Modulation of Neuroinflammation: Advances in Roles and Mechanisms of the IL-33/ST2 Axis Involved in Ischemic Stroke

Ischemic stroke is an acute cerebrovascular disease that is widely recognized as a leading cause of disability and mortality globally. Following an ischemic stroke, a severe reduction in blood supply gives rise to an inadequate supply of oxygen, ultimately leading to the death of neurons. Rapid reperfusion is the clinical common treatment by intravenous thrombolysis or endovascular thrombectomy; however, it has the limited effects for the functional recovery [1]. Further exploration of the underlying mechanisms and the effective treatment strategy in ischemic stroke is urgently needed.

The onset of stroke can initiate an inflammatory cascade in both the systemic immune system and central nervous system (CNS). This cascade plays a crucial role in the progression of cerebral ischemia pathology [2]. Various immune cytokines are involved in this complex neuro-immune crosstalk, such as interleukin (IL)-33 and its receptor ST2 [3]. IL-33 is a multifunctional immunomodulatory cytokine, which may mediate Th2 immune responses via the receptor complex consisting of full-length transmembrane form ST2L and IL-1 receptor accessory protein [4, 5]. As a decoy receptor, soluble ST2 (sST2) can bind to IL-33 to prevent its signaling through ST2L [6]. IL-33 is expressed at a high level in the brains and spinal cords of both humans and rodents, suggesting its specific functions within CNS [4, 7]. Recent research studies have shown that the IL-33/ST2 pathway is involved in neuroinflammation from a variety of CNS disorders, including ischemic stroke, neurodegenerative diseases, infection, tumorigenesis, and injury [3, 8]. For ischemic stroke, the protective effects of the IL-33/ST2 pathway have been gradually reported, and sST2 has also been identified as a potential inflammatory biomarker [9‒13]. This review aimed to highlight the current critical findings for the IL-33/ST2 axis in neuroinflammation, especially focusing on its mechanisms involved in cerebral ischemia. Additionally, specific targeting sST2 for the treatment strategy in ischemic stroke was also discussed.

Via the Il-33-LacZ Gt reporter strain (Il-33 Gt/Gt), IL-33 was demonstrated to be highly expressed in various mouse epithelial barrier tissues (such as the lung, skin, and vagina), lymphoid organs (including lymph nodes and spleen), brain, embryos, and inflamed tissues [14]. IL-33 expression can also be detected in the CNS during post-embryogenesis [15]. Previous studies suggested that astrocytes and oligodendrocytes were responsible for producing the mRNA and protein of IL-33, rather than neurons and microglia [7, 15, 16]. Nevertheless, the inconsistent results were shown in experimental autoimmune encephalomyelitis and subarachnoid hemorrhage that IL-33 protein was expressed both in neuron cells and astrocytes but not in microglia [17, 18]. In the brains of individuals with multiple sclerosis, the protein IL-33 is expressed in various CNS resident cells, including neurons, oligodendrocytes, astrocytes, and microglia cells [19]. Whether the discrepancy is related to experimental methods remains to be further investigated. However, the extensive expression of IL-33 in brain tissues and neural cells has revealed its specific functions in CNS and CNS disorders. In the ischemic brain, the robust expression of IL-33 was found to be produced by glial cells, particularly by oligodendrocytes, suggesting glia cells as the major players in mobilizing and recruiting peripheral leukocytes into the brain via the damaged blood-brain barrier (BBB) [20‒22]. Also, peripheral lymphoid organs and immune cells can also produce IL-33, which may play at least partly immunoregulatory role for ischemic brain injury [14]. And systemic administration of recombinant IL-33 has been reported to be neuroprotective in the animal model of cerebral ischemia [23, 24]. The exact source of IL-33 in ischemic stroke merits further elucidation.

