Roflumilast Ameliorates Isoflurane-Induced Inflammation in Astrocytes via the CREB/BDNF Signaling Pathway.

Background and purpose: Astrocyte-mediated neuroinflammation plays an important role in anesthetic isoflurane-induced cognitive impairment. Roflumilast, a selective inhibitor of phosphodiesterase-4 (PDE-4) used for the treatment of chronic obstructive pulmonary disease (COPD), has displayed a wide range of anti-inflammatory capacity in different types of cells and tissues. In the current study, we aimed to investigate whether roflumilast possesses a protective effect against isoflurane-induced insults in mouse primary astrocytes.
Methods: Primary astrocytes were isolated from the cerebral cortices of immature rats. The production of NO was determined using DAF-FM DA staining assay. QRT-PCR and western blot were used to evaluate the expression levels of iNOS, COX-2, and BDNF in the astrocytes treated with different therapies. The gene expressions and concentrations of IL-6 and MCP-1 released by the astrocytes were detected using qRT-PCR and ELISA, respectively. The expression levels of phosphorylated CREB and PGE2 were determined using western blot and ELISA, respectively. H89 was introduced to evaluate the function of CREB. Recombinant human BDNF and ANA-12 were used to verify the role of BDNF.
Results: The upregulated iNOS, excessive production of NO, IL-6, and MCP-1, and activated COX-2/PGE2 signaling pathways in the astrocytes induced by isoflurane were significantly reversed by the introduction of roflumilast, in a dose-dependent manner. Subsequently, we found that BDNF could be upregulated by roflumilast, which was verified to be related to the activation of CREB and blocked by H89 (a CREB inhibitor). In addition, the COX-2/PGE2 signaling pathway activated by isoflurane can be inactivated by recombinant human BDNF. Finally, the regulatory effect of roflumilast against the isoflurane-activated COX-2/PGE2 signaling pathway was significantly blocked by ANA-12, which is a BDNF inhibitor.
Conclusion: Roflumilast might ameliorate isoflurane-induced inflammation in astrocytes via the CREB/BDNF signaling pathway.

Introduction

Approximately 0.2 billion patients undergo surgery globally every year,[1] among which 31–47% are diagnosed with cognitive disorder after they are discharged from hospital.[2] Inhalational anesthesia has been applied in approximately 80% of general anesthesia operations.[3] Isoflurane (Figure A), a widely used inhalational anesthesia, is reported to impact the development of neurons and long-term cognitive function of rats in different degrees.[4] As the most abundant neuroglia in the brain, astrocytes could induce the proliferation of synapses, promote synaptic plasticity, and impact the transmission of neurotransmitters by secreting different kinds of cytokines,[5,6] which play an important role in the brain structure, development of brain function, and maintenance of the nervous steady state.[7] Recently, it was reported that the toxicity on developing astrocytes could be induced by anesthesia, which contributes to developmental retardation and dysfunction.[8] Furthermore, isoflurane is reported to induce the activation of astrocytes to stimulate the production of inflammatory factors.[9] In addition to severe inflammation, the expression of inducible nitric oxide synthase (iNOS), an important enzyme that catalyzes the synthesis of nitric oxide (NO), is significantly upregulated by the active substances released by activated astrocytes, which induce the excessive production of NO and contribute to neurotoxicity. The excessively produced NO triggers the electron transport reaction with reactive oxygen radicals, such as superoxide anion, or exerts complicated functions by combining with various biological macromolecules containing metal ions or sulfhydryl groups.[10−12] Except for regular inflammatory factors, such as IL-6, TNF-α, and IL-1β, cyclo-oxygenase-2 (COX-2) is another vital inflammatory factor regulated by the activation of astrocytes. It is reported that COX-2 could be upregulated in the LPS-stimulated astrocytes by the MyD88-dependent signaling pathway, which further elevates the level of PGE2 to trigger severe inflammation.[13] Brain-derived neurotrophic factor (BDNF), an important factor in maintaining the proliferation of neurons and brain function, is significantly downregulated in the hippocampal tissues by the activated astrocytes,[14] which results in the apoptosis of neurons. The expression of BDNF is regulated by the phosphorylation of cAMP-response element-binding protein (CREB), which is regarded as a vital target to alleviate the neuron injury caused by the loss of BDNF.[15,16]

Figure 1

Molecular structure of isoflurane and roflumilast. (A) Molecular structure of isoflurane; (B) molecular structure of roflumilast.

