摘要
After completing this article, readers should be able to: Amniotic fluid (AF) is an accepted physical barrier against fetal trauma and a defense against fetal infection. The importance of AF in fetal nutrition also is an acknowledged fact. Fetal malnutrition related to abnormalities of AF and its consequences on neonatal health, however, are not fully appreciated. This review examines the nutritional characteristics of AF.In 1975, fetal swallowing was shown to be the primary method of clearing proteins from AF. ( 1) Using near-term Rhesus monkeys and 35S-labeled proteins injected into AF, 10% to 15% of nitrogen accretion in late pregnancy was shown to be related to fetal swallowing. A decade later, Mulvihill and colleagues ( 2) reported experiments that ligated the esophagi of fetal rabbits to prevent swallowing of AF. These researchers subsequently infused different nutritive solutions into the fetal stomach and compared growth and organ weights. Swallowing of AF enhanced fetal gastrointestinal development, and in late gestation, accounted for 10% to 14% of the nutritional requirements of the normal fetus. In sheep, esophageal ligation causes abnormal enterocyte differentiation and decreases fetal intestinal growth, conditions that are mitigated by re-establishing fetal swallowing. ( 3) Ultrasonography recently defined gastric emptying cycles in human fetuses throughout pregnancy, and near-term, the delay in gastric emptying may be related to satiation. ( 4)The effect of AF on fetal intestinal and somatic growth in humans is more circumstantial (Table 1). In 1994, Surana and Puri ( 5) studied jejunal versus ileal atresia and discovered that the more proximal the intestinal obstruction in the human fetus, the more severe is the fetal growth restriction (FGR). A study of 56 neonates who had duodenal, jejunal, or ileal atresia showed that the incidence of polyhydramnios and preterm birth is reduced and birthweight is increased with more distal obstruction in the small bowel. ( 6) The observation that some infants who have intestinal atresia have normal gastrointestinal development, a normal birthweight, and a normal volume of AF suggests a redundancy in factors that mediate these outcomes.Growth factors and hormones initially were identified in human AF three decades ago, ( 7) but only years later were the principal mediators of fetal intestinal and somatic growth in AF delineated. Hirai and associates ( 8) examined the trophic effects of AF, human milk, and several recombinant growth factors on a human fetal small intestinal cell line. AF and human milk equally promoted cell growth. These researchers noted that cell growth was partially inhibited by antibodies to epidermal growth factor (EGF), insulin-like growth factor-I (IGF-I), fibroblast growth factor (FGF), hepatocyte growth factor (HGF), and transforming growth factor-alpha (TGF-alpha). Recombinant forms of these growth factors stimulated cell growth synergistically, but enhancement was less than with AF alone.The fetal rabbit is a favorite model of FGR because each fetus has its own AF space and there is consistent runting of the fetuses at specific positions in the bicornuate uterus. In 1986, attempts to overcome FGR in a rabbit model with intra-amniotic infusions were first described. ( 9) The studies used infusions of dextrose, dextrose plus amino acids, or a lipid emulsion. None ameliorated FGR compared with controls. In fact, the lipid emulsion resulted in chronic lipid aspiration and additional growth restriction. Other investigators infused bovine AF into the amniotic space of fetal rabbits and observed improved somatic growth and enhanced organ weights in the small bowel, lung, and liver. ( 10) Although not yet used in clinical practice, it has been proposed that the infusion of nutrients or growth factors into AF might mitigate FGR in human pregnancy. ( 11)In late pregnancy, it is assumed that fetal swallowing prepares the gastrointestinal tract for postnatal nutrition. This concept is based on similar nutritive components and growth factors being present in AF and human milk (Table 2). ( 12) Limited information exists, however, regarding which macro- and micronutrients initiate preparation of the fetal intestine for postnatal life.Reports measuring glucose content in AF throughout pregnancy indicate that the concentration of glucose declines significantly during the third trimester. ( 13) This finding is consistent with the placental transfer of glucose and regulation of fetal growth via glucose, insulin, and IGF-I interactions. ( 14) Swallowing AF containing excess glucose may enhance fetal weight in rats. Glucose assimilation in this clinical scenario is not reported in humans. There is little evidence that alternative carbohydrates augment fetal nutrition via the swallowing of AF. Amniotic fluid does contain mucins, but their role in nutrition is uncertain. Oligosaccharides present in human milk and associated with postnatal intestinal health have not been identified in AF. ( 15)Concentrations of amino acids in AF have been reported during early, middle, and late periods of normal and abnormal human pregnancy. ( 