A molecular staging model for accurately dating the endometrial biopsy

Natural variability in menstrual cycle length, coupled with rapid changes in endometrial gene expression, makes it difficult to accurately define and compare different stages of the endometrial cycle. Here we develop and validate a method for precisely determining endometrial cycle stage based on global gene expression. Our 'molecular staging model' reveals significant and remarkably synchronised daily changes in expression for over 3400 endometrial genes throughout the cycle, with the most dramatic changes occurring during the secretory phase. Our study significantly extends existing data on the endometrial transcriptome, and for the first time enables identification of differentially expressed endometrial genes with increasing age and different ethnicities. It also allows reinterpretation of all endometrial RNA-seq and array data that has been published to date. Our molecular staging model will significantly advance understanding of endometrial-related disorders that affect nearly all women at some stage of their lives, such as heavy menstrual bleeding, endometriosis, adenomyosis, and recurrent implantation failure.

Superficial peritoneal endometriotic lesions are histologically diverse and rarely demonstrate menstrual cycle synchronicity with matched eutopic endometrium

STUDY QUESTION

Do menstrual cycle-dependent changes occur in the histological appearance of superficial peritoneal endometriotic lesions, and are they equivalent to those observed in the eutopic endometrium?

SUMMARY ANSWER

Only a small subset of superficial peritoneal endometriotic lesions exhibits some histological features in phase with menstrual cycle-related changes observed in eutopic endometrium.

WHAT IS KNOWN ALREADY

Endometriotic lesions are frequently described as implants that follow menstrual cycle-related changes in morphology, as per the eutopic endometrium. This concept has been widely accepted despite the lack of conclusive published evidence.

STUDY DESIGN, SIZE, DURATION

This was a retrospective cohort study of 42 patients, from across the menstrual cycle, with surgically and histologically confirmed endometriosis. Patients were a subset selected from a larger endometriosis study being conducted at the Royal Women's Hospital, Melbourne since 2012.

PARTICIPANTS/MATERIALS, SETTING, METHODS

Histological features of epithelium, stroma and gland morphology were examined in haematoxylin and eosin stained sections of superficial peritoneal endometriotic lesions and matched eutopic endometrium (menstrual: n = 4, proliferative: n = 11, secretory: n = 17, hormone-treated: n = 10). At least two biopsies (average = 4, range = 2-8 biopsies) and a matched endometrial sample were analysed for each patient and results were presented per endometriotic gland profile (n = 1051). Data were analysed using mixed effects logistic regression to account for multiple patients and multiple endometriotic biopsies, each with multiple endometriotic gland profiles. This model also enabled analysis of endometriotic lesions versus eutopic endometrium.

MAIN RESULTS AND THE ROLE OF CHANCE

There was considerable inter- and intra-patient variability in the morphology of superficial peritoneal endometriotic lesions. Menstrual cycle-associated changes were only observed for some features in a subset of endometriotic gland profiles. The proportion of endometriotic gland profiles with epithelial mitoses significantly increased in the proliferative phase (18% of gland profiles) relative to the menstrual phase (0% of endometriotic gland profiles) (odds ratios (OR) 9.30; 95% confidence intervals (CI) = 3.71-23.32; P < 0.001). Fewer blood-filled gland lumens were observed in the secretory phase (45% of endometriotic gland profiles) compared to the menstrual phase (67% of endometriotic gland profiles) (OR, 0.30; 95% CI = 0.11-0.79; P = 0.015). The features of the eutopic endometrium analysed in this study did not reflect the results in matched endometriotic lesions (P > 0.05).

LARGE SCALE DATA

Not applicable.

LIMITATIONS, REASONS FOR CAUTION

This study focused on features observed in sections of superficial peritoneal lesions and these may differ from features of deep infiltrating endometriosis or ovarian endometriomas. Cycle phases were limited to menstrual, proliferative and secretory phases to allow appropriate statistical modelling.

WIDER IMPLICATIONS OF THE FINDINGS

This study highlights heterogeneity in the histological characteristics of superficial peritoneal lesions. It challenges the assumption that lesion morphology consistently reflects menstrual cycle-associated changes.

STUDY FUNDING/COMPETING INTEREST(S)

Research reported in this publication was supported in part by National Health and Medical Research Council (NHMRC) project grants GNT1012245, GNT1105321 and GNT1026033 (P.A.W.R., J.E.G. and S.J.H.-C.). There are no competing interests.

Genetic risk factors for endometriosis near estrogen receptor 1 and coexpression of genes in this region in endometrium

The etiology and pathogenesis of endometriosis are complex with both genetic and environmental factors contributing to disease risk. Genome-wide association studies (GWAS) have identified multiple signals in the estrogen receptor 1 (ESR1) region associated with endometriosis and other reproductive traits and diseases. In addition, candidate gene association studies identified signals in the ESR1 region associated with endometriosis risk suggesting genetic regulation of genes in this region may be important for reproductive health. This study aimed to investigate hormonal and genetic regulation of genes in the ESR1 region in human endometrium. Changes in serum oestradiol and progesterone concentrations and expression of hormone receptors ESR1 and progesterone receptor (PGR) were assessed in endometrial samples from 135 women collected at various stages of the menstrual cycle. Correlation between hormone concentrations, receptor expression and expression of genes in the ESR1 locus was investigated. The effect of endometriosis risk variants on expression of genes in the region was analyzed to identify gene targets. Hormone concentrations and receptor expression varied significantly across the menstrual cycle. Expression of genes in the ESR1 region correlated with progesterone concentration; however, they were more strongly correlated with expression of ESR1 and PGR suggesting coregulation of genes. There was no evidence that endometriosis risk variants directly regulated expression of genes in the region. Limited sample size and cellular heterogeneity in endometrial tissue may impact the ability to detect significant genetic effects on gene expression. Effects of these variants should be validated in a larger dataset and in relevant individual cell types.

