Immunology
Immune Tolerance
Self-tolerance — central negative selection, AIRE, regulatory T cells, and anergy
Immune tolerance is how the adaptive immune system learns not to attack the body's own tissues — deleting or disarming self-reactive lymphocytes so the roughly 10 million distinct antigen receptors we carry can still tell self from non-self. Because V(D)J recombination builds those receptors at random, a large fraction are inevitably self-reactive and must be removed. Tolerance solves this in two tiers: central tolerance, where developing T cells in the thymus and B cells in the bone marrow that bind self too strongly are deleted by negative selection — a process the AIRE transcription factor makes possible by forcing the thymus to display thousands of tissue-restricted self-proteins — and peripheral tolerance, where FOXP3+ regulatory T cells, anergy, and immune privilege restrain the self-reactive clones that slip through. The concept of acquired tolerance won Burnet and Medawar the 1960 Nobel Prize; when the system fails, autoimmunity follows.
- Two tierscentral (thymus/marrow) + peripheral
- Thymic deletionmost self-reactive T cells removed
- Key geneAIRE — displays tissue antigens
- Treg master TFFOXP3 (~5–10% of CD4+)
- Anergysignal 1 without signal 2
- NobelBurnet & Medawar 1960
Interactive visualization
Press play, or step through manually. The visualization is yours to drive — try it before reading on.
Watch the 60-second explainer
A condensed visual walkthrough — narrated, captioned, under a minute.
Why immune tolerance matters
- It is the price of a random repertoire. V(D)J recombination shuffles gene segments to build T- and B-cell receptors combinatorially, generating on the order of 107 distinct specificities in a person at any moment and a theoretical potential exceeding 1015. That randomness is what lets us recognize pathogens we have never met — but it guarantees that a substantial fraction of new lymphocytes recognize us. Tolerance is the mandatory quality-control step that makes an unbiased repertoire safe to deploy.
- Autoimmunity is the failure mode, and it is common. An estimated 5 to 8 percent of people develop an autoimmune disease, with roughly a 2:1 to 4:1 female predominance across most conditions. Type 1 diabetes, multiple sclerosis, rheumatoid arthritis, lupus, and autoimmune thyroid disease are all, at root, breakdowns of self-tolerance against a specific set of self-antigens.
- Transplantation lives or dies by it. The reason a kidney or heart graft is rejected is that the recipient was never tolerized to the donor's MHC alleles. The entire field of transplant immunology — from calcineurin inhibitors to tolerance-induction protocols — is an attempt to manufacture, pharmacologically, the tolerance that Medawar induced naturally in newborn mice in 1953.
- Pregnancy is a tolerance puzzle. A fetus expresses paternal MHC that is foreign to the mother, yet it is not rejected. Maternal-fetal tolerance depends heavily on regulatory T cells, the non-classical MHC molecule HLA-G at the placental interface, and local immunosuppressive signals — a natural experiment in peripheral tolerance that runs for nine months.
- Cancer immunotherapy is deliberate tolerance-breaking. Checkpoint inhibitors (anti-CTLA-4, anti-PD-1) work by loosening the same peripheral brakes that normally prevent autoimmunity. Their characteristic side effects — colitis, thyroiditis, hypophysitis, type 1 diabetes — are iatrogenic autoimmunity, proving in the clinic that these pathways enforce self-tolerance every day.
- Tolerance is antigen-specific, not general immunosuppression. A tolerant immune system still fights infection normally; it simply omits responses to a defined set of self-antigens. This specificity is the holy grail for treating autoimmune disease and preventing graft rejection without leaving the patient defenseless — the goal of antigen-specific tolerance therapies now in trials.
How immune tolerance works, step by step
1. Random receptors are built — including dangerous ones. In the thymus and bone marrow, developing lymphocytes rearrange their receptor genes by V(D)J recombination. The output is unbiased with respect to self: a newly minted thymocyte is just as likely to recognize insulin as influenza. Before it can be released, it must pass a two-part exam.
