Naïve CD8+ T lymphocytes are mature circulating cells that can acquire various effector functions depending on external clues (38). Their differentiation program is not pre-determined, but is shaped instead by conditions such as inflammatory states and ageing (39). In mice, the naïve phenotype is characterized by the expression of surface markers CD62L (L-selectin) and CD197 (CCR7), while in humans also includes the expression of CD45RA epitope (32, 39, 40) ( Figure 1 ). Upon activation, naïve CD8+ T cells lose the expression of homing receptors, initiating a proliferation and differentiation program resulting in an army of effector CD8+ T lymphocytes (31, 40, 41) ( Figure 1 ). In addition to the traditional method of identifying cells based on their classical surface markers, a new approach has emerged that focuses on their metabolic traits. Recent research has shown that naïve and memory T cells rely on oxidative phosphorylation and fatty acid oxidation, while effector T cells use aerobic glycolysis and amino acid metabolism to maintain their active state (42, 43).

In contrast to naïve CD8+ T cells, the fate of effector T lymphocytes is more clearly defined (44, 45). Indeed, these cells are specifically activated and directed toward pathogen-derived or tumor-derived peptides, (29, 46, 47). Effector CD8+ T lymphocytes classically express CD44 and CD69 surface molecules, allowing them to enter peripheral tissues (32, 48, 49) ( Figure 1 ). Moreover, they can be identified by the expression of CD95 (FasR) molecule, which contributes to the direct CD8-mediated cytotoxic process, and by the Ki-67 proliferation marker, which is important during clonal expansion (50–52). CD8+ T cells also possess a high cytotoxic potential by secreting various effector and cytotoxic molecules, including granzymes, perforins, IFN-γ, TNF-α and interleukin-2 (IL-2), which enable them to effectively combat infections (32, 33, 53, 54) ( Figure 1 ).

Following antigen clearance, a two-tiered contraction occurs in the CD8+ T cell population, defined by the expression of the killer cell lectin-like receptor G member 1 (KLRG1). Short-lived effector KLRG1+ CD127- CD8+ T cells undergo selective apoptosis, whereas effector KLRG1+ CD127+ CD8+ T lymphocytes are preserved and evolve into exKLRG1 long-lived memory T cells, providing immunological memory (37, 41, 44) ( Figure 1 ). However, a subset of effector CD8+ T cells, called memory precursors effector (T_MPE) cells, has been found to lack KLRG1 expression, and this was associated with enhanced survival during the contraction phase and a higher developmental plasticity (37). Accordingly, T_MPE cells retain the capability to differentiate in all the subsets of memory cells, playing a critical role in long-term protective immunity (37).

During the memory phase of the immune response, CD8+ T cells display immunological memory, which enhances their ability to rapidly and effectively respond to previously encountered pathogens, thus safeguarding the body against known threats (55, 56).

The memory compartment of CD8+ T lymphocytes comprises three different cell subsets: (i) T central memory (T_CM), (ii) T effector memory (T_EM), and (iii) tissue-resident memory (T_RM) (55, 57). It is largely known that effector CD8+ T lymphocytes may generate all these memory populations during the “contraction phase” of the immune response and CX3CR1 expression has a key role in this process. Specifically, CX3CR1+ effector T cells appear to preferentially differentiate into T_EM cells, while CX3CR1- precursors give rise to T_CM cells (37, 58, 59) ( Figure 1 ). The different origin of T_CM and T_EM reflects their phenotypical and functional differences. Similarly to naïve cells, T_CM CD8+ T lymphocytes express CD62L and CD197 homing receptors, which are responsible for their recirculating behavior (32, 48, 60). However, T_CM cells also express CD44 molecule in mice, and both CD44 and CD45RO epitopes in humans, indicating their memory-like phenotype (39, 61) ( Figure 1 ). During the second expansion phase, T_CM CD8+ T cells encounter their cognate antigen and differentiate into T_EM CD8+ T lymphocytes, losing the expression of homing receptors and migrating into peripheral tissues, where they can release cytotoxic molecules (39, 62) ( Figure 1 ). When the antigen is effectively cleared, T_EM cells evolve into antigen-specific T_CM CD8+ T lymphocytes, and the ultimate goal of these two populations is to provide systemic immunity.

