Leptin, but not Estradiol, Signaling in PACAP Neurons Modulates Puberty Onset

Abstract

The adipose-derived hormone leptin critically modulates reproductive function, such that its absence results in hypothalamic hypogonadism. Pituitary adenylate cyclase-activating polypeptide (PACAP)-expressing neurons are potential mediators of leptin's action on the neuroendocrine reproductive axis because they are leptin-sensitive and involved in both feeding behavior and reproductive function. In the complete absence of PACAP, male and female mice exhibit metabolic and reproductive abnormalities, yet there is some sexual dimorphism in the reproductive impairments. We tested whether PACAP neurons play a critical and/or sufficient role in mediating leptin's effects on reproductive function by generating PACAP-specific leptin receptor (LepR) knockout and rescue mice, respectively. We also generated PACAP-specific estrogen receptor alpha knockout mice to determine whether estradiol-dependent regulation of PACAP was critically involved in the control of reproductive function and whether it contributed to the sexually dimorphic effects of PACAP. We showed that LepR signaling in PACAP neurons is critically involved in the timing of female, but not male, puberty onset, but not fertility. Rescuing LepR-PACAP signaling in otherwise LepR-deficient mice was unable to rescue the reproductive deficits observed in LepR null mice but led to a marginal improvement in body weight and adiposity in females. Finally, PACAP-specific estrogen receptor alpha knockout did not lead to any changes in body weight or puberty onset compared with control mice. These data highlight that PACAP is a critical mediator of some of leptin's, but not estradiol's, influence on puberty onset in females, but is not critically involved in relaying leptin's effects in males or in adult females.

The adipose-derived hormone leptin plays a critical role in the regulation of metabolic and reproductive function, such that genetic leptin deficiency results in hypothalamic hypogonadism, infertility, severe obesity, and glucose intolerance, which can all be corrected with exogenous leptin administration (1-3). Although the signaling form of the leptin receptor (LepRb) is expressed both peripherally and centrally, leptin exerts its critical control over metabolic and reproductive function by acting centrally (4). Much headway has been made in deciphering the complex neurocircuitry and intracellular signaling pathways whereby leptin exerts its targeted control over metabolic and reproductive function (5, 6), yet identifying the targeted effect(s) of leptin-LepRb signaling in discrete neuronal populations remains an active area of research. Recently, the role of leptin signaling on pituitary adenylate cyclase-activating polypeptide (PACAP)-expressing neurons has come to attention. PACAP has been shown to influence feeding behavior, energy expenditure, and reproductive parameters, and, in turn, PACAP expression is modulated by fasting, high-energy diet-feeding, and, importantly, both leptin deficiency and treatment (7, 8).

In the complete absence of PACAP, male mice exhibit testosterone deficiency and LH insufficiency and female mice exhibit estrous cycle disruption, subfertility, and reduced uterine implantation rates (9, 10). However, PACAP is expressed in the gonads (11, 12) as well as the hypothalamus (13, 14) and it therefore remained unclear whether peripheral and/or central PACAP deficiency was underlying the observed reproductive phenotype in PACAP-deficient mice. Female mice expressing PACAP deficiency from either LepRb-expressing cells or from the ventral premamillary nucleus of the hypothalamus (PMV) were shown to exhibit delayed puberty onset and impaired fecundity (15). Although LepRb-expressing cells can be found in the gonads, PACAP and LepRb are not expressed in the same cell types in the gonads (16-18), suggesting PACAP acts centrally, presumably in LepRb-expressing cells in the PMV, as a neuropeptide to critically regulate the timing of sexual maturation, at least in female mice. What remains unclear is whether leptin signaling in PACAP neurons is critically required or sufficient to induce appropriate levels of PACAP expression for normal reproductive function, or whether other modulators of PACAP expression can compensate in the absence of LepRb-PACAP signaling. For example, long-term estradiol treatment to ovariectomized rats was shown to increase central PACAP gene expression (19), and paraventricular PACAP gene expression has also been shown to follow estrous cycle-driven fluctuations in rats (20). We were therefore interested in further exploring the interaction of LepRb signaling as well as estrogen receptor alpha (ERα) signaling in PACAP-expressing neurons in the control of male and female metabolic and reproductive function. To this end, we generated mice exhibiting the targeted deletion of either LepRb or ERα exclusively from PACAP-expressing cells (PACAP-LepR KO mice and PACAP-ERα KO mice, respectively) as well as LepRb-deficient mice exhibiting LepRb exclusively in PACAP-expressing cells (PACAP-LepR rescue mice) to test whether PACAP-LepRb or PACAP-ERα signaling are critically required or sufficient, respectively, in the control of metabolic and/or reproductive function in mice.

