Which of the following fatty acids is critical to fetal brain and i development

As introduced above, all of the n-6 and n-3 fatty acids accumulated by the fetus are derived by transfer across the placenta and ultimately originate from the maternal diet. The n-6 and n-3 fatty acids may be provided as 20:4n-6 and 22:6n-3, or as their 18:2n-6 and 18:3n-3 precursors, respectively. Experimental studies have shown that placental fatty acid transfer involves diffusion as well as membrane and cytosolic fatty acid binding proteins; membrane binding proteins that favour n-6 and n-3 fatty acids over non-essential fatty acids and 20:4n-6 and 22:6n-3 over 18:2n-6 and 18:3n-3 may be important in facilitating placental transfer of the latter longer chain n-6 and n-3 fatty acids to the fetus [6], [7], [8], [9], [10]. Recent reviews of placental fatty acid transfer have been published [9], [13], [14]. The Δ6 and Δ5 desaturases are present in fetal liver from early in gestation, but the activity of these enzymes appears to be low before birth [1], [15], [16], [17], [18]. Further, experimental studies have clearly shown that preformed 22:6n-3 provided in the maternal diet is much more efficacious than 18:3n-3 as a source of n-3 fatty acids for fetal tissue 22:6n-3 accretion [19], [20], [21], [22], [23]. Similarly, although desaturation of 18:3n-3 to 22:6n-3 occurs in infants and adults, including preterm infants, the activity of the pathway appears to be low with <1–9% 18:3n-3 converted to 22:6n-3 [24], [25], [26], [27]. There is no evidence that the ability to form 20:4n-6 from 18:2n-6 is low in humans, although providing a source of preformed 20:4n-6 in the diet increases plasma and red blood cell phospholipid 20:4n-6 in adults and infants [28], [29], [30], and increases tissue 20:4n-6 in animals [31], [32].

Analyses of the n-6 and n-3 fatty acids in maternal and fetal plasma (cord blood collected immediately following term birth) has shown that the relative proportions of 20:4n-6 and 22:6n-3 are higher, while 18:2n-6 is lower in triglycerides, phospholipids and cholesterol esters in the fetal than maternal plasma (Table 1). Similarly, 20:4n-6 and 22:6n-3 represent a higher proportion while 18:2n-6 is lower in the unsaturated fatty acids of fetal plasma esterified lipids than in the plasma of one-month-old breast-fed infants or infants fed formula (Table 2). However, despite higher proportions of 20:4n-6, 18:2n-6 is clearly transported across the placenta. The contribution of preferential acylation of 20:4n-6 and 22:6n-3 into esterified lipids released by the placenta to the fetal circulation and of specificity of acyltransferases involved in triglyceride and phospholipid synthesis in fetal liver to the high proportions of 20:4n-6 and 22:6n-3 in fetal plasma is not yet known. In this regard, recent studies have shown that the placenta secretes apo B containing particles in the low density lipoprotein (LDL) density range [33], suggesting that the placenta could contribute to the molecular species of phospholipids, triglycerides and cholesterol esters characteristic of fetal plasma lipids. In addition, both Δ6 and Δ5-desaturase have been identified in the placenta [34], [35]; thus it is possible that placental synthesis of 20:4n-6 from 18:2n-6 could contribute to 20:4n-6 in the fetal circulation. The unusual distribution of fatty acids in fetal plasma esterified lipids includes about 40% 20:4n-6 in cholesterol ester fatty acids whereas about 80% of cholesterol esters in maternal plasma are esterified with 18:2n-6 (Figure 1). After birth, plasma cholesteryl esters are derived from the action of lecithin:cholesterol acyltransferase (LCAT), which esterifies unesterified cholesterol with the fatty acid from the sn-2 position of high density lipoprotein (HDL) phospholipid [36]. The phospholipid fatty acid substrate available in utero for LCAT is predominately 20:4n-6 rather than 18:2n-6 (Table 2), however, LCAT activity appears to be low in fetal plasma [37]. In addition, the major of lipoprotein in fetal plasma is HDL, rather than low density lipoprotein (LDL) as in the adult [38], [39]. The high apo E content of fetal HDL has led to the suggestion that HDL could be important in delivering cholesterol to tissues prior to birth [38]. Although the fetal rat brain appears to synthesize cholesterol de novo, rather than acquiring cholesterol for new membrane synthesis by uptake from plasma [40], it is not clear if the developing human brain is similarly autologous with respect to cholesterol synthesis. Several studies have shown that unesterified 20:4n-6 and 22:6n-3 are readily incorporated into the brain [41], [42]. Although the proportions of 20:4n-6 and 22:6n-3 in plasma unesterified fatty acids are relatively low, the high turnover of this pool of fatty acids could support an important role in providing 20:4n-6 and 22:6n-3 for the developing brain [43].

