Dynamic metabolism during early mammalian embryogenesis

Metabolism is the core process underlying all biological phenomena. It can be defined as the sum of biochemical processes in living organisms that either produce or consume energy. Catabolic metabolism refers to biochemical processes that generate energy by breaking down biomass. Generally, glucose carbon is used through glycolysis and the tricarboxylic acid (TCA; Box 1) cycle to serve as fuel for adenosine triphosphate (ATP) production (Wang et al., 2019). In addition, intermediates generated in both glycolysis and the TCA cycle facilitate macromolecular biosynthesis in anabolic metabolism. With energy recycling, cellular metabolism acts as a dynamic orchestration of interconnected cell signalling networks, and dynamic metabolism supplies energy sources and metabolic building blocks ‘on demand’ as required in the developmental process.

Precise metabolism reprogramming is essential to support the energetic and biosynthetic needs of early embryos from the zygotic stage through pre-implantation blastocysts (Kaneko, 2016) and decades of research has established the changing nutrient requirements of the developing embryo (Gardner and Leese, 1990, 1988; Gott et al., 1990; Leese, 2012). Such knowledge has come from a rather small group of laboratories, using specialist, low throughput and labour-intensive methods to determine the metabolism of early embryos. In many cases, this has relied on assessing the metabolic function of groups of embryos; an approach that is powerful in defining overall patterns of biochemical function but lacking the resolution to enable studies on individual embryos to unpick the extent of metabolic heterogeneity. Fortunately, in the past ten years, technological advances have enabled metabolomic studies with a small number of samples obtained directly from early embryos (Li et al., 2020; Sharpley et al., 2021; Zhao et al., 2021). Therefore, there has been a remarkable increase in the knowledge of embryo nutrient handling and metabolic function. The nutrients not only are important for canonical metabolism functions, such as bioenergetics and biosynthesis (Leese, 2015; Lewis and Sturmey, 2015), but also store spatiotemporally regulated signal cues involving important developmental events (Miyazawa and Aulehla, 2018; Tippetts et al., 2023). It is increasingly appreciated that metabolites can be rate-limiting substrates for epigenetic and post-translational modifications during important developmental events (Lu et al., 2021).

In this Review, we present the overall picture of oocyte maturation and pre-implantation embryo metabolism in the context of understanding how the extracellular nutritional environment and intracellular metabolic pathways contribute to developmental processes beyond energetic requirements. We focus largely on evidence from mouse, mentioning other mammals where relevant, and conclude by highlighting key metabolic programmes in the context of development.

Generally, pre-implantation embryos obtain their metabolites from the exogenous environment (i.e. oviduct or uterine fluid, or culture medium in vitro) (Fig. 1), although some endogenous metabolites, such as NAD⁺, are inherited from the oocyte (Hocaoglu et al., 2021; Zhu and Wang, 2021). Studies of in vitro culture systems and the in vivo environment have identified carbohydrates, amino acids and lipids as possible energy sources for embryo development (Lewis and Sturmey, 2015). Furthermore, measurements of nutrient consumption and female reproductive tract fluid composition have revealed the essential metabolites for pre-implantation embryo metabolism including glucose and pyruvate (Leese, 2003). For further discussion of other nutrients involved in mammalian development, we refer the reader to the review by Tu et al. (2023).

Carbohydrates, including sugars, starch and cellulose, are large organic compounds and are the main dietary source of energy in aerobic and anaerobic respiration via glycolysis, the TCA cycle and the pentose phosphate pathway (PPP). Carbohydrates can also be metabolised to provide protein cofactors and precursors for macromolecule modifications (Rinschen et al., 2019). Carbohydrates, such as glucose and pyruvate, are taken up by the oocyte and the embryo from the surrounding environment, such as follicular and uterine fluids, and different developmental stages of embryos exhibit distinct carbohydrate metabolism profiles.

Amino acids are widely recognised as the constituent units of proteins. However, they have several cellular functions in embryos, including energy substrates, signal transduction (González et al., 2012; Martin and Sutherland, 2001), pH regulation, osmoregulation and nitrogen sources (Leese et al., 2021). The amino acids required by early embryos in vivo are mainly derived from oviduct and uterine fluids, which have a dynamic composition, although most abundant in oviduct fluid are glycine and alanine (Hill et al., 1997). An embryo in culture relies on the growth medium to satisfy its nutrient demands and so culture medium should provide all the primary amino acids present in reproductive fluids (Sturmey et al., 2008). Importantly, the one-carbon (1C) metabolic pathway (Box 1) integrates cellular nutrient status by cycling carbon units from amino acids. This process is based on a series of interlinked metabolic pathways, including the folate cycle, the methionine cycle and trans-sulfuration pathway (Clare et al., 2019; Xu and Sinclair, 2015) (Box 1).

