Leo Lefranc¸ois
Development, trafficking, and function of memory T-cell subsets
Authors’ address
Summary: The precise mechanisms that govern memory T-cell lineage
Leo Lefranc¸ois
commitment during an immune response continue to be the subject of intense scrutiny. The existence of memory T-cell subsets defined by location, function, and phenotype adds an additional layer of complexity to the overall memory T-cell population. In this review, the integration of memory subset development and migration and the functional consequences of specific tissue localization are discussed.
University of Connecticut Health Center, Department of Immunology, Farmington, CT, USA. Correspondence to: Leo Lefranc¸ois University of Connecticut Health Center Department of Immunology 263 Farmington Ave. Farmington, CT 06107 USA Tel.: +1 860 6793242 Fax: +1 860 6791868 E-mail:
[email protected] Acknowledgements I thank Lynn Puddington for constant support, insightful discussions, and critical reading of the manuscript and the members of my laboratory past and present for their hard work and dedication toward deciphering the mysteries of T-cell memory. This work was funded by grants from the National Institutes of Health.
Immunological Reviews 2006 Vol. 211: 93–103 Printed in Singapore. All rights reserved
ß 2006 The Author Journal compilation ß 2006 Blackwell Munksgaard
Immunological Reviews 0105-2896
Introduction ‘‘To be, or not to be: that is the question; Whether ‘tis nobler in the mind to suffer the slings and arrows of outrageous fortune, Or to take arms against a sea of troubles, and by opposing end them? To die: to sleep; No more; and, by a sleep to say we end the heartache and the thousand natural shocks that flesh is heir to, ‘tis a consummation devoutly to be wished’’. William Shakespeare, Hamlet (1600–01). As Hamlet contemplates ‘self-murder’ at the beginning of Act III, he begins his soliloquy with ‘To be or not to be, that is the question’. That is also the question that faces each T-cell entering into an immune response. To be, to suffer the slings and arrows of outrageous fortune and survive to see another day and become . . . a memory T-cell! and by surviving, experience the heartache and thousand natural shocks, or, to continue the metaphor, respond to secondary or tertiary antigen challenge or beyond and survive in the face of frequent perturbations of the system. The alternative to survival is to take arms against the sea of troubles and thus end them, by committing suicide. What are the factors governing which path is chosen leading to survival or death? When does the decision occur, and can those precursors destined to become memory T cells be identified? And, once that path leading to memory is chosen, how is the population maintained? Furthermore, to achieve the ultimate goal of protection against invasion, how is the memory T-cell pool integrated
93
Lefranc¸ois Central and effector memory T cells
into the larger immune system and the mammalian body as a whole, given the complexity and the interrelationships of the various organ systems? The answers to these questions are determined by a number of factors including the relationship between proliferation, survival, and migration of memory T cells, and the consequences of the encounter of memory cells with antigen. Recent studies have shed some light on these and other aspects of the memory T-cell subsystem. This review discusses our contributions to the field and their place in the larger landscape of memory T-lymphocyte biology.
Life versus death: an early decision Before discussing the details of the mechanism of memory Tcell formation, some definitions are in order. In this review, the discussion of memory T cells focuses on responses to acute infections or immunizations. When one studies chronic infections or immunizations with adjuvants such as Freund’s (1), antigen may be expressed or retained for protracted periods of time, and this retention will greatly influence the generation of memory as well as the functional and phenotypic characteristics of the memory cells (2–5). The question of how memory cells are defined vis a vis naive and effector T cells, is invariably asked. There are three criteria that can be used to identify bona fide memory cells. (i) The frequency of antigenspecific cells late after immunization is greater than the frequency of naive cells with that specificity prior to immunization. This observation is routinely made with robust infections, where antigen-specific naive cells are undetectable before infection and memory cells make up an easily detectable subset months or years later. Similarly, the immune systems of vaccinated people harbor increased frequencies of antigen-specific T cells for many years (6–8). Thus, in typical healthy individuals, the total number of antigen-specific cells (naı¨ve + memory) is always greater than the naı¨ve number alone. However, thymic atrophy during advancing age or conditions of immunodeficiency can result in a reduced number of naive T cells specific for any given antigen. (ii) T cells with a memory phenotype are small resting cells. Although most phenotypic markers of memory cells can be shared with effector cells, memory cells will generally not be undergoing blastogenesis. Therefore, memory cells will be smaller in size than effector cells, although they may be slightly larger than naive T cells (9, 10). (iii) There is a more rapid response to challenge late after primary infection/immunization compared with the response in naı¨ve animals. The quicker response time of antigen-specific memory cells is due to
94
Immunological Reviews 211/2006
their increased frequency as well as heightened signaling capacity, another hallmark of bona fide memory T cells (11–14).
