Lipopolysaccharide induces mouse translocator protein (18 kDa) expression via the AP-1 complex in the microglial cell line, BV-2

It has been reported that neuroinflammation occurs in the central nervous system (CNS) in patients with neuropathic pain, Alzheimer’s disease and autism spectrum disorder. The 18-kDa translocator protein TSPO is used as an imaging target in positron emission tomography to detect neuroinflammation, and its expression is correlated with microglial activation. However, the mechanism underlying the transcriptional regulation of Tspo induced by inflammation is not clear. Here, we revealed that lipopolysaccharide (LPS) -induced Tspo expression was activated by the AP-1 complex in a mouse microglial cell line, BV-2. Knockdown of c-Fos and c-Jun, the components of AP-1, reduced LPS-induced Tspo expression. Furthermore, the enrichment of Sp1 in the proximal promoter region of Tspo was increased in the presence of LPS. In addition, the binding of histone deacetylase 1 (HDAC1) to the enhancer region, which contains the AP-1 site, was decreased by LPS treatment, but there were no significant differences in HDAC1 binding to the proximal promoter region with or without LPS. These results indicated that HDAC1 is involved not in the proximal promoter region but in the enhancer region. Our study revealed that inflammatory signals induce the recruitment of AP-1 to the enhancer region and Sp1 to the proximal promoter region of the Tspo gene and that Sp1 may regulate the basal expression of Tspo.


Introduction
Microglia are resident immune cells of the central nervous system (CNS) and play a pivotal role in maintaining the neuronal environment. Resting  Animals and primary microglial cell culture Male C57BL/6 mice approximately 12 weeks old were used in this study. Mice were housed as described previously [19]. This study was carried out in accordance with the recommendations the guidelines for animal research issued by the Physiological Society of Japan. The experimental procedures were approved by the Animal Research Committee of Hirosaki University (Approval Number: M12007), and all efforts were made to minimize the number of animals used and their suffering. Isolation of microglia from adult mice and primary microglial cell culture were performed as described before [20]. Primary microglial cells were seeded in 35-mm dishes at a density of 1.0×10 4 cells/dish. Twenty-four hours after seeding, cells were stimulated with 100 ng/ml LPS for 8 hours.

RNA isolation, cDNA synthesis and quantitative real-time reverse transcription PCR (qRT-PCR)
BV-2 cells were seeded in 35-mm dishes at a density of 1.0×10 5 cells/dish. After chemical treatment, cells were harvested by centrifugation. Total RNA was extracted from cells using ISO-GEN II (NIPPON GENE, Tokyo, Japan) according to the manufacturer's instructions. cDNA synthesis and qRT-PCR were performed as described previously [19]. The primer sets used in this study are shown in S1 Table. The expression level of each transcript was measured using the standard curve method, Gapdh as an internal control.

