Development of endometriosis-like lesions after transplantation of human endometrial fragments onto the chick embryo chorioallantoic membrane

Jacques W.M. Maas1,4, Patrick G. Groothuis2, Gerard A.J. Dunselman1, Anton F.P.M. de Goeij2, Harry A.J. Struijker-Boudier3 and Johannes L.H. Evers1

1 Department of Obstetrics and Gynaecology, 2 Department of Pathology, Research Institute Growth and Development (GROW) and 3 Department of Pharmacology, Cardiovascular Research Institute Maastricht (CARIM), Maastricht University, Maastricht, The Netherlands


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
The chick embryo chorioallantoic membrane (CAM) bioassay was used to investigate the early pathogenesis of endometriosis. Endometrial fragments were explanted onto the CAM. The grafts including the surrounding CAM were excised at 24, 48 or 72 h after explantation, fixed and embedded in paraffin. Immunohistochemical analysis was used to distinguish endometrial cells. To identify cells of human origin, in-situ hybridization was performed using a probe specific for human chromosome 1. After 24 h, direct contact between endometrial stromal as well as epithelial cells and the mesenchymal layer of the CAM was observed. Invasion of both stromal cells and intact endometrial glands into the mesenchymal layer was observed after 48 h. At 72 h, endometriosis-like lesions were observed in the mesenchymal layer. Positive staining with antibodies to vimentin and pan-cytokeratin was observed in the invading cells as well as in the lesions. In the lesions these positively stained cells showed in-situ hybridization signals for human chromosome 1, confirming their human origin. In conclusion, after 3 days of incubation, endometriosis-like lesions consisting of human endometrial glands and stromal cells were found in the mesenchymal layer of the CAM. These lesions apparently resulted from the invasion of intact human epithelial structures and stromal cells.

Key words: chorioallantoic membrane/endometriosis/endometrium/in-situ hybridization/invasion


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Endometriosis is the presence of functional endometrial tissue in ectopic locations outside the uterine cavity. Various theories on the pathogenesis of endometriosis exist. The transplantation or implantation theory described by Sampson remains the most widely accepted (Sampson, 1927Go). It presumes retrograde transportation of viable endometrial cells during menstruation, adhesion of these cells onto the peritoneum, with subsequent implantation and proliferation. The presence of viable endometrial cells has been demonstrated in peritoneal fluid (Kruitwagen et al., 1991Go) and the results of our earlier immunohistochemical studies support the contention of transport of menstrual detritus to the peritoneal cavity in women with patent Fallopian tubes (van der Linden et al., 1995Go). Using in-vitro models, the adhesion of endometrial tissue to amnion or peritoneum has been studied. The results strongly suggested that adhesion of endometrial fragments is prevented by intact mesothelium (Groothuis et al., 1999Go; Koks et al., 1999Go). It is still not clear what occurs between the moment of adhesion of endometrial cells to the peritoneum and the existence of endometriotic lesions. These early steps of the pathogenesis of endometriosis are difficult to study in in-vivo models, since the injected or implanted cells cannot be directly observed. The chick embryo chorioallantoic membrane (CAM) model has been introduced for analysis of invasiveness of malignant cells prior to tumour formation (Leighton, 1964Go; Scher et al., 1976Go). The aims of this study were (i) to investigate whether, in analogy to tumour formation, endometriosis-like lesions can be formed in this CAM model after grafting of fresh human endometrial fragments, (ii) to characterize these lesions and (iii) to study the steps prior to lesion development.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Tissue
A Probet® endometrial sampling device (Gynetics, Oisterwijk, Netherlands) was used to collect samples of endometrial tissue from 15 patients with regular menstrual cycles undergoing laparoscopic surgery during fertility investigations. According to the number of days since the last menstrual period, endometrium was obtained at the following cycle days: 5, 6, 7 (three patients), 8, 10 (two patients), 11 (two patients), 12, 15, 21, 22 and 23. During laparoscopy, some endometriotic lesions were observed in three patients. The use of human tissue for this study was approved by the institutional review board of the Maastricht University Hospital and all women included in this study signed a written informed consent. After collection, the endometrium was placed in sterile saline and stripped of blood with a fine forceps. Subsequently, the endometrial tissue was carefully sectioned into uniform fragments of 1.5x2 mm with the aid of a dissecting microscope.

