Foetal cord blood contains higher portions of n-3 and n-6 long-chain PUFA but lower portions of trans C18:1 isomers than maternal blood

Background/objective An adequate supply of long-chain polyunsaturated fatty acids (LC PUFA) promotes foetal health and development, whereas generally, trans fatty acids (tFA) are considered to negatively interfere with LC PUFA metabolism. Nevertheless, to date, limited data concerning separate trans C18:1, such as t9 and t11, are available for maternal and foetal blood. Therefore, in this study the portions of individual trans C18:1, LC n-6, and n-3 PUFA in lipids of maternal and foetal plasma and erythrocyte membranes of German mother and child pairs (n=40) were analysed. Results Portions of linoleic acid and α-linolenic acid as LC precursors were lower (~0.4-fold); whereas the metabolites arachidonic acid (AA, n-6) and docosahexaenoic acid (DHA, n-3) were significantly higher (~2-fold) in foetal than in maternal plasma and erythrocytes. The main tFA in maternal and foetal blood were elaidic acid (C18:1t9; t9) and vaccenic acid (C18:1t11; t11). Portions of t9, t10, t11, and t12 in foetal blood lipids were lower (~0.5-fold) compared with maternal blood. In foetal lipids, t9 was higher than t11. The t9 correlated negatively with eicosapentaenoic acid (n-3) and AA in maternal and foetal lipids; whereas t11 correlated negatively only with foetal total LC n-6 (plasma and erythrocytes) and n-3 PUFA (erythrocytes). No correlation between maternal tFA and foetal PUFA was observed. Conclusions ‘Biomagnification’ of LC n-6 and n-3 PUFA AA and DHA in foetal blood was confirmed, whereas single trans isomers were lower compared with maternal blood. Nevertheless, tFA intake, especially from industrial sources, should be as low as possible.

T he essential dietary fatty acids (FA) linoleic acid (C18:2 n-6; LA) and a-linolenic acid (C18:3 n-3; ALA) are necessary for the endogenous synthesis of the long-chain polyunsaturated fatty acids (LC PUFA) arachidonic acid (C20:4 n-6; AA) and docosahexaenoic acid (C22:6 n-3; DHA) (1). These metabolites, particularly DHA, are important structural components, such as of cell membranes of the retina and the nervous system, suggesting an essential role for foetal and infant brain development (1). Studies also show a positive relationship between AA status and infant growth (2), as well as between the direct intake of LC omega-3 (n-3) PUFA and the length of gestation, birth weight, and a reduced risk of preterm delivery (3,4). Because LA and ALA, as essential precursors of LC PUFA, cannot be synthesised de novo, the human foetus has to obtain these FA exclusively from the maternal diet by placental transfer. Foetal tissue is able to convert ALA into DHA and LA into AA as shown by Carnielli et al. (5) for premature infants, but the activity of elongation and desaturation enzymes is limited (6). Therefore, DHA and AA are predominantly derived by placental transfer from the mother in relation to the maternal FA metabolism and diet and accumulate in the foetus in utero (7,8). It could be shown that DHA and AA are preferentially transferred to the foetus compared with other FA by the placenta (9,10). This 'biomagnification' results in higher relative proportions of these LC PUFA in foetal blood  lipids compared with maternal circulating lipids (11). Most of the DHA is stored in foetal adipose tissue, probably to ensure a constant supply of DHA for the first 2 months after birth (11,12). Compared with the relatively stable AA levels, DHA in human breast milk is subjected to fluctuations depending on the maternal diet (13,14). Trans fatty acids (tFA) in foetal tissue are exclusively derived from the maternal diet, perinatal by placental transport and postnatal by breast milk (7,15). tFA are considered to exhibit adverse effects on essential FA and LC PUFA metabolism. The intake of essential FA can be reduced by the competition of tFA and essential FA or PUFA, respectively, which can result in negative effects regarding foetal and infant development (16). These negative effects are thought to be predominantly associated with a high intake of tFA from industrially produced foods such as hydrogenated vegetable oils, such as elaidic acid (C18:1 t9; t9), C18:1 t10 (t10), and trans C18:2 n-6, compared with trans isomers from natural sources such as dairy products, mainly vaccenic acid (C18:1 t11; t11) (17). In addition, ruminant-derived fat contains conjugated linoleic acids (CLA), which are isomers of LA with conjugated double bonds in trans and cis configurations. Studies show that ruminant-derived t11 and c9,t11 CLA in breast milk are associated with a lower risk of atopic outcomes in breast-fed children (18). In general, t11 and natural c9,t11 CLA exhibit beneficial health-related characteristics, such as anti-inflammatory effects as reviewed by Viladomiu et al. and Kuhnt et al. (19,20).
