Morphological and Phenotypic Analyses of the Human Placenta Using Whole Mount Immunofluorescence

ABSTRACT

The human placenta performs multiple essential functions required for successful pregnancy. Alterations in the placental vasculature have been implicated in severe complications of pregnancy. Despite the importance of placental vascular function during pregnancy, there are gaps in our knowledge regarding the molecular pathways that control vessel development. Furthermore, there are limited tools available to simultaneously examine the morphology, phenotype, and spatial arrangement of cells within intact placental structures. To overcome these limitations, we developed whole mount immunofluorescence (WMIF) of the human placenta. Morphological analyses using WMIF revealed that blood vessel structures were consistent with an immature, angiogenic morphology in first-trimester placentas and mature, remodeled endothelium at term. To investigate placental expression of factors that control blood vessel development, we utilized WMIF to examine gestation age-specific expression of 1) the receptors for vascular endothelial growth factor (VEGFR-1, VEGFR-2, and VEGFR-3), which are required for placental vascular development in mice, and 2) activated, tyrosine phosphorylated STAT3 (pSTAT3), a transcription factor that mediates VEGFR2 signaling. We detected high levels of VEGFR2, VEGFR3, and pSTAT3 expression in early placental blood vessels that were significantly diminished by term. VEGFR1 was expressed primarily in trophoblast and Hofbauer cells throughout gestation. Based on our collective results, we propose that VEGFR2, VEGFR3, and STAT3 play essential roles in the development of the human placental vasculature. In addition, we anticipate that WMIF will provide a powerful approach for comparing placental morphology and protein expression in normal versus pathological pregnancies and for investigating the effects of environmental factors on placental function.

Early human placental blood vessels express high levels of the pro-angiogenic receptors VEGFR2 and VEGFR3, and the activated transcription factor pSTAT3, which suggests that these molecules play a role in regulation of placental vascular development.

INTRODUCTION

The human placenta performs multiple functions that are critical for successful pregnancy, including gas and nutrient exchange, hormone production, and protection of the genetically distinct fetus from both infection and maternal immune system-mediated destruction. The placenta is composed of treelike structures termed villi. Floating villi are in direct contact with circulating maternal blood, and they mediate gas and nutrient exchange between the mother and fetus. Anchoring villi contact both maternal blood and the decidua (i.e., uterine endometrium) and function to attach the placenta to the uterine wall. The outer lining of individual villous trees that is bathed in maternal blood consists of a single layer of multinucleated syncytiotrophoblast, which is responsible for gas and nutrient exchange. The tips of anchoring villi contain extravillous trophoblast cells (EVTs), which attach the placenta to the uterine wall, remodel uterine blood vessels to enable blood flow to the fetus, and mediate histiotrophic nutrition for the embryo. Cytotrophoblast cells serve as progenitor cells for both the syncytiotrophoblast and EVTs. The villous stroma contains fetal macrophages, known as Hofbauer cells, and several types of structural and accessory cells, including an extensive vascular network. Development and remodeling of placental blood vessels is essential for successful pregnancy. Defects in the development and/or functions of the placental vasculature are associated with severe complications of pregnancy, such as preeclampsia (PE), intrauterine growth restriction (IUGR), and preterm birth [1–4]. Furthermore, these complications can directly affect neonatal health and have life-long consequences [5, 6]. Despite these associations, there are significant gaps in our understanding of human placental blood vessel development.

Vascular endothelial growth factor (VEGF) is a potent mitogen and major survival factor for endothelial cells, and it is primarily responsible for the initiation of vasculogenesis and angiogenesis [7, 8]. The VEGF family is composed of several closely related molecules, including VEGF-A, -B, -C, and -D and placental growth factor (PlGF) [9]. VEGFs exert their biological effects through binding one of three receptors (VEGFR1–3), all of which are expressed in the human placenta [3, 8, 10–12]. Knockout studies in mice have highlighted the importance of all three VEGF receptors in the development of the embryo and suggest that these receptors have nonoverlapping functions [13–18]. Altered levels of both the VEGFs and VEGFRs have been detected in the placentas of women suffering from severe complications of pregnancy, which suggests that these molecules are also important for the proper development and function of the human placenta [3, 19–24]. Expression of the VEGFRs has been previously studied in the human placenta [25]; however, comprehensive analyses of the levels and cellular expression patterns of these molecules in placental villi have not been performed over the course of human gestation.

