Hum. Reprod. Advance Access originally published online on October 2, 2006
Human Reproduction 2006 21(12):3068-3080; doi:10.1093/humrep/del310
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Morphological and glycosylation changes associated with the endometrium and ectopic lesions in a baboon model of endometriosis
1 Academic Unit of Obstetrics and Gynaecology, Division of Human Development, University of Manchester, St Mary’s Hospital 2 Division of Laboratory and Regenerative Medicine, University of Manchester, Manchester, UK and 3 Department of Obstetrics and Gynecology, University of Illinois, Chicago, IL, USA
4 To whom correspondence should be addressed at: Room 80 Research Floor, Department of Obstetrics and Gynaecology, University of Manchester, St Mary’s Hospital, Hathersage Road, Manchester M13 0JH, UK. E-mail: carolyn.jones{at}manchester.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
BACKGROUND: Endometriosis is one of the most common causes of infertility and pelvic pain. A baboon model has recently been developed whereby the intrapelvic injection of menstrual endometrium results in the induction of endometriotic lesions. We have used this model to investigate changes in ultrastructure and glycosylation of endometria from normal and diseased baboons. METHODS: Endometriosis was induced in eight female baboons; endometrial tissue and endometriotic lesions were removed on days 9–11 post ovulation between 3 and 16 months of disease and compared with endometrium from 17 control animals, using electron microscopy and lectin histochemistry. RESULTS: Ultrastructurally, diseased endometrial glands showed abnormalities in secretory vacuoles and an intracellular accumulation of glycogen; in later stages of the disease, glands resembled those of the late secretory phase endometrium. The abnormalities were mirrored by changes in glycan expression. In early disease, there was an increased binding of lectin from Dolichos biflorus agglutinin (DBA) to fucosylated N-acetylglucosamine residues, whereas in later stages, this binding generally decreased in association with the appearance of a late secretory phenotype. CONCLUSIONS: Endometriosis is accompanied by progressive changes in the gland architecture and biochemistry resulting in dyssynchrony within the window of uterine receptivity, which may result in the reduced fertility associated with this disease.
Key words: baboon/endometriosis/ultrastructure/lectin histochemistry
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Endometriosis is defined as the presence of ectopic foci of endometrial glands and stroma which tend to respond to ovarian hormones in a similar way to the uterine mucosa (Czernobilsky and Fox, 2003
). As such, it is one of the most common causes of infertility, dysmenorrhea and pelvic pain, affecting 10% of women during their reproductive years (Eskenazi and Warner, 1997). A theory has been advanced (Sampson, 1927) that endometriotic foci develop from the regurgitation of endometrial fragments through the oviduct as a result of retrograde menstruation. However, the mechanism by which these fragments adhere to and proliferate on an ectopic site is still unknown.
Because it is not possible to study these mechanisms in affected women in vivo, a baboon model has been developed in which the intrapelvic injection of menstrual endometrium results in the induction of endometriotic lesions (D’Hooghe et al., 1995a; D’Hooghe, 1997; Fazleabas et al., 2002). This baboon model has, to date, been used to investigate the effects of endometriosis on fertility (D’Hooghe et al., 1996a), the role of immunosuppression on the development and progression of the disease (D’Hooghe et al., 1995b) and changes in the leukocyte population (D’Hooghe et al., 1996b), steroid receptor and aromatase gene expression (Fazleabas et al., 2003) and endometrial gene expression (Gashaw et al., 2006; Hastings et al., 2006); however, little is known about possible alterations in tissue ultrastructure and glycan expression. In this study, we have examined the ultrastructure of both eutopic and endometriotic tissues in the baboon model of endometriosis, as well as the expression of various glycans including that bound by the lectin from Dolichos biflorus (GalNAc
1,3(LFuc1,2)Galß1, 3/4GlcNAcß1-), which has been shown to change over the course of the menstrual cycle in baboons (Jones et al., 1998). It is strongly expressed in the proliferative phase, gradually declining from about day 10 post ovulation onwards. Tissues at days 9–11 post ovulation have been collected from baboons with induced endometriosis over a time period ranging from 3 to 16 months’ duration of disease and compared with normal endometrium from control animals using both electron microscopy and lectin histochemistry.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Endometriosis was induced in eight female baboons as previously described (Fazleabas et al., 2002
, 2003) by the i.p. inoculation of menstrual endometrium on two consecutive menstrual cycles. Menstrual tissue was obtained by means of a Unimar pipelle (Cooper Surgical Inc., Shelton, CT, USA) and deposited into the Pouch of Douglas as well as the broad ligaments near the Fallopian tubes. Laparotomies were performed at various times (3, 6, 7, 9, 12–13, 15–16 months) after the second inoculation to obtain endometrium and endometriotic lesions. Control tissues were obtained from 17 baboons that had not been inoculated with menstrual tissue. The Animal Care Committee of the University of Illinois, Chicago, approved all experimental procedures on baboons. See Tables I–III for details of the animals, their identifiers and tissues.