The ST2 protein can be also generated by different types of cells, such as macrophages, T cells, mast cells, and neural cells [8]. The IL-33 receptor components, ST2L and IL-1RAcP, were found to be expressed in glia, especially astrocytes and microglia, but neurons only express IL-1RAcP, which means that the earliest responders to IL-33 may be astrocytes and microglia [4]. A recent study has suggested that the ST2 protein expression can be found in neurons from mouse spinal cord and a mouse model of sciatic nerve injury [7, 25]. However, the expression of the membrane ST2 differs from the soluble variant of ST2, and it is still unclear where the circulating sST2 in patients with different diseases or normal healthy subjects comes from. In cultured astrocytes, the increased expression and release of IL-33 can be induced by the TLR ligand lipopolysaccharide, then stimulating microglia to secrete a series of inflammatory cytokines for a feedback regulation [7]. Recent studies have shown that IL-33 can enhance the phagocytosis of microglia, leading to the production of chemokines and anti-inflammatory cytokines [16]. The expression of IL-33 and ST2 in the CNS appears to vary depending on physiological and pathological conditions, indicating the diverse effects and functions of the IL-33/ST2 axis in the CNS.

Neuroinflammation is closely associated with numerous neurological diseases, such as degenerative diseases, CNS infections, tumorigenesis, brain injury, and ischemic stroke [26]. As an important neuroinflammatory factor, IL-33 and its pathway have pro- or anti-inflammatory effects in the progression of neuroinflammation-related diseases, with dual roles (Fig. 1).

Inflammation-mediated synaptic plasticity, autophagy, and apoptosis are involved in the pathogenesis of cognitive dysfunction. In animal models of multiple sclerosis and Alzheimer’s disease (AD), moderate levels of IL-33 can promote the polarization of microglia and Th cells toward anti-inflammatory phenotypes as microglia may effectively phagocytose harmful substances, leading to a reduction in nerve damage and disease progression [17, 27, 28]. However, IL-33 may be upregulated by the sustained neuroinflammation to cause the dysregulation of synaptic plasticity, autophagy, and apoptosis, ultimately leading to cognitive decline [29]. It has been speculated that the complicated influence of IL-33 on cognitive impairment may depend on the concentration of IL-33, the degree of neuroinflammation, and differential binding to the different ST2 isoforms [26].

The IL-33/ST2 pathway also plays a critical role in neuroinflammation induced by CNS infections [30]. Take neuro-AIDS as an example, the different HIV-1 clades may stimulate the varying levels of IL-33/ST2L expression. High concentration of IL-33 can decrease the expression of myocyte enhancer factor 2C (MEF2C), which regulates synaptic function, promote apoptosis, and upregulate the gene expression (NOD2, SLC11A1) associated with the initiation of immune response in CNS cells, then leading to HIV-induced neuropathogenesis [31, 32]. In the animal model of viral encephalitis caused by the Rocio virus, the role of the IL-33/ST2 axis appears to be inverse. The absence of IL-33/ST2 signaling may result in increased neuroinflammation due to the production of IFN-γ, which in turn leads to NO production and tissue injury [33]. IL-33 administration or blockade of ST2 may supply the new therapy to viral encephalitis. Similarly, IL-33/ST2 signaling was thought to be critical in preventing the development of encephalitis in mice infected with T. gondii [34]. Using the Plasmodium berghei model of infection, IL-33 can prevent the development of cerebral malaria by coordinating a protective immune response via type-2 innate lymphoid cells (ILC2), M2 macrophages, and regulatory T cells (Tregs) [35]. However, in another cerebral malaria model caused by Plasmodium falciparum infection, IL-33/ST2 signaling was identified to be involved in neurocognitive impairments through glial cells, which may exacerbate neuroinflammation and change neurogenesis [36]. The different types of pathogenic infection may trigger the different mechanisms of inflammatory responses, and the related mechanisms of IL-33/ST2 among CNS cells may be distinctive.

