Adult male and female Sprague-Dawley rats (250–280 g) were obtained from the Experimental Animal Center of Zunyi Medical University, which were breeding and mated in the animal facility of the Department of Physiology in Zunyi Medical University (Zunyi, China). Rats were housed in groups of two per cage in sanitary barrier room in ventilated racks with free access to food and tap water. The room temperature was maintained at 22°C with a 12 h light/dark cycle (lights on/off at 8:00 AM/8:00 PM). Primary dissociated cultures of dorsal spinal cord were prepared from SD rats (≤3 days) in accordance with the National Institutes of Health guidelines in a manner that minimized animal suffering and animal numbers. All experiments were carried out in accordance with China animal welfare legislation and were approved by the Zunyi Medical University Committee on Ethics in the Care and Use of Laboratory Animals.

Primary dissociated cultures of dorsal spinal cord astrocytes were prepared and maintained as described previously (Zeng et al., 2008). No sample calculation was performed to predetermine the sample size. For the experiments of this work, we used 8 litters of 12 pups each of both sexes. Briefly, SD rat pups (≤3 days) were anesthesia with ether vapors, killed by rapid decapitation using sterile scissors, and euthanasia was performed between 8 am to 12 noon to minimize animal suffering. This method of anesthesia was chosen because others common methods (such as isoflurane) were found to decrease primary astrocytic viability (Guo et al., 2017). Their dorsal spinal cord was rapidly cut into pieces and then incubated in 0.125% trypsin (HyClone, SH30042.01) for 30 min at 37 °C. Dulbecco’s modified Eagle’s medium (DMEM/F12 medium, Gibco, C11330500BT) containing 15% (v/v) fetal bovine serum (FBS, MRC, CCS30009.02) was added to stop the action of typsin. The cells were collected after centrifugation at 1,000 g min for 5 min. The supernatants were discarded and the cells were suspended in complete medium containing DMEM/F12 and 15% (v/v) fetal bovine serum. 10⁵ cells/cm² cells were planted in poly-D-lysine-coated 75 cm² flask and cultured in the incubator at 37°C in a humidified 5% CO₂-95% air atmosphere. After 10 days, flasks were shaken (260 rpm/min) overnight to remove non-astroglial cells. Dorsal spinal cord astrocytes stained positively for the astrocytic marker, glial fibrillary acid protein (GFAP).

Lipopolysaccharide (L2630), TAK-242 (TLR4 antagonist; 614316), carbenoxolone (CBX, a Cx43 hemichannel and gap junction inhibitor; C4790) (Yin et al., 2018), Gap19 (inhibiting Cx43 hemichannel activity but not gap junctional communication; SML1426) (Yin et al., 2018), ammonium pyrrolidinedithiocarbamate (PDTC, NF-κBp65 inhibitor; 548000), ethidium bromide (EtBr; E8751) and lucifer yellow (LY; L0144) were purchased from Sigma (St. Louis, MO, United States). Gap26 (a specific gap junction and hemichannel blocker corresponding to residues 63–75 of Cx43; AS-62644) (Yin et al., 2018) was purchased from AnaSpec. SR11302 (AP-1 inhibitor; sc-204295) was purchased from Santa Cruz, Dallas, TX, United States. Among these drugs, LPS, CBX, Gap26, Gap19, PDTC, SR11302, EtBr and LY were dissolved in 0.01M PBS. CBX is a widely used Cx43 gap junction/hemichannel blocker (Ye et al., 2009). TAK-242 was dissolved in 10% dimethyl sulfoxide (DMSO) and diluted in 0.01M PBS. Vehicle (0.01% DMSO diluted in PBS solution) was used in the Veh-treated group and Veh + LPS group. The presence of DMSO (<0.01%) alone did not affect the production and release of chemokine from cultured astrocyte at quiescence state.

