- Email: [email protected]
Regional variations in the levels of zinc, iron, copper, and calcium in the term human placenta
Placenta (1987), 8, 497-502
Regional Variations in the Levels of Zinc, Iron, Copper, and Calcium in the Term Human Placenta
ELIZABETH A. MANCI” & WILL R. BLACKBURN Department of Pathology, University of South Alabama, Fillingim Street, Mobile, AL 36617, USA
2451
a To whom correspondence should be addressed Paper accepted 25.3. I 987
INTRODUCTION Trace metal analysis of the human placenta is particularly useful in studies of the effects of environmental pollution on placental function and birth defects (Karp and Robertson, 1977; Since the volume of tissue required for these studies is small Khera and Wibberley, I&). (I per cent or less of the total weight of the placenta), obtaining a representative sample offers difficulties. Some investigators have resorted to homogenizing the entire placenta; this method is cumbersome and greatly increases the risk of sample contamination (Moody, 1982; Tsuchiya et al, 1984). Other investigators have analysed random samples from unspecified sites within the placental disc (Baglan et ai, 1974; Karp and Robertson, 1977; Tsuchiya et al, 1984); the validity of this randomized sampling has not been tested. Since recent morphological studies of the placenta have emphasized the heterogeneity of the parenchyma, there may be corresponding regional variations in the trace metal content (Bacon, Gilbert and Longo, 1986). The sampling site could be a significant factor influencing the analytical results and, therefore, responsible for discrepancies in the published data. In this study, correlations between the location of the sampling site and the levels of four elements (iron, zinc, copper and calcium) were investigated.
MATERIALS AND METHODS More than 147 tissue samples were collected from 21 uncomplicated term pregnancies (38 to 42 weeks, menstrual age) following review of the medical records for the following clinical data: ’ race, maternal age, menstrual age of gestation, infant sex, gravidity, and birthweight. To avoid the seasonal variations in metal content reported for some elements (e.g., calcium: Dawson et al, 1968), all tissues were obtained from deliveries during a three-week interval in December, 1985. The placentae were placed in polyethylene bags and stored at 4%. Within 12 h, the placentae were weighed and dissected. Contact of tissue with metals was avoided by using plastic knives and polyethylene gloves and work surfaces, acid-washing equipment, and handling the 497
Placenta (1987), Vol. 8
Figure J. Schematic depiction of the areas of the placental disc that were sampled.
tissue as little as possible. Full-thickness tissue samples were obtained from the umbilical cords (midway between insertion on disc and the fetal insertion), fetal membranes (midway between margin of disc and point of rupture, where this distance was greatest), and placental discs. Each placental disc was sampled in several regions: peri-insertion (within 3 cm of the insertion of the umbilical cord), mid-disc (midway between the cord insertion and margin where this distance was greatest), and periphery (the outermost 3 cm of the disc margin) (Figure I). Interlobular spaces were avoided. The middisc samples were divided horizontally into equal maternal and fetal halves and each half was analysed independently. Rather than duplicating a full-thickness mid-disc sample, the levels from the maternal half and the fetal half were added together to obtain the level for the full-thickness mid-disc region in each case. At each sampling site, 20 g of tissue were obtained for ashing. This relatively generous sample size was employed to ensure representative blocks of tissue even though the sensitivity of the assay for calcium was reduced as much as 19 per cent when compared with the levels obtained using parallel sample sizes of 7g. Nevertheless, since the 20-g sample size was used for all