Methods

Animals

All mice were bred on a predominantly C57BL/6 strain background. To generate mice with deletion of LepRb specifically from PACAP neurons, homozygous Lepr flox mice (Lepr^fl/fl; loxP sites flanking Lepr coding exon 17; RRID:MGI:3511747) (21) were bred to PACAP-Cre mice (Jackson Laboratory RRID:IMSR_JAX:030155) (22). The resulting Lepr^fl/wt, PACAP-Cre mice were backcrossed to Lepr^fl/fl mice to generate Lepr^fl/fl, PACAP-Cre conditional knockout mice (referred to as PACAP-LepR KO mice). To generate mice with specific rescue of LepRb only in PACAP neurons, mice heterozygous for a Cre-dependent Lepr expression strategy (Lepr^loxTB/wt; Jackson Laboratory RRID:IMSR_JAX:018989; loxP-flanked transcription blocker sequence between exons 16 and 17 of the Lepr gene prevents transcription of the downstream exons (23)) were bred to PACAP-Cre mice. The resulting Lepr^loxTB/wt, PACAP-Cre mice were bred together to generate Lepr^loxTB/loxTB (LepR Null), PACAP-Cre conditional rescue mice (referred to as PACAP-LepR Rescue mice) (24). To generate mice with deletion of ERα specifically from PACAP neurons, homozygous ERα flox mice (ERα^fl/fl; loxP sites flanking exon 3) (24) were bred to PACAP-Cre mice. The resulting ERα^fl/wt, PACAP-Cre mice were backcrossed to ERα^fl/fl mice to generate ERα^fl/fl, PACAP-Cre conditional KO mice (referred to as PACAP-ERα KO mice). PACAP-Cre was visualized through PACAP-Cre-dependent tau green fluorescent protein (τGFP) expression as a result of crossing PACAP-Cre mice with a τGFP reporter line (MGI ID:4878874) (25, 26), producing PACAP-τGFP mice.

Transgenic mice were identified by PCR analysis of genomic DNA using the following primer sets at annealing temperatures of 59 °C (Lepr flox), 61 °C (LeprloxTB), or 55 °C (PACAP-Cre): for Lepr flox identification, AAT GAA AAA GTT GTT TTG GGA CGA and CAG GCT TGA GAA CAT GAA CAC AAC AAC and CTG ATT TGA TAG ATG GTC TTG AG (200-bp product indicates the wild-type gene, 250 bp indicates the floxed gene); for LeprloxTB identification, TGG CTT TTA AGC TCT GCA GTC and TAG GGC CAA ACC CAC ATT TA and CCC AAG GCC ATA CAA GTG TT (522-bp product indicates the wild-type gene, 360 bp indicates the floxed gene); for ERα flox identification, TAG GCT TTG TCT CGC TTT CC and CCC TGG CAA GAT AAG ACA GC and AGG AGA ATG AGG TGG CAC AG (350 bp indicates the wild-type gene and 480 bp indicates the floxed gene); for PACAP-Cre identification, GAA ATT GCA TCG CAT TGT CT, CAG TTG TTT TCT TAG ATG CCT GAC, CAG ACT CTC AGC CCA GAA GG and AGA CTT GCC CCC AGA TTC TC (291 bp indicates the wild-type gene, 219 bp indicates the Cre gene); for τGFP: CGA AGT CGC TCT GAG TTG TTA TC, GCA GAT GGA GCG GGA GAA AT, GCT CCT ATT GGC GTT ACT ATG (600-bp product indicates the wild-type gene and a 400-bp product indicates the τGFP gene). Animals were group housed or paired with an animal of the opposite sex (for the fecundity experiments). Mice were housed under conditions of controlled lighting (lights on from 6:00 Am to 6:00 Pm) and temperature (22 ± 1 °C). They had ad libitum access from the date of weaning to standard rodent chow, except during overnight fasting as described. All mice were weighed weekly. The University of Otago Animal Ethics Committee approved all animal experimental protocols.

Assessment of Hypothalamic PACAP Expression Across Development

Changes in hypothalamic PACAP expression across development have not been described. One practical reason this is important is that any developmental decreases in PACAP gene (Adcyap1) expression would render the PACAP-Cre mouse less useful for adult studies because Cre-driven gene excision is permanent even if Cre expression declines. To assess this, we generated mice expressing τGFP exclusively in PACAP-expressing cells (PACAP-τGFP mice) by crossing PACAP-Cre mice with a τGFP reporter line. Cre-induced GFP expression was then visualized across the hypothalamus in brain sections collected from female mice postfixed in paraformaldehyde on postnatal day 0 (PND0) or perfused on the day of first estrus (PND33). To better visualize endogenous GFP expression, fluorescent immunohistochemical labeling for GFP was performed. Sections were incubated for 24 to 48 hours in the primary antibody, chicken anti-GFP (1:5000 dilution, Aves Labs #GFP-1020, RRID:AB_10000240). Tissue was then incubated for 1.5 hours in the secondary antibody, AlexaFluor-488 goat anti-chicken immunoglobulins (1:500 dilution; Invitrogen, Thermo Fisher Scientific). Tissues were then washed, mounted onto gelatin-coated microscope slides, and cover slipped with Fluoromount G mounting medium. Because hypothalamic PACAP has been mapped in adults previously, the primary aim of this pilot study was to confirm that Adcyap1 expression does not decrease across postnatal development and the study was limited to 2 to 3 female mice per time point.

