Hum. Reprod. Advance Access originally published online on July 11, 2008
Human Reproduction 2008 23(10):2282-2291; doi:10.1093/humrep/den198
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Trophoblastic invasion in vitro and in vivo: similarities and differences
1 Institute of Cell Biology, Histology and Embryology, Center of Molecular Medicine, Medical University of Graz, Harrachgasse 21, A-8010 Graz, Austria 2 Institute of Biophysics, Center of Physiological Medicine, Medical University of Graz, A-8010 Graz, Austria 3 AplaGen GmbH, D-52499 Baesweiler, Germany
4Correspondence address. Tel: +43-316-380-4256; Fax: +43-316-380-9625; E-mail: christine.helige{at}meduni-graz.at
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionFundingAcknowledgementsReferences |
|---|
BACKGROUND: The basic mechanisms of trophoblast invasion are not completely understood. This may be due to the lack of suitable in vitro models which enable experimental modulation of this complex process. In the present study, a three-dimensional co-culture model is used for comparing two factors considered to be implicated in the regulation of trophoblast invasion, the expression of HLA-G and apoptosis, in vitro and in vivo.
METHODS: Tissue fragments from human first trimester decidua parietalis were put in close contact with spheroids of AC-1M59 trophoblast/choriocarcinoma hybrid cells as a model of the invasive trophoblast. Cryostat sections from these co-cultures were immunohistochemically stained and compared with first trimester placentation sites in vivo. RESULTS: Only the invasive trophoblast-derived cells showed an intensive staining for HLA-G, whereas the cells on the periphery of the confrontation culture exhibited only a weak staining. A similar staining pattern was found in vivo. Both in vitro and in vivo CD45+ apoptotic leukocytes were frequently detected in close proximity to the invasive trophoblastic cells.
CONCLUSIONS: In this co-culture system, key factors considered to be implicated in trophoblast invasion in vivo can also be demonstrated in vitro. Therefore, it may help in finding strategies for the management of diseases associated with deficient trophoblast invasion.
Key words: apoptosis/HLA-G/invasion model/natural killer cells/trophoblastic invasion
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionFundingAcknowledgementsReferences |
|---|
In the first trimester of pregnancy, extravillous trophoblast cells deeply invade the maternal decidua thereby coming in close contact with numerous maternal leukocytes present in the uterine mucosa (Trundley and Moffet, 2004
). Although being semi-allogeneic, they are neither attacked by T lymphocytes nor by natural killer (NK) cells. For these purposes, invading trophoblast cells seem to have developed various strategies to prevent themselves from being killed by maternal leukocytes present in the uterine mucosa. In this context, certain types of putative tolerance signaling molecules have been discussed, including the CD200-CD200R interaction, HLA-G and HLA-E, and apoptosis inducing molecules such as Fas/Fas ligand and indolamine 2,3-dioxygenase, which allow the invasion of the semi-allogeneic fetal cells into the uterus without being rejected (Clark, 2005).
As HLA-G is of particular interest, its potential role in the escape of the cells from immune recognition and destruction has been discussed in detail. The protein has only been detected within the extravillous trophoblast population of first trimester cell columns, in invasive extravillous trophoblast cells, in syncytiotrophoblast at a few special sites at the basis of an anchoring villus as well as in the term chorion laeve and basal plate (Le Bouteiller et al., 1999; Blaschitz et al., 2005), in the endometrial glandular epithelium but not in eutopic endometrium (Barrier et al., 2006) and in a subpopulation of thymic epithelial cells (Mallet et al., 1999). HLA-G interacts directly or indirectly via HLA-E with specific killer inhibitory receptors present on the surface of leukocytes hereby protecting the fetal cells from the maternal immune attacks (Ponte et al., 1999; King et al., 2000; Hunt, 2006). Its major function is to program cells into immunosuppressive phenotypes, although several other targets of HLA-G have also been identified (Hunt et al., 2005, 2007).
In the placenta, in addition to the expression of HLA-G, trophoblast cells have several other defense mechanisms against leukocytes, e.g. the apoptosis-inducing Fas/Fas ligand system. However, there are various different data concerning apoptosis at the feto–maternal interface. Both trophoblast cells and decidual leukocytes have been reported to be involved (Hammer and Dohr, 1999; Emmer et al., 2002; Von Rango et al., 2003; Pongcharoen et al., 2004; Qiu et al., 2005; Huppertz et al., 2006).
Nevertheless, the basic mechanisms of the regulation of trophoblast invasion are far from being completely understood. This may be due to the lack of suitable in vitro models which enable a direct examination and/or a modulation of cellular interactions during invasion. Most of the studies on trophoblast invasion in vitro have used extracellular matrix barriers as substrates of invasion in which cell–cell interactions could not be examined (Trew et al., 2000; Karmakar and Das, 2002; Korff et al., 2004; LaMarca et al., 2005). However, there are also several more complex in vitro systems. Already in 1993 Vicovac et al. used co-cultures of placental villous tissue from 8–12 weeks of gestation and decidua parietalis in order to examine trophoblast differentiation in the extravillous lineage. Carver et al. (2003) established an in vitro implantation model in which human hatched blastocysts were co-cultivated with human endometrial stromal cell monolayers. Interactions of trophoblast cells with decidual spiral arteries were analyzed in models developed by Cartwright et al. (2002) and Dunk et al. (2003). Another three-dimensional organ culture system consisting of co-cultures of rounded fragments of human luteal phase endometrium and multicellular spheroids of choriocarcinoma cells has also been served as a model of trophoblast invasion (Grümmer et al., 1994; Helige et al., 2001).