the samples in this study, all samples were subjected to the same conditions and the data are valid for a comparative study of the elemental levels at various sites within these placentae. The tissue samples were vacuum dried at 18 mmHg/r3o”C for 36 h to a constant weight and ashed in a muffle furnace at 490°C for 24 h. The ashes were dissolved in IO ml concentrated hydrochloric acid (trace metal grade), heated at 90°C for 30 min, and diluted to 25 ml with distilled deionized water. The National Bureau of Standard’s bovine liver (1577a) was employed for standardizing the method; the range of the standard errors for the various metals was O.I to 3-2 per cent. The variability between 15 pairs of duplicate samples analysed independently was less than 17 per cent (range 0.4 to 16.8 per cent); the mean recovery of s&xl (fort&xl) samples was within IO per cent the added quantities for the various metals. The added quantity was determined as 30 per cent of the expected conce~tmtion fsr the sample site and element studied. The levels of metals were measured with a Perkin-Ehner 503 spec&ophotometer
of
Mawr, Blackburn: Regmal Placental Metals
499
using hollow cathode lamps and air-acetylene flames. For zinc, the wavelength of 213.9 nm and slit width of 0.7 nm were used; for copper, 324.8 nm and 0.7 nm; for iron, 248.3 nm and 0.2 nm; for calcium, 21 I nm-vis and I .4 nm. For each sample, three readings were made and averaged. The data, which were calculated as micrograms per gram of dry weight, may be converted to micrograms per gram of wet weight by dividing by the following constants: for placental periinsertion and mid-disc samples (fetal and maternal halves), 6.2; for placental periphery, 5.3; for umbilical cords, 8.9; and for fetal membranes, 9.6. These constants were derived from the mean wet weight: dry weight ratios at each sampling site. For haemoglobin electrophoresis, 36 samples were studied from gestations with normal maternal levels of haemoglobin (IO to 13 g/dl, Miale, 1977) and with no family histories of haemoglobinopathies. A few drops of whole blood were collected from the placental parenchyma through a shallow incision in the fetal surface avoiding the large fetal vessels, and, similarly, a few drops were collected from a shallow incision in the maternal surface. For controls, a few drops of umbilical cord blood and maternal blood were also collected. For each sample, one drop of whole blood was lysed with six drops of 0.005 M ethylenediamine tetraacetic acid (EDTA) reagent (Helena) and centrifuged; the supernatant was electrophoresed at pH 8.2 (cellulose acetate). For each sampling site, the means, standard deviations and standard errors for each element were computed. The regional differences were compared using the Student’s unpaired t-test to determine significance. The null hypothesis (there are no differences between the regions) was rejected at the 5 per cent probability level. RESULTS The study population was 66 per cent Caucasian and 33 per cent Black; infants were 60 per cent male and 40 per cent female, the mean maternal age was 26. I years (range I 5 to 39 years), mean gravidity was 3.8 (range I to 12), mean birthweight was 3402.6 f 523.42 g (range 2556.0 to 4771.2 g), and mean placental weight was 441.96 f 93.55 g (range 279 to 677 g). The levels for zinc, iron, copper, and calcium at the seven sampling sites are summarized in Table I. Iron levels were highest in the peri-insertion and mid-disc regions; the levels in the periphery of the placental plate were 22 per cent lower (P < 0.025) than in the central (periinsertion and mid-disc) regions. Iron levels were 27 per cent lower (P < 0.0025) in the umbilical cords and 22 per cent lower (P < 0.0005) in the fetal membranes than in the central disc regions. The levels of zinc in the peri-insertion, mid-disc, and peripheral regions were not significantly different, but the zinc levels in the fetal half of the mid-disc regions were lower than in the maternal half (P < 0.0025). Zinc levels were 17 per cent lower in the fetal membranes Table I. Mean elemental levels (pg/g dry weight) f standard errors in placentae, fetal membranes and umbilical cords from 21 uncomplicated term gestations Sample site