|
|
|
After laparotomy, we fixed the tissues in 3% (w/v) paraformaldehyde/1% (v/v) glutaraldehyde for 6 h at room temperature. Some samples were bisected, and half the tissue was subsequently post-fixed in 1% osmium tetroxide (v/v)/1.5% potassium ferrocyanide (w/v) in distilled water for 1 h at room temperature, and all tissues were processed into Araldite epoxy resin (Ladd Research Industries, Burlington, VT, USA). Semithin sections 0.5-µm thick from both osmicated and non-osmicated tissues were cut on a Reichert Ultracut ultramicrotome and stained with 1% (w/v) toluidine blue in 1% (w/v) borax for evaluation. Only those tissues containing glandular elements and, in the case of endometriotic specimens, epithelial cystic structures were selected for further investigation (Tables I–
III). For ultrastructural examination, pale gold (70–90 nm) sections of selected areas of osmicated tissues were cut with a diamond knife, mounted on 200-mesh copper grids and contrasted with uranyl acetate/lead citrate before examination with a Philips 301 or CM10 electron microscope. For lectin histochemistry, sections 0.75-µm thick were cut and mounted on aminopropyltriethoxysilane (APES)-coated slides (Maddox and Jenkins, 1987) and stained with a panel of lectins (Table IV) as previously described (Jones et al., 1998). Some sections were also pretreated with 0.1 units/ml neuraminidase (Type VI from Clostridium perfringens, Sigma) for 2 h at 37°C, to cleave off terminal sialic acid, before treatment with the lectins (Jones et al., 1992). Other controls were carried out by incubating the lectin in the presence of a blocking sugar: 2 M N-acetylgalactosamine in the case of D. biflorus agglutinin (DBA) and 2 M N-acetylglucosamine for Phytolacca americana agglutinin (PAA) and Triticum vulgaris agglutinin (Wheatgerm, WGA). Non-specific binding of avidin-peroxidase was controlled by staining the sections with 0.05 M Tris-buffered saline (TBS) substituted for the lectin.
|
Image analysis
Sections stained with DBA were analysed with the Leica Qwin image analysis system to determine whether the differences observed between tissue from animals with endometriosis and control animals were of statistical significance. Images were captured using a Leica DMRB microscope linked to an Infinity X digital camera. Using a x10 objective, images were captured as TIFF files at a resolution of 1024 x 768. The lumen was manually delineated and the area subsequently subtracted from the manually outlined area of the whole gland. The threshold for colour segmentation was set at a predetermined staining density. The area of the thresholded component compared with the area of the whole gland (minus the lumen) was calculated electronically through the Quantimet (Leica, UK) system, and the resulting areas of the whole gland minus lumen, and that occupied by the thresholded positively stained area, were sent to a Microscoft Excel spreadsheet. These values were subsequently expressed as a ratio. The number of readings taken depended entirely on the size of the section and the number of gland profiles present and varied from 3 to 10. The population distribution of data from both diseased and control sections was assessed using the one-sample Kolmogorov–Smirnov test and then analysed using Student’s t-test.
Those images of diseased specimens with a distinct late secretory phenotype were excluded from this statistical analysis as they were felt to be a separate population, as were areas of basal glands near the myometrium. Thus, only diseased glands showing a morphology similar to that of typical mid-secretory glands of the functionalis, but with variable intensity of stain, were analysed, to verify whether differences in the stain in diseased specimens with this phenotype were statistically significant. Specimens that showed only a late secretory phenotype comprised specimens of 9, 12, 15 and 16 months’ duration of disease (PA 2733, PA 7015 [12 and 16 months first specimen], PA 7026), and none of their readings were analysed. Parts of the endometrium showed this late secretory phenotype in two cases of 13 and 16 months’ disease (PAN 2654 and PA 7015, second specimen); two of six and three of seven readings were used from these specimens, respectively. Sections that included basal parts of glands as well as those in the functionalis were found in five cases of endometriosis and one control; these parts of the gland were not analysed as their staining properties were entirely different.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Electron microscopy
Control endometrium
The glands of the functionalis region only will be described as these showed most changes. They were composed of columnar cells that bore microvilli and contained arrays of vesicles, most of which had clear contents with an electron-dense core or finely dispersed glycogen particles (Figure 1A and B). Occasionally, some fluffy material was also present. Nuclei were generally regular in outline and mainly euchromatic with one or more nucleoli. Several Golgi bodies and associated vesicles were in a supranuclear position and the saccules mainly narrow, and numerous smooth membranes of endoplasmic reticulum were also seen. In some cells, stacks of rough endoplasmic reticulum were also evident, and cisternae were often in close association with mitochondria, following their contours (Figure 1B). Mitochondria were small and narrow and tended to cluster basally, with clumps of glycogen and occasional secretory droplets. Lateral membranes were regular with few interdigitations (Figure 1B), and junctional complexes were found at the apical border, generally associated with a terminal web of fine filaments. Occasionally, glycogen aggregates were also present in a supranuclear position, but this was generally uncommon and found only in one case.
