Introduction
To substantiate the Developmental Origins of Health and Disease (DOHaD) theory, which states that the prenatal environment influences future health and disease risk, we reported low birth weight with prenatal undernutrition. Reference Gluckman, Hanson and Buklijas1 In the DOHaD theory, it is believed that the risk of noncommunicable diseases increases when there is a mismatch between the predisposition acquired during the prenatal period and the postnatal environment. However, postnatal environmental factors, in addition to prenatal predisposition, trigger abnormal stress responses. Our previous results suggested that some of our fetal undernourished-induced low-birth-weight (LBW) model rats failed to catch up after growth and showed a small body size even after growth. Reference Nemoto and Kakinuma2 Moreover, the model rats were vulnerable to stress, and blood corticosterone levels in LBW rats were maintained at high levels after stress exposure, regardless of their catch-up growth levels. Reference Nemoto, Kakinuma and Shibasaki3 Downregulation of the corticotropin-releasing factor (CRF) receptor is caused by the negative feedback regulation of glucocorticoids. Reference Sakai, Horiba, Sakai, Tozawa, Demura and Suda4,Reference Iredale and Duman5 We have identified miR-449a as a factor that downregulates the CRF receptor. Reference Nemoto, Mano and Shibasaki6 Our results showed that miR-449a had a sequence that could bind to the 3’-untranslated region (3’-UTR) of the CRF receptor and was induced by glucocorticoids to reduce the CRF receptor mRNA stability and protein translation. Furthermore, we demonstrated that miR-449a overexpression reduced CRF receptor expression levels and that miR-449a silencing partially reduced the dexamethasone-induced downregulation of the CRF receptor expression. Reference Nemoto, Mano and Shibasaki6 Our previous results also demonstrated that the expression of Gas5 long noncoding RNA (lncRNA) was upregulated in a pituitary-specific manner, with elevated expression of Gas5 lncRNA inhibiting glucocorticoid receptor binding to DNA and glucocorticoid-induced gene expressions in the model rats. Reference Nemoto, Kakinuma and Shibasaki3,Reference Nemoto and Kakinuma7 Gas5 is an lncRNA with a sequence homologous to the glucocorticoid-responsive element (GRE) that competitively inhibits the binding of the glucocorticoid receptor (GR) to the GRE. Reference Kino, Hurt, Ichijo, Nader and Chrousos8 These suggest the possibility that after restraint stress exposure, there is a failed induction of miR-449a expression. However, the mechanism by which the expression level of Gas5 lncRNA is upregulated in the pituitary gland of model rats remains unclear. Gas5 lncRNA interacts with various microRNAs (miRNAs). Reference Guo, Song, Sun, Jin and Dai9–Reference Cheng, Zhao, Wang, Wang and Zhu12 Therefore, we measured the expression of these microRNAs in the pituitary gland of the model rats and control rats. Because Gas5 lncRNA expression increases by starvation, Reference Tani, Torimura and Akimitsu13 we simultaneously quantified Gas5 lncRNA and miRNAs, which have been previously reported to interact with Gas5 lncRNA using the mouse adrenocorticotrpic hormone-producing neuroendocrine tumor (ACTHoma) cell line. We also evaluated the primary cultured rats’ anterior pituitary cells after starvation. Subsequently, we investigated whether the suppression of miRNA expression would increase the expression of Gas5 lncRNA.
Several diseases and predispositions are transgenerationally inherited by the next generations without any DNA mutations, many of which are now thought to be epigenome-mediated. When adults are exposed to a stimulus, their germline, as well as the germline of the fetus in pregnant females, is exposed to the same stimuli. Reference Miska and Ferguson-Smith14,Reference Heard and Martienssen15 Therefore, we examined not only offspring exposed to undernutrition during the fetal period but also their offspring and reported that birth weight in our model rats was affected until at least the F4 generation. Reference Nemoto and Kakinuma2 However, the transgenerational effects of impaired-glucocorticoid negative feedback in the pituitary gland remain unclear. Thus, we investigated the transgenerational effects of the impaired-glucocorticoid negative feedback mediated by miRNAs, which we identified above, using offspring obtained from mating model rats.