Differently, local protective immune responses are orchestrated by T_RM CD8+ T lymphocytes, which have a distinct profile from other CD8+ T cell subsets (59, 63). Notably, T_RM CD8+ T cells are not recirculating, but are mainly organized in lymphoid niches close to anatomical and physiological barriers, acting as sentinels and protecting against reinfections (64). While further research is needed to fully understand the microenvironmental signals needed to establish and maintain the population of T_RM CD8+ T cells into different peripheral tissues, it is clear that IL-15, IL-7, TGF-β, IL-21, TNF-α, and IL-33 play a crucial role in the formation of this CD8+ T cell subset (21, 65–68). The retention of T_RM CD8+ T cells in peripheral tissues is mediated by CD69 and CD103 molecules, which are classically expressed on the surface of both human and murine T_RM CD8+ T lymphocytes (22, 37). CD69 inhibits the expression of sphingosine-1-phosphate receptor 1 (S1PR1) molecule, promoting T cell residency, while CD103 integrin, an adhesion receptor for E-cadherin, contributes to CD8+ T cells persistence inside the tissues (69–71). Notably, all T_RM CD8+ T lymphocytes express CD103 integrin in lymphoid tissues, but its expression may be lost in non-lymphoid tissues (64, 69). Accordingly, several studies reported the presence of both CD103+ CD69+ and CD103- CD69+ T_RM CD8+ T lymphocytes in non-lymphoid tissues, such as liver, brain, gut, skin, and lungs (17, 22, 63, 65, 72–76). Recently, it has been demonstrated that CD103+ and CD103- T_RM CD8+ T cells originate from two separate differentiation paths and are characterized by distinct effector functions (22, 37). ExKLRG1 effector CD8+ T cells give rise to CD103- T_RM lymphocytes which can be distinguished from CD103+ counterpart due to their cytotoxic potential (37, 65) ( Figure 1 ). In contrast, the precursors of CD103+ T_RM cells, featuring a lower expression of granzymes and other effector molecules, seem to originate from KLRG1- T_MPE CD8+ T lymphocytes (22, 37, 65) ( Figure 1 ).

It is now well established that the T_RM compartment of CD8+ T lymphocytes upregulates CXCR6 homing receptor gene, and ITGA1 gene, encoding CD49a collagen-binding integrin, while downregulating SELL and CX3CR1 genes, encoding for CD62L and CX3CR1 molecules, respectively (17, 57, 77). Moreover, Runx3, Notch, Bhlhe40, Blimp1 and its homolog Hobit, and the AP-1 family members, including Jun, Junb, Jund, Fos, Fosb, and Batf have been identified as crucial transcription factors (TFs) in the regulation of T_RM cells formation (59). Additionally, TFs induced by interferon signaling, such as Stat1, Irf1, Irf7, and Irf9, or related to the NF-κB signaling pathway, including Bcl3, Rela, Relb, Rel, and Nfkb2 are enriched in T_RM T lymphocytes, adding further markers for this CD8+ T cell population (59). Altogether, these data highlight the heterogenicity and complexity within the T_RM compartment of the CD8+ T cell population.

All stages of the immune response and their players are perfectly coordinated and functioning under acute inflammatory states, when the immune reaction successfully clears antigens. However, persistent antigen stimulation leads to chronic inflammation, disrupting this harmoniously synchronized mechanism (78, 79). Effector T cells in this condition become dysfunctional, undergoing exhaustion, resulting in poor effector functions and reduced proliferative potential (78, 79) ( Figure 1 ; Table 1 ). Importantly, T cell exhaustion is not just an alteration of cell phenotype and functions, but also represents a distinct differentiation state, with different characteristics compared to the memory features (24, 81). Exhausted (T_EX) CD8+ T cells maintain the same characteristics under different inflammatory conditions, with a well-defined gene signature, including TCR-signaling related genes such as Batf, Egr2, Ezh2, Irf4, Nfatc1, Nfatc2, Nr4a1, Nr4a2, and Nr4a3 (94–96), confirming that continuous exposure to persistent antigens is a key factor in T cell exhaustion, whereas short antigen exposures lead to exhaustion recovery (80) (81) ( Table 1 ).

