GnRH Receptor Expression and Reproductive Function Depend on JUN in GnRH Receptor‒Expressing Cells

Abstract

Gonadotropin-releasing hormone (GnRH) from the hypothalamus regulates synthesis and secretion of luteinizing hormone (LH) and follicle-stimulating hormone (FSH) from the anterior pituitary gonadotropes. LH and FSH are heterodimers composed of a common α-subunit and unique β-subunits, which provide biological specificity and are limiting components of mature hormone synthesis. Gonadotrope cells respond to GnRH via specific expression of the GnRH receptor (Gnrhr). GnRH induces the expression of gonadotropin genes and of the Gnrhr by activation of specific transcription factors. The JUN (c-Jun) transcription factor binds to AP-1 sites in the promoters of target genes and mediates induction of the FSHβ gene and of the Gnrhr in gonadotrope-derived cell lines. To analyze the role of JUN in reproductive function in vivo, we generated a mouse model that lacks JUN specifically in GnRH receptor‒expressing cells (conditional JUN knockout; JUN-cKO). JUN-cKO mice displayed profound reproductive anomalies such as reduced LH levels resulting in lower gonadal steroid levels, longer estrous cycles in females, and diminished sperm numbers in males. Unexpectedly, FSH levels were unchanged in these animals, whereas Gnrhr expression in the pituitary was reduced. Steroidogenic enzyme expression was reduced in the gonads of JUN-cKO mice, likely as a consequence of reduced LH levels. GnRH receptor‒driven Cre activity was detected in the hypothalamus but not in the GnRH neuron. Female, but not male, JUN-cKO mice exhibited reduced GnRH expression. Taken together, our results demonstrate that GnRH receptor‒expression levels depend on JUN and are critical for reproductive function.

Knockdown of JUN in GnRH receptor‒expressing cells leads to diminished reproductive capacity, reduced GnRH receptor expression, and lower serum LH in male and female mice.

Mammalian reproduction is regulated by the hypothalamic-pituitary-gonadal axis. The hypothalamic decapeptide gonadotropin-releasing hormone (GnRH) is the final brain output that regulates both expression and secretion of gonadotropins, luteinizing hormone (LH), and follicle-stimulating hormone (FSH) from the anterior pituitary gonadotropes (1). This function is mediated by gonadotrope-specific expression of the GnRH receptor (Gnrhr), which belongs to the rhodopsin family of seven transmembrane G protein‒coupled receptors (2). LH and FSH in turn stimulate steroidogenesis and gametogenesis in the gonads (3, 4).

Gonadotropin levels are regulated primarily by transcription of their unique β-subunits, which provide biological specificity. Alternations in the transcription of β-subunits correlate with changes in the concentration of the mature hormones in the circulation (5, 6). The β-subunits heterodimerize with a common α-subunit to form the mature glycoproteins (7). GnRH induces LHβ (Lhb), FSHβ (Fshb), and Gnrhr transcription via induction of specific immediate-early genes: EGR1, which regulates Lhb transcription, and FOS and JUN, which activate both Fshb and Gnrhr transcription (4). The FOS and JUN transcription factors form the AP-1 heterodimer, which is rapidly and transiently activated (8). Both mouse and human Fshb and Gnrhr genes are induced by GnRH via AP-1 (9-13). Transcriptome analysis demonstrated that AP-1 members are strongly induced by GnRH in LβT2 cells (14) and in primary rat gonadotrope cells (15).

Responsiveness of the Fshb gene to GnRH is conveyed by AP-1 response elements in the proximal promoter (9, 16–19). GnRH induces FOS (c-Fos), FOSB, JUN (c-Jun), and JUNB but not JUND in the LβT2 cell line, a model of mature gonadotropes. A combination of these factors binds the AP-1 site in the Fshb promoter (9). In the αT3 gonadotrope cell line, GnRH regulates Gnrhr expression via AP-1 as well (11, 20). JUN homodimer or a heterodimer with FOS, FOSB, FRA1, or FRA2 binds the mouse Gnrhr promoter at two different sites (13, 21). AP-1 heterodimer of JUN and FOS also regulates expression of the human GNRHR gene by GnRH (22).

Although gonadotrope cell models, such as LβT2 and αT3 cells, facilitated identification of transcription factors that led to induction of gonadotrope genes, it is critical to determine the roles of these transcription factors in vivo. LHβ induction by GnRH is mediated by the EGR1 transcription factor. EGR1 is an immediate-early gene and a member of the zinc finger family of transcription factors. EGR1 plays a nonredundant role in reproduction, and other family members are unable to compensate. Consistent with this, global EGR1 knockout mice are infertile and lack LH expression, resulting in blunted sex steroid hormone synthesis (23, 24). FOS also plays nonredundant roles in reproduction in vivo (25). In the pituitary, FOS is critical for gonadotropin gene expression, whereas expression of another glycohormone subunit, TSHβ (Tshb), is not affected. In the hypothalamus, FOS is expressed in both kisspeptin and GnRH neurons during the preovulatory surge and can be used as a marker of their activation (26–28). FOS is necessary for normal kisspeptin neuron numbers and Kiss1 expression, primarily in the female, whereas GnRH neuron location, axon targeting, or gene expression does not depend on FOS (25).

Because JUN is an obligatory heterodimerization partner of FOS for DNA binding (8), we used c-Jun^flox/flox mice crossed to GnRH receptor Cre animals to create mice that lack JUN specifically in the Gnrhr-expressing cells. These conditional JUN knockout (JUN-cKO) mice were used to analyze the reproductive physiology and determine the cell-specific role of JUN in reproduction.

