Phenotype Shifting in Astrocytes Account for Benefits of Intra-Arterial Selective Cooling Infusion in Hypertensive Rats of Ischemic Stroke

Abstract

The translational failure of neuroprotective therapies in stroke may be influenced by the mismatch of existing comorbidities between animal models and patients. Previous studies found that single-target neuroprotective agents reduced infarction in Sprague–Dawley but not in spontaneously hypertensive rats. It is of great interest to explore whether multi-target neuroprotectants and stroke models with comorbidities should be used in further translational researches. Ischemic stroke was induced in normotensive or hypertensive rats by 90- or 120-min middle cerebral artery occlusion (MCAO) and reperfusion. Intra-Arterial Selective Cooling Infusion (IA-SCI) was started at the onset of reperfusion for 30 minutes. Acute neurological deficits, infarct volumes, gene expression and markers of A1-like and A2-like astrocytes were evaluated. In further analysis, TNFα and IL-1α were administrated intracerebroventricularly, phenotype shifting of astrocytes and infarct volumes were assessed. Normobaric oxygen treatment, as a negative control, was also assessed in hypertensive rats. IA-SCI led to similar benefits in normotensive rats with 120-min MCAO and hypertensive rats with both 90-min and 120-min MCAO, including mitigated functional deficit and reduced infarct volumes. IA-SCI shifted astrocyte phenotypes partly by downregulating A1-like astrocytes and upregulating A2-like astrocytes in both RNA and protein levels. Upregulated A1-type astrocyte markers levels, induced by intracerebroventricular injection of TNFα and IL-1α, were closely related to increased infarct volumes in hypertensive rats, despite receiving IA-SCI treatment. In addition, infarct volumes and A1/A2-like genes were not affected by normobaric oxygen treatment. IA-SCI reduced infarction in both normotensive and hypertensive rats. Our results demonstrated the neuroprotective effects of IA-SCI in hypertensive rats may be related with phenotype shifting of astrocytes.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13311-022-01186-y.

Introduction

Early reperfusion is increasingly prioritized in ischemic stroke care [1]. However, approximately 50% patients still have a severe disability or die at 90 days post-stroke despite endovascular therapy [2–4]. Reperfusion alone may not be sufficient to fully salvage the penumbral tissue [4, 5]. For a long time, reperfusion therapy plus neuroprotective strategies have been explored to further reduce ischemic damage [5]. A recent clinical trial suggested potential neurological benefits when Nerinetide with multiple targets was added to patients receiving endovascular thrombectomy [6]. Therefore, it is deserving to keep on exploring neuroprotectants that harbour multiple mechanisms which may provide more benefits for patients with acute ischemic stroke [5, 7].

Previous researches observed that some neuroprotective agents induced neuroprotection in young healthy rodents but not in those with comorbidities [8, 9]. Recommendations for stroke research emphasize the importance of animal models with comorbidities in translating experimental data from bench to bedside, for example in Spontaneously Hypertensive Rats (SHR) [10]. Therapeutic hypothermia (TH) is a promising multi-target neuroprotective treatment [11–13], especially intra-arterial selective cooling infusion (IA-SCI) when combined with reperfusion therapy [14–16]. Additionally, astrocytes are the major component of the neurovascular unit, providing trophic support for neurons [17–19]. Recent studies indicated that activated astrocytes, with either potentially neurotoxic (A1-like) or beneficial (A2-like) phenotypes, broadly participated in acute central nervous system injury [20, 21]. Here, we hypothesized that TH may provide neuroprotective effects in a SHR model of ischemic-reperfusion and the phenotype shifting of astrocytes may account for potential mechanisms

