N-myc Downstream-Regulated Gene 2 (Ndrg2): A Critical Mediator of Estrogen-Induced Neuroprotection Against Cerebral Ischemic Injury

Growing evidence indicates that estrogen plays a pivotal role in neuroprotection against cerebral ischemia, but the molecular mechanism of this protection is still elusive. N‐myc downstream‐regulated gene 2 (Ndrg2), an estrogen-targeted gene, has been shown to exert neuroprotective effects against cerebral ischemia in male mice. However, the role of Ndrg2 in the neuroprotective effect of estrogen remains unknown. In this study, we first detected NDRG2 expression levels in the cortex and striatum in both female and male mice with western blot analyses. We then detected cerebral ischemic injury by constructing middle cerebral artery occlusion and reperfusion (MCAO-R) models in Ndrg2 knockout or conditional knockdown female mice. We further implemented estrogen, ERα, or ERβ agonist replacement in the ovariectomized (OVX) Ndrg2 knockout or conditional knockdown female mice, then tested for NDRG2 expression, glial fibrillary acidic protein (GFAP) expression, and extent of cerebral ischemic injury. We found that NDRG2 expression was significantly higher in female than in male mice in both the cortex and striatum. Ndrg2 knockouts and conditional knockdowns showed significantly aggravated cerebral ischemic injury in female mice. Estrogen and ERβ replacement treatment (DPN) led to NDRG2 upregulation in both the cortex and striatum of OVX mice. Estrogen and DPN also led to GFAP upregulation in OVX mice. However, the effect of estrogen and DPN in activating astrocytes was lost in Ndrg2 knockout OVX mice and primary cultured astrocytes, but partially retained in conditional knockdown OVX mice. Most importantly, we found that the neuroprotective effects of E2 and DPN against cerebral ischemic injury were lost in Ndrg2 knockout OVX mice but partially retained in conditional knockdown OVX mice. These findings demonstrate that estrogen alleviated cerebral ischemic injury via ERβ upregulation of Ndrg2, which could activate astrocytes, indicating that Ndrg2 is a critical mediator of E2-induced neuroprotection against cerebral ischemic injury.


Introduction
Ischemic stroke is one of the leading causes of death and disability worldwide and is widely accepted to be sexually dimorphic [1]. A number of both clinical trials and basic research studies have shown that the incidence of stroke is higher and stroke outcomes are worse in men compared to those in women [2][3][4]. Women's relative protection against strokes can be explained in part by serum levels of the neuroprotective ovarian hormone 17β-estradiol (estradiol or E2) [5]. Our previous studies demonstrated that estrogen exerted significantly neuroprotective effects against cerebral ischemia in female mice through binding to estrogen receptors α and β (ERα and ERβ) [6][7][8]. However, the underlying mechanisms of estrogen neuroprotection in female mice remain unclear.
N-Myc downstream-regulated gene 2 (Ndrg2), which plays vital roles in cell proliferation, differentiation, apoptosis, and stress responses [9], is widely expressed throughout the brain and is specifically expressed in astrocytes [10,11]. We previously first found that Ndrg2-knockout male mice had much larger cerebral infarction volumes compared with wild-type male mice [12]. We further found that Ndrg2 knockout male mice aggravate injury due to brain edema following cerebral ischemia [13]. These results demonstrate that Ndrg2 could be a potential target for the neuroprotection against cerebral ischemia in male mice. However, the role of Ndrg2 in estrogen neuroprotection in female mice has not been reported.
In a previous study, we first found that estrogen and estrogen receptor β (ERβ) agonist DPN could upregulate NDRG2 mRNA and protein expression in the mouse hippocampus, demonstrating that astrocytic Ndrg2 is a target gene of estrogen and ERβ [14]. The present study aims to explore a previously unstudied area: the role of Ndrg2 in estrogen-induced neuroprotection against cerebral ischemic injury in female mice and the underlying molecular mechanism.