CAM model
Fertile eggs of Lohman-selected White Leghorns were incubated for 3 days at 37°C, 55% relative air humidity, while being rotated hourly. At day 3 of incubation, a rectangular window (1 cmx1.5 cm) was made in the eggshell. Two ml albumen were withdrawn, using a 21 G needle, through the large blunt edge of the egg. The window was covered with scotch tape to prevent dehydration. The eggs were replaced in the incubator without rotation until day 10 of incubation.

The CAM is an impenetrable barrier to invasive cells unless it has first been traumatized by removing the upper peridermal part of the double epithelial layer, leaving the basal cell layer intact. Therefore, just before explantation a small portion of the CAM was gently traumatized by laying a 1 cm wide strip of sterile ether-extracted lens tissue onto the surface of the epithelium and then removing it immediately (Armstrong et al., 1982Go). Subsequently, the dissected fragments were explanted onto the CAM within 2 h of obtaining the tissue. On each CAM only one fragment was layered. Following the explantation the window was covered again and the egg was placed back in the incubator. At 24, 48 or 72 h after explantation the graft including the surrounding CAM was excised, fixed in 3.7% buffered formalin and embedded in paraffin. Paraffin sections (4 µm) were cut and either stained with haematoxylin and eosin (HE) for histological evaluation or stored for later immunohistochemical and in-situ hybridization analysis.

Immunohistochemistry
Paraffin sections were deparaffinized by incubation with xylene for 2x5 min and rehydrated in a graded alcohol series. Endogenous peroxidase activity was blocked by incubation with 0.3% hydrogen peroxide in methanol for 20 min. The sections were rinsed in phosphate buffered saline (PBS) and subsequently digested in 0.1% pepsin in 0.1 mol/l HCl for 30 min. After rinsing again in PBS, sections were incubated overnight at 4°C with a 1:100 dilution of the primary antibody to vimentin (Organon Teknika, Boxtel, Netherlands) or to pan-cytokeratin (Clone MNF 116; DAKO, Glostrup, Denmark). After a PBS rinse, sections were exposed to biotinylated rabbit anti-mouse secondary antibody for 1 h, rinsed and further exposed to avidin-biotinylated peroxidase complex for 30 min. Antibody binding was visualized using 3'-3-diaminobenzidine. Sections were washed and counterstained with haematoxylin, washed, dehydrated and mounted for light microscopy.

In-situ hybridization
The in-situ hybridization procedure was performed as described in detail by Jansen and coworkers (1998). In short, sections were pretreated with 1 mol/l sodium thiocyanate at 80°C for 10 min and digested with 4 mg/ml pepsin in 0.2 mol/l HCl at 37°C. The digoxigenin-labelled DNA probe, used for the in-situ hybridization analysis, hybridizes to the (peri)centromeric regions of human chromosome 1 (pUC1.77, 1.77 kb), but not to chicken chromosome 1. The probe was dissolved in hybridization buffer, containing 60% formamide, 2xsaline sodium citrate (SSC) pH 7.0, 10% dextran sulphate, 0.2 mg/ml herring sperm DNA and yeast tRNA as carrier DNA and RNA respectively. The final probe concentration was 0.4 ng/ml. Ten µl of hybridization mixture were added to the slides under a coverslip. Samples were then denaturated in a moist chamber at 75°C for 3 min. Hybridization was allowed overnight at 37°C. Samples were washed twice for 5 min in 2xSSC/0.05% Tween20 buffer pH 7.0, at 42°C, twice for 5 min in 0.01xSSC/0.05% Tween20 buffer pH 7.0, at 60°C, and once in 4xSSC/0.05% Tween20 buffer for 5 min at room temperature. The digoxigenin-labelled probe was detected in subsequent incubations with mouse anti-digoxine immunoglobulin (Ig)G (1:2000; Sigma, St Louis, MO, USA), horseradish peroxidase conjugated rabbit anti-mouse IgG (1:80; DAKO), and finally with tetramethylrhodamine isothiocyanate-labelled tyramide, diluted 1:100 in PBS/imidazol containing 0.001% hydrogen peroxide. Nuclei were counterstained with 4-,6-diaminodino-2-phenylindole (DAPI; Sigma; 1.25 ng/ml), diluted in glycerol containing 2.3% 1,4-diazobicyclo-(2.2.2)-octane (DABCO; Sigma). Nuclei were examined with a Leica microscope (Leica, Cambridge, Cambs, UK) equipped for fluorescence with DAPI and rhodamine filter sets.