According to previous studies, LPS at dose of 1 μg/ml for 24 h induced astrocytic activation and inflammatory cytokine production (Oliveira-Junior et al., 2019). However, LPS at dose of 2 μg/ml was shown to remarkable decrease cell viability in primary cultured astrocytes (Xian et al., 2019). For this reason, we used 1μg/mL LPS for 24 h in the following experiments. In addition, TAK-242 at 100 nM was used in order to confirm the role of TLR4 in LPS-induced inflammatory response in cultured astrocytes (Du et al., 2017). It is reported that CBX, Gap26, and Gap19 at 100 μM were frequently used in order to establish the involvement of Cx43 in neuroactive substance release from astrocytes (Takaki et al., 2012; Guillebaud et al., 2020). PDTC at 1 μM and SR11302 at 10 μM were frequently used in order to establish the involvement of NF-κBp65 and AP-1 signaling in cellular responses (Zhao et al., 2015; Liddell et al., 2016). Based on these previous experimental reports, in our present experiments, cells were exposed to LPS (1 μg/ml for 24 h) with or without pretreatment of TAK-242 (100 nM for 2 h), PDTC (1 μM for 30 min), SR11302 (10 μM for 2 h), CBX (100 μM for 1 h), Gap26 (100 μM for 1 h), and Gap19 (100 μM for 1 h), respectively. Afterward, the cells and supernatants were harvested separately and processed according to the different assay protocols (Figure 1).

To observe the nuclear translocation of NF-κB or AP-1 in astrocytes, cells grown on glass coverslips were stained for NF-κBp65/p50, AP-1 and DAPI. After treatment with 1 μg/ml LPS for 24 h, the cells were fixed in 4% paraformaldehyde for 20 min, washed three times with 0.01 M PBS and incubated with 10% goat serum for 15 min to block non-specific binding. Astrocytes cultures were stained with rabbit anti-NF-κBp65 antibody (520 μg/ml, 1:350, Proteintech Group, Rosemont, IL, United States), mouse anti-NF-κBp50 antibody (200 μg/ml, 1:350, Santa Cruz, Dallas, TX, United States) or rabbit anti-AP-1 antibody (1200 μg/ml, 1:350, Proteintech Group, Rosemont, IL, United States) overnight. Subsequently, the cultures were incubated with FITC-conjugated goat anti-rabbit IgG (600 μg/ml, 1:350 in PBS, Proteintech Group, Rosemont, IL, United States) or FITC-conjugated goat anti-mouse IgG (600 μg/ml, 1:350 in PBS, Proteintech Group, Rosemont, IL, United States) for 1 h at room temperature. Finally, 4’,6-diamidino-2-phenylindole (DAPI, 1:1000, Roche) was applied for 5 min to label all nuclei. We randomly chose 10 visual fields (nearly 50 cells per visual field) to count the percentage of the positive cells (nuclear staining for NF-κBp65/p50 or AP-1) with Image J software.

In addition, cells grown on glass coverslips were stained for TLR4 and Cx43 by double immunofluorescence. Astroglial cultures were stained with mouse anti-TLR4 antibody (1700 μg/ml, 1:350, Proteintech Group, Rosemont, IL, United States) overnight. Subsequently, the cultures were incubated with CY3-conjugated goat anti-mouse IgG (600 μg/ml, 1:350 in PBS, Proteintech Group, Rosemont, IL, United States) for 1 h, followed by incubations with rabbit anti-Cx43 antibody (700 μg/ml, 1:350, Proteintech Group, Rosemont, IL, United States) overnight. Then, the cultures were incubated with FITC-conjugated goat anti-rabbit IgG (600 μg/ml, 1:350 in PBS, Proteintech Group, Rosemont, IL, United States) for 1 h. At last, DAPI was applied for 5 min to label all nuclei. All incubations were held at 25°C and separated by three 5-min washes in 0.01 M PBS in a dark place. Control experiments were performed using an excess of the appropriate homolog peptide antigen to absorb the primary antibodies and thus confirm a specific immunoreaction. After staining, coverslips were mounted on glass slides and examined using Airyscan2 confocal laser scanning microscope 900 (Zeiss, Germany).

The total RNA of cultured dorsal spinal cord astrocytes was isolated by using Trizol reagent after 24 h of drug administration. Real time quantitative PCR (RT-qPCR) is carried out in iCycler IQ Real-Time PCR Detection System (BIO-RAD Co., Pleasanton, CA, United States) with SYBR Green PCR Master Mix (ABI Co., Foster, CA, United States). PCR conditions were as follows: 95°C 30 s 1 Cycle; 95°C 5 s, 60°C 30 s, 40 Cycles. The cycle threshold (Ct) value represents the cycle number at which a fluorescent signal rises statistically above background. The relative quantification of gene expression was analyzed by the 2^{–Δ Δ} Ct method.