Materials and Methods

Cell lines and transient transfection

LβT2 cells, a gift from Dr. Pamela Mellon (University of California at San Diego), were maintained in Dulbecco’s modified Eagle medium with 10% fetal bovine serum at 37°C and 5% CO₂. The line was authenticated with reverse transcription polymerase chain reaction‒based expression analysis of endogenous gonadotropin β-subunits. For transfection, LβT2 cells were plated in 12-well plates 1 day before transfection with FuGENE 6 transfection reagent (Roche Molecular Biochemicals, Indianapolis, IN), 0.1 µg of expression vectors, 0.5 µg of a luciferase-reporter plasmid [reported previously (17, 29–31)], and 0.1 µg of thymidine kinase β-galactosidase, a reporter plasmid driven by a herpesvirus thymidine kinase promoter as a control for transfection efficiency. Complementary DNA (cDNA) for the AP-1 transcription factors was in the same backbone under the same promoter, and its expression was evaluated by western blot. Forty-eight hours after transfection, cells were lysed with 100 nM of KPO₄ buffer containing 0.2% Triton X-100 luciferase activity measured on a luminometer (Veritas Microplate luminometer; Turner Biosystems, Sunnyvale, CA) by injecting 100 µL of buffer containing 25 mM of Tris pH 7.8, 15 mM of MgSO₄, 10 mM of adenosine triphosphate, and 65 µM of luciferin into each well. Using the Tropix Galacto-Light β-Galactosidase Assay (Applied Biosystems, Foster City, CA) and following the manufacturer’s instructions, β-galactosidase activity was measured subsequently. Transfections were performed in triplicate and repeated a minimum of three times. One-way analysis of variance by the Tukey post hoc test was performed using the JMP program (SAS Institute, Cary, NC) with significance set at P < 0.05.

Animals

Mice lacking c-Jun in GnRH receptor‒expressing cells were obtained by crossing c-Jun^flox/flox mice with GnRH-Receptor-Cre (GRIC) mice. Briefly, c-Jun^flox/flox mice, in which the only coding exon of the c-Jun allele is flanked by LoxP sites (32, 33), were created by Dr. Randall Johnson (University of California, San Diego). Gnrhr^tm1(cre)Uboe mice (GnRH receptor-internal ribosome entry site-Cre, GRIC) carry a knock-in Gnrhr allele fused to an internal ribosome entry site and a Cre transgene. GRIC drives Cre expression in pituitary gonadotrope cells (34). Because some Cre expression is also observed in male germ cells in these animals (35), the GRIC allele was always introduced via the female. Homozygous c-Jun^flox/flox Cre⁺ mice served as experimental mice, whereas Cre⁻ littermates were used as controls. TdTomato reporter mice, Gt(ROSA)26Sor^{tm9(CAG-tdTomato)Hze}/J, were obtained from Jackson Laboratory (strain 007909) and crossed to GRIC mice to analyze Cre activity. Animals were maintained under a 12-hour light, 12-hour dark cycle and received food and water ad libitum. All experiments were performed with approval from the University of California Animal Care and Use Committee and in accordance with the National Institutes of Health Animal Care and Use Guidelines using 8-week-old animals, unless indicated otherwise. Males and females were analyzed separately to determine potential sex differences. At least six animals per sex per genotype were analyzed, and statistical differences between Cre⁺ and Cre⁻ were determined by Student t test and Tukey test for multiple comparisons.

Fertility studies

Eight-week-old Cre⁺ and Cre⁻ male or female mice were individually paired with an adult C57BL/6 mouse of the opposite sex, and the presence of litters was monitored daily over a period of 4 months. In addition, starting at 8 weeks of age, a separate cohort of female mice was assessed for estrous cycle stage with daily vaginal smears for 5 weeks.

Sperm count

The epididymides were dissected, macerated, and incubated in 1 mL of Dulbecco’s modified Eagle medium at room temperature for 30 minutes with shaking. Sperm was cleared with a 70-μm cell strainer, diluted with sterile water, and counted with a hemocytometer.

Histological analyses and immunohistochemistry

Ovaries and testes were fixed overnight at 4°C in 4% paraformaldehyde or Bouin’s fixative, respectively. Tissues were dehydrated in ethanol, embedded in paraffin, cut into 10-μm-thick sections, floated onto UltraClear™ Plus Microslides (Denville Scientific Inc, Holliston, MA), and stained with hematoxylin and eosin.

Pituitaries were fixed in 4% paraformaldehyde, embedded in paraffin, and cut to 10 μm. Slides were deparaffinized in xylene and rehydrated. Antigen unmasking was performed by heating for 10 minutes in Tris-EDTA-0.3% Triton X, and endogenous peroxidase was quenched by incubating for 10 minutes in 0.3% hydrogen peroxide. Slides were then blocked with 20% goat serum and incubated with primary antiserum against LH (1:300 raised in rabbit; National Hormone and Peptide Program, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD) overnight at 4°C. After phosphate-buffered saline washes, slides were incubated with biotinylated goat anti-rabbit immunoglobulin G (1:300, BA-1000; Vector Laboratories, Burlingame, CA) for 30 minutes. The Vectastain ABC Elite Kit (Vector Laboratories) was used per manufacturer’s instructions, after which the DAB peroxidase kit was used for colorimetric staining. Slides were dehydrated in ethanol and xylene and cover-slipped with Vectamount (Vector Laboratories).