In the present study, this slightly modified co-culture system, which consists of rounded fragments of human first trimester decidua parietalis and multicellular spheroids of AC-1M59 choriocarcinoma/trophoblast hybrid cells, is used to analyze two key factors considered to be involved in the regulation of invasion, the expression of HLA-G and apoptosis. The results are compared with the in vivo conditions at the feto–maternal interface.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionFundingAcknowledgementsReferences |
|---|
Cells and culture conditions
The fusion of the human hypoxanthine–guanine–phosphoribosyltransferase-defective choriocarcinoma cell line AC1-1 with normal extravillous human trophoblast cells derived from the chorion laeve of the term placenta resulted in the formation of trophoblast/choriocarcinoma hybrid cells (Funayma et al., 1997
; Gaus et al., 1997). Several monoclonal cell lines with different tumorigenicities derived from those fusions. One of those clones, the non-tumorigenic AC-1M59 cells (commercially available from DSMZ, Braunschweig, Germany), was used as a model for the invasive trophoblast. These cells have already been characterized and their tumorigenicity has previously been tested (Frank et al., 2000; Schmitz, 2002). They were used to imitate the in vivo conditions more closely than by using pure tumor cells. The cells were cultivated in DMEM/Ham's F12 medium (PAA Laboratories, Linz, Austria) containing 10% fetal calf serum (Cambrex, Verwiers, Belgium) and antibiotics (PAA Laboratories). After every fifth passage the cells were cultivated in a medium containing 100 µM hypoxanthine (Sigma, St Louis, USA) and 5.7 µM azaserine (Sigma) in order to eliminate unfused choriocarcinoma cells after the fusion procedure.
Multicellular spheroids of AC-1M59 cells
Exponentially growing monolayers were gently detached by accutase (Innovative Cell Technologies, San Diego, USA; supplied by PAA Laboratories). The resulting single-cell suspension was cultivated in Petri dishes with non-adherent surfaces to form multicellular aggregates. After a cultivation period of 2 days, the round-shaped cell aggregates were transferred into 25-ml spinner flasks and stirred at 
130 rpm using a magnetic stirrer system (Telesystem 06.40 and Telemodul 40C, H+P Labortechnik AG, Oberschleißheim, Germany). Multicellular spheroids with a diameter of 300 µm were selected under a dissecting microscope equipped with an ocular grid.
Fragments of human first trimester decidual tissue
Human first trimester decidual tissue derived from the decidua parietalis was obtained from the Department of Obstetrics and Gynecology, University Hospital, Graz, Austria after legal termination of pregnancy (week 8–10). The procedure was approved by the Ethical Committee of the Medical University of Graz. Decidua basalis had to be excluded as invasive trophoblast cells may be found within the tissue before confrontation with the trophoblast/choriocarcinoma hybrid cells. Prior to the co-culture experiments, the absence of invasive trophoblast cells in decidual tissue was determined by staining cryostat sections from decidual tissues with anti-cytokeratin and/or anti-HLA-G antibodies. Only tissue fragments free of invasive trophoblast cells were used for confrontation cultures.
The tissue was cut into pieces of 1 mm in diameter under a dissecting microscope equipped with an ocular grid, then transferred into 25-ml spinner flasks and stirred at 130 rpm in a humidified CO2 incubator at 37°C and 20% O2. The culture medium (DMEM; PAA Laboratories) was supplemented with 10% fetal calf serum (Cambrex), antibiotics (PAA Laboratories) and 20 ng/ml progesterone, 10 ng/ml gestonoroncapronate (19-nor-17a-hydroxy-progesterone, a progesterone analogon) and 300 pg/ml 17-β-estradiol (kindly provided by Dr Halfbrodt, Schering, Berlin, Germany). In addition to progesterone, gestonoroncapronate serving as a depot compound was included in the culture medium in order to avoid fluctuations of the hormone level. The medium was changed daily and the tissue fragments were cultivated for 3–4 days prior to confrontation with the multicellular spheroids of the AC-1M59 cells. After that time, the decidual tissue fragments were round-shaped and had a diameter of 800 µm.
Confrontation cultures of rounded fragments of human first trimester decidual tissues with multicellular spheroids of AC-1M59 cells
The method used is based on an invasion model previously established (Grümmer et al., 1994; Helige et al., 2001), except that the endometrial tissue was replaced by decidual tissue. In brief, one pre-cultivated fragment of decidual tissue derived from first trimester decidua parietalis, 800 µm in diameter, and one multicellular spheroid of AC-1M59 trophoblast/choriocarcinoma cells were each transferred into one out of 15 small funnels inserted in silicon pads (Fig. 1). The confrontation cultures were covered with DMEM/Ham's F12 medium containing the hormones mentioned above and incubated for at least 24 h in a humidified CO2 incubator at 37°C and 20% O2. This incubation period was required for the formation of stable contacts between the confrontation partners. Thereafter, the co-cultures were transferred into 25-ml spinner flasks and stirred at 130 rpm using the magnetic stirrer system Telesystem 06.40 and Telemodul 40C (H+P Labortechnik AG) for up to 5 days. The culture medium containing the hormones was changed daily. Thereafter, the confrontation cultures were embedded in tissue-freezing medium (Tissue Tek OCT-compound; Sakura Finetek, Inc., Torrance, USA) and processed for immunohistochemistry. Each confrontation culture was sectioned serially. Decidual tissue fragments from 7 different pregnancies were used and at least 10 confrontation cultures from each decidua were analyzed.