Roflumilast (Figure B), a selective inhibitor of phosphodiesterase-4 (PDE-4) used for the treatment of chronic obstructive pulmonary disease (COPD), has displayed a wide range of anti-inflammatory capacity in different types of cells and tissues.[17,18] PDE-4 is a hydrolase for cyclic adenosine monophosphate (cAMP) which regulates the phosphorylation of CREB via the cAMP/PKA signaling pathway.[19] We supposed that by inhibiting PDE-4, the phosphorylation of CREB could be induced to promote the expression of BDNF and to alleviate the injury and apoptosis on neurons caused by anesthesia. In the present study, the protective effect of roflumilast against activated astrocytes induced by isoflurane will be investigated to explore the potential therapeutic property of roflumilast on anesthetic-induced cognitive impairments in the clinic.

Results

Expression of iNOS and Production of NO, Induced by Isoflurane, Were Inhibited by Roflumilast

Successful isolation of astrocytes was verified by staining with GFAP (Supporting Information Figure 1). Astrocytes were stimulated with 1.5% isoflurane in the presence or absence of roflumilast (5, 10 μM) for 24 h. As shown in Figure A,B, the expression of iNOS was significantly promoted by incubation with isoflurane but greatly suppressed by the introduction of roflumilast in a dose-dependent manner. Figure C shows the production of NO by the astrocytes. We found that the elevated release of NO, induced by isoflurane, was significantly reversed by the administration of roflumilast.

Figure 2

Roflumilast reduced isoflurane-induced expression of inducible NOS (iNOS) and production of nitric oxide (NO) in primary astrocytes. Cells were stimulated with 1.5% isoflurane in the presence or absence of roflumilast (5, 10 μM) for 24 h. (A) mRNA of iNOS as measured by real-time PCR; (B) protein of iNOS as measured by western blot; and (C) production of nitric oxide (NO) as measured by DAF-FM DA staining (####, P < 0.0001 vs control group; $$, $$$, P < 0.01, 0.001 vs isoflurane group).

Roflumilast Inhibited Isoflurane-Induced Expressions and Secretions of Proinflammatory Cytokines in Primary Astrocytes

As shown in Figure A, the expression levels of TNF-α, IL-6, MCP-1, and IL-1β were significantly upregulated by stimulation with isoflurane. However, the administration of two doses of roflumilast dose-responsively inhibited the isoflurane-induced expressions. The inhibitory effect of roflumilast on these cytokines was confirmed at their protein levels. As shown in Figure B, exposure to isoflurane induced high levels of TNF-α, IL-6, MCP-1, and IL-1β release in the culture media, but the presence of two doses of roflumilast significantly ameliorated isoflurane-induced release of these factors, respectively.

Figure 3

Roflumilast inhibited isoflurane-induced expression and secretions of proinflammatory cytokines in primary astrocytes. Cells were stimulated with 1.5% isoflurane in the presence or absence of roflumilast (5, 10 μM) for 24 h. (A) mRNA of TNF-α, IL-6, MCP-1, and IL-1β as measured using real-time PCR; (B) secretions of TNF-α, IL-6, MCP-1, and IL-1β (####, P < 0.0001 vs control group; $$, $$$, P < 0.01, 0.001 vs isoflurane group).

