Main

IUGR is associated with increased morbidity and mortality in the newborn(1) and follow-up studies have shown that further problems occur in later childhood(2,3). Low birth weight has also been associated with an increased incidence of hypertension, diabetes, stroke, and myocardial infarction in adult life(4). Some small babies are normal and healthy, but most small babies have suffered an adverse intrauterine environment. While some risk factors for low birth weight such as smoking and preeclampsia are well recognized, the mechanisms to link cause and effect in IUGR are not understood. It is unlikely that there is a single physiological mechanism to explain all cases.

Several barriers to transfer of solute between maternal and fetal circulations exist in the human hemochorial placenta. Two of these are the maternal-facing (microvillous) and the fetal-facing (basal) plasma membranes of the syncytiotrophoblast-the transporting epithelium. The MVM is in direct contact with maternal blood. Several distinct energy-dependent amino acid transporters, with differing substrate preferences, kinetic properties, and membrane orientations are present in the microvillous and basal plasma membranes(5). Fetal umbilical venous plasma amino acid concentrations are higher than maternal arterial concentration in human pregnancy and levels within placental trophoblast are even higher(6). Experiments on isolated human placental cotyledons, simultaneously perfused from the maternal and fetal sides, have shown that the placenta can take up amino acids from either circulation against a concentration gradient(7). These observations make it likely that transporters present in the MVM have a more important effect on net transfer of amino acids from mother to fetus than those in the basal membrane.

In postnatal life, amino acids are used almost exclusively for building proteins in tissues. By contrast, in prenatal life amino acids serve both as a substrate for fetal growth and also as a significant fuel source for general metabolic needs(8). Furthermore, there is evidence that control of the transport of neutral amino acids in mammalian epithelial cells may be a mechanism by which hormonal influences regulate cellular growth(9,10). Small neutral amino acids such as glycine and alanine are transported by the SysA transporter that is present in all human tissues, including the placenta, and subject to a variety of regulatory mechanisms(11–14).

There is now good evidence that placental amino acid transport is lower in the IUGR or SGA, compared with AGA, baby. Cordocentesis data have shown a significantly lower total plasma amino acid concentration in third trimester SGA fetuses than in AGA fetuses, reflecting significantly lower concentrations of several individual amino acids(15). Furthermore, it has been shown that SysA activity in MVM vesicles prepared from the placentas of SGA babies is reduced compared with that from normal birth weight babies(16–18). However, there are presently no published data comparing the relationship between placental SysA activity and measures of fetal growth in AGA versus SGA birth weight groups. The aim of this study therefore was to investigate concurrently the relationship between placental SysA activity and anthropometric measures of size at birth in term AGA and SGA babies.

METHODS

Sample selection. Caucasian babies from singleton pregnancies born at St. Mary's Hospital, Manchester, between 37-42 wk gestation (term) without evidence of major congenital malformation, whose mothers gave informed consent, were included in this study. Approval for this study was given by the Ethics Committee of the Manchester Health Commission. Birth weight category (SGA/AGA) was assigned using Gairdner-Pearson growth charts(19). SGA was defined as less than the third centile. Pregnancies were included only if menstrual dates had been confirmed by an ultrasound scan done between 16-20 wk gestation and a postnatal Dubowitz clinical assessment of gestational age(20) did not differ by >2 wk. Twenty five SGA and 24 AGA babies/placentas were included for study.

Placental microvillous membrane isolation. Placental processing was begun within 1 h of delivery. A membrane fraction enriched for MVM was isolated by an established method involving Mg2+-precipitation and differential centrifugation of homogenized placental tissue(21). The purity of the MVM fraction was assessed by measurement of the enrichment of ALP activity in the MVM sample compared with that in placental homogenate. ALP (EC 3.1.3.1) is known to be abundantly present on and localised to the placental MVM(22). More than 10-fold enrichment was required before a sample was included for study, but no sample was rejected because of inadequate enrichment.

Amino acid uptake studies. SysA activity in MVM vesicles was measured as the initial rate (30-s time point) of Na+-dependent uptake at room temperature of MeAIB, a radiolabeled, nonmetabolizable amino acid substrate specific to SysA, using an established method(17,18,23). Uptakes were corrected for the protein content of the vesicle samples, assessed by the Lowry method(24). 14C-MeAIB uptake was expressed as nmol/mg vesicle protein/30 s. All analyses were completed within 48 h of vesicle preparation.