Iron
Zinc
Placenta peri-insertion mid-disc fetal half
555 f 18 558 f 14 547 f 18
maternal half periphery Fetal membranes Umbilical cord
585 .jq 435 4Og
55.3 * 2.3 53.8 f 0.9 50.3 f 1.4 56.0 f 1.2 56.3 f 2.3 46.0 f 2.6 42.8 f 4. I
f 18 f 22 f 3’ f 42
Calcium
Copper
9.25 f 0.8 9.40 f ‘.4 IO.20
f
IO.70 8.98 17.20 II.10
f I.2 f 0.9 f 2.0 f 2.1
I.8
602
f
76
490 f 46 420 f 60 525 f 79 7’2 f 47 290 f 35 329
f
16
Placenta (1987), Vol. 8
500 Table z. Mean haemoglobinlevels (%) f standarderrorsin umbilical cord blood, maternal blood and placenta parenchymal blood collected from the fetal half and from the maternal half of the mid-disc region in nine uncomplicated term gestations AI
A2
Umbilical cord blood Placental blood
31 f I
of0
fetal half maternal half Maternal blood
69f8 95 f 1 98 f o
2fo If0 2fo
F
69+2 30 f 9 3fI of0
(P < 0.01) and 22 per cent lower in the umbilic$ cords (P < 0.01) than in the placental plate. Calcium levels in the peri-insertion and mid-disc regions were not significantly different, but calcium levels in the periphery were 39 per cent higher than in the central regions (P < 0.0025). Calcium levels in the fetal membranes were 44 per cent lower (P < 0.0025) and in the umbilical cords were 36 per cent lower (P < 0.0025) than in the mid-disc samples. Copper levels did not differ significantly among the five sites sampled within the placental plates or umbilical cords, but were 91 per cent higher in the fetal membranes (P < 0.0025) than in the placental disc. To test for measureable differences in the regional distribution of fetal and maternal blood by another method, haemoglobin electrophoresis was performed on blood aspirated from the pairenchyma of the fetal half of the mid-disc region and compared with that from the maternal half in nine uncomplicated term gestations. Umbilical cord blood and maternal whole blood samples were included for controls; these data confirmed that the infant and maternal haemoglobin levels were normal (Table 2). Using the percentage of haemoglobin F present as a marker for the relative amount of fetal blood present in each sample, significant (P < 0.025) regional variations were seen. The mean level of haemoglobin F in the fetal half of the mid-disc region was 30 f 9 per cent (range o to 61 per cent), and in the maternal half was 3 f I per cent (range 0 to 9 per cent).
DISCUSSION In studies of metals within the placenta the method of tissue sampling is critical because the volume of tissue to be represented is large (normal range at term 400 to 450 g), the volume of tissue required for the assay is small (6 to 8 g), and the substances to be measured are present in trace quantities (parts per million or parts per billion). Furthermore, there is recent evidence of regional anatomical and, probably, functional variations which betray the underlying heterogeneity of this tissue (Bacon, Gilbert and Longo, 1986). Finally, the abundance of these elements in the environment makes inadvertent contamination of these samples a major source of error, and demands meticulous attention to this phase of the assay. For these reasons, this study was initiated to scrutinize carefully the sampling procedure and define optimal methods for procuring representative samples of the placenta. To uncover significant regional variations, the effect of the location of the sampling site on the levels of four elements (zinc, iron, copper, and calcium) was studied. The levels of these elements were measured and compared at seven carefully defined locations within the placentae, umbilical cords, and fetal membranes. Significant regional differences were apparent and appear to correspond with regional differences in blood flow and tissue composition.