Hyporesponsive T_EX cells are defined by their high surface expression of programmed cell death-1 (PD-1), lymphocyte-activation gene 3 (LAG-3), CD244 (2B4), T-cell immunoglobulin and mucin domain-3 (TIM-3), cytotoxic T-lymphocyte-associated protein-4 (CTLA-4), and CD160 inhibitory receptors (24, 82–85) ( Table 1 ). These receptors typically expressed on the surface of T_EX CD8+ T bind to a variety of ligands, suggesting that microenvironment clues, such as ligand availability, may regulate the functionality of T_EX CD8+ T lymphocytes (78, 79).

T_EX lymphocyte compartment is heterogeneous showing three differentiation states: (i) T cell factor 1+ (TCF1+) PD1+ CXCR5+ T_EX precursors (T_EX-Pre), expressing the T-bet TF and showing memory-like features, such as the expression of Sell, Ccr7, Id3, and Bcl6 genes; (ii) PD1+ TIM-3+ TCF1- Intermediate T_EX (T_EX-Int), not expressing Zeb1 gene, which encodes for the ZEB-1 TF; and (iii) PD1+ TIM-3+ CD101+ TCF1- terminally differentiated T_EX (T_EX-Term) expressing the TFs ZEB-1, Blimp-1, and Eomesodermin (EOMES) (24, 97, 98) ( Figure 1 ). Along the differentiation process, T_EX-Pre, T_EX-Int, and T_EX-Term show a hierarchical decrease of effector activity, marked by mitochondrial dysfunctions, and proliferative capacity, while gradually increasing the expression of inhibitory receptors, ultimately leading to cell death (78, 98, 99) ( Figure 1 ). However, T_EX lymphocytes do not entirely lack effector functions, exhibiting instead a gradual increase of (i) IL-2 production, (ii) cytotoxicity mediated by granzyme B and perforin, and (iii) release of pro-inflammatory molecules such as IFNγ and TNF-α (78, 79, 98) ( Figure 1 ; Table 1 ).

Senescent CD8+ T cells exhibit cell cycle arrest and poor proliferation along with severe functional abnormalities similar to those occurring during T cell exhaustion (78, 87, 88) ( Figure 1 ; Table 1 ). They differentiate from effector CD8+ T cells typically occurring during conditions associated to chronic inflammation, including auto-immune diseases, cancer and ageing (88–91) ( Table 1 ). Unlike hypo-functional T_EX lymphocytes, senescent CD8+ T lymphocytes continue to secrete a range of factors including proteases such as cathepsins and serine proteases, and cytokines such as CCL5, CCL16, CCL23, TNF-α, IL-29, and IL-18, which in turn may induce IFN-γ production, suggesting a pro-inflammatory senescent-associated secretory phenotype (SASP) (87, 92) ( Table 1 ). Moreover, CD45RA molecule is re-express on the surface of these hyper-functional CD8+ T lymphocytes, which are commonly referred as effector memory CD45RA+ (T_EMRA) cells (91). T_EMRA cells are characterized by the simultaneous expression of PD-1, KLRG-1, and CD57 on their surface, which also represent classical phenotypical markers of senescent CD8+ T lymphocytes (82, 91, 100–103) ( Figure 1 ). A certain degree of heterogenicity has been also shown among T_EMRA lymphocytes. Accordingly, a recent report revealed that effector CD8+ T cells differentiate into CD27- CD28- or CD27+ CD28+ T_EMRA cells based on the strength of TCR engagement and the immunogenicity of the tumor antigens (91) ( Figure 1 ), while other studies have shown that the lack of CD27 and CD28 surface molecules is associated with the expression of p16 and p21 proteins, causing G1 cell cycle arrest and replicative senescence (86, 87) ( Table 1 ).