To visualize costaining of TdTomato and pituitary hormones, pituitaries were fixed in 4% paraformaldehyde, frozen in OCT, and cut to 12-μm sections using a Leica cryostat. Hypothalami were sectioned to 30-μm sections for GnRH staining. Slides were blocked with 20% goat serum and incubated with primary antibodies against LH or FSH (Table 1; 1:300 raised in rabbit; National Hormone and Peptide Program, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health) or GnRH [provided kindly by Greg Anderson, University of Otago, Dunedin, New Zealand (36)] overnight at 4°C. After phosphate-buffered saline washes, slides were incubated with biotinylated goat anti-rabbit immunoglobulin G (1:300, BA-1000; Vector Laboratories) for 30 minutes, followed by Streptavidin-Cy5 (1:500; Molecular Probes; Thermo Fisher) for 30 minutes. Secondary antibody-only controls were performed, and it was determined that endogenous TdTomato expression was strong for visualization and that its emission in the TdTomato/rhodamine channel overlapped with the FITC/Alexa 488 channel. Thus, Streptavidin-Cy5 was used for visualization of LH-, FSH- or GnRH-expressing cells, and slides were cover-slipped using Vectasheild (Vector Laboratories). To determine percentage of coexpression, we counted how many of the hundred LH- or FSH-containing cells expressed TdTomato and vice versa. We counted at least three nonoverlapping fields of view in three different sections per mouse (= nine fields), and stained pituitaries from three male and three female Cre⁺ mice.

Table 1.

Antibodies

Antibody	Provider, Catalog No.	Species	Dilution	RRID
LH	NHPP, AFP240580Rb	Rabbit	1:300	AB_2665533
FSH	NHPP, AFP-C0972881	Rabbit	1:300	AB_2687903
Prolactin	NHPP, PRL	Rabbit	1:300	AB_2629220
GnRH	Greg Anderson, University of Otago	Rabbit	1:5000	AB_2721118

Abbreviations: NHPP, National Hormone and Peptide Program; RRID, Research Resource Identifier.

Quantitative polymerase chain reaction analyses

Tissues were dissected, and total RNA was extracted and reverse-transcribed using Superscript III (Invitrogen, CA). Quantitative polymerase chain reaction (qPCR) was performed using an iQ SYBR Green supermix and an IQ5 real-time PCR machine (Bio-Rad Laboratories, Hercules, CA) with the primers listed in Table 2 under the following conditions: 95°C for 15 minutes followed by 40 cycles at 95°C for 20 seconds, 56°C for 30 seconds, and 72°C for 30 seconds. A standard curve with dilutions of 10 pg per well, 1 pg per well, 100 fg per well, and 10 fg per well of a plasmid containing LHβ or FSHβ cDNA was generated in each run with the samples. The amount of the gene of interest was calculated by comparing the threshold cycle obtained for each sample with the standard curve generated in the same run. Replicates were averaged and divided by the mean value of the GAPDH housekeeping gene in the same sample using the ^ΔΔCt method. After each run, a melting curve analysis was performed to confirm that a single amplicon was generated in each reaction. Statistical differences in expression between genotypes were determined by Student t test and the Tukey honest significant difference (HSD) test for multiple comparisons using JMP software (SAS Institute).

Table 2.

Primers

Primers	Forward	Reverse
Lhb (LHβ)	CTGTCAACGCAACTCTGG	ACAGGAGGCAAAGCAGC
Fshb (FSHβ)	GCCGTTTCTGCATAAGC	CAATCTTACGGTCTCGTATACC
Cga (aGSU)	ATTCTGGTCATGCTGTCCATGT	CAGCCCATACACTGGTAGATGG
Gnrhr (GnRH receptor)	GCCCCTTGCTGTACAAAGC	CCGTCTGCTAGGTAGATCATCC
Prl (prolactin)	TGTTCCCAGCAGTCACCAT	CAGCAACAGGAGGAGTGTC
Star (StAR)	GGAGGGGTGGTAGTCAGGAGA	TCCCCTGTTCGTAGCTGCTG
Cyp11	AAGTATGGCCCCATTTACAGG	TGGGGTCCACGATGTAAACT
Cyp17a1	ATCCTTGTCACGGTGGGAGA	GGAGGTGAGTCCGGTCATTG
Cyp19a1 (aromatase)	TTCCCATGGCAGATTCTTGTG	CGAATCGGGAGATGTAGTG
Shbg (ABP)	GACATTCCCCAGCCTCATGCA	TGCCTCGGAAGACAGAACCAC
Cldn3 (claudin 3)	AACTGCGTACAAGACGAGACG	GGCACCAACGGGTTATAGAAAT
Gnrh (GnRH)	CTACTGCTGACTGTGTGTTTG	CATCTTCTTCTGCCTGGCTTC
Gapdh	TGCACCACCAACTGCTTAG	GGATGCAGGGATGATGTTC

Abbreviations: ABP, androgen-binding protein, Shbg, sex hormone‒binding globulin; StAR, steroidogenic acute regulatory protein.

Hormone analyses

For serum collection, mice were euthanized between 9 and 11 am by isoflurane inhalation, and blood was obtained from the inferior vena cava. The blood was left to coagulate for 15 minutes at room temperature and then centrifuged at 2000 relative centrifugal force for 15 minutes for serum separation. Hormone assays were performed by the University of Virginia Ligand Core. The University of Virginia Center for Research in Reproduction Ligand Assay and Analysis Core is a fee-for-service core facility that is supported in part by Eunice Kennedy Shriver National Institute of Child Health and Development/National Institutes of Health Grant U54-HD28934.

