Despite the status of stroke as one of the principal causes of death and disability worldwide, there is currently a paucity of treatments available to ameliorate its devastating effects [1, 2]. Over 250 clinical trials have failed [3], suggesting that the discovery of efficacious treatments may require further investigation of the basic biology of stroke pathogenesis [4-6]. One promising focus of basic stroke science investigation is the field of stroke immunobiology. Ischemic stroke pathogenesis is complex, culminating in mitochondrial and DNA damage, release of reactive oxygen species, inflammation, and programmed cell death [7, 8]. Several studies support the status of immune responses and inflammation as important factors in the pathogenic process: inflammation, initiated by stagnant blood flow, activation of intravascular leukocytes, and the release of pro-inflammatory mediators from the ischemic endothelium, has been shown to increase brain injury [9, 10]. The immune system is involved in all stages of stroke. Local inflammatory responses mainly occur through activation of innate and adaptive immunity [11, 12]. After an acute stroke, sensors of the innate immune system such as Toll-like receptors (TLRs) and innate immune cells, are activated by brain ischemia, leading to amplification of the inflammatory cascade. Subsequently, the adaptive immune system mediated by lymphocytes is activated and further amplifies the inflammatory response [13]. In particular, there is a growing body of evidence linking TLRs to the deleterious inflammatory effects of ischemia/reperfusion injury associated with stroke [14, 15]. Moreover, the finding that genes responsible for innate inflammatory regulation comprise the majority of those acutely modulated in the post-stroke circulation suggests the importance of inflammation as an injury mechanism [16-18].

Research has demonstrated that immunoregulation may limit excessive inflammation after reperfusion. Inhibition of post-stroke inflammation, as exemplified by the effect of TAM/Gas6 signaling on the innate immune response promoted by TLRs [19], is recognized as a promising neuroprotective strategy. Additional innate immune-mediated protection against stroke is conferred by anti-inflammatory molecules such as IL-1 receptor antagonist (IL-1Ra), IL-10, and TGF-β. IL-1Ra, for instance, not only limits brain injury by ischemic stroke, but also appears to confer benefit in patients with subarachnoid hemorrhage (SAH) [20-22]. Accordingly, both the TAM/Gas6- and cytokine-mediated pathways will be discussed in this review.

Following innate immune activation, inflammatory mediators produced in this initial response recruit the cellular effectors of the adaptive immune response. T-lymphocytes, in particular, play an integral role in the inflammatory response after stroke [23]. These, and the regulatory subset that has been implicated in neuroprotective immunosuppression, will also be discussed below after the initial section that describes the innate response.

Finally, a discussion of the neuro-immunomodulatory effects of the vagal cholinergic anti-inflammatory pathway, paired with a related discussion regarding the regulatory roles played by activation of the SNS and HPA axes, will conclude the review. Because neuro-immunomodulation negatively correlates with infarct volume [24], this pathway may represent an endogenous target for future pharmacologic neuroprotective strategies.

TLRs and TAM/Gas6

TLRs are transmembrane proteins composed of three structural domains: a leucine-rich repeats (LRRs) motif, a transmembrane domain, and a cytoplasmic Toll/IL-1 receptor (TIR) domain. Thus far, TLR1-TLR10 have been identified in humans [15]. Through the downstream actions of MyD88 and TRIF, these receptors have been shown to activate the innate immune response [25]. The TLR signaling pathway plays a crucial role in the pathogenesis of stroke [26]. TLR2 and TLR4 have been found to be especially important in this context: previous research has reported that these TLRs are widely expressed in the brain following cerebral ischemia, and may exacerbate tissue damage [27]. In a clinical study of ischemic stroke patients, it was reported that increased TLR2 and TLR4 expression was independently associated with poor functional outcome [28]. Recognition of this association may pave the way toward future therapies, as was illustrated by the same study in an in-vitro model of stroke, in which blocking TLR2 and TLR4 reduced the monocytic inflammatory response.

