The potential of CD4 T-cell memory

Abstract

While many aspects of memory T-cell immunobiology have been characterized, we suggest that we know only a fraction of the effector functions that CD4 T cells can bring to bear during secondary challenges. Exploring the full impact of memory CD4 T-cell responses is key to the development of improved vaccines against many prominent pathogens, including influenza viruses, and also to a better understanding of the mechanisms of autoimmunity. Here we discuss factors regulating the generation of memory CD4 T cells during the activation of naïve cells and how the nature of the transition from highly activated effector to resting memory upon the resolution of primary responses might impact memory CD4 T-cell heterogeneity in vivo. We stress that memory CD4 T cells have unique functional attributes beyond the secretion of T helper (Th) subset-associated cytokines that can shape highly effective secondary responses through novel mechanisms. These include the recruitment of innate inflammatory responses at early phases of secondary responses as well as the action of enhanced direct effector functions at later phases, in addition to well-established helper roles for CD8 T-cell and B-cell responses.

Introduction

The generation of T-cell memory is a fundamental goal of vaccine strategies aimed at pathogens against which neutralizing antibody alone does not confer lasting immunity. In addition, T-cell-based vaccines offer the possibility of long-term protection against a range of seasonal and pandemic influenza virus strains as there is far less variation documented in internal viral proteins recognized by T cells compared with the antigenic shift and drift observed in coat proteins recognized by antibody.¹ However, the mechanisms by which memory T cells, especially CD4 T cells, contribute to protection against influenza virus and other pathogens are not fully understood. Here we discuss aspects of the generation and maintenance of memory CD4 T cells, their heterogeneity, and the diverse functions memory CD4 T cells can carry out during secondary immune responses.

Memory CD4 T-cell generation

During primary immune responses, naïve CD4 T cells recognizing antigen expand many fold and differentiate into highly activated effector cells.^2,3 Depending on the impact of several variables during priming, effectors can become polarized into distinct subsets with specialized functions and attributes. The best characterized of such subsets are T helper type 1 (Th1) and Th2, which have long been known to promote effective cell-mediated and humoral responses, respectively.⁴ More recently, additional subsets have been described, including Th17, follicular helper cells (TFH), and various regulatory populations^5–7 (Table 1). In addition, substantial heterogeneity and plasticity, as assessed by cytokine production patterns, have been observed within subsets, especially in vivo.^8,9 This impressive diversity in effector subtype underscores the tremendous potential range of T-cell functions and suggests that there is a corresponding complexity in the orchestration of optimal protective immunity. Effector CD4 T cells influence immune responses through both ‘helper’ and direct ‘effector’ functions.¹⁰ These will be discussed later in more detail.

Table 1.

Characteristics of CD4 T-cell subsets

CD4 T-cell effector subset	Polarizing factors	Transcription factors	Signature cytokines	Susceptibility to AICD
Th1	IL-12, IFN-γ	T-bet	IFN-γ, TNF, IL-2	++++
Th2	IL-4	GATA-3	IL-4, IL-5, IL-13	+
Th17	TGF-β, IL-6	ROR-γt	IL-17, IL-21, IL-22	++
TFh	IL-6, IL-21	Bcl-6	IL-4, IL-21	?
Treg	TGF-β, RA	Foxp3	IL-10	?

AICD, activation induced cell death; IFN, interferon; IL, interleukin; TGF, transforming growth factor; Th, T helper; TNF, tumour necrosis factor; Treg, regulatory T cell.

After the resolution of a primary response, the majority of effector CD4 T cells die via apoptosis, leaving a small population of memory cells. Surviving antigen-specific cells revert to a resting state but at frequencies generally much higher than those found in naïve individuals. A quantitative increase in antigen-specific T cells thus represents a well-documented advantage of the memory state.^1,11–13 In at least some situations, the magnitude of the effector response has been found to correlate with the size of the resulting memory pool.¹⁴ The intracellular pathways and external signals involved in co-ordinating the contraction of effectors in vivo remain poorly understood.¹⁵ For example, signalling through the death receptor Fas has been studied extensively, but its relevance during contraction in vivo remains unclear.¹⁶ The pro-apoptotic molecule Bim (B-cell lymphoma 2 interacting mediator of cell death) seems to play a central role during the contraction of effectors and its expression within a population of responding cells limits the pool of memory-competent cells.¹⁷ Recent studies suggest that Bim expression by effector CD4 T cells is influenced during priming by the strength of T-cell receptor (TCR) stimulation,¹⁸ which may be regulated by competition between responding cells for antigen.