Senescent CD8+ T cells possess the capability to express natural killer (NK)-associated receptors (NKR), such as NKG2D, NKG2C, NKG2A, and killer immunoglobulin-like receptor (KIR) families, allowing them to be reprogrammed into hyper-functional lymphocytes with the ability to recognize and kill target cells through both TCR and NKR recognition mechanisms (93) ( Figure 1 ; Table 1 ). Notably, senescent CD8+ T lymphocytes expressing NKR need to be distinguished from invariant natural killer T (iNKT) cells, which are a subset of natural killer T (NKT) innate immune cells that decrease with age (104, 105). Thus, senescent CD8+ T cells seem to have beneficial as well as detrimental roles during aging, having the potential to retain a broad spectrum of effector functions to kill malignant and infected cells, whilst also having the capacity to induce or sustain autoimmunity and other chronic disorders (106). However, further studies are required to better clarify the involvement of senescent CD8+ T cells during homeostasis and disease.

AD is a progressive neurodegenerative disorder characterized by neuronal death and accumulation of amyloid beta (Aβ) deposits and hyperphosphorylated tau protein in the brain, leading to memory loss and cognitive impairment (121). To date, approximately 35 million people worldwide have been affected by AD, making this disease the most common cause of dementia (122). Approximatively 75% of AD subjects are 75 years old or older, indicating a strong correlation between the development of AD and aging (123). Moreover, a shorter telomere length in CD8+ T cells was correlated to a greater AD severity, together with a lower CD28 expression and an increase of cytotoxic molecules and sensitivity to apoptosis, suggesting the presence of dysfunctional CD8+ T cells in AD (124, 125). The majority of studies performed in AD showed changes in the overall CD8+ T cell population in mice developing amyloid-related pathology, that may not closely reflect the human condition in which tau pathology also represents a disease hallmark (8, 10, 15, 17). Indeed, the association between tau pathology and CD8 T cells was previously suggested, although more recent studies found no correlation between tau hyperphosphorylation and the presence of cytotoxic T cells (15, 126, 127).

The heterogenicity of the CD8+ T compartment has been only recently studied in AD, showing clonally expanded CD8+ T_EMRA cells with cytotoxic potential in the cerebrospinal fluid (CSF) of AD individuals (6). Surprisingly, T_EMRA lymphocytes were not clonally expanded against AD-specific antigens such as Aβ peptides or tau protein, but they were reactive in the presence of Epstein-Barr virus (EBV) antigens (6) ( Figure 2 ). These data do not provide a causal link between EBV infection and AD, but suggest that senescence of CD8+ T lymphocytes may play a role in AD (128). Furthermore, the absence of clonal expansion toward disease-specific antigens suggests that CD8+ T cell trafficking into the CNS could be a stochastic phenomenon (6, 17). Accordingly, the damage of CNS barriers previously described in AD can favor nonspecific CD8+ T cells brain invasion from the blood circulation (129) ( Figure 2 ). The increased expression of lymphocyte function-associated antigen 1 (LFA-1) integrin on infiltrating total T cells in the AD brain, and the augmented expression of Itgb2 gene, encoding for the CD18 subunit of LFA-1 molecule, in the hippocampus of a mouse model of tauopathy compared to controls suggest a role for LFA-1-intercellular adhesion molecule-1 (ICAM-1) interactions in the migration of CD8+ T cells in the CNS during AD (8, 126). This is also supported by the increased expression of ICAM-1 detected on brain endothelial cells of mice with AD-like disease and higher levels of soluble ICAM-1 in the plasma from AD subjects compared to controls (3, 130).

Although ablation of CD8+ T cells in AD mice did not reduce Aβ deposition and cognitive deficits in APP-PS1 mice at later disease stages, recent studies suggested that CD8+ T cells migrate in the brain parenchyma in subjects with AD and its mouse models (15, 127) ( Figure 2 ). The phenotype of CD8+ T lymphocytes in aged and AD transgenic mice have a T_RM gene signature, but the characterization is still in its infancy (17, 20) ( Figure 2 ). In summary, existing literature suggests CD8+ T cell heterogeneity in AD, highlighting a potential role for T_EMRA and T_RM cells subsets in the pathogenesis of this disorder. However, the involvement of CD8+ T cells in AD is still unclear and more studies are needed to identify the molecular mechanisms mediating a potential CD8+ T cell-dependent damage.

