Oxidative Stress, Intrauterine Growth Restriction, and Developmental Programming of Type 2 Diabetes

Abstract

Intrauterine growth restriction (IUGR) leads to reduced birth weight and the development of metabolic diseases such as Type 2 diabetes in adulthood. Mitochondria dysfunction and oxidative stress are commonly found in key tissues (pancreatic islets, liver, and skeletal muscle) of IUGR individuals. In this review, we explore the role of oxidative stress in IUGR-associated diabetes etiology.

Introduction

Intrauterine growth restriction (IUGR) is a common complication of pregnancy, affecting 10–15% of pregnancies. IUGR results in decreased birth weight and the development of diseases such as diabetes, obesity, hypertension, and cardiovascular disease later in life (48, 49, 73, 125, 127, 160). David Barker and Nicholas Hales coined the term “fetal origins of adult disease” based on their studies demonstrating a relationship between low birth weight for gestational age (SGA) and the later development of cardiovascular disease, impaired glucose tolerance, and Type 2 diabetes (T2D) (6, 7, 48, 49). Early studies showed that being born SGA was associated with an odds ratio of >5 for impaired glucose tolerance in men ~64 yr of age (50). Later studies, however, accounted for confounding factors such as obesity, family history of diabetes, cigarette smoking, and socioeconomic status, and showed an odds ratio for Type 2 diabetes of 1.8 in both men (25) and women (129) born with low birth weight. Although these two studies suggest that there are not sex differences in the incidence of Type 2 diabetes in individuals who were born SGA, other studies do show sex differences in incidence of metabolic disorders. For example, females exposed to the Dutch Famine in utero have increased body mass index (BMI), waist circumference, and adiposity as well as a disrupted lipid profile, but not males (93, 126, 145). In contrast, animal models preferentially show a male sex bias in adult metabolic disease associated with IUGR (26). Because men were initially studied and animal models preferentially show a male sex bias in adult metabolic disease associated with IUGR, the vast majority of research in this field has only investigated developmental programming of diabetes in males, and it is these studies in males that are presented in this review. More studies are needed in both humans and animal models to understand how changes to the early life environment impact differently on the long-term health of male and female individuals.

Finally, most human studies use reduced birth weight at term as a proxy for poor intrauterine growth, and it is not known whether these low birth weight individuals are truly growth restricted. However, for the sake of simplicity, we will use the term IUGR.

The period from conception to birth is a time of cellular replication and differentiation, functional maturation of organ systems, and rapid growth. These processes are very sensitive to alterations in nutrient availability, and an abnormal intrauterine metabolic milieu can have long-lasting effects on the offspring. This fetal “malnutrition” may be caused by extreme maternal undernutrition, such as the case of famine (125, 146), or by uteroplacental insufficiency. Uteroplacental insufficiency, characterized by attenuated placental perfusion, is the primary cause of IUGR in the developed world and is commonly associated with morbid obesity (123), maternal smoking (113), diabetes (143), and hypertension (128, 143). Although the mechanism of action is less clear, developmental exposure to environmental contaminants such as bisphenol A (22, 116) and organochlorines is also associated with decreased birth weight (42, 110), and there is some evidence indicating that mitochondrial dysfunction is involved (5).

Oxidative stress is consistently associated with IUGR (46, 70, 98, 117). Oxidative stress results from redox imbalance caused by either overproduction of reactive oxygen species (ROS) or diminished antioxidant activities. Studies in pregnant mothers of growth-restricted fetuses have shown increased plasma protein carbonyls (137) and malondialdehyde (MDA) (70), and reduced plasma total antioxidant capacity (137), indicating maternal oxidative stress. A longitudinal case-control study showed increased urinary 8-oxo-7,8-dihydro-2′-deoxyguanosine (8-oxodG) at 12 and 28 wk of age in mothers whose offspring would become IUGR, and 8-oxodG levels decreased in cases and controls from 12 to 28 wk, suggesting oxidative stress precedes clinical features of IUGR and thus may, to a degree, be causative. The cause of maternal oxidative stress in IUGR is still unknown. However, placental hypoxia at the intervillious space as a result of inadequate perfusion (101) may contribute to both maternal and fetal oxidative stress. Protein carbonylation is increased in IUGR placenta (130), and neonates have increased cord blood MDA (13, 45), increased protein carbonyls (137), and reduced antioxidant capacity (13, 45, 137). These studies indicate oxidative stress in mothers and their IUGR offspring, but ROS sources have yet to be clearly defined.