|
For reference purposes, specimens from the late follicular and late secretory phases (days 12–14) were also examined. The late follicular glands typically had euchromatic nuclei with one or two nucleoli, and these were basally situated, with an occasional mitotic figure evident. Supranuclear stacks of closely packed Golgi saccules, many mitochondria and a variety of vesicles were found in the apical cytoplasm (Figure 1C), ranging from moderate to electron dense. Short microvilli or blunt projections decorated the flat or domed apical surfaces. Occasionally, dispersed glycogen was evident throughout the cells.
The late secretory phenotype was characterized by cuboidal cells, with generally basally situated heterochromatic nuclei containing a nucleolus, and numerous surface microvilli. The most prominent feature was the presence of large, intracellular aggregates of glycogen, particularly in the supranuclear and apical parts of the cell (Figure 1D). Narrow cisternae of endoplasmic reticulum, sometimes forming stacks, and Golgi bodies were evident. Occasional clear vesicles, fat droplets and vesicles packed with glycogen could be found.
Endometriosis
The examination of toluidine blue-stained semithin sections often showed discrepancies between adjacent blocks of the same endometrium (osmicated versus non-osmicated tissue samples), suggestive of considerable gland heterogeneity. Most of the changes were restricted to the glands in the functionalis, and those specimens comprising basalis glands generally showed little alteration in their ultrastructure.
One of the most striking findings in the endometrium of diseased animals was a heavy accumulation of glycogen in the glands of the functionalis, evident to some extent even after 3 months duration (Figure 2A), where glycogen was found both within vesicles and loose in the cytoplasm, and often this was accompanied by the development of a late secretory phenotype, as also seen in the lectin histochemical findings.
|
There was a significant alteration in the ultrastructure of specimens from PA 7026 over the period biopsied, 3 months to 15 months, during which time the appearance of the glands changed to a late secretory phenotype. At 3 months duration of disease, the secretory vesicles contained finely dispersed particles of glycogen (Figure 2A). Basal deposits were not so much in evidence, with the nuclei occupying a basal position and the cytoplasm between the nucleus and basal plasma membrane being packed with mitochondria. Supranuclear regions, as in the normal controls, contained many Golgi bodies and vesicles either with or without glycogen. By 7 months, the vesicles often exhibited central cores of moderate electron density as well as dispersed glycogen (Figure 2B). At 15 months, many glands showed a late secretory phenotype, as seen also with the lectin histochemistry; secretory vesicles became less common and were often somewhat distorted, and the cells contained large aggregates of glycogen especially in apical protrusions (Figure 2C). The surface microvilli were often lost, and apical blebs had a smooth profile. Many autophagic vacuoles were also present.
This progression varied between animals; PA 7024 biopsied at 9, 12 and 15 months did not develop the dilated glands of the late secretory phase, although in some areas glands were wider than normal. Again, the secretory vesicles tended to contain fluffy material that increased in intensity between 12 and 15 months, and there was a lot of dispersed glycogen evident throughout the period under study. PAN 2733, in contrast, had developed a late secretory phenotype by 9 months and PA 7015 by 12 months’ duration of disease (Figure 2D), and, in the latter, this was very prominent by 16 months. Large accumulations of glycogen were evident, although some cells also contained secretory vesicles. Golgi vacuoles were prominent, and intracellular vacuoles were also occasionally seen (Figure 3A), a feature never observed in the normal controls. There was variation, however, in different areas of the endometrium; PA 2733 at 9 months and PAN 2654 at 13 months both showed areas with a late secretory ultrastructure with large clumps of glycogen though in the former the basal glands were little changed; in the latter, most glands had columnar cells containing dispersed glycogen and few vesicles, some of which were filled with fine aggregates of glycogen (Figure 3B). PAN 2640 at 13 months had two biopsies taken, one with mainly narrow and the other dilated glands; the narrow ones had few secretory vesicles, less glycogen and narrow strands of rough endoplasmic reticulum reminiscent of a late follicular phenotype (Figure 3C). The second biopsy with mainly dilated glands was not well fixed and was unsuitable for examination.