PD is the second most common form of neurodegenerative disorder, affecting more than 10 millions people around the world, with an increased prevalence in the aged population (131). The main neuropathological hallmark of the disease is the presence α-synuclein (α-syn) aggregates, referred to as Lewy bodies, and Lewy neurites, observed in neuromelanin-containing neurons of the substantia nigra (SN) (132). This is associated with the classical PD symptoms of bradykinesia, rest tremor, muscular rigidity, motor and cognitive alterations, as well as autonomic dysfunctions (132). Similarly to AD, the initiators of the pathogenic cascade leading to neuronal death and to the related neurological alterations are presently poorly understood in PD.

Recently, a study conducted in PD patients provided novel insights into the T cell-mediated adaptive immune responses by performing single-cell transcriptome and TCR sequencing, revealing a continuous progression of CD8+ T cells from a central memory to a terminal effector phenotype (133). In addition, previous data showed that CD8+ T cells are increased in the substantia nigra of diagnosed PD cases and positively correlate with neuronal death (5). Interestingly, CD8+ T cells infiltration is present since early disease stages, preceding neuronal loss and α-syn aggregation, suggesting a role for these cells in disease development (5). Notably, although a direct demonstration is still lacking, CD8+ T cells with a T_RM phenotype were detected near altered neurons in the substantia nigra, suggesting that CD8+ T cells may exert cytotoxic functions potentially contributing to the pathological changes in PD (5) ( Figure 2 ). Furthermore, a longitudinal case study and analysis of two independent cohorts revealed that elevated α-syn-specific CD8+ T cell responses in the blood were present prior to and after diagnosis of motor PD, and were significantly associated with age (134) ( Figure 2 ). These studies suggest that CD8+ T cell infiltration is an early event in PD, paralleling the progression of neuronal death and synucleinopathy, providing insight into new disease mechanisms and early diagnosis in PD. Accordingly, T cells from PD patients do not react to common antigens, but are activated by α-syn-derived antigens, suggesting that the T cell responses observed in PD may be prevalently directed towards autoantigens (135).

Recent data showed a core gene signature for α-syn-reactive CD8+ T lymphocytes in PD, which includes the expression of CX3CR1, CCR5, CCR1 pro-inflammatory genes, but also genes expressed in neurons such as LRRK2, LAMP3, and aquaporin genes, previously associated with PD worsening (136). Interestingly, the increased expression of the leucine-rich repeat kinase 2 (LRRK2) in CD8+ T lymphocytes from PD patients correlated with an increased secretion of pro-inflammatory molecules and cell activation, suggesting that LRRK2 may represent a therapeutic target in PD (137). Accordingly, increased expression of granzymes and perforin 1 in clonally expanded CD8+ T cells are present in the blood and CSF of PD patients (133) ( Figure 2 ). Also, the increased cytotoxicity of CD8+ T lymphocytes in PD correlates with a terminal effector phenotype and the expression of proteins involved in cell migration, such as CX3CR1 and the adhesion G protein-coupled receptor G1 (ADGRG1) (133). The migration of CD8+ T lymphocytes into the CNS during PD (5) has been associated with the upregulation of ITGAM and ITGB1 genes, suggesting a role for CD11b and CD29, subunits of LFA-1 and very late antigen-4 (VLA-4) integrins, respectively (133) ( Figure 2 ). Moreover, the disruption of the blood-brain barrier (BBB) in PD (138, 139), indicate that vascular phenomena may also contribute to CD8+ T cell infiltration into the CNS ( Figure 2 ). Overall, these findings emphasize an increased immune reactivity of CD8+ T cells against CNS antigens in PD, which is associated with an enhanced clonally expanded T cell infiltration and cytotoxicity in the brain, suggesting a circulating origin of α-syn-reactive CD8+ T lymphocytes in the PD brains (133, 135). Future studies exploring the phenotype and functions of CD8+ T lymphocyte subsets in PD could lead to the identification of novel and specific therapeutic targets for this neurodegenerative disease.