Manci, Blackburn: Regional Placental Metals
50’
Within the placental plate, iron levels were highest in the central (mid-disc and periinsertion) regions where the blood flow is also relatively greater. This finding probably reflects the relative distribution of the total blood volume within the placental plate and, consequently, the greater content of haemoglobin in the central regions. Zinc levels were greater in the maternal half of the mid-disc section than in the fetal half. This difference may reflect the greater levels of zinc reported in maternal blood at term, as compared to the fetal blood (Tscuhiya et al, 1984) and, therefore, the greater content of maternal blood in these samples. This explanation is supported by haemoglobin electrophoresis which demonstrated regional differences in the distribution of haemoglobin F within the placenta. By comparing the haemoglobin F levels of blood aspirated from the maternal half of the mid-disc region with that from the fetal half, the latter was seen to be ten times greater, indicating greater quantities of fetal blood in the fetal half of the placental plate and of maternal blood in the maternal half. Calcium levels were greatest in the periphery of the placental disc where degenerative changes (i.e., hyalinization) are seen normally. This finding suggests that calcium is preferentially deposited in this area. All of these elements (iron, zinc, and calcium) were present in the lowest levels in the fetal membranes and umbilical cord samples. Conversely, the levels of copper did not vary significantly among the five sites sampled in the placental disc, but were highest in the fetal membranes. Possibly, this finding reflects the regional distribution of copper binding proteins within the fetal membranes. This is the first study to challenge the method of randomized sampling and to document regional variations in placental elemental composition. These regional variations in trace elements probably contribute to the large standard deviations and lack of clinicopathological correlations which have limited the usefulness of this assay in previous placental studies. Retrospectively, the effects of the sampling site variations can be gleaned by comparing the variance reported by the rare investigators who controlled for such variations (Dawson et al, 1968) and those who did not (Baglan et al, 1974; Karp and Robertson, 1977; Tsuchiya et al, 1984). These difficulties with sampling are reminiscent of the problems encountered with the placental biopsy in previous years (Fox, 1978). Clearly, careful selection and definition of the sampling site is an essential step in elemental analysis of the placenta.
SUMMARY This study investigated the influence of the location of the sampling site during elemental analyses of 21 human term placentae. The levels of iron, zinc, copper and calcium in fetal membranes, umbilical cords and placental discs were measured by atomic absorption spectrophotometry and compared. The disc samples were obtained from central (peri-insertion and mid-disc fetal and maternal halves), and peripheral regions. Significant variations were found. Copper was present in highest levels (17.2 f 2.0 pg/g dry weight) in the fetal membranes. Calcium levels were highest (712 f 47 pg/g dry weight) in the periphery of the placental disc. Iron levels were highest (558 f 14pg/g dry weight) in the central regions of the disc. Zinc levels were lower (50.3 f I .4 pg/g dry weight) in the fetal half of the mid-disc regions than in the maternal half (56.0 f I .2 pg/g dry weight). This study demonstrates the importance of defining the location of the sampling site in studies involving elemental analysis of the placenta. ACKNOWLEDGEMENTS These investigations South Alabama.
were supported
by an Intramural
Research Grant from the College of Medicine, University
of
Placenta (1987), Vol. 8 REFERENCES Bacon, B. J., Gilbert, R. D. & Longo,
L. D. (1986) Regional anatomy of the term human placenta. Placenta, 7,233
24,.
Baglan, R. J., Brill, A. B., Schulert, A. et al. (1974) Utility of placental posure to adult and fetus. Envrwnmental Research, 8,64470.
tissue as an indicator of trace element ex-
Dawson, E. B., Croft, H. A., Clark, R. R. & McGanity, W. J. (1968) Study of seasonal variations in nine cations of normal term placentas. AmericanJournal of Obstetrics and Gynecology, 102,354-361. Fox, H. (1978) Pathology of the Placenta. Major Problems in Pathology, Volume 7, Placental Biopsy, pp. 48c-481. London: W. B. Saunders. Karp, W. B. & Robertson, A. F. (1977) Correlation of human placental enzymatic activity with trace metal concentration in placentas from three geographical locations. Environmental Research, 13,47~477. Khera, A. K. SCWibberley, D. G. (1983) Placental lead levels. Mutation Research, 113,267-268. Miale, J. B. (1977) Laboratory Medicine Hematology, pp. 425-426. St Louis: C. V. Mosby. Moody, J. R. (1982) The sampling, handling and storage of materials for trace analysis. Philosophical Transactions of the Royal Society of London, 305,669-680. Tsuchiya, H., Mitani, K., Kodama, K. & Nakata, T. (1984) Placental transfer of heavy metals in normal pregnant Japanese women. Archives of Environmental Health, 39, 1t-17.