Despite experimental and clinical evidence consistent with preferential transfer across the placenta, information from both human and animal studies has shown that the maternal dietary intake of n-6 and n-3 fatty acids influences maternal to fetal 20:4n-6 and 22:6n-3 transfer. The relative proportions of 20:4n-6 and 22:6n-3 in maternal plasma are significantly and positively correlated with the proportion of the same fatty acid in fetal plasma [12]. Studies to show that supplementation of pregnant women with 22:6n-3 from fish or fish oils increases 22:6n-3 in plasma and red blood cell lipids of the infant at birth [44], [45], [46] show that placental transfer of 22:6n-3 is dependent on maternal plasma 22:6n-3. However, supplementation of pregnant women with 18:3n-3 does not result in higher umbilical cord blood levels of 22:6n-3 [47]. Information to show that infants born with higher blood levels of 22:6n-3 and 20:4n-6 maintain this advantage for several weeks after birth [48], [49] suggests that these fatty acids must be accumulated in fetal tissue lipids and do contribute to the circulating lipid pool after birth.

In animals, low intakes of n-3 fatty acids in gestation result in decreased 22:6n-3 in neural growth cones, the amoeboid leading edge of the growing neurite in fetal brain and liver, as well as in the placenta [23], [50]. Reduced accretion of 22:6n-3 in the retina and brain during development results in decreased electroretinogram responses, decreased performance in behaviour tests of learning, exploratory activity and auditory brain stem evoked potential responses, and changes in dopamine and serotonin metabolism [23], [29], [51], [52]. High maternal intakes of 20:5n-3 and 22:6n-3, however, also decrease 20:4n-6 in the placenta, as well as in the fetal liver and brain [23], [50]. In humans, several studies have reported a positive association between cord plasma 20:4n-6 and infant birthweight [12], [53] and in preterm infants, plasma 20:4n-6 is positively related to growth [54]. Clinical trials have also noted higher growth in preterm infant fed with formula containing 20:4n-6 and 22:6n-3 [53], [55]. Although a dietary intake of about 0.2% energy as 20:4n-6 fulfills the essential role of n-6 fatty acids for growth [1], [2], an explanation for a positive effect of 20:4n-6 on growth in the presence of a maternal or postnatal dietary supply of 18:2n-6 is not available.

Only limited information is as yet available on the possible physiological significance of differences in maternal dietary 22:6n-3 intake to fetal development in humans. Studies using electroencephalography (EEG) at 2 days after birth as a measure of CNS maturity have reported that infants with a more mature EEG pattern had significantly higher 22:6n-3 in cord plasma phospholipids than infants with a less mature EEG pattern [45]. In the latter study, however, supplementation of pregnant women from 16 weeks of gestation with 22:6n-3 in the form of cod liver oil had no effect on the EEG measures in the infants [45]. Other epidemiologic studies have reported an inverse relationship between maternal plasma 22:6n-3 and active sleep and sleep–wake transitions, and a positive association between 22:6n-3 and wakefulness in 2-day-old infants [56]. Measures of cognitive performance in older children in relation to their cord blood levels of 22:6n-3 have yielded inconsistent data [57], [58], [59]. These types of epidemiological studies will require large cohorts to control for socio-demographic, family and other maternal and early childhood dietary variables that influence early child development, including potential confounding effects of developmental neurotoxins, such as methyl mercury and PCBs in fish.

Information of the importance of 22:6n-3 for the developing human fetal brain and retina is also available from clinical studies on n-3 fatty acid nutrition in preterm infants. Meta-analyses have shown that providing preterm infants with a dietary source of 22:6n-3 results in higher visual acuity during the first months after birth [60]. In addition to higher visual acuity, preterm infants <1250 g birthweight fed formula containing about 1.2% energy as 18.3n-3 supplemented with 22:6n-3 showed an advantage in test scores on the Bayley mental developmental inventories and in the MacArthur Communicative Inventories at 12 months corrected term age [61]. These latter controlled, randomized clinical studies show that the n-3 fatty acid requirements of preterm infants are not met by 1.2% dietary energy as 18:3n-3, and that a small amount of 22:6n-3 facilitates early advantages in visual and neural development. The long-term significance of these early benefits to the developing central nervous system, however, has yet to be demonstrated.

