T cells in the central nervous system: messengers of destruction or purveyors of protection?

Abstract

Although the destructive effects of an overactive adaptive immune system have been well established, especially in the context of autoimmune diseases, recently an understanding of the beneficial effects of the adaptive immunity in central nervous system (CNS) injuries has emerged. CD4⁺ T cells have been shown to benefit injured CNS tissue through various mechanisms; both traditional cytokine signalling and by modulating the phenotype of neural cells in the injury site. One of the major targets of the cytokine signalling in the CNS are myeloid cells, both resident microglia and monocytes, that infiltrate the tissue after injury and whose phenotype; protective or destructive, appears to be directly influenced by T cells. This cross-talk between the adaptive and innate immune systems presents potential new targets that could provide tangible benefits in pathologies that currently have few treatment options.

Introduction

Our understanding of the immune system, and the many components that make up the innate and adaptive arms of this complex system, is constantly evolving. For years, the assumption was that the mere presence of activated immune cells in the central nervous system (CNS) was a hallmark of ongoing pathology.¹ This view was particularly applicable in the injured CNS, where the presence of activated T lymphocytes was consistently associated with a poor prognosis and exacerbated neuronal loss.² However, recently this thinking has been challenged, and it is becoming evident that T cells play a critical role in resolving damage after CNS injury. Early works hinting at this demonstrated that the presence of activated immune cells improved recovery after nerve injury.^3–7 Since these early works, further evidence has begun to accumulate in support of a beneficial role for T cells in a wide range of normal brain functions, and that malfunctions in these homeostatic roles may contribute to neurodegenerative disorders,^8–10 and a range of developmental disorders formerly believed to be purely neurological in nature.^11,12 This review will highlight emerging data in support of a beneficial role for different T-cell subtypes in CNS acute and chronic neurodegeneration.

Neurodegeneration in the CNS

Injury in the CNS, as in the periphery, results in a cascade of cellular and molecular responses that amplify tissue damage beyond that expected from the severity of the initial injury itself.^13,14 This process, called secondary degeneration, can lead to severe neurodegeneration even when the initial insult may have only involved partial injury to the nerve or spinal cord.^9,15 Strikingly, however, the secondary degeneration is more extensive in animals lacking an adaptive immune system than in their wild-type counterparts, suggesting a previously unknown neuroprotective role for immune cells.^9,15 Restoration of the immune system, and particularly of the T-cell compartment, in immune-deficient mice restores their normal response to CNS injury,^16,17 further suggesting that an endogenous immune response to CNS injury is neuroprotective. Importantly, it was discovered that not all T cells can mediate this neuroprotective effect, but that the T cells need to be specific to brain-restricted antigens,⁴ which probably governs their migration to, and accumulation in, the injured CNS.^18,19 Hence, transfer of autoreactive T cells directed against the CNS antigen, myelin basic protein, reduced the secondary degeneration after nerve injury in rats, and this neuroprotection could be provided through both active immunization (via immunization with the myelin basic protein and adjuvant), or passive immunizations (through the transfer of pre-activated myelin basic protein-specific T cells) ^4,5,20 (Fig.1).

Figure 1.

Neuroprotective mode(s) of T-cell action in the injured central nervous system (CNS). T cells in the injured CNS are able to act both directly and indirectly to promote neuroprotection. They produce neuroprotective molecules such as brain-derived neurotrophic factor, which can directly promote neuronal survival. In addition, they are able to potentiate neuronal survival after CNS injury by influencing myeloid cell phenotype at the site of injury by promoting a protective M2 skew though production of interleukin-4 (IL-4) and IL-13. These M2 myeloid cells are then able to promote neuroprotection through production of neuroprotective molecules such as transforming growth factor β (TGF-β), insulin-like growth factor 1 (IGF1) and IL-10.

Regulatory T cells – regulators of the immune response to injury

These findings present something of a quandary because there is significant pathology associated with autoreactive T cells in the brain, in such conditions as multiple sclerosis and neuromyelitis optica.^21,22 These conditions, unlike the CNS injuries described above, are improved or eliminated by removal of autoreactive T cells. Furthermore, many of the treatments that are approved for use in autoimmune diseases have proven to be detrimental in CNS injury.²³ What then causes autoreactive T cells to be beneficial in injury conditions, yet be detrimental in these autoimmune diseases? A key player in controlling autoimmune responses that might hold the answer to this quandary was uncovered some 40 years ago with the discovery that a population of lymphocytes could control adaptive immune responses²⁴ and, some 20 years later, with our understanding of the molecular identity of these CD25⁺ regulatory T (Treg) cells.²⁵