Animals
Thirteen male and twenty-eight female C57BL/6 mice aged 8 months old were provided by the Laboratory Animal Center of Chinese PLA General Hospital. Sixty-four Ndrg2 +/+ , sixty-four Ndrg2 −/− , and one hundred and four Ndrg2 flox/flox female mice were constructed by the Shanghai Model Organisms, China. Ndrg2 flox/flox mice were crossed with B6.C-Tg(CMV-cre)1Cgn/J mice (Jackson Labs) to obtain Ndrg2 −/− mice. The strain was backcrossed to C57BL/6 J more than 20 times. The pAAV-CAG-Cre-3flag virus (2 μL, Hanbio, China) was injected into the right lateral cerebral ventricle of Ndrg2 flox/flox mice to conditionally knock out the expression of NDRG2 in astrocytes (referred to as AAV-Ndrg2). All of the animals were maintained under the following standard conditions: 12/12 h light/dark cycle, 50-60% environmental humidity, 25 ± 1 °C, and free access to food and water. The experimental protocols were reviewed and approved by the Ethics Committee of the Chinese PLA General Hospital, China. The schematic diagram of the detailed animal experiment design and group is shown in Fig. 1.

OVX and Hormone Replacement Treatment
Ovariectomy (OVX) was performed by dorsolateral incisions as previously described [15]. After 1 week, the ovariectomized female mice then received daily subcutaneous injections of 0.1 ml sesame oil, E2 (100 μg/kg/day), estrogen receptor α agonist (PPT) (2 mg/kg/day), or DPN (8 mg/kg/ day) for 3 weeks. These concentrations selected in this study were found to be effective in prior studies [7,14]. The blood samples for estrogen levels detection were collected 24 h after the last estradiol injections. The levels of serum E2 were measured to confirm the effect of E2 replacement. As shown in Supplementary Fig. 1, the serum E2 levels in the Con group were 23.0 ± 2.2 pg/ml; however, serum E2 levels in the OVX group were decreased by 10.5 ± 1.1 pg/ml compared with the Con group ( * p < 0.05), and E2 replacement increased the serum E2 levels to 61.5 ± 2.7 pg/ml ( # p < 0.05).

Middle Cerebral Artery Occlusion and Reperfusion (MCAO-R)
The mice received MCAO-R injury as previously described [16]. Mice were anesthetized with 2% pentobarbital (3 ml/ kg). Following 1 h of transient occlusion, cerebral blood flow was restored by removing the suture for 24 h. Physiological parameters were monitored, including rectal temperature, blood pressure, blood gas, and glucose levels (Supplementary Table 1). Regional cerebral blood flow was monitored by laser-Doppler flowmetry. Only the mice whose mean cortical cerebral blood flow decreased to 15% of the preischemic value during occlusion then recovered to 70% of the baseline after reperfusion were used for further data analysis.

Assessment of Infarct Volume
After neurological scoring, infarct volume was assessed via 2,3,5-triphenyltetrazolium chloride (TTC) staining as previously described (n = 10) [18]. Brain slices were photographed, and infarct volume (the unstained areas) was measured by the ImageJ software. Corrections were made for swelling, and relative infarct size was determined based on the following equation: relative infarct size = (contralateral area − ipsilateral non-infarct area)/contralateral area.

Immunofluorescence (IF) Staining
Immunofluorescence staining was performed on frozen coronal sections of mouse brains (n = 3). Briefly, the mice were anesthetized as described above, and the brains were fixed via transcardial perfusion with 0.9% cold heparinized saline and 4% paraformaldehyde. After post-fixation and concentration gradient dehydration, the brains were cut into 10-μm-thick sections using a Leica CM1900 frozen slicer (Leica, Germany). The sections were washed three times with phosphate-buffered saline (PBS), then incubated overnight at 4 °C in a humidified atmosphere with mouse anti-GFAP antibodies (1:200; Cell Signaling Technology). Samples were then incubated with Alexa-488 (green, Invitrogen, USA)-conjugated donkey anti-mouse secondary antibodies for 2 h in the dark at room temperature. The sections were mounted with 50% glycerol and examined under a fluorescence microscopy (BX51; Olympus, Tokyo, Japan).

