Objective
Mouse embryonic exposure to alcohol, lithium, and homocysteine results in intrauterine growth restriction (IUGR) and cardiac defects. Our present study focused on the placental effects. We analyzed the hypothesis that expression of nonmuscle myosin (NMM)-II isoforms involved in cell motility, mechanosensing, and extracellular matrix assembly are altered by the 3 factors in human trophoblast (HTR8/SVneo) cells in vitro and in the mouse placenta in vivo.
Study Design
After exposure during gastrulation to alcohol, homocysteine, or lithium, ultrasonography defined embryos exhibiting abnormal placental blood flow.
Results
NMM-IIA/NMM-IIB are differentially expressed in trophoblasts and in mouse placental vascular endothelial cells under pathological conditions. Misexpression of NMM-IIA/NMM-IIB in the affected placentas continued stably to midgestation but can be prevented by folate and myoinositol supplementation.
Conclusion
It is concluded that folate and myoinositol initiated early in mouse pregnancy can restore NMM-II expression, permit normal placentation/embryogenesis, and prevent IUGR induced by alcohol, lithium, and homocysteine.
Our previous studies using the mouse model demonstrated that an acute embryonic exposure to the drug lithium (Li ++ ), to an elevated serum level of the natural metabolite homocysteine (HCy), or to a binge-drinking level of alcohol (ethanol [EtOH]) during gastrulation on embryonic day (ED) 6.75 results in similar fetal cardiac and placental anomalies, including intrauterine growth restriction (IUGR). This timing of exposure extrapolates to 16-19 days of human pregnancy. Before the 4-chambered heart forms, a beating, tubular human heart with extra- and intraembryonic circulations is present on day 21 after fertilization, bringing blood to the embryo.
Given that 49% of pregnancies are unintended, the timing and effects of early exposure to environmental factors as alcohol, Li ++ , or HCy, together with prophylactic mechanisms are critical to define. Placental blood flow represents a large fraction of cardiac output at early stages of development. Because nonmuscle myosin-IIs (NMM-IIs) are important in processes associated with placentation (eg, hemodynamic mechanosensing and cell migration ), we analyzed possible expression alterations in this class of proteins in placentas associated with IUGR in the mouse model.
Class II myosins comprise a protein family that includes the sarcomeric, smooth muscle, and the NMM-IIs. Each NMM-II isoform is composed of 2 identical nonmuscle heavy chains (NMHCs) and 2 pairs of light chains. NMM-II is a cytoskeletal protein that interacts with actin, contributing to cell motility, cell polarity, chemotaxis, cell adhesion, cytoskeletal tension sensing, and cytokinesis. In mammals there are 3 NMM-II isoforms, NMM-IIA, NMM-IIB, and NMM-IIC. Three separate genes, MYH9 , MYH10 , and MYH14 , encode the NMHCs, NMHC-IIA, NMHC-IIB, and NMHC-IIC, respectively.
NMM-IIA/IIB are essential during embryogenesis and placentation. Little is known of their role and expression in the placenta. When NMM-IIA is deleted by germline ablation, early embryonic death results from altered cell-cell adhesion, lack of visceral endoderm, and placental defects. NMM-IIA has an important role in placental development, which appears different from its function in cell adhesion. In NMM-IIA mutated embryos displaying embryonic lethality, placental vascularization was abnormal and placentas were more compact. Although NMM-IIB is present in the placenta, it cannot compensate for the novel function of NMM-IIA.
In contrast, when NMM-IIB is deleted, embryos die around ED14.5 because of heart and brain defects. We demonstrated that NMM-IIB is important in cardiac myofibrillogenesis, left-right asymmetry, and heart looping. It has a role in cardiac trabeculation and mechanotransduction of blood flow forces. NMM-II isoforms and smooth muscle myosin were defined biochemically to be present in the human placenta, suggesting our studies are relevant to human pregnancy. This may be of specific importance for individuals that have mutations in NMHC-IIA that comprise a spectrum of human diseases.
Invasion of human trophoblasts is promoted through activation of Wnt signaling, indicating this pathway has a role in placental development and morphogenesis. Lithium mimics the canonical Wnt pathway by inhibiting glycogen synthase kinase-3. Lithium also has been shown to deplete phosphoinositides that act within the Wnt second-messenger phosphatidylinositol signaling pathway. Importantly, phosphatidylinositol 3-kinase/protein kinase B (PI3K/AKT) and phosphatidylinositol-anchored proteins have a role in trophoblast cell differentiation. Thus, we analyzed whether addition of myoinositol (MI) to the media used for culturing the human first-trimester trophoblast cell line (HTR-8/SVneo cells) would show any additional benefit than use of folate (FA) alone. MI supplementation was expected to stabilize normal phosphatidylinositol signaling, preventing depletion by environmental factors and thus normalizing multiple intermediaries and processes important in placentation. Interestingly, during embryonic development, a plateau seemingly has been reached in FA reduction of neural tube defects (NTDs), and a portion of NTDs are resistant to FA prevention. With these resistant forms as well as in diabetic embryopathy, the embryo is protected by myoinositol supplementation.