Finally, we have also previously reported that a methyl modulator intervention during the third trimester of pregnancy or during the lactation period restored glucocorticoid responses in fetal undernourished-induced LBW model rats. Reference Nemoto and Kakinuma7 Alterations in gene expression patterns governed by epigenetics can lead to various noncommunicable chronic diseases. Epigenetic mechanisms, including DNA methylation, histone modifications, and miRNAs, can produce heritable phenotypic changes without DNA sequence changes. Reference Zhang, Lu and Chang16 As S-adenosylmethionine, a metabolite of methionine, is a methyl donor for the methylation of DNA and histone proteins, folate-methionine metabolism is widely considered to be important in epigenetic regulation. Reference Mentch and Locasale17 In the present study, we investigated whether a methyl modulator intervention in lactating maternal rats could restore impaired-stress responses in the subsequent generations. After experimenting with the timing of several interventions as described in previous reports, we found that interventions during the third trimester of pregnancy and 1 week immediately after birth could restore negative feedback regulation of glucocorticoids. Reference Nemoto and Kakinuma7 In the present study, due to the number of rats housing and the breeding schedule, we investigated intervention only for one week immediately after birth. To examine the effects of the methyl modulator, we measured systemic corticosterone levels and miR-449a expression levels, which have been previously shown to increase with stress, and proopiomelanocortin (POMC) and CRF receptor expressions in the pituitary. Reference Nemoto, Kakinuma and Shibasaki3,Reference Nemoto and Kakinuma7 We also measured the expression level of Fkbp5 mRNA. Reference Nemoto, Kakinuma and Shibasaki3,Reference Nemoto and Kakinuma7 Fkbp5 is a co-chaperone factor of GR, GR activity directly stimulates Fkbp5 gene transcription, and then Fkbp5 inhibits GR activity. Reference Kageyama, Iwasaki, Watanuki, Niioka and Daimon18 Thus, GR regulation is influenced by Fkbp5 through an ultrashort negative feedback loop, and the expression level of Fkbp5 indicates glucocorticoid action.
Through these experiments, we report here that; 1. the identification of miR-23 as a suppression factor of Gas5 lncRNA expression in the pituitary of LBW; 2. the findings of decreased expression of miR-23 and increased expression of Gas5 lncRNA were transgenerational inherited in the F2 and F3 generations; and 3. methyl modulator intervention in lactating F0 dams restored the expression levels of miR-23 and Gas5 lncRNA, and corticosterone levels after restraint stress to NBW levels in F2 and F3 generations.
Materials and methods
Model rats
Wistar rats were maintained at 23 ± 2°C with a 12:12-h light-dark cycle (lights on at 0800 h, off at 2000 h). The animals had ad libitum access to laboratory chow and sterile water. All experimental procedures were reviewed and approved by the Laboratory Animals Ethics Review Committee of the Nippon Medical School (#27-067 and #2020-003). All experiments were performed in accordance with the relevant guidelines and regulations. Reference Nemoto, Mano and Shibasaki19 We previously generated fetal low-carbohydrate and calorie-restricted rats. Reference Nemoto and Kakinuma2 Briefly, 20 female rats (age, 9 weeks) were mated with normal male rats. The dams were individually housed with free access to water and were divided into two groups: low-carbohydrate and calorie-restricted diet dams, in which the calorie intake was restricted to 60% of the control group (D08021202, Research Diet Inc., New Brunswick, NJ) during the entire gestational period (Fig. S1). The control dams had free access to food during the same period. In total, 12–20 pups were obtained from 10 dams in each group. We excluded rat pups born with a body weight of > 6.0 g, which is the average – 2 standard deviation body weight of the offspring of normal dams. No surrogate mothers were used, and the 10 rat pups were randomly selected and raised under the birth mother rat. After weaning, the rats of the different litters were mixed. We previously created a methyl modulator diet (D15090803, Research Diet) that was reported in a previous study Reference Cordero, Gomez-Uriz, Campion, Milagro and Martinez20 with the addition of zinc, which is required for the metabolism of homocysteine to methionine (Figure S1). We fed lactating mother rats a methyl modulator diet for 1 week immediately after birth. Fathers and mothers of the next generation (F2) offspring were randomly selected from male and female rats, and mated with 10 pairs of NBW, 10 pairs of LBW, and 10 pairs of LBW + Methyl, respectively. After mating, mother rats including LBW were fed a standard chow ad libitum, and they were also fed a standard chow diet ad libitum during the lactating period. As with the F1 generation, the number of pups was adjusted to 10 per dam, and weaned at 21 days. Ten male and female rats were randomly selected from littermates, and were used as the parents of the F3 generation. F3 generation offspring were obtained using the same methodology as F2-generation offspring (Fig. S1).