VEGFRs propagate their biological effects by signaling through the PI3-kinase, MAP kinase, and JAK-STAT pathways [26, 27]. Several studies have demonstrated that VEGF can signal through VEGFR2 to activate signal transduction and activator of transcription 3 (STAT3) [27–29]. Expression of the activated, phosphorylated form of STAT3 (pSTAT3) has been observed on the vascular endothelium of tumors, and inhibition of pSTAT3 activity repressed angiogenesis and blocked tumor growth [28]. STAT3-deficient mice are embryonic lethal, and lethality correlated with defects in embryo implantation and decidualization of the placenta [30, 31]. To date, placental expression of pSTAT3 has not been comprehensively examined over the course of human gestation. Furthermore, the potential role(s) of STAT3 in placental vascular development is currently unknown.

Despite the critical roles of the placental vasculature and trophoblast cells in successful pregnancy, the techniques currently available for examining the human placenta are limited by their inability to simultaneously examine the morphology, phenotype, and spatial arrangement of cells within intact placental structures. Techniques previously utilized for examining placental architecture include stereological methods and immunohistochemical techniques [10, 32–35]. Immunohistochemistry and immunofluorescence of tissue sections have provided invaluable insights into placental structure and protein expression, but three-dimensional spatial information, including morphology, cannot be obtained due to the necessity to section the tissues prior to staining with antibodies. Whole mount approaches have been used extensively to characterize patterns of mRNA and protein expression in embryos from multiple species [36]. Although whole mount approaches have been used in a few reports to examine gene expression and infection in the placenta, overall this type of methodology has not been widely used in studies of the human placenta [37–39]. Furthermore, the levels of mRNA and protein expression were not quantified in the previous studies. We therefore developed whole mount immunofluorescence (WMIF) for the human placenta and utilized this method to examine placental cell expression patterns and quantify the levels of the VEGFRs and pSTAT3 over the course of human gestation. Striking images of the placental vasculature and trophoblast layers were obtained using this method. Furthermore, significant differences in the levels and cellular expression patterns of the VEGFRs and pSTAT3 were detected in the human placenta over the course of pregnancy. Specifically, high levels of VEGFR2, VEGFR3, and nuclear pSTAT3 were detected in early placental blood vessels, but these levels were significantly lower at term. Our collective observations suggest that all of these molecules are involved in vascular development in the human placenta. In addition, based on these studies we anticipate that WMIF will provide a powerful approach for detailed analyses of the placental architecture in normal versus pathological pregnancies and for studying the effects of environmental factors on placental function and morphometry.

MATERIALS AND METHODS

Human Tissue Collection

A total of 35 placentas (16 first trimester, 7 second trimester, and 12 third trimester) were utilized in these studies. All third-trimester placentas (≥39 wk) were collected from normal, uncomplicated cesarean section pregnancies at the University of Rochester Medical Center. First-trimester (6–12 wk) and second-trimester (18–22 wk) placentas were collected from selective pregnancy terminations for psychosocial reasons. Pregnancies with known complications or abnormalities, including infection, drug use, and genetic abnormalities, were excluded from this study. All the participants gave informed consent prior to collection. Upon collection, all the tissues were fixed as soon as possible. This investigation and all tissue collections were approved by the Institutional Human Subjects Review Board at the University of Rochester.

Antibodies

The following primary antibodies were used: anti-human CD31 (1:10 dilution for PE-conjugate and 1:40 for APC-conjugate; clone WM59, BioLegend, San Diego, CA), anti-human CD45 (1:10 dilution; clone H130, BioLegend), anti-human CD68 (1:10; clone Y1/82A, BioLegend), anti-human chorionic gonadotropin (hCG) (1:10; clone A0231, Dako, Glostrup, Denmark), anti-human cytokeratin 7 (1:40; clone LP5K, Millipore, Billerica, MA), anti-human HLA-G (1:100; clone MEM-G/11; Thermo Scientific, Waltham, MA), anti-human VEGFR1 (1:66; clone Y103, Abcam, Cambridge, England), anti-human VEGFR2 (1:10; clone 89106, BD Biosciences, San Jose, CA), anti-human VEGFR3 (1:10; clone 54733, R&D Systems, Minneapolis, MN), and anti-human STAT3 (pY705) (1:40; clone 4/P-STAT3, BD Biosciences). All of the primary antibodies used in this study, with the exception of the anti-hCG antibody, were directly conjugated to fluorochromes. The secondary antibody used for hCG was Cy3-conjugated donkey anti-rabbit IgG (1:200; Jackson ImmunoResearch, West Grove, PA).