Mitochondrial dysfunction is one of the major sources of ROS caused by leakage of electrons from the electron transport chain (ETC) complexes I and III, which then reduces molecular oxygen to the reactive superoxide anion (47, 56, 100). Inhibition of the mitochondrial respiration under normoxia generates superoxide from complex I via reduction of its proximal flavin mononucleotide (FMN) (78, 139) or by its distal coenzyme Q during reverse electron transport (79, 100). Under hypoxic conditions, ROS production from complex III increases cytosolic H₂O₂ levels (47, 100). Finally, α-ketoglutarate dehydrogenase is a significant source of superoxide and H₂O₂, particularly when oxidized nicotinamide adenine dinucleotide (NAD⁺) levels are reduced (100, 144, 157). Mitochondria are not only a source of ROS, but they are also targets of ROS (85, 171, 172), which exacerbates oxidative stress and diminishes oxidative phosphorylation (29). Specifically, H₂O₂ attenuates activity of tricarboxylic acid cycle enzymes aconitase and α-ketoglutarate dehydrogenase (158) and induces complex I H₂O₂ production via flavin-mediated free-radical generation (77). In addition to mitochondrial ROS generation, uncoupling of endothelial nitric oxide synthase (eNOS) also produces superoxide. Oxidative damage to the tetrahydropterin cofactor of eNOS can facilitate uncoupling, thus perpetuating ROS production (115, 132). Finally, superoxide is produced from nicotinamide adenine dinucleotide phosphate (NADPH) oxidase (NOX) and xanthine oxidase enzymes whose over-activations can cause oxidative stress. Superoxide can be metabolized to H₂O₂ by superoxide dismutase (SOD) or may react with nitric oxide (NO) to produce the highly reactive peroxynitrite.

In this review, we discuss evidence that oxidative stress contributes to IUGR-associated metabolic sequelae and explore mechanisms by which oxidative stress impairs gluco-regulatory functions of pancreatic islets, liver, and skeletal muscle.

Oxidative Stress in Pancreatic Islets

In IUGR humans and animals, one of the first indications of ensuing islet dysfunction is capillary rarefaction defined histologically as decreased proportion of blood vessels within an islet (59, 161), which may be caused by oxidative stress (35, 91, 120). Islet β-cell proliferation and function in the neonate are dependent on β-cell–endothelial cross talk, in which secretion of β-cell vascular endothelial growth factor (VEGF) promotes vascularization while endothelial cells secrete the β-cell mitogen hepatocyte growth factor (HGF) (64, 133). NO bioavailability, however, is needed for VEGF function; therefore, uncoupling of eNOS or ROS-mediated degradation of nitric oxide may contribute to IUGR islet dysfunction (3, 36, 37). Indeed, IUGR is associated with reduced β-cell proliferation and age-related reductions in β-cell mass (142). Endothelial dysfunction undoubtedly plays a role in these early and persistent β-cell maladies, and although mechanisms exist by which oxidative stress conceivably confers adverse phenotypes, the contribution of oxidative stress in islet endothelial dysfunction in the context of IUGR and T2D development has not been investigated. Proper endothelial function is important for normal β-cell development and function, but endothelial dysfunction is not solely responsible for diminished glucose-stimulated insulin secretion (GSIS) observed in T2D, nor is it solely responsible for β-cell impairments associated with IUGR.

