Toxicology 321 (2014) 53–61

Contents lists available at ScienceDirect

Toxicology journal homepage: www.elsevier.com/locate/toxicol

Prenatal caffeine ingestion induces aberrant DNA methylation and histone acetylation of steroidogenic factor 1 and inhibits fetal adrenal steroidogenesis Jie Ping a,b,1 , Jian-fei Wang a,1 , Lian Liu a,c , You-e Yan a , Fang Liu a , You-ying Lei a , Hui Wang a,b,∗ a

Department of Pharmacology, Basic Medical School of Wuhan University, Wuhan 430071, China Research Center of Food and Drug Evaluation, Wuhan University, Wuhan 430071, China c Department of Pharmacology, Medical School of Yangtze University, Jingzhou 434000, China b

a r t i c l e

i n f o

Article history: Received 2 February 2014 Received in revised form 29 March 2014 Accepted 30 March 2014 Available online 6 April 2014 Keywords: Intrauterine growth retardation (IUGR) Caffeine Steroidogenic factor-1 (SF-1) Fetal adrenal DNA methylation Histone acetylation

a b s t r a c t Prenatal caffeine ingestion is one of the risk factors for intrauterine growth retardation (IUGR). Adrenal plays a pivotal role, mainly through steroidogenesis, in the regulation of intrauterine homeostasis and in fetal development and maturation. We have shown that prenatal caffeine ingestion can inhibit fetal adrenal corticosterone production, but the underlying mechanism is unknown. This study investigated the effects of prenatal caffeine ingestion on corticosterone and its associated synthesized enzymes (steroidogenic acute regulatory protein, StAR; 3␤-hydroxysteroid dehydrogenase, 3␤-HSD; cytochrome P450 cholesterol side chain cleavage, P450scc; P450c21; and P450c11) in the fetal adrenal in rats and further explored the underlying mechanism by analyzing the epigenetic modification and expression of the key transcription factor steroidogenic factor-1 (SF-1). The pregnant rats were intragastrically treated with 120 mg/kg.d caffeine from gestational day 11–20. The results showed that the IUGR rate was 51.2% after caffeine treatment. The contents of corticosterone and the mRNA levels of StAR, P450scc, P450c21, and P450c11 were decreased significantly in the fetal adrenal. Furthermore, caffeine reduced both the protein and the mRNA expression of SF-1 in the fetal adrenal. The epigenetic analysis showed that caffeine treatment can significantly enhance the mRNA expression of DNA methyltransferase (Dnmt) 1, Dnmt3a, histone deacetylases (Hdac) 1, and Hdac2. The detection of DNA methylation by bisulfite-sequencing PCR uncovered a notably increased total methylation rate in the SF-1 promoter. The ChIP assay showed decreased acetylation levels of H3K9 and H3K14 in the SF-1 promoter. In conclusion, prenatal caffeine ingestion is able to induce aberrant DNA methylation and histone acetylation of the SF-1 promoter in the rat fetal adrenal. These effects may contribute to the inhibition of the expression of SF-1 and its associated steroidogenic enzymes and the production of corticosterone during fetal development. © 2014 Elsevier Ireland Ltd. All rights reserved.

1. Introduction Caffeine is a xanthine alkaloid that is widely consumed in the form of coffee, tea, soft beverages, food, and some analgesic

Abbreviations: 3␤-HSD, 3␤-hydroxysteroid dehydrogenase; IUGR, intrauterine growth restriction; StAR, steroidogenic acute regulatory protein; P450scc, cytochrome P450 cholesterol side chain cleavage; SF-1, steroidogenic factor-1; Dnmt, DNA methyltransferase; Hdac, histone deacetylases; HPA, hypothalamic–pituitary–adrenal; BSP, bisulfite sequencing PCR. ∗ Corresponding author at: 185, Donghu Road, Wuhan 430071, China. Tel.: +86 27 6875 8665; fax: +86 27 8733 1670. E-mail address: [email protected] (H. Wang). 1 Ping and Wang contributed equally to this work. http://dx.doi.org/10.1016/j.tox.2014.03.011 0300-483X/© 2014 Elsevier Ireland Ltd. All rights reserved.

drugs. For example, 52.1% of all individuals in the UK over 10 years of age consume coffee (Barone and Roberts, 1996). It has also become increasingly common for pregnant women to intake caffeine-containing foods and/or beverages. Although caffeine induces several well-documented positive effects, such as increased alertness and cognition, caffeine can induce several potentially negative consequences. Both clinical investigations and animal experiments have demonstrated that caffeine ingestion during pregnancy results in reproductive and embryo toxicities (Brent et al., 2011; Huang et al., 2012; Momoi et al., 2008). Maternal caffeine ingestion is one of the risk factors attributable to intrauterine growth retardation (IUGR), which is defined as a birth weight and/or length below the 10th percentile for their gestational age and an abdominal circumference that is less than the 2.5th