Generation of immortalized human endometrial stromal cell lines with different endometriosis risk genotypes

Endometriotic lesions are composed in part of endometrial-like stromal cells, however, there is a shortage of immortalized human endometrial stromal cultures available for research. As genetic factors play a role in endometriosis risk, it is important that genotype is also incorporated into analysis of pathological mechanisms. Human telomerase reverse transcriptase (hTERT) immortalization (using Lenti-hTERT-green fluorescent protein virus) took place following genotype selection; 13 patients homozygous for either the risk or non-risk 'other' allele for one or more important endometriosis risk single nucleotide polymorphism on chromosome 1p36.12 (rs3820282, rs56318008, rs55938609, rs12037376, rs7521902 or rs12061255). Short tandem repeat DNA profiling validated that donor tissue matched that of the immortalized cell lines and confirmed that cultures were genetically novel. Expression of morphological markers (vimentin and cytokeratin) and key genes of interest (telomerase, estrogen and progesterone receptors and LINC00339) were examined and functional assays for cell proliferation, steroid hormone and inflammatory responses were performed for 7/13 cultures. All endometrial stromal cell lines maintained their fibroblast-like morphology (vimentin-positive) and homozygous endometriosis-risk genotype following introduction of hTERT. Furthermore, the new stromal cultures demonstrated positive and diverse responses to hormones (proliferation and decidualisation changes) and inflammation (dose-dependent response), while maintaining hormone receptor expression. In conclusion, we successfully developed a range of human endometrial stromal cell lines that carry important endometriosis-risk alleles. The wider implications of this approach go beyond advancing endometriosis research; these cell lines will be valuable tools for multiple endometrial pathologies offering a level of genetic and phenotypic diversity not previously available.

Follistatin is essential for normal postnatal development and function of mouse oviduct and uterus

Female mice lacking the follistatin gene but expressing a human follistatin-315 transgene (tghFST315) have reproductive abnormalities (reduced follicles, no corpora lutea and ovarian-uterine inflammation). We hypothesised that the absence of follistatin-288 causes the abnormal reproductive tract via both developmental abnormalities and abnormal ovarian activity. We characterised the morphology of oviducts and uteri in wild type (WT), tghFST315 and follistatin-knockout mice expressing human follistatin-288 (tghFST288). The oviducts and uteri were examined in postnatal Day-0 and adult mice (WT and tghFST315 only) using histology and immunohistochemistry. Adult WT and tghFST315 mice were ovariectomised and treated with vehicle, oestradiol-17β (100ng injection, dissection 24h later) or progesterone (1mg×three daily injections, dissection 24h later). No differences were observed in the oviducts or uteri at birth, but abnormalities developed by adulthood. Oviducts of tghFST315 mice failed to coil, the myometrium was disorganised, endometrial gland number was reduced and oviducts and uteri contained abundant leukocytes. After ovariectomy, tghFST315 mice had altered uterine cell proliferation, and inflammation was maintained and exacerbated by oestrogen. These studies show that follistatin is crucial to postnatal oviductal-uterine development and function. Further studies differentiating the role of ovarian versus oviductal-uterine follistatin in reproductive tract function at different developmental stages are warranted.

Activin and follistatin interactions in the male reproductive tract: activin expression and morphological abnormalities in mice lacking follistatin 288

Activin A is an important regulator of testicular and epididymal development and function, as well as inflammation and immunity. In the adult murine reproductive tract, activin A mRNA (Inhba) expression levels are highest in the caput epididymis and decrease progressively towards the distal vas deferens. The activin-binding protein, follistatin (FST), shows the opposite expression pattern, with exceptionally high levels of the Fst288 mRNA variant in the vas deferens. This unique pattern of expression suggests that activin A and follistatin, in particular FST288, play region-specific roles in regulating the epididymis and vas deferens. The cellular distribution of activin and follistatin and structural organization of the male reproductive tract was examined in wild-type and transgenic (TghFST315) mice lacking FST288. Compared to wild-type littermates, TghFST315 mice showed a 50% reduction in serum follistatin and a significant elevation of both activin A and B. Testicular, epididymal and seminal vesicle weights were reduced, but intra-testicular testosterone was normal. A decrease in the epididymal duct diameter in the corpus and thickening of the peritubular smooth muscle in the cauda, together with increased coiling of the proximal vas deferens, were observed in TghFST315 mice. No immune cell infiltrates were detected. Immunohistochemistry indicated that epithelial cells are the main source of activins and follistatin in the epididymis and vas deferens. Activin A, but not activin B, was also localized to sperm heads in the lumen of the epididymis and vas deferens. Expression of Inhba and another immunoregulatory gene, indoleamine-2,3-dioxygenase (Ido-1), was increased approximately twofold in the TghFST315 caput epididymis, but several other genes associated with immunoregulation, inflammation or fibrosis were unaffected. Our novel data indicate that disruption of follistatin expression has significant effects on the testis and epididymis, and suggest an association between activin A and indoleamine-2,3-dioxygenase in the caput epididymis, with implications for the epididymal immunoenvironment.

