Mammalian embryo development is initiated by the fertilization event. The haploid genome of oocyte and sperm are combined after fertilization and the genomes undergo origin-specific remodeling of their epigenome. After the remodeling, embryonic cells, i.e., blastomeres, start to possess unique epigenomic make-ups at the time of lineage specification. Proper remodeling of the epigenome is necessary for normal development and the dynamic changes are governed by a series of finely orchestrated events. The epigenetic reprogramming that embryos undergo includes changes to chromatin structure due to the remodeling of histone modifications and DNA methylation. This first wave of reprogramming initiates the change from germ cell specific epigenetic marks to embryo specific marks. Also, lineage-specific epigenetic markers are established as embryonic cells differentiate into fetal and placental lineages.

Global reprogramming of DNA methylation marks in early embryos has been intensively studied because disruption to the remodeling can have severe impact on development. Upon fertilization, a rapid decrease in DNA methylome is observed in multiple species, including humans (Dean et al., 2001; Santos et al., 2002; Guo H. et al., 2014). This demethylation event organizes the epigenetic structure of the embryonic genome prior to zygotic genome activation. Recent studies point out that the DNA demethylation is orchestrated by the ten-eleven translocation (TET1/2/3) enzyme family by oxidizing 5-methylcytosine (5mC) into 5-hydroxymethylcytosine (5hmC) (Tahiliani et al., 2009). Then, the 5hmC are ultimately converted into non-methylated cytosine (He et al., 2011; Shen et al., 2013). Demethylation of the genome is important for embryo survival and normal development, but the specific mechanisms underlying the remodeling are still under investigation (Wossidlo et al., 2011).

Disruption of the TET family during early development increases frequency of infertility, neurological defects, and cancer development in animal models (Dawlaty et al., 2011; Dawlaty et al., 2013; Zhang et al., 2013; Fahrner, 2023). The ablation of Tet1 causes neurological defects in spatial learning and short-term memory in mouse models (Rudenko et al., 2013; Zhang et al., 2013). Mouse models lacking functional Tet2 have an increased propensity of hematopoietic cell lineage malignancies (Li et al., 2011). Dysregulation of TET1 or TET2 is also associated with diseases in the clinic (Good et al., 2017; Cao et al., 2019; Jiang, 2020). The mutation or silencing of TET3 causes the onset of Beck-Farhner syndrome, an autosomal dominant epigenetic disorder, in humans (Fahrner, 2023). Clinical data and animal studies highlight conserved roles of the TET family in mammals and their importance for normal development.

In this review, we summarize and discuss recent findings underlying mechanisms regulating epigenetic reprogramming upon fertilization, particularly related with DNA methylation dynamics during preimplantation development in mammals. The mechanism of the reprogramming events will be outlined while emphasizing species specific characteristics. Epigenetic disorders found in patients that are rooted to dysregulated genes involved in the reprogramming process will also be discussed.

Biochemists and cytologists from the late 19th century were the first to observe that a DNA and protein complex existed in the cell’s nucleus, and the complex was named ‘chromatin’ by Walther Flemming in the 1880’s (Olins and Olins, 2003; Deichmann, 2015). Although the role of chromatin was not clear, it was suggested to influence gene expression without changing gene structure. Research on the modification of chromatin structure demonstrated that their conformation is linked to the expression of a gene without changing the DNA sequence, paving the concept of epigenetics.

