In vitro differentiation of the hypothalamic KNDy neuron, a master regulator for reproduction, from mouse embryonic stem cells

Abstract

Background

Hypothalamic kisspeptin neurons are master regulators of reproduction in mammals, including humans and rodents, via direct stimulation of gonadotropin-releasing hormone (GnRH) neurons. Therefore, they play a central role in puberty onset and gonadal function in mammals. Arcuate kisspeptin neurons express both neurokinin B (NKB) and dynorphin A (Dyn). Thus, these neurons are also referred to as KNDy neurons, and accumulating evidence suggests that KNDy neurons are responsible for tonic GnRH/gonadotropin release and the consequent folliculogenesis/steroidogenesis. Notably, mutations in genes encoding kisspeptin (KISS1), kisspeptin receptor (GPR54), or NKB (TAC3) in humans, or deletion of these genes in rodents, cause hypogonadotropic hypogonadism. Therefore, cellular models of KNDy neurons are useful for elucidating the pathogenesis of reproductive disorders and for developing novel therapies for these disorders. Here, we established a method to differentiate KNDy neurons from mouse embryonic stem cells (mESCs).

Methods

mESCs were incubated in culture medium containing agents that enhance Sonic Hedgehog signaling, a ventralizing signal, to induce the differentiation of ventral hypothalamic organoids. Kisspeptin-expressing cells were generated by adding a Notch inhibitor, dispersing the aggregates, and transferring them into a two-dimensional culture system. Kisspeptin-, NKB-, and Dyn-expressing cells were identified by immunohistochemistry, and kisspeptin secretion into the culture medium was quantified by enzyme-linked immunosorbent assay.

Results

Immunoreactivity for kisspeptin, NKB, and Dyn was detected in differentiated organoids derived from mESCs. Furthermore, kisspeptin secretion was evident in culture supernatants of differentiated KNDy neurons.

Conclusions

This is the first report demonstrating the differentiation of KNDy neurons from pluripotent stem cells, which can be applied to create cellular models for various diseases caused by KNDy peptide deficiency and may contribute to the development of novel therapeutic approaches and understanding cellular mechanisms regulating mammalian reproduction.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12958-026-01536-x.

Background

Hypothalamic kisspeptin neurons are master regulators of the hypothalamic-pituitary-gonadal (HPG) axis via the direct stimulation of gonadotropin-releasing hormone (GnRH) neurons in mammals. They play critical roles in reproductive functions, including puberty onset and ovarian function in mammals, including humans. The deletion or mutation of the kisspeptin gene (KISS1 in primates and Kiss1 in rodents) or its receptor (GPR54/Gpr54) leads to pubertal failure [1–5]. Specifically, kisspeptin neurons in the arcuate nucleus (ARC) of the mediobasal hypothalamus are also called KNDy neurons, as they coexpress neurokinin B (NKB) and dynorphin A (Dyn). KNDy neurons are considered GnRH pulse generator [6–10], regulating pulsatile GnRH/gonadotropin secretion, and subsequently enhancing folliculogenesis and steroidogenesis in rodents and ruminants [9–11]. Furthermore, functional hypothalamic amenorrhea (FHA) has been studied in experimental animal models in which malnutrition or stress suppresses KNDy neurons and GnRH/luteinizing hormone (LH) pulses [10, 12–17]. Another population of kisspeptin neurons localized in the anterior hypothalamus, such as the anteroventral periventricular nucleus/preoptic area, is considered the GnRH surge generator that triggers LH surge and ovulation [11, 18, 19].

Dysfunction of KNDy neurons is implicated in various human diseases: Loss-of-function mutations in KISS1, GPR54, TAC3 (encoding NKB), or TACR3 (encoding neurokinin 3 receptor [NK3R]) cause hypogonadotropic hypogonadism and pubertal failure [1, 4, 20]. Conversely, gain-of-function mutations in these genes are associated with central precocious puberty [21–23]. Furthermore, KNDy neurons are involved in various disorders affecting women’s health across the lifespan, including FHA, vasomotor symptoms (VMS), and polycystic ovary syndrome (PCOS). In addition, late-onset hypogonadism (LOH) syndrome, the male counterpart of menopause, has been recognized [24].

Establishing in vitro models of KNDy neurons is valuable for elucidating the pathophysiology of KNDy-related diseases and for developing new therapeutic strategies. Previously, the mHypoA-55 cell line, derived from immortalized mouse hypothalamic cells, was established as a model for KNDy neurons [23]. However, to date, in vitro KNDy neuronal models using pluripotent stem cells such as embryonic stem cells (ESCs) or induced pluripotent stem cells (iPSCs) are yet to be established despite their potential to overcome the disadvantages of immortalized cell models. We previously established a three-dimensional differentiation culture method for ESCs, named serum-free floating culture of embryoid body-like aggregates with quick reaggregation (SFEBq) [25], by which various ectodermal organoids were differentiated from ESCs. Using this method, Wataya et al. successfully differentiated vasopressin neurons, mainly localized in the dorsal hypothalamus in vivo, from mouse ESCs (mESCs). Furthermore, the ventromedial nucleus of the hypothalamus model was differentiated by adding sonic hedgehog (Shh), which promotes ventral nerve differentiation during embryonic development [26].

This study aimed to establish an in vitro differentiation method for KNDy neurons from mESCs. First, we cultured mESCs using a modified Wataya’s method with enhanced Shh signaling to promote differentiation into ARC neurons, such as proopiomelanocortin (POMC) and agouti-related peptide (AgRP) neurons. Next, we inhibited Notch signaling to induce the differentiation of KNDy neurons based on previous reports [27, 28]. Furthermore, we examined the expression of kisspeptin, NKB, and Dyn peptides in the in vitro mouse KNDy neuron model using immunohistochemistry. Finally, we confirmed the secretion of kisspeptin from differentiated KNDy neurons.

