Background
Epidemiological findings and studies with animal models reveal that impaired growth
in utero and size at birth are critical determinants of the onset of various disease processes later in life [
1]. Optimum intrauterine growth depends on several factors including proper nourishment of the developing fetus by maternal blood through the placenta [
2]. Accordingly, proper development of the placenta plays a significant role in determining the health and well being of the offspring. Trophoblasts, the outer layer of cells in the blastocyst, not only facilitate its attachment with the uterine epithelium but play a significant role in the formation of the placenta by invading the uterine tissue and differentiating into several cell types with endocrine, vascular, immunological or transport functions [
3]. Therefore, interference with trophoblast invasion may result in abnormal development of the placenta resulting in suboptimal nourishment to the fetus.
Drugs cross the placental barrier by ultrafiltration, diffusion, active transport or by special processes, such as pinocytosis, or through breaks in placental wall and access the developing fetus [
4]. Valproate sodium (VPA), used in the treatment of epilepsy and bipolar disease, is lipophilic, is actively transported into the trophoblast, and crosses the placenta through passive diffusion and by interacting with placental carnitine transporter [
5‐
8]. Valproate therapy during pregnancy causes a wide range of congenital and behavioral malformations in children [
9]. Animal studies demonstrate histopathological changes in the extraembryonic and embryonic tissues, such as necrosis of cytotrophoblasts and suppressed proliferation of fetal capillaries, following exposure to valproic acid [
10,
11]. The immunomodulatory and anticancer drug thalidomide (THA) also passes through the placental barrier and causes several defects in children including neurobehavioral problems, ear and limb malformations [
12]. Although use of this drug during pregnancy is prohibited in the United States, it remains a potential teratogen in many parts of the world [
13‐
15]. Exposure to thalidomide increases mitotic activity of cytotrophoblast, forming irregular masses of cells with or without syncytiotrophoblast surrounding it, forming structures identical to embryoid bodies [
16,
17]. Alcohol (ethanol; ALC) diffuses across the placenta and affects the developmental program of the fetus at many levels resulting in fetal alcohol syndrome [
18]. Exposure to alcohol during pregnancy increases the number and size of trophoblasts and dilates cisterns of rough endoplasmic reticulum. Alcohol during pregnancy causes hyperplasia of capillary basal lamina, hypertrophy of trophoblastic basal lamina, irregular vascularization and hyperemia at the basal zone and labyrinth of placenta [
19,
20]. All three teratogens change the proliferation and migration of various cell types
in vitro and
in vivo [
21‐
23] and cause placental pathologies [
10,
11,
16,
24‐
26]. Thus it is possible that these teratogens may alter the number and the invasion of trophoblasts, changing placental ultrastructure.
To date, no study has been conducted to examine the effects of these teratogens on the migration and proliferation of human placental trophoblast. Therefore in this study, the effects of valproate, thalidomide and alcohol on the migration and proliferation of first trimester trophoblast cell line were examined
in vitro. Because dose of toxicant is a critical determinant of developmental toxicity and is likely to be a key factor responsible for interspecies variability in response to many test agents [
27], both low and high concentrations of drugs and alcohol were tested on the trophoblast cells. Since trophoblasts invade maternal tissues in multiple layers [
28,
29], migration assays were conducted both in mono and multilayers to examine the effects of drugs and alcohol on the migration of trophoblast in multiple layers. Moreover because changes in α
5β
1 integrin receptor-mediated adhesion are known to alter migration of trophoblast [
30], the effects of valproate, thalidomide or alcohol exposure on the expression levels of β
1 and α
5 integrin subunits were examined in the trophoblast cell line in culture by Western blotting.
Discussion
This study reports a novel way of studying cell migration in vitro using suction-wounds that can be created using a sterile tip and a vacuum unit. This method is superior to conventionally used scratch assays because wounds with comparable perimeter can be created in wells that allow examining cell migration and healing in relatively similar wounds. In addition, the entire wound area can be visualized under low magnification (4×) for study. This obviates measurement errors during imaging at different times of culture that is common with conventional scratch assays and occur because of differences in the width of the wound along the length of the scratch.
