Partial activation of naive CD4 T cells and tolerance induction in response to peptide presented by resting B cells

Tolerance is thought to occur when Ag is presented to T cells in the absence of costimulatory interactions from APC accessory molecules. Of the professional APC, the resting B cell may be the main tolerizing cell in vivo. We have analyzed several aspects of activation of naive transgenic CD4 cells stimulated with resting or activated B cells presenting peptide Ag. Similar results were obtained with stimulation from peptide presenting fibroblast APC lacking or expressing B7-1 with intracellular adhesion molecule-1. TCR ligation with little or no accessory molecule coreceptor engagement induced efficient blastogenesis; up-regulation of CD25, CD44, CD69, CD95 and CD71; and down-regulation of CD62L over a 48-h period. Accessory molecule help enhanced the expression of CD25, CD44, CD69, and CD71, but to very modest degrees. Only two molecules, CD40 ligand and IL-2, were found to be extremely dependent on accessory molecule help, with little or no expression evident with peptide presented on resting B cells or class II-positive fibroblasts. T cells induced on resting B cells expanded minimally over 3 days, and this was followed by extensive cell death and hyporesponsiveness of the resulting cells. These studies suggest that under tolerizing conditions, such as Ag presentation by resting B cells, much of the naive CD4 response is induced efficiently. Partial activation, however, may be the overall result due to the lack of CD40 ligand expression, which may regulate costimulatory activity in APC and, in turn, may contribute to limiting the production of IL-2 required for T cell expansion and survival.

CD40 ligand induction on T cell subsets by peptide-presenting B cells: implications for development of the primary T and B cell response

Recent data suggest that CD40 ligand (CD40L)-CD40 interactions are essential for up-regulation of costimulatory activity on APC and that efficient induction of CD40L may be pivotal to the success of a CD4 T cell response. CD40L is regulated primarily by TCR signaling, but high level expression on a naive T cell appears to require additional interactions between T cell coreceptors and APC accessory molecules. The data reported here show that resting B cells presenting peptide Ag, in contrast to both dendritic cells and preactivated B cells, induce very little CD40L on naive CD4 cells, which in turn is insufficient to promote APC costimulatory activity. We also show, however, that previously activated effector T cells have enhanced responsiveness to Ag when accessory molecule help is limiting and consequently can express high levels of CD40L after interaction with resting B cells. High level CD40L expression correlated with B cell activation and up-regulation of costimulatory activity; however, blocking studies showed that CD40L was only partially responsible for these phenomena. These studies reinforce the notion that resting B cells may be tolerogenic for naive CD4 cells in part because of inefficient CD40L induction. The data also suggest that a successful primary T cell response will only occur if either the initial interaction is with a dendritic cell followed by subsequent interactions of the effector T cells with resting APC or if nonspecific inflammatory stimuli up-regulate accessory molecule expression on resting APC before an encounter with the naive T cell.

CD40 ligand induction on T cell subsets by peptide-presenting B cells: implications for development of the primary T and B cell response

Recent data suggest that CD40 ligand (CD40L)-CD40 interactions are essential for up-regulation of costimulatory activity on APC and that efficient induction of CD40L may be pivotal to the success of a CD4 T cell response. CD40L is regulated primarily by TCR signaling, but high level expression on a naive T cell appears to require additional interactions between T cell coreceptors and APC accessory molecules. The data reported here show that resting B cells presenting peptide Ag, in contrast to both dendritic cells and preactivated B cells, induce very little CD40L on naive CD4 cells, which in turn is insufficient to promote APC costimulatory activity. We also show, however, that previously activated effector T cells have enhanced responsiveness to Ag when accessory molecule help is limiting and consequently can express high levels of CD40L after interaction with resting B cells. High level CD40L expression correlated with B cell activation and up-regulation of costimulatory activity; however, blocking studies showed that CD40L was only partially responsible for these phenomena. These studies reinforce the notion that resting B cells may be tolerogenic for naive CD4 cells in part because of inefficient CD40L induction. The data also suggest that a successful primary T cell response will only occur if either the initial interaction is with a dendritic cell followed by subsequent interactions of the effector T cells with resting APC or if nonspecific inflammatory stimuli up-regulate accessory molecule expression on resting APC before an encounter with the naive T cell.