|
The menopausal specimen with spontaneous disease had a few shrunken glands in a very fibrotic stroma. The cells were low and cuboidal, with no secretory vesicles or glycogen deposits (Figure 3D). Cisternae of endoplasmic reticulum were narrow and were interspersed with rod-shaped mitochondria, occasional Golgi bodies and rare electron dense secretory droplets. Short microvilli covered the apical surfaces, which were slightly domed.
Endometriotic lesions
Various types of lesion were examined, and these differed considerably in their ultrastructure, although the bulk of those excised unfortunately lacked any glandular structures (Table III). A lesion at 3 months showed a cyst wall composed of columnar cells packed with dense secretory droplets and a lumen filled with flocculent, amorphous material. One cell was observed to have numerous basal bodies in the apical cytoplasm (Figure 4A), although cilia were not generally seen either in this or any other specimen, and apical surfaces bore microvilli or blunt processes. Golgi bodies and rod-shaped mitochondria were visible between the masses of droplets. Glands away from the cyst had a more normal architecture, though flat topped and with glycogen clumps in the apical cytoplasm. Another cystic structure lined with epithelial cells, from a case with 6 months’ disease, had no apical secretory droplets and just a few clear vesicles, with flat topped cells and sparse microvilli together with a prominent terminal web (Figure 4B). Basal deposits of glycogen were present in some areas, often with small, dense secretory droplets, and mitochondria. Lack of osmication made the membranous components of this specimen difficult to identify.
|
The examination of non-cystic lesions at 7 and 13 months showed tall, columnar surface epithelial cells with a normal architecture, sometimes with subnuclear glycogen deposits, and glands composed of narrow columnar cells with few vesicles or glycogen (Figure 4C) but longitudinally orientated, sometimes branched mitochondria and many Golgi bodies. Narrow cisternae of rough endoplasmic reticulum were closely associated with the mitochondria, as observed previously.
In contrast to these, spontaneous menopausal lesions contained thin-walled glands with flattened cells, heterochromatic nuclei and little if any visible glycogen (Figure 4D). Occasional microvilli were present on the apical surface and some secretory droplets near the basal face. Absence of osmication made membranous structures difficult to visualize.
Lectin histochemistry
Binding of the lectin from D. biflorus to the baboon endometrium has previously been shown to alter with different phases of the menstrual cycle (Jones et al., 1998) and is therefore under hormonal control, with heavy staining in the late proliferative phase fading to a thin residual binding to the apical plasma membrane in the late secretory phase. Binding of the lectin was therefore of prime interest in the cases of endometriosis to correlate the morphology with the biochemical status of the tissues.
All but one of the control tissues from days 9–11 post ovulation showed a very consistent staining pattern, with variable binding by the glands of the functionalis. The apical surface bound the lectin heavily, and there was an area apical to the nucleus that showed strong to weak staining that probably corresponded to the Golgi bodies and membranes of the rough endoplasmic reticulum, which we refer to as the supranuclear complex (Figure 5A). The basal areas of the cells usually contained a sparse population of granules giving a speckled appearance. In the basalis, staining tended to fade away, with narrow columnar cells without a distinct supranuclear complex but sometimes binding the lectin weakly on the upper half of the cells.
|
In the cases of endometriosis, staining often varied according to the histological appearance of the glands. Usually, glands with columnar cells and a narrow lumen, typical for days 9–11 post ovulation, stained darkly, but some specimens showed areas where the glands followed a tortuous path and were dilated, and the cells were more cuboidal with a scalloped profile. This pattern was more characteristic of the late secretory phase of the cycle, although all specimens were taken at the mid-secretory phase, and this was often, but not always, associated with low-level binding of DBA. In these cases, staining was usually restricted to the apical surface and supranuclear complex, although sometimes more of the apical area of the cell was involved.
