Identification of gene targets of developmental neurotoxicity focusing on DNA hypermethylation involved in irreversible disruption of hippocampal neurogenesis in rats.

We have previously found that maternal exposure to 6-propyl-2-thiouracil (PTU), valproic acid (VPA), or glycidol (GLY) has a sustained or late effect on hippocampal neurogenesis at the adult stage in rat offspring. Herein, we searched for genes with hypermethylated promoter region and downregulated transcript level to reveal irreversible markers of developmental neurotoxicity. The hippocampal dentate gyrus of male rat offspring exposed maternally to PTU, VPA, or GLY was subjected to Methyl-Seq and RNA-Seq analyses on postnatal day (PND) 21. Among the genes identified, 170 were selected for further validation analysis of gene expression on PND 21 and PND 77 by real-time reverse transcription-PCR. PTU and GLY downregulated many genes on PND 21, reflecting diverse effects on neurogenesis. Furthermore, genes showing sustained downregulation were found after PTU or VPA exposure, reflecting a sustained or late effect on neurogenesis by these compounds. In contrast, such genes were not observed with GLY, probably because of the reversible nature of the effects. Among the genes showing sustained downregulation, Creb, Arc, and Hes5 were concurrently downregulated by PTU, suggesting an association with neuronal mismigration, suppressed synaptic plasticity, and reduction in neural stem and progenitor cells. Epha7 and Pvalb were also concurrently downregulated by PTU, suggesting an association with the reduction in late-stage progenitor cells. VPA induced sustained downregulation of Vgf and Dpysl4, which may be related to the aberrations in synaptic plasticity. The genes showing sustained downregulation may be irreversible markers of developmental neurotoxicity.

INTRODUCTION

The hippocampal dentate gyrus of the mammalian brain is crucial for higher brain functions, such as learning and memory, which are closely related to adult hippocampal neurogenesis and activity‐dependent synaptic plasticity (Vivar et al., 2013). Adult neurogenesis starts to produce new neurons from the subgranular zone (SGZ) of the dentate gyrus during postnatal life (Zhao et al., 2008). In the hilus of the dentate gyrus, the γ‐aminobutyric acid (GABA)ergic interneurons innervate granule cell lineage populations to control SGZ neurogenesis (Masiulis et al., 2011). In addition to inputs by GABAergic interneuron subpopulations, various neuronal populations outside the SGZ create synaptic connections with neurons in the dentate gyrus, such as cholinergic neurons and glutamatergic neurons that are important for maintaining proper proliferation and differentiation of granule cell lineages (Fonnum et al., 1979; Cameron et al., 1995; Zhu et al., 2008).

All the cell populations and their inherent processes involved in the adult neurogenesis may be sensitive targets of developmental neurotoxicity (DNT). Especially, self‐renewal of stem cells, proliferation and migration of progenitor cells, neuritogenesis, synaptogenesis, and myelinogenesis may be the developmental processes vulnerable to chemical toxicity. Based on this hypothesis, previously, we have examined the effect of maternal exposure to a number of neurotoxicants on hippocampal neurogenesis and found that all the known neurotoxicants targeted hippocampal neurogenesis under an exposure regimen in accordance with the guidelines for chemicals testing developed by the Organisation for Economic Co‐operation and Development (OECD; Test No. 426: Developmental Neurotoxicity Study; OECD, 2007) in rodent animals (Shibutani, 2015). Importantly, some neurotoxicants, such as manganese, aluminum, 6‐propyl‐2‐thiouracil (PTU), glycidol (GLY), and valproic acid (VPA), had a sustained or late effect on neurogenesis in terms of granule cell lineage subpopulations, interneuron subpopulations, and/or other regulatory systems (Wang et al., 2012; Akane, Shiraki et al., 2013; Shiraki et al., 2016; Watanabe et al., 2017; Inohana et al., 2018). If the effect on neurogenesis is permanent in nature, the concern arises regarding learning and memory formation later in life.

Recent studies have indicated that various epigenetic mechanisms, such as alterations in DNA methylation, are involved in the regulation of different aspects of adult neurogenesis (Sun et al., 2011). The idea that DNA methylation is a long‐lasting cellular memory necessary for keeping a cellular phenotype has been challenged by discoveries of its dynamic nature (Covic et al., 2010). This relationship is typically seen in CpG sequences at the promoter regions, where DNA methylation can directly interfere with transcription factor binding to DNA or indirectly suppress transcription through methylated DNA‐binding proteins that recruit histone deacetylases (HDAC), leading to chromatin condensation and subsequent gene silencing (Jones et al., 1998). Although the effects of epigenetic alterations on neurogenesis have remained unexplored, environmentally induced disruption of DNA methylation warrants further study (Ceccatelli et al., 2013), given the clear importance of DNA methylation to neuronal development. In fact, exposure to stress (Mueller & Bale, 2008), neurotoxicants (Kundakovic et al., 2013), and maternal neglect (Weaver et al., 2004) in early life have been shown to disrupt epigenetic programming involving DNA methylation in the brain, with lasting consequences for brain gene expression and behavior.

We have recently examined genes showing promoter region hypermethylation in the hippocampal dentate gyrus of mice exposed maternally to manganese, 3,3′‐iminodipropionitrile, or hexachlorophene by CpG island (CGI) microarray analysis or methyl‐capture sequencing analysis (Wang et al., 2013; Watanabe et al., 2018; Tanaka et al., 2019). Among the genes identified, a number of genes showed sustained hypermethylation and transcript downregulation after exposure to these compounds. Interestingly, genes expressed in neural stem cells (NSCs) or neural progenitor cells showed sustained downregulation, whereas those expressed in interneurons mostly showed transient downregulation.

The present study was performed to reveal irreversible DNT markers that can be applied to DNT studies. For this purpose, we focused on hypermethylated genes during the disruption of hippocampal neurogenesis. Because rats are recommended to use in the aforementioned guidelines for testing DNT, we selected PTU, GLY, and VPA as model compounds that have shown irreversible or late effects on hippocampal neurogenesis in rats (Table 1). Hippocampal dentate gyrus was subjected to next generation sequencing analyses in terms of DNA methylation and gene expression at the end of maternal exposure. Selected genes were further examined for transcript expression level changes by real‐time reverse transcription‐PCR (RT‐PCR) analysis to confirm irreversibility of transcript downregulation.

TABLE 1

Immunohistochemical changes of cellular populations related to hippocampal neurogenesis after maternal exposure to PTU, VPA, or GLY in rats

Compound	Target	PND 21	PND 77	References
PTU	Granule cell lineages			Shiraki et al., 2016
		Type‐1 NSCs↓
		Type‐2a progenitor cells↓	Type‐2a progenitor cells↓
		Type‐3 progenitor cells↓
		SGZ cell apoptosis↑
	GABAergic interneurons
		RELN+ interneurons↑
		PVALB+ interneurons↓	PVALB+ interneurons↓ b
		CALB2+ interneurons↑	CALB2+ interneurons↑
		SST+ interneurons↑	SST+ interneurons↑
	Synaptic plasticity
		EPHA4+ GCL cells↓	EPHA4+ GCL cells↑
		ARC+ GCL cells↓	ARC+ GCL cells↑
VPA	Granule cell lineages			Watanabe et al., 2017
			NeuN+ GCL cells↑
			SGZ cell proliferation↑
	GABAergic interneurons
		RELN+ interneurons↓
		PVALB+ interneurons↓
		GAD67+ interneurons↓	GAD67+ interneurons↓
	Synaptic plasticity
			ARC+ GCL cells↑
			COX2+ GCL cells↑
GLY	Granule cell lineages			Akane, Shiraki et al., 2013
		Immature granule cells↓		Akane et al., 2014
	GABAergic interneurons
		NeuN+ interneurons↑	NeuN+ interneurons↑
		RELN+ immature interneurons↑	RELN+ immature interneurons↑
		CALB2+ interneurons↑	CALB2+ interneurons↑
	Synaptic plasticity
			ARC+ GCL cells↑
			FOS+ GCL cells↑

Abbreviations: ARC, activity‐regulated cytoskeleton‐associated protein; CALB2, calbindin‐d‐29 k (calretinin); COX2, cyclooxygenase 2; EPHA4, EPH receptor A4; FOS, FBJ osteosarcoma oncogene; GABA, γ‐aminobutyric acid; GAD67, glutamate decarboxylase 67; GCL, granule cell layer; GLY, glycidol; NeuN, neuronal nuclei; NSC, neural stem cell; PND, postnatal day; PTU, 6‐propyl‐2‐thiouracil; PVALB, parvalbumin; RELN, reelin; SGZ, subgranular zone; SST, somatostatin; VPA, valproic acid.

Changes observed at least at the high‐dose group are listed in each study.

Statistically nonsignificant change as compared with the untreated controls.

MATERIALS AND METHODS

Chemicals and animals

PTU (CAS No. 51‐52‐5, purity ≥ 99%), VPA (sodium salt; CAS No. 1069‐66‐5, purity ≥ 98%), and GLY (CAS No. 556‐52‐5, purity ≥ 96%), were purchased from Sigma‐Aldrich Japan Co., Inc. (Tokyo, Japan).

Forty‐nine mated female Slc : SD rats were purchased from Japan SLC, Inc. (Hamamatsu, Japan) at gestational day (GD) 1 (appearance of the vaginal plug was designated as GD 0). Rats were individually housed with their offspring in plastic cages with paper bedding until Day 21 post‐delivery (where Day 0 is the day of delivery). Animals were maintained in an air‐conditioned animal room (temperature: 23 ± 2°C, relative humidity: 55 ± 15%) with a 12‐h light/dark cycle, and provided pelleted basal diet (CRF‐1; Oriental Yeast Co. Ltd. Tokyo, Japan) throughout the experimental period and tap water until the start of exposure to chemicals ad libitum. From postnatal day (PND) 21 onwards, offspring were reared two or three animals per cage and provided pelleted CRF‐1 basal diet and tap water ad libitum.