Restraint stress exposure
Six-week-old rats were wrapped in a flexible wire mesh (12 mm × 12 mm) and maintained for 120 min between 0900 h and 1200h in an isolated room. Reference Nemoto, Iwasaki-Sekino, Yamauchi and Shibasaki21 The rats were sacrificed in an adjacent room immediately after restraint, and their trunk blood and anterior pituitaries were collected. Non-stressed control rats were housed in a separate rooms, sacrificed in the same manner, and subjected to identical procedures.
Cell culture
Mouse adrenocorticotropic hormone-producing neuroendocrine tumor (ACTHoma) AtT-20 cells were obtained from Japanese Collection of Research Bioresources Cell Bank. The cells were maintained in F-10 minimum essential medium supplemented with 15% donor horse serum and 2.5% fetus bovine serum in a humidified 5%CO2–95% air at 37°C. On the day of each experiment, the culture media were changed to HBSS (#14060040, Thermo Fisher Scientific) and incubated for 4–24 hrs. After incubation, the cells were assayed for miRNA and Gas5 lncRNA expression.
Primary culture of pituitary cells
Six-week-old male rats (n = 30) were sacrificed by decapitation, and each respective pituitary was removed under sterile conditions. All anterior pituitaries were collected, pooled together, and then mechano-enzymatically dispersed as previously described. Reference Nemoto, Iwasaki-Sekino, Yamauchi and Shibasaki22 Briefly, the tissues were washed twice in phosphate-buffered saline (PBS) and then incubated at room temperature in the PBS containing 0.01% dispase (Godoshusei, Tokyo, Japan) with constant stirring for 30 min. After three washes with PBS, the cells were plated in 24-well plates and cultured in DMEM/F10 HAM culture medium (Sigma-Aldrich Co., St Louis, MO) supplemented with 10% FBS and an antibiotics/antimycotic solution (Thermo Fisher Scientific, Carlsbad, CA). The cells were subsequently allowed to attach to the plating surfaces for 4 days in a humidified 5% CO2–95% air incubator set at 37°C. On the day of each experiment, the culture media were changed to HBSS (#14060040, Thermo Fisher Scientific) and incubated for 1 h. After incubation, the cells were assayed for miRNA and Gas5 lncRNA expression.
miRNA knockdown
mirVana miRNA inhibitor system for miR-23 was purchased from Thermo Fisher Scientific. AtT-20 cells were transfected with miR-23 inhibitor using Lipofectamine RNAiMAX (Thermo Fisher Scientific), according to the manufacturer’s instructions. After 48 h, culture cells were harvested for RNA extraction.
RNA extraction and real-time RT-PCR
We quantified the mRNA and miRNA levels, as previously reported. Reference Nemoto, Kakinuma and Shibasaki3 Total RNA was extracted from pituitaries using RNAiso Plus (Takara, Shiga, Japan). The absorbance of each sample was assayed at 260 nm and 280 nm, and determined 260/280 nm ratio (the 260 /280 nm ratio of all samples used in this study was > 1.7). For miRNA (miR-23, miR-103, miR-182, miR-222 and miR-449a) expression analyses, first-strand cDNA was synthesized at 37°C for 1 h using 500 ng of denatured total RNA and then terminated at 85°C for 5 min using a Mir-X® miRNA First-Strand Synthesis and SYBR® qRT-PCR kit (Clontech Laboratories Inc., Mountain View, CA). For mRNA expression analyses, first-strand cDNA was generated using 250 ng of denatured total RNA; the reaction mixture was incubated at 37°C for 15 min, 84°C for 5 s, and 4°C for 5 min using a PrimeScript® RT reagent kit with gDNA Eraser (Takara). PCR was performed by denaturation at 94°C for 5 sec with annealing-extension at 60°C for 30 s for 40 cycles using SYBR premix Ex Taq (Takara) and specific primer sets for rat Crhr1 (RA025834, Takara), POMC (RA013901, Takara), Fkbp5 (RA049269, Takara), GAPDH (RA015380, Takara) or Gas5 lncRNA. Reference Nemoto, Kakinuma and Shibasaki3 To normalize each sample for RNA content of each sample, GAPDH and a housekeeping gene were used for mRNA expression analysis, respectively. The second derivative method was used as the standard to calculate Ct values. Reference Nolan, Hands and Bustin23
Measurement of blood corticosterone levels
Corticosterone levels were measured in the blood serum from decapitated rats and were measured using a rat corticosterone ELISA kit (#501320, Cayman Chemical).