Lipids, including triglycerides, phospholipids and fatty acids, play a pivotal role in the regulation of embryonic development. Lipolysis and lipophagy degrade lipids to yield fatty acids and glycerol. Subsequently, fatty acids are oxidized by β-oxidation into acetyl-CoA, which can enter the TCA cycle, thus serving as a source of energy. In addition, polar lipids, such as phospholipids and sphingolipids, contribute to cellular membrane composition while facilitating crucial cell signalling events essential for embryo development and implantation (Saint-Dizier et al., 2019). The fluid in both the oviduct and uterus contains cholesterol, triglycerides and fatty acids in bovine and mouse (Saint-Dizier et al., 2019; Wang et al., 1998; Warzych et al., 2017). Myristic acid, palmitic acid and oleic acid are generally most abundant in the oviduct and uterine fluids, as well as in mouse embryos (Wang et al., 1998). Isotope tracing has shown that exogenous fatty acids, such as palmitic and oleic acids, are absorbed from their microenvironment and incorporated into both neutral lipids and polar lipids in pre-implantation embryos (Gongjin and Hirotada, 1998; Flynn and Hillman, 1980). The role of lipids in early embryo development has been described extensively (Shi and Sirard, 2022).

Cellular metabolism is coordinated by metabolite sensors through cell signalling pathways that regulate gene expression (Wang and Lei, 2018). Metabolite sensing and signalling play crucial roles in regulating the activity, stability, localization and interactions of proteins (Deribe et al., 2010). Several metabolic factors can function as sensors of cellular or extracellular nutrients to adapt or even influence the developmental programme. Two key sensors include adenosine 5′-monophosphate (AMP)-activated protein kinase (AMPK), which senses the AMP:ATP ratio (Prastowo et al., 2016), and mammalian target of rapamycin (mTOR), which primarily senses glucose and amino acids (Liu and Sabatini, 2020).

AMPK serves as a highly conserved intracellular adenosine nucleotide level sensor. It is activated in response to ATP depletion and subsequently promotes catabolic pathways for more ATP generation. In the oocyte, for example, it has been suggested that the AMPK pathway prevents excessive lipolysis by creating a negative feedback mechanism controlled by the AMP:ATP ratio. This mechanism ensures economic use of the embryonic internal energy storage during mouse oocyte meiotic resumption (Chen et al., 2006; Prastowo et al., 2016).

mTOR is a crucial nutrient sensor for the control of cell size and proliferation. It senses extracellular nutrition, such as glucose or certain amino acids, and integrates intracellular signals to serve as a central regulator of cell metabolism, growth, proliferation and survival (Huynh et al., 2023; Murakami et al., 2004). For example, in the mouse follicle, mTOR facilitates the interaction between the oocyte and granulosa cells while regulating oocyte and follicle development (Su et al., 2022). mTOR signalling is facilitated by two mTOR complexes (mTORCs). The mTORC1 complex senses nutrients, whereas mTORC2 is regulated via PI3K and growth factor signalling (Battaglioni et al., 2022).

Healthy, mature oocytes are imperative for successful fertilization. Following fertilization, the diploid totipotent zygote undergoes three rounds of cleavage, compacts into a morula and differentiates into a blastocyst containing inner cell mass (ICM) and trophectoderm (TE) cells (Bedzhov et al., 2014; Rossant, 2018). As we discuss here, both oocyte maturation and embryonic development are intricately linked to metabolic processes.

DNA methylation is a covalent modification on the fifth carbon of cytosines (5-methylcytosine, 5mC). In mammals, 5mC is predominantly located within high-density CpG dinucleotides (CpG islands, CGIs) (Eckersley-Maslin et al., 2018). DNA is methylated by DNA methyltransferases (DNMT3A/B/C) and methylation is erased by ten-eleven translocation enzymes (TETs) (Greenberg and Bourc'his, 2019). DNA methylation plays essential roles in mammalian gene regulation and genome stability during development (Wei and Wu, 2022) (Fig. 2). The amino acid derivative, S-adenosyl-methionine (SAM) is a crucial molecule produced in the methionine cycle and serves as the principal substrate for the methylation of DNA, as well as proteins (such as histones, discussed below) and RNA (Jones and Takai, 2001; Meyer et al., 2012; Yu et al., 2019).