Experimental design

Mated female rats were randomly divided into four groups of 12–13 animals and were either left untreated (untreated controls; n = 13) or treated with PTU at 10 ppm (PTU group; n = 12), GLY at 1,000 ppm (GLY group; n = 12), or VPA at 2,000 ppm (VPA group; n = 12) in drinking water from GD 6 to Day 21 post‐delivery. The doses and experimental design were identical to those reported previously (Shiraki et al., 2016; Watanabe et al., 2017; Akane, Shiraki et al., 2013). On PND 4, the litters were randomly culled, leaving eight male offspring per dam. If dams had fewer than eight male pups, more female pups were included to maintain a total of eight pups per litter. Dams and female offspring were euthanized by exsanguination through the abdominal aorta under CO2/O2 anesthesia on Day 21 post‐delivery. Male offspring were used for gene selection in the hippocampal dentate gyrus, because neurogenesis is influenced by circulating levels of steroid hormones during the estrous cycle (Pawluski et al. 2009). Half of male offspring were subjected to necropsy on PND 21, and the remaining male offspring were maintained without chemical exposure until PND 77 and subjected to necropsy. Body and brain weights of offspring were measured on the necropsy day. At each time point, 10 or more animals in each group (one or two pups per litter) were selected for future immunohistochemical assessment, and the remaining animals were used for gene selection by means of analyses of DNA methylation and gene expression of transcript levels. During the experimental period, all animals were checked for general conditions regarding nutritional state and signs of abnormal behavior in their home cage in gross observation. Body weight and food and water consumption were measured once or twice per week throughout the study period.

All procedures of the animal experiment were conducted in compliance with the Guidelines for Proper Conduct of Animal Experiments (Science Council of Japan, June 1, 2006) and according to the protocol approved by the Animal Care and Use Committee of Tokyo University of Agriculture and Technology (Approved no.: 30–60). All efforts were made to minimize animal suffering.

DNA and RNA extraction

For gene screening and following validation analysis of gene transcript levels on PND 21 and PND 77, 27–32 male offspring per group (2–4 male offspring per dam) were euthanized by exsanguination through the abdominal aorta under CO2/O2 anesthesia and subjected to necropsy, and then brains were removed, fixed in methacarn solution for 5 h at 4°C, and then dehydrated in ice‐cold absolute ethanol overnight at 4°C, as described previously (Akane, Saito et al., 2013). After dehydration, 2‐mm‐thick coronal cerebral slices were prepared at the position of −3.5 mm from the bregma. Tissues of the hippocampal dentate gyrus were collected from the slice using a punch biopsy device with a pore‐size diameter of 1 mm (Kai Industries Co., Ltd., Gifu, Japan) and stored in ethanol at −80°C until extraction. For analyses of DNA methylation and expression of gene transcript levels, genomic DNA (gDNA) and total RNA were extracted from tissue samples using an Allprep DNA/RNA Mini Kit (Qiagen, Hilden, Germany). Extracted gDNA was used for Methyl‐Seq analysis (n = 5 from different dams/group, pooled as one sample). Extracted total RNA was used for RNA‐Seq analysis (n = 5 from different dams/group, pooled as one sample) and real‐time RT‐PCR analysis (n = 6 from different dams/group).

Methyl‐Seq analysis

To identify PTU, VPA, or GLY‐induced DNA methylation changes on PND 21, SureSelect Target Enrichment System (Rat Methyl‐Seq; Agilent Technologies, Santa Clara, CA, USA) was implemented according to the manufacturer's protocol (SureSelectXT Methyl‐Seq Target Enrichment System, version E0, April 2018). Using the publicly available databases from the University of California Santa Cruz (UCSC) Genome Browser (http://genome.ucsc.edu), genomic coordinates for all known CGIs, shores, and shelves in the rat genome were obtained. Briefly, 0.8 μg of gDNA from each animal was pooled from five animals of each group to prepare one sample. Each pooled gDNA sample was fragmented using a Covaris sonicator (Covaris, Woburn, MA, USA). These fragments were end‐repaired, 3′‐adenylated, and further ligated with methylated primers. Following hybridization to biotinylated, plus‐strand DNA‐complementary RNA library ‘baits’, precipitation from the solution using streptavidin‐coated magnetic beads, and RNase‐digestion of the baits, captured DNA was bisulfite‐converted using the EZ‐DNA Methylation‐Gold Kit (Zymo Research, Irvine, CA, USA). Subsequently, DNA samples were PCR‐amplified using sample‐specific indexed (‘barcoding’) primers to allow for multiplexing and sequenced by 150 bp paired‐end sequencing by Illumina NovaSeq 6000 (Illumina, Inc., San Diego, CA, USA) as described in the manufacturer's protocol. After sequencing, the raw sequence reads were filtered based on quality by FastQC v0.11.5 (Babraham Institute, Cambridgeshire, UK). The adapter sequences were also trimmed off the raw sequence reads by Trimmomatic v0.32 (RWTH Aachen University, Aachen, Germany). The trimmed reads were mapped to the reference genome, the Rattus norvegicus.rn4.fa (with GA, CT converted Index), which was produced by the Rat Genome Reference Consortium, with BSMAP version 2.87 (https://code.google.com/archive/p/bsmap/), which was based on the SOAP (Short Oligo Alignment Program). The only uniquely mapped reads were selected to sort and index, and PCR duplicates were removed with SAMBAMBA version 0.5.9 (http://lomereiter.github.io/sambamba/). The methylation ratio of every single cytosine location within the on‐target region was extracted from the mapping results using ‘methylatio.py’ script in BSMAP. The results of the coverage profiles were calculated as number of C/effective CT counts for each cytosine in CpG.

Genes showing CpG site hypermethylation up to 2‐kb upstream from the transcription start site were selected with the criterion of the methylation ratio of ≥0.2 in each treatment group sample as compared with the untreated control sample.

RNA‐Seq analysis

Whole transcriptome sequencing was performed to identify PTU, VPA, or GLY‐induced transcript‐level expression changes on PND 21. TruSeq Stranded mRNA LT Sample Prep Kit (Illumina, Inc.) was used to sample preparation according to the manufacturer's protocol (TruSeq Stranded mRNA Sample Preparation Guide Part #1531047 Rev. E, October 2013). Using the publicly available databases from the UCSC Genome Browser (http://genome.ucsc.edu), genomic coordinates in the rat genome were obtained. Briefly, 0.4 μg of total RNA from each animal was pooled from five animals of each group to prepare one sample. DNA contamination was eliminated using DNase. The poly‐A containing mRNA molecules were purified using poly‐T oligo‐attached magnetic beads. The fragments of purified RNA were reverse‐transcribed into complementary DNA (cDNA). Adapters were ligated onto both ends of the cDNA fragments. After amplifying fragments using PCR, fragments with insert sizes between 200 and 400 bp were selected. Subsequently, cDNA fragments were PCR‐amplified using sample‐specific indexed (barcoding) primers to allow for multiplexing and sequenced by 100 bp paired‐end sequencing by Illumina NovaSeq 6000 as described in the manufacturer's protocol. After sequencing, the raw sequence reads were filtered based on quality by FastQC version 0.11.7 (Babraham Institute). The adapter sequences were also trimmed off the raw sequence reads by Trimmomatic version 0.32 (RWTH Aachen University). The trimmed reads were assembled by StringTie version 13.3b (Johns Hopkins University, Baltimore, MD, USA) and mapped to the reference genome, the Rattus norvegicus.rn6.fa (with GA, CT converted Index), which was produced by the Rat Genome Reference Consortium with HISAT2 version 2.0.5, Bowtie 2 version 2.3.4.1 (Johns Hopkins University) splice‐aware aligner. Expression profiles were represented as read count and normalized value as fragments per kilobase of transcript per million mapped (FPKM) reads, which was based on transcript length and depth of coverage. The difference per comparison pair was expressed as ‘fc.’, the fold change between the log2 (1 + FPKM) value of the each chemical exposure group and that of the untreated controls.

Genes showing downregulation were selected with the criterion of the fc. of ≤ −1 as compared with the untreated controls, and gene ontology‐based functional annotation analysis was performed using the Database for Annotation, Visualization and Integrated Discovery (DAVID), version 6.7 (Huang et al., 2009a, b) to clarify biological functions.

Real‐time RT‐PCR analysis

Real‐time RT‐PCR quantification of transcript level was performed for genes selected as being hypermethylated in Methyl‐Seq analysis, downregulated in RNA‐Seq analysis, and related to the nervous system in functional annotation analysis. Around 50 to 60 genes that have been investigated the functional role in nervous system development and differentiation were selected per compound from those showing profound downregulation of transcript level in RNA‐Seq analysis. For genes whose downregulation of transcript levels was confirmed on PND 21, expression on PND 77 was also analyzed. First‐strand complementary DNA was synthesized using SuperScript® III Reverse Transcriptase (Thermo Fisher Scientific, Waltham, MA, USA) in a 20‐μl total reaction mixture with 0.6 μg of total RNA. Analysis of the transcript levels for the candidate genes shown in Table S1 was performed using PCR primers designed with Primer Express software version 3.0 (Thermo Fisher Scientific). Real‐time PCR with Power SYBR® Green PCR Master Mix (Thermo Fisher Scientific) was conducted using a StepOnePlus™ Real‐time PCR System (Thermo Fisher Scientific). The relative differences in gene expression between the untreated controls and each treatment group were calculated using threshold cycle (C T) values that were first normalized to that of Gapdh or Hprt1, which served as endogenous controls in the same sample, and then relative to a C T value of the untreated controls using the 2−ΔΔ T method (Livak and Schmittgen, 2001).