Signals impacting the effector-to-memory transition

While the signals involved in the activation of naïve cells and their transition into effectors have been studied extensively,^3,19,20 comparatively little is known regarding the cessation of effector responses and the effector-to-memory transition. Understanding this process in greater detail holds the potential for improving memory generation and survival (Fig. 1). After optimal priming, we and others have shown that removal of antigen and inflammatory cytokines marks a critical starting point for the generation of memory.^14,21,22 In fact, few other extrinsic signals are required, although interleukin (IL)-7 signalling has been shown to support the survival of effectors during the transition to memory.^23,24 The lack of signalling in this process is in contrast to the tightly regulated set of cell-surface molecules and cytokines required for the full activation of naïve CD4 T cells. This suggests that critical aspects of the effector-to-memory transition are programmed at earlier stages of the CD4 T-cell response.

Figure 1.

Many factors can influence the generation of memory CD4 T cells. Variables present during T-cell activation, the effector phase, and after the resolution of a primary response can all impact the size and quality of the resulting memory population. APC, antigen-presenting cell; IL, interleukin; TCR, T-cell receptor.

In support of this theory, several factors associated with T-cell activation and effector generation have been reported also to impact the generation of memory. For example, IL-2 signalling during priming has been shown to enhance long-term survival of CD4 T cells,²⁵ and a similar role for CD28-dependent signalling during priming has been reported for optimal memory generation.²⁶ The results of several studies also support an important role for OX-40-dependent signalling in maximizing memory, at least under Th2-polarizing conditions.^27–30 Finally, the level of or duration of antigen stimulation during priming can impact memory CD4 T-cell generation, as studies suggest that limiting TCR signalling can inhibit memory formation.³¹

Rapid effector-to-memory transition

We have recently investigated the kinetics of the effector-to-memory transition by comparing the phenotypes, functional attributes, and gene expression profiles of well-defined effector and memory CD4 T cells to those of transitional populations of effectors rested for short periods of time in the absence of antigen and inflammatory cytokines.²² Surprisingly, we found the effector-to-memory transition to be quite rapid – effectors rested for only 3 days closely resembled canonical long-term memory cells by all criteria tested. The rapid kinetics of the effector-to-memory transition seem to be part of a default program requiring few external signals, as similar results were obtained using both Th1 and Th2 lineage effectors rested both in vitro and in vivo and primed both in vitro and in vivo.³² Thus, the immune system is poised to abruptly shift from the effector mode, which can itself be pathogenic, to the more benign memory mode shortly after the resolution of a primary response.

Differential maintenance of CD4 and CD8 T-cell memory

While an important difference distinguishing the naïve from the memory state is an increase in antigen-specific cells, attrition of memory T-cell populations occurs with time. There also appear to be very different rules for how memory CD4 and CD8 T cells are maintained in the long term. As compared with memory CD8 T-cell populations, which tend to remain relatively stable, memory CD4 T-cell numbers tend to gradually decrease.^33–35 Several factors could account for the differential homeostasis of antigen-experienced CD4 and CD8 T cells. A primary mechanism could involve the ability to access critical survival factors. In support of this hypothesis, long-term CD4 T-cell survival has been shown to be positively affected by IL-15 signalling,³⁶ and IL-15 is a cytokine thought to be primarily involved in memory CD8 T-cell homeostasis. Thus, memory CD8 T cells may out-compete memory CD4 T cells for IL-15 via expression of a higher affinity receptor,³⁷ thereby causing the latter population to be less fit for long-term survival. A larger population of antigen-experienced CD4 as compared to CD8 T cells may also require cognate interactions with antigen.^38–40 Residual antigen depots have been reported to be present after the resolution of primary responses in diverse models,^41–44 and their gradual waning over time might constrict survival niches. For example, following influenza infection, residual viral antigen presentation seems differentially capable of generating and perhaps maintaining CD4 and CD8 T-cell memory.^41,45

