Autoimmune effector memory T cells: the bad and the good

Abstract

Immunological memory is a hallmark of adaptive immunity, a defense mechanism endowed to vertebrates during evolution. However, an autoimmune pathogenic role of memory lymphocytes is also emerging with accumulating evidence, despite reasonable skepticism on their existence in a chronic setting of autoimmune damage. It is conceivable that autoimmune memory would be particularly harmful since memory cells would constantly “remember” and attack the body's healthy tissues. It is even more detrimental given the resistance of memory T cells to immunomodulatory therapies. In this review, we focus on self-antigen-reactive CD4⁺ effector memory T (T_EM) cells, surveying the evidence for the role of the T_EM compartment in autoimmune pathogenesis. We will also discuss the role of T_EM cells in chronic and acute infectious disease settings and how they compare to their counterparts in autoimmune diseases. With their long-lasting potency, the autoimmune T_EM cells could also play a critical role in anti-tumor immunity, which may be largely based on their reactivity to self-antigens. Therefore, although autoimmune T_EM cells are “bad” due to their role in relentless perpetration of tissue damage in autoimmune disease settings, they are unlikely a by-product of industrial development along the modern surge of autoimmune disease prevalence. Rather, they may be a product of evolution for their “good” in clearing damaged host cells in chronic infections and malignant cells in cancer settings.

Introduction

Autoimmune diseases, wherein the body's immune system attacks self-tissues, collectively afflict 5–10 % of the world's developed population. The incidence of many autoimmune diseases, including type 1 diabetes (T1D) [1], system lupus erythematosus (SLE) [2] and multiple sclerosis (MS) [3] have been increasing over the past decade and is estimated to increase further in the coming years [4]. The exact cause of this surge remains unclear, but environmental changes associated with industrialization have long been suspected. Those changes include sanitization from parasitic and microbial agents, as formulated in the hygiene hypothesis [5], and industrial pollutants that may alter the differentiation of immune cells [6–8]. Despite advances in treating autoimmune diseases, many of them involve general immunosuppression, leading to adverse side effects. A better understanding of the immunological causes of autoimmune diseases is needed for developing therapies that specifically target the pathogenic immuno-logical subsets responsible for the autoimmune attack.

Over the past decade, it has come to light that immunological memory can exist in the context of autoimmunity as well. It represents a “constant-remembrance” of self-antigen that may account for the persistence of autoimmune attack. Combating autoimmune memory has been a challenge not just in autoimmune diseases but also in transplantation, where the autoimmune memory cells attack the transplanted tissue. This review gives a brief overview of the different subsets of memory T cells and discusses in detail the effector memory T (T_EM) cell subset that is emerging as a major contributor to autoimmune pathogenicity. We will also explore the prospect that autoimmune memory responses, while pathogenic in autoimmune diseases, could be put to good use in anti-tumor responses (Fig. 1).

Fig. 1.

Predominant subsets of memory T cells in infections, tumors and autoimmune diseases. (1) Persistence of antigens in chronic infections, tumor microenvironment and autoimmune disease settings leads to the increased formation of the effector memory T (T_EM) cell subset. (2) T_EM cells are exhausted in settings of chronic infections and tumors. (3) In autoimmune diseases, T_EM cells are not exhausted, making them highly pathogenic. Their longevity and active functionality perpetuate autoimmune damage. (4) Persistence of antigens in chronic infections, tumors and autoimmune disease settings diminishes the formation of the central memory T (T_CM) cells and possibly the tissue-resident memory T (T_RM) cell subsets. (5) Immune control genes such as CTLA4 are highly polymorphic, and the polymorphisms that may promote T_EM cells are well preserved in human populations despite their deleterious potential in causing autoimmune diseases. These genetic variations suggest that the differentiation of autoimmune T_EM cells may have evolved for their beneficial potential in clearing unhealthy cells in chronic infections or boosting anti-tumor immunity. Further studies are needed to bridge the genetic discoveries to immunobiology and pathophysiology