WMIF

At least three to four intact segments of placental villi (2 × 2 × 1 mm) from each sample were fixed in 90% methanol or 2% paraformaldehyde for 2 h at 4°C, blocked with normal serum of the primary or secondary antibody species, and subsequently stained overnight with antibodies against specific cell surface or intracellular proteins in 1× PBS, 1% bovine serum albumin, 0.1% sodium azide, and 0.3% Triton-X. Samples were prepared as whole mounts as originally described for mouse tumors [40]. Methanol fixation was necessary to detect staining for nuclear-localized pSTAT3, but it worked for all of the other markers examined in this study with the exception of HLA-G. Imaging of stained placental structures was performed using an Olympus BX43 conventional fluorescence microscope or an Olympus FV1000 laser scanning confocal microscope (Olympus America, Center Valley, PA). Images were taken with the same field of view for conventional microscopy using a defined set of fluorescent cubes designed to detect specific fluorochromes. Confocal images were taken using a 20× UPlanS-Apo oil objective (NA 0.85), a Kalman setting of 2–3, and sequential scanning option in the University of Rochester School of Medicine and Dentistry Confocal and Conventional Microscopy core. For the generation of images, the microscope control settings and exposure times were established for each antibody and used throughout the remainder of the study. Monochrome images were then analyzed using ImagePro Plus software (Media Cybernetics, Rockville, MD).

Image Analysis

An average of 6.4 fields of view was randomly chosen for image analysis of each placental sample. The percentages of VEGFR2- and VEGFR3-positive vessels were determined using ImagePro software by manually thresholding the CD31 image and the corresponding VEGFR2 or VEGFR3 images to create binary image masks that included only the portions of the images positive for CD31, VEGFR2, or VEGFR3, respectively. Thresholding values for VEGFR2 and VEGFR3 were kept constant for all the images analyzed. The areas for each positive stain (VEGFR2 or VEGFR3) were then divided by the total CD31 area to define the percentage of vessels positive for either VEGFR2 or VEGFR3 in each image. The mean pixel intensity of the VEGFR2 or VEGFR3 regions was also determined by averaging over all the pixels positive for both CD31 and either VEGFR2 or VEGFR3.

The pSTAT3 positive nuclei were quantified by first manually thresholding the corresponding CD31 and 4′,6-diamidino-2-phenylindole (DAPI) images to create binary image masks of blood vessels and nuclei, respectively. The overlap between these masks (obtained using an AND operation) thus provided a mask for the nuclei of CD31-positive blood vessels. The overlap between the CD31-nuclear mask and the pSTAT3 image (using another AND operation) then produced an image of the pSTAT3 intensities within each endothelial nucleus, from which mean nuclear pSTAT3 intensity was finally determined.

Statistical Analysis

Statistical data analyses were conducted with GraphPad Prism software using one-way analysis of variance (ANOVA) followed by the Bonferroni post hoc test. Data are presented as means ± standard error of mean (SEM). Differences were considered statistically significant at values of P < 0.05.