16) Diabetes, aminoacidopathies, and other abnormal pregnancies can alter the normal amino acid content of AF. Maternal caloric deprivation profoundly alters amino acid composition in AF. ( 17) Starvation elevates valine, leucine, isoleucine, and taurine concentrations and lowers alanine and citrulline concentrations. Several reports suggest the existence of a transport mechanism that enhances amino acids in AF. Consistent with this concept, pregnant sheep that undergo placental embolization to damage this fetal organ and create FGR have a decline in the concentrations of amino acids in AF. ( 18)Two amino acids in AF, glutamine and arginine, are of particular interest in fetal intestinal growth and maturation. Glutamine is considered conditionally essential for normal gastrointestinal growth and function. ( 19) In fetal sheep, glutamine in swallowed AF is absorbed by the intestine and used as a fuel in the intestine, and the infusion of IGF-I into the AF appears to increase intestinal glutamine uptake. ( 18) Whether injections of glutamine and IGF-I into the amniotic space could partially overcome FGR is an intriguing question. Of interest, preterm infants who develop necrotizing enterocolitis have lower circulating levels of glutamine and arginine for 7 days before the development of symptoms. ( 20)Arginine is the amino acid precursor for nitric oxide and polyamine synthesis. ( 21) Whether arginine in AF is swallowed and absorbed by the fetal intestine, subsequently acting as a substrate for nitric oxide synthesis, thereby increasing blood flow in the fetal intestine, is undetermined. In the ovine model of FGR, ( 18) citrulline content of the fetal intestine is reduced. Citrulline is the other end product of nitric oxide synthesis from the amino acid l-arginine. Arginine also is hydrolyzed to ornithine by arginase. Ornithine decarboxylase then converts ornithine into putrescine, which subsequently is metabolized to spermidine and spermine. Putrescine and spermidine increase in AF until the 30th week of human pregnancy, while spermine remains high throughout gestation. ( 22) In preterm infants, arginine and polyamines are associated with intestinal growth and function, ( 23) but the role of arginine and polyamines in the growth of fetal bowel is based on inferred evidence.AF contains many proteins, some of which are important hormones and growth regulators. Transferrin and lactoferrin are iron-binding glycoproteins present in AF during the latter half of pregnancy. ( 24) These proteins may deliver ferric iron to intestinal epithelia. Lactoferrin has growth factor-promoting qualities, and it accelerates the incorporation of thymidine into enterocytes. ( 25) Several hormones, separate from protein-related growth factors, also are found in AF. Growth hormone, prolactin, placental lactogen, and parathyroid hormone-related protein are identified in AF and likely are derived from the placenta. ( 26)( 27) There are disparate reports regarding the levels of the aforementioned hormones in AF of infants who have FGR or macrosomia, and the roles that these hormones play in the ontogeny and growth of the fetal intestine are inadequately studied. Gastric inhibitory polypeptide, a potential regulatory protein in glucose-insulin homeostasis, and the calcium-binding protein S100B have been identified in AF, but the effect of these two proteins on fetal growth regulation is uncertain. Albumin and fibrinogen, which also are present in AF, may have nutritive value for the fetus, but they are widely considered inhibitors of surfactant-related protein activity. Other proteins in AF are used to identify fetal anomalies, the health of the pregnancy or the fetus, lung maturation, and infection/inflammation, but their role in fetal nutrition probably is limited.AF is considered aqueous liquor with suspended fetal cells. The cells may degrade and supply lipids for fetal growth because there is a sharp increase in cholesterol and glyceride concentrations after the 37th week of gestation. ( 28) Total lipids in AF increase progressively with advancing gestational age, reaching maximum concentrations near term. ( 29) This finding is related to surfactant production and the release of lipids from pulmonary epithelia and their efflux from the fetal airways into the AF. Additional research will determine whether AF-related lipids enhance intestinal and somatic development in the fetus.Studies of vitamins, minerals, and other micronutrients in AF related to either intestinal or somatic growth in the fetus are limited. Given the large volumes of AF swallowed in utero, ( 30) even small concentrations of these nutrients may have an impact. Vitamin concentrations in the AF are defined. Ascorbic acid and alpha-tocopherol are found in the human extracoellomic cavity in the first trimester, arriving in the fetal gut and circulation via the secondary yolk sac. ( 31) They likely play an important antioxidant role in early pregnancy. The same is true of vitamin C in AF of the near-term infant, in whom levels are decreased in mothers who are tobacco smokers. ( 32) Other vitamins also may serve as antioxidants. Retinol and retinol-binding protein