Corticosterone rather than ethanol epigenetic programmed testicular dysplasia caused by prenatal ethanol exposure in male offspring rats

ABSTRACT

Prenatal ethanol exposure (PEE) could affect offspring’s testicular development. This study aimed to illuminate its intrauterine origin and the programming mechanism caused by PEE. Pregnant Wistar rats were given ethanol (4 g/kg.d) by gavage administration during gestational days (GD) 9–20. Serum samples and testes of male offspring rats were collected on GD20, postnatal week (PW) 6, and PW12. We found that PEE induced testicular morphological abnormality, low serum testosterone levels, expressive suppression of 3β-hydroxysteroid dehydrogenase (3β-HSD), and low acetylation levels of histone 3 lysine 14 (H3K14ac) of 3β-HSD before and after birth. In utero, when fetal rats were overexposed to corticosterone by PEE, the expression levels of testicular glucocorticoid receptor (GR) and histone deacetylase 2 (HDAC2) were increased, while that of steroidogenic factor 1 (SF1) was decreased. In vitro, corticosterone (rather than ethanol) at 500 to 2,000 nM concentration decreased testosterone production and 3β-HSD expression in a concentration-dependent manner. Moreover, corticosterone downregulated SF1 and upregulated HDAC2 via activating GR, accompanied by a low H3K14ac level of 3β-HSD; SF1 overexpression could reverse the increased HDAC2 expression, and knockdown of HDAC2 could partially reverse the inhibitory effects of corticosterone on H3K14ac level and 3β-HSD expression but not on SF1 expression. Taken together, PEE caused testicular dysplasia in male offspring rats, which was associated with corticosterone-induced low-functional programming of 3β-HSD through the GR/SF1/HDAC2/H3K14ac pathway. This study provides new academic perspectives to illuminate the theory of ‘Developmental Origins of Health and Disease.’

Introduction

Ethanol is one of the most widely consumed substances of abuse [1]. According to a report from the Centers for Disease Control, among pregnant women, the prevalence of any alcohol use and binge drinking was 10.2% and 3.1%, respectively. Among women of childbearing age, the prevalence of any alcohol use and binge drinking was 53.6% and 18.2%, respectively [2]. Epidemiologic evidence suggested that the correlation between maternal ethanol consumption and increased incidence of hypospadias and cryptorchidism in adult men [3,4]. Experimental studies also showed that prenatal ethanol exposure (PEE) affected offspring’s development and reproductive system function in adulthood, such as decreased activities of testicular steroidogenic enzymes, low testosterone levels, and shortened anogenital distance [5]. Despite the well-known long-term testicular developmental toxicity of PEE on male offspring, there appears to be a significant knowledge gap regarding the effect of PEE during the intrauterine period and its programming mechanism.

Testosterone is known to play a key role in maintaining male reproductive system function [6]. In the testis, testosterone is synthesized from cholesterol, and a series of steroidogenic enzymes are responsible for testosterone biosynthesis. 3β-Hydroxysteroid dehydrogenase (3β-HSD) directly participates in the oxidation and isomerization of delta(5)-3beta-hydroxysteroid precursors into delta(4)-ketosteroids, thus catalyzing an essential step in the formation of all classes of active steroid hormones [7,8]. Studies have reported that the expression of 3β-HSD is regulated by a variety of transcriptional factors, including steroidogenic factor 1 (SF1) [9,10]. Studies on maternal phthalate exposure showed that the suppression of nuclear receptor SF1 was involved in the reduced 3β-HSD expression and testosterone biosynthesis in fetal rat testes [11]. These findings suggested that PEE-induced testicular development toxicity might be related to the low 3β-HSD expression mediated by SF1.

‘Intrauterine programming alteration’ implies that the structure and function of tissues are altered permanently by adverse environments during early life stages [12]. Aberrant epigenetic modifications are associated with the intrauterine programming alterations [13]. Xenobiotic exposure during early life has been shown to promote aberrant epigenetic modifications of steroidogenic enzymes [14]. Hong et al showed that exposure to bisphenol A in preimplantation embryo reduced histone acetylation of steroidogenic acute regulatory protein (StAR) to decrease testicular testosterone synthesis [15]. Arsenic exposure induced 3β-HSD upregulation by suppressing H3K9me2/3 status in Leydig cells and caused male reproductive dysfunction [16]. Increasing studies have reported that glucocorticoid overexposure is one of the main initiating factors of epigenetic processes, which likely involves intrauterine programming alterations [13,17]. Previous studies have shown that PEE could inhibit the placental 11β-hydroxysteroid dehydrogenase 2 (11β-HSD2) expression and thus induce fetal overexposure to maternal glucocorticoids [18,19]. Therefore, we proposed that PEE affected epigenetic programming of 3β-HSD mediated by glucocorticoid overexposure, which might lead to testicular dysplasia throughout the life.

In the present study, pregnant rats were treated with ethanol (4 g/kg.d) during middle and late pregnancy as reported previously in our study [20]. First, to confirm testicular dysplasia induced by PEE in male offspring rats, we detected the testicular morphological and functional changes before and after birth. Next, we detected histone acetylation modifications of 3β-HSD and investigated its possible effect and mechanism. Finally, we determined the effects of corticosterone and ethanol on testosterone synthesis in vitro and verified its molecular mechanism. This study provides experimental evidence and new academic perspectives to illuminate the theory of ‘Developmental Origins of Health and Disease (DOHaD).’

