C57Bl/6J (B6) mice were housed in a 12:12 h light:dark cycle. Sham, OVX and 17β-estradiol pellet implantation were performed as previously described (20,21). All the OVX (n=12) and Sham (n=6) mice were 2 months old. The OVX mice had access to phytoestrogen-free food, receiving either a placebo pellet (OVX + placebo; n=6) or a 17β-estradiol pellet (OVX + E2; n=6). The pellets contained 0.18 mg of 17β-estradiol or placebo in a matrix that is designed to release contents (17β-estradiol or placebo) over a 60-day period (Innovative Research of America, Sarasota, FL, USA). Sham operations were performed (Sham; n=6) on an additional group of mice. Treatment with placebo or 17β-estradiol lasted for 4 weeks. Unless otherwise indicated, the reported results were obtained from homozygous mutants. The animal procedures were approved by the Institutional Animal Care and Use Committee of Hunan Agriculture University (approval no. 09-01).

Hippocampal neuronal cells were prepared from mouse embryos using the methods described previously (22). Briefly, female C57Bl/6J (B6) mice with 18.5 days of gestation were anesthetized with isoflurane, and embryonic mouse pups were surgically removed and decapitated. Hippocampuses were harvested under sterile conditions. The hippocampal tissue was enzymatically dissociated in 0.125% trypsin II (Sigma-Aldrich, St. Louis, MO, USA). Isolated neural cells were placed in poly-D-lysine-coated 35-mm plastic culture dishes containing 3 ml of neurobasal medium to a culture surface cell density of 5×10⁵/ml (400–500/mm²). The cultures were maintained in neurobasal medium supplemented with B27 (2%, v/v; Invitrogen, Carlsbad, CA, USA), glutamine (0.5 mM) and penicillin/streptomycin (100 units) for at least 7–10 days prior to being used for experiments. To test the potential function of ERs, hippocampal neuronal cultures were treated with 1 μM ER antagonist ICI182,780 (Sigma-Aldrich), or 1 μM selective ERα antagonist methyl-piperidino-pyrazole (MPP; Sigma-Aldrich), or 5 μM selective ERβ antagonist 4-[2-phenyl-5,7-bis(trifluoromethyl)pyrazolo[1,5-a]pyrimidin-3-yl] phenol (PHTPP; Sigma-Aldrich), with the presence of 10 nM 17β-estradiol (E2; Sigma-Aldrich). The antagonists and E2 were dissolved in dimethylsulfoxide (DMSO). Vehicle-treated controls received the same volume of DMSO. Medium was changed with fresh medium every 4 days.

Total RNA was prepared using TRIzol (Invitrogen), and an equal amount of RNA from each tissue (50–100 ng) was reverse transcribed using the SuperScript III first-strand synthesis system (Invitrogen). SYBR-Green PCR Master Mix for qRT-PCR was purchased from ABI (Applied Biosystems, Foster City, CA, USA). qRT-PCR was performed according to the manufacturer’s instructions, with minor modifications based on cDNA samples and primers. The primers used in this study were: Tnfaip1 forward, 5′-gcgtgctcttcatcaaggat-3′ and reverse, 5′-ttccaccttggtctgcttct-3′; β-actin forward, 5′-cggttccgatgccctgaggctctt-3′ and reverse, 5′-cgtcacacttcatgatggaattga-3′; and 40 cycles of amplification was achieved in a 96-well plate consisting of SYBR-Green Master Mix, primers and cDNA. The standard mode of an ABI 7500 Fast Sequence detection system was used. β-actin was amplified in parallel as an endogenous control to standardize the amount of sample and for the quantification of the relative gene expression within and between samples. Reactions were performed in triplicate.

Frozen tissue sections (10 mm) were treated with blocking solution containing 5% goat serum and 0.2% Triton X-100 at room temperature for 30 min. Primary antibodies were added at a dilution of 1:200, and incubated at 4°C overnight. After three washes in 1X phosphate-buffered saline (PBS), Alexa-568 (Molecular Probes, Carlsbad, CA, USA) conjugated secondary antibodies were added at a dilution of 1:600, together with DAPI (1:10,000), and incubated at room temperature for 2 h in the dark. Non-immune sera (instead of primary antibodies) were used as negative controls in the preliminary experiment. Slides were mounted in Fluoromount-G (Southern Biotech, Birmingham, AL, USA) and viewed using a Carl Zeiss Microscope Objective (Microscope Axio Observer.A1; Carl Zeiss Microscopy, LLC, San Diego, CA, USA).

To study the effect of estrogen on the expression of Tnfiap1 in vitro and in vivo, western blot assay was performed as previously described (8). The proteins were extracted with RIPA buffer and boiled together with loading buffer. After separation with SDS/PAGE, the proteins were transferred to a polyvinylidene difluoride membrane (Millipore), blocked in 5% non-fat milk for 1 h, and probed overnight with the rabbit polyclonal anti-Tnfaip1 (Abcam) in the presence of 5% non-fat milk. Immunoblots were washed 3 times (5–10 min each) in PBS containing 0.1% Tween-20 and incubated with the goat anti-rat-HRP (Amersham Biosciences) (1:10,000) for 1 h. Blots were washed 3 times (10 min each) in PBS containing 0.1% Tween-20, developed in ECL reagent (Millipore) for 2–5 min, and exposed to Blue XB-1 film (Kodak).