The family of proteins called histones are the primary protein component of chromatin (Li, 2002). An octamer of these small and highly conserved proteins is associated with approximately 200 base pairs of DNA to form a nucleosome (Li, 2002). The lysine, arginine, and serine residues on the amino-terminal tail of histones can be modified to influence the accessibility of chromatin, thus highlighting the impact of histone tails on gene expression (Li, 2002). For example, transcriptional activation is associated with methylation of lysine on histone H3 at the fourth residue (H3K4me3), H3K36me, and H3K79me (Miao and Natarajan, 2005; Morgan et al., 2005). Alternatively, transcription suppression is accompanied by H3K9me3, H3K27me3, and H4K20me3 marks (Miao and Natarajan, 2005; Morgan et al., 2005). Other post-translational modifications such as phosphorylation, acetylation, and ubiquitylation of histones can also regulate gene activity (Li, 2002). Chromatin structure is also influenced by direct methylation of the nucleic acids within the DNA strand. The 5-methylcytosine (5mC) was first observed by Johnson and Coghill in 1925 (Johnson and Coghill, 1925); however, it wasn’t until 1975 when DNA methylation was labeled as an epigenetic mark (Holliday and Pugh, 1975; Riggs, 1975). After identifying presence of 5mC in mammalian, insect, and plant DNA in 1950, their distribution patterns on the genome were confirmed 4 years later (Wyatt, 1950; Sinsheimer et al., 1954). The 5mC marks were specifically detected before guanine (CpG dinucleotides), rather than being randomly distributed throughout the genome (Sinsheimer et al., 1954). In mammalian cells, DNA methylation is predominantly found at CpG dinucleotides, and methylation of CpG dinucleotides in the promoters of genes is typically associated with epigenetic silencing of gene transcription (Weber et al., 2007; Meissner et al., 2008; Jones, 2012). Histone modification and DNA methylation mechanisms are linked together (Li, 2002; Morgan et al., 2005) as DNA methylation marks are added to the genome by DNA methyltransferases (DNMT) and their activity are highly correlated to the local chromatin states that are controlled by histone modifications (Burgers et al., 2002).

Epigenetic marks are in general maintained consistently during cell division, i.e., mitosis. Interestingly, during preimplantation development, highly methylated genomes, inherited from germ cells, are dramatically demethylated until blastocyst stage except for imprinting control regions (ICRs) and certain retrotransposons (Borgel et al., 2010; Smallwood et al., 2011; Kobayashi et al., 2012; Smith et al., 2012; Smith et al., 2014). The maternal and paternal genomes exhibit different rates of DNA demethylation during preimplantation development and distinct demethylation pathways are involved; the maternal genome is passively demethylated, whereas the paternal genome is actively demethylated (Borgel et al., 2010; Smallwood et al., 2011; Kobayashi et al., 2012; Smith et al., 2012; Smith et al., 2014). Specifically, the paternal pronuclei are demethylated by the TET3 enzymes while the maternal pronuclei are protected by the protein STELLA (Wossidlo et al., 2011; Nakamura et al., 2012). Later in development, DNA methylation marks of the genes that are critical for pluripotency of embryos are re-established by de novo methylation (Li et al., 2007; Meissner, 2010). Although the key sequential reprogramming events are conserved across most mammals, species specific differences such as degree of demethylation and timing of onset in de novo methylation have been reported (Young and Beaujean, 2004; Hou et al., 2007). Mechanistic actions driving the reprogramming of gamete genome after fertilization have not been fully elucidated.

In mammalian cells, the presence of methylated CpG dinucleotides, specifically 5-methylcytosine (5mC), in the promoter regions is in general interpreted as silencing of genes (Weber et al., 2007; Meissner et al., 2008; Jones, 2012). Non-CpG methylation has recently been identified in other cell types such as oocytes, embryonic stem cells (ESCs), and neurons (Lister et al., 2009; Xie et al., 2012; Lister et al., 2013; Shirane et al., 2013). Methylation of cytosine preceding cytosine (CpC), thymine (CpT), and adenine (CpA) accounts for approximately 15% of cytosine DNA methylation (Ziller et al., 2011) but the epigenetic role of such methylation marks remains to be identified. The role of non-CpG methylation during early embryo development is largely unknown. Interestingly, an average of non-CpG methylation level, mostly CpA bases, is greater in oocytes (∼3%) than any other stage of development, including post-implantation (∼1%) (Li C. et al., 2018). There is little evidence it has a widespread impact on gene expression during embryogenesis. DNA methylation marks are established by two major de novo methyltransferases, DNA methyltransferase 3A (DNMT3A) and DNA methyltransferase 3B (DNMT3B) (Okano et al., 1998; Okano et al., 1999). A catalytically in-active DNMT, DNMT3L, is also involved in de novo methylation specifically in germline by stimulating activities of DNMT3A and DNMT3B through direct interaction (Suetake et al., 2004; Ooi et al., 2007). Maintenance of DNA methylation marks is led by DNA methyltransferase 1 (DNMT1). Since DNA replication occurs in semiconservative manner, newly synthesized complimentary sequence on daughter DNA strands lacks DNA methylation. Because DNMT1 preferably bind to hemi-methylated CpG dinucleotides, it functions as a key regulator, which maintain symmetrical DNA methylation levels throughout cell divisions (Bestor et al., 1988; Bestor, 1992). A depletion of DNMT1 results in the passive demethylation of the genome as DNA methylation is diluted after DNA replication (Holliday and Pugh, 1975; Hermann et al., 2004; Nishiyama et al., 2013).