The TAM receptor class consists of three receptors: Tyro3, Axl, and MerTK. Growth arrest-specific gene 6 (Gas6) is a common ligand of these TAM receptors [29]. All three TAM receptors have similar extracellular domain structures, including two tandem N-terminal immunoglobulin-like domains (IGs) and two membrane-proximal fibronectin type III-like (FNIII) domains. TAM receptors and their ligands exert anti-inflammatory action and may be thought of as a countervailing force against the effects of TLRs [30]. TAM receptor activation may inhibit TLR activation and the associated signal transduction cascades, including those involving NF-κB and MAPK, and has been shown to inhibit the level of TLR-induced pro-inflammatory cytokines, such as TNF, IL-6, and IL-12 [31].

TAM/Gas6-mediated immunosuppression

As previously mentioned, TLRs are critical initiators of post-stroke inflammation. They are activated by the release of danger-/damage-associated molecular patterns (DAMPs), resulting in the production of pro-inflammatory mediators [15]. Activation of TLRs initiates signal transduction cascades that involve kinases, including the transcription factors activator protein-1 (AP-1) and NF-κB, which induce the expression of inflammation-associated molecules and cytokines [32]. To avoid cerebral damage secondary to chronic inflammation, the TLR signaling pathway must be tightly regulated. Recent studies have demonstrated that TAM receptors act as innate immune system regulators by inhibiting the TLR-mediated pro-inflammatory response [30], and they have been identified as a potential therapeutic target in the setting of stroke [33].

Under both physiological and pathological conditions, TAM receptors are widely expressed in cells of the immune and nervous systems [34, 35]. The level of TAM receptor expression is significantly increased after birth and remains high in adults, suggesting the importance of these receptors in normal physiology. This was further confirmed by a study of TAM knockout mice, in which peripheral lymphoid organs began to enlarge three weeks after birth. The spleens and lymphoid cell populations of these mice demonstrated excessive proliferation at 6 months, resulting in an imbalanced immune response [36]. The negative immunoregulatory mechanism of TAM receptors proceeds through downstream signaling cascades involving SOCS1 and SOCS3 E3 ubiquitin ligases. These, in turn, inhibit inflammatory responses mediated by regulatory signaling molecules such as TLRs and NF-κB [37]. TAM receptors are critical to the maintenance of immunohomeostasis: excessive inflammation mediated by the TLR signaling pathway would upregulate TAM/Gas6, yielding a TAM-initiated damping of TLR inflammatory signaling. In addition, the function of TAM receptor signaling has been demonstrated in central nervous system physiology. Tyro3, Axl, and MerTK are involved in the early development of the nervous system [38]. They have also been linked to CNS pathophysiology: the endogenous expression of Gas6 and Axl decreased significantly 24 h after middle cerebral artery occlusion (MCAO), and recombinant Gas6 reduced brain injury and inhibited the TLR/TRAF/NF-κB pathway [39].

Microglia, which act as resident macrophages within the central nervous system and are known mediators of neuroinflammation [40], express cytokine receptors, such as TLRs [41].Activation of microglia results in neuronal damage through the release of proinflammatory cytokines. Several reports have shown that TAM receptors may regulate the function of microglia [42]. Microglia express all three TAM receptors [43], and microglial neuroinflammation is subject to inhibition by the TAM/Gas6 system, which has been shown to maintain the cells’ phagocytic ability while inhibiting LPS-induced IL-1β expression. A recent study showed how this pathway may be exploited in stroke therapy [39]. Further evidence comes from studies that have investigated microglial function in the context of TAM signaling deficiencies. For instance, microglia lacking TAM receptors were found to produce a large number of proinflammatory cytokines after activation [43]. Additionally, deficient activity of both Axl and MerTK was found in adult mice to result in reduced microglial activity and impaired apoptotic cell clearance [42]. Finally, MerTK expression was found to be stimulated by immunosuppressive drugs, such as dexamethasone [44]. Taken together, this body of research suggests that TAM receptors participate in the negative regulation of microglial innate immune responses.

Astrocytes represent another type of immunomodulating cell capable of producing inflammatory neurotoxic mediators [45]. They have been shown to exhibit strong expression of Tyro3 and Axl, along with low expression of the MerTK receptor [46]. Analogous to the aforementioned microglia research, TAM-deficient astrocytes were found to release higher levels of IL-6 after LPS activation than was observed in wild type (WT) counterparts. Additionally, TAM-deficient astrocytes produced a stronger IL-1β expression response to pro-inflammatory stimulation, than WT cells [45] (Fig. 1).