Several observations support the hypothesis that niches for memory populations are indeed limited in T-cell-replete mice, suggesting an environment of intense competition amongst antigen-experienced T cells.⁴⁶ It is currently unclear what defines such niches in vivo. An obvious possibility is access to major histocompatibility complex class II (MHC II)–peptide complexes recognized by TCRs above a certain threshold⁴⁷ (although these need not be complexes for which the TCR is specific). This hypothesis is supported by observations showing that long-lived memory CD4 T cells can express higher affinity TCRs than naïve cells of the same specificity.^43,48 A hallmark of memory CD4 T cells, however, is their ability to survive for extended periods of time upon transfer to MHC II-deficient hosts, while naïve cells rapidly disappear.⁴⁹ Alternatively, survival niches may be based on the ability of memory cells to better access sources of IL-7 and/or other survival signals through interactions between tumour necrosis factor (TNF) family receptors and their ligands.⁵⁰

Functional properties of memory CD4 T cells

While a quantitative increase in antigen-specific CD4 T cells represents a major advantage of the memory state, we believe that a more important feature of T-cell memory is the enhanced qualitative attributes of memory as compared with naïve and effector T cells. These changes affect homeostasis and longevity as well as functional attributes (Table 2). Most of the functional changes that distinguish memory CD4 T cells from naïve cells can be viewed as endowing memory cells with enhanced and more rapid responsiveness upon antigen recognition.

Table 2.

Properties of CD4 T-cell differentiation stages

Functional characteristics	Naïve	Effector	Memory	Secondary effector
Effector function	–	++++	+++	+++++
IL-2:IL-10 ratio	+++	+	+++++	++++
AICD	–	++++	++	?
Adhesion	–	++++	++	+++++
Migration	Lymphoid	Non-lymphoid	Lymphoid	Non-lymphoid

AICD, activation induced cell death; IL, interleukin.

Most strikingly, while naïve CD4 T cells primarily produce IL-2 shortly after TCR triggering, memory CD4 T cells retain the ability to produce effector-associated cytokines upon re-stimulation.⁵¹ Also, while naïve cells respond to antigenic stimulation only after a lag period, memory cells are capable of more rapid cytokine production and proliferation.^52,53 The intracellular signalling components responsible for the rapid response kinetics of memory CD4 T cells are now being investigated.⁵⁴ Finally, memory CD4 T cells display less stringent activation requirements, and respond optimally to lower doses of antigen and lower levels of costimulation than naïve cells.^{22,52,53,55–59} More recent studies, however, suggest that costimulation (in particular through CD28 and CD40L) is important in regulating certain aspects of memory responses.^58,60 Combined, these observations point towards a ‘faster, bigger and better’ response of memory as compared with naïve CD4 T cells.

Additional functional qualities distinguish memory CD4 T cells from highly activated effectors (Table 2). Several groups have demonstrated that memory CD4 T cells lose susceptibility to the cytokine-induced apoptosis⁶¹ and activation-induced cell death that characterizes activated effectors.^21,22,62 In several instances, effectors lose the ability to produce IL-2, while its expression is often observed in responding memory cells.^{12,21,22,32,63} We and others have also observed that IL-10 production is significantly reduced in memory as compared with effector CD4 T cells.^22,64,65 It is interesting to speculate that production of IL-10, a factor associated with suppression of immune responses,⁶⁶ is involved in a type of immune modulation acting during the peak of primary but not secondary T cell responses.⁶⁵ Memory CD4 T cells also down-regulate expression of a myriad of activation and adhesion molecules highly expressed on effectors, consistent with a return to a resting state.²² However, expression of certain molecules, such as CD44⁶⁷ and CD54,⁵⁵ remains up-regulated on resting memory cells, and may facilitate enhanced interactions with antigen-bearing antigen-presenting cells (APCs). Finally, the in vivo migration of memory CD4 T cells differs from that of effector CD4 T cells under steady-state conditions. While highly activated effectors can traffic throughout lymphoid organs as well as the periphery, trafficking of memory cells resembles the more restricted pattern of recirculation through lymphoid organs that characterizes naïve populations,⁶⁸ and can even be more restricted than that of naïve cells.⁶⁹ The wider trafficking pattern of effectors correlates with their increased expression of several adhesion molecules and loss of expression of CD62L.⁶⁸