Accessory molecule and costimulation requirements for CD4 T cell response

T cell activation is brought about by recognition of peptide/MHC complexes on an antigen-presenting cell (APC) by the T cell receptor (TCR). However, in general this appears to be insufficient for the full development of T cell responses and therefore additional signals are required, provided by ligation of counter-receptors on the T cell by APC accessory molecules. Although many studies have suggested that B7 molecules (CD80/CD86) binding to CD28 induce this second signal, it is now evident that any one of a number of molecules may provide accessory function and that efficient response is only generated following multiple interactions. It has also become clear that T cells exist in varying states of activation or differentiation, and that requirements for accessory molecules and costimuli are not always equivalent. This review covers much of the recent data regarding accessory molecule regulation of T cell responses. A modified version of the two signal model is presented, suggesting that the major function of accessory molecules during the initial stages of activation is to augment the ability to signal through the TCR, and that the primary role of costimulatory signals is to allow IL-2 secretion and growth. The requirement for multiple accessory molecules interactions is discussed in relation to activation of naive T cells and how such interactions are less critical at the memory and effector stages. Finally, this new information is related to how T cells interact with varying APC and how these interactions may modulate T cell response.

Naive and effector CD4 T cells differ in their requirements for T cell receptor versus costimulatory signals

We used naive CD4 cells and in vitro-derived Th1 and Th2 effectors from TCR transgenic mice to investigate the requirements of these subsets for TCR signaling and interactions with accessory molecules. Peptide Ag and immobilized anti-CD3 were used to provide different TCR signals. Anti-CD28 Ab or a panel of class II+ fibroblasts, expressing no accessory molecules or expressing intracellular adhesion molecule-1, B7-1, or both molecules, were used as APC or accessory cells (AC). An efficient naive T cell response required a strong TCR signal (high dose anti-CD3 or peptide) and high levels of multiple synergizing costimulatory signals, while effector cells responded efficiently to anti-CD3 alone. Addition of AC only slightly augmented the effector response. Effectors responded to lower doses of peptide than naive cells. However, when peptide-pulsed APC were used to stimulate effectors, requirements varied with the cytokine measured. The production of IL-4 did not require accessory molecules on APC. IL-2 production required interacting APC to express accessory molecules, but was little augmented by AC not presenting Ag, suggesting a requirement for noncostimulatory interactions. Proliferation of effectors closely paralleled IL-2 production. Production of IFN-gamma was intermediate in dependence on accessory molecules, and production of IL-5 was nearly as dependent as IL-2. These results establish major differences between the induction of naive and effector responses and document differential requirements for the induction of distinct cytokines, indicating that different cytokines may be produced depending on the context of effector restimulation.

Naive and effector CD4 T cells differ in their requirements for T cell receptor versus costimulatory signals

We used naive CD4 cells and in vitro-derived Th1 and Th2 effectors from TCR transgenic mice to investigate the requirements of these subsets for TCR signaling and interactions with accessory molecules. Peptide Ag and immobilized anti-CD3 were used to provide different TCR signals. Anti-CD28 Ab or a panel of class II+ fibroblasts, expressing no accessory molecules or expressing intracellular adhesion molecule-1, B7-1, or both molecules, were used as APC or accessory cells (AC). An efficient naive T cell response required a strong TCR signal (high dose anti-CD3 or peptide) and high levels of multiple synergizing costimulatory signals, while effector cells responded efficiently to anti-CD3 alone. Addition of AC only slightly augmented the effector response. Effectors responded to lower doses of peptide than naive cells. However, when peptide-pulsed APC were used to stimulate effectors, requirements varied with the cytokine measured. The production of IL-4 did not require accessory molecules on APC. IL-2 production required interacting APC to express accessory molecules, but was little augmented by AC not presenting Ag, suggesting a requirement for noncostimulatory interactions. Proliferation of effectors closely paralleled IL-2 production. Production of IFN-gamma was intermediate in dependence on accessory molecules, and production of IL-5 was nearly as dependent as IL-2. These results establish major differences between the induction of naive and effector responses and document differential requirements for the induction of distinct cytokines, indicating that different cytokines may be produced depending on the context of effector restimulation.