Statistical analysis

Numerical data were presented as the mean ± SD. Body weights, brain weights, and transcript expression data of offspring were analyzed using the litter as the experimental unit. Other data were analyzed using the individual animal as the experimental unit. Data were analyzed using Levene's test for homogeneity of variance. If the variance was homogenous in the analysis of data from four groups, that is, body and brain weights, food and water consumption, and reproductive parameters, numerical data were evaluated using Dunnett's test for comparisons between the untreated controls and each treatment group. For heterogeneous data, Aspin–Welch's t test with Bonferroni correction was used. If the variance was homogenous in the analysis of data from two groups, that is, transcript‐level expression, numerical data were evaluated using Student's t test. When data were heterogeneous, Aspin–Welch's t test was used. All analyses were performed using the IBM SPSS Statistics version 25 (IBM Corporation, Armonk, NY, USA), and p < 0.05 was considered statistically significant.

RESULTS

Maternal parameters

In the present study, there were two and one nonpregnant females in the untreated controls and GLY group, respectively.

In the PTU group, there were no abnormal clinical signs during the experimental period. Body weight was significantly decreased on PND 12 compared with the untreated controls (Figure S1A). Food consumption was significantly decreased on GD 20 and from PND 5 to PND 20 compared with the untreated controls (Figure S1B). Water consumption was significantly decreased from GD 20 to PND 20 compared with the untreated controls (Figure S1C). Maternal exposure to PTU did not affect reproductive parameters (Table S2).

In the VPA group, there were no abnormal clinical signs during the experimental period. Body weight was significantly decreased on PND 12 compared with the untreated controls (Figure S1A). Food consumption was significantly decreased on PND 9 and PND 16 compared with the untreated controls (Figure S1B). Water consumption was significantly decreased from GD 14 to GD 20 and from PND 5 to PND 20 compared with the untreated controls (Figure S1C). Maternal exposure to VPA did not affect reproductive parameters (Table S2).

Dams exposed to GLY exhibited abnormal gait after PND 12. Body weight was significantly decreased from GD 20 to PND 21 compared with the untreated controls (Figure S1A). Food consumption was significantly decreased from GD 20 to PND 20 compared with the untreated controls (Figure S1B). Water consumption was significantly decreased from GD 7 to PND 20 compared with the untreated controls (Figure S1C). Maternal exposure to GLY did not affect reproductive parameters (Table S2).

Clinical observations and necropsy data of offspring

In the PTU group, there were no abnormal clinical signs during the experimental period. Body weight was significantly decreased from PND 4 to PND 77 compared with the untreated controls (Figure S2A). Food and water consumption were significantly decreased from PND 28 to PND 56 compared with the untreated controls (Figure S2B,C). Body and brain weights were decreased at necropsies on PND 21 and PND 77 compared with the untreated controls (Table S3).

In the VPA group, there were no abnormal clinical signs during the experimental period. Body weight was significantly decreased on PND 4 and PND 8 and from PND 15 to PND 28 compared with the untreated controls (Figure S2A). There were no significant changes in food and water consumption between the VPA group and the untreated controls (Figure S2B,C). Body weight was decreased at necropsy on PND 21 compared with the untreated controls (Table S3).

In the GLY group, there were no abnormal clinical signs during the experimental period. Body weight was significantly decreased from PND 4 to PND 77 compared with the untreated controls (Figure S2A). Food and water consumption were significantly decreased from PND 28 to PND 42 compared with the untreated controls (Figure S2B,C). Body and brain weights were decreased at necropsies on PND 21 and PND 77 compared with the untreated controls (Table S3).

Hypermethylated and downregulated genes in the hippocampal dentate gyrus

The metrics for the high‐throughput sequencing of samples were shown in Tables S4 and S5. The numbers of genes showing hypermethylation and downregulation by Methyl‐Seq and RNA‐Seq analyses were shown in Figure 1. Total of 1,220, 429, and 805 genes showed promotor‐region hypermethylation (difference in the methylation ratio ≥ 0.2 as compared with the untreated controls) and downregulation of the transcript level (difference in the fc. ≤ −1 as compared with the untreated controls) on PND 21 by maternal exposure to PTU, VPA, and GLY, respectively. There were 77 genes in common with all 3 compounds, 84 genes in common with both PTU and VPA, 313 genes in common with both PTU and GLY, and 40 genes in common with both VPA and GLY (Figure 1A). Among the hypermethylated and downregulated genes, 249, 81, and 183 genes were related to the nervous system in the PTU, VPA, and GLY by functional annotation investigation using DAVID version 6.7, respectively. There were 18 genes in common with all 3 compounds, 9 genes in common with both PTU and VPA, 65 genes in common with both PTU and GLY, and 7 genes in common with both VPA and GLY (Figure 1B).

FIGURE 1

Venn diagram denoting the number of genes showing promotor‐region hypermethylation and transcript downregulation by means of Methyl‐Seq and RNA‐Seq analyses on PND 21. (A) Total number of genes obtained. (B) Number of nervous system‐related genes selected by functional annotation

Transcript expression changes of candidate genes

Among the hypermethylated and downregulated genes, 61, 50, and 59 genes were selected for validation analysis of transcript levels by real‐time RT‐PCR in comparison with the untreated controls in the PTU, VPA, and GLY, respectively (Table 2 and S6–S11). Among them, transcript downregulation was confirmed with 39, 7, and 23 genes on PND 21 in the PTU, VPA, and GLY, respectively (Tables 2–5). On PND 77, the downregulation was further sustained with 12, 3, and 0 genes in the PTU, VPA, and GLY, respectively. In the PTU group, transcript levels of Arc, Bmp3, Cebpb, Creb1, Epha7, Fgf13, Fzd9, Hes5, Mas1, Pvalb, Reln, and Sema3c were significantly decreased compared with the untreated controls on both PND 21 and PND 77 (Table 3). In this group, transcript levels of Acsl4, Arhgef2, Atp2b2, Baiap2, Cntn2, Dab2ip, Dao, Dll4, Epha5, Fgf2, Id2, Igf1, Ndel1, Neurod6, Numbl, Ptpru, Robo3, Ror2, Shh, Six4, Tfap2c, Thrsp, Vegfa, Wnt7b, Wnt8b, Wnt9a, and Wnt16 were significantly decreased only on PND 21 compared with the untreated controls. In the VPA group, transcript levels of Dpysl4, Sox2, and Vgf were significantly decreased compared with the untreated controls on both PND 21 and PND 77 (Table 4). In this group, transcript levels of Pvalb, Snw1, Tead3, and Wdpcp were significantly decreased only on PND 21 compared with the untreated controls. In the GLY group, transcript levels of Acsl4, Afdn, Arc, Arx, Baiap2, Cxcr4, Dgkg, Epha5, Fgf2, Fgf13, Gabrb1, Grin2b, Gsk3b, Hes5, Kalrn, Mas1, Neurod2, Ptpru, Sema3c, Snw1, Sorl1, Stx3, and Tcf4 were significantly decreased only on PND 21 compared with the untreated controls (Table 5).

TABLE 2

Number of genes validated for downregulation of transcript level by real‐time reverse transcription‐PCR (RT‐PCR)

Variable	PTU	VPA	GLY	Common to all three chemicals	Common to PTU and VPA	Common to PTU and GLY	Common to VPA and GLY
Selected for validation analysis	61	50	59	13	6	21	6
Downregulated at PND 21	39	7	23	0	1	10	1
Downregulated at PND 77	12	3	0	0	0	0	0

Abbreviations: GLY, glycidol; PTU, 6‐propyl‐2‐thiouracil; RT, reverse transcription; VPA, valproic acid.

TABLE 5

List of validated genes for transcript‐level downregulation on PND 21 and transcript levels of expression on PND 21 and PND 77 after developmental exposure to GLY

