JMJD3 Is Crucial for the Female AVPV RIP-Cre Neuron-Controlled Kisspeptin–Estrogen Feedback Loop and Reproductive Function

Abstract

The hypothalamic-pituitary-gonadal axis controls development, reproduction, and metabolism. Although most studies have focused on the hierarchy from the brain to the gonad, many questions remain unresolved concerning the feedback from the gonad to the central nervous system, especially regarding the potential epigenetic modifications in hypothalamic neurons. In the present report, we generated genetically modified mice lacking histone H3 lysine 27 (H3K27) demethylase Jumonji domain-containing 3 (JMJD3) in hypothalamic rat-insulin-promoter-expressing neurons (RIP-Cre neurons). The female mutant mice displayed late-onset obesity owing to reduced locomotor activity and decreased energy expenditure. JMJD3 deficiency in RIP-Cre neurons also results in delayed pubertal onset, an irregular estrous cycle, impaired fertility, and accelerated ovarian failure in female mice owing to the dysregulation of the hypothalamic–ovarian axis. We found that JMJD3 directly regulates Kiss1 gene expression by binding to the Kiss1 promoter and triggering H3K27me3 demethylation in the anteroventral periventricular (AVPV) nucleus. Further study confirmed that the aberrations arose from impaired kisspeptin signaling in the hypothalamic AVPV nucleus and subsequent estrogen deficiency. Estrogen replacement therapy can reverse obesity in mutant mice. Moreover, we demonstrated that Jmjd3 is an estrogen target gene in the hypothalamus. These results provide direct genetic and molecular evidence that JMJD3 is a key mediator for the kisspeptin–estrogen feedback loop.

The hypothalamus controls multiple physiological functions, including reproduction and metabolic homeostasis, through the neuroendocrine system (1, 2). In the arcuate nucleus (ARC) of the hypothalamus, rat-insulin-promoter–expressing neurons (RIP-Cre neurons) were identified as essential for regulating both food intake and energy expenditure (3–8). In the ARC and the anteroventral periventricular (AVPV) nuclei, kisspeptin (encoded by the Kiss1 gene) expressing neurons are vital for the initiation of puberty and the maintenance of the female reproductive cycle (2, 9, 10). These neurons can sense a multitude of peripheral signals, such as adipocyte-secreted leptin and gonadal steroid estrogen, and coordinate effective feedback action aimed at maintaining normal physiological homeostasis. For example, ovarian estrogen exerts important feedback action in the reproductive axis through estrogen receptor-α (ERα) signaling in kisspeptin neurons (11–14). Deletion of ERα signaling in kisspeptin neurons results in impaired estrogen feedback and reproductive defects in female mice (14, 15). Estrogen controls puberty onset and the preovulatory luteinizing hormone surge by upregulating kisspeptin expression in AVPV neurons and regulates gonadotropin-releasing hormone pulse generation by downregulating kisspeptin expression in ARC (16). Furthermore, estrogen plays a vital role in the regulation of energy balance through ERα signaling in the hypothalamic neurons (17, 18). Estrogen deficiency in postmenopausal women or in ovariectomized (OVX) mice is associated with obesity, which occurs primarily because of reduced locomotor activity and energy expenditure (19–21). Nevertheless, the direct target genes of estrogen in hypothalamic neurons are largely unknown.

DNA methylation and histone modifications control gene expression by altering chromatin structure or affecting the recruitment of transcription factors to the gene’s promoter region (22, 23). Accumulating evidence has suggested that these epigenetic modifications regulate neuroendocrine-related reproductive and metabolic functions. The initiation of puberty and estrogen-positive feedback action are accompanied by altered histone acetylation/methylation and cytosine-phosphate-guanine island methylation in the promoter of the Kiss1 gene (24–27). Furthermore, different diets and nutritional status invoke epigenetic changes in the hypothalamus in terms of DNA promoter methylation and histone modification of specific genes encoding orexigenic and anorexigenic hypothalamic neuropeptides (28–30). Abundance of H3 lysine 27 (H3K27)me3 at the Kiss1 promoter markedly decreases during the pubertal process and fasting has a stimulatory effect on trimethylation of H3K27 in hypothalamic neurons (25, 30). It is reasonable to assume, therefore, that the enzymes responsible for H3K27 methylation and demethylation might participate in the neuroendocrine regulation of reproduction and metabolic homeostasis.

Jumonji domain-containing 3 (JMJD3), also known as lysine-specific demethylase 6B is a histone demethylase that specifically catalyzes the demethylation of H3K27me2/3 methyl marks and transcriptionally activates gene expression (31–34). Recent studies have demonstrated that JMJD3 is required for several cellular processes, including inflammation (32, 35–37), reprogramming (38), neurogenesis (39–41), senescence (42–44), and organogenesis (45–48). Nevertheless, the role and mechanism of JMJD3 in the neuroendocrine control of reproduction and energy homeostasis remain unclear.

In the present study, we generated a genetically modified mouse model with JMJD3 deleted in hypothalamic RIP-Cre neurons and found that JMJD3 ablation led to delayed puberty development, reproductive dysfunction, and female-specific late-onset obesity. We identified the Kiss1 gene as a direct target of JMJD3 by mediating the methylation status of H3K27 in the Kiss1 promoter. In addition, we found that Jmjd3 was an estrogen target gene in the hypothalamus.

Materials and Methods

Animals

All the mice were housed in an Association for Assessment and Accreditation of Laboratory Animal Care International–accredited specific-pathogen-free animal facility with a 12-hour light/dark cycle and had ad libitum access to water and a standard mouse chow diet. All experiments were approved by the Institutional Animal Care and Use Committee of the Model Animal Research Center of Nanjing University. The Jmjd3^loxp/loxp mice with two loxP sites flanking exons 14 to 20 of the Jmjd3 gene and the Rip-Cre (B6.Cg-Tg (Ins2-cre) 25Mgn/Nju), ROSA26-mT/mG (Gt(ROSA)26Sor^{tm4(ACTB-tdTomato,-EGFP)Luo}/JNju), and ROSA26-LacZ (Gt(ROSA)26Sor^tm1/Nju) reporter transgenic mice were obtained from the Model Animal Research Center, Nanjing University. Rip-Cre transgenic mice were crossed with Jmjd3^loxp/loxp mice to generate Rip-Cre; Jmjd3^loxp/loxp mice (Rip-Jmjd3cKO) and their littermates Rip-Cre; Jmjd3^+/+ (also known as control). All mice had the C57BL/6J genetic background. Genotyping of the mice was performed using polymerase chain reaction (PCR) analyses of genomic DNA isolated from mouse tails. Primer sequences are listed in Supplemental Table 1.