Materials and Methods

Experimental Animals

All experimental procedures were approved by the Institutional Animal Investigation Committee of Capital Medical University, and all animal treatments were performed in strict accordance with the Care and Use of Laboratory Animals guidelines from the National Institutes of Health. To adhere to the Transparency and Openness Promotion Guidelines, all the data supporting the findings of this study are available from the corresponding authors on reasonable request. The utmost effort was made to minimize the number of animals used and their suffering. Adult male Sprague–Dawley (SD) rats (280–300 g; SPF grade; Vital River Laboratory Animal, Co) and adult male SHR rats (280–300 g; SPF grade; Vital River Laboratory Animal, Co) were used in this study. All rats were housed under a 12-h light/dark cycle and provided ad libitum access to food and water. The temperature of the rearing room was 23–25 ℃ and the relative humidity was 40–80%. The rats were randomly divided into 8 groups: a SHR-sham group (hypertensive rats subjected to surgery without disturbing the arteries; n = 8), a SD-normothermia group [normotensive rats underwent 120-min transient middle cerebral artery occlusion (tMCAO) followed immediately by room temperature saline infusion; n = 17], a SD-hypothermia group (normotensive rats underwent 120-min tMCAO followed immediately by 4 ℃ saline infusion; n = 17), a SHR-normothermia group (hypertensive rats underwent 90-min tMCAO followed immediately by room temperature saline infusion; n = 17), a SHR-hypothermia group (hypertensive rats underwent 90-min tMCAO followed immediately by 4 ℃ saline infusion; n = 17), a SHR-normothermia-120 min group (hypertensive rats underwent 120-min tMCAO followed immediately by room temperature saline infusion; n = 17), a SHR-hypothermia-120 min group (hypertensive rats underwent 120-min tMCAO followed immediately by 4 ℃ saline infusion; n = 17) and a SHR-hypothermia-T/I group (recombinant TNFα and IL-1α were administrated to hypertensive rats intracerebroventricularly before MCAO, and 4 ℃ saline was infused immediately after the reperfusion; n = 12). Rats in the experiments of normobaric oxygen (NBO) treatment were divided into 2 groups: SHR-MCAO group (hypertensive rats underwent 90-min MCAO) and SHR-NBO group (hypertensive rats were treated with NBO 15 min post ischemia and stopped at the onset of reperfusion). Detailed distribution of animals of each group are listed in Table 1 in Supplemental Data [8].

MCAO Procedures

Adult male SD and SHR rats were subjected to MCAO by using the intraluminal suture occlusion method. Rats were anesthetized in an induction chamber with 3–4% isoflurane in 70% N₂ and 30% O₂. After approximately 2 min, when rats respirations have slowed to 1 per second, remove the animal from the induction chamber and place its nose in the anesthesia mask with the isoflurane maintained at a concentration of 1–1.5%. The right middle cerebral artery was occluded for 90/120 min, and subsequent reperfusion was achieved with careful withdrawal of the filament [22]. Throughout the experimental procedure, the rectal temperature was maintained at 37 ± 0.5 ℃ with a feedback-regulated heating pad (69,020, RWD, Shenzhen, China) by using a rectal probe (69,022, RWD, Shenzhen, China). The surgeon who performed the MCAO surgery monitored the body temperature. Animals from each group were placed under warm conditions for an additional 1 h after the surgery. In the sham group, an identical surgical procedure was performed without disturbing the arteries. Cerebral artery occlusion was accomplished by using the intraluminal suture occlusion method under a laser speckle blood flow imaging system (LSF, Perimed, Jarfalla, Sweden) monitoring. Rats would be excluded from further analysis if they died before the end of the study, or if they did not exhibit any signs of ischemic injury based on LSF monitoring or Longa score. There were not important adverse events and death of animals. All rats with Longa scores ≥ 1 were included in our further analysis.

Cerebral Blood Flow Measurement

 A midline incision was made in the scalp and the skull surface was cleaned with sterile normal saline. Both ipsilateral and contralateral skull bones were carefully transparentized using a hand-held cranial drill (78,001, RWD, Shenzhen, China) with a flat head. Laser scanning imaging measurements were performed on the intact skull. Real-time cerebral blood flow (CBF) changes were recorded with a CCD camera and a Pericam PSI System (Perimed) that was placed roughly 10 cm above the brain. CBF was monitored pre stroke, after ischemia, and after reperfusion. CBF in ischemic brain tissue was calculated using the following formula: (CBF of ischemic hemisphere/CBF of nonischemic hemisphere)/(baseline CBF of ischemic hemisphere/baseline CBF of nonischemic hemisphere) × 100% [23].

