Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Morphological Analysis of the Axonal Projections of EGFP-Labeled Esr1-Expressing Neurons in Transgenic Female Medaka

Morphological Analysis of the Axonal Projections of EGFP-Labeled Esr1-Expressing Neurons in... Abstract Some hypothalamic neurons expressing estrogen receptor α (Esr1) are thought to transmit a gonadal estrogen feedback signal to gonadotropin-releasing hormone 1 (GnRH1) neurons, which is the final common pathway for feedback regulation of reproductive functions. Moreover, estrogen-sensitive neurons are suggested to control sexual behaviors in coordination with reproduction. In mammals, hypothalamic estrogen-sensitive neurons release the peptide kisspeptin and regulate GnRH1 neurons. However, a growing body of evidence in nonmammalian species casts doubt on the regulation of GnRH1 neurons by kisspeptin neurons. As a step toward understanding how estrogen regulates neuronal circuits for reproduction and sex behavior in vertebrates in general, we generated a transgenic (Tg) medaka that expresses enhanced green fluorescent protein (EGFP) specifically in esr1-expressing neurons (esr1 neurons) and analyzed their axonal projections. We found that esr1 neurons in the preoptic area (POA) project to the gnrh1 neurons. We also demonstrated by transcriptome and histological analyses that these esr1 neurons are glutamatergic or γ-aminobutyric acidergic (GABAergic) but not kisspeptinergic. We therefore suggest that glutamatergic and GABAergic esr1 neurons in the POA regulate gnrh1 neurons. This hypothesis is consistent with previous studies in mice that found that glutamatergic and GABAergic transmission is critical for estrogen-dependent changes in GnRH1 neuron firing. Thus, we propose that this neuronal circuit may provide an evolutionarily conserved mechanism for regulation of reproduction. In addition, we showed that telencephalic esr1 neurons project to medulla, which may control sexual behavior. Moreover, we found that some POA-esr1 neurons coexpress progesterone receptors. These neurons may form the neuronal circuits that regulate reproduction and sex behavior in response to the serum estrogen/progesterone. Reproduction in vertebrates is regulated by the hypothalamo-pituitary-gonadal (HPG) axis in a well-coordinated manner. Here, gonadotropin-releasing hormone (GnRH), which is now often called gonadotropin-releasing hormone 1 (GnRH1) for evolutionary reasons (1), acts on the pituitary to facilitate the release of gonadotropins, luteinizing hormone (LH), and follicle-stimulating hormone (FSH). LH and FSH, in turn, stimulate gonads and induce their maturation (2, 3). The mature gonads release sex steroids into the circulation. Then, the sex steroids, especially estrogen, stimulate or inhibit the hypothalamus to upregulate or downregulate release of GnRH1. This entire process is called steroid feedback. Recently, the importance of each component of HPG axis regulation in teleosts (GnRH1, LH, and FSH) has been scrutinized by the use of gene knockout techniques in zebrafish and medaka (4, 5), suggesting the importance and the evolutionary conservation of basic HPG axis regulation mechanisms in vertebrates, although there are some important differences between mammals and teleosts. In HPG axis regulation, gonadal estrogen may be considered one of the most important signals that transmit the reproductive status of the gonad to the central nervous system, and its mechanisms of action have been analyzed intensively. It has been generally suggested that there are two types of estrogen receptors (Esrs): Esr1 (or ERα, coded by esr1) and Esr2 (or ERβ, coded by esr2). Furthermore, data from knockout mice suggest that Esr1 but not Esr2 is essential for reproduction (6, 7). A growing body of evidence in mammals suggests that Esr1 is expressed in kisspeptin neurons rather than in GnRH1 neurons (8–10) and that kisspeptin neurons regulate the firing activity of GnRH1 neurons and GnRH1 release in the median eminence in response to serum estrogen levels. Although the expression of esr1 in kisspeptin neurons has been reported to be widely conserved in vertebrates including teleosts (11, 12), accumulating evidence argues against the presence of robust regulation of GnRH1 neurons by kisspeptin neurons and hence their role in regulation of reproduction in nonmammalian vertebrates. For example, GnRH1 neurons in teleosts do not express kisspeptin receptors (neither GPR54 paralogs gpr54-1 and gpr54-2) (13–15), and teleosts with kiss1 and kiss2 double knockout show normal fertility (16, 17). Furthermore, it is already known that avian species completely lack kisspeptin genes (18). Therefore, presumably unidentified estrogen-sensitive neurons other than kisspeptin neurons play important roles in the HPG axis regulation in nonmammalian vertebrates. On the other hand, it is also well known that sex steroids modulate specific behaviors depending on reproductive conditions. For example, most vertebrates show species-specific courtship behavior only after sexual maturation and during the breeding season. Thus, hormonal signals from the gonads have been suggested to control neural circuits for the modulation of sexual behaviors (19, 20). However, the nature of such estrogen-sensitive neurons remains to be elucidated. With these unanswered questions in mind, we aimed to anatomically identify the location and axonal projections of Esr1-expressing neurons mediating estrogen feedback signals to GnRH1 neurons, as well as those mediating sex steroid–sensitive signals to circuits that modulate sex behaviors. Here, we generated transgenic (Tg) animals in which esr1-expressing neurons are labeled with enhanced green fluorescent protein (EGFP) in the live brain for later detailed anatomical and physiological analyses. We chose medaka because of the following experimental advantages. First, molecular genetic tools can be applied easily. Second, female medaka spawn regularly every day under long-day conditions, and their breeding and nonbreeding states can be experimentally controlled by changing the day length. In addition, we have already described the anatomical distribution of esr messenger RNA (mRNA) expression in the medaka brain by using in situ hybridization (21). In the current study, we further detailed the axonal projections of esr1-expressing neurons (esr1 neurons) by using immunohistochemical methods to detect EGFP. Because Esrs are nuclear receptors, it is not possible to visualize projections of Esr1 neurons by using antibody directed against the Esr peptide. Therefore, we generated a Tg medaka line that expresses EGFP specifically in esr1 neurons. In these animals, EGFP that is expressed under the regulation of the esr1 promoter passively diffuses in the entire cytoplasm, enabling visualization of the entire neuronal morphology of esr1 neurons, including axons and dendrites. Materials and Methods Animals All the experiments including the generation of Tg line were conducted with d-rR strain medaka (Oryzias latipes), a teleost fish. For each experiment, sexually mature, gonadally intact female wild-type or Tg animals were used. Medaka were maintained under a 14-hour light/10-hour dark photoperiod at 27°C. The fish were fed twice daily with live brine shrimp and flake food. All experiments were conducted in accordance with the protocols approved by the Animal Care and Use Committee of the University of Tokyo (permission number 15–3). Generation of Tg medaka We generated constructs for esr1:enhanced green fluorescent protein (EGFP) Tg medaka by using a double-promoter approach for efficient screening. Using the zebrafish cardiac myosin light chain 2 (cmlc2) promoter, we visualized the heart for screening transgene-positive embryos. We carried out polymerase chain reaction (PCR) by using PrimeSTAR polymerase (Takara, Shiga, Japan) to amplify a medaka genomic DNA fragment containing the 5′-flanking region (3.7 kb) of exon 2 of the esr1 gene from a bacterial artificial chromosome clone (clone number Golwb109_F22) from a medaka strain, HdrR, by using a forward/reverse primer pair, Pesr1-F (5′-ACAGGATGGAGGTCAAAAGC-3′)/Pesr1-R (GACCCCCTCGGTGACATGTATCCACCGGTCGCCACCATGG-3′); Pesr1-F (5′-CAGAACTTCCTTGCTCATGCTCACC-3′)/Pesr1-R LINKEGFPhind (5′-GAGAAGCTTCAGAGCCCTTCCCCTGTGCTCAGGC-3′). For gene IDs, see Table 1. Then, we fused this amplicon with an EGFP open reading frame by overlap extension PCR. Downstream of the coding sequence of EGFP, we fused the cmlc2 promoter region and I-SceI restriction enzyme site. A fragment of ~6 kb was then cloned into the TOPO-XL cloning vector (Invitrogen, Carlsbad, CA). For microinjection and screening, we followed a protocol that has been described previously (22). After generation of the Tg line, we confirmed the specificity of EGFP expression in esr1 neurons in brain sections by double labeling with esr1 in situ hybridization and EGFP immunohistochemistry. The brains of fish from the esr1:EGFP Tg line were fixed with 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS). The fixed brains were frontally cryosectioned at 20 μm with a cryostat (CM 3050S; Leica Microsystems, Wetzlar, Germany) and mounted onto MAS-GP type A coated glass slides (Matsunami, Osaka, Japan). Sections were incubated with anti-EGFP antibody raised in rabbit (generous gift from Drs. Kaneko and Hioki, Kyoto University, Kyoto, Japan; see also Table 2) diluted 1:1000 with phosphate-buffered saline containing 0.3% Tween 20 (PBST) overnight, rinsed twice with PBST, and incubated with biotinylated anti-rabbit immunoglobulin G (Invitrogen) (diluted 1:200 with PBST) for 2 hours. Then, sections were fixed by 4% PFA in PBS for 15 minutes and rinsed with PBS containing 0.2% glycine. Next, we performed in situ hybridization with these slides to detect esr1, estrogen receptor β1 (esr2a), or estrogen receptor β2 (esr2b) mRNA by using esr1-,esr2a-, or esr2b-specific digoxigenin (DIG)-labeled probes that were described in a previous report (11, 21), following a protocol described previously (21). Briefly, slides were incubated with 0.25% acetic anhydride in 0.1 M triethanolamine for 10 minutes. Sections were then washed with PBS and prehybridized at 58°C for ≥30 minutes in a hybridization buffer containing 50% formamide, 3× saline sodium citrate (SSC), 0.12 M phosphate buffer (pH 7.4), 1× Denhardt solution (Sigma, St. Louis, MO), 125 μg/mL transfer RNA, 0.1 mg/mL calf thymus DNA (Invitrogen), and 10% dextran sulfate (Sigma). Slides were incubated at 58°C overnight in the same solution containing 100 ng/mL denatured riboprobe. After hybridization, the sections were washed twice with 50% formamide and 2× SSC for 15 minutes each at 58°C. Then, the sections were immersed in 10 mM Tris-HCl, pH 7.5; 500 mM NaCl; and 1 mM EDTA, pH 8.0 (TNE) for 10 minutes at 37°C. The sections were incubated with 20 μg/mL ribonuclease A (Sigma) in TNE for 30 minutes at 37°C and then washed with TNE for 10 minutes at 37°C to remove the ribonuclease A. They were then washed with 2× SSC twice, followed by 0.5× SSC twice for 15 minutes each at 58°C. The slides were immersed in DIG-1 (0.1 M Tris-HCl, pH 7.5; 0.16 M NaCl; and 0.1% Tween 20) for 5 minutes, followed by 1.5% blocking reagent with DIG-1 for 30 minutes and DIG-1 for 15 minutes, and then incubated with avidin-biotin complex (ABC) reagents (1% A solution and 1% B solution in DIG-1 buffer; Vector Laboratories, Burlingame, CA) for 1 hour. Sections were rinsed twice with DIG-1 buffer, incubated with Alexa Fluor 488 conjugated streptavidin (diluted 1:500 with DIG-1 buffer; Invitrogen) and alkaline phosphatase–conjugated anti-DIG antibody (diluted 1:1000 with DIG-1 buffer; Roche, Molecular Biochemicals GmbH, Mannheim, Germany) for 2 hours. Sections were then rinsed twice with DIG-1 buffer. After detection of EGFP signals, alkaline phosphatase activity, which was used to label mRNA, was detected with a Fast-Red substrate kit (Roche) according to the manufacturer’s instructions. Incubation for this substrate was carried out until visible signals were detected and was stopped by washing in PBS containing 0.5 mM EDTA. Then, the sections were coverslipped with CC/Mount (Diagnostic BioSystems, Pleasanton, CA). Fluorescence was observed with a confocal laser-scanning microscope (LSM-710; Carl Zeiss, Oberkochen, Germany). We calculated the percentage of EGFP-labeled esr1-expressing cells by counting the number of both EGFP-positive and EGFP-negative esr1-expressing cells. Numbers of cells were expressed as mean ± standard deviation (SD). Table 1. List of Gene IDs Gene  Ensembl Gene ID  vglut1.1  ENSORLG00000011296  vglut1.2  ENSORLG00000007982  vglut2.1  ENSORLG00000006157  vglut2.2  ENSORLG00000005791  vglut3  ENSORLG00000011203  gad1.1  ENSORLG00000009208  gad1.2  ENSORLG00000017268  gad2  ENSORLG00000012966  esr1  ENSORLG00000014514  esr2a  ENSORLG00000017721  esr2b  ENSORLG00000018012  pr  ENSORLG00000002651  Gene  Ensembl Gene ID  vglut1.1  ENSORLG00000011296  vglut1.2  ENSORLG00000007982  vglut2.1  ENSORLG00000006157  vglut2.2  ENSORLG00000005791  vglut3  ENSORLG00000011203  gad1.1  ENSORLG00000009208  gad1.2  ENSORLG00000017268  gad2  ENSORLG00000012966  esr1  ENSORLG00000014514  esr2a  ENSORLG00000017721  esr2b  ENSORLG00000018012  pr  ENSORLG00000002651  View Large Table 2. Antibody Table Peptide/Protein Target  Name of Antibody  Names of Individuals Providing the Antibody  Species Raised in; Monoclonal or Polyclonal  Dilution Used  RRID  EGFP  Anti-EGFP antibody  Drs. Kaneko and Hioki (Kyoto University)  Rabbit; polyclonal  1:1000  AB_2716624  Peptide/Protein Target  Name of Antibody  Names of Individuals Providing the Antibody  Species Raised in; Monoclonal or Polyclonal  Dilution Used  RRID  EGFP  Anti-EGFP antibody  Drs. Kaneko and Hioki (Kyoto University)  Rabbit; polyclonal  1:1000  AB_2716624  Abbreviation: RRID, Research Resource Identifier. View Large RNA sequencing Brains of sexually mature female esr1:EGFP Tg medaka were dissected out under an upright fluorescent microscope (Eclipse E600FN; Nikon, Tokyo, Japan) via a method described in our previous study (23). Brains were kept in a chamber filled with artificial cerebrospinal fluid, and a constant flow (1 mL/min) of artificial cerebrospinal fluid was applied 15 minutes before and during cell harvesting via a peristaltic pump to clean debris from dead or unhealthy cells. Patch pipettes of borosilicate glass capillaries with an outer diameter of 1.5 mm (GD-1.5; Narishige, Tokyo, Japan) were pulled to produce a 3- to 5-µm tip with a micropipette puller (P-97; Sutter Instruments, Novato, CA). The tip of the capillary was backfilled by capillarity with nuclease-free water (<1 μL). Slight positive pressure was applied to patch pipettes while they approached the targeted EGFP-positive neurons, and then EGFP-positive neurons were carefully pulled into the capillary by negative pressure. The contents of the pipette were expelled into a 0.2-μL tube containing 10 μL Buffer RLT (Qiagen, Hilden, Germany) with 1% 2-mercaptoethanol. Using this approach, we prepared five samples consisting of five neurons each from five different fish. Samples were frozen at −80°C immediately after harvesting. Total RNA was purified with Ampure XP RNA (Beckman Coulter, Brea, CA). For reverse transcription, whole-transcript amplification, and RNA sequencing (RNA-seq) library preparation, we followed the Quartz-seq protocol (24). We performed RNA-seq with HiSeq1500 and the HiSeq Rapid SBS Kit version 1 (Illumina, San Diego, CA), following manufacturer’s instructions. For data analysis, we used the medaka genome (ftp://ftp.ensembl.org/pub/release-79/gtf/oryzias_latipes). At least 13 million sequences of 50-base single reads were mapped onto the genome for each sample. Calculations of read mapping and reads per kilobase of transcript per million mapped reads (RPKM) were performed in CLCbio Genomics workbench software (CLC Bio, Aarhus, Denmark). Double labeling for EGFP and mRNA of gnrh1 or progesterone receptor We performed double labeling for EGFP and mRNA of gnrh1 or progesterone receptor (pr) as described earlier. To detect pr mRNA, we prepared a specific DIG-labeled riboprobe. Nested primers used to amplify medaka-pr-complementary DNAs were as follows: first round, pr-F1 5′-ATGGAGAGTAAAATGAACGGAAAGCTGG-3′ and pr-R1 5′-GTCATAGCCGGAGTACACGGTCT-3′; second round, pr-F2 5′-CCGAGTCCAGAGTAAATGGCTTGATCGA-3′ and pr-R2 5′-CGGCTCGATGTTCTCCAGAATGT-3′. The template used to make the pr probe was 1.1 kb in length (DNA Data Bank of Japan/European Molecular Biology Laboratory/GenBank accession no. AB360545.1). The probe for pr was synthesized from the medaka brain with a labeling kit (Roche). We performed in situ hybridization by using a gnrh1-specific DIG-labeled probe that was generated previously (23). We calculated the percentage of pr-expressing cells that were labeled with EGFP by counting the number of both EGFP-positive and EGFP-negative pr-expressing cells. We also calculated the percentage of gnrh1 neurons that received projections from EGFP fibers by counting the number of cell bodies of gnrh1 neurons that were surrounded by EGFP-labeled fibers within 1 μm. For both calculations, number of cells was expressed as mean ± SD. Immunohistochemistry for tract tracing We carried out immunohistochemistry for EGFP to analyze axonal projections of esr1 neurons in detail. All the procedures were performed as described earlier up to the secondary antibody, which was a biotinylated anti-rabbit immunoglobulin G. Sections were washed twice with PBST and then incubated with ABC reagents (1% A solution and 1% B solution in PBST buffer; Vector) for 1 hour. Then, the slides were washed twice with PBST and reacted with 3, 3′-diaminobenzidine. For some slides, we performed Nissl counterstaining by using a 0.1% cresyl violet solution. The slides were then dehydrated, cleared, and coverslipped. Photographs were taken with a digital camera (DFC310FX; Leica Microsystems) on a Leica DM5000B microscope (Leica Microsystems). Double-label in situ hybridization for esr1 mRNA and vglut or gad mRNA We performed a double-label in situ hybridization by using a mixture of DIG-labeled esr1 and fluorescein-labeled vesicular glutamate transporter (vglut) 2.1, glutamic acid decarboxylase (gad)1.1, gad1.2, and gad2 probes that were synthesized as described in a previous report (25). Medaka complementary DNAs were PCR amplified with forward/reverse primer pairs: vglut2.1-F 5′-GAGATCAACCTGCGCTCACCACA-3′/Vglut2.1-R 5′-TGAATACTGAACCAGGATCCCAG-3′ gad1.1-F 5′-GAGGCTGTGACTCATGCGTG-3′/gad1.1-R 