Scanning electron microscopy
Membranes were processed for electron microscopic analysis to check if the basal epithelial cell layer was still intact after removing the peridermal layer. Membranes were fixed in 2.5% glutaraldehyde in phosphate buffer (pH 7.4). The membranes were dehydrated in alcohol. The samples were critical-point dried, sputtered with gold, and studied under a Philips 505 scanning electron microscope (Philips, Eindhoven, The Netherlands).


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
The CAM consists of an outer chorionic epithelium (CE) of ectodermal cells, a mesodermal layer consisting of loose connective tissue and blood vessels and an inner allantoic epithelium (AE) of endodermal cells. The CE is a continuous double layer of cells with a superficial sheet of flat peridermal cells and a basal cell layer, which is an impenetrable barrier to invasive cells. Therefore the upper peridermal part was removed, leaving the basal epithelial cell layer intact (Figure 1Go). Endometrial fragments, consisting of glands and stroma, were grafted onto the CAM (Figure 2Go). Already after 24 h, endometrial stroma and occasionally intact glands were found in direct contact with the mesenchymal layer of the CAM (Figure 3A, BGo). After 48 h both stromal cells and intact epithelial glands surrounded by stromal cells invaded across the epithelial layer into the mesenchymal layer (Figure 3C, DGo). Invasion of single cytokeratin positive cells was not observed. After 3 days of incubation, endometriosis-like lesions, which consisted of dilated glands and stromal cells in the mesenchymal layer at a distance from the original graft (Figure 3EGo), were observed. In each distant lesion at least one dilated gland was observed. The vital lesions appeared to receive their blood supply from CAM vessels, which contained nucleated chick erythrocytes. This is illustrated in Figure 3FGo, which shows a branch from a CAM vessel penetrating the endometriosis-like lesion. Furthermore, cross-sections of CAM vessels can be seen inside the lesion.



View larger version (150K):
[in this window]
[in a new window]
 
Figure 1. Scanning electron micrograph of the partly traumatized chorionic epithelium of the chick embryo chorioallantoic membrane (CAM): the superficial sheet of flat peridermal cells (PD) has been removed exposing the basal epithelial cell layer (BL) (bar = 11 µm).

 


View larger version (143K):
[in this window]
[in a new window]
 
Figure 2. Light micrograph of a cross-section of the CAM containing an endometrial fragment (E) of cycle day 7, that consists of glands (G) and stromal cells (S). CE = chorionic epithelium; M = mesenchymal layer; AE = allantoic epithelium (cytokeratin; bar = 160 µm).

 


View larger version (196K):
[in this window]
[in a new window]
 
Figure 3. Development of endometriosis-like lesion after grafting of an endometrial fragment onto the CAM. (A) Twenty-four hours after grafting: direct contact endometrial glands and stromal cells with mesenchymal layer [haematoxylin and eosin (HE); bar = 55 µm]. (B) Staining of endometrial gland, which is in direct contact with mesenchymal layer, with pan-cytokeratin antibody (bar = 40 µm). (C) Forty-eight hours after grafting: invasion of endometrial glands and stromal cells across the epithelial layer into the mesenchymal layer (HE; bar = 80 µm). (D) Invasion of vimentin positive stromal cells through the opening in the chorionic epithelium into the mesenchymal layer (vimentin; bar = 80 µm). (E) Seventy-two hours after grafting: lesion in mesenchymal layer of CAM, containing glands and stromal cells (HE; bar = 80 µm). (F) Branch of CAM vessel penetrates the lesion, and in the lesion cross-section of CAM vessels can be observed (HE; bar = 80 µm). CE = chorionic epithelium; E = endometrial fragment; G = endometrial gland; M = mesenchymal layer; S = stromal cells; V = CAM vessel.

 
Figure 4Go presents the histochemical and in-situ hybridization analysis of the endometriosis-like lesions observed 72 h after grafting. The centre of the lesions consists of vimentin positive cells (Figure 4BGo). Furthermore, intense staining of the gland-like structures with human pan-cytokeratin antibody was observed (Figure 4CGo). In both vimentin positive and cytokeratin positive cells in-situ hybridization signals were found for human chromosome 1 (Figure 4DGo); no signals were observed in the rest of the CAM.



View larger version (103K):
[in this window]
[in a new window]
 
Figure 4. (A) Endometriosis-like lesion in mesenchymal layer (M) of the CAM at a distance from the original endometrial graft (E), containing dilated glands (G) and stromal cells (S). CE = chorionic epithelium; V = CAM vessel (haematoxylin and eosin; bar = 275 µm). (B) Staining of stromal cells with vimentin antibody (bar = 80 µm). (C) Staining of glandular epithelial cells with pan-cytokeratin antibody (bar = 80 µm). (D) Fluoresence in-situ hybridization using probe for human chromosome 1 (red); signal is visible in nuclei of both stromal and epithelial cells (bar = 80 µm).