LH was analyzed using a sensitive two-site sandwich immunoassay (37), and mouse LH reference prep (AFP5306A; provided by Dr. A.F. Parlow and the National Hormone and Peptide Program) was used as a standard. FSH was assayed by radioimmunoassay using reagents provided by Dr. A.F. Parlow and the National Hormone and Peptide Program, as previously described (38). Mouse FSH reference prep AFP5308D was used for assay standards. Steroid hormone levels were analyzed using validated commercially available assays, information for which can be found on the core’s website: http://www.medicine.virginia.edu/research/institutes-and-programs/crr/lab-facilities/assay-methods-page and as previously reported (39). Limits of detection were 0.24 ng/mL for LH, 2.4 ng/mL for FSH, 3 pg/mL for estradiol, and 10 ng/dL for testosterone (T). Intra-assay and interassay coefficients of variation were 6.4%/8.0%, 6.9%/7.5%, 6.0%/11.4%, and 4.4%/6.4% for LH, FSH, estrogen (E2), and T, respectively. For the assays used for this manuscript, interassay coefficients of variation data are the result of 30 assays for LH and FSH and 60 assays for E2 and T. Six animals per group were used for each hormone analysis. Statistical differences in hormone levels between wild-type and null groups were determined by Student t test and Tukey-Kramer HSD for multiple comparisons using JMP software (SAS Institute).

Results

JUN induced FSHβ and Gnrhr reporters in the LβT2 gonadotrope cell line

Given that the AP-1 family of transcription factors is composed of four FOS members [FOS (c-Fos), FOSB, FRA1, and FRA2] and three JUN members [JUN (c-Jun), JUNB, and JUND], combinatorial heterodimerization of these provides a variety of different factors that can induce target genes. GnRH induces all family members in gonadotropes, except for JUND (9). Because AP-1 heterodimers bind FSHβ (Fshb) and Gnrhr promoters using the electromobility shift assay (9, 13, 21), we first analyzed the level of induction of these target genes in gonadotropes with different combinations of AP-1 factors. cDNAs for the AP-1 transcription factors were cloned in the same vector backbone under the same promoter, and their expression was confirmed by western blot (data not shown). We also compared the induction with AP-1 overexpression with the induction by GnRH (G in Fig. 1). Because Gnrhr reporter induction by GnRH was previously analyzed using αT3-1 cells, a model of immature gonadotrope, we determined the level of induction in the model of mature gonadotropes, LβT2 cells. GnRH induced FSHβ reporter 6.2-fold, and Gnrhr reporter 2.4-fold. FRA1 and FRA2 did not induce FSHβ (Fig. 1A) or Gnrhr (Fig. 1B) expression either alone or more highly in combination with either JUN or JUNB compared with JUN or JUNB alone. JUN in combination with FOS or FOSB induced FSHβ reporter to similar levels compared with those observed with GnRH treatment (Fig. 1A). JUN heterodimers induced FSHβ to higher levels compared with JUNB heterodimers with FOS or FOSB. Gnrhr, on the other hand, was induced to similar levels by either JUN or JUNB heterodimers with FOS or FOSB (Fig. 1B). In LβT2 cells, Gnrhr was induced by GnRH 2.4-fold, a level similar to that observed with AP-1 overexpression. Because JUN induced both AP-1 gene targets in gonadotrope-derived cell lines, we next crossed c-JUN^flox/flox mice with GRIC animals to analyze the role of JUN in gonadotropes in vivo.

Figure 1.

JUN-containing heterodimers induced both FSHβ and Gnrhr reporters in LβT2 cells. Expression vectors for AP-1 isoforms were cotransfected with the mouse (A) FSHβ and (B) Gnrhr reporters. In the separate sets of samples, cells transfected with the reporters were treated with vehicle (V) or GnRH (G; 10 nM of GnRH, 5 hours). Data represent a mean of three independent experiments, each performed in triplicate. Asterisk (*) indicates significant induction compared with the empty vector control. Error bars represent standard error of the mean. Dashed line delineates lack of induction. cF, cFOS; ev, empty vector; F1, Fra1; F2, Fra2; FB, FosB.

Reduced reproductive capacity but normal gonadotrope numbers in mice lacking JUN in GnRH receptor‒expressing cells

Previous studies successfully used the GRIC allele to express Cre recombinase in gonadotropes to analyze the roles of transcription factors in gonadotropin gene expression (40–42). We used the GRIC allele to knock down JUN and create a JUN-cKO. Because JUN is an immediate-early gene that is expressed at a very low basal level, undetectable by immunostaining, we were unable to reliably demonstrate JUN knockdown in the gonadotrope. Thus, to analyze Cre activity in the gonadotrope and coexpression of Cre and gonadotropin hormones, we used TdTomato reporter mice in which TdTomato is specifically induced in Cre-expressing cells, following Cre-mediated excision of the stop cassette. Immunohistochemisty of frozen pituitary sections with antibodies to gonadotropin hormones revealed faithful expression of the TdTomato fluorescence [i.e., that 98% of TdTomato-expressing cells also express LH or FSH, consistent with the previous report, (34)]. Furthermore, 88% of cells containing LH expressed TdTomato (Fig. 2A; white arrowheads indicate LH-containing cells lacking TdTomato expression). Seventy-six percent of FSH-containing cells expressed TdTomato (Fig. 2B).

Figure 2.

TdTomato (TdTom) coexpressed with gonadotropin hormones. TdTomato reporter mice were crossed with GRIC mice to allow TdTomato expression in a Cre-dependent manner. Pituitaries from GRIC⁺/TdTomato⁻ mice (not shown) and from GRIC⁺/TdTomato⁺ mice from four separate litters were sectioned and stained for LH and FSH. (A) In all, 88% of LH (green) cells coexpressed TdTomato (red; arrowheads indicate LH cells that did not express TdTomato). (B) In all, 76% of FSH cells (green) coexpressed TdTomato (red; arrowheads indicate FSH cells that lack TdTomato expression). One hundred gonadotropin hormone‒containing cells were counted in three nonoverlapping fields of view in three sections from three different male and three female mice.