5′-CCTTCTTTATGGAATAGTGGC-3′ gad1.2-F 5′-GCCAGATCCACGCTGGTGGAC-3′/gad1.2-R 5′-CATTAGCACAAAAACTGGAG-3′ gad2-F 5′-AAACAGCCCATCCCAGGTAC-3′/gad2-R 5′-AGCAGCGGGATTTGAGATGAC-3′ These amplified fragments were used to generate probes with fluorescein RNA labeling mixtures (Roche) according to the manufacturer’s protocol. After brains were fixed by transcardial perfusion, all procedures up to the blocking step were performed as described earlier. The sections were then incubated with a horseradish peroxidase conjugated antifluorescein antibody (diluted 1:500 with DIG-1; PerkinElmer, Foster City, CA) for ≥1 hour. Sections were washed twice with DIG-1 for 10 minutes, incubated for 30 minutes with the Biotinyl Tyramide Amplification Kit (NEL700A; PerkinElmer) diluted 1:50 in dilution buffer (PerkinElmer), washed with DIG-1 twice for 10 minutes each, and incubated with ABC reagents (Vector) for 1 hour. Sections were washed twice with DIG-1 for 10 minutes and then incubated with Alexa Fluor 488 conjugated streptavidin (diluted 1:500 with DIG-1) and alkaline phosphatase conjugated anti-DIG antibody (diluted 1:1000 with DIG-1) for 2 hours. Then, the sections were washed twice with DIG-1 for 10 minutes. After detection of positive signals, alkaline phosphatase activity was detected with a Fast-Red substrate kit (Roche) according to the manufacturer’s instructions. The incubation for this substrate was carried out until visible signals were detected and was stopped by washing in PBS containing 0.5 mM EDTA. Sections were then coverslipped with CC/Mount (Diagnostic BioSystem). Fluorescent signals were observed and documented with a LSM-710 confocal laser-scanning microscope (Carl Zeiss). We calculated the percentage of coexpression between esr1-expressing cells and vglut- or gad-expressing cells by counting the number of esr1-expressing cells and both esr1- and vglut- or gad-positive cells. Numbers of cells were expressed as mean ± SD. Results Histological analysis of the expression of esr1 in gnrh1 neurons We examined the expression of esr1 in hypophysiotropic gnrh1 neurons in the preoptic area (POA) by using in situ hybridization of adjacent sections [Fig. 1(a)–1(d)]. We found that esr1 mRNA was localized in the medial POA [Fig. 1(c)]. On the other hand, gnrh1 mRNA were localized in the lateral POA [Fig. 1(d)]. Therefore, to determine whether gnrh1 neurons express esr1, we performed double in situ hybridization for esr1 and gnrh1 mRNA [Fig. 1(e) and 1(f)]. We found 516 ± 10 esr1-expressing cells and 32 ± 4 gnrh1-expressing cells in the POA (means ± SD, n = 2 fish). However, we did not find any coexpression of gnrh1 and esr1 mRNA [Fig. 1(f)]. These results indicate that esr1 is not expressed in gnrh1 neurons. Figure 1. View largeDownload slide gnrh1 neurons do not express esr1 in medaka POA. (a, b) Schematic illustrations of (a) lateral (left is rostral) and (b) frontal sections of the medaka brain showing the plane of section corresponding to the photographs in (c) and (d). The boxed area in the POA is shown in (c) and (d). (c, d) Light photomicrographs showing in situ hybridization for (c) esr1 mRNA and (d) gnrh1 mRNA in adjacent sections. These photographs show that gnrh1 neurons are localized in the region lateral to esr1 neurons in POA. (e) Schematic illustrations of a frontal section of the medaka brain, showing the plane of section corresponding to the photographs in (f). The boxed area in (e) is shown in (f). (f) Photographs showing double labeling for esr1 mRNA (magenta) and gnrh1 mRNA (green), which demonstrate that POA gnrh1 neurons do not coexpress esr1. ca, anterior commissure; dDm, dorsal region of the medial part of dorsal telencephalic area; DI, lateral part of dorsal telencephalic area. Scale bars: 25 μm. Figure 1. View largeDownload slide gnrh1 neurons do not express esr1 in medaka POA. (a, b) Schematic illustrations of (a) lateral (left is rostral) and (b) frontal sections of the medaka brain showing the plane of section corresponding to the photographs in (c) and (d). The boxed area in the POA is shown in (c) and (d). (c, d) Light photomicrographs showing in situ hybridization for (c) esr1 mRNA and (d) gnrh1 mRNA in adjacent sections. These photographs show that gnrh1 neurons are localized in the region lateral to esr1 neurons in POA. (e) Schematic illustrations of a frontal section of the medaka brain, showing the plane of section corresponding to the photographs in (f). The boxed area in (e) is shown in (f). (f) Photographs showing double labeling for esr1 mRNA (magenta) and gnrh1 mRNA (green), which demonstrate that POA gnrh1 neurons do not coexpress esr1. ca, anterior commissure; dDm, dorsal region of the medial part of dorsal telencephalic area; DI, lateral part of dorsal telencephalic area. Scale bars: 25 μm. Establishment of esr1:EGFP Tg line We generated a Tg medaka line that expresses EGFP specifically in esr1 neurons [Fig. 2(a)–2(c)]. We found that the EGFP-positive cells were distributed mainly in the supracommissural part of ventral telencephalic area (Vs), POA, and ventral tuberal nucleus (NVT) region, all of which have previously been reported to contain esr-expressing neurons in medaka (21). To examine specificity, we analyzed the results of double labeling with EGFP immunohistochemistry and esr1 mRNA in situ hybridization [Fig. 2(b) and 2(c)]. We found that 44% of POA-esr1 neurons (201 ± 23/460 ± 51 cells, n = 3 fish) and about 70% of Vs-esr1 neurons (n = 2) and NVT-esr1 neurons (n = 2) were labeled with anti-EGFP antibody. Although some EGFP-negative esr1 cells were observed in these areas, all the EGFP positive cells expressed esr1, at least in these areas. Thus, the specificity of EGFP labeling is satisfactory for performing anatomical analysis of the axonal projections of these neurons. Figure 2. View largeDownload slide Establishment of esr1:EGFP Tg medaka. (a) The construct used to generate the esr1:EGFP Tg medaka. The EGFP-coding sequence was fused to the 3.7-kb DNA fragment containing the 5′-flanking region of exon1, intron1, and part of exon2, upstream of the first methionine of the esr1 gene. This construct contains the cardiac myosin light chain 2 (cmlc2) promoter region to express EGFP in the heart for screening. (b) Schematic illustrations of lateral (left is rostral) and frontal sections of the medaka brain, showing the plane of section corresponding to the panels in (c)–(e). The boxed area indicates the POA population of esr1 neurons shown in (c); note that the midline is located in the center of the pictures in (c). The red box indicates the POA population of esr2a and esr2b neurons shown in (d) and (e). (c) Photographs showing double labeling for EGFP immunohistochemistry (green) and esr1 mRNA (magenta) in the POA, which demonstrates that EGFP specifically labeled esr1 neurons. In addition to the POA, we confirmed specificity of EGFP labeling in the Vs and NVT, which also express esr1 (data not shown). (d) Photographs showing double labeling for EGFP immunohistochemistry (green) and esr2a mRNA (magenta) in the POA, which demonstrates that EGFP labeled esr2a neurons (arrowhead). Approximately 8% of esr2a neurons were labeled by EGFP (9 ± 3/115 ± 33 cells, n = 2 fish). (e) Photographs showing double labeling for EGFP immunohistochemistry (green) and esr2b mRNA (magenta) in the POA, which demonstrates that EGFP labeled esr2b neurons (arrowhead). Approximately 10% of esr2b neurons were labeled by EGFP (41 ± 11/424 ± 57 cells, n = 2 fish). ca, anterior commissure; dDm, dorsal region of the medial part of dorsal telencephalic area; DI, lateral part of dorsal telencephalic area. Scale bars: 10 μm. Figure 2. View largeDownload slide Establishment of esr1:EGFP Tg medaka. (a) The construct used to generate the esr1:EGFP Tg medaka. The EGFP-coding sequence was fused to the 3.7-kb DNA fragment containing the 5′-flanking region of exon1, intron1, and part of exon2, upstream of the first methionine of the esr1 gene. This construct contains the cardiac myosin light chain 2 (cmlc2) promoter region to express EGFP in the heart for screening. (b) Schematic illustrations of lateral (left is rostral) and frontal sections of the medaka brain, showing the plane of section corresponding to the panels in (c)–(e). The boxed area indicates the POA population of esr1 neurons shown in (c); note that the midline is located in the center of the pictures in (c). The red box indicates the POA population of esr2a and esr2b neurons shown in (d) and (e). (c) Photographs showing double labeling for EGFP immunohistochemistry (green) and esr1 mRNA (magenta) in the POA, which demonstrates that EGFP specifically labeled esr1 neurons. In addition to the POA, we confirmed specificity of EGFP labeling in the Vs and NVT, which also express esr1 (data not shown). (d) Photographs showing double labeling for EGFP immunohistochemistry (green) and esr2a mRNA (magenta) in the POA, which demonstrates that EGFP labeled esr2a neurons (arrowhead). Approximately 8% of esr2a neurons were labeled by EGFP (9 ± 3/115 ± 33 cells, n = 2 fish). (e) Photographs showing double labeling for EGFP immunohistochemistry (green) and esr2b mRNA (magenta) in the POA, which demonstrates that EGFP labeled esr2b neurons (arrowhead). Approximately 10% of esr2b neurons were labeled by EGFP (41 ± 11/424 ± 57 cells, n = 2 fish). ca, anterior commissure; dDm, dorsal region of the medial part of dorsal telencephalic area; DI, lateral part of dorsal telencephalic area. Scale bars: 10 μm. In addition, we analyzed the results of double labeling analysis with EGFP immunohistochemistry and esr2a/esr2b mRNA in situ hybridization [Fig. 2(d) and 2(e)], because we previously suggested that esr1, esr2a (or erβ1), and esr2b (or erβ2) were widely distributed in POA and that their distribution patterns in the POA were similar to each other. We found that EGFP also labeled esr2a and esr2b in this Tg line [Fig. 2(d) and 2(e)] and that 8% of POA-esr2a neurons (9 ± 3/115 ± 33 cells, n = 2 fish) [Fig. 2(d)] and ~10% of POA-esr2b neurons (41 ± 11/424 ± 57 cells, n = 2 fish) [Fig. 2(e)] were labeled with anti-EGFP antibody (Table 3). Table 3. Percentage of EGFP-Labeled esrs-Expressing Cells   Average Number of EGFP-Labeled Cells/esr Neurons in POA  % of EGFP-Labeled Cells  esr1  201 ± 23/460 ± 51 (n = 3 fish)  44  esr2a  9 ± 3/115 ± 33 (n = 2 fish)  8  esr2b  41 ± 11/424 ± 57 (n = 2 fish)  10    Average Number of EGFP-Labeled Cells/esr Neurons in POA  % of EGFP-Labeled Cells  esr1  201 ± 23/460 ± 51 (n = 3 fish)  44  esr2a  9 ± 3/115 ± 33 (n = 2 fish)  8  esr2b  41 ± 11/424 ± 57 (n = 2 fish)  10  View Large Anatomical analysis of the axonal projections of the esr1 neurons We performed EGFP immunohistochemistry in the brain of esr1:EGFP Tg medaka to analyze axonal projections of esr1 neurons (Fig. 3). Immunohistochemistry for EGFP, which enhances EGFP signals, enabled precise morphological analysis of axonal projections. Most of the POA-esr1 neurons projected their axons to the lateral POA. These fibers formed a thick bundle and projected caudally [Fig. 3(b)]. Frontal sections show that fibers project caudally, passing through the ventrolateral region of the telencephalon and hypothalamus [Fig. 3(c)]. In the hypothalamus, axons of the POA-esr1 neurons coursed laterally to the anterior tuberal nucleus (NAT) and NVT. These bundles of fibers innervated the pituitary, and although these results do not exclude the possibility that some POA-esr1 neurons project to other regions, we may safely conclude that the majority of axons of the POA-esr1 neurons projected to the pituitary. Axons of esr1 neurons entered the rostral pituitary and projected widely to the rostral half of the pituitary, whereas only a few EGFP immunoreactive (ir) fibers were observed in the caudal region (data not shown). Figure 3. View largeDownload slide Light photomicrographs showing the (a, b) sagittal and (c) frontal sections of the esr1:EGFP neurons visualized with anti-EGFP antibody in the brain of female esr1:EGFP Tg medaka. Sections were counterstained with cresyl violet. (a) The boxed area in the schematic illustration (lateral view of the brain) is shown in (b). (b) Sagittal sections showing EGFP-ir cell bodies in POA (arrowhead) and fiber bundles (arrow). Each photograph shows different mediolateral levels of the brain. Most fibers of the POA-esr1 neurons project caudally and reach the ventral region of hypothalamus. (c) Frontal sections showing axons of the esr1 neurons. The left column contains schematic drawings of the boxed areas in the photographs of the right column. POA-esr1 neurons project their axons (arrows) through the lateral region of the POA and run caudally, passing the ventrolateral region of hypothalamus, and project to the pituitary. dDI, dorsal region of DI; DI, lateral part of dorsal telencephalic area; DM, dorsomedial thalamic nucleus; Dp, posterior part of dorsal telencephalic area; GR, corpus glomerulosus; nII, optic nerve; NPPv, periventricular posterior nucleus; POm, magnocellular preoptic nucleus; POp, parvocellular preoptic nucleus; TO, optic tectum; TS, torus semicircularis; VM, ventromedial thalamic nucleus; Vp, postcommissural part of V. Scale bars: 100 μm. Figure 3. View largeDownload slide Light photomicrographs showing the (a, b) sagittal and (c) frontal sections of the esr1:EGFP neurons visualized with anti-EGFP antibody in the brain of female esr1:EGFP Tg medaka. Sections were counterstained with cresyl violet. (a) The boxed area in the schematic illustration (lateral view of the brain) is shown in (b). (b) Sagittal sections showing EGFP-ir cell bodies in POA (arrowhead) and fiber bundles (arrow). Each photograph shows different mediolateral levels of the brain. Most fibers of the POA-esr1 neurons project caudally and reach the ventral region of hypothalamus. (c) Frontal sections showing axons of the esr1 neurons. The left column contains schematic drawings of the boxed areas in the photographs of the right column. POA-esr1 neurons project their axons (arrows) through the lateral region of the POA and run caudally, passing the ventrolateral region of hypothalamus, and project to the pituitary. dDI, dorsal region of DI; DI, lateral part of dorsal telencephalic area; DM, dorsomedial thalamic nucleus; Dp, posterior part of dorsal telencephalic area; GR, corpus glomerulosus; nII, optic nerve; NPPv, periventricular posterior nucleus; POm, magnocellular preoptic nucleus; POp, parvocellular preoptic nucleus; TO, optic tectum; TS, torus semicircularis; VM, ventromedial thalamic nucleus; Vp, postcommissural part of V. Scale bars: 100 μm. On the other hand, analysis of sagittal sections showed that Vs-esr1 neurons projected mainly to caudal regions of the brain [Fig. 4(a) and 4(b)]. Analysis of frontal sections showed that these EGFP-positive axons of the Vs-esr1 neurons projected caudally, passing near the midline region of the hypothalamus bilaterally (data not shown). These axons passed through the periventricular region of the hypothalamus, in the region lateral to the periventricular posterior nucleus, and coursed further caudally (data not shown). Finally, axons of Vs-esr1 neurons terminated in the region ventral to medial reticular formation [Fig. 4(c)]. In this Tg line, no EGFP-ir axons were observed projecting to the spinal cord. Figure 4. View largeDownload slide Light photomicrographs showing (a, b) sagittal and (c) frontal sections of esr1:EGFP neurons visualized with EGFP antibody in the brain of female esr1:EGFP Tg medaka. Sections were counterstained with cresyl violet. (a) Sagittal section (right) showing EGFP-ir cell bodies and axons. The boxed area in the schematic illustration (left, lateral view of the brain) is shown in the photographs in (b). Arrowheads indicate cell bodies in Vs region. The EGFP-ir neurons in Vs (Vs-EGFP-ir neurons) project their axons to the caudal region. (b) Sagittal sections more lateral to the photograph of (a), showing that axons from Vs region project caudally, which suggests that these fibers project to the medulla, passing through the diencephalon. Arrows indicate fibers that originate from Vs-esr1 neurons. (c) Frontal section showing fibers of the Vs EGFP-ir neurons in the ventral region of medulla at the level indicated in the lateral view of the brain (top left). The boxed area in the schematic illustration (top right) is shown in the photograph below. Arrowheads indicate EGFP-ir axons. EGFP-ir fibers were not observed in more caudal regions. nII, optic nerve; RFm, medial reticular formation; TE, telencephalon. Scale bars: (a, b) 100 μm, (c) 50 μm. Figure 4. View largeDownload slide Light photomicrographs showing (a, b) sagittal and (c) frontal sections of esr1:EGFP neurons visualized with EGFP antibody in the brain of female esr1:EGFP Tg medaka. Sections were counterstained with cresyl violet. (a) Sagittal section (right) showing EGFP-ir cell bodies and axons. The boxed area in the schematic illustration (left, lateral view of the brain) is shown in the photographs in (b). Arrowheads indicate cell bodies in Vs region. The EGFP-ir neurons in Vs (Vs-EGFP-ir neurons) project their axons to the caudal region. (b) Sagittal sections more lateral to the photograph of (a), showing that axons from Vs region project caudally, which suggests that these fibers project to the medulla, passing through the diencephalon. Arrows indicate fibers that originate from Vs-esr1 neurons. (c) Frontal section showing fibers of the Vs EGFP-ir neurons in the ventral region of medulla at the level indicated in the lateral view of the brain (top left). The boxed area in the schematic illustration (top right) is shown in the photograph below. Arrowheads indicate EGFP-ir axons. EGFP-ir fibers were not observed in more caudal regions. nII, optic nerve; RFm, medial reticular formation; TE, telencephalon. Scale bars: (a, b) 100 μm, (c) 50 μm. Morphological analysis of axonal projections of esr1 neurons to gnrh1 neurons To assess the participation of Esr1 neurons in mediating estrogen feedback to GnRH1 neurons, we analyzed the results of double labeling with EGFP immunohistochemistry and gnrh1 mRNA in situ hybridization (Fig. 5). EGFP-positive axons from POA-esr1 neurons surrounded gnrh1-expressing neuronal cell bodies localized in the ventrolateral POA. Seventy-seven percent of gnrh1 neurons (30 ± 6/39 ± 6 cells, n = 3 fish) appeared to receive axonal projections from POA-esr1 neurons. On the other hand, very few axons were observed around the dorsal and medial groups of gnrh1 neurons, which have been suggested not to