Thinning of the placental villous membrane during maintenance in hypoxic organ culture: structural adaptation or syncytial degeneration?

Hypoxia enhances the expression of follistatin-like 3 in term human trophoblasts

Hypoxic injury hinders placental differentiation and alters trophoblast gene expression. We tested the hypothesis that the expression of follistatin-like 3 (FSTL3), a member of the follistatin family of proteins, is modulated by hypoxia in primary human trophoblast (PHT). Using immunofluorescence of human term placental villi we detected the expression of FSTL3 protein in placental villi, primarily in trophoblasts. We verified this finding in cultured term PHT cells. Basal expression of FSTL3 transcript in cultured PHT cells, determined using quantitative PCR, was stable over the culture period. Importantly, when compared to culture in FiO(2)=20% or FiO(2)=8%, PHT cells cultured in FiO(2) <1% exhibited a 4-6 fold increase in FSTL3 mRNA expression as early as 4h in hypoxia. Whereas cellular FSTL3 protein was unchanged in hypoxia, we found that hypoxia increased the level of FSTL3 in the medium. Lastly, the exposure of PHT cells to either the hypoxia-mimetic cobalt chloride or the proline hydroxylase inhibitor dimethyloxaloylglycine upregulated the expression of FSTL3 transcript. Our data indicate that hypoxia enhances the expression of FSTL3 and its release from PHT cells. Our finding that hypoxia-mimetic agents enhance FSTL3 expression implicates HIF1alpha in this process.

Adriana and Luisa Castellucci Award lecture 2001. Hypoxia inducible factor-1: oxygen regulation of trophoblast differentiation in normal and pre-eclamptic pregnancies--a review

During early pregnancy, trophoblast differentiation occurs in an environment of relative low oxygen tension which is essential for normal embryonic and placental development. At around 10-12 weeks' gestation, when the intervillous space opens to maternal blood, there is an increase in Po(2). This increase correlates with the time of maximal trophoblast invasion into the maternal decidua, which allows extravillous trophoblast cells to access and remodel the maternal spiral arteries. Hypoxia Inducible Factor 1(HIF-1) is a transcription factor which activates gene transcription in response to varying oxygen concentration of cells. HIF-1 is a heterodimer composed of the inducible HIF-1alpha and the constitutively expressed HIF-1beta/ARNT subunits. Using villous explants, we have demonstrated that the oxygen-regulated events of early trophoblast differentiation are in part mediated by TGFbeta(3), an inhibitor of trophoblast differentiation, via HIF-1alpha. Pre-eclampsia is a disease of pregnancy that is characterized by shallow trophoblast invasion. Recently, we have reported that TGFbeta(3) is over-expressed in pre-eclamptic pregnancy and that its down-regulation restores invasive capability to trophoblast cells. Because TGFbeta(3) is downstream of HIF-1alpha, in the present study we investigated the expression of HIF-1alpha in pre-eclamptic placentae and age-matched controls using in situ hybridization and histochemical analyses. We found that HIF-1alpha mRNA and protein expression are abnormally elevated in pre-eclamptic placental tissue when compared to normal placental tissue. We conclude that pre-eclampsia may result from a developmental failure of oxygen to increase or of trophoblast cells to respond and/or sense an increase in oxygen. This will prevent the normal TGFbeta3 down-regulation and will lead to poor trophoblast invasion predisposing the pregnancy to pre-eclampsia.

Hypobaric hypoxia and villous trophoblast: evidence that human pregnancy at high altitude (3600 m) perturbs epithelial turnover and coagulation-fibrinolysis in the intervillous space

Spatial relationships between fibrin-type fibrinoid and regions of villous trophoblast were examined in order to address two main questions: [1] is high-altitude pregnancy accompanied by changes in the sizes of trophoblast compartments (cytotrophoblast, syncytiotrophoblast, denudation sites)?, and [2] do highland placentae differ in the amounts and distribution patterns of perivillous fibrin-type fibrinoid? Placentae were collected from two ethnic groups completing term pregnancies at low (400 m above sea level; n=25) and high (3600 m; n=45) altitude in Bolivia. Masson trichrome-stained sections were sampled randomly and analysed stereologically to estimate compartment volumes and surfaces. Comparisons were drawn using variance, Chi-squared and contingency table analyses. At high altitude, birthweights were 265 g lower and placentas had a larger intervillous space (270 cf 181 cm(3)), less fibrin-type fibrinoid (4.1 cf 8.4 cm(3) by volume; 2570 cf 4430 cm(2) by surface area), less villous trophoblast (50 cf 73 cm(3)) and a smaller villous surface (5.6 cf 7.0 m(2)). Volumes were reduced in all syncytiotrophoblast compartments (with and without nuclear aggregations). Cytotrophoblast was maintained and its relative volume increased significantly (from 2.7 to 3.6 per cent of trophoblast volume). Decreases in villous surface area affected primarily thinner (nuclear aggregate-free) regions of syncytium. Regardless of altitude, fibrin-type fibrinoid was deposited non-randomly: it was preferentially located at sites of trophoblast denudation. Although no altitudinal differences in fibrin-type fibrinoid patterns were detected, absolute surfaces were diminished on denuded and thinner regions of trophoblast but not on syncytial knots or bridges. Ethnic differences at low altitude (relatively greater deposits on denudations in Amerindians) were minimized at high altitude. We conclude that pregnancy at high altitude alters the epithelial steady state (towards cytotrophoblast and away from syncytiotrophoblast) and the coagulation-fibrinolysis steady state in the intervillous space (to favour fibrinolysis over coagulation). Thinner regions of syncytiotrophoblast may be the main sites of greater fibrinolytic or anticoagulatory activity. The findings are partly consistent with results from in vitro studies which indicate that hypoxia stimulates proliferation of cytotrophoblast but impairs fusion into syncytium.