Physiological studies

The blood glucose levels with ad libitum feeding and 16 hours of overnight fasting conditions were measured from tail vein blood using a glucometer (Arkray, Tokyo, Japan). Glucose tolerance tests were performed after a 16-hour overnight fast. The blood glucose levels were immediately monitored before (0) and 15, 30, 60, 90, and 120 minutes after intraperitoneal injection of d-glucose (2 g/kg; Sigma-Aldrich, St. Louis, MO). For the serum insulin and leptin measurements, the blood samples were collected by retro-orbital puncture, and the serum was isolated after cold centrifugation. For serum estradiol measurements, the blood samples were collected by retro-orbital puncture at 10:00 to 11:00 am after collection and analysis of vaginal cytology. The serum insulin, leptin, and estradiol levels were measured using mouse insulin (EMD Millipore, Billerica, MA), leptin (EMD Millipore), and estradiol (Fengxiang Bio, Shanghai, China) enzyme-linked immunosorbent assay kits according to the manufacturer’s instructions.

Histologic examination

The pancreas and ovaries collected at diestrus were removed and fixed in 4% paraformaldehyde (PFA) overnight at 4°C. They were dehydrated in a graded ethanol series and embedded in paraffin wax. Serial sectioning of the pancreas and ovaries was performed at 10-µm thickness using a microtome (Leica Biosystems, Nussloch, Germany). The sections were deparaffinized in xylene, rehydrated through a graded ethanol series, stained with hematoxylin and eosin, and examined with a microscope (Nikon, Tokyo, Japan).

Body weight, fat pad, and reproductive organ measurements

The body weights of the male and female mice were monitored weekly from the time the mice were 4 weeks old. Female mice at the age of 4 months were euthanized by decapitation and the inguinal, reproductive adipose tissue, and scapular brown adipose tissue (BAT) were rapidly dissected and weighed. Female mice at diestrus were euthanized by decapitation and the ovaries and uterus were removed and weighed.

Food intake and energy expenditure measurements

Daily food intake and indirect calorimetry were measured using a comprehensive laboratory animal monitoring system (Columbus Instruments, Columbus, OH). The mice were individually housed in calorimeter chambers. After 24 hours to allow for acclimation, the food consumption, oxygen consumption, CO₂ exhalation, heat production, total activity, and ambulatory activity were continuously recorded for ≥3 days.

Ovariectomy and estrogen replacement therapy

Bilaterally ovariectomy or sham operations were performed at 6 to 7 weeks of age in female mice. The mice were anesthetized by intraperitoneal injection of 0.01 mL/g 2.5% tribromoethanol and secured on a surgical table with a heating pad to maintain the core body temperature. A 1- to 2-cm midline dorsal incision was made, and the subcutaneous connective tissue was separated from underlying muscle. A 1-cm incision was made in the muscle layer on both sides to provide entry into the peritoneal cavity. The ovaries were tied off with absorbable sterile suture and then excised. The muscle and skin incisions were closed with sterile suture, and the mice were kept on the heated pad until they had recovered from the anesthetic. The sham operations were performed using the same procedures, except for removal of the ovaries. For estrogen replacement therapy, the female mice received daily subcutaneous injections of 2 μg/mouse 17β-estradiol (Sigma-Aldrich) dissolved in olive oil (Sangon Biotech, Shanghai, China) beginning at 6 to 7 months of age; the body weight was monitored daily.

Puberty and estrous cycle analysis

All female mice were weaned at 3 weeks of age, and the vagina was inspected daily for the initial vaginal opening. After vaginal opening, vaginal lavages of the female mice were performed daily in the morning by flushing the vagina gently with 10 μL of 0.9% saline for ≥2 weeks. Vaginal smears were collected, mounted on a poly-l-lysine–coated slide (Citoglas, Haimen, China) and then stained with Wright-Giemsa dye (Jiancheng Biotech Co., Ltd., Nanjing, China) (18). The phase of the estrous cycle was assessed using the vaginal cytology: a predominance of nucleated epithelial cells indicated the proestrus stage, cornified squamous epithelial cells indicated the estrus stage, a predominance of leukocytes and a few nucleated epithelial cells or cornified squamous epithelial cells indicated the metestrus stage, and an abundance of leukocytes indicated the diestrus stage. To examine the effects of genotype on estrous cyclicity, vaginal lavages from 3-month-old mice were performed daily in the morning for ≥3 weeks to assess the estrous cycle.

Fertility assessment

Eight-week-old control and Rip-Jmjd3cKO female mice were paired with proven fertile C57BL/6J adult males for 4 months (18). The number of litters and pups per litter were recorded according to the number of pups observed on the morning after delivery. Three-month-old male control and Rip-Jmjd3cKO mice were paired with C57BL/6J adult females for 3 weeks. Vaginal plug formation was checked every morning, and the numbers of litters born were recorded.

Immunofluorescence

The mice were anesthetized and transcardially perfused with 0.9% saline solution, followed by ice cold 4% PFA. The brains were immediately dissected out and postfixed in 4% PFA overnight at 4°C and immersed in 30% sucrose for 48 hours. The samples were then embedded in OCT compound and stored at −80°C. Coronal brain sections 12 µm in thickness were cut using a microtome (Leica Biosystems). The slides were warmed to room temperature for 30 minutes and then rinsed in PBS. For JMJD3 and KISS1 immunofluorescence, the sections were blocked with blocking solution (2% bovine serum albumin and 2% Tween-20 in PBS) at room temperature for 20 minutes and then incubated with rabbit anti-JMJD3 [catalog no. ab85392; Research Resource Identifier (RRID), AB_1860714; Abcam, Cambridge, UK] for 1 hour at room temperature. The sections were rinsed in PBS and incubated with Cy3-conjugated sheep anti-rabbit IgG (catalog no. C2306; RRID, AB_258792; Sigma-Aldrich,) at room temperature for 30 minutes. The sections were washed and then blocked with normal mouse serum. The sections were incubated with rabbit anti-KISS1 (catalog no. AB9754; RRID, AB_2296529; EMD Millipore) for 1 hour at room temperature, rinsed in PBS, and then incubated with Cy5-conjugated donkey anti-rabbit IgG (code, 711-175-152; RRID, AB_2340607; Jackson ImmunoResearch Laboratories, West Grove, PA) and 4′,6-diamidino-2-phenylindole (Sigma-Aldrich), which were diluted in PBS for 1 hour at room temperature. The sections were washed with PBS, mounted on slides with glycerol, and photographed using confocal microscopy (Olympus Corp., Tokyo, Japan).