T lymphocytes in autoimmune diseases

The involvement of the adaptive immune system in auto-immune diseases has been extensively characterized. T cells are critical contributors to autoimmune diseases, including T regulatory (T_reg) cells that inhibit disease development and conventional T (T_conv) cell subsets that play a role in B cell activation and differentiation, produce various inflammatory cytokines and destroy target cells with direct cytotoxicity. An important characteristic of the adaptive immune response is the formation of immunological memory after initial antigen exposure that helps the immune system learn with experience. Naïve T_conv cells, upon antigen exposure, can differentiate into effector T (T_EFF) cells, T_EM cells, tissue-resident memory T (T_RM) cells and central memory T (T_CM) cells. The T cell clones carrying the T cell receptor (TCRs) that recognize antigens most effectively are preserved in the form of long-lived memory T cells. On secondary antigen exposure, these expanded clones help mount a quicker and stronger immune response against invading pathogens. However, immunological memory is a double-edged sword. In an autoimmune response, when memory cells are formed against the “self,” they help mount a highly efficient pathogenic response against the body's own tissues. These memory cells, by virtue of being long lived also become very difficult to eliminate. It is believed that these memory cells play a critical role in making the autoimmune response persistent.

A case of systemic autoimmune disease: systemic lupus erythematosus

System lupus erythematosus is an autoimmune disease characterized by immune tolerance breakdown in both T and B lymphocyte compartments. Extensive studies have implicated both the innate and adaptive immune branches in SLE pathogenesis [9–13], but mechanisms that drive sustained systemic autoimmune damage remain unclear. Decades of research has unequivocally established a major role of B lymphocytes in SLE pathogenesis, culminating in the recent FDA approval of SLE treatment by neutralizing BlyS (BAFF), a B cell survival cytokine [12, 14]. CD4⁺ T cells are believed to play an important role in helping the activation of autoreactive B cells and their further differentiation into antibody-secreting cells [15–17] which produce an array of self-antigen-specific antibodies, in particular those against cell nuclear materials. Those antibodies cause multi-organ damage in advanced stages of SLE [9–12]. Dysfunction in a special subset of T helper cells, T follicular helper (T_FH) cells, which play an important role in B cell differentiation and maturation in the germinal centers, has also been implicated in SLE pathogenesis in both mouse models and human patients [18–21]. Whether T cells continue to play a pathogenic role in late stage SLE remains to be clarified, an issue carrying substantial relevance in SLE treatment. There is evidence suggesting the involvement of CD4⁺ T_EM cells in SLE pathogenesis [22]. Understanding the role of CD4⁺ T_EM cells in advanced SLE could lead to new therapeutic targets.

A case of organ-specific autoimmune disease: type 1 diabetes

Type 1 diabetes is amongst the best studied T cell-driven autoimmune diseases where pathogenic T cells target the insulin-producing beta cells of pancreatic islets. The MHC class II haplotype is the major genetic contributor to T1D susceptibility [23–26] which indicates the importance of the CD4⁺ T cell compartment in the autoimmune response in T1D. Studies involving MHC class I transgenic mice and CD8-restricted TCR transgenic models have also shown the involvement of CD8⁺ T cells in the response [27, 28]. CD4⁺ T_reg cells help control the autoimmune pathogenic T cell responses by regulating antigen presentation, direct cell–cell inhibition of T effectors and by production of anti-inflammatory mediators like IL10 and TGF-β. Their dys-function has been observed as a potential contributor to T1D pathogenesis, in experimental mouse models of T1D and in human patients with T1D [29–33]. Recent studies have also shown that CD4⁺ T_reg cells can further be divided functionally into effector and memory subsets [34], showing that regulatory memory T cells in the tissue are more potent at suppressing the autoimmune response when compared to regulatory T_EFF cells.

A number of studies suggest that the autoimmune CD4⁺ memory T cells are pathogenic in T1D [35–37], although the exact contribution of each subset remains to be further studied. These cells have the potential to secrete cytokines like IFN-γ and IL17 which have been shown to promote T1D pathogenesis [38]. Autoimmune memory T cells persist even after complete destruction of pancreatic islets of the patient. They may also be the major cause of transplant rejection in T1D patients with islet transplants [39, 40].

T cell memory

Upon antigen exposure, Naïve T cells undergo clonal expansion to become activated T_EFF cells. Most of T_EFF cells migrate to the site of infection where they tackle the invading pathogens. Once the infection is cleared, the majority of the T_EFF cells generated by clonal expansion die. Along the process, some of the antigen-activated cells become long-lived memory T cells. Whether and to what extent the memory compartment is derived from T_EFF cells remain debated.