JUN-cKO animals exhibited profound changes in their reproductive physiology; females had significantly longer estrous cycles, 7.4 days, than controls, 4.4 days [(Fig. 3A) representative females’ stage of the estrous cycle over a 33-day period; (Fig. 3B) days per cycle in six females per genotype]. Male JUN-cKO mice had a 43% lower sperm count than that of controls (Fig. 3C). JUN-cKO mice also displayed longer intervals between litters when paired with wild-type C57BL/6 mice of the opposite sex (Fig. 3D; female cKO data presented, male data not shown).

Figure 3.

Fertility was profoundly affected in JUN-cKO animals. (A) Representative estrous cycle changes in Cre⁻ control females (top) and JUN-cKO females (bottom) were assessed by vaginal smears for 33 days starting at 8 weeks of age. (B) JUN-cKO animals (cKO; black bars) had increased average cycle length (six females per group) compared with that of Cre⁻ controls (Ctrs; gray bars). (C) Sperm count indicated 43% lower numbers in 8-week-old JUN-cKO mice compared with control littermates. (D) Animals were continuously present in the cages with wild-type mice of the opposite sex and were monitored daily for litters. JUN-cKO mice had a longer time interval between litters. Differences (*) between Ctr and JUN-cKO animals were determined by Student t test followed by the Tukey HSD test. Error bars represent standard error of the mean. D/M, diestrus/metestrus; E, estrus; P, proestrus.

To assess the role of JUN in gonadotrope differentiation, we stained pituitaries from JUN-cKO and control mice for LH to determine the number of gonadotropes. The morphology and size of JUN-cKO (cKO, Cre⁺) and control (Ctr, Cre⁻) pituitaries were indistinguishable (Fig. 4A). We then counted gonadotropes and determined that animals of both sexes and both genotypes contained the same numbers (Fig. 4B). Therefore, the lack of JUN in the gonadotropes did not affect gonadotrope numbers. The JUN-cKO and control animals were of the same size and weight (data not shown). Therefore, despite the same number of gonadotropes, the lack of JUN in gonadotrope cells resulted in subfertility in both sexes.

Figure 4.

JUN was not required for gonadotrope differentiation. (A) Pituitaries of control (Ctr; Cre⁻, cJun^flox/flox homozygous without Cre recombinase) and JUN-cKO (cKO; Cre⁺, cJun^flox/flox homozygous with Cre recombinase) animals were subjected to immunohistochemistry for LH to analyze number of gonadotrope cells. (B) Quantification of gonadotropes in males and females indicates that the lack of JUN had no effect on gonadotrope cell number (Ctr, Cre⁻: gray bars; cKO, Cre⁺: black bars). Error bars represent standard error of the mean.

JUN-cKO mice had reduced LH levels

Analyses of gonadotropin levels in the circulation revealed that JUN-cKO males exhibited 49% lower serum LH level than control males, whereas the LH concentration in JUN-cKO diestrus females was reduced by 56% compared with that of control females in diestrus (Fig. 5A). Although GnRH induced FSHβ via FOS and JUN in the gonadotrope-derived cell line, FSH levels were the same in both JUN-cKO and control males and in JUN-cKO and control diestrus females (Fig. 5B). Steroid hormone levels were reduced; T was lower in males, whereas estradiol was lower in females (Fig. 5C), likely because of reduced LH levels in the circulation.

Figure 5.

Lower LH levels in JUN-cKO animals. Six 8-week-old Cre⁻ controls (Ctrs; gray bars) and six JUN-cKO littermates (cKO; black bars) were analyzed for serum gonadotropin concentration. Females were monitored for the estrous cycle stage and analyzed in diestrus. (A) Male and female JUN-cKO mice had lower levels of LH than Cre⁻ controls, (B) whereas FSH level was unchanged. (C) Consequently, sex steroid levels were lower. Differences (*) between control (gray bars) and JUN-cKO (black bars) mice were determined by Student t test followed by the Tukey HSD test. Error bars represent standard error of the mean.

We also analyzed gonadotrope gene expression at 8 weeks of age. Consistent with the reduction in LH concentration in the circulation, Lhb messenger RNA (mRNA) level was 29% lower in JUN-cKO males (Fig. 6A) and 62% lower in JUN-cKO diestrus females than in Cre⁻ littermate controls (Fig. 6B). Consistent with unaltered FSH levels, there was no difference in Fshb expression between genotypes (Fig. 6C and 6D). Expression of the Gnrhr mRNA, however, was reduced by 28% in JUN-cKO males (Fig. 6E) and by 56% in JUN-cKO females (Fig. 6F). Expression of the common Cga subunit (αGSU) that heterodimerizes with both LHβ and FSHβ was unaffected (Fig. 6G and 6H). Previous studies analyzing Lhb expression did not reveal a role for the FOS and JUN AP-1 family, whereas the importance of AP-1 in Gnrhr induction is well established (13, 43). Our results may point to a role for AP-1 in Lhb expression. On the other hand, concomitant reduction of both Lhb and Gnrhr expression in both JUN-cKO males and females may implicate diminished Gnrhr levels in lower Lhb mRNA. This is consistent with previous studies postulating that the receptor concentration correlates with LHβ levels (44).

Figure 6.

Reduced LHβ and Gnrhr expression in JUN-cKO mice. Pituitaries from six 8-week-old Cre⁻ controls (Ctrs; gray bars) and six JUN-cKO littermates (cKO; black bars) were analyzed for expression of gonadotrope genes by qPCR: (A, B) Lhb (LHβ; males, females), (C, D) Fshb (FSHβ; males, females), (E, F) Gnrhr (males, females), and (G, H) Cga (common αGSU; males, females). Statistical significance (*) between Ctr and cKO was determined by Student t test followed by the Tukey post hoc test. Error bars represent standard error of the mean. GAPDH, glyceraldehyde 3-phosphate dehydrogenase.