Preparation of whole-cell extracts and cell fractionation, Western blotting and densitometric analysis
Whole-cell extracts were prepared as described previously [21]. BV-2 cells were seeded at a density of 1.0×10 5 cells/35-mm dish (for whole-cell extracts) or at a density of 1.0×10 6 cells/ 100-mm dish (for cell fractionation). The mitochondrial, membrane, cytosolic and nuclear fractions were prepared according to the following method. After washing the cells with cold PBS (-) containing 1 mM PMSF, cells were harvested via centrifugation. Cells were suspended in hypotonic buffer (10 mM HEPES [pH7.9], 10 mM KCl, 1 mM MgCl 2 , 0.5 mM DTT and Complete™ Protease Inhibitor Cocktail: Roche, Mannheim, Germany) and incubated on ice for 15 min. Cells were homogenized using a Dounce Tissue Grinder with a tight pestle (WHEATON, Millville, NJ, USA). Nuclei and unhomogenized cells were precipitated via centrifugation at 3,300×g for 5 min at 4˚C, and then, the supernatants were then centrifuged at 8,000×g for 15 min at 4˚C to obtain the mitochondrial fraction. In addition, the supernatants were centrifuged at 100,000×g for 60 min at 4˚C, and the pellets were recovered as the membrane fraction. The resulting supernatants were used as the cytosolic fraction. The nuclear, mitochondrial and membrane fractions were washed 3 times with isotonic buffer (20 mM HEPES [pH7.9], 150 mM KCl, 3 mM MgCl 2 , 0.5 mM DTT and Complete™ Protease Inhibitor Cocktail) and were then suspended in isotonic buffer containing 1% SDS. The protein concentrations were determined by Bradford protein assay (Bio-Rad, Hercules, CA, USA). Western blotting and densitometric analysis were performed as described previously [19]. The antibodies used in this study are shown in S2 Table. Immunocytochemistry BV-2 cells were seeded on cover glasses at a density of 2.0×10 4 cells/well (Nunc™ Cell-Culture Treated Multidishes, Thermo Fisher Scientific, Waltham, MA, USA). After chemical treatment, cells were fixed with 4% paraformaldehyde/PBS (-) for 15 min at room temperature and were then permeabilized with 0.1% Triton X-100/PBS (-) for 15 min at room temperature.
After blocking with 1% BSA/PBS (-) for 60 min at room temperature, cells were incubated  with the primary antibody at 4˚C overnight. Cells were incubated with the secondary antibody  for 2 hours at room temperature, and nuclei were then stained with DAPI for 15 min at room  temperature. Cells on the cover glasses were mounted with VECTASHIELD1 Mounting  Medium (VECTOR LABORATORIES, Burlingame, CA, USA). All steps were carried out in the dark and cells were washed 3 times with PBS (-) or 0.1% Tween 20/PBS (-) after incubation. The antibodies and dilution ratios used in this study are shown in S2 Table. Images were acquired using a confocal laser scanning microscope and analyzed with NIS-Elements AR Analysis software (Nikon, Tokyo, Japan). To label mitochondria, cells were stained with Mito-Tracker Orange1 CMTMRos (Thermo Fisher Scientific) in accordance with the manufacturer's instructions.

Reporter gene assay
The regulatory region of the mouse Tspo gene was amplified using forward primers and a common reverse primer with the addition of the Spe I (forward) or Bam HI (reverse) sites as described in S1 Table. The products were ligated to the pMCS-Gaussia Luc Vector (Thermo Fisher Scientific), and the sequences were confirmed via sequencing (S1 Table) using a Big-Dye1 Terminator v3.1 Cycle Sequencing Kit (Thermo Fisher Scientific); the reaction products were then purified using a Gel Filtration Cartridge (Edge Bio, San Jose, CA, USA). The sample sequences were analyzed using an ABI PRISM1 3100 Genetic Analyzer (Applied Biosystems, Waltham, MA, USA). Various constructs were transfected into BV-2 cells using Attractene (QIAGEN, Hilden, Germany) in accordance with the manufacturer's instructions. A reporter gene assay was performed using a Gaussia Luciferase Glow Assay Kit (Thermo Fischer Scientific) according to the manufacturer's instructions.

Chromatin immunoprecipitation (ChIP) assay and real-time PCR
A ChIP assay was performed as described previously [21]. The antibodies used in this study are shown in S2 Table. Purified DNA was subjected to real-time PCR with primer pairs specific for the Tspo enhancer region containing the AP-1 binding site or the proximal promoter region containing the Sp1 binding site, as described in S1 Table. All quantified values were normalized to that of the corresponding input for each sample using the standard curve method.

Statistical analyses
All numerical data are presented as the means ± S. Ds, with n = 3 (qRT-PCR, Western blotting and ChIP assay) or n = 4 (reporter gene assay). Comparisons were assessed using Student's ttest or ANOVA. A value of p<0.05 was considered statistically significant. Statistical analyses were performed using StatPlus:mac Pro software (AnalystSoft, Walnut, CA, USA).