project to the pituitary (22, 23, 26). Figure 5. View largeDownload slide Double labeling analysis with esr1:EGFP Tg medaka suggests that POA-esr1 neurons directly contact the cell bodies of the gnrh1 neurons. (a) Illustration of a frontal section of the medaka brain, showing the plane of section corresponding to the photographs in (b). The boxed area is shown in (b). (b) Photographs showing double labeling using EGFP immunohistochemistry (green) and gnrh1 mRNA in situ hybridization (magenta) demonstrate that the cell bodies of gnrh1 neurons are surrounded by EGFP-ir fibers. Taken together with the distribution of axons of esr1 neurons, POA-esr1 neurons are suggested to make direct contacts on gnrh1 neurons. dDI, dorsal region of DI; DI, lateral part of dorsal telencephalic area; Dp, posterior part of dorsal telencephalic area; nII, optic nerve; POm, magnocellular preoptic nucleus; POp, parvocellular preoptic nucleus; Vp, postcommissural part of V. Scale bars: 25 μm. Figure 5. View largeDownload slide Double labeling analysis with esr1:EGFP Tg medaka suggests that POA-esr1 neurons directly contact the cell bodies of the gnrh1 neurons. (a) Illustration of a frontal section of the medaka brain, showing the plane of section corresponding to the photographs in (b). The boxed area is shown in (b). (b) Photographs showing double labeling using EGFP immunohistochemistry (green) and gnrh1 mRNA in situ hybridization (magenta) demonstrate that the cell bodies of gnrh1 neurons are surrounded by EGFP-ir fibers. Taken together with the distribution of axons of esr1 neurons, POA-esr1 neurons are suggested to make direct contacts on gnrh1 neurons. dDI, dorsal region of DI; DI, lateral part of dorsal telencephalic area; Dp, posterior part of dorsal telencephalic area; nII, optic nerve; POm, magnocellular preoptic nucleus; POp, parvocellular preoptic nucleus; Vp, postcommissural part of V. Scale bars: 25 μm. Neurotransmitter candidates for the POA-esr1 neurons We next examined neurotransmitter candidates for POA-esr1 neurons. First, we collected EGFP-positive neurons from the POA of female esr1:EGFP Tg medaka and performed RNA-seq of those neurons. It should be noted that we found no trace of kiss1 or kiss2 gene expression, contrary to what we may expect from results in mammals (8, 27). These results strongly support the results of previous in situ hybridization studies demonstrating that neither kiss1 nor kiss2 neurons are localized in the POA in medaka (11, 12, 28). Therefore, we examined the expression of marker genes for classic transmitters, glutamate and γ-aminobutyric acid (GABA). As markers for glutamatergic and γ-aminobutyric acidergic (GABAergic) transmission, medaka has been reported to possess five subtypes of vglut (vglut1.1, vglut1.2, vglut2.1, vglut2.2, and vglut3) and three subtypes of gad (gad1.1, gad 1.2, and gad2), respectively (25). We evaluated the expression level of vglut and gad genes based on the average RPKM; in this approach, only vglut2.1,gad1.1, and gad2 were detected from the EGFP-positive neurons (RPKM for vglut2.1, 2.13 ± 3.9; gad1.1, 95.1 ± 86; gad2, 23.6 ± 14; others, RPKM < 1). This result suggests that POA-esr1 neurons include both glutamatergic and GABAergic neurons. Next, we analyzed colocalization of glutamatergic or GABAergic markers in the POA-esr1 neurons by double in situ hybridization for mRNA of esr1 and either vglut2.1, gad1.1, gad1.2, or gad2 (Fig. 6). These results are summarized in Table 4. More than 95% of POA-esr1 neurons coexpressed vglut 2.1 [(Fig. 6(b)] (326 ± 81/342 ± 71 cells, n = 3 fish). In addition, 36% of POA-esr1 neurons coexpressed gad 1.1 [(Fig. 6(c)] (148 ± 23/410 ± 42 cells, n = 2 fish). Because most esr1 neurons were shown here to be glutamatergic by double labeling of esr1 and vglut 2.1, we did not examine other subtypes of vglut. Among the GABA-synthesizing enzymes, gad1.1 showed the highest percentage of colocalization with esr1 mRNA (Fig. 6). However, the colocalization ratio for gad1.1 was much lower than that of vglut2.1, which suggests that glutamate is the main classic neurotransmitter in POA-esr1 neurons. In addition, RNA-seq results suggested that EGFP-positive POA neurons from esr1:EGFP medaka express not only esr1 (RPKM 188 ± 81) but also esr2a and esr2b (RPKM 28.2 ± 19 and 3.02 ± 4.8). These results are consistent with our histological results [Fig. 2(d) and 2(e)]. Moreover, among other sex steroid receptor genes, we found that POA-esr1 neurons highly express pr (RPKM 155 ± 113). Therefore, we analyzed double labeling with EGFP immunohistochemistry and pr mRNA in situ hybridization. We found that ~68% of EGFP-expressing neurons in POA coexpressed pr (167 ± 33/246 ± 33 cells, n = 3 fish) (Fig. 7). Figure 6. View largeDownload slide Double in situ hybridization for esr1 and marker genes for glutamate or GABA in POA. (a) Schematic illustration of a frontal section of the medaka brain, showing the plane of section corresponding to the photographs. The boxed area corresponds to the photographs in (b)–(e). (b) Photographs showing double labeling for vglut2.1 mRNA (green) and esr1 mRNA (magenta), demonstrating that POA-esr1–expressing cells express vglut2.1. These results suggest that the majority of the POA-esr1 neurons are glutamatergic. (c) Photographs showing double labeling for gad1.1 mRNA (green) and esr1 mRNA (magenta), which demonstrate that some POA-esr1–expressing cells coexpress gad1.1. About 36% of POA-esr1–expressing neurons expressed gad1.1 (148 ± 23/410 ± 42 cells, n = 2 fish). These results suggest that most POA-esr1–expressing neurons produce glutamate, and some POA-esr1–expressing neurons produce GABA. (d) Photographs showing merged images of double in situ hybridization for gad1.2 mRNA (green) and esr1 mRNA (magenta), demonstrating that very few POA-esr1 cells express gad1.2 (arrowhead). (e) Photographs showing merged images of double in situ hybridization for gad2 mRNA (green) and esr1 mRNA (magenta), demonstrating that some POA-esr1 cells express gad2 (arrowhead). ca, anterior commissure; DI, lateral part of dorsal telencephalic area; dDm, dorsal region of the medial part of dorsal telencephalic area. Scale bars: 25 μm. Figure 6. View largeDownload slide Double in situ hybridization for esr1 and marker genes for glutamate or GABA in POA. (a) Schematic illustration of a frontal section of the medaka brain, showing the plane of section corresponding to the photographs. The boxed area corresponds to the photographs in (b)–(e). (b) Photographs showing double labeling for vglut2.1 mRNA (green) and esr1 mRNA (magenta), demonstrating that POA-esr1–expressing cells express vglut2.1. These results suggest that the majority of the POA-esr1 neurons are glutamatergic. (c) Photographs showing double labeling for gad1.1 mRNA (green) and esr1 mRNA (magenta), which demonstrate that some POA-esr1–expressing cells coexpress gad1.1. About 36% of POA-esr1–expressing neurons expressed gad1.1 (148 ± 23/410 ± 42 cells, n = 2 fish). These results suggest that most POA-esr1–expressing neurons produce glutamate, and some POA-esr1–expressing neurons produce GABA. (d) Photographs showing merged images of double in situ hybridization for gad1.2 mRNA (green) and esr1 mRNA (magenta), demonstrating that very few POA-esr1 cells express gad1.2 (arrowhead). (e) Photographs showing merged images of double in situ hybridization for gad2 mRNA (green) and esr1 mRNA (magenta), demonstrating that some POA-esr1 cells express gad2 (arrowhead). ca, anterior commissure; DI, lateral part of dorsal telencephalic area; dDm, dorsal region of the medial part of dorsal telencephalic area. Scale bars: 25 μm. Table 4. Colocalization of Glutamatergic or GABAergic Markers in the POA-esr1 Neurons   Average Number of Cells in POA  % of Each Neurotransmitter Expressing POA-esr1 Neurons  vglut2.1 and esr1-expressing neuron  326 ± 81 (in 342 ± 71 esr1 neurons, n = 3 fish)  95  gad1.1 and esr1-expressing neuron  148 ± 23 (in 410 ± 42 esr1 neurons, n = 2 fish)  36  gad1.2 and esr1-expressing neuron  20 ± 4 (in 515 ± 28 esr1 neurons, n = 2 fish)  4  gad2 and esr1-expressing neuron  110 ± 18 (in 430 ± 56 esr1 neurons, n = 2 fish)  26    Average Number of Cells in POA  % of Each Neurotransmitter Expressing POA-esr1 Neurons  vglut2.1 and esr1-expressing neuron  326 ± 81 (in 342 ± 71 esr1 neurons, n = 3 fish)  95  gad1.1 and esr1-expressing neuron  148 ± 23 (in 410 ± 42 esr1 neurons, n = 2 fish)  36  gad1.2 and esr1-expressing neuron  20 ± 4 (in 515 ± 28 esr1 neurons, n = 2 fish)  4  gad2 and esr1-expressing neuron  110 ± 18 (in 430 ± 56 esr1 neurons, n = 2 fish)  26  View Large Figure 7. View largeDownload slide Double labeling analysis with esr1:EGFP Tg medaka suggests that POA-esr1 neurons coexpress pr. Photographs showing double labeling EGFP immunohistochemistry (green) and pr mRNA in situ hybridization (magenta), demonstrating that POA-esr1 neurons express pr (arrowhead). ca, anterior commissure; DI, lateral part of dorsal telencephalic area; dDm, dorsal region of the medial part of dorsal telencephalic area. Scale bars: 20 μm. Figure 7. View largeDownload slide Double labeling analysis with esr1:EGFP Tg medaka suggests that POA-esr1 neurons coexpress pr. Photographs showing double labeling EGFP immunohistochemistry (green) and pr mRNA in situ hybridization (magenta), demonstrating that POA-esr1 neurons express pr (arrowhead). ca, anterior commissure; DI, lateral part of dorsal telencephalic area; dDm, dorsal region of the medial part of dorsal telencephalic area. Scale bars: 20 μm. Discussion In the current study, we generated Tg medaka in which esr1-expressing neurons are labeled with EGFP and examined the anatomy of these neurons. We identified the neuronal pathways mediating estrogen feedback signals to gnrh1 neurons, as well as those mediating sex steroid–sensitive signals to a circuit that modulates sex behaviors. We found that POA-esr1 neurons project their axons to gnrh1 neurons and that the former comprise glutamatergic/GABAergic neurons, which may be widely conserved among mammals and teleosts. We also found a possible neural substrate for estrogen-sensitive modulation of sexual behaviors. Taken together with results of previous behavioral studies suggesting involvement of Vs neurons in the regulation of sexual behaviors (29–31), the present results suggest that the caudally projecting esr1 neurons in Vs are involved in regulating sexual behaviors with respect to the level of circulating estrogen. In addition, we found that some POA-esr1 neurons coexpress pr, which suggests that progesterone, in addition to estrogen, may regulate reproduction or sexual behaviors by controlling POA-esr1/pr neurons in accordance with serum hormone levels. gnrh1 neurons do not express esr1 in medaka In the current study, we analyzed the neuronal circuits involved in transmission of estrogenic signals to gnrh1 neurons in medaka. Esr1 has been demonstrated to be an essential factor for regulation of the HPG axis in mammals (6, 8). Therefore, we first examined whether gnrh1 neurons express esr1 in medaka (Fig. 1). Double in situ hybridization for gnrh1 and esr1 mRNA disproved the expression of esr1 in gnrh1 neurons [Fig. 1(f)]. Thus, as in mammals, gnrh1 neurons do not directly receive estrogenic signals via Esr1 in medaka (32, 33). Specificity of EGFP labeling in the brain of esr1:EGFP Tg medaka In the brains of esr1:EGFP Tg medaka, we found that EGFP-labeled cell bodies were distributed in the POA, Vs, and NVT, and we demonstrated the specificity of EGFP labeling in these regions (Fig. 2). It should be noted that not all esr1 neurons were labeled with EGFP; on the other hand, EGFP labeling was found in nucleus diffusus tori lateralis and the optic tectum, regions in which esr1 mRNA was not detected by in situ hybridization. So far, it is technically impossible to determine whether this signal indicates ectopically labeled neurons or neurons labeled more sensitively than with in situ hybridization. Either way, these labeled neurons projected locally and did not interfere with analysis of axonal projections of POA or Vs neurons in the current study. Interestingly, we also demonstrated, by double labeling of esr2a/esr2b mRNA in situ hybridization and EGFP immunohistochemistry, that some EGFP-positive esr1 neurons coexpressed esr2a or esr2b [Fig. 2(d) and 2(e)]. This finding is consistent with RNA-seq results, which suggested that the esr1 neuronal population includes esr2a or esr2b mRNA. Thus, a small population of esr1 neurons in this Tg line can be considered esr1/2a or esr1/2b neurons. Functions of POA-Esr1 neurons Using our Tg medaka line, we obtained morphological evidence that glutamatergic and GABAergic POA-esr1 neurons directly regulate hypophysiotropic gnrh1 neurons (Figs. 5 and 6 ). Many close contacts were observed between the gnrh1 neuronal cell bodies and the nerve terminals of POA-esr1 neurons, most of which were proven to be glutamatergic, and some of them were GABAergic. In medaka, a previous study demonstrated that expression of esr1,esr2a, and esr2b in the POA region was not changed by either ovariectomy or estrogen replacement (34). Thus, esr1 is considered to be stably expressed in the POA and to regulate the release of glutamate or GABA in accordance with the serum estrogen level. Because glutamate and GABA are the major excitatory synaptic transmitters (GABA is also suggested to be excitatory to GnRH1 neurons), it follows that estrogen-dependent modulation of glutamatergic and GABAergic transmission may affect GnRH1 firing activity. In medaka, the firing activity of GnRH1 neurons changes toward puberty (35); therefore, it is possible that the system for regulating GnRH1 neuron activity changes during pubertal development. After sexual maturation, adult female medaka show daily spawning, and our previous study has documented time-of-day–dependent changes in the firing activity of GnRH1 neurons and in expression of LH and FSH (23). However, the neuronal mechanisms underlying puberty onset or time-of-day–dependent changes in GnRH1 neuronal activities have not yet been clarified. In mammals, glutamate/GABA is thought to play important roles in the regulation of puberty onset and the estrous cycle; the proportion of glutamatergic agonist α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid–responsive GnRH1 neurons changes across puberty (36). Moreover, it has been reported that the number of fast synaptic transmissions (Glu/GABA) to GnRH1 neurons changes in accordance with the serum estrogen level in mammals (37, 38). Finally, a recent study suggested that glutamatergic Esr1 neurons in the mouse limbic forebrain play a key role in the negative/positive steroid feedback regulation of GnRH neurons, whereas GABAergic Esr1 neurons play a role in the positive feedback (39). Thus, estrogen-dependent alteration of glutamatergic and GABAergic inputs may be important for the reproductive regulation of vertebrates in general, and the glutamatergic and GABAergic neurons that project to the gnrh1 neurons morphologically identified here in medaka are candidates for such regulatory neurons. In the current study, we also found that POA-esr1 neurons coexpress pr (Fig. 7). Interestingly, in mice, arcuate GABA/PR neurons have been suggested to project heavily to GnRH1 neurons (40). Results of the current study suggested that 36% of POA-esr1 neurons are GABAergic. Thus, some POA-esr1/pr neurons that express GABA probably project to gnrh1 neurons in medaka. Taken together with the results of previous study in mice, our results show that the innervation of gnrh1 neurons by gaba/pr neurons seems to be conserved among mammals and teleosts, and the gaba/pr neurons may be important for the regulation of the HPG axis in vertebrates. Medaka brains, even in adults, are small and are advantageous for analyzing EGFP-labeled POA-esr1 neurons in whole brain in vitro preparations, in which we can maintain the neuronal circuitries regulating the GnRH1 neuron activities almost intact. Therefore, the Tg medaka we established in the current study should provide a good model system for studying estrogen-dependent changes in glutamatergic and GABAergic synaptic inputs that regulate GnRH1 neuron activities and hence the regulation of reproduction. In addition to these neuronal circuits, mammals have developed a Kiss1-mediated HPG axis regulation system for the fine tuning of reproduction; for example, follicular development and ovulation are suppressed during lactation in mammals, mainly because of the suppression of pulsatile GnRH/LH secretion (41). On the other hand, in teleosts it has been recently demonstrated that GnRH is necessary for ovulation but not folliculogenesis (5), suggesting that glutamatergic innervation of GnRH1 neurons may be important for LH surges and ovulation rather than LH pulses. Interestingly, in a daily surge model of OVX+E mouse (42), spontaneous glutamatergic excitatory postsynaptic potentials from GnRH neurons drastically increase in the afternoon (37). These lines of evidence suggest that glutamatergic transmission may be important for LH surges in both species, although other possibilities cannot be excluded. We also examined the projections from esr1 neurons to the pituitary. Results of the present anatomical analysis suggested that POA-esr1 neurons may directly regulate activity of gonadotropes. However, it should be noted that axons of the esr1 neurons were distributed broadly in the rostral half of the pituitary gland, which suggests that esr1 neurons may also regulate the release of hormones other than gonadotropins. Further investigation with this Tg line is necessary to clarify whether the POA-esr1 neurons directly regulate gonadotropin release. Regulation of sexual behaviors by Vs-Esr1 neurons and POA-Esr1/PR neurons Among rodents, estrogen is suggested to activate lordosis in female rats (19). However, the site of estrogen action in the neuronal circuits of this sexual behavior has not been clarified so far. In teleosts, previous studies reported that electrical stimulation of POA or Vs acutely induced sexual behavior in hime salmon (30). Thus, neurons in POA and Vs may be involved in the regulation of motor control of sexual behavior. In the current study, we showed that Vs-esr1 neurons project as far as the ventral region of medulla. However, we did not find neural fibers of Vs-esr1 neurons in the spinal cord in our Tg line. Therefore, uncharacterized neurons in the medulla that receive axonal projections from the Vs-esr1 neurons may relay this estrogenic modulation