The darkly staining phenotype was seen in 7 of the 18 cases, including a menopausal animal with spontaneous disease (PAN 5655). All biopsies from PA 7024 (7, 9, 12 and 15 months) were deeply stained (Figure 6A) as well the first two biopsies from PA 7026 at 3 and 7 months’ disease (Figure 6B). Cells were tall and columnar, more in keeping with the appearance of glands during the late follicular and early mid-secretory phases, which stain more strongly than mid-late secretory phase. The third biopsy of PA 7026 taken at 15 months, however, had a late secretory phenotype, although binding was still heavy on the apical membrane and only slightly lower in upper parts of the cells (Figure 6C). The same phenotype but with low levels of stain was seen in PA 7015 at 12 and 16 months (Figure 6D), although one of the 16-month biopsies was only partly affected. PAN 2733 at 9 months and PAN 2654 at 13 months, both of which were biopsied only once, also exhibited this late secretory phenotype, the latter in part only, with low levels of stain. Of the two 13 month biopsies from the same animal (PAN 2640), one was poorly fixed and the other showed little change from normal. The application of neuraminidase before incubation with DBA did not affect the staining to any significant degree in any of the specimens (Figure 5B).
|
The lesions showed a variety of staining results. A cystic structure from a case of 3 months’ disease (PA 7026) showed strong binding of the epithelial cyst wall and contents (Figure 7A), whereas an adjoining gland showed a basalis phenotype with variable but rather weak staining. Both blue and brown lesions from 6 and 7 months’ disease [PA 7022 (Figure 7B) and PA 7024], and a chocolate/red lesion of 15 month’s disease (PA 7024, Figure 7C) also showed variably stained basalis type glands while a blue/chocolate lesion (PA 7022) had a surface epithelium that stained very strongly with DBA but glands that showed a range of binding properties. A lesion from a menopausal animal that had spontaneous disease (PAN 5655) contained thin-walled glands with moderate staining (Figure 7D).
|
The other lectins used did not show significant changes and so will be only briefly described. Sialic acid-binding lectins from Sambucus nigra (SNA-1) and Maackia amurensis (MAA) did not bind strongly to the endometrial glands of either control or diseased specimens (Figure 5C and D), although with both lectins there was a gradual increase in binding from glands in the functionalis to those near the myometrium. In the basalis glands, vesicles were often outlined by both lectins though their contents were unstained. There was a marginal increase in SNA-1 binding to glands in some diseased specimens, especially those in the basalis, and sometimes lateral cell membranes were weakly stained. MAA bound more weakly than SNA-1 but was also increased basally in some diseased cases. With both lectins, there was an increase in the gland surface plasma membrane staining with the onset of disease.
Pokeweed mitogen (PAA) stained luminal epithelium quite strongly, but glands in the functionalis were generally weak with patchy stain or negative (Figure 5E), with some increase in the basalis (Figure 7F). Only one or two cases showed an increase over that seen in controls, although one animal stained slightly more strongly at 7, 12 and 15 months’ duration of disease (PA 7024), showing some fine supranuclear binding in the functionalis and more generalized stain basally. Lesions with a basal gland phenotype also showed increased binding in some areas, and two cystic structures (PA 7026 and PA 7022) had some strong staining of the epithelial cell wall (Figure 7E). Pretreatment with neuraminidase generally resulted in a slightly increased density of stain, especially on the apical plasma membrane and in the basal regions. WGA bound the supranuclear region strongly in the functionalis in most control cases (Figure 5F), with more homogeneous staining of the supranuclear cytoplasm in the basalis. There was also some staining of basal granules and lateral membranes, and binding of this lectin tended to be weaker in cases showing a more advanced phenotype. The examination of diseased cases and lesions showed little significant change compared with the range of staining densities seen on the normal controls. Neuraminidase had minimal effect, although there was marginal loss in a few cases.