2. Positive selection — can you read self-MHC? In the thymic cortex, a developing T cell must bind self-peptide–MHC with at least weak affinity, or it dies by neglect. This ensures the mature repertoire is MHC-restricted — able to see peptides only when displayed on the body's own MHC molecules. Roughly 90 percent of thymocytes fail this step and are discarded.
3. Negative selection — but not too well. Cells that survive move to the thymic medulla, where medullary thymic epithelial cells (mTECs) and dendritic cells present a broad sample of self-peptides. A thymocyte whose receptor binds self-peptide–MHC too strongly is triggered to undergo apoptosis. This is central deletion — the single most important tolerance mechanism — and it removes the bulk of overtly self-reactive T cells before they leave.
4. AIRE paints a molecular self-portrait. The catch with step 3 is that most self-proteins live in distant organs and never reach the thymus. The AIRE (autoimmune regulator) transcription factor solves this: it drives mTECs to promiscuously express thousands of tissue-restricted antigens — insulin, thyroglobulin, retinal S-antigen, myelin proteins — genes that have no business being on in an epithelial cell. Each mTEC expresses a shifting subset, so across the medullary population the whole body's proteome is displayed, and T cells specific for peripheral tissues are deleted in the thymus. Without AIRE, those cells escape, and multi-organ autoimmunity (APECED) results.
5. B cells are edited and deleted in the marrow. Immature B cells that bind self-antigen in the bone marrow do not immediately die; they first get a second chance through receptor editing, re-rearranging their light-chain gene to change specificity. If editing fails to remove self-reactivity, the cell is deleted (clonal deletion) or rendered anergic. This is central B-cell tolerance.
6. Some escapees become the enforcers. A small fraction of thymocytes with intermediate self-affinity are not killed but diverted into the regulatory T-cell (Treg) lineage, switching on the master transcription factor FOXP3. These natural Tregs leave the thymus with the express job of suppressing self-reactive cells in the periphery — turning a potential liability into a safeguard.
7. Peripheral anergy disarms the survivors. Self-reactive T cells inevitably slip out. In the tissues, a T cell that meets its antigen on a resting, non-inflamed cell gets signal one (TCR–peptide-MHC) without signal two (CD28 costimulation from CD80/CD86). Rather than activating, it becomes anergic — functionally paralyzed, unable to make IL-2 or proliferate even if later fully stimulated. Two-signal dependence is the elegant logic that keeps quiescent tissues off-limits.
8. Tregs actively suppress, IL-2 gets rationed, CTLA-4 disarms APCs. FOXP3+ Tregs patrol lymph nodes and tissues, suppressing autoreactive cells by secreting IL-10, TGF-β, and IL-35; by soaking up local IL-2 through their high-affinity CD25 receptors so effector cells starve; and by using CTLA-4 to physically strip the costimulatory ligands CD80/CD86 off antigen-presenting cells, denying signal two to everyone nearby.
9. Immune privilege and ignorance cover the rest. Some self-antigens are simply hidden. Sites like the eye, brain, and testis are immune-privileged — protected by physical barriers, local immunosuppressive factors, and Fas ligand that kills incoming lymphocytes. Antigens the immune system never encounters at sufficient concentration are said to be ignored. This is why trauma that releases sequestered antigen (for example, sympathetic ophthalmia after eye injury) can suddenly trigger autoimmunity against a tissue that was tolerated only by being invisible.
Common misconceptions
- "Tolerance means the immune system ignores self." Passive ignorance is only a minor mechanism. Real tolerance is active: the thymus deliberately manufactures self-antigens via AIRE to delete reactive cells, and Tregs actively suppress others. The system does work to tolerate self, not merely overlook it.
- "Central tolerance is enough." It cannot be. Not every self-protein is expressed in the thymus at adequate levels, thymic deletion is incomplete, and some self-antigens (like gut microbiota-modified proteins or developmentally late antigens) are never seen there. Peripheral tolerance is a mandatory second layer, which is why FOXP3 loss alone (IPEX) is lethal despite intact thymic selection.