Quantitative real-time PCR

The mice were euthanized by decapitation, and their brains were immediately removed from the skull and placed on ice. The hypothalami were cut out along the following boundaries: anterior border of the optic chiasm, posterior border of the mammillary body, lateral 2 mm of either side of the third ventricle, and the thalamus dorsally. To separate the ARC from the AVPV, the hypothalami were cut in the coronal plane anterior to the infundibulum stalk. The anterior dissection contained the AVPV, and the posterior dissection contained the ARC. The tissues were snap-frozen in liquid nitrogen and were maintained at −80°C. After stimulation with 17β-estradiol with or without wortmannin, as previously described. MCF-7 cells were harvested and maintained at −80°C. Total RNA was extracted using RNAiso plus reagent (TaKaRa Japan, Tokyo, Japan). Complementary DNA was synthesized with PrimeScript RT Reagent Kit (TaKaRa Japan) according to the manufacturer’s instructions. SYBR Premix Ex Taq (TaKaRa Japan) was used for quantitative real-time PCR analysis. All assays were performed using an ABI 7300 sequence detection system (Applied Biosystems, Foster City, CA) using the following conditions: 95°C for 30 seconds, followed by 40 cycles of 5 seconds at 95°C and 31 seconds at 60°C. The relative messenger RNA (mRNA) levels were normalized to 36B4 mRNA levels. The primer sequences are shown in Supplemental Table 1. All reactions were performed in triplicate for each sample. The relative mRNA levels were calculated using the 2^−ΔΔCt method, where Ct is the cycle threshold: ΔΔ(Ct) = sample 1 Δ(Ct) − sample 2 Δ(Ct); Δ(Ct) = 36B4 (Ct) − testing gene (Ct).

Cell culture and hormone treatments

The 293T and MCF-7 cell lines were cultured in Dulbecco's modified Eagle medium (DMEM; Hyclone; GE Healthcare Life Sciences, Little Chalfont, UK) with 10% fetal bovine serum (FBS; Gibco; Thermo Fisher Scientific Life Sciences, Waltham, MA), penicillin 100 U/mL, and streptomycin 100 U/mL (HyClone; GE Healthcare Life Sciences). Cells were incubated at 37°C in a humidified atmosphere of 5% CO₂. For estrogen stimulation experiments, MCF-7 cells were grown in phenol red-free DMEM medium (Hyclone; GE Healthcare Life Sciences) supplemented with 5% dextran-coated charcoal-stripped FBS for 3 days. The cells were then serum starved in phenol red-free DMEM for 24 hours before treatment with 10⁻⁸ M 17β-estradiol (Sigma-Aldrich). PIK3 inhibitor wortmannin (1 μM; Cell Signaling Technology, Danvers, MA) or vehicle was administered 1 hour before treatment with 17β-estradiol.

Transfection and luciferase reporter assays

The 293T cell line was cultured in DMEM (Hyclone; GE Healthcare Life Sciences) with 10% FBS (Gibco; Thermo Fisher Scientific Life Sciences), penicillin 100 U/mL, and streptomycin 100 U/mL (HyClone; GE Healthcare Life Sciences). The cells were incubated at 37°C in a humidified atmosphere of 5% CO₂. The 293T cells were transfected in 24-well plates using Lipofectamine 2000 (Invitrogen; Thermo Fisher Scientific Life Sciences) according to the manufacturer’s instructions. The cells were cotransfected with pRL-TK plasmid to control the transfection efficiency. Approximately 24 hours after transfection, the cells were harvested, and luciferase activity was measured using the dual luciferase reporter assay (Promega, Madison, WI) using a luminometer (Turner BioSystems, Sunnyvale, CA). For estrogen treatment, MCF-7 cells were grown in phenol red-free DMEM supplemented with 5% dextran-coated charcoal-stripped FBS before transfection. The MCF-7 cells were then cotransfected with reporter constructs and pRL-TK plasmid, as previously described. Twelve hours after transfection, the cells were serum starved in phenol red-free DMEM for 24 hours and treated with 10⁻⁸ M 17β-estradiol for 3 hours. The JMJD3 expression plasmid pCS2-Jmjd3-F was a gift from Kai Ge (Addgene plasmid no. 17440; Addgene, Cambridge, MA). Reporter constructs containing Jmjd3 promoter regions, Jmjd3 3′ untranslated region (UTR) and different lengths of 5′ proximal promoter of Kiss1 were cloned from mouse genomic DNA using the primers shown in Supplemental Table 2.

Western blot

After stimulation with 17β-estradiol, with or without wortmannin, as previously described, the MCF-7 cells were harvested and lysed in radioimmunoprecipitation assay buffer [50 mM Tris-HCl, pH 8.0; 150 mM NaCl, 1% NP40, 0.1% sodium dodecyl sulfate (SDS)], followed by the addition of a protease inhibitor cocktail (Sigma-Aldrich). Next, 20 μg of whole-cell lysate was separated by electrophoresis in a 7.5% SDS-polyacrylamide gel and transferred to polyvinylidene difluoride membrane (EMD Millipore). After blocking with 5% bovine serum albumin, the membranes were incubated with specific primary antibodies against JMJD3 (1:1000; RRID, AB_1860714; Abcam), phospho-ser473 AKT (1:1000; RRID, AB_2315049; Cell Signaling Technology), and β-actin (1:5000; RRID, AB_476744; Sigma-Aldrich). Horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG (1:20,000; RRID, AB_258284; Sigma-Aldrich) and HRP-conjugated goat anti-mouse IgG (1:20,000; RRID, AB_228307; Pierce; Thermo Fisher Scientific Life Sciences) were used as the secondary antibody. The immunoreactive bands were visualized using the Immobilon Western HRP Substrate (EMD Millipore).

