GENETIC ORIGIN AND PROPORTION OF BASAL PLATE SURFACE-LINING CELLS IN NORMAL AND ABNORMAL PREGNANCIES

Abstract

The human placenta is a transient organ, the villous surface of which is in direct contact with the maternal circulation during pregnancy. Thus, the syncytiotrophoblast and the basal plate-lining cells are considered continuous with the endothelial layer of the maternal vasculature. Two types of cells are found on the surface of the basal plate: trophoblasts (of fetal origin) and endothelial cells of putative maternal origin. Histologic abnormalities have been described in the basal plate of the placenta obtained from patients with preeclampsia and intrauterine growth restriction. Moreover, endothelial cell dysfunction and intravascular inflammation are key features of preeclampsia. The objectives of this study were to: 1) determine the origin of the endothelial cells located in the basal plate surface of the placenta (from male fetuses); and 2) analyze the relative proportion of the intervillous surface of the basal plate occupied by trophoblasts and endothelial cells. Immunohistochemistry and morphometry were performed in placentas from women in the following clinical groups: 1) normal term pregnancies (n=15); 2) severe preeclampsia at term (n=15); 3) small for gestational age (SGA) neonates delivered at term (n=15); 4) preterm deliveries (<37 weeks) without inflammation (n=5); and 5) preterm preeclampsia (n=5). Laser capture microdissection and polymerase chain reaction (PCR) were used to determine the allelic pattern of the amelogenin gene of the endothelial cells on the intervillous surface of the basal plate. Our results showed that: 1) the endothelial cells lining the basal plate in placentas of male fetuses were uniformly of maternal origin; and 2) in placentas from uncomplicated pregnancies, the median proportion of trophoblasts and endothelial cells covering the surface of the basal plate were 27.7% and 46.5%, respectively. The remaining area of the intervillous surface of the basal plate was composed of fibrin and anchoring villi. Of interest, placentas from women who delivered an SGA neonate had a higher proportion of trophoblasts and a lower proportion of endothelial cells lining the basal plate than those from normal pregnancies (p < 0.05). The same tendency was observed in placentas from patients with preeclampsia. This study demonstrates that endothelial cells of maternal origin cover the intervillous surface of the basal plate of the placenta, along with trophoblasts of fetal origin. The proportion of this surface lined by trophoblasts is greater in placentas from SGA and preeclampsia than in normal pregnancy. We propose that this change reflects a compensatory mechanism whereby the basal plate surface covered by injured endothelial cells is replaced by trophoblasts or results from a failure of trophoblastic involution in abnormal pregnancies. Our observations also suggest that the lining of the basal plate can provide information about the pathology of endothelial cells in complications of pregnancy.

INTRODUCTION

The basal plate of the placenta, defined as the maternal floor of the intervillous space, is part of the feto-maternal junction that is delivered with the placenta [1]. Major components of the basal plate include decidua, fibrinoid, connective tissue, extravillous trophoblasts, and residues of degenerating villi and maternal vessels [2]. Compared with normal pregnancies, pregnancy complications, such as preeclampsia and intrauterine growth restriction, are associated with a higher proportion of failure of physiologic transformation of the spiral arteries and infarcts in the basal plate [3–8].

The intervillous space of the placenta is a compartment of the maternal circulation. The cells covering the basal plate surface, therefore, are the anatomical counterpart of the maternal endothelium. The surface of the basal plate facing the intervillous space was initially described as a unicellular layer composed of trophoblasts and fibrin [9]. Further ultrastructural and immunohistologic studies have revealed the presence of endothelial cells lining the surface of the basal plate, in addition to trophoblasts and fibrin [2,10–14]. Indeed, the endothelial and trophoblast cells lining the basal plate form a single layer and share junctional organelles [10]. The origin of these endothelial cells remains unclear. Although some have proposed that these cells are of maternal origin [10,11,13,14], it is possible that they represent transformed endovascular trophoblasts of a fetal source.

Recently, Smith et al. [13] estimated the surface area of the basal plate facing the intervillous space lined by trophoblasts and endothelial cells. The authors reported that the proportion of this area covered by endothelial cells was higher than that covered by trophoblasts, and that this difference was exaggerated in preeclampsia [13]. Whether changes in the proportion of cells lining the intervillous surface of the basal plate are the cause or consequence of preeclampsia (a disease characterized by generalized endothelial cell dysfunction [15–19]) is unclear. To the extent that the endothelial cells on the lining of the basal plate are maternally derived, it is possible that changes in maternal endothelium can be reflected on the basal plate lining.