DNA methylation is established in germ-cell precursors by DNMTs in an asymmetrical fashion. In males, spermatogonia stem cells undergo DNA methylation during the perinatal period (Bourc'his and Bestor, 2004; Dura et al., 2022; Shirane et al., 2020). In females, primary oocytes enter meiosis and arrest in the diplotene stage of the first meiotic prophase I coincident with low DNA methylation after birth. At puberty, DNA methylation is initiated in oocytes and established during the follicular growth phase and oocyte growth (Sendzikaite and Kelsey, 2019; Smallwood and Kelsey, 2012; Smallwood et al., 2011). Dynamic changes in SAM production as a methyl group donor meet the demands of epigenetic reprogramming during oocyte maturation and embryo development (Guo et al., 2017; Smallwood et al., 2011; Xu et al., 2020). In mature oocytes, the depletion of serine hydroxymethyltranferase 2 (SHMT2), which can support SAM synthesis, limits the availability of one-carbon units and reduces SAM levels, therefore decreasing mean genome methylation (Li et al., 2020).

In addition to SAM, which is produced by 1C metabolism (Box 1), redox metabolites, such as nicotinamide adenine dinucleotide (NAD/NAD⁺), play a crucial role in ensuring oocyte quality. Treatment with the NAD⁺ metabolic precursor nicotinamide mononucleotide (NMN) rejuvenates oocyte quality in aged mice; however, the exact mechanism underlying this effect remains unclear (Bertoldo et al., 2020). Several studies have demonstrated that NAD⁺ is essential for energy metabolism and epigenetic homeostasis (Canto et al., 2015).

Lipid metabolism plays a crucial role in regulating oocyte maturation and is closely associated with oocyte developmental competence. Some lipids are stored in droplets in various species and the lipid droplets (LDs) in different mammalian species exhibit notable variations. For example, porcine and bovine oocytes contain abundant lipids, whereas mouse and human oocytes possess fewer LDs (Jasensky et al., 2016; Khan et al., 2021). LDs also re-localise during oocyte maturation; staining with a lipid-specific probe has demonstrated that LDs are distributed throughout the cytoplasm in immature oocytes, whereas in maturate oocytes LDs are more peripheral, associating with the mitochondria (Sturmey et al., 2006). The number of LDs decreases during maturation but can enlarge through merging processes observed by various imaging techniques (Bradley and Swann, 2019; Brusentsev et al., 2019) (Fig. 1). As fatty acids are stored in LDs, fatty acid β-oxidation following lipolysis is involved in oocyte maturation in cumulus-oocyte complexes (Dunning et al., 2010, 2014).

Between the zygote and morula stages of embryonic development, the carboxylic acids pyruvate and lactate, derived from oviduct fluid, are the preferred energy substrates in mouse, rabbit and monkey embryos (Lane and Gardner, 2000). Pyruvate is essential for the first cleavage division to form a two-cell-stage embryo. In an elegant set of experiments, Nagaraj and colleagues observed that omission of pyruvate from culture medium between 24 and 54 h induced developmental arrest and failure of zygotic genome activation (ZGA) in mouse embryos (Nagaraj et al., 2017). They further observed that adding α-ketoglutarate (α-KG) to medium without pyruvate overcomes the two-cell blockade. The proposed mechanisms are discussed below. In addition, two-cell embryos favour methionine, polyamine and glutathione metabolism and exist in a reductive redox state through the ratio of reduced (GSH) and oxidized (GSSG) glutathione (Hansen et al., 2020; Zhao et al., 2021). Glutathione is produced via intracellular NADPH transport from granulosa cells during oogenesis and through the PPP (Box 1) (Sutton-McDowall et al., 2010). Metabolic flux analysis has revealed that neither glucose nor the carboxylic acids pyruvate and lactate serves as sources of glutathione in the embryo. Therefore, the entire pool of glutathione might be derived maternally until at least the morula stage (Chi et al., 2020).