Intracerebroventricular Injection

Guide cannulas were implanted stereotaxically (-1.5 mm posterior to bregma, 1.0 mm lateral to the sagittal suture, and 4 mm in depth) into the right lateral cerebral ventricle of the brain under isoflurane anesthesia (3–4% in 70% N₂ and 30% O₂) 5 days before MCAO to permit administration of substances intracerebroventricularly. TNFα (10 μg in 0.9% NaCl; HY-P70697, MedChemExpress) and IL-1α (10 μg in 0.9% NaCl; HY-P7096, MedChemExpress) were administered into the cerebral ventricles before MCAO [24–26]. For each injection, a volume of 1 μl was infused over 5 min.

Physiological Parameters

Whole-blood samples were taken from the tail arteries of rats for the arterial blood gas analysis before and during focal ischemia to determine blood pH, PaCO₂ and PaO₂ (VetStat, IDEXX, USA). Whole blood (200 μl) was collected with a Lithium-Heparin Syringe and proceeded immediately (within 5 min) to analysis. Blood pressure was measured by a noninvasive blood pressure system for rats (BP-2010A, Softron Biotechnology, Beijing, China).

IA-SCI

IA-SCI was induced by infusion of 6 ml of 4 ℃ isotonic saline, which was administered at the onset of reperfusion [22, 27, 28]. In all 4 groups, a small incision was made in the right external carotid artery to allow saline to enter the middle cerebral artery. A modified PE-10 catheter (0.2-mm outer diameter and 0.1-mm inner diameter) was inserted through the incision in the carotid artery. Six milliliters of 4 ℃ isotonic saline were then slowly and constantly (0.2 ml/min for 30 min) infused, approximately 0.25 ml/g brain tissue per minute. To avoid biases, different treatments were performed on the same day.

Brain and Body Temperature Monitoring

The rectal, cortex and striatum temperature were monitored before, during (for 30 min) and after IA-SCI (for another 40 min). Brain temperature was monitored ipsilaterally in the area supplied by the middle cerebral artery. Needle thermistor probes (Harvard Apparatus, Holliston, MA) were placed into cortex (3 mm lateral to bregma) and striatum (3 mm posterior and 5 mm lateral to bregma) through holes [22]. The rectal temperature was sustained at 37 ± 0.5 ℃ with a feedback-regulated heating pad throughout the MCAO procedure.

Brain Injury Determined by Infarct Volume and Neurological Deficits

Infarct volume was evaluated at 72 h after reperfusion in all rats. Five coronal brain slices with a 2-mm thickness were cut, then treated with 2, 3, 5-triphenyltetrazolium chloride (Sigma-Aldrich) at 37 ℃ for 20 min [29]. All slices were then photographed with a digital camera, and the infarct volumes were measured with NIH Image J software (Version 1.52, National Institutes of Health, Bethesda, Md). The non-infarcted region of the ipsilateral hemisphere was subtracted from that of the contralateral hemisphere to give an infarct size, which was presented as a percentage of the contralateral hemisphere. Following surgery, the neurological deficits were assessed using the Zea-Longa 5-point scoring [30] and tape removal test [31]. Longa test has been used for confirming a successful occlusion of the middle cerebral artery after completion of surgery. Longa scores were graded as 0 = no apparent deficit; 1 = slight deficit; 2 = moderate deficit; 3 = circling; 4 = severe circling; or 5 = no spontaneous movement or sever deficit. As for tape removal test, small adhesive backed (sticky) labels were placed on the distal-radial aspects of both forelimbs. The contact and removal times were recorded. Rats performed these tests at 24 h and 72 h after reperfusion, and all tests were overseen by the same person, who was blinded to the treatment groups.