54

J. Ping et al. / Toxicology 321 (2014) 53–61

percentile with pathologic restriction of fetal growth (Valsamakis et al., 2006). Increasing epidemiological evidence supports the hypothesis that adverse events in fetal life, such as IUGR, permanently alter the structure and physiology of the adult offspring; this phenomenon is denoted “fetal programming” (Guilloteau et al., 2009; Habib et al., 2011). We have previously reported that prenatal caffeine ingestion is associated with increased risk of IUGR and susceptibility to adult diseases, particularly metabolic syndrome (Huang et al., 2012; Xu et al., 2012a). Fetal programming of key endocrine systems, particularly the hypothalamic–pituitary–adrenal (HPA) axis, has been proposed as a potential intermediary that links IUGR to adult metabolic dysfunction (Kanaka-Gantenbein, 2010; Xita and Tsatsoulis, 2010). Adrenal is the terminal effector organ of the HPA axis and plays a pivotal role, mainly through steroidogenesis, in the regulation of intrauterine homeostasis and in fetal development and maturation (Harris and Seckl, 2011; Xu et al., 2012b). The long-term consequences of low birth weight on the secretion of adrenal glucocorticoids (cortisol in humans and primates and corticosterone in rodents) contribute to increased risks for the development of metabolic syndrome later in life, which suggests a central role for adrenocortical steroidogenesis in intrauterine fetal programming (Marciniak et al., 2011; Ong, 2005). Similar to the adult adrenal, the fetal adrenal expresses a series of enzymes involved in the biosynthesis of active steroid hormones from cholesterol. In the rat fetus, steroidogenic acute regulatory protein (StAR) mediates the translocation of cholesterol from the outer to the inner mitochondrial membrane, which is the initial and rate-limiting step in adrenocortical steroid biosynthesis, whereas the cytochrome P450 cholesterol side chain cleavage enzyme (P450scc) cleaves the cholesterol side chain to convert cholesterol to pregnenolone, which is the precursor of steroid hormones. Pregnenolone is then metabolized into corticosterone by the sequential catalysis of 3␤hydroxysteroid dehydrogenase (3␤-HSD), P450c21, and P450c11 (Miller and Auchus, 2011; Ishimoto and Jaffe, 2011). These steroidogenic enzymes are the key determinants of steroid biosynthesis and are vulnerable to be affected by environmental factors (Kraugerud et al., 2010; Lavoie and King, 2009; Sanderson, 2006). Our previous studies have shown that the fetal adrenal may be an important xenobiotic-metabolizing organ in fetal development and may play a potential role in xenobiotic-induced fetal development toxicity (Wang et al., 2008). We recently showed that prenatal caffeine ingestion can inhibit adrenal StAR and P450scc expression and corticosterone production in fetal rats (Xu et al., 2012b). However, how caffeine alters the steroidogenic enzymes in the fetal adrenal is unknown. Steroidogenic factor-1 (SF-1, also known as NR5A1 or AD4BP) is a nuclear receptor that is expressed in all steroidogenic tissues. It is a master regulator of steroidogenesis whose expression is critical for normal adrenal development and function (Gardiner et al., 2012; Hatano et al., 1996). In mice, a homozygous null mutation in SF-1 results in the loss of the adrenal glands, the initial primordium undergoing apoptosis, and the pups die shortly after birth due to adrenal insufficiency (Gardiner et al., 2012; Luo et al., 1994). SF-1 has been shown to increase the expression of the steroidogenic machinery by binding to its response element site found in the promoter regions of the genes encoding StAR, CYPs, and 3␤HSD (Morohashi et al., 1992; Parker et al., 2002). Epigenetics can be described as a stable alteration in gene expression potential that takes place during development and cell proliferation without any change in the gene sequence (Klein, 2005). Aberrant epigenetic modifications occur in human adrenocortical tumorigenesis and are often accompanied by abnormal hormone production (Liu et al., 2004; Martinez-Arguelles and Papadopoulos, 2010). Kwon et al. reported that caffeine treatment can affect the DNA methylation status and the in vitro development of porcine nuclear transfer

embryos in relation to nuclear reprogramming (Kwon et al., 2008). Our recent study showed that caffeine can decrease insulin-like growth factor-1 expression by decreasing histone methylation in the fetal liver in vivo (Tan et al., 2012) and can enhance 11␤hydroxysteroid dehydrogenase 2 expression by increasing DNA methylation in the fetal hippocampus in vitro (Xu et al., 2012b). These data suggest that caffeine can change the epigenetic modifications of genes during fetal development. In this study, we investigated the effects of caffeine on corticosterone and its associated synthesized enzymes in the fetal adrenal in rats and further explored the underlying mechanisms by observing the expression of and epigenetic modification changes in SF-1. The present study will provide evidence for elucidating the fetal adrenal development toxicity of caffeine and for exploring the fetal origin of adult HPA axis dysfunction and the metabolic syndrome susceptibility of IUGR offspring induced by caffeine. 2. Materials and methods 2.1. Chemicals Caffeine was purchased from Sigma–Aldrich Co. (St Louis, MO, USA). Ex TaqTM DNA Polymerase and TRIZOL were purchased from Invitrogen Co. (Carlsbad, CA, USA). The pGEM-T Easy vector was purchased from Promega Co. (Madison, WI, USA). Polyclonal rat SF-1 antibody was purchased from Santa Cruz Biotechnology Co. (Santa Cruz, CA, USA). Goat anti-rabbit IgG was obtained from Pierce Biotechnology Inc. (Rockford, IL, USA). Protein G beads were purchased from GE Healthcare (Little Chalfont, United Kingdom). Histone H3 (acetyl-K9) antibody and histone H3 (acetyl-K14) antibody were purchased from Abcam (Cambridge, MA, USA). All of the primers were synthesized by Sangon Biotech Co., Ltd. (Shanghai, China). Isoflurane was purchased from Baxter Healthcare Co. (Deerfield, IL, USA). The rat CORT ELISA kit was obtained from AssayPro Co. (MO, USA). Reverse transcription and real-time reverse-transcription PCR (RT-PCR) kits were purchased from Takara Biotechnology Co., Ltd. (Dalian, China). The total protein detection kit was obtained from Bio-Rad Laboratories, Inc. (Hercules, CA, USA). RNA-Solv Reagent and HiBindTM PCR DNA extraction kits were obtained from Omega Bio-Tek Inc. (Norcross, GA, USA). Genomic DNA isolation kits (DNeasy Blood and Tissue Kit) were obtained from Qiagen Co., Ltd. (Duesseldorf, Germany). The EZ DNA methylation kit was obtained from Zymo Research Co. (Orange, CA, USA). All of the other chemicals and reagents were of analytical grade. 2.2. Animals and treatment Specific pathogen-free Wistar rats weighing 180–220 g (female) and 260–300 g (male) were obtained from the Experimental Center of the Hubei Medical Scientific Academy (No. 2006-0005, Hubei, China). The animal work described in this study was performed in the Center for Animal Experimentation of Wuhan University (Wuhan, Hubei, PR China), which has been accredited by the Association for the Assessment and Accreditation of Laboratory Animal Care International (AAALAC International). All of the experimental procedures involving animals were approved by the Chinese Animal Welfare Committee and performed in accordance with the committee’s Guidelines for the Care and Use of Laboratory Animals. The animals were allowed to acclimate for at least one week before being subjected to the experimental conditions. The rats were treated as described before (Xu et al., 2012b). Briefly, for mating, two female rats were placed together with one male rat overnight. The day in which the evidence of mating was observed (vaginal plug or vaginal smear with sperm cells) was designated gestational day (GD) 0. The animals were housed under standard conditions and allowed free access to standard chow and water. From GD11 to GD20, the pregnant rats were intragastrically treated with 120 mg/kg caffeine once a day, and the control rats were sham-treated with the same volume of the vehicle (physiological saline solution). On GD20, to minimize the effects of serum collection through the sacrificing of the animals on the corticosterone levels, inhaled isoflurane was applied to maintain a silent status. The animals were rapidly sacrificed by a blow to the back of the neck after their righting reflexes disappeared. Each feto-placental unit was removed quickly from the uterus, and the fetuses were weighed after being dried on filter papers. The number of pregnant rats in each group was set to 8 (the litter size of each pregnant rat was maintained at 8–14 at birth). IUGR was diagnosed when the body weight of an animal was two standard deviations lower than the mean body weight of the control group (Engelbregt et al., 2001). The IUGR rate was obtained by dividing the number of IUGR rats in each litter by the total pups of the litter. The pups were randomly chosen from each litter for subsequent analysis. Fetal adrenals were collected and randomly selected for fixing in phosphate-buffered 4% paraformaldehyde solution for 24 h before being dehydrated in alcohol and embedded in paraffin. The remaining adrenals of the littermates were pooled together, immediately frozen in liquid nitrogen, and stored at −80 ◦ C for PCR, DNA methylation, and CHIP assays.