Expressions of COX-2 and PGE2 Induced by Isoflurane Were Significantly Reversed by Roflumilast

As shown in Figure A,B, the elevated expression level of COX-2 in the astrocytes induced with 1.5% isoflurane was significantly suppressed by the introduction of roflumilast. Additionally, the concentrations of PGE2 released by the astrocytes (Figure C) incubated with blank medium, 1.5% isoflurane, 1.5% isoflurane in the presence of 5 μM roflumilast, and 1.5% isoflurane in the presence of 10 μM roflumilast were 86.5, 321.7, 225.4, and 166.9 pg/mL, respectively.

Figure 4

Roflumilast ameliorated isoflurane-induced expression of COX-2 and production of prostaglandin E2 (PGE2) in primary astrocytes. Cells were stimulated with 1.5% isoflurane in the presence or absence of roflumilast (5, 10 μM) for 24 h. (A) mRNA of COX-2; (B) protein of COX-2; and (C) production of prostaglandin E2 (PGE2) as measured by ELISA (####, P < 0.0001 vs control group; $$, $$$, P < 0.01, 0.001 vs isoflurane group).

Roflumilast Increased the Expression of BDNF in Primary Astrocytes by Activating CREB

Following stimulation with roflumilast (5, 10 μM) for 24 h, the expression level of BDNF was evaluated using qRT-PCR and western blot. As shown in Figure A,B, we found that the mRNA and protein expressions of BDNF were significantly upregulated by the introduction of roflumilast in a dose-dependent manner. These results were further verified using enzyme-linked immunosorbent assay (ELISA) (Figure C). As shown in Figure D–F, the exposure to 1.5% isoflurane reduced cellular BDNF expression, but the addition of roflumilast (10 μM) dramatically mitigated this decrease.

Figure 5

Roflumilast increased the expression of BDNF in primary astrocytes. (A–C) Cells were stimulated with roflumilast (5, 10 μM) for 24 h. mRNA of BDNF, the total level of BDNF, and secretion of BDNF were measured (##, ###, P < 0.01, 0.001 vs vehicle group). (D–F) Cells were stimulated with 1.5% isoflurane in the presence or absence of roflumilast (10 μM) for 24 h. mRNA of BDNF, the total level of BDNF, and secretion of BDNF were measured (####, P < 0.0001 vs control group; $$$, P < 0.01, 0.001 vs isoflurane group).

Subsequently, we measured the level of phosphorylated CREB (p-CREB) at Ser133 and total CREB. The level of p-CREB (Ser133) was significantly elevated by the administration of roflumilast (Figure A) in a dose-dependent manner. Meanwhile, isoflurane treatment reduced p-CREB (Ser133) to about half, but the addition of roflumilast recovered the majority of p-CREB (Ser133) (Figure B). To further confirm the function of CREB in the regulatory effect of roflumilast on BDNF, the astrocytes were incubated with roflumilast (10 μM) in the presence or absence of H89 (10 μM), which is an inhibitor of the phosphorylation of CREB. As shown in Figure C–E, we found that the elevated expression level of BDNF induced by roflumilast was significantly blocked by H89, suggesting that the effects of roflumilast are mediated by CREB. Additionally, we knocked down the expression of CREB by transduction with lentiviral shRNA to CREB. Successful knockdown of CREB is shown in Supporting Information Figure 2A. Importantly, our results indicate that the effects of roflumilast on the expression of BDNF were abolished by knockdown of CREB (Supporting Information Figure 2B,C).

Figure 6

Effects of roflumilast on BDNF expression is mediated by CREB. (A). Cells were stimulated with roflumilast (5, 10 μM) for 2 h. Phosphorylated and total levels of CREB were measured by western blot analysis. (B) Cells were stimulated with 1.5% isoflurane in the presence or absence of roflumilast (10 μM) for 24 h. Phosphorylated (Ser133) and total levels of CREB were measured using western blot analysis. (C–E) Cells were stimulated with roflumilast (10 μM) in the presence or absence of H89 (10 μM). mRNA of BDNF, total level of BDNF, and secretion of BDNF were measured (##, ###, ####, P < 0.01, 0.001, 0.0001 vs vehicle group; $$$, P < 0.001. vs isoflurane or roflumilast group).