LPS treatment induced Tspo expression in BV-2 cells
To determine the optimum assay conditions, BV-2 cells were treated with LPS for various durations and at various concentrations. The expression of Tspo mRNA was induced by a high concentration of LPS (100 or 500 ng/ml), peaked at 8 hours and gradually returned to normal. By contrast, the expression of Tspo mRNA did not increase at low concentrations of LPS (1 or 10 ng/ml). Under the same conditions, the expression of pro-inflammatory mediators such as Tnf-α, Il-1β and Il-6 was also upregulated by LPS stimulation (Fig 1A). Furthermore, the transcriptional activation of Tspo was induced by LPS stimulation in primary microglial cells from adult mice (Fig 1B).
Subsequent to analysis of the Tspo mRNA expression levels, Western blotting and densitometric analysis were also performed to the measure protein levels (Fig 1C and 1D). The protein expression of TSPO was correlated with its mRNA expression. In addition, the results of Western blotting and immunocytochemistry revealed that TSPO expression was increased at high concentrations of LPS ( Fig 1E and S1 Fig). These results indicated that LPS was a potent inducer of Tspo in mouse microglial cells, as reported previously [22]. In addition, we examined whether proinflammatory mediators such as IFN-γ or IL-1β induce the expression of Tspo. BV-2 cells were stimulated with 100 ng/ml IFN-γ and/or IL-1β for 8 hours. The expression of Tspo mRNA was also induced by administration of IFN-γ, IL-1β, or both (S2 Fig).

LPS treatment increased TSPO expression in the mitochondrial and membrane fractions
The TSPO protein is mainly localized in mitochondria in the adrenal gland homogenate [23]. To investigate whether TSPO in microglial cells is localized in mitochondria and whether the intracellular localization of TSPO is affected by LPS treatment, subcellular fractionation analysis was performed. BV-2 cells were collected, and cell fractions were prepared for Western blotting after stimulation with 100 ng/ml LPS for 8 hours. First, the purity of the fractions was confirmed by fraction markers, such as COXIV, the alpha1 subunit of the Na + /K + ATPase, GAPDH and Lamin-B1, and there was no contamination or variation in expression between the control and LPS-treated groups in the same fractions. Similar to the fractionation analysis results, TSPO expression was significantly increased in the mitochondrial fraction by LPS treatment (Fig 2A and 2B, p = 0.03). Interestingly, TSPO was also localized in the membrane fraction in the control group and was 1.94-fold higher in the LPS-treated group than in the control group (p = 0.005). To verify whether TSPO is localized in the plasma membrane, TSPO, the Na + /K + ATPase and mitochondria were stained by immunocytochemistry and analyzed by laser confocal microscopy. As previously reported, almost all TSPO protein was localized in the mitochondrial outer membrane. However, some fluorescence signals corresponding to TSPO protein were detected in non-mitochondrial regions in both the absence and presence of LPS (Fig 2C, upper, arrowhead). In addition, some TSPO signals were colocalized with or very close to Na + /K + ATPase signals (Fig 2C, lower, arrowhead). These results indicated that a small amount of TSPO was localized in the plasma membrane.

The transcription factor AP-1 is the candidate transcriptional activator of Tspo after LPS treatment
Prior to the construction of vectors for the reporter gene assay, the Tspo regulatory sequence was analyzed using ChIP-seq data in the UCSC Genome Browser (http://genome.ucsc.edu/ ENCODE/) and prediction tools for transcription binding sites, such as STAMP (http://www. benoslab.pitt.edu/stamp/), JASPAR (http://jaspardev.genereg.net/) and DBCLS Galaxy (http:// galaxy.dbcls.jp/). The presumed transcription factors and their binding sites were identified via a comprehensive assessment using several tools and are shown in Fig 3A. To determine which transcription factor and cis element regulates Tspo expression during LPS treatment, serial deletion constructs were designed for the Tspo reporter gene assay according to the schematic in Fig 3A. Forty-eight hours after transfection of each vector, BV-2 cells were treated with vehicle or 100 ng/ml LPS for an additional 8 hours, and a reporter gene assay was then performed. As shown in Fig 3B, deletion of the AP-1 site (-517) significantly reduced luciferase activity; however, basal expression was not altered. These results indicated that AP-1 was the candidate transcriptional activator of Tspo during LPS stimulation. (C) BV-2 cells were treated as described above. The protein expression level of TSPO was measured using Western blot analysis. (D) Densitometric analysis of TSPO expression using ImageJ. The expression value was measured and normalized to that of GAPDH as the internal control. The asterisks indicate statistically significant differences (n = 3, � p<0.05, �� p<0.01). (E) Microscopy images showing TSPO expression. BV-2 cells were stimulated with 1, 10, 100 or 500 ng/ml LPS for 8 hours and immunocytochemistry was then carried out. Images of nuclear (DAPI, blue) and TSPO (green) staining were acquired at 40× magnification. Scale bar: 100 μm. https://doi.org/10.1371/journal.pone.0222861.g001

