The change from proliferative to secretory endometrium, described in detail in Chapter 4, is an essential part of achieving the receptive conditions required for implantation. This change is the histologic expression of many biochemical and molecular events. The endometrium is 10-14 mm thick at the time of implantation in the midluteal phase. By this time, secretory activity has reached a peak, and the endometrial cells are rich in glycogen and lipids. Understanding the dynamic endocrine behavior of the endometrium (Chapter 4) increases the appreciation for its active participation in the implantation process. The window of endometrial receptivity is restricted to days 16-19 (of a 28-day cycle).

Early pregnancy factor (EPF) can be detected in the maternal circulation within 1-2 days after coitus results in a pregnancy. EPF prior to implantation is apparently produced by the ovary in response to a signal from the embryo. After implantation, EPF is no longer secreted by the ovary but is derived from the embryo. EPF has immunosuppressive properties and is associated with cell proliferation and growth. Many other proteins, such as pregnancy-associated plasma protein-A and pregnancy associated endometrial protein, have been identified in trophoblast and the endometrium, but the biologic roles for these proteins remain to be determined.

Blastocysts grown in culture produce and secrete human chorionic gonadotropin (HCG). Messenger RNA for HCG can be found in 6 to 8-cell human embryos. Because the 8 to 12-cell stage is achieved about 3 days after fertilization, it is believed that the human embryo begins to produce HCG before implantation when it can be detected in the mother (about 6-7 days after ovulation). The embryo is capable, therefore, of preimplantation signaling, and higher levels of estradiol and progesterone can be measured in the maternal circulation even before maternal HCG is detectable, presumably due to stimulation of the corpus luteum by HCG delivered directly from the uterine cavity to the ovary. Function of the corpus luteum is crucial during the first 7-9 weeks of pregnancy, and luteectomy early in pregnancy can precipitate abortion. Similarly, early pregnancy loss in primates can be induced by injections of anti-HCG serum.

In rodents and rabbits, implantation can be interrupted by injection of prostaglandin inhibitors. Indomethacin prevents the increase in endometrial vascular permeability normally seen just prior to implantation. Additional evidence for a role by prostaglandins in the earliest stages of implantation is the finding of increased concentrations at implantation sites, similar to any inflammatory response. The blastocysts of mice, rabbits, sheep, and cows produce prostaglandins, and prostaglandin release from human blastocysts and embryos has been demonstrated. The endometrial cells are also a likely source of prostaglandin, and its synthesis may be stimulated by the tissue response that accompanies implantation. However, decidual synthesis of prostaglandins is significantly reduced compared to proliferative and secretory endometrium, apparently a direct effect of progesterone activity and perhaps a requirement in order to maintain the pregnancy. Nevertheless, prostaglandin synthesis is increased at the implantation site, perhaps in response to blastocyst factors, e.g. platelet activating factor. In the rabbit, platelet activating factor also induces the production of early pregnancy factor (discussed above).

As discussed in Chapter 4, the many cytokines, peptides, and lipids secreted by the endometrium are inter-related through the stimulating and inhibiting actions of estrogen and progesterone, as well as the autocrine/paracrine activities of these substances on each other. The response to implantation certainly involves the many members of the growth factor family. Epidermal growth factor, for example, is highly concentrated in the implantation site in the mouse.

Implantation is defined as the process by which an embryo attaches to the uterine wall and penetrates first the epithelium and then the circulatory system of the mother to form the placenta. It is a process that is limited in both time and space. Implantation begins 2-3 days after the fertilized egg enters the uterus on day 18 or 19 of the cycle (3 or 4 days after ovulation). Thus, implantation occurs 5-7 days after fertilization. The implantation site in the human uterus is usually in the upper, posterior wall in the midsagittal plane.

The human blastocyst remains in the uterine secretions for approximately 72 hours and then hatches from its zona pellucida in preparation for attachment. Implantation is marked initially by apposition of the blastocyst to the uterine epithelium. A prerequisite for this contact is a loss of the zona pellucida, which, in vitro, can be ruptured by contractions and expansions of the blastocyst. In vivo this activity is less critical, because the zona can be lysed by ingredient of the uterine fluid. The exact nature and function of these ingredient and related proteins that are thought to mediate the implantation process (implantation-initiating factor, fibronectin, uteroglobin and blasto-kinin) are uncertain. Their production is, however, known to be dependent upon the secretion of ovarian steroid hormones. Even if the hormonal milieu and protein composition of the uterine fluid are hospitable to the implantation, it may not occur if the embryo is not at the proper stage of development. It has been inferred from this information that there must be developmental maturation of the surface of the embryo before it is able to achieve attachment and implantation.