Together, these observations support the theory that effector CD4 T cells progress beyond an early ‘helper’ stage, where they enhance responses of B cells and CD8 T cells, towards an end stage where they are more likely to be of a restricted subset. Memory cells, in contrast, seem to retain important effector attributes but also re-acquire others associated with naïve CD4 T cells and polyfunctional potential. Gene expression analysis confirms this, as long-term memory populations closely resemble naïve CD4 T cells.²²

Heterogeneity

One of the most striking elements of T-cell memory is heterogeneity. Several groups have demonstrated a great deal of diversity, and have defined memory CD4 T-cell subsets based on functional capability, surface marker expression, and anatomical location.^70–73 The root of this diversity is not well understood, nor is its long-term maintenance.^74–77 The rapid kinetics of the effector-to-memory transition may help to explain some elements of heterogeneity. For example, we have previously shown that, based on dilution of carboxyfluorescein succinimidyl ester (CFSE) and surface marker expression, a wide range of heterogeneity characterizes influenza-specific effectors during a primary infection.⁷⁸ Surprisingly, much of this diversity is retained in long-term memory populations.¹⁰ This suggests that, at least in terms of the number of cell divisions and diverse cell-surface markers, no obvious pre-memory effector state is required to facilitate the effector-to-memory transition (Fig. 2).

Figure 2.

Effector heterogeneity is influenced by several factors. While some effectors progress to a terminally differentiated state, the rapid kinetics of the effector-to-memory transition upon resolution of a primary immune response enables a great deal of heterogeneity to be retained in the memory phase. APC, antigen-presenting cell; TCR, T-cell receptor.

An intriguing possibility is that some elements of heterogeneity are attributable to differential kinetics of entry of naïve T cells into a dynamic immune response. Observations show that T cells respond to antigen throughout the primary response, even during its resolution, and that naïve cells responding later, when levels of antigen and inflammation are lower, can give rise to effectors that are phenotypically and functionally distinct from effectors generated at earlier times.^{41,45,79–81} This dynamic variable, coupled with T-cell-intrinsic variables such as TCR affinity, and T-cell-extrinsic variables such as the type of APC, could then potentially lead to a spectrum of responding antigen-specific T cells with regard to activation status, phenotype, function and longevity. It is possible that different subsets of responding T cells thus created during the primary response might represent populations of specialized cells with optimal helper, killer and regulatory functions.^10,82 Whatever the mechanism, diversity in T-cell effectors seems to be fostered and may thus be important in orchestrating and affecting optimal responses. We suggest that, if CD4 T-cell effectors in different relative states of activation gain and/or lose functional and/or regulatory attributes, then it is likely that an equally diverse memory pool would provide optimal protection during secondary challenges.

Potential impact of memory CD4 T cells

How memory CD4 T cells protect during secondary challenge is not fully understood. This is in part a consequence of the fact that CD4 T cells and their products influence immune responses against different pathogens differently. For example, CD4 T-cell help is required for optimal primary CD8 T-cell responses in some but not all challenge models and interferon (IFN)-γ produced by memory CD4 T cells is required for protection against some but not all pathogens that elicit a Th1-response.^83–86 Also, as already discussed, a great deal of heterogeneity characterizes memory CD4 T cells, with different subsets displaying different functions. Thus, it is increasingly difficult to ascribe protective function to memory CD4 T cells in general, rather than to discrete subsets that might not be present or play an important role in all situations. Finally, the ‘quality’ of memory CD4 T cells can vary dramatically, impacting their protective potential. For example, the ability of both mouse and human memory CD4 T cells to produce multiple cytokines has recently received considerable attention, and this phenotype has been correlated with superior functional ability and protective capacity compared with populations displaying a more restricted cytokine secretion pattern.^87–92