|
Analysis of apoptosis and immunofluorescence of decidual leukocytes
In order to detect apoptotic nuclei, the TUNEL assay (Ylem, Rome, Italy) was performed as described previously (Hammer and Dohr, 1999
). In brief, 5-µm cryostat sections were fixed for 15 min at room temperature with 4% buffered paraformaldehyde and, after rinsing in phosphate-buffered saline (PBS), immersed in a solution of 0.1% Triton X-100 and 0.1% sodium citrate (w/v) for 2 min at 4°C. After rinsing in PBS, the TUNEL assay was performed using the APOPTOSIS-I.S.-kit following the manufacturer's instructions excepting that streptavidin-peroxidase was replaced by streptavidin-FITC in a dilution of 1:15. After performing the TUNEL assay, the sections were rinsed in PBS, incubated with the UV blocking solution (LabVision, Fremont, USA) for 10 min and with a mouse anti-human cytokeratin antibody for 30 min at room temperature, rinsed in PBS and thereafter incubated with a CY3-labeled goat anti-mouse antibody. After blocking free binding sites of the anti-mouse antibody with normal mouse serum (1:150) for 15 min, a mouse anti-human CD45 antibody labeled with a CY5 labeling kit was applied according to the manufacturer's instructions. Thereafter, the sections were rinsed in PBS, mounted with Moviol (Calbiochem-Novabiochem, La Jolla, USA) and sequentially analyzed on a confocal laser scanning microscope (Leica SP2, Leica Lasertechnik GmbH, Heidelberg, Germany) using the 488 nm laser line for the excitation of FITC, 543 nm for CY3 and 633 nm for CY5, respectively. Detection settings were: 500–535 nm for FITC, 555–620 nm for CY3 and 665–750 nm for CY5. Negative controls were performed by omitting the enzymes of the TUNEL assay and by replacing the antibodies by IgG1.
Immunohistochemistry
Cryostat sections of 5-µm from confrontation cultures were thawed, air dried and fixed in acetone for 5 min at room temperature. After rinsing in tris-buffered saline (TBS), pH 7.2, the sections were incubated with an UV block containing 10% human serum for 10 min. Thereafter, the sections were incubated with the appropriate concentrations of the primary antibodies for 30 min at room temperature, rinsed thoroughly in TBS and exposed to the Ultravision horse-radish peroxidase (HRP)-labeled streptavidin biotin visualization system (LabVision) following the manufacturer's instructions. The peroxidase was developed with 3-amino-9-ethyl-carbazole/dimethylformamide (AEC) in acetate buffer in the dark for 4 min. The reaction was stopped by rinsing in distilled water. The sections were counterstained with Mayer's hemalum and mounted with Kaiser's glycerol gelatine (Merck, Vienna, Austria).
For double-labeling the Ultravision/HRP visualization system was followed by the alkaline phosphatase anti-alkaline phosphatase (APAAP) system (Dako, Glostrup, Denmark). For this purpose, after the development of the peroxidase with AEC, the sections were incubated with the second primary antibody for 30 min at room temperature. Thereafter, the sections were exposed to the APAAP link antibody for 25 min, rinsed in TBS and incubated with the APAAP (1:50) for 25 min. After rinsing in TBS, the alkaline phosphatase was developed with 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium (Dako) under microscopical control. The reaction was stopped by rinsing in distilled water. The sections were counterstained with Mayer's hemalum and mounted with Kaiser's glycerol gelatine (Merck).
Negative controls were incubated with the appropriate IgG fractions as isotype controls.
Sections were viewed through an Axiophot microscope and photographs were taken using an AxioCam HRc digital camera (Zeiss, Oberkochen, Germany).
The following antibodies were used:
- monoclonal mouse anti-human cytokeratin (clone MNF116; Dako): 1:100;
- monoclonal mouse anti-human CD56 (clone MY31; Becton Dickinson, San Jose, USA): 1:50;
- monoclonal mouse anti-human HLA-G (clone MEM-G/1; Exbio, Praha, Czech Republic): 1:100;
- monoclonal mouse anti-human HLA-G (clone 4H84, kindly provided by Dr. Olga Genbacev, University of California, San Francisco, USA): 1:1000;
- monoclonal mouse anti-human Ki-67 (clone MIB-1; Dako): 1:100;
- monoclonal mouse anti-human CD45 (BD PharMingen, San Diego, USA) labeled with a CY5 labeling kit (Amersham Biosciences Europe GmbH, Freiburg, Germany): protein concentration: 8 µg/ml;
- polyclonal goat anti-mouse-CY3 (Jackson ImmunoResearch Laboratories, Inc., Baltimore Pike, USA): 1:300.
Semi-quantitave evaluation of the HLA-G expression during trophoblastic invasion in vitro
Digital image processing was used to evaluate the HLA-G expression semi-quantitatively. Immunohistochemically stained median sections from the co-cultures were used for the semi-quantitative analysis. The images of the co-cultures were segmented by color thresholding in order to achieve an image of the stained AC-1M59 cells without segmenting the stained decidual component which was set to the background color of the images (white). Color thresholding was done by converting the individual color channels (R, G and B, respectively) into the equivalent HLS (Hue, Lightness, Saturation) color model. The HLS model enabled a more accurate extraction and segmentation of the AC-1M59 cells compared to the common RGB color model.