When? True memory cells (as defined at the molecular level by a stable genetic profile and at the cellular level by the ability to optimally respond to antigenic challenge) appear not to develop until a few weeks after the contraction phase of the T-cell response to acute infection (15). However, the immunization regimen can shorten or lengthen the time required for memory T-cell development (16). The factors controlling the kinetics of progression to memory are likely linked to the rapidity with which naive T cells encounter antigen. Thus, if sufficient antigen and the correct costimulators are immediately available (e.g. peptide vaccination with adjuvant), the response kinetics may be hastened, resulting in more rapid memory T-cell formation (16). How quickly antigen is removed from the system may also affect the timing of memory development (17). Nevertheless, precisely when a T-cell receives the stamp of approval to proceed into the memory lineage is unclear. The determination of whether memory T cells will be formed appears to occur early in the immune response, perhaps even within the first few hours after a naı¨ve T cell encounters an antigen-presenting cell (APC) bearing antigen (18, 19). Early events also regulate the magnitude of the overall expansion of the responding cells, which is in turn linked to memory T-cell generation (20). That is, in general, when the initial growth of the responding population is limited, smaller numbers of memory cells are produced (21). Although this correlation is not absolute, it is clear that the quality of the initial interaction of the T cell with an antigen bearing APC can regulate both the magnitude of the response and the size of the resultant memory T-cell population (20, 22). For example, immunization with heatkilled (HK) Listeria monocytogenes (LM) results in rapid but abortive expansion of CD8+ T cells and inefficient memory T-cell generation, while infection with live LM induces more sustained T-cell blastogenesis, robust expansion, and efficient production of memory CD8+ T cells (20). These results are probably due, at least in part, to reduced Toll-like receptormediated triggering of the innate immune system by HKLM versus live LM (23). Thus, with HKLM immunization, ineffective dendritic cell (DC) activation occurs (24, 25), resulting in transient APC – T-cell conjugate formation and poor T-cell activation. There are several possibilities regarding the timing of ‘memory commitment’. First, if the initial encounter of a
Lefranc¸ois Central and effector memory T cells
naı¨ve T-cell with an APC occurs within the first 48 h after infection and if the interaction of the naive T cell with an antigen-bearing APC is stable for up to 24 h or more (2628), then it might be supposed that during this time frame the initial decision is made to shunt a subset of cells into the memory lineage. Second, it can be proposed that while an optimal initial encounter is necessary for memory development, it is not by itself sufficient. That is, downstream events, such as cytokine availability (29–33), will determine the fate of the effector population and the subsequent development of memory T cells. For example, even though high numbers of adoptively transferred T-cell receptor (TCR) transgenic CD4+ T cells vigorously respond to a virus infection in which robust expansion occurs, memory T-cell development is not assured (Blair and Lefranc¸ois, unpublished data), perhaps as a result of competition for APC and/or survival signals. In another scenario, originally proposed by Sprent and Tough (34), naı¨ve T cells entering the response late (presumably >48 h after immunization) would preferentially generate memory T cells. This theory is based on the thinking that very strong stimulation, as encountered early in the primary response, leads to effector cell production but eventual deletion through clonal exhaustion. As antigen concentration wanes, T-cell activation becomes more favorable for memory T-cell production. Precisely when ‘late’ in the response occurs is unclear but would no doubt be dependent on the type of immunization or infection and the availability of antigen. Regardless of the time at which a naı¨ve T-cell enters the response, the data suggest that a sustained interaction with an appropriately activated APC (or perhaps with multiple APCs) is essential to memory T-cell development.
How? Although the quality of the initial interaction of a naı¨ve T-cell with APC determines the efficacy of memory development (35), the question remains as to how the lucky survivors are singled out from the majority of effectors that will die during the contraction phase. In this context, one can speculate whether survival is the result of the absence of a deathinducing signal or the presence of survival signals, although these are not mutually exclusive possibilities. The fact that antiapoptotic molecules, such as Bcl-2, are downregulated in effector cells (29, 36) but are more highly expressed in memory cells (29, 36, 37) suggests that active signaling is needed to promote survival of a subset of T cells. Regulation of the antiapoptotic molecule Bcl-3 (38, 39) and the proapoptotic molecule bim (40, 41) also appears to be involved in the transition from effector to memory T-cell. In terms of
selection of individual cells into the memory pool, the process could be stochastic, whereby 5–10% of responding cells escape deletion and enter the memory lineage. In this scenario, all responding cells are given the correct signals early on, which empower them to become memory cells, but perhaps due to competition for resources, only a subset survives. Alternatively, memory T-cell selection could be deterministic such that only a subpopulation of cells initially receives the correct signals which allow that subset to survive to become memory cells.
Who? The Holy Grail of memory found? Finding a marker able to identify memory T-cell precursors early during an immune response has been a longstanding quest in the study of memory T-cell biology. Not surprisingly, it was reasoned that molecules involved in promoting T-lymphocyte survival would be potential candidates. Furthermore, memory cells would express such a molecule while the vast majority of effector cells would not. Our initial approach to identify possible memoryspecific genes was to perform subtractive hybridization in which mRNA sequences from effector CD8+ T cells are subtracted from memory T-cell mRNA sequences. In this way, the mRNA encoding interleukin-7 receptor (IL-7R) was shown to be more highly expressed in memory versus effector CD8+ T cells (Schluns and Lefranc¸ois, unpublished results). This finding led to the demonstration that IL-7R, which is expressed on the surface of naive T cells and is required for their survival (42, 43), is downregulated upon T-cell activation but gradually increases following the contraction phase, resulting in high-level IL-7R expression by memory CD8+ T cells (29). Moreover, expression of IL-7R by naı¨ve TCR transgenic CD8+ T cells is essential to their survival as well as for efficient memory T-cell development after activation in vivo. IL-7R expression also correlates with Bcl-2 levels, further suggesting that active receptor signaling determines memory commitment (27). Following on from these findings, Kaech et al. (44) demonstrated that a subset of effector cells responding to lymphocytic choriomeningitis virus (LCMV) infection retains IL-7R expression and that this population preferentially survives to become memory cells. Also, in a system using LM infection in which the contraction phase is largely absent, CD8+ T-cell survival correlates with IL-7R expression (45). However, while expression of IL-7R may be necessary to generate memory, IL-7R alone is insufficient to guarantee the production of memory T cells. Thus, activation of CD8+ T cells with antigen delivered in Immunological Reviews 211/2006
95
Lefranc¸ois Central and effector memory T cells
complete Freund’s adjuvant or via injection of peptidecoated DCs does not induce substantial downregulation of IL-7R, although memory cells develop (46). In other words, the number of cells expressing high levels of IL-7R does not correlate with the size of the resulting memory population, indicating that in these cases IL-7R expression does not identify a subset of memory cell precursors, although the results do not indicate whether IL-7R is necessary for memory development. In the case of CD4+ T cells, IL-7R in conjunction with TCR-mediated signals is critical for memory cell homeostasis (47). It should be noted that many of these studies were performed using adoptive transfer of relatively large numbers of TCR transgenic T cells, which may affect the expression of molecules regulated by activation (see below). In addition, although IL-7 promotes survival of memory CD4+ and CD8+ T cells (29, 44, 48–50), it remains to be seen whether IL-7 ‘selects’ the subset of responders destined for the memory pool or whether IL-7R expression is an epiphenomenon linked to the expression of other survivalpromoting factors.