IL-1 receptor antagonist

IL-1 is an important pro-inflammatory mediator of brain injury, and its inhibition has yielded neuroprotective efficacy in experimental models over the past two decades [47, 48]. There is a growing body of evidence indicating that rapid up-regulation of the two principal IL-1 ligands, IL-1α and IL-1β, occurs after stroke [49, 50]. Recently, a third ligand was discovered, and found to be a competitive antagonist: IL-1Ra [51]. This antagonist has been investigated in the context of subarachnoid hemorrhage: Greenhalgh et al. [52] reported that IL-1Ra is beneficial in a rat SAH model, and a phase II randomized controlled trial further demonstrated that IL-1Ra is safe for SAH patients. In the latter study, IL-1Ra was found to cross the blood–brain barrier (BBB) and appeared to decrease (albeit not to a degree that reached statistical significance) the concentration of IL-6, an inflammatory marker and downstream product of IL-1, in the cerebrospinal fluid (CSF) [53]. The therapeutic potential of IL-1Ra was also demonstrated in a mouse model of ischemic stroke, in which increased expression of IL-1Ra in leukocytes conferred neuroprotection [21]. These findings were echoed by another recent experiment by Pradillo et al. [54], in which IL-1Ra administration reduced ischemic brain injury in rats. In humans, intravenous IL-1Ra has been found in Phase II trials to be safe for use in ischemic stroke [55]; additionally, a recent Phase II trial of subcutaneous IL-Ra in ischemic stroke patients demonstrated reductions in inflammatory markers associated with poor acute post-stroke outcome [56].

IL-10 and TGF-β

Like IL-1Ra, IL-10 acts in the context of brain injury as an anti-inflammatory cytokine, promoting neuronal survival and suppressing inflammatory responses by producing inhibition at a number of steps in cytokine signal transduction, including cytokine synthesis, cytokine receptor expression, and cytokine receptor activation [57, 58]. It is mainly produced by astrocytes and microglia [59, 60]. The protection conferred by this cytokine against ischemic brain damage is illustrated by an experiment that compared response to permanent MCAO between IL-10^-/- mice and wild-type mice. Compared with the latter, knockout mice had 30% larger infarct volumes 24 h post-stroke [61]. Similarly, IL-10 has been reported to improve neurological outcomes and limit infarct volume in experimental stroke [62, 63], while IL-10 polymorphisms have been implicated in increased susceptibility to stroke [64]. A study that employed intracerebroventricular injection in a mouse ischemic stroke model highlighted the genetic component of the effects of IL-10, showing that this method produced downregulation of over 300 genes that were upregulated by ischemia; moreover, most of these genes were associated with inflammation [65]. Additionally, several epidemiologic studies have demonstrated a relationship between IL-10 and ischemic stroke risk. A study in a south Indian population, for instance, showed that possession of the A allele of the IL-10 promotor SNP rs1800896, which is associated with low IL-10 production, conferred an increased risk of ischemic stroke [66]. Another study, in this case of a population of Eastern Finnish origin, found that plasma IL-10 correlated with high-risk sources of cardioembolic stroke, which suggested its utility in improving identification of stroke etiology [67].

TGF-β is another neuroprotective and anti-inflammatory mediator that shows promise as an effective therapeutic agent in stroke. TGF-β is mainly produced by astrocytes and microglia. Like IL-10, blocking TGF-β exacerbates brain damage [68], while TGF-β overexpression leads to a decreased inflammatory response and reduced brain injury in mice following MCAO [69]. In rats, TGF-β antagonism can aggravate brain damage caused by focal cerebral ischemia [16]. Pretreatment with TGF-β, however, attenuated the activation of NF-κB and upregulation of IL-1 mRNA levels, thus reducing production and release of proinflammatory cytokines [70]. Additionally, TGF-β has been reported to function on the cellular level by decreasing the chemotactic activity of microglia [71].