The nucleotide sequences of the primers were synthesized by TaKaRa Biological Engineering Company. Primer preparations were devised according to the sequence searched on GeneBank. The nucleotide sequences of the primers used in this experiment were as follows: (1) MCP-1 (genebank: NC-005109.4): F: 5′-CAGGTC TCTGTCACGCTTC-3′; R: 5′-AGTTCTCCAGCCGACTCA-3′. (2) CXCL1 (genebank: NC-005113.4): F: 5′-AACCGAAGTCATAGCCACA-3′; R: 5′-GGGGA CACCCCTTTAGCA-3′; (3) Cx43 (genebank: NC-051355.1): F: 5′-ATCATCC TCGACTGGTCCT-3′; R: 5′-ACATCCACACGTAGAAGC-3′. (4) NF-κBp65 (genebank: NC-051336.1): F: 5′-CCAAAGACCCACCTCACC-3′; R: 5′-CGCTTCTTCACACACTGGA-3′; (5) NF-κBp50 (genebank: NC-051337.1): F: 5′-GGGCAGAAGTCAACG-3′; R: 5′-TGTCGTCCCATCGTAGGT-3′; (6) c-Jun (genebank: NC-051340.1): F: 5′-TGTCTGTATGCTGGGGTGA-3′; R: 5′-GGTTGCTGGGGAGAGAGA-3′; (7) β-actin (genebank: NC-005111.4): F: 5′GGCTGTATTCCCCTCCATCG-3′; R: 5′-CCAGTTG GTAACAATGCCATGT-3′.

Total proteins were extracted from cultured astrocytes lysed by radioimmunoprecipitation (RIPA) buffer containing protease and phosphatase inhibitor cocktails (Solarbio Science & Technology Co., Ltd, Beijing, China), and the protein concentration was determined by using a BCA protein assay kit (Solarbio Science & Technology Co., Ltd, Beijing, China). Then 20 μg equal amounts of denatured proteins were subjected to SDS-PAGE gel electrophoresis and transferred onto polyvinylidenefluoride (PVDF) membranes (Merck Millipore Ltd, Cork, Ireland). Next, the transferred membranes were blocked with 5% Albumin Bovine V (Solarbio, A8020) for 2 h and incubated with the primary antibodies (Table 1) overnight at 4°C. The following day, the membranes were washed three times with 0.1% Tween-Tris-buffered saline (TBST, Servicebio, G0001), and then incubated with secondary antibodies (Table 1) at room temperature for 1 h. Finally, the membranes were imaged using an Electro-Chemi-Luminescence-Kit. Images of the blots were captured using the G: BOX Chemi XX9 system (Synoptics Group, Frederick, MD, United States). The image was scanned, and band intensity was semi-quantified using Image J (version 1.8.0, National Institutes of Health). β-actin or glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as the internal reference, and the gray value of the target protein was divided by the gray value of the internal reference to correct the error. The result represents the relative content of the target protein in the sample. The protein levels were normalized to β-actin or GAPDH protein levels and the results are shown as the percentage of Vehicle (% of Vehicle).

For dye uptake experiments, astrocytes were stimulated with LPS for 24 h and exposed to 0.5 μM EtBr for 10 min at 37°C. EtBr is impermeable through membrane but can transit through hemichannels and becomes more fluorescent after binding to DNA. After 10 min exposure to EtBr, astrocytes were washed with PBS and stained with DAPI (Feng et al., 2018). Fluorescence photomicrographs were captured with a digital video camera (Olympus Corporation, Tokyo, Japan) connected to an inverted fluorescent microscope equipped with the appropriate filters (Olympus Corporation, Tokyo, Japan). We randomly chose 10 visual fields (nearly 50 cells per visual field) to count the percentage of the EtBr-positive cells by Image J software.

Gap junction permeability was determined by the scrape-loading/dye transfer technique as previously described by Gangoso et al. (2017). LY is a fluorescent dye that can pass through the gap junctions of loaded cells to their neighbors. Astrocytes were stimulated with LPS for 24 h. Then, scrape-loading was performed by scraping the cell layer with a broken razor blade in PBS containing LY (1 mg/ml). After 5 or 10 min, the dye solution was removed and the cells were carefully washed. Subsequently, fluorescence photomicrographs were captured and the diffusion distance of the dye was quantified using Image J software.

