Estrogen receptor α phosphorylated at Ser216 confers inflammatory function to mouse microglia

Abstract

Background: Estrogen receptor α (ERα) has been suggested to regulate anti-inflammatory signaling in brain microglia, the only resident immune cells in the brain. ERα conserves the phosphorylation motif at Ser216 within the DNA binding domain. Previously, Ser216 was found to be phosphorylated in neutrophils infiltrating into the mouse uterus and to enable ERα to regulate migration. Given the implication of this phosphorylation in immune regulation, ERα was examined in mouse microglia to determine if Ser216 is phosphorylated and regulates microglia's inflammation. It was found that Ser216 was constitutively phosphorylated in microglia and demonstrated that in the absence of phosphorylated ERα in ERα KI brains microglia inflamed, confirming that phosphorylation confers ERα with anti-inflammatory capability. ERα KI mice were obese and weakened motor ability.

Methods: Mixed glia cells were prepared from brains of 2-days-old neonates and cultured to mature and isolate microglia. An antibody against an anti-phospho-S216 peptide of ERα (αP-S216) was used to detect phosphorylated ERα in double immunofluorescence staining with ERα antibodies and a microglia maker Iba-1 antibody. A knock-in (KI) mouse line bearing the phosphorylation-blocked ERα S216A mutation (ERα KI) was generated to examine inflammation-regulating functions of phosphorylated ERα in microglia. RT-PCR, antibody array, ELISA and FACS assays were employed to measure expressions of pro- or anti-inflammatory cytokines at their mRNA and protein levels. Rotarod tests were performed to examine motor connection ability.

Results: Double immune staining of mixed glia cells showed that ERα is phosphorylated at Ser216 in microglia, but not astrocytes. Immunohistochemistry with an anti-Iba-1 antibody showed that microglia cells were swollen and shortened branches in the substantial nigra (SN) of ERα KI brains, indicating the spontaneous activation of microglia as observed with those of lipopolysaccharide (LPS)-treated ERα WT brains. Pro-inflammatory cytokines were up-regulated in the brain of ERα KI brains as well as cultured microglia, whereas anti-inflammatory cytokines were down-regulated. FACS analysis showed that the number of IL-6 producing and apoptotic microglia increased in those prepared from ERα KI brains. Times of ERα KI mice on rod were shortened in Rotarod tests.

Conclusions: Blocking of Ser216 phosphorylation aggravated microglia activation and inflammation of mouse brain, thus confirming that phosphorylated ERα exerts anti-inflammatory functions. ERα KI mice enable us to further investigate the mechanism by which phosphorylated ERα regulates brain immunity and inflammation and brain diseases. Video abstract.

Keywords: Brain; Estrogen receptor; Inflammation; Microglia; Nuclear receptor; Phosphorylation.

Fig. 1

ERα phosphorylated at Ser216 in mouse brain microglia. a Microglia. Mixed glia cells were prepared, cultured for 4-days and double-stained with either an anti-ERα (in green) or an anti-P-S216 (in green) antibody and an anti-Iba-1 (in red) antibody for microglia marker and visualized as described in the Methods section. Nuclei were stained by DAPI in blue. b Enriched microglia. Microglia was enriched from glia cells as described in the Methods section for double staining. c Astrocytes. Mixed glia cells were double-stained with either an anti-ERα (in green) or an anti-P-S216 (in green) antibody and an anti-GFAP (in red) antibody for astrocyte marker and were visualized as described in the Methods section. Nuclei were stained by DAPI in blue. d Glia cells obtained from 2-days-old neonates of Ex3-ERα KO and wild type (WT) mice were cultured and double-stained by an anti-ERα or an anti-P-Ser-216 (P-S216) (in green) and an anti-Iba-1 (in red) antibodies and visualized as described in the Methods section. Nuclei were stained by DAPI in blue. e Whole lysates prepared from enriched microglia were subjected to Western blots by an anti-ERα or an anti- P-S216 antibody. An an anti-β-Actin antibody was utilized to verify equal amounts of proteins in an each well

