The serotonin type 1A receptor (5-HT_1AR) is considered to be an important specific therapeutic target for depression. 5-HT_1AR and serotonin type 2C receptor (5-HT_2CR) modulate reward behavior by modulating dopamine release in the NAc (Leventopoulos et al., 2009). Repeated early deprivation leads to reduce reward motivation and a decrease in 5-HT_1AR binding in the anterior cingulate cortex (ACC), CA1, and dorsal raphe nucleus (DRN) in adulthood (Leventopoulos et al., 2009). Epigenetic mechanisms of anxiety- or depression-like behavioral changes are associated with maternal 5-HT_1AR deficiency. In the offspring, the immobility behavior of the second-filial generation of male mice was particularly reduced in the forced swimming test. In addition, the behavioral changes in the first- and second-filial generation of female mice were in opposite directions. These suggest that there may be sex differences in epigenetic mechanisms resulting from maternal 5-HT_1AR deficiency (Mitchell et al., 2016). Immune system dysregulation in 5-HT_1AR ^+/− heterozygote and first-filial generation females was associated with immune activation in their offspring and the transmission of somatic anxiety trait. Non-genetic traits of complex psychiatric-like phenotypes were independently transmitted across multiple generations through parallel non-genetic mechanisms. The features of anxiety and hypoactivity were transmitted through somatic mechanisms, while the altered stress-reactivity was transmitted through gamete mechanisms (Mitchell et al., 2016). In vitro experiments further found that differentially methylated regions existed in the first-, second-, and third-filial generations of neurons, and 95% of the methylation changes occurred in CpG dinucleotides (among them, hypomethylation accounts for 55% and hypermethylation accounts for 45%), and 87% of differentially methylated sites were unidirectional (Mitchell et al., 2016).

In clinical studies, ELS induced hypermethylation of 5-HT _1A R, whereas patients with high 5-HT _1A R methylation from −340 to −149 bp upstream of the transcription start site (TSS) showed a decrease in 5-HT_1AR expression (Xu et al., 2022), which further reduced 5-HT_1AR availability (Xu et al., 2011; David and Gardier, 2016). These are consistent with the results of animal studies.

Maternal separation induces the development of adult depression and increases 5-HT _2C R pre-mRNA editing significantly (Bhansali et al., 2007). Fluoxetine administration in adolescence reduced depression-like behaviors and suppressed the increase in the phenotype of 5-HT _2C R pre-mRNA editing; however, fluoxetine administration in adult mice did not affect either depression-like behaviors or the 5-HT _2C R pre-mRNA editing phenotype (Bhansali et al., 2007). Mice exposed to ELS showed significantly increased expression of mRNA and protein-encoding the Gαq subunit of G-protein that couples to 5-HT_2A/2CRs. The aforementioned results suggest that compensatory changes in Gαq expression occur in mice with persistent changes in 5-HT _2C R pre-mRNA editing (Bhansali et al., 2007).

5-HT _2A R genotype was associated with the methylation of 5-HT _2A R at CpG-1420 and CpG-1224 in a sample of preschoolers with ELS (Parade et al., 2017). Contextual stress was positively correlated with the methylation of A homozygotes at the CpG-1420 site and negatively correlated with the methylation of G homozygotes at the CpG-1420 site. Depression-like behaviors were negatively correlated with methylation of CpG-1420 and positively correlated with methylation of CpG-1224. Collectively, environmental factors and DNA variation influence the epigenetic process of 5-HT _2A R (Parade et al., 2017).

Serotonin type 3A receptor (5-HT_3AR) is required for exercise-induced neurogenesis and antidepressant effects, and it modulates cortical interneuron migration and dendritic morphology in pyramidal neurons (Murthy et al., 2014; Kondo et al., 2015; Perroud et al., 2016). In clinical studies, early life trauma interacts with 5-HT _3A R and brain-derived neurotrophic factor (Bdnf) genes to exacerbate the risk for depression (Gatt et al., 2010a; Gatt et al., 2010b). The study found that emotional neglect in children was inversely correlated with methylation levels of CpG1_I (located in the GRE element upstream of 5-HT _3A R). In addition, a functional 5HT _3A R single nucleotide polymorphism (SNP) (rs1062613) selectively affects the methylation of a CpG located at 1 bp of the SNP (Perroud et al., 2016). However, the relationship between depression-like outcomes produced by different types of ELS and the epigenetic modification of 5HT _3A R and its mechanism still needs to be explored.