Methods

mESC maintenance culture and hypothalamic neural differentiation

We used Rax-enhanced green fluorescent protein (EGFP) knock-in male mESCs generated from the EB5 parental line (CVCL_J648; [29, 30]) and female mESCs, the BRC6 (feeder-free) line (CVCL_RH44; [31]), which do not carry EGFP labeling. Both the cell lines were provided by the RIKEN BioResource Research Center (Tsukuba, Japan). Undifferentiated mESCs were maintained on gelatin-coated dishes at 37 °C in 5% CO₂ and passaged every 2–3 days [25, 26, 32]. The maintenance medium for EB5 was Glasgow modified Eagle medium (11710-035; Gibco, Waltham, MA) supplemented with 1% fetal bovine serum (FBS) (172012; Sigma–Aldrich, St. Louis, MO), 10% KnockOut serum replacement (KSR) (1386120; Invitrogen, San Diego, CA), 0.1 mM nonessential amino acids (11140-050; Gibco), 0.1 mM sodium pyruvate (25030-081; Gibco), 0.1 mM 2-mercaptoethanol (131-14572; FUJIFILM Wako Pure Chemical, Osaka, Japan), 2,000 U/mL of recombinant mouse leukemia inhibitory factor (ESG1107; EMD Millipore, Bedford, MA), and 20 mg/mL blasticidin S (KK400; Kaken Pharmaceutical, Tokyo, Japan). The maintenance medium for BRC6(feeder-free) was Dulbecco’s modified Eagle medium (DMEM) (11995-065; Gibco) supplemented with 10% KSR, 0.1 mM nonessential amino acids, 0.1 mM 2-mercaptoethanol, and 1,000 U/mL of recombinant mouse leukemia inhibitory factor.

For hypothalamic induction, mESCs were dissociated into single cells using TrypLE Express (12605-010; Gibco) for EB5 or 0.25% Trypsin EDTA (25200056; Gibco) for BRC6(feeder-free) and quickly reaggregated in a 96-well low-cell-adhesion plate with U-bottomed wells (174929; Thermo Fisher Scientific, Waltham, MA; 3,000 cells per 100 mL per well). mESCs were cultured in a growth factor-free chemically defined medium (gfCDM) comprising a 1:1 mixture of Iscove modified Dulbecco’s medium GlutaMAX (31980-030; Gibco) and F-12 GlutaMAX (31765-035; Gibco) containing 5 mg/mL purified bovine serum albumin (A3156; Sigma–Aldrich), 1% chemically defined lipid concentrate (11905-031; Gibco), and 450 mM monothioglycerol (M6145; Sigma–Aldrich) for 9 days. On day 10, DMEM/F-12 supplemented with glucose, N-2 supplement, and B27 supplement (DFNB) comprising DMEM/F-12 (D8900; Sigma–Aldrich) containing 3.85 g/L glucose (07-0680-5; Sigma–Aldrich), 1.2 g/L sodium hydrogen carbonate (28-1850-5; Sigma–Aldrich), 50 U/mL penicillin-streptomycin (15140-122; Gibco), 1% N-2 supplement (175020-01; Gibco), 2% B27 supplement (125870-10; Gibco), and 10 ng/mL ciliary neurotrophic factor (257-NT; R&D Systems, Minneapolis, MN), were added to each well (100 µL per well). On day 13, the aggregates were transferred from the 96-well plate to a collagen-coated Millicell^® culture insert (MCSP06H48; EMD Millipore) and cultured in DFNB. The medium was changed every other day until subsequent experiments were performed.

To induce differentiation into the ventral hypothalamus, cultured mESCs were treated from day 4 onward with or without 30 nM recombinant mouse Shh N-terminal protein (461-SH; R&D Systems), referred to as Wataya’s method [26], or with 10, 100, or 300 nM Smoothened agonist (SAG, a chemical Shh agonist) (ALX-270-426; Enzo Life Sciences, Farmingdale, NY).

To stabilize the aggregates, mESCs were treated with 10 µM Y-27632, a Rho-associated coiled-coil forming kinase inhibitor (034-24024; FUJIFILM Wako Pure Chemical), 30 min before the single-cell dissociation on Day 0. Cells were treated with 10 µM Y-27632 and 0.5 or 0.1% KSR in 96-well plates for 4 days. On day 4, to remove Y-27632, 50 µL of supernatant was aspirated from each well and replaced with gfCDM containing 1% KSR. This procedure was repeated four times.

To induce differentiation into KNDy neurons, the cells were treated with 10 µM N-[N-(3, 5-difluorophenacetyl)-L-alanyl]-S-phenylglycine t-butyl ester (DAPT), a Notch signaling inhibitor (049-33583; Wako), from day 10 for 1, 3, 5, or 7 days.

Dissociated culture of mESC aggregates

To further promote KNDy neuronal differentiation, mESC aggregates were dispersed and transferred to an adhesive culture as previously described [33]. On days 41–43, 10 µM Y-27632 was added to the culture. On the following day, the aggregates on Millicell^® inserts were collected and dissociated into single cells using neuron dissociation solution S (297-78101; FUJIFILM Wako Pure Chemical), seeded on poly-D-lysine (PDL)-coated glass coverslips in 24-well plates (174930; Thermo Fisher Scientific) at a density of 50,000 cells/mL/well in dissociation medium, which comprises DFNB, 10% FBS (SFBM30-2537; Equitech-Bio, Kerrville, TX), recombinant human brain-derived neurotrophic factor (BDNF) (028-16451; Wako Pure Chemical), recombinant human neurotoropin-3 (146–09231; FUJIFILM Wako Pure Chemical) and LM22A-4, a BDNF mimic, (SML0848; Sigma–Aldrich), with 10 µM Y-27632, and cultured at 37 °C in a 5% CO2. The medium without Y-27632 was changed every other day.