to the central pattern generator in the spinal cord that induces sexual behavior. Our present results clearly demonstrated that some of the esr1-expressing neurons in the POA coexpress pr [Fig. 7(b)], which is partly consistent with the results of previous studies in mice (43, 44). In mice, it has been shown that ablation of Esr1/PR neurons in the ventromedial hypothalamus diminishes sexual behaviors (31). Progesterone is suggested to induce sexual behavior in teleosts as well as in mammals (45). As suggested in rodent studies, it is possible that these Esr1/PR neurons are involved in sexual behavior. It should be noted that an immunohistochemical study in mammals demonstrated that PRs were detected only in Esr1 neurons (44). However, we suggest that some pr neurons in medaka do not express esr1. Although not all esr1 neurons were labeled by EGFP in the Tg line we established here, as shown in Fig. 7, many EGFP-negative pr neurons (326 ± 44/460 ± 42 cells, n = 3 fish) were found in the POA. Thus, it is possible that prs are expressed not only in esr1 neurons but also in esr1-negative neurons in medaka. The function of progesterone and PR is likely to differ between placental mammals and nonplacental teleosts, and thus additional studies using diverse mammals and teleosts are necessary to elucidate the reasons for this difference. A recent study in mice suggested that optogenetic activation of Esr1 neurons in the ventrolateral subdivision of the ventromedial hypothalamus induced sexual behavior (46). Additional investigation using genetic tools, such as Tg medaka that express optogenetic tools specifically in Esr1 neurons, may enable us to analyze neuronal circuits underlying estrogen-primed modulation of sexual behavior. In conclusion, in the current study we succeeded in anatomically identifying and visualizing two kinds of esr1 neurons: glutamatergic and/or GABAergic esr1 neurons in the POA that may relay gonadal estrogen feedback signals to gnrh1 neurons and esr1 neurons in the Vs that may modulate sexual behavior by acting on medullary circuits (Fig. 8). The Tg medaka line established in the current study will contribute to our understanding of the estrogen regulation of reproduction and behavior in vertebrates. Figure 8. View largeDownload slide Illustration of a working hypothesis concerning Esr1-mediated regulation of reproduction and sexual behavior. Glutamatergic and GABAergic POA-Esr1 neurons receive estrogenic feedback signals and directly regulate GnRH1 neurons in accordance with the gonadal status. Glutamatergic and GABAergic synaptic transmission to GnRH1 neuron is altered by estrogen levels. Estrogen also acts on Vs-Esr1 neurons and possibly modulates expression of neurotransmitters or firing activity in accordance with gonadal status. Unidentified neurons in the medulla that receive axonal projections from Vs-Esr1 neurons are suggested to relay this estrogenic modulation to the motor pattern generator in the spinal cord to modulate sexual behavior. Although the projection has not been clarified, Esr1/PR neurons are also considered to be involved in the regulation of sexual maturation and sexual behaviors. E2, estrogen; P, progesterone. Figure 8. View largeDownload slide Illustration of a working hypothesis concerning Esr1-mediated regulation of reproduction and sexual behavior. Glutamatergic and GABAergic POA-Esr1 neurons receive estrogenic feedback signals and directly regulate GnRH1 neurons in accordance with the gonadal status. Glutamatergic and GABAergic synaptic transmission to GnRH1 neuron is altered by estrogen levels. Estrogen also acts on Vs-Esr1 neurons and possibly modulates expression of neurotransmitters or firing activity in accordance with gonadal status. Unidentified neurons in the medulla that receive axonal projections from Vs-Esr1 neurons are suggested to relay this estrogenic modulation to the motor pattern generator in the spinal cord to modulate sexual behavior. Although the projection has not been clarified, Esr1/PR neurons are also considered to be involved in the regulation of sexual maturation and sexual behaviors. E2, estrogen; P, progesterone. Abbreviations: ABC avidin-biotin complex DIG digoxigenin EGFP enhanced green fluorescent protein Esr estrogen receptor Esr1 estrogen receptor α Esr2 estrogen receptor β Esr2a estrogen receptor β1 Esr2b estrogen receptor β2 FSH follicle-stimulating hormone GABA γ-aminobutyric acid GABAergic γ-aminobutyric acidergic GnRH gonadotropin-releasing hormone GnRH1 gonadotropin-releasing hormone 1 HPG hypothalamo-pituitary-gonadal ir immunoreactive LH luteinizing hormone mRNA messenger RNA NAT anterior tuberal nucleus NVT ventral tuberal nucleus PBS phosphate-buffered saline PBST phosphate-buffered saline/Tween PCR polymerase chain reaction PFA paraformaldehyde POA preoptic area PR progesterone receptor RNA-seq RNA sequencing RPKM reads per kilobase of transcript per million mapped reads SD standard deviation SSC saline sodium citrate Tg transgenic TNE Tris-HCl, NaCl, and EDTA Vs ventral telencephalic area. Acknowledgments We thank Drs. Martin Kelly, Oline Rønnekleiv, and Martha Bosch (Oregon Health and Science University) for advice on the cell harvesting procedure for deep sequencing; Dr. Heather Eisthen (Michigan State University) for help in editing the manuscript; Drs. Takeshi Kaneko and Hiroyuki Hioki (Kyoto University) for the gift of anti-EGFP antibody; Drs. Min Kyun Park (The University of Tokyo) and Kataaki Okubo (The University of Tokyo) for helpful advice and discussion; Ms. Maiko Matsuda (The University of Tokyo) for support during the preparation of probes; and Ms. Miho Kyokuwa and Hisako Kohno (The University of Tokyo) for excellent care of the fish used in this study. Financial Support: This work was supported by Japan Society for the Promotion of Science (http://dx.doi.org/10.13039/501100001691) Grants 26221104 (to Y.O.) and 13J10475 (to B.Z.). Disclosure Summary: The authors have nothing to disclose. References 1. Okubo K, Nagahama Y. Structural and functional evolution of gonadotropin-releasing hormone in vertebrates. Acta Physiol (Oxf) . 2008; 193( 1): 3– 15. Google Scholar CrossRef Search ADS PubMed  2. Childs G. Gonadotropes and lactotropes. In: Neill JD, ed. Knobil and Neill’s The Physiology of Reproduction . 3rd ed. New York, NY: Academic Press; 2006: 1492– 1539. 3. Karigo T, Aikawa M, Kondo C, Abe H, Kanda S, Oka Y. Whole brain-pituitary in vitro preparation of the transgenic medaka (Oryzias latipes) as a tool for analyzing the differential regulatory mechanisms of LH and FSH release. Endocrinology . 2014; 155( 2): 536– 547. Google Scholar CrossRef Search ADS PubMed  4. Chu L, Li J, Liu Y, Cheng CH. Gonadotropin signaling in zebrafish ovary and testis development: insights from gene knockout study. Mol Endocrinol . 2015; 29( 12): 1743– 1758. Google Scholar CrossRef Search ADS PubMed  5. Takahashi A, Kanda S, Abe T, Oka Y. Evolution of the hypothalamic-pituitary-gonadal axis regulation in vertebrates revealed by knockout medaka. Endocrinology . 2016; 157( 10): 3994– 4002. Google Scholar CrossRef Search ADS PubMed  6. Dorling AA, Todman MG, Korach KS, Herbison AE. Critical role for estrogen receptor alpha in negative feedback regulation of gonadotropin-releasing hormone mRNA expression in the female mouse. Neuroendocrinology . 2003; 78( 4): 204– 209. Google Scholar CrossRef Search ADS PubMed  7. Couse JF, Yates MM, Walker VR, Korach KS. Characterization of the hypothalamic-pituitary-gonadal axis in estrogen receptor (ER) null mice reveals hypergonadism and endocrine sex reversal in females lacking ERalpha but not ERbeta. Mol Endocrinol . 2003; 17( 6): 1039– 1053. Google Scholar CrossRef Search ADS PubMed  8. Smith JT, Cunningham MJ, Rissman EF, Clifton DK, Steiner RA. Regulation of Kiss1 gene expression in the brain of the female mouse. Endocrinology . 2005; 146( 9): 3686– 3692. Google Scholar CrossRef Search ADS PubMed  9. George JT, Seminara SB. Kisspeptin and the hypothalamic control of reproduction: lessons from the human. Endocrinology . 2012; 153( 11): 5130– 5136. Google Scholar CrossRef Search ADS PubMed  10. Pielecka-Fortuna J, Chu Z, Moenter SM. Kisspeptin acts directly and indirectly to increase gonadotropin-releasing hormone neuron activity and its effects are modulated by estradiol. Endocrinology . 2008; 149( 4): 1979– 1986. Google Scholar CrossRef Search ADS PubMed  11. Mitani Y, Kanda S, Akazome Y, Zempo B, Oka Y. Hypothalamic Kiss1 but not Kiss2 neurons are involved in estrogen feedback in medaka (Oryzias latipes). Endocrinology . 2010; 151( 4): 1751– 1759. Google Scholar CrossRef Search ADS PubMed  12. Kanda S, Akazome Y, Matsunaga T, Yamamoto N, Yamada S, Tsukamura H, Maeda K, Oka Y. Identification of KiSS-1 product kisspeptin and steroid-sensitive sexually dimorphic kisspeptin neurons in medaka (Oryzias latipes). Endocrinology . 2008; 149( 5): 2467– 2476. Google Scholar CrossRef Search ADS PubMed  13. Kanda S, Akazome Y, Mitani Y, Okubo K, Oka Y. Neuroanatomical evidence that kisspeptin directly regulates isotocin and vasotocin neurons. PLoS One . 2013; 8( 4): e62776. Google Scholar CrossRef Search ADS PubMed  14. Grone BP, Maruska KP, Korzan WJ, Fernald RD. Social status regulates kisspeptin receptor mRNA in the brain of Astatotilapia burtoni. Gen Comp Endocrinol . 2010; 169( 1): 98– 107. Google Scholar CrossRef Search ADS PubMed  15. Escobar S, Servili A, Espigares F, Gueguen MM, Brocal I, Felip A, Gómez A, Carrillo M, Zanuy S, Kah O. Expression of kisspeptins and kiss receptors suggests a large range of functions for kisspeptin systems in the brain of the European sea bass. PLoS One . 2013; 8( 7): e70177. Google Scholar CrossRef Search ADS PubMed  16. Tang H, Liu Y, Luo D, Ogawa S, Yin Y, Li S, Zhang Y, Hu W, Parhar IS, Lin H, Liu X, Cheng CH. The kiss/kissr systems are dispensable for zebrafish reproduction: evidence from gene knockout studies. Endocrinology . 2015; 156( 2): 589– 599. Google Scholar CrossRef Search ADS PubMed  17. Nakajo M, Kanda S, Karigo T, Takahashi A, Akazome Y, Uenoyama Y, Kobayashi M, Oka Y. Evolutionally conserved function of kisspeptin neuronal system is nonreproductive regulation as revealed by nonmammalian study. Endocrinology . 2018; 159( 1): 163– 183. Google Scholar CrossRef Search ADS PubMed  18. Akazome Y, Kanda S, Okubo K, Oka Y. Functional and evolutionary insights into vertebrate kisspeptin systems from studies of fish brain. J Fish Biol . 2010; 76( 1): 161– 182. Google Scholar CrossRef Search ADS PubMed  19. Pfaff DW, Sakuma Y. Facilitation of the lordosis reflex of female rats from the ventromedial nucleus of the hypothalamus. J Physiol . 1979; 288: 189– 202. Google Scholar PubMed  20. Kim Y, Ko E, Lappan S, Kezele A, Kwon SW, Choe JC. Changes in fecal estrogen levels and sexual behavior in captive sichuan snub-nosed monkeys (Rhinopithecus roxellana) following a male replacement. Zoo Biol . 2013; 32( 3): 342– 346. Google Scholar CrossRef Search ADS PubMed  21. Zempo B, Kanda S, Okubo K, Akazome Y, Oka Y. Anatomical distribution of sex steroid hormone receptors in the brain of female medaka. J Comp Neurol . 2013; 521( 8): 1760– 1780. Google Scholar CrossRef Search ADS PubMed  22. Okubo K, Sakai F, Lau EL, Yoshizaki G, Takeuchi Y, Naruse K, Aida K, Nagahama Y. Forebrain gonadotropin-releasing hormone neuronal development: insights from transgenic medaka and the relevance to X-linked Kallmann syndrome. Endocrinology . 2006; 147( 3): 1076– 1084. Google Scholar CrossRef Search ADS PubMed  23. Karigo T, Kanda S, Takahashi A, Abe H, Okubo K, Oka Y. Time-of-day-dependent changes in GnRH1 neuronal activities and gonadotropin mRNA expression in a daily spawning fish, medaka. Endocrinology . 2012; 153( 7): 3394– 3404. Google Scholar CrossRef Search ADS PubMed  24. Sasagawa Y, Nikaido I, Hayashi T, Danno H, Uno KD, Imai T, Ueda HR. Quartz-Seq: a highly reproducible and sensitive single-cell RNA sequencing method, reveals non-genetic gene-expression heterogeneity [published correction appears in Genome Biology201718:9]. Genome Biol . 2013; 14( 4): R31. Google Scholar CrossRef Search ADS PubMed  25. Akazome Y, Kanda S, Oka Y. Expression of vesicular glutamate transporter-2.1 in medaka terminal nerve gonadotrophin-releasing hormone neurones. J Neuroendocrinol . 2011; 23( 7): 570– 576. Google Scholar CrossRef Search ADS PubMed  26. Takahashi A, Islam MS, Abe H, Okubo K, Akazome Y, Kaneko T, Hioki H, Oka Y. Morphological analysis of the early development of telencephalic and diencephalic gonadotropin-releasing hormone neuronal systems in enhanced green fluorescent protein-expressing transgenic medaka lines. J Comp Neurol . 2016; 524( 4): 896– 913. Google Scholar CrossRef Search ADS PubMed  27. Adachi S, Yamada S, Takatsu Y, Matsui H, Kinoshita M, Takase K, Sugiura H, Ohtaki T, Matsumoto H, Uenoyama Y, Tsukamura H, Inoue K, Maeda K. Involvement of anteroventral periventricular metastin/kisspeptin neurons in estrogen positive feedback action on luteinizing hormone release in female rats. J Reprod Dev . 2007; 53( 2): 367– 378. Google Scholar CrossRef Search ADS PubMed  28. Kitahashi T, Ogawa S, Parhar IS. Cloning and expression of kiss2 in the zebrafish and medaka. Endocrinology . 2009; 150( 2): 821– 831. Google Scholar CrossRef Search ADS PubMed  29. Koyama Y, Satou M, Oka Y, Ueda K. Involvement of the telencephalic hemispheres and the preoptic area in sexual behavior of the male goldfish, Carassius auratus: a brain-lesion study. Behav Neural Biol . 1984; 40( 1): 70– 86. Google Scholar CrossRef Search ADS PubMed  30. Satou M, Oka Y, Kusunoki M, Matsushima T, Kato M, Fujita I, Ueda K. Telencephalic and preoptic areas integrate sexual behavior in hime salmon (landlocked red salmon, Oncorhynchus nerka): results of electrical brain stimulation experiments. Physiol Behav . 1984; 33( 3): 441– 447. Google Scholar CrossRef Search ADS PubMed  31. Yang CF, Chiang MC, Gray DC, Prabhakaran M, Alvarado M, Juntti SA, Unger EK, Wells JA, Shah NM. Sexually dimorphic neurons in the ventromedial hypothalamus govern mating in both sexes and aggression in males. Cell . 2013; 153( 4): 896– 909. Google Scholar CrossRef Search ADS PubMed  32. Laflamme N, Nappi RE, Drolet G, Labrie C, Rivest S. Expression and neuropeptidergic characterization of estrogen receptors (ERalpha and ERbeta) throughout the rat brain: anatomical evidence of distinct roles of each subtype. J Neurobiol . 1998; 36( 3): 357– 378. Google Scholar CrossRef Search ADS PubMed  33. Hrabovszky E, Shughrue PJ, Merchenthaler I, Hajszán T, Carpenter CD, Liposits Z, Petersen SL. Detection of estrogen receptor-beta messenger ribonucleic acid and 125I-estrogen binding sites in luteinizing hormone-releasing hormone neurons of the rat brain. Endocrinology . 2000; 141( 9): 3506– 3509. Google Scholar CrossRef Search ADS PubMed  34. Hiraki T, Takeuchi A, Tsumaki T, Zempo B, Kanda S, Oka Y, Nagahama Y, Okubo K. Female-specific target sites for both oestrogen and androgen in the teleost brain. Proc Biol Sci . 2012; 279( 1749): 5014– 5023. Google Scholar CrossRef Search ADS PubMed  35. Arai Y, Kanda S, Karigo T, Hasebe M, Oka Y. Changes in GnRH sensitivity of LH cells play a key role in seasonal regulation of reproduction. In: Abstract of the Annual Meeting of the Society for Neuroscience 2015; 2015; Chicago, IL. Program number 431.404 (poster number W425). 36. Constantin S, Jasoni CL, Wadas B, Herbison AE. Gamma-aminobutyric acid and glutamate differentially regulate intracellular calcium concentrations in mouse gonadotropin-releasing hormone neurons. Endocrinology . 2010; 151( 1): 262– 270. Google Scholar CrossRef Search ADS PubMed  37. Christian CA, Pielecka-Fortuna J, Moenter SM. Estradiol suppresses glutamatergic transmission to gonadotropin-releasing hormone neurons in a model of negative feedback in mice. Biol Reprod . 2009; 80( 6): 1128– 1135. Google Scholar CrossRef Search ADS PubMed  38. Christian CA, Moenter SM. Estradiol induces diurnal shifts in GABA transmission to gonadotropin-releasing hormone neurons to provide a neural signal for ovulation. J Neurosci . 2007; 27( 8): 1913– 1921. Google Scholar CrossRef Search ADS PubMed  39. Cheong RY, Czieselsky K, Porteous R, Herbison AE. Expression of ESR1 in glutamatergic and GABAergic neurons is essential for normal puberty onset, estrogen feedback, and fertility in female mice. J Neurosci . 2015; 35( 43): 14533– 14543. Google Scholar CrossRef Search ADS PubMed  40. Moore AM, Prescott M, Marshall CJ, Yip SH, Campbell RE. Enhancement of a robust arcuate GABAergic input to gonadotropin-releasing hormone neurons in a model of polycystic ovarian syndrome. Proc Natl Acad Sci USA . 2015; 112( 2): 596– 601. Google Scholar CrossRef Search ADS PubMed  41. Yamada S, Uenoyama Y, Kinoshita M, Iwata K, Takase K, Matsui H, Adachi S, Inoue K, Maeda KI, Tsukamura H. Inhibition of metastin (kisspeptin-54)-GPR54 signaling in the arcuate nucleus-median eminence region during lactation in rats. Endocrinology . 2007; 148( 5): 2226– 2232. Google Scholar CrossRef Search ADS PubMed  42. Christian CA, Mobley JL, Moenter SM. Diurnal and estradiol-dependent changes in gonadotropin-releasing hormone neuron firing activity. Proc Natl Acad Sci USA . 2005; 102( 43): 15682– 15687. Google Scholar CrossRef Search ADS PubMed  43. Gréco B, Allegretto EA, Tetel MJ, Blaustein JD. Coexpression of ER beta with ER alpha and progestin receptor proteins in the female rat forebrain: effects of estradiol treatment. Endocrinology . 2001; 142( 12): 5172– 5181. Google Scholar CrossRef Search ADS PubMed  44. Blaustein JD, Turcotte JC. Estradiol-induced progestin receptor immunoreactivity is found only in estrogen receptor-immunoreactive cells in guinea pig brain. Neuroendocrinology . 1989; 49( 5): 454– 461. Google Scholar CrossRef Search ADS PubMed  45. O’Connell LA, Ding JH, Hofmann HA. Sex differences and similarities in the neuroendocrine regulation of social behavior in an African cichlid fish. Horm Behav . 2013; 64( 3): 468– 476. Google Scholar CrossRef Search ADS PubMed  46. Lee H, Kim DW, Remedios R, Anthony TE, Chang A, Madisen L, Zeng H, Anderson DJ. Scalable control of mounting and attack by Esr1+ neurons in the ventromedial hypothalamus. Nature . 2014; 509( 7502): 627– 632. Google Scholar CrossRef Search ADS PubMed  Copyright © 2018 Endocrine Society http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Endocrinology Oxford University Press

Morphological Analysis of the Axonal Projections of EGFP-Labeled Esr1-Expressing Neurons in Transgenic Female Medaka

Download PDF
14 pages

Loading next page...