The aim of the present study was to gain more insight into the LC PUFA and CLA status of foetal and maternal lipids as well as the profile of individual trans C18:1 isomers, such as t9, t10, t11, and t12. Therefore, the FA profiles including single trans C18:1 isomers were determined in maternal and foetal plasma and the respective erythrocyte membranes as well as in milk samples.

Study subjects
The study protocol was approved by the ethics committee of the Friedrich Schiller University Jena and was part of the study no. 1354-06/04. Several cohorts were recruited in this study, for example, Enke et al. (21) referred to (2007Á2008).
The present study was conducted from May to November 2005 prior to that of Enke et al. (21) who analysed a different study population, at the Jena University Hospital. Existing data from this earlier cohort were reanalysed to gain more insight into the distribution of individual trans isomers representing the different sources. In addition to the study of Enke et al. (21) breast milk samples of a subgroup of the cohort were also analysed.
Written informed consent was obtained from all subjects. The study population consisted of 40 mothers with their term-born infants (Table 1); however, breast milk samples could only be obtained from a subgroup of 20 mothers. All mothers and infants appeared healthy and there was no indication of diseases that could affect placental function.
Collection of maternal and cord blood, and breast milk samples Blood from mothers was taken from the vena basilica at the time of birth and cord blood was obtained from the infants directly after its clamping. Blood was collected into the EDTA tubes. To separate erythrocytes and plasma, the blood was centrifuged at 1,500 )g for 10 min at 88C. In addition, breast milk ( Â15 ml) was obtained from a subgroup of mothers (n020) sampled at 3Á5 days postpartum [also called colostrum (22)]. Erythrocytes, plasma, and breast milk samples were stored at (808C. FA distribution in all samples was analysed immediately after taking the samples of all the 40 mother/child pairs simultaneously (storage up to a maximum of 6 months).
Fatty acid analysis FA analysis of purified FAME extracts was conducted via gas chromatography (GC, GC-17 V3; Shimadzu, Japan) equipped with an autosampler (AOC-5000) and a flame ionisation detector, as described by Kuhnt et al. (26). Two different GC procedures were performed using two  (27). The injector and detector temperatures were constant at 260 and 2708C, respectively, for both GC methods, using H 2 as the carrier gas, as described by Kuhnt et al. (28). In all analysed materials, the same 55 FA were integrated including trans C18:1 isomers, using the GC solution software (Shimadzu). Individual FAME were expressed as a percentage of total identified FAME peak areas (% of total FAME).

Statistical analysis
Statistical analysis of the data was performed using IBM SPSS statistics 19 (SPSS, Inc., Chicago, IL). Values are presented as mean with standard deviation (SD). P 50.05 indicates significant difference. The KolmogorovÁ Smirnov test was used to test the distribution of the data. Comparisons of two groups were performed with Student's t-test if the data were normally distributed. To calculate correlations, the Pearson correlation analysis was performed.

Results
Fatty acid profiles in maternal and foetal plasma and erythrocyte membranes Regarding relative proportions of PUFA in total maternal and foetal blood lipids, AA was the major LC n-6 PUFA, whereas DHA was the major LC n-3 PUFA ( Table 2). Levels of the precursors LA (n-6) and ALA (n-3) were significantly lower in foetal plasma and erythrocyte membranes (0.3-to 0.5-fold) compared with maternal plasma and erythrocytes (Table 2 and Fig. 1). In contrast, the respective LC metabolites such as g-linolenic acid (GLA), dihomo-g-linolenic acid (DHGLA), and AA (n-6), as well as docosapentaenoic acid (C22:5 n-3; DPA; only plasma) and DHA (n-3) showed up to 2.7-fold elevated levels in foetal plasma and erythrocyte membranes compared with maternal lipids ( Table 2); except DPA and eicosapentaenoic acid (C20:5 n-3; EPA) in erythrocytes (Fig. 1). A significant correlation could be detected between maternal and foetal lipids for AA (plasma and erythrocytes) and for DHA (erythrocytes; Table 2). The major trans C18:1 isomers in maternal blood lipids were t9 and t11 (Table 2). Interestingly, the proportions of trans C18:1 isomers such as t9, t10, t11, and t12 were significantly lower in foetal than in maternal plasma (0.5-to 0.7-fold) as also in erythrocyte membranes (0.3-to 0.6-fold; Fig. 2). In foetal plasma and erythrocyte lipids, t11 was lower than t9, resulting in a higher foetal t9/t11 index compared with the maternal t9/t11 index. In plasma, t9 and t10 were significantly correlated between maternal and foetal lipids, but not t11 ( Table 2).