Our objectives were to define in trophoblast HTR-8/SVneo cell changes in cell proliferation and migration that may occur with Li, HCy, and EtOH exposure, and alterations in the expression of NMM-IIA or NMM-IIB. Importantly, NMM-IIA and -IIB are associated with mechanotransduction, cell motility, and extracellular matrix (ECM) assembly, all integral to placenta development.
We subsequently addressed the changes in the NMM-II isoform expression patterns that occur in vivo in mouse placentas of embryos exposed to Li, elevated HCy, and alcohol; and analyzed specifically those embryos displaying IUGR, abnormal hemodynamics, and cardiac anomalies.
Lastly, we compared the abnormal NMM-II expression patterns from the fetuses described in earlier text with placentas of FA-rescued mouse embryos in which pregnancy complications were prevented. Although the mouse placenta is considered an imperfect model for studies relating to the human placenta, many similarities exist on the molecular and cellular levels. The different cell types of the murine placenta described in our study all have their counterparts in the human placenta. The analyses of placental alterations experimentally induced in the mouse are expected to enable characterization of mechanisms involved in development of human placental abnormalities.
Material and Methods
A human first-trimester trophoblast cell line HTR-8/SVneo has been well characterized and was used in this study. This cell line was established from explant culture of first-trimester human placenta and was immortalized by transfection with the gene encoding simian virus 40 large T antigen. The resulting cell line (HTR-8/SVneo) shares phenotypic similarities with the progenitor cells; and proliferation, migration, and invasiveness were shown to be regulated by the same signaling molecules that modulate extravillous trophoblast cells. Generally, this cell line has been considered a highly useful model to study mechanisms regulating normal human trophoblast proliferation, migration, and invasiveness in response to environmental factors. The cells were cultured in an atmosphere of 5% CO 2 /95% air in RPMI 1640 medium containing 5% fetal bovine serum (FBS) and 1% penicillin and streptomycin. The medium was changed every 3 days.
Cell growth curve experiments
HTR-8/SVneo cells (5 × 10 3 cells/well) were seeded into 96 well plates (Corning 3610; Corning, New York, NY). Cells were treated with the following substances added to the culture medium: ethanol (100 mM), FA (75 μM), MI (50 μM), HCy (50 μM), and lithium chloride (LiCl [50 μM]). The control group was HTR-8/SVneo cells incubated in culture medium without test substances. The cell culture media was replaced on day 3 and assayed on day 4.
The number of viable HTR-8 cells was determined colorimetrically using the CellTiter 96 AQueous One Solution cell proliferation assay (Promega, Madison, WI). After 6 hours of treatment, the CellTiter 96 AQueous solution (20 μL) was added to each well and incubated for an additional 2 hours. The absorbance (490 nm) was recorded using Gen 5 Biotek Synergy 2 Microplate Data Collection (Biotek, Winsookie, VT). These procedures were conducted daily for 5 days to obtain the proliferation curves. The experiment was repeated 3 times for a total of 10 samples for the following groups: EtOH, FA, EtOH plus FA, MI, EtOH plus MI, EtOH plus combination of MI/FA, and FA alone; quadruplicates were obtained for HCy, HCy plus FA, and duplicates for LiCl and LiCl plus FA.
Correlation between cell numbers and color formation was determined using cell concentrations (1 × 10 4 , 1.2 × 10 5 , 2.3 × 10 5 , 3.4 × 10 5 , 5.6 × 10 5 , 6.7 × 10 5 , 7.8 × 10 5 , 8.9 × 10 5 , 1 × 10 6 ) in 96 well plates with RPMI 1640 medium containing FBS and antibiotics. After 1 hour equilibration, the Cell Titer 96 Aqueous solution (Promega) was added to each well. After 2 hours, the absorbance (490 nm) was recorded as above.