The results suggest that valproate, thalidomide and alcohol may influence migration of human first trimester trophoblast. Therefore, exposure to these teratogens during the first trimester of pregnancy may interfere with the normal development of placenta. This may cause suboptimal nourishment of developing embryos resulting in developmental defects. Results demonstrate that the changes in the migration rate of human first trimester trophoblast after drug and alcohol treatments may result from the alteration in the expression levels of β1 and α5 integrin subunits.
Data presented here reveal for the first time that the migration rates of cells in monolayer, and as reported in several studies using scratch assays, differ from those of cells in multilayers, a situation that is relatively more realistic to what is seen
in vivo [
28,
29]. Although experiments conducted do not explain the mechanisms for the differences in the migration rate of cells between mono or multilayer wounds, it is likely that the accelerated migration of cells in multilayers wounds is due to better enrichment of medium with the cell migration enhancing factors secreted by relatively larger number of cells per well. These molecules may include extracellular matrix proteins regulating trophoblast migration, cytokines (TGF-β), growth factors (IGF-II), decorin, plaminogen activators, endothelin-1 and hormones regulating trophoblast migration [
40‐
48]. Therefore, migration of invading trophoblast
in vivo, where they invade in multiple layers [
28,
29], may be favored by the sufficiency of these molecules.
All three teratogens tested in this study are reported to alter the migration of different cell types, albeit differently.
In vitro tests show that valproate may increase or decrease the migration of different glioma cell lines [
49,
50] and neural crest cells individually or in sheets [
49]. Thalidomide changes the migration of cells differently at different concentrations [
51] and cell type. It increases migration of multiple myeloma cells and inhibits migration of human keratinocytes [
52,
53]. Valproate and thalidomide exposure during prenatal development causes abnormal positioning of serotonergic neurons in rats [
22], and migration of cortical neurons in rat fetal brains are delayed following prenatal exposure to alcohol [
21]. There are numerous reports on alcohol's ability to alter the migration of various cell types, including mouse trophoblast, in culture differently [
23,
54‐
58]. Therefore, it is obvious that these teratogens target cell migration machinery differently in different cell types depending upon the concentration and cellular milieu. Data presented here also demonstrate that drugs and alcohol change the migration rate of trophoblast differently depending upon concentration and cell density (Table
1). Therefore, it is possible that exposure to these teratogens during placentogenesis may influence invasion of first trimester trophoblast differently depending upon the dosage and the cellular surroundings. The disparity in the effects of low concentrations of alcohol on the migration of cells between mono and multilayers supports this possibility. Differences in the drug- or alcohol-induced changes in the migration rates between the mono and multilayer wounds may derive from the capacity of each drug to influence secretion of ECM and factors influencing migration machinery, and possibly may be due to their ability to change the cell-plate and cell-cell interactions.
Expression of integrin subunit α
6, α
5 and β
1 mRNA, and the roles of α
6β
1, α
6β
4, α
1β
1, α
2β
1, α
5β
1, α
vβ
1 and α
vβ
3 integrin receptors regulating the migration of trophoblast were reported earlier [
31,
59‐
61]. Data presented here provide evidence for the expression of α
3A, α
4, and α
IIb integrin subunit mRNA and splicing of α
6 integrin transcripts in a first trimester human trophoblast cell line. Detection of α
IIb mRNA in first trimester trophoblast cell line and a recent report describing the role of α
IIbβ
3 integrin receptor in trophoblast migration in mice [
28] implies possible involvement of this receptor in the migration of human trophoblast. Additionally, the existence of α
6A and α
6B mRNA splice variants in the human trophoblast cell line hint for additional regulatory controls of trophoblast invasion mediated by α
6β
1 or α
6β
4 receptors [
61‐
63].