On day 6 of the dissociated culture (corresponding to day 50 of the overall differentiation protocol), DFNB medium containing 100 nM 17β-estradiol (E2) was added to the dissociated cultures. Controls received DFNB medium only. The cells were collected 4 h later and used for RNA extraction and polymerase chain reaction (PCR), as described below. The concentration and duration of E2 treatment were selected based on previous in vitro studies demonstrating the negative feedback action of E2 on Kiss1 expression in the mHypoA-55 mouse immortalized KNDy neuron cell line [23, 34].

Immunohistochemistry for KNDy and other hypothalamic peptides

Fluorescence immunohistochemistry was performed on brains obtained from embryonic day 11.5 (E11.5) mouse embryos, postnatal day 14 (P14) mice, and 15-week-old female mice that had undergone bilateral ovariectomy, as well as on aggregates derived from mESCs and cells from dissociated cultures. Pregnant ICR mice (SLC: ICR) and adult virgin C57BL/6 mice (C57BL/6JJmsSlc) were purchased from Japan SLC, Inc. (Hamamatsu, Japan). Some pregnant mice were euthanized by cervical dislocation under isoflurane anesthesia at day 11.5 of pregnancy, and embryos were collected by laparotomy and fixed in 4% paraformaldehyde (PFA) for 3 h at 4℃. Other pregnant mice delivered at term. P14 mice were transcardially perfused with 4% PFA, and their brains were collected and subsequently fixed in 4% PFA overnight at 4℃. 13-week-old female mice were anesthetized with isoflurane, and bilateral ovariectomy was performed. At 15 weeks of age, the mice were transcardially perfused with 4% PFA, and their brains were dissected, followed by fixation in 4% PFA overnight at 4℃. All animal experiments were approved by the Animal Experimental Committee of the Nagoya University Graduate School of Medicine (M250004-003 and M250013-003). Animal care and use complied with institutional guidelines and relevant regulations. The aggregates were fixed in 4% PFA for 15 min. Cells on PDL-coated glass coverslips were fixed with 2% PFA for 10 min, followed by 4% PFA for 15 min. Aggregates were immersed in 20% sucrose (in 30% for brain) and embedded in optimal cutting temperature compound (4583; Sakura Finetek, Tokyo, Japan) and cut into 10-µm-thick sections using a cryostat (HM525, Thermo Fisher Scientific). Sections were permeabilized by washing thrice (15 min each) with 0.3% Triton X-100/phosphate-buffered saline (PBS), followed by three washes with Triton X-free PBS (5 min each). Subsequently, sections were blocked by incubation in 2% (w/v) dry skimmed milk/PBS for 1 h at 25 °C. Sections were then incubated overnight at 4 °C, with primary antibodies diluted in 2% dry skimmed milk/PBS. The next day, sections were washed thrice (15 min each) in 0.05% Tween 20/PBS and incubated with secondary antibodies and 4’,6-diamidino-2-phenylindole (DAPI) (D523; Dojindo, Kumamoto, Japan) diluted in 2% dry skimmed milk/PBS for 1.5 h at 25 °C. Subsequently, the sections were washed thrice (15 min each) in 0.05% Tween 20/PBS and mounted in Slow-Fade™ Diamond (S36972; Thermo Fisher Scientific). For cells on PDL-coated glass coverslips, ProLong™Diamond (P36970; Thermo Fisher Scientific) was used. The following primary antibodies were used: AgRP (GT15023; goat; 1:250; Neuromics, Edina, MN), copeptin (CPP; sc-7812; goat; 1:100, Santacruz, Dallas, TX), kisspeptin (C2; rabbit; 1:2,000 [Okamura, Yamamura, and Wakabayashi 2017]), NKB (NB300-201; rabbit; 1:1,000, Novus Biologicals, Centennial, CO), neurophysin II (NPII; MABN845; mouse; 1:100, Millipore), neuropeptide Y (NPY; ab30914; rabbit; 1:1,000; abcam, Cambridge, UK), NK2 homeobox1 (Nkx2.1; NCL-L-TTF-1; mouse; 1:100; Leica Biosystems, Nussloch, Germany), paired box 6 (Pax6; ab195045; rabbit; 1:350; abcam), POMC (H-02930; rabbit; 1:400; Phoenix Pharmaceutics, Burlingame, CA), POU domain, class 3, transcription factor 2 (Pou3f2; sc6029; goat; 1:100; Santacruz), prodynorphin (GP10110; guinea pig; 1:500, Neuromics), retina and anterior neural fold homeobox (Rx; M229; guinea pig; 1:1,000; TAKARA, Kusatsu, Japan), and Tubulin β3 (TUBB3; MMS-435P; mouse; 1:2,000; BioLegend, San Diego, CA). Positive controls for AgRP, kisspeptin, NKB, NPY, Nkx2.1, POMC, Pou3f2, prodynorphin, and Rx are shown in Additional file 1. For NPII and copeptin, we used previously validated antibodies, with positive-control staining in mouse brain tissue [33]. Secondary antibodies were used at 1:1,000 and included Alexa Fluor 488-labeled anti-goat immunoglobulin (Ig)G (A11055; donkey; Invitrogen), Alexa Fluor 647-labeled anti-goat IgG (A21447; donkey; Invitrogen), Alexa Fluor 488-labeled anti-guinea pig IgG (A11073; goat; Invitrogen), Alexa Fluor 647-labeled anti-guinea pig IgG (A21450; goat; Invitrogen), Alexa Fluor 546-labeled anti-mouse IgG (A10036; donkey; Invitrogen), Alexa Fluor 488-labeled anti-rabbit IgG (A21205; donkey; Invitrogen), Alexa Fluor 546-labeled anti-rabbit IgG (A10040; donkey; Invitrogen), and Alexa Fluor 488-labeled anti-mouse IgG (A21202; donkey; Invitrogen). For the costaining of kisspeptin and NKB, anti-NKB antibodies were labeled using the HiLyte Fluor™ 555 Labeling Kit-NH₂ (LK14; Dojindo).