- "Self vs non-self is a molecular label the body reads." There is no chemical stamp for 'self.' Tolerance is learned negatively by subtracting self-reactive clones during development. Foreignness is inferred from the absence of prior tolerization plus innate danger signals — the reason Janeway's 'infectious non-self' and Matzinger's 'danger' models refined the original self/non-self picture.
- "Regulatory T cells just fail to respond." Tregs are not merely anergic bystanders. They are a dedicated, FOXP3-defined lineage that actively suppresses other cells through cytokines, IL-2 consumption, CTLA-4-mediated stripping of costimulation, and cytolysis. Deleting them (as in scurfy mice or IPEX patients) unleashes fatal multi-organ autoimmunity.
- "Anergy and deletion are the same thing." They are distinct outcomes. Deletion kills the self-reactive cell; anergy keeps it alive but functionally silenced; suppression is imposed from outside by Tregs. Anergy is also reversible under strong costimulation plus IL-2, which is one route by which infections occasionally break self-tolerance.
- "Autoimmunity means the immune system is weak." The opposite — autoimmunity is a failure of restraint, not of firepower. The effector machinery works fine; what has broken is the tolerance system that should have deleted or muzzled the offending clones. This is why treatment aims to restore regulation, not to boost immunity.
Central vs peripheral tolerance
| Feature | Central tolerance | Peripheral tolerance |
|---|---|---|
| Location | Thymus (T cells), bone marrow (B cells) | Blood, lymph nodes, tissues |
| Timing | During lymphocyte development, before release | On mature lymphocytes already in circulation |
| Main mechanism | Negative selection (clonal deletion); B-cell receptor editing | Treg suppression, anergy, deletion, immune privilege |
| Key molecule/gene | AIRE (drives tissue-antigen display in mTECs) | FOXP3 (Treg identity); CTLA-4, IL-2, PD-1 |
| Fate of self-reactive cell | Apoptosis, editing, or diversion to Treg lineage | Survives but suppressed, anergized, or later deleted |
| Coverage gap | Antigens not expressed in thymus escape | Backup for escapees; reversible under strong stimulation |
| Signature disease when it fails | APECED / APS-1 (AIRE mutation) | IPEX (FOXP3 mutation); ALPS (Fas mutation) |
Tolerance mechanisms compared
| Mechanism | Where | What happens to the self-reactive cell | Reversible? |
|---|---|---|---|
| Negative selection (deletion) | Thymus / bone marrow | Killed by apoptosis | No |
| Receptor editing | Bone marrow (B cells) | Re-rearranges light chain, changes specificity | N/A (specificity changes) |
| Treg diversion | Thymus | Becomes a FOXP3+ suppressor instead of dying | No (lineage-committed) |
| Anergy | Periphery | Alive but functionally paralyzed (no IL-2) | Yes (with signal 2 + IL-2) |
| Treg suppression | Periphery | Actively muzzled by another cell | Yes (if Tregs removed/blocked) |
| Immune privilege / ignorance | Eye, brain, testis | Never effectively meets its antigen | Yes (if antigen exposed by trauma) |
A famous experiment and history
- Owen's cattle twins (1945). Ray Owen noticed that non-identical (dizygotic) cattle twins sharing a placenta permanently carried each other's red blood cells as lifelong chimeras and never rejected each other's tissue. Exposure to foreign cells before birth had made the animals tolerant — the first concrete hint that self-recognition is learned, not innate.
- Burnet's prediction (late 1940s). Frank Macfarlane Burnet, developing his clonal selection theory, argued that the immune system must eliminate or silence clones that react to self during a critical developmental window. Tolerance, he proposed, is acquired — a testable claim that reframed immunology.