Materials and methods

Chemicals and reagents

Ethanol was purchased from Zhen Xin Co., Ltd. (Shanghai, China). Isoflurane was purchased from Baxter Healthcare Co. (Deerfield, IL, USA). A rat testosterone enzyme-linked immunosorbent assay (ELISA) kit was obtained from R&D Systems, Inc. (Minneapolis, MN, USA). Rat/mouse iodine [¹²⁵I] testosterone radioimmunoassay kits (S1094093) were purchased from the North Institute of Biological Technology (Beijing, China). The antibody of 3β-HSD (sc30820) was purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). Antibodies such as β-actin (AC004), HDAC2 (A2084), glucocorticoid receptor (GR) (A2164), and anti-acetyl histone 3 Lysine 14 (H3K14ac) (A7254) were purchased from Abclonal Technology Co., Ltd. (Wuhan, China). Antibodies such as immunoglobulin G (IgG) (ab172730) and Ki67 (ab15580) were purchased from Abcam Technology Co., Ltd. (Cambridge, UK). 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-Tetrazolium, inner salt (MTS) assay kit was purchased from Cayman Chemical Co. (Ann Arbor, Michigan, USA). Reverse transcription and quantitative real-time polymerase chain reaction (qRT-PCR) kits (Q223) were purchased from Takara Biotechnology Co., Ltd. (Dalian, China). A DNA purification kit (Q5314) was purchased from TIANGEN Biotech Co., Ltd. (Beijing, China). Mifepristone (RU486) (ODR4395) and proteinase K (ST533) were purchased from Kori Biotech Co., Ltd. (Wuhan, China). HDAC2 siRNA and plasmid SF1 pcDNA3.1(+) vector (CN14379-1) were purchased from Biosci Biotech Co., Ltd. (Wuhan, China). The other reagents for experiments were of analytical grade.

Animals and treatment

Specific pathogen-free Wistar rats (No. 2012–2014, license number: SCXK (Hubei), certification number: 42000600002258) weighing 280 ± 20 g (males) and 200 ± 20 g (females) were purchased from the Experimental Center of the Hubei Medical Scientific Academy (Wuhan, China). Animal experiments were performed in the Center for Animal Experiments of Wuhan University (Wuhan, China), which is accredited by the Association for the Assessment and Accreditation of Laboratory Animal Care International (AAALAC International). The protocol was approved by the Committee on the Ethics of Animal Experiments of the Wuhan University School of Medicine (permit number: 14016). All animal experimental procedures were performed in accordance with the guidelines of the Chinese Animal Welfare Committee. Wistar rats were housed in cages with wire-mesh floors in standard conditions and allowed a normal diet after one week of acclimation, and two female rats were mated with one male rat overnight. Upon confirmation of mating by the appearance of sperm in a vaginal smear, the day was taken as gestational day (GD) 0. According to our previously describled [20], a 60-kg pregnant woman drinking 960 mL of beer 351 or 320 mL of red wine daily is equivalent to ethanol exposure of 0.65 g/kg.d. Considering the dose conversion between humans and rats (human: rat = 1:6.17) [21], the dose of ethanol exposure roughly corresponds to 4 g/kg.d for a pregnant rat. Thus, the pregnant rats of the PEE group (n= 20) were given 4 g/kg.d ethanol by gavage administration during GD9 to GD20, and those of the control group (n= 20) were administered the same volume of saline. The flowchart of animal treatment is shown in Figure 1.

Figure 1.

Timetable and schematic procedure of rat treatment from gestational day (GD) 0 to postnatal week (PW) 12.

Figure 2.

Effects of prenatal ethanol exposure (PEE, 4 g/kg.d) on testicular morphology in fetal rats.

(a, b) Testicular morphological analysis by H&E staining (40× for A, 200× for B, n = 5). (c) Ki67 protein expression by IHC (black represent Ki67 positive cells, 200×, 400×, n= 5). (d) Leydig cells structure by transmission electron microscopy (5000×, n = 5). Comparisons between groups were performed using Student’s t-test, mean ± S.E.M. *P < 0.05 vs. control. Red arrow: interstitial region; yellow arrow: mitochondria.

Part of pregnant rats were anesthetized with isoflurane and sacrificed on GD20 (n= 10 pregnant rats per group), and the blood sample was collected from the carotid artery. Those dams with litter sizes of 10–14 were considered to qualify for the experiment. The male fetuses were decapitated immediately to collect blood sample and testes, and the blood samples of different male fetuses from each litter were pooled as one sample. Three whole fetal testes in three different fetal rats from each litter were randomly counted as one sample and processed for gene and protein analysis. The other 10 pregnant rats in each group were allowed to deliver normally. On postnatal day (PD) 1, the number of pups was normalized to 8 per litter (each litter retained 4 male and 4 female pups, n = 8 litters per group), and we provided them adequate and standardized nutrition until weaning at postnatal week (PW) 4. After weaning, the rats were given a normal diet. At PW6 and PW12, one male rat was randomly obtained from each litter and anesthetized with isoflurane and sacrificed. Then, the blood sample from the carotid artery was collected，the serum was isolated, and the testes and epididymis tissues were removed rapidly. Except for epididymis tissues that were immediately used for the sperm analysis, portions of the male right testes obtained from different litters were fixed for morphological observations, and the other collected samples were immediately frozen at – 80°C for subsequent analysis.