A direct role for IFN-gamma in regulation of Th1 cell development

IL-12 has been identified as a major cytokine influencing the differentiation of CD4 cells to a Th1 phenotype, whereas a role for IFN-gamma is controversial. We investigated the interrelationship between IL-12 and IFN-gamma in promoting Th1 responses using naive CD4 cells reactive with pigeon cytochrome c from TCR transgenics and memory CD4 cells derived by in vivo priming with KLH. Without exogenous rIL-12 or rIFN-gamma, primary and memory effectors induced by Ag or anti-CD3 and anti-CD28 secreted variable levels of IL-2 and IFN-gamma. The level of IFN-gamma secreted by effectors correlated with endogenous IFN-gamma produced in primary cultures, and anti-IFN-gamma largely inhibited the development of effectors producing IFN-gamma. With optimal TCR stimulation and costimulation, endogenous IFN-gamma, without IL-12, was sufficient to elicit Th1 cells via an autocrine mechanism, whereas with suboptimal stimulation, exogenous rIFN-gamma or rIL-12 was required for Th1 development. However, rIL-12 was more effective than rIFN-gamma, partially because rIL-12 greatly enhanced autocrine production of IFN-gamma, and optimal development of the Th1 phenotype was mediated by the synergistic actions of both cytokines. Thus, both IFN-gamma and IL-12 can independently regulate Th1 development, but because of IFN-gamma-mediated feedback, their relative contributions are determined by the conditions of T cell stimulation. The extent of differentiation to a Th1 phenotype may, therefore, depend on the availability of both APC-derived IL-12 and autocrine IFN-gamma consequent to the overall strength of T cell stimulation.

A direct role for IFN-gamma in regulation of Th1 cell development

IL-12 has been identified as a major cytokine influencing the differentiation of CD4 cells to a Th1 phenotype, whereas a role for IFN-gamma is controversial. We investigated the interrelationship between IL-12 and IFN-gamma in promoting Th1 responses using naive CD4 cells reactive with pigeon cytochrome c from TCR transgenics and memory CD4 cells derived by in vivo priming with KLH. Without exogenous rIL-12 or rIFN-gamma, primary and memory effectors induced by Ag or anti-CD3 and anti-CD28 secreted variable levels of IL-2 and IFN-gamma. The level of IFN-gamma secreted by effectors correlated with endogenous IFN-gamma produced in primary cultures, and anti-IFN-gamma largely inhibited the development of effectors producing IFN-gamma. With optimal TCR stimulation and costimulation, endogenous IFN-gamma, without IL-12, was sufficient to elicit Th1 cells via an autocrine mechanism, whereas with suboptimal stimulation, exogenous rIFN-gamma or rIL-12 was required for Th1 development. However, rIL-12 was more effective than rIFN-gamma, partially because rIL-12 greatly enhanced autocrine production of IFN-gamma, and optimal development of the Th1 phenotype was mediated by the synergistic actions of both cytokines. Thus, both IFN-gamma and IL-12 can independently regulate Th1 development, but because of IFN-gamma-mediated feedback, their relative contributions are determined by the conditions of T cell stimulation. The extent of differentiation to a Th1 phenotype may, therefore, depend on the availability of both APC-derived IL-12 and autocrine IFN-gamma consequent to the overall strength of T cell stimulation.

Accessory molecule regulation of naive CD4 T cell activation

Naive CD4 T cell activation is a complex process involving many steps. T cell receptor (TCR) signals, provided by interaction with peptide/MHC on antigen-presenting cells (APC), control many events associated with activation. The extent of TCR signaling and the magnitude of the T cell response is in turn controlled by accessory molecules on APC, which stabilize T-APC interactions. Full T cell activation additionally requires multiple costimulatory signals, generated upon ligation of T cell coreceptors by accessory molecules, and these lead to IL-2 production, proliferation and differentiation of the naive cell into an effector state. This review summarizes the role played by accessory molecules in naive CD4 activation and discusses how integration of signals from these molecules, with signals from the TCR, may determine the outcome of T-APC interaction. The available data provide explanations for why only APC which express high levels of multiple costimulatory/adhesion molecules, such as dendritic cells and activated B cells, induce efficient naive T cell responses, and suggest that ICAM-1/LFA-1 and B7/CD28 interactions are major pathways used to initiate naive T cell activation.