Statistical analysis
Unpaired t tests, one-way analysis of variance (ANOVA) followed by Tukey’s multiple comparisons test, two-way ANOVA followed by Tukey’s multiple comparisons test, or Pearson’s correlation test were used for each statistical analysis. Prism 6.0 software (GraphPad Software, Inc., La Jolla, CA) was used for all calculations. Real-time RT-PCR results are expressed as percent ± standard error of the means, with the control set to 100. P < 0.05 was considered statistically significant.
Results
Expression of miR-23 is downregulated in the model rats and negatively correlated with Gas5 long noncoding RNA (lncRNA) in the pituitary gland of model rats
Although evaluations of miRNA expression have reported finding interactions with Gas5 lncRNA in the pituitary gland of model rats (LBW) and control rats (normal birth-weight [NBW]) rats, there was only a significant decrease in the expression level of miR-23 in the LBW rats (Fig. 1a). However, no statistical difference was observed in the expression of miR-103 (Fig. 1b), miR-182 (Fig. 1c), and miR-222 (Fig. 1d) between NBW and LBW. miR-23 expresssion consistently showed a significant negative correlation with the expression level of Gas5 lncRNA (Fig 1a).
Figure 1. Expression of miR-23, miR-103, miR-182 and miR-222 in the pituitary of LBW and NBW rats. The expression levels of miR-23 (A), miR-103 (B), miR-182 (C), and miR-222 (D) in the pituitary of control rats (NBW) and LBW rats were quantified. The miRNA expression level is a ratio obtained by correcting the ΔΔCT values of miRNAs with the ΔΔCT values of U6 and setting the value of NBW to 100. Values plotted are mean ± S.E.M. (n = 5). Statistical analyses were performed using and unpaired t-test and Pearson’s correlation test. NBW, normal birthweight; LBW, low birthweight.
miR-23 expression is induced by starvation and suppression of miR-23 expression increases the expression of Gas5 lncRNA in the pituitary cells
Starved pituitary adrenocorticotropin (ACTH)-producing tumor AtT-20 cells significantly decreased miR-23 expression and increased Gas5 lncRNA expression (Fig. 2a). In addition, the transfection of the miR-23 inhibitor at a dose of 10 nM into AtT-20 cells significantly increased Gas5 lncRNA expression (Fig. 2b). Similar results were obtained in Hanks’ balanced salt solution (HBSS)-starved rat primary pituitary cells (Fig. 2c).
Figure 2. Suppression of Gas5 lncRNA expression by miR-23 in mouse ACTHoma cells and primary cultured pituitary cells. The expression levels of miR-23 and Gas5 lncRNA in the starvation-exposed AtT-20 mouse ACTHoma (A) or primary rat pituitary (C) cells were quantified (n = 6). Gas5 lncRNA expression in miR-23-inhibited AtT-20 cells was quantified (B) (n = 3). GAPDH or U6 snRNA was used to normalized the RNA content in each sample. Values plotted are mean ± S.E.M. Statistical analysis was performed using unpaired t test. *P < 0.05 vs 0h, **P < 0.01 vs 0h, †P < 0.05 vs growth medium, ##P < 0.01 vs Neg cont (negative control).
Impaired-negative feedback regulation of glucocorticoids in the pituitary gland due to fetal undernutrition is the transgenerational effect
miR-23 expression was significantly lower and Gas5 lncRNA expression was significantly higher in the F2-generation LBW rats compared with the NBW rats. Moreover, a methyl modulator diet in the lactating F0 maternal rats restored the LBW-induced lower miR-23 expression and higher Gas5 lncRNA expression to NBW levels (Fig. 3a, b). Surprisingly, the diet intervention in lactating maternal rats of the F0 generation led to restoration in the pituitary gland of the F3-generation LBW rats with the levels reaching the same levels as those found for NBW rats (Fig. 4a, b).