Gene	PND 21				PND 77
	Control		GLY		Control		GLY
	Normalized by		Normalized by		Normalized by		Normalized by
	Gapdh	Hprt1	Gapdh	Hprt1	Gapdh	Hprt1	Gapdh	Hprt1
Acsl4	1.01 ± 0.15	1.01 ± 0.15	0.79 ± 0.25	0.80 ± 0.08*	1.01 ± 0.17	1.00 ± 0.11	0.89 ± 0.07	0.94 ± 0.09
Afdn	1.03 ± 0.27	1.02 ± 0.18	0.68 ± 0.27	0.69 ± 0.16**	1.01 ± 0.17	1.01 ± 0.12	0.93 ± 0.30	0.96 ± 0.24
Arc	1.03 ± 0.27	1.03 ± 0.29	0.37 ± 0.15**	0.42 ± 0.17**	1.12 ± 0.47	1.11 ± 0.45	0.87 ± 0.54	0.92 ± 0.60
Arx	1.05 ± 0.34	1.01 ± 0.14	0.56 ± 0.30*	0.58 ± 0.25**	1.00 ± 0.08	1.01 ± 0.14	0.86 ± 0.34	0.89 ± 0.32
Baiap2	1.03 ± 0.27	1.02 ± 0.22	0.62 ± 0.30*	0.62 ± 0.22*	1.01 ± 0.17	1.01 ± 0.16	0.89 ± 0.22	0.92 ± 0.16
Cxcr4	1.02 ± 0.20	1.03 ± 0.23	0.64 ± 0.25*	1.02 ± 0.28	1.01 ± 0.16	1.00 ± 0.06	0.85 ± 0.17	0.88 ± 0.11
Dgkg	1.08 ± 0.42	1.03 ± 0.27	0.59 ± 0.21*	0.63 ± 0.13**	1.00 ± 0.07	1.00 ± 0.11	0.88 ± 0.22	0.94 ± 0.31
Epha5	1.03 ± 0.27	1.04 ± 0.30	0.64 ± 0.34*	0.55 ± 0.21**	1.02 ± 0.20	1.01 ± 0.14	0.89 ± 0.29	0.92 ± 0.25
Fgf2	1.08 ± 0.48	1.07 ± 0.42	0.53 ± 0.18*	0.54 ± 0.10*	1.10 ± 0.62	1.12 ± 0.71	0.95 ± 0.15	1.01 ± 0.26
Fgf13	1.04 ± 0.33	1.01 ± 0.17	0.60 ± 0.24*	0.65 ± 0.23*	1.01 ± 0.16	1.02 ± 0.23	0.90 ± 0.13	0.95 ± 0.18
Gabrb1	1.01 ± 0.18	1.02 ± 0.20	0.67 ± 0.26*	0.60 ± 0.12**	1.01 ± 0.13	1.01 ± 0.14	0.88 ± 0.14	0.92 ± 0.08
Grin2b	1.02 ± 0.24	1.03 ± 0.25	0.79 ± 0.32	0.71 ± 0.17*	1.01 ± 0.18	1.01 ± 0.16	0.87 ± 0.22	0.91 ± 0.16
Gsk3b	1.00 ± 0.07	1.00 ± 0.07	0.78 ± 0.16*	1.28 ± 0.21	1.00 ± 0.10	1.00 ± 0.06	0.91 ± 0.18	0.95 ± 0.16
Hes5	1.02 ± 0.26	1.02 ± 0.25	0.54 ± 0.17**	0.88 ± 0.15	1.03 ± 0.29	1.02 ± 0.20	0.92 ± 0.18	0.96 ± 0.11
Kalrn	1.06 ± 0.43	1.06 ± 0.41	0.47 ± 0.14*	0.48 ± 0.07*	1.02 ± 0.23	1.02 ± 0.19	0.86 ± 0.11	0.90 ± 0.11
Mas1	1.02 ± 0.23	1.02 ± 0.23	0.68 ± 0.29*	0.70 ± 0.12*	1.01 ± 0.11	1.00 ± 0.03	0.95 ± 0.23	0.99 ± 0.18
Neurod2	1.05 ± 0.35	1.02 ± 0.21	0.55 ± 0.14*	0.62 ± 0.20**	1.01 ± 0.12	1.00 ± 0.05	0.94 ± 0.31	0.96 ± 0.25
Ptpru	1.11 ± 0.57	1.11 ± 0.57	0.43 ± 0.12*	0.69 ± 0.10	1.16 ± 0.53	1.18 ± 0.57	0.81 ± 0.41	0.83 ± 0.37
Sema3c	1.02 ± 0.19	1.01 ± 0.16	0.61 ± 0.23**	0.97 ± 0.25	1.04 ± 0.32	1.02 ± 0.21	0.95 ± 0.23	0.99 ± 0.19
Snw1	1.00 ± 0.05	1.00 ± 0.08	0.83 ± 0.18*	1.35 ± 0.15	1.01 ± 0.12	1.00 ± 0.07	0.96 ± 0.17	1.00 ± 0.11
Sorl1	1.07 ± 0.40	1.03 ± 0.26	0.52 ± 0.22*	0.53 ± 0.10**	1.01 ± 0.15	1.01 ± 0.15	0.99 ± 0.13	1.05 ± 0.17
Stx3	1.01 ± 0.19	1.02 ± 0.21	0.73 ± 0.12*	1.20 ± 0.16	1.01 ± 0.18	1.03 ± 0.24	1.30 ± 0.60	1.36 ± 0.65
Tcf4	1.04 ± 0.31	1.01 ± 0.12	0.67 ± 0.20*	0.75 ± 0.25*	1.00 ± 0.09	1.01 ± 0.12	0.87 ± 0.22	0.91 ± 0.17

Note. Data are expressed as the mean ± SD (n = 6/group).

Abbreviations: Acsl4, acyl‐CoA synthetase long‐chain family member 4; Afdn, afadin, adherens junction formation factor; Arc, activity‐regulated cytoskeleton‐associated protein; Arx, aristaless‐related homeobox; Baiap2, BAR/IMD domain containing adaptor protein 2; Cxcr4, C‐X‐C motif chemokine receptor 4; Dgkg, diacylglycerol kinase, gamma; Epha5, EPH receptor A5; Fgf2, fibroblast growth factor 2; Fgf13, fibroblast growth factor 13; Gabrb1, gamma‐aminobutyric acid type A receptor subunit beta1; Gapdh, glyceraldehyde‐3‐phosphate dehydrogenase; GLY, glycidol; Grin2b, glutamate ionotropic receptor NMDA type subunit 2B; Gsk3b, glycogen synthase kinase 3 beta; Hes5, hes family bHLH transcription factor 5; Hprt1, hypoxanthine phosphoribosyltransferase 1; Kalrn, kalirin, RhoGEF kinase; Mas1, MAS1 proto‐oncogene, G protein‐coupled receptor; Neurod2, neuronal differentiation 2; PND, postnatal day; Ptpru, protein tyrosine phosphatase, receptor type, U; Sema3c, semaphorin 3C; Snw1, SNW domain containing 1; Sorl1, sortilin‐related receptor 1; Stx3, syntaxin 3; Tcf4, transcription factor 4.

p < 0.05.

p < 0.01, significantly different from the untreated controls by Student's or Aspin–Welch's t test.

TABLE 3

List of validated genes for transcript‐level downregulation on PND 21 and transcript levels of expression on PND 21 and PND 77 after developmental exposure to PTU

