T-Cell Activation and the Two-Signal Hypothesis

CD4⁺ and CD8⁺ T cells leave the thymus and enter the circulation as naïve T cells. Although relatively mature (see Chapter 8), they have not yet encountered antigen. Their chromatin is condensed, they have very little cytoplasm, and they exhibit little transcriptional activity. However, they are mobile cells and recirculate continually among the blood, lymph, and secondary lymphoid tissues, including the lymph nodes, browsing for antigen. It is estimated that each naïve T cell recirculates from blood through lymph nodes and back again every 12 to 24 hours. Because only about 1 in 10⁵ naïve T cells is likely to be specific for any given antigen, this large-scale recirculation increases the chances that a particular T cell will find “its” antigen.

If a naïve T cell does not bind any of the MHC-peptide complexes it encounters as it browses the surfaces of antigen-presenting cells in the secondary lymphoid tissue, it exits and rejoins the circulation to try again in another tissue. If it has been browsing through a lymph node or Peyer’s patch, it exits via the efferent lymphatics, ultimately draining into the thoracic duct and reentering the blood via the vena cava (see Figures 2-12 and 2-13). If it has been browsing the T-cell zones of the spleen, it exits directly into the blood (see Figure 2-15). However, if a naïve T cell does encounter an APC expressing an MHC-peptide complex to which it binds with high affinity, it stops migrating and initiates an activation program that produces a diverse array of cells that orchestrate the short-term and long-term responses to infection.

Early investigations revealed the need for not just one, but two signals to activate naïve cells (the two-signal hypothesis). Signal 1 is triggered by TCR engagement and Signal 2 by engagement of costimulatory molecules, such as CD28. Even though it is often still referred to as the two-signal hypothesis, full T-cell activation actually requires a third set of signals, provided by local cytokines (Signal 3), that directs the differentiation of T cells into distinct effector cell types (Overview Figure 10-2).

OVERVIEW FIGURE 10-2

Three Signals Are Required for Activation of a Naïve T Cell

The TCR/MHC-peptide interaction, along with CD4 and CD8 coreceptors and adhesion molecules, provide Signal 1. Costimulation by a separate set of molecules, including CD28 (or ICOS, not shown) provides Signal 2. Together, Signal 1 and Signal 2 initiate a signal transduction cascade that results in activation of transcription factors and cytokines (Signal 3) that direct T-cell proliferation (IL-2) and differentiation (polarizing cytokines). Cytokines can act in an autocrine manner, by stimulating the same cells that produce them, or in a paracrine manner, by stimulating neighboring cells.

A na�ve T cell binds to an APC. The first is the TCR signaling which is initiated when the TCR of the na�ve T cell binds to the MHC protein of APC. The second signal is initiated through costimulatory interaction when the CD28 coreceptor of the na�ve T cell binds to the CD80 or CD86 coreceptor of the APC. The third is the cytokine signaling which occurs in a paracrine or in an autocrine manner. In paracrine signaling, cytokines such as IL-12 from the APC stimulate the na�ve T cell. In autocrine signaling, cytokines such as IL-2 generated by the na�ve T cell, bind to its own receptors and stimulate the na�ve T cell. The signals are transmitted to the nucleus of the na�ve T cell, leading to gene expression.

A successful T cell–APC interaction results in the stable organization of signaling molecules into an immune synapse (Figure 10-3). The TCR/MHC-peptide complexes, which deliver Signal 1, aggregate in the central part of this synapse (the central supramolecular activating complex, or cSMAC). The intrinsic affinity between the TCR and MHC-peptide surfaces is actually quite low (K_d ranges from 10^-4 M to 10^-7 M). Signal 1, in fact, is stabilized by the activity of several other molecules, which together increase the avidity (the combined affinity of all cell-cell interactions) of the cellular interaction. The coreceptors CD4 and CD8, which are also found in the cSMAC, are key participants and bind MHC class II and MHC class I molecules, respectively.

FIGURE 10-3 Surface interactions responsible for T-cell activation. (a) A successful T-cell/dendritic-cell interaction results in the organization of signaling molecules into an immune synapse. A scanning electron micrograph (left) shows the binding of a T cell (artificially colored yellow) and dendritic cell (artificially colored blue). A fluorescence micrograph (right) shows a cross-section of the immune synapse, where the TCR is stained with fluorescein (green) and adhesion molecules (specifically LFA-1) are stained with phycoerythrin (red). Other molecules that can be found in the central part of the synapse (central supramolecular activation complex [cSMAC]) and the peripheral part of the synapse (pSMAC) are listed. (b) A schematic of the interactions between a CD4⁺ T cell (left) or CD8⁺ T cell (right) and its activating dendritic cell. Dendritic cells (on the right in each diagram) process antigens and present peptides associated with MHC class II to CD4⁺ T cells and present peptides associated with MHC class I to CD8⁺ T cells. Binding of TCR to MHC-peptide is enhanced by the binding of coreceptors CD4 and CD8 to MHC class II and class I, respectively. CD4 and CD8 also associate with Lck, which helps initiate signaling. CD28 interactions with CD80/86 provide the required costimulatory signals. Adhesion molecule interactions, two of which (LFA-1/ICAM-1, CD2/LFA-3) are depicted, markedly strengthen the connection between the T cell and APC or target cell so that signals can be sustained.

A scanning electron micrograph on the left shows a small, round, yellow cell bound to a large, blue, irregularly-shaped antigen-presenting cell that has numerous fingerlike protrusions on its surface. A fluorescence micrograph on the right shows a small, yellow region at the center, surrounded by a green region. A large red region surrounds the yellow and green regions. A callout points to the yellow region and reads "cSMAC: TCR or CD3; Coreceptors CD4 or CD8; Costimulatory receptors (CD28)." A callout points to the red region and reads "pSMAC: Adhesion molecules LFA-1 or ICAM-1, LFA-3 or CD2."

A CD4 superscript plus T cell binds to an antigen-presenting cell. The CD2 of the CD4 superscript plus T cell binds to the LFA-3 of the antigen-presenting cell. The TCR or CD3 complex and CD4 of the CD4 superscript plus T cell bind to the MHC class II of the antigen-presenting cell. An Lck protein is between the CD4 coreceptor and the TCR or CD3 receptors. The CD28 coreceptor of the CD4 superscript plus T cell binds to the CD80 or CD86 coreceptor of the antigen-presenting cell. The LFA-1 of the CD4 superscript plus T cell binds to the ICAM-1 of the antigen-presenting cell. TCR or CD3, CD4, and CD28 belong to the cSMAC. LFA-1, ICAM-1, LFA-3, and CD2 belong to the pSMAC and are located on either side of the receptors belonging to cSMAC. The antigen is presented as a peptide to the CD4 superscript plus T cell through the interaction between the TCR or CD3 receptor and the MHC class II molecule.