1
 
/lp/ou_press/morphological-analysis-of-the-axonal-projections-of-egfp-labeled-esr1-1TD28XcU54

References (84)

Publisher
Oxford University Press
Copyright
Copyright © 2018 Endocrine Society
ISSN
0013-7227
eISSN
1945-7170
DOI
10.1210/en.2017-00873
pmid
29300923
Publisher site
See Article on Publisher Site

Abstract

Abstract Some hypothalamic neurons expressing estrogen receptor α (Esr1) are thought to transmit a gonadal estrogen feedback signal to gonadotropin-releasing hormone 1 (GnRH1) neurons, which is the final common pathway for feedback regulation of reproductive functions. Moreover, estrogen-sensitive neurons are suggested to control sexual behaviors in coordination with reproduction. In mammals, hypothalamic estrogen-sensitive neurons release the peptide kisspeptin and regulate GnRH1 neurons. However, a growing body of evidence in nonmammalian species casts doubt on the regulation of GnRH1 neurons by kisspeptin neurons. As a step toward understanding how estrogen regulates neuronal circuits for reproduction and sex behavior in vertebrates in general, we generated a transgenic (Tg) medaka that expresses enhanced green fluorescent protein (EGFP) specifically in esr1-expressing neurons (esr1 neurons) and analyzed their axonal projections. We found that esr1 neurons in the preoptic area (POA) project to the gnrh1 neurons. We also demonstrated by transcriptome and histological analyses that these esr1 neurons are glutamatergic or γ-aminobutyric acidergic (GABAergic) but not kisspeptinergic. We therefore suggest that glutamatergic and GABAergic esr1 neurons in the POA regulate gnrh1 neurons. This hypothesis is consistent with previous studies in mice that found that glutamatergic and GABAergic transmission is critical for estrogen-dependent changes in GnRH1 neuron firing. Thus, we propose that this neuronal circuit may provide an evolutionarily conserved mechanism for regulation of reproduction. In addition, we showed that telencephalic esr1 neurons project to medulla, which may control sexual behavior. Moreover, we found that some POA-esr1 neurons coexpress progesterone receptors. These neurons may form the neuronal circuits that regulate reproduction and sex behavior in response to the serum estrogen/progesterone. Reproduction in vertebrates is regulated by the hypothalamo-pituitary-gonadal (HPG) axis in a well-coordinated manner. Here, gonadotropin-releasing hormone (GnRH), which is now often called gonadotropin-releasing hormone 1 (GnRH1) for evolutionary reasons (1), acts on the pituitary to facilitate the release of gonadotropins, luteinizing hormone (LH), and follicle-stimulating hormone (FSH). LH and FSH, in turn, stimulate gonads and induce their maturation (2, 3). The mature gonads release sex steroids into the circulation. Then, the sex steroids, especially estrogen, stimulate or inhibit the hypothalamus to upregulate or downregulate release of GnRH1. This entire process is called steroid feedback. Recently, the importance of each component of HPG axis regulation in teleosts (GnRH1, LH, and FSH) has been scrutinized by the use of gene knockout techniques in zebrafish and medaka (4, 5), suggesting the importance and the evolutionary conservation of basic HPG axis regulation mechanisms in vertebrates, although there are some important differences between mammals and teleosts. In HPG axis regulation, gonadal estrogen may be considered one of the most important signals that transmit the reproductive status of the gonad to the central nervous system, and its mechanisms of action have been analyzed intensively. It has been generally suggested that there are two types of estrogen receptors (Esrs): Esr1 (or ERα, coded by esr1) and Esr2 (or ERβ, coded by esr2). Furthermore, data from knockout mice suggest that Esr1 but not Esr2 is essential for reproduction (6, 7). A growing body of evidence in mammals suggests that Esr1 is expressed in kisspeptin neurons rather than in GnRH1 neurons (8–10) and that kisspeptin neurons regulate the firing activity of GnRH1 neurons and GnRH1 release in the median eminence in response to serum estrogen levels. Although the expression of esr1 in kisspeptin neurons has been reported to be widely conserved in vertebrates including teleosts (11, 12), accumulating evidence argues against the presence of robust regulation of GnRH1 neurons by kisspeptin neurons and hence their role in regulation of reproduction in nonmammalian vertebrates. For example, GnRH1 neurons in teleosts do not express kisspeptin receptors (neither GPR54 paralogs gpr54-1 and gpr54-2) (13–15), and teleosts with kiss1 and kiss2 double knockout show normal fertility (16, 17). Furthermore, it is already known that avian species completely lack kisspeptin genes (18). Therefore, presumably unidentified estrogen-sensitive neurons other than kisspeptin neurons play important roles in the HPG axis regulation in nonmammalian vertebrates. On the other hand, it is also well known that sex steroids modulate specific behaviors depending on reproductive conditions. For example, most vertebrates show species-specific courtship behavior only after sexual maturation and during the breeding season. Thus, hormonal signals from the gonads have been suggested to control neural circuits for the modulation of sexual behaviors (19, 20). However, the nature of such estrogen-sensitive neurons remains to be elucidated. With these unanswered questions in mind, we aimed to anatomically identify the location and axonal projections of Esr1-expressing neurons mediating estrogen feedback signals to GnRH1 neurons, as well as those mediating sex steroid–sensitive signals to circuits that modulate sex behaviors. Here, we generated transgenic (Tg) animals in which esr1-expressing neurons are labeled with enhanced green fluorescent protein (EGFP) in the live brain for later detailed anatomical and physiological analyses. We chose medaka because of the following experimental advantages. First, molecular genetic tools can be applied easily. Second, female medaka spawn regularly every day under long-day conditions, and their breeding and nonbreeding states can be experimentally controlled by changing the day length. In addition, we have already described the anatomical distribution of esr messenger RNA (mRNA) expression in the medaka brain by using in situ hybridization (21). In the current study, we further detailed the axonal projections of esr1-expressing neurons (esr1 neurons) by using immunohistochemical methods to detect EGFP. Because Esrs are nuclear receptors, it is not possible to visualize projections of Esr1 neurons by using antibody directed against the Esr peptide. Therefore, we generated a Tg medaka line that expresses EGFP specifically in esr1 neurons. In these animals, EGFP that is expressed under the regulation of the esr1 promoter passively diffuses in the entire cytoplasm, enabling visualization of the entire neuronal morphology of esr1 neurons, including axons and dendrites. Materials and Methods Animals All the experiments including the generation of Tg line were conducted with d-rR strain medaka (Oryzias latipes), a teleost fish. For each experiment, sexually mature, gonadally intact female wild-type or Tg animals were used. Medaka were maintained under a 14-hour light/10-hour dark photoperiod at 27°C. The fish were fed twice daily with live brine shrimp and flake food. All experiments were conducted in accordance with the protocols approved by the Animal Care and Use Committee of the University of Tokyo (permission number 15–3). Generation of Tg medaka We generated constructs for esr1:enhanced green fluorescent protein (EGFP) Tg medaka by using a double-promoter approach for efficient screening. Using the zebrafish cardiac myosin light chain 2 (cmlc2) promoter, we visualized the heart for screening transgene-positive embryos. We carried out polymerase chain reaction (PCR) by using PrimeSTAR polymerase (Takara, Shiga, Japan) to amplify a medaka genomic DNA fragment containing the 5′-flanking region (3.7 kb) of exon 2 of the esr1 gene from a bacterial artificial chromosome clone (clone number Golwb109_F22) from a medaka strain, HdrR, by using a forward/reverse primer pair, Pesr1-F (5′-ACAGGATGGAGGTCAAAAGC-3′)/Pesr1-R (GACCCCCTCGGTGACATGTATCCACCGGTCGCCACCATGG-3′); Pesr1-F (5′-CAGAACTTCCTTGCTCATGCTCACC-3′)/Pesr1-R LINKEGFPhind (5′-GAGAAGCTTCAGAGCCCTTCCCCTGTGCTCAGGC-3′). For gene IDs, see Table 1. Then, we fused this amplicon with an EGFP open reading frame by overlap extension PCR. Downstream of the coding sequence of EGFP, we fused the cmlc2 promoter region and I-SceI restriction enzyme site. A fragment of ~6 kb was then cloned into the TOPO-XL cloning vector (Invitrogen, Carlsbad, CA). For microinjection and screening, we followed a protocol that has been described previously (22). After generation of the Tg line, we confirmed the specificity of EGFP expression in esr1 neurons in brain sections by double labeling with esr1 in situ hybridization and EGFP immunohistochemistry. The brains of fish from the esr1:EGFP Tg line were fixed with 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS). The fixed brains were frontally cryosectioned at 20 μm with a cryostat (CM 3050S; Leica Microsystems, Wetzlar, Germany) and mounted onto MAS-GP type A coated glass slides (Matsunami, Osaka, Japan). Sections were incubated with anti-EGFP antibody raised in rabbit (generous gift from Drs. Kaneko and Hioki, Kyoto University, Kyoto, Japan; see also Table 2) diluted 1:1000 with phosphate-buffered saline containing 0.3% Tween 20 (PBST) overnight, rinsed twice with PBST, and incubated with biotinylated anti-rabbit immunoglobulin G (Invitrogen) (diluted 1:200 with PBST) for 2 hours. Then, sections were fixed by 4% PFA in PBS for 15 minutes and rinsed with PBS containing 0.2% glycine. Next, we performed in situ hybridization with these slides to detect esr1, estrogen receptor β1 (esr2a), or estrogen receptor β2 (esr2b) mRNA by using esr1-,esr2a-, or esr2b-specific digoxigenin (DIG)-labeled probes that were described in a previous report (11, 21), following a protocol described previously (21). Briefly, slides were incubated with 0.25% acetic anhydride in 0.1 M triethanolamine for 10 minutes. Sections were then washed with PBS and prehybridized at 58°C for ≥30 minutes in a hybridization buffer containing 50% formamide, 3× saline sodium citrate (SSC), 0.12 M phosphate buffer (pH 7.4), 1× Denhardt solution (Sigma, St. Louis, MO), 125 μg/mL transfer RNA, 0.1 mg/mL calf thymus DNA (Invitrogen), and 10% dextran sulfate (Sigma). Slides were incubated at 58°C overnight in the same solution containing 100 ng/mL denatured riboprobe. After hybridization, the sections were washed twice with 50% formamide and 2× SSC for 15 minutes each at 58°C. Then, the sections were immersed in 10 mM Tris-HCl, pH 7.5; 500 mM NaCl; and 1 mM EDTA, pH 8.0 (TNE) for 10 minutes at 37°C. The sections were incubated with 20 μg/mL ribonuclease A (Sigma) in TNE for 30 minutes at 37°C and then washed with TNE for 10 minutes at 37°C to remove the ribonuclease A. They were then washed with 2× SSC twice, followed by 0.5× SSC twice for 15 minutes each at 58°C. The slides were immersed in DIG-1 (0.1 M Tris-HCl, pH 7.5; 0.16 M NaCl; and 0.1% Tween 20) for 5 minutes, followed by 1.5% blocking reagent with DIG-1 for 30 minutes and DIG-1 for 15 minutes, and then incubated with avidin-biotin complex (ABC) reagents (1% A solution and 1% B solution in DIG-1 buffer; Vector Laboratories, Burlingame, CA) for 1 hour. Sections were rinsed twice with DIG-1 buffer, incubated with Alexa Fluor 488 conjugated streptavidin (diluted 1:500 with DIG-1 buffer; Invitrogen) and alkaline phosphatase–conjugated anti-DIG antibody (diluted 1:1000 with DIG-1 buffer; Roche, Molecular Biochemicals GmbH, Mannheim, Germany) for 2 hours. Sections were then rinsed twice with DIG-1 buffer. After detection of EGFP signals, alkaline phosphatase activity, which was used to label mRNA, was detected with a Fast-Red substrate kit (Roche) according to the manufacturer’s instructions. Incubation for this substrate was carried out until visible signals were detected and was stopped by washing in PBS containing 0.5 mM EDTA. Then, the sections were coverslipped with CC/Mount (Diagnostic BioSystems, Pleasanton, CA). Fluorescence was observed with a confocal laser-scanning microscope (LSM-710; Carl Zeiss, Oberkochen, Germany). We calculated the percentage of EGFP-labeled esr1-expressing cells by counting the number of both EGFP-positive and EGFP-negative esr1-expressing cells. Numbers of cells were expressed as mean ± standard deviation (SD). Table 1. List of Gene IDs Gene  Ensembl Gene ID  vglut1.1  ENSORLG00000011296  vglut1.2  ENSORLG00000007982  vglut2.1  ENSORLG00000006157  vglut2.2  ENSORLG00000005791  vglut3  ENSORLG00000011203  gad1.1  ENSORLG00000009208  gad1.2  ENSORLG00000017268  gad2  ENSORLG00000012966  esr1  ENSORLG00000014514  esr2a  ENSORLG00000017721  esr2b  ENSORLG00000018012  pr  ENSORLG00000002651  Gene  Ensembl Gene ID  vglut1.1  ENSORLG00000011296  vglut1.2  ENSORLG00000007982  vglut2.1  ENSORLG00000006157  vglut2.2  ENSORLG00000005791  vglut3  ENSORLG00000011203  gad1.1  ENSORLG00000009208  gad1.2  ENSORLG00000017268  gad2  ENSORLG00000012966  esr1  ENSORLG00000014514  esr2a  ENSORLG00000017721  esr2b  ENSORLG00000018012  pr  ENSORLG00000002651  View Large Table 2. Antibody Table Peptide/Protein Target  Name of Antibody  Names of Individuals Providing the Antibody  Species Raised in; Monoclonal or Polyclonal  Dilution Used  RRID  EGFP  Anti-EGFP antibody  Drs. Kaneko and Hioki (Kyoto University)  Rabbit; polyclonal  1:1000  AB_2716624  Peptide/Protein Target  Name of Antibody  Names of Individuals Providing the Antibody  Species Raised in; Monoclonal or Polyclonal  Dilution Used  RRID  EGFP  Anti-EGFP antibody  Drs. Kaneko and Hioki (Kyoto University)  Rabbit; polyclonal  1:1000  AB_2716624  Abbreviation: RRID, Research Resource Identifier. View Large RNA sequencing Brains of sexually mature female esr1:EGFP Tg medaka were dissected out under an upright fluorescent microscope (Eclipse E600FN; Nikon, Tokyo, Japan) via a method described in our previous study (23). Brains were kept in a chamber filled with artificial cerebrospinal fluid, and a constant flow (1 mL/min) of artificial cerebrospinal fluid was applied 15 minutes before and during cell harvesting via a peristaltic pump to clean debris from dead or unhealthy cells. Patch pipettes of borosilicate glass capillaries with an outer diameter of 1.5 mm (GD-1.5; Narishige, Tokyo, Japan) were pulled to produce a 3- to 5-µm tip with a micropipette puller (P-97; Sutter Instruments, Novato, CA). The tip of the capillary was backfilled by capillarity with nuclease-free water (<1 μL). Slight positive pressure was applied to patch pipettes while they approached the targeted EGFP-positive neurons, and then EGFP-positive neurons were carefully pulled into the capillary by negative pressure. The contents of the pipette were expelled into a 0.2-μL tube containing 10 μL Buffer RLT (Qiagen, Hilden, Germany) with 1% 2-mercaptoethanol. Using this approach, we prepared five samples consisting of five neurons each from five different fish. Samples were frozen at −80°C immediately after harvesting. Total RNA was purified with Ampure XP RNA (Beckman Coulter, Brea, CA). For reverse transcription, whole-transcript amplification, and RNA sequencing (RNA-seq) library preparation, we followed the Quartz-seq protocol (24). We performed RNA-seq with HiSeq1500 and the HiSeq Rapid SBS Kit version 1 (Illumina, San Diego, CA), following manufacturer’s instructions. For data analysis, we used the medaka genome (ftp://ftp.ensembl.org/pub/release-79/gtf/oryzias_latipes). At least 13 million sequences of 50-base single reads were mapped onto the genome for each sample. Calculations of read mapping and reads per kilobase of transcript per million mapped reads (RPKM) were performed in CLCbio Genomics workbench software (CLC Bio, Aarhus, Denmark). Double labeling for EGFP and mRNA of gnrh1 or progesterone receptor We performed double labeling for EGFP and mRNA of gnrh1 or progesterone receptor (pr) as described earlier. To detect pr mRNA, we prepared a specific DIG-labeled riboprobe. Nested primers used to amplify medaka-pr-complementary DNAs were as follows: first round, pr-F1 5′-ATGGAGAGTAAAATGAACGGAAAGCTGG-3′ and pr-R1 5′-GTCATAGCCGGAGTACACGGTCT-3′; second round, pr-F2 5′-CCGAGTCCAGAGTAAATGGCTTGATCGA-3′ and pr-R2 5′-CGGCTCGATGTTCTCCAGAATGT-3′. The template used to make the pr probe was 1.1 kb in length (DNA Data Bank of Japan/European Molecular Biology Laboratory/GenBank accession no. AB360545.1). The probe for pr was synthesized from the medaka brain with a labeling kit (Roche). We performed in situ hybridization by using a gnrh1-specific DIG-labeled probe that was generated previously (23). We calculated the percentage of pr-expressing cells that were labeled with EGFP by counting the number of both EGFP-positive and EGFP-negative pr-expressing cells. We also calculated the percentage of gnrh1 neurons that received projections from EGFP fibers by counting the number of cell bodies of gnrh1 neurons that were surrounded by EGFP-labeled fibers within 1 μm. For both calculations, number of cells was expressed as mean ± SD. Immunohistochemistry for tract tracing We carried out immunohistochemistry for EGFP to analyze axonal projections of esr1 neurons in detail. All the procedures were performed as described earlier up to the secondary antibody, which was a biotinylated anti-rabbit immunoglobulin G. Sections were washed twice with PBST and then incubated with ABC reagents (1% A solution and 1% B solution in PBST buffer; Vector) for 1 hour. Then, the slides were washed twice with PBST and reacted with 3, 3′-diaminobenzidine. For some slides, we performed Nissl counterstaining by using a 0.1% cresyl violet solution. The slides were then dehydrated, cleared, and coverslipped. Photographs were taken with a digital camera (DFC310FX; Leica Microsystems) on a Leica DM5000B microscope (Leica Microsystems). Double-label in situ hybridization for esr1 mRNA and vglut or gad mRNA We performed a double-label in situ hybridization by using a mixture of DIG-labeled esr1 and fluorescein-labeled vesicular glutamate transporter (vglut) 2.1, glutamic acid decarboxylase (gad)1.1, gad1.2, and gad2 probes that were synthesized as described in a previous report (25). Medaka complementary DNAs were PCR amplified with forward/reverse primer pairs: vglut2.1-F 5′-GAGATCAACCTGCGCTCACCACA-3′/Vglut2.1-R 