Gene	PND 21				PND 77
	Control		PTU		Control		PTU
	Normalized by		Normalized by		Normalized by		Normalized by
	Gapdh	Hprt1	Gapdh	Hprt1	Gapdh	Hprt1	Gapdh	Hprt1
Arc	1.03 ± 0.27	1.03 ± 0.29	0.33 ± 0.22**	0.33 ± 0.20**	1.00 ± 0.10	1.00 ± 0.08	0.53 ± 0.20**	0.52 ± 0.12**
Bmp3	1.10 ± 0.51	1.04 ± 0.30	0.50 ± 0.26*	0.39 ± 0.20**	1.03 ± 0.29	1.01 ± 0.17	0.59 ± 0.22*	0.83 ± 0.39
Cebpb	1.02 ± 0.23	1.03 ± 0.28	0.58 ± 0.14**	0.55 ± 0.14**	1.02 ± 0.19	1.01 ± 0.16	0.77 ± 0.15*	0.79 ± 0.16*
Creb1	1.02 ± 0.21	1.01 ± 0.16	0.83 ± 0.14	0.79 ± 0.13*	1.01 ± 0.15	1.00 ± 0.05	0.79 ± 0.09*	1.08 ± 0.12
Epha7	1.02 ± 0.23	1.02 ± 0.24	0.46 ± 0.13**	0.43 ± 0.09**	1.02 ± 0.20	1.01 ± 0.19	0.72 ± 0.16*	0.73 ± 0.16*
Fgf13	1.04 ± 0.33	1.01 ± 0.17	0.69 ± 0.26	0.54 ± 0.20**	1.00 ± 0.08	1.00 ± 0.10	0.64 ± 0.03**	0.87 ± 0.08*
Fzd9	1.01 ± 0.17	1.01 ± 0.16	0.67 ± 0.17**	1.06 ± 0.20	1.01 ± 0.16	1.00 ± 0.09	0.75 ± 0.28	0.74 ± 0.11**
Hes5	1.02 ± 0.26	1.02 ± 0.25	0.62 ± 0.10*	1.01 ± 0.15	1.00 ± 0.07	1.01 ± 0.12	0.82 ± 0.20	0.83 ± 0.13*
Mas1	1.02 ± 0.23	1.02 ± 0.23	0.68 ± 0.42	0.52 ± 0.31*	1.01 ± 0.18	1.01 ± 0.13	0.77 ± 0.09*	1.05 ± 0.12
Pvalb	1.05 ± 0.40	1.06 ± 0.47	0.26 ± 0.08**	0.25 ± 0.06**	1.01 ± 0.12	1.00 ± 0.06	0.59 ± 0.14**	0.59 ± 0.06**
Reln	1.04 ± 0.31	1.03 ± 0.31	0.77 ± 0.11	0.72 ± 0.11*	1.02 ± 0.19	1.01 ± 0.12	0.73 ± 0.09**	0.99 ± 0.17
Sema3c	1.02 ± 0.19	1.01 ± 0.16	0.09 ± 0.03**	0.15 ± 0.05**	1.01 ± 0.17	1.01 ± 0.15	0.47 ± 0.18**	0.46 ± 0.07**
Acsl4	1.01 ± 0.15	1.01 ± 0.15	0.78 ± 0.10*	0.74 ± 0.08**	1.01 ± 0.16	1.01 ± 0.13	0.97 ± 0.17	0.99 ± 0.08
Arhgef2	1.03 ± 0.30	1.00 ± 0.10	0.73 ± 0.16	0.59 ± 0.22**	1.00 ± 0.07	1.00 ± 0.04	1.12 ± 0.20	1.15 ± 0.15
Atp2b2	1.01 ± 0.19	1.02 ± 0.23	0.58 ± 0.05**	0.55 ± 0.07**	1.01 ± 0.11	1.00 ± 0.05	1.16 ± 0.21	1.18 ± 0.21
Baiap2	1.03 ± 0.27	1.02 ± 0.22	0.69 ± 0.20*	0.65 ± 0.18**	1.01 ± 0.15	1.01 ± 0.12	0.88 ± 0.12	0.90 ± 0.12
Cntn2	1.03 ± 0.30	1.03 ± 0.29	0.45 ± 0.09**	0.42 ± 0.08**	1.01 ± 0.15	1.00 ± 0.08	1.54 ± 0.40*	1.55 ± 0.09**
Dab2ip	1.02 ± 0.23	1.02 ± 0.22	0.78 ± 0.17	0.72 ± 0.09*	1.01 ± 0.16	1.00 ± 0.09	0.99 ± 0.07	1.36 ± 0.15
Dao	1.06 ± 0.42	1.04 ± 0.32	0.45 ± 0.31*	0.43 ± 0.23**	1.19 ± 0.61	1.17 ± 0.59	1.74 ± 1.18	2.38 ± 1.61
Dll4	1.02 ± 0.21	1.01 ± 0.18	0.74 ± 0.14*	0.69 ± 0.15**	1.03 ± 0.26	1.02 ± 0.25	0.96 ± 0.19	1.01 ± 0.31
Epha5	1.03 ± 0.27	1.04 ± 0.30	0.45 ± 0.23**	0.42 ± 0.19**	1.00 ± 0.08	1.01 ± 0.13	0.99 ± 0.23	1.00 ± 0.14
Fgf2	1.08 ± 0.48	1.07 ± 0.42	0.49 ± 0.08*	0.46 ± 0.05*	1.02 ± 0.20	1.02 ± 0.20	0.99 ± 0.24	1.00 ± 0.15
Id2	1.01 ± 0.17	1.01 ± 0.14	0.73 ± 0.21*	0.66 ± 0.11**	1.01 ± 0.19	1.01 ± 0.14	0.86 ± 0.22	0.86 ± 0.12
Igf1	1.01 ± 0.16	1.01 ± 0.14	0.77 ± 0.08**	1.24 ± 0.18*	1.02 ± 0.21	1.01 ± 0.19	1.65 ± 0.31**	1.72 ± 0.44**
Ndel1	1.01 ± 0.14	1.01 ± 0.16	0.81 ± 0.13*	0.75 ± 0.04*	1.00 ± 0.05	1.00 ± 0.03	0.90 ± 0.14	0.92 ± 0.11
Neurod6	1.01 ± 0.17	1.02 ± 0.23	0.48 ± 0.40*	0.44 ± 0.34**	1.01 ± 0.13	1.01 ± 0.18	0.85 ± 0.25	0.85 ± 0.08
Numbl	1.04 ± 0.30	1.03 ± 0.25	0.89 ± 0.17	0.70 ± 0.16*	1.00 ± 0.06	1.00 ± 0.06	1.03 ± 0.14	1.06 ± 0.13
Ptpru	1.11 ± 0.57	1.11 ± 0.57	0.45 ± 0.10*	0.73 ± 0.17	1.01 ± 0.18	1.01 ± 0.17	0.86 ± 0.20	0.88 ± 0.14
Robo3	1.02 ± 0.22	1.02 ± 0.24	0.06 ± 0.02**	0.06 ± 0.03**	1.01 ± 0.18	1.01 ± 0.17	0.91 ± 0.1	0.94 ± 0.17
Ror2	1.06 ± 0.40	1.04 ± 0.34	0.63 ± 0.14*	0.58 ± 0.07*	1.02 ± 0.23	1.02 ± 0.23	1.06 ± 0.24	1.08 ± 0.20
Shh	1.07 ± 0.42	1.06 ± 0.38	0.53 ± 0.16*	0.51 ± 0.21*	1.03 ± 0.28	1.03 ± 0.30	1.50 ± 0.51	1.57 ± 0.60
Six4	1.10 ± 0.51	1.15 ± 0.66	0.53 ± 0.16*	0.51 ± 0.14*	1.02 ± 0.24	1.02 ± 0.22	1.43 ± 0.40	1.46 ± 0.36*
Tfap2c	1.05 ± 0.34	1.04 ± 0.32	0.37 ± 0.16**	0.60 ± 0.27*	1.09 ± 0.52	1.08 ± 0.49	0.95 ± 0.27	1.27 ± 0.31
Thrsp	1.05 ± 0.34	1.01 ± 0.16	0.47 ± 0.09**	0.37 ± 0.09**	1.01 ± 0.18	1.01 ± 0.14	1.01 ± 0.24	1.02 ± 0.11
Vegfa	1.10 ± 0.58	1.03 ± 0.31	0.84 ± 0.14	0.68 ± 0.19*	1.02 ± 0.21	1.01 ± 0.17	1.25 ± 0.25	1.27 ± 0.12*
Wnt7b	1.10 ± 0.54	1.03 ± 0.28	0.66 ± 0.10	0.53 ± 0.14**	1.02 ± 0.21	1.01 ± 0.15	1.25 ± 0.23	1.28 ± 0.17*
Wnt8b	1.07 ± 0.43	1.06 ± 0.43	0.62 ± 0.17	0.50 ± 0.22*	1.02 ± 0.22	1.03 ± 0.27	1.01 ± 0.57	0.97 ± 0.32
Wnt9a	1.13 ± 0.66	1.06 ± 0.40	0.56 ± 0.11	0.45 ± 0.12**	1.07 ± 0.43	1.07 ± 0.41	0.79 ± 0.10	0.81 ± 0.14
Wnt16	1.06 ± 0.36	1.05 ± 0.31	0.51 ± 0.20**	0.41 ± 0.23**	1.03 ± 0.26	1.02 ± 0.23	1.04 ± 0.24	1.09 ± 0.35

Note. Data are expressed as the mean ± SD (n = 6/group).

Abbreviations: Acsl4, acyl‐CoA synthetase long‐chain family member 4; Arc, activity‐regulated cytoskeleton‐associated protein; Arhgef2, Rho guanine nucleotide exchange factor 2; Atp2b2, ATPase plasma membrane Ca2+ transporting 2; Baiap2, BAR/IMD domain containing adaptor protein 2; Bmp3, bone morphogenetic protein 3; Cebpb, CCAAT/enhancer‐binding protein beta; Cntn2, contactin 2; Creb1, cAMP‐responsive element‐binding protein 1; Dab2ip, DAB2‐interacting protein; Dao, d‐amino‐acid oxidase; Dll4, delta‐like canonical Notch ligand 4; Epha5, EPH receptor A5; Epha7, Eph receptor A7; Fgf2, fibroblast growth factor 2; Fgf13, fibroblast growth factor 13; Fzd9, frizzled class receptor 9; Gapdh, glyceraldehyde‐3‐phosphate dehydrogenase; Hes5, hes family bHLH transcription factor 5; Hprt1, hypoxanthine phosphoribosyltransferase 1; Id2, inhibitor of DNA‐binding 2; Igf1, insulin‐like growth factor 1; Mas1, MAS1 proto‐oncogene, G protein‐coupled receptor; Ndel1, nudE neurodevelopment protein 1‐like 1; Neurod6, neuronal differentiation 6; Numbl, NUMB‐like, endocytic adaptor protein; PND, postnatal day; Ptpru, protein tyrosine phosphatase, receptor type, U; PTU, 6‐propyl‐2‐thiouracil; Pvalb, parvalbumin; Reln, reelin; Robo3, roundabout guidance receptor 3; Ror2, receptor tyrosine kinase‐like orphan receptor 2; Sema3c, semaphorin 3C; Shh, sonic hedgehog signaling molecule; Six4, SIX homeobox 4; Tfap2c, transcription factor AP‐2 gamma; Thrsp, thyroid hormone responsive; Vegfa, vascular endothelial growth factor A; Wnt7b, Wnt family member 7B; Wnt8b, Wnt family member 8B; Wnt9a, Wnt family member 9A; Wnt16, Wnt family member 16.

p < 0.05

p < 0.01, significantly different from the untreated controls by Student's or Aspin–Welch's t test.

TABLE 4

List of validated genes for transcript‐level downregulation on PND 21 and transcript levels of expression on PND 21 and PND 77 after developmental exposure to VPA

Gene	PND 21				PND 77
	Control		VPA		Control		VPA
	Normalized by		Normalized by		Normalized by		Normalized by
	Gapdh	Hprt1	Gapdh	Hprt1	Gapdh	Hprt1	Gapdh	Hprt1
Dpysl4	1.02 ± 0.19	1.02 ± 0.24	0.71 ± 0.14*	1.21 ± 0.26	1.01 ± 0.17	1.02 ± 0.19	0.73 ± 0.09**	0.67 ± 0.02**
Sox2	1.04 ± 0.32	1.04 ± 0.34	0.53 ± 0.33*	0.49 ± 0.31*	1.06 ± 0.33	1.01 ± 0.18	0.58 ± 0.10**	0.54 ± 0.05**
Vgf	1.00 ± 0.08	1.01 ± 0.14	0.86 ± 0.24	0.75 ± 0.20*	1.07 ± 0.35	1.02 ± 0.21	0.72 ± 0.18	0.66 ± 0.12**
Pvalb	1.05 ± 0.40	1.06 ± 0.47	0.65 ± 0.15*	0.65 ± 0.07	1.03 ± 0.25	1.00 ± 0.11	0.94 ± 0.59	0.85 ± 0.50
Snw1	1.00 ± 0.05	1.00 ± 0.08	0.76 ± 0.06**	1.28 ± 0.13**	1.02 ± 0.20	1.01 ± 0.17	0.94 ± 0.17	0.86 ± 0.08
Tead3	1.02 ± 0.26	1.01 ± 0.15	0.80 ± 0.24	0.79 ± 0.11*	1.02 ± 0.21	1.00 ± 0.06	0.97 ± 0.28	0.88 ± 0.22
Wdpcp	1.02 ± 0.20	1.02 ± 0.20	0.68 ± 0.07**	1.16 ± 0.17	1.03 ± 0.28	1.01 ± 0.16	1.03 ± 0.27	0.93 ± 0.17

Note. Data are expressed as the mean ± SD (n = 6/group).