Similarly, a CD8 superscript plus T cell binds to an antigen-presenting cell. The CD2 of the CD8 superscript plus T cell binds to the LFA-3 of the antigen-presenting cell. The TCR or CD3 complex and CD8 of the CD8 superscript plus T cell bind to the MHC class I of the antigen-presenting cell. The CD28 coreceptor of the CD8 superscript plus T cell binds to the CD80 or CD86 coreceptor of the antigen-presenting cell. An Lck protein is between the CD8 and the CD28 coreceptors. LFA-1 of the CD8 superscript plus T cell binds to the ICAM-1 of the antigen-presenting cell. The antigen is presented as a peptide to the CD8 superscript plus T cell through the interaction between the TCR or CD3 receptor and the MHC class I molecule.

In contrast, adhesion molecules and their ligands (e.g., LFA-1/ICAM-1 and LFA-3/CD2) organize themselves around the perimeter of the central aggregate, forming the peripheral supramolecular activating complex, or pSMAC. Interactions between these molecules help to sustain the signals generated by allowing long-term cell interactions.

Even the increased avidity offered by CD4 (or CD8) and adhesion molecules is still not sufficient to fully activate a naïve T cell. Interactions between costimulatory receptors on T cells, including CD28, and costimulatory ligands on antigen-presenting cells, including CD80/86, provide a required second signal (Signal 2).

TCR Signaling Provides Signal 1 and Sets the Stage for T-Cell Activation

How do TCR-MHC interactions generate signals that trigger T-cell proliferation and differentiation? What do costimulatory molecules contribute? Signaling cascades can be at once daunting and dull to study, but they are a vital part of the immune response, which is shaped by the information it receives through these pathways. Here, we describe a few general features of TCR signaling (Signal 1). This outline also provides a useful framework for understanding many other receptor signaling pathways. The next section will address the contributions of Signal 2, the costimulatory signals that are also required for full T-cell activation. We will also discuss the contributions of coinhibitory signals, which provide the T cell with a way to turn off signaling cascades.

When trying to make sense of signaling, it helps also to be guided by a keen observation once made by a young professor: Signaling is all about location, location, location!

Tyrosine Kinases and Initiation of TCR Signaling

TCR signaling begins with the activation of a tyrosine kinase known as Lck, a member of the Src family (see Chapter 3). Once a TCR engages MHC-peptide on the surface of an APC, the coreceptor CD4 or CD8 steps in to stabilize this interaction by binding invariant regions of MHC. CD4 and CD8 also provide another key function. Their cytoplasmic tails associate with Lck, which comes along for the ride (see Figure 10-3b). Now in the direct vicinity of the TCR-CD3 complex, Lck can phosphorylate the tyrosines in the ITAMs found in the tails of the CD3 molecules (see Chapter 3).

(An important side note is that Lck, itself, needs to be activated before it is able to act. The details behind this activation are still debated, but, essentially, clustering of molecules triggered by TCR engagement also brings Lck closer to the membrane-associated tyrosine phosphatase CD45. CD45 removes an inhibitory phosphate group on Lck, which is then phosphorylated at an activating tyrosine site by neighboring Lck proteins. Only then can it phosphorylate the CD3 tails.)

Tyrosine phosphorylation of ITAMs generates new docking sites for proteins with SH2 domains, including ZAP-70, another T cell–specific tyrosine kinase. ZAP-70 joins the signaling complex and is phosphorylated and activated by Lck, its new neighbor. ZAP-70 is now in a position to phosphorylate multiple neighboring proteins. The adapter molecule LAT (linker protein of activated T cells) and its associate, SLP-76, are among its most important substrates (Figure 10-4).

FIGURE 10-4 Schematic of T-cell receptor signaling. The tyrosine kinases Lck and ZAP-70 initiate TCR signaling and phosphorylate the CD3 tail as well as adapter molecules LAT and SLP-76. Phosphorylation generates docking sites for the assembly and organization of signaling molecules, which lead to the activation of transcription factors. Signaling routes initiated by PLCγ and Ras are highlighted here (see text), but many others are triggered, too.

TCR or CD3 complex and CD4 or CD8 receptor are on the T-cell membrane. In the T-cell cytoplasm, the CD4 or CD8 cytoplasmic tail binds with Lck. Lck along with ZAP-70 act as a phosphorylate on the ITAMs found in the domains of the CD3 molecules. ZAP-70 also phosphorylates SLP76 and PLC gamma associated with LAT. SLP76 interacts with PLC gamma which in turn acts on PIP subscript 2 and cause it to form IP subscript 3 and DAG. IP subscript 3 interacts with the endoplasmic reticulum and induces the release of Ca superscript 2 plus. Ca superscript 2 plus binds to calmodulin and activates calcineurin which in turn activates NFAT.

The DAG created by PIP subscript 2 binds PKC theta. PKC theta phosphorylates and activates NF-kappa beta. LAT also associates with Grb-2 which binds SOS and Ras. DAG binds RasGRP which acts on Ras and initiates the MAP kinase cascade. Phosphorylation in the MAP kinase pathway produces the AP-1 dimer. The AP-1 dimer, the NF-kappa B, and NFAT move into the nucleus and promote activation of genes that regulate survival, proliferation, and effector function.

By phosphorylating LAT and SLP-76, ZAP-70 generates yet another set of new docking sites, permitting the assembly of new complexes of proteins. Phosphorylated adapter molecules provide a critical physical framework for the initiation of the next set of downstream signaling events. The network of signals initiated by the adapter molecule assemblies are varied and complex. Here, we will describe two of the major signaling cascades that contribute to activation of transcription factors responsible for many of the changes associated with T-cell activation. Many of the molecules in these TCR signaling pathways are common to all cells, and may be familiar.