Experiment 1: Are Leptin Actions on PACAP Neurons Required for Normal Body Weight, Adiposity, Puberty Onset, and Fertility?

PACAP-LepR KO and LepR^fl/fl littermate controls were used to evaluate the requirement of leptin signaling in PACAP neurons for energy homeostasis, puberty onset, and subsequent fertility. All mice were weighed weekly from weaning, and at the end of the study, subcutaneous and abdominal fat mass was assessed. In female mice, puberty onset was measured by assessing the age of vaginal opening and first estrus. From 21 days of age, all mice were checked daily for vaginal opening. Once this had occurred, vaginal cytology was used to detect occurrence of first estrus. Estrous cyclicity of adult females was then assessed for 14 consecutive days, starting at least 10 days after first estrus. Experimental and control female animals were paired with adult wild-type C57BL/6J males between 60 and 140 days of age to assess their fecundity (body weight measurements were not obtained from females over this time because of pregnancies). Cages were checked daily for the presence of pups, and the date and size of the litter were recorded before pups were removed and culled. Male puberty progression was assessed visually based on the date of separation of the prepuce from the glans penis. From 21 days of age, all male mice were checked daily for preputial separation; assessment of male fecundity was not performed. At the end of the experiment, the animals were fasted overnight to reduce the concentration of endogenous circulating leptin and given injections of recombinant leptin (1 mg/kg, subcutaneously; National Hormone and Peptide Program) at 9:00 Am. Two hours after injection, they were anesthetized with sodium pentobarbital (240 mg/kg, IP) and transcardially perfused with 4% paraformaldehyde in 0.1 M PBS, pH 7.4. Visceral and subcutaneous fat masses were measured at this time. Coronal (30-μm thick) sections were cut from bregma +0.62 mm to −2.80 mm for each brain on a sliding microtome to be used for immunohistochemical staining. To visualize leptin-responsive cells in PACAP-LepR KO and PACAP-LepR rescue mice and their respective control groups, immunohistochemical labeling of phosphorylated signal transducer and activator of transcription 3 (pSTAT3) was performed. Antigen retrieval was performed by incubating for 15 minutes in 1 mM EDTA, pH 8.0, at 90 °C. To quench endogenous peroxidase activity, the tissues were incubated in 1% H₂O₂ for 30 minutes. Sections were incubated for 24 hours in the primary antibody, monoclonal rabbit anti-pSTAT3 (Tyr705, D3A7 XP; 1:1000 dilution, Cell Signaling Technology, RRID:AB_2491009). Tissue was then incubated for 1 hour in the secondary antibody, biotinylated goat anti-rabbit Ig (1:1000 dilution, Vector Laboratories). The signal was amplified by incubating in Vector Elite avidin–biotin peroxidase (Vector Laboratories) and was stained in diaminobenzidine solution to visualize pSTAT3 immunoreactivity. Omission of the primary antibody resulted in a complete absence of staining, and replacement of the leptin challenge with saline before perfusion resulted in very few stained cells. Stained cells in the ventromedial preoptic nucleus (VMPO), arcuate nucleus (Arc), ventromedial hypothalamic nucleus (VMH), and PMV (at least 3 sections per area from each animal) were counted.

Experiment 2: Are Leptin Actions on PACAP Neurons Sufficient for Normal Body Weight, Adiposity, Puberty Onset, and Reproductive Cycles?

To test the sufficiency of leptin signaling in PACAP neurons for puberty onset and subsequent fertility, we generated mice with specific rescue of LepR only in PACAP neurons (PACAP-LepR rescue mice). LepR^loxTB/loxTB and LepR^wt/wt littermates were used for the 2 control groups (referred to as LepR-null and LepR-intact controls, respectively). The LepR-null control group has previously been shown to be essentially infertile (27) and was included as a baseline reference against which any improvement of fertility could be compared in the PACAP-LepR rescue group. In male and female PACAP-LepR rescue, LepR-null, and LepR-intact control mice, body weight, adiposity, puberty onset, and estrous cycles were assessed as described for experiment 1. To assess the fecundity of PACAP-LepR rescue mice in comparison to LepR-null and LepR-intact controls, adult males and females were paired with reproductively capable mates for 35 days. At the end of the experiment, mice were perfused, visceral and subcutaneous fat masses weighed, and the brains processed for pSTAT3 as described previously.

Experiment 3: Is ERα Signaling in PACAP Neurons Required for Normal Body Weight and Puberty Onset?

PACAP-ERα KO and ERα^fl/fl littermate controls were used to evaluate the requirement of ERα signaling in PACAP neurons for body weight regulation and puberty onset. Body weight and puberty onset were assessed as described in experiment 1. In situ hybridization histochemistry using an RNAscope kit (Advanced Cell Diagnostics USA) was used to confirm ERα (Esr1 RNA) deletion in PACAP (Adcyap1 RNA) neurons. Brains from PACAP-ERα KO and controls mice (n = 4-5 per group) were collected and immediately frozen. Brains were embedded in optimal cutting temperature compound. Frozen sections were then cut into 3 identical sets of 16-μM coronal sections containing the paraventricular nucleus (selected because it strongly expresses both Esr1 and Adcyap1) using a cryostat (Leica CM1950) and mounted onto Superfrost slides and stored at −80 °C. Commercially available RNA probes for Esr1 (Mm-Esr1-C2; #478201-C2; Esr1 probe target region 678-1723bp; accession number NM_007956.5) and Adcyap1 (Mm-Adcyap1; #405911; Adcyap1 probe target region 676-1859bp; accession number NM_009625.2), and a duplex mouse reagent kit (#322436) were purchased from Advanced Cell Diagnostics and used according to the manufacturer's instructions. At least 3 sections between −0.6 to −1.1 mm rostral to bregma were averaged for each mouse.