The choice of venue: designation of memory T-cell sublineages The memory T-cell pool is comprised of phenotypically and functionally distinct subpopulations. For example, CD4+ memory T cells geared toward producing T-helper 1 or Th2-type cytokines can be produced depending on the nature of the initial immunization (51–53). In addition, both CD4+ and CD8+ memory T cells can be subdivided into at least two broad lineages based on phenotype, function, and location. These designations, originally derived from analysis of memory T cells in human blood (54), describe central memory T cells (Tcm) and effector memory T cells (Tem). Tcm preferentially migrate to lymph nodes (LNs) by virtue of expression of CD62L and CCR7, two homing molecules required for T cells to cross high endothelial venules (HEVs) and enter the LN from the bloodstream (55). Tcm are also present in blood and spleen but are relatively deficient in non-lymphoid tissues, although the bone marrow has recently emerged as an important site of Tcm localization and proliferation (56–59). Functionally, human CD8+ and CD4+ Tcm generally are poor producers of effector cytokines, such as interferon-g (54, 60), but produce IL-2, although mouse but not human CD8+ Tcm produce effector cytokines following in vitro peptide stimulation (10, 54). Both human and mouse CD8+ Tcm express low levels of lytic activity and cytotoxic effector molecules such as granzyme B and perforin (10, 54). Conversely,
96
Immunological Reviews 211/2006
constitutively lytic Tem are highly enriched in non-lymphoid tissues including the lung, liver, and the intestinal mucosal effector sites, the lamina propria (LP) and the epithelium (10). It should be noted that expression of molecules other than CD62L and CCR7 can further subdivide memory T-cell populations. For example, the heterogeneity of the CD8+ memory cell population is further exemplified by CD27 expression, which appears to be linked to lytic subsets in non-lymphoid tissues (61). The origin of the memory cell lineages based on CD62L expression has been a matter of debate (62–64). One of the important questions is whether each lineage represents a phenotypically stable subset or whether the subpopulations can interconvert. Naı¨ve T cells invariably express CCR7 (65–67) and CD62L (68, 69), both of which are downregulated as a result of T-cell activation (66, 67, 70). Recent data indicate that the initial frequency of naive CD8+ T cells responding to virus infection has dramatic effects on the memory cells subsequently produced (64). When high numbers of TCR transgenic CD8+ T cells are transferred to mice and activated by LCMV or vesicular stomatitis virus (VSV) infection, some CD62L– cells isolated several weeks after infection are able to re-express CD62L following transfer into naive hosts (63, 64). In contrast, when CD62L– memory cells are the progeny of low numbers of naive CD8+ T cells, as in a normal primary response, antigen-independent conversion from CD62L– to CD62L+ cells does not occur (64). Moreover, initial precursor frequency determines the proliferative potential of the derived CD62L– memory CD8+ T cells. When generated from a high precursor frequency response, a much higher proportion of CD62L– memory cells proliferate than when they are derived from a low precursor frequency response. In fact, the CD62L– memory cells produced from a large number of T-cell precursors proliferate to a similar extent as CD62L+ memory cells, irrespective of whether the latter are derived from a high or low precursor frequency response. These findings led us to propose that ‘transitional’ Tem cells are produced when the initial CD8+ T-cell precursor frequency is high, such that the cells initially lack CD62L but are able to proliferate at the level of a Tcm cell and re-express CD62L (64). To extrapolate to immune responses in general, it can be proposed that strong stimulation (as in high antigen : T-cell situations) will result in the development of distinct non-interconvertible CD8+ memory T-cell lineages, at least based on CD62L expression (Fig. 1). Conversely, weak stimulation (as in low antigen : T-cell situations) will drive development of transitional CD62L– memory cells with high proliferative capacity and the ability to re-express CD62L. In this
Lefranc¸ois Central and effector memory T cells
Strong stimulation
CD62L–
CD62L+
CD62L–
99%
99% X
Naïve CD8 T cell
1% CD62L+ Effector cells
Weak stimulation CD62L+
Naïve CD8 T cell
CD62L–
1% CD62L+ Memory cells
CD62L–
70%
70%
30%
30%
CD62L+ Effector cells
Time
CD62L+ Memory cells
Time
Fig. 1. Effect of stimulation strength on the contribution of CD62L+ and CD62L– CD8+ T cells to the memory population. With strong stimulation (top) of naive CD62L+ CD8+ T cells, few CD62L+ effector cells and early memory cells are produced. With increasing time, the contribution of the CD62L+ subset to the overall memory pool enlarges due to a higher proportion of the CD62L+ cells dividing compared with the CD62L– cells. With less stimulation (bottom), a larger fraction of
CD62L+ effector and early memory cells are generated. Transitional CD62L– Tem are also produced that are able to re-express CD62L, which will contribute to an increased proportion of CD62L+ cells. Because a greater number of CD62L+ memory cells are produced, the conversion of the overall memory pool to Tcm occurs more rapidly with a weak than with a strong initial stimulus. In either case, the total memory cell number remains constant.
scenario, fewer CD62L– cells may also be generated at the effector stage, which will result in a greater percentage of CD62L+ memory cells downstream. In vivo proliferation studies show that a higher percentage of CD62L+ versus CD62L– memory CD8+ T cells incorporate bromodeoxyuridine over a several week labeling period (64). This phenomenon may be the result of the integration of several cellular events including the following: increased proliferation of a subset of CD62L+ cells, a higher proportion of proliferating CD62L+ cells, a higher proportion of non-dividing CD62L– cells or slower division of CD62L– cells, or the preferential death of dividing CD62L– cells or non-dividing CD62L+ cells. Regardless of the underlying reasons (63, 64), the net effect in both cases is that the ratio of CD62L+ to CD62L– cells will increase over time. Thus, the initial ratio of CD62L+ : CD62L– cells will determine the rapidity with which the memory cell pool converts to a predominantly
CD62L+ phenotype (Fig. 1). This situation is a hypothetical one, which may not apply to memory T cells in certain nonlymphoid tissues. For example, in the LP of the small intestine, few CD8+CD62L+ memory cells are found long after infection (unpublished results). The complexity of CD62L expression should also be taken into account when studying memory T-cell lineages. One mechanism of CD62L downregulation occurs through proteolytic cleavage, which results in reduced levels of cell-surface expression and release of a soluble form (71–75). Loss of CD62L due to cleavage appears to be an acute event following T-cell activation, whereas transcriptional downregulation occurs several days after activation (74). In any case, in those cells in which cleavage has occurred, CD62L could presumably be re-expressed from newly produced or existing mRNA (71) (Fig. 2). The timing of such re-expression or the relative contribution of this type of regulation during an in Immunological Reviews 211/2006
97
Lefranc¸ois Central and effector memory T cells
CD62L– Strong Stimulation
CD62L– from “gene silenced” cells
CD62L–
CD62L–
99%
CD62L+
CD62L–
99% From CD62L “shed” cells
90%
Short time
10%
Naive CD8 T cell
CD62L+ Memory cells
sing tim
CD62L+ Effector cells
Increa
1%
1%
CD62L+ CD62L+
CD62L+
Recall CD62L–
e
CD62L
CD62L–
–
80%
50% 50%
20%
CD62L+ CD62L+ 1° Memory cells Fig. 2. Primary memory subset distribution determines the subset composition of secondary memory cells. During the primary response and early memory development, CD62L– cells are established as a distinct non-interconvertible memory lineage via silencing of the CD62L gene. Some cells that lost CD62L via proteolytic cleavage will re-express CD62L during the transition to memory cells. As shown in Fig. 1, with increasing time after immunization, a greater proportion of the memory population
vivo T-cell immune response to infection is not known. Alternatively, the gene encoding CD62L could be silenced by epigenetic effects such as DNA methylation patterns (Fig. 2). Thus, CD62L expression on the cell surface could conceivably be either transiently or permanently downregulated. Given these facts concerning CD62L regulation, it is possible that during initiation of a CD8+ T-cell response from a high naive precursor frequency or in the case of weak antigenic stimulation, the CD62L gene may remain active in the responding cells, but the initial downregulation of CD62L expression may be due to proteolytic cleavage. Such events could lead to the development of transitional Tem with the ability to re-express CD62L. In contrast, endogenous or TCR transgenic virus-specific CD8+ T cells at low precursor frequency generate apparently stable CD62L– and CD62L+ memory subpopulations that do not interconvert. In either case, some CD62L+ memory cells may have arisen early in the response from CD62L– activated T cells that re-expressed CD62L following proteolytic cleavage (Fig. 2).