 
Endometrial fragments were explanted onto 59 CAMs. By means of extensive observations both at regular time intervals and in series of parallel sections throughout the site of explantation, endometriosis-like lesions could be found after grafting of endometrium obtained from nine out of 15 women. Although lesions could not be found after grafting of endometrial fragments at several CAMs from six women, invasion of endometrial cells obtained from two of these women into the mesenchymal layer was observed. Therefore, endometrial fragments from only four (cycle day 5, 10, 12 and 22) out of 15 women showed neither invasion nor formation of a distant lesion after grafting onto the CAM. During laparoscopy, some endometriotic lesions were observed in three patients. The endometrial cells obtained from these three patients were invasive in this CAM model, and endometriosis-like lesions were found after grafting of endometrium of two of these women.


    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
The chick embryo CAM has been used to study tumour transplantation and invasion since the first description of successful transplantation of heterologous tumours to the CAM was published (Murphy, 1912Go). With the aid of scanning electron microscopy it became clear that the CAM is an impenetrable barrier to neoplastic cells unless it has first been traumatized by removing the upper peridermal part of the double layer of chorionic epithelium (McCormick et al., 1984Go). The basal cell layer of this epithelium is resistant to mechanical damage and can be used to distinguish invasive from non-invasive cells (Schroyens et al., 1989Go). Using this CAM model, it was demonstrated in the present study that both endometrial stromal and epithelial cells are invasive. Invasion of single cytokeratin positive epithelial cells could not be observed, only invasion of intact endometrial glands has been seen. These invading glands were surrounded by vimentin positive cells, indicating that these cells were either endometrial stromal cells or fibroblasts. Using a collagen gel invasion assay to assess the invasive potential of endometrial cells, Gaetje and coworkers (1997) concluded that only vimentin positive cells were invasive in vitro and not epithelial cells (Gaetje et al., 1997Go). An explanation for the discrepancy between their findings and those presented here may be that in their study endometrial biopsies were digested and the cells were cultured. They plated single cells onto the collagen gels, excluding the possibility of studying the invasive potential of intact fragments consisting of stroma and glands. It appears that for epithelial cells the mutual contact, leaving the glandular structure intact, as well as contact with surrounding stromal cells is a prerequisite for invasive potential.

Endometriosis-like lesions could already be found in the mesenchymal layer 72 h after grafting of endometrial fragments onto the CAM. These lesions in the chick embryo CAM consist of both endometrial stromal and epithelial cells, as was demonstrated by the positive staining for vimentin and cytokeratin respectively. In addition, some staining of CAM epithelium as well as some mesenchymal cells was seen occasionally. Therefore, fluorescence in-situ hybridization analysis using a probe specific for human chromosome 1 was applied to make a positive distinction between chick embryo and human cells. It was demonstrated clearly that the lesions in the chick embryo CAM were of human origin. Occasionally vessels, containing nucleated chick erythrocytes, could be observed in the lesions. Apparently, the CAM blood vessels branched and penetrated the endometriosis-like lesions.

Recently, Ohtake and coworkers (1999) concluded that endometriotic lesions arise through a process of metaplasia from ovarian surface epithelial cells. Their study was based on a previously published induction theory (Levander and Normann, 1955Go) which was then further developed (Merrill, 1966Go). According to this theory, the endometrial fragments should produce or release factors, which activate the chick embryo chorionic epithelium to form glandular structures in the mesenchymal layer. In the current study it was demonstrated that the epithelial lining of the glands in the endometriosis-like lesions is of human origin, which is not consistent with the induction theory.

In the current study lesions and site of invasion could not always be found. This either means that an appropriate section of the very small lesion was not obtained or that the endometrial cells were not invasive. Furthermore it has already been described that results can be varied depending on the tumour used and, even within the same experiment, divergent results were sometimes observed (Leighton, 1964Go).

Fragments used in this study may not be truly representative of the in-vivo situation. Therefore, it may also be relevant to use endometrial tissue isolated from menstrual effluent or cell suspensions. Based on the findings of the present study, it is postulated that the presence of intact glandular structures with stromal components is required for the invasive potential of endometrial cells and the formation of endometriosis-like lesions.