PLCγ Signaling and Activation of Transcription Factors NF-κB and NFAT

One of the proteins that binds phosphorylated LAT via its SH2 domain is the enzyme phospholipase C gamma (PLCγ), which is now in proximity to its substrates, phospholipids in the plasma membrane. PLCγ splits the “head” of the phospholipid PIP₂ from its lipid tails, generating two new signaling molecules: soluble IP₃ and membrane-bound diacylglycerol (DAG). IP₃ induces the release of calcium from multiple stores. Ca²⁺ is a common second messenger in signaling networks, binding calmodulin and activating the phosphatase calcineurin. Calcineurin dephosphorylates and activates the transcription factor NFAT (nuclear factor of activated T cells). In T cells, DAG created by PIP₂ hydrolysis binds a specialized form of protein kinase C called PKC-θ (theta). This part of the signaling cascade culminates in the degradation of the inhibitors of transcription factor NF-κB and the translocation of the active transcription factor into the nucleus (see Figure 10-4).

Ras-ERK Signaling and Activation of Transcription Factor AP-1

Phosphorylated LAT also associates with the SH2 domain of Grb2, the adapter molecule that recruits components of the Ras pathway to the signaling complex. In T cells, the Ras pathway triggers the sequential activation of MAP kinases or MAPKs. These kinases phosphorylate serine residues and include RAF, MEK, and ERK. ERK has many targets, but is particularly important in the activation of the transcription factor AP-1 (see Figure 10-4).

As we have mentioned, successful activation of naïve T cells requires simultaneous engagement of the TCR and costimulatory molecules like CD28. As long as costimulatory molecules are engaged, just a few TCR-MHC interactions are needed to trigger a signaling cascade that culminates in (1) cell survival, (2) proliferation, and (3) cell differentiation.

Just as the adapter proteins provide a scaffold for the immune system to organize its signaling proteins, we hope that this section provides a similar scaffold for the organization of the reader’s thoughts on the complexities of the first signal required for T-cell activation. We now turn our attention to Signal 2, which coordinates with TCR-generated signals in both time and space.

Key Concepts:

• Three distinct signals are required to induce naïve T-cell activation, proliferation, and differentiation. Signal 1 is generated by the interaction of the TCR-CD3 complex with an MHC-peptide complex on an antigen-presenting cell.
• TCR signaling is initiated by two tyrosine kinases called Lck and ZAP-70, which phosphorylate multiple proteins, generating new docking sites for signaling proteins. Adapter proteins like LAT and SLP-76 use these docking sites to organize new signaling cascades, which activate transcription factors.
• Transcription factors downstream of TCR signaling regulate expression of genes that induce T-cell survival, proliferation, and differentiation into effector cells.

Costimulatory Signals Are Required for Optimal T-Cell Activation Whereas Coinhibitory Signals Prevent T-Cell Activation

What evidence pointed to a requirement for a second signal, the costimulatory signal? In 1987, Helen Quill and Ron Schwartz recognized that high-affinity TCR-MHC interactions in the absence of functional APCs led to T-cell nonresponsiveness rather than activation—a phenomenon they called T-cell anergy. They advanced the simple, but powerful two-signal hypothesis. We now know that Signal 2 is generated by interactions between specific costimulatory receptors on T cells and costimulatory ligands that are expressed only by professional antigen-presenting cells (Table 10-1). When a T cell receives both Signal 1 and Signal 2, it produces cytokines that enhance entry into the cell cycle and proliferation (see Overview Figure 10-2).

TABLE 10-1 T-cell costimulatory and coinhibitory receptors and their ligands

Receptor on T cell	Ligand	Activity
Costimulatory receptors
CD28	CD80 (B7-1) or CD86 (B7-2) Expressed by professional APCs (and medullary thymic epithelium)	Activation of naïve T cells
ICOS	ICOS-L Expressed by B cells, some APCs, and T cells	Maintenance of activity of differentiated T cells; a feature of T-/B-cell interactions
Coinhibitory receptors
CTLA-4	CD80 (B7-1) or CD86 (B7-2) Expressed by professional APCs (and medullary thymic epithelium)	Negative regulation of the immune response (e.g., maintaining peripheral T-cell tolerance; reducing inflammation; contracting T-cell pool after infection is cleared)
PD-1	PD-L1 or PD-L2 Expressed by professional APCs, some T and B cells, and tumor cells	Negative regulation of the immune response, regulation of TREG differentiation
BTLA	HVEM Expressed by some APCs, and T and B cells	Negative regulation of the immune response, regulation of TREG differentiation (?)

Recall from Chapter 4 that antigen-presenting cells, including dendritic cells (DCs), are activated by antigen binding to pattern recognition receptors (PRRs) to express costimulatory ligands (e.g., CD80 and CD86) and produce cytokines that enhance their ability to activate T cells. CD28 is the most commonly cited example of a costimulatory receptor, but other, related molecules that provide costimulatory signals during T-cell activation have since been identified and are also described here.

Coinhibitory receptors, once referred to as negative costimulatory receptors, counter the activity of costimulatory receptors and inhibit T-cell activation. They play important roles in (1) maintaining peripheral T-cell tolerance and (2) reducing inflammation both after the natural course of an infection and during responses to chronic infection. In this section we introduce several coinhibitors that have a major impact on T-cell biology.

Costimulatory Receptors: CD28

CD28, a 44-kDa glycoprotein expressed as a homodimer, was the first costimulatory molecule to be discovered (Classic Experiment Box 10-1). Expressed by all naïve and activated human and murine CD4⁺ T cells, all murine CD8⁺ T cells, and, interestingly, only 50% of human CD8⁺ T cells, it markedly enhances TCR-induced proliferation and survival by cooperating with T-cell receptor signals to induce expression of the pro-proliferative cytokine IL-2 and the prosurvival Bcl-2 family member, Bcl-x_L.

CLASSIC EXPERIMENT BOX 10-1

Discovery of the First Costimulatory Receptor: CD28

In 1989, Navy immunologist Carl June took the first step toward filling (and revealing) the considerable gaps in our understanding of T-cell proliferation and activation by introducing a new actor: CD28. CD28 had been recently identified as a dimeric glycoprotein expressed on all human CD4⁺ T cells and half of human CD8⁺ T cells, and preliminary data suggested that it enhanced T-cell activation. June and his colleagues specifically wondered if CD28 might be related to the Signal 2 that was known to be provided by APCs (sometimes referred to as “accessory cells” in older literature).