Cultured astrocytes were seeded in 12-well plates (1 × 10⁶ cells/well) pretreated with various antagonists followed by stimulation with LPS (1 μg/ml, 24 h). After stimulation, culture media were collected and centrifuged at 10,000 rpm for 5 min. The amounts of MCP-1 and CXCL1 in the culture medium were measured with commercial ELISA kits obtained from Shanghai Hushang Biological Technology Co., Ltd. China.

The supernatant was discarded from each group of cells, and the nuclear protein samples were obtained according to the manufacturer’s instructions of Minute™ Cytoplasmic and Nuclear Extraction Kit (cat. no. SC-003; Invent Biotechnologies, Inc, Dallas, TX, United States). Cells were washed twice with cold PBS and suspended in 120 μl of cytoplasmic extraction buffer (pH 7.9, 10 mM Hepes, 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM dithiothreitol, 0.5 mM PMSF). The samples were vigorously vortexed for 15 s and lysed on ice for 2 min (repeat this for five times). Following lysis, the extract was centrifuged at 10,000 × g for 5 min to transfer the supernatant as the cytosol fraction. The nuclei were suspended in 60 μl of nuclear extraction buffer (20 mM Hepes, 20% glycerol, 0.4 M NaCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM DTT, 0.2 mM PMSF). The samples were vigorously vortexed for 15 s and lysed on ice for 2 min (repeat this for five times). The extract was centrifuged at 10,000 × g for 5 min and the white microtransparent precipitation at the bottom was the nuclear protein. The nuclear protein was determined by using a BCA protein assay kit.

It is reported that the NF-κBp65 binding site in the promoter of MCP-1 gene (254 bp to 261 bp, p65-1: 5′-AGTTGAGGGGGACTTTCCCAGGC-3′) and the AP-1 binding site in the promoter of Cx43 gene (251 bp to 257 bp, AP-1-1: 5′-CGCTTGATGAGTCAGCCGGAA-3′) (Liu and Quinn, 2002; Teunissen et al., 2003). In addition, as shown in Table 2, the JASPAR and PROMO websites predicted the possible binding sequences of transcription factors (NF-κBp65 or AP-1) to the promoter region of MCP-1, CXCL1, or Cx43. We found that only one sets of high comparability DNA alignment with sequence logo for NF-κBp65 in the promoter region of MCP-1 and 2 binding sites in CXCL1, while with three binding sites for NF-κBp65 in the promoter region of Cx43. In addition, we found three sets of high comparability DNA alignment with sequence logo for AP-1 in the promoter region of MCP-1 and three binding sites in CXCL1, while with two binding sites for AP-1 in the promoter region of Cx43. According to the above information, to detected the interactions of transcription factors (NF-κBp65 or AP-1) and MCP-1 gene promoter, the following oligonucleotides were used: biotin 3’ end DNA-labeled NF-κBp65 probes (p65-1: 5′-AGTTGAGGGGGACTTTCCCAGGC-3′; p65-2: 5′-TAAAATGTGATTTCCCTTCAGT-3′) or AP-1 probes (AP-1-2: 5′-TATATCCCTGACACACCTGGGG-3′; AP-1-3: 5′-CCTCCTGCTGGGTCA GTTCTCCG-3′; AP-1-4: 5′-TGGGATGCTCAGTCATAGAGATG-3′). To detected the interactions of transcription factors (NF-κBp65 or AP-1) and CXCL1 gene promoter, the following oligonucleotides were used: biotin 3′ end DNA-labeled NF-κBp65 probes (p65-3: 5′-CTGTCCTGGAATGTCCTTGTCC-3′; p65-4: 5′-TTTACCTGGGATGTCCTCTCCT-3′) and AP-1 probes (AP-1-5: 5′-GGTCCTGACACAGAGTCACTGT-3′; AP-1-6: 5′-TTGGGATATGACTCTGGG GACA-3′; AP-1-7: 5′-AAATCTTTTGACTTATTCCATT-3′).