Reduced LH levels target genes in the gonads of JUN-cKO mice

We next analyzed potential downstream effects of reduced LH levels in the gonads in both males and females at 12 weeks of age. Male JUN-cKO mice had a 22% reduction in seminal vesicle weight compared with that of controls (Fig. 7A), which is consistent with reduced intratesticular T levels (Fig. 7B). We also examined the expression of steroidogenic enzymes, which are induced by LH signaling. Although Star (steroidogenic acute regulatory protein) expression was unchanged, expression of Cyp11 and Cyp17 was reduced by 20% and 25%, respectively (Fig. 7C‒7E). Expression of the FSH target gene in the testis, Shbg (sex hormone‒binding globulin; androgen-binding protein), was unaffected, consistent with unperturbed FSH levels in the circulation (Fig. 7F). We observed lower sperm numbers, as shown previously. T levels, which were reduced because of the reduction in LH concentration, are necessary for spermatogenesis and for the maintenance of the blood-testis barrier. The blood-testis barrier is established via expression of tight junction proteins from the claudin family (45, 46). Expression of claudin 11 (Cldn11) did not change (data not shown). Claudin 3 (Cldn3) expression is regulated by androgens (47); however, despite a decrease in T, expression of Cldn3 was not significantly reduced (Fig. 7G) (P = 0.1).

Figure 7.

Reduced seminal vesicle weight and spermatogenesis in male JUN-cKO mice. (A) Seminal vesicles were dissected and measured to reveal reduced weight in 12-week-old JUN-cKO males. (B) Testes were homogenized and intratesticular T was measured. (C‒J) Testes were homogenized, and mRNA was extracted using Trizol. qPCR revealed lower expression of (D) CYP11 and (E) CYP17 steroidogenic enzymes and (I, J) lower levels of mRNA for late-stage spermatogenesis markers. (K, L), Histological analyses of testes after hematoxylin and eosin staining exhibited some abnormal seminiferous tubules in JUN-cKO males. Differences (*) between control (Ctr; gray bars) and JUN-cKO (cKO; black bars) mice were determined by Student t test followed by the Tukey HSD test. Error bars represent standard error of the mean. GAPDH, glyceraldehyde 3-phosphate dehydrogenase.

Given that sperm numbers in the epididymides were diminished, we assessed markers for several stages of spermatogenesis (48) and determined that the early stage spermatogenesis marker Sycp3 was unchanged, whereas later stage markers such as Spert and Elp were reduced in JUN-cKO males by 31% and 36%, respectively, compared with controls (Fig. 7H‒7J). Histological analyses of the testes uncovered a small number of abnormal seminiferous tubules (∼5%) lacking mature sperm in the JUN-cKO males (Fig. 7L). Thus, lack of JUN in GnRH receptor‒expressing cells in JUN-cKO males caused lower expression of steroidogenic enzymes and reduced levels of the late-stage spermatogenesis markers, corresponding to a reduced sperm count.

The ovaries of JUN-cKO mice weighed 37% less than control ovaries and contained fewer corpora lutea (Fig. 8A‒8C). JUN-cKO females expressed a 43% lower level of the LH target gene Cyp17a1 (Fig. 8D), whereas the FSH target gene Cyp19a1 (aromatase) was unchanged in the ovaries (Fig. 8E). Given that the antral stage of folliculogenesis is not affected corresponding to unaltered FSH levels, fewer corpora lutea may stem from reduction in prolactin levels, as prolactin is necessary for corpus luteum function in rodents (49, 50). We measured the expression of prolactin (Prl) in the pituitary and determined that Prl mRNA was reduced by 67% in JUN-cKO female mice (Fig. 8F). Therefore, female as well as male gonads from JUN-cKO animals exhibited a phenotype corresponding to diminished reproductive capacity.

Figure 8.

Reduced expression of LH target gene CYP17 and fewer corpora lutea in JUN-cKO females. (A) Histological analyses of ovaries after hematoxylin and eosin staining illustrates lower numbers of corpora lutea in JUN-cKO females. (B) Ovaries from JUN-cKO mice were smaller and (C) had fewer corpora lutea. Ovaries were homogenized, and mRNA was extracted using Trizol. qPCR revealed lower expression of (D) CYP17 but not (E) CYP19 steroidogenic enzymes in 12-week-old JUN-cKO female mice. (F) Prolactin (Prl) expression in the pituitary was reduced. Statistical significance (*) between control (Crt; gray bars) and JUN-cKO (cKO; black bars) mice was determined by Student t test followed by the Tukey post hoc test. Error bars represent standard error of the mean. GAPDH, glyceraldehyde 3-phosphate dehydrogenase.

Cre activity in the hypothalamus

The lack of an effect on Fshb expression and FSH levels in the circulation was unexpected, given previous evidence in the literature. In addition to inadequate Cre activity in a portion of gonadotrope cells, other JUN family members such as JUNB, which is also induced by GnRH (9), may compensate for the loss of JUN. Although JUN and JUNB exert nonoverlapping functions in other tissues, as evidenced by the different phenotypes of the respective knockout mice (32, 51), they may be able to substitute for each other in this scenario. To assess a possible compensatory increase in JUNB expression, we analyzed the level of Junb mRNA in the pituitaries of JUN-cKO and control males and females. In both sexes, JUN-cKO animals exhibited an increase in JUNB expression in the pituitary (Fig. 9). Therefore, JUNB increase may compensate for the loss of JUN for Fshb but not for Gnrhr expression.

Figure 9.

Increased JUNB expression. JUNB expression in the pituitaries of the 8-week-old male and female controls and JUN-cKO mice was analyzed to determine whether JUNB expression was elevated to compensate for the lack of JUN. Statistical significance (*) was determined by Student t test and Tukey post hoc analysis. Error bars represent standard error of the mean. GAPDH, glyceraldehyde 3-phosphate dehydrogenase.