The mechanistic requirement for pyruvate, besides energy production during ZGA, could be for histone acetylation, in which a negatively charged acetyl group is added to lysine residues on one end of a histone molecule via histone acetyl transferases (HATs) (Ryall et al., 2015) (Fig. 2A). Generally, acetylated histones open chromatin to enable transcription, whereas histone deacetylation by histone deacetylases (HDACs) is associated with transcriptional inactivation (Ruthenburg et al., 2007). Abundant pyruvate facilitates the transient localization of TCA cycle enzymes (such as pyruvate dehydrogenase, acetyl-CoA synthetase and ATP-citrate lyase) in the nucleus of the two-cell embryos (Nagaraj et al., 2017), where oxidative decarboxylation of pyruvate produces acetyl-CoA, a substrate required for a post-translational modification known as acetylation (Sutendra et al., 2014). Thus, nuclear acetyl-CoA potentially facilitates HAT-dependent acetylation of histones (Weinert et al., 2014; Wellen et al., 2009; Zhou et al., 2020), specifically H3K4 and H3K27, for chromatin remodelling during ZGA in mice (Nagaraj et al., 2017) and pigs (Zhou et al., 2021, 2020). Indeed, H3K4ac and H3K27ac modifications are permissive to gene expression and considered markers of active transcription (Zhang et al., 2018). This model explains why growth medium lacking pyruvate cannot support growth beyond the two-cell stage and suggests that the compartmentalization and regulation of acetyl-CoA and pyruvate have important implications for chromatin dynamics and gene expression (Brown and Whittingham, 1991; Gardner and Leese, 1988).

In addition to DNA methylation discussed above, histones can be methylated via histone methyltransferases (HMTs). Broad histone H3K4me3 domains are established in mouse oocytes but are removed by Jumonji C (Jmjc) domain-containing histone demethylases, KDM5A and KDM5B, at the two-cell stage, which enables the erasure of the H3K4me3 after fertilization (Dahl et al., 2016; Liu et al., 2016; Zhang et al., 2016). JmjC-domain-containing proteins use α-KG as a cofactor (Fig. 2A) (Klose et al., 2006; Tarhonskaya et al., 2017), which can be derived from the oxidation of L-2-hydroxyglutarate (L-2-HG) by L-2-hydroxyglutarate dehydrogenase (L2hgdh). L-2-HG inhibits histone demethylases and is abundant at and before two-cell-stage embryos but levels decrease during development into blastocysts (Zhao et al., 2021). In the absence of L2hgdh, or when supplementing cells with exogenous L-2-HG, histones become hypermethylated (Zhao et al., 2021). Therefore, high L-2-HG could hinder embryo development by impeding histone demethylation. However, the source of L-2-HG during embryo development remains unclear and whether L-2-HG function regulates differentiation in vivo remains to be shown.

Blastocysts have higher levels of metabolites related to the mitochondrial TCA cycle (Box 1), including succinate, fumarate, malate and α-KG, than two-cell embryos and exist in a more oxidative state (Sharpley et al., 2021; Zhao et al., 2021).

Although glucose cannot substitute for pyruvate during the earliest stages of mouse development, it is essential for the eight-cell stage, compaction and transition to the blastocyst (Brown and Whittingham, 1991; Gott et al., 1990; Hardy et al., 1989) (Fig. 1). Glucose likely makes only a minimal contribution to energy after compaction because glucose enters into glycolysis only in the absence of pyruvate (Chi et al., 2020); however, exposure to glucose for at least 1 min is essential for blastocyst formation (Chatot et al., 1994). It has been suggested that the hexosamine biosynthesis pathway (HBP; Box 1) acts as a glucose sensor in pre-blastula embryos and activates the expression of Slc2a3, a glucose transporter, to increase glucose uptake and facilitate blastocyst formation (Pantaleon et al., 2008). Glucose then enters the PPP (Box 1) to allow the nuclear localization of Yes-associated protein 1 (YAP1), a transcriptional coregulator of the Hippo pathway (Nishioka et al., 2009), as well as activating transcription factor AP-2 gamma (TFAP2C) and mTOR (Chi et al., 2020). Nuclear YAP1 and TFAP2C interact with the YAP1-binding factor TEAD4 to form a complex and facilitate TE specification (Chi et al., 2020). This requirement for glucose cannot be replaced by pyruvate and lactate (Martin and Leese, 1995); increasing lactate availability in the culture medium of blastocysts elevates the percentage of pyruvate oxidized, but does not affect glucose metabolism (Lane and Gardner, 2000). The research findings demonstrate that glucose plays a crucial role in the formation of blastocysts.