Imaging of immunopositive-hypothalamic marker peptides in mESC-derived hypothalamic cells

Phase-contrast microscopy (BZ-X710; KEYENCE, Osaka, Japan) was used to observe the aggregates. Immunohistological imaging of hypothalamic marker peptides, such as AgRP, POMC, CPP, kisspeptin, NKB, Dyn, NPII, NPY, Nkx2.1 (a marker of the ventral hypothalamus), Pax6, Pou3f2, prodynorphin, and Rx, was conducted using an FV3000 laser scanning microscope (Evident, Tokyo, Japan). The area of the immunopositive regions was measured using cellSens software (Evident).

RT-PCR for Kiss1, Tac2 (encoding mouse NKB), and Pdyn (encoding prodynorphin) in the differentiated cells from mESCs

Total RNA was extracted from cultured cells on day 10 (negative control) using 10 aggregates per sample and on day 47 using one well of cells (seeded at 5,000 cells per well) as one sample, using an RNeasy kit (Qiagen), following the manufacturer’s protocol. Reverse transcription reaction with 1 µg of total RNA was performed using a first-strand cDNA synthesis kit, ReverTraAce^® qPCR RT Master Mix (TOYOBO, Osaka, Japan). The resulting cDNA was diluted 1:10, and PCR was performed using Blend Taq^® -Plus (TOYOBO) and Applied Biosystems^® 2720 Thermal Cycler (Thermo Fisher Scientific) following the manufacturer’s protocol. The protocol involved an initial denaturation at 98℃ for 120 s, denaturation at 98℃ for 10 s, annealing at 60℃ for 10 s, and extension at 68℃ for 30 s (45 cycles), followed by melting at 95℃ for 10 s, 65℃ for 60 s, and 97℃ for 1 s. The quantitative RT-PCR was performed in duplicates using KOD SYBR^® qPCR Mix (TOYOBO) in 96-well plates with 0.2 mL thin-walled PCR tubes using the LightCycler^® 96 system (Roche, Basel, Switzerland). Quantification was performed by calculating the ratio of gene of interest expression to that of glyceraldehyde-3-phosphate dehydrogenase (Gapdh) using the comparative C_t method. The specificity of Kiss1 qRT-PCR amplification was evaluated using melting-curve analysis. The qualitative PCR protocol involved an initial denaturation at 94℃ for 5 min, denaturation at 94℃ for 30 s, annealing at 60℃ for 30 s, and extension at 72℃ for 30s (40 cycles). The PCR products were electrophoresed on a 1.5% agarose gel to confirm amplicon specificity.

The primers and the size of the amplicon were as follows: mouse Gapdh (forward 5’ -

ATGAATACGGCTACAGCAACAGG − 3’; reverse 5’ -

CTGTTGCTCAGTGTCCTTGCTG-3’; 102 bps), mouse Kiss1 (forward 5’-

CCCAGAATGATCTCAATGGCTTCT − 3’; reverse 5’-

CTCTCTGCATACCGCGATTCCT − 3’; 161 bps), mouse Pdyn (forward 5’-

TGCAGTGAGGATTCAGGATG − 3’; reverse 5’- AGAGACCGTCAGGGTGAGAA − 

3’; 136 bps), and mouse Tac2 (forward 5’- GCTCCACAGCTTTGTCCTTC − 3’;

reverse 5’- GCTAGCCTTGCTCAGCACTT − 3’; 234 bps).

Measurement of in vitro kisspeptin secretion from the differentiated KNDy neurons

On day 47, the culture supernatant (500 µL) and dissociation medium (negative control) were collected, mixed with 4x volume of cooled acetone, incubated at -80 °C for 1 h, and centrifuged at 15,000 g for 20 min at 4 °C. After discarding the supernatant, the resulting pellet was kept on ice for 30 min and subsequently thawed with 140 µL of PBS. Kisspeptin concentrations were measured using a kisspeptin 1 enzyme-linked immunosorbent assay Kit (EKU05505; Biomatik, Kitchener, Ontario, Canada) following the manufacturer’s protocol.

Statistical analysis

Statistical analyses were performed using the IBM SPSS Statistics software (IBM, Armonk, NY, USA). The χ² test was used to compare aggregate formation rates between BRC6(feeder-free) treated with KSR and Y-27632 and the control group. One-way analysis of variance (ANOVA) followed by Tukey’s honest significant difference (HSD) test was used to assess differences in kisspeptin-immunopositive areas among the four groups treated with DAPT for 1, 3, 5, or 7 days. Two-tailed unpaired t-tests were used for two-group comparisons of Kiss1, Tac2, and Pdyn mRNA expression, and kisspeptin concentration in the culture supernatant. P < 0.05 was considered statistically significant.

Results

Differentiation of ventral hypothalamic neurons from male mESCs

Fig. 1A shows the culture protocol for inducing ventral hypothalamic differentiation of mESCs using Wataya’s method with enhanced Shh signaling, according to a previous study, showing that the concentration gradient of bone morphogenetic proteins secreted from the roof plate and Shh secreted from the notochord determines the positional information of the ventral dorsal axis during neural tube development [35]. Based on these findings, male mESCs (EB5) were differentiated using Wataya’s method with the addition of Shh or SAG starting on day 4 (Fig. 1A). Treatment with a higher concentration (100 nM) of SAG resulted in the appearance of more Nkx2.1-immunopositive cells (a marker of the ventral hypothalamus) and fewer Pou3f2-positive cells (a marker of the dorsal hypothalamus) than the other groups, including the lower concentration (10 nM) of SAG on day 13 (Fig. 1B), indicating the induction of ventral hypothalamic progenitor cells. Furthermore, immunohistochemistry revealed the emergence of a small number of kisspeptin-positive cells in the 10 nM SAG-treated group on day 25 (Fig. 1C), indicating partial differentiation of mESCs into kisspeptin neurons.