Anti-inflammatory regulation of Hsp70

Heat Shock Proteins (Hsps) are evolutionarily conserved molecules that reduce brain injury. They have been divided into six major families, including Hsp100, Hsp90, Hsp70, Hsp60, Hsp40, and small Hsps. Among these, it has been reported that Hsp70 confers an anti-inflammatory effect through inhibition of both pro-inflammatory transcription factor NF-κB activation, and of ROS [72]. Hsp70 has been detected in neurons, astrocytes, microglia, and endothelial cells after brain infarction, and appeared in the context of a stroke model to function through direct association with NF-κB and IκB, whereupon it prevents IκB phosphorylation [73]. Other experimental findings that have helped to delineate the function of Hsp70 include that intracellular overexpression of Hsp70 decreases NF-κB activation in astrocytes [74]; overexpression in an experimental stroke model produced downregulation of TNF-α and IL-1β [75]; extracellular Hsp70 modulates immune responses [76]; and intravenous injection of Hsp70 in rat model reduced the volume of infarction produced by experimental stroke induction [77].

T cells are rapidly recruited to the ischemic brain within 24 h post-stroke, at which point they generally function to aggravate brain damage [78, 79]. This is not invariably the case, however, as is illustrated by the Treg subset of T cells, which plays an important protective role in immunohomeostasis and the pathophysiology of ischemic stroke [80]. Treg cells accumulate within the ischemic hemisphere, spleen, and proximal and distal lymph nodes in experimental models of stroke [81], and Dolati et al. [82] demonstrated that Treg depletion may promote stroke. Conversely, the augmentation of Treg activity was shown to decrease both the volume of an experimentally-induced infarct, and subsequent post-stroke deficit [81]. There are many subpopulations of Tregs, including Th3, Trl, CD8 Tregs, natural killer Tregs, and CD4⁺CD25⁺FoxP3⁺Tregs [83]. Among these, CD4⁺CD25⁺FoxP3⁺ Tregs are the most well-characterized. CD4⁺CD25⁺FoxP3⁺Tregs mainly arise from progenitor cells in the bone marrow and develop in the thymus through the processes of positive and negative selection (Fig. 2) [84]. CD4⁺CD25⁺FoxP3⁺Tregs can be activated via T cell receptors [85]. In the context of stroke, a growing body of evidence implicates CD4⁺CD25⁺FoxP3⁺Tregs as important neuroprotective immunomodulators, but the mechanism by which this effect proceeds remains unclear [86, 87]. One experiment observed that, in mice, CD4⁺CD25⁺Tregs were beneficial to neuronal survival after an ischemic insult by modulating autoimmunity [88]. In another mouse study, these Tregs were reported to be associated with cocaine-and-amphetamine-regulated-transcript-mediated neuroprotection after stroke [89]. A third study, in this case performed on patients following acute ischemic stroke, reported both an increase of CD4⁺CD25⁺FoxP3⁺Tregs in the peripheral blood, and a reduction in the suppressive effects of these cells on T cell proliferation [90].

Another manner by which the function of Treg cells has been investigated is through their interactions with microglia. Foxp3, which functions as a repressor of microglial activation, plays an important role in reducing microglia-mediated neuroinflammation. Cerebral Foxp3⁺ Tregs were shown to inhibit the LPS- induced inflammatory response of microglia in vitro [91], while a Foxp3-mutation appeared to increase microglial release of pro-inflammatory factors, such as CXCL10 and MCP-1, in a mouse model [92]. Mechanistically, Tregs appear able to promote polarization of microglia toward an M2 phenotype accompanied by lowered IL-6 and TNF-α expression in vitro [93]. In summary, these experimental results demonstrate the beneficial consequences of Treg activation both in vitro and in vivo (Fig. 2).

Finally, yet another line of evidence involves programmed death-1 (PD-1), a T cell regulatory molecule with two ligands, PD-L1 and PD-L2. PD-L1 is widely expressed on T cells, B cells, monocytes, and dendritic cells (DCs); PD-L2 is located on macrophages, certain B cells, and DCs [94]. PD-L1 is expressed on Tregs and functions to inhibit the proliferation of T cells. Through inhibition of neutrophil MMP-9 expression, Treg-mediated PD-1/PD-L1 interaction has been demonstrated to reduce brain damage after ischemic stroke [95].