Roflumilast Suppressed the Expressions of COX-2 and PGE2 by Activating BDNF Signaling

To further investigate the mechanism underlying the protective effect of roflumilast against isoflurane-induced injury on astrocytes, we evaluated the relationship between BDNF signaling and the COX-2/PGE2 pathway. As shown in Figure , the astrocytes were stimulated with 1.5% isoflurane in the presence or absence of recombinant human BDNF (50, 100 ng/mL) for 24 h. We found that the elevated expression level of COX-2 by the stimulation with 1.5% isoflurane was significantly suppressed by introducing recombinant human BDNF. The concentrations of PGE2 produced by the astrocytes incubated with blank medium, 1.5% isoflurane, 1.5% isoflurane in the presence of 50 ng/mL recombinant human BDNF, and 1.5% isoflurane in the presence of 100 ng/mL recombinant human BDNF were 95.2, 342.1, 236.8, and 173.2 pg/mL, respectively. The roflumilast-treated condition was used as a positive control.

Figure 7

Recombinant human BDNF inhibited isoflurane-induced expression of COX-2. Cells were stimulated with 1.5% isoflurane in the presence or absence of recombinant human BDNF (50, 100 ng/mL) or roflumilast (10 μM) for 24 h. (A) mRNA of COX-2; (B) production of prostaglandin E2 (PGE2) as measured by ELISA (####, P < 0.0001 vs control group; $$, $$$, P < 0.01, 0.001 vs isoflurane group).

To further verify the function of BDNF in the regulatory effect of roflumilast on the isoflurane-induced injury on astrocytes, the cells were stimulated with 1.5% isoflurane in the presence or absence of roflumilast (10 μM) or the BDNF receptor TrkB inhibitor ANA-12 (10 μM) for 24 h, which blocked the action of BDNF. As shown in Figure A, we found that the suppressed expression level of COX-2 induced by roflumilast was significantly reversed by the introduction of ANA-12. In addition, as shown in Figure B, the concentrations of PGE2 produced by the astrocytes incubated with blank medium, 1.5% isoflurane, 1.5% isoflurane in the presence of 10 μM roflumilast, and 1.5% isoflurane in the presence of both 10 μM roflumilast and 10 μM ANA-12 were 95.2, 342.1, 236.8, and 173.2 pg/mL, respectively. These findings suggest that the protective effects of roflumilast against isoflurane are mediated by BDNF.

Figure 8

Blockage of BDNF signaling abolished the beneficial effects of roflumilast against isoflurane. Cells were stimulated with 1.5% isoflurane in the presence or absence of roflumilast (10 μM) or ANA-12 (10 μM) for 24 h. (A) mRNA of COX-2; (B) production of prostaglandin E2 (PGE2) (####, P < 0.0001 vs control group; $$$, P < 0.001 vs isoflurane group; ***, P < 0.001 vs isoflurane + roflumilast group).

Discussion

Cyclic nucleotide (cAMP) signaling is significantly involved in the brain’s metabolic biological functions, such as neuronal activity and energy production, metabolic processes, and synaptic physiology. The inactivation of protein kinase (PKA) is blocked by the suppression of cAMP, which further hinders the activation of phosphorylated CREB.[20,21] It is reported that the CREB family transcription factors are major mediators of BDNF transcriptional autoregulation in cortical neurons.[22] Ebrahimzadeh also reported that BDNF induced by the phosphorylation of CREB mediates the neuroprotective effects of crocin against MPH-induced neurotoxicity.[23] Therefore, the cAMP/CREB/BDNF signaling pathway might be an effective target for the neuroprotective purpose.