June and his colleagues isolated T cells from human blood by density gradient centrifugation and by depleting a T-cell–enriched population of cells that did not express CD28 (negative selection). They then measured the response of these CD28⁺ T cells to TCR stimulation in the presence or absence of CD28 engagement (Figure 1a).

FIGURE 1 Evidence that CD28 is costimulatory ligand for T cell proliferation. (a) June’s experimental setup used anti-CD3 monoclonal antibodies or a mitogen, PMA, to provide Signal 1, and anti-CD28 antibodies to provide Signal 2. (b) June measured [³H]uridine incorporation, an indicator of RNA synthesis, in response to various treatments. Addition of stimulating anti-CD28 antibody increased RNA synthesis in response to activation by PMA or anti-CD3.

The illustration shows a T cell. PMA is bound to the T cell. Anti-CD3 is bound to the CD3 receptor. Anti-CD28 is bound to the CD28 receptor. Signal 1 from the TCR and signal 2 from the CD28 receptor is transmitted to the nucleus, leading to gene expression.

The bar graph shows the [superscript 3 H]-uridine incorporation by the T cells when exposed to various stimuli. Without stimulation or when exposed to PHA, or treated with anti-CD28, the [superscript 3 H]-uridine incorporation was at 0.3 counts per minute times 10 to the negative power 3. When exposed to PMA, the [superscript 3 H]-uridine incorporation was at 0.5 counts per minute times 10 to the negative power 3. When treated with PMA and anti-CD28, the [superscript 3 H]-uridine incorporation was at 4.9 counts per minute times 10 to the negative power 3. When treated with anti-CD3, the [superscript 3 H]-uridine incorporation was at 3.7 counts per minute times 10 to the negative power 3. When treated with anti-CD3 and anti-CD28, the [superscript 3 H]-uridine incorporation was at 7 counts per minute times 10 to the negative power 3.

To mimic the TCR-MHC interaction, June and colleagues used either monoclonal antibodies to the CD3 complex or the mitogen phorbol myristate acetate (PMA), a protein kinase C (PKC) activator. To engage the CD28 molecule, they used an anti-CD28 monoclonal antibody. They included two negative controls: one population that was cultured in growth medium with no additives and another population that was exposed to phytohemagglutinin (PHA), which was known to activate T cells only in the presence of APCs. This last control was clever and would be used to demonstrate that the researchers’ isolated populations were not contaminated with APCs, which would express their own sets of ligands and confound interpretation.

June and colleagues measured the proliferation of each of these populations by measuring incorporation of a radioactive (tritiated) nucleotide, [³H]uridine, which is incorporated by cells that are synthesizing new RNA (and, hence, are showing signs of activation). Their results were striking, particularly when responses to PMA were examined (Figure 1b). As expected, T cells grown without stimulation or with incomplete stimulation (PHA) remained quiescent, exhibiting no nucleotide uptake. Cell groups treated with stimuli that were known to cause T-cell proliferation—anti-CD3 and PMA—showed evidence of activation, with CD3 engagement producing relatively more of a response than PMA at the time points examined.

The cells treated with anti-CD28 only were just as quiescent as the negative control samples, indicating that engagement of CD28, alone, could not induce activation. However, when CD28 was engaged at the same time cells were exposed to PMA, incorporation of [³H]uridine increased markedly. T cells cotreated with anti-CD28 and anti-CD3 also took up more [³H]uridine than those treated with anti-CD3 alone.

What new RNA were these cells producing? Using Northern blot analysis and functional assays to characterize the cytokines in culture supernatant of the stimulated cells, June and colleagues went on to show that CD28 stimulation induced anti-CD3–stimulated T cells to produce higher levels of cytokines involved in antiviral, antitumor, and proliferative activity, generating an increase in T-cell immune response.

When this article was published, June did not know the identity of the natural ligand for CD28, or even if the homodimer could be activated in a natural immune context. We now know that CD28 binds to CD80/86 (B7), providing the critical Signal 2 during naïve T-cell activation, a signal required for optimal up-regulation of IL-2 and the IL-2 receptor. Finding this second switch capable of modulating T-cell activation was only the beginning of a landslide of discoveries of additional costimulatory signals—positive and negative—involved in T-cell activation, and the recognition that T cells are even more subtly perceptive than we once appreciated.

Based on a contribution by undergraduate Harper Hubbeling, Haverford College, 2011.

REFERENCE

• Thompson, C. B., et al. 1989. CD28 activation pathway regulates the production of multiple T-cell-derived lymphokines/cytokines. Proceedings of the National Academy of Sciences USA 86:1333.

CD28 binds to two distinct ligands of the B7 family of proteins: CD80 (B7-1) and CD86 (B7-2). These are members of the immunoglobulin superfamily, which have similar extracellular domains. Interestingly, their intracellular regions differ, suggesting that they might not simply act as passive ligands; rather, they may have the ability to generate signals that influence the APC, a view that has some experimental support.

Although most T cells express CD28, most cells in the body do not express its ligands. In fact, only professional APCs have the capacity to express CD80/86. Mature dendritic cells, the best activator of naïve T cells, appear to constitutively express CD80/86, and macrophages and B cells have the capacity to up-regulate CD80/86 after they are activated by an encounter with pathogen (see Chapter 4).

How does CD28 signaling add to the effects of TCR signaling described previously? It appears to make both quantitative and qualitative contributions—enhancing the strength of signal, but also assembling a unique group of signaling molecules. PI3 kinase (PI3K) is one of the most important molecules recruited by CD28. Once brought in proximity to the membrane, it activates other downstream kinases that, in turn, regulate many aspects of cell metabolism, survival and division. These pathways work in concert with those generated by the TCR to prepare a T cell for its role as an effector cell.

Costimulatory Receptors: ICOS

Since the discovery of CD28, several other structurally related receptors have been identified. Like CD28, the closely related inducible costimulator (ICOS) provides positive costimulation for T-cell activation. However, rather than binding CD80 and CD86, ICOS binds to another member of the growing B7 family, ICOS-ligand (ICOS-L), which is also expressed on a subset of activated APCs.

Differences in expression patterns of CD28 and ICOS indicate that these positive costimulatory molecules play distinct roles in T-cell activation. Unlike CD28, ICOS is not expressed on naïve T cells; rather, it is expressed on memory and effector T cells. Investigations suggest that CD28 plays a key costimulatory role during the initiation of activation and ICOS plays a key role in maintaining the activity of already differentiated effector and memory T cells.