SAM is derived from methionine metabolism as a substrate by HMT for histone methylation (Clare et al., 2019) (Fig. 2A). SAM conversion from endogenous methionine is, in turn, limited by methionine and homocysteine levels, and SAM inhibitors block blastocyst formation (Menezo et al., 1989). In contrast, increasing SAM synthesis or supplementing SAM significantly increases bovine embryo hatching rate (Shojaei Saadi et al., 2016). Changes in SAM availability can result in alterations in proper epigenetic modification. Depletion of SAM induced by removing threonine from culture medium reduces levels of H3K4me3, a marker of active transcription, and promotes differentiation in mouse embryonic stem cells (mESCs) (Shyh-Chang et al., 2013).

Beyond the need for provision of appropriate substrates for methylation, proper histone and DNA demethylation promotes blastocyst formation, as well as maintaining naïve pluripotency (Tischler et al., 2019; Zhang et al., 2019b). The metabolic intermediary α-KG is a cofactor for JmjC domain-containing histone demethylases and TET enzymes, which catalyze DNA demethylation (Xu et al., 2011). By activating TET enzymes, α-KG enhances the cell number in the ICM and the competence for blastocyst formation in vitro (Zhang et al., 2019b). In naïve mESCs, which resemble the peri-implantation stage ICM, high levels of α-KG promote low H3K27me3 and DNA demethylation, contributing to pluripotency maintenance (Carey et al., 2015). Similarly, increasing the production of α-KG mediated by phosphoserine aminotransferase 1 (Psat1) maintains mESC self-renewal and proper differentiation (Hwang et al., 2016). In primed human and mESCs, however, α-KG promotes early differentiation by decreasing global histone methylation and increasing the 5hmC-to-5mC ratio (TeSlaa et al., 2016). α-KG may also extend primordial germ cell (PGC) specification competence by sustaining an appropriate balance between H3K9me2 acquisition and H3K27me3 depletion, which is key to maintaining the developmental competence for PGC fate (Lu and Teitell, 2019; Tischler et al., 2019). Some metabolites may competitively inhibit α-KG-dependent dioxygenases, because they are structurally similar to α-KG; these metabolites include succinate, fumarate and 2-HG (Xiao et al., 2012). These metabolites inhibit demethylases when they accumulate to a specific level (Kinnaird et al., 2016; Laukka et al., 2018). Fumarate and succinate production induced by fumarate hydratase and succinate dehydrogenase mutations in tumours suppressed histone demethylation and DNA hydroxymethylation by functioning as competitive inhibitors of a-KG (Xiao et al., 2012).

In addition to histone methylation, the regulation of histone acetylation is important for pluripotency because the inhibition of glycolysis leads to deacetylation and induces differentiation in ESCs (Moussaieff et al., 2015). Also, the rate of glycolysis mediates specific histone acetylation rates at certain sites in some cancer cell lines (Cluntun et al., 2015). Glycolysis may generate acetyl-CoA for histone acetylation of early-specification genes. However, it remains unclear whether these metabolites function as regulators of differentiation in vivo.

The balance between the rate of histone acetylation and deacetylation is essential for chromatin condensation and the regulation of gene expression (Roach and Mostoslavsky, 2021). In addition to histone acetylation during ZGA, the erasure of zygotic H3K27ac promotes mouse embryo pre-implantation (Li et al., 2022). One category of HDACs, deacetylase sirtuins (SIRT), requires a high level of the metabolic cofactor NAD⁺ for mediating deacetylation (Rajman et al., 2018; Schwer and Verdin, 2008) (Fig. 2). Indeed, embryo medium supplemented with the NAD⁺ precursor NMN increases blastocyst quality in aged animals (Bertoldo et al., 2020).

Amino acids in the culture medium exert beneficial effects on pre-implantation embryo development (Gardner and Lane, 1996; Lane and Gardner, 1997; Martin and Sutherland, 2001; Summers and Biggers, 2003; van der Weijden et al., 2021). For example, arginine is the most depleted amino acid during mouse early development and arginine and leucine supplementation have a positive effect on pre-implantation development (González et al., 2012). Conversely, alanine is the most abundant amino acid during early development and evidence suggests that it is important for the disposal of ammonium to protect the embryo, because embryos appear to lack a functional urea cycle (Orsi and Leese, 2004). These studies indicate that amino acids are required during embryo development.

