disease who needed kidney transplantation (see Renal Transplantation) and found no evidence of chronic rejection in the longest surviving patient over 5 years.  However, they modified the regimen to replace total-body irradiation with 2 doses of cyclophosphamide 60 mg/kg given intravenously before transplantation, without splenectomy. Immunosuppressive therapy was withdrawn.
This strategy was also attempted in patients with end-stage renal disease without malignancy who received HLA-mismatched kidneys; although allograft tolerance due to mixed chimerism could be achieved in these patients, intense immunosuppression of humoral responses was necessary. 
A caveat to these interesting reports is that some were described in meeting abstracts, and only some of the approaches have been reported in full papers.
At present, partial tolerance or minimal immunosuppression is possible.  This allows minimal use of immunosuppressive drugs, which results in fewer risks of infection, neoplasia, or drug-related adverse effects. This partial, or incomplete, donor-specific tolerance has been termed prope tolerance  (from Latin prope “near”) or minimal immunosuppression tolerance. 
Professor Sir Roy Calne postulated that prope tolerance preserves some of the transplant recipient’s immune responses to infection and other antigens, reducing morbidity and mortality due to immunosuppressive effects.  In the late 1990s, his group reported 31 patients treated with Campath-1H 20 mg given 1 day after renal transplantation to reduce T and B cells by 1-2 logs. This was followed by half-dose cyclosporine monotherapy started on the third day. Trough cyclosporine levels were kept between 50 and 100 mg/dL.
In the study by Calne et al, 29 patients had good graft function for over 5 years, and 3 had rejection necessitating additional immunosuppressive therapy.  Six other steroid-responsive graft rejections were reported. Long-term results were similar to those observed in a comparator full-immunosuppression group.
Kirk et al, seeking to induce tolerance in renal allografts by inducing profound lymphocyte depletion with alemtuzumab (Campath-1H), found that this therapy achieved acute lymphocyte depletion but failed to induce tolerance.  . All patients had reversible rejection episodes in the first 3 weeks. They exhibited predominant monocyte graft infiltration, and all were responsive to steroids and/or sirolimus. Reduced immunosuppression, usually with sirolimus, was continued without further rejection episodes, despite a recovery of normal lymphocyte counts.
Knechtle et al studied alemtuzumab (Campath-1H) and low-dose sirolimus and determined that about 30% of kidney recipients had early rejection.  Most cases were secondary to humoral responses.
Starzl, Shapiro, and colleagues compared alemtuzumab (Campath 1H) with tacrolimus monotherapy and reported a low rate of rejection episodes. 
These data indicate that alemtuzumab-related lymphocyte depletion at time of transplantation followed by calcineurin-inhibitor monotherapy (with either cyclosporine or tacrolimus) was a successful minimization strategy at 2 transplant centers. It did not increase rejection episodes and did not impair graft survival in the short-to-medium term.
Although researchers at numerous laboratories are investigating assays to monitor the degree of immunosuppression, no assays or tests are currently available to monitor tolerance. Dr. Sarwal at Stanford University is currently using exciting microarray technology to identify genetic identifiers of allograft recipients that are rendered tolerant. [33, 34] This technology may very well be what is required to overcome the barriers to allograft tolerance.
Identifying either prope or complete tolerance depends on the elimination, withdrawal, or reduction of maintenance immunosuppression and on the observation of a favorable response. Protocol allograft biopsy may or may not be helpful in identifying rejection at an early stage if the strategy is unsuccessful. Indeed, a specific directive of granting agencies, such as the National Institutes of Health and the Immune Tolerance Network, is to fund research to develop tolerance assays. Biomarkers capable of identifying and predicting operational tolerance in candidates for liver and kidney transplants have been identified, and biomarker-led prospective immunosuppression withdrawal trials are in progress. 
The demonstration of immune tolerance induction in many rodent models stands in stark contrast to the lack of success in humans and primates, with the exception of myeloablative therapy followed by donor-derived stem cell infusion. The specific pathogen–free environment in which rodents are housed for their lifetimes limits the number of memory T cells that such animals generate.
On the other hand, humans and primates are exposed to many viruses during their long and less pathogen-free lives. In addition, they generate a considerable pool of self-renewing memory T cells (nearly half of circulating T cells in humans). Therefore, they are less immunologically naïve than experimental rodents.
Many of these memory T cells can cross-react with foreign major histocompatibility complex (MHC). Therefore, the translation of tolerance induction strategies from the rodent laboratory to large animals and then to humans may need to account for previous specific and net immunologic memory.