Regional Variations in the Levels of Zinc, Iron, Copper, and Calcium in the Term Human Placenta
ELIZABETH A. MANCI” & WILL R. BLACKBURN Department of Pathology, University of South Alabama, Fillingim Street, Mobile, AL 36617, USA
2451
a To whom correspondence should be addressed Paper accepted 25.3. I 987
INTRODUCTION Trace metal analysis of the human placenta is particularly useful in studies of the effects of environmental pollution on placental function and birth defects (Karp and Robertson, 1977; Since the volume of tissue required for these studies is small Khera and Wibberley, I&). (I per cent or less of the total weight of the placenta), obtaining a representative sample offers difficulties. Some investigators have resorted to homogenizing the entire placenta; this method is cumbersome and greatly increases the risk of sample contamination (Moody, 1982; Tsuchiya et al, 1984). Other investigators have analysed random samples from unspecified sites within the placental disc (Baglan et ai, 1974; Karp and Robertson, 1977; Tsuchiya et al, 1984); the validity of this randomized sampling has not been tested. Since recent morphological studies of the placenta have emphasized the heterogeneity of the parenchyma, there may be corresponding regional variations in the trace metal content (Bacon, Gilbert and Longo, 1986). The sampling site could be a significant factor influencing the analytical results and, therefore, responsible for discrepancies in the published data. In this study, correlations between the location of the sampling site and the levels of four elements (iron, zinc, copper and calcium) were investigated.
MATERIALS AND METHODS More than 147 tissue samples were collected from 21 uncomplicated term pregnancies (38 to 42 weeks, menstrual age) following review of the medical records for the following clinical data: ’ race, maternal age, menstrual age of gestation, infant sex, gravidity, and birthweight. To avoid the seasonal variations in metal content reported for some elements (e.g., calcium: Dawson et al, 1968), all tissues were obtained from deliveries during a three-week interval in December, 1985. The placentae were placed in polyethylene bags and stored at 4%. Within 12 h, the placentae were weighed and dissected. Contact of tissue with metals was avoided by using plastic knives and polyethylene gloves and work surfaces, acid-washing equipment, and handling the 497
Placenta (1987), Vol. 8
Figure J. Schematic depiction of the areas of the placental disc that were sampled.
tissue as little as possible. Full-thickness tissue samples were obtained from the umbilical cords (midway between insertion on disc and the fetal insertion), fetal membranes (midway between margin of disc and point of rupture, where this distance was greatest), and placental discs. Each placental disc was sampled in several regions: peri-insertion (within 3 cm of the insertion of the umbilical cord), mid-disc (midway between the cord insertion and margin where this distance was greatest), and periphery (the outermost 3 cm of the disc margin) (Figure I). Interlobular spaces were avoided. The middisc samples were divided horizontally into equal maternal and fetal halves and each half was analysed independently. Rather than duplicating a full-thickness mid-disc sample, the levels from the maternal half and the fetal half were added together to obtain the level for the full-thickness mid-disc region in each case. At each sampling site, 20 g of tissue were obtained for ashing. This relatively generous sample size was employed to ensure representative blocks of tissue even though the sensitivity of the assay for calcium was reduced as much as 19 per cent when compared with the levels obtained using parallel sample sizes of 7g. Nevertheless, since the 20-g sample size was used for all the samples in this study, all samples were subjected to the same conditions and the data are valid for