RESULTS

Analysis of Human Placental Morphology Using WMIF

The techniques currently utilized for visualizing the human placenta lack the ability to simultaneously analyze the morphology, phenotype, and spatial arrangement of placental cell populations. Thus, we developed placental WMIF to overcome this limitation. Small, intact segments of first-trimester placenta were dissected and immediately fixed, stained overnight with fluorochrome-conjugated antibodies to identify villous trophoblast cell populations and stromal cells, and subsequently imaged using multicolor WMIF (Fig. 1). We were able to distinguish the outer syncytiotrophoblast layer, the underlying cytotrophoblast cells, and the EVTs using antibodies specific to human chorionic gonadotropin (hCG), cytokeratin-7 (CK7), and human leukocyte antigen (HLA)-G, respectively. The syncytiotrophoblast layer stained for both hCG (Fig. 1A) and CK7 (data not shown), and displayed a honeycomblike morphology due to its multinucleated nature (Fig. 1A). The underlying cytotrophoblast cells stained positively for CK7 (Fig. 1B) and negatively for hCG (data not shown), and were present in large numbers in the first trimester. EVTs exhibited a HLA-G⁺/CK7^bright phenotype (Fig. 1C; data not shown). Additionally, a substantial number of fetal macrophages, or Hofbauer cells, were detected within the placental villi in samples that were concurrently stained for CD45 and CD68 (Fig. 1D). Using WMIF, we obtained unparalleled images of villous blood vessels with the endothelial cell marker CD31, identifying potential arterioles, venules, and extensive capillary networks showing vascular hierarchy (Fig. 1E). No staining was observed when placental samples were incubated with fluorescently-conjugated secondary antibodies alone or in control samples incubated in the absence of antibody (data not shown). In order to define the three-dimensional spatial arrangement of placental cell populations relative to each other, and to colocalize the markers of interest, we further adapted this technique for confocal microscopy. We were able to demonstrate the close proximity of the syncytiotrophoblast layer and underlying cytotrophoblast cell layer utilizing WMIF in conjunction with confocal microscopy (Fig. 2A and Supplemental Movie S1; all the supplemental data is available online at www.biolreprod.org). Furthermore, confocal analysis demonstrated that CD45⁺ immune cells were uniformly distributed throughout the stroma (Fig. 2B and Supplemental Movie S2). Thus, multiple different cell populations could be simultaneously visualized within the same tissue sample, including trophoblast cells, immune cells and blood vessels. Taken together, our results demonstrate the capacity of WMIF, in conjunction with either conventional or confocal microscopy, to simultaneously characterize the morphology, phenotype, and spatial arrangement of several distinct cell populations within an intact placental microenvironment.

FIG. 1.

Identification of distinct cell populations within intact placental villous structures using whole mount immunofluorescence (WMIF). Intact placental villi from first-trimester placentas were fixed and stained with antibodies against hCG (A), cytokeratin-7 (B), HLA-G (C), CD45/CD68 (D), or CD31 (E), and imaged using conventional microscopy. The schematic shows first-trimester villous structures and the relative locations of defined cell populations. Bars = 50 μm except in the CD31 image, which is 100 μm.

FIG. 2.

Imaging of first-trimester placental samples using WMIF and confocal microscopy. A) Still frame captured from Supplemental Movie S1 demonstrating the proximity of the syncytiotrophoblast layer and underlying cytotrophoblast cells in first-trimester placenta. Tissues were stained with antibodies to distinguish the hCG⁺/CK7⁺ syncytiotrophoblast layer, hCG⁻/CK7⁺ cytotrophoblast cells, and CD45⁺ immune cells, and subjected to WMIF in conjunction with confocal microscopy. Bars = 50 μm. B) Still frame captured from Supplemental Movie S2 demonstrating the spatial arrangement of trophoblast cells, immune cells, and blood vessels in first-trimester placental villi. Placental samples were stained with antibodies to distinguish CK7⁺ cytotrophoblast cells, CD45⁺ immune cells, and CD31⁺ blood vessels, and imaged as described for A.

Examination of Placental Blood Vessel Morphology Throughout Human Gestation Using WMIF

The striking images of first-trimester vasculature obtained using WMIF led us to further explore placental blood vessel morphology and phenotype throughout the course of gestation. Visualization of placental villous blood vessels from first-, second-, and third-trimester explants by conventional and confocal microscopy revealed drastic changes in vessel morphology (Fig. 3). Blood vessels in the first trimester were largely of a webbed morphology (Fig. 3A), indicative of immature angiogenic blood vessels, which have yet to undergo vascular remodeling such as pruning. In contrast, blood vessels in the third trimester presented with more straight and tortuous capillaries (Fig. 3A), indicative of a mature, remodeled phenotype. The vasculature in the second trimester had an intermediate phenotype, exhibiting both webbing as well as narrow, looping capillaries (Fig. 3A). These observations were confirmed using confocal microscopy (Fig. 3B). Overall, these results are consistent with angiogenesis in the human placenta during normal pregnancy, which begins with a complex network of immature blood vessels during the first trimester. After remodeling, tightly looped capillaries are formed by the third trimester [32].

FIG. 3.

Placental blood vessel morphology changes during the course of gestation. To investigate the morphology of placental vessels throughout gestation using WMIF, intact placental villi from first-, second-, and third-trimester placentas were stained with antibodies against CD31 and subsequently imaged using either conventional (A, upper panel) or confocal (B, lower panel) microscopy (n = 7–16 placentas per gestational age group). Bars = 50 μm.