Insulin secretion from pancreatic islets begins the process of modulating postprandial plasma glucose levels. Glucose uptake, glycolysis, and oxidative metabolism generate the rise in cytosolic adenosine triphosphate (ATP) needed to depolarize the cell membrane and mobilize intracellular Ca²⁺ required for insulin granule exocytosis (FIGURE 1). β-Cell nutrient metabolism needed for insulin secretion is associated with an intracellular rise in ROS (12) derived from mitochondria (84, 136) as well as NADPH oxidase (43, 57, 148) (FIGURE 1). H₂O₂ generation in β-cells in response to glucose occurs rapidly and participates in stimulus-secretion coupling and can induce insulin secretion at basal glucose levels (114). Interestingly, decreasing ROS with antioxidants attenuates glucose-stimulated insulin secretion (84, 114). Cell-permeable catalase attenuated GSIS, but cell-permeable SOD did not, pointing to H₂O₂ rather than superoxide as the ROS facilitating stimulus-secretion coupling (114). Taken together, these studies show H₂O₂ is a second messenger in GSIS, and abrogating its actions with antioxidants is deleterious, which helps explain why β-cells have very low expression of antioxidant enzymes in normal conditions (86, 156). However, diminished ability to neutralize ROS makes β-cells particularly susceptible to oxidative stress and its damaging effects.

FIGURE 1.

Pancreatic β-cell oxidative stress

Insulin secretion in β-cells is coupled to glucose metabolism, leading to subsequent increase in the ATP-to-ADP ratio. Mitochondria are pivotal for producing ATP required for nutrient-induced insulin secretion. ATP binds and closes the ATP-sensitive K⁺ (K_ATP) channels. This depolarizes the plasma membrane, opening voltage-gated Ca⁺² channels with the influx of Ca⁺² stimulating secretion of insulin. NOX activation also contributes to insulin secretion, although the mechanism is unclear. These events observed in GSIS are depicted with black lines, and mechanisms by which oxidative stress is generated and contributes to altered GSIS are depicted with red lines and red Xs. ROS can damage mitochondrial components, including mtDNA and protein, which could result in reduced mitochondrial ATP production and increase mitochondrial-derived ROS. ROS consequently activates redox-responsive signaling pathway JNK, leading to nuclear exclusion of PDX-1, whereas activation of redox-regulated NF-κB upregulates pro-oxidant enzymes NOX-2 and iNOS. Both NF-κB and NRF2 increase gene expression of antioxidant defense genes. GSIS, glucose-stimulated insulin secretion; ROS, reactive oxygen species; TCA, tri-carboxylic acid cycle; Cyt C, cytochrome C; ATP, adenosine triphosphate; Glut2, glucose transporter type 2; NOX, NADPH oxidase; iNOS, inducible nitric oxide synthase; PDX-1, pancreatic and duodenal homeobox 1; JNK, c-Jun NH₂-terminal kinase; NRF2, nuclear factor (erythroid-derived 2)-like 2; NF-κB, nuclear factor kappa light-chain-enhancer of activated B cells.

Oxidative stress in β-cells blunts GSIS by a variety of mechanisms, including mitochondrial dysfunction and modulation of redox-regulated signaling pathways (FIGURE 1). Oxidative stress induced by H₂O₂ has been shown to cause mitochondrial Ca²⁺ overload, severely reducing mitochondrial membrane potential and consequently diminishing ATP levels (94). Because mitochondrial ATP production facilitates the rise in ATP-to-ADP ratio and is obligate for proper GSIS, β-cell mitochondrial dysfunction contributes to hyperglycemia and T2D. H₂O₂-induced activation of c-jun NH₂-terminal kinase (JNK) and p21 mitogen-activated protein kinase (MAPK) reduces insulin gene expression via modulating DNA binding of a key β-cell transcription factor, pancreas, and duodenal homeobox 1 (67, 68). Furthermore, transcriptional upregulation of pro-oxidant and antioxidant enzymes by redox-regulated transcription factors nuclear factor kappa-light-chain-enhancer of activated B cells (NF-kB) and nuclear factor (erythroid-derived 2)-like 2, disrupts H₂O₂ participation in glucose-stimulated insulin secretion (114). Together, these pathways diminish β-cell insulin expression, nutrient metabolism, and signaling pathways facilitating glucose-stimulated insulin secretion, signifying the multifaceted deleterious effects β-cell oxidative stress have on insulin secretion and consequently whole body glucose homeostasis.