Sperm count and motility examinations

Samples were obtained from the right caudal epididymides of rats at W12. Epididymal sperm count and motility were analyzed as previously described [22]. The right caudal epididymides were cleaned and minced in 1 mL of phosphate-buffered saline (PBS, pH 7.4, 37°C), and they were then pierced four to six times using a scalpel blade. The sperms were allowed to fully swim out from the caudal epididymides for 15 min and subjected diagnostic sperm analysis according to World Health Organization (2010) guidelines. 10 μL of sperm suspension was placed on a preheated (37°C) slide, covered with a cover slip, and immediately examined using a light microscope at 100 × . The sperm count was determined by a hemocytometer and expressed as × 10⁶ cells/mL. Sperm motility was evaluated by counting the motile and nonmotile sperms and assessed up to 200 sperms under a 400 × magnification of a light microscope. Ten microscopic fields were selected randomly, and the average percentage of motile spermatozoa was determined in each group.

Serum sample analysis

To measure the intratesticular testosterone contents, testicular samples were placed on the ice to unfreeze. Three whole fetal testes were pooled as one sample. A 0.5-mg adult testicular tissue was cut as one sample and homogenized individually in 0.5 mL PBS. Testosterone contained in testicular tissue extracts, testosterone contents of adult rats, and that in culture medium were detected using [¹²⁵I] testosterone radioimmunoassay kit. Testosterone contents of fetal rats were measured using a testosterone ELISA kit, and all assay procedures followed the manufacturer’s protocols; intratesticular testosterone content was expressed as ng/mg. The intra- and inter-assay coefficients of variation were 10.0% and 15.0% for testosterone, respectively.

The concentration of fetal serum corticosterone was determined using an ELISA assay kit according to the manufacturer’s instructions [23]. The intra- and inter-assay coefficients of variation were 5.0% and 7.2% for corticosterone, respectively.

Histological and structural analysis

The right testes were fixed in 4% paraformaldehyde overnight and processed with the paraffin sectioning technique. Sections (5-μm thick) were rehydrated and stained with hematoxylin-eosin (H&E) and then observed and photographed with an Olympus AH-2 light microscope (Olympus, Tokyo, Japan). The sections with the maximum area were obtained from fixed random right testes per group (n= 5), and five random fields of each section were observed under the microscope by two laboratory members blind to the exposure group. The maximal diameter and area of fetal testes were measured using the image analysis system (Olympus, Tokyo, Japan); the diameter of seminiferous tubule and the thickness of the seminiferous epithelium were also determined. Finally, we evaluated relevant examination indexes from each photomicrograph and took the average value per section.

Fetal testes were fixed with 0.1 M of PBS containing 4% formaldehyde, 2.5% glutaraldehyde formaldehyde, and 1% osmium tetroxide. The samples were dehydrated through a concentration gradient ethanol and embedded in EPON812. Ultra-thin sections (50-nm thick) were cut using an LKB-V ultramicrotome (Bromma, Sweden), stained with uranyl acetate and lead citrate, and examined with an H-600 transmission electron microscope (TEM) (Hitachi, Tokyo, Japan).

Immunohistochemistry (IHC) and immunofluorescence (IF) measurements

For IHC analysis, the sections (n= 5) were incubated overnight at 4°C with the anti-Ki67 (1:1,000). Immunohistochemical analysis was performed using a 3,3′-diaminobenzidine (DAB) staining kit to determine the expression levels of Ki67 proteins in the testes. Immunostaining for the negative control was performed on a parallel section, in which the primary antibody was replaced with nonimmune rabbit anti-IgG. The intensity of staining was determined by measuring the mean optical density in five random fields for each section.

For IF analysis, the sections (n= 5) were incubated overnight at 4°C with the primary antibody (3β-HSD 1:500). After rewarming for 15 min, corresponding fluorescent secondary antibodies (1:400) were added to the sections, and incubated at room temperature for 1 h in the dark. Nuclear counterstain (4′,6-diamidino-2-phenylindole (DAPI); Sigma-Aldrich) was diluted 1:500 in Tris-buffered saline and incubated for 10 min. As described previously [20], the number of Leydig cells per unit square of interstitial tissue areas (10⁴ μm²) were calculated by examining 25 randomly selected sites in each group to avoid a sampling bias. All images were captured using an Olympus AH-2 Light Microscope (Olympus, Tokyo, Japan). Analysis of the stained images was performed using Olympus software.

Cell culture, transfection, and treatment

The mouse Leydig tumor cell line (MLTC-l) was purchased from Pro-cell Co., Ltd. (Wuhan, China). The cells were cultured in 1640 medium with 10% fetal bovine serum (Hyclone, China) and 0.1% penicillin/streptomycin and incubated in a 5% CO₂ humidified incubator at 37°C. According to the above experimental data in vivo, ethanol and corticosterone concentrations of fetal serum were approximately 60 mM and 1,224 nM, respectively [23]. Thus, we treated the MLTC-1 cells with ethanol (0, 30, 60, and 120 mM) or corticosterone (0, 500, 1,000, and 2,000 nM) for 48 h. The MTS assay was conducted to detect the cytotoxicity of ethanol and corticosterone on the MLTC-1 cells following the manufacturer’s protocol. Absorption intensity was measured at 490 nm using a microplate reader (TECAN, Australia). With the same treatment, the testosterone concentration in culture medium and mRNA expression of 3β-HSD in MLTC-1 cells were measured. To confirm that the inhibitory effects of corticosterone were mediated by upregulating GR expression, the cells were co-treated with 1,000 nM of corticosterone and 2.5 μM of RU486 (a GR inhibitor) for 48 h. Additionally, for HDAC2 knockdown, RNA interference technology was used. MLTC-1 cells were transiently transfected with the siRNAs targeting mouse HDAC2 or with nontargeting control siRNA using 10 μL of Lipofectamine 3000 (Invitrogen, Carlsbad, CA, USA) at a final concentration of 20 nM. After 12 hr, the medium was exchanged for a fresh medium, and the cells were treated with 1,000 nM of corticosterone. The cells were harvested for further analysis after 48 h. To overexpress SF1, MLTC-1 cells were plated and transfected using 5 μg plasmid pcDNA3.1(+) vector in combination with 5 μL of Lipofectamine 3000 and 10 μL of P3000 according to the manufacturer’s instructions. After 12 h of transfection, the cells were treated as described above.

