Defining immune memory resilience: implications for vaccine development

Figure 1. Concept of immune memory resilience and its importance for protection against pathogens with immunomodulatory properties.

The ability of a vaccine to induce primary immune responses has relatively little effect on its effectiveness as the primary immune response will have largely abated by the time the pathogen infects the host. This is illustrated by the symmetrical nature of the schematic above and below the dashed line. In the absence of immune memory there is no protection irrespective of the type of pathogen. In the presence of immune memory there is always protection when the pathogen has little immunomodulatory ability. By contrast, protection against highly immunomodulatory pathogens is dependent on the presence of resilient immune memory responses.

The generation of antigen (Ag)-specific immune memory that can be restimulated upon subsequent encounter of the same Ag is one of the hallmarks of adaptive immunity. Vaccines are generally designed to induce such immune memory, which is recalled during infection, resulting in a rapid and augmented immune restimulation that protects the host. Hence, the effectiveness of the immune memory induced, rather than the magnitude of the primary immune response, is of utmost importance for the designs of effective vaccines.

Defining immune memory resilience

Immune responses are classically characterized by their magnitude (i.e., frequency of Ag-specific cells), quality (i.e., phenotype and effector function of immune cells), breadth (i.e., number of epitopes recognized) and avidity (i.e., ability of receptors to bind Ag). The quality of the immune memory response is often overlooked, despite its importance for protection.

There is a rapidly growing body of evidence suggesting that pathogens have developed ways of manipulating the immune system. Viruses including Leporipoxvirus, poxviruses, Epstein–Barr virus, human herpes virus-8, cytomegalovirus and vaccinia have all acquired molecules homologous with a cytokine or cytokine receptor that have the ability to manipulate the immune system. For example, orf virus has developed ways of inducing immune responses that are detrimental to protection by manipulating the balance between Th1- and Th2-like immunity [1]. More complex pathogens such as Chlamydia psittaci are able to manipulate infected macrophages into producing a range of cytokines [2]. These pathogens take advantage of the fact that cytokines produced by Th1, such as IFN-γ, inhibit the induction of cytokines, such as IL-4 and IL-10, made by Th2 cells and vise versa (reviewed in [3]). Other immunomodulatory mechanisms used by pathogens include cytokine agonists or antagonists [4] and immunomodulatory peptides [5], or regulatory T cells (Tregs) [6] able to suppress immune responses in an Ag-specific manner [7].

When considering vaccines against this growing list of immunomodulatory pathogens, the ability of the immune memory induced following vaccination to withstand immunomodulation during subsequent in vivo restimulation by the pathogen becomes critical (Figure 1). We will refer to this new characteristic as the ‘resilience’ of the immune memory.

Role of adjuvants & vaccination strategy in inducing resilient immune memory responses

In the initial phase of immune induction, Ags and adjuvants, which are transported to the draining lymph node via afferent lymphatics and naive T cells, are activated by professional Ag-presenting cells [8]. While the Ag is essential for specificity, the adjuvant largely determines the cytokine environment in which this interaction takes place [9] and hence the type of immune response induced. As a result, different adjuvants such as Alum are known to induce predominantly T helper (Th)2-type immune responses (characterized by the secretion of type-2 cytokines such as IL-4 and IL-10), while Quil A-based adjuvants or killed Corynebacterium parvum, for example, induces predominantly Th1-type immune responses (characterized by production of IFN-γ). Other parameters important for determining the type of immune response induced include: the nature and amount of Ag delivered, the route of delivery, and the genetic background of the vaccinated host. Combinations of these factors results in the induction of immune responses over a wide spectrum, ranging from predominantly Th2- to Th1-dominated responses.