A gray level image was developed, in which the gray level represented the staining saturation. Since the invasion process often leads to a radial symmetric staining pattern in rather round-shaped confrontation cultures maintained in spinner culture, the radial distribution of the staining saturation represents the spatial distribution of the HLA-G expression. Starting from the center of confrontation cultures, the gray level intensities were measured along 360 directions towards the periphery of the images. These 360 directions were orientated with an angular difference of 1° over the image (Fig. 2). The average distribution over one image and the average of seven sample images (corresponding to seven confrontation cultures) were plotted. Image processing was done with IDL (Creaso, Gilching, Germany) and an interactive image processing tool (IQM, developed by Dr Ahammer; http://www.meduni-graz.at/qmnm/iqm/index.htm).
|
Statistics
The trend to staining saturation was determined by linear regression analysis. Radial symmetric gray level distribution was measured for 360 individual directions leading to 360 individual distributions. The mean distribution was calculated, representing the radial symmetric distribution of a confrontation's image. Seven sample images were used and for each radius (radial distance from the confrontation's center), the mean, the standard deviation, the confidence interval for the mean and the correlation coefficient were calculated. The confidence level was set to 0.95 (significance level
= 0.05). |
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionFundingAcknowledgementsReferences |
|---|
HLA-G expression of AC-1M59 cells
The HLA-G expression of AC-1M59 cells cultivated as monolayers and as multicellular spheroids was determined using the monoclonal HLA-G-specific antibody MEM-G/1 which recognized the
1 domain of HLA-G. Many, but not all, of the AC-1M59 cells cultivated as monolayers expressed HLA-G (Fig. 3a). The HLA-G+ cells were distributed randomly within the culture. There was a remarkable heterogeneity in the staining intensity. Some cells were intensively stained by the antibody, whereas others only showed a faint staining or were not stained at all.
|
AC-1M59 cells cultivated as multicellular spheroids exhibited a similar HLA-G-staining pattern to the one observed with monolayer cultures, although there were less HLA-G+ cells found in multicellular spheroids and the staining intensity was lower compared with the cells cultivated as monolayers. Intensively stained cells were found both on the periphery and in the inner parts of the multicellular spheroids (Fig. 3b). Usually, there was no area of central necrosis in multicellular spheroids of 300 µm in diameter.
Invasion in vitro and in vivo
After more than one week of cultivation, the decidual tissue was well preserved and did not show any necrotic areas in the central parts of the tissue fragments. Within the tissue fragments, several decidual NK cells expressed the Ki-67 antigen demonstrating proliferative activity without any exogenous stimulus (Fig. 4). A similar number of proliferating NK cells was found in freshly prepared tissue (data not shown). Apart from several glands, neither cytokeratin+, nor HLA-G+ cells were found in the stroma prior to confrontation with the AC-1M59 cells, indicating that no normal trophoblast cells had invaded the decidual tissue prior to the invasion experiment (data not shown). AC-1M59 cells were highly invasive in the three-dimensional organ culture model. The cells started to invade the rounded fragments of decidual tissue after 2 days of co-cultivation. After 5 days, several finger-like protrusions of the AC-1M59 cells became obvious and some single cells had already detached from the multicellular aggregates and were found lying deeply in the stromal component of the confrontation culture (Fig. 5a and b). In the decidual tissue fragments, numerous CD56+ NK cells were present even after several days of cultivation. At the invasion front those NK cells were often seen in close proximity to the AC-1M59 cells (Fig. 6).
|
|
|
Almost all of the AC-1M59 cells, including the highly HLA-G+ invasive ones, expressed the Ki-67 antigen indicating that both proliferation and invasion could appear on identical cells in this trophoblast-derived cell line (Fig. 7a and b).
|
At the feto–maternal interface in vivo, however, the invasive trophoblast cells were negative for Ki-67 (Fig. 7c).
Expression of HLA-G during invasion in vitro and in vivo
In co-cultures of multicellular spheroids of AC-1M59 cells with rounded fragments of human first trimester decidua, the HLA-G expression was demonstrated using two monoclonal antibodies, clones 4H84 and MEM-G/1, both recognizing the free heavy chain of HLA-G. AC-1M59 cells located directly at the invasion front as well as the cells which had already penetrated deeply into the host tissue showed the most intensive staining for HLA-G. An intensive staining was particularly found in cells being in close contact to the decidual cells. Most of the AC-1M59 cells at the periphery of the confrontation culture or cells not in contact with the decidual tissue were only stained weakly (Figs 5c and d and 6b). That staining pattern was found with both HLA-G-specific antibodies and with several analyzed co-cultures, although more positive cells and a higher staining intensity became visible after staining with 4H84 compared with MEM-G/1. Particularly noticeable was the close contact of highly HLA-G+ AC-1M59 cells with CD56+ NK cells at the invasion front (Fig. 6c).
A similar pattern of the HLA-G staining intensity was found during trophoblast invasion in vivo. Trophoblast cells of the cell columns showed a rather weak staining, whereas the cells invading the decidual tissue were more intensively stained (Fig. 5e).
Semi-quantitative evaluation of the HLA-G expression during invasion in vitro
Many confrontation cultures were round shaped after 4 days in spinner culture and only those co-cultures were used for the measurement (Fig. 2). The different expression of HLA-G by AC-1M59 cells during invasion, based on the radial symmetric staining saturation, was evaluated semi-quantitatively using computerized image processing. It showed increased expression rates in the inner parts of the confrontation cultures, where the highly invasive cells which had already invaded deeply into the host tissue were found (near the center of the confrontation culture as well as at a distance of 120–170 µm from the center) which seemed to correspond to the invasion front (Fig. 8). The staining saturation decreased in the peripheral parts of the confrontation culture where the AC-1M59 cells were not in contact with the decidual tissue. The calculated confidence intervals and the high correlation coefficient (R2 = 0.95) for the linear regression model strengthened the statistical significance.