98
Immunological Reviews 211/2006
CD62L+ 2° Memory cells
expresses CD62L, although the total number of memory cells remains constant. The ratio of CD62L+ to CD62L– cells at the time of reactivation influences the ratio of Tcm to Tem produced by the secondary response. Thus, depending on the time after the primary response that a challenge is received, the number of secondary memory cells expressing CD62L could be dramatically different. The timing of subsequent immunizations would also affect the relative development of the memory cell subtypes.
Migration of memory CD8+ T cells: one or many memory T-cell pools? The expression of a particular pattern of homing molecules will clearly dictate the ability of a memory cell to enter a given tissue (76). Effector CD8+ T cells are highly promiscuous with respect to their migratory capacity and are able to enter many, if not all, non-lymphoid tissues (77, 78). However, CD8+ memory T-cell migration appears to be much more restricted and tissue-dependent. Using parabiosis as a method to study memory T-cell migration allows the analysis of trafficking blood-borne cells (79–81). In this system, memory CD8+ T cells raised by systemic infection with VSV or LM migrate freely into the lung and the liver within a few days after parabiosis (81). By 8 days after surgery, the memory cell populations in these organs are completely equilibrated (i.e. same number of memory cells in the donor and host tissue). In contrast, migration into the intestinal LP and the brain is highly restricted and never reaches equilibrium over the time
Lefranc¸ois Central and effector memory T cells
course of the experiments. Migration into the peritoneal cavity is also slow but appears to reach equilibrium at approximately 2 weeks after parabiosis. Adoptive transfer studies further show that migration into the peritoneal cavity is blocked by pertussis toxin treatment of the cells, while migration to the lung and liver is unaffected. Migration to the bone marrow is also partially inhibited by pertussis toxin treatment. These findings indicate that chemokine receptors or other G-protein-coupled receptors are involved in memory CD8+ Tcell entry into some tissues but not others. Because memory cells express heightened levels of many adhesion receptors (CD11a, CD44, a4 and a1 integrins, etc.), perhaps such molecules and others are important for movement into non-lymphoid tissues whether or not chemokines also participate in the process. Egress from non-lymphoid tissues is also an important step in lymphocyte migration. The draining lymphatics of tissues contain substantial numbers of lymphocytes, some of which are memory phenotype (82, 83). Recent data suggest that the entry of effector T cells into the draining LN of the lung (84) or the entry of B cells and naı¨ve or memory phenotype T cells into the popliteal LN from the footpad (85) requires the expression of CCR7. Of course, under most circumstances, effector T cells might not be expected to express CCR7. Thus, downregulation of CCR7 is a mechanism for retention of effector cells in non-lymphoid tissues where they are most needed. These findings are in keeping with the previous demonstration that CCR7 is essential for DCs to enter draining LNs from the lymphatics (86–88). If Tem require CCR7 to enter the draining LNs of non-lymphoid tissue, then it would not be expected that CCR7– Tem would exit tissues. However, this idea is difficult to reconcile with the parabiosis studies in which lung and liver Tem rapidly equilibrate via the blood (81). This result suggests that memory CD8+ T cells are readily exiting these tissues. Although CCR7 expression was not examined in our study, a subset of CD62L– lung CD8+ memory T cells can express CCR7 (89). Thus, further analysis is needed to determine whether antigen-specific nonlymphoid memory cells utilize CCR7 to migrate to the draining LN of non-lymphoid tissues. In those tissues in which entry of memory cells is highly restricted, exit may also be limited. For example, when VSV-specific memory CD8+ T cells enter the intestinal LP and the epithelium of an uninfected mouse, both CD103 and CD69 are upregulated (81, 90). Very few memory cells outside the intestine express these molecules, suggesting that these cells may be retained in the intestinal mucosa, perhaps through the action of these receptors. Whether exit of memory T cells occurs in other ‘privileged’ sites, such as the brain, remains to be determined. It
should also be noted that in infections or in situations in which antigen is retained long-term (e.g. chronic infections), it is likely that memory T-cell migration is affected. For example, herpes simplex virus-specific CD8+ T cells are preferentially retained in the sensory ganglia, the site of infection and virus reactivation (5). In addition, following intranasal infection with influenza virus, antigen-specific memory CD8+ T cells are preferentially retained in the lung-draining mediastinal LNs, which also harbor antigen for several weeks postinfection (91, Zammit et al., Immunity, in press). Taken together, the results suggest the potential existence of multiple memory T-cell pools. One, a blood-borne migrating population with the ability to enter multiple tissues (e.g., lung and liver), an additional subset or subsets with the capacity to enter specific tissues [e.g. a4b7+ CD8+ memory T cells able to enter the intestinal mucosa (81)], and tissue-specific memory populations that are retained in a particular tissue long-term. Given differences in tissue structure and cellular makeup, the question of how memory populations are maintained throughout the body is important (32). IL-15 drives CD8+ memory T-cell homeostatic proliferation and is essential for long-term maintenance of memory T-cell numbers, while IL7 is important for survival of T cells in general, including memory CD4+ and CD8+ T cells (29, 30, 48–50). IL-15 is presented to CD8+ T cells by an unusual mechanism termed transpresentation (92). In this system, the high-affinity IL-15Ra ‘presents’ IL-15 to CD8+ memory T cells and to natural killer cells expressing the IL-2/IL-15Rb and the common g-chain (gc) (93–97). There is no requirement for the memory cells to express IL-15Ra for IL-15, which is not thought to be produced by T cells, to induce their proliferation. The question of how memory T cells acquire maintenance factors during their travels has been recently discussed in detail elsewhere (32).