In animals, n-3 fatty acid deficiency results in decreased 22:6n-3 and a marked increase in docosapentaenoic acid (22:5n-6), formed from 18:2n-6 in a parallel pathway to that involved in synthesis of 22:6n-3 from 18:3n-3 [1], [2]. This leads to a decrease in the ratio of 22:6n-3/22:5n-6 in membrane PE in n-3 fatty acid deficient animals, and suggests that the ratio of 22:6n-3/22:5n-3 in phospholipids could be a useful biochemcial marker of inadequate n-3 fatty acid nutrition. However, 22:6n-3 and 22:5n-6 are positively, not inversely related in young children; similarly in infants fed formula with no 22:6n-3, neither plasma nor red blood cell PE show an increase in 22:5n-6 when compared to breast-fed infants or infants fed formula containing 22:6n-3 [62]. A similar situation occurs in the fetus; despite a wide range of 22:6n-3 in red blood cell membrane PE of 2.2–12.8% total fatty acids, 22:6n-3 and 22:5n-6 are significantly and positively, not inversely correlated, r = 0.3, P < 0.001, n = 148, (unpublished data). These findings suggest that the desaturation of both n-6 and n-3 fatty acids in humans may be slow beyond the Δ5 desaturase step [62], which may result in important differences in the functional effects of n-3 fatty acid deficiency in humans from those effects demonstrated in n-3 fatty acid deficient rodents.

Premature infants are particularly vulnerable to nutritional deficiency because of their limited adipose tissue mass and immaturity in many metabolic and physiologic pathways at birth. In addition, the growth of dendritic arbors and peak formation of synapses, which are enriched in 22:6n-3, extends from about 34 weeks of gestation through 24 months after birth, during which time new connections form at rates up to almost 40,000 synapses/s [63]. Several approaches can be used to address the n-6 and n-3 fatty acid requirements of infants born during the third trimester of gestation. One approach is to match the blood levels of 20:4n-6 and 22:6n-3 of the fetus in utero. As described, the fetal plasma lipids are characterized by an abundance of 20:4n-6 and 22:6n-3, relatively low 18:2n-6, and triglyceride-rich lipoproteins of alimentary origin are very low. Following birth, the high 20:4n-6 and 22:6n-3 characteristic of fetal plasma lipids decline (Figure 2). Fat is provided as major source of energy in order to attain growth rates approaching that of the third trimester fetus, and the initiation of triglyceride-rich lipoproteins as a major lipid transport particle occurs, irrespective of whether nutritional support is provided by parenteral nutrition with intravenous lipids, human milk or infant formula. Although the increase in 18:2n-6 in infant plasma lipids may be partly explained by the considerable abundance of 18:2n-6 over 20:4n-6 and 22:6n-3 in human milk, infant formula and intravenous lipids, other physiological changes that accompany the provision of triglycerides as a major source of energy are likely to be involved.

Another approach is to derive an estimate of need based on accretion of fatty acids in fetal tissue. Analyses of autopsy tissue from fetus from 22 weeks gestation to term birth has estimated intrauterine accretion of 552 mg/day n-6 fatty acids and 67 mg/day n-3 fatty acids (mostly 22:6n-3) during the last trimester of gestation, most of which is accumulated in white adipose tissue (368 mg n-6 and 52 mg n-3 fatty acids/day) [64], [65]. In these estimates, fetal brain accretion amounted to 5.8 mg n-6 and 3.1 mg n-3 fatty acids/day, equivalent to about 1.1% and 4.65% of the total body accretion of the n-6 and n-3 fatty acids, respectively. Whether the brain is protected during limited availability of 22:6n-3 is not known; however, the ease with which fetal brain 22:6n-3 is altered by maternal dietary n-3 fatty acid intakes suggests that the fetal brain is sensitive to perturbation of membrane lipid composition by changes in 22:6n-3 supply [19], [23]. Assuming an average of 3.7 g fat/dL human milk, milk with 0.2–0.4% fatty acids as 22:6n-3 [66] would provide the 1 kg preterm infant fed at full enteral feeds of 180 mL/day with only 13–25 mg 22:6n-3/day, an amount clearly below the in utero rate of accretion of 52 mg/day. Dose response studies to determine whether the provision of higher intakes of 22:6n-3 for preterm infants would facilitate development and tissue accretion of 22:6n-3 which more closely resembles those of the third trimester fetus and thus of infants at term birth is a subject for future research.