Background
Neuroinflammation caused by ischemic stroke comprises a pathogenic process, leading to a worsened local neurological dysfunction and metabolic disorders. This neuroinflammation may represent a therapeutic target in terms of ameliorating the effects of ischemic stroke [
1,
2]. Astrocytes, the most abundant cell type in the CNS, are key components of the neuroinflammatory response in concert with microglia [
3]. Activated astrocytes differentiate into reactive astrocytes in a process known as reactive astrogliosis, a ubiquitous but poorly understood hallmark of all CNS pathologies, including ischemic stroke. Reactive astrogliosis is characterized by the upregulation of glial fibrillary acidic protein (GFAP) and cellular hypertrophy [
4,
5]. Reactive astrocytes can produce either harmful or beneficial effects during CNS recovery following ischemic stroke, depending on their polarization subtype: anti-inflammatory phenotype or pro-inflammatory phenotype [
5,
6]. A better understanding of the molecular mechanisms underlying the transition between different phenotypes in reactive astrocytes may contribute to the development of novel therapies for improving neurological function after ischemic stroke.
In the pathological conditions underlying ischemic stroke, a severe delay exists in the oxygen and glucose delivery to the injured portion of the brain. Lactate produced through the breakdown of glucose and glycogen stored intracellularly in astrocytes accumulates in the surrounding tissues [
7]. Our previous research demonstrated that the tissue lactate content on the ipsilateral side in a rat model with middle cerebral artery occlusion (MCAO) was significantly increased during the early phases and that this increased lactate content acted on astrocytes, thereby causing astrogliosis and inducing a meaningful increase in axon guidance function [
8]. Moreover, several recent studies have similarly reported that, under hypoxic conditions, lactate acts as a signaling molecule that plays an inhibitory role in the inflammatory response by modulating polarization and transcription in macrophages [
9,
10]. However, the regulation of lactate in the neuroinflammatory response and immune regulation under conditions of cerebral ischemia has not been reported to date.
The
NDRG (
N-Myc downstream-regulated gene) family is one of the downstream regulatory genes of N-Myc (i.e., the N-Myc proto-oncogene protein), consisting of four family members (
NDRG1,
NDRG2,
NDRG3, and
NDRG4). All these members are highly conserved in structure, containing 57–65% of the same amino acid sequences [
11]. Among them,
NDRG2 is specifically expressed in astrocytes and plays an important role in the physiological and pathological processes in the CNS [
12‐
16]. Studies have shown that
NDRG2 plays an important role in regulating the polarization and inflammatory response in tumor-associated macrophages (TAMs). A lack of
NDRG2 in TAMs promotes M1 polarization, producing IL1, IL12, and TNFα, thereby inhibiting the tumor growth [
17,
18]. This result suggests that
NDRG2 may function as an important regulator in cells with immunomodulatory functions, such as astrocytes.
In this study, we employed a model of a rat with MCAO and found that the NDRG2 expression increased along with reactive astrogliosis. Moreover, we showed that the presence of lactate maintained the astrocytic NDRG2 stability by inhibiting its ubiquitination. Genetic inhibition of NDRG2 in astrocytes resulted in a strong increase in inflammation, upregulation of immune-related genes, and promotion of signaling pathways in vitro, with TNFα as the core. Additionally, we showed that the NDRG2 knockdown induced the TNFα expression and secretion by increasing c-Jun phosphorylation. Hence, this study provides new insights into targeting NDRG2 as a potential therapeutic and neuroinflammatory strategy for ameliorating the adverse effects of ischemic stroke.
Discussion
In this study, we evaluated NDRG2-mediated lactate signaling and its respective neuroprotective roles in astrocytes under OGD conditions. We found that high levels of lactate increased the expression levels of intracellular NDRG2. Additionally, we found that NDRG2 deficiency triggered the dysregulation of inflammatory and immunomodulatory activity and led to elevated levels of TNFα, thus leading to additional destruction of the local microenvironment.