Isolation of Astrocytes from Study Subjects

Brains were harvested from rats at 72 h after reperfusion, infarct regions were finely minced and then incubated at 37 ℃ with 0.125% trypsin (Gibco) and 100 μg/mL DNase I (Sigma-Aldrich) for 5 min, pipetting up and down several times during incubation, and centrifuged for 10 min at 1100 rpm at room temperature [32]. After removing supernatant, cells were resuspended in PBS (Gibco), and then filtered using a 70 μm cell strainer (Miltenyi Biotec). Subsequently, astrocytes were isolated from single-cell suspensions by using Anti-glutamate-aspartate transporter (GLAST) MicroBeads (Miltenyi Biotec) and magnetic activated cell sorting (MACS) [32]. Finally, isolated cells were resuspended in TRIzol reagent (Thermo Fisher Scientific, USA) for the total RNA extraction and was stored at -80 ℃.

Quantitative Real-Time Polymerase Chain Reaction

RNA was further purified from the TRIzol (Thermo Fisher Scientific, USA) extract with chloroform, centrifuged and precipitated from the resulting water phase with isopropanol. Subsequently, the RNA was washed with 75% ethanol, the RNA pellet was resuspended in RNase-free water and the RNA concentration was measured with a spectrophotometer (NanoDrop, Thermo Fisher Scientific, USA). Reverse transcription of RNA was performed with the HiScript III RT SuperMix for qPCR (Vazyme, Nanjing, China) according to supplier’s protocol. Quantitative RT-PCR was run using 2 μl cDNA and SYBR green chemistry (Vazyme, Nanjing, China) using the supplier’s protocol and a cycling program of 10 min at 95 ℃ followed by 40 cycles of 95 ℃ for 3 s and 60 ℃ for 30 s on a LightCycler 480 II (Roche). After completion of qPCR, a melting curve of amplified products was determined. All primer sequences are listed in Table 2 in Supplemental Data [20].

Immunofluorescence

Brain frozen sections on coverslips were blocked with 3% bovine serum albumin (BSA) in PBS containing 0.3% Triton X-100 for 1 h at room temperature. The sections were then incubated with primary antibodies overnight at 4 ℃. Two antibodies were added simultaneously for double-immunofluorescence staining. The following antibodies were used: mouse anti-C3 (1:100; Santa Cruz, USA), rabbit anti-acyl-coA synthetase long-chain family member 5 (ACSL5; 1:80; Aviva Systems Biology), mouse anti-GFAP (1:300; Jackson ImmunoResearch), rabbit anti-GFAP (1:300; Cell Signaling Technology, USA), rabbit anti-S100A10 (1:500; Invitrogen, USA). The samples were then incubated with mixtures of Alexa-488 (green, Invitrogen, Carlsbad, CA, USA) and Cy3 (red, Invitrogen, Carlsbad, CA, USA) conjugated donkey anti-mouse and goat anti-rabbit secondary antibodies for 1 h in the dark at RT. Mounting solution with DAPI (ab104139; AbCam, Cambridge, UK) was used to mount the sections. Sections were examined under a fluorescence microscope. The total number of C3/GFAP-positive cells, ACSL5/GFAP-positive cells and S100A10/GFAP-positive cells in the peri-infarct region were counted respectively in 5 different fields of view for each section.

Statistical Analysis

None of the investigators knew the groups assignment. The statistics were done on coded data. All values are expressed as mean ± SD. Statistical analyses were performed using the IBM “Statistic Package for Social Sciences” SPSS program for Windows (version 25.0). Differences between multiple groups were evaluated using one-way ANOVA and Double factor variance analysis (two-way ANOVA). D'Agostino-Pearson normality test and Shapiro–Wilk test were used to test normality. Post hoc comparison between groups was conducted using the least significant difference method. P-values of less than 0.05 were considered to show statistically significant differences between groups.