AP-1, Sp1 and HDAC1 regulated Tspo expression
Based on the results of the reporter gene assay and previous studies [24,25], we examined via a ChIP assay whether the enrichment of AP-1 at the binding site in the Tspo regulatory region was promoted by LPS treatment. A previous study showed that TSPO expression is regulated by the binding of Sp1, Sp3 and/or Sp4 to the proximal promoter region in human breast cancer cell lines, and a similar sequence was observed in the mouse Tspo gene [26]. In addition, there is the possibility that HDAC1 acts as an inhibitory factor of Tspo expression by binding to Sp1 in a transcriptionally repressed state [21,27]. Furthermore, since few studies have examined the interaction between HDAC1 and Sp3 or Sp4, we excluded Sp3 and Sp4 from this study and focused on the interaction between Sp1 and HDAC1 in the proximal promoter region of the Tspo gene. Based on these observations, the primer sets used for the ChIP assay were designed as shown in the schematic in Fig 4A. The ChIP experiment revealed that the enrichment of the AP-1 components c-Fos and c-Jun at the AP-1 binding site was increased by approximately 3.83-and 2.88-fold, respectively, in the presence of LPS (Fig 4B). In the proximal promoter region, the enrichment of RNA polymerase II (RNAP II) and Sp1 was increased at the time of transcriptional activation compared to that in control cells (Fig 4C). These results clearly showed that the binding of AP-1 and Sp1 the upstream region of the Tspo gene was increased by LPS treatment. Interestingly, after LPS treatment, enrichment of HDAC1 was decreased in the enhancer region, which contains the AP-1 site, but enrichment of HDAC1 was not altered in the proximal promoter region.

Silencing of c-Fos and c-Jun decreased Tspo expression
To confirm the involvement of AP-1 in Tspo expression, silencing of c-Fos and c-Jun was performed via RNAi. First, Western blot analyses were performed to evaluate the knockdown efficiency of c-Fos and c-Jun. Forty-eight hours after transfection of c-Fos siRNA, c-Jun siRNA or siCont, cells were harvested and subjected to Western blot analysis of c-Fos and c-Jun. As shown in Fig 5A and 5B, the expression of c-Fos and c-Jun was clearly decreased by siRNA compared to siCont. Under these conditions, cells were stimulated with 100 ng/ml LPS for an additional 8 hours and analyzed to measure the expression level of Tspo mRNA. The results of qRT-PCR analysis revealed that knockdown of c-Fos or c-Jun repressed the LPS-promoted expression of Tspo (Fig 5C). Similar results were obtained at the protein level for TSPO ( Fig  5D and 5E). Therefore, c-Fos and c-Jun are essential transcriptional activators of LPS-induced Tspo regulation in mice.