Total RNA extraction, reverse transcription, and qrt-pcr

Testicular tissues or cultured cells were homogenized in Trizol reagent. Total RNA was single-strand cDNA prepared from 2 μg of total RNA according to the protocol of the Applied Biosystems TaqMan Reverse Transcription reagent kit. Primers of StAR, cytochrome P450 cholesterol side chain cleavage (P450scc), 3β-HSD, 17 alpha-hydroxysteroid dehydrogenase 1 (17α-HSD₁), 17 beta-hydroxysteroid dehydrogenase 3 (17β-HSD₃), glyceraldehyde 3-phosphate dehydrogenase (GAPDH), and other primers were designed using Primer Premier 5.0 (PREMIER Biosoft International, CA). The sequences of primers used in this experiment are shown in Table 1. The cycle thresholds (Ct) were detected, and the relative expression of genes were determined using the 2^△△Ct method with normalization to GAPDH expression that was used as a quantitative control.

Table 1.

Oligonucleotide primers and PCR conditions for real-time quantitative PCR.

Species	Genes	Forward primer	Reverse primer	Annealing (°C)
Rat	GAPDH	GCAAGTTCAATGGCACAG	AAGTTCTTCCTGGCCGGTAT	63(30 s)
	StAR	GGGAGATGCCTGAGCAAAGC	GCTGGCGAACTCTATCTGGGT	65(30 s)
	P450scc	GCTGCCTGGGATGTGATTTTC	GATGTTGGCCTGGATGTTCTTG	63(30 s)
	3β-HSD	TCTACTGCAGCACAGTTGAC	ATACCCTTATTTTTGAGGGC	58(30 s)
	17α-HSD1	CAATCTCTGGGCACTGCATC	ACTCTGCGTGGGTGTAATGA	60(30 s)
	17β-HSD3	CCACTGCAACATTACCTCCG	CTATACAGAGGCCAGGGACG	60(30 s)
	GR	CACCCATGACCCTGTCAGTC	AAAGCCTCCCTCTGCTAACC	63(30 s)
	HDAC2	GGACAAGAGGACAGATGTTAAGG	GGGTTGTTGAGTTGTTCTGATTT	60(30 s)
	SF1	CTGAGGGAGACTCCTGGAAA	GTGAAATTGGTTAAGGGCATG	60(30 s)
Mouse	GAPDH	GGCTCTTCCAGAACAGATTAG	CCGAGGCCAAGTTAAGAATAG	60(30 s)
	3β-HSD	TATCGCAGACCCAGATAGAG	GCTGAAGATGGACAGACTTG	60(30 s)
	GR	ACCTGGAAGCTCGAAAAACGA	CAGCAGTGACACCAGGGTAG	60(30 s)
	HDAC2	CTATTCCAGAGGATGCTGTTC	GCCTTCTTTGCTCCTTTCT	60(30 s)
	SF1	GGGAAAGTTGGGCGTAAA	TAGATCGTTCTGGTCCTCTG	60(30 s)

GAPDH, glyceraldehyde phosphate dehydrogenase; StAR, steroidogenic acute regulatory protein; P450scc, cytochrome P450 cholesterol side chain cleavage; 3β-HSD, 3β-hydroxysteroid dehydrogenase; 17α-HSD1, 17α-hydroxysteroid dehydrogenase 1; 17β-HSD3, 17β-hydroxysteroid dehydrogenase 3; GR, glucocorticoid receptor; HDAC2, histone deacetylase 2; SF1, steroidogenic factor 1.

Chromatin immunoprecipitation-polymerase chain reaction (chip-pcr)

The homogenate of testicular tissues or scraped cells were fixed with 1% formaldehyde for 15 min at 37°C to cross-link DNA and its associated proteins. Glycine (0.125 M final concentration) was added to terminate the reaction for 8 min. The lysates were then sonicated to shear the DNA to a size of 200–800 bp. After sonication, the samples were collected by centrifugation and diluted with dilution buffer. After mixing, 10 μL of the supernatant was saved as input DNA for the normalization of chromatin input. The remainder was divided into 200 μL per Eppendorf tube and subjected to clear with bovine serum albumin (BSA)-treated protein agarose beads; 2 μL of anti-H3K14 or IgG was added, and the mixture was incubated overnight at 4°C with rotation. The immunoprecipitated DNA-protein complex that was linked to beads was collected by centrifugation and washed sequentially with low-salt, high-salt, LiCl immune complex, and Tris-EDTA washing buffer. Prepared elution buffer was used to elute the DNA-protein complex, and each elution was repeated twice. Samples were incubated overnight at 65°C with 200 µg/mL of proteinase K and were subsequently purified using a DNA purification kit, following the manufacturer’s protocol. Purified DNA was finally dissolved in 100 μL of elution buffer.