MS is a chronic inflammatory and autoimmune disorder of the CNS affecting approximatively 2.8 million people worldwide (140). Multifocal inflammatory lesions develop in both brain and spinal cord (SC), resulting in demyelination and neurodegeneration and leading to a progressive decline in motor, sensory, and cognitive functions (141). In most patients, MS is characterized by a relapsing and remitting onset, followed by a chronic, prolonged, and progressive inflammatory state during which the neurological symptoms gradually worse (141). Several studies have shown that CD8+ T lymphocytes predominate in active MS lesions (142–144). Furthermore, CD8+ T cells from in MS lesions are clonally expanded against common antigens and have been probably recruited from the periphery (12, 142, 145) ( Figure 2 ). Indeed, it has been demonstrated that blood, CSF and brain CD8+ T cell clones share a high degree of phenotypic similarity further supporting the view that CD8+ T cells invading the MS brain originate from the periphery and contribute to MS progression (144, 145). The homing of clonally expanded CD8+ T lymphocytes to the CNS of MS patients may be further promoted by BBB damage and increased expression of adhesion molecules and chemoattractants on brain endothelial cells (143, 146) ( Figure 2 ). Moreover, previous studies have shown a role for platelet and endothelial cell adhesion molecule 1 (PECAM1), P-selectin glycoprotein ligand-1 (PSGL-1), vascular cell adhesion protein-1 (VCAM-1), intracellular adhesion molecule-1 (ICAM-1), VLA-4 and LFA-1 in the migration of CD8+ T cells into the CNS (143, 144, 147, 148) ( Figure 2 ). Once infiltrated to the CNS, CD8+ T cells express higher levels of cytotoxic molecules (19, 143, 144), thus suggesting their detrimental role in the progression of MS-related neurological alterations. Indeed, CD49d+ CD8+ T lymphocytes invading the MS brain exhibited a pro-inflammatory effector phenotype, expressing CD137 and CD95L, as well as inhibitory receptors TIM-3 and PD-1 on their surface, in addition to the transcriptional factor EOMES (12, 14, 19, 143) ( Figure 2 ). Importantly, an increased production of granzyme B by CD8+ T cells in MS lesions, has been suggested to contribute to neuronal alterations (143, 144) ( Figure 2 ). In line with these findings, in early MS brain lesions the majority of CD8+ T lymphocytes express CD69, but not CD103, and were shown to contain granzyme B (12). Although a link has been suggested between T cell exhaustion and the progression of chronic neuroinflammation in MS, the contribution of T_EX CD8+ T lymphocytes in MS course is unclear (149). EBV infection is now considered a key environmental factor for chronic CNS inflammation during MS and CD8+ T cells are clonally expanded against EBV-derived antigens presented by B cells in MS patients (12). EBV can establish a prolonged latent and intermittent reactivation within B cells (150), potentially resulting in a series of CD8+ T cells immune responses, suggesting that B cells may represent essential players in promoting chronic CD8+ T lymphocyte activity in MS (151). Overall, MS represents the neurodegenerative disease in which CD8+ T cells were studied more into detail, although the molecular mechanisms leading to brain damage are not yet fully understood.

ALS is an incurable and devastating neurodegenerative disorder, that causes a progressive degeneration of motor neurons leading to a loss of voluntary muscle control and, in severe cases, to respiratory failure (152). Most commonly, symptoms of ALS appear between the ages of 40 and 70, and several genes were associated to the development of this disease, including superoxide dismutase 1 (SOD1) (152–155). The normal function of SOD1 protein is to protect cells from oxidative damage, and its alterations lead to increased oxidative stress and mitochondrial dysfunctions (156). In addition, oxidative stress levels are substantially increased by neuroinflammation, which has been recently included among the hallmarks of ALS (157, 158).