The serotonin transporter is encoded by a single gene, SERT (also known as 5-HTT or SLC6A4) (Caspi et al., 2003), located in the presynaptic 5-HT nerve terminal, axons, and cell bodies (Blakely et al., 1998). In the brain, SERT modulates the intensity and duration of serotonergic neurotransmission (Gaspar et al., 2003). Clinical study results show that methylation of the SLC6A4 promoter is associated with increased susceptibility to depression (Olsson et al., 2010), and higher SLC6A4 promoter methylation is significantly associated with childhood adversity (Kang et al., 2013). Methylation of SLC6A4 was positively associated with depression severity in women but not with depression severity and age of onset (Sanwald et al., 2021).

There is a common polymorphic region in the SERT gene, the 5-HT transporter-linked polymorphic region (5-HTTLPR), and carriers of the 5-HTTLPR short (S) variant have an increased susceptibility to depression under adversity (Uher and McGuffin, 2010). Interestingly, one study found that sex determined neonatal SLC6A4 methylation independent of ELS and 5-HTTLPR genotype, and the methylation was higher in females than males (Dukal et al., 2015). The short allele of the 5-HTTLPR polymorphism and maternal prenatal stress/child maltreatment showed an additive relationship to the reduction of SLC6A4 mRNA expression in vivo (Wankerl et al., 2014). Deletion of the SERT led to depression-like behavior, which may be associated with reduced neuronal plasticity (Lira et al., 2003). Female offspring of mice after maternal immune activation have enhanced anhedonia behavior, as manifested by a reduced preference for sucrose (Reisinger et al., 2016). In addition, in the hippocampus, the binding of acH3 and acH4 histones to the SERT promoter was increased nearly twofold, suggesting that SERT is a specific target for the regulation of epigenetic changes induced by maternal immune activation (Reisinger et al., 2016).

In the prefrontal cortex, Bdnf mRNA levels were more sensitive to the changes in SERT. SERT deficiency significantly reduced Bdnf mRNA expression in rat prefrontal cortex during the first week of life, whereas changes in the ventral hippocampus were not seen until the second week (Pezawas et al., 2008; Calabrese et al., 2013). This may be related to the degree of promoter methylation of Bdnf exon IV in SERT ^−/− rats. Further study found that the level of DNA methyltransferase, DNA (cytosine-5)-methyltransferase 1 (Dnmt1) was increased in SERT ^−/− rats, while the level of demethylase growth arrest and DNA-damage-inducible beta (Gadd45β) was decreased. In addition, the induction of depression- or anxiety-like behaviors by SERT deletion is closely associated with increased DNA methylation of Bdnf and decreased availability of transcription factors such as cAMP-response element-binding protein (Creb), neuronal PAS domain protein 4 (Npas4), and calcium-responsive transcription factor (Carf) (Molteni et al., 2010; Calabrese et al., 2013). Furthermore, the study found that serum SERT mRNA levels were reduced in individuals exposed to maternal prenatal stress or child maltreatment, but this phenomenon was not observed in stressed adults, suggesting that this change occurs during a sensitive period of early development (Wankerl et al., 2014). In addition, depressive symptoms in second-trimester women were positively associated with methylation within subregions of SERT CpG island (Devlin et al., 2010), and the methylation status of the gene promoter was closely related to the volume of the hippocampus, especially in the dentate gyrus, CA2, and CA3 of in the hippocampus (Booij et al., 2015).