PDEs are the vital enzymes that regulate the hydrolyzation and inactivation of cAMP, which is of great interest in the search for memory-enhancing drugs.[24] PDE4 is one of the novel enzymes that mediate the rate of cAMP hydrolysis and is widely expressed in brain tissues.[25,26] In the present study, roflumilast, a specific PDE4 inhibitor, was used to investigate its potential protective effects against isoflurane-induced injuries and damages on astrocytes. We first incubated the rat astrocytes with isoflurane to explore its proinflammatory effect, and it was verified by the elevated expression of iNOS, promoted the production of inflammatory factors, increased the release of COX-2 and PGE2, and reduced the expression of BDNF. These inflammatory changes were found to be significantly reversed by the introduction of roflumilast, indicating a promising anti-inflammatory property against injury caused by anesthetics. We also found that the phosphorylated CREB was significantly elevated by roflumilast and the effects of roflumilast on the expression of BDNF could be blocked by the inhibitor of CREB. These data imply that roflumilast might regulate the expression of BDNF by activating the cAMP/CREB signaling pathway, via the inhibition of the activity of PDE-4.

COX-2 is an important proinflammatory enzyme and regarded as a biomarker of multiple inflammation-related diseases.[27−29] PGE2 is one of the main catalysates of COX-2 and has been reported to upregulate the MAPK, AP-1, and NF-κB signaling pathways.[30] Recently, COX-2/PGE2 has been widely regarded for its function in regulating neuroinflammation. Song[31] reported that hippocampal CA1 βCaMKII mediates neuroinflammatory responses via the COX-2/PGE2 signaling pathway in depression. By activating the COX-2/PGE2 pathway, neuronal death could be triggered by TDP-43-depleted microglia through regulating the MAPK/ERK pathway.[32] The relationship between the COX-2/PGE2 pathway and BDNF has been widely explored.[33,34] In the present study, we found that the activated COX-2/PGE2 signaling pathway by the introduction of isoflurane could be blocked by recombinant human BDNF, and blockage of BDNF signaling abolished the inhibitory effects of roflumilast against the isoflurane-activated COX-2/PGE2 signaling pathway. These data indicate that roflumilast, a PDE4 inhibitor, might inhibit the inflammation in the astrocytes by suppressing the COX-2/PGE2 signaling pathway through regulating the cAMP/CREB/BDNF axis.

A recent study reveals that the HDAC4 (histone deacetylase 4)-induced transcriptional inactivation of CREB is responsible for isoflurane-induced impairment in the brain. The exposure to isoflurane promotes nuclear translocation of HDAC4, which interacts with and disrupts CREB’s binding to its co-activator CBP for activation in the nucleus. As a result of CREB inactivation, BDNF expression is reduced after exposure to isoflurane.[35] The study by Huang et al. reports that isoflurane exposure induces neuronal apoptosis pathways by epigenetic modulation of MAPK signaling via the HADC family factors.[36] BDNF-CREB is also responsible for sevoflurane-mediated neurotoxicity.[37] A just-published work by Tang et al. shows that SIRT-1-mediated CREB inactivation is involved in the regulation of sevoflurane.[38] Therefore, the epigenetic modification leading to CREB inactivation could be the key mechanism of inhalational anesthetics-induced neurotoxicity. In our future work, the regulatory effect of roflumilast on the cAMP/CREB/BDNF axis and COX-2/PGE2 signaling pathway will be further investigated in an in vivo isoflurane-induced cognitive disorder animal model to better understand the potential therapeutic property of roflumilast on anesthetics-triggered cognitive impairments.