Results

IA-SCI Treatment Significantly Reduced the Brain Temperature

Significant difference was found between normotensive and hypertensive groups in mean arterial blood pressure levels pre-stroke, after MCAO and after saline infusion (P < 0.001 each; Fig. 1B). Blood pH, PaCO₂ and PaO₂ were maintained within similar ranges for all animals (Table 3 in Supplemental Data). There were no significant differences in CBF levels among the four groups at three time points of MCAO, respectively (Fig. 1C). Within 5 min, local infusion of 4 ℃ saline solution significantly reduced the temperature of the cerebral cortex from 37.5 ± 0.1 ℃ to 34.5 ± 0.1 ℃ in SD rats, from 37.4 ± 0.1 ℃ to 34.6 ± 0.2 ℃ in SHR rats (P < 0.05 each, Fig. 1D). Striatal temperatures exhibited a similar pattern of focal hypothermia, from 37.5 ± 0.1 ℃ and 37.0 ± 0.4 ℃ to 33.8 ± 0.4 ℃ and 33.6 ± 0.1 ℃ in SD and SHR rats, respectively. Hypothermic states remained for up to 60 min in all groups (P < 0.05 each, Fig. 1D). Rectal temperature remained close to the baseline (Fig. 1D).

Fig. 1.

Schematic design, cerebral blood flow monitoring, and hypothermic treatment. A, Schematic design for the experiment. B, Mean arterial pressure measured before focal ischemia, after MCAO and after saline infusion. C, Cerebral blood flow monitored before MCAO, during MCAO and after reperfusion (Bar = 2 mm). D, Temperatures in the cortex, striatum, and rectum were measured in a time-dependent manner. *: SD-normothermia vs SD-hypothermia; #: SHR-normothermia vs SHR-hypothermia. MCAO, middle cerebral artery occlusion. RT, room temperature. Data are shown as mean ± SD. n = 8 per group. */#P < 0.05; *** P < 0.001

IA-SCI Treatment Significantly Reduced the Infarct Size

We observed that infarcts were larger in SHR rats than in SD rats (49.63 ± 3.79% vs 39.08 ± 6.08%; P = 0.035) subjected to 120 min transient occlusion (Fig. I in Supplemental Data). In that case, we firstly shortened the period of occlusion in SHR rats to 90 min and assessed the efficacy of IA-SCI in SD and SHR rats when they had a comparable infarct size. We found that 90-min tMCAO in SHR rats resulted in comparable ischemia–reperfusion profiles and infarct sizes to those in SD rats with 120-min tMCAO (38.80 ± 5.92% vs 39.08 ± 6.08%; P = 0.649; Fig. 2A, B). More importantly, IA-SCI significantly reduced the infarct volumes in both SD- and SHR-hypothermia groups, compared with SD- (P = 0.019) and SHR-normothermia groups (P = 0.009; Fig. 2A, B).

Fig. 2.

Lesion analysis and neurofunctional testing in SD (120-min MCAO) and SHR (90-min MCAO) rats. A, Triphenyltetrazolium chloride staining demonstrated reduced infarct volumes in the ischemic territory supplied by the middle cerebral artery. B, IA-SCI significantly reduced the 72-h infarct volumes when quantified as a ratio of the infarcted tissue volume to the volume of the contralateral hemisphere. Longa score assessed with the 5-score system (C) and adhesive-removal test (D) were recorded at the defined time points. Both 1-way and 2-way ANOVA analyses indicated that IA-SCI significantly reduced the neurological deficits of hypothermia groups compared with normothermia groups. Data are shown as mean ± SD. n = 8 per group. *P < 0.05

IA-SCI Treatment Led to Mitigated Neurological Deficit

At 24 and 72 h after ischemia/reperfusion, neurological deficits were significantly ameliorated in hypothermia groups (both SD with 120-min MCAO and SHR rats with 90-min MCAO) when compared with normothermia groups (P < 0.05 each), based on the Longa score (Fig. 2C). As for tape removal test, SD-hypothermia group showed significant recovery compared with SD-normothermia group in both the first touch time and the removal time (P < 0.05 each; Fig. 2D). Significant difference was only found in the first touch time between SHR-normothermia and SHR-hypothermia groups (P < 0.05; Fig. 2D), but not in the removal time between the two groups.

IA-SCI Inhibited A1-type Astrocytes and Promoted A2-type Astrocytes in the Peri-infarct Region

As IA-SCI appeared to be neuroprotective in SHR rats with 90-min MCAO, we asked whether these effects were mirrored in the balance between potentially damaging A1-like genes and beneficial A2-like genes [20]. Astrocytes were isolated, and qPCR results showed that IA-SCI decreased all three A1-like genes and increased all three A2-like genes in SHR-hypothermia group when compared with SHR-normothermia group (P < 0.05 each; Fig. 3A, Fig. II in Supplemental Data).