Fig. 1.

Differentiation of ventral hypothalamic neurons from male mESCs. A, Schematic of the protocol for differentiation of ventral hypothalamus by modified Wataya’s method using male mESCs, EB5, with Shh or SAG. B, Fluorescence immunostaining of Nkx2.1 (red) and Pou3f2 (white) and nuclear counterstaining with DAPI (blue) in differentiated male mESC aggregates, treated with or without 30 nM Shh, 10 nM, or 100 nM SAG from day 4–13. Bars: 100 μm. C, Fluorescence immunostaining of KP (red) and nuclear counterstaining with DAPI (blue) in differentiated male mESC aggregate, treated with 10 nM SAG from day 4–25. Bars: 100 μm (low magnification), 10 μm (high magnification). Abbreviations: Shh, Sonic Hedgehog; mESCs, mouse embryonic stem cells; SAG, Smoothened agonist; gfCDM, growth factor-free chemically defined medium; DFNB, DMEM/F-12 supplemented with glucose, N-2 supplement, and B27 supplement; Nkx2.1, NK2 homeobox 1; Pou3f2, POU domain, class 3, transcription factor 2; DAPI, 4',6-diamidino-2-phenylindole; KP, kisspeptin

Differentiation of hypothalamic ARC neurons using female mESCs

Female mESCs, BRC6(feeder-free), cultured under the aforementioned conditions, failed to form stable aggregates after 13 days of culture (Fig. 2A). To stabilize aggregates, BRC6(feeder-free) cells were treated with Y-27632 and/or KSR, as shown in Fig. 2B. Notably, supplementation with 1% KSR significantly increased aggregate formation on day 13 compared to control cultures without Y-27632 and with 0.5% KSR (92.7% vs. 43.9%, χ² test, P < 0.05) (Fig. 2C). To optimize the differentiation conditions, the timing and duration of Y-27632 treatment were evaluated (Fig. 2B). Continuous treatment with Y-27632 for 13 days resulted in small aggregates and a limited number of Rx/Pax6-positive hypothalamic progenitor cells (Fig. 2D and E). Conversely, washing out Y-27632 on day 4 resulted in larger, more stable aggregates and increased Rx/Pax6-positive cell proportions (Fig. 2D and E), indicating that the optimal condition was 1% KSR from day 0 to 13 and 10 µM Y-27632 from day 1 to 4 (Fig. 2F and G). Under optimal conditions, BRC6(feeder-free) cells differentiated into CPP- and NPII-co-positive vasopressin neurons in the absence of SAG (Fig. 2H). In contrast, BRC6(feeder-free) cells differentiated into ARC neurons, such as POMC/Nkx2.1-co-positive (Fig. 2I) and AgRP/NPY-co-positive neurons (Fig. 2J), after treatment with 100 nM SAG. Few kisspeptin-positive cells were observed under these conditions (Fig. 2K). These results demonstrated the successful differentiation of ARC-localized neurons from female mESCs.

Fig. 2.

Phase-contrast microscope images of female (A) mESCs cultured by modified Wataya’s method on days 4, 7, 10, and 13. Bars: 100 μm. Female mESCs, BRC6 (feeder-free), did not form aggregates by day 13. B, Schematic of protocols to stabilize the aggregates using female mESCs, BRC6 (feeder-free) with several doses of Y-27632 and KSR. C, Phase-contrast microscope images of female mESCs on day 13 cultured with 0.5% or 1.0% KSR, without Y-27632. Bars: 200 μm. Aggregates were formed in 43.9% (0.5% KSR) and 92.7% (1.0% KSR) wells, respectively. Phase-contrast images (D) and fluorescence immunostaining of Rx (green), Pax6 (magenta), and nuclear counterstaining with DAPI (blue) (E) of 1% KSR-primed female mESC aggregates, treated with or without Y-27632 and its washout. Bars: 200 μm. F, Schematic of the protocol to stabilize the aggregates and for ventralization of hypothalamic precursors using female mESCs, BRC6 (feeder-free) with Y-27632 and KSR. G, Phase-contrast images of female mESC aggregates with 1% KSR and 10 µM Y-27632 on days 4, 7, 10, 12. Bars: 100 μm. H-K, Fluorescence immunostaining of CPP (green) and NPII (magenta) in differentiated female mESC aggregates without SAG treatment from day 4–49 (H), POMC (green) and Nkx2.1 (magenta) (I), AgRP (green) and NPY (magenta) (J), and KP (red) (K) with DAPI (blue) in differentiated female mESC aggregates treated with 100 nM SAG from day 4–42. Arrows indicate cells that coexpress these markers. Bars: 50 μm (low magnification), 20 μm (high magnification). Abbreviations: KSR, KnockOut™ serum replacement; Rx, retina and anterior neural fold homeobox; Pax6, paired box 6; Shh, Sonic Hedgehog; SAG, smoothened agonist; Y, Y-27632; CPP, copeptin; NPⅡ, neurophysin II; POMC, pro-opiomelanocortin; Nkx2.1, NK2 homeobox 1; AgRP, agouti-related protein; NPY, neuropeptide Y

Differentiation of hypothalamic KNDy neurons from female mESCs by inhibiting Notch signaling