|
Analysis of apoptosis during invasion in vitro and in vivo
The TUNEL assay revealed some sparse apoptotic cells throughout the whole confrontation culture. Many cells belonging to decidual glands, identified by staining with the anti-cytokeratin antibody and by their specific morphology, showed a positive TUNEL staining. This seems to be a normal part of decidualization as that staining was also found in the decidua basalis, where trophoblast cells normally invade. Some single apoptotic cells which showed neither a staining with the anti-cytokeratin nor with the anti-CD45 antibody, the marker of leukocytes, were identified as decidual stromal cells. Apoptosis of these cells is considered to be part of the normal cell turnover. Some of the apoptotic cells were identified as AC-1M59 cells by both morphology and the anti-cytokeratin antibody. However, those cells were usually not lying close to the leukocytes, but were rather found in the central parts of the AC-1M59 component of the confrontation cultures. At the invasion front of the AC-1M59 cells, CD45+ apoptotic leukocytes were frequently found (Fig. 9a).
|
At the feto–maternal interface in vivo, it was the leukocyte population again showing fragmented DNA. Numerous double TUNEL+ and CD45+ cells were found in close proximity to the invasive trophoblast cells (Fig. 9b).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionFundingAcknowledgementsReferences |
|---|
Several tissue culture systems have been used to study trophoblast invasion and its regulation in vitro. In most of the experiments, extracellular matrix barriers have served as a substrate for invasion (e.g. Trew et al., 2000
; Karmakar and Das, 2002; Korff et al., 2004; LaMarca et al., 2005) although more complex in vitro co-culture techniques have also been established (Vicovac et al., 1993; Grümmer et al., 1994; Helige et al., 2001; Cartwright et al., 2002; Dunk et al., 2003). Hanna et al. (2006) presented a combined in vitro–in vivo model of trophoblast invasion in which decidual NK cells embedded in Matrigel were injected into nude mice. The cells were able to grow in vivo within those palpable plugs. Thereafter, trophoblast cells were injected which invaded the Matrigel and interacted with the existing leukocytes. That system was used for studying the chemokine receptor expression of invasive trophoblast cells and their specific receptors on decidual NK cells. Both the trophoblast cells and the decidual NK cells, however, were isolated and not maintained in their natural environment. In the present study, tissue fragments from first trimester decidua parietalis were used as a substrate for trophoblastic invasion (Helige et al., 2006). In this model the interactions of invasive trophoblast-derived cells with decidual leukocytes which are considered to play an important role during trophoblast invasion, can be examined within the tissue, where they are normally present, without any isolation procedures. Under suitable experimental conditions, decidual NK cells were even able to proliferate within cultivated tissue fragments without any exogenous stimulation by cytokines thus confirming the results found for cultivated endometrial tissue (Helige et al., 2001). In contrast, a suboptimal dose of interleukin (IL)-2 and a co-stimulation with decidual stromal cells were required for the proliferation of isolated decidual NK cells (King et al., 1999). Our system also had the advantage that decidual lymphocytes were not subjected to disaggregation or purification which may alter their phenotype and behavior. The AC-1M59 cells serving as a model of the invasive extravillous trophoblast were highly invasive. In contrast to normal trophoblast cells, these cells grow permanently and they form multicellular spheroids. This is of particular importance, since normal extravillous trophoblast cells only have a limited life span and do not proliferate in vitro. Cell proliferation, however, is a prerequisite for the formation of multicellular spheroids which were used in the co-culture experiments. The hybrid cells exhibit many characteristics of normal extravillous trophoblast including the expression of cytokeratin, extracellular matrix molecules, oncofetal proteins (Schmitz, 2002), and HLA-G which is demonstrated in the present study. Several trophoblast-derived cell lines have been used for studies on placental biology. However, most of them have not shown any or shown only a very weak expression of HLA-G, e.g. the extravillous trophoblast cell lines HRT-8/SVneo, HT-116 and SGHPL-4 (Zdravkovic et al., 1999; Lash et al., 2006). Therefore, for studies of the interactions of HLA-G-expressing trophoblastic cells with decidual leukocytes, or whenever in vitro studies on the implication of HLA-G during trophoblast invasion are performed, AC-1M59 trophoblast/choriocarcinoma hybrid cells can be used as a good model of the invasive HLA-G-expressing extravillous trophoblast. Nevertheless, one major difference remains: the co-existence of proliferation and invasion in the trophoblast/choriocarcinoma cells. In the three-dimensional trophoblast invasion model, almost all of the invasive AC-1M59 cells also expressed the Ki-67 antigen indicating that the invasive pathway did not exclude cellular proliferation. Therefore, in this regard the hybrid cells showed a similar behavior to choriocarcinoma cells. In contrast, normal trophoblast cells show a separation between proliferation and invasion. Their invasive character only appears after the cessation of proliferation.