Anthropometry. Babies were measured on the second postnatal day (24-48 h old) using a nonstretch paper tape for circumferential measurements, Harpenden skin calipers (British Indicators Ltd., Burgess Hill, UK) for skinfold thicknesses, and a Harpenden neonatometer (Holtain Ltd., Crymych, UK) for length measurement. Measurements of head circumference, abdominal circumference, midarm circumference, subscapular and triceps skinfold thicknesses, and supine length were taken. Birth weight, placental weight, and maternal height were obtained from the maternal case notes. Placental ratio is placental weight/birth weight, and ponderal index is birth weight/length3.

Statistics. The mean values of maternal and neonatal characteristics, and the similarity of the MVM preparations between the AGA and SGA groups were compared by independent t tests unless otherwise stated. Correlations between anthropometric measurements and SysA activity were assessed by Spearman rank correlation testing. Results with a probability of >0.05 were not considered significant. Statistical analyses were performed using SPSS for Windows, version 6.1.3 (SPSS Inc., Chicago, IL), or InStat, version 2.04a (GraphPad Software, San Diego, CA).

RESULTS

Microvillous membrane vesicle preparations. The enrichments of ALP activity in the MVM vesicles compared with placental homogenates and the protein recoveries did not differ, at the 5% level of significance, between SGA and AGA groups (Table 1). Mean SysA activity was significantly lower in the SGA group compared with the AGA group (Table 1).

Table 1 Characteristics of microvillous membrane vesicle preparations from SGA and AGA study groups (mean ± SEM)
Full size table

Maternal and neonatal characteristics. The mean gestation of the two groups of pregnancies was not significantly different: AGA babies were born at 276.8 ± 6.8 d of gestation and SGA babies at 280 ± 8.7 d gestation. Seventeen AGA babies and 23 SGA babies were born by vaginal delivery and 7 and 2, respectively, by Caesarean section; there was no significant difference in mode of delivery between the two groups (Fisher's exact test). Mothers of SGA babies were more likely to be smokers (p = 0.005) but were not otherwise different in terms of age, height, gravidity, or parity (Table 2). A repeated measures analysis of variance test with Bonferroni correction showed no difference between SGA smoking (0.020 ± 0.003 nmol/mg vesicle protein/30 s, n = 16) and nonsmoking (0.040 ± 0.008 nmol/mg vesicle protein/30 s, n = 9) subgroups or between AGA smoking (0.036 ± 0.007 nmol/mg vesicle protein/30 s, n = 4) and nonsmoking (0.047 ± 0.006 nmol/mg vesicle protein/30 s, n = 20) subgroups, with respect to mean SysA activities, but SGA/smoking and AGA/nonsmoking subgroups were different (p < 0.01). No mothers from either group had preeclampsia. Babies in the SGA group were smaller on all measurements at birth and had smaller placentas, but placental ratios were not different (Table 3).

Table 2 Comparison of maternal characteristics of AGA and SGA groups (mean ± SEM)
Full size table
Table 3 Comparison of neonatal characteristics of AGA and SGA groups (mean ± SEM)
Full size table

Correlations between SysA activity and measurements of size at birth (Table 4). In SGA babies, subscapular skinfold thickness, triceps skinfold thickness, placental weight, and placental ratio were positively correlated with placental SysA activity. In AGA babies, only the placental ratio was significantly correlated with placental SysA activity, and the correlation was negative. Thus correlations in opposite directions were found for the relationship of placental ratio to SysA activity in the SGA group compared with the AGA group.

Table 4 Spearman rank correlation coefficients for anthropometric measures against microvillous membrane system A amino acid transporter activity within birth weight categories
Full size table

DISCUSSION

The data in this study confirm and extend previous reports(16–18) that mean placental MVM SysA amino acid transporter activity is lower in SGA babies than in AGA babies. This is therefore a robust finding, likely to be of some importance to a full understanding of the pathophysiology of growth restriction in utero. The current study provides the first information on associations between SysA activity and a variety of anthropometric measures within birth weight categories (SGA versus AGA). The thinness and "head sparing" characteristic of asymmetric IUGR was evident in the lower MAC:HC ratio and ponderal index of the SGA group. This suggests that the majority of SGA babies whose placentas we studied suffered an adverse intrauterine environment which prevented them achieving their true growth potential rather than suffering from any intrinsic cause of small stature at birth, i.e. they were IUGR, not just SGA.