Statistical Analysis

Values are presented as mean ± standard error of the mean. Differences were considered significant at P < .05. In experiment 1, unpaired Student t tests were used to identify significant differences between control and PACAP-LepR KO animals when group sizes were greater than n = 10. In experiment 2, 1-way ANOVA followed by the post hoc Holm–Sidak test was used to identify significant differences between the LepR-null, PACAP-LepR rescue, and LepR-intact groups when group sizes were greater than n = 10. For smaller sample sizes, the nonparametric Kruskal–Wallis H test was used, followed by the Dunn post hoc test. Body weight data were analyzed using a repeated-measures 2-way ANOVA followed by the Holm–Sidak test or a mixed-model analysis if there were any missing data points. All statistical analyses were performed using GraphPad PRISM 9, GraphPad Software Inc.

Results

Assessment of Hypothalamic PACAP Expression Across Postnatal Development

As seen in Fig. 1, PACAP-Cre-GFP was abundantly expressed in the paraventricular hypothalamic nucleus (PVH), VMH, and PMV in pubertal (PND33) mice. Loosely scattered cell populations were observed in the suprachiasmatic nucleus, retrochiasmatic nucleus, and dorsomedial hypothalamus. Almost no stained cells were observed in the Arc (Fig. 1). This PACAP distribution is in agreement with that reported previously for adult male PACAP:EGFP (14) and adult female PACAP-Cre:GFP (15) mouse lines. In PND0 mice, PACAP-Cre-GFP expression was markedly less abundant than in PND33 mice in all hypothalamic regions (Fig. 1). This postnatal developmental increase in hypothalamic PACAP cell numbers confirms that the PACAP-Cre induced receptor knockout or rescue in floxed mouse lines is unlikely to be affecting cells that later cease to produce PACAP.

Figure 1.

Postnatal development of PACAP/Adcyap1 expression in the hypothalamus. Representative images from female mice from brains collected on the day of birth (top row) or on the day of first estrus (postnatal day 35) showing PACAP-Cre-GFP expression throughout hypothalamic regions. Numerous PACAP-GFP cells were evident in the paraventricular nucleus (PVH), suprachiasmatic and retrochiasmatic nucleus, ventromedial nucleus (VMH), and ventral premammillary nucleus (PMV). PACAP-GFP abundance generally increased across postnatal development; this was particularly evident in the PVH, VMH, and PMV. Images are representative of 2 to 3 mice collected at each time point. Scale bars indicate 200 μm.

Experiment 1: Are Leptin Actions on PACAP Neurons Required for Normal Body Weight, Adiposity, Puberty Onset, and Fertility?

Phosphorylated STAT3 immunohistochemistry was used to detect the presence of leptin-induced pSTAT3 signaling, which is a functional indicator of leptin responsiveness, to validate that the PACAP-LepR KO mice did indeed exhibit a reduction in leptin responsiveness in hypothalamic nuclei known to express LepR and PACAP, including the VMPO, VMH, and PMV. Counting of pSTAT3 immunolabeled cells was also performed in the Arc as an internal negative control because the Arc expresses LepR but not PACAP. The PVH was not counted, despite the dense PACAP expression in this region, as leptin receptors are rarely expressed in the rodent PVH (28, 29). As seen in Fig. 2A, a significant reduction in leptin-induced pSTAT3 was observed in the PACAP-LepR KO mice in comparison to control mice in all the regions known to widely express both LepR and PACAP (VMPO, VMH, and PMV). As expected, no differences in leptin-induced pSTAT3 immunoreactivity were observed between PACAP-LepR KO and control mice in the Arc. Representative examples of pSTAT3 staining in mediobasal hypothalamic nuclei are shown in Fig. 2B.

Figure 2.

PACAP-LepR KO mice exhibit reduced leptin-induced pSTAT3 signaling compared with control mice in brain regions with PACAP-LepR colocalization. (A) Numbers of recombinant leptin (1 mg/kg subcutaneously)-induced pSTAT3 immunoreactive cells per section in hypothalamic regions of PACAP-LepR KO vs control mice (n = 2-4 mice per group; *P < .05, analyzed using Student unpaired t tests). (B) Representative examples from control and PACAP-LepR KO mice showing leptin-induced pSTAT3 in mediobasal hypothalamic regions. Arc, arcuate nucleus; PMV, ventral premammillary nucleus; VMH, ventromedial hypothalamus; VMPO, ventromedial preoptic area. Scale bars indicate 200 μm.

