Brain-derived estrogens facilitate male-typical behaviors by potentiating androgen receptor signaling in medaka

  1. Yuji Nishiike
  2. Shizuku Maki
  3. Daichi Miyazoe
  4. Kiyoshi Nakasone
  5. Yasuhiro Kamei
  6. Takeshi Todo
  7. Tomoko Ishikawa-Fujiwara
  8. Kaoru Ohno
  9. Takeshi Usami
  10. Yoshitaka Nagahama
  11. Kataaki Okubo  Is a corresponding author
  1. Department of Aquatic Bioscience, Graduate School of Agricultural and Life Sciences, The University of Tokyo, Japan
  2. Optics and Bioimaging Facility, Trans-Scale Biology Center, National Institute for Basic Biology, Japan
  3. Department of Genome Biology, Graduate School of Medicine, Osaka University, Japan
  4. Division of Reproductive Biology, National Institute for Basic Biology, Japan
5 figures, 1 table and 3 additional files

Figures

Figure 1 with 1 supplement
cyp19a1b-deficient males exhibit severely impaired male-typical mating and aggressive behaviors.

(A, B) Levels of E2, testosterone, and 11-ketotestosterone (11KT) in the brain (A) and periphery (B) of adult cyp19a1b+/+, cyp19a1b+/−, and cyp19a1b−/− males (n=3 per genotype). (C) Brain cyp19a1b transcript levels in cyp19a1b+/+, cyp19a1b+/−, and cyp19a1b−/− males (n=6 per genotype). Mean value for cyp19a1b+/+ males was arbitrarily set to 1. (D) Setup for testing the mating behavior of cyp19a1b+/+, cyp19a1b+/−, and cyp19a1b−/− males. (E) Latency of cyp19a1b+/+, cyp19a1b+/−, and cyp19a1b−/− males (n=18 per genotype) to initiate each mating act toward the stimulus female. (F) Setup for testing mating behavior using an esr2b-deficient female as the stimulus. (G) Latency of cyp19a1b+/+, cyp19a1b+/−, and cyp19a1b−/− males (n=16, 16, and 17, respectively) to initiate each mating act toward the esr2b-deficient female. (H) Number of each mating act performed. (I) Latency of cyp19a1b+/+ and cyp19a1b−/− males and cyp19a1b+/+ females (n=15 each) to receive courtship displays from the esr2b-deficient female. (J) Setup for testing aggressive behavior among grouped males. (K) Total number of each aggressive act performed by cyp19a1b+/+, cyp19a1b+/−, and cyp19a1b−/− males. Each data point represents the sum of acts recorded for the 4 males of the same genotype in a single tank (n=6, 7, and 6 tanks, respectively). Statistical differences were assessed by Bonferroni’s or Dunn’s post hoc test (A, B, C, H, K) and Gehan-Breslow-Wilcoxon test with Bonferroni’s correction (E, G, I). Error bars represent SEM. *p<0.05, **p<0.01, ***p<0.001.

Figure 1—figure supplement 1
Generation of cyp19a1b-deficient medaka.

(A) Schematic of the cyp19a1b locus showing the location of the mutation identified by TILLING (targeting-induced local lesions in genomes). Each exon is numbered. Nucleotide sequences of the wild-type (cyp19a1b+) and mutant (cyp19a1b) alleles at the mutation site are shown, with the substituted nucleotide in red. (B) Comparison of the amino acid sequences deduced from the cyp19a1b+ and cyp19a1b alleles. Amino acid numbers are shown on the right. Putative functional domains (membrane-spanning region [residues 21–34 of wild-type Cyp19a1b], I-helix region [287–321], Ozol’s peptide region [346–368], aromatic region [404–415], and heme-binding region [429–452]) are shaded in gray. Asterisks denote stop codons. (C) Predicted three-dimensional structures of wild-type (left) and mutant (right) Cyp19a1b proteins. Key structural features are annotated as follows: membrane helix (blue), aromatic region (red), and heme-binding loop (orange).

Figure 2 with 2 supplements
Brain-derived estrogens facilitate male-typical behaviors probably by stimulating brain AR expression.