HLA-G is considered to be implicated in trophoblast invasion. In the present study, we show that the expression of this molecule seems to be regulated by decidual tissue or by factors derived from it. AC-1M59 cells cultivated as monolayers or as multicellular spheroids without any contact to decidual tissue expressed HLA-G. The highly positive cells were randomly distributed in both cultures. However, in co-cultures of multicellular spheroids of AC-1M59 cells with fragments of first trimester decidua parietalis, the AC-1M59 cells directly at the invasion front as well as the cells which had already invaded deeply into the host tissue showed a more intensive staining for HLA-G compared with the cells lying within the multicellular spheroid thus being without any contact to the decidual tissue. According to these results, it can be assumed that the decidual tissue may up-regulate HLA-G expression of these cells, at least in this co-culture system. In human placental trophoblast in vivo, a similar staining pattern was observed. There was a gradient of increasing staining intensity for HLA-G beginning at the cell column, where the trophoblast cells showed a rather weak staining for HLA-G towards the decidua, where highly HLA-G+ invasive trophoblast cells were found. These data confirm previous studies (Goldman-Wohl et al., 2000; Emmer et al., 2002). The factors up-regulating the expression of HLA-G during trophoblast invasion have not been determined so far. Two potential candidates may be cytokines produced by uterine NK cells or trophoblasts themselves. In vitro experiments have shown that IL-10 produced by both NK cells and human trophoblast (Roth et al., 1996; Vigano et al., 2001), induces HLA-G mRNA transcription in trophoblast organ culture (Moreau et al., 1999) and therefore, may also stimulate HLA-G antigen expression. IFN-
increases HLA-G mRNA and HLA-G antigen expression in macrophages and enhanced the cell surface expression of HLA-G in JEG-3 choriocarcinoma cells and in macrophage cell lines (Yang et al., 1996; Lefebvre et al., 1999). Under the experimental conditions during co-cultivation, decidual NK cells may maintain the production of these cytokines in vitro, thus up-regulating HLA-G expression of invading trophoblast-like cells. Also progesterone has been shown to have an up-regulating effect on HLA-G gene and protein expression in first trimester cytotrophoblasts and JEG-3 choriocarcinoma cells (Yie et al., 2006). However, the AC-1M59 cells at the periphery of the co-culture, where the cells were in direct contact with the progesterone-containing medium, were less intensively stained for HLA-G than the cells which were lying inside the decidual tissue. Therefore, despite progesterone being included in the tissue culture medium during the invasion experiment, the up-regulation of HLA-G of the AC-1M59 cells at the invasion front was more likely due to factors derived from the decidua than to progesterone in the culture medium.
Apoptosis at the feto–maternal interface has already been studied in detail leading to contradictory results on the type of apoptotic cells. Several authors have reported on apoptosis of trophoblast cells in normal pregnancies (Chan et al., 1999; Von Rango et al., 2003; Pongcharoen et al., 2004; Qiu et al., 2005; Huppertz et al., 2006). Our immunohistochemical staining of sections from both confrontation cultures and from the decidua basalis at the feto–maternal interface demonstrate a close contact of invasive trophoblastic cells with decidual leukocytes. However, that interaction did not result in apoptosis of trophoblast-derived cells. In another contact-independent co-culture system in which first trimester villous explants were cultivated in the presence of free decidual NK cells, there was also no induction of apoptosis of the trophoblast cells not even by the IL-15-stimulated NK cells (Hu et al., 2006). It was rather the leukocyte population that exhibited the fragmented DNA. This makes sense, keeping in mind that extravillous trophoblast cells in the decidua express the Fas ligand and that also activated NK cells were found in this area (Hammer and Dohr, 1999). On activation, NK cells express the Fas receptor, which makes them prone to apoptosis induced by the Fas/Fas ligand system. Indeed, an induction of apoptosis of CD56+ NK cells upon an interaction with HLA-G expressing trophoblast cells has already been reported previously (Emmer et al., 2002). Yet our data are in contradiction to the results of Von Rango et al. (2003) and Pongcharoen et al. (2004) who did not detect apoptotic CD45+ leukocytes in decidua parietalis or in decidua basalis. This may be due in part to different methods for the detection of apoptosis in the tissues. We performed the TUNEL assay on cryostat sections, whereas in most other studies paraffin embedded sections were used. In general, TUNEL assays on paraffin sections may cause problems due to the digestion with trypsin or heating necessary for the test. When performed not intensively or not long enough, false negative results may be possible, whereas treatment of the sections for too long may cause DNA fragmentation resulting in false positive results.
In conclusion, our data demonstrate that certain important processes considered to be implicated in trophoblast invasion in vivo, e.g. the expression of HLA-G and apoptosis, can also be imitated in an appropriate three-dimensional in vitro model. For certain experimental studies of the regulation of trophoblastic invasion, the permanently growing AC-1M59 trophoblast/choriocarcinoma hybrid cells represent an acceptable model of the normal invasive trophoblast, despite containing a tumor component. Both in vitro and in vivo decidual NK cells interact with HLA-G+ trophoblastic cells which may result in apoptosis of several leukocytes. This system can be used for studying the effects of cytokines and growth factors or oxidative stress in an experimental system which closely resembles the in vivo conditions during placentation. However, the limitation of this model is due to the use of a trophoblastic cell line instead of primary extravillous trophoblast. Another more efficient model, the co-cultivation of decidual tissue fragments with first trimester villous explants, may enable an assessment of the interactions between extravillous trophoblast and decidua pairs from concordant pregnancies, which is also our future perspective. Nevertheless, the present system may help to find strategies for the management of diseases associated with deficient trophoblast invasion.
|
Funding |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionFundingAcknowledgementsReferences |
|---|
The study was supported by the Kamillo Eisner-Foundation, Hergiswil, Switzerland and by the Franz Lanyar-Foundation, Graz, Austria.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionFundingAcknowledgementsReferences |
|---|
We would like to thank Mrs Christine Daxböck and Mrs Gabriele Hagendorfer for their excellent technical assistance, Mrs Astrid Blaschitz for numerous valuable hints on immunohistochemical double-staining techniques and Mr Helmut Kolaric for the layout of the figures. We are most grateful to Dr Wolfgang Walcher, Department for Obstetrics and Gynecology, Medical University of Graz, for providing the first trimester decidual tissue.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionFundingAcknowledgementsReferences |
|---|
Barrier BF, Kendall BS, Ryan CE, Sharpe-Thimms KL. HLA-G is expressed by the glandular epithelium of peritoneal endometriosis but not in eutopic endometrium. Hum Reprod (2006) 21:864–869.