This study was conducted to determine the origin of the endothelial cells lining the surface of the basal plate and to analyze the relative proportion of trophoblasts and endothelial cells lining the basal plate in normal and complicated pregnancies.

MATERIALS AND METHODS

Study design

Placentas were obtained from women in the following clinical groups: 1) normal term pregnancies (n=15); 2) severe preeclampsia at term (n=15); 3) small for gestational age (SGA) neonates delivered at term (n=15); 4) spontaneous preterm deliveries without inflammation (n=5); and 5) severe preterm preeclampsia (<34 weeks; n=5). Placentas were collected following vaginal delivery or cesarean section. Inclusion criteria for normal pregnancy included the absence of medical, obstetrical, or surgical complications and the delivery of a term infant with an adequate for birthweight for gestational age. Severe preeclampsia was defined as severe hypertension (diastolic blood pressure ≥ 110 mm Hg) and proteinuria or mild hypertension plus severe proteinuria (a 24-hour urine sample that contained 3.5 g protein or one urine specimen of ≥ 3+ protein by dipstick measurement). Neonates from preeclamptic patients had a birthweight above the 10^th percentile for gestational age. An SGA neonate was defined as one having a neonatal birthweight below the 5^th percentile for gestational age. All women who delivered SGA neonates were normotensive. Exclusion criteria included multiple gestations, histologic chorioamnionitis, and chromosomal abnormalities. Eligible patients were approached at the Detroit Medical Center/Hutzel Women’s Hospital, Detroit, Michigan, and specimens were collected after written informed consent was obtained. The use of tissue for research purposes was approved by the Institutional Review Boards of Wayne State University and the National Institute of Child Health and Human Development, National Institutes of Health, Department of Health and Human Services (NICHD/NIH/DHHS).

Laser capture microdissection/polymerase chain reaction for sex typing

Five placentas from women who delivered male neonates were used to determine the origin (maternal vs. fetal) of the endothelial cells lining the intervillous space surface of the basal plate. Immunostaining for CD31 (1:50; DakoCytomation, Glostrup, Denmark) was performed to locate endothelial cells among the basal plate lining cells. CD31-positive cells lining the basal plate were microdissected using the AutoPix™ Automated Laser Capture Microdissection System (Arcturus Bioscience, Inc., Mountain View, CA). Isolation and amplification of genomic DNA from microdissected endothelial cells were performed according to the manufacturers’ instructions using the PicoPure™ DNA Extraction Kit (Arcturus Bioscience, Inc.) and the GenomiPhi™ DNA Amplification Kit (Amersham Biosciences UK Limited, Buckinghamshire, UK), respectively. PCR for the amelogenin gene was performed. The sequence of the primers used were 5’-ACCTCATCCTGGGCACCCTGG-3’ (forward) and 5’-AGGCTTGAGGCCAACCATCAG-3’ (reverse), and the size of the amplicons were 212 bp for the X allele and 218 bp for the Y allele [20]. PCR was performed utilizing a PCR Master Mix (Roche Diagnostics GMBH, Penzberg, Germany). The reaction was composed of an initial denaturation for 2 minutes at 94°C, 25 cycles of 30 seconds at 94°C, 45 seconds at 64°C, 45 seconds at 72°C, and a final extension at 72°C for 7 minutes. The amplified products were analyzed by 15% polyacrylamide gel electrophoresis, followed by ethidium bromide staining.

Immunohistochemistry

Immunohistochemistry was performed to examine the composition and proportion of the cells lining the basal plate. Five placental sections were taken from the following areas of the placenta: center (n=1), paracenter (n=2), and periphery (n=2) regions. Tissues were fixed for 24 hours in 10% neutral formalin and 5-μm-thick paraffin sections were obtained. Double immunohistochemical staining was performed using an automatic immunostainer (Dako Corp., Carpinteria, CA). The primary antibodies used were murine monoclonal anti-human cytokeratin-7 (1:1000; DakoCytomation) and rabbit monoclonal anti-human vimentin (1:400; NeoMarkers, Fremont, CA) for identification of trophoblasts and endothelial cells, respectively. Secondary antibodies used were anti-mouse HRP-conjugated and anti-rabbit alkaline phosphatase-conjugated immunoglobulins (Vector Laboratories, Inc., Burlingame, CA). The chromogens used were Vector blue and 3’3’-diaminobenzidine for cytokeratin-7 and vimentin, respectively. Digital images of the entire basal plate in a given section (approximately 20 mm²) were obtained with the Automated Cellular Imaging System^® (ChromaVision System, CLARIENT Inc., Irvine, CA). A 5-μm section of the basal plate surface was randomly selected to measure the length occupied by endothelial and trophoblast cells using Image-Pro Plus 5.0 software (Media Cybernetics, Inc., Silver Spring, MD). The area occupied by each cell type was calculated by squaring the values obtained and the ratio of the area occupied by each cell type was calculated. All study specimens were analyzed by one observer blinded to the clinical information.