The role and maintenance of 5mC have been intensively studied. Other modifications to the cytosine have been reported; however, their functions are largely unknown. For instance, the hydroxylated form of 5mC, 5-hydroxymethylcytosine (5hmC) was first reported in 1972; and it was originally reported that 5hmC accounted for ∼15% of total cytosines in DNA isolated from brain tissues of rat, mouse, and frog (Penn et al., 1972). The presence of 5hmC in mammalian DNA could not be confirmed by other studies until it was robustly detected in the mouse cerebellum and ESCs by two research groups in 2009 (Kriaucionis and Heintz, 2009; Tahiliani et al., 2009). The 5hmC residue accounted for 0.6% of the total nucleotides in mouse Purkinje neurons and 0.03% of the total nucleotides in mouse ESCs (Kriaucionis and Heintz, 2009; Tahiliani et al., 2009). Another study identified 5hmC in various mouse and human tissues with high levels in the central nervous system (Song et al., 2011). Other cytosine bases such as 5-formylcytosine (5fC) and 5-carboxycytosine (5caC) have also been reported. However, the role of the bases is not clearly understood.

Discovery of the diverse cytosine bases prompt investigations on how the bases are converted. Identifying enzymes responsible for converting 5mC into 5hmC was inspired by the production of base J (β-D-glucosyl hydroxymethyluracil) in trypanosomes (Borst and Sabatini, 2008). Base J is a modified thymine associated with gene silencing, similar in function to 5mC in mammals and is synthesized by the hydroxylation of a methyl group of thymine (Borst and Sabatini, 2008). It was suggested enzymes JBP1 and JBP2 catalyzed the hydroxylation of the methyl group of thymine as a part of the 2-oxoglutarate (2OG)- and Fe(II)-dependent oxygenase superfamily (Yu et al., 2007; Cliffe et al., 2009). Following the studies, Tahiliani et al. (2009) identified the ten-eleven translocation (TET) proteins as mammalian homologs of the trypanosome proteins JBP1 and JBP2 and demonstrated that the TET enzymes are 2OG- and Fe(II)-dependent enzymes that catalyze the conversion of 5mC to 5hmC (Tahiliani et al., 2009). Subsequent studies revealed that 5hmC can be further oxidized to 5-formylcytosine (5fC) and 5-carboxycytosine (5caC) by the TET enzymes (He et al., 2011; Ito et al., 2011). The discovery and conversion of the three 5mC derivatives suggested a new demethylation mechanism orchestrated by the TET enzymes. Recent studies found that thymine DNA glycosylase (TDG), an enzyme mediating base excision repair of DNA, has direct activity with 5fC and 5caC, implying that these two cytosine bases are intermediates of the active demethylation process (He et al., 2011; Shen et al., 2013). In addition, conversion of 5mC to 5hmC by the TET enzymes may aid in the acceleration of passive demethylation because affinity of DNMT1 towards 5hmC is much lower than for 5mC in hemi-methylated DNA strands, thus preventing the addition of 5mC to a newly synthesized strand (Hashimoto et al., 2012; Ji et al., 2014). The discovery of the DNA demethylation pathways greatly expanded our understanding of how DNA methylation marks are regulated. Conventionally, it was hypothesized that the lack of DNMT would derive DNA demethylation in cells; however, the theory could not explain some of rapid changes in DNA methylation marks seen in cells especially during embryogenesis. The TET enzyme-mediated DNA demethylation pathway offers more detailed explanation to the dynamic changes in epigenetic marks and provide clues for successful cellular reprogramming and mechanistic understanding to certain diseases associated with aberrant DNA methylation.