Discussion
In this study, LPS-induced gene expression of Tspo in a microglial cell line was examined. Our experiments clearly demonstrated that AP-1, Sp1 and HDAC1 are involved in the transactivation of Tspo expression. Specifically, the release of HDAC1 from the enhancer region of the Tspo gene in the presence of LPS was a novel finding.
The transcription factors c-Jun and STAT3 bind to the enhancer region of the Tspo gene in the mouse cell lines MA-10 and NIH-3T3 via PKCε activation through phorbol 12-myristate 13-acetate (PMA) stimulation [25]. PKCε activation via PMA activates the Raf1, MEK1/2 and ERK1/2 pathways. This signal transduction promotes Tspo gene expression following the activation and DNA binding of STAT3 and c-Jun. Our results indicating that LPS stimulation led to the binding of c-Fos and c-Jun to the enhancer region were partially consistent with those of previous studies. In general, LPS binds directly to Toll-like receptor 4 (TLR4) [28] and activates downstream factors such as mitogen-activated protein kinases (MAPKs) [29,30]. The MKK3/6 and p38 MAPK pathways induce phosphorylation of c-Fos and enhance transcriptional activity [31]. Additionally, c-Jun phosphorylation is induced by TLR4 signaling via the MKK4/7 and JNK pathways, and dimerization of c-Jun and c-Fos is promoted [32,33]. Collectively, these results indicate that LPS treatment is likely to activate the AP-1 complex in a manner mediated by several transduction pathways, such as the MAPK pathway, and that Tspo expression is consequently promoted by the binding of the AP-1 complex to DNA. In contrast to the facilitation of AP-1 binding, the enrichment of HDAC1 at the AP-1 binding site was decreased by LPS treatment. The reduction in HDAC1 enrichment in the enhancer region by LPS treatment possibly led to hyperinduction of Tspo expression via recruitment of AP-1 or another transcriptional activator. A recent study revealed that deletion of sequences between -593 and -520 bp from the TSS, which contains the AP-1, Stat3 and Ets.2 cis elements, abolished the transactivation of Tspo gene expression in BV-2 and Raw-264.7 macrophages [34]. Since our results were consistent with those of that recent report, it is convincingly concluded that the AP-1 and Sp1 are involved in the LPS-induced transcriptional regulation of Tspo. In addition to the involvement of AP-1 and Sp1, Rashid et al reported that the Pu.1 and its binding motif, Ets, play an important role in the transactivation of Tspo gene expression. Prediction tools for transcription binding sites were used in this study; however, Pu.1/Ets was not identified as a candidate transcription factor for the regulation of the Tspo gene. Contrary to the prediction results, the Ets sequence (5'-GGAA-3') is located 3 bp from the AP-1 binding site and acts as a transcriptional activator of Tspo in response to LPS stimulation.
In addition to LPS stimulation, treatment with IFN-γ and IL-1β also induced Tspo transcription. Generally, IFN-γ activates the JAK-STAT pathway, and STAT acts as a transcriptional activator of inflammatory genes. The consensus sequence of STAT3 is located the upstream region of the Tspo gene, which overlaps with the AP-1 and Ets sites [25]. The IL-1β signaling pathway activates several transcription factors, including AP-1, via MAPKs such as p38, JNK and ERK1/2 [35]. These lines of evidences indicate that the general inflammatory response induces the expression of Tspo in microglial cells via the activation of transcription factors such as AP-1 or STATs.
Sp1, which belongs to the C 2 H 2 -type zinc-finger protein family, binds to GC-rich motifs and regulates numerous genes [36]. A previous study showed that there are several GC-rich sequences in the proximal promoter region of the mouse Tspo gene [26]. In this study, LPS treatment of BV-2 cells increased the enrichment of Sp1 in the proximal promoter region of Tspo. This evidence suggests that Sp1 binds to GC-rich sequences to regulate various genes by interacting with a transcriptional activator or repressor such as CBP/p300 or HDAC1 [37,38]. In the proximal promoter region of the Tspo gene, Sp1 enrichment was increased by LPS treatment, but HDAC1 enrichment was not affected. This result indicated that HDAC1 localized in the proximal promoter region was not involved in the pathway downstream of TLR4 activation by LPS. Furthermore, the finding that Tspo was expressed to some extent even under the control condition suggested that HDAC1 might minimally bind to the proximal promoter region under physiological conditions.