The purified DNA was assayed using qRT-PCR. The sequences of the primers spanning the 3β-HSD binding region are as follows: rat: GAAGGGGAAGGGGTTCAGAA (forward) and GGCATGAATTAGGGAGGGGT (reverse), mice: GCAGCTGTTGTCCATCTTGT (forward) and TTCTTGTGAGGTCCCAGTGG (reverse). With the IgG-negative control values subtracted as background, the input values were used for quantification and normalized to their corresponding values of immunoprecipitation (IP) samples following formula (IP/input = 2 ^{Ct input DNA – Ct IP DNA}).

Western blotting

Protein concentration was determined using the BCA Protein Assay Kit (Sigma, St. Louis, MO). Western blotting was performed as described previously. In brief, protein samples were separated on 12% sodium dodecyl sulphate-polyacrylamide gels before transferring to polyvinylidene difluoride membranes. Then, membranes were blocked with 5% non-fat milk powder for 1.5 h, and incubated with various primary antibodies: 3β-HSD (1:1,000), GR (1:1,000), HDAC2 (1:1,000), and β-actin (1:5,000), respectively, overnight at 4°C followed by secondary antibodies anti-rabbit 3β-HSD and HDAC2 (1:3,000), and anti-mouse β-actin and GR (1:5,000) conjugated with horseradish peroxidase (HRP). Finally, the protein bands were detected using the ECL Plus Kit, and the mean optical density represented the relative protein levels. Detection of β-actin was used as a loading control.

Statistical analysis

SPSS 19 (SPSS Science Inc., Chicago, Illinois) and Prism 6.0 (Graph Pad Software, La Jolla, CA, USA) were used to perform data analysis. Quantitative data were expressed as mean ± S.E.M. Two-tailed Student’s t-test was used for comparisons between control and treatment groups, and for studies involving more than two groups, data were evaluated with one-way analysis of variance (ANOVA) followed by Tukey’s post hoc test. Statistical significance was defined as P < 0.05.

Results

Effects of PEE on testicular morphology in fetal rats

We first performed H&E staining to observe fetal testicular morphological changes. The analysis showed that the fetal testicular maximum area and maximum diameter of the PEE group were decreased (P< 0.05, Figure 2(a)). The interstitial region among the seminiferous tubules appeared slightly enlarged (red arrowhead), but there were no other significant pathological changes (Figure 2(b)). Moreover, immunolocalization for Ki67 demonstrated that the cell proliferation was reduced (P< 0.05, Figure 3(c)). TEM images showed obvious mitochondrial vacuolation in Leydig cells of the PEE group (yellow arrowhead) (Figure 2(d)). These results suggested that PEE could inhibit testicular morphological development and alter the ultrastructure of Leydig cells in the fetal rats.

Figure 3.

Effects of prenatal ethanol exposure (PEE, 4 g/kg.d) on testosterone levels and testicular steroidogenic enzymes’ expression in fetal rats.

(a, b) Testosterone levels by ELISA assay (n = 8). (c) Testicular steroidogenic enzymes’ expression by qRT-PCR (n= 8). (d) 3β-HSD protein expression by Western blotting (n= 3). (e) Leydig cells numbers by IF (red represent 3β-HSD positive cells, 400×, n= 5). Comparisons between groups were performed using Student’s t-test, mean ± S.E.M. *P < 0.05, **P < 0.01 vs. control.

Effects of PEE on testosterone synthetic function in fetal rats

Next, we detected testosterone levels and expression of steroidogenic enzymes to evaluate testosterone synthetic function in the fetal rats. As shown in Figure 3, fetal serum testosterone concentration and intratesticular testosterone content in the PEE group were decreased (P< 0.05, Figure 3(a,b)). In the PEE fetal testis, when compared with the control, the mRNA expression of steroidogenic enzymes (StAR, 3β-HSD and 17α-HSD₁) was inhibited (P< 0.05, P< 0.01, Figure 3(c)), and the protein expression of 3β-HSD reduced (P< 0.01, Figure 3(d)). Measurement of Leydig cell numbers using the optical dissector method (3β-HSD⁺, a specific biomarker of Leydig cells) showed a significant reduction in the number of Leydig cells in the PEE group (P< 0.05, Figure 3(e)). The above results indicated that PEE could inhibit the testosterone synthetic function in the fetal rats, including reducing the expression levels of steroidogenic enzymes and the number of Leydig cells.

Effects of PEE on serum corticosterone level, testicular GR/SF1/HDAC2 expression, and H3K14ac level of 3β-HSD in fetal rats

We further investigated the potential mechanism of the inhibited testosterone synthesis by PEE in the fetal rats. The results showed that the serum corticosterone level was dramatically increased in the PEE group (P< 0.01, Figure 4(a)). The expression of testicular GR did not alter in the mRNA level but increased significantly in the protein level, and the expression of testicular SF1 was decreased in both mRNA and protein levels (P< 0.05, Figure 4(b,c)). Because the modification of histone acetylation could be catalyzed by histone deacetylases (HDACs), we screened multiple HDACs via the mRNA expression detection, and found that the mRNA and protein expression levels of HDAC2 were increased significantly (P< 0.01, Figure 4(b,c)). The result further indicated that the H3K14ac level of 3β-HSD promotor in the PEE group was lower than that of the control (P< 0.05, Figure 4(d)). These results suggested that PEE induced fetal corticosterone overexposure, and increased the expression levels of GR and HDAC2, while it decreased the expression of SF1 and the H3K14ac level of 3β-HSD.

Figure 4.

Effects of prenatal ethanol exposure (PEE, 4 g/kg.d) on serum corticosterone level, testicular GR, SF1 and HDAC2 expression, and H3K14ac level of 3β-HSD in fetal rats.