Progesterone stimulates expression of follistatin splice variants Fst288 and Fst315 in the mouse uterus

Follistatin, an inhibitor of activin A, has key regulatory roles in the female reproductive tract. Follistatin has two splice variants: FST288, largely associated with cell surfaces, and FST315, the predominant circulating form. The mechanism regulating uterine expression of these variants is unknown. Quantitative RT-PCR was used to measure expression of follistatin splice variants (Fst288, Fst315), the activin bA subunit (Inhba) and the inhibin a subunit (Inha) in uterine tissues during early pregnancy (days 1–4, preimplantation) and in response to exogenous 17b-oestradiol (single s.c. injection) and progesterone (three daily s.c. injections) in ovariectomized mice. Uterine Fst288, Fst315 and Inhba expression increased during early pregnancy, with greater increases in Fst315 relative to Fst288 suggesting differential regulation of these variants. Fst288, Fst315, Inhba and Inha all increased in response to progesterone treatment. Fst288, but not Fst315, mRNA decreased in response to 17b-oestradiol treatment, whereas Inhba increased. A comparison of the absolute concentrations of uterine follistatin mRNA using crossing thresholds indicated that both variants were more highly expressed in early pregnancy in contrast to the hormone treatment models. It is concluded that progesterone regulates uterine expression of both follistatin variants, as well as activin A, during early pregnancy in the mouse uterus

157. LYMPHATICS IN THE HUMAN PLACENTAL BED AND SURROUNDING THE SPIRAL ARTERIOLES DISAPPEAR DURING ENDOMETRIAL DECIDUALISATION

The mammalian placenta plays central roles in maternal tolerance of the semi-allogeneic fetus and fluid balance between maternal and fetal compartments. The lymphatics play a role in both immune function and fluid balance. The aim of this study was to describe the distribution of lymphatic vessels in human placental bed and decidua, with particular focus on the lymphatics that surround the remodelling spiral arteries during decidualisation and trophoblast invasion. Placental bed, non-placental bed and decidual biopsies were obtained from women undergoing elective pregnancy termination (6–18 weeks gestational age) and from women undergoing elective caesarean section at term. Double immunohistochemical labeling was performed on serial sections to identify lymphatic vessels in conjunction with blood vessels, smooth muscle, epithelial and trophoblast cells, or proliferating cells. Using representative photomicrographs, descriptive findings of the organisation of the human placental bed lymphatics were made. Lymphatic vessels positive for podoplanin (D2-40) were abundant in hypersecretory endometrium (lacking stromal decidualisation) at all stages of gestation. By contrast, the decidua was nearly always devoid of lymphatics. In some samples, structures that appeared to be regressing lymphatics were observed at the boundary between hypersecretory and decidual tissues. Lymphatic vessels were notably absent from the vicinity of spiral arteries that were surrounded by decidualised stromal cells. Lymphatic vessels in hypersecretory endometrium appeared larger and more elongated as gestation progressed. Proliferating lymphatic vascular endothelial cells were identified in both large vessels, and in streaks of D2-40 positive cells that could have been newly forming lymphatic vessels. In conclusion endometrial stromal cell decidualisation results in a loss of lymphatics, with this phenomenon being particularly apparent around the spiral arterioles; the functional consequences of this loss have yet to be elucidated.

308. FOLLISTATIN SPLICE VARIANTS FST288 AND FST315 INCREASE DURING EARLY MOUSE PREGNANCY: REGULATION BY PROGESTERONE?

Follistatin acts by binding and neutralising the activity of activin-A, which has important regulatory roles in development, reproduction and inflammation. There are two isoforms of follistatin comprising 288 and 315 amino acids (FST288 and FST315), resulting from alternative gene splicing. FST288 binds spontaneously to heparan sulphate and is largely bound to cell surface proteoglycans. FST315 is the predominant circulating form and can only bind to heparan sulphate after binding activin-A. The regulation of these splice variants in the female reproductive tract have not been investigated in detail. In this study, our aim was to quantify the expression of FST288 and FST315 mRNA in the mouse uterus during early pregnancy (days 1–4, pre-implantation), and in response to exogenous oestradiol-17b (100 ng × single s.c. injection, dissection after 24 h) and progesterone (1 mg × three daily s.c. injections, dissection 24 h after last injection) in ovariectomised mice. Gene expression was analysed using quantitative RT-PCR. Primers amplifying a product from exon 5 to 6a (unique to FST288) or from exon 5 to 6b (unique to FST315) were used to discriminate the isoforms. In early pregnancy, expression of both FST288 and FST315 increased significantly (approximately 35-fold and 100-fold, respectively) on days 3–5, relative to days 1–2, corresponding with the increase in circulating progesterone levels that occurs at day 3. A significant increase in FST288 and FST315 mRNA expression (both approximately 35-fold) was also observed in ovariectomised mice in response to exogenous progesterone, but there was no increase in response to oestradiol-17β. In contrast to the similar rate of increase in response to exogenous progesterone, FST315 mRNA expression increased more rapidly than FS288 in early pregnancy, indicating that differential regulation of the two isoforms also occurs. We conclude that progesterone regulates both FST288 and FST315 mRNA expression during early pregnancy in the mouse uterus.

