Introduction

Allopregnanolone (ALLO; IUPAC: 1-[(3R,5S,8R,9S,10S,13S,14S,17S)-3-hydroxy-10,13-dimethyl-2,3,4,5,6,7,8,9,11,12,14,15,16,17-tetradecahydro-1H-cyclopenta[a]phenanthren-17-yl]ethenone) is an endogenous steroid synthesized de novo or by progesterone (P4) metabolism in the nervous system, adrenal glands, ovaries and placenta in the female1,2,3,4,5. ALLO is mostly known for its anxiolytic, antidepressant, anticonvulsant and neuroprotective actions at the central nervous system (CNS) level6,7,8,9. Moreover, it can modulate a variety of behaviors, such as reproductive10,11,12,13, social, exploratory and affective conducts14,15,16. Previous work from our laboratory reported ALLO actions over the hypothalamic–pituitary–gonadal axis when administered via ICV17,18,19 and its effects on the ovarian peripheral innervation using an ex vivo culture of the superior mesenteric ganglion- ovarian nervous plexus—ovarian system20,21,22. These findings led us to consider ALLO not only as a modulator of central function but also of ovarian physiology.

The ovary is one of the major sources of circulating ALLO23,24, and the ovarian cells express many membrane receptors that could be action sites for its in situ reproductive modulation. ALLO is synthesized by ovarian interstitial cells, follicles and corpora lutea (CL)25. Its serum concentrations vary along the rat estrous cycle24, human menstrual cycle and gestation26, following P4 levels24,27, and its ovarian synthesis is stimulated by LH28. The changes in steroid hormones serum concentration is also accompanied by an increase in their ovarian concentration24. However, the direct actions of ALLO over the ovarian physiology are not known.

ALLO is considered a potent positive allosteric modulator of GABAA receptors (GABAAR), enhancing GABAergic activity at nanomolar concentrations and directly activating the receptor at micromolar concentrations29,30,31,32,33. There are few reports of the presence and function of GABAAR in non-neural tissue34. The presence of GABAAR, GABA and the enzyme glutamate acid decarboxylase (GAD) has been previously demonstrated in the ovary34,35,36. Ovarian GABA concentrations change throughout the estrous cycle37 and gestation38,39. An example of non-neural GABAergic action occurs in the uterus of pregnant rats, where ALLO exerts its effects by regulating the contractions of the rat’s myometrium40,41,42. In addition, in pseudopregnant rats, an intrabursal perfusion of GABA (0.5 µM) produces vasoconstriction, a decrease in P4 secretion, and an immediate increase in estrogen secretion43. The importance of P4 metabolite-mediated regulation of non-neural GABAAR present in the rat reproductive system remains to be fully elucidated.

To evaluate the action of a GABAAR modulator, it is important to test its functional activity by using an antagonist. Briefly, in a receptor functional study, the antagonist prevents the action of the agonist, thus avoiding its effect. This confirms that the receptor is indeed involved in the agonist’s actions. Bicuculline (BIC) is a benzylisoquinoline alkaloid and a competitive antagonist of GABAAR44. In a primary culture of astrocytes, the half-maximal inhibitory concentration (IC50) of bicuculline on GABAARs is 3.3 μM for a concentration of 30 μM GABA. It shows a Hills coefficient of − 1.8 (± 0.06), indicating negative cooperativity at the receptor (initial binding inhibits subsequent binding to other receptor sites)45. Bicuculline sensitivity is defined by IUPHAR as an important criterion in the definition of GABAARs46,47,48, and it is the most frequently used antagonist for GABAAR functional studies.

Bearing in mind the previous findings, we analyzed ALLO effects on the local regulation of ovarian key features using an in vivo intrabursal administration model. We evaluated (a) changes in ovarian morphometry, weight and ovulation, (b) ovarian angiogenesis, (c) apoptosis and proliferation of granulosa and luteal cells, d) steroidogenesis and hormone secretion. Finally, in order to determine whether the actions of ALLO involve this receptor, we performed a GABAAR functional experiment with bicuculline. This is the first evidence of the local effect of ALLO on the female rat ovarian function and its eventual pathological consequences.

Materials and methods

Reagents

Allopregnanolone (C21H34O2, 3α-hydroxy-5α-pregnan-20-one, Sigma Chemical Co., St. Louis, MO, USA, CAS Number 516-54-1), Ketamine HCl (Ketonal 50 mg/ml, Richmond Laboratories, Veterinary Division, Buenos Aires, Argentina) and Xylazine (Sedomin 100 mg/ml, König Laboratories, Buenos Aires, Argentina) were used for experimental and surgical procedures.

To obtain the stock solution, ALLO was first dissolved in propylene glycol at a concentration of 3.14 mM. Working concentration was obtained after serial dilutions in sterile physiological solution; the resulting propylene glycol concentration was less than 0.2%.

Bicuculline methiodide [1(S),9(R)-(-)-Bicuculline methiodide, Sigma, USA] was used to antagonize GABAARs. A stock solution was prepared in DMSO at an initial concentration of 1 × 10−2 M. Working concentrations were obtained by dilution of the initial one in physiological solution; the resulting DMSO concentration at working solutions was less than 0.2%.

The control group received physiological solution (BRAUN, Buenos Aires, Argentina) with a similar concentration of propylene glycol/DMSO.