J. Ping et al. / Toxicology 321 (2014) 53–61 2.3. Immunohistochemistry

55

The immunohistochemical stainings for SF-1 in the paraformaldehyde-fixed rat adrenals were assessed through the routine immunohistochemistry SP method. Sections approximately 4 ␮m thick were stained with an antibody specific to SF1 (diluted 1:400). At least five random fields from each section were examined at a magnification of 400× and analyzed using the Medical Color Image Analysis System (HPIAS-2000, Guangzhou, China).

purified in parallel with the immunoprecipitated samples as a control, and ChIP reactions were also performed in the absence of antibody to detect the occurrence of any non-specific binding. After centrifugation, DNA was recovered by DNA binding buffer, and quantitative real-time PCR was then performed with primers for sequences located in the promoter of SF-1. The primers were designed using the Vector NTI software, and their sequences are the following: sense, 5 -TCCAGGTTTTCCGTTTGTG-3 ; antisense, 5 -GAAAGACCAGTG ATGATGGG3 (Ta = 57 ◦ C).

2.4. Corticosterone content

2.8. Statistical analysis

The corticosterone synthesized in the adrenal was determined using a commercial ELISA assay kit. Briefly, the adrenals were homogenized with 20% ethanol in PBS and then centrifuged at 1200 × g and 4 ◦ C for 5 min (Wotus et al., 1998). The supernatants were collected for corticosterone analysis, and the pellets were resuspended in 1 N NaOH for measurement of the protein content. The cross-reactivity for the corticosterone ELISA, as defined by the manufacturer, was 2% for progesterone and 2% for aldosterone. The corticosterone content was expressed relative to the adrenal protein measured using the total protein detection kit (Bio-Rad Laboratories, Hercules, CA, USA).

The Statistical Packages for Social Sciences (SPSS 13) was used for the data analysis. The comparison of the means between the control and the caffeine group was performed using Student’s t-test. The mean weights for each litter were calculated and statistically analyzed. The enumeration data, such as the IUGR rates, the IUGR rates of each litter were first calculated; subsequently, they were arcsine squareroot transformed to make the data follow a normal distribution (Tan et al., 2012; Vidmar et al., 2011) before being analyzed through t-test. The methylation frequencies of the two groups were compared statistically by Pearson’s 2 test. The level of statistical significance was set to P < 0.05 for all cases.

2.5. RNA extraction and real-time RT-PCR

3. Results

To determine the mRNA levels of StAR, P450scc, P450c21, P450c11, 3␤-HSD, SF-1, DNA methyltransferase 1 (Dnmt1), Dnmt3a, Dnmt3b, histone deacetylases 1 (Hdac1), and Hdac2, a quantitative real-time PCR assay was performed using a SYBR Green PCR Core Reagents kit. The total RNA was extracted from the adrenals using the TRIZOL reagent following the manufacturer’s protocol. The tissues of each littermate were pooled for homogenization. The concentration and purity of the RNA were determined using a spectrophotometer (NanoDrop 2000), and the RNA concentration was adjusted to 1 ␮g/␮l. Single-strand cDNA was prepared from 1 ␮g of total RNA according to the protocol of the kit and was stored at −20 ◦ C until use. All of the cDNA sequences were obtained from the NCBI Entrez nucleotide databases, and the primers were designed using Primer Premier 5.0 (PREMIER Biosoft International, CA). The sequences of each of the designed primers were queried using the NCBI BLAST database for homology comparison and are listed in Table 1. A relative standard curve was constructed for the target genes and the housekeeping gene (GAPDH) using their corresponding RT-PCR products isolated by the DNA extraction kit with different concentrations ranging from 10 to 10000 pg per reaction. The PCR assays were performed in 36-well optical reaction plates using the RG-3000 Rotor-Gene 4 channel Multiplexing System (ABI Stepone, USA) in a total volume of 20 ␮l containing 1 ␮l of cDNA template, 0.4 ␮l of 10 ␮M each primer, 10 ␮l of 2× Premix Ex Taq, 0.4 ␮l ROX, and 7.8 ␮l of DEPC-H2 O. To precisely quantify the transcripts of the genes, the GAPDH mRNA level was measured as the quantitative control, and each sample was normalized to the GAPDH mRNA content. The PCR cycling conditions were the following: 30 s at 95 ◦ C for pre-denaturation, 5 s at 95 ◦ C for denaturation, and the annealing conditions for each gene are listed in Table 1. The specificity of the produced amplification products was confirmed through agarose electrophoresis examination. 2.6. Genomic DNA sodium bisulfite modification and bisulfite sequencing PCR (BSP)

