|Home | About | Journals | Submit | Contact Us | Français|
In mammals, the extraembryonic tissues, which include the placenta, are crucial for embryonic development and growth. Because the placenta is no longer needed for postnatal life, however, it has been relatively understudied as a tissue of interest in biomedical research. Recently, increased efforts have been placed on understanding the placenta and how it may play a key role in human health and disease. In this review, we discuss two very different types of environmental exposures--assisted reproductive technologies and in utero exposure to endocrine disrupting chemicals. We summarize the current literature on their effects on placental development in both rodent and human, and comment on the potential use of placental biomarkers as predictors of offspring health outcomes.
The vitality of the placenta is essential to normal development of mammalian offspring. The placenta is the gestational interface between mother and fetus, being required for exchange of gases, nutrients and waste products. The placenta also provides pregnancy-induced hormones, growth factors, and immune protection for the fetus. It is well-recognized that compromised maternal health and genetic lesions of the conceptus can result in placental insufficiency and associated fetal growth defects. Moreover, environmental insults, the topic of this review, can also contribute to placental dysfunction and compromised offspring outcomes.
More recently, it has become appreciated that a healthy placenta is essential for the lifelong health of the mother and offspring. In fact, the Developmental Origins of Health and Disease (DOHaD) hypothesis, which posits that environmental stresses or exposures during development can increase the disease risk later in life, likely has equal contributions from the fetus and placenta (1). With the growing interest in DOHaD and the recognition that the placenta may play an important role in offspring programming, it is not surprising that there are increased efforts focused on placental biology (2). Studies on human placenta, however, lag behind those in other mammals. Given its critical role in mammalian development, the placenta is remarkably variable among mammals (3). Nevertheless, there are notable similarities between the human and rodent placenta, which enables studies of rodent and human placenta to be compared (4). This review will highlight how the environment can modify the placenta in both rodent and human. Specifically, we will focus on the changes iatrogenically induced by Assisted Reproductive Technologies (ART) and changes that occur after developmental exposure to endocrine disrupting chemicals (EDCs). Whether any of the reported changes following either environmental perturbation can serve as a reliable biomarker, and surrogate for fetal health, requires further investigation.
Morphological, physiological, and/or molecular changes in the placenta may serve as biomarkers of offspring health and risks. Currently, morphology and function are assessed throμgh gross morphological assessments, histology, and imaging techniques. Advanced imaging could allow for more accurate, real-time assessments of placental function and fetal health, possibly improving prognoses and informing interventions. Moreover, investigations of molecular biomarkers have relied on -omics-based techniques. In addition to transcriptomic and proteomic characterization, epigenomic studies of normal and abnormal placentas will provide insight into mechanisms underlying altered gene expression in functionally compromised placentas. Stepping beyond the central dogma, metabolomic and secretomic studies will also add to a more comprehensive understanding of placental function.
Because the environment is composed of a broad array of potential insults, identifying a specific biomarker associated with a given insult is daunting. Instead, identifying biomarkers of response that reflect changes in biological function as a consequence of an exposure may ultimately be more informative (5). Althoμgh response biomarkers mask the specific insult that is responsible for the placental change, the tendency for different environmental stimuli to converge and affect a more limited number of molecular pathways sμggests biomarkers of response may ultimately be more relevant and predictive of fetal health.
ART is not traditionally considered an environmental exposure. Nevertheless, ART procedures consist of fertilization outside the body in an artificial culture environment and many other procedures such as hormone stimulation, gamete/embryo freezing, embryo culture, cell biopsy, and embryo transfer procedures that are not inherent to mammalian reproduction. These manipulations have the potential to disrupt the normal biological processes in embryos and cause damage via oxidative, thermal, and mechanical stress (6). One major biological process that coincides with the timing of ART procedures is the epigenetic reprogramming of the genome. It has been proposed that ART procedures disrupt this crucial reprogramming, leading to epigenetic perturbations that can affect the embryo and the extraembryonic tissues.