Bouin solution (Biopur Diagnostics, Santa Fe, Argentina), hematoxylin, eosin (Merk, Germany) and Canada Balsam Synthetic (Biopack, Buenos Aires, Argentina) were used for histological procedures.

Animals

Adult female Sprague Dawley rats (Rattus norvegicus) of 200–250 g were maintained under controlled conditions of temperature and light (12 h light/12 h darkness photoperiod). They were housed in groups of four animals per cage, with food (standard rat chow, Cargill, Córdoba, Argentina) and water available ad libitum. Vaginal smears from each rat were observed daily (07:00–09:00 am) with a light microscope (Zeiss, Germany) to determine the stage of the estrous cycle. Only those animals exhibiting two or more consecutive 4 or 5-day cycles were used. All protocols were previously approved by the Institutional Committee for the Care and Use of Laboratory Animals of the National University of Cuyo (CICUAL aval: 154/2019). The animals for these experiments were kept and handled according to the National Institutes of Health Guide for the care and use of laboratory animals, 8th edition (National Research Council). The study was performed, and results reported in accordance to ARRIVE guidelines. The total number of animals used for the experiments was 35.

Anesthesia

A combination of ketamine (80 mg/kg) and xylazine (10 mg/kg) administered intraperitoneally was used. Confirmation of the level of anesthesia was assessed by loss of postural reflex, palpebro-palpebral reflex, corneo-palpebral reflex, podal reflex and lack of response to tail pinch.

Surgical procedure

Access to each ovary was made through two paramedian ventral incisions. This procedure allowed gentle handling of the tissue and correct visualization of the inoculation site, and thus to evaluate the complete discharge of the inoculum and the integrity of the post-manipulation bursa.

With the animal in lateral decubitus, a skin incision was made no more than 10 mm caudal to the last rib (Fig. 1), taking as anatomical references the greater trochanter of the femur and the vertebral column. The ovary was carefully lifted with forceps, taking it by one of the fat pads that surround it. Once outside the abdominal cavity, the ovary was placed on a pad of wet gauze adjacent to the incision site to avoid dehydration of the structures. The corresponding treatment (5 µl of total volume) was then injected using a Hamilton type syringe. Correct inoculation was corroborated by visualizing the needle and dilation of the post-injection bursa. The needle was quickly withdrawn with care not to tear the ovarian bursa. Then, the ovary was carefully reintroduced into the abdominal cavity and each of the incised planes was sutured. The described procedure was performed on each abdominal side, taking less than 8 min in total. After surgery, the animals were kept under observation in a heated room until they fully recovered from anesthesia. After 24 h, the animals were in perfect condition with no signs of pain or lethargy.

Figure 1
figure 1

Experimental design. On the day of proestrus the rats were subjected to an intrabursal administration surgery. On the day of estrous, 24 h later, they were euthanized by decapitation. Blue lines show the anatomical references used for the surgery: the greater trochanter of the femur, the vertebral column and the last rib. The red line shows the area of incision.

The ovarian intrabursal drug delivery model was used to analyze the local effects of ALLO on ovarian physiology. This model allows in vivo topical administration of drugs to the ovary, keeping the anatomy, innervation, and blood flow intact49. Treatments inoculated into the bursal cavity rapidly penetrate the ovarian parenchyma50,51. In this way, the ovulation rate and in situ response to the applied drugs can be assessed while the ovary is under systemic regulation.

Experimental design

On the morning of proestrus, 5 µl of ALLO at a concentration of 6 µM and controls were administered intrabursally (Fig. 1). The ALLO concentration of 6 µM was chosen for these studies based on a concentration–response curve (data not shown) and previous works. It was the pharmacological concentration with the greatest effects on ovarian physiology19,22.

The inoculation volume was determined based on the work of Abramovich et al.52. On the morning of estrus, 24 h later, the animals were euthanized by guillotine decapitation. Confirmation of estrous was performed prior to euthanasia. The ovaries were removed and placed in glass Petri dishes with chilled saline solution for subsequent assays. Serum samples were collected and stored at − 80 °C until processing. All the counts were recorded by a researcher blinded to the group.

Experiment 1 A paired model was used to reduce the variability between groups. Thus, the left ovary of each animal received ALLO 6 µM and the right ovary (control) received saline. A paired study is possible since the changes produced by the inoculation of treatments in one ovary are not compensated by the contralateral49, at least until 24 h after their injection53. Also, the intrabursal administration of saline solution, commonly used in the control ovary, does not alter ovarian physiology54,55. For serum hormone determinations, both ovaries of two groups of animals were treated with saline or ALLO 6 µM (not paired experiment). The total number of animals used for this experiment was 15.

Experiment 2 To evaluate whether some of the ALLO actions on ovarian morphometry were mediated by GABAAR, a functional assay was performed with bicuculline, a competitive GABAAR antagonist. We tested 3 commonly used doses: 5, 10 and 100 µM56,57,58. None of the BIC concentrations per se induced significant changes in the evaluated parameters (data not shown). Therefore, the 10 µM concentration was selected for this experiment. In this case, both ovaries were injected with the treatments (n = 5 animals/group). For this experiment the total number of animals used was 20.

Evaluation of ovulation and ovarian weight

After removal of the ovaries, the distended portion of the ampullae was located and punctured with a 21-gauge needle under a stereoscopic magnifying glass (Zeiss, Germany). The number of oocytes released were then counted for each ovary. To clean the ovary, the adhering tissue and fat were dissected out and discarded. The ovaries were weighted on an analytical scale. The number of animals used was 5, with 5 ovaries per experimental group.