3.1. Occurrence of IUGR The fetal body weight was an important index for the diagnosis of IUGR (Engelbregt et al., 2001). As shown in Fig. 1A, after the treatment of pregnant rats with 120 mg/kg caffeine, the offspring had significantly lower average body weights at GD20 than the control group (P < 0.01). The IUGR rate increased to 51.2% in the caffeine group (Fig. 1B). 3.2. Corticosterone content and steroidogenic enzyme expression in fetal adrenal As shown in Fig. 2A, the levels of corticosterone in the fetal adrenal were decreased significantly to 56.2% compared with the control (P < 0.05). To determine whether steroidogenic enzymes are involved in the caffeine-induced decrease in corticosterone production in the fetal adrenal, the mRNA expression of corticosterone-synthesized enzymes was detected. Fig. 2B–F shows that the mRNA expression of StAR, P450scc, P450c21, and P450c11 decreased significantly after prenatal caffeine ingestion (P < 0.05). The expression of 3␤-HSD also showed a repressed tendency (P > 0.05). 3.3. mRNA and protein expression of SF-1 in fetal adrenal

The methylation status of the SF-1 promoter was quantitated using the BSP method. Genomic DNA samples were extracted from the adrenals using the DNeasy Blood and Tissue kit and then subjected to bisulfite modification using the EZ DNA methylation kit according to the manufacturer’s instructions. The bisulfitetreated genomic DNA was amplified by Ex TaqTM DNA Polymerase using primers that cover the CpG-rich region of the proximal SF-1 promoter (−280 bp to 60 bp). These primers were designed using Methyl Primer Express 1.0, and their sequences are the following: sense, 5 -ATTTGATTTTTTTAGAATCGGGGTTTTGT-3 ; antisense, 5 CACCTTATCACCAC ACACTAAACACAACT-3 . The thermocycler parameters included an initial 5-min denaturation step at 95 ◦ C, followed by 30 cycles of denaturation at 95 ◦ C for 30 s, annealing at 58 ◦ C for 30 s, and extension at 72 ◦ C for 30 s. A final 10-min extension was performed before the reaction was cooled to 4 ◦ C. All of the amplification steps were completed using a Thermal Cycler (Applied Biosystems, USA). The PCR products were electrophoretically separated on a 1.5% agarose gel and visualized with ethidium bromide to determine the success of the amplification and the presence of an adequate signal. The bisulfite PCR products were purified with a PCR purification kit and cloned into the pGEM-T easy vector for sequencing. Ten clones from each product were sequenced. The percentage of methylation was calculated from the number of methylated CpG divided by the total CpG loci. 2.7. Chromatin immunoprecipitation (ChIP) assay Seventy-five milligrams of adrenal tissue was fixed by immersion in 1% formaldehyde for 10 min. The extracted DNA/protein complex was then sheared by sonication to a size range of 200–600 bp. The complex was then immunoprecipitated by challenging with Protein G beads and 8 ␮g of H3K9 and H3K14 antibodies overnight. A 100× diluted input DNA (Input) obtained from 1 ml of extract was

SF-1 is a key nuclear receptor that controls the expression of steroidogenic enzymes. We determined the protein and mRNA expression of SF-1 by immunohistochemistry and real-time PCR, respectively. The immunohistochemical analysis with a specific anti-SF-1 antibody showed positive staining in the cytoplasm and nucleus (Fig. 3A–D). However, the signal in the nucleus was stronger than that in the cytoplasm. Compared with the control (Fig. 3A and C), there were considerably fewer SF-1-positive cells in the fetal adrenal of the caffeine group (Fig. 3B and D). The results of the semiquantitative assay also showed a statistically significant difference between the two groups (P < 0.01, Fig. 3E). As demonstrated in Fig. 3F, the treatment of pregnant rats with 120 mg/kg caffeine significantly inhibited the SF-1 mRNA levels in the fetal adrenal (P < 0.01). 3.4. Expression of Dnmts and methylation rate of the SF-1 promoter CpG methylation of the proximal promoter is one of the wellknown epigenetic mechanisms that regulate gene expression. First, we analyzed the expression levels of three Dnmts (Fig. 4A).

56

J. Ping et al. / Toxicology 321 (2014) 53–61

Fig. 1. Effects of prenatal caffeine (120 mg/kg) ingestion on the fetal body weight and intrauterine growth restriction (IUGR) rate of fetal rats on gestational day (GD) 20. From GD11 to GD20, the pregnant rats were intragastrically treated with 120 mg/kg caffeine once per day, and the control rats were sham-treated with vehicle. On GD20, the pregnant rats were anesthetized with isoflurane and sacrificed. IUGR was diagnosed when the body weight of each individual animal of the treated group was two standard deviations less than the mean body weight of the control group. The IUGR rates were calculated by dividing the number of IUGR rats in each litter by the total pups of the litter. (A) Fetal weights; (B) IUGR rates. As enumeration data, the IUGR rates were arcsine square-root transformed to make the data follow a normal distribution before being subjected to t-test evaluations. The data are presented as the means ± SEM; n = 8. ** P < 0.01 vs. the control.

Fig. 2. Effects of prenatal caffeine ingestion on mRNA expression of steroidogenic enzymes and corticosterone content in the fetal adrenal in rat. (A–E) The mRNA expression of acute regulatory protein (StAR), cytochrome P450 cholesterol side chain cleavage (P450scc), 3␤-hydroxysteroid dehydrogenase (3␤-HSD), P450c21, and P450c11 was detected by real-time PCR. GAPDH was used as an internal control. Four pairs of adrenals from each littermate were pooled for homogenization into one sample; n = 5. (F) The pregnant rats were intragastrically treated with 120 mg/kg caffeine from gestational day 11–20, and the levels of corticosterone in the fetal adrenals were detected by ELISA. Five pairs of adrenals from each littermate were pooled for homogenization into one sample; n = 6. The results are expressed as the means ± SD; * P < 0.05 vs. the control.