In the United States, the number of ART cycles has increased by 25% in the last decade, and the number of babies born has risen from 49,458 born in 2004 to 66,706 born in 2013 (7). However, even as reproductive medicine advances, ART is still associated with pregnancy complications, including low birth weight and abnormal placentation (e.g., 2). Althoμgh the patients’ infertility status may contribute to these complications, there is now abundant evidence supporting that ART procedures themselves can cause these effects. Indeed, experimental studies using the gametes and uterine environment of fertile animals provide evidence that ART procedures can induce adverse effects.
While the vast majority of ART babies are born healthy, there is concern that ART children display symptoms indicative of increased risk for diseases later in life (9). As ART continues to advance and become more available, iatrogenic effects should be minimized to ensure healthy pregnancies and limit adverse health outcomes for ART-conceived children. In the following section, we summarize the current knowledge of identifiable changes in the placenta with ART in both experimental and clinical studies.
Several studies have shown that ART increases term placental weight in mice (10–15). Intriguingly, the procedure of transferring blastocyst stage embryos alone can increase overall placental weight, strongly sμggesting that relatively brief procedures can contribute to placental phenotypes (14). In addition to overall increased placental weight, mouse IVF placentas also exhibit disproportionate overgrowth of the junctional zone, a tissue that secretes factors that may influence the growth of the placental vasculature (12–16). This disproportion is observed in E12.5 IVF concepti, meaning this morphological change is already detectable soon after placental formation (16). It is unclear if increases in placental weight and changes in placental cell composition are evidence of a direct adverse effect of ART procedures impairing placental development or are evidence of compensatory mechanisms. If changes in cell composition are the result of compensatory mechanisms, the fact that they occur so early sμggests that stress signals are already affecting the early embryo prior to placental formation. Importantly, increased placental weight with ART was almost always associated with a significant reduction in fetal weight (11–15). In some instances, fetal weight was unchanged, which may be influenced by the genetic background of the embryo (14,15). Althoμgh reduced fetal weight may be indicative of reduced placental efficiency, it is not clear if reduced fetal weight is due to impaired placental function or if ART procedures also directly impact fetal growth processes.
Most studies in humans report at least one type of morphological difference in ART placentas, but the exact phenotypes are not consistent among studies. Increased placental weight, placental:fetal ratio, placental thickness, abnormal umbilical cord insertion and abnormal placental shape have been observed with ART in term singletons (17–22). In a study comparing placentas from ART or spontaneously conceived pregnancies from either fertile or subfertile patients, only ART placentas exhibited increased placental thickness and increased incidence of hematomas (22), sμggesting that the placental pathologies are the result of ART procedures rather than patients’ infertility status. One study also noted edema and microcalcifications in ART placenta (23). The incidence of other placental pathologies, including gross malformations, infarcts, fibrin deposition, chorioamnionitis, fibrinoid necrosis, lesions, or syncytial knots have not been observed (18,20). Importantly, in addition to differences in gross morphology and noticeable placental pathologies, ART placentas may display more subtle degenerative alterations. For example, in one study, ultrastructural changes in syncytiotrophoblasts were only detectable by transmission-electron microscopy (24).
There is compelling work that shows ART can affect genomic imprinting in both mice and humans. Imprinted genes, which are genes that display monoallelic expression dependent on the parent-of-origin, are regulated by DNA methylation at discrete cis-acting elements known as imprinting control regions (ICRs). The placenta, relative to the embryo, is particularly sensitive to these epigenetic changes (14,25–28). In mid-gestation and term mouse placenta, ART procedures result in ICR hypomethylation and biallelic expression of many disease-relevant imprinted genes (14,25–28). In a different study, the relative expression of imprinted genes was significantly increased in mid-gestation placenta from ART and embryo culture groups by comparison with controls althoμgh no differences in ICR methylation were observed (29). Yet in a different study of ART placentas, methylation status at select ICRs was dependent on gestational age and was associated with altered total expression (13,30).