(A) Setup for testing the mating behavior of E2-treated cyp19a1b+/+ and cyp19a1b−/− males. (B) Latency of cyp19a1b+/+ males (n=12) to initiate courtship displays toward the stimulus female before (day 0) and after (day 4) E2 treatment. (C) Number of courtship displays performed by cyp19a1b+/+ males. (D) Latency of cyp19a1b−/− males to initiate courtship displays before (day 0; n=12) and after (day 4; n=11) E2 treatment. (E) Number of courtship displays performed by cyp19a1b−/− males. (F) Total area of ara expression signal in each brain nucleus of cyp19a1b+/+ (n=6 except for pPPp, where n=5) and cyp19a1b−/− (n=7) males. The data are displayed in two graphs for visual clarity. (G) Representative images of ara expression in the PPa, pPPp, and NVT. (H) Total area of arb expression signal in each brain nucleus of cyp19a1b+/+ (n=6) and cyp19a1b−/− (n=7) males. The data are displayed in two graphs for visual clarity. (I) Representative images of arb expression in the PMp, aPPp, and NPT. (J) Total area of ara expression signal in the PPa, pPPp, and NVT of cyp19a1b+/+ and cyp19a1b−/− males treated with vehicle alone or E2 (n=5 per group except for NVT of E2-treated cyp19a1b+/+ males, where n=4; and PPa of vehicle-treated cyp19a1b+/+, E2-treated cyp19a1b+/+, and vehicle-treated cyp19a1b−/− males, where n=3). (K) Total area of arb expression signal in the PMp, aPPp, and NPT of cyp19a1b+/+ and cyp19a1b−/− males treated with vehicle alone or E2 (n=5 per group except for NPT of vehicle-treated cyp19a1b+/+ males, where n=4). Scale bars represent 50 μm. For abbreviations of brain nuclei, see Supplementary file 1. Statistical differences were assessed by Gehan-Breslow-Wilcoxon test (B, D) and unpaired t test, with Welch’s correction where appropriate (C, E, F, H, J, K). Error bars represent SEM. *p<0.05, **p<0.01, ***p<0.001.

Figure 2—figure supplement 1
Effect of estrogen replacement on aggression and brain ara and arb expression in cyp19a1b-deficient males.

(A) Total number of each aggressive act observed among cyp19a1b+/+ or cyp19a1b−/− males before (day 0; n=6 and 5, respectively) or after (day 4; n=5 and 4, respectively) E2 treatment. (B) Total area of ara expression signal in the Vs/Vp, PMp, PMm/PMg/aPPp, VM, NAT, and NPT of cyp19a1b+/+ and cyp19a1b−/− males treated with vehicle alone or E2 (n=5 per group except for Vs/Vp of vehicle- and E2-treated cyp19a1b+/+ males, PMp and NPT of E2-treated cyp19a1b+/+ males, and VM of all males, where n=4). (C) Total area of arb expression signal in the Vv, Vs/Vp, PPa, pPPp, and NVT of cyp19a1b+/+ and cyp19a1b−/− males treated with vehicle alone or E2 (n=5 per group except for NVT of vehicle-treated cyp19a1b+/+ males, where n=4; and Vv of vehicle- and E2-treated cyp19a1b−/− males, where n=3). For abbreviations of brain nuclei, see Supplementary file 1. Statistical differences were assessed by unpaired t test, with Welch’s correction where appropriate (A, B, C). Error bars represent SEM. *p<0.05, **p<0.01.

Figure 2—figure supplement 2
Effect of estrogen replacement on brain ara and arb expression in cyp19a1b-deficient males.

(A) Representative images of ara expression in the Vs/Vp, PMp, PPa, PMm/PMg/aPPp, pPPp, NAT, NVT, and NPT of cyp19a1b+/+ and cyp19a1b−/− males treated with vehicle alone or E2. (B) Representative images of arb expression in the Vs/Vp, PMp, PPa, aPPp, pPPp, NVT, and NPT of cyp19a1b+/+ and cyp19a1b−/− males treated with vehicle alone or E2. Scale bars represent 50 μm. For abbreviations of brain nuclei, see Supplementary file 1.

cyp19a1b deficiency impairs behaviorally relevant signaling pathways downstream of ARs.