Moreover, numerous studies have identified that IL-33 has a dual function as a coordinator of the glioblastoma microenvironment, which ultimately contributes to tumorigenesis [37‒40]. Its nuclear and secreted functions may promote chemokine recruitment and activate innate immune cells, establishing a pro-tumorigenic environment in glioblastoma [38]. Specialized tissue Treg cells have been shown to express α-chain of ST2, which suggests that the IL-33/ST2 axis plays a significant role in the accumulation of Treg cells in the tumor microenvironment, finally leading to the suppression of antitumor immunity [41, 42]. However, the relationship between Treg effects and IL-33/ST2 signaling in glioblastoma has not yet been confirmed. A previous study found that ST2 protein can suppress anchorage-independent growth of a glioblastoma cell line, T98G, suggesting its negative effect on malignancy [43]. Whether the ST2 expression in different cells determines the different effects of the IL-33/ST2 axis in CNS tumorigenesis needs to be further elucidated.

Neuroinflammation is also one of the essential mechanisms of brain tissue damage. The increasing evidence have revealed the protective role of the IL-33/ST2 pathway in brain injury, including ischemic stroke, traumatic brain injury, and neonatal hypoxic-ischemic brain injury [44‒46]. Reversely, in subarachnoid hemorrhage, IL-33 seems to be pro-inflammatory via conjuncting with IL-1β [18, 47]. It should be concerned that there may be a functional transition in IL-33/ST2 signaling during the different phase after CNS injury [46]. For the most common CNS injury in clinic, the expression profile, potential mechanisms, and therapeutic prospects of the IL-33/ST2 pathway in ischemic stroke would be highlighted in the following content.

The initial data linking IL-33 to ischemic stroke was from a case-control study in 2013 in the north Chinese population through the single nucleotide polymorphism assay, finding that the genetic variation of rs4742170 in the IL33 gene was a moderate protective factor for ischemic stroke [48]. Subsequently, another study showed that serum IL-33 concentration was remarkedly increased with the infarction volume in patients with acute cerebral infarction, suggesting that IL-33 was involved in the pathogenesis and progression of cerebral ischemia, but the beneficial or detrimental role of IL-33 was still confusing [49]. Our laboratory reported in 2015 that IL-33 may ameliorate ischemic brain injury in the mouse model of middle cerebral artery occlusion (MCAO), first verifying its protective role in ischemic stroke. In the ischemic brain tissues, we also found that the mRNA and protein levels of IL-33 were significantly decreased [12]. Then, numerous studies have further clarified the protective role of the IL-33/ST2 pathway and revealed the underlying mechanisms in ischemic stroke [9‒11]. Alleles of four single nucleotide polymorphisms within the IL-33/ST2 axis (rs10435816, rs7025417, rs11792633, rs7044343) were also demonstrated to be associated with a decreased risk of large-artery atherosclerosis stroke [50].

The immunomodulatory mechanism of the IL-33/ST2 axis in alleviating ischemic brain injury has been gradually recognized. The onset of cerebral ischemia induces neurotoxic Th1-type response. In the mouse MCAO model, IL-33 administration can promote the transfer of T cell to the protective Th2-type, the polarization of macrophage from M1-type to M2-type, and increase the secretion of anti-inflammatory cytokines such as IL-4, similar to the mechanism of IL-33 in atherosclerosis [11, 12, 51]. Given this classical function in immunity, it is not surprising for regulating Th1/Th2 balance of the IL-33/ST2 pathway in ischemic stroke. As a nuclear factor, IL-33 has also been reported to inhibit Th17-type immune response in experimental cerebral ischemia by regulating gene transcription [12, 23]. Recently, electroacupuncture combined with human stem cell-derived small extracellular vesicles suppressed Th1 and Th17 responses by regulating the IL-33/ST2 signaling, showing neuroprotective effects on cerebral ischemia in the mice model of stroke [52].