Changes in imprinted gene methylation have also been reported in human ART placentas (31–37). Similar to mice, human ART placenta exhibit hypomethylation of the H19/IGF2 ICR and a corresponding increase in H19 expression (36). In general, epigenetic changes in human placenta include both hypomethylation and hypermethylation of differentially methylated regions. In one study, a proportion of detectable changes at imprinted genes were also detected in sperm from the father, sμggesting that some of the defects in imprinted gene methylation are already present in the sperm and inherited rather than induced by ART procedures (32). Katagiri and colleagues specifically compared the expression of select imprinted genes in placenta from concepti exhibiting intrauterine growth restriction conceived spontaneously or via ART. They observed that while growth restriction affects imprinted gene expression regardless of mode of conception, ART growth-restricted placentas were still significantly different from growth-restricted placentas from spontaneous conceptions (34). Because imprinted genes are epigenetically regulated and have known functions in placental development, they may serve as ideal biomarkers—they can be informative of both epigenetic changes and potential functional changes in the placenta.
Several studies have also demonstrated more widespread changes in DNA methylation in ART placentas. A global reduction in DNA methylation was detected in ART term mouse placentas using luminometric methylation assay (LUMA), an assay that measures the relative amount of global CpG methylation using methylation sensitive and methylation insensitive restriction enzymes (14). A global reduction in DNA methylation was also detected in ART extraembryonic tissues at E7.5 using methylated DNA immunoprecipitation (MeDIP) (12). Similar to results of imprinted genes, human ART placentas also demonstrate lower CpG methylation at select non-imprinted genes (31,38).
In addition to changes in imprinted genes, changes in the expression of non-imprinted genes important to placental development have been detected. One important group of genes, nutrient transporters, has been shown by several groups to have reduced expression in ART placenta (11,14,15,39). To determine how ART procedures affect genome-wide changes during placental development, two groups have performed microarray and RNA-seq analyses. Examining gene changes at E10.5 by microarray, Fauque et al. observed dysregulation of genes involved in cell growth, protein modifications, immunity, and angiogenesis (40). Using RNA-seq analyses, Tan and colleagues showed that ART significantly altered the expression of many genes in extraembryonic tissue and placenta at E7.5 and E10.5 (16). Genes that were affected at both time points were associated with genetic information processing, cytoskeleton, energy and amino acid metabolism, and vasculogenesis/angiogenesis. A proportion of the genes identified with altered expression exhibited hypomethylation or hypermethylation at promoters. In both studies, most genes were downregulated in ART tissues, but some were upregulated. In humans, genome-wide changes in gene expression in term ART placentas were investigated by microarray (41). Overrepresented biological pathways included cell cycle, metabolism, immunity, among others.
Sui and colleagues have conducted proteomic analyses of E7.5 and E10.5 extraembryonic tissues of ART mouse concepti (12). The majority of genes that were affected were downregulated by comparison with control placenta. Approximately one quarter of the genes identified were dysregulated at both times. Similar to what was observed in gene expression changes (16), the altered proteins were associated with regulation of gene and protein expression, transport, hematopoiesis/angiogenesis, cytoskeleton, energy metabolism, and immunity (12). In human ART placenta, proteome analyses identified genes associated with membrane traffic, metabolism, nucleic acid processing, stress response, and cytoskeletal functions (12). Also, ART placentas have been shown to have significantly higher steroid metabolizing enzyme activity, resulting in a higher clearance capability (10,42). Althoμgh not entirely clear, this may impact hormone signaling and fetal growth.
Despite the extensive characterization of epigenetic gene regulation, to our knowledge, only one study has directly tested the impact of ART procedures on placental function. Amino acid and glucose transport was evaluated in mouse fetuses that were spontaneously conceived and transferred to pseudopregnant recipients or that were conceived via IVF. Amino acid transport was reduced by 58% and associated with reduced levels in the fetus, while glucose transport and overall glucose levels in the fetus were unchanged (11). This demonstrates that the larger placental phenotype observed with ART is associated with reduced efficiency. Amino acid deficiency during prenatal development did not impact term fetal weight, but it remains to be determined if it may contribute to adverse health effects later in life.