The expression of TSPO can be upregulated in noninflammatory conditions. According to the Human Brain Transcriptome database (http://hbatlas.org/) [39], human TSPO expression increases during the embryonic to the prenatal period and decreases during the neonatal period, and the TSPO expression level in the neonatal period is maintained throughout life. These data suggest that the functions of TSPO, such as increasing cholesterol transport from the cytosol to mitochondria and promoting pregnenolone production, might be important for fetal development. In addition, there is a possibility that the increase in TSPO expression during the neural developmental stage might be reflective of activated microglia. During neural development, microglia are activated and maintain CNS homeostasis by removing apoptotic cells, myelin debris and other endogenous redundant debris [40]. It is unknown which transcription factor or factors regulate the transcription of the TSPO gene in the fetal period; however, the unknown pathway might activate AP-1 and Sp1 or a different transcription factor.
In this study, we found that TSPO was not only localized in mitochondria but also in the plasma membrane fraction, as evidenced by the cell fractionation, Western blotting and immunocytochemical analysis results. The results of Western blot analysis indicated that not only the mitochondrial fraction but also in the membrane fraction contained TSPO. Consistent with this finding, the immunocytochemical results revealed that TSPO was mainly colocalized with mitochondria and that small amounts of TSPO signals were not distant from but were closely localized with Na + /K + ATPase signals regardless of LPS stimulation. Generally, the Na + /K + ATPase localizes to the transmembrane region and is used as a marker of the plasma membrane. However, the plasma membrane is not solely constituted by the Na + /K + ATPase. Consequently, the TSPO signals were not completely colocalized with the plasma membrane signals. In addition to mitochondria and the plasma membrane, other organelles appeared to feature TSPO immunostaining signals. Thus, as it is reported that not all of the TSPO protein colocalizes with mitochondria, it is possible that TSPO localizes in organelles other than mitochondria [22]. Further research is needed to assess the localization and physiological function of TSPO induced by LPS stimulation.
In patients with ASD, the binding potential of [ 11 C] (R) PK-11195 to TSPO is increased in multiple brain regions, such as the cerebellum, brainstem and anterior cingulate cortex (ACC) [9]. These results indicate that the activation of microglia, i.e., neuroinflammation, and the expression of TSPO are increased. However, the conditions in our ex vivo study using a microglial cell line and LPS differed from the those of the CNS in patients with ASD; therefore, further investigation is needed to clarify the transcriptional regulation of Tspo using an animal model of ASD. It is necessary to elucidate whether this alteration in TSPO is due to transcriptional activation via AP-1 or a different transcriptional factor or factors in patients with ASD or animal model of ASD.
It is also important to clarify whether expressed TSPO leads to biological protection against or promotion of inflammation in the CNS. Because TSPO expression is increased under inflammatory conditions, it is thought that TSPO may act as a mediator of inflammation [41]. However, a recent study reported that TSPO acts as a negative regulator of inflammation [22]. Furthermore, TSPO overexpression rescues LPS-induced cognitive impairment. TSPO reduced inflammatory cytokines production and microglial activation, and neurogenesis in the hippocampus was promoted. Consequently, LPS-induced cognitive dysfunction was ameliorated [42]. TSPO may play a neuro-protective role in inflammation, and thus, it could be expected that elucidating the mechanism of TSPO induction under conditions other than inflammation could lead to the development of molecular targeted drugs.
Supporting information S1 Table. Oligonucleotide sequences used in this study. Oligonucleotide sequences used for real-time PCR and vector construction; sequencing primers used for the vectors and ChIP assay. (XLSX) S2 Table. List of antibodies used in this study. Antibodies used in this study are shown with their supplier, product number and experimental procedure. WB: Western blotting, ICC: Immunocytochemistry, ChIP: Chromatin immunoprecipitation (XLSX) S1 Fig. Microscopy images showing TSPO expression. BV-2 cells were stimulated with 1, 10, 100 or 500 ng/ml LPS for 4, 8, 12 or 24 hours and immunocytochemistry was then carried out. Images of nuclear (DAPI, blue) and TSPO (green) staining were acquired at 40× magnification. Scale bar: 100 μm. (TIFF) S2 Fig. IFN-γ and IL-1β induced Tspo expression in the microglial cell line, BV-2. BV-2 cells were treated with 100 ng/ml IFN-γ and/or IL-1β for 8 hours and RNA was then extracted and subjected to qRT-PCR. The mRNA levels were measured and normalized to those of Gapdh mRNA. The asterisks indicate statistically significant differences compared to the control (n = 3, � p<0.05, �� p<0.01). (TIFF)