Differential Expression of VEGFRs Within Human Placental Villi Throughout the Course of Gestation

Based on the dynamic changes in placental vascular morphology observed over the course of gestation, we utilized WMIF to compare placental expression of VEGFR1, VEGFR2, and VEGFR3 in the first, second, and third trimesters of pregnancy, respectively. VEGFR1 was predominately expressed in the trophoblast layer and stromal immune cells throughout pregnancy, with little to no detectable expression on CD31-positive blood vessels (Fig. 4 and Supplemental Fig. S1). The levels of VEGFR1 expression did not change appreciably over the course of gestation (data not shown). In contrast, expression of VEGFR2 was restricted primarily to placental endothelial cells in all three trimesters of pregnancy (Fig. 5A). Quantification, as described in Materials and Methods, revealed an approximately 2-fold decrease in vascular VEGFR2 expression in the third trimester compared to the first and second trimesters (Fig. 5B). In addition, a small, but statistically significant ∼8% decrease in the percent coverage of VEGFR2 on villous blood vessels was detected in the third trimester (Fig. 5C). Similar to VEGFR2, high levels of VEGFR3 were observed on early villous endothelial cells (Fig. 6). VEGFR3 expression decreased approximately 1.5-fold on villous blood vessels in the third trimester compared to the first trimester (Fig. 6B), while the percent coverage of VEGFR3 decreased ∼50% (Fig. 6C). In addition, VEGFR3 was also expressed on the trophoblast layer; however, the expression levels on trophoblast cells did not significantly change over the course of gestation (Supplemental Fig. S1). These collective results demonstrate that VEGFR1, VEGFR2, and VEGFR3 have distinct patterns of cellular expression in the human placenta over the course of pregnancy. VEGFR2 and VEGFR3, but not VEGFR1, are abundantly expressed on placental blood vessels during the first trimester, but expression levels decrease significantly at term.

FIG. 4.

VEGFR1 is weakly expressed on placental blood vessels throughout the course of gestation. Explants from first-, second-, and third-trimester placentas were stained with antibodies to VEGFR1 and CD31 and subjected to WMIF in conjunction with confocal microscopy. Weak/no staining of VEGFR1 on CD31⁺ blood vessels is shown with an arrowhead. Bars = 50 μm.

FIG. 5.

Expression of the angiogenic receptor VEGFR2 on placental blood vessels significantly decreases during the third trimester. A) WMIF was performed on first-, second-, and third-trimester placentas using antibodies to CD31 and VEGFR2 and imaged using confocal microscopy. B) The intensity of VEGFR2 staining on placental blood vessels was quantified as described in Materials and Methods and represented as mean pixel intensity (arbitrary units: A.U.). C) Quantification of the percent coverage of VEGFR2-positive vessels. Each symbol on the graphs in B and C represent data obtained from one placenta (n = 7–11 placentas per gestational age group; error bars in graph are defined as mean ± SEM; **P < 0.01 vs. second-trimester placentas and ***P < 0.001 vs. first-trimester placentas). Bars = 50 μm.

FIG. 6.

VEGFR3 expression is significantly decreased on placental blood vessels during the third trimester. A) Images of VEGFR3-positive vessels in the first, second, and third trimester using WMIF. Arrowheads show costaining of VEGFR3 and CD31 on endothelial cells. Quantification of the intensity of VEGFR3 on vessels (B) and the percentage of VEGFR3-positive vessels (C) was performed as described for Figure 5. Symbols in B and C represent data from one placenta (n = 7–16 placentas per gestational age group; error bars in graphs are represented as mean ± SEM; ***P < 0.001 vs. first-trimester and/or second-trimester placentas). Bars = 50 μm.

