Chapter 3
The Role of Hexosamine Biosynthesis and Signaling in Early Development Marie Pantaleon
Abstract Although the culture requirements and the metabolic profile of the preimplantation embryo have been thoroughly investigated since their first successful culture in a defined medium, now more than 50 years ago (Whitten, Nature 177:96, 1956), it is only recently that we have begun to appreciate the impact of the environment on life-course trajectory. The mechanisms involved in how nutrient availability may potentially modulate developmental potential are consequently not well defined. Originally thought of as simple energy substrates and biosynthetic precursors, the currently emerging paradigm suggests that nutrients may act in non-classical roles to impact on developmental potential. This is now an area of considerable activity thanks to pioneering epidemiological studies (Barker et al., BMJ 298:564– 7, 1989) that have led to the establishment of the Developmental Origins of Health and Disease (DoHAD) hypothesis and a whole new field of research activity. The period prior to implantation is of particular interest as this has been identified as a critical window of developmental sensitivity to environmental or nutrient stress (Fleming et al., Biol Reprod 71:1046–54, 2004a). This review seeks specifically to explore the pivotal role of glucose in early mouse development and the mechanisms by which it may impact on the cellular functions of the developing embryo. The emerging paradigm suggests that this humble hexose sugar may be at the heart of a rather sophisticated mechanism of cellular control that not only impacts on cellular proliferation and viability in the short term but on cellular memory through to the next generation. Keywords Early development · Gene regulation · Glucose signaling · Hexosamine biosynthesis · Hexosamine signaling · Embryo metabolism · Nutrient sensing · O-linked glycosylation · Embryo viability · Embryonic programming
M. Pantaleon () School of Biomedical Sciences, The University of Queensland, St Lucia, Brisbane, Qld 4072, Australia e-mail: [email protected] © Springer Science+Business Media New York 2015 H. J. Leese, D. R. Brison (eds.), Cell Signaling During Mammalian Early Embryo Development, Advances in Experimental Medicine and Biology 843, DOI 10.1007/978-1-4939-2480-6_3
53
54
M. Pantaleon
3.1 The Embryo and its Environment During the preimplantation period the embryo undergoes a series of critical developmental events, which set up its future growth trajectory. These include, one of two genome-wide demethylation episodes immediately following fertilization in both male and female pronuclei (Dean 2014), zygotic genome activation (ZGA), as well as morphologic and metabolic differentiative events that result in blastocyst formation. ZGA occurs during the second cell cycle in the mouse and follows a wave of maternal mRNA degradation initiated during the first cell cycle. This is a critical event in development in the absence of which, the embryo fails to develop further (Schultz 1993). DNA methylation marks are re-established concomitant with differentiation and blastocyst formation, thus forming an epigenetic barrier as cells become developmentally more restricted (Seisenberger et al. 2013). The erasure of the previous generation’s epigenetic signature and the re-setting of epigenetic marks at this stage based on nutrients available at this time, confers a flexibility to the preimplantation embryo that allows it to respond to environmental challenges which could otherwise have a major impact on it future growth trajectory. Although physically independent of the mother, the preimplantation embryo is nonetheless reliant on maternal physiology for a regulated supply of the requisite nutrients and removal of end products. Glucose levels are low in the oviduct during the early cleavage phase but increase as the embryo enters the uterus to levels more in line with altered embryonic needs following compaction as the embryo increases both its reliance on glucose as an energy substrate and its ability to respond to exogenous growth factors (Kaye 1997). Oviductal and uterine epithelial secretions provide the early embryo with essential nutrients, electrolytes, macromolecules and growth factors required for optimal growth prior to implantation. Whilst it has long been accepted that these factors potentially can influence development, they have classically been viewed as facilitative rather than obligatory determinants of early development (Leese 1995) since preimplantation embryos can be cultured to the blastocyst stage in relatively simple defined media. Nonetheless, the contribution of nutrients and maternal growth factors to optimal growth can be significant as demonstrated by the difference in growth rates between in vivo-matured embryos and their in vitro counterparts (Bowman and McLaren 1970; Kaye and Gardner 1999). Indeed, metabolic control at this stage can be flexible and adaptive, as demonstrated by numerous studies inspired by the DOHaD hypothesis in various mammalian animal models (for review see Fleming et al. 2012), leaving its mark and setting growth trajectories based on the maternal nutrient environment and in anticipation of future nutrient availability. When the postnatal nutrient environment conflicts with this early setting and growth trajectories are unmatched there is the potential for disorders in the offspring (Fleming et al. 2004b). The critical question is how these adaptive mechanisms are set, what are the triggers and effector pathways that potentially lead to these adaptive responses that control growth and metabolism not only in utero but into the next generation. The mechanisms by which early embryos sense and respond to their nutrient
3 The Role of Hexosamine Biosynthesis and Signaling in Early Development
55
environment in order to bring about a biosynthetic, proliferative and potentially epigenetic programme of events, in an attempt to maintain developmental potential are therefore central to understanding this phenomenon.
3.2 A Role for Glucose in Early Development? Early experimentation aimed at understanding the nutrient requirements of preimplantation embryos established that pre-compacted embryos from a number of species, unlike other somatic cells, are unable to utilise glucose as an energy substrate. Instead they rely on monocarboxylates, pyruvate (especially for the first division) and lactate to support early cleavage (Brinster 1965; Brinster and Thomson 1966; Wales and Whittingham 1967; Biggers et al. 1967). Experiments through the late 1980s and early 1990s aimed at developing successful culture formulations for assisted reproductive technologies, further claimed that glucose was in fact inhibitory during cleavage stage development in a number of species (Schini and Bavister 1988; Chatot et al. 1989; Diamond et al. 1991; Lawitts and Biggers 1991; De Hertogh et al. 1991; Scott and Whittingham 1996; Thompson et al. 1992; Conaghan et al. 1993), (though this was later disputed by (Summers and Biggers 2003) but that nutrients such as glutamine exerted considerable beneficial effects (Chatot et al. 1989). Despite strong data indicating that glucose does not in fact support development energetically prior to compaction and may possibly be inhibitory, a subset of studies suggested that glucose was nonetheless required at some stage, raising considerable debate and confusion about the role of glucose during cleavage stage development and at what appeared to be a temporal sensitivity to that requirement (Leppens-Luisier and Sakkas 1997; Brown and Whittingham 1991, 1992; Chatot et al. 1989, 1994). Whilst glucose is not preferentially utilized for energy during early cleavage, it is nonetheless utilised via the pentose phosphate pathway to produce ribose-5-phosphate for nucleic acid synthesis and also for the generation of reducing power in the form of NADPH (O’Fallon and Wright 1986). Glucose becomes the embryo’s fuel of choice over pyruvate following compaction at the 8-cell stage in the mouse (Leese and Barton 1984), when it is required to fuel the high energy requiring Na + /K + ATPases that facilitate cavitation (Watson and Barcroft 2001). It would appear therefore that this metabolic differentiation helps facilitate the morphological changes associated with blastocyst formation. This acquisition of metabolic competence to utilize glucose as an energy substrate correlates with the apical expression of the high-affinity high-capacity glucose transporter GLUT3 or SLC2A3 at compaction (Pantaleon et al. 1997). Inhibition of GLUT3 using antisense oligonucleotides not only reduced blastocyst glucose transport as anticipated, but inhibited blastocyst formation (Pantaleon et al. 1997) linking blastocyst formation with metabolic differentiation.
56
M. Pantaleon
This was intriguing because notwithstanding a preference for glucose post compaction, mouse and human embryos cultured from the 2-cell or the 2–4-cell stage respectively in the absence of glucose could adapt to its absence by increasing utilization of pyruvate to form blastocysts (Martin and Leese 1995; Conaghan et al. 1993). Moreover these embryos already possessed, or developed, the capacity to transport glucose at high rates should it become available (Martin and Leese 1995).