Ovarian morphometry

After ovarian weight analysis, the ovaries were immediately fixed in Bouin solution for 12 h. They were then dehydrated in a series of increasing concentrations of ethyl alcohol and finally embedded in paraffin. Paraffin sections of 5 μm were taken every 40 μm and mounted onto microscope slides to prevent counting the same structure twice, according to the method described by Woodruff et al.53. Finally, the slides were stained with hematoxylin–eosin and mounted with mounting medium.

Follicles were classified as primary (a single layer of cuboidal granulosa cells), secondary (more than one layer of granulosa cells with an incipient antrum), tertiary (more than three granulosa layers and a clearly defined antral space) or Graafian follicles (polarized oocyte, defined cumulus granulosa layer and a cavity occupying most of the follicular volume). Corpora lutea were identified by the presence of large and small luteal cells with their characteristic cytoplasmic eosinophilia59. Corpora albicans were scarce, and they were not taken into account. Morphological characteristics of atretic follicles (ATF) include pyknotic nuclei (cell degeneration) in more than 5% of the total number of granulosa cells, detachment of the granulosa cell layer from the basal laminae, hypertrophy of external theca and oocyte degeneration60. Follicular cysts were identified by a thin granulosa layer (≤ 3 mm) and a large antral cavity (the oocyte may be absent or present)61. The number of ovarian structures was determined in four ovarian sections from each ovary (n = 5 animals/group). The structure diameters were measured using Image J software (Image Processing and Analysis in Java; National Institutes of Health, Bethesda, MD, USA). Results were expressed as mean diameter ± SEM.

In situ cell death detection (TUNEL)

For quantification of apoptosis in follicles and CL, 3 µm thick paraffin sections were processed for in situ localization of nuclei showing DNA fragmentation by the TUNEL technique (terminal-end-labeling with dUTP-digoxigenin using the enzyme deoxynucleotidyl transferase (TdT)). Ovarian sections were deparaffinized and hydrated by washing in ethyl alcohol of decreasing concentration (100°–70°). They were then incubated with Proteinase K 20 μg/ml (Promega, Madison, WI, USA) for 10 min at room temperature, followed by incubation in 0.01 M citrate buffer pH 6 for 30 min at 90–100 °C. The In-situ Cell Death Detection KIT, POD (Roche, Mannheim, Germany) was used. The labeling reaction was performed according to the manufacturer’s protocol.

The preparations were incubated with digoxigenin-dUTP and TdT enzyme (labeling solution and enzyme solution) for 1 and 30 min, respectively. They were then incubated with peroxidase-conjugated anti-digoxigenin monoclonal antibody for 30 min in a humid chamber at 37 °C. Finally, apoptotic positive cells were visualized after reaction with diaminobenzidine (SIGMAFAST™ 3,3′-diaminobenzidine tablets, Sigma, Darmstadt, Germany). Negative controls were obtained by replacing the TdT solution with phosphate buffer pH 7. The sections were counterstained with hematoxylin. The number of apoptotic cells was determined by counting labeled cells in follicles in randomly selected fields at 400 × magnification (n = 5 ovaries/group). The apoptotic index was calculated as the percentage of apoptotic cells relative to the total number of cells.

Immunohistochemistry

Ovarian tissue sections were cut at 3 µm and mounted on positively charged slides. Tissue sections were deparaffinized in xylene and rehydrated by washing with alcohols. Antigen retrieval was performed in 0.01 M sodium citrate buffer at pH 6 for 10 min in a microwave oven at maximum power (800 W).

Endogenous peroxidase enzyme activity was blocked with 3% hydrogen peroxide solution in PBS buffer for 10 min. Blocking of non-specific binding sites was performed in a humidity chamber by incubating the sections with 2% BSA (bovine serum albumin) for 20 min at room temperature. The histological sections were then incubated with 20 µl of the primary antibodies listed in Table 1 overnight at 4 °C in a humidity chamber.

Table 1 Antibodies used for immunohistochemistry.

A negative control mouse IgG (I-2000-1, Vector Laboratories, Burlingame, CA, USA) was used instead of the primary mouse monoclonal antibody to evaluate non-specific binding. A secondary negative control was performed by omitting incubation of the sections with the primary antibody. The next day, after successive washes with PBS, the sections were incubated with the appropriate secondary antibodies listed in Table 1 for 2 h. They were then incubated with ABC: avidin-biotinylated peroxidase complex (Vectastain ABC system, Vector Laboratories, Burlingame, CA, USA) for 30 min. Positive labeling was visualized by precipitation of the peroxidase enzyme substrate, diaminobenzidine (DAB), in developer buffer (Roche Diagnostics, Germany). The reaction was stopped with distilled water and counterstained with hematoxylin for 1 min. Upon completion of immunolabeling, photographs were taken with a digital camera attached to a light microscope (Nikon, Melville, NY, USA) for further processing. Image J software was used to count positively labeled cells and total cells per structure.

Cell proliferation was evaluated by PCNA (proliferating cell nuclear antigen) in secondary, tertiary and de Graaf follicles as well as in CL. This group of follicles is the most susceptible to degeneration by atresia, since they depend on gonadotrophin stimulation for their survival and can sense the balance between external pro/anti-apoptotic factors62,63. The proliferation index was calculated for each field using the following formula: number of immunolabeled cells/total number of cells counted per field. The number of apoptotic cells was determined by counting labeled cells in follicles in randomly selected fields at 400 × magnification (n = 5 ovaries/group). Results were expressed as mean ± SEM of cells positive for the selected antigen.