Communication between the immune and nervous systems is essential in host defense against inflammatory diseases. There is a growing body of evidence confirming that neuroendocrine pathways, such as the hypothalamopituitary-adrenal (HPA) axis and the sympathetic division of the autonomic nervous system (SNS), act as anti-inflammatory balancing mechanisms to regulate the inflammatory response [12, 96, 97]. Neurohormonal negative regulation of immune function by the HPA axis has been widely studied [98]. HPA axis activation results in the release of glucocorticoids, which exert their manifold effects on many different effectors through their action on transcription-modulating intracellular receptors. In immune cells such as macrophages, neutrophils, T cells, and B cells, cortisol is a potent anti-inflammatory molecule that acts to downregulate the production of pro-inflammatory cytokines through interaction with the NF-κB transcription factor that controls the production of these cytokines [99]. The SNS has been shown to possess both pro- and anti-inflammatory potential [100]. All lymphoid organs receive input from postganglionic sympathetic neurons; norepinephrine released through the termini of these nerves regulates the activity of the cellular effectors of immune function. Catecholamines released through this pathway bind to receptors (primarily the β₂ adrenergic receptor) on the surface of the cellular effectors of immunity, whereupon they act through intracellular signaling cascades to influence gene expression relevant to the regulation of immune responses [101]. It has been suggested that cocaine- and amphetamine-regulated transcript-mediated neuroprotection may occur through SNS regulation [89]. Down-regulation of inflammation by the SNS proceeds mainly through β-adrenoceptors. One study addressed the relevance of this mechanism to stroke pathogenesis by reporting that stroke-induced activation of the SNS results in the secretion of catecholamines, causing a β-adrenergic receptor-mediated reduction in TNF-α and concomitant promotion of IL-10 production [102].

The cholinergic anti-inflammatory pathway is classified as a neuro-immunomodulatory pathway [103]. As compared to the neuroendocrine mechanisms mentioned above, neuro-immunomodulation is distinguished by its rapid action. When pro-inflammatory cytokines are released after immune response activation, sensory vagal afferents and regulatory vagal efferents form an inflammatory reflex arc that continuously monitors the response [104]. The cholinergic anti-inflammatory pathway can be activated in this setting to counteract the release of excessive TNF-α; the ability of this pathway to inhibit the release of cytokines has been shown to promote neuroprotection [105]. The central function of the pathway may be mediated through stimulation of the α7 nACh receptors (α7 nAChRs) responsible for microglial activation: intraperitoneal injection of PNU-120596, an allosteric modulator of α7 nAChRs, decreased infarct size and improved neurological tests results in a mouse MCAO model [106, 107]. Subsequent work confirmed that intranasal administration of PNU-120596 produced a similar therapeutic effect in a rat model [108]. Likewise, treatment with the selective α7 nAChR agonist PHA 568487 was found to be associated with a decrease in the number of microglia expressing the M1 phenotype and an increase in number of M2 microglia in a pMCAO model [109]. PHA 568487 administration was also found to reduce brain injury after experimental stroke in rodents [110].

The inflammatory response, by disrupting immunohomeostasis and aggravating brain damage, constitutes a major contributory factor to the pathobiology of stroke. Previous studies have highlighted the neuroprotective effects of immunosuppression, which is achievable through both innate and adaptive mechanisms. Various regulatory pathways, such as the TAM/Gas6 pathway, have been shown to be involved in modulating post-stroke inflammation, and therefore in the reduction of post-stroke brain injury. Furthermore, a growing body of research describes the use of anti-inflammatory regulators such as IL-1Ra in clinical trials for ischemic stroke. In addition to the traditional immunosuppressive molecules, CD4⁺CD25⁺FoxP3⁺Tregs play a unique protective role in stroke pathogenesis. Treg-mediated immunoregulation following stroke should be a focus of future research. Finally, the cholinergic anti-inflammatory pathway provides an additional target for post-stroke pharmacologic intervention. Taken together, these findings (Table 1) lead us to conclude that immunoregulation may provide a promising approach both to the study of stroke pathophysiology, and to the discovery of treatments that limit the destructive effects of cerebrovascular disease.

The authors declare no conflict of interests.

This work was partially supported by the National Nature Science Foundation of China (no. 81871838 and 82072549); the Science and Technology Plan of Beijing Tongzhou District (KJ2020CX002); and the Beijing Tongzhou District Financial Fund.

Immunosuppression mechanisms studied in stroke