The establishment of DNA methylation marks has been extensively studied and its impact on gene regulation is generally understood. Yet, mechanistic pathways of active DNA demethylation and its influence on gene regulation are still under investigation. Active DNA demethylation in mammalian tissues is mediated by the TET enzymes which oxidize the 5mC mark into 5hmC (Kriaucionis and Heintz, 2009; Tahiliani et al., 2009). This discovery quickly advanced our understanding of epigenetic regulation during early embryo development in mammals. Following fertilization, genome wide demethylation process rapidly erases DNA methylation marks, which were inherited from germ cells, and starts to establish embryonic DNA methylation patterns.

Amino acid sequences of TET family are highly conserved among species and share key domains in all species (Table 1). The conserved sequences highlight the importance of TET enzymes in leading epigenetic reprogramming for successful embryo development in mammals. Three TET enzymes, TET1/2/3, are differentially expressed throughout the stages of preimplantation development and responsible for the formation of 5hmC in different stages (Tahiliani et al., 2009; Ito et al., 2010). The TET3 protein is detected in oocytes and zygotes and highly expressed in neurons (Iqbal et al., 2011; Wossidlo et al., 2011; Perera et al., 2015; Yu et al., 2015). The Tet1 and Tet2 proteins are highly expressed in the inner cell mass (ICM) of blastocysts, ESCs, and PGCs (Ito et al., 2010; Vincent et al., 2013). The TET2 is also highly expressed in hematopoietic stem cells (HSCs) and essential in hematopoiesis, including HSCs self-renewal and lineage commitment (Solary et al., 2014). A Tet1 knockout (KO) resulted in reduced birth weight and subfertility in both male and female mice, but the modification did not lead to embryo lethality (Dawlaty et al., 2011; Yamaguchi et al., 2012). The mild impact of Tet1 KO on development has been thought to be due to functional redundancy between Tet1 and Tet2, as they possess overlapping expression patterns in ESCs and HSCs (Costa et al., 2013; Zhao et al., 2015). A double KO of Tet1 and Tet2 did not cause any visible abnormalities in mice while its influence on epigenome remains unclear (Wang et al., 2013) Homozygous Tet3 KO resulted in neonatal lethality (Wang et al., 2013; Tsukada et al., 2015), and the loss of maternal Tet3 caused embryonic sub-lethality (Gu et al., 2011).

The conversion of 5mC to 5hmC is directly mediated by the catalytic domain of TET enzymes. The catalytic domain is located at the carboxyl terminal region of TET enzymes and consists of a cysteine-rich domain (CRD) and two double-stranded beta-helixes (DSBH) (Iyer et al., 2009; Tahiliani et al., 2009) (Figure 1). Two zinc fingers bind the CRD and the DSBH domain to each other to form a compact catalytic domain (Hu et al., 2013). While the catalytic domain is present in all members of the TET family (TET1/2/3), the CXXC domain, which is located at the amino terminal region, is not present in TET2. The CXXC domain mediates specific binding of TET enzymes to DNA containing CpG dinucleotides (Xu et al., 2011; Xu et al., 2012). TET1 and TET3 have a tendency to bind preferentially to CpG-rich promoter regions, and the DNA binding property of the CXXC domain is thought to confer this property (Williams et al., 2011; Wu et al., 2011; Xu et al., 2012). The CXXC domain of an ancestral TET2 was lost through chromosomal inversion during evolution and then became a separate gene, IDAX (also called CXXC4) (Ko et al., 2013). In contrast to TET1 and TET3, 5hmC regulated by TET2 is abundant in gene body and exon regions rather than in promoter regions (Huang et al., 2014). The structural similarities, especially between TET1 and TET3, indicate their overlapping function (Figure 1). However, it is unclear why the TET family genes are differentially expressed during early embryo development.

Dynamic changes in DNA methylation levels during preimplantation embryo development exhibit similar patterns in mammalian species (Figure 2). For example, a rapid demethylation of the paternal pronucleus occurs in mice, rats, pigs, and cattle (Dean et al., 2001). However, quantitative differences in the level of DNA methylation exist among the species (Kobayashi et al., 2012; Guo H. et al., 2014; Okae et al., 2014; Smith et al., 2014; Ivanova et al., 2020). For instance, differences in DNA methylation levels between species are apparent in cleavage stage embryos. Pig embryos have higher CpG methylation levels (55%) (Ivanova et al., 2020) than mouse embryos (23%–28%) (Altschuler et al., 1985; Smith et al., 2012) at 8–16 cells stage. The higher DNA methylation level in pig embryos decreases dramatically at the blastocyst stage, reaching the lowest level (13%) (Ivanova et al., 2020) compared to other species (19%–29%) (Altschuler et al., 1985; Smith et al., 2012; Smith et al., 2014; Zhu et al., 2018). In contrast to other species, dramatic DNA demethylation is not observed during preimplantation development in sheep (Beaujean et al., 2004; Zhang et al., 2021). During development from gametes to the 16-cells stage, very limited demethylation occurs, then the level of methylation increases until the blastocyst stage (Beaujean et al., 2004; Zhang et al., 2021).

