Plasma Soluble ST2 Levels Are Higher in Neurodegenerative Disorders and Associated with Poorer Cognition

INTRODUCTION

Neurodegenerative disease is the 5th leading cause of mortality worldwide according to the World Health Organization and its incidence has been increasing exponentially in part due to the rapidly aging global population [1–3]. Early detection of neurodegenerative diseases facilitates earlier intervention, delaying disease progression [1]. Thus, it is crucial to explore possible diagnostic biomarkers for neurodegenerative diseases.

The interleukin-33 (IL-33)/Suppressor of tumorgenicity 2 (ST2) system is primarily involved in inflammatory conditions such as asthma, pulmonary fibrosis, and autoimmune diseases [4, 5]. ST2 and IL-33 are highly expressed in the central nervous system (CNS), where IL-33 is expressed in oligodendrocytes [6] and ST2 is expressed mainly in microglia and astrocytes [7]. ST2 is a member of the Toll-like/IL-1-receptor super-family. ST2 ligand (ST2L) binds to IL-33, which in turn induces the Th2 immune response, leading to polarization of microglia toward an anti-inflammatory M2 phenotype [6]. Several studies have suggested that the IL-33ST2 system may be implicated in neurodegenerative diseases [8, 9]. IL-33 can induce regulatory T cells (Treg) to produce anti-inflammatory cytokines, which have been reported to alleviate cognitive decline in an AD mouse model [10]. In contrast, soluble ST2 (sST2) acts as a decoy receptor for IL-33, attenuating the neuroprotective Th2 inflammatory response [5]. The imbalance between the TH1 and TH2/Treg can promote the production of pro-inflammatory cytokines and exacerbate neurodegeneration [11, 12].

ST2 is highly expressed throughout the hippocampus [13]. Neuronal IL-33 augments hippocampal synaptic plasticity and induces microglial phagocytosis of the extracellular matrix, which regulates synapse remodeling and memory formation [14]. Blockade of IL-33/ST2 signaling in vivo by intracerebroventricular administration of sST2 inhibits homeostatic synaptic plasticity in hippocampal CA1 neurons and impairs spatial memory formation in mice [15]. The binding of IL-33 to ST2L triggers downstream signaling pathways that result in clearing of Aβ plaques and rescue of the AD phenotype [5, 16]. In mouse AD models, high sST2 was associated with amyloid plaque accumulation, with administration of IL-33 showing rescue of AD phenotype and improvement in cognitive performance by microglia/macrophages [8, 17].

Single nucleotide polymorphisms (SNPs) within IL-33 are associated with Alzheimer’s disease (AD) risk in Caucasian and Han Chinese. Patients with mild cognitive impairment (MCI) and AD showed lower IL-33 and higher sST2 levels compared to healthy controls [18]. Patients with higher IL-33 expressing cells preserved their cognitive function and hippocampal volume, compared with IL-33 non-expressing patients [19, 20]. However, whether peripheral sST2 levels correlates with cognitive function, disease state and/or severity remains unclear.

To address this gap, this study aimed to investigate plasma levels of sST2 across neurodegenerative diseases compared to healthy controls (HC) and to assess the association between sST2 levels and cognitive and motor scores.

MATERIALS AND METHODS

RESULTS

Plasma sST2 levels in neurodegenerative diseases and healthy controls

Plasma sST2 levels were significantly different across all neurodegenerative diseases (p < 0.001; Fig. 1A), controlling for age and gender. Within the dementia cohorts, post-hoc analysis revealed higher sST2 levels in AD (p = 0.002, Bonferroni corrected p = 0.012) and FTD (p < 0.001, corrected p < 0.001) versus HC. Higher sST2 levels were also seen in FTD versus MCI (p = 0.004, corrected p = 0.026), controlling for age and gender. There was no significant difference between AD versus FTD (p = 0.217, corrected p = 1.0). Within the PD cohort, plasma sST2 levels were significantly higher in PD versus HC (p = 0.041; Fig. 1B), controlling for age and gender. When the two cohorts were analyzed, higher sST2 levels were observed in AD versus PD (p = 0.034, Bonferroni corrected p = 0.338) and FTD versus PD (p = 0.001, Bonferroni corrected p = 0.007), correcting for age and gender. In a subset of 22 patients (6 MCI, 8 AD, and 8 FTD) with available CSF samples, plasma sST2 showed significant correlation with CSF sST2 levels (R-sq = 0.684, p < 0.001; Fig. 2).

OPEN IN VIEWER

Fig. 1

Plasma sST2 levels across disease groups A) In the Dementia cohort; B) In the early PD cohort ^*p < 0.05, ^**p < 0.01, p values are adjusted for age, gender, and multiple comparisons via Bonferroni.

OPEN IN VIEWER

Fig. 2

Plasma and CSF sST2 levels.