Next, we investigated the role of Notch signaling inhibition in the differentiation of KNDy neurons using additional supplementation with DAPT, a Notch signaling inhibitor, for varying durations (1, 3, 5, and 7 days) (Fig. 3A). Kisspeptin immunoreactivity was observed in DAPT-treated cells across all treatment durations (Fig. 3B). The kisspeptin-immunopositive area was significantly larger in cultures treated with DAPT for 3 days from day 10 than in those treated for 1, 5, or 7 days (P < 0.01; one-way ANOVA followed by Tukey’s HSD test) (Fig. 3C). In addition, kisspeptin-immunopositive cells were confirmed to be neurons, as they coexpressed TUBB3 (Fig. 3D). To further promote neuronal maturation, aggregates were dispersed and transferred to adherent cultures (Fig. 3E). Some cells exhibited immunoreactivity for kisspeptin, NKB, and Dyn on day 47 (Fig. 3F), with 67 of 141 kisspeptin-positive cells (47.5%) coexpressing both NKB and Dyn across the three fields of view. In addition, quantitative RT-PCR revealed that Kiss1 and Tac2 expression tended to increase (P = 0.090 and 0.104, respectively, two-tailed unpaired t-test) and that Pdyn was significantly upregulated (P < 0.05) compared to day 10 aggregates (Fig. 3G). Day 47 samples displayed a single sharp melting peak, confirming specific Kiss1 amplification, whereas day 10 aggregates, used as a biological negative control, showed low intensity and broad peaks consistent with nonspecific background or primer–dimer formation (Additional file 2). RT-PCR further revealed Kiss1 expression in cells cultured for 47 days, but not in those cultured for 10 days (Fig. 3H; see also Additional file 3 for the full, uncropped gels/blots). The amplicon size was identical to that of the placental positive control. Notably, kisspeptin concentration was significantly higher in the culture supernatant than in the control medium (culture supernatant 7.69 ± 1.92 pg/mL vs. dissociation medium 3.74 ± 0.085 pg/mL, mean ± SD, two-tailed unpaired t-test, t₍₈₎ = -2.776, P = 0.05) (Fig. 3I). To assess the functional properties of differentiated KNDy neurons, 100 nM E2 was added on day 50 of the overall differentiation protocol (day 6 post-dissociation; n = 4 per group). However, E2 did not decrease relative Kiss1 expression (control 1.00 ± 0.51 vs. E2 1.03 ± 0.78, mean ± SD, two-tailed unpaired t-test, t₍₆₎ = -0.092, P = 0.93).

Fig. 3.

Differentiation of hypothalamic KNDy neurons from female mESCs by regulating Notch signaling. A, Schematic of the protocol for differentiation of KP neurons from female mESCs treated with DAPT, a Notch signaling inhibitor, for 1, 3, 5, or 7 days from day 10. B, Fluorescence immunostaining of KP (red) with DAPI (blue) in female mESC aggregates on day 43. Bars: 50 μm. C, KP-immunopositive (+) area/DAPI-positive (+) area (%) in the female mESC aggregates treated with DAPT from day 10 for 1 (n = 9), 3 (n = 10), 5 (n = 10), and 7 (n = 10) days. Different letters indicate significant differences (P < 0.01, one-way ANOVA with Tukey's HSD test). D, Fluorescence immunostaining of TUBB3 (green) and kisspeptin (magenta) with DAPI (blue) in differentiated female mESCs on day 44. Bars: 10 μm. Arrows indicate cells that coexpress kisspeptin and TUBB3. E, Schematic of the protocol for differentiation of KNDy neurons from female mESC aggregates dispersed and transferred to adhesive culture. F, Fluorescence immunostaining of KP (green), NKB (magenta), and PDYN (yellow) with DAPI (blue) in differentiated female mESCs on day 47. Bars: 100 μm (low magnification), 20 μm (high magnification). Arrows indicate cells that coexpress all three markers (kisspeptin, NKB, and Pdyn), whereas arrowheads indicate cells that are kisspeptin-positive and PDYN-negative. G, Quantitative RT-PCR analysis of Kiss1, Tac2, and Pdyn mRNA expression in differentiated mESCs on days 47 (with DAPT treatment, n = 3) and 10 (undifferentiated aggregates as negative controls, n = 2), *P < 0.05 by two-tailed unpaired t-test. H, Qualitative RT-PCR analysis of Kiss1 and Gapdh mRNA expression in differentiated female mESCs on days 47 and 10. The placenta was used for positive control of Kiss1 expression. I, Kisspeptin concentration in culture supernatants of day 47 aggregates (n = 8) and control medium (n = 2), *P < 0.05 by two-tailed unpaired t-test. Abbreviations: DAPT, N-[N-(3, 5-difluorophenacetyl)-L-alanyl]-S-phenylglycine t-butyl ester; TUBB3, Tubulin β 3; NKB, neurokinin B; PDYN, ProDynorphin; pla, placenta

Differentiation of hypothalamic Kisspeptin neurons from male mESCs

Furthermore, EB5 (male mESCs) were cultured under the same conditions (300 nM SAG and 3-day DAPT treatment, Fig. 4A) as BRC6 female mESCs, resulting in the differentiation of kisspeptin/Dyn-immunopositive (Fig. 4B) and kisspeptin/NKB-immunopositive cells (Fig. 4C). Quantitatively, 52 of 128 kisspeptin-positive cells (40.6%) coexpressed dynorphin, and 142 of 314 cells (45.2%) coexpressed neurokinin B across the three fields of view.

Fig. 4.