are elevated in AF of mothers whose pregnancies are complicated by birth defects, diabetes, and gestational hypertension. ( 33)( 34) One third of pregnancies have increased 1,25-dihydroxy-vitamin D concentrations in AF at term, paralleling the increased fetal demand for calcium assimilation in late pregnancy. ( 35) Vitamin B12 and folate levels in AF have been studied as markers of neural tube defects. ( 36) Vitamin B12 concentrations consistently are increased in AF throughout pregnancy, and it is interesting to speculate on the vitamin's activity in the distal small bowel of the fetus. Whether swallowing of vitamins in AF affects fetal nutrition remains unclear.Finally, minerals and trace metals are found in AF. Sodium, potassium, calcium, and phosphorus are found in AF throughout pregnancy, but their kinetics and nutritional value have not been investigated. A recent study of nonprotein-bound iron concentration in AF demonstrated an interesting correlation between AF iron and fetal growth parameters. ( 37) Human AF also contains magnesium, copper, zinc, cadmium, lead, selenium, and mercury. Zinc alterations in AF, usually seen as elevations, have been noted consistently in abnormal pregnancies or gestations associated with fetal malformations. ( 38) Comparisons of maternal and fetal plasma and liver concentrations of zinc, copper, and iron demonstrate the high rates of transfer of these metals from mother to fetus. How the AF content of these metals affects fetal nutrition is uncertain.AF and human milk contain growth factors that exert pre- and postnatal effects on intestinal and somatic growth in the fetus and the newborn, respectively.IGF-I is the primary mediator of both intrauterine and postnatal growth in mammals. IGF-I infused enterally into sheep that have undergone esophageal ligation increases somatic growth and bowel wall thickness. ( 39) The concentration of IGF-I in human AF can reach 20 ng/mL, and fetal swallowing near term means that up to 20 mcg of IGF-I may be ingested daily. ( 40) When 125I-IGF-I is injected into the amniotic sac of late gestation fetal sheep, it is distributed rapidly throughout the AF volume, is swallowed by the fetus, and is absorbed intact across the gut into the portal venous circulation. ( 41) Infusions of IGF-I into the AF of sheep have resulted in increased fetal intestinal growth, increased fetal plasma IGF-I concentrations, and increased maturity of fetal enterocytes. ( 42) Reversal of the intestinal and somatic findings in the fetal ovine model of FGR by infusing IGF-I into the AF or the gut lumen have produced mixed, but promising results, and active research continues in this area.IGF-II is a major modulator of early embryonic and second trimester fetal growth. IGF-II is synthesized by the fetal lung and likely enters the AF via the efflux of fetal lung fluid. There is threefold greater concentration of IGF-II in AF during the early second semester than IGF-I. IGF-II concentration in AF peaks at 19 weeks' gestation, declining dramatically thereafter, when mRNA expression for IGF-II in fetal lung tissue wanes. ( 43) The IGF-II gene is expressed paternally in the fetus and placenta. Deletion from the IGF-II gene of a transcript specifically expressed in the labyrinthine trophoblast of the placenta leads to reduced placental growth, followed several days later by FGR. ( 44) The two human IGF-II-binding proteins (IGF BP1 and IGF BP3) also are found in AF and have a regulatory role. Specifically, measurements of IGF-II, IGF BP1, and IGF BP3 in second trimester AF were compared recently with birthweight with interesting results. Higher levels of IGF BP1 correlated with lower birthweight, and higher levels of IGF BP3 correlated with macrosomia. ( 45)EGF is a small peptide that stimulates cell mitosis and differentiation. Significant quantities of EGF are found in human AF, and the content increases as pregnancy progresses. ( 46) Alternatively, lower concentrations are seen with FGR. Measurements of EGF are fourfold higher in AF than in fetal urine, suggesting that the site of production is either the lung or the amnionic membranes. The impact of EGF in AF on fetal intestinal growth is an area of active research. EGF increases DNA and glycoprotein synthesis in cultured human fetal gastric cells. ( 47) In fetal rabbits, enteral infusions of EGF reverse the effects of esophageal ligation, and amniotic infusions of EGF increase small intestinal length, lactase and maltase activity, and villus height as well as correct naturally occurring FGR. ( 48) EGF receptors are present in the human stomach as early as 12 weeks of gestation, ( 49) but EGF is not produced there, suggesting that any effects of this growth factor on the fetal intestine tract are from swallowed EGF.TGF-alpha has a structure similar to EGF and binds to the same receptor. TGF-alpha is present in AF and human milk, and like EGF, is found in higher concentrations in human milk from women delivering prior to 27 weeks of gestation compared with those delivering after 27 weeks. ( 50) TGF-alpha, and to lesser extent EGF, is expressed in the fetal intestine. Recombinant TGF-alpha elicits a