Recent research studies also showed that IL-33 may be released from CNS cells quickly after ischemic brain injury, which contributes to activate the immune responses in lesion areas and affect the functions of CNS cells [53]. During the acute phase after stroke, IL-33 may reduce the astrocytic activation in the peri-ischemic area, and IL-33-treated human T-cells can increase the release of IL-4 and reduce the secretion of IL-6 from astrocytes, playing a protective role [11]. To the subacute phase after stroke, the metabolic shift is one of the important features of reactive astrocytes, and inhibition of lipogenesis in astrocytes can decrease IL-33 production, deteriorate BBB damage, and interfere with long-term functional recovery [54, 55]. As such, IL-33 supplementation in this “sensitive period” of poststroke showed a rather different mechanism for astrocytic function. Moreover, microglia are the first line of defense against ischemic brain injury, and they differentiate into a range of phenotypes, playing different roles in different stages after stroke. Among them, the activated neurotoxic M1 and the alternatively activated beneficial M2 phenotypes are the most extensively studied [56]. Recent studies have suggested that IL-33/ST2 engagement can stimulate microglia to produce IL-10, thus enhancing neuronal survival upon ischemic conditions. ST2 deficiency can shift microglia toward a M1 phenotype, showing the importance of IL-33/ST2-dependent microglial response in cerebral ischemia [9, 10]. Celastrol, a traditional oriental medicine, was recently found to be protective against ischemic stroke by promoting M2 polarization of microglia mediated by the IL-33/ST2 pathway [57]. Recent single-cell RNA sequencing analysis defined at least 9 subpopulations of microglia localized in the mouse brain with unique markers. Among that, several clusters may be the major source of inflammatory signals in the brain, such as the Ccl4+ subpopulation, cluster 8/OA2/IR2.2/IR2.3, and so on [58]. Precise targeting of Ccl4+ subpopulation in disease may provide a safe and effective way to maintain the beneficial functions and limit the side effects of microglial activation [58, 59]. Additionally, single-cell RNA sequencing analysis in AD mouse models revealed the AD-acquired disease-associated microglia or activated-response microglia, and a subpopulation of IL-33-responsive microglia (IL-33RM) showed the enhanced phagocytic activity and Aβ clearance [60, 61]. It seems that whether ischemic stroke could induce nontraditional disease-associated microglia activation and IL-33 could affect these unique subpopulations would be a meaningful research in the future.

White matter damages are also critical components of poststroke brain lesion, including the death of oligodendrocyte precursor cells and myelin-producing oligodendrocytes, which can lead to long-term disability [62]. A recent study identified signal transducer and activator of transcription 6 (STAT6) as a key molecule, which can mediate the protective effect of IL-33/ST2 on ischemic brain oligodendrocytes, suggesting that the IL-33/ST2/STAT6 signaling is involved in improving white matter integrity and long-term neurological functions after ischemic stroke [63]. Furthermore, the anti-apoptotic effect of IL-33 took part in the mechanism of cerebral ischemia. Using a brain hypoxic-ischemic model in neonatal mice, IL-33 treatment apparently inhibited astrocyte apoptosis by inhibiting PUMA, a transcriptional target of p53, which promotes ischemia/reperfusion-induced astrocytic apoptosis [45, 64]. For this neuroprotective effect, IL-33/ST2 signaling was found to be essential in vitro [45]. Under hypoxic conditions, the abundance of apoptosis and the apoptosis-related proteins that cleaved caspase-3 and BAX were significantly reduced in neurons treated with IL-33, while the anti-apoptotic protein Bcl-2 expression was increased remarkably [65].