HTR-8/SVneo-scratch wound assay for cell migration and immunostaining
Cells (5 × 10 5 /mL) were directly cultured using treated glass culture slides, 8 chambers, in RPMI 1640 medium as described above and incubated until 90% confluence. The analysis was carried out in triplicate for each experimental condition. Culture medium was changed to expose the cells to HCy, HCy/ FA, EtOH, EtOH/FA, or FA alone and incubated for 2 hours. The monolayer of cells then was scratched once from the left to the right wall of the well with a sterile plastic tip, resulting in a cell-free cleft. The cells were fixed at 8 hours after wounding and at 24 hours. This scratch assay permitted following specific cell markers during the early steps of directed migration of the trophoblasts at the cell-free boundary as cells begin to invade the gap at 8 hours and to analyze the same markers after a 24-hour incubation in the different experimental media. Results are shown after 24 hours because those differences were most notable.
Mitotic cells were immunostained with Histone 3 antibody (Cell Signaling Technology, Inc, Beverly, MA) and Cy-3-conjugated secondary antibody. Other antibodies used included a rabbit polyclonal antibody for fibronectin (Sigma, St. Louis, MO), and affinity-purified rabbit antibodies for NMM-IIA and NMM-IIB (Covance, Emeryville, CA) characterized by Western blotting. An antibody for active β-catenin was purchased from Millipore (Temecula, CA). The horseradish peroxidase-conjugated secondary goat antirabbit antibody was from Vector Laboratories (Burlingame, CA) and the Cy3-conjugated secondary goat antirabbit antibodies were from Jackson Labs (West Grove, PA).
Ultrasound analysis of placental blood flow
The protocol for use of animals was approved by the University of South Florida Institutional Animal Care and Use Committee. On ED 6.75, pregnant mice were given a single intraperitoneal (ip) injection of lithium, homocysteine, or a binge drinking level of alcohol (the latter by 2 ip injections at 3:00 and 6:00 pm ). This dose was used in our previous study and was based on the well-described and generally used model of binge alcohol dose for pregnant mice using 2 ip injections. Echocardiography on each ED 15.5 embryo was carried out noninvasively using Vevo 770 instrumentation (VisualSonics, Inc, Toronto, Canada) to analyze embryonic cardiac function and umbilical artery blood flow.
Mouse placental tissue
Control placentas and placentas from embryos displaying abnormal blood flow patterns and cardiac defects were fixed in paraformaldehyde, embedded in paraffin, and sectioned. Mouse placental tissue was immunostained as published.
Statistical analysis
The statistical analysis of ultrasound parameters showing significant values of experimental exposure in comparison with control group was based on nonparametric Kruskal Wallis test. Significance is based on P < .05.
Results
Using HTR-8/SVneo trophoblast cells, cell proliferation was not altered after EtOH, Li, or HCy exposure, with and without FA/MI supplementation of media. Figure 1 , A shows that EtOH, FA only, EtOH plus FA, EtOH plus MI, or MI only did not significantly affect cell proliferation in vitro of this trophoblast cell line. Similarly, HCy or Li, with or without FA supplementation, by day 5 did not affect proliferation ( Figure 1 , B). The decline in Figure 1 , B appearing on day 4 in our assay may be due to change of culture medium on day 3. The decrease was within the SD. For detection of mitotic cells after HTR-8/SVneo cell exposure to EtOH, HCy, or Li, with and without FA, we immunostained cells using histone 3 antibody ( Figure 2 ). Confirming the proliferation assay results cited in earlier text, no significant difference in numbers of mitotic cells was observed. We next analyzed effects on cell migration.
HTR-8/SVneo cell migration scratch assay model
Using an in vitro scratch model of wounding, wound-induced cell migration was assayed at 24 hours as cells migrate to close the wound gap. β-Catenin, an important intermediary within the canonical Wnt pathway, showed nuclear localization in cells both near the edge of the wound as well as away from the edge ( Figure 3 , A-C ), indicating active Wnt signaling. The ECM protein fibronectin (FN) is highly expressed and assembled by the HTR-8/SVneo trophoblast cells ( Figure 3 , D-F). Twenty-four hours after wounding, FN is expressed ( white arrows ) in all cells greater than 106 μm ( green line ) away from the wound edge and also present near the free edge ( Figure 3 , F). Similarly, NMHC-IIA and NMHC-IIB are expressed by HTR-8/SVneo cells. Cells near the free edge express higher levels of NMHC-IIA and NMHC-IIB as cells migrate to fill the gap ( Figure 3 , G-I, and Figure 3 , J-L, respectively).