- Medawar's acquired tolerance (1953). Rupert Billingham, Leslie Brent, and Peter Medawar injected spleen cells from one mouse strain into fetal or neonatal mice of another strain. Once grown, those recipients accepted skin grafts from the donor strain indefinitely — grafts that normal adults reject in days. Published in Nature (1953), this was direct proof of actively acquired immunological tolerance. Burnet and Medawar shared the 1960 Nobel Prize.
- AIRE cloned (1997). Two groups identified the gene mutated in the rare autoimmune syndrome APECED as a transcriptional regulator. Later work in AIRE-knockout mice showed the mechanism: mTECs use AIRE to express tissue-restricted antigens, and without it, T cells specific for peripheral organs escape the thymus and cause multi-organ autoimmunity — establishing that central tolerance depends on the thymus building a self-portrait.
- Sakaguchi's suppressor cells and FOXP3 (1995–2003). Shimon Sakaguchi showed in 1995 that a CD4+CD25+ T-cell subset prevents autoimmunity — removing them from mice caused organ-specific autoimmune disease. In 2003 three groups identified FOXP3 as the master transcription factor defining these regulatory T cells, and connected its loss to scurfy mice and human IPEX syndrome, cementing Tregs as a distinct, essential arm of peripheral tolerance.
Frequently asked questions
What is the difference between central and peripheral tolerance?
Central tolerance happens in the primary lymphoid organs while lymphocytes are still developing — T cells in the thymus, B cells in the bone marrow. A developing T cell whose receptor binds self-peptide on MHC too strongly is killed by apoptosis (negative selection) or, in a small fraction, diverted into the regulatory T-cell lineage. B cells that bind self-antigen either edit their light-chain receptor or are deleted. Central tolerance is powerful but incomplete, because not every self-protein is displayed in the thymus at high enough levels. Peripheral tolerance is the backup system that operates in blood and tissues on mature lymphocytes that already escaped. It relies on regulatory T cells actively suppressing autoreactive cells, anergy (functional paralysis when a cell sees antigen without a costimulatory second signal), deletion by repeated stimulation, and immune privilege at sites like the eye and testis. Both layers are needed: mice or humans lacking either one develop autoimmunity.
What does the AIRE gene do?
AIRE (autoimmune regulator) is a transcriptional regulator expressed almost exclusively by medullary thymic epithelial cells. It forces those cells to transcribe thousands of tissue-restricted antigens — proteins that would normally appear only in the pancreas, thyroid, retina, or other organs — so that developing T cells encounter and are screened against them in the thymus. This 'promiscuous gene expression' lets negative selection delete T cells specific for insulin, thyroglobulin, or myelin before they ever leave. Loss-of-function mutations in AIRE cause APECED (also called APS-1), a rare monogenic autoimmune disease in which patients attack multiple endocrine organs — hypoparathyroidism, adrenal failure, and chronic candidiasis are the classic triad. AIRE was cloned in 1997, and its mechanism proved that the thymus does not simply present blood-borne proteins; it deliberately manufactures a molecular self-portrait of the entire body.
How do regulatory T cells prevent autoimmunity?
Regulatory T cells (Tregs) are a CD4+ subset defined by the master transcription factor FOXP3 and, usually, high CD25 (the IL-2 receptor alpha chain). They make up roughly 5 to 10 percent of circulating CD4+ T cells and actively police self-reactive cells rather than just failing to respond. They suppress by several routes: secreting inhibitory cytokines IL-10, TGF-beta, and IL-35; consuming local IL-2 so effector cells starve; expressing CTLA-4, which strips the costimulatory ligands CD80/CD86 off antigen-presenting cells; and delivering direct cytolysis via granzyme. Their importance is stark in disease: mutations in FOXP3 cause IPEX syndrome (immune dysregulation, polyendocrinopathy, enteropathy, X-linked), a fatal early-childhood autoimmunity, and the same lesion in mice produces the scurfy phenotype. Blocking CTLA-4 with checkpoint-inhibitor cancer drugs deliberately loosens this brake, which is why those therapies frequently trigger autoimmune side effects.