Coinhibitory Receptors: CTLA-4 and PD-1

The discovery of CTLA-4 (CD152), the second member of the CD28 family to be identified, caused a stir. Although closely related in structure to CD28 and also capable of binding both CD80 and CD86, CTLA-4 did not act as a positive costimulator. Instead, it antagonized T-cell–activating signals and was the first coinhibitory receptor to be identified.

CTLA-4 is not expressed constitutively on resting T cells. Rather, it is induced within 24 hours after activation of a naïve T cell and peaks in expression within 2 to 3 days post-stimulation. Peak surface levels of CTLA-4 are lower than peak CD28 levels, but because it binds CD80 and CD86 with markedly higher affinity, CTLA-4 competes very favorably with CD28.

Interestingly, CTLA-4 expression levels increase in proportion to the amount of CD28 costimulation, suggesting that CTLA-4 acts to “put the brakes on” the pro-proliferative influence of TCR-CD28 engagement. The importance of this inhibitory function is underscored by the phenotype of CTLA-4 knockout mice, whose T cells proliferate without control, leading to lymphadenopathy (greatly enlarged lymph nodes), splenomegaly (enlarged spleen), autoimmunity, and death within 3 to 4 weeks after birth.

Programmed cell death-1 (PD-1 or CD279) is a coinhibitory receptor expressed by both B and T cells. It binds to two ligands, PD-L1 (B7-H1) and PD-L2 (B7-DC), which are also members of the CD80/86 family. PD-L2 is expressed predominantly on APCs; however, PD-L1 is expressed more broadly and may help to mediate T-cell tolerance in nonlymphoid tissues. Recent data suggest that interactions between PD-1 and PD-L1/2 may also regulate the differentiation of regulatory T cells.

The discovery of CTLA-4 and PD-1 inspired the development of checkpoint inhibitors, a very promising new immunotherapy for cancer (Clinical Focus Box 10-2). Molecules that block the interaction between these coinhibitory receptors and their ligands have been shown to enhance latent T-cell activity against tumors and are now being used in the clinic.

CLINICAL FOCUS BOX 10-2

Checkpoint Inhibitors: Breakthrough in Cancer Therapy

One of the most exciting and promising breakthroughs in our centuries-long effort to combat cancer has been the development of checkpoint inhibitors, a clinical advance that took full advantage of the remarkable fruits of basic research.

Because tumors are variations of our own cells, the immune response to them is typically not as robust as it is to infection by pathogens. As you have seen in Chapters 8 and 9, our immune system has evolved multiple mechanisms to enforce tolerance to self. All of these, including central tolerance, regulatory T cells, and immune ignorance, impede the response to tumors.

Promisingly, over the last few decades some investigators have successfully identified tumor-specific antigens (TSAs) that theoretically could provoke an immune response. Several experiments in mice revealed the potential for the immune system to react to tumors. However, it was difficult to translate these observations into clinically useful results; most tumors remained stubbornly resistant to immune cell activity and investigators began to wonder whether our immune system could ever be a useful partner in efforts to treat cancer. And frankly, many wondered whether a cure for cancer, which includes a very heterogeneous set of diseases, was even a realistic goal.

The discovery of coinhibitor molecules in the 1990s not only inspired breakthroughs in T-cell research but also inspired a new excitement, even among the most skeptical, about the possibility of treating cancer. Dr. Jim Allison, one of the discoverers of CTLA-4, had witnessed the ability of the immune system to attack tumors in some of his earlier experiments. He wondered if tumors may be using T-cell coinhibitors to their advantage. If tumor cells or cells associated with tumors expressed ligands for T-cell coinhibitory molecules, they might be inactivating existing tumor-specific T cells. If you removed this “brake,” you might be able to unleash T-cell activity against the tumor.

Allison and his colleagues first tested the possibility in mice. They injected antibodies that blocked CTLA-4 interactions into mice with tumors. The results, reported first in Science in 1996, were dramatic. Although it took time to work, the antibody enhanced rejection of both developing and established tumors. Tumors shrunk and, in some cases, even disappeared, whether they expressed CTLA-4 ligands or not.

The remarkable success of CTLA-4 blockade in the treatment of many different tumors in mice led Allison and colleagues to shift their attention to human patients. They persuaded one brave company, Medarex, to take a chance on the approach and developed a humanized antibody against CTLA-4 (Figure 1). This antibody, now called ipilimumab (or ippy), was entered into a clinical trial and at first did not look promising. But patience was rewarded when it became clear in subsequent trials that, just as in mice, it took time for the antibody to re-activate T cells.

FIGURE 1 How the checkpoint inhibitor ipilimumab works. Naïve T cells are activated when they engage both TCR and CD28 costimulatory receptors on antigen-presenting cells. Activated T cells up-regulate the coinhibitory molecule CTLA-4, which binds CD28 ligands even more strongly. This interaction inactivates the T cell. However, when the CTLA-4 engagement is blocked with the antibody ipilimumab, the T cell is reactivated.

The illustration shows a T cell and an APC. The TCR of the T cell binds to the MHC of the APC. The CD28 coreceptor of the T cell binds to the CD80 or CD86 coreceptor of the APC. In this scenario, cosimulation via CD28 leads to T-cell activation. When CTLA-4 of the T cell binds to the CD80 or CD86 coreceptor of the APC, the CTLA-4 blocks costimulation, inhibiting T-cell activation. When ipilimumab (Y-shaped) blocks CTLA-4 of the T cell, the CD28 coreceptor of the T cell binds to the CD80 or CD86 coreceptor of the APC, leading to T-cell activation.

The antibody was first tested in patients with malignant melanoma, a fatal cancer of pigment-producing skin cells. Although not all responded to the treatment, those who did experienced long-lasting and often dramatic remission. It became the first therapy for the deadly cancer that showed a survival advantage. The U.S. Food and Drug Administration approved the antibody for the treatment of melanoma in 2011. It now joins an arsenal of other antibodies that block the engagement of coinhibitory molecules, including antibodies to PD-1, one of the other potent coinhibitors. Together these therapeutic antibodies are referred to as checkpoint inhibitors, to describe their ability to inhibit the brakes (or “checkpoints”) on the immune system. They represent the best of bench-to-bedside (translational) research, as well as the triumph of imagination, innovation, and sheer doggedness by individuals in the research community. In 2015, Jim Allison was the recipient of one of the most prestigious scientific honors, the Lasker Award.