At last, to detected the interactions of transcription factors (NF-κBp65 or AP-1) and Cx43 gene promoter, the following oligonucleotides were used: biotin 3′ end DNA-labeled NF-κBp65 probes (p65-5: 5′-TGTGTTTGGATATTCCGTGTTT-3′; p65-6: 5′-TCTTGGGGGATTTTTCCTTTGA-3′; p65-7: 5′-ACGTTTGGT GTTTTCCTCTGGC-3′) and AP-1 probes (AP-1-1: 5′-CGCTTGATGAGTCAG CCGGAA-3′; AP-1-8: 5′-TTGACAGTTGAGTCAATGATTTC-3′; AP-1-9: 5′-ATCAACATTTAATCATCTCCTCA-3′). In our study, these portions of the transcription factors/promoter region were PCR amplified and all labeled with biotin at 3′ end DNA.

It is reported that the p50–p65 heterodimers play a key role for MCP-1 gene expression in rat aortic endothelial cells (Panicker et al., 2010). Then, as shown in Table 2, the JASPAR and PROMO websites predicted the possible binding sequences of p50 to the promoter region of MCP-1. We found that two sets of high comparability DNA alignment with sequence logo for NF-κBp50 in the promoter region of MCP-1. In addition, the combination of p50 with the promoter regions of CXCL1 or Cx43 was no predicted with online websites JASPAR and PROMO. Furthermore, the combination of p100 with the promoter region of MCP-1, CXCL1 or Cx43 was also no detected with online websites JASPAR and PROMO. For this reason, to detect the interactions of p50 and MCP-1 gene promoter, the following oligonucleotides were used: biotin 3′ end DNA-labeled NF-κBp50 probes (p50-1: 5′-TGTGAGGTGACATCCCCAGATT-3′; p50-2: 5′-TAAAATGTGATTTCCCTTCAGT-3′).

The Thermo Scientific LightShift Chemiluminescent EMSA Kit (Thermo Scientific Pierce, Shanghai, China) was used to detect the interactions of transcription factors (NF-κBp65, p50 or AP-1) and gene promoter (MCP-1, CXCL1, or Cx43). Nuclear proteins were incubated with biotin-labeled oligonucleotide probes containing the binding sequence for MCP-1, CXCL1 or Cx43 (biotin 3′ end DNA labeling kit, Viagene Biotech Inc.). The binding reaction was performed by adding 8 μg of nuclear extracts, 20 fmol of biotin end-labeled DNA, and 1 μg/μl poly (dI-dC) to the end volume of 20 μl. For competition, a 100-fold molar excess of unlabeled cold NF-κBp65, p50 or AP-1 probe had the same sequence. After incubation for 20 min at room temperature, the reaction is then subjected to gel electrophoresis on a 5% native polyacrylamide gel and transferred to a nylon membrane. The biotin end-labeled DNA was detected using the streptavidin-horseradish peroxidase conjugate and the chemiluminescent substrate. The DNA binding activity was quantified by optical densitometry using Image J software. The binding activity was normalized to control and the results are shown as the percentage of control (% of control).

All data were presented as mean ± standard deviation (SD). For this study, randomization and blinding were not performed. No statistical methods were used to determine the sample numbers in advance. No outlier test was conducted, thus, all data were included in the analysis. Hence, this should be considered an exploratory study.

Shapiro–Wilk tests were performed to examine normality of each data set. Since the distribution was even in most cases, parametric statistics were used to analyze the data. Data analysis was performed using the software Statistical Package for Social Sciences version 17.0 (SPSS Inc., Chicago, IL, United States). The One-Way ANOVA or Two-Way ANOVA plus a post hoc test were used for statistical analysis to compare the differences among treatment groups. Statistical Power analysis was performed by using the ‘General Linear model-Univariate’ function in SPSS 17.0 software. Where indicated, ‘n’ stands for the number of separate experiments carried out and, within each experiment, usually 4–5 technical replicates were performed for each condition/drug-treatment. For rat astrocyte cultures, a separate experiment is counted as cells extracted from different neonatal rat dorsal spinal cord. Differences at the p < 0.05 level were considered statistically significant. The statistical powers at the p > 0.8 level were considered as a real difference in different groups.