Chromatin immunoprecipitation assay

The ARC and AVPV hypothalamic nuclei were dissected as described previously. ARC and AVPV samples of hypothalami from 5 female mice were minced roughly into 1-mm pieces and then cross-linked at room temperature for 10 minutes by adding 1% formaldehyde to the tissues (49). Quenching the cross-linking was achieved by adding glycine to a final concentration of 125 mM. The samples were ground using a Dounce tissue grinder (Sigma-Aldrich), pelleted, and then washed with cold PBS containing a cocktail of protease inhibitors (Sigma-Aldrich). The extracts were resuspended in cold cell lysis buffer (10 mM NaCl, 0.2% NP40, and 10 mM Tris-HCl; pH 8.0) with protease inhibitor cocktail, which was then pelleted and lysed in cold nuclei lysis buffer (10 mM EDTA, 1% SDS, and 50 mM Tris-HCl; pH 8.0) with protease inhibitor cocktail. The lysate was sonicated 210 times for 1 second, with 2-second intervals, using an ultrasonic cell disruptor (Sonics & Materials, Inc., Newtown, CT). Cell debris was removed by centrifugation, and the sonicated cell supernatant was diluted 10-fold in chromatin immunoprecipitation (ChIP) dilution buffer (167 mM NaCl, 0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA, 16.7 mM Tris-HCl; pH 8.0) with the addition of the protease inhibitor cocktail. Protein G agarose was used to preclear the solution (GE Healthcare Life Sciences)/salmon sperm DNA (Sigma-Aldrich) to block nonspecific DNA binding for 1 hour at 4°C. At this point, the solution was separated into multiple aliquots. To each aliquot of solution, 1 to 2 μg of primary antibody was added and incubated overnight at 4°C. Rabbit IgG was used as a negative control. The following antibodies were used: antitrimethyl-H3-lysine 27 (H3K27me3; catalog no. 07-449; RRID, AB_310624; EMD Millipore) and anti-JMJD3 (catalog no. ab85392; RRID, AB_1860714; Abcam). The immune complexes obtained using protein G agarose/salmon sperm DNA beads were washed with low-salt immune complex wash buffer (150 mM NaCl, 0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl; pH 8.0) twice and then washed with a high-salt immune complex washing buffer (500 mM NaCl, 0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl; pH 8.0) two more times before being washed with LiCL immune complex wash buffer (0.25 M LiCl, 1% deoxycholate sodium, 1 mM EDTA, 10 mM Tris-HCl; pH 8.0; EMD Millipore). A final wash with TE (Tris plus EDTA) buffer (1 mM EDTA, 10 mM Tris-HCl; pH 8.0) was performed twice. The complexes were eluted with freshly prepared ChIP elution buffer (1% SDS, 0.1 M NaHCO₃) and then reversed by incubating at 65°C overnight. The samples were incubated with ribonuclease A and digested with 0.5 M EDTA, 1 M Tris-HCl (pH 6.5), and 10 mg/mL proteinase K at 45°C for 1 hour. The DNA was purified by phenol-chloroform extraction, and the PCR was performed under the following conditions: 95°C for 5 minutes; 35 cycles of 95°C for 30 seconds, 58°C for 30 seconds, and 72°C for 30 seconds; and final extension at 72°C for 5 minutes. The PCR products were analyzed by agarose gel electrophoresis. The primer sequence is shown in Supplemental Table 2. After DNA purification, the occupancy of H3K27me3 at the kiss1 gene was assessed by real-time PCR, as described previously. PCR signals were normalized to the input samples.

Statistical analysis

All data are presented as mean ± standard error of the mean, and statistical significance was analyzed using the two-tailed unpaired Student t test. P < 0.05 was considered to indicate a statistically significant difference. All statistical analyses were performed using Prism, version 5.0 (GraphPad, San Diego, CA).

Results

Lack of JMJD3 in RIP-Cre neurons causes female-specific metabolic disturbances

To elucidate the physiological roles of JMJD3 in the neuroendocrine control of metabolic and reproductive homeostasis, we generated hypothalamic RIP-Cre neuron Jmjd3-deficient mice by crossing loxp-flanked Jmjd3 allele (Jmjd3^loxp/loxp) mice with Rip-Cre transgenic mice (50). We confirmed that the RIP-Cre–mediated recombination was restricted to the hypothalamus and islets and was not in other peripheral tissues by crossing Rip-Cre transgenic mice with R26R-lacZ reporter mice [Gt(ROSA)26Sor^tm1/Nju] and PCR for Cre recombinase [Supplemental Fig. 1(A–D)]. To evaluate the efficiency of the RIP-Cre–mediated Jmjd3 deletion in RIP-Cre neurons, we generated Rip-Cre; Jmjd3^loxp/loxp; mT/mG mice by crossing Rip-Cre; Jmjd3^loxp/+ to the mT/mG reporter mice [Gt(ROSA)26Sor^{tm4(ACTB-tdTomato,-EGFP)Luo}/JNju]. Double-immunofluorescence showed the reduction of JMJD3 in the RIP-Cre neurons [Supplemental Fig. 1(F) and 1(G)]. The resulting Rip-Cre; Jmjd3^loxp/loxp (Rip-Jmjd3cKO) and littermate Rip-Cre; Jmjd3^+/+ (also known as control) mice were born at the expected Mendelian ratios and survived to adulthood. Although Rip-Cre transgenic lines exhibit Cre recombinase in pancreatic β-cells [Supplemental Fig. 1(D)], we found perfectly normal β-cell development and glucose homeostasis after the deletion of Jmjd3 in pancreatic β cells (Supplemental Fig. 2).

However, the female Rip-Jmjd3cKO mice displayed a substantial increase in body weight beginning at 16 to 17 weeks of age [Fig. 1(a) and 1(b)]. The body weight of female Rip-Jmjd3cKO mice was approximately 33% greater than that of the control littermates by 24 weeks of age. This obesity phenotype was not evident in the male Rip-Jmjd3cKO mice [Fig. 1(a) and 1(b)]. Dual energy X-ray absorptiometry analysis revealed that the increased body weight was due to an elevated whole body fat mass without any change in lean mass [Fig. 1(c)]. The weight of inguinal and reproductive fat pads directly dissected from female Rip-Jmjd3cKO mice was significantly elevated compared with that of the control mice [Fig. 1(d)]. Further histological analyses of white adipose tissue and BAT using hematoxylin and eosin staining revealed that adipocyte cell size of the inguinal fat pad were visually larger in the Rip-Jmjd3cKO mice [Fig. 1(e)]. Consistent with the increased fat content, the basal serum leptin concentration was greatly elevated in the female Rip-Jmjd3cKO mice [Fig. 1(f)]. Taken together, these findings show that JMJD3 ablation in the RIP-Cre neurons leads to adult-onset obesity in female mice.

Figure 1.

Female-specific obesity in Rip-Jmjd3cKO mice. (a) Body weight in male and female littermates fed a regular chow diet from 4 to 28 weeks of age (male, n = 8 to 9/genotype; female, n = 8/genotype). (b) Gross appearance of control (Ctrl) and Rip-Jmjd3cKO female mice at 7 months of age. (c) Body composition of control (n = 7) and Rip-Jmjd3cKO (n = 7) mice at 7 months of age. (d) Inguinal (ING) and gonadal (GON) fat mass in control (n = 6) and Rip-Jmjd3cKO (n = 5) mice at 4 months of age. (e) Representative photomicrographs of white adipose tissue (WAT) and BAT sections of control and Rip-Jmjd3cKO female mice stained with hematoxylin and eosin. Scale bars = 100 μm (WAT) and 50 μm (BAT). (f) Serum leptin levels in control (n = 9) and Rip-Jmjd3cKO (n = 6) mice at 7 months of age. Data are presented as mean ± standard error of the mean; *P < 0.05, **P < 0.01, unpaired t test compared with control mice.

Next, we monitored the daily food intake of the mutated mice. The food intake in the female mice before the onset of obesity was measured to avoid the secondary effect that might be observed in obese Rip-Jmjd3cKO mice. However, the average daily food intake was not different among the genotypes [Fig. 2(a)], suggesting that the decreased energy expenditure might be primary responsible for the obesity. This was confirmed by direct assessment of the energy expenditure. Rip-Jmjd3cKO female mice consumed significantly less oxygen and generated less CO₂ at night [Fig. 2(b–d)]. Consistently, the physical activity of the female Rip-Jmjd3cKO mice was dramatically decreased during the dark phase [Fig. 2(e) and 2(f)], although heat production was comparable in both groups [Fig. 2(g)]. In addition, the respiratory exchange ratio, an indicator for the preference of metabolizing fat or carbohydrates, was not different among the genotypes [Fig. 2(h)]. The male Rip-Jmjd3cKO mice showed no difference in either food intake or energy expenditure (Supplemental Fig. 3). These results indicate that adult-onset obesity in female Rip-Jmjd3cKO mice results entirely from the reduction in locomotor activity and energy expenditure.