Figure 3. Gene expression and blood corticosterone levels in restraint stress-exposed F2-generation rats with methyl modulator nutritional intervention. The expression levels of miR-23 (A) and Gas5 lncRNA (B) in the pituitary of control (NBW) rats, non-intervened LBW rats or offspring from the 1-week intervention in the lactating F0 maternal rats immediately after delivery (LBW + methyl) rats were quantified. (C) Blood corticosterone levels in restraint stress-exposed NBW, LBW, and LBW + methyl rats (n = 5). The expression levels of POMC mRNA (D), Crhr1 mRNA (E), miR-449a (F) and Fkbp5 mRNA (G) in the pituitary of restrained stress-exposed NBW, LBW, and LBW + methyl rats were quantified. Values plotted are the mean ± S.E.M. (n = 6). Statistical analysis was performed using two-way ANOVA followed by Turkey’s post hoc test for multiple comparisons. *P < 0.05 vs non-stressed, **P < 0.01 vs non-stressed, ****P < 0.001 vs non-stressed, ††P < 0.01 vs NBW, and ††††P < 0.001 vs NBW. NBW, normal birthweight; LBW, low birthweight.
Figure 4. Gene expression and blood corticosterone levels in restraint stress-exposed F3-generation rats with methyl modulator nutritional intervention. The expression levels of miR-23 (A) and Gas5 lncRNA (B) in the pituitary of control (NBW) rats, non-intervened LBW rats or offspring from the 1-week intervention in the lactating F0 maternal rats immediately after delivery (LBW + methyl) rats were quantified. (C) Blood corticosterone levels in restraint stress-exposed NBW, LBW, and LBW + methyl rats (n = 5). The expression levels of POMC mRNA (D), Crhr1 mRNA (E), miR-449a (F) and Fkbp5 mRNA (G) in the pituitary of restrained stress-exposed NBW, LBW, and LBW + methyl rats were quantified. Values plotted are the mean ± S.E.M. (n = 6). Statistical analysis was performed using two-way ANOVA followed by Turkey’s post hoc test for multiple comparisons. *P < 0.05 vs. non-stressed, **P < 0.01 vs non-stressed, ****P < 0.001 vs non-stressed, ††P < 0.01 vs NBW, and ††††P < 0.001 vs NBW. NBW, normal birthweight; LBW, low birthweight.
Restraint stress-exposed F2- (Fig. 3c) or F3- (Fig. 4c) generation LBW rats had significantly higher blood corticosterone levels compared with NBW rats. Blood corticosterone levels in the NBW rats recovered to basal levels 120 min after the stress, whereas these levels remained significantly elevated in the F2- and F3-generation LBW rats. Similarly, restraint stress-exposed F2- (Fig. 3d) or F3- (Fig. 4d) generation LBW rats had significantly higher expression levels of POMC mRNA in the pituitary gland compared with the NBW rats. The expression levels of POMC mRNA in the pituitary gland of NBW rats recovered to basal levels 120 min after the stress, whereas these levels remained significantly elevated in the F2- and F3-generation LBW rats. In NBW rats, Crhr1 mRNA expression in the pituitary gland was decreased at 30 min after exposure to restraint stress, followed by a significant decreased at 120 min. In contrast, it was slowly downregulated in the F2- and F3-generation LBW rats, with significantly higher levels compared with the NBW rats. The methyl modulator downregulated Crhr1 mRNA expression to the same level as that in the NBW rats (Figs. 3e and 4e). There was no induction of the expression of miR-449a, whose expression level increased after stress exposure in the pituitary gland of the NBW rats, in either the F2- or the F3-generation in LBW rats (Figs. 3e and 4e). However, the methyl modulator in lactating maternal rats of the F0 generation definitely induced this expression (Figs. 3f and 4f). The expression level of Fkbp5 mRNA, which is induced by glucocorticoids, significantly increased at 60 and 120 min after exposure to restraint stress in the pituitary gland of the NBW rats (Figs. 3g and 4g). However, in the LBW rats, there was no significant increase in the expression levels until 120 min. The methyl modulators induced Fkbp5 mRNA expression levels in the pituitary gland, comparable to those observed for the NBW rats in the F2- and F3 -generation LBW rats (Figs. 3g and 4g).