EDCs are both natural and synthetic chemicals capable of perturbing the endocrine system (43). EDCs accomplish this by mimicking or blocking the action of endogenous hormones, resulting in the inappropriate activation or inactivation of downstream pathways critical for maintaining cellular homeostasis (44). Given the high production and use of EDCs in many every day products, chronic exposure to these ubiquitous compounds occurs throμgh multiple routes (43). Because endogenous hormones have various target tissues, it is not surprising that reports in the literature have demonstrated the effects of EDCs on numerous tissues and physiological endpoints as well (43). Given that a hallmark of these compounds is their ability to elicit non-monotonic dose responses, identifying the mechanism(s) by which these compounds act will not necessarily be straight-forward (45). The mechanisms are likely to be complex, possibly with multiple targets and pathways affected in the same tissue or a given EDC exhibiting tissue-specific responses. Prenatal development represents a particularly vulnerable window to EDC exposure – aside from the aforementioned epigenetic reprogramming events that occur at this time, the reduced activity of enzymes required for xenobiotic metabolism and clearance can result in the accumulation of EDCs to levels sufficient to cause detrimental effects in target organs of the developing fetus (46). Select studies reporting morphological and molecular changes in EDC-exposed rodent placentas are summarized in Table 1.
In contrast to ART, studies are more limited regarding the consequences of EDC exposure on placental development and subsequent fetal health. In rodent models, the description of placental morphology after early life exposure to EDCs has varied according to laboratory. Prenatal exposure to the estrogenic EDC, bisphenol A (BPA), has been associated with increased as well as reduced placental size at mid-gestation (47–49). Growth of both the labyrinth, which is the site for maternal-fetal nutrient exchange, and the junctional zone appear to be susceptible to BPA exposure, changing in size relative to the whole placenta. The size changes, however, varied by study (47–49). In terms of fetal health, Susiarjo et al. and Tait et al. did not observe significant weight differences in mid-gestation embryos, while Tachibana et al. reported significant reductions in the number of live embryos at mid-gestation (47–49). Differences in these studies could be attributed to dissimilar exposure paradigms (i.e. rodent strain, doses, timing of exposure, route of exposure). Differences aside, all three studies did speculate on BPA exposure jeopardizing placental nutrient transport. Susiarjo et al. observed increased red blood cell accumulation in the labyrinth of BPA-exposed placentas, sμggestive of improper oxygenation (47). Tait et al. and Tachibana et al. observed altered vasculature in the form of compromised maternal blood spaces in the labyrinth (48,49). Whether these morphological changes actually impact placental function, however, remains to be tested.
In support of the estrogenic role of BPA, exogenous administration of estradiol benzoate resulted in diminished development of the labyrinth and its associated fetal vessels in term placentas (50). Consequently, fetal weight and survival were also reduced at term (50). When mice were exposed to diethylstilbestrol (DES), a potent synthetic estrogen, the labyrinth layer and vascular network were underdeveloped (51,52). Similar to the observations following exposure to estradiol benzoate, DES-exposed fetuses exhibited an increased mortality rate (50–52). While Kagawa et al. observed a significant reduction in male and female embryonic weight at mid-gestation, Nagao et al. did not observe significant changes in fetal or placental weight following DES administration (51,52).
Other non-estrogenic EDCs have been shown to disrupt placental morphology. Exposure to supraphysiological doses of di(2-ethylhexyl)phthalate (DEHP), best known for its anti-androgenic activity, reduced the total area as well as relative areas of the junctional zone and labyrinth in a dose-dependent manner in mid-gestation mouse placentas (53). In comparison, exposure to excess testosterone, an endogenous androgen, resulted in reduced placental weight, total placental area, and red blood cells in the labyrinth, while the relative area of the labyrinth was not significantly affected in term placentas (54). From these studies, it is not clear whether DEHP is unexpectedly behaving like an androgen or if the observed changes are due to a general toxicity response given the high doses of DEHP used in the study. The reduction in red blood cell mass upon testosterone administration sμggests that excess androgen may be impacting placental blood flow and subsequent nutrient exchange between mother and fetus, consistent with the observation that fetal size was reduced in pregnant dams exposed to testosterone (54). Finally, exposure to the dioxin 2,3,7,8-tetrachlorodibenzodioxin (TCDD), a known ligand of the aryl hydrocarbon receptor (AhR), constricted maternal vessels in term placentas while the density of fetal capillaries and the thickness of the labyrinth layer were unchanged (55). An increased rate of fetal death was reported following exposure to the higher dose of TCDD, while the surviving concepti did not demonstrate significantly altered placental or fetal weights, sμggestive of a bimodal response to TCDD in early development (55). Investigations identifying the contributing factors to this seemingly stochastic response to dioxins and other EDCs remains open (47,55).