A role for CD8+ T lymphocytes in SOD1-associated ALS form was recently suggested (159–161). Particularly, peripheral CD8+ T cell ablation has been shown to increase motoneuron survival in a mouse model of ALS, whereas in vitro studies showed that SOD-1 expressing CD8+ T lymphocytes recognize and selectively kill motoneurons via binding MHC-I molecules expressed on these cells, suggesting a possible autoimmune origin for ALS (159) ( Figure 2 ). Moreover, activated CD8+ T lymphocytes expressing mutant SOD-1 produce high levels of IFNγ and eliminate a subset of motoneurons in ALS through an antigen restricted, MHC-I-dependent cytotoxic pathway, suggesting a neurotoxic role for self-reactive CD8+ T cells in ALS (159) ( Figure 2 ). These data were supported by studies performed in ALS patients showing an increased activation of peripheral and intrathecal CD8+ T cells, with the activation status of CD8+ lymphocytes in the blood being higher in ALS compared to MS and dementia, further suggesting a role for cytotoxic lymphocytes in ALS (160).

Clonally expanded T_EMRA CD8+ T cells originating from the circulation have been reported in the CNS of Setx knock-in (KI) mice developing ALS4-like disease, as well as in ALS4 patients, who typically have mutations on senataxin (SETX) gene (18) ( Figure 2 ). These data point to T_EMRA cells as negative contributors during ALS and are further supported by a recent retrospective study demonstrating higher frequencies of senescent-like T cells in ALS individuals, suggesting that lymphocyte senescence may drive disease progression (162). Furthermore, the immunophenotyping of T_EMRA CD8+ T lymphocytes detected in the CNS of Setx KI mice revealed a CD49d+ PD-1+ CD103- profile (18), which is consistent with their peripheral origin in ALS4-like mice, suggesting pathological changes of the T_RM CD8+ T cell compartment similar to those observed in MS (12) ( Figure 2 ). Nevertheless, differently from MS, the severity of ALS does not appear to be correlated with BBB leakage, suggesting that the infiltration of CD8+ T cells into the CNS of ALS patients and animal models could represent a tightly regulated process, rather than being favored by stochastic events (163, 164) ( Figure 2 ). Recent data obtained in two ALS mouse models showed that blocking α4-integrins reduces the migration of peripheral immune cells into the CNS and decreases IFNγ, which is primarily produced by CD8+ T lymphocytes and NK cells, further supporting a role for peripheral cytotoxic T cells in ALS (159, 161, 165). However, more research is needed to identify the molecular mechanisms governing the extravasation of CD8+ T cells into the CNS during ALS.

Limbic encephalitis (LE)-induced temporal lobe epilepsy (TLE) is a rare form of epilepsy characterized by different types of epileptic seizures, including focal seizures, secondarily generalized seizures, and status epilepticus (166). This subtype of encephalitis is also characterized by a potent inflammatory reaction predominantly against neurons in the grey matter of the medial temporal lobes of the brain, leading to the generation of recurrent seizures (167, 168). Previous studies associated neuroinflammation, BBB dysfunction and leukocyte migration to the induction of seizures (169–172).

LE is triggered by an autoimmune response favored by various underlying causes, including viral infections and paraneoplastic syndromes (173–175). Indeed, the presence of autoantibodies directed toward neuronal surface antigens (NSAbs) was well described in LE patients (176–180). Moreover, it was suggested that IFNγ-producing CD8+ T lymphocytes promote MHC-I upregulation on LE neurons, thus supporting a neuron-directed CD8+ T cell attack, similar to the immune reactivity demonstrated in ALS (16, 181) ( Figure 2 ). CD8-mediated neurotoxicity has been suggested to contribute to neuronal excitability and acute seizure generation, promoting psychiatric disturbances, memory impairment and behavioral changes (182). Furthermore, it has been hypothesized that chronic neuroinflammation can induce persistent changes in the structural and electrical properties of certain neuronal networks, resulting in the development of chronic spontaneous seizures and epilepsy (16, 181, 182) ( Figure 2 ). Notably, infiltration of granzyme B-producing CD8+ T lymphocytes, which may attack and destroy neurons expressing MHC-I, was found among other leukocyte populations in the hippocampi of TLE patients (2, 16, 183–188) ( Figure 2 ). Interestingly, cytotoxic CD8+ T cells accumulate in the brain of LE patients, suggesting that self-reactive CD8+ T cells can directly cause neurotoxicity, but can also contribute to the conversion of LE into TLE (178) ( Figure 2 ).