Differentiation of hypothalamic Kisspeptin neurons from male mESCs. A, Schematic of the protocol for KNDy neuron differentiation from male mESCs with DAPT. B, C, Fluorescence immunostaining of KP (magenta), PDYN (green) (B), KP (green), and NKB (magenta) (C) with DAPI (blue) in differentiated male mESC aggregates on day 44. Arrows indicate cells that coexpress the markers (kisspeptin, PDYN, or NKB), whereas arrowheads indicate cells that were kisspeptin-positive and PDYN- or NKB-negative. Bar: 50 μm (low magnification), 20 μm (high magnification).Abbreviations: DAPT, N-[N-(3, 5-difluorophenacetyl)-l-alanyl]-S-phenylglycine t-butyl ester; NKB, neurokinin B; PDYN, ProDynorphin

Discussion

This study demonstrates the successful differentiation of KNDy neurons, which are essential for GnRH/gonadotropin secretion and subsequent reproductive function in mammals. Cells differentiated from female mESCs exhibited Kiss1 expression and immunoreactivity for kisspeptin, NKB, and Dyn. Notably, kisspeptin release from differentiated cells was also evident. To the best of our knowledge, this is the first study demonstrating the differentiation of mESCs into KNDy neurons. This method can be applied to create cellular models for various diseases associated with kisspeptin, NKB, and Dyn peptide deficiencies, contributing to the development of novel therapeutic approaches and an understanding of the cellular mechanisms regulating mammalian reproduction.

This study suggests that Shh and Notch signaling play critical roles in inducing the differentiation of ESCs into KNDy neurons. Specifically, KNDy neurons were differentiated by adding Shh from day 4 and inhibiting Notch signaling from days 10 to 13. These findings align with those of previous studies reporting that Shh induces the development of the central nervous system and patterns it dorsoventrally through a Shh concentration gradient [26, 36] and that Notch signaling is involved in the maintenance of neural stem cells and glial cell differentiation during mammalian neurogenesis [37]. In support of these findings, in vivo studies using transgenic mice have shown that Notch inhibition promotes the emergence of POMC-positive progenitors in the ARC and that Notch signaling is required for their subsequent differentiation into KNDy neurons [27, 28]. Taken together, our results suggest that the inhibition of temporal Notch signaling (specifically for 3 days, from day 10 after culture onset) is important for inducing the differentiation of KNDy neurons. Several modifications were required to stabilize aggregate formation from BRC6 (feeder-free). Previous studies using male mESCs (EB5) reported robust SFEBq-based hypothalamic induction, but BRC6 (feeder-free), the female mESC line, had not been applied to SFEBq and tended to collapse. We modified culture conditions by supplementing with KSR and Y-27632. KSR was added for its role in ES/iPS cell proliferation and differentiation, especially in human pluripotent stem cells. While Wataya et al. [26] omitted KSR to minimize nutrients, several mESC protocols use 15–20% KSR for cortical or retinal induction. Limited KSR supplementation stabilized the aggregates effectively. Y-27632, a ROCK inhibitor used to enhance ES/iPS cell survival, was added to reduce apoptosis [38]. Despite fewer reports in mESCs, 20 µM Y-27632 improved neural differentiation [39]. Its addition was beneficial for aggregate stability. These adjustments reflected cell line-specific characteristics rather than sex differences. Optimized conditions allowed stable aggregates and efficient KNDy neuron differentiation from BRC6 (female), comparable to EB5 (male) cultures.

In this study, we successfully differentiated KNDy neurons from female mESCs. Idiopathic hypogonadotropic hypogonadism and central precocious puberty caused by mutations in the kisspeptin gene KISS1 and its receptor gene GPR54 have been reported in both sexes of humans [1, 22]. Therefore, KNDy neurons differentiated from male and female mESCs may serve as valuable tools to understand reproductive diseases in both sexes. Moreover, ESC-derived KNDy neurons may have an advantage over immortalized cell-derived KNDy neurons in duplicating in vivo properties because pluripotent stem cells are differentiated by reproducing the in vivo developmental process. Thus, ESC-derived KNDy neurons may be applied in drug screening before testing in animal disease models, leading to clinical applications. Specifically, female ESC-derived KNDy neurons could be used for drug screening in female-specific diseases, such as PCOS, which is often associated with a risk of perinatal complications, uterine cancer, diabetes, and other diseases [40, 41], as well as menopause. In support of this interpretation, a previous study using immortalized cell lines of KNDy neurons from the mouse hypothalamus [42] showed that changes in Kiss1 expression in response to E2 stimulation were more pronounced in female-derived cells than in male-derived cells. In addition, KNDy neurons derived from male mESCs may serve as useful therapeutic targets for male reproductive disorders. Menopause has also been recognized in men, as the LOH syndrome. Testosterone replacement therapy is a typical treatment for LOH syndrome in men. However, it suppresses the HPG axis and impair spermatogenesis [24]. Thus, these findings are expected to contribute to the establishment of new treatments in clinical practice for male and female reproductive diseases. In particular, the use of mESCs provides a foundation for elucidating the central mechanisms underlying these disorders.

This method can also be applied to generate in vitro models of human KNDy cells from human ESCs or iPSCs, which are useful for developing novel therapies for other diseases. Kisspeptin is reported to be involved in metabolic diseases, including diabetes [43]; moreover, KNDy neuron dysfunction has been linked to individual variations in weight loss, amenorrhea, and heterogeneity in the resumption of menstruation following weight recovery [44]. Furthermore, FHA, the most frequent cause of secondary amenorrhea, is associated with eating disorders, excessive exercise, and stress [45]. It is often associated with bone loss, psychiatric symptoms, dementia, and cardiovascular disorders because of estrogen depletion [46]. Amenorrhea may persist for several years after weight recovery [47]. Several studies have reported the administration of kisspeptin and the kisspeptin receptor (GPR54) agonist MVT-602 for FHA treatment. Although these agents have been shown to increase LH levels in patients with FHA, chronic administration may cause tachyphylaxis [48, 49]. In addition, the high blood kisspeptin concentration and increased LH pulse frequency observed in patients with PCOS, a major cause of infertility, suggest the involvement of KNDy neurons in its pathogenesis [50–52]. A commonly used treatment for infertility in patients with PCOS is gonadotropin, which directly stimulates the ovaries. However, this method often causes ovarian hyperstimulation syndrome [53]. Case studies of PCOS therapies targeting KNDy neurons have been reported; however, their clinical use has not yet been established [54]. Thus, differentiated KNDy neurons may shed light on PCOS pathogenesis and help develop safer and more physiological treatments that directly stimulate the hypothalamus, which is the center of the HPG axis.