A stereological method for testing whether or not there is random deposition of perivillous fibrin-type fibrinoid at the villous surface: description and pilot applications to term placentae

We present a stereological method for testing whether or not there is random deposition of fibrin-type fibrinoid (FTF) at the villous surface of human placenta. The method requires random sampling of tissue with test lattice lines superimposed on microscopic fields at random positions and orientations. Test lines are used to generate chance intersections with specified sub-domains of the villous surface. At least three sub-domains are distinguishable: non-syncytial knots (nonSK), syncytial knots (SK) and areas of trophoblast de-epithelialization (DEP). Other sub-domains may be included to suit individual circumstances and project aims. The relative numbers of intersections with sub-domains provide the basis for an 'expected' distribution. Subsequently, this is compared with an 'observed' distribution which can be calculated from empirical estimates of the numbers of intersections with sub-domains associated with perivillous FTF (e.g. nonSK+FTF, SK+FTF and DEP+FTF). Expected and observed distributions can be compared by a chi-squared analysis. If the null hypothesis (no difference) is rejected, chi-squared values for individual sub-domains can be analysed in order to localize and interpret sites of preferential deposition. Comparisons may be drawn for individual placentae as well as a group of placentae, thereby permitting assessment of inter-placental variability. Finally, between-group comparisons may be drawn in order to test whether or not FTF deposition patterns differ in control and other pregnancies. Worked examples of the statistical procedures are provided. Preliminary results of applications to placentae from normal and complicated (hypobaric hypoxia) pregnancies are presented. They show that FTF deposition is non-random and preferentially located at sites of de-epithelialization. De-epithelialization may be a consequence of syncytial degeneration but also, at least in part, of continuous trophoblast turnover in which syncytial fragments rich in (pre-) apoptotic nuclei detach from the epithelium and are deported from the maternal intervillous space. The nascent detachment site is immediately covered by FTF prior to repair by re-epithelialization.

A quantitative study on the effects of maternal smoking on placental morphology and cadmium concentration

The aim of this study was to quantify the effects of maternal cigarette smoking on placental morphology, paying particular attention to variables known to be influential in facilitating oxygen diffusion. Structural quantities were estimated by stereological analyses of placental samples drawn from non-smoking and smoking women whose smoking habits were assessed both subjectively (from volunteered cigarette consumption) and objectively (by determining levels of plasma cotinine, a major metabolite of nicotine). Concentrations of placental cadmium were also measured. In the smoking group, maternal and fetal haematocrits were elevated and mean birthweight was reduced. Within placentae, the most significant alterations were increases in cadmium levels, the relative volumes of maternal intervillous space, the relative surface areas of fetal capillaries and decreases in the relative and absolute volumes of fetal capillaries. Findings indicate that changes in capillary volume are the result of a decrease in mean capillary diameter rather than total length. The mean thickness of the trophoblast component of the villous membrane was also increased in the smoking group. Although increased haematocrits suggest that fetuses of smoking mothers suffer hypoxic stress, these morphological changes are likely to compromise, rather than assist, transplacental oxygen transfer. This is in marked contrast to the adaptive changes seen in pregnancies associated with preplacental hypoxia and suggests that other factors might be compromising the fetoplacental unit. Finally, although the morphological changes associated with maternal smoking seem to be the result of an all-or-none, rather than dose-dependent, effect, the available evidence is not conclusive.

Thinning of the intervascular tissue layers of the human placenta is an adaptive response to passive diffusion in vivo and may help to predict the origins of fetal hypoxia

OBJECTIVES

To test the hypothesis that thinning of the placental intervascular layers, and greater variability in thickness, are positive adaptations to facilitate passive diffusion and may help to resolve different categories of fetal hypoxia.

STUDY DESIGN

Placentas from 12-41 weeks of normal gestation and from pregnancies associated with fetal hypoxic stress (high altitude, diabetes mellitus) were sampled systematically, fixed in formalin, wax-embedded and quantified using stereological methods. Arithmetic and harmonic mean distances across villous trophoblast, stroma and fetal plasma were estimated by measuring randomly sampled intercept lengths. In each case, an index of variability of layer thickness was calculated by dividing the arithmetic by harmonic mean distance. This index has the value I when a tissue layer is uniformly thick but increases in value as local layer thickness becomes more variable. Comparisons between groups were drawn using variance and regression analysis.