In the present study, dorsal spinal cord astrocytes were identified with immunofluorescence stained with GFAP polyclonal rabbit antibody. The immunofluorescence staining after two generation showed that the purity of the astrocytes was more than 98% (Figure 2A). PCR and WB assays showed that the expression of MCP-1 and CXCL1 at the mRNA and protein levels were significantly increased in Veh + LPS group compared with Veh-treated group (p < 0.05, Power = 1, Figures 2B–D), whereas the effect was completely reversed by pretreatment with TAK-242 (100 nM for 2 h, p < 0.05, Power = 1). Pretreatment with TAK-242 significantly suppressed LPS-induced TLR4 and GFAP expression (TLR4: p < 0.05, Power = 0.935; GFAP: p < 0.05, Power = 1), which indicated that TLR4 activation is one of the key factors in LPS-induced astrocytic activation.

We detected the protein level and activity of NF-κB and AP-1 in astrocytes before and after LPS stimulation. Immunofluorescence staining results showed that NF-κBp65 (Figures 3A,B), p50 (Figures 4A,B), and AP-1 (Figures 5A,B) were weakly expressed and mainly localized in the cytoplasm at quiescence state. After LPS stimulation, the expression of NF-κBp65, p50 and AP-1 were mainly localized in the nucleus (p < 0.05, Power = 1). The nuclear translocation rate of NF-κBp65, p50 and AP-1 in TAK-242 + LPS group was suppressed to 24.6, 25.2, and 19.9% of that in LPS-treated cells, respectively. It seems that TAK-242 pretreatment markedly reduced but not totally abolished the LPS-induced nuclear translocation of NF-κBp65, p50 and AP-1.

WB results showed that the expression of NF-κBp65 and p-NF-κBp65 were significantly increased in Veh + LPS group compared with Veh-treated cells, whereas the effect was completely reversed by TAK-242 (NF-κBp65: p < 0.05, Power = 0.998; p-NF-κBp65: p < 0.05, Power = 1, Figures 3C,D). Furthermore, LPS application increased the level of NF-κBp50 and IKKα protein and decreased the level of IκB-α in Veh + LPS group compared with Veh-treated group, whereas the effect was also completely reversed by TAK-242 (NF-κBp50: p < 0.05, Power = 0.998; IKKα: p < 0.05, Power = 1; IκB-α: p < 0.05, Power = 1, Figures 4C,D). In addition, WB results showed that the LPS-stimulated astrocytes exhibited a significant increase in AP-1 and p-AP-1 expression compared with Veh-treated group. Pretreatment with TAK-242 nearly completely inhibited the increased expression of AP-1 and p-AP-1 induced by LPS (AP-1: p < 0.05, Power = 0.999; p-AP-1: p < 0.05, Power = 1, Figures 5C,D). These results indicated that TLR4 activation is required for LPS-induced NF-κB and AP-1 activity, as well as their nuclear translocation.

To confirm the role of NF-κB or AP-1 in LPS-induced the expression of MCP-1 and CXCL1 in astrocytes, PDTC and SR11302 were used. As shown in Figure 6, both PCR and WB experiments demonstrated that, pretreatment with PDTC (1 μM, 30 min) or SR11302 (10 μM, 2 h) completely inhibited the expression of MCP-1 and CXCL1 at the mRNA (MCP-1: p < 0.05, Power = 1; CXCL1: p < 0.05, Power = 0.995) and protein levels (p < 0.05, Power = 1). It appears that LPS-induced increased expression of MCP-1 and CXCL1 is associated with the activation of the NF-κB or AP-1.