Memory cell reactivation and lineage commitment: resetting the clock? The function of primary memory cells is to protect against subsequent infection. In addition, the prime-boost strategy has substantial merit for promoting vaccine efficacy (98). The character of the response of the primed host is dependent on the efficacy of the initial priming or infection in generating memory T cells. Although the number of memory cells present at the time of boosting or secondary infection is likely to be important in mounting a robust secondary response and in providing protection, the type and the location of memory T cells present in the host will also influence the outcome of the Immunological Reviews 211/2006
99
Lefranc¸ois Central and effector memory T cells
secondary response. For example, CD62L+ memory CD8+ T cells respond more robustly and provide better protection against LCMV infection than do CD62L– memory cells (63). However, Tem are more effective than Tem in protection against to Leishmania major infection of the skin (99). Similarly, in response to intranasal Sendai virus infection, CD62L– memory cells show a heightened response and more effectively generate secondary memory cells than do CD62L+ memory CD8+ T cells. However, over several months, it appears that CD62L+ memory CD8+ T cells play an increasing role in the secondary response to Sendai virus infection (100), perhaps in keeping with their increasing contribution to the total memory pool as discussed above (Fig. 1). CD62L+ and CD62L– memory CD8+ T cells also differentially respond to secondary infections with LM or VSV. Despite intravenous infection in both cases, the recall response to low-dose LM infection was mainly focused in the spleen, while the response to VSV infection occurred in the LN with little contribution from the spleen (Klonowski and Lefrancois, unpublished results). This dichotomy in the location of response initiation resulted in differential reactivation of CD8+ memory T-cell subsets. With secondary LM infection, both CD62L+ and CD62L– memory cells mounted a robust proliferative response and generated similar numbers of secondary memory cells. However, CD62L+ memory cells responded much more vigorously to VSV infection than did CD62L– memory cells (Marzo and Lefrancois, unpublished results). Thus, the location of priming correlated with the location(s) of the memory cell subtype participating in the recall response. Because both CD62L+ and CD62L– memory cells are present in the spleen, both reacted to LM infection. But, because the vast majority of memory cells present in the LNs are CD62L+, the anti-VSV response was dominated by this subset. These and the above findings demonstrate that both Tcm and Tem were able to participate extensively in secondary responses. The nature, dose, and route of infection of the pathogen dictated the degree of involvement of particular subsets (CD62L+ versus CD62L–) in the response. In this way, more effective protection can be provided, because all fronts will be covered. It should also be noted that only the proliferative component of the recall response is being measured in these studies. While the level of cellular expansion is a clear indication of the initiation and strength of a response, memory cells in the non-lymphoid tissues may mediate functional and protective activity in the absence of a robust proliferative phase. Indeed, protection against secondary infection may not require an increase in antigen-specific cell numbers, providing sufficient numbers of functional memory cells are initially present (101).
100
Immunological Reviews 211/2006
The makeup of the primary memory T-cell pool with respect to the different memory T-cell subclasses will not only influence the secondary effector response but will also affect the composition of the secondary memory pool. As discussed above, the ratio of Tcm to Tem generated in the primary response is dependent on the naı¨ve T-cell precursor frequency, perhaps reflecting the ‘strength of signal’ (35), and on the length of time after the initial immunization. As a consequence of the increased proportion of proliferating CD62L+ memory cells, with increasing time there is continual skewing of the overall CD8+ T-cell memory population toward the CD62L+ subset (Fig. 2). The composition of the primary memory population in turn will affect the downstream production of secondary memory cell subsets following a recall response. Therefore, if boosting occurs a relatively short time after memory is established, fewer CD62L+ cells will be available to respond. As a consequence of the secondary response, CD62L expression will be downregulated, resulting in the appearance of very few CD62L+ memory cells. This law of diminishing returns with respect to Tcm loss would continue as additional immunizations/infections occurred, thereby further diminishing the number of CD62L+ memory cells and driving the reaction toward CD62L– memory CD8+ T-cell production. In terms of vaccination, such an event may or may not be desirable, depending on the nature of the pathogen being combated. Conversely, with increasing time after initial priming, a greater percentage of the memory population will express CD62L (Fig. 2). A secondary infection or immunization will then potentially result in the production of a secondary memory population consisting of a higher percentage of CD62L+ cells than the response generated earlier after establishment of memory. Scenarios such as these may hold important implications for vaccination, particularly with a prime-boost regimen. For example, if a particular vaccination scheme is dependent on a strong response by CD62L+ memory cells (i.e. VSV), then boosting early after priming may not be efficacious, whereas allowing additional time to pass prior to boosting may provide additional potency. Although such designer vaccination schemes are hypothetical, the available data provide ample rationale to support testing of the proposed concepts.
Future directions Although evolution has played a critical role in the phylogeny of immunological memory (102), intelligent design will be needed to devise systems capable of providing definitive analyses of the development, migration, and function of memory
Lefranc¸ois Central and effector memory T cells
T-cell subsets. An ongoing goal is the further delineation of memory cell subsets predicated not only on phenotype, function, and location, but also on defined molecular signatures. Relating the development of memory T-cell subsets to protection against infection or cancer remains one of the most important questions extant. While work has begun in this area, much additional research is needed to determine the protective capacity of each subset in vivo against a variety of pathogens and tumors. In such studies, the pathogen, the dose, and the route of infection require consideration when drawing conclusions regarding relative differences in responses mediated by distinct memory cell subtypes. With
respect to memory T-cell migration, the available data are consistent with a dynamic and continuous movement of memory cells throughout most of the body, but the tissues of origin of blood-borne memory T-cell migrants remain unknown, and the mechanisms by which memory cells enter and exit tissues require further analysis. Thus, imaging systems are needed in which entry and egress of single cells from tissues can be observed in real time in vivo. Functional assays including the visualization of early signal transduction events (103) in conjunction with sophisticated tracking systems will eventually provide a holistic view of memory cell movement and function.