Abbreviations: Dpysl4, dihydropyrimidinase‐like 4; Gapdh, glyceraldehyde‐3‐phosphate dehydrogenase; Hprt1, hypoxanthine phosphoribosyltransferase 1; PND, postnatal day; Pvalb, parvalbumin; Snw1, SNW domain containing 1; Sox2, SRY‐box transcription factor 2; Tead3, TEA domain transcription factor 3; Vgf, VGF nerve growth factor inducible; VPA, valproic acid; Wdpcp, WD repeat containing planar cell polarity effector.

p < 0.05.

p < 0.01, significantly different from the untreated controls by Student's or Aspin–Welch's t test.

Functional annotation clustering of genes that have been confirmed to show transcript downregulation on PND 21 revealed diverse gene functions by PTU exposure, such as neuron differentiation, regulation of apoptotic process, neuron migration, neurogenesis, neuron projection development, and axon guidance (Table 6). With regard to VPA exposure, finally, selected genes were small in number; however, diverse neural functions were found to be involved, such as neuron differentiation and regulation of neurogenesis (Table 7). Similarly, GLY exposure involved diverse neural functions, such as neuron differentiation, synaptic plasticity, hippocampus development, and neuron migration (Table 8).

TABLE 6

Classification of hypermethylated and downregulated genes by means of functional annotation clustering in the hippocampal dentate gyrus of offspring on PND 21 after maternal exposure to PTU

Gene function	Number of genes	Gene symbol
Neuron differentiation	15	Cebpb a , Hes5 a , Acsl4, Arhgef2, Atp2b2, Baiap2, Id2, Neurod6, Ptpru, Shh, Tfap2c, Wnt7b , Wnt8b, Wnt9a, Wnt16
Apoptotic process	11	Bmp3 a , Cebpb a , Creb1 a , Epha7 a , Fgf13 a , Fzd9 a , Arhgef2, Dab2ip, Igf1, Six4, Vegfa
Neuron migration	8	Fgf13 a , Reln a , Sema3c a , Arhgef2, Cntn2, Dab2ip, Ndel1, Robo3
Neurogenesis	6	Hes5 a , Arhgef2, Dll4, Numbl, Thrsp, Wnt7b
Neuron projection development	6	Reln a , Baiap2, Cntn2, Dab2ip, Ndel1, Wnt7b
Axon guidance	5	Reln a , Epha5, Robo3, Sema3c a , Vegfa
Hippocampus development	4	Fgf13 a , Mas1 a , Reln a , Epha5
Neuroblast proliferation	4	Fzd9 a , Numbl, Shh, Vegfa
Synaptic plasticity	4	Arc a , Atp2b2, Baiap2, Cntn2
Axonogenesis	4	Creb1 a , Cntn2, Ndel1, Numbl
Astrocyte differentiation, astrocyte development	4	Hes5 a , Cntn2, Id2, Ror2
Oligodendrocyte differentiation, oligodendrocyte development, and myelination	4	Hes5 a , Cntn2, Id2, Shh
Neural precursor cell proliferation	3	Fzd9 a, Dll4, Id2
Glial cell differentiation	3	Reln a , Fgf2, Igf1
Dendrite development	2	Baiap2, Dab2ip
Ephrin signaling pathway	2	Epha7 a , Epha5
Regulation of cytosolic calcium ion concentration	2	Pvalb a , Atp2b2
Synaptic transmission, glutamatergic	2	Reln a , Ror2
Neuronal stem cell population maintenance	1	Hes5 a
Radial glial cell differentiation	1	Hes5 a
Establishment of neuroblast polarity	1	Fgf13 a
Asymmetric neuroblast division	1	Arhgef2
Cerebral cortex GABAergic interneuron migration	1	Cntn2
Dopamine biosynthetic process	1	Dao

Abbreviations: Acsl4, acyl‐CoA synthetase long‐chain family member 4; Arc, activity‐regulated cytoskeleton‐associated protein; Arhgef2, Rho guanine nucleotide exchange factor 2; Atp2b2, ATPase plasma membrane Ca2+ transporting 2; Baiap2, BAR/IMD domain containing adaptor protein 2; Bmp3, bone morphogenetic protein 3; Cebpb, CCAAT/enhancer‐binding protein beta; Cntn2, contactin 2; Creb1, cAMP‐responsive element‐binding protein 1; Dab2ip, DAB2‐interacting protein; Dao, d‐amino‐acid oxidase; Dll4, delta‐like canonical Notch ligand 4; Epha5, EPH receptor A5; Epha7, Eph receptor A7; Fgf2, fibroblast growth factor 2; Fgf13, fibroblast growth factor 13; Fzd9, frizzled class receptor 9; Gapdh, glyceraldehyde‐3‐phosphate dehydrogenase; Hes5, hes family bHLH transcription factor 5; Hprt1, hypoxanthine phosphoribosyltransferase 1; Id2, inhibitor of DNA‐binding 2; Igf1, insulin‐like growth factor 1; Mas1, MAS1 proto‐oncogene, G protein‐coupled receptor; Ndel1, nudE neurodevelopment protein 1‐like 1; Neurod6, neuronal differentiation 6; Numbl, NUMB‐like, endocytic adaptor protein; PND, postnatal day; Ptpru, protein tyrosine phosphatase, receptor type, U; PTU, 6‐propyl‐2‐thiouracil; Pvalb, parvalbumin; Reln, reelin; Robo3, roundabout guidance receptor 3; Ror2, receptor tyrosine kinase‐like orphan receptor 2; Sema3c, semaphorin 3C; Shh, sonic hedgehog signaling molecule; Six4, SIX homeobox 4; Tfap2c, transcription factor AP‐2 gamma; Thrsp, thyroid hormone responsive; Vegfa, vascular endothelial growth factor A; Wnt7b, Wnt family member 7B; Wnt8b, Wnt family member 8B; Wnt9a, Wnt family member 9A; Wnt16, Wnt family member 16.

Downregulated on both PND 21 and PND 77.

TABLE 7

Classification of hypermethylated and downregulated genes by means of functional annotation clustering in the hippocampal dentate gyrus of offspring on PND 21 after maternal exposure to VPA

Gene function	Number of genes	Gene symbol
Neuron differentiation	2	Sox2 a , Vgf a
Regulation of neurogenesis	2	Sox2 a , Snw1
Neuronal stem cell population maintenance	1	Tead3
Asymmetric neuroblast division	1	Tead3
Neuroblast proliferation	1	Sox2 a
Neuron projection guidance	1	Dpysl4 a
Synaptic plasticity	1	Vgf a
Regulation of cytosolic calcium ion concentration	1	Pvalb
Nervous system development	1	Wdpcp

Abbreviations: Dpysl4, dihydropyrimidinase‐like 4; Gapdh, glyceraldehyde‐3‐phosphate dehydrogenase; Hprt1, hypoxanthine phosphoribosyltransferase 1; PND, postnatal day; Pvalb, parvalbumin; Snw1, SNW domain containing 1; Sox2, SRY‐box transcription factor 2; Tead3, TEA domain transcription factor 3; Vgf, VGF nerve growth factor inducible; VPA, valproic acid; Wdpcp, WD repeat containing planar cell polarity effector.

Downregulated on both PND 21 and PND 77.

TABLE 8

Classification of hypermethylated and downregulated genes by means of functional annotation clustering in the hippocampal dentate gyrus of offspring on PND 21 after maternal exposure to GLY

Gene function	Number of genes	Gene symbol
Neuron differentiation	7	Acsl4, Baiap2, Gsk3b, Hes5, Neurod2, Ptpru, Tcf4
Synaptic plasticity	6	Arc, Baiap2, Grin2b, Gsk3b, Kalrn, Neurod2
Hippocampus development	5	Epha5, Fgf13, Grin2b, Gsk3b, Mas1
Neuron migration	5	Arx, Cxcr4, Fgf13, Gsk3b, Sema3c
Axon guidance	3	Arx, Epha5, Sema3c
Neuron projection development	3	Baiap2, Gsk3b, Stx3
Dendrite development	3	Baiap2, Gsk3b, Kalrn
Oligodendrocyte differentiation, oligodendrocyte development, and myelination	3	Afdn, Cxcr4, Hes5
Neuron development	3	Dgkg, Epha5, Gabrb1
Neurogenesis	2	Cxcr4, Hes5, Snw1, Sorl1
Radial glial cell differentiation	2	Afdn, Hes5
Apoptotic process	2	Fgf13, Gsk3b
Axonogenesis	2	Gsk3b, Kalrn
Neuronal stem cell population maintenance	1	Hes5
Establishment of neuroblast polarity	1	Fgf13
Neural precursor cell proliferation	1	Cxcr4
Neuron maturation	1	Gsk3b
Cerebral cortex GABAergic interneuron migration	1	Arx
Glial cell differentiation	1	Fgf2
Astrocyte differentiation	1	Hes5
Ephrin signaling pathway	1	Epha5

Abbreviations: Acsl4, acyl‐CoA synthetase long‐chain family member 4; Afdn, afadin, adherens junction formation factor; Arc, activity‐regulated cytoskeleton‐associated protein; Arx, aristaless‐related homeobox; Baiap2, BAR/IMD domain containing adaptor protein 2; Cxcr4, C‐X‐C motif chemokine receptor 4; Dgkg, diacylglycerol kinase, gamma; Epha5, EPH receptor A5; Fgf2, fibroblast growth factor 2; Fgf13, fibroblast growth factor 13; Gabrb1, gamma‐aminobutyric acid type A receptor subunit beta1; Gapdh, glyceraldehyde‐3‐phosphate dehydrogenase; GLY, glycidol; Grin2b, glutamate ionotropic receptor NMDA type subunit 2B; Gsk3b, glycogen synthase kinase 3 beta; Hes5, hes family bHLH transcription factor 5; Hprt1, hypoxanthine phosphoribosyltransferase 1; Kalrn, kalirin, RhoGEF kinase; Mas1, MAS1 proto‐oncogene, G protein‐coupled receptor; Neurod2, neuronal differentiation 2; PND, postnatal day; Ptpru, protein tyrosine phosphatase, receptor type, U; Sema3c, semaphorin 3C; Snw1, SNW domain containing 1; Sorl1, sortilin‐related receptor 1; Stx3, syntaxin 3; Tcf4, transcription factor 4.