Inhalational anesthetics have been shown to affect the biochemical and morphological changes in neurons in vitro, and preclinical data suggest that the anesthesia exposure is associated with significant cognitive and behavioral impairments.[39] Notably, recent findings suggest that anesthetics-induced neurotoxicity could be related to Alzheimer pathogenesis.[40] Therefore, the preventive measures to ameliorate this side effect of anesthetics could have important clinical implications. Roflumilast has displayed potent anti-inflammatory properties in several cell types, suggesting its pleiotropic role in alleviating inflammation.[41] In the CNS, astrocytes play a critical role in protecting neurons from injury and promoting revasculization.[42] Therefore, roflumilast-mediated astrocyte protection against isoflurane-induced injury could potentially ameliorate anesthetics-induced neurotoxicity. Our findings are in-line with those reported in patients with COPD who take the once-daily dosage (500 μg tables). However, it is regrettable that the treatment by PDE4 inhibitors such as roflumilast show dose-dependent side effects in humans.[43] The major adverse effects of PDE4 inhibitors, such as roflumilast, are emesis and nausea. The mechanism of these adverse responses associated with PDE4 inhibitors is associated with their potent inhibition of PDE4 in emetic centers in the medulla oblongata.[44,45] It should be noted that these side effects could hamper the translational value of PDE4 inhibitors to human studies. Furthermore, the adverse effects could interfere with the mechanism of isoflurane anesthesia. A recent study suggests that the emetic effect of PDE4 inhibitors can reduce anesthesia duration in ketamine xylazine tests.[46] Therefore, the future study of PDE4 inhibitors on amelioration of anesthetics-caused side effects should consider the co-administration of antiemetic compounds.

The dual effect of inhalational anesthetics on neuroinflammation has to be mentioned based on recent studies. Commonly used anesthetics such as isoflurane and sevoflurane inhibit the inflammatory processes in ischemia and induced inflammatory disease animal.[47] However, other studies indicate that these types of anesthetics induce inflammation and neurotoxicity without additional stimuli.[48] Therefore, the effects of inhalational anesthetics on inflammation may be dependent on the context of disease models.[49]

Taken together, our data indicate that roflumilast ameliorated isoflurane-induced inflammation in astrocytes via the CREB/BDNF signaling pathway. The elucidation of the molecular mechanism with which roflumilast acts in the CNS could shed light on its amelioration against anesthetics-induced neurotoxicity.

Materials and Methods

Cell Isolation and Treatments

Newborn Sprague–Dawley rats (1–3 days, SPF grade) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. The rat astrocytes were isolated from the cerebral cortices of immature rats as described previously.[35] Briefly, the meninges, blood vessels, and white matters were removed before mincing the cerebral cortex into small pieces, followed by washing with precold DMEM medium. Subsequently, the tissues were digested with 0.25% trypsin at 37 °C for 30 min and then gently dissociated by trituration. The cell suspension was filtered with the 70 cell strainer, followed by centrifugation at 500g for 5 min. Then, the cells were resuspended, and the disassociated astrocytes were incubated with a completed DMEM containing 10% FBS and 1% antibiotics (penicillin/streptomycin). After being cultured for 3 days, the isolated cell purity was confirmed using immunofluorescence staining with astrocyte markers GFAP and ALDH1L1. All the isolated astrocytes were verified to have about 95% GFAP or ALDH1L1 staining.

To expose the astrocytes to isoflurane, the cells on the dishes were incubated for 24 h in a sealed humidified modular chamber, which was connected to a calibrated vaporizer to deliver 1.5% isoflurane mixed with 5% CO2/95% air gas. The control for the exposure was the 5% CO2/95% air gas without isoflurane. To examine the effect of roflumilast, the compound was simultaneously added to the culture media at the final concentrations of 5 and 10 μM, when cells were being exposed to isoflurane.