In the human IUGR fetus, there is oxidative stress (45, 82) and attenuated first-phase insulin secretion (104), and decreased pancreatic β-cell function is observed in IUGR newborns as well (140). Decreased β-cell function persists in children (97) and young adults who had reduced weight at birth (63, 87) and contribute to diabetes susceptibility later in life. Comparable to developmental programming of β-cell dysfunction in humans, impaired fetal and postnatal pancreatic β-cell function is supported by various experimental animal models of IUGR (44, 89, 90, 141, 142).

In a rat model of uteroplacental insufficiency, IUGR impaired glucose tolerance and diminished insulin sensitivity beginning at 1 wk of age, each declining further with age, such that, by 6 mo, glucose-stimulated insulin secretion is all but extinguished (141, 142). Leucine-induced insulin secretion was also diminished throughout the life course, demonstrating a mitochondrial defect (141). Indeed, isolated IUGR islets exhibited abrogated mitochondrial complex I and III activity and subsequent ATP production. Concomitantly, IUGR islets displayed signs of oxidative stress. ROS production, evidenced by dichlorofluorescine fluorescence, was augmented under basal and stimulatory glucose concentrations in IUGR islets compared with controls (141). 4-Hydroxy-2-nonenal (HNE) protein modification and MnSOD protein levels were increased in IUGR islets, providing further evidence of oxidative stress (141). In placental insufficiency models, RNA sequencing of islets in both rats and sheep revealed significantly reduced antioxidant gene expression. In IUGR rats, glutathione-S-transferases were primarily affected (124), and in fetal sheep, peroxiredoxins were preferentially decreased (72). IUGR-mediated oxidative stress and islet dysfunction are supported by the maternal low protein diet-induced IUGR rat model, showing attenuated islet mitochondrial ATP production, increased uncoupling protein 2 (UCP-2) expression, augmented basal and glucose-stimulated ROS production (153), increased xanthine oxidase expression (151), and increased islet SOD activity (154).

Although we know there is modulation of cellular redox status and mitochondrial function in IUGR islets, there are other mediators that need to be investigated, such as expression and function of NOX enzymes and involvement of redox-regulated signaling pathways. Moreover, islets are a mixed cell population consisting of immune, endothelial, and mesenchymal cell types, in addition to the five endocrine cell types, and studies are needed to delineate the cellular sources of ROS as well as the cellular and molecular targets of ROS.

Taken together, intrauterine perturbations associated with IUGR cause islet mitochondrial dysfunction, oxidative stress, and diminished β-cell function even before birth, and these defects persist throughout life and contribute to developmental programming of T2D. Pancreatic insulin secretion, however, is only one factor contributing to glucose homeostasis. Insulin action in peripheral tissues is primarily responsible for lowering postprandial plasma glucose levels. Diminished capacity for insulin to mediate peripheral glucose uptake, also known as insulin resistance, is evident in human T2D both with and without IUGR.

Oxidative Stress in Insulin-Responsive Peripheral Tissues

Hepatic and skeletal muscle insulin resistance precedes the development of hyperglycemia in humans (168). Binding of insulin to insulin receptors initiates a series of tyrosine phosphorylation events leading to activation of phosphatidylinositol-3-kinase (PI3K), protein kinase B (AKT), and glycogen synthase. In the liver, insulin signaling suppresses hepatic glucose production (HGP) and stimulates glycogen synthesis, and in skeletal muscle, insulin signaling results in translocation of glucose transporter 4 (GLUT4) to the plasma membrane enabling glucose uptake (FIGURE 2). In prediabetes and T2D, oxidative stress is present in insulin-resistant tissues (14, 102, 103). Here, we describe studies that demonstrate a correlation between oxidative stress and insulin resistance following IUGR and give insight into possible mechanisms facilitating insulin resistance in liver and skeletal muscle. Of importance, the development of insulin resistance in IUGR has been shown to be independent of obesity in most human and animal studies (11).

FIGURE 2.