It should be concerned that systemic changes may also occur in conjunction with brain inflammation during cerebral ischemia [24]. The spleen is the main mediator of immune response to ischemic injury in all peripheral lymphoid organs examined, and splenectomy prior to MCAO surgery is neuroprotective and significantly reduces neuroinflammation [66, 67]. In the recent research studies, peripheral administration of IL-33 was shown to increase the secretion of IL-4 in the spleen, and the protective mechanism of IL-33 in cerebral ischemia may be related to the regulation of splenic T-cell immune responses by inhibiting Th1 response and promoting Treg response [11, 23]. In the ischemic brain, the elevated expression of CXCL16, as a chemoattractant for activated CD8⁺ T cells and NKT cells, may be attenuated by IL-33, contributing to reduce damage of lymphoid cells in the ischemic brain [24, 68]. Despite its beneficial effects, IL-33 administration in the acute phase after stroke can also exacerbate systemic immunosuppression, leading to bacterial infection in the lungs, due to an accelerated systemic switch from Th1 to Th2 response [24]. Thus, the mechanisms and effects of the IL-33/ST2 pathway in ischemic stroke were pleiotropic and complex, closely linked to immune response, CNS functions in different nerve cells, apoptosis, and peripheral regulation (Fig. 2).

Treg cells play a crucial role in maintaining immune homeostasis and have been characterized as protective cells in primary CNS inflammatory diseases, such as ischemic stroke [69]. Recent advances have further clarified tissue-resident Tregs with tissue-specific functions [70]. Among that, brain Treg cells that expressed unique genes associated with the nervous system have been recently identified in the brain tissues of humans and rodents [71, 72]. During neuroinflammation, brain Treg cells may regulate astrogliosis through producing amphiregulin (AREG), polarizing microglia into a neuroprotective state and limiting inflammatory responses by releasing IL-10, providing therapeutic opportunities for neuronal protection against stroke [72].

Several studies have shown that Treg cells mediate the immunomodulatory effects of IL-33, which induces ST2-dependent proliferation and accumulation of Foxp3⁺ Tregs in multiple organs [73, 74]. After ischemic stroke, IL-33 can increase the proportion of Tregs in the ischemic brain and spleen, and its neuroprotective effect is related to the reduction of neuronal apoptosis-associated proteins and the production of Treg-associated cytokines [23, 65]. In the previous study, we found that IL-33 has the ability to activate Foxp3 via ST2, which can increase the proportion of Treg in the ischemic brain. The elevated Treg cells then produce AREG, which activates epidermal growth factor receptor (EGFR) located in neurons. This process is beneficial in improving the prognosis [75]. Maybe IL-33 can be used as a promising immune modulatory agent by enhancing protective ST2-dependent Tregs for the treatment of stroke. Nevertheless, identifying the brain tissue-specific Tregs and circulatory Tregs as the main effector of IL-33 was still difficult, and whether brain Tregs took part in the regulatory functions of IL-33 for astroglial activation and microglial polarization remains to be elucidated (Fig. 3).

Due to the significant role of neuroinflammatory pathophysiology of ischemic stroke, anti-inflammatory or pro-inflammatory cytokines biomarkers can be used as early diagnostic and prognostic indicators of stroke. The initial finding in patients with acute ischemic stroke (AIS) showed that the level of serum IL-33 was significantly increased with the infarction volume [49]. However, the subsequent studies suggested that the lower serum IL-33 level was related to the larger infarct size, higher stroke severity, and poorer outcome [76, 77]. For the poststroke depression patients, serum IL-33 was an independent predictor and protective prognosis factor [78]. In a study for hemorrhage transformation (HT) in AIS patients, the serum IL-33 concentration was higher than that in healthy subjects, while the lower IL-33 levels were associated with HT and the unfavorable outcomes [79]. The conflicting findings may depend on the different study groups, sample sizes, and testing methods. In order to verify the value of IL-33 as an AIS biomarker, additional larger studies from multiple centers are needed.