What is anergy in immunology?
Anergy is a state of long-lived functional unresponsiveness in a lymphocyte that has seen its antigen but not the full activating context. A naive T cell needs two signals to switch on: signal one is its T-cell receptor binding peptide-MHC, and signal two is costimulation, chiefly CD28 engaging CD80/CD86 on a properly activated antigen-presenting cell. If a T cell receives signal one without signal two — as happens when it meets self-antigen on a resting, non-inflamed tissue cell — it does not die but becomes anergic: it cannot make IL-2 or proliferate even if later given both signals. Molecularly, anergy involves calcium-driven NFAT activation without AP-1, inducing genes like GRAIL, Cbl-b, and DGK that raise the activation threshold. Anergy is a peripheral-tolerance mechanism, distinct from deletion (the cell survives) and from Treg suppression (it is cell-intrinsic). It can be reversed if the cell is later exposed to strong costimulation plus IL-2, which is one way infections can occasionally break self-tolerance.
How does the immune system tell self from non-self?
There is no single 'self' molecular tag. Instead the immune system learns self negatively, by education. Lymphocytes generate antigen receptors at random through V(D)J recombination, producing an enormous, unbiased repertoire that recognizes essentially any shape — including the body's own. Tolerance then subtracts the self-reactive members. In the thymus, T cells are positively selected to recognize the body's own MHC (or they are useless) and then negatively selected to not bind self-peptide too tightly (or they are dangerous). What survives is a repertoire tuned to see foreign peptides presented on self-MHC. This is why 'self vs non-self' is better stated as 'tolerated vs not-yet-tolerated': the system does not detect foreignness directly, it detects the absence of prior tolerization, supplemented by innate danger signals (PAMPs and DAMPs) that license a response only when tissue damage or infection is genuinely present — the core of Polly Matzinger's danger model and Janeway's infectious-non-self model.
What happens when immune tolerance fails?
When tolerance fails, self-reactive lymphocytes attack the body's own tissues, causing autoimmune disease — a category affecting an estimated 5 to 8 percent of people, with a strong female bias. The target organ depends on which self-antigen is attacked: pancreatic beta cells in type 1 diabetes (anti-insulin and anti-GAD65 T cells), the myelin sheath in multiple sclerosis, thyroid peroxidase and the TSH receptor in Hashimoto's and Graves' disease, joint synovium in rheumatoid arthritis, and nuclear antigens systemically in lupus. Failure can be monogenic and catastrophic — AIRE mutations cause APECED, FOXP3 mutations cause IPEX, Fas mutations cause ALPS — or, far more commonly, polygenic, where dozens of risk alleles (especially particular HLA/MHC haplotypes) combine with environmental triggers like infection, which can mimic self through molecular mimicry or expose sequestered antigens. Modern checkpoint-inhibitor cancer drugs deliberately break peripheral tolerance and produce autoimmune toxicities as proof of the mechanism.
How was immunological tolerance discovered?
Frank Macfarlane Burnet predicted in the late 1940s that the ability to distinguish self from non-self must be learned during development, not inherited fixed — a corollary of his clonal selection theory. Peter Medawar's group tested it experimentally: in 1953 they injected cells from one mouse strain into fetal or newborn mice of a second strain, and those recipients, once grown, permanently accepted skin grafts from the donor strain that unmanipulated adults would have rejected. This 'actively acquired tolerance' showed the developing immune system can be taught to accept foreign tissue as self if it meets the antigen early enough. Ray Owen's 1945 observation that non-identical cattle twins sharing a placenta carried each other's blood cells for life foreshadowed it. Burnet and Medawar shared the 1960 Nobel Prize in Physiology or Medicine. The mechanistic layers — thymic negative selection, AIRE (1997), and FOXP3 Tregs (Sakaguchi's 1995 CD25 work; FOXP3 identified 2003) — were filled in over the following decades.