Furthermore, the current model may also help investigate VMS, represented by hot flashes, and thought to be caused by KNDy neuron hyperactivity and thermoregulatory center dysregulation via NK3R [55]. VMS significantly affect the life events of women, including social and leisure activities and sex life, by affecting sleep, mood disturbances, concentration, general energy levels, and the overall quality of life [56]. VMS appears in 33–79% of menopausal women [57], and the recent approval of an NK3R antagonist, fezolinetant, for VMS treatment [58] underscores the relationship between VMS and KNDy neurons. Notably, the results of the MOONLIGHT I study in Asians showed that treatment was not as effective in Asians as in Westerners [59]. Therefore, further in vitro and in vivo studies on KNDy neurons are warranted for VMS treatment.

This study has some limitations. Kiss1 mRNA expression did not show a significant increase compared to the negative control (day 10), and E2 treatment as a functional analysis did not produce detectable changes in Kiss1 mRNA levels. These results may partly reflect the limited number of samples available for qPCR, as adhesive culture requires a relatively large number of aggregates (approximately 2–15 per well), thereby restricting the number of wells available for mRNA analysis. Furthermore, KNDy neuron differentiated from ESCs may not have matured sufficiently, resulting in neurons exhibiting limited responsiveness to Kiss1 mRNA expression.

Another limitation is that the male ESCs used in this study carried an EGFP knock-in at the Rax locus, which prevented triple immunostaining. Although the coexpression of kisspeptin/Pdyn and kisspeptin/NKB was confirmed, we could not definitively determine whether all three peptides were simultaneously expressed in individual cells. To address this limitation, future studies will need to examine differentiation using ESC lines without EGFP knock-in, as well as additional cell lines with different genetic backgrounds.

Despite these limitations, our findings provide a foundational framework for generating KNDy neurons from pluripotent stem cells and highlight important considerations for refining future differentiation protocols.

Conclusions

This study successfully established a method for the differentiation of mESCs into hypothalamic KNDy neurons. The resulting cellular model could contribute to the understanding of the cellular mechanisms regulating mammalian reproduction, the pathogenesis of diseases caused by KNDy neuron dysfunction, and the development of novel therapeutic approaches for reproductive and nonreproductive disorders in humans.

Abbreviations

AgRP

Agouti-related peptide

ANOVA

Analysis of variance

ARC

Arcuate nucleus

BDNF

Brain-derived neurotrophic factor

CPP

Copeptin

DAPI

4',6-diamidino-2-phenylindole

DAPT

N-[N-(3, 5-difluorophenacetyl)-L-alanyl]-S-phenylglycine t-butyl ester

DFNB

DMEM/F-12 supplemented with glucose, N-2 supplement, and B27 supplement

DMEM

Dulbecco’s modified Eagle medium

Dyn

Dynorphin A

ESCs

Embryonic stem cells

FBS

Fetal bovine serum

FHA

Functional hypothalamic amenorrhea

Gapdh

Glyceraldehyde-3-phosphate dehydrogenase

gfCDM

Growth factor-free chemically defined medium

GnRH

Gonadotropin-releasing hormone

HPG axis

Hypothalamic-pituitary-gonadal axis

HSD

Honest significant difference

iPSCs

Induced pluripotent stem cells

KSR

KnockOut serum replacement

LH

Luteinizing hormone

LOH

Late-onset hypogonadism

E2

17β-estradiol

mESCs

Mouse ESCs

NK3R

Neurokinin 3 receptor

NKB

Neurokinin B

Nkx2.1

NK2 homeobox 1

NPII

Neurophysin II

NPY

Neuropeptide Y

Pax6

Paired box 6

PBS

Phosphate-buffered saline

PCOS

Polycystic ovary syndrome

PCR

Polymerase chain reaction

PDL

Poly-D-lysine

PFA

Paraformaldehyde

POMC

Proopiomelanocortin

Pou3f2

POU domain, class 3, transcription factor 2

Rx

Retina and anterior neural fold homeobox

SAG

Smoothened agonist

SFEBq

Serum-free floating culture of embryoid body-like aggregates with quick reaggregation

Shh

Sonic Hedgehog

TUBB3

Tubulin β 3

VMS

Vasomotor symptoms

Y

Y-27632

Authors’ contributions

NM, HS, SO, and HK conceived and designed the study. NM performed the experiments with technical assistance from HS, SO, B, TM, MSoen, MSakakibara, SG, NI, YU, and HT. NM, HS, SO, TS, RS, AM, TN, SH, NI, YU, HT, HA, and HK interpreted the data. NM, HS, SO, YU, and HT wrote the manuscript. All the authors reviewed the manuscript.

Funding

The following projects supported this research: AMED Practical Research Project for Rare/Intractable Diseases (JP24ek0109702, Japan); the JST FOREST Program (JPMJFR200N, Japan); JSPS KAKENHI (Grant-in-Aid for Scientific Research(C), JP23K08005, Japan); Ichihara International Scholarship Foundation (Japan) to Hidetaka Suga, Nagoya University Hospital Funding for Clinical Research (Japan) to Satoko Osuka, MEXT KAKENHI (Grant-in-Aid for Early-Career Scientists, JP23K15807, Japan); Sugiyama Memorial Foundation for Public Interest (Japan) to Natsuki Miyake.

Data availability

The datasets generated and/or analyzed in the current study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

All animal experiments were approved by the Animal Experimental Committee of the Nagoya University Graduate School of Medicine (M250004-003 and M250013-003).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.