As mentioned previously, leptin has been shown to modulate PACAP expression, and, in turn, PACAP has been shown to reduce feeding behavior and increase energy expenditure. We were therefore interested in determining whether the absence of leptin signaling in PACAP neurons would result in an increased body weight phenotype. As seen in Fig. 3, no differences in postweaning body weight or adiposity were observed in either male (Fig. 3A-B) or female (Fig. 3C-D) PACAP-LepR KO mice in comparison to control mice, suggesting leptin-LepR signaling in PACAP-expressing cells does not play a critical role in the regulation of energy homeostasis.

Figure 3.

PACAP-LepR KO mice exhibit normal metabolic function (n = 8-11 mice per group). No differences in male or female weight gain (A, C) or subcutaneous and visceral adiposity (B, D) were observed between PACAP-LepR KO and control littermates. Body weight gain data were analyzed using 2-way ANOVA and adiposity data were measured using unpaired Student t tests.

Female mice exhibiting PACAP deficiency either globally or in LepR-expressing cells or in the PMV have delayed puberty onset, dysregulated estrous cyclicity, and reduced fecundity (10, 15), highlighting that PACAP release from LepR-expressing cells in the PMV plays a critical role in regulating female reproductive function. We therefore wanted to know whether the absence of leptin-induced PACAP release would recapitulate the reproductive phenotype seen in PMV-PACAP-deficient female mice, or whether other activators of PACAP expression could sufficiently compensate for the absence of leptin-induced PACAP release. As seen in Fig. 4A-B, female PACAP-LepR KO mice exhibited delayed vaginal opening (P < .01) and first estrus (P < .05). This delay in puberty onset is consistent with what was observed in female mice exhibiting PACAP deficiency from the PMV- or LepR-expressing cells (15). However, in contrast to PMV- or LepR-PACAP-deficient mice, PACAP-LepR KO mice exhibited normal estrous cyclicity and fecundity (Fig. 4C, E-F). We also wanted to assess the impact of PACAP-specific LepR KO on male reproductive function. Unlike the females, PACAP-LepR KO mice exhibited normal puberty onset because no differences in the age of preputial separation were observed between male PACAP-LepR KO and control mice (Fig. 4D).

Figure 4.

Female PACAP-LepR KO mice exhibit delayed vaginal opening (A) and first estrus (B) but normal estrous cyclicity (C) compared with LepR intact littermate controls. No differences in preputial separation (D) were observed between PACAP-LepR KO and control mice. Last, no differences in female fecundity (E) or litter size (F) were observed between PACAP-LepR KO and control mice (n = 9-12 mice per group). *P < .05, **P < .01, analyzed using 2-way ANOVA (cycle stage frequency and female fecundity) and unpaired Student t tests.

Experiment 2: Are Leptin Actions on PACAP Neurons Sufficient for Normal Body Weight, Adiposity, Puberty Onset, and Estrous Cyclicity?

To test the degree to which leptin's critical actions on energy homeostasis and reproductive function can be carried out when leptin signaling is restricted to PACAP-expressing cells, we assessed body weight, adiposity, and estrous cycles in mice expressing LepRb exclusively in PACAP-expressing cells (PACAP-LepR rescue mice) in comparison to LepR-intact and LepR-null control mice. We first validated these mice using leptin-induced pSTAT3, and as seen in Fig. 5A, LepR-rescue mice exhibited restored leptin-induced pSTAT3 signaling in regions with high PACAP + LepR coexpression. No differences in leptin-induced pSTAT3 were observed between LepR-intact and LepR-rescue mice in the VMPO and VMH, and LepR-rescue mice exhibited significantly increased leptin signaling vs LepR-null mice in the VMPO, VMH (P < .001), and PMV (P < .01). In the Arc, where PACAP neurons are rare, pSTAT3 cell counts remained low in LepR rescue mice relative to LepR intact control mice (P < .001). Representative examples of pSTAT3 staining in mediobasal hypothalamic nuclei are shown in Fig. 5B. However, as seen in Fig. 6B-E, this restored PACAP LepR signaling did not rescue the obese phenotype in males and only slightly reduced obesity in females. There were significant treatment main effects on bodyweight (males: F_(1,16) = 285.8; P < .001; females: F_(2,21) = 232.4; P < .0001) and adiposity (males: subcutaneous fat F_(2,10) = 24.42; P = .0001 and visceral fat F_(2,10) = 255.58; P = .0001; females: subcutaneous fat F_(2,21) = 45.32; P < .0001 and visceral fat F_(2,21) = 68.80; P < .0001). Both male and female PACAP-LepR rescue mice exhibited significantly increased body weight gain and adiposity in comparison to LepR-intact control mice, as did LepR-null mice (P < .001), highlighting that leptin signaling restricted to PACAP-expressing cells cannot markedly rescue leptin's central actions on the regulation of energy homeostasis. Indeed, male PACAP-LepR rescue mice exhibited a small but significant increase in body weight compared with LepR-null mice (P < .05) over the first 8 weeks of age, but not thereafter (Fig. 6B). In contrast, female PACAP-LepR rescue mice exhibited a significant decrease in body weight gain (P < .01 at 10 weeks of age) and in subcutaneous (but not visceral) fat mass (P < .05), compared with LepR-null mice (Fig. 6D-E), suggesting a small, sex-specific role for leptin acting via PACAP neurons in body weight regulation.