The main CLA isomer c9,t11 CLA was generally higher in plasma compared with erythrocyte lipids. In addition, the maternal c9,t11 CLA portion was higher compared with the corresponding foetal lipids and showed a positive correlation. Generally, in maternal and foetal plasma, c9,t11 CLA portions were clearly higher compared with t11, but in erythrocytes both levels were fairly similar ( Table 2).
Correlation of trans C18:1 isomers with PUFA Within maternal and foetal plasma and erythrocyte lipids, t9 was consistently inversely correlated with total LC n-3 PUFA and EPA as well as with LC n-6 PUFA and AA; whereas t10 only in maternal plasma. Negative correlations for t11 with total LC n-6 PUFA in plasma and erythrocytes and with total LC n-3 PUFA in plasma were exclusively detectable within foetal lipids, and for t12 only with EPA in maternal erythrocytes (Table 3).

Discussion
DHA and AA are important for foetal growth and development. They can be transferred to the foetus primarily via placental transfer or breastfeeding after birth (28,29). Although AA and DHA can also be synthesised from precursors in foetal tissue, the foetal availability of these FA, especially DHA, depends on the maternal diet.
The concentration of maternal plasma triglycerides elevates progressively through pregnancy (29). Compared with the foetus, the mother has a higher absolute concentration of all plasma FA, including DHA and AA; however, the relative proportion of DHA and AA in FA of total circulating lipids, as we analysed, is consistently higher in the neonate (11). This was called 'biomagnification' (30). In the present study, it was clearly shown that the precursor portions of LA and ALA were higher in maternal compared with foetal plasma, whereas their LC metabolites, especially the final products such as AA and DHA, were elevated in foetal plasma. Similar results were found in lipids of isolated cell membranes of erythrocytes [consisting of 85% phospholipids (31)]. The measurement of FA in erythrocytes, however, may best reflect a longterm accumulation of FA indicating dietary intake and endogenous metabolism (31). In addition, lower precursor/ product ratios (e.g. LA/AA, ALA/DHA) were found in foetal lipids in both plasma and erythrocyte membranes. These recent results confirmed the 'biomagnification' process, resulting from a preferential transfer of AA and DHA to the foetus (11,30). Moreover, foetal tissue is able to synthesise AA with an enhanced capacity compared with DHA (11) as indicated due to higher foetal plasma portion of AA compared with DHA in relation to maternal plasma (2.7-fold vs. 2.0-fold, respectively). Positive correlations between maternal and foetal PUFA in blood lipids were found for AA, DHA, DPA, and for total LC n-3 and n-6 PUFA (not consistent for plasma and erythrocytes). Similar results were obtained by Elias and Innis (7) and Enke et al. (21).  In contrast, tFA are not synthesised in foetal tissue but are exclusively transported by placental transfer or postnatal by breastfeeding (16). Therefore, foetal blood tFA originate from a maternal diet containing industrial hydrogenated vegetable oils (mainly t9 and t10) and ruminant fat (mainly t11) (7,16,32). Several studies suggest that tFA interfere with LC PUFA metabolism and, thereby, compromise foetal development (16,32), whereas only a few studies distinguish between single trans isomers. For example, high concentration of t9 in maternal phospholipids was associated with lower birth weight in Dutch pregnant women (33). In another study from the Netherlands, lower birth weight and head circumference were significantly and negatively related to the t9 concentration in cord lipids (34), whereas t11 data were not available.
The present total trans C18:1 levels in maternal plasma lipids were higher compared with foetal plasma lipids, ranging from 0.47 to 0.60% FAME. These results can be associated with low maternal tFA intake (B1% of energy intake, En%) (35,36), which is also supported by data presented by the Federal Institute of Risk Assessment Germany who estimated that the tFA intake in Germany (2005Á2007) was below 1 En% as reviewed by Kuhnt et al. (20). The results of the present study are in line with other studies that revealed higher proportions of tFA in maternal lipid fractions (21,37,38). The portions of individual trans C18:1 isomers (t9, t10, t11, and t12) were significantly lower in foetal plasma compared with maternal plasma (Â0.5-fold), confirmed by lower portion in the membrane lipids of erythrocytes (Fig. 2). Enke et al. (21) detected comparable portions of single trans C18:1 isomers in maternal and foetal lipids, also with generally lower portions in foetal lipids ( Â0.5-fold). The t11 in maternal lipids increased in relation to a high dairy fat consumption, but without increasing t11 in foetal lipids (21). Taken together, this indicates a reduced transfer of trans C18:1 to the foetus. Additionally, tFA might be located in an unfavourable position in triglycerides (sn-1 and sn-3) to be hydrolysed and released by placental lipoprotein lipase, which predominantly hydrolyses FA in sn-2 position as also discussed by Mü ller et al. for CLA (37). In the present study, t9 and t10 correlated positively between maternal and foetal plasma, whereas t11 showed no association (Table 2). On the contrary, Enke et al. (21) showed that independent of dairy consumption, maternal t11 was highly correlated with foetal t11, whereas t9 was not.