Fig. 3.

Expression of A1/A2-like astrocytes markers in SD and SHR rats. A, Gene expression of A1-like and A2-like astrocytes in the peri-infarct region (n = 4/group; 3 time points). Values plotted are the natural logarithm transformed proportion of each gene. *: SHR-normothermia vs SHR-hypothermia. B-C, Representative immunofluorescence images and percentage of positive cells (n = 5/group; 5 sections in each rat per group). Bar = 50 μm on the first row. Bar = 15 μm on the second row. Data are shown as mean ± SD. *P < 0.05

Immunofluorescence also confirmed the IA-SCI led to diminishment of ACSL5⁺/GFAP⁺ and C3⁺/GFAP⁺ A1-type astrocytes and enhancement of S100A10⁺/GFAP⁺ A2-type astrocytes in SHR-hypothermia group when compared with SHR-normothermia rats in the peri-infarct region (P < 0.05 each; Fig. 3B, C).

Potential Association Between Shifts of A1/A2-type Astrocytes and Benefits of Hypothermia

We resorted to two independent experiments to explore whether shifts of A1/A2-type astrocytes were associated with benefits of hypothermia (Fig. 4A, C). Firstly, administration of TNFα and IL-1α significantly upregulated A1-like genes in SHR rats (Fig. 4B). More importantly, the reduced infarct volumes in SHR rats receiving IA-SCI treatment were reversed in those receiving IA-SCI, TNFα, and IL-1α (Fig. 4B). Secondly, we reproduced our previous results, in which NBO treatment did not alter infarct volumes in SHR rats after MCAO compared with control group (P = 0.598; Fig. 4D). As expected, NBO treatment increased PaO₂ above 400 mmHg during ischemia, compared with controls (Table 3 in Supplemental Data). More importantly, there were no significant differences in gene expressions of A1/A2-type astrocytes between SHR rats receiving NBO or not (Fig. 4D).

Fig. 4.

Neuroprotective effects of hypothermic treatment and phenotype shifting of astrocytes. A, Study design of TNFα/IL-1α injection. B, Gene expression of A1-like astrocytes after intracerebroventricular injection of TNFα/IL-1α (n = 4/group; 3 time points). Triphenyltetrazolium chloride staining and quantification of the relative ratio of the infarcted tissue volume to the volume of the contralateral hemisphere (n = 8 per group). C, Study design of the NBO treatment. D, Gene expression of A1-like and A2-like astrocytes in the peri-infarct region after NBO treatment (n = 4/group; 3 time points). Triphenyltetrazolium chloride staining and quantification of the relative ratio of the infarcted tissue volume to the volume of the contralateral hemisphere (n = 8 per group). SHR-N, SHR-normothermia; SHR-H, SHR-hypothermia; SHR-H-T/I, SHR-hypothermia- TNFα/IL-1α injection. Data are shown as mean ± SD. *P < 0.05

IA-SCI Treatment Partially Reversed the Increased Brain Damage and Neurological Deficits Because of Hypertension

Since IA-SCI treatment showed significant neuroprotective efficacy in SD and SHR rats when they had comparable infarct sizes, we asked whether these benefits still exist when SHR rats have the same period of occlusion (120 min), as compared with SD rats.

The SHR-normothermia rats had larger infarct sizes when compared with the SD-normothermia rats at 72 h after 120-min MCAO (P < 0.001; Fig. 5A). IA-SCI treatment decreased infarct volumes both in SHR and SD rats when compared with their respective normothermia group at 72 h after stroke onset (Fig. 5A, B). SHR-normothermia-120 min group exhibited more severe neurological deficits when compared with SD-normothermia group as measured by Longa score both at 24 h and 72 h after MCAO (P < 0.05 each; Fig. 5C). Furthermore, IA-SCI treatment resulted in significant improvements in functional outcomes in SD-hypothermia rats at both 24 h and 72 h post MCAO compared with SD-normothermia rats (P < 0.05 each; Fig. 5C), but improved functional outcome was only found at 72 h after MCAO in SHR-hypothermia-120 min rats compared to the SHR-normothermia-120 min group (P < 0.05; Fig. 5C).