Blaschitz A, Juch H, Volz A, Hutter H, Daxböck C, Desoye G, Dohr G. The soluble pool of HLA-G produced by human trophoblasts does not include detectable levels of the intron 4-containing HLA-G5 and HLA-G6 isoforms. Mol Hum Reprod (2005) 11:699–710.
Cartwright JE, Kenny LC, Dash PR, Crocker IP, Aplin PN, Baker PN, Whitley GS. Trophoblast invasion of spiral arteries: a novel in vitro model. Placenta (2002) 23:232–235.[CrossRef][Web of Science][Medline]
Carver J, Martin K, Spyropoulou I, Barlow D, Sargent I, Mardon H. An in vitro model for stromal invasion during implantation of the human blastocyst. Hum Reprod (2003) 18:283–290.
Chan CC, Lao TT, Cheung AN. Apoptotic and proliferative activities in first trimester placentae. Placenta (1999) 20:223–227.[CrossRef][Web of Science][Medline]
Clark DA. Tolerance signaling molecules. Chem Immunol Allergy (2005) 89:36–48.[Medline]
Dunk C, Petkovic L, Baczyk D, Rossant J, Winterhager E, Lye S. A novel in vitro model of trophoblast-mediated decidual blood vessel remodeling. Lab Invest (2003) 83:1821–1828.[CrossRef][Web of Science][Medline]
Emmer PM, Steegers EAP, Kerstens HMJ, Bulten J, Nelen WL, Joosten I. Altered phenotype of HLA-G expressing trophoblast and decidual natural killer cells in pathological pregnancies. Hum Reprod (2002) 17:1072–1080.
Frank H-G, Gunawan B, Ebeling-Stark I, Schulten HJ, Funayama H, Cremer U, Huppertz G, Gaus G, Kaufmann P, Füzesi L. Cytogenetic and DNA fingerprint characterization of choriocarcinoma cell lines and a trophoblast/choriocarcinoma cell hybrid. Cancer Genet Cytogenet (2000) 116:16–22.[CrossRef][Web of Science][Medline]
Funayma H, Gaus G, Ebeling I, Takayama M, Füzesi L, Huppertz B, Kaufmann P, Frank H-G. Parent cells for trophoblast hybridization II: AC1 and related trophoblast cell lines, a family of HGPRT-negative mutants of the choriocarcinoma cell line Jeg-3. Trophoblast Res (1997) 10:191–201.
Gaus G, Funayama H, Huppertz B, Kaufmann P, Frank H-G. Parent cells for trophoblast hybridization I: Isolation of extravillous trophoblast cells from human term chorion laeve. Trophoblast Res (1997) 10:181–190.
Goldman-Wohl DS, Ariel I, Greenfield C, Hochner-Celniker D, Cross J, Fisher S, Yagel S. Lack of human leukocyte antigen-G expression in extravillous trophoblasts is associated with pre-eclampsia. Mol Hum Reprod (2000) 6:88–95.
Grümmer R, Hohn H-P, Mareel MM, Denker HW. Adhesion and invasion of three human choriocarcinoma cell lines into human endometrium in a three-dimensional organ culture system. Placenta (1994) 15:411–429.[Web of Science][Medline]
Hanna J, Goldman-Wohl D, Hamani Y, Avraham I, Greenfield G, Natanson-Yaron S, Prus D, Cohen-Daniel L, Arnon T I, Manaster I, et al. Decidual NK cells regulate key developmental processes at the human fetal-maternal interface. Nat Med (2006) 12:1065–1074.[CrossRef][Web of Science][Medline]
Helige C, Hagendorfer G, Smolle J, Dohr G. Uterine natural killer cells in a three-dimensional tissue culture model to study trophoblast invasion. Lab Invest (2001) 81:1153–1162.[Web of Science][Medline]
Helige C, Hammer A, Dohr G. Trophoblastic invasion in vitro and in vivo: A comparison. Am J Reprod Immunol (2006) 56:44.
Hammer A, Dohr G. Apoptotic nuclei within the uterine decidua of first trimester pregnancy arise from CD45 positive leukocytes. Am J Reprod Immunol (1999) 42:88–94.[Medline]
Hu Y, Dutz JP, MacCalman CD, Yong P, Rusung T, von Dadelszen P. Decidual NK cells alter in vitro first trimester extravillous cytotrophoblast migration: A role for IFN-. J Immunol (2006) 177:8522–8530.
Hunt JS. Strangers in a strange land. Immunol Rev (2006) 213:36–47.[CrossRef][Web of Science][Medline]
Hunt JS, Petroff MG, McIntire RH, Ober C. HLA-G and immune tolerance in pregnancy. FASEB J (2005) 19:681–693.
Hunt JS, Morales PJ, Pace JL, Fazleabas AT, Langat DK. A commentary on gestational programming and functions of HLA-G in pregnancy. Trophoblast Res (2007) 28(Suppl_A):S57–S63. Placenta.
Huppertz B, Kadyrov M, Kingdom JCP. Apoptosis and its role in the trophoblast. Am J Obstet Gynecol (2006) 193:1–11.[CrossRef][Web of Science]
Karmakar S, Das C. Regulation of trophoblast invasion by IL-1beta and TGF-beta1. Am J Reprod Immunol (2002) 48:210–219.[CrossRef][Web of Science][Medline]
King A, Gardner L, Loke YW. Co-stimulation of human decidual natural killer cells by interleukin-2 and stromal cells. Hum Reprod (1999) 14:656–663.