Regulation of CD40 ligand expression on naive CD4 T cells: a role for TCR but not co-stimulatory signals

We have investigated the roles of TCR and accessory co-stimulatory signals in the induction of CD40 ligand (CD40L) on CD4 cells. Using naive T cells form TCR transgenic mice, specific for a peptide of pigeon cytochrome c, we show that in contrast to IL-2 secretion, CD40L expression is regulated primarily by signalling through the TCR, is enhanced by accessory molecule interactions, but co-stimulatory signals play little if any role. CD40L was induced at high levels on naive T cells, peaking at 5 h, by class II MHC+ fibroblast antigen-presenting cells (APC) which expressed either ICAM-1, B7-1 or both molecules, whereas only low levels were induced by fibroblasts which did not express any accessory molecules. Differences in intensity and duration of expression were seen following stimulation with ICAM- and B7-expressing APC, with the presence of ICAM resulting in greater and longer expression, although both molecules together were most efficient. The involvement of co-stimulatory signals delivered from accessory molecules was investigated in systems where there was no effect on TCR signalling from adhesive interactions. Anti-CD3, or antigen-pulsed APC lacking accessory molecules, were used to provide the TCR signal, with co-stimulus from either anti-CD28 or accessory molecule-expressing fibroblasts not presenting antigen. Anti-CD3 in the absence of co-stimuli induced high CD40L expression but no IL-2 production and provision of co-stimulatory signals, although inducing large quantities of IL-2, did not increase CD40L expression. In addition, low CD40L expression induced by antigen presented in the absence of accessory molecules was not enhanced by co-stimulation, although IL-2 was strongly up-regulated. These studies suggest that efficient expression of CD40L on naive CD4 cells does require accessory molecules on APC. However, the role of these molecules for CD40L induction, as opposed to IL-2 secretion, is not one of co-stimulation but one of adhesion, presumably allowing stronger or more prolonged signals to be generated through the TCR. The synergistic role of ICAM and B7 during naive CD4 activation was confirmed using dendritic cells as APC, with nearly complete inhibition of CD40L expression as well as IL-2 secretion being seen when both CTLA-4-Ig and anti-LFA-1 were used to block these molecules.

Costimulatory requirements of naive CD4+ T cells. ICAM-1 or B7-1 can costimulate naive CD4 T cell activation but both are required for optimum response

Efficient initiation of a CD4 T cell response requires both activation through the TCR and costimulation provided by molecules on APC with counterreceptors on the T cell. We investigated the relative contribution of the ICAM-1:LFA-1 and B7:CD28/CTLA-4 costimulatory pathways in naive T cell activation, using either anti-CD28 Ab or fibroblast cell lines transfected with I-Ek, which express either no costimulatory molecules, ICAM-1 alone, B7-1 alone, or ICAM-1 and B7-1 together. Peptide Ag or immobilized anti-CD3 was used to provide the TCR signal. CD4 T cells from mice transgenic for the V beta 3/V alpha 11 TCR, which recognize a peptide of pigeon cytochrome c complexed to I-Ek, were used as a source of naive T cells. Naive T cells stimulated with Ag or anti-CD3 responded well to high numbers of APC expressing either ICAM-1 alone or B7-1 alone. However, APC expressing both ICAM-1 and B7-1 were much better stimulators of proliferation and IL-2 secretion at low cell numbers, and were far superior inducers of IL-2 at higher numbers, indicating a synergy between the two pathways. Stimulation provided by ICAM-1 could not be solely attributed to adhesive strengthening of other pathways, since costimulation was seen when immobilized anti-CD3 was used and when ICAM-1 only APC were added, indicating that ICAM-1 was in fact acting as a classic costimulatory molecule. Both the magnitude of the response and the amount of costimulation required for response were dependent on the intensity of TCR interaction. These results suggest that an efficient naive T cell response requires both a strong TCR signal and more than one costimulatory signal that will synergize with the TCR signal. This offers an explanation as to why APC such as dendritic cells and activated B cells, which express high levels of multiple costimulatory/adhesion molecules, are the only APC that elicit naive T cell responses.

Costimulatory requirements of naive CD4+ T cells. ICAM-1 or B7-1 can costimulate naive CD4 T cell activation but both are required for optimum response