a comparative study of the elemental levels at various sites within these placentae. The tissue samples were vacuum dried at 18 mmHg/r3o”C for 36 h to a constant weight and ashed in a muffle furnace at 490°C for 24 h. The ashes were dissolved in IO ml concentrated hydrochloric acid (trace metal grade), heated at 90°C for 30 min, and diluted to 25 ml with distilled deionized water. The National Bureau of Standard’s bovine liver (1577a) was employed for standardizing the method; the range of the standard errors for the various metals was O.I to 3-2 per cent. The variability between 15 pairs of duplicate samples analysed independently was less than 17 per cent (range 0.4 to 16.8 per cent); the mean recovery of s&xl (fort&xl) samples was within IO per cent the added quantities for the various metals. The added quantity was determined as 30 per cent of the expected conce~tmtion fsr the sample site and element studied. The levels of metals were measured with a Perkin-Ehner 503 spec&ophotometer
of
Mawr, Blackburn: Regmal Placental Metals
499
using hollow cathode lamps and air-acetylene flames. For zinc, the wavelength of 213.9 nm and slit width of 0.7 nm were used; for copper, 324.8 nm and 0.7 nm; for iron, 248.3 nm and 0.2 nm; for calcium, 21 I nm-vis and I .4 nm. For each sample, three readings were made and averaged. The data, which were calculated as micrograms per gram of dry weight, may be converted to micrograms per gram of wet weight by dividing by the following constants: for placental periinsertion and mid-disc samples (fetal and maternal halves), 6.2; for placental periphery, 5.3; for umbilical cords, 8.9; and for fetal membranes, 9.6. These constants were derived from the mean wet weight: dry weight ratios at each sampling site. For haemoglobin electrophoresis, 36 samples were studied from gestations with normal maternal levels of haemoglobin (IO to 13 g/dl, Miale, 1977) and with no family histories of haemoglobinopathies. A few drops of whole blood were collected from the placental parenchyma through a shallow incision in the fetal surface avoiding the large fetal vessels, and, similarly, a few drops were collected from a shallow incision in the maternal surface. For controls, a few drops of umbilical cord blood and maternal blood were also collected. For each sample, one drop of whole blood was lysed with six drops of 0.005 M ethylenediamine tetraacetic acid (EDTA) reagent (Helena) and centrifuged; the supernatant was electrophoresed at pH 8.2 (cellulose acetate). For each sampling site, the means, standard deviations and standard errors for each element were computed. The regional differences were compared using the Student’s unpaired t-test to determine significance. The null hypothesis (there are no differences between the regions) was rejected at the 5 per cent probability level. RESULTS The study population was 66 per cent Caucasian and 33 per cent Black; infants were 60 per cent male and 40 per cent female, the mean maternal age was 26. I years (range I 5 to 39 years), mean gravidity was 3.8 (range I to 12), mean birthweight was 3402.6 f 523.42 g (range 2556.0 to 4771.2 g), and mean placental weight was 441.96 f 93.55 g (range 279 to 677 g). The levels for zinc, iron, copper, and calcium at the seven sampling sites are summarized in Table I. Iron levels were highest in the peri-insertion and mid-disc regions; the levels in the periphery of the placental plate were 22 per cent lower (P < 0.025) than in the central (periinsertion and mid-disc) regions. Iron levels were 27 per cent lower (P < 0.0025) in the umbilical cords and 22 per cent lower (P < 0.0005) in the fetal membranes than in the central disc regions. The levels of zinc in the peri-insertion, mid-disc, and peripheral regions were not significantly different, but the zinc levels in the fetal half of the mid-disc regions were lower than in the maternal half (P < 0.0025). Zinc levels were 17 per cent lower in the fetal membranes Table I. Mean elemental levels (pg/g dry weight) f standard errors in placentae, fetal membranes and umbilical cords from 21 uncomplicated term gestations Sample site