3.3 Glucose Primes Embryos to Adapt to Their Environment Unlike 2-cell embryos, newly fertilized zygotes collected and cultured without glucose do not progress beyond the morula stage. Moreover, these embryos do not develop the capacity to transport glucose should it become available, or the adaptive capacity to cope with its absence (Martin and Leese 1995). Thus, zygotes cultured under these conditions display reduced proliferative capacity and higher levels of apoptosis (Pantaleon et al. 2008). Moreover they undergo oxidative stress and exhibit elevated levels of reactive oxygen species (ROS) despite a supply of sufficient monocarboxylates (Jansen et al. 2009). This suggests that in the complete absence of glucose the embryo is unable to respond to environmental stress, making it more susceptible to developmental arrest and cell death. Cellular ability to respond to multiple forms of stress is in fact associated with the ability to increase glucose uptake rapidly. This has classically been thought to be an essential survival response of adaptive importance in a healthy organism. Remarkably, brief exposure to glucose prior to the morula stage was sufficient to “prime” these naïve embryos and permit subsequent blastocyst formation (Brown and Whittingham 1991, 1992; Chatot et al. 1994). So despite an inability to derive metabolic energy from glucose prior to compaction, these experiments suggested the presence of adaptive responses that were nonetheless activated by glucose availability (Martin and Leese 1995). The ability of this brief glucose exposure to modulate utilization of pyruvate and glucose in response to an adverse environment is linked to activation of GLUT3 expression and maintenance of expression of the transporter for pyruvate and lactate, H + monocarboxylate cotransporter 1 (MCT1) post compaction (Pantaleon et al. 2008; Jansen et al. 2006). Non-metabolisable glucose analogues such as 3-O methyl glucose and 2-deoxy glucose were unable to replace glucose and stimulate blastocyst formation (Chatot et al. 1994) suggesting that the role of glucose is to provide a transcriptionally active downstream metabolite. Moreover, glutamine is also unable to substitute for glucose (Brown and Whittingham 1991), although it is noteworthy that its incorporation increases 10-fold between the zygote and morula stages (Brinster 1971) facilitated by a 3-fold increase in transport capacity between the 2-cell and blastocyst stages (Lewis and Kaye 1992). Fructose was also shown to facilitate blastocyst formation but not to the same degree as glucose although it did appear to have a beneficial effect on cell number (Chatot et al. 1994). The fact that hexokinase and
3 The Role of Hexosamine Biosynthesis and Signaling in Early Development
57
glucose phosphate isomerase are expressed in mouse oocytes (el Mouatassim et al. 1999) and active throughout cleavage stage development in both mouse and human spare embryos (Martin et al. 1993; West et al. 1989) suggested that glucose conversion to fructose-6-phosphate was involved. As an aside, the reduced efficacy of fructose to stimulate blastocyst formation may be due to reduced uptake from the external environment rather than metabolism per se. GLUT1, the glucose transporter expressed during cleavage stage development in the mouse (Pantaleon et al. 2001a; Pantaleon and Kaye 1998), transports glucose and other hexoses with varying efficacy but not fructose which is transported by the cytochalasin B-insensitive Class II mammalian monosaccharide carriers, GLUT2 and GLUT5 (Cura and Carruthers 2012). Whilst there are some reports of GLUT2 expression in later stage blastocysts (Aghayan et al. 1992; Hogan et al. 1991) expression of neither transporter has been described in mouse cleavage stage embryos to date. Moreover, functional studies, which failed to demonstrate residual saturable glucose transport activity following cytochalasin B inhibition (Gardner and Kaye 1995), do not support expression of any cytochalasin B-insensitive Class II carriers through preimplantation development in the mouse and thus an efficient conduit for fructose uptake from the external environment.
3.4 Hexosamine Biosynthesis: An Embryonic NutrientSensing Pathway The increased rate of glutamine incorporation and transport capacity during the cleavage stages (Brinster 1971; Lewis and Kaye 1992) coupled with its beneficial effect on early development as discussed above, led us to question whether glucose metabolism through hexosamine biosynthesis might be involved in conferring on the embryo the adaptive capacity to develop in an adverse environment. In support of this hypothesis, early experiments demonstrated that glucosamine, can substitute for glucose and facilitate both blastocyst formation and GLUT3 expression (Pantaleon et al. 2001b). Over 90 % of 14C-1-glucosamine initially taken up by somatic cells in culture is found in the UDP-N-acetylhexosamine pool (Kornfeld and Ginsburg 1966) suggesting that glucose metabolism through to this activated sugarnucleotide may be implicated in this function for glucose in setting up adaptive responses at this stage in development. Significantly, glucose flux through this pathway appears to be restricted to glucose transported by GLUT1 (Buse et al. 1996). Conclusive evidence to support a role for the hexosamine biosynthetic pathway (HBP) in glucose sensing in the early embryo, came from pharmacological studies. Inhibition of the first and rate-limiting enzyme of the pathway, glutamine: fructose-6-phosphate amidotransferase (GFPT) using the competitive analogue azaserine inhibited glucose activated blastocyst formation and GLUT3 expression (Pantaleon et al. 2008). Moreover, these perturbations could be rescued with glucosamine supplementation, which feeds into hexosamine biosynthesis below the level of GFPT (Fig. 3.1), thus supporting the hypothesis that the HBP may be the
58
M. Pantaleon
Fig. 3.1 Glucose is phosphorylated upon cellular entry and isomerized to fructose-6-phosphate (F-6-P) prior to utilization through glycolysis and pentose phosphate pathway (PPP). Up to 5 % of F-6-P however is converted to glucosamine-6-phosphate (GlcN-6-P) by the rate-limiting enzyme GFPT and subsequently acetylated by glucosamine-6-phosphate acetyl transferase (Emeg32) and converted into UDP-N-GlcNAc via the UDP-GlcNAc pyrophosphorylase (uap1) for use as a donor substrate in multiple biosynthetic reactions including the O-linked modification of a diverse group of nucleoplasmic proteins by N-acetylglucosamine (GlcNAc). Addition of GlcNAc to ser/ thr residues of target proteins is catalysed by O-GlcNAc transferase (OGT) and its removal by an O-GlcNAc-selective β-N-acetylglucosaminidase ( O-GlcNAcase or OGA). These reactions can be manipulated using specific inhibitors, BADGP and PUGNAc to evaluate the role of O-linked GlcNAcylation systematically, whilst the rate-limiting step in UDP-GlcNAc formation catalysed by GFPT, can be manipulated experimentally using the glucosamine analogue azaserine
pathway through which glucose signals impact on early embryos (Pantaleon et al. 2008). The importance of this pathway is highlighted by loss/gain of function models for two of the key enzymes of this pathway. Thus, homozygous null mutants of glucosamine-6-phosphate acetyl transferase (also known as Emeg32) abort at E7.5 with pronounced developmental delays (Boehmelt et al. 2000), whilst overexpression of GFPT leads to peripheral insulin resistance (Hebert et al. 1996). Indeed, the HBP has been shown to be important in the development of insulin resistance in somatic cells (Yang et al. 2008) which underscores the importance of this pathway as a nutrient sensor (Marshall et al. 1991; Traxinger and Marshall 1991; Hawkins et al. 1996; Wang et al. 1998). Unlike other nutrient-response systems that respond to specific stimuli, for example AMPK, which responds to the AMP/ATP ratio, and mTOR which responds
3 The Role of Hexosamine Biosynthesis and Signaling in Early Development
59
to amino acid concentrations as well as to the insulin/PI3K system, the HBP integrates several nutrient sources and nutrient sensing pathways, making it acutely sensitive to all nutrient flux (Hanover et al. 2010). Indeed, the HBP integrates the metabolism of carbohydrates (glucose), amino acids (glutamine), fat (acetyl CoA) and nucleotides (uridine-diphosphate) with energy charge (ATP) in the synthesis of UDP-GlcNAc, making this activated sugar-nucleotide potentially the most general sensor of cellular nutritional status (Hanover et al. 2010). UDP-GlcNAc then acts as the donor substrate for a number of biosynthetic reactions including GPI lipid anchor biosynthesis, sugar nucleotide, glycoside and ganglioside biosynthesis and N-linked glycosylation for the production of glucosaminoglycans, glycolipids and membrane and secretory glycoproteins. Additionally UDP-GlcNAc acts as a precursor to modify regulatory proteins post-translationally in the nucleus and cytoplasm through O-linked glycosylation with N-acetyl glucosamine ( O-GlcNAcylation) (Fig. 3.1). Not surprisingly, the HBP is highly active during oocyte maturation where it is utilised for the biosynthesis of complex glycan structures such as hyaluronic acid in the extra cellular matrix of cumulus-oocyte complexes (COCs). Indeed, it is estimated that 25 % of glucose consumed by gonadotropin stimulated bovine COCs is incorporated into the extracellular matrix via the hexosamine biosynthetic pathway to facilitate cumulus expansion (Sutton-McDowall et al. 2004).