References 1. Lipton MM, Freund J. The formation of complement fixing and neutralizing antibodies after the injection of inactivated rabies virus with adjuvants. J Immunol 1950;64:297–303. 2. Wherry EJ, Blattman JN, Murali-Krishna K, van der MR, Ahmed R. Viral persistence alters CD8 T-cell immunodominance and tissue distribution and results in distinct stages of functional impairment. J Virol 2003;77:4911–4927. 3. Fuller MJ, Khanolkar A, Tebo AE, Zajac AJ. Maintenance, loss, and resurgence of T cell responses during acute, protracted, and chronic viral infections. J Immunol 2004;172:4204–4214. 4. Cauley LS, et al. Virus-specific CD4+ memory T cells in nonlymphoid tissues express a highly activated phenotype. J Immunol 2002;169:6655–6658. 5. Khanna KM, Bonneau RH, Kinchington PR, Hendricks RL. Herpes simplex virus-specific memory CD8+ T cells are selectively activated and retained in latently infected sensory ganglia. Immunity 2003;18:593–603. 6. Hammarlund E, et al. Duration of antiviral immunity after smallpox vaccination. Nat Med 2003;9:1131–1137. 7. Slifka MK. Immunological memory to viral infection. Curr Opin Immunol 2004;16:443–450. 8. Beverley PC. Kinetics and clonality of immunological memory in humans. Semin Immunol 2004;16:315–321. 9. Zimmermann C, Brduscha-Riem K, Blaser C, Zinkernagel RM, Pircher H. Visualization, characterization, and turnover of CD8+ memory T cells in virus-infected hosts. J Exp Med 1996;183:1367–1375. 10. Masopust D, Vezys V, Marzo AL, Lefranc¸ois L. Preferential localization of effector memory cells in nonlymphoid tissue. Science 2001;291:2413–2417.
11. Curtsinger JM, Lins DC, Mescher MF. CD8+ memory T cells (CD44high, Ly-6C+) are more sensitive than naive cells (CD44low, Ly-6C-) to TCR/CD8 signaling in response to antigen. J Immunol 1998;160:3236–3243. 12. Farber DL. Differential TCR signaling and the generation of memory T cells. J Immunol 1998;160:535–539. 13. Hussain SF, Anderson CF, Farber DL. Differential SLP-76 expression and TCRmediated signaling in effector and memory CD4 T cells. J Immunol 2002;168:1557–1565. 14. Bachmann MF, et al. Developmental regulation of Lck targeting to the CD8 coreceptor controls signaling in naive and memory T cells. J Exp Med 1999;10:1521–1530. 15. Kaech SM, Hemby S, Kersh E, Ahmed R. Molecular and functional profiling of memory CD8 T cell differentiation. Cell 2002;111:837–851. 16. Badovinac VP, Messingham KA, Jabbari A, Haring JS, Harty JT. Accelerated CD8+ T-cell memory and prime-boost response after dendritic-cell vaccination. Nat Med 2005;11:748–756. 17. Williams MA, Bevan MJ. Shortening the infectious period does not alter expansion of CD8 T cells but diminishes their capacity to differentiate into memory cells. J Immunol 2004;173:6694–6702. 18. Van Stipdonk MJ, Lemmens EE, Schoenberger SP. Naive CTLs require a single brief period of antigenic stimulation for clonal expansion and differentiation. Nat Immunol 2001;2:423–429. 19. Kaech SM, Ahmed R. Memory CD8+ T cell differentiation: initial antigen encounter triggers a developmental program in naive cells. Nat Immunol 2001;2:415–422.
20. Lefranc¸ois L, Marzo A, Williams K. Sustained response initiation is required for T cell clonal expansion but not for effector or memory development in vivo. J Immunol 2003;171:2832–2839. 21. Hou S, Hyland L, Ryan KW, Portner A, Doherty PC. Virus specific CD8+ T-cell memory determined by clonal burst size. Nature 1994;369:652–654. 22. Curtsinger JM, Johnson CM, Mescher MF. CD8 T cell clonal expansion and development of effector function require prolonged exposure to antigen, costimulation, and signal 3 cytokine. J Immunol 2003;171:5165–5171. 23. Edelson BT, Unanue ER. MyD88-dependent but Toll-like receptor 2-independent innate immunity to Listeria: no role for either in macrophage listericidal activity. J Immunol 2002;169:3869–3875. 24. Feng H, et al. Listeria-infected myeloid dendritic cells produce IFN-b, priming T cell activation. J Immunol 2005;175:421–432. 25. Muraille E, et al. Distinct in vivo dendritic cell activation by live versus killed Listeria monocytogenes. Eur J Immunol 2005;35:1463–1471. 26. Bousso P, Robey E. Dynamics of CD8(+) T cell priming by dendritic cells in intact lymph nodes. Nat Immunol 2003;4:579–585. 27. Mempel TR, Henrickson SE, von Andrian UH. T-cell priming by dendritic cells in lymph nodes occurs in three distinct phases. Nature 2004;427:154–159. 28. Cahalan MD, Parker I. Close encounters of the first and second kind: T-DC and T-B interactions in the lymph node. Semin Immunol 2005;17:442–451. 29. Schluns KS, Kieper WC, Jameson SC, Lefranc¸ois L. Interleukin-7 mediates the homeostasis of naive and memory CD8 T cells in vivo. Nat Immunol 2000;1:426–432.