Elucidating the mechanisms during early-stage of ischemic stroke injury is important to develop new strategies capable of minimizing detrimental side effects and improving patient outcomes following stroke. Previous studies in humans indicate that the lactate signal within the infarct area is initially significantly elevated and then decreases sharply. Finally, lactate levels appeared to normalize at 2 weeks [
21,
22]. Similarly, studies using rat MCAO models suggest that lactate levels are significantly higher than baseline within a few minutes after MCAO onset and return to normal at 48 h [
23‐
25]. Therefore, we speculate that lactate may play an important role, mainly in the early stages of the onset of ischemic stroke, especially in the first few hours. However, the accumulation of lactate may take longer to significantly affect the associated phenotype of astrocytes. In combination with our previous findings from model of a rat with MCAO, we demonstrated that
NDRG2 protein levels in ipsilateral tissue of MCAO models show an increasing trend over time, but the difference detected by western blot compared with the contralateral side is statistically significant, at least 8 h later, during which it is accompanied by astrocyte activation [
8]. The lactate content of the injured side is also significantly higher than that of the contralateral side at 8 h [
8]. Therefore, we believe that 8 h is the key time point for the detection of related indicators in vivo and in vitro, and, in the subsequent experiments in this manuscript, were uniformly detected at 8 h.
In a high-lactate microenvironment, lactate either mediates signaling via its receptor expressed on the cell surface or is transported within the cells where it mediates signal transduction. GPR81 (also named hydroxycarboxylic acid receptor 1 [HCAR1]), an orphan G-protein-coupled receptor, is a sensor of extracellular lactate and modulates cell functions during oxygen-deficient conditions [
26]. Moreover, it is also reported to suppress inflammation (even the production of TNFα and IL6) and is involved in the protection of the brain from the deleterious effects of ischemic stroke [
27,
28]. qRT–PCR analysis revealed the presence of
GPR81 mRNA in astrocytes, although at low levels compared to neurons and other tissues such as fat [
29]. Therefore, we cannot rule out that the effects of lactate may also involve the potential contribution it plays through GPR81 function. Further, our previous studies demonstrated that extracellular lactate might execute its role under OGD conditions by transport into astrocytes (mediated via MCT1) [
8]. In our current study, we showed that the
NDRG2 mRNA levels were genetically independent of lactate. In addition
, NDRG2 protein levels were upregulated by lactate treatment and decreased when the lactate uptake was suppressed by siMCT1. Therefore, we speculate that intercellular lactate plays a major role in regulating
NDRG2 expression at the translational level or post-translational level.
To date, direct protein targets of intercellular lactate have rarely been reported. However, a previous study identified that
NDRG3, another member of the
NDRG family, acts as a lactate sensor whose ubiquitination and degradation are inhibited by direct lactate binding through hydrogen bonds, thus triggering downstream kinase signaling and providing the genetic basis for lactate-induced hypoxia responses [
30]. Given the high degree of primary amino acid sequence and structural similarities among
NDRG family members, we also studied the effects of lactate on the ubiquitination of
NDRG2 [
31]. The present study demonstrates that lactate suppresses
NDRG2 ubiquitination, leading to the accumulation of intercellular
NDRG2; this process mostly occurs through direct binding (which mainly comprised hydrogen bonds). These results indicate that the existence of
NDRG2-mediated signaling is enhanced under OGD conditions along with lactate-induced astrogliosis. This occurs even in the energy-restricted state of cerebral ischemia since the NDRG2 protein, once stabilized by lactate binding, remains quite stable. However, recently, accumulating evidence has proven that, at high lactate levels, lactylation occurs on the lysine residues of cellular histone proteins (i.e., in macrophages and other cancer cells) [
32‐
34]. This new type of post-translational protein modification plays a crucial biological function during physiological and pathological conditions. Hence, we cannot exclude the possibility that lactate also exerts this function through the NDRG2 lactylation. Additional experiments are needed to explore the lactylation of NDRG2 and its resulting functions.