3.5 The Response Path: N-Linked Vs. O-Linked Glycosylation? Early studies in embryos using tunicamycin, showed that this specific inhibitor of N-linked glycosylation suppressed glucose incorporation into glycoproteins and inhibits compaction (Surani 1979; Wales and Hunter 1990). However, it did not affect other metabolic pathways including glycolysis (Wales and Hunter 1990) suggesting that metabolic differentiation proceeds normally in these embryos in direct contrast to the effects of glucose deprivation. Interestingly [14C]-amino acid incorporation into the embryo glycoprotein fraction was found to be insensitive to tunicamycin treatment (Wales and Hunter 1990). Moreover an earlier study using 3H-glucosamine showed that 40 % of the incorporated glycoprotein label in blastocysts is also tunicamycin insensitive (Surani 1979). Taken together these data suggest that N-linked glycosylation is unlikely to be the responsive mechanism at the heart of glucose activated blastocyst formation. The most likely downstream effector pathway in activating an adaptive response and metabolic differentiation is therefore most likely to be O-linked modification of cellular proteins by N-acetylglucosamine. Studies in bovine oocytes implicating O-linked glycosylation in oocyte developmental competence (Sutton-McDowall et al. 2004) were also consistent with this pathway playing a critical role in glucose signaling in early development and metabolic differentiation associated with blastocyst formation.
60
M. Pantaleon
In contrast to the Golgi associated N-glycosylation of membrane associated or secreted glycoproteins and other forms of glycosylation, O-linked glycosylation (or O-GlcNAcylation) is a nuclear and cytosolic modification of an increasingly diverse range of nuclear and cytoplasmic proteins (Holt and Hart 1986). It is a short (single residue) regulatory modification exhibiting properties similar to phosphorylation in contrast to other typical forms of glycosylation (Kreppel et al. 1997) which typically exhibit multiple, often branched glycosidic residues and are associated with cell-cell interactions. Target proteins are modified by a dynamic process involving cyclic addition and removal of a single moiety of N-acetylglucosamine (GlcNAc) to the hydroxyl groups of serine or threonine residues. This latter, terminal signal-transducing arm of the hexosamine biosynthetic pathway is now known as the hexosamine signaling pathway (HSP). In terms of high-energy compounds, the intracellular concentration of the activated donor, UDP-GlcNAc, is second only to ATP, making O-GlcNAcylation one of the most common cellular post-translational modifications (Hart et al. 2007).
3.6 Hexosamine Signalling: A Nutrient Response Pathway 3.6.1 The Enzymes The enzyme involved in the terminal transfer of the GlcNAc moiety to target proteins is the O-linked N-acetylglucosaminyltransferase (OGT) while the removal of GlcNAc is catalyzed by the β-selective N-acetylglucosaminidase ( O-GlcNAcase or OGA- annotated as meningioma-expressed antigen mgea5 and characterized as a hyaluronidase (Heckel et al. 1998)) (Wells and Hart 2003; Zachara et al. 2004). These form a single cooperatively regulated enzyme complex whose activity is exquisitely sensitive to UDP-GlcNAc and hence nutrient supply (Kreppel and Hart 1999). Both enzymes are products of single, highly conserved genes with differentially spliced variants in mammals. OGT maps to Xq13 on the mammalian X-chromosome and encodes three variants which differ in the length of their amino terminal tetratricopeptide (TRP) repeats (which vary from 3–12 repeats) and in their subcellular distribution: short OGT, mitochondrial OGT (mOGT) and nuclear/cytoplasmic OGT (ncOGT) (Lubas et al. 1997; Kreppel et al. 1997; Kreppel and Hart 1999; Lubas and Hanover 2000). The TRP domains produce a superhelical structure containing an asparagine ladder that is critical for protein recognition (Jinek et al. 2004). Alternate splicing of these differentially targeted variants therefore produces isoforms with varying TRPs thus allowing each isoform to modify a select subset of substrates, conferring on this single gene enzyme the ability to modify many targets (Lubas et al. 1997; Kreppel et al. 1997; Lazarus et al. 2006). OGA exists as two isoforms: the full-length isoform incorporates a histone acetyltransferase (HAT) domain in its C-terminus and a short isoform lacking the
3 The Role of Hexosamine Biosynthesis and Signaling in Early Development
61
C-terminal HAT domain. The shorter isoform is associated with lipid droplets and thought to have a role in lipid droplet assembly and mobilization, since it promotes proteasomal degradation of surface lipid droplet proteins (Keembiyehetty et al. 2011). The longer isoform, on the other hand, is present throughout the nucleus and cytoplasm. Moreover, this bifunctional nuclear/cytoplasmic OGA associates with OGT through specific domains into a single O-GlcNAczyme complex (Whisenhunt et al. 2006). Basal OGT and OGA activities are modulated by their own intrinsic O-GlcNAcylation status (Kreppel et al. 1997; Gloster and Vocadlo 2010) supporting the idea of cooperative activity between the two enzymes. Ultimately this cooperativity results in a finely balanced equilibrium whereby even small deviations in O-linked glycosylation impact on cellular signalling activities and therefore cellular homeostasis in response to a nutrient signal. OGT and OGA appear to regulate attachment and removal of O-GlcNAc in much the same way that kinases and phosphatases regulate phosphorylation. OGT was originally believed specifically to target serine/threonine residues on a number of nucleoplasmic proteins to alter their activity and stability in response to nutrient availability (Wells et al. 2001). Indeed, for a subset of proteins, OGT competes dynamically with protein kinases leading to the idea that such modification is functionally reciprocal to phosphorylation at the same sites (Kelly et al. 1993; Comer and Hart 2001; Chou et al. 1995). Whereas protein phosphorylation is achieved by the coordinate regulation of approximately 500 protein kinases and ~ 100 phosphatases (Forrest et al. 2003; Manning et al. 2002) the regulation of O-GlcNAcylation is achieved by the concerted action of the two highly conserved enzymes, OGT and OGA, highlighting their potential significance as regulators of kinase dependent cellular pathways.
3.6.2 The Targets Whilst this reciprocity certainly exists for a subset of proteins and is well characterized for some, such as the RNA polymerase II C-terminal domain (Comer and Hart 2001) and Sp1 (Kudlow 2006), for other proteins such as p53, O-GlcNAcylation can sterically hinder nearby phosphorylation sites and impact on their activity (Yang et al. 2006). The number of proteins whose activity and stability is modified by Olinked glycosylation is extensive and increasing and includes transcription factors, nuclear pore proteins, cytoskeletal components, metabolic enzymes as well as phosphatases and kinases that are known regulators of early developmental processes. There is in fact extensive cross-talk between OGT and the canonical nutrient sensing kinase cascades such as Insulin-Akt, MAPK, mTOR and AMPK (Hanover et al. 2010). Whilst O-linked glycosylation often results in decreased protein activity this is not a generalised phenomenon since in some proteins such as the global transcription factor Sp1 GlcNAc modification can increase protein activity (Goldberg et al. 2006; Wells and Hart 2003). Moreover, studies also show that O-linked glycosylation can both prevent and increase target susceptibility to proteasomal degradation (Han and Kudlow 1997). The 26S proteasome is a proteolytic organelle present
62
M. Pantaleon
in both the cytoplasm and nucleus responsible for degradation of proteins that are marked for turnover by polyubiquitination, thus controlling protein half-life, maintenance of metabolic proteins and elimination of damaged proteins. It is present throughout preimplantation development and its function is essential for early embryonic development (McCue et al. 2008). In the case of Sp1, under conditions of inadequate nutrition, O-GlcNAcylation of this transcription factor is reduced and this correlates with increased degradation that can be inhibited by proteasomal inhibition (Han and Kudlow 1997). Conversely hyperglycosylated Sp1 is protected from proteasomal degradation leading to the suggestion that O-GlcNAcylation of Sp1 may play a role as a nutritional checkpoint by generally reducing transcription in the absence of adequate nutrition (Han and Kudlow 1997). Consistent with this hypothesis, Sp1 promoter activation is reduced in an O-GlcNAc dependent manner as a result of an interaction between OGT and the transcriptional repressor mSin3A (Yang et al. 2002). Inhibition of the proteasome itself by O-linked glycosylation also couples protein turnover to metabolic state allowing cells to control the availability of amino acids and regulatory proteins (Zhang et al. 2003). O-GlcNAcylation and/or the enzyme OGT, can also interact with other posttranslational modifications such as ubiquitination (Guinez et al. 2008), which targets cellular proteins to the proteasome, and nitrosylation (Ryu and Do 2011) although the mechanisms by which these interactions occur are not clearly defined. Moreover, the OGT/OGA complex associates with histone deacetylases (HDACs) in nuclear transcriptional co-repression complexes to regulate transcription. Disruption of the enzyme complex regulating O-GlcNAcylation in these complexes interferes with transcriptional repression (Whisenhunt et al. 2006), thus implicating the enzyme complex in still broader roles. More recently OGT and O-GlcNAcylation have also become linked to the multi-faceted “histone code” with recent findings suggesting that all four core histones are modified by O-GlcNAc (Sakabe et al. 2010). The sites of O-GlcNAc-histone modification hint at a role in chromatin remodeling and add to a mounting body of evidence linking O-GlcNAc cycling to higher-order chromatin organization and epigenetic memory (Hanover et al. 2012). The identity and function of these targets and partners would suggest that HSP plays a critical role in controlling all aspects of cellular function ranging from protein activity and stability, fuel metabolism, cytoskeletal organization, cell growth, gene expression and differentiation as well as longer-term transcriptional regulation of developmental processes in response to nutrient availability.