The promoter region of mouse Tnfaip1 (GenBank accession no. NM_009395.4) was predicted using the Promoter Inspector program (www.genomatix.de) and ProScan (www.bimas.cit.nih.gov). Transcription factor binding sites were predicted by online software JASPAR (http://jaspar.cgb.ki.se/). The CpG were also predicted by online software (http://www.bio-soft.net/sms/cpg_island.html). The promoter region (−2166/+61) was amplified from mouse genomic DNA using primer pairs with KpnI site at the forward primers and XhoI site at the reverse primers. A series of fragments including (−1072/+61), (−734/+61) and (−273/+61) were amplified from the promoter region. The primers used were: forward primers with KpnI site, for −2166/+61, 5′-ggggtacccagacacgtgaacctgaagga-3′; −1072/+61, 5′-ggggtacctgataaccgctagcagtggat-3′; −734/+61, 5′-ggggtaccgatgggtaagatagatatggt-3′; −273/+61, 5′-ggggtaccgtcccaagtcctccacg-3′; the common reverse primer with XhoI site, 5′-ccgctcgagagcagtgggtcgaaaacttgac-3′. The letters in bold italic refer to the recognition site of the restriction enzyme. These amplified fragments were inserted into the promoterless-luciferase gene upstream of plasmid pTAL-luc, separately denoted p(−2166/+61)-Luc, p(−1072/+61)-Luc, p(−734/+61)-Luc and p(−273/+61)-Luc. Site-directed mutagenesis for the ERβ binding site construct was performed by overlapping extension PCR as previously described (2) and termed p(−Δ734/+61)-Luc.

The primary hippocampal cells were plated in 6-well tissue culture dishes at 9×10⁴ cells/well and used for transient transfection at day 7 in vitro (DIV). Cells were cotransfected with 1.5 μg of luciferase reporter plasmid and 0.5 μg of β-gaglactosidase expression vector pCMVβ as an internal control. At 36 h after transfection, the cells were lysed to measure the luciferase activity using the Luciferase Assay System (Promega, Madison, WI, USA).

To study the regulation of ERβ on Tnfaip1 using the luciferase assay, the coding sequences of mouse ERβ were amplified from mouse hippocampal cDNA and cloned into the eukaryotic expression vector pCMV-Myc. The construct was designated as pCMV-Myc-ERβ. The luciferase reporter plasmid including the site-directed mutagenesis for the ERβ binding site and pCMV-Myc-ERβ were cotransfected.

To detect the association of ERβ with Tnfaip1, chromatin immunoprecipitation (ChIP) was performed using overexpressed ERβ protein. The primary hippocampal cells were cultured and transfected with an ERβ-expressing pCMV-Myc vector as previously described (23). ChIP was conducted following the recommended protocols (Active Motif). Specifically, enzymatic shearing was conducted for 10 min and IP was performed with the ERβ rabbit polyclonal IgG (Sigma-Aldrich) at a concentration of 3 μg/100 μl or the pre-immune mouse IgG as the negative control. Potential ERβ binding sites were predicted using the JASPAR TF Search program. PCR primers were designed flanking putative ERβ binding sites (forward, 5′-tcatgtggaaacagaaggctg-3′ and reverse primer, 5′-agttcttcctgtgtctacgcc-3′) and ChIP pull-down products were amplified for 30 cycles with a 60°C annealing temperature and imaged by agarose gel electrophoresis.

To study the function of Tnfaip1 in the nerve system, we compared the expression of Tnfaip1 in nerve system using quantitative PCR (qPCR). The 1.5 M mouse tissues including cerebral cortex, hippocampus, cerebellar, hypothalamus, and nerve in spine were dissected and qPCR was performed as described in Materials and methods. Positive signals were detected in all the tissues. In these tissues, the a significant difference in Tnfaip1 expression in hippocampus and cerebellar with a 45-fold difference was identified (Fig. 1A). This result suggested that Tnfaip1 may play a critical role in the nerve system, particularly in the hippocampus.

Previously it was reported that Tnfaip11 was involved in response to β-amyloid peptide accumulation in a transgenic C. elegans AD model (3). However, little is known with regard to the exact role of Tnfaip1 in these diseases. Evidence that the parahippocampal cortex (including the entorhinal cortex) undergoes pathological changes during the early stages of the disease is theoretically significant (24). To investigate the role of Tnfaip1 in these diseases, three check points were selected: embryonic (E17.5), postnatal (P6), and adult (1.5 M), with a focus on the expression of Tnfaip1 in hippocampus. The hippocampal tissues were dissected from mice of different age and gender. The results of qPCR showed that Tnfaip1 expression was low in E17.5, high in P6, and lowest in 1.5 M. Of note, at the 1.5 M stage, male mice had a stronger expression than the female mice (Fig. 1B).