Western blotting data show that valproate, as well as thalidomide and alcohol, changes the expression levels of β
1 and α
5 integrin subunits. Therefore treatments with these drugs or alcohol may alter the migration of trophoblasts by changing the availability of subunits for the formation of active α
5β
1 integrin receptors on the cell surface. Western blot data presented in the tabulated format (Table
2) reveal that the pattern of changes in the expression levels of β
1 and α
5 subunits by thalidomide or alcohol are almost similar, whereas valproate altered the expression levels of β
1 and α
5 integrin subunits differently. Nonetheless, it is likely that exposure to these drugs and alcohol during first trimester of pregnancy may alter the invasion of trophoblasts by disturbing the subunit stoichiometry during the formation of functional α
5β
1 receptors.
Enhanced adhesion of trophoblasts by α
5β
1 receptor is shown to inhibit the invasion of human trophoblasts [
30], but this model may not completely explain changes in the migration rate of trophoblasts following alcohol treatments. This is because alcohol increased α
5 and β
1 integrin subunit levels at both low and high concentrations but accelerated migration of trophoblasts in monolayer and inhibited in multilayer wounds. These results indicate that expression levels of β
1 and α
5 integrin subunits may not be the sole determining factor for the migration rate of cells. Obviously, changes in the expression pattern of other integrin subunits α
1, α
2, α
v, α
6, α
IIb, β
3 and β
4 that are known to be expressed in trophoblast [
28,
30,
62,
63], and down-stream molecules regulating integrin-mediated migration may also account for these differences. Besides, because cell migration is a dynamic process, changes in the expression of these molecules may not be steady during the course of migration. Therefore, studying steady state levels of integrin subunit expression pattern may not completely explain the migratory behavior of cells. Studies of molecules regulating cell migration in real-time in untreated and treated cells may be required to further clarify the mechanisms.
Valproate, thalidomide and alcohol influence proliferation of various cells [
53,
64‐
68]. Therefore, changes in the number of trophoblasts caused by these teratogens during the migration assay may influence the healing rates of the wounds. To examine the extent of this possibility, efficacy of low and high concentrations of teratogens on the number of trophoblasts was tested in a time frame relevant to the early phase of healing (3 and 12 h), when the migration rates were rapid. Data obtained indicated that changes in cell numbers due to treatments were mild and significant only at low concentrations of thalidomide or both at low and high concentrations of alcohol. Therefore, treatment of cells with these drugs is not likely to influence the migration rate of trophoblasts robustly. Besides, changes in the number of cells did not correlate with changes in the migration rate of cells. For instance, 12 h of treatment with 25 μM thalidomide increased the cell numbers significantly, but the migration of cells at this concentration was inhibited. Similarly, alcohol at low concentration increased the number of cells more than those treated with high concentration for 12 h, while the migration rate of cells in monolayer wounds was accelerated more by high concentration of alcohol during 3 to 12 h. These observations suggest that changes in the migration rate of cells were not influenced significantly by differences in proliferation rates induced by treatments, at least during the initial 12 h of incubation.
It is possible that
in vivo effects of drugs on the migration and proliferation of trophoblasts may not be the same as those observed in this cell line tested
in vitro. For instance, valproate inhibits human sperm motility differently
in vivo as compared to
in vitro conditions [
69]. Maternal genotype regulating metabolism of drugs and alcohol may also influence the outcome [
47,
70]. In addition, the direct action of teratogens tested on cells in culture may not be same as those occurring due to the drug or alcohol and their metabolites
in vivo. Therefore, data obtained
in vitro will require verification in control and drug-exposed human placentas of different genotypes.
One of the likely consequences of poor trophoblast migration and invasion is the development of preeclampsia. Increased risk of preeclampsia due to valproate therapy is reported [
71], but no association between thalidomide exposure or maternal alcohol drinking with preeclampsia is known so far. Therefore, it is likely that disturbances in the development of placenta due to changes in the invasiveness of trophoblast may not be the sole determining factors for the risk of preeclampsia. Of the three teratogens tested in this study, only valproate may influence these additional factor/s.
Authors' contributions
UK Rout planned the project, conducted experiments, analyzed data and wrote the manuscript.