Consistent with the DNA demethylation process, similar expression patterns of TET enzymes are observed in preimplantation embryos of different species (Figure 2). Transcripts of TET3 are abundant in oocytes and zygotes, while TET1 is highly expressed in the relatively late-stage embryos in mice, pigs, cows, and sheep (Ito et al., 2010; Gu et al., 2011; Iqbal et al., 2011; Wossidlo et al., 2011; Lee et al., 2014; Jafarpour et al., 2017; Duan et al., 2019; Uh et al., 2020; Uh and Lee, 2022). In mouse zygotes, the role of Tet3 in the demethylation and formation of 5hmC in paternal pronuclei has been identified (Gu et al., 2011; Iqbal et al., 2011; Wossidlo et al., 2011). Similarly, disruption of TET3 impaired 5hmC formation in porcine zygotes and bovine 2-cell embryos (Lee et al., 2014; Zhang et al., 2017; Cheng et al., 2019), suggesting a conserved role of TET3 in the demethylation of fertilized mammalian embryos. Furthermore, the fine-tuning of post-fertilization demethylation by TET3 is required for the regulation of pluripotency gene expression in mammalian embryos. Impaired DNA methylation and expression levels of pluripotency genes due to TET3 disruption have been detected in developing embryos of mice, pigs, and cows (Gu et al., 2011; Lee et al., 2014; Zhang et al., 2017; Cheng et al., 2019; Uh and Lee, 2022). Genetic ablation of Tet3 leads to perinatal death without pregnancy failure in mice (Gu et al., 2011), but the impact of TET3 deficiency beyond the preimplantation stage has not been reported in other species. TET1, which is enriched in pluripotent stem cells and in the inner cell mass of blastocysts, plays an important role in the regulation of pluripotency. Although Tet1 is dispensable for maintaining pluripotency of mouse embryonic stem cells (Dawlaty et al., 2011), the disruption of TET1 leads to impaired lineage specification, particularly of the inner cell mass in both mouse and pig blastocysts (Ito et al., 2010; Uh et al., 2020). Conservation of embryonic reprogramming among our domesticated species allows greater understanding of the impact that TET family disruption has on embryo development and their epigenetic landscape (Table 2).

Disruption of epigenetic machinery is known to be linked to several neurodevelopmental and growth abnormalities in the clinic. For example, immunodeficiency-centromeric instability-facial anomalies (ICF) syndrome and Tatton-Brown-Rahman syndrome are caused by mutations in DNMT3B and DNMT3A, respectively (Vukic and Daxinger, 2019; Kumps et al., 2023). Patients with either syndrome show signs of epigenetic abnormalities. There is global hypomethylation of the genome in patients with Tatton-Brown-Rahman syndrome (Kumps et al., 2023) while regions surrounding the centromeres are hypomethylated in patients with ICF syndrome (Vukic and Daxinger, 2019). Beck-Fahrner syndrome is an autosomal dominant disorder with mutations in the catalytic domain to reduce TET3 enzymatic activity (Beck et al., 2020; Fahrner, 2023). Individuals suffering from Beck-Fahrner syndrome commonly have intellectual disability, autistic traits, movement disorders, hypotonia, and facial dysmorphism (Beck et al., 2020; Fahrner, 2023). There are no consensus diagnostic criteria for Beck-Fahrner syndrome except modifications to the TET3 gene (Fahrner, 2023).