Iron
Zinc
Placenta peri-insertion mid-disc fetal half
555 f 18 558 f 14 547 f 18
maternal half periphery Fetal membranes Umbilical cord
585 .jq 435 4Og
55.3 * 2.3 53.8 f 0.9 50.3 f 1.4 56.0 f 1.2 56.3 f 2.3 46.0 f 2.6 42.8 f 4. I
f 18 f 22 f 3’ f 42
Calcium
Copper
9.25 f 0.8 9.40 f ‘.4 IO.20
f
IO.70 8.98 17.20 II.10
f I.2 f 0.9 f 2.0 f 2.1
I.8
602
f
76
490 f 46 420 f 60 525 f 79 7’2 f 47 290 f 35 329
f
16
Placenta (1987), Vol. 8
500 Table z. Mean haemoglobinlevels (%) f standarderrorsin umbilical cord blood, maternal blood and placenta parenchymal blood collected from the fetal half and from the maternal half of the mid-disc region in nine uncomplicated term gestations AI
A2
Umbilical cord blood Placental blood
31 f I
of0
fetal half maternal half Maternal blood
69f8 95 f 1 98 f o
2fo If0 2fo
F
69+2 30 f 9 3fI of0
(P < 0.01) and 22 per cent lower in the umbilic$ cords (P < 0.01) than in the placental plate. Calcium levels in the peri-insertion and mid-disc regions were not significantly different, but calcium levels in the periphery were 39 per cent higher than in the central regions (P < 0.0025). Calcium levels in the fetal membranes were 44 per cent lower (P < 0.0025) and in the umbilical cords were 36 per cent lower (P < 0.0025) than in the mid-disc samples. Copper levels did not differ significantly among the five sites sampled within the placental plates or umbilical cords, but were 91 per cent higher in the fetal membranes (P < 0.0025) than in the placental disc. To test for measureable differences in the regional distribution of fetal and maternal blood by another method, haemoglobin electrophoresis was performed on blood aspirated from the pairenchyma of the fetal half of the mid-disc region and compared with that from the maternal half in nine uncomplicated term gestations. Umbilical cord blood and maternal whole blood samples were included for controls; these data confirmed that the infant and maternal haemoglobin levels were normal (Table 2). Using the percentage of haemoglobin F present as a marker for the relative amount of fetal blood present in each sample, significant (P < 0.025) regional variations were seen. The mean level of haemoglobin F in the fetal half of the mid-disc region was 30 f 9 per cent (range o to 61 per cent), and in the maternal half was 3 f I per cent (range 0 to 9 per cent).
DISCUSSION In studies of metals within the placenta the method of tissue sampling is critical because the volume of tissue to be represented is large (normal range at term 400 to 450 g), the volume of tissue required for the assay is small (6 to 8 g), and the substances to be measured are present in trace quantities (parts per million or parts per billion). Furthermore, there is recent evidence of regional anatomical and, probably, functional variations which betray the underlying heterogeneity of this tissue (Bacon, Gilbert and Longo, 1986). Finally, the abundance of these elements in the environment makes inadvertent contamination of these samples a major source of error, and demands meticulous attention to this phase of the assay. For these reasons, this study was initiated to scrutinize carefully the sampling procedure and define optimal methods for procuring representative samples of the placenta. To uncover significant regional variations, the effect of the location of the sampling site on the levels of four elements (zinc, iron, copper, and calcium) was studied. The levels of these elements were measured and compared at seven carefully defined locations within the placentae, umbilical cords, and fetal membranes. Significant regional differences were apparent and appear to correspond with regional differences in blood flow and tissue composition.