Image analysis
The population distribution of data from the DBA staining of control and diseased animals, both with and without the secretory gland phenotypes, was found to be normal using the one-sample Kolmogorov–Smimov test. When DBA staining was compared between the two populations, excluding the late secretory phenotype, a highly significant difference was observed (P < 0.0001), thus confirming as significant the more densely stained appearance of the early/mid-secretory phenotype glands from some of the diseased tissues (Figure 8). It has already been shown that the there is a significant decrease in DBA binding in the mid-late secretory compared with the proliferative phase in control animals (Jones et al., 1998). In contrast, during the mid-secretory phase in animals with disease, the DBA staining was more reflective of the follicular phase during the early stages of the disease, and the late secretory phenotype was mainly seen as the disease progressed.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
The parallel examination of ultrastructure together with the expression of GalNAc-containing sequences bound by DBA allowed us to draw some very interesting conclusions about the responses of the endometrium to ectopic lesions seen in this model of endometriosis. In many cases, there was clear evidence of asynchrony between the estimated day of the menstrual cycle and the observed histological/ultrastructural appearances of the glands. The examination of biopsies from control animals in late follicular as well as the late luteal phase of the cycle enabled us to match the appearance of the glands from diseased animals with those at different stages of the normal cycle, and it was apparent that many of the affected animals showed a late secretory phenotype with respect to both DBA binding and also ultrastructure. This phenotype was found to develop at later stages of the disease, that is at 9 months or later, and such a progression was demonstrated very clearly in an animal biopsied at 3, 7 and 15 months (PA 7026), with the late secretory phenotype appearing by 15 months. Late secretory endometrial gland epithelial cells are characteristically more cuboidal than at earlier stages, with domed apices. They often express little of the glycan bound by DBA and also contain heavy deposits of glycogen in their apices, such as have been illustrated in Figure 1D. Specimens from animals with 9, 12, 13, 15 and 16 months’ disease showed both morphological and histochemical evidences of a late secretory phenotype in some glands, with low levels of DBA staining and massive amounts of glycogen in the somewhat cuboidal cells. Sometimes, this phenotype was apparent in the whole of the specimen, but at other times, only part of the sample showed this, indicating that there was heterogeneity within the endometrium. As well as this, often the depth of the endometrium was much shallower than that of true late luteal specimens, suggesting that there is an aberrant maturation process. This glandular variation was also reported in two ultrastructural studies of human endometriosis (Schweppe and Wynn, 1984
; Schweppe et al., 1984), although an earlier study by the same authors (Schweppe and Wynn, 1981) concluded that the ultrastructural features of most of the glands from the ectopic lesion resembled those of the corresponding uterine tissue.
It was not clear what caused the intense DBA staining found in some of the early specimens with endometriosis. However, this staining, which is more reflective of the proliferative stage of a normal menstrual cycle, is in agreement with our recent molecular and cellular studies. Early in the disease process (1–6 months) oestrogen-regulated genes such as c-fos and Cyr61 are up-regulated in the eutopic endometrium of baboons with disease, and the expression of these molecules decreases as the disease progresses (Gashaw et al., 2006; Hastings et al., 2006). The DBA staining further supports these molecular observations. Most of the early disease specimens showed a more intense staining pattern that was comparable with that seen in the late follicular stage of the normal cycle (Jones et al., 1998). Another feature of the presence of endometriotic lesions is the development of progesterone resistance in both the ectopic and eutopic endometria (Bulun et al., 2006). The suppression of progesterone action by progesterone receptor antagonists in the baboon also results in the continued staining for DBA in the mid-secretory phase of the cycle (Jones et al., 1998), which further supports our observations that endometriosis results in development of a predominantly proliferative morphological phenotype in the mid-secretory phase. The late secretory phenotype became more evident with extended periods of disease. These time points in the disease process are associated with a decrease in c-fos and Cyr61 and altered expression patterns of steroid receptors and HOXA 10 (Hastings and Fazleabas, 2006).
The examination of sections under high power indicated that the clear vesicles were often unstained—they could be visualized as small dots—but they were bound by an area of intense DBA expression. Possibly, this was due to the intracellular accumulation of glycans within membranous compartments such as the endoplasmic reticulum or Golgi apparatus. Foci of stain were often elongated and in a supranuclear position, suggesting that this might be so, especially as, in later stages, the supranuclear organelles and the apical membrane are the last to retain DBA binding. There are reports in the literature of a developmental delay in endometriosis as seen by electron microscopy (Schweppe et al., 1984) and similar findings are suggested by the DBA staining in this study. Previous work from our laboratory has shown strong DBA staining in normal endometrium at the late follicular stage (Jones et al., 1998) which may be maintained for some days after ovulation.
There are few descriptions of the ultrastructure of baboon endometrium, and those that are available are poorly illustrated (MacLennon et al., 1971; Dollar et al., 1978, 1981), and so it was important to have adequate controls with which to compare the endometrium from diseased animals. Having examined material from control day 14 post-ovulation animals, it became abundantly clear that a proportion of the later stage disease samples taken at days 9–11 post ovulation were showing a late secretory phenotype, with the characteristic accumulations of glycogen in the apices of the gland cells which also showed reduced height and a typically domed profile. This accelerated maturation may account in part for the reported reduction in fecundity of baboons (D’Hooghe, 1997) with endometriosis. There is also evidence that markers of uterine receptivity are absent from eutopic endometrium during the receptive period in baboons with endometriosis (Fazleabas et al., 2003), and some genes are also down-regulated or abnormally expressed (Hastings and Fazleabas, 2006), which may also affect fertility. The altered synchrony of the glandular structures in baboons with endometriosis during the window of uterine receptivity may further contribute to the inability of these cells to appropriately express the genes that are thought to be important for the establishment of a successful pregnancy.