J. Ping et al. / Toxicology 321 (2014) 53–61

57

Table 1 Oligonucleotide primers and PCR conditions used in quantitative real-time PCR. Genes

Forward primer

Reverse primer

Product (bp)

Annealing

StAR P450scc P450c21 P450c11 3␤-HSD SF-1 Dnmt1 Dnmt3a Dnmt3b Hdac1 Hdac2 GAPDH

GGGAGATGCCTGAGCAAAGC GCTGCCTGGGATGTGATTTTC AGGAGCTGAAGAGGCACAAG CCCCTTTGTGGATGTGGTAG TCTACTGCAGCACAGTTGAC CCAGTACGGCAAGGAAGA GCTAAGGACGATGATGAGACGC CAGCGTCACACAGAAGCATATCC AGACCTGGCTGCTTAGAGT AATGCTAATGTTGGGAGG CAGTTGCCCTTGATTGTGAGATTC GCAAGTTCAACGGCACAG

GCTGGCGAACTCTATCTGGGT GATGTTGGCCTGGATGTTCTTG GAGGTAGCTGCATTCGGTTC CACGCTCTCAGGTTTCAGGT ATACCCTTATTTTTGAGGGC GAGGCTGAAGAGGATGAGGA CTTTTTGGGTGACGGCAACTC GGTCCTCACTTTGCTGAACTTGG GGTCCCTTGAGTTTGTTCG ATTGGAAGGGCTGATGTG GCTATCCGCTTGTCTGATGCTC GCCAGTAGACTCCACGACA

188 156 306 190 271 193 447 433 273 166 295 107

65 ◦ C, 20 s 63 ◦ C, 20 s 63 ◦ C, 30 s 61 ◦ C, 30 s 58 ◦ C, 30 s 63 ◦ C, 30 s 65 ◦ C, 30 s 65 ◦ C, 30 s 58 ◦ C 30 s 55 ◦ C, 30 s 68 ◦ C, 30 s 63 ◦ C, 30 s

StAR, steroidogenic acute regulatory protein; P450scc, cytochrome P450 cholesterol side chain cleavage enzyme; 3␤-HSD: 3␤-hydroxysteroid dehydrogenase; SF-1, steroidogenic factor-1; Dnmt, DNA methyltransferase; Hdac, histone deacetylases.

The results showed that caffeine treatment can significantly enhance both Dnmt1 and Dnmt3a expression (P < 0.05, P < 0.01, respectively). The expression of Dnmt3b was also increased by 38.3% after caffeine treatment. Then, to determine whether CpG

methylation changes are involved in the caffeine-induced decrease in SF-1 expression, the proximal promoter of the rat SF-1 gene was selected for methylation analysis. The methylation status within the CpG-rich region (−280 bp to 60 bp) was detected using BSP

Fig. 3. Effects of prenatal caffeine ingestion on the protein and mRNA expression of steroidogenic factor-1 (SF-1) in fetal rat adrenals. The SF-1 protein expression was measured by immunohistochemistry and semiquantitative analysis. Three fetal adrenals were randomly selected for fixing in phosphate-buffered 4% paraformaldehyde solution for 24 h before being dehydrated in alcohol and embedded in paraffin. (A) Control (200 × ); (B) 120 mg/kg caffeine (200 × ); (C) control (400 × ); (D) 120 mg/kg caffeine (400 × ); (E) semiquantitative measurement of SF-1 protein expression in the fetal adrenal; and (F) SF-1 mRNA expression. Four pairs of adrenals from each littermate were pooled for homogenization into one sample; n = 5. The results are expressed as the means ± SD; ** P < 0.01 vs. the control.

58

J. Ping et al. / Toxicology 321 (2014) 53–61

Fig. 4. Effects of prenatal caffeine ingestion on the mRNA expression of DNA methyltransferases (Dnmts) and the methylation status of the steroidogenic factor-1 (SF-1) promoter in fetal adrenals. (A) The mRNA expression of Dnmt1, Dnmt3a, and Dnmt3b was assayed by real-time PCR. (B) The total methylation rate within the CpG-rich region (nt −280 bp to 60 bp) was detected through the bisulfite-sequencing PCR (BSP) method. (C) Methylation maps of some CpG sites on the SF-1 promoter region. The transcription initiation site was designated +1. Five pregnant rats from each group were analyzed, and the adrenals from each littermate were pooled for homogenization into one sample. The results are expressed as the means ± SD; n = 5. * P < 0.05 and ** P < 0.01 vs. the control.

assays. Fig. 4B shows that caffeine can notably increase the total methylation rate within nt −280 bp to 60 bp of the SF-1 promoter (P < 0.05). Among the 36 CpG sites, nt −162, −139, −108, −104, −91, −53, −50, and −14 also showed higher frequency of single CpG methylation after caffeine treatment (Fig. 4C). 3.5. Histone deacetylation of the SF-1 promoter Similarly to DNA methylation, histone acetylation is another important epigenetic mechanism regulating gene expression profiles. As shown in Fig. 5A, the mRNA expression levels of Hdac1 and Hdac2 are markedly elevated in the fetal adrenal of the caffeine group compared with the control (P < 0.05). We then determined whether the histone acetylation level of the SF-1 promoter was altered in the fetal adrenal of the caffeine group. The acetylations of H3K9 and H3K14 in the SF-1 promoter were assayed through the ChIP method. The comparison of the “Input” and “IgG” revealed that prenatal caffeine ingestion decreased the acetylation level of H3K9 and H3K14 to 26.85% and 27.45%, respectively, compared with the control levels (Fig. 5B). 4. Discussion Caffeine is consumed daily by many pregnant women (Sengpiel et al., 2013). Recently, Sengpiel et al. examined the association between maternal caffeine intake from different sources and the risk for low birth weight in a large epidemiological study. These researchers concluded that caffeine intake is consistently associated with decreased birth weight and increased odds of small for gestational age (SGA) (Sengpiel et al., 2013). In our previous study (Xu et al., 2012b; Huang et al., 2012), several adverse reproductive and developmental parameters were observed after female rats

were treated with caffeine (20, 60, and 180 mg/kg per day) from GD11 to GD20. These parameters include the number of absorbed and stillborn fetuses, the body weights of the live fetuses, the IUGR rates, and the body and tail lengths. Similar results were observed in mice after caffeine treatment. To achieve the typical IUGR model, the middle dose of 120 mg/kg d caffeine was used in the present study. We determined the caffeine concentration in maternal blood and showed it was 254 ± 11 ␮M (49 ± 2 ␮g/ml) after intake of 120 mg/kg d caffeine (Wang et al., 2014), which is 1.6 times higher than the clinical signs of intoxication (∼30 ␮g/ml), but did not reach the dose with fatalities (∼80 ␮g/ml). Therefore, the dose of caffeine applied in this study should be reasonable. As the terminal effector organ of the HPA axis, the adrenal gland secretes glucocorticoid, which plays a crucial role in the maturity of fetal organs. It has been reported that the adrenal is the earliest organ in the development of the embryo HPA axis. In human, the primordium of the adrenal glands can be recognized at 3–4 weeks of gestation. The large eosinophilic cells of the fetal adrenal are well differentiated by 9–12 weeks of gestation and are capable of active steroidogenesis. The fetal adrenal grows rapidly and progressively in mass. The weight of the fetal adrenal during midgestation is 10–15% of the fetal liver weight, although there is a 100-fold difference in the adult organ weights between the livers and adrenals. In newborns, the weight of the fetal adrenal is 20-fold higher than that of the adult (Ishimoto and Jaffe, 2011; Yan, 1983). Thus, adrenal toxicity may have a greater impact on growth and development in fetuses than in adults (Rainey et al., 2004; Wang et al., 2008). Any interference in the fetal adrenal development by xenobiotics may retard the development of the fetal HPA axis function and consequently have a profound effect on intrauterine homeostasis and fetal growth (Kraugerud et al., 2010; Lavoie and King, 2009; Sanderson, 2006). Our findings on the effects of caffeine