Manci, Blackburn: Regional Placental Metals
50’
Within the placental plate, iron levels were highest in the central (mid-disc and periinsertion) regions where the blood flow is also relatively greater. This finding probably reflects the relative distribution of the total blood volume within the placental plate and, consequently, the greater content of haemoglobin in the central regions. Zinc levels were greater in the maternal half of the mid-disc section than in the fetal half. This difference may reflect the greater levels of zinc reported in maternal blood at term, as compared to the fetal blood (Tscuhiya et al, 1984) and, therefore, the greater content of maternal blood in these samples. This explanation is supported by haemoglobin electrophoresis which demonstrated regional differences in the distribution of haemoglobin F within the placenta. By comparing the haemoglobin F levels of blood aspirated from the maternal half of the mid-disc region with that from the fetal half, the latter was seen to be ten times greater, indicating greater quantities of fetal blood in the fetal half of the placental plate and of maternal blood in the maternal half. Calcium levels were greatest in the periphery of the placental disc where degenerative changes (i.e., hyalinization) are seen normally. This finding suggests that calcium is preferentially deposited in this area. All of these elements (iron, zinc, and calcium) were present in the lowest levels in the fetal membranes and umbilical cord samples. Conversely, the levels of copper did not vary significantly among the five sites sampled in the placental disc, but were highest in the fetal membranes. Possibly, this finding reflects the regional distribution of copper binding proteins within the fetal membranes. This is the first study to challenge the method of randomized sampling and to document regional variations in placental elemental composition. These regional variations in trace elements probably contribute to the large standard deviations and lack of clinicopathological correlations which have limited the usefulness of this assay in previous placental studies. Retrospectively, the effects of the sampling site variations can be gleaned by comparing the variance reported by the rare investigators who controlled for such variations (Dawson et al, 1968) and those who did not (Baglan et al, 1974; Karp and Robertson, 1977; Tsuchiya et al, 1984). These difficulties with sampling are reminiscent of the problems encountered with the placental biopsy in previous years (Fox, 1978). Clearly, careful selection and definition of the sampling site is an essential step in elemental analysis of the placenta.
SUMMARY This study investigated the influence of the location of the sampling site during elemental analyses of 21 human term placentae. The levels of iron, zinc, copper and calcium in fetal membranes, umbilical cords and placental discs were measured by atomic absorption spectrophotometry and compared. The disc samples were obtained from central (peri-insertion and mid-disc fetal and maternal halves), and peripheral regions. Significant variations were found. Copper was present in highest levels (17.2 f 2.0 pg/g dry weight) in the fetal membranes. Calcium levels were highest (712 f 47 pg/g dry weight) in the periphery of the placental disc. Iron levels were highest (558 f 14pg/g dry weight) in the central regions of the disc. Zinc levels were lower (50.3 f I .4 pg/g dry weight) in the fetal half of the mid-disc regions than in the maternal half (56.0 f I .2 pg/g dry weight). This study demonstrates the importance of defining the location of the sampling site in studies involving elemental analysis of the placenta. ACKNOWLEDGEMENTS These investigations South Alabama.
were supported
by an Intramural
Research Grant from the College of Medicine, University
of
Placenta (1987), Vol. 8 REFERENCES Bacon, B. J., Gilbert, R. D. & Longo,
L. D. (1986) Regional anatomy of the term human placenta. Placenta, 7,233
24,.
Baglan, R. J., Brill, A. B., Schulert, A. et al. (1974) Utility of placental posure to adult and fetus. Envrwnmental Research, 8,64470.
tissue as an indicator of trace element ex-
Dawson, E. B., Croft, H. A., Clark, R. R. & McGanity, W. J. (1968) Study of seasonal variations in nine cations of normal term placentas. AmericanJournal of Obstetrics and Gynecology, 102,354-361. Fox, H. (1978) Pathology of the Placenta. Major Problems in Pathology, Volume 7, Placental Biopsy, pp. 48c-481. London: W. B. Saunders. Karp, W. B. & Robertson, A. F. (1977) Correlation of human placental enzymatic activity with trace metal concentration in placentas from three geographical locations. Environmental Research, 13,47~477. Khera, A. K. SCWibberley, D. G. (1983) Placental lead levels. Mutation Research, 113,267-268. Miale, J. B. (1977) Laboratory Medicine Hematology, pp. 425-426. St Louis: C. V. Mosby. Moody, J. R. (1982) The sampling, handling and storage of materials for trace analysis. Philosophical Transactions of the Royal Society of London, 305,669-680. Tsuchiya, H., Mitani, K., Kodama, K. & Nakata, T. (1984) Placental transfer of heavy metals in normal pregnant Japanese women. Archives of Environmental Health, 39, 1t-17.