Several human and animal studies have identified associations between PAE and alterations in epigenetic modifications such as DNA methylation, histone modifications, and non-coding RNAs, including microRNAs (miRNAs)
[44,45,46][44][45][46]. However, the molecular mechanisms underlying these alcohol-induced epigenetic effects are still poorly understood. It has been suggested that early PAE may interfere with epigenetic reprogramming and could change the establishment of epigenetic marks. One suggested mechanism is the alcohol-induced reduction in folate, B6, or B12 vitamins in the methionine cycle, which decreases the amount of S-adenosylmethionine (SAM) in the cells
[47,48,49][47][48][49]. As SAM is a donor of methyl groups for both DNA and histone methylation, alcohol could change the establishment of epigenetic marks in a developing embryo. Furthermore, acetyl groups from alcohol were rapidly incorporated in histone acetylation in the brain in a mouse model and changes have been observed also in a prenatally alcohol exposed fetal brain
[46]. Epigenetic changes in the first embryonic cells could be fixed in persistent cellular memory and mitotically transmitted to different cell and tissue types. Consequently, these alterations may affect gene regulation and depending on the function of the cell types or tissues, they will contribute to the complex phenotype of FASD.
2.1. Effects of Early PAE on Epigenome in Mouse Models
Only a few studies have examined the epigenetic effects of in vivo alcohol exposure during the first trimester thus far. The first global methylation study was performed by Garro et al.
[50]. Acute alcohol administration twice a day during GD9–11 (in total of five 3 g/kg doses by gavage) resulted in global hypomethylation in DNA methylation profiles in mouse fetuses, potentially by inhibiting
DNA methyltransferase 1 (
Dnmt1) activity. In later studies, early PAE-induced epigenetic changes were detected in the placenta, but the embryos remained unaffected. Haycock and Ramsay
[51] studied the effects of alcohol in GD10.5 mouse embryos and placentas exposed during preimplantation development (2.9 g/kg ethanol injections in females on GD1.5 and GD2.5) and found decreased DNA methylation at the imprinting control region of the
Igf2/
H19-imprinted locus in the placentas.
By using a mouse whole-embryo culture, Liu et al.
[52] investigated the effects of PAE (88 mM ethanol exposure on GD8.5 throughout 44 h) at early embryonic neurulation. They showed that early PAE causes changes in DNA methylation with associated changes in gene expression and found significant methylation changes in imprinted genes and genes known to have roles in growth, cell cycle, apoptosis, cancer, and olfaction. In this study, specific regions became less methylated and others more methylated in response to alcohol exposure, suggesting that some regions may be more sensitive to the effects of alcohol-induced alterations. In addition, they observed delayed growth and reduced overall growth with significant alteration in the development of the heart, caudal neural tube, brain vesicles, optic system, and limb buds of the embryos treated with alcohol.
By using our mouse model of early PAE (maternal ad libitum ingestion of 10% (
v/v) ethanol during GD0.5–8.5, approximately 12 g/kg/day), we have shown for the first time that alcohol could affect the adult phenotype by altering the epigenotype of the early mouse embryo
[53]. We studied the effects of early PAE by using a metastable epiallele
Agouti viable yellow (
Avy), which is a dominant mutation of the murine
Agouti (
A) locus, caused by the insertion of an intracisternal A-particle (IAP) retrotransposon upstream of the
Agouti coding exons. The activity of
Avy is variable among genetically identical mice, resulting in mice with a range of coat colors; from yellow to mottled to agouti (termed pseudoagouti)
[54].
Avy is a widely used “biosensor” to study the effects of environmental exposures on the epigenome
[55,56][55][56]. Our results demonstrated that early PAE increases the DNA methylation level at the
Avy allele and, consequently, alters the coat color of offspring
[53]. Furthermore, we found similar changes in gene expression in the hippocampus, olfactory epithelium, and mesodermal bone marrow of adolescent mice, suggesting that changes in gene regulation may have already occurred in the first cells of the embryo
[57].
2.2. Embryonic Stem Cells as a Model for Early Alcohol-Induced Effects
Most of the early alcohol exposure studies have focused on exposure in in vitro models using undifferentiated or differentiating stem cells. Embryonic stem cells (ESCs) derived from the inner cell mass of a developing blastocyst are pluripotent, having the developmental potential to give rise to all three embryonic germ layers and eventually differentiate into all cells and tissues of an adult organism. The pluripotent identity of ESCs is governed by a network of transcriptional factors, signaling pathways, epigenetic regulators, and structurally open chromatin that holds ESCs in an undifferentiated state. In response to differentiation signals, the transcriptome, epigenome, and chromatin structure in the differentiating cells undergo rapid global changes that silence the pluripotency genes and activate selected lineage-specific genes
[58,59][58][59].