J. Ping et al. / Toxicology 321 (2014) 53–61

59

Fig. 5. Effects of prenatal caffeine ingestion on the mRNA expression of histone deacetylases (Hdacs) and the histone acetylation level of the steroidogenic factor-1 (SF-1) promoter in fetal adrenals. (A) The mRNA expression levels of Hdac1 and Hdac2 were assayed by real-time PCR. Four pairs of adrenals from each littermate were pooled for homogenization into one sample; n = 5. The results are expressed as the means ± SD; * P < 0.05 vs. the control. (B) The acetylations of H3K9 and H3K14 in the SF-1 promoter were assayed through the chromatin immunoprecipitation (ChIP) method.

on fetal adrenal steroidogenesis showed a significant decrease in corticosterone production, which may contribute to the occurrence of prenatal caffeine ingestion-induced IUGR. Steroid hormones are synthesized by the catalysis of a series of enzymes. Our study demonstrated that the expressions of all of the enzymes in the pathway from cholesterol to corticosterone, including StAR, P450scc, P450c21, P450c11, and 3␤-HSD, are inhibited in the fetal adrenals of the caffeine-exposed offspring. Of all the transcriptional factors that regulate steroidogenic enzymes expression, SF-1 is most important because the promoter region of all steroidogenic enzymes has the specific SF-1 binding site (Lavoie and King, 2009; Morohashi et al., 1992; Parker et al., 2002). In the human adrenal corticocarcinoma cell line NCI-H295R, a reduction in SF-1 decreases StAR protein expression (Sugawara et al., 2006). In our study, the treatment of pregnant rats with caffeine resulted in a significant decrease in both SF-1 mRNA and protein expression. Thus, we hypothesized that caffeine can inhibit SF-1 expression and thus reduce the expression of the steroidogenic enzymes and further suppress the function of corticosterone synthesis in the fetal adrenal. Epigenetics is an important mechanism regulating the gene expression throughout the process of fertilization and embryo development. DNA methylation is a major epigenetic mechanism that controls developmental gene expression. Coordinated waves of demethylation and de novo methylation establish the genomewide methylation pattern during embryogenesis (Reik et al., 2001). The role of DNA methylation in steroidogenesis has been a topic of considerable interests in the last few years. Several studies have shown that aberrant DNA methylation plays an important

role in the expression of some key steroidogenic genes (MartinezArguelles and Papadopoulos, 2010). Mammalian Dnmts are divided into two groups based on their preferred DNA substrate. Dnmt1 copies the methylation pattern during DNA replication, whereas de novo methylation is conducted by Dnmt3a and Dnmt3b (Leonhardt et al., 1992). The genetic manipulation of the Dnmt genes has demonstrated that appropriate DNA methylation is required for normal mammalian development (Chen and Li, 2006). Our data showed that caffeine treatment can enhance the expressions of these three Dnmts, which indicates that DNA methylation may be involved in the caffeine-induced decrease in SF-1 expression. DNA methylation has been shown to control SF-1 gene expression, as observed by the finding that the methylation of the SF-1 promoter is associated with the lack of SF-1 expression in normal endometrial tissue, whereas promoter hypomethylation is associated with SF-1 expression in endometrial stromal cells (Xue et al., 2007). In this study, we observed that caffeine can notably increase the total methylation rate within the SF-1 promoter, which certifies our hypothesis that DNA methylation participates in the inhibitory effect of prenatal caffeine ingestion on SF-1 expression. In addition to DNA methylation, histone acetylation is also an epigenetic modification that regulates the gene expression profiles in embryogenesis. Histone acetylation aids in chromatin remodeling and is maintained by the activity of histone acetyl transferases (Hats) and Hdacs (Henikoff and Ahmad, 2005; Wade et al., 1997). Evidence suggests that epigenetic modifications can independently regulate cellular mechanisms (Bestor, 1998), or the joint activities of DNA methylation and histone deacetylation could result in a common outcome (Murr, 2010). Jacob et al. reported that the