(A) Total area of vt expression signal in the PMp/PPa/PMm/PMg, SC/aNVT, and pNVT of cyp19a1b+/+ and cyp19a1b−/− males (n=4 per genotype). (B) Representative images of vt expression in the pNVT. (C) Total area of gal expression signal in the aPMp/PPa, pPMp, PPp, and NAT/NVT/NRL of cyp19a1b+/+ and cyp19a1b−/− males (n=3 and 4, respectively). (D) Representative images of gal expression in the pPMp. Scale bars represent 50 μm. For abbreviations of brain nuclei, see Supplementary file 1. Statistical differences were assessed by unpaired t test, with Welch’s correction where appropriate (A, C). Error bars represent SEM. **p<0.01.

Figure 4 with 1 supplement
Estrogens directly stimulate the transcription of ARs through ESRs.

(A, B) Schematic of ara (A) and arb (B) loci showing the location of the canonical bipartite ERE-like sequences. Bent arrows mark the transcription initiation sites. Nucleotides of the ERE-like sequences are denoted by capital letters, and those identical to the consensus ERE (AGGTCAnnnTGACCT) are gray-shaded. (C) Ability of E2 to directly activate ara transcription. Cultured cells were transfected with a luciferase reporter construct containing a genomic fragment upstream of exon 3 of ara, together with an Esr1, Esr2a, or Esr2b expression construct. The cells were stimulated with different concentrations of E2, and luciferase activity was measured. (D) Effect of mutations in the ERE-like sequences on the E2-induced activation of ara transcription. Cultured cells were transfected with a wild-type luciferase construct or a construct carrying a mutation in the ERE-like sequence at position +1699 (mut +1699) or +2050 (mut+2050), together with an Esr1, Esr2a, or Esr2b expression construct. The cells were stimulated with or without E2, and luciferase activity was measured. (E) Ability of E2 to directly activate arb transcription. The assay described in C was performed with a luciferase construct containing a genomic fragment upstream of the first methionine codon of arb. (F) Effect of mutations in the ERE-like sequences on the E2-induced activation of arb transcription. The assay described in D was performed with luciferase constructs, each carrying a mutation in the ERE-like sequence at position –1272 (mut–1272), –2180 (mut–2180), or –3327 (mut–3327). Values are expressed as a fold change relative to a control without E2 stimulation (C, E) or a control using the wild-type construct without E2 stimulation (D, F). Statistical differences were assessed by Dunnett’s post hoc test (C, E) and unpaired t test with Bonferroni-Dunn correction (D, F). Error bars represent SEM. **p<0.01; ***p<0.001.

Figure 4—figure supplement 1
Effect of mutations in each half-site of the identified estrogen-responsive element (ERE) on the E2-induced activation of ara and arb transcription.

(A) Effect of mutations in each half-site of the ERE at position +2050 of the ara locus on the E2-induced activation of ara transcription. Cultured cells were transfected with a wild-type luciferase reporter construct or a construct carrying a mutation in the 5′ (mut+2050–5′) or 3′ (mut+2050–3′) half-site of the ERE, together with an Esr1, Esr2a, or Esr2b expression construct. The cells were stimulated with or without E2, and luciferase activity was measured. (B) Effect of mutations in each half-site of the ERE at position –3327 of the arb locus on the E2-induced activation of arb transcription. Cultured cells were transfected with a wild-type luciferase construct or a construct carrying a mutation in the 5′ (mut–3327–5′) or 3′ (mut–3327–3′) half-site of the ERE, together with an Esr1, Esr2a, or Esr2b expression construct. The cells were stimulated with or without E2, and luciferase activity was measured. Values are represented relative to the wild-type construct without E2 stimulation. Statistical differences were assessed by unpaired t test with Bonferroni-Dunn correction (A, B). Error bars represent SEM. ***p<0.001.

Figure 5 with 4 supplements
Brain-derived estrogens stimulate ara and arb expression in behaviorally relevant brain regions primarily through Esr2a and Esr1, respectively.