Skeletal muscle oxidative stress

In skeletal muscle, insulin stimulates translocation and fusion of intracellular GLUT4-containing vesicles to the plasma membrane where it facilitates glucose uptake. Insulin mediates translocation via signal transduction involving IR and IRS tyrosine phosphorylation and PI3K activation. From here, signaling bifurcates activating Rac1 and AKT. Rac1 activation compartmentalizes insulin signaling and participates in GLUT4 translocation, whereas AKT activity leads to nNOS activation and AS160 inhibition, both participating in GLUT4 translocation. Also through unknown mechanisms, insulin stimulates ROS production from NOX, which contributes to signal transduction by locally and temporally decreasing PTP activity. Finally, calcium release from intracellular stores, mediated through PLC generation of IP3 and S-glutathionylation of RyRs, is required for translocation. These events observed in insulin-stimulated glucose uptake are depicted with black lines, and mechanisms by which oxidative stress contributes to altered GLUT4 translocation are depicted with red lines. Oxidative stress can activate serine/threonine protein kinases JNK or IKKβ that, respectively, inhibit IRS tyrosine phosphorylation and activate NF-κB. NF-κB activation is associated with modulation in cellular redox status and diminished insulin-mediated GLUT4 translocation. Oxidative stress also inhibits Rac1 activation and its downstream events. Finally, mitochondria-specific generation of oxidative stress abrogates insulin signaling. Disrupted insulin signaling led to insulin resistance and T2D. GLUT4, glucose transporter type 4 (SLC2A4); IRS, insulin receptor substrate; PI3K, phosphatidylinositol-4,5-bisphosphate 3-kinase; PIP3, phosphatidylinositol (3-5)-trisphosphate; PLC, phospholipase C; IP3, inositol 1,4,5-trisphosphate; IP3R, IP3 receptor; RyR, ryanodine receptor; Rac1, Ras-related C3 botulinum toxin substrate 1; GTP, guanosine triphosphate; AKT, protein kinase B; AS160, AKT substrate 160-KD (TBC1D4); nNOS, neuronal nitric oxide synthase; PTP, protein-tyrosine phosphatase.

Liver

Mitochondrial dysfunction is not limited to the β-cell in the IUGR animal. IUGR animals at an early age exhibit marked insulin resistance before the onset of hyperglycemia, characterized by blunted whole body glucose disposal in response to insulin (142) and impaired insulin suppression of hepatic glucose production (HGP) (163). Basal HGP is also increased (163). Four weeks after birth, isolated hepatocytes from IUGR rats exhibited mitochondrial dysfunction, specifically, reduced rates of pyruvate, succinate, and α-ketoglutarate oxidation (112). Rotenone-sensitive NADH-O₂ oxidoreductase activity in IUGR mitochondria was similar to controls, showing disrupted mitochondrial function was proximal to the ETC (112). Indeed, activity of pyruvate and α-ketoglutarate dehydrogenases were diminished in IUGR hepatocytes and were associated with concomitant increase in covalent HNE modification of E2 catalytic and E3 binding protein subunits (112). A compensatory increase in MnSOD protein levels in IUGR hepatocytes was observed, further indicating oxidative stress (112). Pertaining to HGP, mRNA expression of the gluconeogenesis rate-limiting enzyme, phosphoenolpyruvate carboxykinase (PEPCK), was increased 250% in IUGR hepatocytes compared with controls (112). In adult IUGR rats before hyperglycemia, hyperinsulinemic-euglycemic clamp studies demonstrated increased HGP at baseline and during insulin infusion (163). These perturbations were correlated, respectively, with increased PEPCK and glucose-6 phosphatase (G6PC) expression, and attenuated insulin-stimulated insulin receptor substrate (IRS)-2 and AKT phosphorylation (163).

Hyperinsulinemic-euglycemic clamp studies in IUGR fetal sheep confirm findings in IUGR rats that HGP is increased and associated with upregulation of PEPCK and G6PC (155). Oxidative stress and mitochondrial function, however, were not assessed in IUGR sheep livers. Notwithstanding, diminished insulin-mediated suppression of HGP was associated with increased activation of redox-responsive JNK and forkhead box protein O1 (FOXO1) (155). Phosphorylation of FOXO1 by JNK causes its activation and results in increased expression of key genes involved in HGP such as PEPCK and G6PC (121, 155). Insulin sensitivity progressively declines with age, especially if IUGR offspring exhibit rapid catch-up growth and obesity in adulthood (32, 34, 65, 142).