The percentage of positively labeled endothelial (von Willebrand factor-labeled) and peri-endothelial (α-actin-labeled) areas relative to the total ovarian area was quantified using Image J software. The examined areas were determined in 3 ovarian sections from each ovary. The results were expressed as mean ± SEM of the positively labeled endothelial area.

Corpora lutea vascular area

Lectin BS-1 (from Bandeiraea simplicifolia, 20 mg/ml), a proven constitutive endothelial cell marker64,65, was used to assess CL blood vessels. Ovarian sections were incubated with biotinylated lectin BS-1 overnight at 4 °C. After washing, the slides were incubated with avidin biotinylated horseradish peroxidase complex (Vectastain ABC system) for 30 min. Protein expression was revealed with diaminobenzidine (DAB) staining. Hematoxylin was used as a counterstain. The slides were then dehydrated before mounting (Canada Balsam Synthetic, Biopack, Argentina). Negative controls were obtained in the absence of lectin BS-1.

Vascular area (lectin BS-1-positive cells) was determined by thresholding the positive stained area and calculated by relativization to the total luteal area of the captured photograph using the Image J software (n = 3 animals/group, 16 ovarian sections/group).

Hormone determinations by chemiluminescent immunoassay (CLIA)

For hormone determination, one group of animals received saline or ALLO 6 µM in both ovaries (n = 5 animals/group). Serum estrogen (E2), P4 and FSH concentrations were determined by CLIA using a Mindray CL-1200i analyzer (Mindray Laboratories, Shenzhen, China) and its specific reagents, with analytical sensitivity of ≤ 0.2 mIU/mL for FSH, ≤ 25 pg for E2, and ≤ 0.1 ng/mL for P4.

Real-time PCR

Total RNA was obtained from ovarian fragments using the TRIZOL method according to the manufacturer’s instructions (n = 4 animals/group). Quantification of the RNA obtained was performed in a Jenway spectrophotometer at 260 nm. RNA integrity was verified by electrophoresis in a 1% agarose gel stained with SYBR Gold (Thermofisher Scientific Technologies, Buenos Aires, Argentina). Five micrograms of total RNA were retrotranscribed at 37 °C using random hexamer primers (Random Hexamers, Thermofisher) and murine leukemia virus retrotranscriptase enzyme (M-MLV, Thermofisher) in a reaction mixture with a final volume of 20 µl, according to previously published steps21.

Reactions were performed in a real-time Corbett Rotor Gene 6000 thermal cycler (Corbett Research Pty Ltd, Sydney, Australia), with a final volume of 20 µl. The reaction mixture consisted of 2 µL of 10 × PCR buffer, 1 µL of 50 mM MgCl2, 0.4 µL of 10 mM dNTPs, 1 µL of 20 × Eva Green (Biotium, Hayward, CA, USA), 0.25 µL of 5 U/µL Taq DNA Polymerase (Thermofisher), 0.1 µL of each primer (sense and antisense, detailed in Table 2) 2.5 mM, and 10 µL of cDNA diluted 1/20 in sterile RNAase-free H2O. PCR reactions were initiated with a 5-min incubation at 95 °C, followed by 40 cycles of 95 °C for 30 s, 60 °C for 30 s and 72 °C for 30 s.

Table 2 Primers used for quantitative polymerase chain reaction amplification.

Melt curve analysis was used to verify that a single specific amplified product was generated. Real-time quantification was monitored by measuring the increase in fluorescence caused by the binding of EvaGreen dye to double-stranded DNA at the end of each amplification cycle. Relative expression was determined using the comparative quantitation method of normalized samples relative to the expression of a calibrator sample according to the manufacturer’s instructions66. Each PCR run included a non-template control and a sample without reverse transcriptase.

All measurements were performed in duplicate. Reaction conditions and amounts of cDNA added were calibrated so that the assay response was linear with respect to the amount of cDNA input for each pair of primers (Table 2). RNA samples were tested for DNA contamination by performing the respective PCR reactions without prior reverse transcription.

The reference gene used to normalize the results was ribosomal S16. For the selection of this gene, the expression stability of 3 candidate genes, β-actin, S16 and HPRT1 (hypoxanthine guanine phosphoribosyltransferase 1 enzyme), was estimated using the free software BestKeeper® version 1 [http://gene-quantification.com/bestkeeper.html]. S16 showed the lowest coefficient of variation between groups and the highest stability between treatments.

Statistical analysis

The raw data obtained were analyzed using Graph Pad Prism V.8.0.1 for Windows (Graph Pad Software, La Jolla, CA, USA). Shapiro–Wilk normality test was performed before statistical analysis. Paired Student’s t-test was used to determine differences between the means of paired normally distributed groups. Unpaired Student’s t-test was used for hormone assays. Wilcoxon signed ranked test was used to determine differences between the means of paired non-normally distributed groups. The comparisons between the means of different groups were carried out using one-way ANOVA followed by Tukey’s post hoc. The Kruskal–Wallis test followed by Dunn’s multiple comparison method was used to analyze nonparametric results. Two-tailed values of p < 0.05 were considered significant.