Discussion
The results of the current study demonstrated the possibility that miR-23 negatively regulates Gas5 lncRNA expression, thereby leading to a negative correlation between these miRNAs in the pituitary gland of LBW rats. Furthermore, the decreased miR-23 expression affected at least the F3-generation rats, indicating that the abnormal stress response caused by malnutrition during the prenatal period is transmitted between generations. We also found that nutritional intervention in maternal rats during the lactation period improved the stress responses caused by the dysregulation of gene expression in the pituitary gland.
Epigenetics is one of the mechanisms that can influence the environment during prenatal period that persists and appears during subsequent development. Egger et al. suggested that epigenetics was a transcriptional control mechanism that occurred via acquired modifications to chromatin formed by genomic DNA wrapped around histones. Reference Egger, Liang, Aparicio and Jones24 In contrast, noncoding RNAs that do not encode proteins, especially miRNAs, which are other factors that control the expression of genomic information, are beginning to attract increasing research attention. Furthermore, miRNAs regulate epigenetic regulation. Reference Lewis, Shih, Jones-Rhoades, Bartel and Burge25–Reference Szulwach, Li and Smrt27 Although little is known about miRNA regulation and transcription, the best-characterized miRNAs originate from independent RNA polymerase II transcripts Reference Ason, Darnell and Wittbrodt28,Reference Cai, Hagedorn and Cullen29 or from introns or exons of protein-coding or nonprotein-coding genes. Reference Lagos-Quintana, Rauhut, Meyer, Borkhardt and Tuschl30 miRNAs have attracted attention as possible mediators of the epigenetic inheritance of transgenerational changes. For example, in worms, various phenotypes, such as viral immunity, nutritional status, and aging, have been affected over several generations through the induction of specific miRNAs in the parents. Reference Greer, Maures and Ucar31–Reference Rechavi, Houri-Zeʼevi and Anava33 Moreover, in mice, both maternal separations early in life and chronic fluctuating stress in late adulthood alter the levels of several miRNAs in sperm. Reference Rodgers, Morgan, Bronson, Revello and Bale34,Reference Gapp, Jawaid and Sarkies35 An injection of stress-regulated miRNAs into normally fertilized eggs recapitulates the transgenerational transmission of behavioral, hormonal, and gene expression defects in the offspring. Reference Rodgers, Morgan, Leu and Bale36 Although our results have shown that miR-23 downregulation can have a continued effect until at least the F3-generation due to fetal undernutrition, the mechanism by which miR-23 downregulation occurs remains unclear. Further studies are needed to clarify the effects of regulatory factors in miR-23 expression, which is involved in epigenetic regulation. On the other hand, miR-23 expression was negatively correlated with Gas5 lncRNA expression, and inhibition of miR-23 expression increased Gas5 lncRNA expression. This result is supported by previous reports. In fact, Zhuo, et al. reported an inverse expression association between Gas5 lncRNA and miR-23. Reference Zhou, Jiang, Lin, Li and Li37 They found a possible miR-23 binding site was located in the 3’-UTR of Gas5 lncRNA, and overexpression of miR-23 decreased the activity of the Gas5 reporter activity using the dual luciferase reporter assay. Moreover, miR-23 coprecipitated with Gas5 lncRNA, confirming the potent physical interactions between Gas5 lncRNA and miR-23. Our and their results indicate that Gas5 lncRNA has a direct relationship with miR-23.