Immunological Reviews 211/2006
101
Lefranc¸ois Central and effector memory T cells
30. Schluns KS, Williams K, Ma A, Zheng XX, Lefranc¸ois L. Cutting edge: requirement for IL-15 in the generation of primary and memory antigen-specific CD8 T cells. J Immunol 2002;168:4827–4831. 31. Schluns KS, Lefrancois L. Cytokine control of memory T-cell development and survival. Nat Rev Immunol 2003;3:269–279. 32. Klonowski KD, Lefrancois L. The CD8 memory T cell subsystem: integration of homeostatic signaling during migration. Semin Immunol 2005;17:219–229. 33. Yajima T, et al. IL-15 regulates CD8+ T cell contraction during primary infection. J Immunol 2006;176:507–515. 34. Sprent J, Tough DF. T cell death and memory. Science 2001;293:245–248. 35. Gett AV, Sallusto F, Lanzavecchia A, Geginat J. T cell fitness determined by signal strength. Nat Immunol 2003;4:355–360. 36. Grayson JM, Zajac AJA, Ahmed R. Increased expression of Bcl-2 in antigen-specific memory CD8+ T cells. J Immunol 2000;164:3950–3954. 37. Marrack P, Kappler J. Control of T cell viability. Annu Rev Immunol 2004;22:765–787. 38. Mitchell TC, et al. Immunological adjuvants promote activated T cell survival via induction of Bcl-3. Nat Immunol 2001;2:397–402. 39. Valenzuela JO, Hammerbeck CD, Mescher MF. Cutting edge: Bcl-3 up-regulation by signal 3 cytokine (IL-12) prolongs survival of antigenactivated CD8 T cells. J Immunol 2005;174:600–604. 40. Hildeman DA, et al. Activated T cell death in vivo mediated by proapoptotic bcl-2 family member bim. Immunity 2002;16:759–767. 41. Pellegrini M, Belz G, Bouillet P, Strasser A. Shutdown of an acute T cell immune response to viral infection is mediated by the proapoptotic Bcl-2 homology 3-only protein Bim. Proc Natl Acad Sci USA 2003;100:14175–14180. 42. Sudo T, et al. Expression and function of the interleukin 7 receptor in murine lymphocytes. Proc Natl Acad Sci USA 1993;90:9125–9129. 43. Maraskovsky E, et al. Impaired survival and proliferation in IL-7 receptor-deficient peripheral T cells. J Immunol 1996;157:5315–5323. 44. Kaech SM, Tan JT, Wherry EJ, Konieczny BT, Surh CD, Ahmed R. Selective expression of the interleukin 7 receptor identifies effector CD8 T cells that give rise to long-lived memory cells. Nat Immunol 2003;4:1191–1198. 45. Badovinac VP, Porter BB, Harty JT. CD8+ T cell contraction is controlled by early inflammation. Nat Immunol 2004;5:809–817.
102
46. Lacombe MH, Hardy MP, Rooney J, Labrecque N. IL-7 receptor expression levels do not identify CD8+ memory T lymphocyte precursors following peptide immunization. J Immunol 2005;175:4400–4407. 47. Seddon B, Tomlinson P, Zamoyska R. Interleukin 7 and T cell receptor signals regulate homeostasis of CD4 memory cells. Nat Immunol 2003;4:680–686. 48. Goldrath AW, et al. Cytokine requirements for acute and basal homeostatic proliferation of naive and memory CD8+ T cells. J Exp Med 2002;195:1515–1522. 49. Li J, Huston G, Swain SL. IL-7 promotes the transition of CD4 effectors to persistent memory cells. J Exp Med 2003;198:1807–1815. 50. Kondrack RM, Harbertson J, Tan JT, McBreen ME, Surh CD, Bradley LM. Interleukin 7 regulates the survival and generation of memory CD4 cells. J Exp Med 2003;198:1797–1806. 51. Mosmann TR, Coffman RL. TH1 and TH2 cells: different patterns of lymphokine secretion lead to different functional properties. Annu Rev Immunol 1989;7:145–173. 52. Swain SL. Generation and in vivo persistence of polarized Th1 and Th2 memory cells. Immunity 1994;1:543–552. 53. Boursalian TE, Bottomly K. Stability of naive and memory phenotypes on resting CD4 T cells in vivo. J Immunol 1999;163:9–16. 54. Sallusto F, Lenig D, Forster R, Lipp M, Lanzavecchia A. Two subsets of memory T lymphocytes with distinct homing potentials and effector functions. Nature 1999;401:708–712. 55. Warnock RA, Askari S, Butcher EC, von Andrian UH. Molecular mechanisms of lymphocyte homing to peripheral lymph nodes. J Exp Med 1998;187:205–216. 56. Parretta E, Cassese G, Barba P, Santoni A, Guardiola J, Di Rosa F. CD8 cell division maintaining cytotoxic memory occurs predominantly in the bone marrow. J Immunol 2005;174:7654–7664. 57. Mazo IB, et al. Bone marrow is a major reservoir and site of recruitment for central memory CD8+ T cells. Immunity 2005;22:259–270. 58. Becker TC, Coley SM, Wherry EJ, Ahmed R. Bone marrow is a preferred site for homeostatic proliferation of memory CD8 T cells. J Immunol 2005;174:1269–1273. 59. Di Rosa F, Pabst R. The bone marrow: a nest for migratory memory T cells. Trends Immunol 2005;26:360–366. 60. Sallusto F, Geginat J, Lanzavecchia A. Central memory and effector memory T cell subsets: function, generation, and maintenance. Annu Rev Immunol 2004;22:745–763.
Immunological Reviews 211/2006
61. Baars PA, et al. Properties of murine (CD8+) CD27- T cells. Eur J Immunol 2005;35:3131–3141. 62. Baron V, et al. The repertoires of circulating human CD8(+) central and effector memory T cell subsets are largely distinct. Immunity 2003;18:193–204. 63. Wherry EJ, et al. Lineage relationship and protective immunity of memory CD8 T cell subsets. Nat Immunol 2003;4:225–234. 64. Marzo AL, Klonowski KD, Le Bon A, Borrow P, Tough DF, Lefrancois L. Initial T cell frequency dictates memory CD8+ T cell lineage commitment. Nat Immunol 2005;6:793–799. 65. Yoshida R, et al. Molecular cloning of a novel human CC chemokine EBI1-ligand chemokine that is a specific functional ligand for EBI1, CCR7. J Biol Chem 1997;272:13803–13809. 66. Potsch C, Vohringer D, Pircher H. Distinct migration patterns of naive and effector CD8 T cells in the spleen: correlation with CCR7 receptor expression and chemokine reactivity. Eur J Immunol 1999;29:3562–3570. 67. Sallusto F, et al. Switch in chemokine receptor expression upon TCR stimulation reveals novel homing potential for recently activated T cells. Eur J Immunol 1999;29:2037–2045. 68. Gallatin WM, Weissman IL, Butcher EC. A cell-surface molecule involved in organspecific homing of lymphocytes. Nature 1983;304:30–34. 69. Jalkanen S, Reichert RA, Gallatin WM, Bargatze RF, Weissman IL, Butcher EC. Homing receptors and the control of lymphocyte migration. Immunol Rev 1986;91:39–60. 70. Jung TM, Gallatin WM, Weissman IL, Dailey MO. Down-regulation of homing receptors after T cell activation. J Immunol 1988;141:4110–4117. 71. Jung TM, Dailey MO. Reversibility of loss of homing receptor expression following activation. Adv Exp Med Biol 1988;237:519–524. 72. Jung TM, Dailey MO. Rapid modulation of homing receptors (gp90MEL-14) induced by activators of protein kinase C. Receptor shedding due to accelerated proteolytic cleavage at the cell surface. J Immunol 1990;144:3130–3136. 73. Chao CC, Jensen R, Dailey MO. Mechanisms of L-selectin regulation by activated T cells. J Immunol 1997;159:1686–1694. 74. Venturi GM, et al. Leukocyte migration is regulated by L-selectin endoproteolytic release. Immunity 2003;19:713–724. 75. Smalley DM, Ley K. L-selectin: mechanisms and physiological significance of ectodomain cleavage. J Cell Mol Med 2005;9:255–266.