Memory T cells have been classically delineated into two subsets: T_CM and T_EM cells. T_CM cells are considered to be long-lived memory T cells with greater proliferation potential and are predominant in the secondary lymphoid organs. Compared to T_CM cells, T_EM cells are relatively short-lived, have lesser proliferation potential but possess a range of effector functions and are predominant in the target tissues. They are in a state of readiness to respond to antigen re-challenge at the tissue much faster than T_CM cells. In typical settings of acute infectious diseases, the T_EM compartment declines with time after antigen exposure, but a stable quiescent population of T_CM can give rise to a secondary immune effector response even decades after initial antigen exposure [41]. On the other hand, in autoimmune disease settings, as in chronic infections, the constant presence of antigen at the tissue may preclude the development of traditional memory response, particularly the formation of T_CM compartment [42, 43].

While they are a long-lived subset, quiescent T_CM cells need to be re-activated on antigen encounter, in order to produce a T_EFF population and migrate to the target tissue to execute their functions. An expedited memory response could be mounted if memory cells are present in lymph nodes that drain the target tissues, or within the target tissue itself, with an effector arsenal ready to function. Such subsets of memory cells indeed exist in the form of T_EM and T_RM cells.

A well-justified doubt on existence of memory T cells in autoimmune settings: lack of T_RM and T_CM cells?

T_RM cells are a new subset of memory T cells, recently identified as the population of T cells that permanently resides in the tissue even after the infection is cleared [44– 47]. These memory cells have poor proliferation potential and survival when compared to T_CM and T_EM cells and appear to be terminally differentiated [48]. Besides these differences, the functional difference between T_EM and T_RM cells at the tissue is unclear. The basic distinction seems to be that while the T_EM cells can recirculate between the spleen, blood and tissues, the T_RM cells can migrate only within the tissue [44–47]. CD69 and CD103 are the phenotypic markers that define T_RM cells. These molecules may also be involved in the formation/persistence of the T_RM compartment. Hence, the three different memory subsets are phenotypically defined as—T_RM: CD44^hiCD62L⁻CD69⁺CD103⁺, T_CM: CD44^hiCD62L⁺ CD69⁻CD103⁻CD127⁺, and T_EM: CD44^hiCD62L⁻CD69-CD103⁻CD127⁺ [48]. T_EFF cells generated during the response at the tissue phenotypically (CD44^hiCD62L CD69⁺CD103⁺CD127⁻) resemble T_RM cells, but it is assumed that antigen-specific cells present in the tissue long after antigen clearance belong to the T_RM compartment since T_EFF cells are short-lived cells.

One of the first studies identifying CD8⁺ T_RM cells in a HSV infection model showed that the CD8⁺ T_RM antigen-specific cells persisted for more than 100 days after primary infection in the absence of antigen. This T_RM subset formed the first line of defense against reinfection and provided site-specific immunity at the particular tissue up to 100 fold better than in a site without previously generated T_RM cells [45]. Similar results were obtained in a CD4⁺ T cell response, where CD4⁺ T_RM cells were far superior in terms of protection from influenza reinfection [47]. Thus, protection mediated by both CD4⁺ and CD8⁺ T_RM subsets has been shown in various models of infection in various tissues [44–49]. The T_RM subset has also been identified in humans [50]. Though there is strong evidence for the existence of T_RM cells in many different tissues in different settings of infection, their existence in the context of autoimmune diseases has yet to be clearly demonstrated. Akin to how the constant presence of antigen may preclude the development of T_CM cells, constant self-antigen exposure may preclude the formation of T_RM cells as well [51].

T_EM defined in infectious disease settings

T_EM cells are relatively long-lived and have higher proliferation potential when compared to T_RM cells. They can undergo self-renewal using IL7 and IL15 with their CD127 and CD122 receptors like T_CM cells [52] and also possess immediate effector functionality like the T_EFF subset. They circulate between the spleen, blood and peripheral tissue and are thus able to mediate the first line of defense against pathogens at the site of tissue invasion. It has been shown that CD4⁺ T_EM cells that persist in lung tissues and airways after a respiratory virus infection can substantially protect against secondary virus infections [53]. In a model of intranasal parainfluenza virus infection, it was shown that the CD8⁺ T_EM cells played a more prominent role than the CD8⁺ T_CM cells in recall responses in the lung [54]. Another subset that forms a part of the CD8⁺ T_EM compartment, the CD8⁺CD44⁺CD27^loCD43^lo subset, localizes to the tissue compartment. This compartment has been shown to have immediate perforin-dependent effector functionality and is thus able to mediate superior protection against bacterial and viral infections [55]. Thus, even in acute infections where a functional T_CM subset is formed, the T_EM subset plays an important role at the front line of immune defense until the T_CM subset is able to generate a second wave of effectors.