Figure 2.

Reduced energy expenditure in female Rip-Jmjd3cKO mice. (a) Food intake in control (Ctrl; n = 9) and Rip-Jmjd3cKO (n = 7) female mice at 15 weeks of age. (b–g) Oxygen consumption, oxygen, carbon dioxide production, total activity, heat production, and respiratory exchange ratio in control (n = 9) and Rip-Jmjd3cKO (n = 7) mice at 15 weeks of age. Data are presented as mean ± standard error of the mean; *P < 0.05, **P < 0.01, unpaired t test compared with control mice.

Lack of JMJD3 in RIP-Cre neurons disrupts female reproduction

To determine the effects of insufficient Jmjd3 expression in hypothalamic RIP-Cre neurons on female reproductive development and function, we assessed the pubertal maturation in Rip-Jmjd3cKO mice. Vaginal opening and the onset of first estrus reflect activation of the hypothalamic-pituitary-ovarian axis and were used to indicate pubertal onset in female mice. The age of vaginal opening was delayed by approximately 4 days in the Rip-Jmjd3cKO mice compared with that of the control mice. Also, Rip-Jmjd3cKO mice exhibited a substantial delay of first estrus compared with the controls, indicating significantly delayed pubertal maturation in the absence of JMJD3 [Fig. 3(a)]. Next, we analyzed the initiation and maintenance of female estrous cyclicity by examining the vaginal cytology daily of adult female mice for a period of 21 days. The results demonstrated that Rip-Jmjd3cKO mice exhibited lengthened and irregular estrous cycles compared with those of the control mice. The average estrous cycle length was 4 to 5 days in the control mice but was extended to 7 to 8 days in the Rip-Jmjd3cKO mice [Fig. 3(b)]. During the 21 days of estrous cycle monitoring, only 11% of Rip-Jmjd3cKO mice had three complete cycles. In contrast, 67% of the control littermates displayed three complete cycles [Fig. 3(c)]. We also observed that Rip-Jmjd3cKO mice spent more time in diestrus but a shorter time in the proestrus and metestrus stages than did the controls [Fig. 3(d)].

Figure 3.

Lack of JMJD3 in RIP-Cre neurons disrupts female reproduction. (a) Age at vaginal opening in control (Ctrl; n = 10) and Rip-Jmjd3cKO (n = 8) mice and day of first estrus in control (n = 8) and Rip-Jmjd3cKO (n = 7) mice. (b) Average estrous cycle length in control (n = 9) and Rip-Jmjd3cKO (n = 9) mice. (c) Representative 21-day estrous cycle profiles of control (n = 9) and Rip-Jmjd3cKO (n = 9) mice beginning at 12 weeks of age. The four phases of the estrous cycles are proestrus (P), estrus (E), metestrus (M), and diestrus (D). (d) Percentage of time spent in each estrous cycle phase in control (n = 9) and Rip-Jmjd3cKO (n = 9) mice. (e) Average number of litters in control (n = 10) and Rip-Jmjd3cKO (n = 9) female mice during the 4-month mating period. (f) Ovarian sections stained with hematoxylin and eosin at 6 months of age. Scale bars = 200 μm. (g) Representative view of uteri and ovary from Rip-Jmjd3cKO and control mice at 4 to 6 months of age. (h) Uterine and ovarian weight of control (n = 14) and Rip-Jmjd3cKO (n = 10) mice at 4 to 8 months of age. Data presented as mean ± standard error of the mean; *P < 0.05, **P < 0.01, unpaired t test compared with control mice. CL, corpora lutea.

Impaired reproductive function is expected from the delayed pubertal maturation and irregular estrous cycles of Rip-Jmjd3cKO mice. We then assessed the reproductive capability of mice by mating the Rip-Jmjd3cKO mice and controls to wild-type males. During 4 months of mating, the female Rip-Jmjd3cKO mice had significantly fewer litters than did the control mice, although the numbers of pups per litter were comparable [Fig. 3(e)]. Consistent with the irregular estrous cyclicity and impaired fertility of Rip-Jmjd3cKO mice, the ovarian histologic examination revealed that Rip-Jmjd3cKO mice had a markedly reduced number of corpora lutea at 4 months and an almost complete absence of corpora lutea at 6 months, although normal folliculogenesis was observed [Fig. 3(f) and Supplemental Fig. 4]. The uterus was thread-like in the Rip-Jmjd3cKO mice relative to that of the controls, with a noticeably narrower endometrial layer [Fig. 3(g)]. Also, the uterus and ovarian weights of the Rip-Jmjd3cKO mice were profoundly decreased compared with those of the control mice [Fig. 3(h)]. Taken together, these results indicate that Jmjd3 ablation in the RIP-Cre neurons leads to delayed puberty, irregular estrous cycles, and an impaired fertility in female mice and further supports the notion that JMJD3 is essential for neuroendocrine control of female reproduction.

Metabolic dysfunction is caused by estrogen deficiency in female Rip-Jmjd3cKO mice

The characteristic metabolic defects, with reduced locomotor activity and energy expenditure and unaltered energy intake in female Rip-Jmjd3cKO mice, closely resemble those of postmenopausal obesity, ovariectomy-induced obesity, or mice deficient in ERα (17–21, 51). Thus, we explored whether estrogen deficiency or dysregulation of estrogen signaling in the hypothalamus could contribute to the metabolic abnormality in female Rip-Jmjd3cKO mice. We found that the serum estradiol levels were significantly downregulated in the female Rip-Jmjd3cKO mice [Fig. 4(a)]. Next, female Rip-Jmjd3cKO and their control littermates underwent OVX or a sham operation at 6 to 7 weeks of age. The ovary-intact Rip-Jmjd3cKO mice gained more weight than the intact control mice. However, the weight gain after OVX was not substantially different between the Rip-Jmjd3cKO mice and controls [Fig. 4(b)]. However, the effects of OVX on the major parameters of energy homeostasis in the Rip-Jmjd3cKO mice were comparable to those in the controls (Supplemental Fig. 5). Because estrogen can prevent or reverse body weight gain in postmenopausal women and in OVX mice, we tested whether estrogen treatment could rescue the body energy imbalance in Rip-Jmjd3cKO mice. Thus, the female Rip-Jmjd3cKO and control mice at 6 to 7 months of age received exogenous estradiol by intraperitoneal injection. The obese phenotype of the Rip-Jmjd3cKO mice was significantly reversed by systemic 17β-estradiol replacement therapy [Fig. 4(c)]. The body weight decrease in the Rip-Jmjd3cKO mice was approximately 13% after treatment with estradiol compared with only a 2.5% change in control group [Fig. 4(d)]. Collectively, these results strongly suggest that adult-onset obesity in female Rip-Jmjd3cKO mice is due to a deficiency of estrogen resulting from neuroendocrine dysregulation.