Efficient initiation of a CD4 T cell response requires both activation through the TCR and costimulation provided by molecules on APC with counterreceptors on the T cell. We investigated the relative contribution of the ICAM-1:LFA-1 and B7:CD28/CTLA-4 costimulatory pathways in naive T cell activation, using either anti-CD28 Ab or fibroblast cell lines transfected with I-Ek, which express either no costimulatory molecules, ICAM-1 alone, B7-1 alone, or ICAM-1 and B7-1 together. Peptide Ag or immobilized anti-CD3 was used to provide the TCR signal. CD4 T cells from mice transgenic for the V beta 3/V alpha 11 TCR, which recognize a peptide of pigeon cytochrome c complexed to I-Ek, were used as a source of naive T cells. Naive T cells stimulated with Ag or anti-CD3 responded well to high numbers of APC expressing either ICAM-1 alone or B7-1 alone. However, APC expressing both ICAM-1 and B7-1 were much better stimulators of proliferation and IL-2 secretion at low cell numbers, and were far superior inducers of IL-2 at higher numbers, indicating a synergy between the two pathways. Stimulation provided by ICAM-1 could not be solely attributed to adhesive strengthening of other pathways, since costimulation was seen when immobilized anti-CD3 was used and when ICAM-1 only APC were added, indicating that ICAM-1 was in fact acting as a classic costimulatory molecule. Both the magnitude of the response and the amount of costimulation required for response were dependent on the intensity of TCR interaction. These results suggest that an efficient naive T cell response requires both a strong TCR signal and more than one costimulatory signal that will synergize with the TCR signal. This offers an explanation as to why APC such as dendritic cells and activated B cells, which express high levels of multiple costimulatory/adhesion molecules, are the only APC that elicit naive T cell responses.

Recently activated naive CD4 T cells can help resting B cells, and can produce sufficient autocrine IL-4 to drive differentiation to secretion of T helper 2-type cytokines

Development of T cells during primary responses was investigated using pigeon cytochrome C-specific naive Th from TCR transgenic mice. Naive CD4 cells did not activate and help resting B cells. This failure was found to be primarily because the resting B cells were incapable of stimulating the naive Th. Provision of a costimulatory signal such as anti-CD28, or addition of APCs that express costimulatory molecules, such as dendritic cells, activated B cells, and B7+ and B7+ICAM(+)-expressing fibroblasts, induced naive Th activation and promoted T cell-dependent help for IgM secretion. T cell activation for as little as 24 h promoted helper activity, and Ig secretion required production of small amounts of IL-4 by the activated naive Th. On initial stimulation, naive Th secrete only IL-2. By mRNA analysis, activated naive Th were also shown to produce IL-4, however induction of IL-4 message only occurred 24 h after initial activation and required additional stimulation with Ag. A single exposure of naive CD4 to Ag/APC followed by 4 to 12 days in culture led to generation of effector Th which secreted IL-2 and some IFN-gamma, and no detectable IL-4 or IL-5, and which could only help B cells to IgM secretion. In contrast, similar cultures that received Ag/APC one or more times during this period generated effector cells capable of secreting easily detectable titers of IL-4 and IL-5, as well as IL-2 and IFN-gamma, and able to now promote IgG1 and IgE responses. Generation of these Th0-like effectors was accompanied by increasing amounts of IL-4 secreted during the culture period after each restimulation, and addition of anti-IL-4 in culture inhibited development of the capacity to produce Th2 cytokines. These studies reinforce the notion that naive CD4 must interact with a costimulatory professional APC, rather than a resting B cell, for initiation of the primary response, but show that such an interaction can result in rapid development of the ability to interact with and provide cognate help to B cells. They also suggest that if activated naive CD4 cells receive multiple stimulations from Ag/APC, enough endogenous IL-4 can be produced to drive differentiation into effectors secreting type 2 cytokines. The existence of such an autocrine feedback mechanism suggests that the amount and availability of Ag could influence the nature and polarization of the Th response.

Recently activated naive CD4 T cells can help resting B cells, and can produce sufficient autocrine IL-4 to drive differentiation to secretion of T helper 2-type cytokines