Studies from chronic viral infections like CMV infections have shown that persistence of high levels of antigen favors the development of the T_EM cells [42, 56]. These chronic infections are characterized by a life-long persistence of a high frequency of CD4⁺ and CD8⁺ T_EM cells that prevent disease progression though they cannot completely eliminate the pathogen [56, 57]. This inability to eliminate pathogen has been attributed to the exhaustion of the T_EM compartment due to the persistence of antigen in chronic infections, which has lower functionality when compared to T_EM cells from acute infections [58]. It has been shown that CD4⁺ and CD8⁺ T_EM responses help combat HIV so much so that current vaccine strategies aim to specifically increase the HIV-epitope-specific CD4⁺ T_EM and CD8⁺ T_EM cells. This vaccine strategy that elicited continuous T_EM immune surveillance has been effective in clearing highly pathogenic SIV infection in 50 % of rhesus macaques [59, 60].

Thus, the T_EM population presents itself as a highly effective, antigen-specific memory subset capable of eliciting effector functions on pathogen encounter at the tissue in both chronic and acute infections. The T_EM subset represents the balance between the long-lived quiescent T_CM subset, which require re-activation despite their superior proliferation potential, and the T_EFF subset which posses potent effector function but are short-lived and have poor proliferation potential. These T_EM cells form the front line of defense against many invading pathogens at the tissue and thus often determine the final outcome of the immune response.

T_EM as the main subset of memory T cells in autoimmune conditions

T cell-mediated pathogenicity in autoimmune diseases is most likely brought about by both T_EFF and T_EM subsets because the persistence of auto-antigen precludes the formation of autoimmune T_CM and T_RM cells. Between the T_EFF and T_EM cells, T_EM cells likely drive the persistence of autoimmune diseases because of their ready effector functionality and relative longevity. As shown in studies of chronic infections, persistent antigen increases the pool of T_EM cells, which would be the case in autoimmune disease settings as well with persistence of self-antigens.

Recent evidence from gene expression studies suggests substantial contribution of the CD4⁺ T_EM compartment to autoimmune disease pathogenesis. Genome-wide association studies (GWAS) have implicated many candidate immune genes in autoimmune pathogenesis, including CTLA4, BACH2 and PD-1. However, it remains unknown as to which specific cell compartment is affected by the genetic variations. The identified gene may have functions in many different cell types, making it hard to associate the genetic polymorphism identified from the GWAS with a mechanism of disease pathogenesis. In this regard, genome-wide gene expression analyses of distinct cell subsets, like the immunological genome project [61], could offer helpful insights. In particular, studies can be conducted to link the pool of disease-susceptible gene polymorphisms identified with profiles of genes expressed in distinct cell types. One such recent study analyzing gene expression data from pathogenic cell types in auto-immune diseases has been able to show the enrichment of CD4⁺ T_EM-cell-associated genes within SLE loci, Crohn's loci and rheumatoid arthritis (RA) loci [62]. In another study, RA susceptibility loci identified by high-density genetic mapping contained genes that were most significantly expressed in CD4⁺ T_EM cells [63]. When such bioinformatics approaches using large datasets from large populations involving genes expressed in a broad range of cell types converge on a single subset, CD4⁺ T_EM cells, the evidence lends further support to the hypothesis that CD4⁺ T_EM cells play a crucial role in autoimmune disease pathogenesis.

Studies on the involvement of memory cells in autoimmunity have been hindered by technical difficulties in identifying the actual autoimmune memory population. In many infectious disease studies, the memory cells are not necessarily phenotypically defined because their presence long after antigen clearance is sufficient to classify them as memory T cells. After antigen clearance, the CD44^hiCD62L^low subset is phenotypically defined as the T_EM subset, because the T_EFF cells that are also CD44^hiCD62L^low are assumed to be short-lived. This method of identifying the T_EM cells after antigen clearance is convincing in context of acute infectious diseases. However, in context of persisting self-antigen in autoimmune diseases, the CD44^hiCD62L^low subset will include a signifi-cant number of short-lived T_EFF cells as well. To resolve the two populations, additional markers like CD69 and CD127 are required.