To further explore the DNA-binding activity of NF-κB or AP-1 to the promoter regions of MCP-1 and CXCL1 in astrocytes, EMSA was used to localize the transcription factors binding site (p65, p50, or AP-1) in the promoter regions of MCP-1 and CXCL1. As shown in Figures 7A,B, a single shift was observed in the presence of increasing amounts of purified nuclear extracts of cells activated for 24 h in the presence of LPS (p65-1: biotin 3’ end DNA-labeled probe containing a NF-κBp65-activated site element). Addition of excess cold unlabeled oligonucleotide completely abolished p65 bands confirming the specificity of the NF-κBp65 binding activity (p < 0.05, Power = 1). It appears that NF-κBp65 in astrocytes recognizes the sequence from bp 254 to 261 (5′-AGTTGAGGGGGACTTTCCCAGGC-3′) and may play a key role in the expression of MCP-1 in LPS-treated astrocytes. Although we tested several conditions, we could not observe a gel shift of the other probes (p65-2, AP-1-2, AP-1-3, and AP-1-4) with nuclear extracts from any groups (data not shown). The result does not support the idea that the binding of AP-1 to MCP-1 promoter is induced in LPS-treated astrocytes. In addition, as shown in Figures 7C,D, a single shift was observed in the presence of increasing amounts of purified nuclear extracts of cells activated for 24 h in the presence of LPS (p50-1: biotin 3’ end DNA-labeled probe containing a NF-κBp50-activated site element). Addition of excess cold unlabeled oligonucleotide completely abolished p50 bands confirming the specificity of the NF-κBp50 binding activity (p < 0.05, Power = 1). It appears that NF-κBp50 in astrocytes recognizes the sequence from bp 389 to 398 (5′-TGTGAGGTGACATCCCCAGATT-3′) and may be involved in LPS-induced MCP-1 gene expression.

Second, as shown in Figures 7E,F, a single shift was observed in the presence of increasing amounts of purified nuclear extracts of cells activated for 24 h in the presence of LPS (AP-1-6: biotin 3’ end DNA-labeled probe containing an AP-1-activated site element). Addition of excess cold unlabeled oligonucleotide completely abolished AP-1 bands confirming the specificity of the AP-1 binding activity (p < 0.05, Power = 1). It appears that AP-1 in astrocytes recognizes the sequence from bp 323 to 329 (5′-TTGGGATATGACTCTGGGGACA-3′) and may play a key role in the expression of CXCL1 in LPS-treated astrocytes. We could not observe a gel shift of the other probes (p65-3, p65-4, AP-1-5, and AP-1-7) with nuclear extracts from any groups (data not shown). This result seems does not support the idea that the binding of NF-κBp65 to CXCL1 promoter is induced in LPS-treated astrocytes.

As shown in Figures 7B,D,F, LPS application increased NF-κ Bp65-, NF-κ Bp50-, and AP-1-specific binding activities in cultured astrocytes. However, this activation was obviously diminished when cells were pretreated with PDTC or SR11302 (p < 0.05, Power = 1). It looks likely that LPS-induced MCP-1 and CXCL1 gene expression is related to the activation of NF-κB or AP-1.

Thus, to confirm whether SR11302 can suppress LPS-induced NF-κBp65 activation in astrocytes, we evaluated NF-κBp65 nuclear translocation rate. As shown in Figures 8A,B, immunofluorescence staining results showed that LPS-induced NF-κBp65 nuclear translocation was markedly reduced but not totally abolished by SR11302 (p < 0.05, Power = 1). Subsequently, compared with the Veh-treated astrocytes, the mRNA expression of NF-κBp65 and p50 in the LPS-treated astrocytes were significantly increased, while SR11302 significantly reversed this situation (p < 0.05, Power = 1, Figures 8C,D). The WB data confirmed this result. LPS-induced the increased expression of NF-κBp65, p-NF-κBp65, NF-κBp50, IKKα was significantly suppressed by SR11302 (p < 0.05, Power = 1, Figures 8E,F). In addition, pretreatment with SR11302 significantly inhibited LPS-induced IκB-α degradation (p < 0.05, Power = 1, Figures 8E,F).

To confirm the role of NF-κBp65 activation in LPS-induced AP-1 activity, we evaluated AP-1 nuclear translocation rate. Compared with the Veh + LPS-treated group, PDTC treatment nearly completely abolished LPS-induced AP-1 nuclear translocation (p < 0.05, Power = 1, Figures 9A,B). Meanwhile, compared with the Veh-treated astrocytes, LPS increased the expression of AP-1 (c-Jun) at the mRNA level (p < 0.05, Power = 1), while PDTC significantly reversed this situation (p < 0.05, Power = 1, Figure 9C). Moreover, LPS-induced the increased expression of AP-1 and p-AP-1 was significantly suppressed by PDTC (AP-1: p < 0.05, Power = 0.852; p-AP-1: p < 0.05, Power = 1, Figures 9D,E).