Primary Astrocyte Culture and Drug Treatment
Primary mouse astrocytic cultures were harvested from the cortex of 1-to 2-day-old Ndrg2 −/− pups and their wild-type littermates (Ndrg2 +/+ pups). Briefly, after removal of the meninges and hippocampus, the cortical tissues were subjected to enzymatic digestion and mechanical isolation. The resulting mixed cortical cells were passed through a 70-μm nylon mesh cell strainer and seeded into a cell culture flask in Dulbecco's modified Eagle's medium (DMEM, HyClone, USA) containing 10% fetal bovine serum (Gibco, USA) and 1% penicillin/streptomycin. When cells had reached confluence, the mixed glial cultures were shaken on an orbital shaker at 220 rpm for 18 h. The cultures were incubated at 37 °C in a 95/5% mixture of atmospheric air/CO 2 . The purity of the astrocyte culture was greater than 95%, as confirmed by staining with the astrocytic marker GFAP (Supplementary Fig. 2).

Statistical Analysis
All data were analyzed by an observer who was blind to the experimental protocol. Statistical calculations were performed with GraphPad Prism software, version 8.0. Comparisons between two groups were performed using Student's t test. Comparisons among multiple groups were performed using one-way analysis of variance (ANOVA) followed by Tukey's post hoc test. Neurological deficit scores are presented as median with interquartile range and were analyzed using two-tailed Mann-Whitney U test, and the other values are presented as the mean ± SD. p < 0.05 were considered statistically significant.

Sex Differences in Cerebral Ischemic Injury and NDRG2 Expression
First, we assessed infarct volume and neurological scores to verify gendered differences in cerebral ischemic injury. As shown in Fig. 2A(a) and (b), the mean infarct volume in male mice was 50.2 ± 1.7%, whereas female mice exhibited a smaller mean infarct volume of 32.2 ± 1.9% ( * p < 0.05). Neurological scores were also significantly lower in male than in female mice, demonstrating more severe neurological damage ( * p < 0.05). We next detected NDRG2 expression in the cortex and striatum of male and female mice. As shown in Fig. 2B(a) and (b), NDRG2 expression was significantly higher in female mice than in male mice in both the cortex ( * p < 0.05) and striatum ( ## p < 0.01).

Ndrg2 Deficiency Significantly Aggravates Cerebral Ischemic Injury in Female Mice
To investigate whether Ndrg2 is necessary for neuroprotective effects against cerebral ischemic injury in female mice, we used Ndrg2 systemic knockout female mice to examine infarct volume and neurological scores after MCAO-R injury. First, immunoblotting was performed to detect NDRG2 expression and PCR was used to confirm the genotypes in the brain of Ndrg2 +/+ and Ndrg2 −/− mice. As shown in Fig. 3A(a) and (b), Ndrg2 +/+ mice did and Ndrg2 −/− mice did not express NDRG2 in the brain. Moreover, following cerebral ischemic injury, the infarct volume in the Ndrg2 +/+ mice was 32.3% ± 1.8%, whereas Ndrg2 knockouts significantly increased the infarct volume to 48.0% ± 1.6% (Fig. 3B(a) and (b), * p < 0.05). Compared with the Ndrg2 +/+ mice, the Ndrg2 −/− mice exhibited worse neurological deficit scores (Fig. 3B(c), * p < 0.05).