DISCUSSION

We have previously found that maternal exposure to PTU, an antithyroid agent that causes hypothyroidism (Shibutani et al., 2009), had diverse effects on hippocampal neurogenesis including aberrations in granule cell lineages, GABAergic interneurons, and synaptic plasticity in rat offspring and that the disruption was mostly sustained through the adult stage on PND 77 (Table 1; Shiraki et al., 2016). However, VPA only affected interneuron subpopulations at the end of maternal exposure and typically showed a late effect on granule cell lineages involving synaptic plasticity at the adult stage (Table 1; Watanabe et al., 2017). GLY had a reversible effect on granule cell lineages at the end of maternal exposure and a sustained effect on interneuron subpopulations (Table 1; Akane, Shiraki et al., 2013). However, in nature, the effect on interneurons may be transient and disappear later on, because the change in granule cell lineages disappeared along with a compensatory increase in synaptic plasticity at the adult stage (Table 1; Akane, Shiraki et al., 2013, Akane et al., 2014). In the present study, validation analysis of transcript levels of selected genes confirmed the downregulation of many genes on PND 21 induced by exposure to PTU or GLY, reflecting the diverse effects on neurogenesis by these compounds. Furthermore, genes showing sustained downregulation were found with PTU or VPA exposure, reflecting sustained or late effect on neurogenesis by these compounds. VPA is a HDAC inhibitor, which may cause suppression of DNA methylation, because HDAC activation and DNA methylation usually cooperate in transcriptional gene silencing (Sasidharan Nair et al., 2020). This may be the reason for the lower number of validated genes showing transcript downregulation on PND 21 with VPA compared with PTU and GLY. In contrast, GLY did not induce sustained downregulation, probably because of the reversible nature of its effect on neurogenesis.

In the present study, PTU downregulated Creb1 on both PND 21 and PND 77. The translated product, cAMP‐responsive element‐binding protein (CREB), is a transcription factor in the nervous system that is involved in diverse processes such as neurodevelopment, synaptic plasticity, and neuroprotection (Sakamoto et al., 2011). The irreversible reduction in CREB protein was observed in the dentate gyrus, cornu Ammonis (CA) 1, and CA3 of the hippocampus in rats after maternal PTU exposure (Dong et al., 2009). Reduction in the expression of CREB or in the active phosphorylated CREB was also observed in the hippocampus in rat offspring after maternal thyroidectomy (Zhang et al., 2015). CREB is closely related in structure and function to cAMP‐response element modulator (CREM), and mice with a Crem −/− background and lacking CREB in the brain during prenatal life show extensive apoptosis of postmitotic neurons (Mantamadiotis et al., 2002), as well as delayed migration of neural and glial progenitors (Díaz‐Ruiz et al., 2008). By contrast, mice in which both Creb1 and Crem are disrupted in the postnatal forebrain show progressive neuronal loss (Mantamadiotis et al., 2002). Maternal PTU exposure in our previous study induced reductions in NSCs and neural progenitor cells by apoptosis at the end of exposure and sustained reduction in early progenitor cells later on (Table 1; Shiraki et al., 2016). However, the number of postmitotic immature granule cells was unaffected. Moreover, developmental hypothyroidism, such as that induced by PTU exposure, causes neuronal mismigration in the hippocampus (Shibutani et al., 2009), suggesting involvement of CREB/CREM downregulation in the neuronal mismigration. Although data are not included in the list of selected genes, we found PTU exposure caused promoter region hypermethylation and downregulation of transcript level of Crem by analysis of Methyl‐Seq and RNA‐Seq, respectively.

CREB also plays a role in synaptic plasticity and acts as a transcription factor of immediate‐early genes (IEGs) related to synaptic plasticity (Kaldun & Sprecher, 2019). Arc is one of these target IEGs; we observed a decrease in activity‐regulated cytoskeleton‐associated protein (ARC)+ granule cells at the end of maternal PTU exposure on PND 21, suggestive of suppressed synaptic plasticity (Table 1; Shiraki et al., 2016). In the present study, Arc was hypermethylated on PND 21 and downregulated on both PND 21 and PND 77 due to maternal PTU exposure. These results suggest the involvement of multiple mechanisms in the transcriptional suppression of Arc.

In the current study, PTU downregulated Hes5 on both PND 21 and PND 77. The translated product, hes family bHLH transcription factor 5 (HES5), is expressed in NSCs to suppress the transcription and function of activator‐type bHLH factor as an effector of the Notch signal, and thus maintains NSCs and early progenitor cells (Lugert et al. 2010). Notch activation‐mediated promotion of hippocampal neurogenesis involves the activation of downstream molecules, such as CREB (Baik et al., 2020). As aforementioned, PTU in this study caused sustained Creb1 downregulation. Therefore, it is reasonable to consider that PTU‐mediated developmental hypothyroidism causes suppression of Notch signaling through downregulation of HES5 and CREB1, resulting in decreased numbers of NSCs and early progenitor cells on PND 21 and a sustained decrease in early progenitor cells on PND 77 as observed in our previous study (Table 1; Shiraki et al., 2016).

Herein, PTU downregulated Tfap2c on PND 21. Transcription factor AP‐2 gamma (TFAP2C) is expressed in type‐2b and ‐3 neural progenitor cells in the hippocampal dentate gyrus and plays a role in promoting proliferation and differentiation to produce postmitotic granule cells (Mateus‐Pinheiro et al., 2017). Maternal PTU exposure in our previous study caused reductions in type‐1 NSCs and type‐2a and type‐3 progenitor cells in the SGZ on PND 21 (Table 1; Shiraki et al., 2016), suggesting that Tfap2c downregulation reflects the reduction in type‐3 progenitor cells at this time point.

In the present study, PTU also downregulated Robo3 on PND 21, which is consistent with our previous expression microarray data (Shiraki et al., 2014). Roundabout guidance receptor 3 (ROBO3) and its ligand, neural EGFL‐like 2, regulate axon guidance, and neuronal migration in the developing brain (Pak et al., 2020). Furthermore, PTU downregulated Sema3c on both PND 21 and PND 77 in this study, and this result was consistent with our previous expression microarray data on PND 21 (Shiraki et al., 2014). Semaphorins and their receptors, neuropilins, are expressed in the developing hippocampus and are associated with axon tract and synapse formation through cytoskeleton reorganization of the axonal growth cone (Gil and Del Río, 2019). In the developing brain, semaphorin 3C (SEMA3C) was found to be a critical component required for radial migration of developing neurons (Wiegreffe et al., 2015). These results suggest an involvement of ROBO3 or SEMA3C downregulation in the neuronal mismigration and axon guidance induced by developmental hypothyroidism, as well as the involvement of CREB/CREM downregulation as aforementioned.

In our study, PTU downregulated Epha7 on both PND 21 and PND 77 and Epha5 on PND 21. The promoter region hypermethylation and downregulation of transcript level have already been reported for Epha5 in rats with developmental hypothyroidism (Wu et al., 2015). In our previous microarray analysis, maternal PTU exposure caused downregulation of Epha4, Epha6, Epha8, Ephb2, and Ephb6 in the hippocampal dentate gyrus (Shiraki et al., 2014), as well as a decrease in Eph receptor A4 (EPHA4)+ cells in the hilus of the dentate gyrus (Table 1; Shiraki et al., 2016). Epha receptors and ephrin‐A ligands play important roles in neural development and synaptic plasticity, and among them, EPHA7 gene mutations have been reported to be linked to developmental neurological delays (Wurzman et al., 2015). EPHA7 is required for connectivity stabilization of parvalbumin (PVALB)‐expressing GABAergic interneurons with granule cells of the dentate gyrus (Beuter et al., 2016). We have previously found that maternal PTU exposure caused a sustained decrease in PVALB+ interneurons in the hilus of the dentate gyrus (Table 1; Shiraki et al., 2016), suggesting a causal relationship of sustained Epha7 downregulation with the decrease in PVALB+ interneurons. With regard to EPHA5, binding of ephrin‐A5 to this receptor activates the CREB signal transduction pathway to induce spine maturation and filopodia formation in synaptogenesis (Akaneya et al., 2010). As aforementioned, the present study also revealed the downregulation of Creb1 and Epha5 by PTU on PND 21, suggesting suppression of the signaling cascade in synaptic plasticity, as evident by the decrease in ARC+ granule cells at this time point (Table 1; Shiraki et al., 2016).

PTU also downregulated Pvalb on both PND 21 and PND 77, in accordance with the decrease in PVALB+ interneurons as aforementioned (Table 1; Shiraki et al., 2016). We have also previously reported promoter region hypermethylation and downregulation of Pvalb, accompanying reduction in PVALB+ interneurons in the dentate gyrus upon maternal manganese exposure in mice (Wang et al., 2013). GABAergic interneurons that promote differentiation of intermediate progenitor populations are considered to be basket cells or axo‐axonic cells (Tozuka et al., 2005). Considering that subpopulations of basket cells or axo‐axonic cells express PVALB (Freund and Buzsáki, 1996), reduction in PVALB+ interneurons by maternal PTU exposure may also be caused by promoter region hypermethylation, leading to the reduction in type‐3 progenitor cells (Table 1; Shiraki et al., 2016).