An RNA spiking method demonstrates that 18S rRNA is regulated by progesterone in the mouse uterus

Identifying suitable housekeeping genes for quantitative RT-PCR in the uterus is problematic, as this tissue undergoes significant structural and functional alterations during the oestrous cycle and pregnancy in response to circulating hormones. The suitability of 18S rRNA as a housekeeping gene in mouse uterus was investigated by introducing an 'RNA spike' standard into the reverse transcription reaction. 18S rRNA levels increased by Day 4 of pregnancy and after progesterone administration in ovariectomized mice. We conclude that 18S rRNA is not a suitable housekeeping gene for quantitative RT-PCR analysis in progesterone-responsive tissues, and the RNA spiking method provides a suitable alternative.

527. EXTENSIVE LYMPHATIC REMODELLING OCCURS IN HUMAN PLACENTAL BED DURING THE FIRST 20 WEEKS OF PREGNANCY

The aim of this study was to describe the distribution of lymphatics in human placental bed from 6–20 weeks of pregnancy, with emphasis on the lymphatics surrounding the spiral arterioles as they undergo remodelling associated with trophoblast invasion. Placental bed biopsies were obtained following elective termination (n=41, 6–18 weeks gestation) and at term (n=5). Samples were routinely processed for serial sectioning at 3um to allow structural relationships to be compared in consecutive sections. Double immunohistochemistry protocols were used to identify lymphatic vessels in conjunction with one of: blood vessels, smooth muscle, epithelium and trophoblast, or proliferating cells. Immunostaining demonstrated that lymphatic vessels were present in abundance in the hypersecretory endometrium throughout all stages of gestation, and were prominent encircling spiral arterioles and adjacent to the endometrial glands. In contrast, the decidua was nearly always devoid of lymphatics, and in particular they were absent from the vicinity of spiral arterioles that were surrounded by decidual cells. In hypersecretory endometrium lymphatic vessel profiles varied greatly in architecture and size, and included focal accumulations of endothelial cells, through to fully formed lymphatic vessels. There was evidence of lymphatic vessel endothelial cell proliferation in all the different vessel types. The trophoblast appeared to have no overt influence on the structure or appearance of the lymphatic endothelium, with no evidence of incorporation into, or distortion of, the lymphatic vessel walls. Obvious increases were noted in the size of the lymphatic vessel profiles in hypersecretory endometrium as gestation progressed. The study provides the first detailed description of the placental bed lymphatics and their relationship to other endometrial structures during human pregnancy. We have shown that lymphatic vessels are absent from the decidua, apparently regressing as surrounding stromal cells decidualise. In contrast, the underlying hypersecretory endometrium contains abundant lymphatics which increase in size as pregnancy progresses.

531. Lyve-1 AND Vegfr-3 POSITIVE LYMPHATIC VESSELS ARE LARGELY RESTRICTED TO THE MYOMETRIUM IN THE MOUSE UTERUS

The human endometrium contains a distinct population of lymphatic vessels, with reduced numbers of vessel profiles in the functionalis relative to the basalis and underlying myometrium. The mechanisms restricting lymphatic vessel development within the functionalis and the consequences of this distribution for endometrial function have not been investigated. To determine how the distribution of endometrial lymphatics is regulated, an in vivo animal model is required for mechanistic studies. As an initial step, we described the distribution of lymphatic vessels and the expression of Vegf-C and Vegf-D within the mouse uterus using immunohistochemistry. Uterine tissues were collected from mice during the oestrus cycle (diestrus, proestrus and oestrus, n=7–9 per group), during early pregnancy (days 1-4, n = 4–5 per group), and from ovariectomised mice treated with vehicle, estradiol-17β or progesterone treatment (n=7–8 per group). Uterine sections were immunostained with antibodies against Lyve-1, Vegfr3, Vegf-C and Vegf-D. Lyve-1 positive lymphatic vessels were almost exclusively observed in the connective tissue between the longitudinal and circular muscle layers of the myometrium. Lymphatic vessel profiles were rare within the endometrium and were only observed on 24 % of the sections examined. The Vegfr3 immunostaining was less robust than Lyve-1 with considerable variation in non-lymphatic staining among the different sections (which did not relate to a particular reproductive state or treatment group). Despite the variability, the pattern of lymphatic vessels was as seen with the Lyve-1 antibody. Vegf-C and Vegf-D immunostaining was present in all uterine compartments (epithelium, stroma, myometrium), however, only minimal changes were noted in expression across the oestrus cycle, during early pregnancy, or in hormone treated mice. In conclusion, despite the presence of the key lymphangiogenic growth factors Vegf-C and Vegf-D, there are minimal lymphatic vessels present within the mouse endometrium. This implicates other factors in the mechanism restricting endometrial lymphatic vessel growth.

240. Reduced endometrial angiogenesis during early pregnancy in relaxin-deficient mice

Relaxin is a peptide hormone with important roles in the reproductive tract, including the growth and remodelling of endometrial vasculature. It has been shown to stimulate VEGF secretion from human endometrial stromal cells in vitro and increases endometrial vascularisation in ovariectomised steroid-primed primates in vivo. We have used mouse models to show that oestrogen and progesterone stimulate angiogenesis (new blood vessel growth) within the endometrium. Endometrial angiogenesis also occurs in the early stages of mouse pregnancy, which coincides with an increase in circulating progesterone. To date, no studies have investigated the effects of relaxin on endometrial angiogenesis in early pregnancy. Our aim was to test the hypothesis that endometrial angiogenesis would be reduced in relaxin-deficient mice (Rln−/−) in comparison to their wildtype (Rln+/+) counterparts. Uterine tissues were collected from Rln−/− and Rln+/+ mice on days 1 to 4 of pregnancy, before implantation. All mice were treated with BrdU before dissection to allow the number of blood vessel profiles containing proliferating endothelial cells (PVPs) to be quantified by double CD31/BrdU immunohistochemistry. Consistent with published studies, PVPs were first observed on days 3 and 4 of pregnancy. However, the percentage of PVPs was reduced in Rln−/− mice compared with Rln+/+ mice (Day 3: median = 4.4% v. 19.6%, Day 4: 9.6% v. 22.2%). We subsequently identified relaxin and relaxin receptors in the mouse endometrium in early pregnancy. Our data suggest that locally synthesised relaxin acts in synergy with progesterone to initiate endometrial angiogenesis in early pregnancy.