(a) The serum corticosterone level by ELISA assay (n= 8). (b) The mRNA expression of GR, SF1 and HDAC2 by qRT-PCR (n= 8). (c) The protein expression of GR, SF1 and HDAC2 by Western blotting (n= 3). (d) The H3K14ac level of 3β-HSD by ChIP-PCR (n= 3). Comparisons between groups were performed using Student’s t-test, mean ± S.E.M. *P < 0.05, **P < 0.01 vs. control. CORT, corticosterone.

Effects of PEE on testicular morphology in offspring rats after birth

The testicular morphological changes after birth in offspring rats are shown in Figure 5. The testicular volume and index were persistently decreased by PEE from puberty (PW6) to adulthood (PW12) (P< 0.05, Figure 5(a,b)). H&E staining showed that the interstitial regions among seminiferous tubules were enlarged, and Leydig cells were accumulated (red arrowhead) (Figure 5(c)). The diameter of seminiferous tubules and the thickness of seminiferous epithelium exhibited significant reductions in the PEE offspring from PW6 to PW12 (P< 0.05, Figure 5(d)). These findings showed that PEE-induced testicular morphological abnormality persistently existed after birth in the offspring rats.

Figure 5.

Effects of prenatal ethanol exposure (PEE, 4 g/kg.d) on testicular morphology in offspring rats after birth.

(a, b) Testicular volume and index (n= 8). (c) Testicular morphological analysis by H&E staining (200×, n= 5). (d) Seminiferous epithelium thickness and seminiferous tubule diameter by H&E staining. Comparisons between groups were performed using Student’s t-test, mean ± S.E.M., *P < 0.05 vs. control. PW, postnatal week. Red arrow: interstitial region.

Effects of PEE on testosterone synthetic function and H3K14ac level of 3β-HSD in offspring rats after birth

To investigate the effects of PEE on testosterone synthetic function, we also examined the expression of steroidogenic enzymes and the number of Leydig cells in PW6 and PW12. Our results showed that the serum testosterone level and intratesticular testosterone content of the PEE offspring were both decreased in PW6 and PW12 (P< 0.05, P< 0.01, Figure 6(a,b)). Remarkably, the mRNA expression of StAR, P450scc, and 3β-HSD was decreased in PW6, but only the obvious reduction of 3β-HSD still existed in PW12 (P< 0.05, P< 0.01, Figure 6(c)). In both PW6 and PW12 of the PEE group, the western blotting showed the sustained inhibition of 3β-HSD protein levels (P< 0.05, P< 0.01, Figure 6(e)); the IF measurement indicated the reduced numbers of Leydig cells (3β-HSD+) (P< 0.05, Figure 6(d)), and the ChIP-PCR assay showed the decreased H3K14ac levels of 3β-HSD (P< 0.05, Figure 6(f)). Additionally, we found that sperm count was decreased (P< 0.01), whereas sperm motility did not present a significant change in the PEE adult offspring rats (Figure 6(g)). Collectively, these results indicated that PEE could decrease testicular testosterone synthetic function after birth in offspring rats.

Figure 6.

Effects of prenatal ethanol exposure (PEE, 4 g/kg.d) on testosterone levels, testicular steroidogenic enzymes’ expression and H3K14ac levels of 3β-HSD in rats after birth.

(a, b) Serum and intratesticular testosterone levels by iodine [¹²⁵I] testosterone radioimmunoassay kit (n= 8). (c) Testicular steroidogenic enzymes’ mRNA expression by qRT-PCR (n= 8). (d) Testicular 3β-HSD protein expression by Western blotting (n= 3). (e) The number of Leydig cells by IF (red represent 3β-HSD positive cells, 400×, n= 5). (f) The acetylation level of 3β-HSD by ChIP-PCR (n= 3). (g) Sperm counts and motility through the hemocytometer and optical microscope (n= 8). Comparisons between groups were performed using Student’s t-test, mean ± S.E.M. *P < 0.05, **P < 0.01 vs. control. PW, postnatal week.

Effects of ethanol and corticosterone on testosterone concentration, 3β-HSD, and GR expression in MLTC-1 cells

To clarify whether the reduced testosterone synthetic function in the PEE offspring was caused by ethanol or corticosterone, we carried out further experiments in MLTC-1 cells. No cytotoxicity was observed in up to 120 mM of ethanol or 2,000 nM corticosterone treatment for 48 h (Figure 7(a,b)). Furthermore, ethanol (≤ 60 mM) promoted testosterone production and 3β-HSD expression, especially at concentrations of 60 mM (P< 0.05, P< 0.01, Figure 7(c,d)). In contrast, these indicators were inhibited in a concentration-dependent manner following the corticosterone treatment (P< 0.05, P< 0.01, Figure 7(e,f)). Additionally, we detected GR expression and found that it was upregulated with corticosterone treatment (P< 0.05, Figure 7(g)). By administering 2.5 μM of RU486 (a GR inhibitor), the inhibitory effects of corticosterone on 3β-HSD expression and testosterone production were markedly reversed (P< 0.05, Figure 7(h,i)). All these findings showed that corticosterone (rather than ethanol) inhibited testosterone synthetic function in MLTC-1 cells.

Figure 7.

Effects of ethanol and corticosterone on testosterone concentration, 3β-HSD and GR expression in the MLTC-1 cells.