The effectiveness of checkpoint inhibition continues to impress the medical community. However, not every patient and not every tumor responds and investigators are working furiously to understand why. The identification of other coinhibitory molecules expressed by T cells, including Lag-3, Tim-3, and TIGIT, may offer new hope and new targets for immunotherapy.

REFERENCE

• Leach, D. R., M. F. Krummel, and J. P. Allison. 1996. Enhancement of antitumor immunity by CTLA-4 blockade. Science 271:1734.

A Coinhibitory Receptor Expressed by Many Immune Cell Types: BTLA

B and T lymphocyte attenuator (BTLA or CD272) is a coinhibitory receptor that is more broadly expressed: not only has it been found on conventional T_H cells as well as γδ T cells and regulatory T cells, but it is also expressed on NK cells, some macrophages and dendritic cells, and most highly on B cells. Interestingly, BTLA’s primary ligand appears not to be a B7 family member, but a TNF receptor family member known as herpes virus entry mediator (HVEM), which is also expressed on many cell types. Studies on the role of this interesting coinhibitory receptor-ligand pair are ongoing, but there are indications that BTLA-HVEM interactions also play a role in down-regulating inflammatory and autoimmune responses.

As you can imagine, the expression and activity of costimulatory and coinhibitory molecules must be carefully regulated, both temporally and spatially. Naïve T cells, for example, do not express coinhibitory receptors, allowing them to be activated in secondary lymphoid tissue during the initiation of an immune response. On the other hand, effector T cells up-regulate coinhibitory receptors at the end of an immune response, when proliferation is no longer advantageous. However, these generalizations belie the complexity of regulation of this highly important costimulatory network, and investigators are still working to understand the details.

As the genome continues to be explored, additional costimulatory and coinhibitory molecules—both positive and negative in influence—have been identified. Understanding their regulation and function will continue to occupy the attention of the immunological community and has already provided the clinical community with new tools for manipulating the immune response during transplantation and disease (see Clinical Focus Box 10-2).

Key Concepts:

• Signal 2 is produced by costimulatory ligands expressed by APCs interacting with costimulatory receptors (CD28 or ICOS) expressed by T cells.
• Coinhibitory receptors, such as PD-1 and CTLA-4, are expressed by activated T cells and bind to ligands on antigen-presenting cells. When coengaged with the TCR, they send signals that inhibit T-cell activation.

Clonal Anergy Results If a Costimulatory Signal Is Absent

Experiments with cultured cells show that if a naïve T cell’s TCR is engaged (Signal 1) in the absence of a suitable costimulatory signal (Signal 2), that specific T cell clone becomes unresponsive to subsequent stimulation, a state referred to as clonal anergy (Figure 10-5). There is good evidence that both CD4⁺ and CD8⁺ T cells can be anergized, but most studies of anergy have been conducted with CD4⁺ T_H cells.

FIGURE 10-5 Signals that lead to clonal anergy versus clonal expansion. (a) Resting T cells that engage antigen and costimulatory ligands CD80/86 are activated and expand. (b) T cells that engage antigen on the surface of a non-antigen-presenting cell (such as a pancreatic islet beta cell) will not receive costimulatory signals and, instead, are inactivated or anergized. (c) Anergy can also be induced when a T cell engages antigen and receives inhibitory rather than costimulatory signals through coinhibitory receptors such as CTLA-4 and PD-1.

The illustration shows a T cell and an APC. The TCR of the T cell binds to the MHC of the APC. The CD28 coreceptor of the T cell binds to the CD80 or CD86 coreceptor of the APC. Signals from the TCR and the CD28 receptors are transmitted to the nucleus of the T cell. This results in an activated T cell that releases cytokines. When the TCR of T cell binds to the MHC of a non-APC such as a pancreatic islet beta cell, the coreceptor CD28 coreceptor of the T cell does not bind to any receptors on the non-APC cell. The nucleus of the T cell receives signals only from the TCR. This results in an anergic T cell. The T cell also becomes anergic when the TCR of the T cell binds to the MHC of the APC, the CTLA-4 receptor binds to the CD80 or CD86 coreceptor of APC, the PD-1 receptor of T cell binds to the PD-L1 or PD-L2 receptor of the APC, and the CD28 coreceptor of T cell does not bind to any receptor on the APC. In this case, the inhibitory signals from CTLA-4 and PD-1 are stronger than the signal from the TCR and the T cell is inactivated.

Anergy can be demonstrated experimentally with systems designed to engage the TCR in the absence of costimulatory molecule engagement. For instance, T cells specific for an MHC-peptide complex can be induced to proliferate in vitro by incubation with activated APCs that express both the appropriate MHC-peptide combination and CD80/86. However, exposing T cells to fixed APCs that cannot up-regulate CD80/86, or physically blocking the CD28-CD80/86 interaction with antibodies, renders T cells unresponsive. Anergic T cells are no longer able to secrete cytokines or proliferate in response to subsequent stimulation.

The requirement for costimulatory ligands to activate a T cell decreases the probability that autoreactive T cells that have escaped the thymus will be activated and become dangerous. For instance, a naïve T cell expressing a T-cell receptor specific for an MHC class I–insulin peptide complex would be rendered nonresponsive if it encountered a pancreatic islet beta cell expressing this MHC class I–peptide complex. Why? Islet beta cells cannot be induced to express costimulatory ligands, and the encounter would result in T-cell anergy, preventing an immune attack by these T cells (see Figure 10-5b).

Interactions between coinhibitory receptors and ligands can also induce anergy (see Figure 10-5c). This phenomenon, which applies only to activated T cells that have up-regulated coinhibitory receptors, could help curb T-cell proliferation when antigen is cleared. Coinhibition may also contribute to the T-cell “exhaustion” during chronic infection, such as that caused by mycobacteria, HIV, hepatitis virus, and even exposure to tumor antigens. T cells specific for these pathogens and antigens express high levels of PD-1 and BTLA, and become functionally anergic. Recent therapies designed to block coinhibitory interactions have successfully reactivated T cells and are the basis for checkpoint inhibitor drugs that are considered one of the most exciting recent breakthroughs in cancer therapy (see Clinical Focus Box 10-2).