3.7 O-GlcNAcylation in Development Unlike many other glycosyltransferases, OGT is a soluble protein predominantly found in the nucleus but associated with nuclear pores, mitochondria and cytoplasm of all tissues studied to date (Kreppel et al. 1997; Haltiwanger et al. 1992; Hanover et al. 2003). Consistent with this distribution, mouse embryos pulselabelled with 14C-glucosamine display 14C-label grains in the nucleolus, regions of
3 The Role of Hexosamine Biosynthesis and Signaling in Early Development
63
nucleolar-associated chromatin and the nuclear envelope (Acey et al. 1977). This is consistent with observations of predominantly nuclear but also cytoplasmic associated O-GlcNAc in early mouse embryos using antisera that recognize the β-OGlcNAc linkage (Pantaleon et al. 2010). Interestingly GLUT1 is also nuclear at this stage in development (Pantaleon et al. 2001a) and may provide a potential conduit for hexosamines to this location, although direct evidence for this is currently not available. Deletion of key enzymes involved in the HSP is found to be lethal in mammals (Boehmelt et al. 2000; O’Donnell et al. 2004; Shafi et al. 2000; Yang et al. 2012). OGT is an X-linked gene, and since most mouse embryonic stem cell lines are XY, early attempts to generate knockout mice were unsuccessful as they resulted in stem cell lethality (Shafi et al. 2000). This necessitated the use of Cre-recombinase technology to generate conditional, tissue-specific OGT deletion thus highlighting the fundamental importance of O-GlcNAcylation in embryo viability. Tissue-specific knockouts of OGT lead to profound changes in all cells types examined including thymocytes, and fibroblasts, resulting in T-cell apoptosis, and fibroblast growth arrest, respectively. Moreover, the neuronal specific knockout results in smaller pups with abberant locomotor activity. These animals did not nurse well and died about 10 days post-natally (O’Donnell et al. 2004). Targeted OGT deletion in the ooyte is also embryonic lethal resulting in peri-implantational (E5) abortion (O’Donnell et al. 2004) clearly indicating that preimplantation development is severely compromised. Taken together these outcomes suggest essential, pleiotropic roles for this enzyme during development.
3.8 The HSP: Sensor of an Adverse Environment? Given the ability of glucose signaling through this pathway to impart adaptive capacity to the developing embryo as discussed earlier and the functional diversity of the potential targets, the critical role played by this signaling pathway and OGlcNAc modification mostly likely relates to maintenance of cellular homeostasis in response to stress as part of a pro-survival signaling program. As such it may be viewed as a potential sensor of an adverse environment. Indeed, blocking hexosamine biosynthesis and hense the sensor component of this pathway eliminates stress responsiveness leading to a decrease in cellular survival in numerous systems (Zachara and Hart 2006). This is consistent with observations in the early embryo discussed earlier, where inhibition of GFPT ablates the ability of glucose to activate embryonic adaptive responses (Pantaleon et al. 2008). Blocking or reducing O-GlcNAcylation and thus cellular capacity to respond to stress also renders cells more vulnerable to stress and decreases cellular survival, whilst increasing levels of O-GlcNAcylation appear to have a protective effect by promoting cell survival (Zachara et al. 2004). Indeed, in somatic cells, cellular stress is intimately linked with changes in O-GlcNAcylation which, in response to multiple forms of cellular stress, appears
64
M. Pantaleon
to increase in a dose dependent manner in cultured cells (Zachara et al. 2004). Significantly, increasing levels of O-GlcNAc prior to (Zachara et al. 2004) or immediately following, cellular injury promotes survival whilst suppressing them sensitizes cells to death suggesting that this a key regulator of the cellular stress response (Zachara and Hart 2004). Stress induced O-GlcNAcylation promotes cell and tissue survival by regulating a multitude of biological processes which include but are not limited to, the phosphoisitide/Akt pathway, central to survival signalling, heat shock protein expression, calcium homeostasis and reactive oxygen species (ROS) generation as well as mitochondrial dynamics (Zachara and Hart 2006). These are all aimed at maintaining ATP levels, mitochondrial membrane potential and stabilizing redox state: all critical elements of early embryo viability (Lane and Gardner 2005). Ubiquitination and proteasomal degradation of p53, the most extensively studied tumor suppressor, is blocked by O-GlcNAcylation (Yang et al. 2006). This observation is relevant to survival signaling in the preimplantation embryo, which involves stimulation of prosurvival factor expression (mitochondrial Bcl-2) and ubiquitination/degradation of p53 via the phosphoisitide/Akt pathway (O’Neill et al. 2012). Moreover, O-GlcNAc has also been implicated in the regulation of mitochondrial Bcl-2 levels following ischemia-reperfusion injury (Champattanachai et al. 2008). It is not surprising then that the inability to regulate these processes is incompatible with embryo viability.
3.9 Perturbed O-GlcNAcylation and Embryo Development Whilst the capacity of the embryo to respond to stress by the HSP is essential to adaptation and survival, significant pertubations in O-GlcNAcylation arising from perturbed glucose flux through this pathway would have profound effects on development and may underlie the glucotoxic effects of hyperglycemia. Consistent with this hypothesis all treatments that perturb levels of O-GlcNAc either during oocyte maturation alone or through early development have a negative impact on early developmental outcomes. Increased activity of the HSP as a result of glucosamine supplementation during mouse and bovine oocyte IVM has detrimental effects on subsequent developmental competence (Sutton-McDowall et al. 2006; Schelbach et al. 2010). Moreover, pharmacological manipulation of the key enzymes of OGlcNAcylation, OGT and OGA as well as nutrient manipulation of nucleoplasmic O-GlcNAc levels in early mouse embryos lead to decreased rates of development, reduced cellular proliferation and increased levels of apoptosis thus confirming that this system is critical for embryonic cellular homeostasis (Pantaleon et al. 2010). Intriguingly, glucose deprivation does not cause a reduction in nucleoplasmic O-GlcNacylation as might have been anticipated given that there is no glucose flux through the HSP under these conditions. Indeed, levels of nucleoplasmic O-GlcNAc are increased significantly in response to embryo glucose deprivation (Pantaleon et al. 2010), consistent with data indicating that diverse forms of cellular stress which include heat stress, oxidative stress, ethanolic stress and osmotic stress all
3 The Role of Hexosamine Biosynthesis and Signaling in Early Development
65
elevate the levels of O-GlcNAcylation (Zachara and Hart 2004). Inhibition of OGT using BADGP was not able to reverse this level of O-GlcNAcylation despite its efficacy on glucosamine-induced O-GlcNAc upregulation. This suggests that in the complete absence of glucose, and hence limited production of UDP-GlcNAc and reduced OGT activity, O-GlcNAcase activity may also be inhibited, highlighting the cooperative nature of the two enzymes acting in concert to maintain O-GlcNAcylation at levels required for cellular survival, hence the reason why diverse stress signals can act to increase levels of O-GlcNAc. However despite this apparent maintenance of O-GlcNAc levels in response to various stressors, continued absence of glucose during early development and thus disruption of cycling of O-GlcNAc on and off cellular proteins, is incompatible with cellular and embryo viability.