There are limited data in other animal models of IUGR, but low birth weight pigs and IUGR rat offspring resulting from maternal protein deprivation show signs of oxidative stress in liver (152, 169). Although these studies in IUGR animals support an early and persistent role for oxidative stress in HGP, studies in humans show that diminished suppression of HGP is not observed in young adults but rather becomes impaired with aging (60, 63, 118, 159), and it remains to be determined whether this aging-related decline in hepatic glucose homeostasis is due to mitochondrial dysfunction and oxidative stress.

Skeletal Muscle

During IUGR, there is selective redistribution of blood flow away from skeletal muscle resulting in decreased relative muscle mass (28). Because skeletal muscle is responsible for nearly 80% of insulin-mediated glucose uptake (166), IUGR-associated skeletal muscle perturbations can have significant ramifications for glucose homeostasis. In skeletal muscle, insulin receptor activation leads to IRS and PI3K activation, at which point signaling bifurcates activating ras-related C3 botulinum toxin substrate 1 (Rac1) in one pathway and AKT in the other (58) (FIGURE 2). AKT activity facilitates both plasma membrane GLUT4 translocation and glycogen synthase activity. Translocation, rather than GLUT4 content, appears to be the critical step in insulin signaling facilitating glucose uptake. Rac1 activation is important for compartmentalization of insulin signaling events and is required for translocation of GLUT4 to the plasma membrane (21, 61, 62, 75, 149). In muscle, like pancreatic β-cells and other cell types, ROS are important mediators of signal transduction. In response to insulin receptor ligand activation, H₂O₂ production locally and temporally reduces protein-tyrosine phosphatase activity, thereby supporting the cascade of phosphorylation events involved in signal transduction (95, 167). NOX-derived H₂O₂ participates in insulin-mediated GLUT4 translocation through redox-regulation of ryanodine receptors and modulation of intracellular calcium concentrations (24) (FIGURE 2). Moreover, translocation of GLUT4 to the plasma membrane in response to insulin requires nitric oxide downstream of AKT, presumably by serine phosphorylation and activation of neuronal NOS (71). Oxidative stress in insulin-responsive tissues such as skeletal muscle uncouples ROS involvement in insulin signaling and subsequently abrogates GLUT4 translocation and glucose uptake (14).

Oxidative stress-induced insulin resistance is often associated with defects in intracellular compartmentalization of insulin signaling second messengers (14). For instance, hyperinsulinemic-euglycemic clamp studies in rats after glutathione depletion-induced oxidative stress revealed skeletal muscle insulin resistance and NF-κB-dependent disruption of phosphorylated IRS-1 and PI3K intracellular localization without reductions in GLUT4 levels (107). Skeletal muscle oxidative stress in hypertensive rats is associated with NOX-dependent ROS-mediated NF-κB activation and attenuation of AKT phosphorylation and GLUT4 translocation (164). Because mitochondrial defects are associated with developmental programming of T2D, it is noteworthy that mitochondrial superoxide production per se, in the absence of associated reduction in respiration and without affecting insulin signaling, decreases insulin-stimulated GLUT4 translocation to the plasma membrane and abrogates glucose uptake in myotubes and skeletal muscle in vitro and ex vivo, respectively (33, 131). These studies indicate that oxidative stress-related mechanisms of skeletal muscle insulin resistance may contribute to insulin resistance associated with IUGR programming of T2D.

Multiple studies have shown that IUGR humans exhibit diminished peripheral glucose uptake (54, 60, 118). Studies of IUGR young adults show attenuated skeletal muscle glucose uptake with reduced expression of protein kinase C (PKC), PI3K, and GLUT4 (159). Interestingly, a study showed that both IUGR rats and IUGR humans exhibit muscle insulin resistance that is associated with decreased protein expression of PKCζ, PI3K, and GLUT4 (108), suggesting that similar mechanisms underlie IUGR diabetic etiology across species, and thus animal models of T2D developmental programming have clinical relevance. The decreased expression of GLUT4 in these IUGR studies is inconsistent with studies in humans with T2D (31). In fact, studies in healthy or mildly insulin-resistant humans show GLUT4 levels do not correlate with skeletal muscle insulin sensitivity (52, 53). However, some more recent studies do show mild reductions in skeletal muscle GLUT4 in T2D (40, 66), but it is unlikely that such mild reductions alone could confer the level of insulin resistance observed in T2D, which together support skeletal muscle GLUT4 plasma membrane translocation and not total content as the determining factor mediating glucose uptake.