On the other hand, reduced expression of Gnrhr and Lhb may stem from extrapituitary sites. Gnrhr is expressed in several hypothalamic nuclei and may be expressed in GnRH neurons themselves (52–58). We used TdTomato reporter mice to determine the activity of Cre recombinase in the hypothalamus. We also performed immunostaining for GnRH to detect GnRH neurons and determine whether TdTomato is expressed in GnRH neurons after Cre excision of the stop cassette. Coronal sections of the mediobasal hypothalamus demonstrated that TdTomato was expressed in the arcuate nucleus in GRIC⁺ animals, whereas GnRH axon terminals were located in the median eminence (Fig. 10A; GnRH, green; TdTomato, red). Staining of the preoptic area detected GnRH neurons in their expected location, whereas TdTomato-expressing cells were situated more laterally (Fig. 10B; GnRH, green; TdTomato, red). There was no overlap of green and red fluorescence in any section from either male or female mice.

Figure 10.

Cre activity in the hypothalamus. (A) Coronal sections at the level of the mediobasal hypothalamus demonstrate TdTomato reporter expression (red) and Cre activity in the arcuate nucleus and GnRH axon terminals staining in the median eminence (green). (B) Coronal section of the preoptic area shows that GnRH neurons (green) did not express TdTomato reporter (red). Male and female mice from four separate litters were used, and no sex differences were detected. (C) GnRH expression (Gnrh) in the hypothalami of male mice was not altered. (D) Reduced expression of Gnrh gene in the female JUN-cKO mice. Statistical significance (*) was determined by Student t test and Tukey post hoc analysis. Error bars represent standard error of the mean. cKO, JUN-cKO; Ctr, control; GAPDH, glyceraldehyde 3-phosphate dehydrogenase.

We also performed qPCR on the biopsy punched preoptic area and analyzed Gnrh expression. Gnrh expression did not differ in control and JUN-cKO male mice (Fig. 10C). However, Gnrh expression was reduced by 56% in female JUN-cKO mice (Fig. 10D). Given that GnRH neurons of either sex did not express TdTomato reporter, we hypothesize that lower Gnrh mRNA levels in the female may stem from the upstream regulatory neurons that may be affected by either lack of JUN in GnRH receptor‒expressing cells or by lower E2 levels.

Discussion

The molecular mechanisms of GnRH regulation of its target genes in pituitary gonadotropes have been previously examined primarily in cell lines and in primary cultures; however, the role of GnRH-induced transcription factors regulating gonadotrope genes in vivo is just beginning to emerge. As an immediate-early gene, JUN is rapidly induced in gonadotrope cells after GnRH treatment both in vivo (59) and in model cell lines (14, 60). In these, JUN mediates GnRH induction of the FSHβ (Fshb) gonadotropin subunit (9) and of the Gnrhr (12) by binding to the AP-1 site in the proximal promoters of these genes after dimerization with FOS. Herein, we examined the role of JUN in the hypothalamic-pituitary-gonadal axis gene expression in vivo, using c-Jun^flox/flox mice crossed to GRIC, in which Cre expression is driven by the Gnrhr promoter. We demonstrated that JUN expression in the GnRH receptor‒expressing cells is necessary for normal reproductive function.

Mice lacking JUN in GnRH receptor‒expressing cells exhibit a number of reproductive defects. Males have decreased Lhb and Gnrhr expression, which results in a decline in LH concentration in the circulation and consequent reduction in testicular function, including lower expression of several steroidogenic enzymes, leading to reduced T levels, smaller seminal vesicles, and fewer mature spermatozoa. Females as well have lower LH levels, which results in longer estrous cycles, reduced expression of the Cyp17 steroidogenic enzyme, and fewer corpora lutea in the ovaries. Reduced numbers of corpora lutea despite normal numbers of antral follicles may stem from diminished intraovarian steroid hormone levels due to lower expression of Cyp17. Alternatively, fewer corpora lutea may be a result of abrogated prolactin levels. Prolactin has a critical permissive role for LH action in the ovary and is necessary for luteinization and corpus luteum function in rodents (49, 50). Reduced prolactin expression likely derives from decreased levels of steroid hormones. Estrogen strongly upregulates prolactin in females (61, 62). In males, expression of aromatase in the pituitary allows for T conversion to E2, which then increases prolactin levels (63). Therefore, decreased E2 may contribute to diminished prolactin expression and reduced numbers of corpora lutea. Previous studies analyzing regulation of Lhb expression failed to find a role for JUN, whereas JUN is involved in Gnrhr induction. Because Lhb expression is dependent on Gnrhr numbers at the surface of gonadotropes (44), we believe that reduced levels of Gnrhr are a cause of diminished LH levels. On the other hand, it is possible that AP-1 has a role in Lhb expression in vivo.

Unexpectedly, FSH levels were unchanged in the cKO animals, although in the LβT2 model cell line JUN mediated GnRH induction of the Fshb subunit (9). This may illustrate a discrepancy between cell models and in vivo function, as suggested before (64). The GRIC model was used in the recent literature to analyze the role of transcription factors in the gonadotrope (40–42). We determined a significant overlap between LH and TdTomato expression. Although the difference in the percentage coexpression of the reporters and LH between previously reported results (34) and results reported herein is small, it may stem from different fluorescent reporters used or from heterogeneous expression in gonadotrope cells, whereby weakly fluorescent cells might have escaped detection in the absence of anti-TdTomato staining. Twelve percent of LH-expressing cells lacked TdTomato expression, demonstrating insufficient Cre activity in these cells. The number of FSH-expressing cells that did not have sufficient Cre expression was higher at 24%. It is possible that FSH-containing cells that do not express functional Cre are sufficient to maintain normal levels of FSH in the circulation, especially because FSH can be constitutively secreted (65, 66) and thus would be less dependent on the level of Gnrhr expression. Lack of Gnrhr expression in a portion of the FSH-containing cells was reported previously (35), although it was postulated that this population is present only during development. Data presented herein imply that FSH-containing cells without the Gnrhr persist in adulthood, which is consistent with several previous studies (67, 68).