In conclusion, 3 days after grafting of endometrial fragments, endometriosis-like lesions, consisting of human endometrial glands and stromal cells, were detected in the mesenchymal layer of the CAM. These lesions resulted from the invasion of intact human epithelial structures surrounded by human endometrial stromal cells.


    Acknowledgements
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
The authors thank Lilian Kessels from the Department of Pharmacology for excellent technical assistance and Professor Dr F. Ramaekers and Monique Ummelen from the Department of Molecular Cell Biology and Genetics, Maastricht University, for the hybridization procedure. Part of this investigation was supported by an unrestricted scientific grant of the Searle-Monsanto company b.v., Maarssen, The Netherlands.


    Notes
 
4 To whom correspondence should be addressed at: PO Box 5800, 6202 AZ Maastricht, The Netherlands. E-mail: jmaa{at}sgyn.azm.nl Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Armstrong, P.B., Quigley, J.P. and Sidebottom, E. (1982) Transepithelial invasion and intramesenchymal infiltration of the chick embryo chorioallantois by tumour cell lines. Cancer Res., 42, 1826–1837.[Abstract]

Gaetje, R., Kotzian, S., Herrmann, G. et al. (1997) Nonmalignant epithelial cells, potentially invasive in human endometriosis, lack the tumour suppressor molecule E-cadherin. Am. J. Pathol., 150, 461–467.[Abstract]

Groothuis, P.G., Koks, C.A.M., de Goeij, A.F.P.M. et al. (1999) Adhesion of human endometrial fragments to peritoneum in vitro. Fertil. Steril., 71, 1119–1124.[ISI][Medline]

Jansen, M.P.H.M., Hopman, A.H.N., Haesevoets, A.M. et al. (1998) Chromosome abnormalities in Hodgkin's disease are not restricted to Hodgkin/Reed–Sternberg cells. J. Pathol., 185, 145–152.[ISI][Medline]

Koks, C.A.M., Groothuis, P.G., Dunselman, G.A.J. et al. (1999) Adhesion of shed menstrual tissue in an in-vitro model using amnion and peritoneum: a light and electron microscopic study. Hum. Reprod., 14, 816–822.[Abstract/Free Full Text]

Kruitwagen, R.F.P.M., Poels, L.G., Willemsen, W.N.P., et al. (1991) Endometrial epithelial cells in peritoneal fluid during the early follicular phase. Fertil. Steril., 55, 297–303.[ISI][Medline]

Leighton, J. (1964) Invasion and metastasis of heterologous tumours in the chick embryo. Progr. Exp. Tumor Res., 4, 98–125.[ISI]

Levander, G. and Normann, P. (1955) The pathogenesis of endometriosis. Acta Obstet. Gynecol. Scand., 34, 366–398.

McCormick, J.F., Nassauer, J., Bielunas, J. et al. (1984) Anatomy of the chick chorioallantoic membrane relevant to its use as a substrate in bioassay systems. Scan. Electron Microsc., 4, 2023–2030.

Merrill, J.A. (1966) Endometrial induction of endometriosis across Millipore filters. Am. J. Obstet. Gynecol., 94, 780–790.[ISI][Medline]

Murphy, J.B. (1912) Transplantability of malignant tumours to the embryos of foreign species. J. Amer. Med. Assoc., 59, 874–875.

Ohtake, H., Katabuchi, H., Matsuura, K. et al. (1999) A novel in vitro experimental model for ovarian endometriosis: the three-dimensional culture of human ovarian surface epithelial cells in collagen gels. Fertil. Steril., 71, 50–55.[ISI][Medline]

Sampson, J.A. (1927) Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am. J. Obstet. Gynecol., 14, 422–469.

Scher, C.D., Haudenschild, C. and Klagsbrun, M. (1976) The chick chorioallantoic membrane as a model system for the study of tissue invasion by viral transformed cells. Cell, 8, 373–382.[ISI][Medline]

Schroyens, W., Schroyens, E. and Bielunas, J. (1989) Different invasion capacity of NBTII and MDCK in the chick embryo chorioallantois. Anticancer Res., 9, 1665–1668.[ISI][Medline]

Van der Linden, P.J.Q., Dunselman, G.A.J., de Goeij, A.F.P.M. et al. (1995) Epithelial cells in peritoneal fluid: of endometrial origin? Am. J. Obstet. Gynecol., 173, 566–570.[ISI][Medline]

Submitted on August 29, 2000; accepted on January 2, 2001.