5′-TGAATACTGAACCAGGATCCCAG-3′ gad1.1-F 5′-GAGGCTGTGACTCATGCGTG-3′/gad1.1-R 5′-CCTTCTTTATGGAATAGTGGC-3′ gad1.2-F 5′-GCCAGATCCACGCTGGTGGAC-3′/gad1.2-R 5′-CATTAGCACAAAAACTGGAG-3′ gad2-F 5′-AAACAGCCCATCCCAGGTAC-3′/gad2-R 5′-AGCAGCGGGATTTGAGATGAC-3′ These amplified fragments were used to generate probes with fluorescein RNA labeling mixtures (Roche) according to the manufacturer’s protocol. After brains were fixed by transcardial perfusion, all procedures up to the blocking step were performed as described earlier. The sections were then incubated with a horseradish peroxidase conjugated antifluorescein antibody (diluted 1:500 with DIG-1; PerkinElmer, Foster City, CA) for ≥1 hour. Sections were washed twice with DIG-1 for 10 minutes, incubated for 30 minutes with the Biotinyl Tyramide Amplification Kit (NEL700A; PerkinElmer) diluted 1:50 in dilution buffer (PerkinElmer), washed with DIG-1 twice for 10 minutes each, and incubated with ABC reagents (Vector) for 1 hour. Sections were washed twice with DIG-1 for 10 minutes and then incubated with Alexa Fluor 488 conjugated streptavidin (diluted 1:500 with DIG-1) and alkaline phosphatase conjugated anti-DIG antibody (diluted 1:1000 with DIG-1) for 2 hours. Then, the sections were washed twice with DIG-1 for 10 minutes. After detection of positive signals, alkaline phosphatase activity was detected with a Fast-Red substrate kit (Roche) according to the manufacturer’s instructions. The incubation for this substrate was carried out until visible signals were detected and was stopped by washing in PBS containing 0.5 mM EDTA. Sections were then coverslipped with CC/Mount (Diagnostic BioSystem). Fluorescent signals were observed and documented with a LSM-710 confocal laser-scanning microscope (Carl Zeiss). We calculated the percentage of coexpression between esr1-expressing cells and vglut- or gad-expressing cells by counting the number of esr1-expressing cells and both esr1- and vglut- or gad-positive cells. Numbers of cells were expressed as mean ± SD. Results Histological analysis of the expression of esr1 in gnrh1 neurons We examined the expression of esr1 in hypophysiotropic gnrh1 neurons in the preoptic area (POA) by using in situ hybridization of adjacent sections [Fig. 1(a)–1(d)]. We found that esr1 mRNA was localized in the medial POA [Fig. 1(c)]. On the other hand, gnrh1 mRNA were localized in the lateral POA [Fig. 1(d)]. Therefore, to determine whether gnrh1 neurons express esr1, we performed double in situ hybridization for esr1 and gnrh1 mRNA [Fig. 1(e) and 1(f)]. We found 516 ± 10 esr1-expressing cells and 32 ± 4 gnrh1-expressing cells in the POA (means ± SD, n = 2 fish). However, we did not find any coexpression of gnrh1 and esr1 mRNA [Fig. 1(f)]. These results indicate that esr1 is not expressed in gnrh1 neurons. Figure 1. View largeDownload slide gnrh1 neurons do not express esr1 in medaka POA. (a, b) Schematic illustrations of (a) lateral (left is rostral) and (b) frontal sections of the medaka brain showing the plane of section corresponding to the photographs in (c) and (d). The boxed area in the POA is shown in (c) and (d). (c, d) Light photomicrographs showing in situ hybridization for (c) esr1 mRNA and (d) gnrh1 mRNA in adjacent sections. These photographs show that gnrh1 neurons are localized in the region lateral to esr1 neurons in POA. (e) Schematic illustrations of a frontal section of the medaka brain, showing the plane of section corresponding to the photographs in (f). The boxed area in (e) is shown in (f). (f) Photographs showing double labeling for esr1 mRNA (magenta) and gnrh1 mRNA (green), which demonstrate that POA gnrh1 neurons do not coexpress esr1. ca, anterior commissure; dDm, dorsal region of the medial part of dorsal telencephalic area; DI, lateral part of dorsal telencephalic area. Scale bars: 25 μm. Figure 1. View largeDownload slide gnrh1 neurons do not express esr1 in medaka POA. (a, b) Schematic illustrations of (a) lateral (left is rostral) and (b) frontal sections of the medaka brain showing the plane of section corresponding to the photographs in (c) and (d). The boxed area in the POA is shown in (c) and (d). (c, d) Light photomicrographs showing in situ hybridization for (c) esr1 mRNA and (d) gnrh1 mRNA in adjacent sections. These photographs show that gnrh1 neurons are localized in the region lateral to esr1 neurons in POA. (e) Schematic illustrations of a frontal section of the medaka brain, showing the plane of section corresponding to the photographs in (f). The boxed area in (e) is shown in (f). (f) Photographs showing double labeling for esr1 mRNA (magenta) and gnrh1 mRNA (green), which demonstrate that POA gnrh1 neurons do not coexpress esr1. ca, anterior commissure; dDm, dorsal region of the medial part of dorsal telencephalic area; DI, lateral part of dorsal telencephalic area. Scale bars: 25 μm. Establishment of esr1:EGFP Tg line We generated a Tg medaka line that expresses EGFP specifically in esr1 neurons [Fig. 2(a)–2(c)]. We found that the EGFP-positive cells were distributed mainly in the supracommissural part of ventral telencephalic area (Vs), POA, and ventral tuberal nucleus (NVT) region, all of which have previously been reported to contain esr-expressing neurons in medaka (21). To examine specificity, we analyzed the results of double labeling with EGFP immunohistochemistry and esr1 mRNA in situ hybridization [Fig. 2(b) and 2(c)]. We found that 44% of POA-esr1 neurons (201 ± 23/460 ± 51 cells, n = 3 fish) and about 70% of Vs-esr1 neurons (n = 2) and NVT-esr1 neurons (n = 2) were labeled with anti-EGFP antibody. Although some EGFP-negative esr1 cells were observed in these areas, all the EGFP positive cells expressed esr1, at least in these areas. Thus, the specificity of EGFP labeling is satisfactory for performing anatomical analysis of the axonal projections of these neurons. Figure 2. View largeDownload slide Establishment of esr1:EGFP Tg medaka. (a) The construct used to generate the esr1:EGFP Tg medaka. The EGFP-coding sequence was fused to the 3.7-kb DNA fragment containing the 5′-flanking region of exon1, intron1, and part of exon2, upstream of the first methionine of the esr1 gene. This construct contains the cardiac myosin light chain 2 (cmlc2) promoter region to express EGFP in the heart for screening. (b) Schematic illustrations of lateral (left is rostral) and frontal sections of the medaka brain, showing the plane of section corresponding to the panels in (c)–(e). The boxed area indicates the POA population of esr1 neurons shown in (c); note that the midline is located in the center of the pictures in (c). The red box indicates the POA population of esr2a and esr2b neurons shown in (d) and (e). (c) Photographs showing double labeling for EGFP immunohistochemistry (green) and esr1 mRNA (magenta) in the POA, which demonstrates that EGFP specifically labeled esr1 neurons. In addition to the POA, we confirmed specificity of EGFP labeling in the Vs and NVT, which also express esr1 (data not shown). (d) Photographs showing double labeling for EGFP immunohistochemistry (green) and esr2a mRNA (magenta) in the POA, which demonstrates that EGFP labeled esr2a neurons (arrowhead). Approximately 8% of esr2a neurons were labeled by EGFP (9 ± 3/115 ± 33 cells, n = 2 fish). (e) Photographs showing double labeling for EGFP immunohistochemistry (green) and esr2b mRNA (magenta) in the POA, which demonstrates that EGFP labeled esr2b neurons (arrowhead). Approximately 10% of esr2b neurons were labeled by EGFP (41 ± 11/424 ± 57 cells, n = 2 fish). ca, anterior commissure; dDm, dorsal region of the medial part of dorsal telencephalic area; DI, lateral part of dorsal telencephalic area. Scale bars: 10 μm. Figure 2. View largeDownload slide Establishment of esr1:EGFP Tg medaka. (a) The construct used to generate the esr1:EGFP Tg medaka. The EGFP-coding sequence was fused to the 3.7-kb DNA fragment containing the 5′-flanking region of exon1, intron1, and part of exon2, upstream of the first methionine of the esr1 gene. This construct contains the cardiac myosin light chain 2 (cmlc2) promoter region to express EGFP in the heart for screening. (b) Schematic illustrations of lateral (left is rostral) and frontal sections of the medaka brain, showing the plane of section corresponding to the panels in (c)–(e). The boxed area indicates the POA population of esr1 neurons shown in (c); note that the midline is located in the center of the pictures in (c). The red box indicates the POA population of esr2a and esr2b neurons shown in (d) and (e). (c) Photographs showing double labeling for EGFP immunohistochemistry (green) and esr1 mRNA (magenta) in the POA, which demonstrates that EGFP specifically labeled esr1 neurons. In addition to the POA, we confirmed specificity of EGFP labeling in the Vs and NVT, which also express esr1 (data not shown). (d) Photographs showing double labeling for EGFP immunohistochemistry (green) and esr2a mRNA (magenta) in the POA, which demonstrates that EGFP labeled esr2a neurons (arrowhead). Approximately 8% of esr2a neurons were labeled by EGFP (9 ± 3/115 ± 33 cells, n = 2 fish). (e) Photographs showing double labeling for EGFP immunohistochemistry (green) and esr2b mRNA (magenta) in the POA, which demonstrates that EGFP labeled esr2b neurons (arrowhead). Approximately 10% of esr2b neurons were labeled by EGFP (41 ± 11/424 ± 57 cells, n = 2 fish). ca, anterior commissure; dDm, dorsal region of the medial part of dorsal telencephalic area; DI, lateral part of dorsal telencephalic area. Scale bars: 10 μm. In addition, we analyzed the results of double labeling analysis with EGFP immunohistochemistry and esr2a/esr2b mRNA in situ hybridization [Fig. 2(d) and 2(e)], because we previously suggested that esr1, esr2a (or erβ1), and esr2b (or erβ2) were widely distributed in POA and that their distribution patterns in the POA were similar to each other. We found that EGFP also labeled esr2a and esr2b in this Tg line [Fig. 2(d) and 2(e)] and that 8% of POA-esr2a neurons (9 ± 3/115 ± 33 cells, n = 2 fish) [Fig. 2(d)] and ~10% of POA-esr2b neurons (41 ± 11/424 ± 57 cells, n = 2 fish) [Fig. 2(e)] were labeled with anti-EGFP antibody (Table 3). Table 3. Percentage of EGFP-Labeled esrs-Expressing Cells   Average Number of EGFP-Labeled Cells/esr Neurons in POA  % of EGFP-Labeled Cells  esr1  201 ± 23/460 ± 51 (n = 3 fish)  44  esr2a  9 ± 3/115 ± 33 (n = 2 fish)  8  esr2b  41 ± 11/424 ± 57 (n = 2 fish)  10    Average Number of EGFP-Labeled Cells/esr Neurons in POA  % of EGFP-Labeled Cells  esr1  201 ± 23/460 ± 51 (n = 3 fish)  44  esr2a  9 ± 3/115 ± 33 (n = 2 fish)  8  esr2b  41 ± 11/424 ± 57 (n = 2 fish)  10  View Large Anatomical analysis of the axonal projections of the esr1 neurons We performed EGFP immunohistochemistry in the brain of esr1:EGFP Tg medaka to analyze axonal projections of esr1 neurons (Fig. 3). Immunohistochemistry for EGFP, which enhances EGFP signals, enabled precise morphological analysis of axonal projections. Most of the POA-esr1 neurons projected their axons to the lateral POA. These fibers formed a thick bundle and projected caudally [Fig. 3(b)]. Frontal sections show that fibers project caudally, passing through the ventrolateral region of the telencephalon and hypothalamus [Fig. 3(c)]. In the hypothalamus, axons of the POA-esr1 neurons coursed laterally to the anterior tuberal nucleus (NAT) and NVT. These bundles of fibers innervated the pituitary, and although these results do not exclude the possibility that some POA-esr1 neurons project to other regions, we may safely conclude that the majority of axons of the POA-esr1 neurons projected to the pituitary. Axons of esr1 neurons entered the rostral pituitary and projected widely to the rostral half of the pituitary, whereas only a few EGFP immunoreactive (ir) fibers were observed in the caudal region (data not shown). Figure 3. View largeDownload slide Light photomicrographs showing the (a, b) sagittal and (c) frontal sections of the esr1:EGFP neurons visualized with anti-EGFP antibody in the brain of female esr1:EGFP Tg medaka. Sections were counterstained with cresyl violet. (a) The boxed area in the schematic illustration (lateral view of the brain) is shown in (b). (b) Sagittal sections showing EGFP-ir cell bodies in POA (arrowhead) and fiber bundles (arrow). Each photograph shows different mediolateral levels of the brain. Most fibers of the POA-esr1 neurons project caudally and reach the ventral region of hypothalamus. (c) Frontal sections showing axons of the esr1 neurons. The left column contains schematic drawings of the boxed areas in the photographs of the right column. POA-esr1 neurons project their axons (arrows) through the lateral region of the POA and run caudally, passing the ventrolateral region of hypothalamus, and project to the pituitary. dDI, dorsal region of DI; DI, lateral part of dorsal telencephalic area; DM, dorsomedial thalamic nucleus; Dp, posterior part of dorsal telencephalic area; GR, corpus glomerulosus; nII, optic nerve; NPPv, periventricular posterior nucleus; POm, magnocellular preoptic nucleus; POp, parvocellular preoptic nucleus; TO, optic tectum; TS, torus semicircularis; VM, ventromedial thalamic nucleus; Vp, postcommissural part of V. Scale bars: 100 μm. Figure 3. View largeDownload slide Light photomicrographs showing the (a, b) sagittal and (c) frontal sections of the esr1:EGFP neurons visualized with anti-EGFP antibody in the brain of female esr1:EGFP Tg medaka. Sections were counterstained with cresyl violet. (a) The boxed area in the schematic illustration (lateral view of the brain) is shown in (b). (b) Sagittal sections showing EGFP-ir cell bodies in POA (arrowhead) and fiber bundles (arrow). Each photograph shows different mediolateral levels of the brain. Most fibers of the POA-esr1 neurons project caudally and reach the ventral region of hypothalamus. (c) Frontal sections showing axons of the esr1 neurons. The left column contains schematic drawings of the boxed areas in the photographs of the right column. POA-esr1 neurons project their axons (arrows) through the lateral region of the POA and run caudally, passing the ventrolateral region of hypothalamus, and project to the pituitary. dDI, dorsal region of DI; DI, lateral part of dorsal telencephalic area; DM, dorsomedial thalamic nucleus; Dp, posterior part of dorsal telencephalic area; GR, corpus glomerulosus; nII, optic nerve; NPPv, periventricular posterior nucleus; POm, magnocellular preoptic nucleus; POp, parvocellular preoptic nucleus; TO, optic tectum; TS, torus semicircularis; VM, ventromedial thalamic nucleus; Vp, postcommissural part of V. Scale bars: 100 μm. On the other hand, analysis of sagittal sections showed that Vs-esr1 neurons projected mainly to caudal regions of the brain [Fig. 4(a) and 4(b)]. Analysis of frontal sections showed that these EGFP-positive axons of the Vs-esr1 neurons projected caudally, passing near the midline region of the hypothalamus bilaterally (data not shown). These axons passed through the periventricular region of the hypothalamus, in the region lateral to the periventricular posterior nucleus, and coursed further caudally (data not shown). Finally, axons of Vs-esr1 neurons terminated in the region ventral to medial reticular formation [Fig. 4(c)]. In this Tg line, no EGFP-ir axons were observed projecting to the spinal cord. Figure 4. View largeDownload slide Light photomicrographs showing (a, b) sagittal and (c) frontal sections of esr1:EGFP neurons visualized with EGFP antibody in the brain of female esr1:EGFP Tg medaka. Sections were counterstained with cresyl violet. (a) Sagittal section (right) showing EGFP-ir cell bodies and axons. The boxed area in the schematic illustration (left, lateral view of the brain) is shown in the photographs in (b). Arrowheads indicate cell bodies in Vs region. The EGFP-ir neurons in Vs (Vs-EGFP-ir neurons) project their axons to the caudal region. (b) Sagittal sections more lateral to the photograph of (a), showing that axons from Vs region project caudally, which suggests that these fibers project to the medulla, passing through the diencephalon. Arrows indicate fibers that originate from Vs-esr1 neurons. (c) Frontal section showing fibers of the Vs EGFP-ir neurons in the ventral region of medulla at the level indicated in the lateral view of the brain (top left). The boxed area in the schematic illustration (top right) is shown in the photograph below. Arrowheads indicate EGFP-ir axons. EGFP-ir fibers were not observed in more caudal regions. nII, optic nerve; RFm, medial reticular formation; TE, telencephalon. Scale bars: (a, b) 100 μm, (c) 50 μm. Figure 4. View largeDownload slide Light photomicrographs showing (a, b) sagittal and (c) frontal sections of esr1:EGFP neurons visualized with EGFP antibody in the brain of female esr1:EGFP Tg medaka. Sections were counterstained with cresyl violet. (a) Sagittal section (right) showing EGFP-ir cell bodies and axons. The boxed area in the schematic illustration (left, lateral view of the brain) is shown in the photographs in (b). Arrowheads indicate cell bodies in Vs region. The EGFP-ir neurons in Vs (Vs-EGFP-ir neurons) project their axons to the caudal region. (b) Sagittal sections more lateral to the photograph of (a), showing that axons from Vs region project caudally, which suggests that these fibers project to the medulla, passing through the diencephalon. Arrows indicate fibers that originate from Vs-esr1 neurons. (c) Frontal section showing fibers of the Vs EGFP-ir neurons in the ventral region of medulla at the level indicated in the lateral view of the brain (top left). The boxed area in the schematic illustration (top right) is shown in the photograph below. Arrowheads indicate EGFP-ir axons. EGFP-ir fibers were not observed in more caudal regions. nII, optic nerve; RFm, medial reticular formation; TE, telencephalon. Scale bars: (a, b) 100 μm, (c) 50 μm. Morphological analysis of axonal projections of esr1 neurons to gnrh1 neurons To assess the participation of Esr1 neurons in mediating estrogen feedback to GnRH1 neurons, we analyzed the results of double labeling with EGFP immunohistochemistry and gnrh1 mRNA in situ hybridization (Fig. 5). EGFP-positive axons from POA-esr1 neurons surrounded gnrh1-expressing neuronal cell bodies localized in the ventrolateral POA. Seventy-seven percent of gnrh1 neurons (30 ± 6/39 ± 6 cells, n = 3 fish) appeared to receive axonal projections from POA-esr1 neurons. On the other hand, very few axons were