238. Endometrial vessel morphology is altered following progestin treatment in a mouse xenograft model

Human endometrium undergoes cyclic changes under the influence of oestrogen and progesterone. When progestins are used for contraception, the endometrium regresses and breakthrough bleeding often occurs. The aim of this study was to investigate the short-term effects of progestin on human endometrium in a mouse xenograft model. Uterine tissue was placed subcutaneously into NOD/SCID mice (n = 12). Mice were given estradial valerate every fourth day for two weeks. Mice then received an implant containing medroxyprogesterone acetate (MPA) or an empty implant. After two weeks, mice were dissected and the xenografts formalin fixed and serially sectioned (5µm) for immunohistochemical analysis. Sections were double immunostained for α-smooth muscle actin and either FVIII (blood vessels) or D2–40 (lymphatic vessels). The endometrium from the progestin treated group contained decidual-like stroma cells and glandular epithelium with morphology ranging from squamous to columnar. The endometrium from the control group also contained fibroblast-like stromal cells and glandular epithelium with tall columnar epithelium. The endometrial blood vessel density was significantly reduced in the progestin-treated group (156.3 ± 13.4 vessel profiles/mm2) compared with the control group (273.5 ± 41.5 vessel profiles/mm2) (P = 0.02); there was no significant difference in lymphatic vessel density (progestin: 43.5 ± 5.9 v. control: 35.6 ± 9.6 vessel profiles/mm2). Blood vessel area was significantly increased in the progestin-treated group (3.7x10−4 ± 1.7x10−5 mm2) compared with controls (1.8x10−4 ± 1.2x10−5 mm2) (P = 0.0001) and the lymphatic v essel area was also significantly increased in the progestin-treated group (8.8x10−4 ± 7.8x10−5 mm2) compared with controls (2.9x10−4 ± 5.7x10−5 mm2) (P = 0.0001). This work has provided a model for the study of human endometrial vasculature, illustrating a significant increase in blood and lymphatic vessel size during progestin treatment. The increase in blood vessel size was associated with a significant reduction in blood vessel density in progestin treated samples.

410. Gene expression analysis in endometriotic lesions using laser capture microdissection

Previous studies examining gene expression profiles in normal endometrium and endometriotic lesions have used RNA extracted from whole tissue samples. Results from these studies can be difficult to interpret as they reflect expression averaged across several different cell types that may be functionally quite different. The aim of this study was to establish laser capture microdissection (LCM) as a technique to examine gene expression in stromal and epithelial cells from normal and ectopic endometrium. We hypothesised that genes associated with inflammation would be elevated in cells from endometriotic lesions. Full thickness uterine samples were collected during abdominal hysterectomy from normal cycling premenopausal women. Endometriotic lesions were collected during abdominal laparoscopy. Samples were either frozen in OCT or stored in RNAlater for 12 h before freezing. Tissues were immunostained with an antibody against CD10 to identify ectopic endometrial stromal cells before LCM. Endometrial epithelial and stromal cells were collected using the PALM MicroLaser System. RNA quality was accessed using Experion. TGFβ1, MMP1, αSMA, SMAD2 and NFκB mRNA was analysed using real-time RT–PCR. Of the endometriotic samples stored in OCT (n = 58), only 14% (n = 8) had visible endometrial glands. Of these, only 37% (n = 3) had RNA of an acceptable quality for further analysis. However, RNA quality and quantity were dramatically improved in 3 of 5 samples collected in RNAlater. In preliminary studies, expression of TGFβ1 and αSMA mRNA was elevated in endometriotic lesions in comparison to the normal endometrium, whereas NFκB expression did not change. We have shown that RNAlater solution is useful to preserve RNA quality for small clinical endometriotic samples and that immuno-guided LCM-generated homogenous cell populations coupled with real-time RT–PCR can provide valuable insights into cell and disease-specific gene expression in endometriotic lesions.

Influence of different hormonal regimens on endometrial microvascular density and VEGF expression in women suffering from breakthrough bleeding

BACKGROUND

The aim of this study was to quantify blood vessel density (BVD) and immunoreactive vascular endothelial growth factor (VEGF) levels in endometrial biopsies taken from women suffering breakthrough bleeding (BTB) under different exogenous hormonal regimes.

METHODS

Endometrial biopsies from women in Melbourne with BTB were divided into four groups: combined-continuous hormone therapy (HT) (estrogen and progestin taken daily), cyclical HT (daily estrogen with progestin for 14 days each cycle), progestin-only, or no HT. Subjects from Barcelona were using the Mirena intrauterine levonorgestrel-releasing system for contraceptive purposes, with menstrual diaries for classification into four groups (amenorrhea, infrequent, regular and prolonged). Control biopsies from Melbourne were included in the study. Endometrial samples were immunostained for VEGF and blood vessel localization using an antibody to CD34.