Estrogen and ERβ Agonist DPN Treatment Upregulate the Expression of NDRG2 in Cortex and Striatum in the OVX Mice
As shown in Fig. 4A(a) and (b), the female mice were divided into five groups: Con (Control), OVX (mice that had undergone ovariectomy for 4 weeks), OVX + E2 (mice treated with estrogen for 3 weeks, beginning 1 week after ovariectomy), OVX + PPT (mice treated with the estrogen receptor α agonist PPT for 3 weeks, beginning 1 week after ovariectomy), and OVX + DPN (mice treated with the estrogen receptor β agonist DPN for 3 weeks, beginning 1 week after ovariectomy). We found that ovariectomy significantly decreased NDRG2 expression in both the cortex and striatum ( * p < 0.05 vs. Con-Cortex group, & p < 0.05 vs. Con-Striatum group). Furthermore, E2 and DPN treatment increased NDRG2 expression in both the cortex and striatum ( # p < 0.05 vs. OVX-Cortex group, ^p < 0.05 vs. OVX-Striatum group). However, PPT treatment did not increase NDRG2 protein expression.
We then quantified astrocyte activation in the region of non-ischemic mice corresponding to the penumbra region of ischemic mice by observing astrocyte morphology. As shown in Fig. 4E(a) and (b), the astrocyte (GFAP-positive) in the Ndrg2 +/+ -Con group displayed a normal activated morphology, which is characterized by a large soma and cytoplasmic processes. However, the astrocyte in the Ndrg2 +/+ -OVX group showed much smaller soma and thinner, shorter processes than those in the Ndrg2 +/+ -Con group ( * p < 0.05). Following ovariectomy, E2 or DPN treatment significantly increased astrocyte activation ( # p < 0.05). Interestingly, the astrocytes in the Ndrg2 −/− -Con, Ndrg2 −/− -OVX, Ndrg2 −/− -OVX + E2, and Ndrg2 −/− -OVX + DPN groups demonstrated small soma and short processes, demonstrating a lack of normal activation ( * p < 0.05 vs. Ndrg2 +/+ -Con group, ^p < 0.05, & p < 0.05). These results provide strong support for the importance of Ndrg2 in activating astrocytes.