In the present study, immunofluorescence staining results indicated the expression of TLR4 on Cx43-labeled dorsal spinal cord astrocytes (Figure 10A). Compared with the Veh-treated astrocytes, LPS increased Cx43 expression at the mRNA and protein level, while the effect were significantly reversed by TAK-242 (p < 0.05, Power = 1, Figures 10B,D). Previously studies suggested that Cx43 phosphorylation at Ser368 is related to the dysregulation of cell-to-cell communication (Zhong et al., 2018). We found that TAK-242 treatment significantly inhibited the LPS-induced the increased p-Cx43 expression (p < 0.05, Power = 0.999, Figure 10D).

To confirm the role of TLR4-mediated NF-κBp65 and AP-1 activation in LPS-induced Cx43 expression, PDTC and SR11302 were used. As shown in Figures 10C,E, compared with the Veh + LPS-treated astrocytes, PDTC or SR11302 treatment significantly inhibited the LPS-induced Cx43 expression at mRNA (p < 0.05, Power = 1) and protein (p < 0.05, Power = 0.999) level, and its phosphorylation (p < 0.05, Power = 1).

Electrophoretic mobility shift assays were used to localize the transcription factors (NF-κBp65 or AP-1) binding site in Cx43 promoter. As shown in Figures 10H,I, a single shift was observed in the presence of increasing amounts of purified nuclear extracts of cells activated for 24 h in the presence of LPS (AP-1-1: biotin 3’ end DNA-labeled probe containing an AP-1-activated site element). Addition of excess cold unlabeled oligonucleotide completely abolished AP-1 bands confirming the specificity of the AP-1 binding activity (p < 0.05, Power = 1). It appears that AP-1 recognizes the sequence from bp 251 to 257 (5′-CGCTTGATGAGTCAGCCGGAA-3′) and may play a key role in Cx43 transcription in LPS-treated astrocytes. We could not observe a gel shift of the other probes (p65-5, p65-6, p65-7, AP-1-8, and AP-1-9) with nuclear extracts from any groups. This result seems does not support the idea that the binding of NF-κBp65 to Cx43 promoter is induced in LPS-treated astrocytes. Both PDTC and SR11302 significantly inhibited LPS-induced AP-1-specific binding activities in Cx43 promoter (p < 0.05, Power = 1). The NF-κB activation may indirectly affect LPS-induced AP-1-specific binding activities in LPS-induced Cx43 transcription.

EtBr uptake assay was used to measure the opening of Cx43 hemichannels in astrocytes. As shown in Figures 11A,C, LPS-treated astrocytes exhibited significantly greater EtBr uptake than that of the Veh-treated group (p < 0.05, Power = 1). However, pretreatment with TAK-242 (100 nM, 2 h), CBX (100 μM, 1 h), Gap26 (100 μM, 1 h) or Gap19 (100 μM, 1 h) all completely suppressed EtBr uptake than that of Veh + LPS-treated group (p < 0.05, Power = 1). It means that LPS application enhanced hemichannel activity in cultured astrocytes, whereas the effect was nearly totally reversed by TAK-242, CBX, Gap26, or Gap19.

The effect of LPS on the communication through gap junctions was explored by using a simple scrape loading/dye transfer assay. The diffusion distance of LY via gap junction migrates to neighboring cells and is considered as a marker to evaluate the activity of gap junction. As shown in Figures 11B,D, the diffusion distance of LY was longer in Veh-treated astrocytes. However, compared with the Veh-treated cells, the diffusion distance in Veh + LPS-treated cells was shortened upon LPS exposure (p < 0.05, Power = 1). It seems that LPS application significantly inhibited gap junction intercellular communication between astrocytes.

At last, the release of MCP-1 and CXCL1 from astrocytes was explores by using ELISA assay. As shown in Figures 11E,F, compared with Veh-treated astrocytes, LPS-induced the release of MCP-1 and CXCL1 was significantly increased (p < 0.05, Power = 1). Furthermore, LPS-induced the release of MCP-1 and CXCL1 were almost totally abolished in cultures co-incubated with TAK-242 (p < 0.05, Power = 1), CBX (p < 0.05, Power = 1), Gap26 (p < 0.05, Power = 1) or Gap19 (p < 0.05, Power = 1). It looks likely that TLR4-mediated Cx43 hemichannel opening is associated with LPS-induced the release of MCP-1 and CXCL1 from cultured astrocytes.