King A, Allan DS, Bowen M, Powis SJ, Joseph S, Verma S, Hiby SE, McMichael AJ, Loke YW, Brand VM. HLA-E is expressed on trophoblast and interacts with CD94/NKG2 receptors on decidual NK cells. Eur J Immunol (2000) 30:1632–1631.[CrossRef][Web of Science][Medline]
Korff T, Krauss T, Augustin H. Three-dimensional spheroidal culture of cytotrophoblast cells mimics the phenotype and differentiation of cytotrophoblasts from normal and preeclamptic pregnancies. Exp Cell Res (2004) 297:415–423.[CrossRef][Web of Science][Medline]
LaMarca HL, Ott CM, Höner zu Bentrup K, LeBlanc CL, Pierson DL, Nelson AB, Scandurro AB, Whitley GS, Nickerson CA, Morris CA. Three-dimensional growth of extravillous cytotrophoblasts promotes differentiation and invasion. Placenta (2005) 26:709–720.[CrossRef][Web of Science][Medline]
Lash GE, Hornbuckle J, Brunt A, Kirkley M, Searle RF, Robson SC, Bulmer JN. Effect of low oxygen concentrations on trophoblast-like cell line invasion. Placenta (2006) 28:390–398.[CrossRef][Web of Science][Medline]
Le Bouteiller P, Solier C, Pröll J, Aguerre-Girr M, Fournel S, Lenfant F. Placental HLA-G protein expression in vivo: where and what for? Hum Reprod Update (1999) 5:223–233.
Lefebvre S, Moreau P, Guiard V, Ibrahim EC, Adrian-Cabestre F, Menier C, Dausset J, Carosella ED, Paul P. Molecular mechanisms controlling constitutive and IFN-gamma-inducible HLA-G expression in various cell types. J Reprod Immunol (1999) 43:213–224.[CrossRef][Web of Science][Medline]
Mallet V, Blaschitz A, Crisa L, Schmitt C, Fournel S, King A, Loke YW, Dohr G, Le Bouteiller P. HLA-G in the human thymus: a subpopulation of medullary epithelial but not CD83+ dendritic cells express HLA-G as a membrane-bound and soluble protein. Int Immunol (1999) 11:889–898.
Moreau P, Adrian-Cabestre F, Menier C, Guiard V, Gourand L, Dausset J, Carosella ED, Paul P. IL-10 selectively induces HLA-G expression in human trophoblasts and monocytes. Int Immunol (1999) 11:803–811.
Pongcharoen S, Bulmer JN, Searle RF. No evidence for apoptosis of decidual leucocytes in normal and molar pregnancy: implications for immune privilege. Clin Exp Immunol (2004) 138:330–336.[CrossRef][Web of Science][Medline]
Ponte M, Cantoni C, Biassoni R, Tradori-Cappai A, Bentivoglio G, Vitale S, Bertone A, Moretta A, Moretta L, Mingari MC. Inhibitory receptors sensing HLA-G1 molecules in pregnancy: Decidua-associated natural killer cells express LIR-1 and CD94/NKG2A and acquire p49, an HLA-G1-specific receptor. Proc Natl Acad Sci USA (1999) 96:5674–5679.
Qiu Q, Yang M, Tsang BK, Gruslin A. Fas ligand expression by maternal decidual cells is negatively correlated with the abundance of leukocytes present at the maternal-fetal interface. J Reprod Immunol (2005) 65:121–132.[CrossRef][Web of Science][Medline]
Roth I, Corry DB, Locksley RM, Abrams JS, Litton MJ, Fisher SJ. Human placental cytotrophoblasts produce the immunosuppressive cytokine interleukin 10. J Exp Med (1996) 184:539–548.
Schmitz U. Subtraktiver Vergleich der Differenzierungsmuster menschlicher invasiver Trophoblastzellen und maligner Choriokarzinomzellen mit Hilfe von Choriokarzinom/Trophoblast-Hybriden. (2002).
Trew AJ, Lash GE, Baker PN. Investigation of an in vitro model of trophoblast invasion. Early Pregnancy (2000) 4:176–190.[Medline]
Trundley A, Moffet A. Human uterine leukocytes and pregnancy. Tissue Antigens (2004) 63:1–12.[CrossRef][Web of Science][Medline]
Vicovac L, Papic N, Aplin J. Tissue interactions in first trimester trophoblast-decidua co-cultures. Trophoblast Res (1993) 7:223–236.
Vigano P, Gaffuri B, Somigliana E, Infantino M, Vignali M, BiBlasio AM. Interleukin-10 is produced by human uterine natural killer cells but does not affect their production of interferon-gamma. Mol Hum Reprod (2001) 7:971–977.
Von Rango U, Krusche CA, Kertschanska S, Alfer J, Kaufmann P, Beier HM. Apoptosis of extravillous trophoblast cells limits the trophoblast invasion in uterine but not in tubal pregnancy during first trimester. Placenta (2003) 24:929–940.[CrossRef][Web of Science][Medline]
Yang Y, Chu W, Geraghty DE, Hunt JS. Expression of HLA-G in human mononuclear phagocytes and selective induction by interferon-gamma. J Immunol (1996) 156:4224–4231.[Abstract]
Yie SM, Li LH, Xiao R, Librach CL. Progesterone enhances HLA-G gene expression in JEG-3 choriocarcinoma cells and human cytotrophoblast cells in vitro. Hum Reprod (2006) 21:46–51.
Zdravkovic M, Aboagye-Mathiesen G, Giumond MJ, Hager H, Ebbesen P, Lala PK. Susceptibility of MHC class I expressing extravillous trophoblast cell lines to killing by natural killer cells. Placenta (1999) 20:431–440.[CrossRef][Web of Science][Medline]
Submitted on January 14, 2008; resubmitted on April 23, 2008; accepted on April 28, 2008.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||