60

J. Ping et al. / Toxicology 321 (2014) 53–61

acetylation of SF-1 stimulates its transcriptional activity. The inhibition of deacetylation by trichostatin A, a histone deacetylase inhibitor, increases SF-1-mediated transactivation and induces the nuclear export of the SF-1 protein (Jacob et al., 2001). In this study, we found that the expression levels of both Hdac1 and Hdac2 are markedly elevated in the fetal adrenal of the caffeine group, which suggests that histone deacetylation also participate in the caffeine-induced decrease in SF-1 expression. Histone tails, mainly H3 and H4, can be modified by acetylation. The most studied histone modification is the acetylation of H3K9, which is associated with transcriptional activation through inactivation of the positive lysine charge (Wade et al., 1997). We then detected the acetylation status of H3K9 and H3K14 in the SF-1 promoter. The results showed that prenatal caffeine ingestion decreased the acetylation levels of H3K9 and H3K14 to 26.85% and 27.45% compared with the control, respectively. Thus, our study demonstrates that it is possible that the epigenetic effects of caffeine are due to its effect on DNA methylation or its combined action on DNA methylation and histone deacetylation to regulate SF-1 expression. In conclusion, the present study demonstrated that prenatal caffeine ingestion is able to induce aberrant DNA methylation and histone acetylation in the promoter of the key transcriptional factor SF-1, which could inhibit the expression of steroidogenic enzymes and the production of corticosterone in the rat fetal adrenal. We recently reported (Xu et al., 2011, 2012a,b) that prenatal caffeine ingestion induces a HPA axis-associated neuroendocrine metabolic programmed alteration in IUGR fetal rats and that the underlying mechanism for these observations may involve over-exposure to maternal glucocorticoid. The HPA axis of these IUGR offspring exhibits a low basal activity and an enhanced sensitivity to chronic stress. Based on our previous reports and the current study, we propose that maternal prenatal caffeine ingestion at a clinically toxic dose retards the development of the fetal HPA axis function by directly injuring fetal adrenal steroidogenesis and by fetal overexposure to maternal glucocorticoid levels. These effects can conjointly lead to retarded fetal development and high susceptibility to adult metabolic syndrome. Our current data provide insights into an epigenetic mechanism through which caffeine reduces steroidogenesis in the fetal adrenal and leads to the occurrence of IUGR. Thus, this study may provide evidence for the intrauterine origins of HPA axis dysfunction, particularly dysfunction of the adrenal, in response to the poor environment in adult IUGR offspring induced by prenatal caffeine consumption. Funding This work was supported by Grants from the National Natural Science Foundation of China (No. 30830112, 30800931, 81220108026 and 81173138), the Key Grant Project of the Chinese Ministry of Education (No. V200801) and the National Science & Technology Pillar Program of China (No. 2013BAI12B01-3). Conflict of interest statement All authors have no conflicts of interest. Transparency document The Transparency document associated with this article can be found in the online version. References Barone, J.J., Roberts, H.R., 1996. Caffeine consumption. Food Chem. Toxicol. 34, 119–129.

Bestor, T.H., 1998. Gene silencing. Methylation meets acetylation. Nature 393, 311–312. Brent, R.L., Christian, M.S., Diener, R.M., 2011. Evaluation of the reproductive and developmental risks of caffeine. Birth Defects Res. B: Dev. Reprod. Toxicol. 92, 152–187. Chen, T., Li, E., 2006. Establishment and maintenance of DNA methylation patterns in mammals. Curr. Topics Microbiol. Immunol. 301, 179–201. Engelbregt, M.J., van Weissenbruch, M.M., Popp-Snijders, C., Lips, P., Delemarrevan, D.W.H., 2001. Body mass index, body composition, and leptin at onset of puberty in male and female rats after intrauterine growth retardation and after early postnatal food restriction. Pediatr. Res. 50, 474–478. Gardiner, J.R., Shima, Y., Morohashi, K., Swain, A., 2012. SF-1 expression during adrenal development and tumourigenesis. Mol. Cell. Endocrinol. 351, 12–18. Guilloteau, P., Zabielski, R., Hammon, H.M., Metges, C.C., 2009. Adverse effects of nutritional programming during prenatal and early postnatal life, some aspects of regulation and potential prevention and treatments. J. Physiol. Pharmacol. 60, 17–35. Habib, S., Gattineni, J., Twombley, K., Baum, M., 2011. Evidence that prenatal programming of hypertension by dietary protein deprivation is mediated by fetal glucocorticoid exposure. Am. J. Hypertens. 24, 96–101. Harris, A., Seckl, J., 2011. Glucocorticoids, prenatal stress and the programming of disease. Horm. Behav. 59, 279–289. Hatano, O., Takakusu, A., Nomura, M., Morohashi, K., 1996. Identical origin of adrenal cortex and gonad revealed by expression profiles of Ad4BP/SF-1. Genes Cells 1, 663–671. Henikoff, S., Ahmad, K., 2005. Assembly of variant histones into chromatin. Annu. Rev. Cell Dev. Biol. 21, 133–153. Huang, J., Zhou, S., Ping, J., Pan, X., Liang, G., Xu, D., Kou, H., Bao, C., Wang, H., 2012. Role of p53-dependent placental apoptosis in the reproductive and developmental toxicities of caffeine in rodents. Clin. Exp. Pharmacol. Physiol. 39, 357–363. Ishimoto, H., Jaffe, R.B., 2011. Development and function of the human fetal adrenal cortex: a key component in the feto-placental unit. Endocr. Rev. 32, 317–355. Jacob, A.L., Lund, J., Martinez, P., Hedin, L., 2001. Acetylation of steroidogenic factor 1 protein regulates its transcriptional activity and recruits the coactivator GCN5. J. Biol. Chem. 276, 37659–37664. Kanaka-Gantenbein, C., 2010. Fetal origins of adult diabetes. Ann. N. Y. Acad. Sci. 1205, 99–105. Klein, G., 2005. Epigenetics: surveillance team against cancer. Nature 434, 150. Kraugerud, M., Zimmer, K.E., Dahl, E., Berg, V., Olsaker, I., Farstad, W., Ropstad, E., Verhaegen, S., 2010. Three structurally different polychlorinated biphenyl congeners (Pcb 118, 153, and 126) affect hormone production and gene expression in the human H295R in vitro model. J. Toxicol. Environ. Health A 73, 1122–1132. Kwon, D.J., Park, C.K., Yang, B.K., Cheong, H.T., 2008. Control of nuclear remodelling and subsequent in vitro development and methylation status of porcine nuclear transfer embryos. Reproduction 135, 649–656. Lavoie, H.A., King, S.R., 2009. Transcriptional regulation of steroidogenic genes: STARD1, CYP11A1 and HSD3B. Exp. Biol. Med. 234, 880–907. Leonhardt, H., Page, A.W., Weier, H.U., Bestor, T.H., 1992. A targeting sequence directs DNA methyltransferase to sites of DNA replication in mammalian nuclei. Cell 71, 865–873. Liu, J., Li, X.D., Vaheri, A., Voutilainen, R., 2004. DNA methylation affects cell proliferation, cortisol secretion and steroidogenic gene expression in human adrenocortical NCI-H295R cells. J. Mol. Endocrinol. 33, 651–662. Luo, X., Ikeda, Y., Parker, K.L., 1994. A cell-specific nuclear receptor is essential for adrenal and gonadal development and sexual differentiation. Cell 77, 481–490. Marciniak, B., Patro-Malysza, J., Poniedzialek-Czajkowska, E., Kimber-Trojnar, Z., Leszczynska-Gorzelak, B., Oleszczuk, J., 2011. Glucocorticoids in pregnancy. Curr. Pharm. Biotechnol. 12, 750–757. Martinez-Arguelles, D.B., Papadopoulos, V., 2010. Epigenetic regulation of the expression of genes involved in steroid hormone biosynthesis and action. Steroids 75, 467–476. Miller, W.L., Auchus, R.J., 2011. The molecular biology, biochemistry, and physiology of human steroidogenesis and its disorders. Endocr. Rev. 32, 81–151. Momoi, N., Tinney, J.P., Liu, L.J., Elshershari, H., Hoffmann, P.J., Ralphe, J.C., Keller, B.B., Tobita, K., 2008. Modest maternal caffeine exposure affects developing embryonic cardiovascular function and growth. Am. J. Physiol. Heart Circ. Physiol. 294, H2248–H2256. Morohashi, K., Honda, S., Inomata, Y., Handa, H., Omura, T., 1992. A common trans-acting factor, Ad4-binding protein, to the promoters of steroidogenic P-450s. J. Biol. Chem. 267, 17913–17919. Murr, R., 2010. Interplay between different epigenetic modifications and mechanisms. Adv. Genet. 70, 101–141. Ong, K., 2005. Adrenal function of low-birthweight children. Endocr. Dev. 8, 34–53. Parker, K.L., Rice, D.A., Lala, D.S., Ikeda, Y., Luo, X., Wong, M., 2002. Steroidogenic factor 1: an essential mediator of endocrine development. Recent Prog. Horm. Res. 57, 19–36. Rainey, W.E., Rehman, K.S., Carr, B.R., 2004. The human fetal adrenal: making adrenal androgens for placental estrogens. Semin. Reprod. Med. 22, 327–336. Reik, W., Dean, W., Walter, J., 2001. Epigenetic reprogramming in mammalian development. Science 293, 1089–1093. Sanderson, J.T., 2006. The steroid hormone biosynthesis pathway as a target for endocrine-disrupting chemicals. Toxicol. Sci. 94, 3–21. Sengpiel, V., Elind, E., Bacelis, J., Nilsson, S., Grove, J., Myhre, R., Haugen, M., Meltzer, H.M., Alexander, J., Jacobsson, B., Brantsaeter, A.L., 2013. Maternal caffeine intake during pregnancy is associated with birth weight but not with gestational