Figure 4.

Estrogen deficiency-dependent metabolic dysfunction in female Rip-Jmjd3cKO mice. (a) Serum estradiol levels in control (Ctrl; n = 6) and Rip-Jmjd3cKO (n = 6) mice. (b) Body weight gain in control and Rip-Jmjd3cKO female mice after OVX or sham operation. (c) Body weight loss in control (n = 11) and Rip-Jmjd3cKO (n = 5) female mice after 18 days of 17β-estradiol treatment (subcutaneous injection of 2 μg/mouse/d). (d) Body weight loss in control (n = 11) and Rip-Jmjd3cKO (n = 5) female mice who received 18 days of 17β-estradiol treatment (subcutaneous infection 2 μg/mouse/d). Data are presented as mean ± standard error of the mean; *P < 0.05, **P < 0.01, unpaired t test compared with control mice.

JMJD3 regulates Kiss1 gene expression in the AVPV of female mice

Kisspeptin is a product of the Kiss1 gene that regulates reproduction by stimulating gonadotropin-releasing hormone neurons via the kisspeptin receptor (also known as GPR54). The abundance of H3K27me3 associated with the Kiss1 promoter decreases markedly during the pubertal process (25). We suspected that the direct target gene for JMJD3 in the hypothalamus would be Kiss1 because the reproductive and metabolic phenotypes of Rip-Jmjd3cKO mice resemble the defects in Kiss1 mutant mice (52–56). We found that the expression of the Kiss1 gene in the AVPV nucleus was markedly reduced in the Rip-Jmjd3cKO mice compared with that in the control littermates at 4 weeks of age [Fig. 5(a) and Supplemental Fig. 6]. To determine whether JMJD3 is required for removing the H3K27me3 repressive modification at the Kiss1 gene promoter, we performed dual immunofluorescent staining and found that RIP-Cre neurons in the AVPV and ARC could express kisspeptin (Supplemental Fig. 7). Next, dual immunofluorescent staining revealed a nuclear localization of JMJD3 in kisspeptin neurons of female mice [Fig. 5(b)]. Next, we performed a luciferase reporter assay in vitro. Five luciferase reporter constructs containing different lengths of the Kiss1 promoter region were constructed and cotransfected into HEK293 cells in the presence or absence of a JMJD3 expression vector. Overexpression of JMJD3 significantly increased the activity of all five Kiss1 promoter fragments. A nearly identical fold increase in reporter activity was observed between the shorter 197-bp Kiss1 promoter and other longer Kiss1 promoter segments, indicating that the region modulating the activation of Kiss1 was located within the proximal promoter [Fig. 5(c)]. To determine whether JMJD3 can be recruited to the promoter region of Kiss1 in vivo, we performed ChIP assays using hypothalamic tissue from female mice and observed that endogenous JMJD3 bound to the Kiss1 proximal promoter [Fig. 5(d)]. Realizing that female Rip-Jmjd3cKO mice displayed AVPV region-specific alterations in Kiss1 expression, we next performed a ChIP assay with an anti-H3K27me3 antibody to determine the hypothalamic region-specific alteration of H3K27me3 abundance at the Kiss1 gene promoter. Consistent with the AVPV-specific downregulation of Kiss1 in Rip-Jmjd3cKO mice, the genetically induced deficiency of Jmjd3 increased the H3K27me3 abundance at the Kiss1 promoter region only in the AVPV nucleus [Fig. 5(e)]. Together, these results demonstrated that JMJD3 can directly bind to the Kiss1 core promoter to activate Kiss1 gene expression in the AVPV region of the hypothalamus via demethylation of the repressive epigenetic modification H3K27me3 and suggest that deletion of Jmjd3 impairs kisspeptin signaling in the hypothalamus, which might contribute to reproductive system disorders and obesity in female Rip-Jmjd3cKO mice.

Figure 5.

JMJD3 regulates Kiss1 gene expression in the AVPV of female mice. (a) Relative mRNA levels of Kiss1 in AVPV and ARC hypothalamic nucleus in control (Ctrl; n = 7) and Rip-Jmjd3cKO (n = 4) female mice at 4 weeks of age. (b) Immunostaining for JMJD3 (red), kisspeptin (green), and 4′,6-diamidino-2-phenylindole (blue) in female ARC and AVPV hypothalamic nuclei. Scale bars = 20 μm. (c) Kiss1 promoter activity was determined by luciferase assays performed in the 293T cells. (d) ChIP assays with antibody against JMJD3 were performed using AVPV and ARC from female mice at proestrus. (e) H3K27 methylation status of the Kiss1 locus in the AVPV and ARC nuclei between control and Rip-Jmjd3cKO female mice at 4 weeks of age (5 mice per pool; n = 3 pools per genotype). Data presented as mean ± standard error of the mean; *P < 0.05, **P < 0.01, unpaired t test compared with control mice.

Estrogen regulates Jmjd3 gene expression in the hypothalamus

In female mice, the Kiss1 gene is positively regulated by estrogen in the AVPV, which is required for normal reproductive function (11, 12). Our findings showed that the expression of Jmjd3 in the hypothalamus increased gradually during puberty and fluctuated during the course of the estrous cycle and correlated positively with ERα signaling in the hypothalamus during these periods [Fig. 6(a) and 6(b) and Supplemental Fig. 8(A)]. Thus, we hypothesized that Jmjd3 is also a crucial estrogen-response gene in estrogen feedback action. To test this hypothesis, we performed ovariectomy of female mice during puberty, and the results showed that the increase in Jmjd3 in the hypothalamus during puberty was inhibited in the ovariectomized mice [Supplemental Fig. 8(B)]. Next, we investigated the temporal expression patterns of JMJD3 in the hypothalamus of OVX mice treated with estrogen. As expected, Jmjd3 expression in the hypothalamus was rapidly and transiently induced by estrogen treatment, peaking at 2 hours after treatment [Fig. 6(c)]. However, Kiss1 expression was upregulated in the AVPV and downregulated in the ARC by short-term estradiol treatment [Fig. 6(d)]. Furthermore, ChIP assays showed that estradiol treatment significantly decreased H3K27 methylation within in the Kiss1 promoter in the AVPV but not in the ARC [Fig. 6(e)]. A similar result was observed in an ERα-positive MCF-7 cell line, where we observed that estrogen treatment could also rapidly and transiently induce the expression of the JMJD3 gene, peaking at 30 minutes after treatment [Fig. 6(f)]. Estrogen can modulate gene expression through genomic and nongenomic pathways. To gain mechanistic insight into estrogen-induced JMJD3 expression, we investigated the 5′-flanking region of the Jmjd3 gene and found the promoter region to be devoid of detectable estrogen-responsive sites. It is well known that ERK1/2 and AKT can be rapidly activated by estrogen through nongenomic pathways, which is involved in estrogen-mediated gene expression. We observed that estrogen stimulation of JMJD3 was PI3K/AKT dependent, as pretreatment with the PI3K/AKT inhibitor wortmannin attenuated estrogen-induced JMJD3 expression [Fig. 6(g)]. To further elucidate the molecular mechanism of estrogen-induced JMJD3 expression, we performed promoter and 3′UTR luciferase reporter assays in MCF-7 cells. Treatment of the cells with estrogen significantly increased Jmjd3 3′UTR reporter activity, but not Jmjd3 promoter activity, suggesting that estrogen induced Jmjd3 expression via the 3′UTR [Fig. 6(h) and 6(i)]. These results indicate that the Jmjd3 is an estrogen target gene in the hypothalamus, and suggest that hypothalamic JMJD3 contributes to the regulation of estrogen feedback action in female mice.