Recent studies showed that activated CD44+ ovalbumin (OVA)-specific CD8+ T cells, directed against the “SIINFEKL” chicken peptide (OVA 257-264) expressed exclusively by hippocampal neurons, migrate into the brain and promote seizures (16). Furthermore, CD8+ T lymphocyte infiltration leads to prolonged epileptic activity and memory deficits, suggesting that circulating CD8+ T cells may play a role in the induction and progression of LE-induced TLE (16) ( Figure 2 ). Brain invasion by circulating CD8+ T lymphocytes correlate to BBB dysfunction in experimental models of TLE and magnetic resonance imaging revealed BBB alterations after the transfer of OVA-specific CD8+ T cells, followed by a size reduction and degeneration of the hippocampus (2, 16, 169) ( Figure 2 ). Moreover, brain-invading CD8+ T lymphocytes showed an increased expression of CD69 surface marker, suggesting that these cells may acquire a T_RM phenotype once migrated into the brain (16), similarly to what has been previously described in MS (12). This may be supported by previous studies showing that ICAM-1 and VCAM-1 are upregulated in TLE hippocampi and after seizure induction and that preventive blockade of α4 integrins or ICAM-1 adhesion receptor abrogates seizure induction (2, 169). Of note, TLE patients display a T cell activated phenotype in peripheral blood (185). Thus, CD8+ T lymphocytes could migrate into the inflamed brain in TLE using LFA-1 and/or α-4 integrins, as shown in ALS and as previously suggested in patients with MS and epilepsy (189) ( Figure 2 ).

Sus is a rare disorder characterized by neuroinflammation and CNS dysfunction, due to focal microangiopathy affecting small and medium size vessels of the brain, retina and the inner ear (190, 191). The etiology of Sus remains largely unknown, and the role of neuroinflammation has only recently started to be addressed. However, the successful use of immunosuppressive and immunoregulatory drugs in Sus cases, support an autoimmune origin for this disease (192, 193). Accordingly, an increased immune cell infiltration into the brain, with the majority of cells being CD8+ T lymphocytes, has been found in Sus patients compared to healthy controls (9, 194). Also, it has been suggested that circulating self-reactive CD8+ T lymphocytes may promote endothelial cell alterations and BBB dysfunction during Sus, potentially leading to microinfarcts and leukocyte transmigration across the BBB (13) ( Figure 2 ). Moreover, Sus patients treated with natalizumab, which inhibits lymphocyte migration into the CNS, or mice treated with an antibody blocking anti-α4 integrins, showed a reduction in disease severity compared to controls, further supporting the idea that Sus shares several key immunopathological elements with MS (13) ( Figure 2 ). Indeed, VLA-4 neutralization was found to restrict infiltration of CD8+ T lymphocytes into the CNS of EAE mice (143, 195), suggesting this may be the case also in Sus. Despite several common disease mechanism between MS and Sus, the total number of leukocytes detected in the CSF is in the normal range in Sus patients, whereas it is significantly increased in MS cases (13). However, the proportion of CD8+ T cells is selectively and significantly increased in the CSF of Sus subjects, suggesting a role for these cells in Sus (13). Sus patients have a larger proportion of clonally expanded CD57+ CD8+ T_EMRA lymphocytes of circulating origin close to disease onset, with lower clonal expansion levels during remission, further supporting a relationship between Sus pathogenesis and CD8+ T cells (13). Moreover, immunophenotyping of CD8+ T lymphocytes in brain of Sus patients revealed a cytotoxic capacity, as evidenced by the higher expression of granzyme B lytic molecule compared to healthy controls (13) ( Figure 2 ). Overall, several lines of evidence point to a key role for CD8+ cells in Sus, but more studies are needed to understand how these cells contribute to BBB dysfunction and neuronal damage.