No other genetic diseases are known to have a direct link to mutations in the TET family. However, there are clinical cases that are rooted to abnormal TET activity or somatic mutation of TET family. These clinical cases are difficult to pinpoint clear association of abnormal TET family to their pathogenesis; however, animal studies indicate potential outcomes if the level of TET family is misregulated (Table 2). Mice with Tet1 KO had a reduction in birthweight and increased infertility (Dawlaty et al., 2011). Reduced follicle size and follicle number contributed to the subfertile phenotype, but progeny was still viable (Dawlaty et al., 2011). It was observed that the ablation of TET1 caused several neurological defects such as reduced spatial learning and impaired short-term memory although their brain weight, neuron number, synaptic activity, or eyesight memory seemed to be normal (Rudenko et al., 2013; Zhang et al., 2013). The disruption of TET1 caused promoter hypermethylation and subsequent downregulation of genes associated with neural progenitor proliferation, neuroprotection, and mitochondrial function (Rudenko et al., 2013; Zhang et al., 2013). A different phenotype emerges when TET2 was disrupted. The Tet2 KO mice had expanded hematopoietic stem cell and progenitor cell populations (Li et al., 2011; Moran-Crusio et al., 2011; Ito et al., 2019), which induced disorders in myeloid and lymphoid cell lineages mimicking chronic myelomonocytic leukemia (CMML) (Moran-Crusio et al., 2011; Ito et al., 2019). Similar phenotypes are seen in the human clinic as several myeloid malignancies have loss-of-function mutations in TET2 (Li et al., 2011). There are fewer functional studies of TET3 as the knockout mice model appears to be neonatal lethal (Wang et al., 2013; Tsukada et al., 2015). It was demonstrated that a conditional Tet3 knockout in brain neurons increased anxiety-like behavior and cortisol production that’s similar to Cushing’s disease (Antunes et al., 2021).

Various somatic-cell diseases are associated with disruption of the TET family. For example, it has been shown that TET1 is downregulated in many tumor cell lines (Li et al., 2016). In breast cancer, uterine cancer, or glioblastoma, there is an enrichment of a truncated TET1 isoform that lacks the CXXC domain (Good et al., 2017), thus presumably inhibit normal action of TET1. The promoter methylation of tumor suppressing genes in these cells remained to be high and silenced these genes (Li et al., 2016; Good et al., 2017). Proper regulation of the TET family is necessary to avoid disease onset. It is well established that deregulation of WNT signaling pathway induces early-stage cancer formation (Klaus and Birchmeier, 2008). Normal TET1 expression supports WNT pathway inhibitors that inhibit cell migration, cell division, and epithelial-mesenchymal transition (Duan et al., 2017; Fan et al., 2018; Wu et al., 2019). This disruption of TET2 is common in hematopoietic cancers as its enzymatic activity mediates the formation of immune cells, especially T cells, B cells, and macrophages. Specifically, the loss of TET2 causes hypermethylation in enhancer regions within immune progenitor cells which induces their tumorigenesis (Jiang, 2020). Somatic-cell cancers such as ovarian, head/neck squamous cell carcinoma, gastric, and colorectal cancers have been implicated with TET3 disruption but the mechanistic influence of TET3 remains unclear (Misawa et al., 2018; Cao et al., 2019; Mo et al., 2020).

Abnormal TET family disruption is linked to several neurological diseases found in the clinic. The level of 5hmC is enriched in mammalian brain compared to other tissues (Tahiliani et al., 2009; Globisch et al., 2010; Song et al., 2011). As one of the main regulators of 5hmC, it is not a surprise TET3 is highly present in many brain regions, including the cerebral cortex, hippocampus and cerebellum, and its expression level is stable across different brain regions (Szwagierczak et al., 2010). As aforementioned, mutation in TET3 causes Beck-Fahrner syndrome where patients present intellectual disability and developmental delay ranging from mild to severe affecting both motor and speech abilities (Beck et al., 2020; Fahrner, 2023). To date, studies using animal models carrying neurons with a disrupted TET3 gene have reported increased anxiety-like behavior, impaired spatial orientation and short-term memory (Antunes et al., 2021; Antunes et al., 2022; Fan et al., 2022), suggesting potential risks of TET3-related neurological disorders in humans. Rare variations of TET2 were often found in patients with early onset Alzheimer’s diseases (AD) and frontotemporal dementia (FTD) (Cochran et al., 2020). Similarly, a mouse model carrying repressed level of Tet2 accelerated at presenting age-related neurogenic decline and the overexpression of Tet2 could rescue the abnormalities by protecting the animals from age-related neurodegenerative decline and enhanced their cognitive function (Gontier et al., 2018). Patients with psychosis, such as schizophrenia or bipolar disorders, do have increased expression of TET1 and altered levels of 5hmC in their brain (Dong et al., 2012; Kwon et al., 2018). The level of 5hmC is enriched in various neuronal cells (Kriaucionis and Heintz, 2009; Globisch et al., 2010; Ruzov et al., 2011; Mellén et al., 2012) and maintains for months without further modifications in the brain (Bachman et al., 2014). Unlike 5mC levels, which are relatively consistent across different tissue types (Li and Liu, 2011), the level of 5hmC is highly tissue specific, ranging from 0.03% of all cytosines in the spleen to 0.7% in the brain (Globisch et al., 2010; Li and Liu, 2011). The enrichment of 5hmC indicates the base is more than an intermediate during DNA demethylation process, but rather a stable epigenetic mark that regulates activity of nervous system (Gontier et al., 2018).