Discussion
Ischemic stroke is a cerebrovascular disease with different effects in each sex [19][20][21]. Epidemiological studies have reported that the incidence of stroke is higher in men than in women and that men have worse stroke outcomes [2][3][4]. Although such gendered differences in stroke have been observed, the molecular mechanisms are incompletely characterized.
N-myc downstream-regulated gene 2 (Ndrg2) is a member of the NDRG family, which is involved in cell proliferation, differentiation, and stress responses [9]. Growing evidence indicates that Ndrg2 plays a pivotal role in the pathogenesis of cerebral ischemic injury [22,23]. A previous study suggested that Ndrg2-deficient male mice in models of cerebral ischemic injury induced by MCAO displayed exacerbated infarct damage [12]. A previous study in our lab demonstrated that Ndrg2 deficiency in male mice aggravated brain edema and elevated permeability of the blood-brain barrier (BBB) at an early phase after ischemia [13]. These results suggest that Ndrg2 is indispensable for neuroprotection in cerebral ischemic injury. In this study, we found that NDRG2 expression in both the cortex and striatum was significantly higher in female mice than in male mice. These data suggest that gendered differential expression of NDRG2 may be one of the reasons for the gender-based difference in cerebral ischemic injury. Nevertheless, the role of Ndrg2 in cerebral ischemic injury in female mice has not been explored before. In the current study, we found that either systemic deletion or conditional knockdown of Ndrg2 in female mice led to significantly increased brain infarct damage and lower neurological scores, consistent with the aforementioned findings in male mice. The results reflect that Ndrg2 is a key molecule in neuroprotection against cerebral ischemic injury in female mice. However, further studies are required to clarify the molecular mechanism underlying the regulation of NDRG2 expression.
We previously found that 17β-estradiol (E2) induced NDRG2 mRNA and protein expression in a dose-and timedependent manner in the mouse hippocampus [14], from which we inferred that Ndrg2 could be regulated by E2. E2 signaling is primarily mediated by the estrogen-binding receptor proteins estrogen receptor (ER) α and ERβ, which act as nuclear transcription factors [24]. Several studies, including ours, found that estrogen exerted significant neuroprotective effects against cerebral ischemia via activating ERα and ERβ [7,25,26]. In the current study, we found that ovariectomy significantly decreased NDRG2 expression in both the cortex and striatum; furthermore, E2 or estrogen receptor β agonist (DPN) treatment, but not estrogen receptor α agonist (PPT) treatment, significantly restored the NDRG2 expression in OVX mice. These results support Ndrg2 in the brain as the regulatory target of E2 and ERβ.
Astrocytes are an indispensable participant in neuropathological process, especially hypoxia/ischemia, inflammation, and repair. Increased GFAP expression in astrocytes indicates the activation of astrocytes. A previous study found that the activation of astrocytes provides strong support for A(b) NDRG2 protein expression levels. * p < 0.05 vs. Con-Cortex group, # p < 0.05 vs. OVX-Cortex group, & p < 0.05 vs. Con-Striatum group, ^p < 0.05 vs. OVX-Striatum group (n = 3 per group). B(a) Western blot showing GFAP protein expression in the region of non-ischemic female mice corresponding to the penumbra region of ischemic mice in different groups. B(b) GFAP protein expression levels. * p < 0.05 vs. Ndrg2 +/+ -Con group, # p < 0.05 vs. Ndrg2 +/+ -OVX group, ^p < 0.05 vs. Ndrg2 +/+ -OVX + E2 group, & p < 0.05 vs. Ndrg2 +/+ -OVX + DPN group (n = 3 per group). C(a) Western blot showing GFAP protein expression in primary astrocytes in different groups. C(b) GFAP protein expression levels. * p < 0.05 vs. Ndrg2 +/+ -Ast-Con group, # p < 0.05 vs. Ndrg2 +/+ -Ast-E2 group, ^p < 0.05 vs. Ndrg2 +/+ -Ast-DPN group (n = 3 per group). D(a) Western blot showing GFAP protein expression in the region of non-ischemic female mice corresponding to the penumbra region of ischemic mice in different groups. D(b) GFAP protein expression levels. * p < 0.05 vs. Ndrg2 flox/flox + AAV-Con-Con group, # p < 0.05 vs. Ndrg2 flox/flox + AAV-Con-OVX group, & p < 0.05 vs. Ndrg2 flox/flox + AAV-Cre-Con group, ^p < 0.05 vs. Ndrg2 flox/flox + AAV-Cre-OVX group (n = 3 per group). E(a) Representative immunofluorescence images showing the morphology of astrocytes labeled with GFAP in the region of nonischemic female mice corresponding to the penumbra region of ischemic mice in different groups. Scale bar = 20 μm. E(b) The average area of single astrocytic cells labeled by GFAP in the region of non-ischemic female mice corresponding to the penumbra region of ischemic mice. * p < 0.05 vs. Ndrg2 +/+ -Con group, # p < 0.05 vs. Ndrg2 +/+ -OVX group, ^p < 0.05 vs. Ndrg2 +/+ -OVX + E2 group, & p < 0.05 vs. Ndrg2 +/+ -OVX + DPN group (n = 5 per group) ◂ ischemic preconditioning-mediated neuroprotection [27]. Activated astrocytes improved the ability of neurons to eliminate excitatory neurotransmitters and ions such as glutamate, H + , and K + , and the viability of neurons co-cultured with astrocytes was greater than neurons cultured alone [28]. In our study, we found that E2 and DPN preconditioning could activate astrocytes both in OVX mice and in primary cultured astrocytes. Thus, we propose that the neuroprotective effects of E2 and DPN treatment against cerebral ischemic injury in female mice could be mediated by preactivation of astrocytes. In agreement with these results, our previous study showed that E2 and ERβ treatment alleviate global cerebral ischemia (GCI) and reperfusion in female mice by the activation of astrocytes [6]. Ndrg2 has been identified as a specific marker for astrocytes; it is mainly expressed in central nervous system astrocytes and plays a pivotal role in astrocyte function [10,11]. In this study, we found that the effect of E2 and ERβ in activating astrocytes disappeared both in Ndrg2 systemic knockout female mice and primary cultured astrocytes, but was partially retained in Ndrg2 conditional knockdown female mice. Altogether, the findings suggest that Ndrg2 plays an irreplaceable role in the neuroprotection of cerebral ischemia via pre-activation of astrocytes by E2 and ERβ treatment.
In summary, we conclude that E2-induced neuroprotection in female mice against cerebral ischemic injury involves activation of astrocytes via ERβ upregulating NDRG2. These findings provide insight into the roles of Ndrg2 in cerebral ischemic injury in female mice and provide a novel target of estrogen neuroprotection against cerebral ischemic injury.