Figure 5.

PACAP-LepR rescue mice have restored leptin-induced pSTAT3 signaling compared with LepR null mice in brain regions with PACAP-LepR colocalization. (A) Numbers of recombinant leptin (1 mg/kg subcutaneously)-induced pSTAT3 immunoreactive cells per section in hypothalamic regions (n = 2-5 mice per group). (B) Representative examples from LepR Intact, LepR Null, and PACAP-LepR Rescue mice showing leptin-induced pSTAT3 in mediobasal hypothalamic regions. Arc, arcuate nucleus; PMV, ventral premammillary nucleus; VMH, ventromedial hypothalamus; VMPO, ventromedial preoptic area. Scale bars indicate 200 μm. **P < .01, ***P < .001, analyzed using 1-way ANOVA for each brain region.

Figure 6.

PACAP-LepR rescue mice retain an obese phenotype (females, n = 7-10 mice per group; males, n = 3-9 mice per group). Male PACAP-LepR Rescue mice exhibited slightly exacerbated weight gain over the first 2 months of life (A), but comparable adiposity (B) compared with LepR Null mice. Female PACAP-LepR Rescue mice exhibited attenuated weight gain (C) and reduced subcutaneous adiposity (D) compared with LepR Null mice. Weight gain analyzed using 2-way ANOVA followed by the Holm–Sidak multiple comparison test. *P < .05, **P < .01 LepR Null vs LepR Rescue, ###P < .001 LepR Null and LepR Rescue vs LepR Intact, adiposity analyzed using 1-way ANOVA for each fat depot; ***P < .001.

To test whether leptin signaling exclusively in PACAP neurons is sufficient to rescue or partially improve the hypothalamic hypogonadism observed in LepR-null mice, we assessed puberty onset and estrous cyclicity in PACAP-LepR rescue mice compared with LepR-null and LepR-intact mice. There was a significant treatment main effects on age at vaginal opening (F_(2,22) = 21.51; P < .001). As seen in Fig. 7A, the delayed vaginal opening was not improved in female PACAP-LepR rescue mice compared with LepR-null mice (both P < .001 compared with LepR-intact controls). Although the female PACAP-LepR rescue mice and LepR-null mice did exhibit vaginal opening (Fig. 7A), no mice from either group ever exhibited estrus in vaginal cytology (Fig. 7B) and they remained completely acyclic (all cytology smears diestrus-like in appearance; Fig. 7D), suggesting leptin signaling exclusively in PACAP-expressing cells is unable to relay leptin's critical actions in the central control of reproductive function. In males, preputial separation was not detected in either the LepR-null or LepR rescue groups during the period of monitoring (Fig. 7C).

Figure 7.

PACAP-LepR rescue mice do not exhibit improved reproductive function. No improvements in puberty onset were observed in PACAP-LepR Rescue vs LepR Null mice as assessed by vaginal opening (A) and first estrus (B) in females (n = 7-10 mice per group) and by preputial separation (C) in males (n = 4-9 per group). No differences in estrous cyclicity (D) were observed in PACAP-LepR Rescue vs LepR Null mice. ***P < .001, analyzed using 1-way ANOVA followed by the Holm–Sidak multiple comparison test.

Experiment 3: Is ERα Signaling in PACAP Neurons Required for Normal Body Weight and Puberty Onset?

We were interested in determining if estrogenic influences on metabolic and reproductive function involved PACAP neurons and whether the sex differences we observed in female vs male PACAP-LepR KO mice were due in part to PACAP-estradiol signaling. To this end, we generated PACAP-ERα KO mice to investigate the effects of this knockout on metabolic and reproductive function. We first validated the PACAP-ERα KO mice using double-label RNAscope in situ hybridization histochemistry for Esr1 (ERα) and Adycap genes in the PVH because, unlike the PMV and VMH, this region expresses both ERα (30) and Adycap (20). As seen in Fig. 8, there was a significant reduction in PACAP + ERα colocalization in both male and female KO vs control mice (P < .001). However, female mice exhibited much higher PACAP + ERα colocalization in comparison to males, and in both sexes some apparent colocalization remained in the PACAP-ERα KO groups. We next assessed whether PACAP-ERα KO mice exhibited any differences in body weight and/or puberty onset. However, as seen in Fig. 9, no differences were observed in any parameters we measured.

Figure 8.

PACAP-ERα KO female (A) and male (B) mice (n = 4-5 mice per group) exhibit significantly reduced Esr1 (red labeling) and Adcyap (blue labeling) RNA coexpression in the paraventricular nucleus of the hypothalamus (PVN), as seen in the representative photomicrographs (C-F). Dashed line insets (150 µm wide) show magnified cells. Scale bar, 100 µm. **P < .01, analyzed using a mixed effects model.