Differential Expression of the Activated Transcription Factor STAT3 on Human Placental Villous Blood Vessels During Normal Human Pregnancy

VEGFR2 was previously shown to promote angiogenesis through activation of STAT3 [27–29]. To investigate the potential role of activated pSTAT3 in human placental blood vessel development, we adapted WMIF to detect intracellular signaling molecules and subsequently examined pSTAT3 expression in first-, second-, and third-trimester placental tissues. Staining for pSTAT3 was detected primarily in the nuclei of endothelial cells in first- and second-trimester placentas, but the intensity of pSTAT3 staining was significantly diminished in the blood vessels by the third trimester (Fig. 7A). Quantification of nuclear pSTAT3 expression in placental endothelial cells revealed an approximately 2-fold decrease in the intensity of pSTAT3 levels in the third trimester compared to the first (Fig. 7B). Taken together, our studies reveal a direct correlation among the levels of VEGFR2, VEGFR3, and pSTAT3 expression in placental endothelial cells over the course of pregnancy.

FIG. 7.

Expression of the activated transcription factor pSTAT3 on placental blood vessels decreases during the course of pregnancy. A) WMIF images of pSTAT3 staining in first-, second-, and third-trimester placenta. Arrowheads demonstrate colocalization of pSTAT3 with DAPI on CD31⁺ endothelial cells. B) Intensity of nuclear pSTAT3 staining on vessels (n = 6–13 placentas per gestational age group) was quantified as described in Materials and Methods and represented as mean pixel intensity (arbitrary units: A.U.). Error bars are defined as mean ± SEM; *P < 0.05 versus first-trimester placentas. Symbols on the graph in B represent data from one placenta. Bars = 50 μm.

DISCUSSION

In this study, we developed WMIF in order to simultaneously visualize multiple parameters of the human placenta, including the morphology, phenotype, and spatial arrangement of cell populations. Utilizing this technique, we demonstrated that the morphology and phenotype of the human placental vasculature dynamically changes over the course of gestation. Early placental blood vessels exhibited a webbed morphology, indicative of an immature state of angiogenesis, and by the third trimester, the morphology of these vessels consisted mainly of long, looping capillaries, corresponding to mature, remodeled endothelium. Placental endothelial cells expressed high levels of the proangiogenic receptors VEGFR2 and VEGFR3 during the first and second trimesters of pregnancy, along with high levels of the activated transcription factor pSTAT3. The expression of all of these molecules significantly decreased on placental blood vessels in the third trimester. Additionally, VEGFR1, which is predominately a negative regulator of angiogenesis during development, was expressed at very low levels on the placental vasculature throughout the course of gestation. We also observed the expression of VEGFRs and pSTAT3 on nonendothelial cells within placental villi, including trophoblast and stromal cells, using WMIF. Thus, this study provides both a qualitative and quantitative analysis of VEGFR and pSTAT3 expression in human placental villi during the course of gestation. However, it is important to note that although the levels of VEGFR and pSTAT3 expression were comparable within different villi from the same placental sample, and among different placentas from the same stage of gestation, the current study did not compare expression levels in terminal versus intermediate and stem villi. Future studies utilizing WMIF to examine markers of interest should address this important issue.

Although placental vascular development is critical during pregnancy, there have been limited methods available to simultaneously analyze vessel morphology, phenotype, and spatial arrangement within an intact microenvironment. To date, development of the human placental vascular system has been examined mainly by immunohistochemical and stereological techniques [3, 10, 32–35]. While these methods have provided valuable insights into placental vascular biology, they lack the ability to provide detailed morphology in conjunction with phenotypic and dimensional information. Recent observations suggesting that several cell types within the placenta, including trophoblast cells, macrophages, and stromal cells, function together to promote placental development and angiogenesis [41–43] underscore the importance of being able to simultaneously examine all of these critical parameters. Using WMIF, we were able to overcome these limitations and obtain unparalleled images of the morphology, phenotype, and spatial orientation of several placental cell populations, including endothelial cells. Our current study, in conjunction with the recent application of WMIF to study mouse implantation sites [44], conclusively demonstrates the power of this approach for examining both human and animal placental tissues.

In addition to detecting expression of VEGFRs on placental vascular endothelium, we also observed the expression of VEGFR1 and VEGFR3 on nonendothelial placental cells. Trophoblast cells expressed VEGFR1 and VEGFR3 throughout the course of gestation. Previous studies suggest that the expression of both VEGFR1 and VEGFR3 on trophoblast cells plays a role in trophoblast invasion and survival [45]. It has also been suggested that activation of VEGFR3 on trophoblast cells stimulates proliferation and may contribute to transport between the maternal and fetal circulation [3]. In addition, the expression of VEGFRs on trophoblast cells allows them to respond to VEGF and PlGF, which have been shown to act as chemoattractants, promoting syncytialization of trophoblast cells and regulating proliferation and apoptosis [46–48]. Similar to what has been observed on other macrophages, we observed expression of VEGFR1 on Hofbauer cells [49, 50]. Although the exact function of VEGFR1 on Hofbauer cells is not known, previous studies in nonplacental macrophages suggest that it may function in regulating chemotaxis [49, 50]. Thus, it is likely that the expression of VEGFRs on nonendothelial cells plays a different role than VEGFRs on endothelial cells.