It is also possible that the dense granules seen in some of the endometriotic lesions, especially the cysts, may reflect an earlier phenotype; dark granules were seen in our late follicular specimens, and similar dense granules were described by Dollar et al. (1978) in their normal late follicular endometria in baboon. However, they also resemble the granules seen in endocervix (Rhodin, 1977), which crowd the upper part of the cells in a similar fashion. Schweppe and Wynn (1984) found a similar appearance in laparoscopically obtained human endometriotic tissue. The lesion of 3 months’ disease contained such granules, although similar granules were not detected in the endometrial glands, despite heavy staining with DBA, suggesting that the cells of the cyst were following an aberrant differentiation pathway.
In summary, by using the induced baboon endometriosis model that permits sequential sampling of both eutopic and ectopic tissues, we have been able to demonstrate progressive changes in glandular architecture. These morphological changes, to a great extent, reflect the cellular and molecular changes associated with the disease process. We propose that this dyssynchrony within the window of uterine receptivity may contribute to the reduced fecundity associated with this disease.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
These studies were supported by a U54 grant HD 40093 from the National Institutes of Health to A.T.F.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Bulun SE, Cheng YH, Yin P, Imir G, Utsunomiya H, Attar E, Innes J, Kim JJ. (2006) Progesterone resistance in endometriosis: link to failure to metabolize estradiol. Mol Cell Endocrinol 248:94–103.[CrossRef][Web of Science][Medline]
Czernobilsky B and Fox H. (2003) Endometriosis. In Fox H and Wells M (Eds.). Haines and Taylor Obstetrical and Gynaecological Pathology 5th edn. (M. Churchill Livingstone, Edinburgh, UK) pp. 963–987.
D’Hooghe TM. (1997) Clinical relevance of the baboon as a model for the study of endometriosis. Fertil Steril 68:613–625.[CrossRef][Web of Science][Medline]
D’Hooghe TM, Bambra CS, Raeymaekers BM, De Jong I, Lauweryns JM, Koninckx PR. (1995a) Intrapelvic injection of menstrual endometrium causes endometriosis in baboons (Papio cyncephalus and Papio anubis). Am J Obstet Gynecol 173:125–134.[CrossRef][Web of Science][Medline]
D’Hooghe TM, Bambra CS, Raeymaekers BM, De Jong I, Hill JA, Koninckx PR. (1995b) The effects of immunosuppression on development and progression of endometriosis in baboons (Papio anubis). Fertil Steril 64:172–178.[Web of Science][Medline]
D’Hooghe TM, Bambra CS, Raeymaekers BM, Riday AM, Suleman MA, Koninckx PR. (1996a) The cycle pregnancy rate is normal in baboons with stage I endometriosis but decreased in primates with stage II and stage III–IV disease. Fertil Steril 66:809–813.[Web of Science][Medline]
D’Hooghe TM, Hill JA, Oosterlynck DJ, Koninckx PR, Bambra CS. (1996b) The effect of endometriosis on white blood cell populations in peripheral blood and peritoneal fluid in baboons. Hum Reprod 11:1736–1740.
Dollar JR, Boots LR, Bagwell JN. (1978) Histochemical and ultrastructural effects of Enovid E on the endometrium of the baboon. Cell Tissue Res 192:451–460.[Web of Science][Medline]
Dollar JR, Boots LR, Santolucito KA, Bagwell JN. (1981) Morphological and histochemical study of the endometrial effects of Ovral in the baboon. Cell Tissue Res 217:611–624.[CrossRef][Web of Science][Medline]
Eskenazi B and Warner ML. (1997) Epidemiology of endometriosis. Obstet Gynecol Clin North Am 24:235–258.[CrossRef][Web of Science][Medline]
Fazleabas AT, Brudney A, Gurates B, Chai D, Bulun S. (2002) A modified baboon model for endometriosis. Ann N Y Acad Sci USA 55:308–317.
Fazleabas AT, Brudney A, Chai D, Langoi D, Bulun SE. (2003) Steroid receptor and aromatase expression in baboon endometriotic lesions. Fertil Steril 80:Suppl 2),, 820–827.
Gashaw I, Hastings JM, Jackson K, Winterhager E, Fazleabas AT. (2006) Induced endometriosis in baboons (Papio anubis) increases the expression of the proangiogenic factor Cyr61 (CCN1) in the eutopic and ectopic endometria. Biol Reprod 74:1060–1066.