The MLTC-1 cells were cultured in the presence of 0–2000 nM corticosterone or 0–120 mM ethanol, and then co-treated with 1000 nM corticosterone and 2.5 μM RU486 for 48 h. (a, b) Cytotoxicity of corticosterone by MTS assays. (c, e, h) Testosterone concentration in cell culture medium by Iodine [¹²⁵I] testosterone radioimmunoassay kit. (d, f, i) 3β-HSD mRNA expression by qRT-PCR. (g) GR mRNA expression by qRT-PCR. Comparisons between groups were performed using Student’s t-test (g) and ANOVA (A-F, H-I), mean ± S.E.M., n= 6 for all examination indexes. *P < 0.05, **P < 0.01 vs. respective controls. CORT, corticosterone; RU486, mifepristone.

Effects of HDAC2 siRNA and SF1 pcDNA3.1(+) plasmid vector on the expression of HDAC2, SF1 and 3β-HSD, and H3K14ac levels of 3β-HSD in MLTC-1 cells

On the basis of the results in vivo, the following experiments were performed to confirm the molecular mechanism of 3β-HSD downregulation by corticosterone in vitro. We found that corticosterone could decrease the H3K14ac level of 3β-HSD, while increasing the expression of HDAC2 (P< 0.05, Figure 8(a,b)). To further confirm that the increased HDAC2 expression was involved in the H3K14ac modification of 3β-HSD, MLTC-1 cells were transfected with HDAC2 siRNA and its nontargeting control. The knockdown efficiency of HDAC2 was confirmed (P< 0.01, Figure 8(c)) and the knockdown of HDAC2 could partially reverse the suppressive effects of corticosterone on H3K14ac and expression of 3β-HSD (P< 0.01, P< 0.05, Figure 8(d,e)). These results suggested that HDAC2 was involved in the downregulation of 3β-HSD.

Figure 8.

Effects of HDAC2 siRNA and SF1 pcDNA3.1(+) plasmid vector on the expression of HDAC2, SF1 and 3β-HSD, and H3K14ac level of 3β-HSD in MLTC-1 cells.

The MLTC-1 cells were cultured in the presence of corticosterone (1000 nM) with HDAC2 siRNA (20 nM), SF1 pcDNA3.1(+) plasmid vector (5 μg) or RU486 (2.5 μM) for 48 h. (a, d) The level of H3K14ac in the 3β-HSD promoter by ChIP-PCR (n= 3). (b, c, h) HDAC2 mRNA expression by qRT-PCR (n= 6). (E) 3β-HSD mRNA expression by qRT-PCR (n= 6). (F, G, I, J) SF1 mRNA expression by qRT-PCR (n= 6). Comparisons between groups were performed using Student’s t-test (a–c, f–g) and ANOVA (D-E, H-J), mean ± S.E.M. *P < 0.05, **P < 0.01 vs. respective controls. NC, negative control; CORT, corticosterone; RU486, mifepristone.

In addition, corticosterone could downregulate the expression of SF1 in MLTC-1 cells (P< 0.05, Figure 8(f)). To verify the role of SF1 in the pathway, SF1 pcDNA3.1(+) plasmid vectors were used to overexpress the SF1 gene. The results revealed that SF1 expression was significantly elevated after SF1 plasmid vector infection (P< 0.05, Figure 8(g)), and the overexpression of SF1 could completely reverse the activation effect of corticosterone on HDAC2 expression (P< 0.05, Figure 8(h)). However, the knockdown of HDAC2 could not alter the inhibitory action of corticosterone on SF1 expression (P< 0.01, Figure 8(i)). Furthermore, we found that the inhibitory effect of corticosterone on SF1 expression could be completely reversed by RU486 (P< 0.05, Figure 8(j)). Taken together, all these findings suggested that corticosterone decreased H3K14ac levels of 3β-HSD and expression of 3β-HSD through the GR/SF1/HDAC2 pathway.

Discussion

Pee-induced testicular dysplasia after birth originated from the intrauterine period

Studies found that PEE could induce reproductive dysfunction in adulthood males, such as abnormal morphology of seminiferous tubules, lower weight of testes, and delayed spermatogenesis [24,25]. In the present study, when pregnant rats were treated with ethanol (4 g/kg.d) during middle and late pregnancy, we found decreased testicular volume and index, enlarged interstitial region among seminiferous tubules, and reduced serum and intra-testicular testosterone levels from puberty to adulthood. Testosterone biosynthesis is known to occur primarily in Leydig cells, and the number of Leydig cells partially determines the level of circulating testosterone [26,27]. We found that the number of Leydig cells was significantly reduced by PEE, which might be partially responsible for the low testosterone production. Studies reported that the low testosterone level could lead to the dysfunction of Sertoli cells, which might be a vital factor contributing to the retardation of spermatogenesis [28,29]. Similarly, our results showed that the diameter and epithelium height of seminiferous tubules and the sperm counts of adult rats were reduced by PEE.

What do these postnatal changes have to do with early testicular development? Previous studies have indicated that exposure to internal and external environmental stimuli in early life had pronounced effects on developmental trajectories of individuals, which may permanently affect the morphology and function of multiple organs [30,31]. The increased incidence of male reproductive dysfunction was believed to result from xenobiotic-induced perturbations of gonadal development in fetuses [32]. Hence, we traced the testicular morphology and function of PEE offspring rats back to the intrauterine period. Similar to our findings in puberty and adulthood, the testicular size and cell proliferation capacity were decreased in fetal rats, together with the number of Leydig cells and testosterone production. Taken together, PEE-induced testicular dysplasia after birth was found to originate from the intrauterine period.