observed around the dorsal and medial groups of gnrh1 neurons, which have been suggested not to project to the pituitary (22, 23, 26). Figure 5. View largeDownload slide Double labeling analysis with esr1:EGFP Tg medaka suggests that POA-esr1 neurons directly contact the cell bodies of the gnrh1 neurons. (a) Illustration of a frontal section of the medaka brain, showing the plane of section corresponding to the photographs in (b). The boxed area is shown in (b). (b) Photographs showing double labeling using EGFP immunohistochemistry (green) and gnrh1 mRNA in situ hybridization (magenta) demonstrate that the cell bodies of gnrh1 neurons are surrounded by EGFP-ir fibers. Taken together with the distribution of axons of esr1 neurons, POA-esr1 neurons are suggested to make direct contacts on gnrh1 neurons. dDI, dorsal region of DI; DI, lateral part of dorsal telencephalic area; Dp, posterior part of dorsal telencephalic area; nII, optic nerve; POm, magnocellular preoptic nucleus; POp, parvocellular preoptic nucleus; Vp, postcommissural part of V. Scale bars: 25 μm. Figure 5. View largeDownload slide Double labeling analysis with esr1:EGFP Tg medaka suggests that POA-esr1 neurons directly contact the cell bodies of the gnrh1 neurons. (a) Illustration of a frontal section of the medaka brain, showing the plane of section corresponding to the photographs in (b). The boxed area is shown in (b). (b) Photographs showing double labeling using EGFP immunohistochemistry (green) and gnrh1 mRNA in situ hybridization (magenta) demonstrate that the cell bodies of gnrh1 neurons are surrounded by EGFP-ir fibers. Taken together with the distribution of axons of esr1 neurons, POA-esr1 neurons are suggested to make direct contacts on gnrh1 neurons. dDI, dorsal region of DI; DI, lateral part of dorsal telencephalic area; Dp, posterior part of dorsal telencephalic area; nII, optic nerve; POm, magnocellular preoptic nucleus; POp, parvocellular preoptic nucleus; Vp, postcommissural part of V. Scale bars: 25 μm. Neurotransmitter candidates for the POA-esr1 neurons We next examined neurotransmitter candidates for POA-esr1 neurons. First, we collected EGFP-positive neurons from the POA of female esr1:EGFP Tg medaka and performed RNA-seq of those neurons. It should be noted that we found no trace of kiss1 or kiss2 gene expression, contrary to what we may expect from results in mammals (8, 27). These results strongly support the results of previous in situ hybridization studies demonstrating that neither kiss1 nor kiss2 neurons are localized in the POA in medaka (11, 12, 28). Therefore, we examined the expression of marker genes for classic transmitters, glutamate and γ-aminobutyric acid (GABA). As markers for glutamatergic and γ-aminobutyric acidergic (GABAergic) transmission, medaka has been reported to possess five subtypes of vglut (vglut1.1, vglut1.2, vglut2.1, vglut2.2, and vglut3) and three subtypes of gad (gad1.1, gad 1.2, and gad2), respectively (25). We evaluated the expression level of vglut and gad genes based on the average RPKM; in this approach, only vglut2.1,gad1.1, and gad2 were detected from the EGFP-positive neurons (RPKM for vglut2.1, 2.13 ± 3.9; gad1.1, 95.1 ± 86; gad2, 23.6 ± 14; others, RPKM < 1). This result suggests that POA-esr1 neurons include both glutamatergic and GABAergic neurons. Next, we analyzed colocalization of glutamatergic or GABAergic markers in the POA-esr1 neurons by double in situ hybridization for mRNA of esr1 and either vglut2.1, gad1.1, gad1.2, or gad2 (Fig. 6). These results are summarized in Table 4. More than 95% of POA-esr1 neurons coexpressed vglut 2.1 [(Fig. 6(b)] (326 ± 81/342 ± 71 cells, n = 3 fish). In addition, 36% of POA-esr1 neurons coexpressed gad 1.1 [(Fig. 6(c)] (148 ± 23/410 ± 42 cells, n = 2 fish). Because most esr1 neurons were shown here to be glutamatergic by double labeling of esr1 and vglut 2.1, we did not examine other subtypes of vglut. Among the GABA-synthesizing enzymes, gad1.1 showed the highest percentage of colocalization with esr1 mRNA (Fig. 6). However, the colocalization ratio for gad1.1 was much lower than that of vglut2.1, which suggests that glutamate is the main classic neurotransmitter in POA-esr1 neurons. In addition, RNA-seq results suggested that EGFP-positive POA neurons from esr1:EGFP medaka express not only esr1 (RPKM 188 ± 81) but also esr2a and esr2b (RPKM 28.2 ± 19 and 3.02 ± 4.8). These results are consistent with our histological results [Fig. 2(d) and 2(e)]. Moreover, among other sex steroid receptor genes, we found that POA-esr1 neurons highly express pr (RPKM 155 ± 113). Therefore, we analyzed double labeling with EGFP immunohistochemistry and pr mRNA in situ hybridization. We found that ~68% of EGFP-expressing neurons in POA coexpressed pr (167 ± 33/246 ± 33 cells, n = 3 fish) (Fig. 7). Figure 6. View largeDownload slide Double in situ hybridization for esr1 and marker genes for glutamate or GABA in POA. (a) Schematic illustration of a frontal section of the medaka brain, showing the plane of section corresponding to the photographs. The boxed area corresponds to the photographs in (b)–(e). (b) Photographs showing double labeling for vglut2.1 mRNA (green) and esr1 mRNA (magenta), demonstrating that POA-esr1–expressing cells express vglut2.1. These results suggest that the majority of the POA-esr1 neurons are glutamatergic. (c) Photographs showing double labeling for gad1.1 mRNA (green) and esr1 mRNA (magenta), which demonstrate that some POA-esr1–expressing cells coexpress gad1.1. About 36% of POA-esr1–expressing neurons expressed gad1.1 (148 ± 23/410 ± 42 cells, n = 2 fish). These results suggest that most POA-esr1–expressing neurons produce glutamate, and some POA-esr1–expressing neurons produce GABA. (d) Photographs showing merged images of double in situ hybridization for gad1.2 mRNA (green) and esr1 mRNA (magenta), demonstrating that very few POA-esr1 cells express gad1.2 (arrowhead). (e) Photographs showing merged images of double in situ hybridization for gad2 mRNA (green) and esr1 mRNA (magenta), demonstrating that some POA-esr1 cells express gad2 (arrowhead). ca, anterior commissure; DI, lateral part of dorsal telencephalic area; dDm, dorsal region of the medial part of dorsal telencephalic area. Scale bars: 25 μm. Figure 6. View largeDownload slide Double in situ hybridization for esr1 and marker genes for glutamate or GABA in POA. (a) Schematic illustration of a frontal section of the medaka brain, showing the plane of section corresponding to the photographs. The boxed area corresponds to the photographs in (b)–(e). (b) Photographs showing double labeling for vglut2.1 mRNA (green) and esr1 mRNA (magenta), demonstrating that POA-esr1–expressing cells express vglut2.1. These results suggest that the majority of the POA-esr1 neurons are glutamatergic. (c) Photographs showing double labeling for gad1.1 mRNA (green) and esr1 mRNA (magenta), which demonstrate that some POA-esr1–expressing cells coexpress gad1.1. About 36% of POA-esr1–expressing neurons expressed gad1.1 (148 ± 23/410 ± 42 cells, n = 2 fish). These results suggest that most POA-esr1–expressing neurons produce glutamate, and some POA-esr1–expressing neurons produce GABA. (d) Photographs showing merged images of double in situ hybridization for gad1.2 mRNA (green) and esr1 mRNA (magenta), demonstrating that very few POA-esr1 cells express gad1.2 (arrowhead). (e) Photographs showing merged images of double in situ hybridization for gad2 mRNA (green) and esr1 mRNA (magenta), demonstrating that some POA-esr1 cells express gad2 (arrowhead). ca, anterior commissure; DI, lateral part of dorsal telencephalic area; dDm, dorsal region of the medial part of dorsal telencephalic area. Scale bars: 25 μm. Table 4. Colocalization of Glutamatergic or GABAergic Markers in the POA-esr1 Neurons   Average Number of Cells in POA  % of Each Neurotransmitter Expressing POA-esr1 Neurons  vglut2.1 and esr1-expressing neuron  326 ± 81 (in 342 ± 71 esr1 neurons, n = 3 fish)  95  gad1.1 and esr1-expressing neuron  148 ± 23 (in 410 ± 42 esr1 neurons, n = 2 fish)  36  gad1.2 and esr1-expressing neuron  20 ± 4 (in 515 ± 28 esr1 neurons, n = 2 fish)  4  gad2 and esr1-expressing neuron  110 ± 18 (in 430 ± 56 esr1 neurons, n = 2 fish)  26    Average Number of Cells in POA  % of Each Neurotransmitter Expressing POA-esr1 Neurons  vglut2.1 and esr1-expressing neuron  326 ± 81 (in 342 ± 71 esr1 neurons, n = 3 fish)  95  gad1.1 and esr1-expressing neuron  148 ± 23 (in 410 ± 42 esr1 neurons, n = 2 fish)  36  gad1.2 and esr1-expressing neuron  20 ± 4 (in 515 ± 28 esr1 neurons, n = 2 fish)  4  gad2 and esr1-expressing neuron  110 ± 18 (in 430 ± 56 esr1 neurons, n = 2 fish)  26  View Large Figure 7. View largeDownload slide Double labeling analysis with esr1:EGFP Tg medaka suggests that POA-esr1 neurons coexpress pr. Photographs showing double labeling EGFP immunohistochemistry (green) and pr mRNA in situ hybridization (magenta), demonstrating that POA-esr1 neurons express pr (arrowhead). ca, anterior commissure; DI, lateral part of dorsal telencephalic area; dDm, dorsal region of the medial part of dorsal telencephalic area. Scale bars: 20 μm. Figure 7. View largeDownload slide Double labeling analysis with esr1:EGFP Tg medaka suggests that POA-esr1 neurons coexpress pr. Photographs showing double labeling EGFP immunohistochemistry (green) and pr mRNA in situ hybridization (magenta), demonstrating that POA-esr1 neurons express pr (arrowhead). ca, anterior commissure; DI, lateral part of dorsal telencephalic area; dDm, dorsal region of the medial part of dorsal telencephalic area. Scale bars: 20 μm. Discussion In the current study, we generated Tg medaka in which esr1-expressing neurons are labeled with EGFP and examined the anatomy of these neurons. We identified the neuronal pathways mediating estrogen feedback signals to gnrh1 neurons, as well as those mediating sex steroid–sensitive signals to a circuit that modulates sex behaviors. We found that POA-esr1 neurons project their axons to gnrh1 neurons and that the former comprise glutamatergic/GABAergic neurons, which may be widely conserved among mammals and teleosts. We also found a possible neural substrate for estrogen-sensitive modulation of sexual behaviors. Taken together with results of previous behavioral studies suggesting involvement of Vs neurons in the regulation of sexual behaviors (29–31), the present results suggest that the caudally projecting esr1 neurons in Vs are involved in regulating sexual behaviors with respect to the level of circulating estrogen. In addition, we found that some POA-esr1 neurons coexpress pr, which suggests that progesterone, in addition to estrogen, may regulate reproduction or sexual behaviors by controlling POA-esr1/pr neurons in accordance with serum hormone levels. gnrh1 neurons do not express esr1 in medaka In the current study, we analyzed the neuronal circuits involved in transmission of estrogenic signals to gnrh1 neurons in medaka. Esr1 has been demonstrated to be an essential factor for regulation of the HPG axis in mammals (6, 8). Therefore, we first examined whether gnrh1 neurons express esr1 in medaka (Fig. 1). Double in situ hybridization for gnrh1 and esr1 mRNA disproved the expression of esr1 in gnrh1 neurons [Fig. 1(f)]. Thus, as in mammals, gnrh1 neurons do not directly receive estrogenic signals via Esr1 in medaka (32, 33). Specificity of EGFP labeling in the brain of esr1:EGFP Tg medaka In the brains of esr1:EGFP Tg medaka, we found that EGFP-labeled cell bodies were distributed in the POA, Vs, and NVT, and we demonstrated the specificity of EGFP labeling in these regions (Fig. 2). It should be noted that not all esr1 neurons were labeled with EGFP; on the other hand, EGFP labeling was found in nucleus diffusus tori lateralis and the optic tectum, regions in which esr1 mRNA was not detected by in situ hybridization. So far, it is technically impossible to determine whether this signal indicates ectopically labeled neurons or neurons labeled more sensitively than with in situ hybridization. Either way, these labeled neurons projected locally and did not interfere with analysis of axonal projections of POA or Vs neurons in the current study. Interestingly, we also demonstrated, by double labeling of esr2a/esr2b mRNA in situ hybridization and EGFP immunohistochemistry, that some EGFP-positive esr1 neurons coexpressed esr2a or esr2b [Fig. 2(d) and 2(e)]. This finding is consistent with RNA-seq results, which suggested that the esr1 neuronal population includes esr2a or esr2b mRNA. Thus, a small population of esr1 neurons in this Tg line can be considered esr1/2a or esr1/2b neurons. Functions of POA-Esr1 neurons Using our Tg medaka line, we obtained morphological evidence that glutamatergic and GABAergic POA-esr1 neurons directly regulate hypophysiotropic gnrh1 neurons (Figs. 5 and 6 ). Many close contacts were observed between the gnrh1 neuronal cell bodies and the nerve terminals of POA-esr1 neurons, most of which were proven to be glutamatergic, and some of them were GABAergic. In medaka, a previous study demonstrated that expression of esr1,esr2a, and esr2b in the POA region was not changed by either ovariectomy or estrogen replacement (34). Thus, esr1 is considered to be stably expressed in the POA and to regulate the release of glutamate or GABA in accordance with the serum estrogen level. Because glutamate and GABA are the major excitatory synaptic transmitters (GABA is also suggested to be excitatory to GnRH1 neurons), it follows that estrogen-dependent modulation of glutamatergic and GABAergic transmission may affect GnRH1 firing activity. In medaka, the firing activity of GnRH1 neurons changes toward puberty (35); therefore, it is possible that the system for regulating GnRH1 neuron activity changes during pubertal development. After sexual maturation, adult female medaka show daily spawning, and our previous study has documented time-of-day–dependent changes in the firing activity of GnRH1 neurons and in expression of LH and FSH (23). However, the neuronal mechanisms underlying puberty onset or time-of-day–dependent changes in GnRH1 neuronal activities have not yet been clarified. In mammals, glutamate/GABA is thought to play important roles in the regulation of puberty onset and the estrous cycle; the proportion of glutamatergic agonist α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid–responsive GnRH1 neurons changes across puberty (36). Moreover, it has been reported that the number of fast synaptic transmissions (Glu/GABA) to GnRH1 neurons changes in accordance with the serum estrogen level in mammals (37, 38). Finally, a recent study suggested that glutamatergic Esr1 neurons in the mouse limbic forebrain play a key role in the negative/positive steroid feedback regulation of GnRH neurons, whereas GABAergic Esr1 neurons play a role in the positive feedback (39). Thus, estrogen-dependent alteration of glutamatergic and GABAergic inputs may be important for the reproductive regulation of vertebrates in general, and the glutamatergic and GABAergic neurons that project to the gnrh1 neurons morphologically identified here in medaka are candidates for such regulatory neurons. In the current study, we also found that POA-esr1 neurons coexpress pr (Fig. 7). Interestingly, in mice, arcuate GABA/PR neurons have been suggested to project heavily to GnRH1 neurons (40). Results of the current study suggested that 36% of POA-esr1 neurons are GABAergic. Thus, some POA-esr1/pr neurons that express GABA probably project to gnrh1 neurons in medaka. Taken together with the results of previous study in mice, our results show that the innervation of gnrh1 neurons by gaba/pr neurons seems to be conserved among mammals and teleosts, and the gaba/pr neurons may be important for the regulation of the HPG axis in vertebrates. Medaka brains, even in adults, are small and are advantageous for analyzing EGFP-labeled POA-esr1 neurons in whole brain in vitro preparations, in which we can maintain the neuronal circuitries regulating the GnRH1 neuron activities almost intact. Therefore, the Tg medaka we established in the current study should provide a good model system for studying estrogen-dependent changes in glutamatergic and GABAergic synaptic inputs that regulate GnRH1 neuron activities and hence the regulation of reproduction. In addition to these neuronal circuits, mammals have developed a Kiss1-mediated HPG axis regulation system for the fine tuning of reproduction; for example, follicular development and ovulation are suppressed during lactation in mammals, mainly because of the suppression of pulsatile GnRH/LH secretion (41). On the other hand, in teleosts it has been recently demonstrated that GnRH is necessary for ovulation but not folliculogenesis (5), suggesting that glutamatergic innervation of GnRH1 neurons may be important for LH surges and ovulation rather than LH pulses. Interestingly, in a daily surge model of OVX+E mouse (42), spontaneous glutamatergic excitatory postsynaptic potentials from GnRH neurons drastically increase in the afternoon (37). These lines of evidence suggest that glutamatergic transmission may be important for LH surges in both species, although other possibilities cannot be excluded. We also examined the projections from esr1 neurons to the pituitary. Results of the present anatomical analysis suggested that POA-esr1 neurons may directly regulate activity of gonadotropes. However, it should be noted that axons of the esr1 neurons were distributed broadly in the rostral half of the pituitary gland, which suggests that esr1 neurons may also regulate the release of hormones other than gonadotropins. Further investigation with this Tg line is necessary to clarify whether the POA-esr1 neurons directly regulate gonadotropin release. Regulation of sexual behaviors by Vs-Esr1 neurons and POA-Esr1/PR neurons Among rodents, estrogen is suggested to activate lordosis in female rats (19). However, the site of estrogen action in the neuronal circuits of this sexual behavior has not been clarified so far. In teleosts, previous studies reported that electrical stimulation of POA or Vs acutely induced sexual behavior in hime salmon (30). Thus, neurons in POA and Vs may be involved in the regulation of motor control of sexual behavior. In the current study, we showed that Vs-esr1 neurons project as far as the ventral region of medulla. However, we did not find neural fibers of Vs-esr1 neurons in the spinal cord in our Tg line. Therefore, uncharacterized