RESULTS

Results showed that BVD was significantly reduced in the progestin-only treated group compared with the other three treatment groups (P = 0.028). In addition, all four Mirena BTB groups had significantly reduced BVD compared with controls. Considerable heterogeneity was observed in VEGF immunostaining within and between individual samples with no major differences between HT or Mirena.

CONCLUSION

These results provide strong evidence that unopposed progestins reduce endometrial BVD and that there is no link between VEGF immunostaining and BVD or BTB.

272. Progesterone stimulates endothelial cell proliferation, but not stromal cell proliferation, in mouse endometrium

Previous studies have suggested that progesterone stimulates stromal cell (SC) proliferation in the mouse endometrium1. However, these studies have not differentiated endothelial cells (EC) from other SC. In this study, we investigated the effects of progesterone on cellular proliferation in ovariectomised mouse endometrium. We hypothesised that progesterone would stimulate both SC and EC proliferation. One group of CBA × C57 mice (n = 6) were treated with a single injection of 100 ng of estradiol on day eight following ovariectomy, followed by a day with no treatment and three consecutive daily injections of 1 mg progesterone. Other groups were treated with either the vehicle (n = 5), estradiol (n = 4) or progesterone (n = 5) injections only. All groups were dissected on day 13 after ovariectomy, 4 h following a BrdU injection. CD31/BrdU double staining immunohistochemistry allowed proliferating EC to be differentiated from proliferating SC. Mice treated with progesterone only had significantly higher EC proliferation in comparison to females treated with progesterone following oestrogen priming (P = 0.05) or vehicle only (P = 0.01) (progesterone only: median=97.3 proliferating EC (PEC)/mm2 [range = 60.8–203.4]; oestrogen plus progesterone: 41.0 PEC/mm2 [8.9–86.9]; vehicle only: 0.0 PEC/mm2 [0.0–3.1]). Unexpectedly, there was no significant difference in SC proliferation among the treatment groups (progesterone only: 50.1 PSC/mm2 [39.2–102.6]; oestrogen plus progesterone: 46.1 PSC/mm2 [12.6–120.8]; vehicle only: 44.8 PSC/mm2 [17.3–68.4]). To determine if VEGF had a role in the progesterone-induced EC proliferation, the previous experiment was repeated with the inclusion of mice treated with VEGF anti-serum. The addition of VEGF anti-serum significantly inhibited progesterone-induced EC proliferation (46.8 PEC/mm2 [38.9–128.0]; P = 0.04], but had no effect on SC proliferation (P = 0.3). These results demonstrate that progesterone stimulates endometrial EC proliferation, but not SC proliferation as reported by earlier studies1. Studies are currently underway to further investigate the role of VEGF in mediating progesterone effects on endometrial EC. (1)Clarke, C.L. and Sutherland, R.L. (1990) Endocrine Reviews 11, 266–301.

239. Hormonal control of vascular mural cell recruitment in the mouse endometrium

The human endometrium undergoes regular periods of growth and regression, including concomitant changes in the vasculature, and is one of the few adult tissues where significant angiogenesis (new blood vessel formation) and arteriogenesis (recruitment of vascular smooth muscle cells (VSMC) and pericytes) occurs on a routine, physiological basis. In this study, mouse models were used to investigate the effects of oestrogen and progesterone on endometrial vascular mural cell recruitment. The aim was to quantify changes in the proportion of vessels covered by α-smooth muscle actin (α-SMA, a marker of VSMC and pericytes) in hormone-treated ovariectomised mice. We hypothesised that relative vessel α-SMA coverage would increase following progesterone treatment (in conjunction with endothelial cell (EC) proliferation), but not following oestrogen treatment (when EC proliferation also occurs). Ovariectomised mice were given a single oestradiol (100 ng) or vehicle injection, before dissection 24 h later, or three consecutive daily injections of progesterone (1 mg) or vehicle. The percentage of vessel profiles with no, minimal, extensive or complete α-SMA coverage were quantified after CD31/α-SMA double immunostaining. There was a significant decrease in the percentage of vessel profiles with no α-SMA coverage following progesterone treatment (20±4.3 % [mean±SE] v. 57±4.6 %, t(7)=12.5, P<0.001), and a significant increase in the percentage of vessels with minimal or extensive α-SMA coverage (44±3.4 % v. 27±3.7%, t(7)=4.7, P<0.001 and 27±4.3% v. 5±0.5%, t(7)=5.8, P<0.001, respectively), in comparison to vehicle-treated mice. The percentage of vessels with complete α-SMA coverage, representing vessels with a coat of VSMC, did not change significantly in comparison to vehicle-treated mice (8±2.3% v. 10±1.2%, t(7)=0.6, P=0.55). There were no significant changes in the percentage of vessels with differing α-SMA coverage in oestrogen-treated mice. In continuing studies, we will quantify the proportion of proliferating α-SMA positive cells and examine mouse endometrial tissues using a pericyte-specific marker.