Improper maintenance of DNA methylation is linked to clinical diseases, and abnormal levels of TET family are often associated with the diseases. Abnormal global epigenetic reprogramming during development or somatic mutations on key epigenetic regulators can cause diseases including neurological disorders and cancers (Figure 3). Understanding mechanistic actions of TET family using cell and animal models will lead to the development of clinical interventions that can be used to design customized treatment options against these epigenetic disorders.

Assisted reproductive technologies (ART) are widely used in the clinic and as a technology to sustain food production around the world. In the clinic, it is estimated that over eight million children have been born via ART in western countries to date, and that approximately 6% of births in Europe in 2014 were conceived through ART (Wennerholm and Bergh, 2020). Although commonly applied, the long-term impact of ART has not been clear. It is estimated that children born using ART have a 1.4 to 1.9-fold higher risk of birth defects than children born naturally (Olson et al., 2005; El-Chaar et al., 2009; Giorgione et al., 2018). A retrospective study also states ART-pregnancies has a risk factor for epigenetic diseases three times higher than normal pregnancy (Uk et al., 2018). Beckwith-Wiedemann and Silver-Russel syndrome, both of which occur due to disruption of epigenetic imprints, have a 10-fold greater frequency in ART-conceived children compared to natural conception (Hiura et al., 2012). In addition, the methylation of CpG sites in placental and umbilical appears to be dysregulated (Katari et al., 2009).

Due to the lack of in vitro models for human implantation and placentation, studies in animal models reflect how ART may impact embryo quality and subsequent development. When compared to embryo development in vivo, ART-derived blastocysts contain a lower total number of cells and a reduced ratio of inner cell mass to trophectoderm cell numbers (Macháty et al., 1998; Yoshioka et al., 2002; Canovas et al., 2017). Similar to humans, the DNA methylomes are disrupted in domesticated species by the in vitro culture system (Barrera et al., 2017; Mani and Mainigi, 2018; Salilew-Wondim et al., 2018). In addition, ART-derived embryos would have a relatively hypermethylated global genome (Deshmukh et al., 2011; Cao et al., 2014; Zhang et al., 2016; Han et al., 2018). ART-induced epigenetic abnormalities are also well-documented in rodent models (Doherty et al., 2000; Market-Velker et al., 2010; Market Velker et al., 2012; Vrooman et al., 2022). Aberrant methylation, in ART-derived embryos, often occurs at promoter and intergenic regions (Canovas et al., 2017; Cao et al., 2020), but some imprinted regions can be disrupted as well (Canovas et al., 2017). Interestingly, a large proportion of embryos generated from somatic cell nuclear transfer (SCNT) present epigenetic abnormalities because of incomplete epigenetic reprogramming of donor cell DNA (Rideout et al., 2001; Mann and Bartolomei, 2002; Bonk et al., 2007). Embryos derived from SCNT have greater global DNA methylation than other ART-derived embryos (Kwon et al., 2008; Huan et al., 2014). In addition, SCNT-derived embryos that are arrested during early cleavage stages have greater genomic methylation than those that developed into blastocyst (Gao et al., 2018). The hypermethylation of SCNT-embryonic genome appears to disrupt transcription of genes essential for zygotic-genome activation, such as Dppa2 and Dppa4 (Cao et al., 2020). Aberrant de novo methylation has been linked to the transcriptional dysregulation of SCNT-embryos (Song et al., 2017; Gao et al., 2018). However, embryonic reprogramming of SCNT-embryos appears to be locus-specific (Zhang et al., 2016), implicating multiple epigenetic modulators being dysregulated.