Figure 9.

PACAP-ERα KO female (A) and male (B) mice exhibit normal body weight gain, as well as normal pubertal timing, as assessed by vaginal opening (C) and first estrus (D) in females and by preputial separation (E) in males compared with control mice (n = 5-6 mice per group). Body weight gain data were analyzed using 2-way ANOVA and adiposity data were measured using unpaired Student t tests.

Discussion

PACAP is widely distributed both centrally and peripherally and exerts a multitude of actions. Importantly, global knockout of PACAP or its receptor results in metabolic disorders and reduced fertility in females (9, 10, 31). Complicating the potential actions of this peptide, central neonatal PACAP administration delays puberty onset and can exert both stimulatory and inhibitory actions in rats (32-34). This neonatal effect may not reflect endogenous PACAP actions at this developmental stage because we show that hypothalamic PACAP neurons are underdeveloped in newborn mice and central injections on day 7 after birth were without effect (34). We generated PACAP-specific LepR KO mice, as well as PACAP-specific LepR rescue mice, to investigate whether leptin signaling in PACAP cells is either critically required and/or sufficient, respectively, for normal pubertal timing and fertility. We observed a 70% reduction in PMV leptin-induced pSTAT3 staining in our PACAP-LepR KO mice and a similarly pronounced rescue of PMV pSTAT3 staining in our PACAP-LepR Rescue mice (to 50% of LepR Intact mouse values). This matches well with a previous report (15) showing that 70% of PMV leptin-induced pSTAT3 cells colocalize with PACAP in mice. Leptin-induced pSTAT3 also colocalizes with most PACAP cells in the VMH (15), which again matches our KO and rescue data for this region. In contrast, PACAP is strongly expressed in the PVH but LepR expression in this region is minimal, whereas the ARC is heavily populated with LepR cells (29) but does not appear to express PACAP.

We were interested in investigating PACAP's role in the metabolic control of puberty onset and fertility and determining whether PACAP neurons mediate leptin's permissive effects on the hypothalamic-pituitary-gonadal (HPG) axis. We showed that female mice exhibiting PACAP-specific deletion of LepR signaling displayed delayed puberty onset, but normal estrous cyclicity and fertility, in comparison to their LepR-intact control littermates. These data suggest leptin-PACAP signaling plays a key role in the metabolic control of pubertal timing in females but is not critically required for adulthood reproductive function. In contrast, restoring LepR signaling exclusively in PACAP neurons in otherwise LepR null mice was unable to rescue or improve pubertal timing or any other metabolic or reproductive parameter assessed. These data suggest leptin-PACAP signaling alone cannot sufficiently relay leptin's permissive effects on the neuroendocrine control of metabolic or reproductive function. In contrast to this finding for PACAP neurons, AgRP neuronal LepR signaling has been shown to be sufficient to permit fertility in otherwise LepR null mice (35). Given that LepR expression is more widespread in PACAP-LepR rescue mice than AgRP-LepR rescue mice, the overall number of cells with LepR expression restored appears to matter less than the identity of the cells themselves. In this regard, we have previously shown that deleting LepR from glutamatergic neurons does not affect pubertal timing or fertility, which make the pubertal delay in response to PACAP-LepR deletion somewhat unexpected considering the VMH and PMV populations of PACAP neurons are also largely glutamatergic. However, there remains a significant number of PMV glutamatergic neurons that do not express PACAP (15), which would not have been targeted in the current LepR KO study. These neurons may have contributed to masking a puberty phenotype in the glutamate neuron-LepR KO mice. The present data reveal an effect a PACAP neuron-specific action of leptin that was not observed in the more widespread glutamatergic-LepR knockout study.

Interestingly, only female mice exhibited a delay in puberty onset in the absence of PACAP-LepR signaling, suggesting leptin actions via PACAP neurons are not critically involved in the timing of male puberty onset. This result is similar to the female-specific pubertal delay that has been reported in AgRP-LepR KO mice (35). The energetic cost of reproduction is much higher for females than males, which may underlie this observed sex difference. Leptin is a metabolic signal that conveys information centrally about peripheral energy status, which is not as critically important for successful male reproduction. PACAP may be acting through kisspeptin to differentially influence pubertal timing because Kiss1 mRNA levels are increased in response to PACAP38 administration in hypothalamic cell lines (36), and the kisspeptin system is sexually dimorphic in mice (37). In further support of PACAP acting via the kisspeptin system to influence HPG axis function, it was recently shown that PACAP knockout female mice exhibit decreased Kiss1 mRNA levels in the Arc (10), and PMV PACAP-expressing neurons project to, and make direct contact with, kisspeptin neurons in the Arc and anterioventral periventricular nucleus (15).