Lefranc¸ois Central and effector memory T cells
76. Campbell DJ, Kim CH, Butcher EC. Chemokines in the systemic organization of immunity. Immunol Rev 2003;195:58–71. 77. Lefranc¸ois L, Masopust D. T cell immunity in lymphoid and non-lymphoid tissues. Curr Opin Immunol 2002;14:503–508. 78. Masopust D, et al. Activated primary and memory CD8 T cells migrate to nonlymphoid tissues regardless of site of activation or tissue of origin. J Immunol 2004;172:4875–4882. 79. Bunster E, Meyer RK. An improved method of parabiosis. Anat Rec 1933;57:339–343. 80. Donskoy E, Goldschneider I. Thymocytopoiesis is maintained by bloodborne precursors throughout postnatal life. A study in parabiotic mice. J Immunol 1992;148:1604–1612. 81. Klonowski KD, Williams KJ, Marzo AL, Blair DA, Lingenheld EG, Lefrancois L. Dynamics of blood-borne CD8 memory T cell migration in vivo. Immunity 2004;20:551–562. 82. Mackay CR, Marston WL, Dudler L. Naive and memory T cells show distinct pathways of lymphocyte recirculation. J Exp Med 1990;171:801–817. 83. Mackay CR, Andrew DP, Briskin M, Ringler DJ, Butcher EC. Phenotype, and migration properties of three major subsets of tissue homing T cells in sheep. Eur J Immunol 1996;26:2433–2439. 84. Bromley SK, Thomas SY, Luster AD. Chemokine receptor CCR7 guides T cell exit from peripheral tissues and entry into afferent lymphatics. Nat Immunol 2005;6:895–901. 85. Debes GF, et al. Chemokine receptor CCR7 required for T lymphocyte exit from peripheral tissues. Nat Immunol 2005;6:889–894.
86. Forster R, et al. CCR7 coordinates the primary immune response by establishing functional microenvironments in secondary lymphoid organs. Cell 1999;99:23–33. 87. MartIn-Fontecha A, et al. Regulation of dendritic cell migration to the draining lymph node: impact on T lymphocyte traffic and priming. J Exp Med 2003;198:615–621. 88. Randolph GJ, Angeli V, Swartz MA. Dendritic-cell trafficking to lymph nodes through lymphatic vessels. Nat Rev Immunol 2005;5:617–628. 89. Unsoeld H, Pircher H. Complex memory T-cell phenotypes revealed by coexpression of CD62L and CCR7. J Virol 2005;79:4510–4513. 90. Kim SK, Schluns KS, Lefranc¸ois L. Induction and visualization of mucosal memory CD8 T cells following systemic virus infection. J Immunol 1999;163:4125–4132. 91. Jelley-Gibbs DM, Brown DM, Dibble JP, Haynes L, Eaton SM, Swain SL. Unexpected prolonged presentation of influenza antigens promotes CD4 T cell memory generation. J Exp Med 2005;202:697–706. 92. Dubois S, Mariner J, Waldmann TA, Tagaya Y. IL-15Ralpha recycles and presents IL-15 In trans to neighboring cells. Immunity 2002;17:537–547. 93. Becker TC, et al. Interleukin 15 is required for proliferative renewal of virus-specific memory CD8 T cells. J Exp Med 2002;195:1541–1548. 94. Burkett PR, et al. IL-15Ralpha expression on CD8+ T cells is dispensable for T cell memory. Proc Natl Acad Sci USA 2003;100:4724–4729.
95. Koka R, et al. Interleukin (IL)-15R[alpha]deficient natural killer cells survive in normal but not IL-15R[alpha]-deficient mice. J Exp Med 2003;197:977–984. 96. Schluns KS, Klonowski KD, Lefranc¸ois L. Trans-regulation of memory CD8 T cell proliferation by IL-15Ra+ bone marrow-derived cells. Blood 2004;103:988–994. 97. Sandau MM, Schluns KS, Lefrancois L, Jameson SC. Cutting edge: transpresentation of IL-15 by bone marrow-derived cells necessitates expression of IL-15 and IL-15R alpha by the same cells. J Immunol 2004;173:6537–6541. 98. Woodland DL. Jump-starting the immune system: prime-boosting comes of age. Trends Immunol 2004;25:98–104. 99. Zaph C, Uzonna J, Beverley SM, Scott P. Central memory T cells mediate long-term immunity to Leishmania major in the absence of persistent parasites. Nat Med 2004;10:1104–1110. 100. Roberts AD, Ely KH, Woodland DL. Differential contributions of central and effector memory T cells to recall responses. J Exp Med 2005;202:123–133. 101. An LL, Rodriguez F, Harkins S, Zhang J, Whitton JL. Quantitative and qualitative analyses of the immune responses induced by a multivalent minigene DNA vaccine. Vaccine 2000;18:2132–2141. 102. Flajnik MF, Du PL. Evolution of innate and adaptive immunity: can we draw a line? Trends Immunol 2004;25:640–644. 103. Zell T, Khoruts A, Ingulli E, Bonnevier JL, Mueller DL, Jenkins MK. Single-cell analysis of signal transduction in CD4 T cells stimulated by antigen in vivo. Proc Natl Acad Sci USA 2001;98:10805–10810.
Immunological Reviews 211/2006
103