3.9.1 Embryotoxic Effects of Hyperglycemia and O-GlcNAcylation Significantly, the toxicity associated with exposure of embryos to a hyperglycemic environment on a number of morphological parameters of early development including blastocyst formation (Fig. 3.2a), cell number and apoptosis is either
Fig. 3.2 a Effect of hyperglycemia in the presence and absence of the OGT inhibitor BADGP on mouse blastocyst formation. Zygotes were cultured from 18 h to 90 h post hCG in control KSOM, KSOM-glucose (-Glu) or KSOM- supplemented with 27 mM glucose (++Glu), or 27 mM glucose supplemented with 2 mM BADGP. 26.8 mM sucrose in KSOM (containing 0.2 mM glucose) was used as an osmotic control. Bars represent means ± SEM from three separate experiments each with 30 zygotes per treatment. Factorial ANOVA indicated no inter-experimental variation and no interaction ( P > 0.05). Means with the same superscript are statistically different. (a–g, P 0.05). Means were further analysed to determine differences between different treatments. Means with the same superscript are statistically different. (a–i, k: P
The Role of Hexosamine Biosynthesis and Signaling in Early Development Marie Pantaleon
Abstract Although the culture requirements and the metabolic profile of the preimplantation embryo have been thoroughly investigated since their first successful culture in a defined medium, now more than 50 years ago (Whitten, Nature 177:96, 1956), it is only recently that we have begun to appreciate the impact of the environment on life-course trajectory. The mechanisms involved in how nutrient availability may potentially modulate developmental potential are consequently not well defined. Originally thought of as simple energy substrates and biosynthetic precursors, the currently emerging paradigm suggests that nutrients may act in non-classical roles to impact on developmental potential. This is now an area of considerable activity thanks to pioneering epidemiological studies (Barker et al., BMJ 298:564– 7, 1989) that have led to the establishment of the Developmental Origins of Health and Disease (DoHAD) hypothesis and a whole new field of research activity. The period prior to implantation is of particular interest as this has been identified as a critical window of developmental sensitivity to environmental or nutrient stress (Fleming et al., Biol Reprod 71:1046–54, 2004a). This review seeks specifically to explore the pivotal role of glucose in early mouse development and the mechanisms by which it may impact on the cellular functions of the developing embryo. The emerging paradigm suggests that this humble hexose sugar may be at the heart of a rather sophisticated mechanism of cellular control that not only impacts on cellular proliferation and viability in the short term but on cellular memory through to the next generation. Keywords Early development · Gene regulation · Glucose signaling · Hexosamine biosynthesis · Hexosamine signaling · Embryo metabolism · Nutrient sensing · O-linked glycosylation · Embryo viability · Embryonic programming
M. Pantaleon () School of Biomedical Sciences, The University of Queensland, St Lucia, Brisbane, Qld 4072, Australia e-mail: [email protected] © Springer Science+Business Media New York 2015 H. J. Leese, D. R. Brison (eds.), Cell Signaling During Mammalian Early Embryo Development, Advances in Experimental Medicine and Biology 843, DOI 10.1007/978-1-4939-2480-6_3
53
54
M. Pantaleon
3.1 The Embryo and its Environment During the preimplantation period the embryo undergoes a series of critical developmental events, which set up its future growth trajectory. These include, one of two genome-wide demethylation episodes immediately following fertilization in both male and female pronuclei (Dean 2014), zygotic genome activation (ZGA), as well as morphologic and metabolic differentiative events that result in blastocyst formation. ZGA occurs during the second cell cycle in the mouse and follows a wave of maternal mRNA degradation initiated during the first cell cycle. This is a critical event in development in the absence of which, the embryo fails to develop further (Schultz 1993). DNA methylation marks are re-established concomitant with differentiation and blastocyst formation, thus forming an epigenetic barrier as cells become developmentally more restricted (Seisenberger et al. 2013). The erasure of the previous generation’s epigenetic signature and the re-setting of epigenetic marks at this stage based on nutrients available at this time, confers a flexibility to the preimplantation embryo that allows it to respond to environmental challenges which could otherwise have a major impact on it future growth trajectory. Although physically independent of the mother, the preimplantation embryo is nonetheless reliant on maternal physiology for a regulated supply of the requisite nutrients and removal of end products. Glucose levels are low in the oviduct during the early cleavage phase but increase as the embryo enters the uterus to levels more in line with altered embryonic needs following compaction as the embryo increases both its reliance on glucose as an energy substrate and its ability to respond to exogenous growth factors (Kaye 1997). Oviductal and uterine epithelial secretions provide the early embryo with essential nutrients, electrolytes, macromolecules and growth factors required for optimal growth prior to implantation. Whilst it has long been accepted that these factors potentially can influence development, they have classically been viewed as facilitative rather than obligatory determinants of early development (Leese 1995) since preimplantation embryos can be cultured to the blastocyst stage in relatively simple defined media. Nonetheless, the contribution of nutrients and maternal growth factors to optimal growth can be significant as demonstrated by the difference in growth rates between in vivo-matured embryos and their in vitro counterparts (Bowman and McLaren 1970; Kaye and Gardner 1999). Indeed, metabolic control at this stage can be flexible and adaptive, as demonstrated by numerous studies inspired by the DOHaD hypothesis in various mammalian animal models (for review see Fleming et al. 2012), leaving its mark and setting growth trajectories based on the maternal nutrient environment and in anticipation of future nutrient availability. When the postnatal nutrient environment conflicts with this early setting and growth trajectories are unmatched there is the potential for disorders in the offspring (Fleming et al. 2004b). The critical question is how these adaptive mechanisms are set, what are the triggers and effector pathways that potentially lead to these adaptive responses that control growth and metabolism not only in utero but into the next generation. The mechanisms by which early embryos sense and respond to their nutrient
3 The Role of Hexosamine Biosynthesis and Signaling in Early Development
55
environment in order to bring about a biosynthetic, proliferative and potentially epigenetic programme of events, in an attempt to maintain developmental potential are therefore central to understanding this phenomenon.
3.2 A Role for Glucose in Early Development? Early experimentation aimed at understanding the nutrient requirements of preimplantation embryos established that pre-compacted embryos from a number of species, unlike other somatic cells, are unable to utilise glucose as an energy substrate. Instead they rely on monocarboxylates, pyruvate (especially for the first division) and lactate to support early cleavage (Brinster 1965; Brinster and Thomson 1966; Wales and Whittingham 1967; Biggers et al. 1967). Experiments through the late 1980s and early 1990s aimed at developing successful culture formulations for assisted reproductive technologies, further claimed that glucose was in fact inhibitory during cleavage stage development in a number of species (Schini and Bavister 1988; Chatot et al. 1989; Diamond et al. 1991; Lawitts and Biggers 1991; De Hertogh et al. 1991; Scott and Whittingham 1996; Thompson et al. 1992; Conaghan et al. 1993), (though this was later disputed by (Summers and Biggers 2003) but that nutrients such as glutamine exerted considerable beneficial effects (Chatot et al. 1989). Despite strong data indicating that glucose does not in fact support development energetically prior to compaction and may possibly be inhibitory, a subset of studies suggested that glucose was nonetheless required at some stage, raising considerable debate and confusion about the role of glucose during cleavage stage development and at what appeared to be a temporal sensitivity to that requirement (Leppens-Luisier and Sakkas 1997; Brown and Whittingham 1991, 1992; Chatot et al. 1989, 1994). Whilst glucose is not preferentially utilized for energy during early cleavage, it is nonetheless utilised via the pentose phosphate pathway to produce ribose-5-phosphate for nucleic acid synthesis and also for the generation of reducing power in the form of NADPH (O’Fallon and Wright 1986). Glucose becomes the embryo’s fuel of choice over pyruvate following compaction at the 8-cell stage in the mouse (Leese and Barton 1984), when it is required to fuel the high energy requiring Na + /K + ATPases that facilitate cavitation (Watson and Barcroft 2001). It would appear therefore that this metabolic differentiation helps facilitate the morphological changes associated with blastocyst formation. This acquisition of metabolic competence to utilize glucose as an energy substrate correlates with the apical expression of the high-affinity high-capacity glucose transporter GLUT3 or SLC2A3 at compaction (Pantaleon et al. 1997). Inhibition of GLUT3 using antisense oligonucleotides not only reduced blastocyst glucose transport as anticipated, but inhibited blastocyst formation (Pantaleon et al. 1997) linking blastocyst formation with metabolic differentiation.
56
M. Pantaleon
This was intriguing because notwithstanding a preference for glucose post compaction, mouse and human embryos cultured from the 2-cell or the 2–4-cell stage respectively in the absence of glucose could adapt to its absence by increasing utilization of pyruvate to form blastocysts (Martin and Leese 1995; Conaghan et al. 1993). Moreover these embryos already possessed, or developed, the capacity to transport glucose at high rates should it become available (Martin and Leese 1995).