Results

Experiment 1

Effect of ALLO 6 µM on ovarian steroidogenesis

Intrabursal administration of ALLO 6 µM induced a significant increase in 3β-hydroxysteroid dehydrogenase (3β-HSD2) (p = 0.0451) and a significant decrease in 20α-hydroxysteroid dehydrogenase (20α-HSD) enzyme gene expression (p = 0.0137). No significant changes in 3α-hydroxysteroid oxidoreductase (3α-HSOR) mRNA expression were observed (p = 0.9608, Fig. 2).

Figure 2
figure 2

Gene expression of ovarian 3β-HSD2, 20α-HSD and 3α-HSOR after ALLO 6 µM intrabursal treatment. Results are expressed as mean ± SEM. Paired Student’s t test (*p < 0.05; n = 4 animals/experimental group).

ALLO administration significantly increased P4 (p = 0.0010) and E2 (p = 0.0079) serum concentration (Fig. 3).

Figure 3
figure 3

Effect of intrabursal administration of ALLO 6 µM on serum progesterone (A), estrogen (B) and FSH concentration. Results are expressed as mean ± SEM. (A) and (C) Unpaired Student’s t test, (B) Mann–Whitney test (**p < 0.01, ***p < 0.001; n = 5 animals/experimental group).

We also measured FSH serum levels. ALLO treatment induced a significant increase in this hormone (p = 0.0010, Fig. 3).

Effects of ALLO 6 μM on ovarian morphometry, weight and ovulation

ALLO intrabursal administration did not induce changes in ovarian weight (p = 0.7807, Fig. 4A) or in the number of ovulated oocytes (p = 0.5457; Fig. 4B).

Figure 4
figure 4

Ovarian weight in grams (A) and number of ovulated oocytes (B) in the ampullae after ALLO 6 µM intrabursal treatment. Results are expressed as mean ± SEM. Paired Student’s t test (ns > 0.05; n = 5 animals/experimental group).

ALLO administration resulted in a significant increase in the number of ATF (p = 0.0077). However, none of the other ovarian structures were significantly affected by the treatment (Fig. 5).

Figure 5
figure 5

Effect of intrabursal administration of ALLO 6 µM on the number of structures per ovary. Primary follicles (P.F), secondary follicles (S.F), tertiary follicles (T.F), de Graaf follicles (G.F), atretic follicles (AT.F.), cysts, total number of corpora lutea (T.CL.), new corpora lutea (N.CL.) and old corpora lutea (O.CL.). Results are expressed as mean ± SEM. Two-way ANOVA with Bonferroni post hoc (n = 4 animals/experimental group, **p < 0.01).

Regarding the diameter of the measured structures, ALLO 6 µM induced a significant increase (p < 0.0001) in the largest diameter of CL. The diameter of tertiary follicles was not affected (Fig. 6).

Figure 6
figure 6

Effect of intrabursal administration of ALLO 6 µM on tertiary follicle (T.F.) and corpora lutea (CL) diameter. Results are expressed as mean ± SEM. Paired Student’s t-test and Wilcoxon signed ranked test (n = 5 animals/experimental group, ***p < 0.0001).

Ovarian BAX, BCL-2, FAS, FAS-L and cyclin D1 gene expression

ALLO 6 µM induced a significant increase in gene expression of both BAX (p = 0.0303) and BCL-2 (p = 0.0236), but without significant changes in the BAX/BCL-2 ratio (p = 0.2778, Fig. 7).

Figure 7
figure 7

Effect of intrabursal administration of ALLO 6 µM on the expression of BAX, BCL-2, FAS, FAS-L and the BAX/BCL-2 and FAS/FAS-L ratio (AF). (G) ALLO effect on cyclin D1 gene expression. Results are expressed as mean ± SEM. Paired Student’s t test (* < 0.05; n = 4 animals/experimental group).

Regarding the extrinsic pathway of apoptosis, ALLO induced a significant increase in the FAS/FAS-L ratio (p = 0.0213, Fig. 7). ALLO administration did not significantly affect FAS (p = 0.0983) and FAS-L (p = 0.1714) gene expression.

Cyclin D1 gene expression was analyzed as a marker of ovarian cell proliferation. Intrabursal administration of 6 µM ALLO produced a significant decrease in this marker (p = 0.0465; Fig. 7).

Effect of ALLO 6 µM on follicular granulosa and theca cell proliferation and apoptosis

Administration of ALLO 6 µM induced a significant decrease in the proliferation index of theca (p = 0.0312) and granulosa (p < 0.0003) cells, accompanied by a significant increase in the apoptotic index (theca: p = 0.0019 and granulosa: p < 0.0001; Fig. 8).

Figure 8
figure 8

Upper panel: Representative microphotographs of tertiary follicles present in control (left panel) and ALLO 6 µM-treated (right panel), PCNA-labeled (A and B) and TUNEL-labeled (C and D) ovaries. Light microscopy, 100 x, bar = 200 µm. Lower panel: Proliferation (E, F) and apoptosis (H, I) rates of granulosa and theca cells of ovarian follicles. Number of positively immunolabeled cells per field. Results are expressed as mean ± SEM. Paired Student’s t test and Wilcoxon matched pairs signed rank test (granulosa PCNA) (*p < 0.05/**p < 0.01/***p < 0.001; PCNA: n = 16 sections/ovary for granulosa cells, n = 12 sections/ovary for theca cells, TUNEL: n = 10 sections/ovary for granulosa cells, n = 10 sections/ovary for theca cells, total of 5 ovaries/group).