Since differentiation can be experimentally induced in vitro, and given their indefinite self-renewing capacity, ESCs are a useful model to study cellular identity and early developmental events during embryogenesis. Their environment can be easily manipulated, which makes them a valuable tool to clarify the interaction between genes and environment in the beginning of embryonic development. ESCs have been used in toxicological research and several human and mouse stem cell models have been established to study the early effects of alcohol. As alcohol strongly disrupts the development of the nervous system
[60], several models of neural differentiation have also been developed to study the effects of alcohol on early human neural and brain development. To date, studies on ESCs have observed alterations in the proliferation and differentiation properties as well as changes in molecular mechanisms such as in the levels of core pluripotency factors and epigenetic marks
[61,62][61][62].
2.3. Effects of Alcohol on Embryonic Stem Cells and Differentiation
Studies on mouse (mESC) and human (hESC) embryonic stem cells have shown that differentiating cells and developing tissues are more vulnerable to alcohol than already differentiated cells. Arzumnayan et al.
[13] showed that 80–84 mM ethanol exposure for 1–6 days affected neither the proliferation nor the expression of pluripotency markers of undifferentiated mESCs, but triggered apoptosis during embryonic body (EB) differentiation. Nash et al.
[63], in turn, showed that a low dose of ethanol (20 mM for one week) increases cell proliferation and induces larger colonies, and simultaneously increases cell apoptosis in undifferentiated cells and ethanol-exposed hESC-derived neural progenitor cells. Moreover, Taléns-Visconti et al.
[64] showed that ethanol exposure (25 and 50 mM in proliferating or differentiating media) not only impairs neural progenitor cell survival, but also the differentiation of hESCs into neural progenitors and further into mature neurons and astrocytes. Ethanol exposure also induced expression changes of neural differentiation-associated genes and disrupted the actin cytoskeleton of neural progenitors
[64].
There is only scarce evidence of the molecular mechanisms behind the alcohol-induced effects on cell differentiation. One potential mechanism is the effects on the signaling pathways that have important roles in the regulation of stem cell differentiation and the control of embryonic development, such as the WNT pathway
[65]. Ethanol exposure (25, 50, and 100 mM after definitive endoderm stage until harvesting) has been shown to suppress the early hepatic differentiation of hESC-derived hepatic progenitor cells in a dose-dependent manner by inhibiting WNT as well as the MAPK/ERK pathway
[66]. Furthermore, alcohol-induced inhibition in WNT signaling was also observed during human neural stem cell (NSC) differentiation
[67] and the cardiac differentiation of mESCs
[68,69][68][69].
Another possible underlying mechanism of the alcohol-induced effects on differentiation is alterations in the levels of core pluripotency factors
POU class 5 homeobox 1 (
Oct4),
SRY-box transcription factor 2 (
Sox2), and
Nanog Homeobox (
Nanog) that control the pluripotent state of ESCs
[70]. In mESCs, ethanol exposure has been seen to inhibit the loss of core pluripotency markers during both EB (81–84 mM ethanol for 1–6 days)
[13] and cardiac differentiation (17.1–51.4 mM for 14 days)
[68], suggesting that alcohol may delay the ESC differentiation.
In addition to maintaining pluripotency, the core pluripotency factors are involved in lineage selection during early differentiation events. The dosage of
Oct4 and
Sox2 in differentiating cells determines the lineage commitment—an increased
Oct4/Sox2 ratio induces ESCs to differentiate into mesoendoderm (ME) and, by contrast, a decreased
Oct4/Sox2 ratio into neuroectoderm (NE)
[71,72,73][71][72][73]. Interestingly, alcohol has been shown to affect the balance of these lineage specifiers and alter the differentiation trajectory into specific lineages, especially when differentiating into NE. Ogony et al.
[74] found that ethanol exposure (25, 50, and 100 mM for 0–6 days) increases the expression of
Oct4 in a dose- and time-dependent manner, elevates the overall
Oct4/Sox2 ratio, and misleads the cells into an ME cell fate during ESC differentiation into NE. Furthermore, they investigated the effects of ethanol (100 mM for 0–6 days) on the
Oct4/Sox2 ratio as well as on the expression of other downstream genes involved in pluripotency, differentiation, and signaling during the mESC differentiation into NE
[62]. Alcohol exposure was shown to downregulate 19 pluripotency genes and upregulate 14 differentiation-associated genes. The changes in the NE differentiation-associated genes altered the overall gene expression dynamics, which could explain the different trajectory observed during the differentiation of alcohol-exposed cells
[62].
These results collectively suggest that alcohol reprograms the lineage specification by changing the balance of core pluripotency factors
Oct4 and
Sox2, thereby forcing ESCs away from neuroectodermal cell fate. This is in line with previous animal studies in which alcohol exposure during gastrulation caused neural progenitor pool reduction
[15] as well as long-term effects on the forebrain and mature brain stem nuclei structures
[75]. The effect of alcohol on lineage specification is well illustrated by Waddington’s epigenetic landscape model, which describes how development is unidirectional, meaning that embryonic stem cells develop into a mature differentiated state
[76]. If alcohol disrupts the balance of the core pluripotency factors, it could force stem cells down the differentiation trajectory away from neuroectoderm (
Figure 2). This misguidance could result in developmental delay and defects of the nervous system underlying FASD phenotypes.