Compared to IL-33, its decoy receptor sST2 as a hazardous biomarker in AIS from the different research groups seems to be consistent (Table 1). It has been demonstrated that the elevated serum sST2 level was related with poorer 90-day prognosis, higher mortality, and increased risk of HT in AIS patients [80]. The previous results from our laboratory have also shown that the serum sST2 levels in AIS patients were apparently higher than those in healthy individuals and increased with the infarct volume and the National Institutes of Health Stroke Scale (NIHSS) score [13]. Data from the mechanical reperfusion therapy in AIS patients revealed a prognostic value of the sST2 level at 24 h, which was associated with an increased risk to adverse clinical events [81]. Recently, sST2 was suggested to be a useful biomarker for grading the severity of cerebral-cardiac syndrome in patients with AIS [82]. In patients with transient ischemic attack/ischemic stroke, the serum sST2 level was identified as a potential long-term prognostic biomarker, and higher serum sST2 was associated with increased risk of a poor outcome within 90 days and 1 year [83]. Moreover, several studies found that elevated concentrations of plasma sST2 in ischemic stroke patients were significantly related with the increased risk of poststroke cognitive impairment and depression, independent of conventional risk factors [84, 85]. As an increasing recognized marker for myocardial injury, whether sST2 has high specificity as a promising biomarker used for AIS needs further verification. Of note, the changes of IL-33 and sST2 in the cerebrospinal fluid from AIS patients have not yet been reported.

The IL-33/ST2 signaling pathway may serve as a viable new target for the treatment of ischemic stroke. Given the broad role of IL-33 in immune responses in multiple tissues and organs, the optimal strategy to suppress IL-33-related neuroinflammation in ischemic stroke without side effects requires further investigation [8]. Specific targeting sST2 may be a more rational choice to attenuate proinflammatory responses and promote CNS repair after stroke [46]. The crystal structures of IL-33 and its receptor complexes have been solved, providing powerful information for blocking the binding of sST2 to IL-33 [86]. The use of Tat-mediated cell-permeable peptide delivery across the BBB that interferes with the protein-protein interaction has been instrumental in preclinical studies for neurological diseases [87‒89]. Thus, the design of the sST2 Tat peptide based on the binding site between sST2 and IL-33 may have potential to be developed as a neuroprotective agent for treating ischemic stroke. Aptamers, which are small, non-immunogenic nucleic acid molecules that can be easily chemically modified and cross the BBB, may offer potential therapeutic and diagnostic applications in neuroscience [90, 91]. Maybe, ssDNA/RNA aptamers against sST2 would be an available approach to detect sST2 concentrations and interfere with IL-33/ST2 effect in future research.

In summary, recent studies have highlighted the relationship between the IL-33/ST2 axis and neuroinflammation, with anti- or pro-inflammatory effects due to the different CNS diseases. For ischemic stroke, the IL-33/ST2 pathway tends to play the protective effect to attenuate brain damage, and the mechanisms were pleiotropic, involved in the immune response modulation, CNS function regulation in different nerve cells, apoptosis, and peripheral regulation. Notably, the novel IL-33/ST2/Treg/AREG/EGFR signaling merits to be further investigated as a immunomodulatory target for the treatment of cerebral ischemia. Given the blocking effect of sST2 on the protective IL-33/ST2 axis, the design of specific targeting sST2 based on the binding site between sST2 and IL-33 may be a rational promising strategy. Clinical trials about the IL-33/ST2 signaling and related biomarkers in ischemic stroke are still lacking. Elevated serum ST2 levels are associated with disease severity in patients with ischemic stroke, while serum IL-33 levels are negatively correlated with infarction volume and stroke severity in these patients. Undoubtedly, the combined measurement of IL-33 and sST2 in peripheral circulation or cerebrospinal fluid may provide a wider insight for the diagnosis and prognosis of ischemic stroke.

The authors declare that they have no conflict of interest.

This study was supported by the National Natural Science Foundation of China (No. 82071324) and the National College Students’ Innovation and Entrepreneurship Traning Program of Wuhan University (202210486120).

All the authors contributed to the review, conception, and design. The literature search and data analysis were performed by Shuang Guo, Junlong Cai, and Wenfeng Li. The first draft of the manuscript was written by Shuang Guo and Chengli Qian. Yi Luo, Junlong Cai, and Zhikun Zeng critically revised the work. All the authors read and approved the final manuscript.

Shuang Guo and Chengli Qian contributed equally to this work.