neurons in the medulla that receive axonal projections from the Vs-esr1 neurons may relay this estrogenic modulation to the central pattern generator in the spinal cord that induces sexual behavior. Our present results clearly demonstrated that some of the esr1-expressing neurons in the POA coexpress pr [Fig. 7(b)], which is partly consistent with the results of previous studies in mice (43, 44). In mice, it has been shown that ablation of Esr1/PR neurons in the ventromedial hypothalamus diminishes sexual behaviors (31). Progesterone is suggested to induce sexual behavior in teleosts as well as in mammals (45). As suggested in rodent studies, it is possible that these Esr1/PR neurons are involved in sexual behavior. It should be noted that an immunohistochemical study in mammals demonstrated that PRs were detected only in Esr1 neurons (44). However, we suggest that some pr neurons in medaka do not express esr1. Although not all esr1 neurons were labeled by EGFP in the Tg line we established here, as shown in Fig. 7, many EGFP-negative pr neurons (326 ± 44/460 ± 42 cells, n = 3 fish) were found in the POA. Thus, it is possible that prs are expressed not only in esr1 neurons but also in esr1-negative neurons in medaka. The function of progesterone and PR is likely to differ between placental mammals and nonplacental teleosts, and thus additional studies using diverse mammals and teleosts are necessary to elucidate the reasons for this difference. A recent study in mice suggested that optogenetic activation of Esr1 neurons in the ventrolateral subdivision of the ventromedial hypothalamus induced sexual behavior (46). Additional investigation using genetic tools, such as Tg medaka that express optogenetic tools specifically in Esr1 neurons, may enable us to analyze neuronal circuits underlying estrogen-primed modulation of sexual behavior. In conclusion, in the current study we succeeded in anatomically identifying and visualizing two kinds of esr1 neurons: glutamatergic and/or GABAergic esr1 neurons in the POA that may relay gonadal estrogen feedback signals to gnrh1 neurons and esr1 neurons in the Vs that may modulate sexual behavior by acting on medullary circuits (Fig. 8). The Tg medaka line established in the current study will contribute to our understanding of the estrogen regulation of reproduction and behavior in vertebrates. Figure 8. View largeDownload slide Illustration of a working hypothesis concerning Esr1-mediated regulation of reproduction and sexual behavior. Glutamatergic and GABAergic POA-Esr1 neurons receive estrogenic feedback signals and directly regulate GnRH1 neurons in accordance with the gonadal status. Glutamatergic and GABAergic synaptic transmission to GnRH1 neuron is altered by estrogen levels. Estrogen also acts on Vs-Esr1 neurons and possibly modulates expression of neurotransmitters or firing activity in accordance with gonadal status. Unidentified neurons in the medulla that receive axonal projections from Vs-Esr1 neurons are suggested to relay this estrogenic modulation to the motor pattern generator in the spinal cord to modulate sexual behavior. Although the projection has not been clarified, Esr1/PR neurons are also considered to be involved in the regulation of sexual maturation and sexual behaviors. E2, estrogen; P, progesterone. Figure 8. View largeDownload slide Illustration of a working hypothesis concerning Esr1-mediated regulation of reproduction and sexual behavior. Glutamatergic and GABAergic POA-Esr1 neurons receive estrogenic feedback signals and directly regulate GnRH1 neurons in accordance with the gonadal status. Glutamatergic and GABAergic synaptic transmission to GnRH1 neuron is altered by estrogen levels. Estrogen also acts on Vs-Esr1 neurons and possibly modulates expression of neurotransmitters or firing activity in accordance with gonadal status. Unidentified neurons in the medulla that receive axonal projections from Vs-Esr1 neurons are suggested to relay this estrogenic modulation to the motor pattern generator in the spinal cord to modulate sexual behavior. Although the projection has not been clarified, Esr1/PR neurons are also considered to be involved in the regulation of sexual maturation and sexual behaviors. E2, estrogen; P, progesterone. Abbreviations: ABC avidin-biotin complex DIG digoxigenin EGFP enhanced green fluorescent protein Esr estrogen receptor Esr1 estrogen receptor α Esr2 estrogen receptor β Esr2a estrogen receptor β1 Esr2b estrogen receptor β2 FSH follicle-stimulating hormone GABA γ-aminobutyric acid GABAergic γ-aminobutyric acidergic GnRH gonadotropin-releasing hormone GnRH1 gonadotropin-releasing hormone 1 HPG hypothalamo-pituitary-gonadal ir immunoreactive LH luteinizing hormone mRNA messenger RNA NAT anterior tuberal nucleus NVT ventral tuberal nucleus PBS phosphate-buffered saline PBST phosphate-buffered saline/Tween PCR polymerase chain reaction PFA paraformaldehyde POA preoptic area PR progesterone receptor RNA-seq RNA sequencing RPKM reads per kilobase of transcript per million mapped reads SD standard deviation SSC saline sodium citrate Tg transgenic TNE Tris-HCl, NaCl, and EDTA Vs ventral telencephalic area. Acknowledgments We thank Drs. Martin Kelly, Oline Rønnekleiv, and Martha Bosch (Oregon Health and Science University) for advice on the cell harvesting procedure for deep sequencing; Dr. Heather Eisthen (Michigan State University) for help in editing the manuscript; Drs. Takeshi Kaneko and Hiroyuki Hioki (Kyoto University) for the gift of anti-EGFP antibody; Drs. Min Kyun Park (The University of Tokyo) and Kataaki Okubo (The University of Tokyo) for helpful advice and discussion; Ms. Maiko Matsuda (The University of Tokyo) for support during the preparation of probes; and Ms. Miho Kyokuwa and Hisako Kohno (The University of Tokyo) for excellent care of the fish used in this study. Financial Support: This work was supported by Japan Society for the Promotion of Science (http://dx.doi.org/10.13039/501100001691) Grants 26221104 (to Y.O.) and 13J10475 (to B.Z.). Disclosure Summary: The authors have nothing to disclose. References 1. Okubo K, Nagahama Y. Structural and functional evolution of gonadotropin-releasing hormone in vertebrates. Acta Physiol (Oxf) . 2008; 193( 1): 3– 15. Google Scholar CrossRef Search ADS PubMed  2. Childs G. Gonadotropes and lactotropes. In: Neill JD, ed. Knobil and Neill’s The Physiology of Reproduction . 3rd ed. New York, NY: Academic Press; 2006: 1492– 1539. 3. Karigo T, Aikawa M, Kondo C, Abe H, Kanda S, Oka Y. Whole brain-pituitary in vitro preparation of the transgenic medaka (Oryzias latipes) as a tool for analyzing the differential regulatory mechanisms of LH and FSH release. Endocrinology . 2014; 155( 2): 536– 547. Google Scholar CrossRef Search ADS PubMed  4. Chu L, Li J, Liu Y, Cheng CH. Gonadotropin signaling in zebrafish ovary and testis development: insights from gene knockout study. Mol Endocrinol . 2015; 29( 12): 1743– 1758. Google Scholar CrossRef Search ADS PubMed  5. Takahashi A, Kanda S, Abe T, Oka Y. Evolution of the hypothalamic-pituitary-gonadal axis regulation in vertebrates revealed by knockout medaka. Endocrinology . 2016; 157( 10): 3994– 4002. Google Scholar CrossRef Search ADS PubMed  6. Dorling AA, Todman MG, Korach KS, Herbison AE. Critical role for estrogen receptor alpha in negative feedback regulation of gonadotropin-releasing hormone mRNA expression in the female mouse. Neuroendocrinology . 2003; 78( 4): 204– 209. Google Scholar CrossRef Search ADS PubMed  7. Couse JF, Yates MM, Walker VR, Korach KS. Characterization of the hypothalamic-pituitary-gonadal axis in estrogen receptor (ER) null mice reveals hypergonadism and endocrine sex reversal in females lacking ERalpha but not ERbeta. Mol Endocrinol . 2003; 17( 6): 1039– 1053. Google Scholar CrossRef Search ADS PubMed  8. Smith JT, Cunningham MJ, Rissman EF, Clifton DK, Steiner RA. Regulation of Kiss1 gene expression in the brain of the female mouse. Endocrinology . 2005; 146( 9): 3686– 3692. Google Scholar CrossRef Search ADS PubMed  9. George JT, Seminara SB. Kisspeptin and the hypothalamic control of reproduction: lessons from the human. Endocrinology . 2012; 153( 11): 5130– 5136. Google Scholar CrossRef Search ADS PubMed  10. Pielecka-Fortuna J, Chu Z, Moenter SM. Kisspeptin acts directly and indirectly to increase gonadotropin-releasing hormone neuron activity and its effects are modulated by estradiol. Endocrinology . 2008; 149( 4): 1979– 1986. Google Scholar CrossRef Search ADS PubMed  11. Mitani Y, Kanda S, Akazome Y, Zempo B, Oka Y. Hypothalamic Kiss1 but not Kiss2 neurons are involved in estrogen feedback in medaka (Oryzias latipes). Endocrinology . 2010; 151( 4): 1751– 1759. Google Scholar CrossRef Search ADS PubMed  12. Kanda S, Akazome Y, Matsunaga T, Yamamoto N, Yamada S, Tsukamura H, Maeda K, Oka Y. Identification of KiSS-1 product kisspeptin and steroid-sensitive sexually dimorphic kisspeptin neurons in medaka (Oryzias latipes). Endocrinology . 2008; 149( 5): 2467– 2476. Google Scholar CrossRef Search ADS PubMed  13. Kanda S, Akazome Y, Mitani Y, Okubo K, Oka Y. Neuroanatomical evidence that kisspeptin directly regulates isotocin and vasotocin neurons. PLoS One . 2013; 8( 4): e62776. Google Scholar CrossRef Search ADS PubMed  14. Grone BP, Maruska KP, Korzan WJ, Fernald RD. Social status regulates kisspeptin receptor mRNA in the brain of Astatotilapia burtoni. Gen Comp Endocrinol . 2010; 169( 1): 98– 107. Google Scholar CrossRef Search ADS PubMed  15. Escobar S, Servili A, Espigares F, Gueguen MM, Brocal I, Felip A, Gómez A, Carrillo M, Zanuy S, Kah O. Expression of kisspeptins and kiss receptors suggests a large range of functions for kisspeptin systems in the brain of the European sea bass. PLoS One . 2013; 8( 7): e70177. Google Scholar CrossRef Search ADS PubMed  16. Tang H, Liu Y, Luo D, Ogawa S, Yin Y, Li S, Zhang Y, Hu W, Parhar IS, Lin H, Liu X, Cheng CH. The kiss/kissr systems are dispensable for zebrafish reproduction: evidence from gene knockout studies. Endocrinology . 2015; 156( 2): 589– 599. Google Scholar CrossRef Search ADS PubMed  17. Nakajo M, Kanda S, Karigo T, Takahashi A, Akazome Y, Uenoyama Y, Kobayashi M, Oka Y. Evolutionally conserved function of kisspeptin neuronal system is nonreproductive regulation as revealed by nonmammalian study. Endocrinology . 2018; 159( 1): 163– 183. Google Scholar CrossRef Search ADS PubMed  18. Akazome Y, Kanda S, Okubo K, Oka Y. Functional and evolutionary insights into vertebrate kisspeptin systems from studies of fish brain. J Fish Biol . 2010; 76( 1): 161– 182. Google Scholar CrossRef Search ADS PubMed  19. Pfaff DW, Sakuma Y. Facilitation of the lordosis reflex of female rats from the ventromedial nucleus of the hypothalamus. J Physiol . 1979; 288: 189– 202. Google Scholar PubMed  20. Kim Y, Ko E, Lappan S, Kezele A, Kwon SW, Choe JC. Changes in fecal estrogen levels and sexual behavior in captive sichuan snub-nosed monkeys (Rhinopithecus roxellana) following a male replacement. Zoo Biol . 2013; 32( 3): 342– 346. Google Scholar CrossRef Search ADS PubMed  21. Zempo B, Kanda S, Okubo K, Akazome Y, Oka Y. Anatomical distribution of sex steroid hormone receptors in the brain of female medaka. J Comp Neurol . 2013; 521( 8): 1760– 1780. Google Scholar CrossRef Search ADS PubMed  22. Okubo K, Sakai F, Lau EL, Yoshizaki G, Takeuchi Y, Naruse K, Aida K, Nagahama Y. Forebrain gonadotropin-releasing hormone neuronal development: insights from transgenic medaka and the relevance to X-linked Kallmann syndrome. Endocrinology . 2006; 147( 3): 1076– 1084. Google Scholar CrossRef Search ADS PubMed  23. Karigo T, Kanda S, Takahashi A, Abe H, Okubo K, Oka Y. Time-of-day-dependent changes in GnRH1 neuronal activities and gonadotropin mRNA expression in a daily spawning fish, medaka. Endocrinology . 2012; 153( 7): 3394– 3404. Google Scholar CrossRef Search ADS PubMed  24. Sasagawa Y, Nikaido I, Hayashi T, Danno H, Uno KD, Imai T, Ueda HR. Quartz-Seq: a highly reproducible and sensitive single-cell RNA sequencing method, reveals non-genetic gene-expression heterogeneity [published correction appears in Genome Biology201718:9]. Genome Biol . 2013; 14( 4): R31. Google Scholar CrossRef Search ADS PubMed  25. Akazome Y, Kanda S, Oka Y. Expression of vesicular glutamate transporter-2.1 in medaka terminal nerve gonadotrophin-releasing hormone neurones. J Neuroendocrinol . 2011; 23( 7): 570– 576. Google Scholar CrossRef Search ADS PubMed  26. Takahashi A, Islam MS, Abe H, Okubo K, Akazome Y, Kaneko T, Hioki H, Oka Y. Morphological analysis of the early development of telencephalic and diencephalic gonadotropin-releasing hormone neuronal systems in enhanced green fluorescent protein-expressing transgenic medaka lines. J Comp Neurol . 2016; 524( 4): 896– 913. Google Scholar CrossRef Search ADS PubMed  27. Adachi S, Yamada S, Takatsu Y, Matsui H, Kinoshita M, Takase K, Sugiura H, Ohtaki T, Matsumoto H, Uenoyama Y, Tsukamura H, Inoue K, Maeda K. Involvement of anteroventral periventricular metastin/kisspeptin neurons in estrogen positive feedback action on luteinizing hormone release in female rats. J Reprod Dev . 2007; 53( 2): 367– 378. Google Scholar CrossRef Search ADS PubMed  28. Kitahashi T, Ogawa S, Parhar IS. Cloning and expression of kiss2 in the zebrafish and medaka. Endocrinology . 2009; 150( 2): 821– 831. Google Scholar CrossRef Search ADS PubMed  29. Koyama Y, Satou M, Oka Y, Ueda K. Involvement of the telencephalic hemispheres and the preoptic area in sexual behavior of the male goldfish, Carassius auratus: a brain-lesion study. Behav Neural Biol . 1984; 40( 1): 70– 86. Google Scholar CrossRef Search ADS PubMed  30. Satou M, Oka Y, Kusunoki M, Matsushima T, Kato M, Fujita I, Ueda K. Telencephalic and preoptic areas integrate sexual behavior in hime salmon (landlocked red salmon, Oncorhynchus nerka): results of electrical brain stimulation experiments. Physiol Behav . 1984; 33( 3): 441– 447. Google Scholar CrossRef Search ADS PubMed  31. Yang CF, Chiang MC, Gray DC, Prabhakaran M, Alvarado M, Juntti SA, Unger EK, Wells JA, Shah NM. Sexually dimorphic neurons in the ventromedial hypothalamus govern mating in both sexes and aggression in males. Cell . 2013; 153( 4): 896– 909. Google Scholar CrossRef Search ADS PubMed  32. Laflamme N, Nappi RE, Drolet G, Labrie C, Rivest S. Expression and neuropeptidergic characterization of estrogen receptors (ERalpha and ERbeta) throughout the rat brain: anatomical evidence of distinct roles of each subtype. J Neurobiol . 1998; 36( 3): 357– 378. Google Scholar CrossRef Search ADS PubMed  33. Hrabovszky E, Shughrue PJ, Merchenthaler I, Hajszán T, Carpenter CD, Liposits Z, Petersen SL. Detection of estrogen receptor-beta messenger ribonucleic acid and 125I-estrogen binding sites in luteinizing hormone-releasing hormone neurons of the rat brain. Endocrinology . 2000; 141( 9): 3506– 3509. Google Scholar CrossRef Search ADS PubMed  34. Hiraki T, Takeuchi A, Tsumaki T, Zempo B, Kanda S, Oka Y, Nagahama Y, Okubo K. Female-specific target sites for both oestrogen and androgen in the teleost brain. Proc Biol Sci . 2012; 279( 1749): 5014– 5023. Google Scholar CrossRef Search ADS PubMed  35. Arai Y, Kanda S, Karigo T, Hasebe M, Oka Y. Changes in GnRH sensitivity of LH cells play a key role in seasonal regulation of reproduction. In: Abstract of the Annual Meeting of the Society for Neuroscience 2015; 2015; Chicago, IL. Program number 431.404 (poster number W425). 36. Constantin S, Jasoni CL, Wadas B, Herbison AE. Gamma-aminobutyric acid and glutamate differentially regulate intracellular calcium concentrations in mouse gonadotropin-releasing hormone neurons. Endocrinology . 2010; 151( 1): 262– 270. Google Scholar CrossRef Search ADS PubMed  37. Christian CA, Pielecka-Fortuna J, Moenter SM. Estradiol suppresses glutamatergic transmission to gonadotropin-releasing hormone neurons in a model of negative feedback in mice. Biol Reprod . 2009; 80( 6): 1128– 1135. Google Scholar CrossRef Search ADS PubMed  38. Christian CA, Moenter SM. Estradiol induces diurnal shifts in GABA transmission to gonadotropin-releasing hormone neurons to provide a neural signal for ovulation. J Neurosci . 2007; 27( 8): 1913– 1921. Google Scholar CrossRef Search ADS PubMed  39. Cheong RY, Czieselsky K, Porteous R, Herbison AE. Expression of ESR1 in glutamatergic and GABAergic neurons is essential for normal puberty onset, estrogen feedback, and fertility in female mice. J Neurosci . 2015; 35( 43): 14533– 14543. Google Scholar CrossRef Search ADS PubMed  40. Moore AM, Prescott M, Marshall CJ, Yip SH, Campbell RE. Enhancement of a robust arcuate GABAergic input to gonadotropin-releasing hormone neurons in a model of polycystic ovarian syndrome. Proc Natl Acad Sci USA . 2015; 112( 2): 596– 601. Google Scholar CrossRef Search ADS PubMed  41. Yamada S, Uenoyama Y, Kinoshita M, Iwata K, Takase K, Matsui H, Adachi S, Inoue K, Maeda KI, Tsukamura H. Inhibition of metastin (kisspeptin-54)-GPR54 signaling in the arcuate nucleus-median eminence region during lactation in rats. Endocrinology . 2007; 148( 5): 2226– 2232. Google Scholar CrossRef Search ADS PubMed  42. Christian CA, Mobley JL, Moenter SM. Diurnal and estradiol-dependent changes in gonadotropin-releasing hormone neuron firing activity. Proc Natl Acad Sci USA . 2005; 102( 43): 15682– 15687. Google Scholar CrossRef Search ADS PubMed  43. Gréco B, Allegretto EA, Tetel MJ, Blaustein JD. Coexpression of ER beta with ER alpha and progestin receptor proteins in the female rat forebrain: effects of estradiol treatment. Endocrinology . 2001; 142( 12): 5172– 5181. Google Scholar CrossRef Search ADS PubMed  44. Blaustein JD, Turcotte JC. Estradiol-induced progestin receptor immunoreactivity is found only in estrogen receptor-immunoreactive cells in guinea pig brain. Neuroendocrinology . 1989; 49( 5): 454– 461. Google Scholar CrossRef Search ADS PubMed  45. O’Connell LA, Ding JH, Hofmann HA. Sex differences and similarities in the neuroendocrine regulation of social behavior in an African cichlid fish. Horm Behav . 2013; 64( 3): 468– 476. Google Scholar CrossRef Search ADS PubMed  46. Lee H, Kim DW, Remedios R, Anthony TE, Chang A, Madisen L, Zeng H, Anderson DJ. Scalable control of mounting and attack by Esr1+ neurons in the ventromedial hypothalamus. Nature . 2014; 509( 7502): 627– 632. Google Scholar CrossRef Search ADS PubMed  Copyright © 2018 Endocrine Society

Journal

Endocrinology – Oxford University Press

Published: Feb 1, 2018

There are no references for this article.