Animal models of placental insufficiency and maternal caloric restriction show decreased skeletal muscle respiration (10, 80), mitochondrial ATP production and insulin-stimulated glucose uptake in IUGR (138), and increased intramuscular lipid content (81, 170), which also contributes to skeletal muscle insulin resistance (23, 76, 99, 111, 165). Diminished glucose uptake was not associated with changes in GLUT4 levels in these animal models but rather with an impaired ability of insulin to elicit translocation to the plasma membrane (138, 170). Glycogen synthesis was also attenuated in IUGR animals (138). This is consistent with studies in older men, showing skeletal muscle basal and insulin-stimulated glycogen synthesis to be positively associated with birth weight (119). Diminished glycogen synthesis, however, could be a result of decreased insulin stimulation of glycogen synthase or a consequence of reduced glucose uptake. Recent studies using maternal protein restriction in rats show adverse responses to IUGR in skeletal muscle are associated with redox-regulated NF-κB activation, increased xanthine oxidase and NOX-2 expression, and a compensatory increase in antioxidant enzymes, which together signify oxidative stress (151). Interestingly, it is this model of maternal protein restriction that shows reductions in PI3K and GLUT4 levels in IUGR skeletal muscle. Therefore, oxidative stress in IUGR may be involved in diminished PI3K and GLUT4 expression observed with protein restriction and also insulin-mediated suppression of GLUT4 translocation observed in placental insufficiency and caloric restriction IUGR models. Regardless of method of IUGR induction, impairments in skeletal muscle glucose uptake and related oxidative stress in IUGR are comparable with models of oxidative stress-induced skeletal muscle insulin resistance. None of these IUGR studies distinguished between the various muscle fiber types, so it is unknown which muscle fiber type is primarily responsible for skeletal muscle insulin resistance. Studies involving oxidative stress-induced skeletal muscle insulin resistance are advancing our understanding of GLUT4 translocation, but the field of developmental programming has yet to determine the relevance of these mechanisms with regard to IUGR-associated skeletal muscle insulin resistance and T2D.

Antioxidant Interventions

Because oxidative stress has been correlated with IUGR-induced metabolic disorders, studies have investigated the efficacy of antioxidants to ameliorate adverse phenotypes. Recently, pregnant women with fetal growth restriction were shown to have significantly reduced plasma levels of antioxidant vitamins C and E (38), but a meta-analysis showed that supplementation with these antioxidant vitamins did not improve fetal growth (134, 135). In guinea pig, supplementation with glutathione precursor, N-acetylcysteine, normalized endothelial function in IUGR placenta and fetus, and consequently normalized fetal growth (55). However, the ability of these antioxidants to disrupt aging-related metabolic sequelae has not been investigated. Supplementation at weaning of the antioxidant coenzyme Q10 to maternal protein-restricted offspring has been shown to improve insulin sensitivity and reduce markers of hepatic oxidative stress in aged offspring (152). Unfortunately, these data presented here represent the extent of research examining the ability of antioxidant supplementation to quell IUGR and its associated gluco-regulatory perturbations. Conclusions regarding antioxidant effectiveness should not be drawn from such limited studies. Supplementation with compounds with indirect antioxidant activities has been investigated, but delineating oxidative stress-specific involvement is confounded because of their indirect antioxidant activity.