RESULTS

During pregnancy, there were significant negative correlations between layer thickness (trophoblast, stroma) and gestational age and fetal weight, and significant positive correlations between thickness irregularity and age and weight. Compared with lowland controls, high-altitude placentas possessed thinner layers but only the trophoblast and stroma (and not fetal plasma) were more variably thick. In maternal diabetes, only fetal plasma distance was reduced but fetal and stromal layers appeared to be more irregular in thickness.

CONCLUSIONS

Alterations in placental intervascular layer thicknesses occur in normal and abnormal pregnancies and represent real adaptations leading to improved diffusive conductances. Differences in the tissue location of the adaptive response may depend on the nature and origins of the fetal hypoxia.

Oxygen and placental villous development: origins of fetal hypoxia

The increasing practice of preterm delivery in the fetal interest for conditions such as pre-eclampsia or intrauterine growth restriction (IUGR) has provided an opportunity to study placental structure in pregnancies with prenatal evidence of fetal compromise. These data suggest that the origin of fetal hypoxia in IUGR with absent end-diastolic flow in the umbilical arteries is due to a failure of oxygen transport from intervillous space to umbilical vein. Failure of the fetoplacental circulation to extract oxygen from the intervillous space under such circumstances means intervillous PO2 is closer to maternal arterial values than under physiological conditions. Correspondingly the placental villi are chronically exposed to a higher oxygen tension than under normal circumstances--the term ¿hyperoxia', relative to normal intraplacental oxygenation, is proposed to describe this situation. Both the trophoblast and villous core react to increased oxygen despite fetal hypoxia. These results challenge the generally accepted concept of ¿placental hypoxia' in all circumstances where fetal hypoxia might arise. Therefore three categories are proposed for the origins of fetal hypoxia: (1) preplacental hypoxia; (2) uteroplacental hypoxia; and (3) postplacental hypoxia. Examples for these three disease states are listed in this review and the structural reaction patterns of placental villi to these differences in oxygenation are discussed.

Placental cytokines and the pathogenesis of preeclampsia

The authors explore the hypothesis that tumor necrosis factor-alpha (TNF-alpha) and possibly other inflammatory cytokines are overproduced by the placenta in response to local ischemia/hypoxia contributing to increased plasma levels, and subsequent endothelial activation and dysfunction in the pregnancy disorder, preeclampsia. It is widely held that inadequate trophoblast invasion and physiologic remodeling of spiral arteries initiate placental ischemia/hypoxia in preeclampsia. Furthermore, focal areas of placental hypoxia have been implicated in the production of "toxic" factor(s) by the placenta, which circulate and cause maternal disease. Placental trophoblast cells and fetoplacental macrophages normally produce TNF-alpha and interleukin-1 (IL-1), which are capable of producing endothelial cell activation and dysfunction. Hypoxia has recently been reported to increase TNF-alpha and IL-1 production by term villous explants from the human placenta. Placental cells also express erythropoietin (EPO), which is the prototype molecule for transcriptional regulation by hypoxia in mammals. Interestingly, TNF-alpha and IL-1 have DNA sequence homologous or nearly homologous to the hypoxia-responsive enhancer element of the EPO gene, thus providing a potential, but as of yet, untested molecular link between placental hypoxia and stimulation of cytokine production. Inflammatory cytokines overproduced by the placenta in response to hypoxia may then lead to increased plasma levels and endothelial activation and dysfunction in preeclampsia. The purpose of this short review is to critically evaluate the hypothesis that placental cytokines contribute to the pathogenesis of preeclampsia. Of note, the etiology of the disease presumably related to deficient trophoblast invasion is beyond the scope of this work.

Quantitative description of the elaboration and maturation of villi from 10 weeks of gestation to term

Stereological methods have been applied to a cross-sectional sample of human placentae collected at 10-41 weeks of gestation in order to provide a quantitative description of the growth and maturation of villi. Random tissue sections were analysed to derive volumes, surface areas, lengths, diameters and membrane thicknesses for villi and their fetal capillaries. Expansion of the total volume and surface area of villi can be explained by a dramatic linear growth of terminal villi which begins at about the middle of the second trimester. Growth of intermediate villi also occurs but to a more limited extent. Linear growth is accompanied by villous maturation which involves increases in the relative volume of capillaries and in villous capillarization coupled with decreases in villous diameter, capillary diameter and harmonic thickness of the villous membrane. These findings confirm that placental growth and development depend greatly on growth and maturation of terminal villi. They do not confirm sinusoidal dilation of fetal vessels as a generalized phenomenon. They also support the contention that changes in effective diffusion distances across the villous membrane have real adaptive significance.