(A) Total area of arb expression signal in each brain nucleus of esr1+/+ and esr1−/− males (n=5 per genotype). The data are displayed in two graphs for visual clarity. (B) Representative images of arb expression in the PMp and aPPp. (C) Total area of ara expression signal in each brain nucleus of esr2a+/+ and esr2a−/− males (Δ8 line; n=5 per genotype except for NPT of esr2a+/+ males and Vs/Vp and PMp of esr2a−/− males, where n=4). The data are displayed in two graphs for visual clarity. (D) Representative images of ara expression in the PPa, pPPp, and NVT. (E) Setup for testing the mating behavior of esr1+/+ and esr1−/− males using an esr2b-deficient female as the stimulus. (F) Latency of esr1+/+ and esr1−/− males (n=24 and 31, respectively) to initiate each mating act toward the stimulus female. (G) Number of each mating act performed. (H) Setup for testing the mating behavior of esr2a+/+ and esr2a−/− males using an esr2b-deficient female as the stimulus. (I) Latency of esr2a+/+ and esr2a−/− males (Δ8 line; n=23 and 22, respectively) to initiate each mating act toward the stimulus female. (J) Number of each mating act performed. (K, L) Total number of each aggressive act observed among esr1+/+ or esr1−/− males (n=6 per genotype) (K) and among esr2a+/+ or esr2a−/− males (Δ8 line; n=8 and 7, respectively) (L) in the tank. Scale bars represent 50 μm. For abbreviations of brain nuclei, see Supplementary file 1. Statistical differences were assessed by unpaired t test, with Welch’s correction where appropriate (A, C, G, J, K, L) and Gehan-Breslow-Wilcoxon test (F, I). Error bars represent SEM. *p<0.05, **p<0.01.

Figure 5—figure supplement 1
Generation of esr2a-deficient medaka.

(A) Schematic of the esr2a locus showing the location of the CRISPR (clustered regularly interspaced short palindromic repeats) target site. Each exon is numbered. Nucleotide sequences of the wild-type (esr2a+) and mutant (esr2a [Δ8] and esr2a [Δ4]) alleles at the target site are shown, with the CRISPR RNA target sequence underlined. Dashes indicate deleted nucleotides. (B) Comparison of the amino acid sequences deduced from the esr2a+ and esr2a (Δ8 and Δ4) alleles. Amino acid numbers are shown on the right. The DNA-binding domain (residues 161–234 of wild-type Esr2a) and ligand-binding domain (304–514) are shaded in gray. The sequences altered by the frameshift are indicated in red. Asterisks denote stop codons.

Figure 5—figure supplement 2
Expression of ara and arb in the brain of males deficient for each ESR.

(A) Total area of ara expression signal in each brain nucleus of esr1+/+ and esr1−/− males (n=5 per genotype). (B) Representative images showing the coexpression of arb and esr1 in the PMp and aPPp. Left panels show arb expression (magenta), middle panels show esr1 expression (green), and right panels show the merged images with DAPI staining (blue). White arrowheads indicate neurons coexpressing arb and esr1. (C) Total area of arb expression signal in each brain nucleus of esr2a+/+ and esr2a−/− males (Δ8 line; n=4 and 5, respectively, except for NPT of esr2a+/+ males, where n=5). (D) Representative images showing the coexpression of ara and esr2a in the PPa, pPPp, and NVT. Left panels show ara expression (magenta), middle panels show esr2a expression (green), and right panels show the merged images with DAPI staining (blue). White arrowheads indicate neurons coexpressing ara and esr2a. (E) Total area of ara expression signal in each brain nucleus of esr2b+/+ and esr2b−/− males (n=5 per genotype). (F) Representative images of ara expression in the NAT. (G) Total area of arb expression signal in each brain nucleus of esr2b+/+ and esr2b−/− males (n=5 per genotype except for Vv of esr2b+/+ males, where n=4). (H) Representative images of arb expression in the PPa. Each quantitative dataset is displayed in two graphs for visual clarity. Scale bars represent 20 μm (B, D) and 50 μm (F, H). For abbreviations of brain nuclei, see Supplementary file 1. Statistical differences were assessed by unpaired t test, with Welch’s correction where appropriate (A, C, E, G). Error bars represent SEM. *p<0.05.