3.3 Glucose Primes Embryos to Adapt to Their Environment Unlike 2-cell embryos, newly fertilized zygotes collected and cultured without glucose do not progress beyond the morula stage. Moreover, these embryos do not develop the capacity to transport glucose should it become available, or the adaptive capacity to cope with its absence (Martin and Leese 1995). Thus, zygotes cultured under these conditions display reduced proliferative capacity and higher levels of apoptosis (Pantaleon et al. 2008). Moreover they undergo oxidative stress and exhibit elevated levels of reactive oxygen species (ROS) despite a supply of sufficient monocarboxylates (Jansen et al. 2009). This suggests that in the complete absence of glucose the embryo is unable to respond to environmental stress, making it more susceptible to developmental arrest and cell death. Cellular ability to respond to multiple forms of stress is in fact associated with the ability to increase glucose uptake rapidly. This has classically been thought to be an essential survival response of adaptive importance in a healthy organism. Remarkably, brief exposure to glucose prior to the morula stage was sufficient to “prime” these naïve embryos and permit subsequent blastocyst formation (Brown and Whittingham 1991, 1992; Chatot et al. 1994). So despite an inability to derive metabolic energy from glucose prior to compaction, these experiments suggested the presence of adaptive responses that were nonetheless activated by glucose availability (Martin and Leese 1995). The ability of this brief glucose exposure to modulate utilization of pyruvate and glucose in response to an adverse environment is linked to activation of GLUT3 expression and maintenance of expression of the transporter for pyruvate and lactate, H + monocarboxylate cotransporter 1 (MCT1) post compaction (Pantaleon et al. 2008; Jansen et al. 2006). Non-metabolisable glucose analogues such as 3-O methyl glucose and 2-deoxy glucose were unable to replace glucose and stimulate blastocyst formation (Chatot et al. 1994) suggesting that the role of glucose is to provide a transcriptionally active downstream metabolite. Moreover, glutamine is also unable to substitute for glucose (Brown and Whittingham 1991), although it is noteworthy that its incorporation increases 10-fold between the zygote and morula stages (Brinster 1971) facilitated by a 3-fold increase in transport capacity between the 2-cell and blastocyst stages (Lewis and Kaye 1992). Fructose was also shown to facilitate blastocyst formation but not to the same degree as glucose although it did appear to have a beneficial effect on cell number (Chatot et al. 1994). The fact that hexokinase and
3 The Role of Hexosamine Biosynthesis and Signaling in Early Development
57
glucose phosphate isomerase are expressed in mouse oocytes (el Mouatassim et al. 1999) and active throughout cleavage stage development in both mouse and human spare embryos (Martin et al. 1993; West et al. 1989) suggested that glucose conversion to fructose-6-phosphate was involved. As an aside, the reduced efficacy of fructose to stimulate blastocyst formation may be due to reduced uptake from the external environment rather than metabolism per se. GLUT1, the glucose transporter expressed during cleavage stage development in the mouse (Pantaleon et al. 2001a; Pantaleon and Kaye 1998), transports glucose and other hexoses with varying efficacy but not fructose which is transported by the cytochalasin B-insensitive Class II mammalian monosaccharide carriers, GLUT2 and GLUT5 (Cura and Carruthers 2012). Whilst there are some reports of GLUT2 expression in later stage blastocysts (Aghayan et al. 1992; Hogan et al. 1991) expression of neither transporter has been described in mouse cleavage stage embryos to date. Moreover, functional studies, which failed to demonstrate residual saturable glucose transport activity following cytochalasin B inhibition (Gardner and Kaye 1995), do not support expression of any cytochalasin B-insensitive Class II carriers through preimplantation development in the mouse and thus an efficient conduit for fructose uptake from the external environment.
3.4 Hexosamine Biosynthesis: An Embryonic NutrientSensing Pathway The increased rate of glutamine incorporation and transport capacity during the cleavage stages (Brinster 1971; Lewis and Kaye 1992) coupled with its beneficial effect on early development as discussed above, led us to question whether glucose metabolism through hexosamine biosynthesis might be involved in conferring on the embryo the adaptive capacity to develop in an adverse environment. In support of this hypothesis, early experiments demonstrated that glucosamine, can substitute for glucose and facilitate both blastocyst formation and GLUT3 expression (Pantaleon et al. 2001b). Over 90 % of 14C-1-glucosamine initially taken up by somatic cells in culture is found in the UDP-N-acetylhexosamine pool (Kornfeld and Ginsburg 1966) suggesting that glucose metabolism through to this activated sugarnucleotide may be implicated in this function for glucose in setting up adaptive responses at this stage in development. Significantly, glucose flux through this pathway appears to be restricted to glucose transported by GLUT1 (Buse et al. 1996). Conclusive evidence to support a role for the hexosamine biosynthetic pathway (HBP) in glucose sensing in the early embryo, came from pharmacological studies. Inhibition of the first and rate-limiting enzyme of the pathway, glutamine: fructose-6-phosphate amidotransferase (GFPT) using the competitive analogue azaserine inhibited glucose activated blastocyst formation and GLUT3 expression (Pantaleon et al. 2008). Moreover, these perturbations could be rescued with glucosamine supplementation, which feeds into hexosamine biosynthesis below the level of GFPT (Fig. 3.1), thus supporting the hypothesis that the HBP may be the
58
M. Pantaleon
Fig. 3.1 Glucose is phosphorylated upon cellular entry and isomerized to fructose-6-phosphate (F-6-P) prior to utilization through glycolysis and pentose phosphate pathway (PPP). Up to 5 % of F-6-P however is converted to glucosamine-6-phosphate (GlcN-6-P) by the rate-limiting enzyme GFPT and subsequently acetylated by glucosamine-6-phosphate acetyl transferase (Emeg32) and converted into UDP-N-GlcNAc via the UDP-GlcNAc pyrophosphorylase (uap1) for use as a donor substrate in multiple biosynthetic reactions including the O-linked modification of a diverse group of nucleoplasmic proteins by N-acetylglucosamine (GlcNAc). Addition of GlcNAc to ser/ thr residues of target proteins is catalysed by O-GlcNAc transferase (OGT) and its removal by an O-GlcNAc-selective β-N-acetylglucosaminidase ( O-GlcNAcase or OGA). These reactions can be manipulated using specific inhibitors, BADGP and PUGNAc to evaluate the role of O-linked GlcNAcylation systematically, whilst the rate-limiting step in UDP-GlcNAc formation catalysed by GFPT, can be manipulated experimentally using the glucosamine analogue azaserine
pathway through which glucose signals impact on early embryos (Pantaleon et al. 2008). The importance of this pathway is highlighted by loss/gain of function models for two of the key enzymes of this pathway. Thus, homozygous null mutants of glucosamine-6-phosphate acetyl transferase (also known as Emeg32) abort at E7.5 with pronounced developmental delays (Boehmelt et al. 2000), whilst overexpression of GFPT leads to peripheral insulin resistance (Hebert et al. 1996). Indeed, the HBP has been shown to be important in the development of insulin resistance in somatic cells (Yang et al. 2008) which underscores the importance of this pathway as a nutrient sensor (Marshall et al. 1991; Traxinger and Marshall 1991; Hawkins et al. 1996; Wang et al. 1998). Unlike other nutrient-response systems that respond to specific stimuli, for example AMPK, which responds to the AMP/ATP ratio, and mTOR which responds
3 The Role of Hexosamine Biosynthesis and Signaling in Early Development
59
to amino acid concentrations as well as to the insulin/PI3K system, the HBP integrates several nutrient sources and nutrient sensing pathways, making it acutely sensitive to all nutrient flux (Hanover et al. 2010). Indeed, the HBP integrates the metabolism of carbohydrates (glucose), amino acids (glutamine), fat (acetyl CoA) and nucleotides (uridine-diphosphate) with energy charge (ATP) in the synthesis of UDP-GlcNAc, making this activated sugar-nucleotide potentially the most general sensor of cellular nutritional status (Hanover et al. 2010). UDP-GlcNAc then acts as the donor substrate for a number of biosynthetic reactions including GPI lipid anchor biosynthesis, sugar nucleotide, glycoside and ganglioside biosynthesis and N-linked glycosylation for the production of glucosaminoglycans, glycolipids and membrane and secretory glycoproteins. Additionally UDP-GlcNAc acts as a precursor to modify regulatory proteins post-translationally in the nucleus and cytoplasm through O-linked glycosylation with N-acetyl glucosamine ( O-GlcNAcylation) (Fig. 3.1). Not surprisingly, the HBP is highly active during oocyte maturation where it is utilised for the biosynthesis of complex glycan structures such as hyaluronic acid in the extra cellular matrix of cumulus-oocyte complexes (COCs). Indeed, it is estimated that 25 % of glucose consumed by gonadotropin stimulated bovine COCs is incorporated into the extracellular matrix via the hexosamine biosynthetic pathway to facilitate cumulus expansion (Sutton-McDowall et al. 2004).