NDRG2 is emerging as a critical and promising neuroprotection target against cerebral ischemic injury through different mechanisms. A series of studies have suggested that NDRG2 played important roles in maintaining the integrity of the blood–brain barrier (BBB), alleviating brain edema, and inhibiting glutamate excitotoxicity after ischemic stroke [
15,
16,
35]. Our current study complementally confirms the importance of
NDRG2 in attenuating astrocytic neuroinflammation. Our RNA-seq demonstrated that
NDRG2 silencing leads to the upregulation of a large number of inflammatory factors and chemokines, as well as the upregulation of inflammation-associated signaling pathways. C3 and RT1-S3 (H2-T23 [histocompatibility 2, T legion locus 23]), which are markers of the A1 phenotype, are each associated with hub genes in the PPI network of upregulated genes after NDRG2 silencing [
36], in addition to a plethora of pro-inflammatory factors and chemokines, such as CXCL10, CXCl11, IL12a, and TNFα. These effects contribute to a neurotoxic microenvironment. Liddelow et al. previously treated mouse models with systemic injection of lipopolysaccharide (LPS) or induced MCAO to cause ischemia. These two treatments resulted in two different astrocytic phenotypes (termed A1 and A2), each associated with a specific subset of upregulated transcripts [
37]. Given that Sprague–Dawley rats and mice are closely related, we analyzed whether the
NDRG2 knockdown in rats affected the expression of the A1/A2-specific transcripts previously defined in mice. We found a relatively higher ratio of A1-specific transcripts (17.1%) than A2-specific transcripts (10%) in the upregulated genes after
NDRG2 silencing (Additional file
1: Fig. S4). Taken together, our results suggest that
NDRG2 silencing leads astrocytes to an A1-like polarization propensity, if not to a complete A1 subtype. Therefore, we conclude that elevated lactate levels under ischemic conditions prevent A1 subtype astrocyte formation and modulate a beneficial microenvironment to promote cell survival by maintaining
NDRG2 stability.
TNFα is a well-known pro-inflammatory cytokine that plays an important role in inflammatory disorders. Through enrichment analysis of signaling pathways, we identified a key “common signaling node”, namely TNFα, that was associated with all seven classifications of signaling pathways, identifying TNFα as a bridging gene. Therefore, in the current study, we focused on the regulation of TNFα as mediated by NDRG2. TNFα exerts its effects through either pro-death or pro-survival/inflammation-associated pathways [
38,
39]. Moreover, TNFα triggers these effects through two structurally related but functionally distinct receptors: the type 1 receptor (TNF-R1) (p55) and the type 2 receptor (TNF-R2) (p75). TNF-R1 contains a cytoplasmic death domain that recruits downstream signaling proteins through the adaptor protein TRADD and initiates subsequent processes including inflammatory, apoptotic, and degenerative cascade effects [
40,
41]. Our qRT–PCR assay showed that
NDRG2 silencing increased mRNA levels of TNFα, TNFR1 (TNFRSF1a), and TRADD. In addition,
NDRG2 knockdown promotes astrocyte apoptosis through TNFα (Additional file
1: Fig. S2). Our study thus provides evidence that
NDRG2 deficiency is not favorable for cell survival as it induces TNFα expression and secretion. Furthermore, TNF-R1 is distributed throughout the brain and is expressed in many types of cells, including microglia and neurons [
42,
43]. Astrocyte-derived TNFα has been reported to critically modulate microglial activation and M1-like polarization [
44,
45]. In addition, we confirmed by TUNEL assay that the cell supernatant of
NDRG2-silenced astrocytes could promote neuronal apoptosis. These findings suggest that
NDRG2 plays an important role in promoting cell survival and inhibiting inflammation following ischemia, thus contributing to a favorable microenvironment that is beneficial for neural restoration.
However, this study had several limitations. First, although we demonstrated the possibility that lactate inhibits NDRG2 ubiquitination through direct interaction and predicted that Lys176 is the key amino acid residue mediating this interaction, our experimental design did not lend itself to providing direct evidence supporting this putative mechanistic pathway. An in vitro binding experiment evaluating 14C-labeled L-lactate combined with NDRG2 would help clarify these findings. Second, we showed that NDRG2 knockdown upregulated inflammation-associated pathway activation, with TNFα as the core component. However, this finding does not signify that TNFα plays the most important role in mediating the mechanistic effects of NDRG2. Additional systematic experiments will be required to determine the exact functions of NDRG2 and TNFα in regulating astrocytic phenotypes and in the outcome of cerebral ischemia. Third, behavioral and histological data for NDRG2 knockout mice in MCAO models are lacking. Based on our experience, differences between animal models due to responses to NDRG2 might be offset by the effects of serious injury occurring after 8 h of MCAO. Thus, future studies are needed to acquire more thorough information on a range of animal models to evaluate the role of lactate and NDRG2 following acute insult in MCAO (i.e., 1 h or 2 h blocking) and reperfusion conditions.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.