Scratch assay model with HCy exposure and FA/MI supplementation: up-regulation of FN, NMHC-IIA, and NMHC-IIB
FN localization
FN is synthesized by untreated human HTR-8/SVneo extravillous trophoblast control cells at the wound edge ( Figure 4 , A ; cell nuclei in blue shown by overlay of 4′,6-diamidino-2-phenylindole [DAPI] staining). FN localization shows a relatively low level of expression by cells at the wound edge with HCy exposure ( Figure 4 , B). With FA supplementation (HCy/FA; Figure 4 , C), FN expression is slightly elevated approximately 100 μm away from the gap edge with little change at the cell-free edge at gap. With the addition of MI to the incubation medium, with or without HCy and FA, there is a notable up-regulation of FN ( Figure 4 , D and E) with high levels 89 μM to 110 μM away from the wound edge and slightly higher expression seen near the gap.
NMM-IIB localization
NNHC-IIB localization is at a relatively high level in untreated control cells after 24 hours near the edge of the wound gap ( blue DAPI stained nuclei shown as an overlay in Figure 4 , F). With HCy exposure, NMHC-IIB is seen generally at a low level of expression ( Figure 4 , G); an elevation of NMM-IIB is apparent with FA supplementation of the media, Figure 4 , H and J). With the addition of MI to the incubation medium, as well as with FA, NMM-IIB is expressed at control levels ( Figure 4 , I and J). Therefore, the addition of MI and FA up-regulated the FN synthesis and NMHC-IIB expression in the human HTR-8/SVneo cells to levels apparent in saline-treated control cultures without HCy.
Altered umbilical and intraembryonic blood flow observed by doppler ultrasound analysis after exposures to alcohol, Li ++ or HCy
Of these 3 environmental factors assayed, a binge level of alcohol exposure had the most severe effect on embryonic cardiac function when compared with Li ++ exposure or to HCy, the latter having the most moderate effect. This was reflected also in the pulsatility indices of the ductus venosus (DV; Figure 5 , A ), umbilical artery (UA; Figure 5 , B), descending aorta (DA; Figure 5 , C), and in the cardiac outflow velocity (OFV; Figure 5 , D). The morphometric and cardiovascular parameters are summarized in the Table .
| Parameters | Uninjected (n = 24) | Ethanol (n = 30) | Lithium (n = 68) | HCy (n = 59) |
|---|---|---|---|---|
| Morphometric measurements | ||||
| CRL, mm | 15.33 | 12.60 a | 13.80 a | 13.08 a |
| BW, g | 0.44 | 0.37 a | 0.37 a | 0.35 a |
| PW, g | 0.13 | 0.10 a | 0.11 a | 0.11 a |
| Valvular regurgitation | ||||
| SLV, % | 0 | 76.70 a | 51.50 a | 47.50 a |
| AVVR, % | 0 | 6.70 a | 0.00 | 6.80 a |
| Myocardial performance index | ||||
| MPI | 0.47 | 0.65 a | 0.60 a | 0.44 |
| IRT% | 12.04 | 17.20 a | 16.82 a | 11.93 |
| Arterial Doppler | ||||
| UA PI | 1.31 | 1.80 a | 1.75 a | 1.54 |
| DA PI | 1.53 | 1.76 | 2.10 a | 1.98 |
| OFV, cm/s | 35.78 | 28.62 | 41.05 | 39.92 |
| Venous Doppler | ||||
| DV PI | 0.86 | 1.19 | 0.97 | 0.97 |
Mouse embryos exposed acutely during gastrulation to the specified environmental factors also demonstrated significant IUGR and reduced placental weights, when compared with the control group. The placentas of mouse embryos displaying IUGR, cardiac valve regurgitation, and abnormal Doppler waveforms were analyzed for changes in NMHC-IIA and NMHC-IIB expression to compare with control tissue.
A single exposure of the pregnant mouse on ED 6.75 to ethanol, Li ++ or HCy differentially altered placental NMM-IIA and NMM-IIB into midgestation (ED 15.5) of IUGR embryos
Comparing with control placentas ( Figure 6 , A-D ), a single exposure to EtOH during gastrulation on ED 6.75 enhanced NMHC-IIA expression (brown horseradish peroxidase localization) within the maternal decidua and labyrinth layers and to a lesser extent in the fetal side of the chorionic plate. This was detectable more than a week later ( Figure 6 , E-H). There was little signal in the spongiotrophoblast cells. The placentas were smaller than those of control embryos. Dietary supplementation with FA initiated the morning after conception rescued normal embryonic heart development from alcohol exposure effects and led to relatively normal control levels of NMHC-IIA in the maternal decidua and labyrinth layers. With FA, however, maternal decidua layer continued to be smaller than in control placentas. With and without environmental exposure or FA supplementation, the fetal chorionic plate NMHC-IIA expression remained stable.