Fig. 5.

Hypothermic treatment in SD and SHR rats receiving 120-min MCAO. A and B, Triphenyltetrazolium chloride staining and quantification of infarct volume in SHR rats received normothermia or hypothermia after 120-min MCAO. C, Longa score assessed with the 5-score system. D, Representative immunofluorescence images and percentage of positive cells (n = 5/group; 5 sections in each rat per group). E, Gene expression of A1-like (C3, SRGN and Serping1) and A2-like (S100A10, Ptgs2 and Tgm1) astrocytes in the peri-infarct region (n = 4/group; 3 time points). Values plotted are the natural logarithm transformed proportion of each gene. Data are shown as mean ± SD. n = 8 per group. *P < 0.05; *** P < 0.001; *: SD-normothermia vs SD-hypothermia or SHR-normothermia-120 min vs SHR-hypothermia-120 min

IA-SCI Treatment Still Existed Beneficial Effects on A1/A2 Astrocytes Markers in SHR after Receiving 120-min MCAO

IA-SCI treatment decreased A1-type astrocytes signal (ACSL5) in SHR-hypothermia-120 min group compared with SHR-normothermia-120 min group (P = 0.038; Fig. 5D). We did not observe any significant difference in the C3 or S100A10 signals between SHR-hypothermia-120 min and SHR-normothermia-120 min group (Fig. 5D).

qPCR results showed that 72 h after 120-min MCAO, IA-SCI treatment reduced one out of three A1-like genes (SRGN) and increased one out of three A2-like genes (S100A10) compared with normothermia treated SHR rats (P < 0.05 each; Fig. 5E, Fig. III in Supplemental Data).

Discussion

Although hundreds of neuroprotective therapies for ischemic stroke have been evaluated effectively in preclinical models, nearly all prove ineffective in clinical trials [14]. The mismatch between basic and clinical studies highlights the need of a clinically relevant animal model [9, 33]. As comorbidities may exert a detrimental impact on the treatment efficacy of neuroprotective therapies, animal models should closely mimic the clinical situation [10]. Thus, preclinical studies, using animal models with comorbidities, such as hypertension, diabetes, and hyperlipidemia, are critical for future clinical trials [10]. Our findings highlight the importance of using stroke models with comorbidities in translational studies.

Comorbidities may influence responses to potential treatments after stroke [8]. Previous studies indicated that dysfunctions of endothelial cells and impaired collateral supply might account for the effects of hypertension on neuroprotection [34]. However, the influence of hypertension on the protective effects of TH has not been thoroughly investigated in ischemic stroke. TH is well known to affect multifactorial pathways in the neurovascular unit, including maintaining the integrity of the blood–brain barrier, decreasing the expression of matrix metalloproteinases, suppressing inflammatory responses, and inhibiting apoptosis [14]. IA-SCI had advantages of a faster rate of achieving local hypothermia in the brain tissue and a lower risk of producing side effects by directly perfusing the hypothermic fluid into the cerebral arteries when compared with systemic hypothermia [14]. Furthermore, local hypothermia can combine with reperfusion therapy in the potential clinical practice. The present study showed that hypertension increased the infarct volume, which may worsen the neurological deficits.

Protecting not only neurons but other neural cells, which together form the neurovascular unit, may be useful when considering the lessons learned from previous translational efforts [17]. A previous study demonstrated that reactive astrocytes show the greatest resistance to focal ischemia among all elements of the neurovascular unit in oxygen glucose deprivation (OGD) test [17]. They found that TH abrogated the neuroprotective effects of OGD astrocyte-conditioned media on neurons in vitro. However, OGD test pays less attention to the interaction between neurons and astrocytes and maintenance of homeostasis (energy metabolism, neurotransmitter release, and signal transmission) during physiological and pathological events in the brain. Protective effects of astrocytes have been documented by many other in vivo investigations [35]. One potential mechanism of benefits may be the phenotype shifting of astrocytes. Activated astrocytes may assume either potentially harmful (A1-like) or potentially protective (A2-like) phenotypes [21]. Hypertensive rats showed increased neurotoxic astrocytic activation [10]. In our present study, IA-SCI treatment (depth of 35℃, duration of 30 min, no delay) exerted neuroprotective effects partly by inhibiting potentially damaging A1-like astrocytes and promoting beneficial A2-like astrocytes. Furthermore, our results suggested that IA-SCI treatment still had a positive effect on the polarization of astrocytes under a longer ischemia duration (120 min).