Changes in 5-HT content regulate crf mRNA level in the paraventricular nucleus of the hypothalamus (PVN) (Jørgensen et al., 2002), and SSRIs administration reverses stress-induced Crf transcription elevation (Pan et al., 2013). CRF is a 41 amino acid peptide. Hyperactivity of the CRF neuronal system appears to be a pathological hallmark of depression, and CRF is a key mediator of the hypothalamic-pituitary-adrenal (HPA) axis (Hasan and Hasan, 2011). In response to stress, CRF initiates a series of physiological processes and ultimately releases glucocorticoids from the adrenal cortex (Smith and Vale, 2006), and HPA axis hyperactivity returns to normal after antidepressant treatment (Arborelius et al., 1999). It has been found that maternal separation alters CRF expression in brain regions such as the central amygdala (CeA), PVN, and bed nucleus of the stria terminalis (BNST) (Chen et al., 2012). In the female mice which were exposed to impoverished housing, increased risk-taking behavior during a reward-related task (predator-odor risk-taking, PORT) was associated with increased CRF receptor 1 (Crfr1) gene expression in the medial prefrontal cortex. Further studies revealed that the levels of the protein marker histone H3 at arginine 2 (H3R2me2s) in the proximal promoter region of the Crfr1 gene were elevated, whereas these changes were not observed in male mice (Viola et al., 2019). The deposition of histone-modified H3R2me2s results in a stable euchromatin structure that is essentially associated with transcriptional activation (Migliori et al., 2012), which further confirms the phenomenon of elevated Crfr1 mRNA levels after PORT testing (Schreiber et al., 2017; Viola et al., 2019). In addition, maternal separation and 5-HT genotypes affect the epigenetic modification of the urocortin 1 (Ucn1) gene (a member of the CRF peptide family). Studies have found that maternal separation induces methylation at CpG-156 and CpG-49 of the Ucn1 promoter region in the Edinger–Westphal nuclei, and 5-HTT deficiency induces DNA methylation at CpG-171 in the Ucn1 promoter region; however, DNA methylation at other CpG sites was not affected (van der Doelen et al., 2017). The interaction of maternal separation and 5-HTT genotype affects DNA methylation of the Crf gene promoter in adult rat CeA, while DNA methylation at specific sites in the Crf promoter was related to Crf mRNA levels in CeA, and CpG 36 negatively correlated with CeA Crf mRNA levels (van der Doelen et al., 2015).

The tryptophan hydroxylase 2 (TPH2) gene encodes a rate-limiting enzyme involved in the 5-HT synthesis (Tesoro-Cruz et al., 2021). Clinical studies have shown that the association of TPH2 methylation changes with ELS only manifests at specific CpG sites. Men with depression who experienced ELS had hypermethylation at the TPH2-5-203 CpG site, whereas depressed women who experienced ELS had hypermethylation at the TPH2-10-60 CpG site. These methylations alter the transcription of TPH2, further disrupting 5-HT levels, thereby counteracting the effects of antidepressants (Shen et al., 2020).

It has been reported that maternal separation combined with social isolation stress reduced Drd1a mRNA expression, while Drd2 mRNA expression did not change significantly (Sasagawa et al., 2017). In addition, stress increased methylation of the promoter of the Drd1a gene in mice NAc, and 29 of 31 CpG sites (CpG sites 1–5, 8–16, and 18–31) in the Drd1a gene were more frequently methylated than normal mice (Table 1) (Sasagawa et al., 2017). It has been found that maternal separation induces an increase in DNA methyltransferase expression in pups and adult rats NAc (Anier et al., 2014; Todkar et al., 2015). According to a recent study, demethylation of histone H3 on lysine 79 (H3K79me2) was a key regulator of transcriptional abnormalities in the adult NAc induced by maternal separation (Kronman et al., 2021). In addition, short-term ELS increased DRD1 expression in the hippocampus and decreased the expression of histone H3 acetylation and dopamine- and cAMP-regulated neuronal phosphoprotein (DARPP-32) but did not change the level of histone H4 acetylation. In contrast, long-term maternal separation upregulated DARPP-32 expression but did not alter DRD1 expression and histone H3/H4 acetylation (Köhler et al., 2019). Furthermore, it was confirmed that the short-term maternal separation induced a decrease in the expression of DARPP-32 was related to the decrease of acetylation of H3 in the promoter region but not to the acetylation of H4 (Köhler et al., 2019).

H3K79me2 demethylation and the enzymes that control this demethylation [disruptor of telomeric silencing 1-like (DOT1L) and lysine-specific demethylase 2B (KDM2B)] in D2-medium spiny neurons (MSNs) are critical for maternal separation-induced stress sensitivity (Kronman et al., 2021). In adult mice, knockout of the DOT1L or overexpression of the KDM2B in NAc D2-MSN neurons attenuated maternal separation-induced depression-like behavior, including increased social interaction, increased open field exploration, and decreased despair-like behaviors, whereas the same treatment in Nac D1-MSNs had no changes in depression-like behaviors (Kronman et al., 2021). Increased DOT1L after maternal separation is associated with H3K29me2 deposition at numerous genomic loci. Meanwhile, systemic administration of a small-molecule DOT1L inhibitor reversed maternal separation-induced behavioral deficits (Kronman et al., 2021). This provides strong evidence support for the treatment of maternal separation-induced depression.

