Main

PCBs are widespread toxic environmental pollutants. They are polycyclic aromatic compounds, and have a total number of 209 possible congeners, which differ in their degree of chlorination and the molecular position of the chlorine atoms. PCBs have extremely long half-lives and are strongly lipophilic. These substances are resistant to high temperatures, they can easily conduct heat, and have electrical insulating properties. PCBs have been used in a wide range of industrial products, including fire retardants, plasticizers, dielectrical fluids in capacitators and transformers, and hydraulic fluids. Their environmental persistence was recognized in the late 1970s, and PCBs were subsequently banned worldwide. PCBs have, however, been produced until the mid-1980s in, e.g. Eastern European countries. In the industrialized world residues can at present be found in water, soil, and biologic tissue. Due to their chemical stability, these compounds become increasingly concentrated while being transferred through the food chain. High PCB levels can be found in adipose tissue of organisms at the top of the food chain, including human beings(1).

PCBs enter the fetus via the placenta(2,3). In the postnatal period, they are transferred from the nursing mother to her child through breast milk(3). Exposure to "background" levels of PCBs is known to have adverse effects on child development. Although relatively large amounts of PCBs are ingested with breast milk, greatest risks have been associated with exposure during the prenatal period. Higher levels of intrauterine to PCBs, as determined by PCB levels in maternal and cord plasma, have been found to result in deficits in fetal and postnatal growth(4,5), a less optimal neurologic development at the ages of 2 wk(6) and 18 mo(7), a lower score on psychomotor developmental tests up to the age of 2 y(8–10), and a lower intelligence quotient at 11 y of age(11).

There are no data on the PCB body burden of the human fetus at different gestational ages and the distribution of these compounds among fetal organs. We analyzed the PCB contents in s.c. adipose tissue, liver, and brain of nine stillborn infants of varying gestational age.

METHODS

Samples. From March to December 1993, we collected tissue samples of fetuses who died in utero. Each eligible stillborn that was presented for obduction to one of the Public Health Laboratories located within a radius of 60 km from Groningen was included in the study. The Groningen region is a semiurban area in the northeast of The Netherlands. For inclusion, fetuses had to show no signs of serious chromosomal or congenital malformations. Also, the absence of signs of maceration was a prerequisite for inclusion. From each stillborn we collected 10 g each of s.c. white adipose tissue, liver, and brain. Depending on total brain size, tissue was sampled from the parietotemporal or parietal area. The samples were stored at-20°C until analysis. Birth weight, gestational age, presumed cause of death, and the age of the mother were recorded. Informed consent was obtained from the parents. The study protocol was approved by the local medical ethics committees and is in agreement with the Helsinki Declaration of 1975, as revised in 1989.

Analyses. Tissue levels of 26 PCB congeners were determined in the TNO Nutrition and Food Research Institute (Zeist, The Netherlands) as described previously(12). Briefly, tissue samples were weighed, homogenized, and extracted with organic solvent by means of a Soxhlet apparatus. Total fat was determined gravimetrically after evaporation of a part of the organic solvent layer to dryness. Another part of the extract was concentrated and purified by column chromatography on basic alumina that was previously deactivated with 10% water. An aliquot was analyzed for PCBs by gas-liquid chromatography/electron capture detection with the use of two capillary columns of different polarity. The recovery amounts typically were >90%, and the variation coefficient was <10%.

Data analysis and statistics. PCB levels were expressed on the basis of the extractable tissue fat content (ng/g of fat). We report only on the levels of the congeners 118, 138, 153, and 180 (International Union of Pure and Applied Chemistry nomenclature). The levels of these congeners are relatively high and accurately measurable. In Western Europe, they are considered to be markers for the levels of the toxicologic most relevant congeners (i.e. the non- and mono-ortho PCBs)(13). The sum of the levels of the congeners 118, 138, 153, and 180 (ΣPCB) was calculated. For each fetus we established the ratio of ΣPCB in liver and adipose tissue, and the ratio of ΣPCB in brain and adipose tissue. The congeneric distribution patterns of the four PCBs in adipose tissue, liver, or brain were obtained by normalization to 100% (g/100 g). Spearman rank correlation coefficients were calculated to evaluate the relationships between PCB levels in adipose tissue, liver, and brain, and to investigate associations between tissue PCB levels and gestational age. p values of 0.05 or less were considered statistically significant.

RESULTS

During the study period nine stillborns were found to be eligible for inclusion. The characteristics of the nine stillborn infants are presented in Table 1. Their median gestational age was 34 wk (range, 17-40 wk), median birth weight was 2050 g (162-3225 g), and median maternal age was 30 y (18-32 y). Three stillborns (nos. 5, 7, and 9) were considered small for gestational age, i.e. ≤10th percentile for gestational age according to Kloosterman(14). The presumed causes of death were chronic (nos. 4, 6, and 7) and acute (no. 5) placental insufficiency, abortion (no. 1), ventricular bleeding (no. 3), intrauterine pneumonia (no. 2), or unknown (nos. 8 and 9). No adipose tissue could be obtained from the fetus who died at 17 wk (no. 1) and also its liver PCB congener 180 was found to be below the detection limit.