The low-functional programming of 3β-HSD by corticosterone (rather than ethanol) may contribute to testicular dysplasia

Testosterone synthesis is a multistep process involving a series of transporters and enzymes. In this process, pregnenolone is converted to progesterone by the catalytic action of 3β-HSD, followed by androstenedione production with the participation of 17α-HSD, and alteration of any one step in this signaling pathway influences the production of testosterone [33,34]. Kim et al reported that the repetitive testicular heat-treatment in mice induced the downregulation of 3β-HSD and the apoptosis of Leydig cells, which decreased testosterone production [35]. In this study, we found that the testicular steroidogenic enzymes were decreased in different degrees in GD20, PW6, and PW12 of PEE. The expression of 3β-HSD was continuously suppressed at all time points. These results indicated that the sustained inhibition of testosterone synthesis in the PEE offspring was related to the low-functional programming of 3β-HSD.

We previously showed that PEE fetal rats overexposed to maternal glucocorticoids might inhibit the development of the multiple organs, such as adrenal glands and bone [23,36]. Additionally, maternal glucocorticoid exposure had potential impacts on reproductive dysfunction in male offspring. For example, prenatal stress exposure induced low levels of serum testosterone, delayed puberty establishment, and caused abnormal sexual behaviors [37,38]. Glucocorticoids act by binding to intracellular GR [39]. In the testis, the GR is localized to multiple cell types (including Leydig cells). GR regulates testosterone synthesis and sperm maturation [40,41]. In this study, we also found that PEE fetal rats were exposed to high levels of corticosterone, and the expression of GR was activated. In vitro, the results further demonstrated that corticosterone could upregulate GR expression, while inhibiting 3β-HSD expression and the production of testosterone in a concentration-dependent manner; further, GR inhibitor (RU486) could completely reverse these effects. Interestingly, 60 mM of ethanol in serum could significantly increase testosterone production in MLTC-1 cells. Taken together, these results suggested that the corticosterone (rather than ethanol) inhibited testosterone synthetic function, and the testicular dysplasia of PEE offspring rats was related to low-functional programming of 3β-HSD by testosterone.

GR/SF1/HDAC2 pathway-mediated low h3k14ac of 3β-hsd programmed its low expression

The nuclear receptor SF1 plays key roles throughout the reproductive axis, including the gonads [42]. SF1 has been shown to increase expression of the steroidogenic enzymes (including StAR, P450scc, and 3β-HSD) by binding to its response element site found in the promoter regions [43,44]. In the present study, PEE could downregulate SF1 in utero, but the testicular SF1 expression did not show significant changes from puberty to adulthood. Furthermore, corticosterone decreased SF1 expression in vitro. Hence, we hypothesized that SF1 was not the reason for the sustained 3β-HSD downregulation, but it might be involved in the occurrence of intrauterine programming.

Epigenetic modifications can be influenced by harmful environmental factors in early life and permanently impact gene expression [45]. H3K14 is the primary acetylated site of histone H3, and H3K14ac is tightly correlated with transcription activation [15]. Our present study found that the H3K14ac level of 3β-HSD was persistently decreased in GD20, PW6, and PW12, and we obtained similar results in MLTC-1 cells, which were consistent with the inhibited 3β-HSD expression in vivo and in vitro. Thus, we proposed that the low H3K14ac of 3β-HSD might program the downregulation of 3β-HSD. Epigenetic modifications are regulated by a variety of modification enzymes (including HDACs) [46,47]. Hence, we further detected the effects of PEE on the expression of several HDACs in fetal testes, in which HDAC2 was highly expressed. We also found that corticosterone increased HDAC2 expression in vitro, and the knockdown of HDAC2 could reverse the inhibitory effect of corticosterone on H3K14ac and expression levels of 3β-HSD. Regarding interactions between SF1 and HDAC2, our results showed that the overexpression of SF1 could significantly reverse the effect of corticosterone on HDAC2 expression, while the knockdown of HDAC2 could not influence the effect of corticosterone on SF1 expression. We further confirmed that the inhibitory effect of corticosterone on SF1 expression could be completely reversed by RU486. As shown above, we concluded that corticosterone increased the expression of HDAC2 by reducing SF1 expression and induced low H3K14ac of 3β-HSD through the GR/SF1/HDAC2 pathway, which programmed low expression levels of 3β-HSD.

Conclusions

PEE induced testicular dysplasia from puberty to adulthood in male offspring rats. This study proposed for the first time that testicular dysplasia originated from the intrauterine period, and its occurrence is related to the low-functional programming of 3β-HSD induced by corticosterone, rather than by the direct effect of ethanol. Furthermore, the molecular mechanism of programming alteration was that fetal overexposure to corticosterone caused by PEE reduced the expression level of SF1 by activating GR, which further increased HDAC2 expression and then reduced the H3K14ac level of 3β-HSD (Figure 9). This study provides experimental evidence and new academic perspectives to illuminate the theory of DOHaD.

Figure 9.

Intrauterine programming mechanism of testicular dysplasia induced by prenatal ethanol exposure (4 g/kg.d).

GR, glucocorticoid receptor; HDAC2, histone deacetylase 2; SF1, steroidogenic factor 1; 3β-HSD, 3β-hydroxysteroid dehydrogenase; H3K14ac, histone 3 acetylated lysine 14.

Funding Statement

This work was supported by grants from the National Key Research and Development Program of China (2017YFC1001300), the National Natural Science Foundation of China (Nos. 81430089, 81673524), and Hubei Province Health and Family Planning Scientific Research Project (No. WJ2017C0003).

Disclosure statement

No potential conflict of interest was reported by the authors.