Effect of long-term progestin treatment on endometrial vasculature in normal cycling mice

The aim of this study was to develop a mouse model to investigate the effects of long-term progestin-only exposure on endometrial vascular structure. Normal cycling mice received Silastic implants containing either medroxyprogesterone acetate (MPA) or levonorgestrel (LNG) and were dissected after 1, 3 or 6 weeks. Endometrial vascular density increased significantly within 1 week of MPA (482 +/- 40.2 vessels/mm2) or LNG (440 +/- 26.5 vessels/mm2) treatment compared with normal cycling mice (293 +/- 10.5 vessels/mm2). MPA increased stromal cell density within 1 week of treatment (13813 +/- 1450 cells/mm2) compared with normal cycling mice (8256 +/- 928 cells/mm2). However, although LNG significantly increased stromal cell density overall, the increase did not reach significance within the individual weeks examined. There was no significant change in the ratio of vascular to stromal cell density among treated and normal cycling mice. Epithelial cell height significantly decreased within 1 week of LNG (17.6 +/- 1.3 microm) treatment compared with normal cycling mice (23.5 +/- 1.3 microm); epithelial cell height also decreased within 1 week of MPA treatment (16.6 +/- 2.1 microm), although this did not reach statistical significance. VEGF immunostaining increased significantly in luminal epithelium after MPA or LNG treatment, and in glandular epithelium after LNG treatment. These observations are similar to those reported in human endometrium, suggesting that this mouse model may facilitate further investigations into breakthrough bleeding due to long-term progestin use.

Intravascular neutrophils partially mediate the endometrial endothelial cell proliferative response to oestrogen in ovariectomised mice

The aim of this study was to investigate the role of intravascular neutrophils in initiating endothelial cell proliferation following oestrogen treatment in ovariectomised mouse endometrium. Uterine tissues were collected from ovariectomised C57/CBA female mice 24 h after oestrogen treatment with or without systemic neutrophil depletion. Neutropenia was achieved with either an in-house anti-neutrophil serum (ANS) or Gr-1 monoclonal antibody. All mice received an i.p. injection of bromodeoxyuridine (BrdU) 4 h prior to dissection to allow visualisation of proliferating cells using immunocytochemistry. Endometrial sections were immunostained for BrdU, vascular endothelial growth factor (VEGF), and neutrophils (using ANS). Oestrogen treatment of ovariectomised mice significantly increased the number of intravascular neutrophils, whereas induction of neutropenia with either ANS or Gr-1 in conjunction with oestrogen treatment prevented this increase. Oestrogen treatment of ovariectomised mice also significantly increased the number of intravascular VEGF-positive cells; however, whereas induction of neutropenia with ANS significantly reduced this increase, Gr-1 did not. In both studies, neutropenia significantly reduced, but did not eliminate, the amount of endometrial endothelial cell proliferation. These results suggest a role for neutrophils in endometrial angiogenesis following acute oestrogen treatment; however, the presence of VEGF-positive cells even after induction of neutropenia suggests that more than one type of leukocyte may be involved.

Persistence of pregnant mare serum gonadotrophin in plasma of Japanese quail (Coturnix coturnix japonica)

Pregnant mare serum gonadotrophin (PMSG) is currently being used to develop a hormone regime that will stimulate reproductive development in Japanese quail (Coturnix coturnix japonica). In this study, the persistence of PMSG in quail plasma was examined after a single injection of the hormone (500 IU). Plasma concentrations of PMSG increased to a peak 12 h after injection and declined to approximately 50% of peak concentrations 24 h after injection. Pregnant mare serum gonadotrophin concentrations declined gradually thereafter and had not returned to basal levels by 96 h after injection. Cloacal diameter, and ovarian and oviducal mass, had increased significantly by 96 h after injection. The persistence of PMSG in quail plasma has implications for the use of the hormone in future regimes stimulating reproductive activity in birds.

Delayed ovulation and parturition in a viviparous alpine lizard (Niveoscincus microlepidotus): morphological data and plasma steroid concentrations

The southern snow skink, Niveoscincus microlepidotus, exhibits an unusual biennial reproductive cycle with an extended gestation period of approximately 1 year. Morphological data were gathered on a monthly basis, providing a detailed picture of the reproductive cycle. Vitellogenesis begins in spring, immediately after parturition. Maximum follicular diameter is reached before the winter hibernation period and ovulation occurs the following spring. Embryos are fully developed and reach maximum size by early autumn. Yolk reserves are depleted before winter. Birth of between one and four young occurs the following spring. Plasma progesterone concentrations are low (2.7 +/- 0.9 ng mL(-1)) in post-partum females, begin to rise in autumn in vitellogenic females and peak (38.5 +/- 7.9 ng mL(-1)) in pre-ovulatory females after hibernation. Concentrations are high (15.4 +/- 5.9 ng mL(-1)) in early pregnancy and decline to basal levels before winter and well before birth in spring. Plasma oestradiol concentrations peak during vitellogenesis (1.0 +/- 0.3 ng mL(-1)) and decline to basal levels during pregnancy (0.2 +/- 0.03 ng mL(-1)). A second oestradiol peak occurs before parturition (0.7 +/- 0.2 ng mL(-1)). Thus, functional completion of vitellogenesis and gestation is achieved by autumn in successive years. The mechanisms that defer ovulation and parturition by a further six months are unknown.

Administration of pregnant mare serum gonadotropin to Japanese quail (Coturnix coturnix japonica): dose response over seven days and comparison of delivery by daily injection or osmotic pump

AIMS

This study is part of a research programme that aims to develop a method of hormone treatment to stimulate breeding in female birds. The aims of this study were to compare dose rates and two different delivery methods, daily injection or osmotic pump, for hormone treatment of Japanese quail (Coturnix coturnix japonica) with pregnant mare serum gonadotropin (PMSG).