We propose that the unique functional qualities of memory CD4 T cells have the potential to impact a secondary response in multiple ways throughout its continuum. Which aspects of CD4 T-cell responses are critical for protection will depend on the nature of the pathogen and the contribution of other arms of innate and acquired immunity. Below we briefly describe the potential for unique impacts of memory CD4 T cells throughout secondary responses (Fig. 3).

Figure 3.

Memory CD4 T cells impact multiple stages of secondary responses. Memory CD4 T cells rapidly produce an array of cytokines and chemokines upon antigen recognition, and can recruit elements of the innate immune system to do likewise. As they divide and further differentiate, memory CD4 T cells can act as superior helpers for CD8 T-cell and B-cell responses compared with naïve CD4 T cells. Finally, highly activated secondary effectors impact secondary responses through direct effector functions that differ from those of primary effectors. All stages concertedly impact the kinetics of pathogen clearance.

Recruitment of innate immunity

Production of pathogen-appropriate cytokines at the early stages of immune responses represents an important contribution of memory CD4 T cells. For example, it has long been established that Th1-associated cytokines, most notably IFN-γ, can activate macrophages.⁹³ This mode of action represents a critical protective effect of CD4 T cells during experimental tuberculosis (TB) challenge.⁸⁶ IFN-γ production by memory CD4 T cells can also induce an antiviral state within infected cells in addition to other systemic effects.⁹⁴ In addition, IFN-γ produced by CD4 T cells can enhance the production of an array of proinflammatory cytokines and chemokines that can act to recruit diverse innate cellular populations and CD8 T-cell effectors to sites of infection.^95,96 Similarly, the Th2-associated cytokines IL-4 and IL-13 play important roles in protection against a broad range of pathogens.⁹⁷

Our recent studies indicate that, during influenza infection, Th1-polarized virus-specific memory, but not naïve, CD4 T cells induce the expression of a wide array of innate inflammatory cytokines and chemokines (IICs) in an IFN-γ- and TNF-independent manner (Strutt, McKinstry, and Swain, submitted). Furthermore, we observed IIC induction by memory CD4 T cells in the absence of conserved pattern recognition receptor signalling. Importantly, memory CD4 T-cell induction of IIC responses led to early viral control. These results demonstrate that memory CD4 T-cell responses can directly recruit elements of the innate immune system to combat infection at early stages of a recall challenge. However, some elements of the inflammatory responses mediated by memory CD4 T cells may become pathogenic at later stages of infection,⁹⁸ and may thus be involved in certain autoimmune conditions.

Helper functions of memory CD4 T cells

That CD4 T cells are commonly referred to as ‘helper’ T cells underscores their specialized role in the adaptive immune system. CD4 T cells are critical in helping B-cell responses, both through cognate interactions and through secreted mediators. In particular, interactions between CD40L, expressed on activated CD4 T cells, and CD40, constitutively expressed on B cells, are critical for germinal centre formation, isotype switching and affinity maturation.⁹⁹ Other molecules expressed by activated CD4 T cells, such as CD28, OX40 and inducible T-cell costimulator, can also play prominent roles in the generation of optimal B-cell responses.^100–102 In addition, signaling lymphocyte activation molecule (SLAM) associated protein expression by CD4 T cells is required for helper function through a mechanism distinct from ligation of CD40 and OX40L.¹⁰³ Finally, Th1/Th2 cytokines produced by CD4 T cells can dramatically impact B-cell responses.^4,104 More recently, IL-21 and IL-17 have been also reported to profoundly influence the quality of T-cell-dependent B-cell responses.^105,106 Importantly, memory CD4 T cells have been demonstrated to be superior helpers for B-cell antibody production as compared with naïve CD4 T cells, and this is most readily explained by the ability of memory cells to reach a helper-competent activation status more quickly than naive cells, and their ability to produce more cytokines with more rapid kinetics.^12,55,107