We have previously evaluated and demonstrated that intervention with a methyl modulator diet in lactating maternal rats restores the glucocorticoid response in the pituitary of offspring (i.e., post-stress exposure blood corticosterone levels and induction of Fkbp5 mRNA expression in the pituitary gland). Reference Nemoto and Kakinuma7 In this study, we found that when maternal rats were fed a calorie-restricted diet during pregnancy and subsequently a methyl modulator diet during the lactation period, gene expressions and stress responses in the pituitary gland were restored in their offspring. Surprisingly, we found that nutritional intervention during the critical period (1 week after delivery) could normalize pituitary gene expression and stress responses in the next generation. Although we have not succeeded in analyzing the nutritional components in rat breast milk, the results suggest that at least some of the folic acid and vitamin B12 ingested by the maternal rats may have been transferred to their offspring through breast milk. Moreover, this component was able to restore the pituitary gene expression regulatory mechanisms of not only their offspring but also those of the next generations. In this study, we supplemented a methyl modulator diet based on previous reports. Reference Cordero, Gomez-Uriz, Campion, Milagro and Martinez20 The diet included folic acid, vitamin B12, betaine, choline, and zinc. The effects of folate and vitamin B12, known as promethylating vitamins, are related to epigenetic changes in the offspring owing to the accessibility of the DNA during early development. Supplementation with choline and betaine has been studied not only as a potential hepatoprotective substrate but also as an epigenetic controller. Reference Cai, Jia and Song38 Although the details of the restoration mechanism remain unknown, it has been hypothesized that S-adenosylmethionine, a metabolite of the folate-methionine cycle, affects epigenomic modification because it is known to be a methyl donor for DNA and proteins. However, the fact that our methyl modulator intervention method has an appropriate timing and span in which the intervention effect is exerted 7 and that it is inherited by the next generation suggests that methyl modulator intervention affects the epigenomic modification of a specific gene region and the mechanisms underlying the regulation and maintenance of epigenomes. According to the Dutch Hunger Winter study, individuals whose mothers were pregnant during famine had higher methylation of some genes and lower methylation of others as compared with those who were not prenatally exposed to famine. Reference Heijmans, Tobi and Stein39–Reference Tobi, Slieker and Luijk41 These methylation differences may explain the likelihood of these individuals developing certain diseases later in life. Reference Roseboom42–Reference Pidsley, Dempster, Troakes, Al-Sarraj and Mill44 To further examine these changes, our research group will be conducting more wide-ranging and comprehensive epigenomic analyses in order to elucidate the mechanisms by which the effects of the fetal undernutrition are inherited from generation to generation.
In our current study, our results showed that undernutrition during the fetal period adversely affected the negative feedback regulation of the glucocorticoids in the pituitary gland in the subsequent generations. After stress stimulates the release of CRFs in the hypothalamus and ACTH secretion in the pituitary gland, ACTH stimulates the secretion of cortisol (corticosterone) from the adrenal cortex. Cortisol (corticosterone) suppresses CRF production in the hypothalamus and ACTH production in the pituitary gland through negative feedback regulation. Chronically elevated glucocorticoid levels, such as those observed in Cushing’s syndrome, cause psychiatric symptoms including depression and attention deficit hyperactivity disorder, in addition to various metabolic symptoms and hypertension. Our results indicated that not only LBW infants but also their offspring might be susceptible to hypercortisolemia after exposure to some type of stress and that they have a higher risk of chronic noncommunicable diseases. Furthermore, animal studies have recently demonstrated that the impairment of fetal hypothalamic-pituitary-adrenal axis function caused by maternal malnutrition is associated with axis dysregulation in adults. Reference Buhl, Neschen and Yonemitsu45 Moreover, similar epigenetic variations generated in early life play a role in generating individual differences in offspring behavior. It has been hypothesized that exposure to various environmental toxins that affect epigenetic machinery can alter long-established epigenetic programs in the brain. Reference Szyf, Weaver and Meaney46 Therefore, abnormalities in stress responses affect the next generation, and these abnormalities may intervene postnatally. Further studies in this area may lead to the development of new treatments for stress-related diseases and help protect the physical and mental health of the next generation.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S2040174423000363.
Data availability statement
All data presented in this study are available within the figures that are provided in the manuscript.
Author contribution
T. Nemoto designed the work, acquired data, analyzed data, and drafted the manuscript. Y. Morita acquired and analyzed data. Y. Kakinuma interpreted the data and substantively revised the manuscript.
Funding statement
This study was supported in part by JSPS KAKENHI (T.N., grant number 17K10195 and 23K10857).
Competing interests
None.
Ethical standards
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national guides on the care and use of laboratory animals (the Laboratory Animals Ethics Review Committee of Nippon Medical School) and has been approved by the institutional committee (the Laboratory Animals Ethics Review Committee of Nippon Medical School).