Reports on changes in the surface charge of preimplantation embryos differ in their findings, and it is unlikely that changes in surface charge are solely responsible for adherence of the embryo to the surface of epithelial cells. Binding of the lectin concanavalin A to the embryo changes during the preimplantation period, an indication that the surface glycoproteins of the embryo are in transition. It is reasonable to assume that these changes in configuration on the surface occur in order to enhance the ability of the embryo to adhere to the maternal surface.
As the embryo comes into close contact with the endometrium, the microvilli on the surface flatten and interdigitate with those on the luminal surface of the epithelial cells. A stage is reached where the cell membranes are in very close contact and junctional complexes are formed. The embryo can no longer be dislodged from the surface of the epithelial cells by flushing the uterus with physiologic solutions. Three types of subsequent interactions between the implanting trophoblast and the uterine epithelium have been described. In the first, trophoblast cells intrude between uterine epithelial cells on their path to the basement membrane. In the second type of interaction, the epithelial cells lift off the basement membrane, an action which allittles the trophoblast to insinuate itself underneath the epithelium. Last, fusion of trophoblast with individual uterine epithelial cells has been identified by electron microscopy in the rabbit. This latter method of gaining entry into the epithelial layer raises interesting questions concerning the immunologic consequences of mixing embryonic and maternal cytoplasm.

Trophoblast has the ability to phagocytose a variety of cells but in vivo this activity seems largely confined to removal of dead endometrial cells, or cells that have been sloughed from the uterine wall. Similarly, despite the invasive nature of the trophoblast, destruction of maternal cells by enzymes secreted by the embryo does not seem to play a major role in implantation. The embryo does secrete a variety of enzymes (e.g. collagenase and plasminogen activators), and these may be important for digesting the intercellular matrix that holds the epithelial cells together. Studies in vitro have demonstrated the presence of plasminogen activator in mouse embryos and in human trophoblast, and its activity is important in the attachment and early outgrowth stages of implantation. Urokinase and proteases, trophoblastic enzymes which convert plasminogen to plasmin, are inhibited by HCG, indicating regulation of this process by the embryo.

The embryo at a somewhat later stage of implantation can digest, in vitro, a complex matrix composed of glycoproteins, elastin and collagen, all of which are ingredient of the normal intercellular matrix. Additional studies in vitro have shown that cells move away from trophoblast in a process called "contact inhibition. Trophoblast then spreads to fill the spaces vacated by the cocultured cells. Once the intracellular matrix has been lysed, this movement of epithelial cells away from trophoblast would allittle space for the implanting embryo to move through the epithelial layer. Trophoblast movement is aided by the fact that only parts of its surface are adhesive, and the major portion of the surface is nonadhesive to other cells.

It is very likely that the highly proliferative phase of trophoblastic tissue during early embryogenesis is regulated by the many growth factors produced in both fetal and maternal tissues. Further penetration and survival depend upon factors that are capable of suppressing the maternal immune response to paternal antigens. The endometrial tissue makes a significant contribution to growth factor activity and immune suppression by synthesizing proteins in response to the blastocyst even before implantation.'81
One of the great mysteries associated with implantation is the mechanism by which the mother rejects a genetically abnormal embryo or fetus. It is possible that the abnormal embryo cannot produce a signal in early pregnancy that can be recognized by the mother.

The embryonic signals will be effectual only in a proper hormone milieu. Much of the knowledge concerning the hormone requirements for implantation in animals has been gained from studies of animals in delayed implantation. In a number of species, preimplantation embryos normally lie dormant in the uterus for periods of time which may extend for as long as 15 months before implantation is initiated. In other species,

delayed implantation can be imposed by postpartum suckling or by performing ovariectomy on day 3 of pregnancy. This produces a marked decrease in synthesis of DNA and protein by the blastocyst. The embryo can be maintained at the blastocyst stage by injecting the mother with progesterone. Using this model, hormonal requirements for implantation have been determined. In mice there is a requirement for estrogen and progesterone follittleed by estrogen, which initiates implantation. In other species the nidatory stimulus of estrogen is not required, and progesterone alone is sufficient.
Although it is known that the hormone milieu of delayed implantation renders the embryo quiescent, it is not known whether this represents a direct effect on the embryo or whether there is a metabolic inhibitor present in uterine secretions that acts upon the embryo. Removal of the embryo from the uterus to culture dishes allittles rapid resumption of normal metabolism, suggesting that there has, in fact, been a release from the inhibitory effects of a uterine product.