Development of T cells during primary responses was investigated using pigeon cytochrome C-specific naive Th from TCR transgenic mice. Naive CD4 cells did not activate and help resting B cells. This failure was found to be primarily because the resting B cells were incapable of stimulating the naive Th. Provision of a costimulatory signal such as anti-CD28, or addition of APCs that express costimulatory molecules, such as dendritic cells, activated B cells, and B7+ and B7+ICAM(+)-expressing fibroblasts, induced naive Th activation and promoted T cell-dependent help for IgM secretion. T cell activation for as little as 24 h promoted helper activity, and Ig secretion required production of small amounts of IL-4 by the activated naive Th. On initial stimulation, naive Th secrete only IL-2. By mRNA analysis, activated naive Th were also shown to produce IL-4, however induction of IL-4 message only occurred 24 h after initial activation and required additional stimulation with Ag. A single exposure of naive CD4 to Ag/APC followed by 4 to 12 days in culture led to generation of effector Th which secreted IL-2 and some IFN-gamma, and no detectable IL-4 or IL-5, and which could only help B cells to IgM secretion. In contrast, similar cultures that received Ag/APC one or more times during this period generated effector cells capable of secreting easily detectable titers of IL-4 and IL-5, as well as IL-2 and IFN-gamma, and able to now promote IgG1 and IgE responses. Generation of these Th0-like effectors was accompanied by increasing amounts of IL-4 secreted during the culture period after each restimulation, and addition of anti-IL-4 in culture inhibited development of the capacity to produce Th2 cytokines. These studies reinforce the notion that naive CD4 must interact with a costimulatory professional APC, rather than a resting B cell, for initiation of the primary response, but show that such an interaction can result in rapid development of the ability to interact with and provide cognate help to B cells. They also suggest that if activated naive CD4 cells receive multiple stimulations from Ag/APC, enough endogenous IL-4 can be produced to drive differentiation into effectors secreting type 2 cytokines. The existence of such an autocrine feedback mechanism suggests that the amount and availability of Ag could influence the nature and polarization of the Th response.

Generation of polarized antigen-specific CD8 effector populations: reciprocal action of interleukin (IL)-4 and IL-12 in promoting type 2 versus type 1 cytokine profiles

We have generated primary effector populations from naive CD8 T cells in response to antigen and determined their patterns of cytokine secretion upon restimulation. The effect of exogenous factors on the effector generation was examined and compared with responses of antigen-specific CD4 effectors generated under comparable conditions. CD8 cells from bm1 mice were stimulated with C57BL/6 (B6) antigen presenting cells (APCs) bearing allogeneic class I and CD8 cells from female severe combined immunodeficiency (SCID) B6 mice, transgenic for a T cell receptor alpha/beta (TCR-alpha/beta) that recognizes H-Y on Db, were stimulated with APCs from male mice. In parallel, CD4 cells from bm12 mice were stimulated with alloantigen and CD4 cells from V beta 3/V alpha 11 TCR transgenics were stimulated with a peptide of pigeon cytochrome c on IEk. T cells from both transgenic mice were of naive phenotype whereas normal mice contained 10-20% memory cells. Effector CD8 populations generated were L-selectin low, CD45RB high, and CD44 high. Naive CD8 cells from SCID anti-H-Y mice made little or no cytokine immediately upon stimulation in contrast to naive CD4 which produced large amounts of interleukin 2 (IL-2). Both populations, however, generated primary effectors over 4-5 d that made substantial quantities of many cytokines upon restimulation. Both CD8 and CD4 effectors produced similar patterns of cytokines with alloantigen or specific antigen. Cytokines present during naive CD8 stimulation influenced the cytokine secretion profile of the effectors, as previously shown for CD4 cells, although secretion by CD8 effectors was generally lower than that of CD4 effectors. CD8 cells cultured with IL-2 alone made predominantly interferon gamma (IFN-gamma) and no IL-4 or IL-5, similar to CD4 cells. Priming with IFN-gamma increased IFN-gamma secretion from CD4 effectors, but had little if any effect on CD8 cells. In contrast, priming with IL-12 generated CD8 effectors, as well as CD4 effectors, producing elevated quantities of IFN-gamma, with similar levels from both the CD4 and CD8 populations. The presence of IL-4 during effector cell generation promoted synthesis of IL-4 and IL-5 from both CD8 and CD4 cells while downregulating IFN-gamma secretion. CD8 cells made only small amounts of IL-4, more than 100-fold less than CD4 cells, whereas significant levels of IL-5 were induced, only 3-10-fold lower than from CD4.(ABSTRACT TRUNCATED AT 400 WORDS)

Activation of naive, memory and effector T cells

An increased understanding of the types of T-cell subsets that exist in vivo, their relationships to one another, and how to identify and isolate them or effect their generation, has led to a comprehensive view of the antigen-presenting cells (APCs) which may be active and regulatory during the course of an immune response. Recent studies show that naive T cells only respond efficiently to dendritic cells and activated B cells whereas memory and effector cells respond to all APC types to some extent, including resting B cells. High level co-stimulatory molecule expression largely explains why APCs such as dendritic cells are far more effective stimulators than resting B cells. The available data, therefore, suggest that the requirement for co-stimulation, and hence capacity to respond to various APCs, is largely a function of the differentiation state of the T cell, and that previous encounter with antigen fundamentally increases the ability of T cells to subsequently respond to antigen rechallenge.