3.5 The Response Path: N-Linked Vs. O-Linked Glycosylation? Early studies in embryos using tunicamycin, showed that this specific inhibitor of N-linked glycosylation suppressed glucose incorporation into glycoproteins and inhibits compaction (Surani 1979; Wales and Hunter 1990). However, it did not affect other metabolic pathways including glycolysis (Wales and Hunter 1990) suggesting that metabolic differentiation proceeds normally in these embryos in direct contrast to the effects of glucose deprivation. Interestingly [14C]-amino acid incorporation into the embryo glycoprotein fraction was found to be insensitive to tunicamycin treatment (Wales and Hunter 1990). Moreover an earlier study using 3H-glucosamine showed that 40 % of the incorporated glycoprotein label in blastocysts is also tunicamycin insensitive (Surani 1979). Taken together these data suggest that N-linked glycosylation is unlikely to be the responsive mechanism at the heart of glucose activated blastocyst formation. The most likely downstream effector pathway in activating an adaptive response and metabolic differentiation is therefore most likely to be O-linked modification of cellular proteins by N-acetylglucosamine. Studies in bovine oocytes implicating O-linked glycosylation in oocyte developmental competence (Sutton-McDowall et al. 2004) were also consistent with this pathway playing a critical role in glucose signaling in early development and metabolic differentiation associated with blastocyst formation.
60
M. Pantaleon
In contrast to the Golgi associated N-glycosylation of membrane associated or secreted glycoproteins and other forms of glycosylation, O-linked glycosylation (or O-GlcNAcylation) is a nuclear and cytosolic modification of an increasingly diverse range of nuclear and cytoplasmic proteins (Holt and Hart 1986). It is a short (single residue) regulatory modification exhibiting properties similar to phosphorylation in contrast to other typical forms of glycosylation (Kreppel et al. 1997) which typically exhibit multiple, often branched glycosidic residues and are associated with cell-cell interactions. Target proteins are modified by a dynamic process involving cyclic addition and removal of a single moiety of N-acetylglucosamine (GlcNAc) to the hydroxyl groups of serine or threonine residues. This latter, terminal signal-transducing arm of the hexosamine biosynthetic pathway is now known as the hexosamine signaling pathway (HSP). In terms of high-energy compounds, the intracellular concentration of the activated donor, UDP-GlcNAc, is second only to ATP, making O-GlcNAcylation one of the most common cellular post-translational modifications (Hart et al. 2007).
3.6 Hexosamine Signalling: A Nutrient Response Pathway 3.6.1 The Enzymes The enzyme involved in the terminal transfer of the GlcNAc moiety to target proteins is the O-linked N-acetylglucosaminyltransferase (OGT) while the removal of GlcNAc is catalyzed by the β-selective N-acetylglucosaminidase ( O-GlcNAcase or OGA- annotated as meningioma-expressed antigen mgea5 and characterized as a hyaluronidase (Heckel et al. 1998)) (Wells and Hart 2003; Zachara et al. 2004). These form a single cooperatively regulated enzyme complex whose activity is exquisitely sensitive to UDP-GlcNAc and hence nutrient supply (Kreppel and Hart 1999). Both enzymes are products of single, highly conserved genes with differentially spliced variants in mammals. OGT maps to Xq13 on the mammalian X-chromosome and encodes three variants which differ in the length of their amino terminal tetratricopeptide (TRP) repeats (which vary from 3–12 repeats) and in their subcellular distribution: short OGT, mitochondrial OGT (mOGT) and nuclear/cytoplasmic OGT (ncOGT) (Lubas et al. 1997; Kreppel et al. 1997; Kreppel and Hart 1999; Lubas and Hanover 2000). The TRP domains produce a superhelical structure containing an asparagine ladder that is critical for protein recognition (Jinek et al. 2004). Alternate splicing of these differentially targeted variants therefore produces isoforms with varying TRPs thus allowing each isoform to modify a select subset of substrates, conferring on this single gene enzyme the ability to modify many targets (Lubas et al. 1997; Kreppel et al. 1997; Lazarus et al. 2006). OGA exists as two isoforms: the full-length isoform incorporates a histone acetyltransferase (HAT) domain in its C-terminus and a short isoform lacking the
3 The Role of Hexosamine Biosynthesis and Signaling in Early Development
61
C-terminal HAT domain. The shorter isoform is associated with lipid droplets and thought to have a role in lipid droplet assembly and mobilization, since it promotes proteasomal degradation of surface lipid droplet proteins (Keembiyehetty et al. 2011). The longer isoform, on the other hand, is present throughout the nucleus and cytoplasm. Moreover, this bifunctional nuclear/cytoplasmic OGA associates with OGT through specific domains into a single O-GlcNAczyme complex (Whisenhunt et al. 2006). Basal OGT and OGA activities are modulated by their own intrinsic O-GlcNAcylation status (Kreppel et al. 1997; Gloster and Vocadlo 2010) supporting the idea of cooperative activity between the two enzymes. Ultimately this cooperativity results in a finely balanced equilibrium whereby even small deviations in O-linked glycosylation impact on cellular signalling activities and therefore cellular homeostasis in response to a nutrient signal. OGT and OGA appear to regulate attachment and removal of O-GlcNAc in much the same way that kinases and phosphatases regulate phosphorylation. OGT was originally believed specifically to target serine/threonine residues on a number of nucleoplasmic proteins to alter their activity and stability in response to nutrient availability (Wells et al. 2001). Indeed, for a subset of proteins, OGT competes dynamically with protein kinases leading to the idea that such modification is functionally reciprocal to phosphorylation at the same sites (Kelly et al. 1993; Comer and Hart 2001; Chou et al. 1995). Whereas protein phosphorylation is achieved by the coordinate regulation of approximately 500 protein kinases and ~ 100 phosphatases (Forrest et al. 2003; Manning et al. 2002) the regulation of O-GlcNAcylation is achieved by the concerted action of the two highly conserved enzymes, OGT and OGA, highlighting their potential significance as regulators of kinase dependent cellular pathways.
3.6.2 The Targets Whilst this reciprocity certainly exists for a subset of proteins and is well characterized for some, such as the RNA polymerase II C-terminal domain (Comer and Hart 2001) and Sp1 (Kudlow 2006), for other proteins such as p53, O-GlcNAcylation can sterically hinder nearby phosphorylation sites and impact on their activity (Yang et al. 2006). The number of proteins whose activity and stability is modified by Olinked glycosylation is extensive and increasing and includes transcription factors, nuclear pore proteins, cytoskeletal components, metabolic enzymes as well as phosphatases and kinases that are known regulators of early developmental processes. There is in fact extensive cross-talk between OGT and the canonical nutrient sensing kinase cascades such as Insulin-Akt, MAPK, mTOR and AMPK (Hanover et al. 2010). Whilst O-linked glycosylation often results in decreased protein activity this is not a generalised phenomenon since in some proteins such as the global transcription factor Sp1 GlcNAc modification can increase protein activity (Goldberg et al. 2006; Wells and Hart 2003). Moreover, studies also show that O-linked glycosylation can both prevent and increase target susceptibility to proteasomal degradation (Han and Kudlow 1997). The 26S proteasome is a proteolytic organelle present
62
M. Pantaleon
in both the cytoplasm and nucleus responsible for degradation of proteins that are marked for turnover by polyubiquitination, thus controlling protein half-life, maintenance of metabolic proteins and elimination of damaged proteins. It is present throughout preimplantation development and its function is essential for early embryonic development (McCue et al. 2008). In the case of Sp1, under conditions of inadequate nutrition, O-GlcNAcylation of this transcription factor is reduced and this correlates with increased degradation that can be inhibited by proteasomal inhibition (Han and Kudlow 1997). Conversely hyperglycosylated Sp1 is protected from proteasomal degradation leading to the suggestion that O-GlcNAcylation of Sp1 may play a role as a nutritional checkpoint by generally reducing transcription in the absence of adequate nutrition (Han and Kudlow 1997). Consistent with this hypothesis, Sp1 promoter activation is reduced in an O-GlcNAc dependent manner as a result of an interaction between OGT and the transcriptional repressor mSin3A (Yang et al. 2002). Inhibition of the proteasome itself by O-linked glycosylation also couples protein turnover to metabolic state allowing cells to control the availability of amino acids and regulatory proteins (Zhang et al. 2003). O-GlcNAcylation and/or the enzyme OGT, can also interact with other posttranslational modifications such as ubiquitination (Guinez et al. 2008), which targets cellular proteins to the proteasome, and nitrosylation (Ryu and Do 2011) although the mechanisms by which these interactions occur are not clearly defined. Moreover, the OGT/OGA complex associates with histone deacetylases (HDACs) in nuclear transcriptional co-repression complexes to regulate transcription. Disruption of the enzyme complex regulating O-GlcNAcylation in these complexes interferes with transcriptional repression (Whisenhunt et al. 2006), thus implicating the enzyme complex in still broader roles. More recently OGT and O-GlcNAcylation have also become linked to the multi-faceted “histone code” with recent findings suggesting that all four core histones are modified by O-GlcNAc (Sakabe et al. 2010). The sites of O-GlcNAc-histone modification hint at a role in chromatin remodeling and add to a mounting body of evidence linking O-GlcNAc cycling to higher-order chromatin organization and epigenetic memory (Hanover et al. 2012). The identity and function of these targets and partners would suggest that HSP plays a critical role in controlling all aspects of cellular function ranging from protein activity and stability, fuel metabolism, cytoskeletal organization, cell growth, gene expression and differentiation as well as longer-term transcriptional regulation of developmental processes in response to nutrient availability.