These cells are marked by expression of the transcription factor FoxP3^26–28 and have been proposed as the key player in controlling autoimmune responses by the adaptive immune system. This subset of T cells acts not to increase the activity of the immune system to salient stimuli, but instead acts as an endogenous brake to ensure that adaptive immune responses are correctly measured in response to their stimulus. Therefore, because it is their role to suppress autoimmune responses, it stands to reason that Treg cells which are necessary for optimal neuro- protection in CNS injury would be detrimental to the autoreactive T-cell response. However, much controversy still exists about the exact role of Treg cells in CNS injury. For example, early work in models of stroke and Parkinson's disease, showed that depletion of Treg cells led to increased neurodegeneration, but increases in Treg cell numbers and function, improved disease outcome, in contrast to what might have been anticipated.^29,30 More recent work, using the same models, has shown that Treg cells play a detrimental role after CNS injuries,³¹ perhaps more in keeping with the idea that they are suppressing a beneficial autoimmune response. While researchers attempt to resolve these discrepancies, there are several factors that have contributed to the disparate findings. Treg cells from naive mice can exhibit plasticity, especially when placed in lymphopenic or inflammatory conditions,^32–34 probably because of heterogeneity in the fate commitment of the Treg cell population.³⁵ Furthermore, there are two methods commonly used to deplete Treg cells, and both have significant experimental limitations: anti-CD25 antibodies will deplete activated effector cells as well as regulatory T cells,³⁶ while DEREG mice (which express a diphtheria toxin receptor–green fluorescent protein fusion protein under the control of the FoxP3 promoter³⁷) are susceptible to non-specific effects of diphtheria toxin.^38,39 In addition to the experimental challenges that have arisen from working with Treg cells, it has become clear that there exists a balance between Treg cell and effector T-cell CD4⁺ CD25⁺ Foxp3⁻ activity, which makes it an over-simplification to say that Treg cells are only neuroprotective or only neurodestructive – too little Treg cell activity and the immune response cannot resolve properly, leading to autoimmune disease,²⁵ whereas too much Treg cell activity and the immune response is not sufficient to promote optimal survival.¹⁷ It seems that the immune system has evolved not to exert optimal recovery from CNS injuries, but rather to minimize the risk of autoimmune diseases. Hence, the regulation of the immune response to injury appears to be sub-optimal in physiological processes and is a viable therapeutic target given that the risk for autoimmunity can be mitigated.

Mechanistic insight into T-cell-mediated neuroprotection

This still leaves open the question of what functions of T cells mediate this neuroprotection in situations of CNS injury. There are several functions of T cells, both canonical (cytokine-producing) and non-canonical, that can contribute to their neuroprotective phenotype after CNS injury. The non-canonical effects of T cells that have been shown to contribute to the recovery from injury often rely on interactions with neural cells or the injured neurons themselves. For instance, T cells can protect neurons directly through production of neurotrophic factors, such as brain-derived neurotrophic factor (BDNF), when they become activated.^4,40 This immune-derived BDNF plays a functional role in the injured CNS, and mice that lack BDNF in their myeloid and T-cell populations exhibit deficits in functional outcomes after inflammatory injury.⁴¹ Additionally, glial cells are involved in the T-cell protective response, as they can signal to astrocytes to up-regulate the production of protective thiol compounds and increase their buffering of glutamate.^42,43 In addition to these cytokine-independent mechanisms for T-cell-mediated neuroprotection, several recent works have begun to unravel importance cytokine-dependent interactions between T cells and myeloid cells in the outcome of CNS injury. Monocytes that are recruited to the site of injury in a CCR2-dependent manner take on an alternatively activated phenotype as shown by expression of interleukin-10, whereas the resident microglia do not display this alternatively activated phenotype.⁴⁴ Additionally, the peak of alternative activation of myeloid cells after injury coincides with the peak of T-cell infiltration into the CNS, suggesting a beneficial T-cell effect on myeloid phenotype.^45,46 These alternatively activated macrophages are considered tissue building, and can produce trophic molecules, such as insulin-like growth factor 1 and transforming growth factor β, which are important in growth and development and promote recovery in injured CNS tissue.⁴⁷ Importantly, one of the best-studied routes to produce alternatively activated macrophages is through signalling by interleukin-4, a prototypical T helper type 2 (Th2) cytokine.⁴⁸ T cells, then, can infiltrate into the tissue and produce Th2 cytokines, contributing to the protective alternatively activated macrophage skew. This response is controlled by Treg cell suppression of the immune response; slight decreases in Treg cells allows for a more efficient Th2 response, which promotes increased alternative activation of macrophages, while complete depletion of Treg cells permits an uncontrolled immune response that decreases the alternative activation of macrophages at the site of injury (unpublished personal communication). Furthermore, mice deficient in T cells show both fewer infiltrating macrophages to the site of injury and greater pro-inflammatory skew of the cells that do enter. T cells are clearly complex cells that play a multi-faceted role in CNS injury that continues to be an active area of research.

Conclusions

Injury to the CNS spontaneously activates a neuroprotective T-cell response, marked by activation and infiltration of T cells into the injured tissue. The exact identity of the T cells that are mediating this response endogenously remains unknown, but auto-antigen-specific T cells have proven to be particularly potent at promoting neuroprotection. This neuroprotective role of effector T cells is regulated by Treg cells, though the varied physiological roles of Treg cells make studying their involvement in injury difficult. In fact, Treg cells could have different roles at the beginning compared with at the end of the immune response to injury. In the initial stages, Treg cell deactivation would allow autoimmune T cells to mediate their beneficial effects, but later the restoration of their suppressive activity would enable the autoimmune injury response to be controlled or terminated. The mechanisms underlying T-cell-mediated neuroprotection remain unclear, and are likely to be multifaceted, however, their interaction with myeloid cells has recently provided great insights. Clearly further studies are needed, but it is becoming evident that appropriate manipulations of this balance between effector and regulatory T cells may provide powerful new treatment options for an array of neurodegenerative and autoimmune pathologies.

Disclosures

The authors declare no conflict of interests.