Effect of ALLO 6 µM on the process of luteal cell proliferation and apoptosis

Paired intrabursal administration of 6 µM ALLO induced a significant increase in immunolabeling for PCNA in CL (p < 0.0001), accompanied by a significant decrease in TUNEL-positive cells (p < 0.0001; Fig. 9).

Figure 9
figure 9

Rate of proliferation (upper panel) and apoptosis (lower panel) of luteal cells. Number of positively immunolabeled cells per field. Results are expressed as mean ± SEM. Wilcoxon matched pairs signed rank test (PCNA) and Paired Student’s t test (***p < 0.001). Representative microphotographs of CL present in control (left panel) and ALLO 6 µM-treated (right panel) ovaries, stained with PCNA (A and B) and TUNEL (C and D). Light microscopy, 100 x, bar = 200 µm (PCNA: n = 27 sections/ovary, TUNEL: n = 35 sections/ovary, total of 5 ovaries/group).

Ovarian angiogenesis

The effect of ALLO 6 µM on vascular wall markers, α-smooth muscle actin and Von Willebrand factor was evaluated with the aim of analyzing angiogenesis. ALLO treatment significantly increased the percentage of relative vascular area for both markers (α-actin p = 0.0005; Von Willebrand p = 0.0058) compared to control (Fig. 10).

Figure 10
figure 10

Relative periendothelial area labeled with α-actin (αA) and relative vascular area labeled with von Willebrand factor (VW). Results are expressed as mean ± SEM. Paired Student’s t test (*p < 0.05/**p < 0.01; n = 5 ovaries/group). Representative microphotographs of sections from control (A and C) and ALLO 6 µM-treated (B and D) ovaries, stained with α-actin (A and B) and Von Willebrand factor (C and D). Light microscopy, 400 x, bar = 200 µm.

Corpora lutea angiogenesis

Lectin BS-1 staining significantly increased in CL after ALLO 6 µM treatment (Fig. 11).

Figure 11
figure 11

Upper panel: Representative microphotographs of sections from control (A) and ALLO 6 µM-treated (B) ovaries, stained with lectin BS-1. Light microscopy, 40 x, bar = 200 µm; inset 400 x, bar = 50 µm. Lower panel: Vascular area labeled with lectin BS-1. Results are expressed as mean ± SEM. Paired Student’s t test (*p < 0.05/**p < 0.01; n = 3 animals/group, 16 ovarian sections/group).

Ovarian expression of the GABAA Rα1-6 receptor

The expression of the ovarian GABAAR staining increased after ALLO 6 µM treatment (Fig. 12). The positive mark was found in stroma, blood vessels, CL and the theca of healthy and atretic follicles.

Figure 12
figure 12

Effect of intrabursal ALLO 6 µM administration on the ovarian expression of the GABAA receptor. Representative microphotographs of sections from control (left panel) and ALLO 6 µM-treated (right panel) ovaries, stained with anti GABAA receptor Rα1-6. Atretic follicles (A and F), antral follicles (B and G), preantral follicles (C and H), corpora lutea section (D and I), stroma with blood vessels (E and J). Light microscopy, 400 x, bar = 100 µm.

Experiment 2

Number of atretic follicles

BIC administration did not produce significant differences with respect to the control. The administration of BIC prior to ALLO prevented the increase in the number of ATF, showing a significant decrease compared to the administration of ALLO alone (F(3, 16) = 12.10; p = 0.0002) (Fig. 13).

Figure 13
figure 13

Effect of intrabursal ALLO 6 µM and BIC 10 µM administration on the number of atretic follicles (A) and on the diameter (µm) of CL (B). Results are expressed as mean ± SEM. (A) ANOVA I, post hoc Tukey; (B) Kruskal–Wallis with Dunn’s multiple comparison test (**p ≤ 0.01, ***p ≤ 0.0001; n = 5 animals/group, 51 CL/group).

CL diameter

None of the BIC concentrations per se induced significant changes in CL diameter (p < 0.0001). When BIC was administered prior to ALLO, values returned to normal (Fig. 13).

Discussion

Ovarian steroidogenesis and local steroid-mediated signaling is critical for proper ovarian function. An alteration in these processes may be the cause of various reproductive pathologies and infertility67. P4 biosynthesis and its metabolites are known as important regulators of the ovarian physiology. Considering that the concentration of P4 metabolites follows the same pattern as their precursor, and that the enzymes for its synthesis are found in ovarian tissue24, it is likely that derivatives are involved in local regulation. In this work, the effects of direct intrabursal ALLO administration were evaluated. It is important to note that the administration of the treatments was performed during the morning of the proestrus day, prior to the preovulatory LH peak, and at a time when ovarian P4 concentrations are low. The animals were sacrificed 24 h later, on the morning of estrus.

Plasma concentrations of E2 and P4 were measured to evaluate the endocrine response of the ovaries to the treatment. ALLO induced a significant increase in E2 concentration, detected on the morning of estrus. The peak of E2 during the normal estrous cycle of the rat typically occurs during the morning-afternoon of proestrus, with its plasma concentration decreasing towards estrus68,69. The increase in E2 during the follicular phase induces a positive feedback on GnRH neurons70,71,72,73, while a longer-lasting serum increase in E2 sensitizes the pituitary to GnRH action74. This coincides with the significant increase in plasma FSH produced by ALLO. As for the source of circulating E2, although an increase in the number of atretic follicles was observed, many were in early stages, so their granulosa cells (GCs) could still have been functional. Also, the increase in FSH could have stimulated the expression of follicular aromatase in preovulatory and non-ovulatory follicles, increasing secretion75. In addition, the CL of the rat retains its ability to secrete E2; hence, it could also have participated in this effect76,77,78.