Figure 6.

Estrogen regulates Jmjd3 gene expression via PI3K signaling. (a) Relative mRNA levels of Jmjd3 in the whole hypothalamus of female mice at different ages (n = 4/group). (b) Relative mRNA levels of Jmjd3 in the hypothalamus of female mice at the four phases [proestrus (P), estrus (E), metestrus (M), and diestrus (D)] of the estrous cycle (n = 3 to 4/group). (c) Relative mRNA levels of Jmjd3 in hypothalamus of OVX female mice treated with 17β-estradiol (1 μg/10 g intraperitoneally) for the indicated treatment duration or vehicle (Veh). (d) Relative mRNA levels of Kiss1 in AVPV and ARC hypothalamic nucleus of OVX female mice after treatment with 4 hours of vehicle (n = 3) or 17β-estradiol (1 μg/10 g intraperitoneally; n = 3). (e) H3K27 methylation status of the Kiss1 locus in the AVPV and ARC nuclei of OVX female mice after treatment with 4 hours of vehicle or 17β-estradiol (1 μg/10 g intraperitoneally; 4 mice per pool; n = 3 pools per treatment). (f) Relative mRNA levels of JMJD3 in serum-starved MCF-7 cells pretreated for 1 hour with 1 μM wortmannin or vehicle, as indicated, and then with 10⁻⁸ M 17β-estradiol for the indicated treatment duration. (g) Western blot analysis of JMJD3. AKT473 protein levels in serum-starved MCF-7 cells were pretreated for 1 hour with 1 μM wortmannin or vehicle, as indicated, and then with 10⁻⁸ M 17β-estradiol for 30 minutes. β-Actin was used as a control. (h) Schematic representation of the Jmjd3 locus. Promoter regions are presented as gray boxes, and arrows indicate the position of the transcription start site (TSS) used in embryonic stem cells (ESC) and macrophages (MF). (i) Jmjd3 promoter and 3′-UTR activities were determined by luciferase assays performed in MCF-7 cells. (j) A proposed model illustrating the role of hypothalamic JMJD3 in the neuroendocrine control of female reproduction and energy balance. Data presented as mean ± standard error of the mean; *P < 0.05, **P < 0.01, unpaired t test compared with control mice.

Discussion

Although many reports have suggested that epigenetic modifications are associated with hypothalamic neuroendocrine control of the hypothalamic-pituitary-gonadal axis, actual data are lacking. Using a genetically modified mouse model with deletion of Jmjd3 in hypothalamic RIP-Cre neurons, we have confirmed that JMJD3 serves as an important regulator of female reproduction and energy homeostasis by regulating kisspeptin signaling in the hypothalamus.

Kisspeptin has been implicated as a critical regulator of reproductive physiology. Impaired kisspeptin signaling in mice and humans can lead to severe pubertal and reproductive defects and metabolic dysfunction (52–56). The polycomb group protein EED has recently been shown to regulate the timing of female puberty and reproductive function by altering Kiss1 gene expression (25). In addition, a histone H3 acetylation modification at the Kiss1 gene promoter is involved in the estrogen-positive feedback action (27). In the present study, we have demonstrated that the H3K27 demethylase JMJD3 controls female reproduction by regulating hypothalamic Kiss1 gene expression. Consistent with the reproductive phenotypes observed in Kiss1/Kiss1r-deficient mice, Jmjd3 deletion in the hypothalamic neurons resulted in delayed puberty onset, an irregular estrous cycle, impaired fertility, and accelerated degradation of the ovaries in female mice (Fig. 3). Because such phenotypes were observed in young mutant mice before their divergence in body weight, these phenotypes are likely a primary defect caused by a lack of JMJD3 in the RIP-Cre neurons and not secondary to obesity. The precise mechanisms involved in the infertility of male Rip-Jmjd3cKO mice require clarification in the future. It remains to be determined how JMJD3 synchronizes with other genes in this complicated regulatory system, because it is reasonable to assume that JMJD3 regulates multiple genes. It will be also interesting to determine which compensatory mechanisms initiate puberty in the JMJD3 mutant mice. The reproductive development and function of male Rip-Jmjd3cKO mice were also assessed in our study. Although testicular development and spermatogenesis were normal, male Rip-Jmjd3cKO mice were infertile (Supplemental Fig. 9). The possible mechanisms involved in the infertility of male Rip-Jmjd3cKO mice could the dysregulation of male sex behavior (57, 58) and require clarification in the future.

ERα signaling mediates estrogen-positive feedback action by targeting the Kiss1 gene in the AVPV nuclei (11, 59). In addition, previous studies have revealed that the H3K27 demethylase activity of JMJD3 is required for ERα-dependent gene activation after estrogen stimulation in ERα-positive cell lines (44). Our study has shown that Kiss1 is a direct target gene of JMJD3 in the AVPV, suggesting that JMJD3 participates in the estrogen-positive feedback loop. A previous study emphasized that the corecruitment of JMJD3 and ERα to the ERα-binding region is necessary for H3K27me3 demethylation and subsequent gene activation (44). It has also been reported that the classic ERα signaling pathway in AVPV kisspeptin neurons is required for an estrogen-mediated positive feedback effect (12, 13). However, in the present study, we have demonstrated that estrogen rapidly induces expression of JMJD3 via nonclassic ERα/PI3K/AKT signaling, indicating that the transient stimulation of JMJD3 gene expression mediated by nonclassic ERα signaling, involving the estrogen response element-independent mechanism, is also indispensable for estrogen-positive feedback action. Further work is required to understand how such JMJD3–KISS1–estrogen–JMJD3 feedback is established during the onset of puberty.