Table 1 Characteristics of the stillborn infants
Full size table

For each of the fetal compartments, the correlations between the levels of the PCB congeners 118, 138, 153, and 180, and the ΣPCB on the one hand, and the gestational age on the other, were not significant; correlation coefficients varied between 0.22 and 0.47. The PCB levels in adipose tissue, liver, and brain (in ng/g of fat) are presented in Table 2. The median (range) liver/adipose tissue ratio of ΣPCB amounted to 0.8 (0.4-0.9) g/g, and the ΣPCB brain/adipose tissue ratio was 0.2 (0.1-0.3) g/g. There were strong relationships betweenΣPCB in adipose tissue and liver (r = 0.98;p < 0.01), and between ΣPCB in adipose tissue and brain(r = 0.91; p < 0.01). The congeneric distribution patterns of PCBs in s.c. adipose tissue, liver, and brain (in g/100 g) proved similar (Table 3).

Table 2 PCB levels in fetal subcutaneous adipose tissue, liver, and brain*
Full size table
Table 3 PCB congeneric distribution in fetal subcutaneous adipose tissue, liver, and brain*
Full size table

DISCUSSION

We report on the levels of the PCB congeners 118, 138, 153, and 180 in s.c. adipose tissue, liver, and brain of nine stillborn infants of 17-40 gestational weeks. PCB levels in The Netherlands are comparable to those found in other parts of the industrialized world, including the United States(8–10). Nevertheless, due to the small number of fetuses, the reported levels should be regarded as indicative rather than representative for tissue levels of the general population of unborn children.

Although the vulnerability of the fetal organs might differ, fetal organ-specific PCB levels and the accumulation of congeners in the different fetal organs can be used for risk evaluations. We found that fetal liver and brain ΣPCB levels, expressed on a fat weight basis, are 80 and 20%, respectively, of that encountered in s.c. adipose tissue. Congeneric distribution patterns did not differ among the organs. The tissue distribution pattern that we found is in agreement with that reported for other species(15,16). Different levels of the organs are likely to be caused by differences in the polarity of the fats in each of these organs. More specific it indicates the affinity of the highly apolar PCBs for the highly apolar (storage) lipids, notably triglycerides and cholesterol esters, and their moderate affinity for the (structural) amphipathic lipids, notably phospholipids and cholesterol. The cytoplasm of the fat cell is largely composed of a triglyceride droplet(17). The liver can be regarded as an intermediate storage place for triglycerides after the uptake of fatty acids or lipoprotein remnants from the circulation, or local de novo synthesis of fat from polar precursors. The cytoplasm may also contain some redundant cholesterol esters, but compared with adipose tissue there is a relatively higher contribution from cell membrane phospholipids and cholesterol. The fat in the brain derives merely from membranes and is therefore largely composed of phospholipids and cholesterol(18).

Because fat from both adipose tissue and human milk is almost exclusively composed of triglycerides, we investigated whether fetal adipose tissue contains, on a fat basis, similar PCB levels compared with human milk. For this we used our(6) previous PCB data of 93 mature milk samples from Dutch mothers who also lived in the Groningen area(Table 4). Fetal adipose tissue levels of PCB congeners 118, 138, 153, and 180, and also ΣPCB (Table 2) fell well within the corresponding ranges of milk levels, and also the fetal adipose tissue congeneric distribution (Table 4) proved to be comparable with that of human milk. The high degree of similarity between PCBs levels in maternal and fetal fat suggests that PCBs readily cross the placenta, and subsequently equilibrate among lipid compartments according to their high affinity for notably triglycerides and cholesterol esters, and the lower affinity for phospholipids and cholesterol. It should be stressed that, due to ethical reasons, we were not able to draw blood from the mother and investigate the individual ratios for maternal and fetal PCB levels. Therefore, we cannot exclude the existence of a partial placental barrier(2).

Table 4 Dutch mature human milk*: PCB levels (ng/g fat) and PCB congeneric distribution (g/100 g)
Full size table

In contrast to postnatal exposure via breast milk, prenatal exposure to PCBs has been found to be the most critical for future growth and development(4–11). During fetal organ development, the transient period of growth spurt is supposed to be one of special vulnerability to adverse influences. The timing and duration of this period of rapid cell multiplication varies between organs. As the fetal PCB levels were found to bear no relationship with the gestational age of the fetus, all of the developing fetal organs are equally at risk of prenatal PCB-induced toxicity. The independency of PCB levels of gestational age reflects a rapid equilibration between the fetal and maternal PCB stores. In the last 3 mo of gestation, an overproportional increase of adipose tissue relative to the total body weight takes place(19). Because the deposited fat is synthesized in the fetus mainly from polar precursors (glucose, lactate), the absence of any dilution effect suggests that this initially PCB-poor fat is rapidly provided with maternal PCBs. It should be kept in mind that, due to the cross-sectional study design and the small number of stillborn infants, a subtle effect of the gestational age cannot be excluded.

We conclude that maternal PCBs have a tendency to accumulate notably in fetal tissues that contain high levels of storage lipids (i.e. especially triglycerides). Consequently, on a fat basis liver contains 80% and brain 20% of the PCB levels encountered in adipose tissue. Fetal tissue PCB levels are not dependent on gestational age, and there are no organ-dependent differences in the distribution patterns of the PCB congeners 118, 138, 153, and 180. Fetal adipose tissue PCB levels compare well with those in human milk. These data suggest that the flux of maternal PCBs across the placenta is of sufficient magnitude to accomplish similar PCB levels in storage fat that is de novo synthesized in the fetus from polar precursors.