ALLO significantly increased serum P4 concentration. The physiological peak of follicular P4 secretion occurs during the afternoon—night of proestrus, recovering basal values in the morning of estrus68,69. This peak is responsible for facilitating the preovulatory release of LH and occurs due to the action of FSH on the GCs of the preovulatory follicle79. Thus, an increase in P4 serum concentration could bring forward or magnify LH secretion80. It is known that an exogenous dose of subcutaneous P4 administered in the morning of proestrus produces a marked increase in ovarian periovulatory P4 secretion and sensitizes the tissue to the action of gonadotropins. In turn, FSH stimulation during estrus increases P4 synthesis by the GCs of the preovulatory follicle79,81. The results presented here suggest that the administration of ALLO during this period could have similar consequences, stimulating the release of P4 by CL and follicles. The increase in CL diameter produced by ALLO could also support this hypothesis. Another feasible hypothesis is that ALLO administration alters the inhibin-activin-follistatin system, necessary for the paracrine regulation of ovarian steroidogenesis and response to gonadotropins82, although more studies should be performed to confirm this.

There is evidence of the role of the GABAergic system in endocrine organs, where it acts as a modulator of hormone synthesis and release43,83. ALLO could be modulating the release of ovarian E2 and P4 through the GABAAR. Thus, the non-neural GABAergic system of these organs represents a possible therapeutic target for drugs that interact with GABA or its receptor. The changes in ovarian steroidogenesis induced by ALLO show effects that could destabilize the delicate endocrine balance. ALLO can alter hypothalamic regulation and induce pharmacological reprogramming of the CL. Pharmacological studies that include an in-depth analysis of systemic effects and implications are therefore of particular interest.

The intrabursal administration of ALLO 6 µM did not produce significant alterations in preovulatory follicles, thus allowing ovulation. Ovarian weight was also analyzed, and no significant differences were found for any of the treatments.

ALLO 6 µM treatment induced a significant increase in the number of atretic follicles interfering with follicular dynamics, affecting the selection and recruitment of dominant follicles for the following cycle. No significant changes in the diameter of tertiary follicles were observed. As for the CLs present in the treated ovaries, their number was not greater than that of the controls, but their diameter was significantly higher. During estrus, CLs from previous cycles are in different stages of luteolysis, showing areas of cell degeneration, fibrosis, and a reduction in size. At this stage, the newly formed CLs begin to increase in size84. As no changes in the number of ovulated oocytes were recorded, the increase in CL diameter could be interpreted as a delay in luteolysis of CL corresponding to previous cycles.

Pharmacological “reprogramming” of rodent CLs that have already embarked on the path of luteolysis is possible85. A postpartum subcutaneous injection of P4 in rats can inhibit luteal cell apoptosis86. Intrabursal administration of P4 to pregnant rats after activation of luteolysis by LH blocks induction of 20α-HSD activity, increases serum P4 concentrations and decreases ovarian PGF2α concentrations87. Thus, ALLO could interact with CL from previous cycles, recovering them from luteolysis. This finding could be useful for the treatment of reproductive conditions that involve deficiencies in CL functionality, or that are caused by early regression of the CL.

The expression of ovarian steroidogenic enzymes involved in P4 synthesis and metabolism was analyzed by real-time PCR. Treatment with ALLO induced an increase in the enzyme 3β-HSD2, which synthesizes P4, accompanied by a decrease in 20α-HSD, which metabolizes it to a derivative lacking progestational activity88,89. These findings are compatible with the increase in plasma P4 produced by ALLO and a possible luteotrophic effect. There were no significant changes in 3α-HSOR gene expression. ALLO does not seem to regulate mRNA expression of the enzyme responsible for the final step in its synthesis, unlike what happens with P4 in the rat90,91.

The expression of pro- and anti-apoptotic ovarian markers varies cyclically and between ovarian structures, determining their fate. ALLO treatment induced an imbalance in ovarian pro- and anti-apoptotic markers, with a significant effect on the FAS/FAS-L ratio, showing an anti-apoptotic effect that could correlate with the inhibition of luteolysis. During luteolysis, prolactin activates CL apoptosis via the FAS/FAS-L system92, and ALLO administration may override this effect. Cyclin D1 is a marker of cell proliferation that shows higher expression in non-luteinised theca and granulosa cells93. Its gene expression was significantly reduced after ALLO treatment, supporting previous findings of ALLO-mediated follicular atresia.

It is important to highlight that the ovarian parenchyma comprises two histologically and biochemically different structures that undergo cyclic changes. The mRNA expression results presented correspond to the processing of the whole ovary, making it difficult to determine which structure was affected. For the correct interpretation of results, it was necessary to complement them with other techniques to differentiate the effects on follicles and CL. Hence, a TUNEL technique and PCNA immunohistochemistry were performed. There was an increase in apoptosis of GCs and follicular theca cells, accompanied by a significant decrease in PCNA labeling, in concordance with the histological count. In the case of CLs, an opposite picture was observed, with a decrease in TUNEL labeling and an increase in PCNA. This result supports the luteotrophic effect of ALLO.