The biochemical pathways that lead to anergy are still under investigation. Chronic TCR stimulation may lead to intracellular changes in signaling molecules, and regulatory T cells may also contribute. Although the in vivo importance of anergy has been contested, recent studies clearly show that anergic CD4⁺ T cells help maintain maternal tolerance to a fetus during pregnancy. Interestingly, anergic CD4⁺ cells may also differentiate into regulatory T cells, which play a key role in maintaining immune tolerance in all vertebrates (see Chapters 8 and 16).

Key Concept:

• In the absence of costimulatory signals (Signal 2) or in the presence of coinhibitory signals, T-cell receptor engagement results in T-cell inactivity or clonal anergy.

Cytokines Provide Signal 3

We have now seen that naïve T cells are activated when they simultaneously engage both MHC-peptide complexes and costimulatory ligands on antigen-presenting cells, particularly dendritic cells. However, the outcome of T-cell activation is critically shaped by the activity of soluble cytokines produced by both APCs and T cells. These assisting cytokines are referred to, by some, as Signal 3 (see Overview Figure 10-2).

Cytokines bind surface cytokine receptors, stimulating a cascade of intracellular signals that enhance both proliferation and/or survival. IL-2 is one of the best-known cytokines involved in T-cell activation and plays a key role in inducing optimal T-cell proliferation, particularly when antigen and/or costimulatory ligands are limiting. TCR and costimulatory signals induce transcription of genes for both IL-2 and the α chain (CD25) of the high-affinity IL-2 receptor. These signals together also enhance the stability of IL-2 mRNA. The combined increase in IL-2 transcription and improved IL-2 mRNA stability results in a 100-fold increase in IL-2 production by the activated T cell. Secretion of IL-2 and its subsequent binding to the high-affinity IL-2 receptor induces activated naïve T cells to proliferate vigorously.

As we will see shortly, Signal 3 also includes another important set of cytokines, known as polarizing cytokines. These are produced by a variety of cell types, including APCs, T cells, and innate lymphoid cells (ILCs), and play central roles in determining what types of effector cells naïve T cells will become.

Key Concept:

• Signal 3 is provided by soluble cytokines and includes IL-2, which enhances T-cell proliferation, as well as polarizing cytokines, which play a key role in determining the type of effector cell that a T cell becomes.

Antigen-Presenting Cells Provide Costimulatory Ligands and Cytokines to Naïve T Cells

Which cells are capable of providing both Signal 1 and Signal 2 to a naïve T cell? Although almost all cells in the body express MHC class I, only professional APCs—dendritic cells, activated macrophages, and activated B cells—express the high levels of MHC class II molecules that are required for naïve CD4⁺ T-cell activation (Figure 10-6). Professional APCs are also capable of expressing costimulatory ligands. (Only one other cell type, the medullary thymic epithelial cell, is known to share this capacity; see Chapter 8.) In order to initiate a T-cell response, all professional APCs must first be activated by microbial components via their pattern recognition receptors (PRRs). This encounter enhances antigen presentation activity, and up-regulates expression of MHC and costimulatory ligands.

FIGURE 10-6 Comparison of professional antigen-presenting cells that induce T-cell activation. This figure describes general features of three major classes of professional APCs. Dendritic cells are the best activators of naïve T cells. This may be due, in part, to their relatively high levels of expression of MHC and costimulatory molecules when they are mature and activated. Activated B cells interact most efficiently with differentiated T_FH cells that are specific for the same antigen that activated them. Macrophages play several different roles, processing and distributing antigen in secondary lymphoid tissues as well as interacting with effector cells in the periphery. It is important to recognize that the distinctions shown are rules of thumb only. Functions among the APC classes overlap, and investigators now recognize many different subsets within each major group of APCs, each of which may act independently on different T-cell subsets. This diversity may be a consequence of activation by different innate immune receptors or may reflect the existence of independent cell lineages. Note that activation of effector and memory T cells is not as dependent on costimulatory interactions.

The illustration shows the resting and activated states of a dendritic cell, a macrophage, and a B lymphocyte. The dendritic cell has an irregular shape with several finger-like protrusions and an oval nucleus. In its resting state, the dendritic cell has an MHC class I molecule on its surface. PAMPs and cytokines activate the dendritic cell. The activated cell expresses higher levels of MHC class I molecules, MHC class II molecules, and CD80 or CD86 molecules. The macrophage is circular and has a kidney-shaped nucleus. In its resting state, the macrophage has an MHC class I molecule on its surface. PAMPs and T-cell help (IFN-gamma) activate the macrophage. The activated macrophage expresses higher levels of MHC class I molecules, MHC class II molecules, and CD80 or CD86 molecules. The B lymphocyte is circular and has a large, circular nucleus. In its resting state, the B lymphocyte has an MHC class I molecule, an MHC class II molecule, and several BCRs on its surface. Antigens activate the B lymphocyte. The activated B lymphocyte expresses higher levels of MHC class I molecules, MHC class II molecules, and CD80 or CD86 molecules.

The key features of dendritic cells and macrophages are: Phagocytic; Express receptors for apoptotic cells, DAMPs, and PAMPs; Localize to tissues; Localize to T-cell zone of lymph nodes following activation (DCs); Constitutively express high levels of MHC class II molecules and antigen-processing machinery; Express costimulatory molecules following activation; Variety of subtypes, some migrating, some resident, some better at cross-presenting; Specific subtypes may specialize in activating different T cells. The key features of B cells are: Internalize antigens via BCRs; Constitutively express MHC class II molecules and antigen-processing machinery; Express costimulatory molecules following activation.

Professional APCs are more diverse in function and origin than originally imagined, and each subpopulation differs both in the ability to display antigen and in the expression of costimulatory ligands (see Figure 10-6). DCs appear to be the most potent activators of naïve T cells and come in many different varieties. Conventional dendritic cells, which arise from bone marrow stem cells, include migratory cells that are activated by antigen in peripheral tissues and travel to regional (draining) lymph nodes to expose themselves to naïve T cells. Conventional dendritic cells also include cells that do not circulate as vigorously. These resident DCs process antigen that comes into the lymph node from the blood. They can also acquire antigen from other antigen-presenting cells. Langerhans cells are another type of migratory DC found exclusively in the skin and discovered in the nineteenth century by an observant young medical student, Paul Langerhans. They originally arise from embryonic hematopoietic stem cells and replenish themselves not from bone marrow stem cells, but simply by dividing. Once activated by interactions with antigen, Langerhans cells migrate to both local and distal lymph nodes to meet and activate T cells (see Chapter 13).