3.7 O-GlcNAcylation in Development Unlike many other glycosyltransferases, OGT is a soluble protein predominantly found in the nucleus but associated with nuclear pores, mitochondria and cytoplasm of all tissues studied to date (Kreppel et al. 1997; Haltiwanger et al. 1992; Hanover et al. 2003). Consistent with this distribution, mouse embryos pulselabelled with 14C-glucosamine display 14C-label grains in the nucleolus, regions of
3 The Role of Hexosamine Biosynthesis and Signaling in Early Development
63
nucleolar-associated chromatin and the nuclear envelope (Acey et al. 1977). This is consistent with observations of predominantly nuclear but also cytoplasmic associated O-GlcNAc in early mouse embryos using antisera that recognize the β-OGlcNAc linkage (Pantaleon et al. 2010). Interestingly GLUT1 is also nuclear at this stage in development (Pantaleon et al. 2001a) and may provide a potential conduit for hexosamines to this location, although direct evidence for this is currently not available. Deletion of key enzymes involved in the HSP is found to be lethal in mammals (Boehmelt et al. 2000; O’Donnell et al. 2004; Shafi et al. 2000; Yang et al. 2012). OGT is an X-linked gene, and since most mouse embryonic stem cell lines are XY, early attempts to generate knockout mice were unsuccessful as they resulted in stem cell lethality (Shafi et al. 2000). This necessitated the use of Cre-recombinase technology to generate conditional, tissue-specific OGT deletion thus highlighting the fundamental importance of O-GlcNAcylation in embryo viability. Tissue-specific knockouts of OGT lead to profound changes in all cells types examined including thymocytes, and fibroblasts, resulting in T-cell apoptosis, and fibroblast growth arrest, respectively. Moreover, the neuronal specific knockout results in smaller pups with abberant locomotor activity. These animals did not nurse well and died about 10 days post-natally (O’Donnell et al. 2004). Targeted OGT deletion in the ooyte is also embryonic lethal resulting in peri-implantational (E5) abortion (O’Donnell et al. 2004) clearly indicating that preimplantation development is severely compromised. Taken together these outcomes suggest essential, pleiotropic roles for this enzyme during development.
3.8 The HSP: Sensor of an Adverse Environment? Given the ability of glucose signaling through this pathway to impart adaptive capacity to the developing embryo as discussed earlier and the functional diversity of the potential targets, the critical role played by this signaling pathway and OGlcNAc modification mostly likely relates to maintenance of cellular homeostasis in response to stress as part of a pro-survival signaling program. As such it may be viewed as a potential sensor of an adverse environment. Indeed, blocking hexosamine biosynthesis and hense the sensor component of this pathway eliminates stress responsiveness leading to a decrease in cellular survival in numerous systems (Zachara and Hart 2006). This is consistent with observations in the early embryo discussed earlier, where inhibition of GFPT ablates the ability of glucose to activate embryonic adaptive responses (Pantaleon et al. 2008). Blocking or reducing O-GlcNAcylation and thus cellular capacity to respond to stress also renders cells more vulnerable to stress and decreases cellular survival, whilst increasing levels of O-GlcNAcylation appear to have a protective effect by promoting cell survival (Zachara et al. 2004). Indeed, in somatic cells, cellular stress is intimately linked with changes in O-GlcNAcylation which, in response to multiple forms of cellular stress, appears
64
M. Pantaleon
to increase in a dose dependent manner in cultured cells (Zachara et al. 2004). Significantly, increasing levels of O-GlcNAc prior to (Zachara et al. 2004) or immediately following, cellular injury promotes survival whilst suppressing them sensitizes cells to death suggesting that this a key regulator of the cellular stress response (Zachara and Hart 2004). Stress induced O-GlcNAcylation promotes cell and tissue survival by regulating a multitude of biological processes which include but are not limited to, the phosphoisitide/Akt pathway, central to survival signalling, heat shock protein expression, calcium homeostasis and reactive oxygen species (ROS) generation as well as mitochondrial dynamics (Zachara and Hart 2006). These are all aimed at maintaining ATP levels, mitochondrial membrane potential and stabilizing redox state: all critical elements of early embryo viability (Lane and Gardner 2005). Ubiquitination and proteasomal degradation of p53, the most extensively studied tumor suppressor, is blocked by O-GlcNAcylation (Yang et al. 2006). This observation is relevant to survival signaling in the preimplantation embryo, which involves stimulation of prosurvival factor expression (mitochondrial Bcl-2) and ubiquitination/degradation of p53 via the phosphoisitide/Akt pathway (O’Neill et al. 2012). Moreover, O-GlcNAc has also been implicated in the regulation of mitochondrial Bcl-2 levels following ischemia-reperfusion injury (Champattanachai et al. 2008). It is not surprising then that the inability to regulate these processes is incompatible with embryo viability.
3.9 Perturbed O-GlcNAcylation and Embryo Development Whilst the capacity of the embryo to respond to stress by the HSP is essential to adaptation and survival, significant pertubations in O-GlcNAcylation arising from perturbed glucose flux through this pathway would have profound effects on development and may underlie the glucotoxic effects of hyperglycemia. Consistent with this hypothesis all treatments that perturb levels of O-GlcNAc either during oocyte maturation alone or through early development have a negative impact on early developmental outcomes. Increased activity of the HSP as a result of glucosamine supplementation during mouse and bovine oocyte IVM has detrimental effects on subsequent developmental competence (Sutton-McDowall et al. 2006; Schelbach et al. 2010). Moreover, pharmacological manipulation of the key enzymes of OGlcNAcylation, OGT and OGA as well as nutrient manipulation of nucleoplasmic O-GlcNAc levels in early mouse embryos lead to decreased rates of development, reduced cellular proliferation and increased levels of apoptosis thus confirming that this system is critical for embryonic cellular homeostasis (Pantaleon et al. 2010). Intriguingly, glucose deprivation does not cause a reduction in nucleoplasmic O-GlcNacylation as might have been anticipated given that there is no glucose flux through the HSP under these conditions. Indeed, levels of nucleoplasmic O-GlcNAc are increased significantly in response to embryo glucose deprivation (Pantaleon et al. 2010), consistent with data indicating that diverse forms of cellular stress which include heat stress, oxidative stress, ethanolic stress and osmotic stress all
3 The Role of Hexosamine Biosynthesis and Signaling in Early Development
65
elevate the levels of O-GlcNAcylation (Zachara and Hart 2004). Inhibition of OGT using BADGP was not able to reverse this level of O-GlcNAcylation despite its efficacy on glucosamine-induced O-GlcNAc upregulation. This suggests that in the complete absence of glucose, and hence limited production of UDP-GlcNAc and reduced OGT activity, O-GlcNAcase activity may also be inhibited, highlighting the cooperative nature of the two enzymes acting in concert to maintain O-GlcNAcylation at levels required for cellular survival, hence the reason why diverse stress signals can act to increase levels of O-GlcNAc. However despite this apparent maintenance of O-GlcNAc levels in response to various stressors, continued absence of glucose during early development and thus disruption of cycling of O-GlcNAc on and off cellular proteins, is incompatible with cellular and embryo viability.
3.9.1 Embryotoxic Effects of Hyperglycemia and O-GlcNAcylation Significantly, the toxicity associated with exposure of embryos to a hyperglycemic environment on a number of morphological parameters of early development including blastocyst formation (Fig. 3.2a), cell number and apoptosis is either
Fig. 3.2 a Effect of hyperglycemia in the presence and absence of the OGT inhibitor BADGP on mouse blastocyst formation. Zygotes were cultured from 18 h to 90 h post hCG in control KSOM, KSOM-glucose (-Glu) or KSOM- supplemented with 27 mM glucose (++Glu), or 27 mM glucose supplemented with 2 mM BADGP. 26.8 mM sucrose in KSOM (containing 0.2 mM glucose) was used as an osmotic control. Bars represent means ± SEM from three separate experiments each with 30 zygotes per treatment. Factorial ANOVA indicated no inter-experimental variation and no interaction ( P > 0.05). Means with the same superscript are statistically different. (a–g, P 0.05). Means were further analysed to determine differences between different treatments. Means with the same superscript are statistically different. (a–i, k: P