Phenotype shifting of astrocytes has been found in many physiological and pathological events in the brain, such as aging, ischemic stroke, Alzheimer’s disease, and Parkinson’s disease [20]. However, there are no specific A1/A2 modulators. Here, we provided two indirect experiments to further prove the association between phenotype shifting of astrocytes and benefits of TH on SHR rats, including positive controls (TNFα and IL-1α injections) and negative controls (NBO). Previous studies had shown that administration of TNFα and IL-1α could induce A1-type astrocytes [20, 36]. We upregulated the expression of A1-type astrocytes by administrating recombinant TNFα and IL-1α intracerebroventricularly, and the neuroprotective effects of TH on ischemia were partly diminished in SHR models. Furthermore, consistent with our previous study [8], NBO did not affect infarct volume of SHR rats. More importantly, we also found that A1/A2 astrocytes gene expressions were not influenced by NBO treatment. Despite the above evidence to prove the association between phenotype shifting of astrocytes and neuroprotective effects, we should add explicit caveats for these studies. For NBO, it remains possible that the lack of tissue protection indirectly leads to lack of effects on astrocytes rather than the other way around. For example, previous studies have demonstrated that hyperoxia can promote astrocytes cell death and even promote activation of proinflammatory phenotype of astrocytes under some conditions [37, 38]. For TNFα/IL1-α injection, it is also possible that other non-astrocyte effects may have increased injury thus leading to overall worsening of multiple endpoints. For example, TNFα and IL-1α play important roles in the immune system during inflammation after stroke onset [20]. Anyway, phenotype shifting in astrocytes following TH supports the idea that more attentions should be focused on the entire neurovascular unit in translational neuroprotective research, rather than only neurons.

This study has certain limitations. First, we only tested one TH dose and only young male rats were used in this study. Second, behavior was only evaluated at 72 h post-MCAO and long-term behavioral tests were needed. Third, detailed mechanisms of phenotype shifting in astrocytes regarding the effectiveness of TH should be further explored.

This proof-of-concept study emphasizes the importance of assessing and adjusting the neuroprotective therapy strategy in animal models with common comorbidities [39]. This study also provides a hint that the combination of endovascular treatment with TH may improve stroke systems of care and expand the eligibility for patients with comorbidities [3].

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

All authors participated in the study design. All endpoints were measured in a blinded manner as follows: the randomizations were done prior to the surgery by DW. After the randomizations, investigator SX performed MCAO surgery and left the room 15 min after ischemia onset. A second investigator (LW) performed the randomization, hypothermia or normothermia treatment, manipulated gas flow rates (for NBO treatment), obtained arterial blood gases, and adjusted the anesthesia or flow of gas if warranted. The room air and oxygen flowmeters were covered to protect the blind. Intracerebroventricular injection were done by a third investigator YD. All analysis of imaging, histology and behavioral were done by another investigator LFW. LW performed experiments and wrote the paper. DW, YD, LFW, YY, YL, ZG, ZL, SX, CW, JY, and JS performed experiments and analyzed results. DW and XJ conceived research and participated in article revision.

Funding

National Natural Science Foundation of China, 82027802, Xunming Ji, 81871022, Di Wu, 82071466, Di Wu, 82071468, Chuanjie Wu, the "mission" talent project of Beijing Municipal Administration of Hospitals, SML20150802, Xunming Ji, National Key R&D Program of China, 2017YFC1308401, Xunming Ji, Beijing Science and Technology Planning Project, Z181100001918026, Xunming Ji.

Declarations

Conflict of Interest

The authors declare that they have no competing interests.