We also observed a sex difference in the body weights of PACAP-LepR rescue mice. Female mice with LepR expression restricted exclusively to PACAP cells exhibited a mild but significantly attenuated obesity phenotype compared to LepR null mice that became more apparent with increasing age, whereas their male counterparts exhibited a slightly exacerbated obesity phenotype during early adult life compared with LepR null mice. Whether these differences were due to changes in central or peripheral PACAP activity remains unclear. PACAP has been shown to influence growth hormone, prolactin, corticotropin, and gonadotropin release (as reviewed elsewhere (38)), which could all influence energy homeostasis, and it is thus difficult to identify the precise mechanism(s) underlying the sex differences observed in body weight between male and female PACAP-LepR rescue vs LepR null mice. Leptin has been shown, via PACAP neurons in the VMH, to control metabolic (39) function (40). Blockade of PACAP receptors in the VMH prevents the ability of leptin in this region to induce hypophagia, weight loss, and thermogenesis (7). These experiments used male rats, whereas the attenuated obese phenotype observed in our study used PACAP LepR rescue female mice. Despite these differences, the VMH remains a likely locus where leptin-dependent metabolic actions via PACAP neurons occur.

The mechanisms whereby PACAP signaling modulates GnRH release are likewise unclear. A comprehensive review discusses the many possible direct and indirect mechanisms whereby central PACAP may exert its influence on GnRH release, while also addressing the possible mechanisms whereby peripheral PACAP may modulate the HPG axis (38). The key leptin-responsive PACAP neurons mediating this effect may be located in the PMV, which, as noted previously, can act on Arc and anterioventral periventricular nucleus kisspeptin neurons. PACAP knockout specifically from the PMV or from all LepRb-expressing cells led to delayed puberty onset in female, but not male, mice (15), which closely matches the findings of the present study in which LepRb was deleted from PACAP cells.

Interestingly, PACAP is also an important modulator of stress signaling, acting both centrally and peripherally on many stress-related systems (41). Stress system activation also inhibits reproductive function (42); it is therefore a possibility that PACAP also influences HPG axis function indirectly via its effects on the stress axes. Given its wide-ranging actions, PACAP appears to serve as an integrative signal that coordinates an appropriate physiological response to environmental conditions by affecting many systems.

PACAP has been postulated to interact with sex hormones to contribute to sex differences in stress-related disease. For example, circulating gonadal hormones are thought to differentially regulate PACAP neuronal activity in males vs females to influence anxiety-like behavior, such that high levels of estrogens were shown to protect female rodents from PACAP-induced startle (43). In the PVH, where both estrogen receptor subtypes are expressed (30), Adycap expression has been shown to increase just before the preovulatory GnRH/LH surge in rats (20). The present data show that half of PVH PACAP neurons coexpress the ERα gene Esr1 in female mice, and 10% to 15% of PVH PACAP neurons coexpress this receptor in males. To further explore the role of sex hormones in the control of PACAP signaling, and to determine whether the sex differences in puberty onset we observed in PACAP-LepR KO mice were in part from estradiol-dependent regulation of PACAP, we generated mice exhibiting PACAP-specific deletion of ERα. However, we did not observe any differences in male or female body weight gain or timing of puberty onset in the PACAP-ERα KO vs control mice. This may be because some PACAP + ERα coexpression remained in the PACAP- ERα KO mice, which may have enabled sufficient estradiol-PACAP signaling, despite the significant deletion of ERα from PACAP cells. Alternatively, the residual PACAP + ERα coexpression in KO mice might be an artifact from background ESR1 signal, ESR1 labeling of overlying non-PACAP cells in the tissue sections, or interactions of the ESR1 probe with nonexcised regions of ESR1 mRNA. Notwithstanding the possibility of residual PACAP + ERα coexpression in knockout mice, these data suggest ERα-mediated regulation of PACAP is not critically important in metabolic or reproductive control, at least under normal, unstressed conditions.

In summary, PACAP expression is widely distributed in the hypothalamus with increasing expression between birth and puberty. We showed that LepR signaling in PACAP neurons plays a critical role in the timing of female, but not male, puberty onset, but is not critically required for normal estrous cyclicity or fertility. Rescuing LepR-PACAP signaling in otherwise LepR-deficient mice was unable to rescue the reproductive deficits observed in LepR null mice, but it led to a marginal improvement in body weight and adiposity in females. Finally, PACAP-specific ERα KO did not lead to any changes in body weight or puberty onset compared with control mice. Taken together with recent findings that leptin-responsive PMV PACAP appear to influence fertility via actions on kisspeptin neurons whereas leptin enhances VMH PACAP actions on metabolic function, our data help to build an understanding of how leptin acts via PACAP neurons to modulate these physiological functions.

Abbreviations

τGFP

tau green fluorescent protein

Arc

arcuate nucleus

ERα

estrogen receptor alpha

HPG

hypothalamic-pituitary-gonadal

KO

knockout

LepR

leptin receptor

LepRb

signaling form of the leptin receptor

PACAP

pituitary adenylate cyclase-activating polypeptide

PMV

ventral premamillary nucleus

PND

postnatal day

pSTAT3

phosphorylated signal transducer and activator of transcription 3

PVH

paraventricular hypothalamic nucleus

VMH

ventromedial nucleus

VMPO

ventromedial preoptic nucleus

Funding

The present study was funded by the Marsden Fund.

Disclosures

The authors have nothing to disclose.

Data Availability

Some or all datasets generated during and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.