One of the closely related and persistent histone modifications in ELS is a histone deacetylase (HDAC)-mediated histone acetylation (Levine et al., 2012; Valzania et al., 2017; Shepard et al., 2018). Histone acetylation, the most reported histone modification in neurological diseases, is dynamically regulated by two distinct types of enzymes, histone acetyltransferases (HATs) and HDACs (Marmorstein and Zhou, 2014; Alameda et al., 2022). HAT transferred acetyl groups to histone tails, leading to chromatin relaxation and subsequent increase in transcription rate; the latter increased chromatin-histone interactions by removing acetyl groups and ultimately decreased transcription rate (Shepard and Nugent, 2020). Studies have shown that HDAC inhibitors exert antidepressant-like effects and alleviate depression-like and addictive behaviors (Covington et al., 2009). Studies have found that spatial and associative memory functions are impaired in histone acetylation enzyme 5 (HDAC5) knockout mice and in early-life stressed mice (Agis-Balboa et al., 2013; Suri et al., 2014; Valzania et al., 2017). In addition, a heterozygous null mutation in HDAC5 increased the time of conditioned place preference in mice that suffered from social isolation in early life but not in socially threatened mice (Valzania et al., 2017). The dorsal striatum may be involved in mediating this effect (Valzania et al., 2017). In brief, individuals are more susceptible to social isolation-induced adverse effects when heterozygous null mutations in HDAC5 are present in the organism.

In another study, it was found that maternal deprivation-induced GABAergic neuroplasticity (not glutamatergic) and aberrant A-kinase anchoring protein (AKAP) signaling could be reversed by injection of HDAC inhibitors in the ventral tegmental area (VTA), and researchers proposed these effects may be related to dopamine neurons in the VTA (Authement et al., 2015; Shepard et al., 2018). Further studies found that maternal deprivation increased HDAC2 expression in VTA dopaminergic neurons and increased BDNF (a biological indicator closely related to the onset of depression) expression in VTA (Shepard et al., 2018). In addition, prenatal stress regulated gene expression in the hippocampus, such as decreased early growth response 1 (Egr1) and RAC-alpha serine/threonine-protein kinase (Akt1) mRNA expression, which were important for regulating cell proliferation and cell survival (Boulle et al., 2016). On the other hand, maternal deprivation also induced histone hypoacetylation in VTA. In addition, maternal deprivation increased the intrasynaptic AKAP150 level and decreased the protein kinase A (PKA)-RIIβ level, and these effects can be reversed by an HDAC inhibitor (Shepard et al., 2018). Maternal deprivation-induced AKAP150-anchored signaling changes may be closely linked with HDAC2-mediated epigenetic modifications, and these modifications prevent transcription of the AKAP5 gene (Shepard et al., 2018; Shepard et al., 2020). In recent research, the interaction of PKA and AKAP150 regulated the cellular excitability and intrinsic membrane properties of VTA dopaminergic neurons, and the disruption of the AKAP150-PKA complex increased VTA action potential generation in normal animals, and it decreased in maternal deprivation animals (Shepard et al., 2020). Therefore, by targeting AKAP150 and HDACs to reduce changes in histone acetylation, they can modulate not only neuronal excitability through synaptic transmission but also ion channel activity and postsynaptic ion channel transport (Shepard et al., 2020).

ELS enhances stress sensitivity in adulthood by reducing microRNA-9 expression, which, in turn, upregulates DRD2 expression in the NAc (Zhang et al., 2015). It has been found that chronic unpredictable stress increased Drd2 mRNA levels within the NAc and decreased microRNA-9 expression, while maternal deprivation synergistically enhanced the effects of chronic unpredictable stress on Drd2 and microRNA-9 expression. In vitro studies have shown that microRNA-9 directly targets the 3′ untranslated region of Drd2 mRNA and then inhibits DRD2 expression (Zhang et al., 2015).

Exposure to in utero glucocorticoids significantly decreased cell number and dopamine levels and significantly upregulated Drd2 mRNA levels in the NAc of adult rats, while repeated morphine administration downregulated the levels of DRD2 expression while increasing the DNA methylation of the Drd2 gene. Administration of therapeutic doses of levodopa restored a hypodopaminergic state, normalized DRD2 expression, and resisted morphine-induced methylation of the Drd2 promoter in the NAc of animals exposed to in utero glucocorticoids. In addition, levodopa treatment also promoted the dendritic and synaptic plasticity of the NAc (Rodrigues et al., 2012).