There was a higher incidence of smoking in the SGA group versus the AGA group. According to estimates by Liu and Jarvis(25), smoking may have contributed up to 200 g to the 1046 g difference in mean birth weight between the two groups. Nicotine is known to decrease SysA activity(26) so the increased prevalence of maternal smoking in the SGA group could have contributed to the lower 14C-MeAIB transport values. However, the lack of any significant difference in SysA activity between SGA smoking and nonsmoking groups or between AGA smoking and nonsmoking groups did not support this possibility. The significant difference we did find in SysA activity between SGA/smoking and AGA/nonsmoking subgroups was probably due to the known association between birth weight category and SysA activity, rather than an effect of smoking per se. This conclusion is supported by the fact that the difference in SysA activity between SGA and AGA groups found here is similar to that of our previous study(17), where there was no difference in the incidence of maternal smoking between the two groups.

Some consideration needs to be given to the power of this study. Clearly, a sample of 25 babies from 3% of the population, as in the SGA group, is more likely to produce an observed mean closer to the true population mean than is a sample of 24 babies from 94% of the population, as in the AGA group. Not only are the mean values for the SGA group more likely to be representative, but the scatter of the data are likely to be less. The increased scatter of data in the AGA group will have contributed to the weaker correlations found for nearly every measure in that group compared with those for the SGA group. This, of course, does not negate the biological significance of correlations seen within the SGA group but does warrant caution against over interpretation of the lack of significant correlations in the AGA group. This is emphasized by a recent much larger study of the relationship between placental SysA activity and fetal size in normal pregnancy, where significant inverse correlations were found(27). Seen in this light, the situation with placental ratio in our study, where there are significant correlations within both groups but in opposite directions, becomes even more interesting.

Skinfold thicknesses are a surrogate measure of subcutaneous fat stores, which reflect the adequacy of intrauterine nutrition in the third trimester of pregnancy28. The SGA group had significantly less fat stores that the AGA group, indicating they were in nutritional difficulty in utero. In the SGA group, there was a positive correlation between skinfold thicknesses at both sites and SysA activity. This trend was not seen in the AGA group. There is no obvious link between lipid and amino acid transport/metabolism in the placenta or fetus to explain these findings. It may reflect a generalized derangement of anabolic metabolism in the IUGR fetus and/or placenta. Another possibility is that it reflects differences in plasma membrane lipid composition, which might affect transporter activity. There is certainly evidence of differences in fatty acid composition of placentas from SGA babies compared with that of placentas from AGA babies(29), although the SGA group in this study included both preterm and term babies, whereas the AGA group were all term.

Placental ratio was the only measure which correlated significantly with SysA activity in both groups, but the direction of correlation was opposite in the two groups. That is to say, we found that for term SGA babies of a given birth weight, the larger the placenta, the higher the SysA activity but for term AGA babies of a given birth weight, the larger the placenta, the lower the SysA activity. Speculating on the possible reasons for a change in a ratio is difficult when it is not known whether the numerator or denominator exerts the greater influence. We believe the following arguments could be relevant to these findings. If AGA babies are considered to be healthy, i.e. they have attained a weight appropriate for their genetic potential, then it would follow that, as the placenta gets larger for a given birth weight, there is less need for transfer of amino acid (and other nutrients) per gram of placenta, because the genetic target weight would not alter. This is consistent with other data recently reported for normal pregnancy(27). However, many SGA babies are unhealthy, i.e. they have been subjected to an adverse intrauterine environment and are probably below their genetic target weight at birth. The observation of a larger placenta and higher SysA activity at a given SGA birth weight (lower placental ratio) could represent increasingly strenuous placental adaptation to supply more substrate for fetal growth. Those placentas which adapt the most (larger placenta, higher placental ratio) produce the largest babies (still SGA, but perhaps less so than they would have otherwise been).

If different mechanisms operate in appropriately grown fetuses compared with IUGR fetuses, then this would apply to any fetus subjected to an adverse intrauterine environment, even those with an AGA birthweight but who are nevertheless below their genetic potential because of growth restriction in utero. These data suggest there may be a fundamental difference in the way that SysA activity relates to fetal size depending on the adequacy of the intrauterine environment. Not enough is known about regulation of SysA activity in the placenta to allow speculation as to what mechanism(s) could mediate these effects.

Fetal growth is a dynamic process over the length of gestation. Studies at term look at the end-point of that process but may miss important events at earlier stages. There is evidence that SysA activity increases during gestation(18,30) and it will be of importance in the future to repeat the current study at what might be more important stages of fetal growth, such as mid-trimester(31).

In conclusion, these data suggest that the mechanism(s) by which anthropometric measurements in the newborn are related to SysA amino acid transporter activity in placental syncytiotrophoblast microvillous membrane may not operate similarly across the whole range of birth weights. Specifically, mechanisms may operate differently in the placentas of SGA fetuses compared with AGA fetuses.