In dairy products, c9,t11 CLA, as the major CLA, originates mainly from the D9-desaturation of t11 (39). Studies showed that dairy fat intake correlated with CLA concentrations in plasma (21), adipose tissue (40), and breast milk (36). The present study revealed higher levels of c9,t11 CLA compared with its precursor t11 in plasma fractions, which could have resulted from the preferential conversion of t11 to c9,t11 CLA (26). The possibility of D9-desaturation distinguishes the t11 from the t9, t10, and t12 isomers (26). In erythrocytes, c9,t11 CLA was generally lower than in plasma lipids. Elias and Innis (7) showed that CLA could cross the human placenta and could be transferred to the foetus. Also, Mü ller et al. (37) reported a linear relationship between CLA in maternal and foetal blood lipids, as found in the present study, indicating placental CLA transfer. The natural CLA derived from dairy products may exhibit beneficial effects on foetal and infant health (16Á18, 41).
tFA may interfere with LC PUFA metabolism, for example, by blocking the transfer of PUFA to the foetus or by inhibiting the conversion of precursors into LC PUFA (7,16,42,43). The influence of individual tFA on LC PUFA status is not sufficiently analysed. In the present study, only t9 correlated consistently inversely with EPA and AA within both maternal and foetal lipids, whereas t10 only in maternal plasma (Table 3). Similar correlation trends were reported (21). In a recent study, t9 was also inversely associated with LC PUFA, such as AA and DHA in the erythrocytes of pregnant women and in umbilical lipids; unfortunately, t11 was not analysed (44). In a recent study, Chisaguano et al. (45) showed that t9 was negatively and t11 positively correlated to n-3/n-6 PUFA, whereas c9,t11 correlated positively with n-3 PUFA in maternal plasma. However, the present work and that by Enke et al. (21) found no significant correlation between maternal trans C18:1 and foetal n-6 and n-3 PUFA, which strengthens the assumption that the relatively low maternal tFA levels did not directly influence the LC PUFA levels of foetal lipids.
FA composition of breast milk is influenced by a variety of variables (e.g. maternal nutritional status, dietary intake, and length of lactation) (22). tFA (46) and CLA (35) in human milk are elevated by maternal intake of these FA. The analysed breast milk lipids contained 1.4% trans C18:1, which can be associated with low maternal tFA intake ( B1% of energy intake) (35,36). In concordance with reducing the content of industrial tFA in foods, the tFA content in breast milk also decreased (46). The relatively high t9/t11 index and the moderate c9,t11 CLA portion in the breast milk samples indicate a maternal diet relatively low in dairy fat, as previously estimated (41).
In conclusion, the present data confirm higher portions of LC n-3 and n-6 metabolites, whereas portions of individual trans C18:1 isomers such as t9, t10, t11, and t12 were lower in foetal plasma and erythrocyte membranes compared with maternal blood lipids. The generally low trans C18:1 portions in this German cohort indicate weak influence on foetal PUFA status. However, it is generally recommended that tFA intake, especially of industrially derived tFA, should be as low as possible. Further research is necessary to elucidate more precisely the role of single tFA isomers in maternal and foetal metabolism, with special regard to industrial and ruminant origins. AA0arachidonic acid; CLA 0conjugated linoleic acids; EPA 0eicosapentaenoic acid; f0foetal; LC0long-chain ( ]C 20 ); m0maternal; PUFA 0polyunsaturated fatty acids; r0Pearson's correlation coefficient. a Within FA of maternal or foetal plasma and erythrocyte fatty acids (*P B0.05, **PB0.01, ***P B0.001); b between maternal and foetal blood lipids ( # PB0.05, ## PB0.01, ### PB0.001). assistance in participant recruitment and sample collection. In particular, they also thank Peter Mö ckel, Christiane Vollhardt, and Kathrin Kü hnert for their participation in fatty acid analysis.