Real-Time PCR

The Trizol reagent (Thermo, Massachusetts, USA) was used to extract total RNA from the astrocytes, which was reverse-transcribed to cDNA with RT Master Mix (Takara, Tokyo, Japan). Real-time polymerase chain reaction (PCR) was conducted with the SYBR Master Mix utilizing the StepOne-Plus system (Takara, Tokyo, Japan) by denaturing at 95 °C for 30 s, annealing at 60 °C for 1 min, and extending at 95 °C for 5 s. The relative expressions of genes were evaluated using the 2-△△Ct method, with GADPH as the internal negative control. The following primers were used in this study: IL-6 (F: 5′-AACAGCGATGATGCAC-3′, R: 5′-TGGGGTAGGAAG GACT-3′); BDNF (F: 5′-TCAGCAGTCAAGTGCCTTTGG -3′, R: 5′-CGCCGAACCCTCATAGACATG -3′); iNOS (F: 5′-CGAAACGCTTCACTTCCAA -3′, R: 5′-TGAGCCTATATTGCTGTGGCT-3′); MCP-1 (F: 5′-GCATCCACGTGTTGGCTCA-3′, R: 5′-CTCCAGCCTACTCATTGGGATCA-3′); COX-2 (F: 5′-CAGACAACATAAACTGCGCCTT -3′, R: 5′-GATACACCTCTCCACCAATGACC -3′); GAPDH (F: 5′-CCGTGAAAAGAT GACCCAG-3′, R: 5′-TAGCCACGCTCGGTC AGG-3′).

Western Blot Analysis

The RIPA lysis buffer (Beyotime, Shanghai, China) was used to extract proteins from the astrocytes according to the instruction of the manufacturer, followed by protein quantification using a BCA assay kit (Beyotime, Shanghai, China). For each sample, 30 μg of total protein was loaded and separated by 12% SDS-PAGE, followed by being transferred to the polyvinylidene difluoride (PVDF) membrane. The membrane was subsequently incubated with antibodies against iNOS (1:1000, Abcam, USA), COX-2 (1:1000, Abcam, USA), p-CREB (Ser133) (1:1000, Cell Signaling Technologies, USA), CREB (1:1000, Cell Signaling Technologies, USA), and BDNF (1:1000, Abcam, USA), using β-actin as a negative control. The pictures were taken and analyzed using Image J.

DAF-FM DA Staining

DAF-FM DA assay was used to determine the production of NO by the astrocytes. The astrocytes were incubated with 30 μM DAF-FM DA (Sigma-Aldrich, USA) diluted with 10 mM Hepes-KOH pH 7.4 for 45 min in the dark. Subsequently, the cells were washed three times in 10 mM Hepes-KOH 7.4 solution. The images were captured using an inverted fluorescence microscope (Thermo Fisher Scientific, USA). The software Image J was used to quantify DAF-FM DA staining. Regions of interest (ROIs) were defined, and total cells (n) were counted. Then, we calculated integrated density value (IDV) of target cells. Average levels of NO = IDV/n.

Enzyme-Linked Immunosorbent Assay

The concentrations of TNF-α, IL-6, MCP-1, IL-1β, BDNF, and PGE2 in the astrocytes were detected using ELISA. Followed by incubation with 1% BSA to remove the non-specific binding proteins, the samples were incubated for an hour with primary antibodies and then subsequently incubated with streptavidin-horseradish peroxidase (HRP)-conjugated secondary antibodies for 20 min at room temperature. The absorbance of the samples at 450 nm was detected using a microplate spectrophotometer (Thermo, Massachusetts, USA).

Lentiviral shRNA to CREB

Cells plated at 70% confluency in six-well plates were transduced with the lentiviral CREB shRNA (sc-72030-V, Santa Cruz Biotechnology, USA). After 16 h, the virus-containing medium was removed and replaced with a normal growth medium.

Immunostaining

First, cells were fixed by incubation with 10% formalin at room temperature for 10 min. Cells were then permeabilized with 0.1% Triton ×100 for 10 min at room temperature. After being blocked with 10% normal goat serum for 1 h at room temperature, cells were incubated with the primary antibody against GFAP (1:1000, Cell signaling Technologies, #80788) overnight at 4 °C, followed by incubation with donkey anti-rabbit IgG Alexa 488 (Invitrogen, USA). Cells were then mounted with 4′,6-diamidino-2-phenylindole (DAPI). Images were visualized with a fluorescent microscope.

Statistical Analysis

The mean (±standard deviations, SDs) was shown. Differences between the means were analyzed for significance using analysis of variance, ANOVA (Statistica for Windows, Stat-Soft Inc., USA). Duncan’s multiple range test was used to test for significance of differences (P ≤ 0.05).