J. Ping et al. / Toxicology 321 (2014) 53–61 length: results from a large prospective observational cohort study. BMC Med. 11, 42. Sugawara, T., Sakuragi, N., Minakami, H., 2006. CREM confers cAMP responsiveness in human steroidogenic acute regulatory protein expression in NCI-H295R cells rather than SF-1/Ad4BP. J. Endocrinol. 191, 327–337. Tan, Y., Liu, J., Deng, Y., Cao, H., Xu, D., Cu, F.L., Lei, Y.Y., Magdalou, J., Wu, M., Chen, L., Wang, H., 2012. Caffeine-induced fetal rat over-ingestion to maternal glucocorticoid and histone methylation of liver IGF-1 might cause skeletal growth retardation. Toxicol. Lett. 214, 279–287. Valsamakis, G., Kanaka-Gantenbein, C., Malamitsi-Puchner, A., Mastorakos, G., 2006. Causes of intrauterine growth restriction and the postnatal development of the metabolic syndrome. Ann. N. Y. Acad. Sci. 1092, 138–147. Vidmar, T., Freshwater, L., Collins, R., 2011. Understanding statistics in developmental and reproductive toxicology. In: Hood, R.D. (Ed.), Developmental and Reproductive Toxicology: A Practical Approach. Informa Healthcare, New York, pp. 542–562. Wade, P.A., Pruss, D., Wolffe, A.P., 1997. Histone acetylation: chromatin in action. Trends Biochem. Sci. 22, 128–132. Wang, H., Ping, J., Peng, R.X., Yue, J., Xia, X.Y., Li, Q.X., Kong, R., Hong, J.Y., 2008. Changes of multiple biotransformation phase I and phase II enzyme activities in human fetal adrenals during fetal development. Acta Pharmacol. Sin. 29, 231–238. Wang, L.L., Shen, L., Ping, J., Zhang, L., Liu, Z.F., Wu, Y., Liu, Y.S., Huang, H.G., Chen, L.B., Wang, H., 2014. Intrauterine glucose and lipid metabolic programming

61

mediates the fetal origin of high susceptibility to adult fatty liver disease in prenatal caffeine-exposed rat offspring. Toxicol. Lett. 224, 311–318. Wotus, C., Levay-Young, B.K., Rogers, L.M., Gomez-Sanchez, C.E., Engeland, W.C., 1998. Development of adrenal zonation in fetal rats defined by expression of aldosterone synthase and 11beta-hydroxylase. Endocrinology 139, 4397–4403. Xita, N., Tsatsoulis, A., 2010. Fetal origins of the metabolic syndrome. Ann. N. Y. Acad. Sci. 1205, 148–155. Xu, D., Chen, M., Pan, X.L., Xia, L.P., Wang, H., 2011. Dexamethasone induces fetal developmental toxicity through affecting the placental glucocorticoid barrier and depressing fetal adrenal function. Environ. Toxicol. Pharmacol. 32, 356–363. Xu, D., Wu, Y., Liu, F., Liu, Y.S., Shen, L., Lei, Y.Y., Liu, J., Ping, J., Qin, J., Zhang, C., Chen, L.B., Magdalou, J., Wang, H., 2012a. A hypothalamic–pituitary–adrenal axis-associated neuroendocrine metabolic programmed alteration in offspring rats of IUGR induced by prenatal caffeine ingestion. Toxicol. Appl. Pharmacol. 264, 395–403. Xu, D., Zhang, B.J., Liang, G., Ping, J., Kou, H., Li, X.J., Xiong, J., Chen, L.B., Magdalou, J., Wang, H., 2012b. Caffeine-induced activated glucocorticoid metabolism in hippocampus causes hypothalamic–pituitary–adrenal axis inhibition in fetal rats. PLoS ONE 7, e44497. Xue, Q., Lin, Z., Yin, P., Milad, M.P., Cheng, Y.H., Confino, E., 2007. Transcriptional activation of steroidogenic factor-1 by hypomethylation of the 5 CpG island in endometriosis. J. Clin. Endocrinol. Metab. 92, 3261–3267. Yan, G.L., 1983. Fetal Endocrinology. People’s Medical Publishing House, Peking, pp. 24 (in Chinese).