Figure 2. Embryonic stem cell differentiation into germ layers during (
A) normal embryonic development and (
B) altered development due to prenatal alcohol exposure. The Waddington’s epigenetic landscape reflects the differentiation of cells, where the fate of stem cells depends on their history as well as developmental and environmental inputs
[76]. On the highest hill are ESC progenies that progress toward differentiation. Changes in epigenetic modifications in ESC progenies lead to a reconfiguration of the core transcription factors Oct4, Sox2, and Nanog, represented as a Waddington’s landscape
[77]. Coming down to the hill, stem cells first reach a plateau with low Nanog expression that enables them to respond to differentiation signals. (
A) During normal stem cell differentiation, higher Oct4 expression induces stem cells toward the mesoendoderm progenitor cell fate, while higher Sox2 drives stem cells toward the neuroectoderm progenitor cell fate; (
B) During alcohol-altered stem cell differentiation, alcohol affects the balance of the core pluripotency factors, especially Oct4/Sox2 ratio
[74]. Therefore, alcohol may reprogram lineage specification favoring excess of Oct4 relative to Sox2 and forcing stem cells down the differentiation trajectory away from neuroectoderm.
2.4. Epigenetic Effects of Alcohol on Embryonic Stem Cells and Differentiation
The effects of alcohol on the epigenome have been mainly studied in differentiated cells such as NSCs and heart progenitors rather than in undifferentiated ESCs or specifically in gastrulation. However, the results have shown associations between ethanol exposure and alterations in epigenetic mechanisms.
Alcohol-induced changes in the DNA methylation of ESCs and differentiating cells were studied in a genome-wide DNA methylation sequencing analysis, which revealed significant alterations in the methylation and transcriptomic profiles of ethanol-treated (20 mM for 24 h) undifferentiated hESCs, leading to reduced pluripotency, and also ethanol-treated differentiated EBs
[61]. A higher global hypermethylation in undifferentiated ESCs than in EBs was observed at the promoter regions, suggesting that the methylomes of undifferentiated ESCs are more prone to alcohol-induced effects than the methylomes of already differentiated cells. The changes in transcriptomic of undifferentiated hESCs and EBs were associated with oxidative stress, metabolic processes, and neuronal properties
[61].
DNA methylation changes have also been detected in ethanol-exposed mouse NSCs. Hicks et al.
[78] focused on promoter regions and found that ethanol exposure (86.8 mM for 48 h) prolonged the total length of the cell cycle, increased the activity of
Dnmt1 and induced hypermethylation of several cell cycle genes. Considerably, the results of the increased activity of
Dnmt1 in NSCs contradict the findings of Garro et al.
[50], who showed that PAE decreases
Dnmt1 activity in fetal mice. Moreover, Zhou et al.
[79] found that binge-like ethanol exposure (88 mM for 6 h) delayed the migration, neuronal formation, and growth of rat NSCs and prevented the methylation of genes associated with neural development, eye development, and developmental disorders during the reprogramming of quiescent NSCs into differentiation. Ethanol exposure (70 mM for 48 h or 8 days) during NCS differentiation has also been shown to increase the expression of
methyl CpG binding protein 2 (
Mecp2), an important epigenetic factor in the brain, in association with decreased DNA methylation and increased hydroxymethylation at its regulatory elements
[80].
The knowledge of alcohol´s effects on the chromatin structure in ESCs is scant, but a few studies show alterations of histone modifications in NSCs
[81,82][81][82] and heart progenitor cells
[83,84][83][84]. Veazey et al.
[81,82][81][82] studied the effects of alcohol on histone modifications (H3K4me3, H3K9me2, H3K9ac, H3K27me3) at specific genes associated with development in fetal mouse NSCs and showed that ethanol exposure (35 and 52 mM for 3 days and 70 mM for 5 days) exhibits significant alterations in the chromatin structure. The effects were dependent on the gene of interest and the dose of ethanol as well as whether the cells had recovered from the treatment before harvesting
[82]. Interestingly, after the recovery, most of the changes were observed in histone marks associated with a repressive chromatin structure (H3K9me2 and H3K27me3). These changes were also consistent with the repressed chromatin structure observed in the in vivo mouse model (ethanol exposure on GD7), correlating with the development of craniofacial and central nervous system defects as well as a decreased expression of genes that are associated with development and epigenetic reprogramming
[82]. The ethanol exposure (50 and 200 mM for 24 h) of mouse heart progenitor cells has been shown to increase H3K9 acetylation and change the expression of the genes involved in heart development
[83,84][83][84].
Extensive alcohol-mediated changes in miRNA and miRNA-target gene expression have been observed in utero that can be detrimental to the developing embryo
[85]. These alterations are mainly studied in mouse neural progenitor cells in the early embryonic stage. For instance, Sathyan et al.
[86] found that ethanol (70 mM for 5 days) suppresses the expression of miRNAs (miR-21, miR-335, miR-9, and miR153) in cerebral cortical neuroepithelial precursors. Later, they reported that exposure also suppressed miR-140-3p in cultured neurospheres, suggesting that relatively small teratogen-induced changes in miRNAs during stem cell differentiation can result in long-lasting deficits in brain function
[87].