To map Tnfaip1 in hippocampus, the brains including the hippocampus were dissected, the slides were cut, and fluorescent immunostaining was performed using anti-Tnfaip1 antibodies as primary antibodies. Tnfaip1 showed strong staining in the hippocampus of P6 mouse, but faint staining in the 1.5 M female hippocampus. The 1.5 M male exhibited stronger immunoreactivity than the female counterpart (Fig. 2). The results were consistent with qPCR results, showing age- and gender-related differences in the expression of Tnfaip1. The CA3 region can be divided into CA3a,b,c subareas (25). Fig. 2 shows data of only CA3a subareas, with the data of CA3b and c not being shown. Among the slides including all the ages and both genders, the pyramidal cells in CA3 regions including the CA3a,b,c subareas in the adult mouse have the strongest staining, higher than in the subregions of CA1 and DG (Fig. 2). These data suggested that Tnfaip1 plays an important role in short-term memory and pattern separation of the geometry of environment through the CA3 region.

Findings of previous studies suggest that the systemic estradiol treatment of ovariectomized animals caused an increase in the number of dendritic spines in this particular hippocampal region (18) and spine synapse density varied in response to fluctuating estrogen levels during the estrous cycle in female rats (19). To study the effect of systemic estrogen on the expression of Tnfaip1 in hippocampus, the OVX mice were implanted with placebo pellet or 17β-estradiol pellet for 4 weeks. The entire brain tissue was dissected for fluorescent immunostaining and the hippocampus was dissected for western blot or real-time analysis. The Sham mice were used as the control. The results with fluorescent immunostaining showed that Tnfaip1 had stronger staining in the hippocampal CA3 subregion of OVX + placebo mice than the OVX + E2 or Sham mice (Fig. 3A–C). The staining was not different in the CA1 or DG subregions (data not shown). Furthermore, the mRNA level (Fig. 3D) and protein level (Fig. 3E) of Tnfaip1 in hippocampus was higher in the OVX + placebo mice than in the OVX + E2 (P<0.001) or in Sham mice (P<0.001). These results suggested that systemic estrogen inhibits Tnfaip1 expression in hippocampus.

To examine the effect of estrogen on Tnfaip1 expression, hippocampal neuronal cells were prepared from mouse embryos and cultured in the presence or absence of E2. Tnfaip1 expression was lower in the cells treated with estrogen compared to the control (Fig. 4).

The classic estrogen receptors, known as ERα and ERβ, are nuclear transcription factors that activate or repress gene expression (12). To determine whether ERα and ERβ were involved in the inhibition progress, the cells were treated with DMSO (control) or estrogen receptor antagonists (ICI182,780, MPP or PHTPP) in the presence of E2. Inhibition of Tnfaip1 expression by E2 was reversed by ER non-selective antagonist ICI182,780 and ERβ-selective antagonist PHTPP, but not by ERα-selective antagonist MPP. These data suggested that estrogen downregulates Tnfaip1 expression in in vitro hippocampal cells, which was consistent with the in vivo results. Additionally, this function of estrogen may be mediated by ERβ.

To examine the regulation of ERβ on Tnfaip1 expression, a promoter analysis was performed. The promoter region of mouse Tnfaip1 was predicted to be a 2166-bp genomic region. A series of 5′-deletion promoter regions upstream of the first exon of the Tnfaip1 gene were amplified and inserted into the promoterless luciferase reporter construct. Insertion of longer regions (−2166/+61, −1072/+61 and −734/+61) significantly enhanced the promoter activity compared to the promoterless luciferase control (Fig. 5A). The deletion −734/+61 had the strongest enhancement, followed by the −1072/+61 deletion. Both were stronger than the longest insertion (−2166/+61). However, the deletion −237/+61 showed a significant decrease in promoter activity. These results showed that the upstream region −2166/+61 of the mouse Tnfaip1 gene contained a functional promoter and the main regulatory element resided in the range of −734 to +61.

To further determine the potential estrogen response element (ERE), the region −734 to +61 was scanned using the JASPAR and CpG-prediction software. No classical TATA box or CCAAT box was found in the promoter region. Instead, a CpG island was detected from −607 to +21, whose sequence was shaded in Fig. 5B. A potential ERβ (known as ERS2 in the software) binding site was identified in this region. However, the ERα binding site was not found (Fig. 5B). To study the regulation of ERβ on Tnfaip1 expression, the p(−734/+61)-Luc or the mutant p(−Δ734/+61)-Luc was co-transfected with pCMV-Myc-ERβ into cultured cells. The luciferase activity analysis revealed that the overexpression of ERβ enhanced the Tnfaip1 promoter activity, compared to the empty or mutated vector (Fig. 5C). To confirm the involvement of ERβ in the Tnfaip1 promoter regulation, we performed ChIP assay using the hippocampal primary cells. (Fig. 5D) The cross-linked mouse Tnfaip1 promoter-ERβ complexes immunoprecipitated with anti-ERβ antibody were detected by PCR amplification with primers covering the potential binding sites of mouse Tnfaip1 promoter. No band was detected with a non-specific antibody as a negative control. These results demonstrated that ERβ can directly bind to the Tnfaip1 promoter, providing novel evidence for Tnfaip1 transcriptional regulation.