The endometrium and embryo implantation

How embryos attach and implant remains a mystery. Implantation represents the remarkable synchronisation between the development of the embryo and the differentiation of the endometrium. As long as these events remain unexplained, improvement in the success of in vitro fertilisation treatment and the development of contraception that modifies implantation is likely to be difficult.

In most animals, the endometrium undergoes a series of changes leading to a period of uterine receptivity called the “window of implantation.” Outside of this time the uterus is resistant to embryo attachment. In a study by Hertig et al in 1956, women were asked to record their menstrual pattern and dates of unprotected intercourse before they had a hysterectomy for benign gynaecological disease.¹ With their informed consent, their uteruses were carefully examined after operation, and the authors found that a number of them had conceived just before surgery. In these cases, embryos found in the uterus before the 20th day of the menstrual cycle were “free lying”—that is, not attached to the endometrium. Embryos found on or after the 21st day of the menstrual cycle were attached. Naturally, such research would not be performed today, but data from in vitro fertilisation programmes have substantiated these findings.² During in vitro fertilisation treatment embryos replaced before the 20th day may implant; those replaced after the 24th day do not.

The architectural changes that occur to the endometrium during a 28 day menstrual cycle were also investigated in the 1950s using light microscopy.³ Alterations in the endometrium during days 16 to 20 mainly affect the epithelial glands, which show increased secretory activity, prominent subnuclear vacuoles, and a decrease in mitotic activity. The stroma abruptly becomes oedematous on day 21. In the 1980s, electron microscopy studies identified epithelial protrusions into the uterine cavity called pinopodes; these appear between day 19 and day 21.⁴ In animals and humans their appearance coincides with a receptive endometrium.

Changes in the expression of molecules on the cell surface have also been observed in the conversion of the endometrial surface from a non-receptive to a receptive state. The mucins, a group of antiadhesive molecules, seem to have the most profound effect. They make up part of a thick layer, the glycocalyx, on the luminal epithelial surface of the uterus. In mice the glycocalyx prevents the embryo from direct contact with the endometrium, but changes in concentrations of oestrogen and progesterone after ovulation cause this layer to thin. This exposes the endometrium and enables it to react with the cells of the embryo.⁵ In humans the mucins, specifically MUC-1, are also under hormonal control, but in contrast to mice the endometrial epithelium continues producing MUC-1 while it is receptive to embryo implantation.⁶ This suggests that implantation is different in humans. It is conceivable that a decrease in MUC-1 is localised to specific but as yet unidentified receptor sites in humans and that this decrease is directed by the embryo itself. Alternatively, hormonal changes during the receptive phase may cause a subtle alteration in the structure of MUC-1 that allows the embryo to attach and implant. Furthermore, in both animals and humans the MUC-1 mucin has been found in the fallopian tube. Although it is not known whether it is under hormonal control at this site, it clearly could prevent ectopic pregnancy because of its antiadhesive properties.

Although the regulation of antiadhesion molecules, such as mucin glycoproteins, is undoubtedly important, this alone is not sufficient to support the attachment of the embryo to the uterine epithelial cells. The expression of adhesion molecules, such as integrins, selectins, cadherins, and the immunoglobulin superfamily, is also thought to be involved in the development of a receptive state. In the endometrium, the profile of expressed integrins varies according to the phase of the menstrual cycle; the combined presence of certain integrins has been proposed as a means of distinguishing receptive endometrium from non-receptive.⁷ The pattern of temporal expression of the selectins, cadherins, and immunoglobulin superfamily is less well defined in humans because much of the data are derived from animal studies.

Because of the ethical and moral dilemmas faced by researchers investigating embryo implantation, most of the in vivo data are from studies that have examined the endometrium or embryo in isolation. It is therefore not surprising that the coordination of the process of human embryo attachment has been attributed to oestrogen and progesterone and to “quality embryos.”⁸ The embryo is not passive but is an active orchestrator of its attachment and fate. The spatiotemporal expression of embryonic proteins and their influence on the endometrium may prove to be critical. Consequently, co-culture techniques using donor embryos and endometrial epithelial cells with or without their stroma are being developed. Such in vitro approaches will contribute to our understanding of the complex interaction between the embryo and the endometrium. Unravelling the mystery of the mechanisms controlling the receptivity of the human endometrium remains an exciting challenge.