Statistical analysis

The Shapiro-Wilk test was used to determine normality. A Chi square test was employed for comparisons of proportions. Kruskal-Wallis and Mann-Whitney U tests were used to determine the differences in the median proportion of basal plate lining cells among and between the study groups. SPSS version 12.0 (SPSS Inc., Chicago, IL) was used for statistical analysis. A p-value of <0.05 was considered statistically significant. Downward adjustment of the alpha value was performed in order to maintain the overall probability of a type I error at 0.05.

RESULTS

Genetic origin of the endothelial cells

CD31-positive endothelial cells were carefully selected and successfully microdissected (Figure 1A). Amelogenin allelotyping of the endothelial cells lining the basal plate of the placentas from male fetuses showed only 212-bp amplicons, indicating a female genotype (maternal origin) in all cases (Figure 1B).

Figure 1. Laser capture microdissected basal plate endothelial cells and their amelogenin gene allelotype profiles.

A, CD31 positive endothelial cells before (left; red line) and after (cap) microdissection. B, Amelogenin allelotype profiles of the cases analyzed. PCR products were electrophoresed in 15% polyacrylamide gel and stained with ethidium bromide. M, F: amplified products from a normal male and female, respectively. Cases 1, 2, 3, 4, and 5 lacked the 218 bp Y allele. Amplified band of Y allele is readily visible in products from genomic DNA of male (M), but not in female (F) control. L: 100 bp ladder marker (200 bp).

Basal plate lining cell composition

The clinical and obstetrical characteristics of the study groups are presented in Table I. According to the study design, the median birthweight was significantly different among normal pregnancy, SGA, and preeclampsia groups (p < 0.05 for each). Cytokeratin-positive trophoblasts and vimentin-positive endothelial cells were clearly distinguished by double immunohistochemistry. Vimentin-positive endothelial cells were usually flatter than cytokeratin-positive trophoblasts and formed a monocellular layer. Trophoblast cells were generally found adjacent to anchoring villi on the basal plate surface, suggesting that the majority of them had extended from the villous surface (Figure 2). Analysis of the images showed that, in normal placentas, the median proportion of trophoblasts and endothelial cells covering the intervillous surface of the basal plate were 27.7% and 46.5%, respectively (Table II). The remaining portion of the basal plate lining was composed of fibrin and anchoring villi (Table II). The median endothelial/trophoblast cell ratio was 1.2 (Table II).

Table I.

Clinical and obstetrical characteristics of the study groups

	Normal pregnancy n = 15	Preeclampsia n = 15	SGA n = 15	PTD no inflammation n = 5	PTD preeclampsia n = 5
Maternal age (y)	24	20	23	24	22
	(13–36)	(16–29)	(16–39)	(18–31)	(18–36)
Nulliparity	4 (26.7)	11 (73.3)α	7 (46.7)	2 (40)	3 (60)
Smoking	2 (13.3)	0	8 (53.3)	1 (20)	0
Gestational age at delivery (wks)	39	38.8	37.8α	31.7	32
	(38.4–41.5)	(37.3–40.5)	(37–39.7)	(24.4–33.7)	(27–33.1)
Male	10 (66.7)	7 (46.7)	4 (26.7)	2 (40)	1 (20)
Female	5 (33.3)	8 (53.3)	11 (73.3)	3 (60)	4 (80)
Birthweight (g)	3160	3200α	2230αβ	1580	1340
	(2948–3760)	(2770–4320)	(2000–2530)	(600–2490)	(820–1720)

Values expressed as median (range) or number (percent)

PTD: preterm delivery; SGA: small for gestational age

^αStatistically significant compared to normal pregnancy

^βStatistically significant compared to preeclampsia

Figure 2. Basal plate from a case of preeclampsia.

Double immunostaining shows (X200) that cytokeratin-7-positive fetal trophoblasts (blue; arrowheads) form a continuous monolayer with vimentin-positive maternal endothelial cells (brown; arrows).