Figure 5—figure supplement 3
Mating behavior of esr1-deficient males.

Latency of esr1+/+ and esr1−/− males (n=23 and 24, respectively) to initiate each mating act toward the stimulus wild-type female. Statistical differences were assessed by Gehan-Breslow-Wilcoxon test. **p<0.01.

Figure 5—figure supplement 4
Mating and aggressive behaviors of esr2a-deficient males.

(A) Latency of esr2a+/+ and esr2a−/− males (Δ8 line; n=24 and 23, respectively) to initiate each mating act toward the stimulus wild-type female. (B) Latency of esr2a+/+ and esr2a−/− males (Δ4 line; n=23 per genotype) to initiate each mating act toward the stimulus wild-type female. (C) Latency of esr2a+/+ and esr2a−/− males (Δ4 line; n=24 per genotype) to initiate each mating act toward the stimulus esr2b-deficient female. (D) Number of each mating act performed. (E) Total number of each aggressive act observed among esr2a+/+ or esr2a−/− males (Δ4 line; n=8 and 7, respectively) in the tank. Statistical differences were assessed by Gehan-Breslow-Wilcoxon test (A, B, C) and unpaired t test, with Welch’s correction where appropriate (D, E). Error bars represent SEM. *p<0.05.

Tables

Key resources table
Reagent type (species) or resourceDesignationSource or referenceIdentifiersAdditional information
Gene (Oryzias latipes)cyp19a1bGenBankGenBank:AB591736
Gene (O. latipes)esr1GenBankGenBank:XM_020714493
Gene (O. latipes)esr2aGenBankGenBank:NM_001104702
Gene (O. latipes)esr2bGenBankGenBank:XM_020713365
Gene (O. latipes)araGenBank; NBRP MedakaGenBank:NM_001122911; NBRP Medaka clone ID:olova36n18
Gene (O. latipes)arbGenBankGenBank:NM_001104681
Gene (O. latipes)vtGenBankGenBank:NM_001278891
Gene (O. latipes)galGenBankGenBank:LC532140
Gene (O. latipes)actbGenBankGenBank:NM_001104808
Strain, strain background (O. latipes)d-rRNBRP MedakaStrain ID:MT837Maintained in a closed colony over 15 years in Okubo lab
Genetic reagent (O. latipes)cyp19a1b-deficient lineThis paperTILLING ID:57D05Generated and maintained in Okubo lab
Genetic reagent (O. latipes)esr1-deficient linehttps://doi.org/10.1093/pnasnexus/pgad413N/AGenerated and maintained in Okubo lab
Genetic reagent (O. latipes)esr2a-deficient Δ8 lineThis paperN/AGenerated and maintained in Okubo lab
Genetic reagent (O. latipes)esr2a-deficient Δ4 lineThis paperN/AGenerated and maintained in Okubo lab
Genetic reagent (O. latipes)esr2b-deficient linehttps://doi.org/10.1016/j.cub.2021.01.089TILLING ID:46E12Generated and maintained in Okubo lab
Cell line (Homo sapiens)HEK293TRiken BRC Cell Bankcell number:RCB2202; RRID:CVCL_0063
Cell line (H. sapiens)HeLaRiken BRC Cell Bankcell number:RCB0007; RRID:CVCL_0030
AntibodyAlkaline phosphatase-conjugated anti-DIG antibody (sheep polyclonal)Roche Diagnosticscat#:11093274910; RRID:AB_514497(1:500 or 1:2000)
AntibodyAnti-DIG antibody (mouse monoclonal)Abcamcat#:ab420; RRID:AB_304362(1:200)
AntibodyHorseradish peroxidase-conjugated anti-fluorescein antibody (sheep polyclonal)PerkinElmercat#:NEF710001EA; RRID:AB_2737388(1:1000)
Recombinant DNA reagentpcDNA3.1/V5-His-TOPOThermo Fisher Scientificcat#:K480001