DISCUSSION

In this study, we report for the first time that plasma sST2 levels are higher across various neurodegenerative diseases including AD, FTD, and PD compared to MCI and HC. Additionally, higher plasma sST2 levels are associated with poorer cognitive scores in the dementia and PD groups, especially in the cognitive domains of working memory, executive and visuospatial function. This presents sST2 as a candidate biomarker not only of disease state but also cognitive dysfunction across the most common neurodegenerative diseases.

The involvement of immunomodulatory cascades in the pathophysiology of neurodegenerative diseases, particularly AD, have been widely established in literature [34, 35]. In recent years, studies have emerged revealing the IL-33/ST2 pathway as an important component of the immunomodulatory cascade [13, 36–39], including in neuroinflammation and AD [4, 40]. Three single nucleotide polymorphisms (SNPs), rs1157505, rs11792633, and rs7044343, within IL-33 have also been associated with AD risk in a Caucasian cohort [40]. These SNPs were further evaluated in Han Chinese, with the IL-33 rs11792633 polymorphism significantly associated with a reduced risk of late onset AD (LOAD) in patients and the T allele as a protective factor for LOAD [41, 42].

There are two isoforms of ST2, namely transmembrane (ST2L) and soluble (sST2) [5]. ST2L is a receptor for IL-33 [43]. IL-33 binds to ST2L, triggering downstream signaling pathways that result in clearing of Aβ plaques and rescue of the AD phenotype [5]. On the other hand, sST2 is a decoy receptor for IL-33 as it binds competitively to IL-33, preventing the IL-33/ST2L interaction, attenuating neuroprotective effects of IL-33 and resulting in an accumulation of misfolded proteins [5] and potentiation of the AD phenotype [8]. In mouse AD models, higher levels of sST2 were associated with accumulation of amyloid plaques and potentiation of AD, with administration of IL-33 showing rescue of AD phenotype and improvement in cognitive performance by microglia/macrophages [8, 17]. Another study observed that mice lacking IL-33 developed AD pathology with tau abnormality and neurodegeneration in the cerebral cortex and hippocampus, accompanied with cognitive impairment [44].

A more subtle form of neuroinflammation is also present in MCI, defined as having an objective decline in cognitive performance not meeting criteria for a dementia diagnosis [22]. In MCI pathophysiology, studies have demonstrated the involvement of the IL-33/ST2 pathway [18] and illustrated its effect on cognition [19] where patients expressing IL-33 preserved their cognitive function compared with IL-33 non-expressing patients [19]. Another study showed that MCI patients who did not convert to AD had higher IL-33⁺ cells correlating with better preserved hippocampal volumes compared to AD converters [20]. IL-33 has been found to be significantly decreased in the serum and CSF of AD and MCI cases, while sST2 was increased in the serum of AD and MCI cases [18], supporting the hypothesis that IL-33 plays a complex anti-inflammatory role that is attenuated in AD- and MCI-associated neuroinflammation. Decreased circulating IL-33 levels in AD might be attributed to the increased decoy receptor sST2 in the blood of AD patients [18]. Additionally, the IL33/ST2 pathway has also been implicated in behavioral changes in both human [45] and murine [46] models. Mice lacking in IL-33 were found to develop several behavioral deficits including impaired social recognition [46]. In humans, IL-33 levels were affected in affective disorders such as depression and bipolar disorder [45]. Highest levels of sST2 were seen in our FTD cases, most of whom were of the behavioral variant FTD subtype, presenting with prominent behavioral and neuropsychiatric impairment.

Our study is the first to explore the effect of the IL-33/ST2 axis in PD, and we demonstrated a trend towards higher sST2 levels in PD patients with MCI. Plasma sST2 was also associated with worse attention domain at baseline and decline in global cognition, visuospatial and executive functions over time, which are domains known to be affected in PD [47–49]. PD is characterized by an accumulation of Lewy bodies in the brain parenchyma and degeneration of midbrain dopaminergic neurons [50]. PD murine models have shown higher levels of IL-33 in astrocytes in the midbrain and striatum compared to age-matched controls [51, 52], but the corresponding levels of sST2, were not explored in that study. In vitro studies have shown that 1-methyl-4-phenylpyridinium (MPP+), a metabolite of the parkinsonian neurotoxin 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP), induces IL-33 release from astrocytes [52]. A high level of plasma sST2 would act as a decoy receptor and mitigate the effects of the rise in IL-33, inhibiting the IL-33/ST2 axis and possibly resulting in the motor and cognitive symptoms characteristic of the PD phenotype [51]. Therefore, while IL-33 administration could be a potential therapy for AD, the role of IL-33 in PD and PD-MCI remains to be further investigated.

The neuroprotective role of IL-33 has been supported by numerous preclinical studies. Our study demonstrating higher levels of sST2 across various neurodegenerative diseases adds to existing evidence that the IL33/ST2 pathway is implicated in their pathogenesis and may yield new therapeutic options for AD, FTD and PD. Strengths of this study include the fact that this is the first report of sST2 as a potential biomarker in FTD and PD. Additionally, we were able to show significant correlation of plasma with CSF sST2 levels in patients who had corresponding CSF sST2 levels measured. This is important given the more accessible nature of blood compared to CSF biomarkers.