Evidence gathered from experimental studies in animal models and ex vivo using peripheral blood samples from patients, especially in the past few years, indicates that CD4⁺ T_EM subset is emerging as an important contributor to many T cell-mediated autoimmune diseases. For example, in the experimental autoimmune encephalomyelitis (EAE) model of MS, adoptive transfer experiments showed that autoimmune memory was maintained by T_EM cells with intact cytokine production and tissue damage potentials [64]. Another study showed that in autoimmune diabetes, unstable T_reg cells converted to CD4⁺ T_EM cells that were highly pathogenic with disease-causing potential [35]. An increased population of CD4⁺ T_EM cells was found in human patients with SLE, even in a disease that is thought to be primarily B cell mediated [22]. The anti-neutrophil cytoplasmic autoantibody associated systemic vasculitis (AAV) disorders have been thought to be caused by autoantibodies against neutrophil proteins. These disorders are characterized by autoimmune damage of blood vessels that leads to vessel occlusion and systemic organ damage. There is increasing evidence that the immuno-pathogenesis in AAV disorders is mediated by CD4⁺ T_EM cells [65]. It was also shown that there was an increase in CD4⁺ and CD8⁺ T_EM subsets in patients with aplastic anemia [66]. Thus, a large body of evidence from studies encompassing many autoimmune diseases, including T1D, SLE, EAE to AAV and aplastic anemia, strongly suggests that CD4⁺ T_EM cells are a crucial mediators of autoimmune destruction.

Unlike in the chronic infection settings wherein T_EM cells are often exhausted and have reduced levels of cytokine production [58, 67], in context of autoimmune diseases, T_EM cells are potent producers of cytokines despite persistence of self-antigens. Human CD4⁺ T_EM cells from patients with RA have been shown to produce IFN-γ [68]. In EAE, it has also been shown that autoimmune CD4⁺ T_EM cells produced IFN-γ [64]. Whereas in an acute infectious disease setting, it was shown that long-lived Th17 memory did not develop although the primary effectors had no defect in IL17 production [69], in the EAE autoimmune model it was shown that Th17 autoimmune memory cells were more potent at transferring disease than their non-Th17 counterparts [70]. In autoimmune diabetes settings, T_EM cells were shown to produce pathogenic cytokines IL17 and IFN-γ [35]. Thus, in autoimmune disease settings, not only is there an increase in the autoimmune CD4⁺ T_EM compartment correlating with disease pathogenesis, but also these CD4⁺ T_EM cells have the potential to produce likely pathogenic cytokines such as IL17 and IFN-γ. In other words, unlike other settings of persistent antigen such as in chronic infections, the auto-immune T_EM cells are not exhausted in their effector functionality in autoimmune disorders.

Targeting undesirable memories

Autoimmune memory cells have been thought to be a major contributor to the resistance of autoimmune diseases to many immunomodulatory therapies despite that substantial advances have been made in curtailing autoimmune damages. For example, blocking the CD28/B7 co-stimulation pathway [71] or the CD154/CD40 [72] pathway was shown to effectively control autoimmune diseases like EAE [73, 74], T1D [75–78], psoriasis [79] and SLE [80, 81]. However, memory cells are less dependent on co-stimulation than Naïve T cells [82–84], which might account for the limited success of such therapies. T_reg infusion therapies have great promise in controlling adverse immune responses [85]. However, some studies have demonstrated that memory cells are also relatively more resistant than both the Naïve and effector compartments to T_reg suppression therapies [86, 87].

With the memory compartment being resistant to many of the immunosuppression therapies and with the CD4⁺ T_EM compartment emerging as a crucial contributor to autoimmune diseases, there is a need for therapies that target this specific compartment. Two possibilities along these lines have been shown in studies of the potassium channel Kv1.3 and the TNF family receptor Fas/CD95. The Fas death inducing signaling complex is more efficiently formed and enriched in the lipid raft microdomains in CD4⁺ T_EM cells. This makes these cells specifically susceptible to Fas-induced cell death while the CD4⁺ T_CM and CD4⁺ Naïve T cells remain unaffected [88]. Kv1.3 is a calcium-activated potassium-gated ion channel expressed by the KCNN4 gene [89]. Pathogenic CD4⁺ T_EM cells have been shown to express much higher levels of Kv1.3 channels compared to the Naïve and central memory subsets. Kv1.3 inhibitors have been shown to ameliorate autoimmune disease in models of T1D, RA and EAE, likely by a specific effect against the pathogenic CD4⁺ T_EM subset [89–91].