CD4 T-cell help can also be critical in the generation and maintenance of optimal cytotoxic CD8 T-cell responses.^83,108 Several studies support a critical role for CD40L-CD40 signalling in this process, whether through direct interactions with CD8 T cells¹⁰⁹ or through interactions with APCs.^110–112 As for B-cell responses, recent studies have demonstrated an enhanced capability of memory CD4 T cells to help both primary^113,114 and secondary^115,116 CD8 T-cell responses. In at least some situations, IL-2 production by CD4 T cells plays a critical role in helping primary and secondary CD8 T-cell responses.^117,118 Again, the enhanced helper capability of memory CD4 T cells in these models is presumably directly linked to their enhanced functional attributes and their ability to provide help sooner and more effectively than naïve CD4 T cells.

Secondary CD4 T-cell effectors

It is also understood that memory CD4 T cells serve as precursors to a population of activated effectors during secondary responses. Although the emergence of secondary effectors at later stages of responses may represent an important aspect of the continuing impact of memory CD4 T cells, the consequences of generating secondary CD4 T-cell effectors have not been investigated. This is in part because of the fact that primary and secondary effectors appear identical phenotypically,¹¹⁹ causing their discrimination to be difficult if not impossible. However, it was shown, via an adoptive transfer system, that secondary effector CD4 T cells were superior to primary effectors as well as to resting memory cells in terms of IL-4 production and providing help for immunoglobulin G (IgG) antibody responses.¹² Thus, while memory cells provide an increased level of help as compared with naïve cells, their helping capacity can increase during the kinetic development of recall responses.

We have recently investigated further functional attributes of secondary effectors utilizing an influenza infection model. Our laboratory has previously described several aspects of the responses of both naïve and of highly activated effector CD4 T cells responding to influenza virus.^78,120 Our new studies show that resting memory CD4 T cells give rise to a population of secondary effectors with similar kinetics to those of primary effectors arising from naïve precursors and that the two populations express similar levels of multiple adhesion and activation markers. Functionally, however, secondary effectors are superior to primary effectors and exhibit a pronounced multifunctional phenotype both at the per-cell and at the population level, producing higher concentrations of more cytokines.

We tested the impact of secondary CD4 T-cell effectors on survival following lethal influenza challenge by adoptively transferring memory populations into host mice deficient in CD8 T cells or B cells. These transfers allowed a focus on helper-independent functions of memory CD4 T cells. We found that secondary effectors could protect otherwise naïve mice devoid of CD8 T cells or B cells from lethal challenges of influenza via several distinct mechanisms while primary effectors, either generated in vivo from naïve precursors or transferred directly, could not. These results highlight the substantial helper-independent protective impact of memory CD4 T cells occurring at later stages of recall responses after their re-expansion and transition into a population of secondary effector cells.

Concluding remarks

A more complete understanding of memory CD4 T-cell immunobiology has the potential to significantly impact diverse fields, including vaccine design. Although many central questions about how memory CD4 T cells are generated and even how they protect remain unanswered, it is clear that memory CD4 T cells have the potential to impact secondary immune responses in unique and important ways (Fig. 4). A complicating factor in the study of CD4 T-cell memory is potential. Depending on the model system employed or on variables present during priming, the number and functional quality of resulting memory cells can vary greatly. Furthermore, the importance of memory CD4 T cells in orchestrating immune responses can vary depending on the pathogen challenge, and on the quality of memory CD8 T cells and B cells. Thus, the continued integration of observations from a variety of experimental models will be required to further understand and utilize the full potential of memory CD4 T cells.

Figure 4.

Both through cell-to-cell contact and through cytokine and chemokine production, memory CD4 T cells have the potential to impact a broad range of both innate and adaptive immune functions during a secondary response. DC, dendritic cell.

Disclosures

The authors have no competing financial interests nor conflicts of interest to disclose.