Table II.

Basal plate lining components from placentas of term and preterm deliveries

	Normal pregnancy n= 15	Pα	Preeclampsia n= 15	Pβ	SGA n=15	Pγ	PTD no infection n= 5	PTD Preeclampsia n= 5	Pδ
Trophoblast cells (%)	27.7	0.3	45.8	1.5	53.3	0.03	57.6	51.1	0.8
	(21.4–37.5)		(23.2–60.1)		(40.3–61.9)		(30.9–72.4)	(33.3–62.6)
Endothelial cells (%)	46.5	0.1	20.3	1.8	12.4	0.02	12.8	30.8	0.3
	(18.7–62.1)		(4.8–37)		(6.1–33.3)		(3.5–43.4)	(17.5–49.3)
Fibrin (%)	5.1	2	6.1	0.6	11.9	0.3	8	8.3	0.7
	(3.1–11.7)		(1.9–14.1)		(4.3–20.9)		(5.3–22.8)	(4.9–11.8)
Anchoring villi (%)	21.7	1.8	12.9	3	13.3	1.5	12.1	10.2	0.7
	(9.7–30.3)		(10.1–36)		(9.2–25.1)		(4.7–21.2)	(6.9–13)
Endothelium/Trophoblast	1.2	0.03	0.4	1.5	0.1	0.01	0.2	0.6	0.4
	(0.6–3.3)		(0.07–0.8)		(0.08–0.8)		(0.05–1.6)	(0.2–1.7)

Values are expressed as median and quartiles (25–75)

PTD: Preterm delivery; SGA: small for gestational age

Pα: compared to preeclampsia

Pβ: compared to SGA

Pγ: compared to normal pregnancy

Pδ: compared to PTD no infection

Comparisons between the three term groups showed that placentas from SGA pregnancies had a higher percentage of trophoblast and a lower percentage of endothelial cells than those from normal pregnancies (Kruskal-Wallis, p = .03 for each; post-hoc test, p = 0.03 and 0.02, respectively; Table II). In addition, placentas from SGA pregnancies and those from preeclampsia had a lower endothelial/trophoblast cell ratio than those from normal pregnancies (Kruskal-Wallis, p = 0.01; post hoc test, p = 0.01 and p = 0.03, respectively). Although preeclampsia was associated with a higher median proportion of trophoblasts and a lower median proportion of endothelial cells lining the intervillous surface of the basal plate than normal pregnancy, these differences were not statistically significant (p = 0.1 and p = 0.3, respectively; Table II and Figure 3). Differences in the proportions between the groups were significant, even after adjusting for placental weight. The proportion of trophoblast and endothelial cells lining the basal plate in women with preeclampsia and those with SGA neonates was not different (for each comparison, p > 0.05).

Figure 3. Proportions of trophoblasts and endothelial cells lining the basal plate of placentas at term.

A, Placentas from small for gestational age neonates (SGA) had a higher median proportion of trophoblasts lining the basal plate than those from normal pregnancies (Kruskal-Wallis p=0.03, post-hoc p=0.03). B, Placentas from small for gestational age neonates (SGA) had a lower median proportion of trophoblasts lining the basal plate than those from normal pregnancies (Kruskal-Wallis p=0.03, post-hoc p=0.02). Similarly, although not statistically significant, placentas from women who had preeclampsia had a tendency towards a higher median proportion of trophoblasts and a lower median proportion of endothelial cells lining the basal plate than normal pregnant women.

There were no differences in the percentage of trophoblast and endothelial cells lining the basal plate of placentas from preterm deliveries due to severe preeclampsia and those of spontaneous preterm deliveries (Table II). In addition, placentas from spontaneous preterm deliveries showed a higher median proportion of trophoblasts as well as a trend toward a lower median proportion of endothelial cells lining the basal plate than those from normal term deliveries (p = 0.03 and 0.07, respectively; Table II).

DISCUSSION

This study determined for the first time the maternal origin of the endothelial cells lining the intervillous surface of the basal plate. We also demonstrated that SGA is associated with a high median proportion of trophoblasts and a low median proportion of endothelial cells lining the basal plate. In addition, we found that the endothelium/trophoblast ratio is significantly reduced in preeclamptic and SGA placentas when compared to normal term pregnancies. Persistence of the significance in the comparisons when adjusted for placental weight suggests that the total number/volume percentage of cells of each lineage is altered in cases of SGA and preeclampsia.