From a molecular perspective, changes in placental gene expression have been primarily assessed via candidate approaches, althoμgh a limited number of studies have assessed global expression following EDC exposures. While changes in epigenetic regulation have been sμggested to play a role in the mode of action of EDCs, few studies have directly assessed changes in epigenetic marks associated with changes in gene expression. Susiarjo et al. examined the status of imprinted gene regulation in BPA-exposed placentas and observed loss of imprinting that correlated with altered total expression and DNA methylation at ICRs (47). Kang et al. also examined imprinted gene expression in mid-gestation placentas exposed to BPA and reported similar effects at some imprinted genes (56). While there were some inconsistent findings between these two studies, difference could be attributed to experimental design. Kang et al. used a shorter exposure window and different dose of BPA compared to Susiarjo et al. (47,56). Given the non-monotonic nature of BPA and other EDCs, the modes of action and subsequent effects of BPA at each dose could yield different molecular and physiological outcomes. On a global level, placental DNA methylation has also been reported to be reduced by early life exposure to BPA in mice (47).
Although there has been significant attention paid to the role of imprinted genes in placental development, inappropriate expression of non-imprinted genes could also greatly impact placental physiology and subsequent fetal health (57,58). Althoμgh DNA methylation is a well-established mechanism of imprinted gene regulation, the role of epigenetics in proper gene regulation at non-imprinted loci is not as clearly defined (59). Microarray and gene ontology (GO) analyses of mid-gestation placentas exposed to BPA demonstrated an enrichment of differentially expressed genes associated with placental vasculature, including Vegfa, a major regulator of placental vasculogenesis and angiogenesis (48,60). Althoμgh Tait et al. did not assess epigenetic changes associated with the observed expression changes, in vitro and in vivo studies sμggest that DNA methylation in the promoter region of Vegfa and Vegfa-associated factors can affect gene expression, and aberrant methylation at these vasculogenesis-associated genes has been linked to pregnancy complications in humans (48,61–63). Many studies assessing the consequences of EDC exposure also focus on examining expression changes in genes encoding nutrient transporters such as the glucose transporters Glut1 and Glut3, the predominant glucose transporter isoforms expressed in the placenta (64). Because the regulation of nutrient transporters occurs at the level of translation, it is critical to confirm mRNA expression changes at the protein level. Ganguly et al. demonstrated that Glut3 expression and DNA methylation at a CpG island upstream of the Glut3 gene can be affected by environmental cues, while Glut1 expression and DNA methylation were not significantly affected (64). Althoμgh changes in morphology and gene expression point towards compromised nutrient transport associated with EDC exposure, most EDC exposure-based studies have not performed any functional validation. Finally, as EDCs are thoμght to disrupt nuclear hormone signaling pathways, expression levels of these receptors are often examined, with reports of varying levels of expression that are EDC-, tissue-, and sex-dependent (48,52,54,65). Interestingly, studies sμggest that nuclear hormone receptors and epigenetic regulatory enzymes can reciprocally regulate one another as well as cooperatively regulate downstream gene expression (66–68). This further supports the idea of an epigenetic basis mediating the effects of early life exposures to EDCs and later life health outcomes.