Lipids also play an important role in energy storage and cell signalling during pre-implantation embryo development (Arena et al., 2021; McKeegan and Sturmey, 2011). Supplementing with palmitic acid, a saturated fatty acid, in pre-implantation mouse embryos in vitro suggests that fatty acids may be incorporated into embryonic lipids and used for energy production via oxidation (Yamada et al., 2012). It is suggested that fatty acids can be used as an energy source for lipids to support bovine and porcine blastocyst formation (Aardema et al., 2011; Guo et al., 2017). In addition, lipids are predominantly sequestered within LDs in mouse peri-implantation embryos, and enlarge and mobilize in coordination with embryo morphogenesis (Bradley et al., 2016; Mau et al., 2022). Furthermore, LDs are important for the embryo crosstalk with uterine receptivity during early pregnancy in mammals (Matorras et al., 2020; Ye et al., 2021), especially in some large animals, such as cows, sheep and horses (Bradley and Swann, 2019; Leese, 2012). LD abundance is speculated to provide a potential endogenous energy reserve for an extraordinarily long pre-implantation period or diapause duration (Box 2), with mustelids exhibiting extended embryonic diapause periods lasting up to 9 months due to high levels of lipids in their oocytes compared with rodents in which diapause lasts only a few days, with lower quantities of lipids (Arena et al., 2021).

As more metabolic profiling is carried out, an increasing number of functional metabolites are being identified during each stage of early embryo development. Based on these findings, we propose a new class of ‘developmental metabolites’ that may play important roles not only in metabolism per se but also in regulating development. Importantly, ‘oncometabolites’ (Xu et al., 2011) and ‘immunometabolites’ (O'Neill and Artyomov, 2019) are two classes of metabolites that exhibit specific functions in cancer cells and immune cells, respectively. Here, we propose a set of criteria to define ‘developmental metabolites’. First, these metabolites should be specifically present at a certain stage or in a specific type of cell during early embryo development. To meet a strict standard for specificity, the metabolite should be rarely present during other physiological contexts. For example, D-2-HG is a well-known oncometabolite that is specifically present in IDH mutation-carrying cancer cells (Xu et al., 2011). Similarly, itaconate is an immunometabolite that is specifically induced in lipopolysaccharide-stimulated macrophage cells (O'Neill and Artyomov, 2019). For a less stringent standard for specificity, a metabolite may be more broadly present but should exhibit a specific function related to development at a specific time. Second, the metabolite should be involved in regulating development through a clearly defined mechanism or process. Therefore, it will be important to further study the upstream production mechanisms and downstream functional impact of metabolites, such as L-2-HG, in early development. In addition, it will be interesting to employ some untargeted metabolomics approaches to identify more unique and uncharacterized developmental metabolites.

Metabolism during pre-implantation embryo development functions not only in a maintenance role but is also an active participant in the outcome of development, in line with the ‘Barker hypothesis’, also known as the ‘developmental origin of health and diseases (DOHaD)’. According to this hypothesis, adverse environmental factors during specific sensitive periods of intrauterine fetal development or early childhood, such as undernutrition or overnutrition, may increase the risk of chronic metabolic diseases or disorders in adult life (Barker, 1997). It has been reported that pregestational diabetes can exert profound effects on the long-term health of offspring through TET3 insufficiency during the environment-sensitive window of oocyte growth and maturation (Chen et al., 2022). In addition, maternal vitamin C is required for the proper reprogramming of DNA methylation and germline development because it helps maintain the activity of TET1 in a mouse model (DiTroia et al., 2019). These findings demonstrate that TETs function as key sensors of maternal metabolism and nutrition during early development, which may account for long-term effects across generations. Therefore, it is essential for women with metabolic disease to enhance peri-conception health care to protect against disease in their offspring.

Other metabolite-dependent histone modifications have been implicated in early development. For example, histone lactylation is a novel epigenetic modification that regulates various cellular processes under a wide range of pathophysiological conditions, such as infection and cancer (Zhang et al., 2019a). Interestingly, histone lactylation of H3K23 and H3K18 form a dynamic landscape in oocytes and pre-implantation embryos in mice. H3K23la and H3K18la abundance decreases when embryos are cultured in a hypoxic environment, which in turn compromises their developmental potential (Yang et al., 2021). The mechanism and function of histone lactylation in embryo development require further study.

We are grateful to all members of the Zhang lab for helpful feedback.