The application of double immunohistochemical staining for cytokeratin-7 and vimentin rendered possible the morphometric analysis of the basal plate surface in the same histologic preparation. Vimentin is expressed in mesenchymal cells of different origin and is not endothelial cell-specific. Therefore, we conducted a pilot study using monoclonal antibodies against CD31 and CD34, and the basic staining pattern was identical to that of vimentin. As vimentin immunoreactivity was more distinct in double immunohistochemistry with cytokeratin-7, we selected vimentin as an endothelial marker in this study.

The predominance of endothelium lining the basal plate of normal term placentas is consistent with the observations of Smith et al. [13], who reported that 60.8% of the basal plate lining was composed of endothelium and 18.9% by trophoblasts. However, our observation that preeclampsia is associated with a tendency for a higher proportion of trophoblasts and a reduction in the endothelium/trophoblast ratio covering the basal plate than in normal pregnancies is not consistent with this previous report [13]. This discrepancy can be attributed to differences in patient selection and the image analysis system used.

As the endothelial cells of the basal plate surface are maternally derived, possible sources include endothelial cells from the uteroplacental vasculature [13] or endothelial progenitor cells present in the maternal circulation. Defective physiologic transformation of the spiral arteries and the subsequent decrease in placental perfusion can be responsible for the lower percentage of endothelial cells covering the surface of the placental basal plate in women who delivered SGA neonates and in those with preeclampsia. Indeed, there are quantitative differences in the uteroplacental vasculature between normal and complicated pregnancies [3,4,6]. There is compelling evidence that preeclampsia and pregnancies with SGA neonates share many common pathophysiologic features, including the absence of physiological transformation of the uterine arteries [3,21–23], endothelial cell dysfunction [15,18,24,25], and granulocyte and monocyte activation consistent with maternal systemic intravascular inflammation [26–28]. In addition, an antiangiogenic state is present among patients with preeclampsia and those who deliver SGA neonates, as demonstrated by the high vascular endothelial growth factor receptor-1 and low placental growth factor concentrations in maternal plasma and serum.

Spiral and basal arteries from preeclamptic patients are more tortuous or densely distributed with narrower lumens and thicker walls than those from normal pregnancies [22,29,30]. These vessels may be more vulnerable to endothelial damage induced by hemodynamic stress, as in preeclampsia [31]. The possibility that damaged basal plate surface areas are repaired by endothelial cells recruited from circulating maternal endothelial progenitor cells is supported by the observation of Luppi et al [32]. These authors reported that the number of circulating endothelial progenitor cells (KDR⁺/CD34⁺) is increased in normal pregnancy, whereas preeclampsia is associated with reduced numbers of these progenitor cells [32]. Similarly, Sugawara et al. [33] reported that preeclamptic patients had a lower number of endothelial progenitor cell colony forming units than normotensive pregnant women, as well as a higher rate of cellular senescence. Circulating endothelial progenitor cells appear to play an important role in angiogenesis [34] because they are more sensitive to angiogenic factors [35]. Similar mechanisms could also be responsible for the low proportion of endothelial cells lining the basal plate of placentas from SGA neonates. However, to our knowledge, there are no available data regarding the changes in circulating endothelial progenitor cells in women who delivered SGA neonates. This is an area that should be further investigated.

The cellular turnover on the basal plate lining is currently unknown. Expansion of the basal plate lining would require expansion of its cellular components: trophoblasts, endothelial cells, or both. In preeclampsia, a limited number of endothelial progenitor cells may account for the lower proportion of endothelial cells lining the basal plate [32]. Another possible explanation for the increased percentage of trophoblast in SGA pregnancies and preeclampsia at term would be a failure of the lining trophoblast (remnants of the original cytotrophoblast shell) to normally involute. The proportion of cellular components lining the basal plate also may change with gestational age. Indeed, we observed a higher proportion of trophoblasts and a trend toward a lower proportion of endothelial cells on the basal plate lining from preterm placentas than from normal term placentas.

In conclusion, our observations suggest that the basal plate lining can provide information about the pathology of endothelial cells in complicated pregnancies. We propose that the changes in the basal plate lining cells in abnormal pregnancies reflect either the presence of a compensatory mechanism whereby the surface lined by injured endothelial cells is replaced by trophoblast or a failure of the lining trophoblasts to involute. Whether these maternal endothelial cells lining the basal plate share similarities in biologic responsiveness with those in the systemic vasculature, such as in the heart and kidneys, needs to be explored by functional in vitro studies.