To our knowledge, no human studies have reported on associations between EDC exposure and gross placental morphology. Reported levels of EDCs in human maternal, fetal, and/or placental compartments and the associated effects can be found in Table 2. A minimal number of studies have examined changes in placental DNA methylation in human tissue samples. LaRocca et al. identified an inverse relationship between phthalate metabolite levels in maternal urine and DNA methylation at the imprinted H19/IGF2 locus in the placenta (69). Kappil et al. demonstrated a positive association between levels of some organic pollutants measured in the placenta and global DNA methylation, and this result trended towards a male-specific effect (70). At a locus-specific level, elevated H19 expression correlated with increased levels of another organic pollutant, polybrominated diphenyl ether (PBDE), while IGF2 expression was unaffected (70). Althoμgh none of these changes in DNA methylation were correlated with birth weight (69,70), it is unknown whether later life health is affected. Longitudinal studies assessing the growth trajectories of these individuals may provide more information on the role of altered placental epigenetic regulation in health and disease.
In humans, methods to assess total xenobiotic burden of (anti)estrogenic (TEXB) and (anti)androgenic (TEXB-AA) compounds have been developed and utilized as a biomarker of endocrine disruption (71,72). TEXB uses high performance liquid chromatography (HPLC) to separate xenobiotic compounds from endogenous hormones in tissue homogenates for subsequent reporter gene or cell proliferation-based assays as readouts of (anti)estrogenicity and/or (anti)androgenicity (73,74). TEXB and DNA methylation at repetitive elements in human placental samples demonstrated an inverse relationship in male placentas, while no association was observed in female placentas (75). The effect of altered DNA methylation at repetitive elements on health and disease is not well understood. Increased TEXB in placentas has also been associated with increased birth weight in males but not females as well as an increased risk of urogenital malformations in males (72,76). What is not known with TEXB-based studies is whether a single EDC or small number of EDC(s) is sufficient to drive a particular health outcome, or if the combination of various EDCs and greater xenobiotic burden, independent of the EDCs themselves, is responsible for the observed phenotypes. Studies using individual EDC exposures in rodents will help to elucidate the effects of each particular EDC, but future studies identifying the consequences of combination exposures are also necessary for a more accurate representation of human exposure to EDCs.
Studies from rodents and human, as well as large animals (not described here), clearly show an effect of the environment, whether throμgh EDC exposure or techniques used in ART, on the morphology, molecular signature and epigenetics of the placenta. More limited information exists regarding real-time assessments of placental function in rodents and humans. Such measurements would provide more opportunities for immediate therapeutic intervention to ameliorate negative health effects imposed on the fetus by a functionally compromised placenta. For example, high-frequency ultrasound has been utilized to reliably measure blood velocity in pregnant mice and may be a useful tool in future studies (77).
It is also important to note that as transcriptome profiling becomes increasingly employed in exposure-based studies, the importance of assessing morphological state prior to investigating global expression changes cannot be overstated. Given environmental cues can affect placental layer proportions, transcriptome changes may be a consequence of altered cell composition as opposed to a true biological change. Reports on global expression will be more informative if conducted on an enriched cell population.
An additional challenge in these fields is the sex-specificity of health outcomes associated with perinatal exposures. Given this, it will be critical to stratify and report both molecular and physiological endpoints by sex. The placenta is known to differentially adapt to environmental cues given the sex of the associated fetus (78). Therefore, the placenta may be a major contributor to the sexual dimorphism of disease states (79).
Taken together, the available literature sμggests that environmental exposures can disrupt placental morphology, epigenetic regulation, and gene expression. It remains to be definitively shown how and if morphological and molecular changes in the placenta contribute to adverse health outcomes in offspring. Moreover, if so, could the placenta be used as a predictor or ‘biomarker’ of offspring health? As our knowledge of the mechanisms underlying normal placental development and the processes disrupted by the environment improves, prevention and intervention strategies should be investigated and implemented to help facilitate an environment for optimal growth in utero.
The authors wish it to be known that, in their opinion, authors Vrooman and Xin should be considered similar in author order.
This review summarizes what is known from human and mouse studies about the effects of two environmental exposures, assisted reproductive technologies and endocrine disrupting chemicals, on the placenta.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.