As discussed in previous sections, mitochondria are both producers and targets of ROS. Interventional studies targeting mitochondrial function via taurine supplementation has proven effective at mitigating the effects of an adverse intrauterine milieu. Taurine is a nonessential sulfur amino acid synthesized from sulfur-containing amino acids methionine and cysteine. Taurine, whose levels are maintained by unidirectional maternal to fetal placental transport by the taurine transporter (TauT) (69), is essential for proper fetal development (41, 74). Growth-restricted human fetuses have reduced taurine levels caused by diminished TauT activity (30, 106). Taurine indirectly alleviates mitochondrial oxidative stress (2, 109) and has been shown to protect placental trophoblasts from oxidative stress induced by H₂O₂ (105). Because taurine levels are reduced in IUGR offspring and taurine preserves mitochondrial function in oxidative stress conditions, increasing taurine levels during development may ameliorate IUGR-associated pathologies. Indeed, maternal supplementation in models of IUGR prevents islet capillary rarefaction and oxidative stress, and prevents the development of T2D in adulthood (8, 9, 15, 16, 83, 96, 150). Interestingly, normalization of these abnormalities and the prevention of T2D occurred without normalization of fetal growth.

Glucagon-like peptide 1 (GLP-1) is an incretin hormone that potentiates GSIS and augments β-cell stress defense. Exendin-4 is a pharmacological activator of the GLP-1 receptor (GLP-1R). In rat models of placental insufficiency, Exendin-4 administration to IUGR offspring in the first week of life prevents IUGR-associated metabolic derangements. Neonatal Exendin-4 administration specifically was shown to restore islet capillary density, β-cell mass, GSIS, glucose tolerance, insulin sensitivity, and HGP (51, 122, 147).

GLP-1R activation stimulates various signaling pathways, including protein kinase A, MAPK, PI3K, and AKT (4), but is not traditionally thought of as participating in modulation of cellular redox status. Notwithstanding, multiple studies have shown that GLP-1R agonists improve mitochondria dysfunction (19, 20), ameliorate oxidative stress (19, 88), and enhance insulin sensitivity in skeletal muscle (1, 162) and liver (1, 39, 92) in human (27, 88) and animal models of T2D (17, 18, 147). Similarly, GLP-1R activation abrogates oxidative stress in rat models of placental insufficiency (122). Further studies are required to determine whether Exendin-4 reduction of oxidative stress is necessary to confer its protective effects. More recently, utilizing the same neonatal intervention window, immunological neutralization of interleukin-4 in this rat model of placental insufficiency rescued islet capillary rarefaction, prevented β-cell mass reductions, and obviated IUGR-induced T2D (59). Importantly, unlike Exendin-4 administration, this immunological intervention did not attenuate oxidative stress (59), demonstrating the possibility that oxidative stress in IUGR-induced T2D is correlative rather than causative. It also demonstrates the gaps in the current understanding of the IUGR-induced pleiotropic effects and how they interact mechanistically in the pathogenesis of IUGR-associated T2D.

Taken together, these studies suggest that fetal oxidative stress and mitochondrial dysfunction may underpin developmental programming of T2D and that recovering fetal growth is an inaccurate proxy for ameliorating IUGR-associated metabolic sequelae. These studies also point to the ability of maternal and postnatal interventions to prevent developmental programming of T2D, but intervention strategies beginning during prediabetic or diabetic periods are not investigated despite their clinical relevance. Future studies should, therefore, deemphasize antioxidant normalization of fetal growth and birth weight, but rather emphasize the ability of antioxidants to prevent IUGR-induced T2D. Finally, studies are needed to assess whether antioxidant interventions commencing near the diabetic period are efficacious.

Conclusions

ROS play an important role in signal transduction related to insulin secretion, hepatic glucose production, and skeletal muscle glucose uptake. ROS in these normal processes are influenced by their source and are regulated spatially and temporally. In IUGR humans and animals, oxidative stress is evident in utero and persists throughout the life course. There is mitochondrial dysfunction that contributes to oxidative stress in IUGR, but the extent to which it contributes to oxidative stress and the involvement of other ROS generators has yet to be elucidated and should be a priority in future studies. Where possible, genetic manipulation strategies and pharmacological modulators of redox status and redox-regulated signaling pathways should be utilized to empirically determine the mechanism by which oxidative stress engenders metabolic dysfunction, as opposed to the current practice of describing mechanistic associations. Moreover, different animal models have been used to understand how IUGR disrupts glucose homeostasis, and, although oxidative stress and mitochondrial dysfunction are observed in these models, T2D pathogenesis differs among them; therefore, mechanisms mediating glucose intolerance may not be the same in other animal models.