Recombinant DNA reagentpGL4.10Promegacat#:E6651
Recombinant DNA reagentpGL4.74Promegacat#:E6921
Recombinant DNA reagentMedaka bacterial artificial chromosome (BAC) clone containing the ara locusNBRP MedakaNBRP Medaka clone ID:ola1-111G01
Recombinant DNA reagentMedaka BAC clone containing the arb locusNBRP MedakaNBRP Medaka clone ID:ola1-192H15
Sequence-based reagentCRISPR RNA (crRNA) for medaka esr2aFasmacN/ACTACGGCGTGTGGTCATGCGAGG
Sequence-based reagentTrans-activating CRISPR RNA (tracrRNA)Fasmaccat#:GE-002
Peptide, recombinant proteinCas9Nippon Genecat#:316-08651
Commercial assay or kitEstradiol ELISA KitCayman Chemical Companycat#:582251
Commercial assay or kitTestosterone ELISA KitCayman Chemical Companycat#:582701
Commercial assay or kit11-Keto Testosterone ELISA KitCayman Chemical Companycat#:582751
Commercial assay or kitRNeasy Plus Universal Mini KitQIAGENcat#:73404
Commercial assay or kitSuperScript VILO cDNA Synthesis KitThermo Fisher Scientificcat#:11754050
Commercial assay or kitLightCycler 480 SYBR Green I MasterRoche Diagnosticscat#:04707516001
Commercial assay or kitDIG RNA Labeling MixRoche Diagnosticscat#:11277073910
Commercial assay or kitFluorescein RNA Labeling MixRoche Diagnosticscat#:11685619910
Commercial assay or kitT7 RNA polymeraseRoche Diagnosticscat#:10881775001
Commercial assay or kitSP6 RNA polymeraseRoche Diagnosticscat#:10810274001
Commercial assay or kitDual-Luciferase Reporter Assay SystemPromegacat#:E1910
Commercial assay or kitPrimeSTAR Mutagenesis Basal KitTakara Biocat#:R046A
Commercial assay or kitAlexa Fluor 555 Tyramide SuperBoost Kit, goat anti-mouse IgGThermo Fisher Scientificcat#:B40913
Commercial assay or kitTSA Plus Fluorescein SystemPerkinElmercat#:NEL741001KT
Commercial assay or kitSep-Pak C18 Plus Light CartridgeWaters Corporationcat#:WAT023501
Chemical compound, drugEstradiol-17β (E2)Fujifilm Wako Pure Chemicalcat#:058-04043
Chemical compound, drug5-Bromo-4-chloro-3-indolyl phosphateRoche Diagnosticscat#:11383221001
Chemical compound, drugNitro blue tetrazoliumRoche Diagnosticscat#:11383213001
Chemical compound, drugLipofectamine LTXThermo Fisher Scientificcat#:15338100
Chemical compound, drugCharcoal/dextran-stripped fetal bovine serumCytivacat#:SH30068
Software, algorithmAlphaFold 3https://alphafoldserver.com/aboutRRID:SCR_025885
Software, algorithmPyMOLhttps://www.pymol.orgRRID:SCR_000305
Software, algorithmOlyviaOlympusRRID:SCR_016167
Software, algorithmJasparhttp://jaspar.genereg.net/RRID:SCR_003030
Software, algorithmMatchhttp://gene-regulation.com/pub/programs.htmlRRID:SCR_007787
Software, algorithmGraphPad PrismGraphPad SoftwareRRID:SCR_002798

Additional files

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Yuji Nishiike
  2. Shizuku Maki
  3. Daichi Miyazoe
  4. Kiyoshi Nakasone
  5. Yasuhiro Kamei
  6. Takeshi Todo
  7. Tomoko Ishikawa-Fujiwara
  8. Kaoru Ohno
  9. Takeshi Usami
  10. Yoshitaka Nagahama
  11. Kataaki Okubo
(2026)
Brain-derived estrogens facilitate male-typical behaviors by potentiating androgen receptor signaling in medaka
eLife 13:RP97106.
https://doi.org/10.7554/eLife.97106.4