METHODS

PMSG (0, 5, 10, 20, 40 and 80 IU PMSG/day) was administered to 6-week-old Japanese quail housed under short-day, cool-temperature conditions (8L:16D at 7-10 degrees C) by daily injections or osmotic pump for 7 days. Three additional groups were untreated: one group was dissected at Day 0, and two groups were maintained under either short-day, cool-temperature or long-day, warm-temperature (16L:8D, 20 degrees C) conditions for 7 days. Cloacal diameter was measured daily, and ovarian and oviductal mass and plasma oestradiol concentrations measured at the end of the treatment period.

RESULTS

PMSG treatment stimulated ovarian and oviductal growth. After 7 days of treatment with 10-20 IU PMSG, ovarian and oviductal mass were similar to those in birds moved from short to long days. Females treated with the highest doses of PMSG (40 or 80 IU) had significantly larger cloacal diameters and ovarian and oviductal mass than other treated birds or birds maintained under long-day, warm-temperature conditions. Daily injections and osmotic pumps were equally effective methods of delivery. However, there was considerable variation in response to PMSG among individual birds and this was particularly obvious at the higher doses (20-80 IU PMSG). There were no differences in plasma oestradiol concentrations between groups treated using daily injections or osmotic pumps.

CONCLUSIONS

A dose of 10 IU PMSG/day was chosen for use in future experiments with Japanese quail, for the first 7 days of treatment. The delivery method of choice for future studies will depend on the practical considerations of the research in question.

Oviducal structure in four species of gekkonid lizard differing in parity mode and eggshell structure

Oviducal structure was analysed in vitellogenic females from four species of gekkonid lizard exhibiting variation in parity mode and eggshell structure: Hemidactylus turcicus (oviparous) which produces a hard, calcareous eggshell; Saltuarius wyberba (oviparous) which produces a soft, parchment-like eggshell; and Hoplodactylus maculatus and Hoplodactylus duvaucelii (both viviparous). Oviducts were analysed by light, scanning electron and transmission electron microscopy. The uterus exhibited differences among species that were directly attributable to parity mode. H. turcicus and S. wyberba (oviparous) had numerous uterine shell glands; H. maculatus and H. duvaucelii (viviparous) had very few. The uterus also exhibited differences between the two oviparous species (H. turcicus and S. wyberba) which may be related to the type of eggshell produced. Variations were noted in the staining properties of the uterine glandular and epithelial cells. The structure of the infundibulum, uterine tube, isthmus and vagina also differed among species, but differences could not be directly related to parity mode or eggshell structure. Instead, the differences may be related to how prepared the oviduct is for ovulation in individuals analysed from the different species. This study confirms, in the Gekkonidae, aspects of oviducal structure that have been associated with parity mode in other squamate taxa.

Ultrastructure of the uterus in an ovariectomized gecko (Hemidactylus turcicus) after administration of exogenous estradiol

The uterus of an oviparous gecko, Hemidactylus turcicus, was analysed after ovariectomized females underwent a period of treatment (up to 14 days) with exogenous estradiol. Analysis focused on the uterine mucosa, which is made up of an epithelial layer and an underlying lamina propria containing the shell glands. These tissues are thought to be responsible for secretion of the eggshell components and were thus chosen for analysis using transmission electron microscopy. In ovariectomized females, the epithelial layer was low and cuboidal with minimal/no differentiation or secretory activity. Treatment with exogenous estradiol resulted in a significant increase in cell height associated with gradual differentiation of the epithelium. Development of non-ciliated cells included production of secretory granules (low electron density) at the apical cell surface. The shell glands showed less obvious changes over the course of treatment. Shell glands contained two cell types: dark cells with darkly staining nuclei and organelles, and light cells with very indistinct nuclei and organelles, except for prominent rough endoplasmic reticulum and free ribosomes. This study provides results consistent with published light microscopy studies for other reptiles and additionally provides ultrastructural details of reptilian uterine development not previously available.

Plasma corticosterone levels are not significantly related to reproductive stage in female common geckos (Hoplodactylus maculatus)

The relationship between plasma CORT concentration and reproductive state was investigated in New Zealand's common gecko, Hoplodactylus maculatus. This primarily nocturnal gecko has an unusual biennial reproductive cycle in which females are pregnant for about 14 months. In contrast to a previous report for a viviparous lizard, plasma CORT concentrations did not vary significantly in daytime samples among females at four different stages of reproduction (vitellogenic, mid-pregnant, late pregnant prior to winter dormancy, and spent). Significant seasonal variation in concentration (independent of reproductive state) was observed in daytime samples collected over an 8-month period (P < 0.001), perhaps reflecting seasonal patterns of growth and/or lipid deposition. Plasma CORT concentrations in summer were compared between night and day in spent and mid-pregnant females. A significant difference in concentration was detected in mid-pregnant females (P < 0.01), but unexpectedly, the concentration was higher by day than by night. No relationship of CORT with time of day was detected in spent females. There was a significant positive correlation between CORT and body temperature in day samples, but not in night samples. In contrast to many previous reports for reptiles, concentrations in females held in cloth bags showed no significant elevation during up to 2.5 hr in captivity. All groups sampled in this study had relatively low mean concentrations of CORT (< or = 6.4 ng/ml), even those held captive for up to 2.5 hr. We conclude that plasma CORT plays no obvious role in maintaining the long gestation period in H. maculatus and that relationships among CORT, activity period, and body temperature need examination in a wider range of nocturnal and diurnal reptiles before consistent patterns can be identified.