The process of apoptosis in the ovary is closely linked to variations in hormone secretion, driven by the regulation of the estrous cycle94. The increased follicular atresia found could be related to the hormonal imbalance caused by ALLO. Several studies95,96,97,98,99 have shown that the follicular fluid of atretic follicles contains altered levels of E2 and P4. However, elevated circulating concentrations of FSH, E2 and P4 could generate an ovarian trophic microenvironment, as they all possess anti-apoptotic effects on GC63,100,101,102,103,104,105. Thus, the ALLO stimulus on non-dominant follicles is decisive, since it occurs in an environment where other stimuli that inhibit follicular atresia are abundant. It is also possible that GABAAR activation induces an efflux of cellular chloride ions. In non-neural cells, chloride ions flow in both directions, according to the cellular electrochemical gradient, regulating osmotic tension106. Changes in intracellular ion concentration and cell volume could induce apoptosis107.

The development of the ovarian structures, as well as their function, depends on correct blood flow that allows the arrival of endocrine regulatory factors and the transport of the hormones produced. It has been demonstrated that the expression of angiogenic factors in the ovary is hormone-dependent; both the follicles and the CLs produce angiogenic factors108,109, which are then regulated by endocrine and paracrine factors. An increase in follicular vascularization selects dominant follicles, whereas its reduction is one of the first signs of atresia110,111. In the mature CL, approximately 50% of its cells correspond to endothelial cells112, and their blood flow correlates with P4 secretion113. Conversely, the decrease in CL blood flow is associated with luteolysis109,114. ALLO induced a significant increase in ovarian vascular and periendothelial area, immunolabeled with Von Willebrand and α-actin, respectively. This effect could be explained by the interaction between ALLO and the endothelial GABAergic system, where GABAAR activation induces an increase in cell proliferation and GABA synthesis115,116,117. ALLO could also modulate pro-angiogenic factors. We showed an increase in lectin BS-1 staining in the CL of ALLO treated ovaries; this finding could indicate increased blood flow to persistent CLs that have increased in diameter and functionality, thus increasing P4 secretion.

Functional studies determine the action of the ligand on a receptor and aid in the pharmacological characterization of the mechanism of action. In this case, to assess whether the action of ALLO in the ovarian parenchyma is GABAAR-dependent, a functional experiment was performed with BIC, a competitive antagonist of the receptor. This is the assay of choice to evaluate the interaction of agonists with GABAAR44.

GABA not only has functions on the central and peripheral nervous systems but is also present in a wide variety of peripheral tissues such as the intestine, stomach, gonads, and others118. Although the presence of GABAAR and a non-neural GABAergic circuit in the ovary is known, there are no studies that delve into its reproductive functionality. All the available studies correspond to modulation of the CNS and PNS7,9,119,120. The study of the effects of GABAAR system in the ovary is of particular interest due to the possible relationship with certain pathologies, such as polycystic ovary syndrome (PCOS) and ovarian cancer, as well as possible undesirable effects of the systemic application of ALLO treatments. In a previous study, ALLO administration to the superior mesenteric ganglion significantly increased GABAA receptor protein expression in the ovary. In this work, ALLO intrabursal administration had a similar effect, increasing GABAA receptor staining in stroma, blood vessels, CL and the theca of healthy and atretic follicles. This effect might increase ovarian GABAergic effects and thus be responsible for some of ALLO actions.

When BIC 10 µM was administered prior to ALLO, a return to control-like values of the number of atretic follicles and CL diameter was observed. These results indicate that the action of ALLO on these structures is mediated through GABAAR. There is a differential effect between structures: apoptosis in follicles is promoted, while luteolysis is delayed. These opposing effects could be explained by considering the plasticity of the receptor present in the different structures.

In the CNS, GABAAR shows plasticity throughout the estrous cycle, so that its expression and conformation seem to be linked to the circulating concentration of P4 and ALLO121,122,123,124,125,126. The physiological response of GABAAR depends on the subunit composition of the receptor, which determines its functional properties127,128. For example, GABAARs containing the ɗ subunit are more sensitive to the action of neuroactive steroids such as ALLO8,129. Finally, GABAAR distribution in the CNS is region-specific130,131. GABAAR activation may thus vary between receptor composition and location. One might expect that these same changes could occur in non-neural tissues. Fujii and Mellon’s work42 corroborates this theory; they report changes in the subunit composition of the uterine GABAAR during pregnancy and postpartum in the rat, with the subunits exhibiting differing sensitivity to ALLO. Thus, it would also be expected that the effects assessed in this work would be different in the different stages of the rat’s estrous cycle.

In conclusion, the administration of intrabursal ALLO 6 µM alters several processes of the ovarian morpho-physiology of the female rat, such as luteal regression, follicular atresia, angiogenesis and ovarian steroidogenesis, which could be related to oocyte quality and fertility. ALLO 6 µM exerts a pro-apoptotic effect on growing follicles, which is proliferative in CL and pro-angiogenic in the ovary. It also alters steroidogenesis and hormonal balance at the estrous stage. Some ALLO effects occurred through the GABAAR, as they were be inhibited with BIC.

Dysregulations of the processes of apoptosis132, steroidogenesis25,133,134 and angiogenesis135 are related to ovarian pathologies such as PCOS and cancer. The study of the ovarian effects of GABAAR modulation by neuroactive steroids is of particular importance since they modulate key aspects of reproductive physiology.