Fig. 2

Generation and phenotypes of ERα S216A KI (Esr1^S216A) mice. a Map of strategy and Knock-in process. b Southern blot analyses were performed to identify the appropriately targeted allele. Genomic DNAs from mouse ear biopsies were digested with KpnI and BglI. KpnI-digestion generated 10.2- and 6.5-kb bands for knock-in (KI) and wild type (WT), respectively, detected by using the 5′-genomic probe. c Genotyping by PCR. Genomic DNA was isolated from mouse ear biopsies. PCR primers were used to amplify across the region in which the one remaining loxP site was inserted. This gave a band of 430 bp for a knock-in allele and 307 bp for a wild-type allele. Self-excision of the ACN cassette in the mutant allele was confirmed. d Expression of ERα mRNA. RNAs were prepared from WT and KI uteri and subjected to real time PCR analysis. These are averages from duplicates of two independent samples. e Conformation of ERα protein expression in WT and KI uteri. Whole extracts (15 μg proteins) prepared from WT and KI uteri were subjected to Western blots by an anti-ERα, an anti-P-Ser-216 (P-S216) or an anti-β-Actin antibody. f Sequence conformation of the mutation. cDNAs were synthesized from WT and KI uteri and sequenced. g ERα expression in ERα KI microglia. Accumulated microglia in the glia cells, which were prepared from brains of 2 days-old neonates of ERα KI males, were subjected to fluorescence staining with an anti-ERα (in green) antibody or double staining with an anti-Iba-1 (in red) antibody. DAPI stained nuclei in blue. h Obese phenotype of ERα S216A KI mice. Six-month-old mice were weighed; 5 males and 5 females of WT and 8 males and 4 females of KI mice. One Way ANOVA plus post hoc test with Tukey-Kramer’s multiple comparisons test (Version 5.0, Stat view-j) was used for statistical analysis. Values are presented as means ± S.E.. **, p < 0.01

Fig. 3

Microglia in substantia nigra of adult brains. a Saline or LPS (5 mg/kg, ip., 2 h)-treated brains. Brains were removed from 6-month-old ERα WT and KI males, from which frozen sections were prepared for immunostaining with an anti Iba-1 antibody. The brain region showing the microglia morphological changes is substantia nigra (SN). b Three brains were used for each group. Four sections from each brain were stained for analysis. Stained microglia were scanned to semi-quantitate microglia activation. Values are presented as means ± S.E., by taking values of ERα WT as 100%. Statistical analysis was conducted with One Way ANOVA plus post hoc test with Bonferroni’s multiple comparisons test (Version 7.00, GraphPad, San Diego, CA). **, p < 0.01

Fig. 4

Expression of inflammatory factors in adult brains. a qRT-PCR. Adult WT or KI male mice were treated with intraperitoneally injected LPS (5 mg/kg, ip.) or saline for each group for 2 h, from the whole brains of which RNAs were prepared for subsequent analysis. Assay was duplicated for each RNA sample. With 6 RT-PCR data for either WT or KI, Two Way ANOVA plus post hoc test with Bonferroni’s multiple comparisons test (Version 5.0, Stat view-j) was used for statistical analysis. Values are presented as means ± S.D.. *, p < 0.05, n = 3. b Cytokine antibody arrays. From the same brains used in (a), whole extracts were prepared, 300 μg proteins of which were subjected to analysis. c Mean Pixel Densities were obtained from each spot using Image J soft

Fig. 5

Expression of inflammatory factors in cultured microglia. a ELISA. Mixed glia cells were prepared from 5 neonates and cultured for 2 weeks and then treated with 1, 10 or 100 ng/mL of LPS or PBS for an additional 24 or 48 h. Media were recovered and subjected to ELISA for IL-6, PGE2 and IL-10. WT and KI denote ERα wild type and KI mice, respectively. ONE Way ANOVA plus post hoc test with Tukey-Kramer’s multiple comparisons test (Version 5.0, Stat view-j) was used for statistical analysis. Values are presented as means ± S.D.. **, p < 0.01, n = 4. b FACS. Mixed glia cells were separately prepared from male and female neonates, cultured for 2 weeks and then treated with 100 ng/mL of LPS or PBS for an additional 2 or 24 h. IL-6 production was determined by intracellular staining, followed by flow cytometry analysis. Assay was triplicated with three independent. ONE Way ANOVA plus post hoc test with Tukey-Kramer’s multiple comparisons test (Version 5.0, Stat view-j) was used for statistical analysis. Values are presented as means ± S.D.. **, p < 0.01. c Cell death was assessed by Annexin V staining, following by flow cytometry analysis. Assay was performed in triplicate eight times with three independent samples. These were separately prepared from male and female neonates. Two Way ANOVA plus post hoc test with Tukey-Kramer’s multiple comparisons test (GraphPad Prism) was used for statistical analysis. Values are presented as means ± S.D.. *, p < 0.05. d The gating strategy and the representative FACS plots for b and c

Fig. 6

Latency to fall in rotarod test. Three-month-old and 6-month-old male mice were tested as described in the Experimental Procedure section. These values are represented as mean ± S.D. of 7–8 mice for each group. Data were analyzed by Two Way ANOVA plus post hoc test with Bonferroni’s multiple comparisons test (version 7.02, GraphPad Prism) and are shown as scatter plots. *, p < 0.05