Targeting autoimmune T_EM cells could benefit from understanding the molecular basis of their formation and function. Twist1 is highly upregulated in CD4⁺ T_EM cells isolated from patients with RA, Crohn's disease and ulcerative colitis. Twist1 is an endogenous regulator of the Th1 compartment. Its expression in response to repeated antigen exposure may lead to CD4⁺ T_EM cell differentiation [92]. It was suggested that Twist1-expressing Th1 cells in patients undergoing immunosuppressive treatment belonged to a refractory CD4⁺ T_EM compartment [93]. Genetically predisposed reduction in CTLA4 splice variant expression in human subjects has been correlated with T1D susceptibility [94]. Genetic polymorphisms in Bach2 have been associated with numerous autoimmune diseases including T1D, MS and Crohn's disease. Bach2 knock-out mice develop autoimmune disease [95]. One of the mechanisms by which Bach2 is thought to prevent auto-immune disease is by regulating the generation of the pathogenic effector memory compartment [96]. Elucidating such molecular pathways of the pathogenic CD4⁺ T_EM compartment will enable the development of new strategies to modulate the function of this compartment without general immunosuppression.

Putting autoimmune effector memory to good use: the anti-tumor response

It is often said that autoimmunity and anti-tumor immunity are the opposite sides of the same “coin.” In essence, however, we suggest that they are actually on the same side of the “coin,” because reactivity to self-antigens underpins both types of immune responses [97, 98]. Our recent studies have shown that autoimmune responses can indeed be potent mediators of anti-tumor immunity [97]. Considering that the CD4⁺ T_EM subset is emerging as a driving force for autoimmune disease pathogenesis, why is an individual not able to mount a potent anti-tumor response mediated by autoimmune T_EM cells, despite there being persistent self-antigen like in autoimmune diseases? The answer could lie in a parallel with chronic infection settings where persistent antigen causes exhaustion of the T_EM compartment [58] (Fig. 1).

One of the major molecules contributing to exhaustion in chronic infections has been found to be PD-1 [99]. In mice with chronic LCMV infection, treatment with antibodies against PD-L1 (the ligand for PD-1) enhanced clearance of virus [99]. This suggest that PD-L1 blockade somehow restores previously exhausted T_EM cells to help clear virus infection. In concordance with this argument, a recent study showed that PD-1 likely prevented the formation of the T_EM compartment [100]. Studies have also shown that PD-L1 is expressed by most types of tumor cells, and blockade of this leads to a more potent anti-tumor response [101–103]. However, PD-L1 is expressed by many tissues of the body including non-parenchymal liver cells, lung, cornea, vascular endothelium, pancreatic islets and keratinocytes. The PD-L1/PD-1 interaction is a major factor in controlling autoimmunity [104], which may explain why overall PD-L1 or PD-1 blockade could be accompanied by autoimmune toxicity in clinical trials of cancer immunotherapies [102, 103].

Conceivably, anti-tumor effect can be achieved by potentiating tumor-infiltrating autoimmune T_EM cells. Adoptive cell therapy has shown efficacy in this context, in studies where tumor-infiltrating lymphocytes from the patient re-stimulated and expanded were able to mount a potent anti-tumor response when transferred back into the patient [105]. In this regard, the tumor microenvironment could function as an immunoprivileged self [97, 98] and cause exhaustion of self-antigen-reactive T_EM cells that infiltrate tumors. However, if these T_EM cells could be isolated, re-potentiated and used for adoptive cell therapies against tumors, we suggest that it might elicit a better outcome.

Back to the Future: perspective from evolution of CTLA4 genetic variations to genomic medicine

Studies from various autoimmune diseases have implicated T_EM cells in autoimmune pathogenicity. The T_EM subset in such a condition is distinct from the memory counterparts generated in chronic infections, acute infections and tumor settings (Fig. 1). Further studies are needed to characterize this autoimmune pathogenic population in terms of phenotypic markers that define it and genes that regulate it. Equipped with this knowledge, it would be possible to develop strategies that specifically target this specific subset. The involvement of the pathogenic effector memory subset in various autoimmune diseases also presents a common target against many autoimmune diseases. In addition, advances in understanding this pathogenic subset may also propel the development of novel strategies using the autoimmune T_EM subset to combat tumors.