The relationship between primary and memory immune responses is often complex and the focus of vaccine development should be on the induction of immune memory responses rather than the primary immune response which, in most cases, will have abated by the time the pathogen strikes (Figure 1). In some instances, for example following DNA vaccination, it is possible to induce immune memory responses even if the primary response is relatively weak [10]. Nevertheless, memory T cells are derived from the effector clones, and during early immune induction naive T cells can be ‘epigenetically imprinted’ as Th1 or Th2 cells [11]. Thus, the way the immune response is induced can affect the level of immune memory resilience. For example, T cells from mice exposed to ovalbumin (OVA) adjuvanted with Alum did not produce IFN-γ while these same mice exposed to OVA during the course of an infection with either Listeria monocytogenes or influenza virus expressing recombinant OVA did produce IFN-γ [12]. The immune memory generated by Alum–OVA could subsequently be differentiated into Th1 or Th2 cells suggesting that the immune memory response in this case is not resilient. By contrast, mice immunized through infection with OVA-expressing organisms could only be restimulated to become Th1 cells secreting IFN-γ [13]. Thus, immune memory responses induced as a result of active infection with either Listeria monocytogenes or influenza virus seem to be more resilient compared with immune memory responses induced by Alum [12]. In humans there is also circumstantial evidence that vaccination and infection do not lead to the induction of the same type of immune memory [12]. Hence, depending on the context in which the Ag is encountered there seems to be differences in the resilience of the induced immune memory response.

Immune memory resilience & types of memory T cells

Two subsets of memory Th cells have been described based on a combination of cell surface marker and function (reviewed in [14]). These types of memory T cells also differ in their ability to retain the imprinted cytokine phenotype upon subsequent restimulation. T-effector memory cells (T_EM) are characterized by an ability to rapidly produce cytokines and other effector mechanisms, but are relatively short-lived. These cells typically express markers associated with immune memory in general (i.e., CD45RO and CD62L) but do not express the chemokine receptor CCR7. By contrast, central memory T cells (T_CM), while also expressing markers typical for memory T cells (CD45RO), do express CCR7. These T_CM are long-lived, capable of self-renewal and of feeding the pool of T_EM, and can therefore be thought of as the classical memory T cells [15]. They have to further differentiate to T_EM cells before being effective. By virtue of their expression of CCR7, the T_CM cells are able to migrate to the lymph nodes [16] and hence are recirculating in search of MHC-associated Ag in the peripheral lymph nodes.

To date, it was widely assumed that the imprinting is quite absolute and that memory T cells, particularly T_EM cells, are committed to be either Th1 or Th2 memory cells, while T_CM are somewhat more flexible. By virtue of their longevity and the fact that they contribute to the pool of T_EM cells, T_CM cells are most interesting in the context of vaccines. Indeed, over time, it can be expected that the T_CM cells will be determining the recall response, and hence the degree of resilience of these T_CM cells is expected to be critical to vaccine development.

Immune memory resilience: from concept to practical application

The as yet largely unexplored possibility that different adjuvants, routes of immunizations, genetic background of the host and/or infective status could result in immune memory responses that vary in their level of resilience, has important implications for vaccine design. To which degree and for how long this imprinting of the immune memory can be retained in the face of manipulation by subsequent infection is unknown. It is also unclear at what time point in the differentiation of memory cells they acquire the immune bias and at what stage this bias becomes irreversible or even if it is irreversible at all. Without this information the shortcomings of a vaccination strategy could become apparent only during infective challenge, when the pathogen has the ability to manipulate the immune memory response and hence affect the disease outcome following vaccination (Figure 1).

The resilience of the immune memory induced during vaccination needs to be considered from the perspective of the pathogen that will trigger the recall response. As such, the best way to assess immune memory resilience is by infective challenge. However, this may not always be practical and surrogate measures for immune memory resilience will have to be developed. These could include using different adjuvants during the priming and boosting to modulate the immune memory in opposite directions and assess the strength of the stimulus required for reversing the established bias. It is also possible that the immune memory resilience fluctuates over time, including as a result of fluctuating ratios of T_EM and T_CM cells. Thus, immune memory resilience is likely to be a dynamic process susceptible to external factors, necessitating careful analysis at different time points following vaccination.

In conclusion, insight into the parameters that determine the resilience of the immune memory response would have paradigm shifting implications for the design of novel adjuvants and vaccination strategies. Indeed, once established, vaccine developers could rely on these parameters for the development of adjuvants that induce highly resilient immune memory responses instead of finding out during clinical trials that the immune memory responses induced by their vaccination strategy is being subverted by the pathogen.

Financial & competing interests disclosure

This work was supported by the Australian Research Council. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.