The transcription factor STAT3 plays key roles in multiple processes, including embryonic development, tumorigenesis, and responses to cytokines [51]. Furthermore, STAT3 has been shown to be a critical factor for endothelial cell survival and proliferation [51, 52]. Specifically, activation and nuclear translocation of STAT3 in endothelial cells has been shown to promote migration and tube formation, suggesting roles in vascular development [53]. The importance of STAT3 in development is evident in STAT3 knockout mice, which are embryonic lethal [30]. The primary reason for reduced viability in these mice is a defect in trophoblast invasion, leading to failures in implantation and placental development [51, 54–56]. To date, there is only one report examining pSTAT3 expression in the human placenta [57]. This study detected pSTAT3 primarily in trophoblast cells, which is consistent with the interpretation of the mouse knockout studies. Our current study confirmed the expression of pSTAT3 in trophoblast cells in first-trimester human placentas, specifically within the columnar cytotrophoblast cells (Supplemental Fig. S2). In addition, we also detected sporadic pSTAT3 staining within villous stromal cells (data not shown). However, our study significantly extends the previous work by demonstrating that intense staining for nuclear pSTAT3 was detected principally in first- and second-trimester placental blood vessels, which subsequently decreased significantly by term. Thus, our results strongly suggest that STAT3 activity also plays an important role in placental vascular development.

VEGFRs signal through several different cascades, including the MAP kinase, PI3-kinase, and JAK-STAT pathways [26, 27]. VEGF binding to VEGFR2 activates the transcription factor STAT3 in mouse tumor blood vessels and human endothelial cells [27–29]. The importance of nuclear pSTAT3 in angiogenesis was previously demonstrated in tumor vascular endothelium, where treatment with antagonists of STAT3 led to decreased tumor angiogenesis [28]. Our results reveal high levels of VEGFR2 and nuclear pSTAT3 on vascular endothelial cells in the first trimester but not in the third trimester. Furthermore, our results demonstrate for the first time, to our knowledge, a correlation between expression of VEGFR2 and active STAT3 in human placental vasculature. Based on our observations, we hypothesize that STAT3 is also playing a role in vascular development in the human placenta via VEGFR2 and propose the following model. During the first trimester, VEGF is highly expressed [58] and binds to VEGFR2 present on the vascular endothelium. This leads to activation and nuclear translocation of STAT3, which stimulates transcription of target genes involved in proliferation of placental endothelial cells and thus promotes angiogenesis. During this stage, the vessels are immature in nature and have yet to undergo remodeling, as evidenced by the webbed morphological appearance of the vasculature in the first trimester. Toward the end of gestation, VEGF [10, 58] and VEGFR2 levels are low compared to early pregnancy, which result in reduced STAT3 activation, and thus attenuation of the signaling required for active angiogenesis. This is consistent with the morphological analysis demonstrating that the placental vessels at term are visibly more uniform in diameter and thus reminiscent of mature vessels that have undergone pruning and remodeling. To the best of our knowledge, this is the first study that implicates a role for STAT3 in angiogenesis within the human placenta.

Abnormal vascular development has been associated with severe pregnancy complications, including PE, IUGR, preterm birth, and diabetes [1–4]. Furthermore, these complications have been linked with alterations in the levels of VEGF and VEGFRs, which could affect normal angiogenesis in these placentas. Our demonstration that WMIF can be successfully used to simultaneously examine the phenotype, morphology, and spatial arrangement of placental cell populations provides convincing evidence that this method will provide a powerful approach to study these parameters in complications of pregnancy, such as PE, IUGR, and placental infection. Furthermore, ongoing studies in the laboratory indicate that WMIF can be utilized to examine placental responses to soluble factors such cytokines and hormones as well as environmental factors such as toxins and drugs. Thus, we expect that the use of WMIF will ultimately lead to significant advancements in our fundamental understanding of human placental biology.