In the present study, PTU downregulated Reln on both PND 21 and PND 77. However, this result was in contrast to the Reln upregulation and increase in reelin (RELN)+ interneurons induced by maternal PTU exposure in our previous study (Table 1; Shiraki et al., 2016). Although the reason for this discrepancy is not clear, there is a report that showed promoter region hypermethylation and downregulation of transcript level of Reln at early PNDs in pups exposed to PTU from GD 15 to weaning; however, this change gradually disappeared after PND 15 (Sui & Li, 2010). After weaning, the increase in RELN+ interneurons in our previous study is thought to be a compensatory response against neuronal mismigration.

Herein, on PND 21 PTU downregulated Hes5, Cntn2, Id2, and Shh, which are associated with oligodendrocyte development and differentiation, and myelination, as well as Hes5, Cntn2, Id2, and Ror2, which are associated with astrocyte development and differentiation, and finally Reln, Fgf2, and Igf, which are associated with glial cell differentiation. We have previously reported that developmental hypothyroidism causes increases in immature astrocytes immunoreactive for vimentin and glial fibrillary acidic protein in the cerebral white matter, as well as decreases in oligodendrocytes immunoreactive for 2',3'‐cyclic nucleotide 3'‐phosphodiesterase or oligodendrocyte lineage transcription factor 2 in several brain regions (Shibutani et al., 2009; Shiraki et al., 2014). The differentiation of oligodendrocytes is strongly thyroid hormone‐dependent (Rodríguez‐Peña, 1999). Delivery of sonic hedgehog signaling molecule (SHH) into demyelinated lesions in rats following spinal cord injury increased oligodendrocyte precursors and neurons (Bambakidis et al., 2003), suggesting that a decrease in oligodendrocytes due to developmental hypothyroidism may be mediated by SHH downregulation.

The VPA‐induced autism model, as well as other autism models, has shown alterations in GABAergic signals, as evident by reductions in glutamate decarboxylase 67 (GAD67)+ and PVALB+ interneuron subpopulations in the cerebral cortex (Cellot and Cherubini, 2014). We have also previously found decreases in RELN+, PVALB+, or GAD67+ interneurons on PND 21 induced by maternal VPA exposure (Table 1; Watanabe et al., 2017). In the present study, we found DNA hypermethylation and downregulation in the transcript level of Pvalb at the same time point, suggesting the involvement of epigenetic gene regulation in disruptive regulation of GABAergic signals during hippocampal neurogenesis.

VPA also downregulated Vgf and Dpysl4 on PND 21 and PND 77. TLQP‐62, a VGF‐derived peptide, has been shown to enhance dendritic maturation in the rat hippocampus (Behnke et al., 2017) and induce neuritogenesis in human SH‐SY5Y neuroblastoma‐derived cell line (Moutinho et al., 2020). Dihydropyrimidinase‐like 4 is highly expressed in the hippocampus from early postnatal development, playing critical roles in both axonal and dendritic morphogenesis of dentate granule cells (Quach et al., 2018). Therefore, Vgf and Dpysl4 downregulation may suppress synaptic plasticity in the hippocampal dentate gyrus. However, we have previously revealed that VPA exposure increased ARC or cyclooxygenase 2‐mediated synaptic plasticity as a late effect at the adult stage (Table 1; Watanabe et al., 2017). Although we did not examine microRNA expression, microRNA‐mediated homeostatic synaptic plasticity can be induced as a compensatory response to alterations in neuronal activity (Hou et al., 2015).

We have previously reported that maternal GLY exposure decreased immature granule cells in the dentate gyrus and increased GABAergic interneuron subpopulations as represented by immature RELN+ interneurons and calretinin+ interneurons in the hilus on PND 21 (Table 1; Akane, Shiraki et al., 2013). The increases in immature RELN+ interneurons remained until PND 77, although the change in granule cell lineage disappeared (Akane, Shiraki et al., 2013). We have also found that GLY causes axon injury in the central and peripheral nervous systems of adult rats (Akane, Shiraki et al., 2013). In accordance with our previous results, GLY exposure downregulated the transcript levels of genes involved in neuronal migration (Arx, Cxcr4, Fgf13, Gsk3b, and Sema3c), neuron projection development (Baiap2, Gsk3b, and Stx3), dendrite development (Baiap2, Gsk3b, and Kalrn), and axonogenesis (Gsk3b and Kalrn) on PND 21. However, the expression of all the downregulated genes recovered on PND 77. These gene expression changes may support suppressed differentiation and mismigration of late‐stage hippocampal neurogenesis at the end of GLY exposure, targeting the newly generated nerve terminals of immature granule cells (Akane, Shiraki et al., 2013).

Herein, GLY downregulated the transcript levels of genes involved in synaptic plasticity (Arc, Baiap2, Grin2b, Gsl3b, Kalrn, and Neurod2) on PND 21, probably reflecting toxicity to newly generating nerve terminals of immature granule cells (Table 1; Akane, Shiraki et al., 2013). On PND 77 after maternal GLY exposure, we have previously found an increase in ARC+ and FBJ osteosarcoma oncogene (FOS)+ hippocampal granule cells (Table 1; Akane et al., 2014). As well as ARC, FOS is an IEG protein regulating synaptic plasticity (Minatohara et al. 2016), and these late responses may be the result of amelioration from disruption of late‐stage differentiation in hippocampal neurogenesis.

In the present study, 10 genes showed downregulation in the transcript level on PND 21 in common with both PTU and GLY. Among them, five genes showed sustained downregulation through PND 77 only in the PTU group. However, we could not find any genes that showed sustained downregulation through PND 77 in the GLY group. There may be two possibilities on the difference in the response of mRNA expression on PND 77. As one possibility, GLY‐exposed animals allowed active demethylation of methylated cytosine in hypermethylated genes. Another possibility is that the methylation pattern of promoter region is different between the PTU and GLY groups. In the present study, we selected genes that showed CpG site hypermethylation with the ratio of ≥0.2 as compared with the untreated controls on PND 21. Among the five genes showing sustained downregulation by PTU, Fgf13 and Mas1 differed the methylated CpG sites between PTU and GLY (Tables S6 and S8), and this difference may cause different responses in transcription on PND 77 between PTU and GLY. In contrast, three other genes, Arc, Hes5, and Sema3C, showed methylation of identical CpG sites with both PTU and GLY (Tables S6 and S8). However, methylation ratios of the corresponding and nearby CpG sites in the promoter region of these genes were higher with PTU as compared with GLY (Tables S6 and S8). It may be possible that the relatively high methylation ratios of the promoter region cause sustained suppression of transcription of these genes through PND 77. Further studies addressing the stability of promoter region hypermethylation of the identified genes, as well as the relationship with gene expression changes, may be necessary for establishing irreversible markers of DNT.

In the present study, maternal food and water consumption were decreased by exposure to PTU, VPA, or GLY. Offspring of these exposure groups, especially of PTU and GLY, decreased the body weight until PND 77. These results suggest treatment‐derived undernutrition and stress on both dams and offspring. Nutrients involved in one‐carbon metabolism, such as methionine, choline, and folic acid, work as methyl donors for maintaining gDNA methylation (Friso and Choi, 2002). Experimentally, prenatal and/or postnatal nutritional deficiency, as well as prenatal or early‐life postnatal stress, has shown to cause alterations in the methylation status of both genome‐wide and specific gene levels in several brain regions in mice and rats (Weng et al., 2014; Xu et al., 2014; Blaze & Roth, 2015). Among the genes identified as hypermethylated ones in the present study, Reln has previously shown to be examined the methylation status in brain regions. In one study, early‐life postnatal stress caused Reln hypomethylation in the hippocampus of rats (Wang et al., 2018). In contrast, prenatal dietary methyl donor deficiency did not alter the methylation level of Reln, in contrast to the hypermethylation of Nnat encoding neuronatin, in the hippocampus of rats (Konycheva et al., 2011). On the other hand, prenatal stress caused hypermethylation of Reln, as well as Gad1 encoding GAD67 and Bdnf encoding brain‐derived neurotrophic factor, in the frontal cortex by prenatal stress in mice (Dong et al., 2016). These results may suggest possible modifying effects in the methylation status of genes identified as hypermethylated ones in exposure groups especially of PTU and GLY in the present study.

In conclusion, PTU and GLY confirmed the downregulation of many genes on PND 21, reflecting their diverse effects on neurogenesis. Furthermore, genes showing sustained downregulation were found upon PTU and VPA exposure, reflecting sustained or late effect on neurogenesis by these compounds. In contrast, these effects were not observed with GLY, probably because of the reversible nature of the effects. Among the genes showing sustained downregulation, Creb, Arc, and Hes5 were concurrently downregulated by PTU, suggestive of an association with neuronal mismigration, suppressed synaptic plasticity, and reduced neural stem and progenitor cells. Epha7 and Pvalb were also concurrently downregulated by PTU, suggestive of an association with the reduction in late‐stage progenitor cells. VPA induced sustained downregulation of Vgf and Dpysl4, which may be related to aberrations in synaptic plasticity. The genes that showed sustained downregulation, that is, Arc, Bmp3, Cebpb, Creb1, Epha7, Fgf13, Fzd9, Hes5, Mas1, Pvalb, Reln, Sema3c, Dpysl4, Sox2, and Vgf, may be markers of irreversible DNT.

CONFLICT OF INTEREST

All authors declare that there are no conflicts of interest that influenced the outcome of the present study.

Data S1 Supporting Information

Click here for additional data file.