Hypothalamic RIP-Cre neurons play important roles in the regulation of energy homeostasis (3–8). A recent report has demonstrated that hypothalamic RIP-Cre neurons can stimulate BAT thermogenesis and energy expenditure by synaptic release of the γ-aminobutyric acid neurotransmitter (5). Previous work has suggested that different nutritional status is associated with marked epigenetic changes in the hypothalamic energy-regulating pathways (28–30). Fasting had a stimulatory effect on dimethylation and trimethylation of H3K27 in the hypothalamic neurons, suggesting that H3K27me3 demethylase might participate in neuroendocrine regulation of metabolic homeostasis (30). In the present study, we found that selective deletion of Jmjd3 from hypothalamic RIP-Cre neurons leads to late-onset obesity in a sex-dependent manner, which is associated with reduced physical activity but not with hyperphagia or decreased BAT thermogenesis (Fig. 2). The results of our studies suggest that although female mice lacking JMJD3 in RIP-Cre neurons are obese, male mutant mice have normal metabolic homeostasis. This implies that factors other than abnormal survival and differentiation of hypothalamic RIP-Cre neurons account for the obesity in the female mutant mice. Several studies have demonstrated that estrogen plays a vital role in the regulation of systemic energy homeostasis. Estrogen deficiency in aromatase deficiency or OVX mice is associated with the onset of obesity and related metabolic pathologic features, which primarily result from reduced locomotor activity and energy expenditure (20, 21). Consistent with these observations, we performed ovariectomy followed by 17β-estradiol replacement therapy and found that estrogen deficiency contributed to the obesity phenotype (Fig. 4). Although our present findings suggest that JMJD3 in hypothalamic RIP-Cre neurons is not directly involved in the regulation of energy balance, we cannot exclude the possibility that epigenetic modifications mediated by JMJD3 in other populations of neurons, such as NPY and POMC neurons, are involved in the regulation of metabolic homeostasis. Further studies are needed to elucidate the specific neurons that could mediate these effects.

In conclusion, the present study has provided evidence that the histone demethylase JMJD3 plays a pivotal role in regulating female reproduction and metabolism. Our results suggest that JMJD3 is involved in reproduction and energy homeostasis by directly targeting Kiss1 in the AVPV. These findings serve to expand our understanding of the mechanisms safeguarding reproduction and energy homeostasis and might provide a potential therapeutic target for treating reproductive and metabolic diseases.

Appendix.

Antibody Table

Peptide/Protein Target	Antigen Sequence (if Known)	Name of Antibody	Manufacturer, Catalog No., and/or Name of Individual Providing the Antibody	RRID	Species Raised in; Monoclonal or Polyclonal	Dilution Used
JMJD3	Synthetic peptide, corresponding to a region within amino acids 600–650 of mouse JMJD3	Anti-JMJD3	Abcam Ltd., ab85392	AB_1860714	Rabbit; polyclonal	1:1000 (Western), 1:200 (IF), 2 µg (ChIP)
KISS1	Peptide from mouse kisspeptin 10	Anti-kisspeptin	EMD Millipore, AB9754	AB_2296529	Rabbit; polyclonal	1:1000 (IF)
Phospho-Akt (Ser473)	Synthetic phosphopeptide corresponding to residues around Ser473 of human Akt	Phospho-Akt (Ser473)	Cell Signaling Technology, 4060S	AB_2315049	Rabbit; monoclonal	1:1000 (Western)
β-Actin	Slightly modified β-cytoplasmic actin N-terminal peptide, Ac-Asp-Asp-Asp-Ile-Ala-Ala-Leu-Val-Ile-Asp-Asn-Gly-Ser-Gly-Lys, conjugated to KLH	β-Actin	Sigma-Aldrich, A5441	AB_476744	Mouse; monoclonal	1:5000 (Western)
HK27me3	KLH-conjugated, synthetic 2X-branched peptide containing the sequence …AR(me3K)SAP… in which me3K corresponds to trimethyl-lysine at residue 27 of human histone H3.	Anti-H3K27me3	Millipore, 07-449	AB_310624	Rabbit; polyclonal	2 µg (ChIP)
GFP	Amino acids 1–238 representing full length GFP of Aequorea victoria origin	Anti-GFP	Santa Cruz Biotechnology, Inc., SC-9996	AB_627695	Mouse; monoclonal	1:200 (IF)
Rabbit IgG (H+L)		Cy5-AffiniPure anti-rabbit IgG (H+L)	Jackson ImmunoResearch Laboratories, 711-175-152	AB_2340607	Donkey; polyclonal	1:200 (IF)
Rabbit IgG, F(ab')2 fragment, whole molecule rabbit		Anti-rabbit IgG (whole molecule), F(ab)2 fragment-Cy3	Sigma-Aldrich, C2306	AB_258792	Sheep; polyclonal	1:200 (IF)
Rabbit IgG, whole molecule		Anti-rabbit IgG (whole molecule) peroxidase	Sigma-Aldrich, A6154	AB_258284	Goat; polyclonal	1:20,000 (Western)
Mouse IgG (H+L) mouse		Anti-mouse IgG (H+L) secondary antibody, HRP	Thermo Scientific Pierce, 31430	AB_228307	Goat; polyclonal	1:20,000 (Western)

Abbreviations: GFP, green fluorescent protein; H, heavy chain; IF, immunofluorescence; KLH, keyhole limpet hemocyanin; L, light chain.

Abbreviations:

 
• ARC

  arcuate nucleus

 
• AVPV

  anteroventral periventricular

 
• BAT

  brown adipose tissue

 
• ChIP

  chromatin immunoprecipitation

 
• DMEM

  Dulbecco's modified Eagle medium

 
• ERα

  estrogen receptor-α

 
• FBS

  fetal bovine serum

 
• H3K27

  H3 lysine 27

 
• HRP

  horseradish peroxidase

 
• JMJD3

  Jumonji domain-containing 3

 
• mRNA

  messenger RNA

 
• OVX

  ovariectomized

 
• PCR

  polymerase chain reaction

 
• PFA

  paraformaldehyde

 
• RIP-Cre neuron

  rat-insulin-promoter–expressing neuron

 
• RRID

  Research Resource Identifier

 
• SDS

  sodium dodecyl sulfate

 
• UTR

  untranslated region.

Acknowledgments

We thank Drs. Chaojun Li, Jianghuai Liu, Ying Xu, and Dong Kong for their helpful suggestions and Dr. Richard H. Finnell for help in editing our report.

This work was supported by Ministry of Science and Technology of China Grants 2014BAI02B01 and 2015BAI08B02 and National Natural Science Foundation of China Grant 31301217.

Disclosure Summary: The authors have nothing to disclose.

Author notes

Address all correspondence and requests for reprints to: Xiang Gao, PhD, Model Animal Research Center, Nanjing University, Nanjing 210061, China. E-mail: [email protected].