The genetic basis of T_EM differentiation remains to be elucidated. It is curious that heritable genetic polymorphisms, such as those in the CTLA4 gene which predispose an individual to autoimmune diseases, have stood the test of evolutionary selection pressure and been preserved in the human population. As a matter of fact, the frequency of a few autoimmune disease susceptibility alleles may even be increasing in the population [106, 107]. This suggests that the benefits of having an “overactive” immune system balance out the disadvantage of the individual's predisposition to autoimmune diseases. These benefits may include the ability to mount stronger anti-tumor and anti-pathogen responses. In concordance with this thought, recent studies have shown that parasite infections are a driving force for the positive/negative selection of inflammatory bowel disease (IBD)-associated loci [108, 109]. In context of anti-tumor immunity, genetic studies found that CTLA4 polymorphisms that predict reduced CTLA4 expression in mRNA and/or proteins [110–114] were associated with protection from lymphoma, breast cancer and skin cancer in humans [115–119]. Enhancement of anti-tumor immunity by CTLA4 blockade was shown to be associated with increased formation of memory T cells [120]. In our recent study [97], we tested the role of CTLA4 expression levels on autoimmunity-mediated anti-tumor immunity, by using a CTLA4 shRNA transgenic model that was constructed to mimic human CTLA4 genetic variations that predispose to T1D development [121]. Indeed, the modest reduction in CTLA4 overcame tumor-associated immunoprivilege in a lymphoma model and curtailed spontaneous development of breast cancer [97] in a model wherein the cancer development is attributed to immune tolerance [122]. Therefore, one might speculate that this potent CD4⁺ T_EM subset in various autoimmune diseases, as discussed previously, could represent an increasingly potent immune system that may be evolved to combat tumors and infections. Along these lines, in a situation wherein collateral damage to healthy tissue is acceptable, one might envision that autoimmune T_EM cells with their lasting potency, could be a valuable arsenal complementing the current momentum of adoptive T cell therapies against cancer [123, 124].

Undoubtedly, translating the advances from various model systems of T_EM differentiation and function to human disease settings in patients remains a daunting challenge, because of limitations in clinical feasibility and ethical considerations in studying the pathophysiology of human disease development in most cases. For example, during the development of T1D, the autoimmune damage in the pancreas typically remains undetected until most of the pancreatic β cells are destroyed, unless at-risk patients are actively monitored in research settings. Furthermore, peripheral blood is often the sole access for analyses of immune cell activity in human patients, and questions often arise concerning how much the activity and profiles of peripheral blood immune cells reflect immune damage in the pancreas. In this regard, it is worth noting that T_EM cells are thought to traffic between target tissue and systemic circulation [48], although such an analysis of one subset of cells from one site is still no better than a “blind man's” effort to approach a disease “elephant.”

One can envision that the function of T_EM cells is orchestrated by interactions with various types of innate and adaptive immune cells, through both yet-to-be identified and well-known molecular “bridges,” such as CTLA4-B7 [125]. Much akin to cellular networks, functionally related genes may also be organized in networks of coordinated expression and activity. In our study of gene expression profiles of peripheral blood samples from patients with autoimmune diabetes, intriguing patterns of innate and adaptive gene expression were identified in samples from patients at different stages of disease development [126, 127]. Of particular interest, CTLA4 gene expression differentially networked with a set of other adaptive and innate immune genes as the disease progressed from at-risk, to new-onset and to long-term diabetic stages [127]. It should be noted, however, that these studies were done with small subsets of selected innate and adaptive immune genes, with then-available technologies. The rapid development of genome technologies, such as next-generation sequencing and computational methods for data mining, enables the research community to access genome-wide studies of biology and pathophysiology of human diseases that have been beyond the reach of most biomedical researchers. In the case of T_EM cells, for example, one can expect the unfolding of its genome biology, in terms of both the genomic underpinning of its differentiation and genome-wide impact if altered. Akin to “blood work” checkups routinely used in clinics, such genomic datasets could potentially serve to provide not only specific indicators for a particular disease activity, but also could help gauge a patient's overall genomic wellness. This could lead to better clinical management both before and after frank disease development, in the emerging era of genomic medicine.

Biography

Zhibin Chen