synergistic trophic response on cultured intestinal cells when combined with recombinant EGF, IGF-I, FGF, and HGF, ( 8) but this trophic response is not as strong as that seen with AF or human milk. The amnion cells of the umbilical cord express EGF, TGF-alpha, and functional EGF/TGF-alpha receptor(s), suggesting the possibility of a regulating role of the amnion in fetal growth and development. ( 51)TGF-beta1 is one of a family of signaling peptides that influences the differentiation of intestinal stem cells. TGF-beta1 is only found in human AF during the late stages of gestation. It is believed to induce terminal differentiation of intestinal epithelial cells and to accelerate the rate of healing of intestinal wounds by stimulating cell migration. Thus, TGF-beta1 may prepare the fetal intestine for the extrauterine environment. Importantly, this peptide may play a role in preventing necrotizing enterocolitis by inhibiting secretion of proinflammatory cytokines by human fetal intestinal cells. ( 52)Several hematopoietic growth factors are present in human AF, ( 53) and recently, focus has shifted to their ability to promote neonatal intestinal health. Erythropoietin (EPO) is found in human AF, colostrum, and mature milk. The role of swallowed EPO in the human fetus is not clear, but studies of the enteral administration of EPO after birth suggest a reduced incidence of necrotizing enterocolitis in preterm infants. ( 54) Oral EPO is not absorbed, and it does not stimulate erythropoiesis. Elevated levels of EPO in AF and cord blood are markers for chronic fetal hypoxia.Granulocyte colony-stimulating factor (G-CSF) also is found in human AF. When G-CSF is given enterally to suckling mice, it enhances intestinal growth. G-CSF has been administered enterally to human neonates and is well tolerated, but it is not absorbed. ( 55) A local trophic effect in the human fetal intestine has not yet been demonstrated.Polyhydramnios may be the consequence of gastrointestinal anomalies, neurologic abnormalities, genetic disorders, or idiopathic factors. ( 56) Regardless of the cause, fetal swallowing often is abnormal. Moreover, the large amniotic fluid volume may significantly dilute important nutrients and growth factors in AF. Conversely, oligohydramnios may limit the amount of nutrients and growth factors available to the fetus because there is an insufficient volume to drink. A fetus may have growth restriction without polyhydramnios or oligohydramnios, but have suboptimal nutrition because of aberrant fetal swallowing. Whichever scenario is present, the fetal intestine is not likely to grow and mature unless there is an intervention. Addressing the root cause of too much or too little AF or the cause of FGR is at the heart of the problem. A solution, however, may not be possible. One proposed approach is the insertion of a percutaneous catheter into the amniotic fluid sac or into the fetal stomach. This catheter could be used to deliver the desired nutrients and growth factors. Although this was proposed 2 decades ago, ( 11) the approach has not taken hold, perhaps because of infectious risks. Innovative methods or novel technology might make this fetal therapy a reality.In a series of reports, Christensen and colleagues ( 57) have proposed a novel approach to early feeding intolerance in very low-birthweight (VLBW) infants. To decrease the intestinal atrophy that occurs with prolonged fasting, these researchers suggested that a simulated feeding similar to AF might be useful for neonates who were not yet ready for rapidly advancing enteral feedings. The "synthetic" AF was sterile, isotonic, and noncaloric, containing electrolytes, albumin, and two recombinant human growth factors, EPO and G-CSF. The solution was stable and relatively resistant to digestion; the ingested growth factors were not absorbed. The investigators demonstrated that receptors for the two growth factors were found on the luminal surface of enterocytes, suggesting a localized effect. The enteral solution was well tolerated by VLBW infants, and they could assimilate higher caloric intakes compared with historic controls. Larger randomized, blinded trials are proposed using simulated AF.Other researchers have used growth factors present in AF and human milk that are more likely to accelerate intestinal growth, maturation, or repair. Such growth factors include recombinant human IGF-I, EGF, recombinant human HGF, and recombinant human growth hormone. An encouraging report is the use of EGF to reverse the disability seen in pediatric short bowel syndrome. ( 58)The ideal synthetic AF might contain glutamine, recombinant human growth factors such as IGF-I, EGF, and TGF-beta1, and a protective component such as lactoferrin. Whether administered as an in utero infusion to correct FGR or by gavage to high-risk neonates, the solution may improve neonatal health and promote a developmental paradigm that avoids the adult consequences of fetal and neonatal nutritional disorders. ( 59)( 60) Alternatively, El-Haddad and colleagues ( 56) have introduced the idea of fetal programming to establish appropriate set points for adult thirst and appetite.