Resting B cells residing in follicles also gain the capacity to activate T cells, although at the late stages of an immune response. Once they bind antigen through their B-cell receptor (BCR), they up-regulate MHC class II and CD80/86, and present antigen to activated CD4⁺ T cells they encounter at the border between the follicle and T-cell zone (see Chapter 14). Because of their unique ability to internalize pathogen via specific BCRs and present them in MHC class II, B cells are best at activating CD4⁺ T cells that recognize epitopes on the same pathogen. This situation serves the immune response very well, focusing the attention of antigen-specific CD4⁺ T cells activated in the T-cell zone on B cells activated by the same antigen in the neighboring follicle. The pairing of B cells with their helper T cells occurs at the junction between the B- and T-cell zones and allows T cells to deliver the help required for B-cell proliferation, differentiation, and memory generation (see Chapter 11).

Several other antigen-presenting cells play a role in the primary immune response, including macrophages that line the sinuses of lymph node and spleen. These filter antigen and transfer it to other antigen-presenting cells in the T-cell zone. Some of the major antigen-presenting cell types are shown in Table 7-4 (and see Chapter 13). This list is not complete and is likely to be joined by even more subtypes as research continues.

Key Concepts:

• Several different professional antigen-presenting cells can provide Signals 1, 2, and 3 to a naïve T cell, although dendritic cell subsets appear particularly potent.
• Professional APCs differ by location and migratory behavior. Some circulate actively and others reside for long periods of time in specific tissues and organs.
• B cells use their capacity as antigen-presenting cells to solicit help from CD4⁺ T cells during an immune response in secondary lymphoid tissue.

Superantigens Are a Special Class of T-Cell Activators

Superantigens are viral or bacterial proteins that bind simultaneously to specific V_β regions of T-cell receptors and to the _α chain of MHC class II molecules. V_β regions are encoded by over 20 different V_β genes in mice and 65 different genes in humans. Each superantigen displays a “specificity” for one of these V_β versions, which can be expressed by up to 5% of T cells, regardless of their antigen specificity. This clamp-like connection mimics a strong TCR-MHC interaction and induces activation, bypassing the need for TCR antigen specificity (Figure 10-7). Superantigen binding, however, does not bypass the need for costimulation; professional APCs are still required for full T-cell activation by these microbial proteins.

FIGURE 10-7 Superantigen-mediated cross-linkage of T-cell receptor and MHC class II molecules. A superantigen binds to all TCRs bearing a particular V_β sequence regardless of their antigen specificity. Exogenous superantigens are soluble secreted bacterial proteins, including various exotoxins. Endogenous superantigens are membrane-embedded proteins produced by certain viruses; they include Mls antigens encoded by mouse mammary tumor virus.

The illustration shows a T subscript H cell and an APC. The TCR of the T subscript H cell binds to the MHC class II molecule of the APC. The MHC presents a peptide for which the TCR is not specific. An endogenous superantigen, bound to the membrane of the APC, binds the V subscript beta region of the TCR to the alpha chain of the MHC.

Both endogenous superantigens and exogenous superantigens have been identified. Exogenous superantigens are soluble proteins secreted by bacteria. Among them are a variety of exotoxins secreted by gram-positive bacteria, such as staphylococcal enterotoxins, toxic shock syndrome toxin, and exfoliative dermatitis toxin. Each of these exogenous superantigens binds particular V_β sequences in T-cell receptors (Table 10-2) and cross-links the TCR to an MHC class II molecule.

Endogenous superantigens are cell-membrane proteins generated by specific viral genes that have integrated into mammalian genomes. One group, encoded by mouse mammary tumor virus (MMTV), a retrovirus that is integrated into the DNA of certain inbred mouse strains, produces proteins called minor lymphocyte-stimulating (Mls) determinants, which bind particular V_β sequences in T-cell receptors and cross-link the TCR to MHC class II molecules. Four Mls superantigens, originating from distinct MMTV strains, have been identified.

TABLE 10-2 Exogenous superantigens and their V_β specificity

		Vβ SPECIFICITY
Superantigen	Disease*	Mouse	Human
Staphylococcal enterotoxins
SEA	Food poisoning	1, 3, 10, 11, 12, 17	ND
SEB	Food poisoning	3, 8.1, 8.2, 8.3	3, 12, 14, 15, 17, 20
SEC1	Food poisoning	7, 8.2, 8.3, 11	12
SEC2	Food poisoning	8.2, 10	12, 13, 14, 15, 17, 20
SEC3	Food poisoning	7, 8.2	5, 12
SED	Food poisoning	3, 7, 8.3, 11, 17	5, 12
SEE	Food poisoning	11, 15, 17	5.1, 6.1–6.3, 8, 18
Toxic shock syndrome toxin (TSST1)	Toxic shock syndrome	15, 16	2
Exfoliative dermatitis toxin (ExFT)	Scalded skin syndrome	10, 11, 15	2
Mycoplasma arthritidis supernatant (MAS)	Arthritis, shock	6, 8.1–8.3	ND
Streptococcal pyrogenic exotoxins (SPE-A, B, C, D)	Rheumatic fever, shock	ND	ND
*Disease results from infection with bacteria that produce the indicated superantigens. ND = not determined.

Because superantigens bind outside the TCR antigen-binding cleft, any T cell expressing that particular V_β sequence will be activated by a corresponding superantigen. Hence, the activation is polyclonal and can result in massive T-cell activation, resulting in overproduction of T_H-cell cytokines and systemic toxicity. Food poisoning induced by staphylococcal enterotoxins and toxic shock induced by toxic shock syndrome toxin are two examples of disorders caused by superantigen-induced cytokine overproduction.

Given that superantigens directly activate the host T-cell response, it is difficult to imagine what value they have for the pathogens that make them. There is some evidence that such antigen-nonspecific T-cell activation and inflammation hampers the development of a coordinated antigen-specific response. Some speculate that the large-scale proliferation and cytokine production that results from superantigen exposure harms the cells and microenvironments that are required to start a normal response; others argue that these events induce T-cell tolerance to the pathogen.

Key Concepts:

• Superantigens are products of bacteria and viruses that activate T cells in an antigen-nonspecific manner.
• Superantigens bind directly to select TCR β chains. This interaction activates all T cells expressing these chains, regardless of the TCR α chain they express. This large-scale T-cell activation leads to inflammation that can cause disease, such as toxic shock syndrome.