Compensation by JUNB may explain unchanged FSH levels as well, although in most tissues, JUN and JUNB have opposing effects (69, 70). However, FSHβ was more highly induced by JUN heterodimers than by JUNB heterodimers, whereas Gnrhr induction was the same with either JUN or JUNB heterodimers with FOS or FOSB. The effect on Gnrhr expression may indicate that either the Gnrhr is more sensitive to the levels of JUN or that JUNB cannot compensate for JUN to induce Gnrhr expression.

Given that the GRIC allele also drives Cre expression in the testes, in the germ cells (35), there is concern that the gonadal phenotype in the male mice may be caused by a lack of JUN in the testes. However, that is unlikely for several reasons. The specific lack of JUN in male and female JUN-cKO mice resulted in the same outcomes: lower expression of Gnrhr and Lhb mRNA in the pituitary, reduced LH in the circulation, and diminished expression of LH-dependent genes in the gonads, resulting in lower sex steroid levels. In fact, in the female JUN-cKO mice, all these effects were exacerbated compared with the male JUN-cKO mice. Furthermore, known targets of AP-1 in the testes were not affected. Although the AP-1 binding site was identified in the FSH receptor promoter, regulating expression of the FSH receptor by FSH (71), in the testes of JUN-cKO males FSH receptor expression was not affected (data not shown).

AP-1 factors also play roles in tight junction formation and the blood-testis barrier (45, 46). The blood-testis barrier, which is necessary for spermatogenesis and fertility, is established via expression of tight junction proteins, primarily claudin 11 (72–74). Expression of claudin 11 is unaltered in JUN-cKO mice (data not shown). Claudin 3, expression of which is regulated by androgens, forms the stage-specific basal barrier in mice (47). Despite a decrease in circulating T level, expression of claudin 3 was not significantly changed either. Because late-stage spermatogenesis markers were reduced, AP-1 may regulate spermatogenesis directly (75). Because of a lack of known AP-1 target genes in germ cells, we were not able to delineate whether a decrease in late-stage spermatogenesis was due to testicular expression of Cre or to reduced levels of LH and diminished T. Therefore, because JUN-cKO males and females exhibited similar phenotypes and Cre was not expressed in the ovary, the observed effects likely stemmed from the gonadotrope-specific JUN knockdown.

Cre expression is driven by the Gnrhr regulatory region, which is expressed in several other extrapituitary sites. Gnrhr, in addition to pituitary gonadotrope, are expressed in the mediobasal hypothalamus, amygdala, and hippocampus (52), but the specific neuronal populations that express Gnrhr are unknown. Several studies demonstrated that Gnrhr is expressed in about 50% of GnRH neurons (55–58), suggesting that the Gnrhr may contribute to autocrine GnRH pulse generation (76, 77). Using this same GRIC mouse, ablation of GnRH receptor‒expressing neurons resulted in an elevated number of GnRH neurons (78), implying that the Gnrhr is not expressed in GnRH neurons themselves but may be expressed in afferent neurons that regulate GnRH neurons. It was also postulated that central GnRH via hypothalamic Gnrhr upstream of GnRH neurons may participate in the pulsatile release and preovulatory surge (79). Our analyses of GnRH receptor‒driven Cre expression in the hypothalamus demonstrated Cre activity in the arcuate nucleus and in the preoptic area but not in GnRH neurons themselves.

Examination of GnRH expression determined that Gnrh mRNA was significantly reduced specifically in female JUN-cKO mice. We previously observed female-specific effects using FOS null animals (25). Because there was no overlap between GnRH neurons and TdTomato expression, these findings suggest that GnRH expression is mediated in part via activity-regulated gene induction by afferent neurons, which may be affected by reduced E2 levels or by JUN knockdown. A number of previous reports determined that hypothalamic factors involved in reproductive function, such as RFamide-related peptide 3, a mammalian gonadotropin-inhibitory hormone ortholog; senktide, a neurokinin B receptor agonist; and oxytocin, elicit changes in LH serum levels, not only via alterations of GnRH secretion but also by modifications of Gnrh transcription (80–82). Alternatively, diminished Gnrh mRNA transcription may be secondary to reduced LH levels that caused lower E2 levels (83, 84). Similar to other studies using whole animal models in which endocrine loops were dysregulated, we were not able to distinguish between these alternatives. These results may indicate that the observed reproductive phenotype in females may stem from reduced GnRH expression.

In summary, our analyses of the mice that lacked JUN in GnRH receptor‒expressing cells revealed several physiological roles of this gene in the reproductive axis. Reduced Gnrhr and lower LH levels contributed to diminished sex-steroid hormone levels, impaired spermatogenesis, and reduced numbers of corpora lutea. Unchanged FSH levels may be due to the compensatory role of JUNB for this gene target but not for the Gnrhr or to the presence of FSH gonadotropes that lack sufficient Cre activity. We demonstrated that JUN expression in GnRH receptor‒expressing cells is necessary for normal reproductive function.

Glossary

Abbreviations:

cDNA

complementary DNA

E2

estrogen

FSH

follicle-stimulating hormone

GnRH

gonadotropin-releasing hormone

Gnrhr

GnRH receptor

GRIC

GnRH-receptor-Cre

HSD

honest significant difference

JUN-cKO

conditional JUN knockout

LH

luteinizing hormone

mRNA

messenger RNA

qPCR

quantitative polymerase chain reaction

T

testosterone