Hastings JM and Fazleabas AT. (2006) A baboon model for endometriosis: implications for fertility. Reprod Biol Endocrinol 4:suppl 1, S7 (Epub 27 Sept).[CrossRef][Medline]
Hastings JM, Jackson KS, Mavrogianis PA, Fazleabas AT. (2006) The estrogen early response gene FOS is altered in a baboon model of endometriosis. Biol Reprod 75:176–182.
Irimura Y and Nicolson GL. (1983) Interaction of pokeweed mitogen with poly(N-acetyllactosamine)-type carbohydrate chains. Carbohydr Res 120:187–195.[CrossRef][Web of Science][Medline]
Jones CJP, Fazleabas AT, Mcginlay PB, Aplin JD. (1998) Cyclic modulation of epithelial glycosylation in human and baboon (Papio anubis) endometrium demonstrated by the binding of the agglutinin from Dolichos biflorus (DBA). Biol Reprod 58:20–27.
Jones CJP, Morrison CA, Stoddart RW. (1992) Histochemical analysis of rat testicular glycoconjugates. 2) 
-galactosyl residues in O-and N-linked glycans in seminiferous tubules. Histochem J 24:327–336.[CrossRef][Web of Science][Medline]
Jones CJP, Santos TC, Abd-Elnaeim M, Dantzer V, Miglino MA. (2004) Placental glycosylation in peccary species and its relation to that of swine and dromedary. Placenta 25:649–657.[CrossRef][Web of Science][Medline]
MacLennon AH, Harris JA, Wynn RM. (1971) Menstrual cycle of the baboon. II. Endometrial ultrastructure. Obstet Gynecol 38:359–374.[Web of Science][Medline]
Maddox PH and Jenkins D. (1987) 3'Aminopropyltriethoxysilane (APES): a new advance in section adhesion. J Clin Pathol 40:1256–1260.
Rhodin JAG. (1977) Histology: A Text and Atlas(Oxford University Press, New York.).
Sampson JA. (1927) Peritoneal endometriosis due to menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 14:422–469.[Web of Science]
Schweppe KW and Wynn RM. (1981) Ultrastructural changes in endometriotic implants during the menstrual cycle. Obstet Gynecol 58:463–473.
Schweppe KW and Wynn RM. (1984) Endocrine dependency of endometriosis: an ultrastructural study. Eur J Obstet Gynecol Reprod Biol 17:193–208.[CrossRef][Web of Science][Medline]
Schweppe KW, Wynn RM, Beller FK. (1984) Ultrastructural comparison of endometriotic implants and eutopic endometrium. Am J Obstet Gynecol 148:1024–1039.[Web of Science][Medline]
Submitted on April 25, 2006; resubmitted on June 29, 2006; accepted on July 6, 2006.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A.G. Braundmeier and A.T. Fazleabas The non-human primate model of endometriosis: research and implications for fecundity Mol. Hum. Reprod., October 1, 2009; 15(10): 577 - 586. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. J. P. Jones, I. M. Inuwa, L. G. Nardo, P. Litta, and A. T. Fazleabas Eutopic Endometrium From Women With Endometriosis Shows Altered Ultrastructure and Glycosylation Compared to That From Healthy Controls--A Pilot Observational Study Reproductive Sciences, June 1, 2009; 16(6): 559 - 572. [Abstract] [PDF]
|
|
|
|
 |
|
 C. Wang, P. A. Mavrogianis, and A. T. Fazleabas Endometriosis Is Associated with Progesterone Resistance in the Baboon (Papio anubis) Oviduct: Evidence Based on the Localization of Oviductal Glycoprotein 1 (OVGP1) Biol Reprod, February 1, 2009; 80(2): 272 - 278. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. J. P. Jones, L. G. Nardo, P. Litta, and A. T. Fazleabas Ultrastructure of Ectopic Peritoneal Lesions From Women With Endometriosis, Including Observations on the Contribution of Coelomic Mesothelium Reproductive Sciences, January 1, 2009; 16(1): 43 - 55. [Abstract] [PDF]
|
|
|
|
|
|
J.J. Kim, H.S. Taylor, Z. Lu, O. Ladhani, J.M. Hastings, K.S. Jackson, Y. Wu, S.W. Guo, and A.T. Fazleabas Altered expression of HOXA10 in endometriosis: potential role in decidualization Mol. Hum. Reprod., May 1, 2007; 13(5): 323 - 332. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||