It is not clear if abnormal activity of epigenetic modulators, such as TET family, in ART-embryos is directly responsible for the epigenetic abnormalities. However, since expression of these pluripotency factors are decreased in cattle and pig in vitro embryos (Cheng et al., 2019; Uh et al., 2020), and TET family is directly related to pluripotency (Costa et al., 2013), presumably, activity of TET family is associated with the epigenetic abnormalities in ART-embryos. Stress from in vitro culture systems often disrupt reprogramming ability of the TET family and other epigenetic modulators. Mouse and bovine embryos derived from in vitro fertilization have abnormal level of the TET family as it develops into a blastocyst (Chu et al., 2021). Additionally, epigenetic changes occur when embryos are removed from a low oxygen environment (∼5%) (Gaspar et al., 2015; Skiles et al., 2018). Bovine embryos exposed to atmospheric levels oxygen (∼20%) carry disrupted transcript level of epigenetic modulators such as polycomb repressor complex, histone methyltransferases, histone demethylases, and TET family enzymes (Skiles et al., 2018). Cryopreservation of mouse and bovine oocytes results in a global abnormality in 5mC and 5hmC (Chen et al., 2016; Fu et al., 2019). These abnormalities may be corrected by functioning TET family enzymes. The presence of functional TET3 greatly influences the transcript abundance of pluripotent factors, such as NANOG or OCT4, in domesticated species (Gu et al., 2011; Skiles et al., 2018; Cheng et al., 2019; Uh et al., 2020). For instance, cattle and goat embryos that overexpress TET3 have been shown to perform much better during in vitro culture (Han et al., 2018; Zhang et al., 2020). The inclusion of embryokines (such as FGF2 and LIF) does not rescue TET enzyme suppression by in vitro culture (Chu et al., 2021), but the activity of available TET enzymes has been stimulated by Vitamin C supplementation which improved 5mC and 5hmC levels (Zhang et al., 2016; Fu et al., 2019; Zhang et al., 2019; Chu et al., 2021). It may be fruitful to investigate how in vitro culture systems influence the function of the TET enzymes or their cofactors as the TET family appear to be major contributors to ART-derived embryo success.

Use of embryos produced via ART is an important part of sustaining livestock productivity. Elite genetics can be rapidly introduced into the production system and ART-embryos between elite animals can enhance genetic improvement (Vishwanath, 2003; Sirard, 2018). Although successful animal production from ART-embryos, even at industry scale, is possible, the ART-embryos and subsequent animals may have epigenetic abnormalities. For example, in vitro conceived cattle often suffer from large offspring syndrome (LOS), which is caused by aberrant expression of insulin-like growth factor 2 receptor (IGF2R) due to abnormal imprinting errors (Urrego et al., 2014). In addition, loss-of-imprinting was detected in ART-induced LOS fetus tissues (Chen et al., 2015). Oocytes derived via in vitro maturation is known to have differentially methylated regions, compared to their in vivo counterparts (Jiang et al., 2018). Several other developmentally important genes have been shown to be disrupted in in vitro bovine embryos, such as X (inactive)-specific transcript (XIST) and insulin-like growth factor 2 (IGF2) (Urrego et al., 2014). As mentioned earlier, the TET-mediated epigenetic reprogramming is conserved in many species, including livestock. Utilizing mechanistic actions of epigenetic modulators such as the TET family will assist us to correct epigenetic abnormalities associated with ART-embryos.

Proper epigenetic reprogramming upon fertilization is coordinated by the TET family and is essential for successful development. These enzymes are indispensable as several neurological disorders and cancers are associated with the malfunction of TET enzymes. While species differences do exist, the role of TET family is highly conserved among different mammals. Understanding mechanistic actions of the TET family can provide clues to improve the well-being of individuals suffering from epigenetic disorders and maintain epigenetic integrity of embryos produced via assisted reproductive technologies.