Intrasynaptic dopamine transmission is primarily regulated by dopamine transporter (DAT) uptake of released dopamine, and it is sensitive to changes in DAT density and its function (Wightman and Zimmerman, 1990). A lower density of DAT sites in mice NAc may be the reason why maternal separation animals are more susceptible to stress responses (Brake et al., 2004). In addition, in the same experiment, it was found that Drd3 mRNA levels were greatly reduced in the NAc of maternal stress model mice (Brake et al., 2004).

Tyrosine hydroxylase (TH) is regarded as a rate-limiting enzyme in catecholamine biosynthesis. It has been found that stress response is associated with TH activity in the central nervous system (Masserano and Weiner, 1983). A study found that restraint stress increased TH mRNA level and enhanced the rate-limiting enzyme in dopamine synthesis in the midbrain VTA and substantia nigra (SN). In addition, restraint stress-induced increase in TH mRNA expression was significantly attenuated in the VTA and SN of neonatal maternal separation-treated rats (Jahng et al., 2010). However, there are few studies on the mechanism of different types of ELS-induced TH expression changes, and further exploration is needed.

In clinical studies, ELS decreased NMDA receptor expression and increased glutamate-mediated excitotoxicity, which further induced neuronal loss and ultimately reduced NMDA receptor binding (Underwood et al., 2020). In animal studies, it has been found that exposure to maternal separation at 2–3 weeks postnatally altered synaptic plasticity in susceptible rats and inhibited the effect of antidepressants on CA1, accompanied by a lifetime upregulation of synaptic NMDA receptor 1 (NR1) levels in sensitive rats (Ryan et al., 2009). In the epigenome study of children, methylation at CpG sites in DNA-binding protein inhibitor ID-3 (ID3), tubulin polymerization promoting protein (TPPP), and Grin1 were inversely related to depression severity in maltreated children (Weder et al., 2014). Similarly, in maternal separation and early weaning (MSEW)–exposed mice, TPPP gene expression in the prefrontal cortex was inversely proportional to dwell time in the open arm and proportional to dwell time in the closed arm. The expression of Grin1 and ID3 genes in the medial prefrontal cortex were inversely proportional to the immobility time in the forced swimming test (Montalvo-Ortiz et al., 2016), and these were consistent with previous results.

Bisphenol A (BPA) exposure induced sex difference in methylation of the CREB-binding site (CpG1) and a site adjacent to CpG1 of the NMDA receptor 2b subunit (Grin2b) promoter, and BPA also induced hypermethylation of CpG3 in Grin2b promoter in males in the mice hippocampus (Kundakovic et al., 2015). Similarly, the effect of ELS on the methylation of the Grin2b gene has been demonstrated in clinical studies, and the Grin2b regulatory region has a higher degree of methylation in adults who have experienced childhood adversity, especially the changes at the CpG3 site in the Grin2b gene (Engdahl et al., 2021). These studies confirm that the methylation of the Grin2b gene is very sensitive to ELS, and these findings may increase the understanding of the impact of ELS on neurodevelopment at the molecular level.

Studies have found that exposure to early-life social isolation for 6 weeks in 21-day-old rats led to depression-like behaviors, including anhedonia and increased despair-like behavior (Fone and Porkess, 2008; Haj-Mirzaian et al., 2015; Wang et al., 2017). Further studies found that early-life social isolation increased the levels of di-methylation of histone H3 at lysine9 (H3K9me2) in the rat hippocampus, but not tri-methylation of histone H3 at lysine4 (H3K4me3). H3K9me2 has been reported to lead to impairment of synaptic plasticity and glutamatergic neurotransmission and is regarded as a risk factor for major depressive disorder (Muchimapura et al., 2002; Peter and Akbarian, 2011). Furthermore, early-life social isolation reduced the levels of glutamate receptor subunits (NMDA receptor subunits, NR1, and AMPA receptor subunits, GluR1 and GluR2) in the rat hippocampus, which could be rescued by minocycline (Wang et al., 2017). In contrast, maternal isolation increased NMDA receptor 2 expression in adolescent mice (Wieck et al., 2013).