Implantation

Long was I hugg'd closeolong and long.
Immense have been the preparations for me,
Faithful and friendly the arms that have help'd me.
Cycles ferried my cradle, rowing and rowing like friendly boatmen...
They sent influences to look after what was to hold me.
Before I was born out of my mother, generations guided me,
My embryo has never been torpid, nothing could overlay it.
- WALT WHITMAN, "Song of Myself"

Implantation and successful pregnancy demand continuous interactions between the embryo and the mother. In 1905, John Ballantyne, one of the most eminent obstetricians of his day, reflected on the language used to describe the interactions between the uterus and the blastocyst. None of the terms used to describe the relationship were satisfactory. The old view (used by Hunter) was that the uterus underwent a decidual reaction to catch the embryo. The term is still used today (and Ballantyne grudging approved it), but it means the same thing as a decidual tree--one that sheds its leaves. The decidua was that which emerges and is then discarded. Ballantyne sees the metaphors of attachment, anchoring, and mooring as not working well at all, and the metaphors of implanting, and embedding as all being better. Much has happened since then, but the relationship between the blastocyst and uterus is still a controversial one. It differs in different species and at different times in any particular species. In some species and times, anchorage, mooring, or attachment might be the best description. In other species or times, implanting would be an obvious metaphor. In still other species and times, it would not be improper to speak about the invasion of the blastocyst into the uterus. It should be remembered, though, that these are merely attempts to describe, in macroscopic, human terms what is occurring on the microscopic level between entities without volition.

Preparation of the uterus for adhesion and implantation:

The uterus and the blastocyst both have to be made amenable to implantation. This is achieved through the actions of the ovarian hormones, progesterone and estrogen. The priming of the uterus occurs through the actions of both hormones (Figure 1). Progesterone is thought to induce a "prereceptive state" that is responsive to the estrogen. In the uterus, estrogen (estradiol) will bind to its receptor and will initiate a receptive state in the uterus. It triggers the uterus to produce growth factors such as epidermal growth factor, heparin-binding EGF, and leukemia inhibitory factor. These proteins, in the presence of Hoxa-10, allow the expression of the cyclooxygenase (COX) enzymes. The cyclooxygenases are the major enzymes in synthesizing prostaglandins. The COX-2 enzyme appears to be critical in this process, since the uteri in mice defective in this enzyme cannot receive embryos (Lim et al., 1997; Paria et al., 2000). The prostaglandins (especially prostacyclin) are critical for the uterus’ ability to receive the blastocysts, although the mechanisms by which they work is not yet understood.

Figure 1   A schematic diagram representing the possible molecular mechanisms coordinating the receptivity of the uterus and the activation of the blastocyst to achieve implantation. The physiological characteristics of the "prereceptive" and "receptive" states need to be ascertained. E2, estradiol; P4, progesterone. (After Paria et al., 2000.)

HB-EGF may be very important, as it is secreted by the cells only at the potential sites of blastocyst attachment in mice, i.e., on the luminal surface of the uterine epithelium. HB-EGF is able to stimulate trophoblast growth and zona hatching of the blastocyst (Martin et al., 1998). The genes are activated there about seven hours before the blastocysts actually attach.

The activation of the blastocyst is also initiated by estrogen from the ovary, but with a twist. The estrogen is metabolized in the uterus into 4-hydroxyestradiol-17b (catecholestrogen). This modified estrogen makes the trophoblasts competent to implant, probably by inducing the expression of cell adhesion molecules and growth factor receptors on their cell surfaces (Paria et al., 1998).

The trophoblast of the blastocyst, in turn, produces interleukin-1b. The uterine tissues have receptors for this growth factor, and IL-1 is necessary for maintaining the uterus in a state that will receive the embryo. If the IL-1 receptor is absent or malfunctioning, no adhesion will occur between the uterus and the trophoblast.

Trophoblast-uterus adhesion

There are several adhesion systems that hold the trophoblast to the uterus. The initial systems may consist of numerous low strength interactions such as those between heparan sulfate proteoglycans and their receptors. Soon afterward, a stronger contact is made by cadherins, integrins, and the trophinin-tastin-bystin complex. E-cadherin and P-cadherin are found on both the uterine tissue and the trophoblast, and these adhesive molecules presumably find their respective partners. Trophinin on the trophoblast can bind to trophinin on the uterine epithelium as long as the trophinin proteins are embedded in a complex with tastin and bystin (Fukuda and Nozawa, 1999; 1995; Kimber and Spanswick, 2000). HB-EGF can span the membrane as well as diffuse in a soluble form. As a membrane bound protein, it can bind to the EGF receptors and the heparan sulfate proteoglycans of the blastocysts. These latter systems appear to be specific for trophoblast-uterine attachment.

he avb3 integrin on the apical surface of the trophoblast tissue binds to fibronectin and perlecan (heparan sulfate proteoglycan) in the uterine extracellular matrix. In humans, this particular integrin is expressed on both the trophoblast and the uterine tissues. In the uterus, this integrin is only expressed between days 19 and 24 of the menstrual cycle, the period of optimum uterine receptivity for the blastocyst. This may be extremely important, as this integrin does not appear to be present in the uterine epithelium of infertile women who have "luteal phase" abnormalities (Lessey et al., 1992). In the mouse, another integrin, a7b1, becomes expressed shortly after adhesion takes place. This is an "invasive" integrin, as it binds to the laminin within the uterine tissue. It will be needed to bring the blastocyst into the uterine wall.

Blastocyst movement into the uterus

The migration of the blastocyst into the uterus is accomplished (1) by the proliferation of trophoblast tissue, (2) by the ability of the trophoblast to digest a path through uterine tissue, (3) by changing the population of integrins on the trophoblast such that the embryo always moves further inside the uterus, and (4) by the "decidual reaction," whereby the uterus can bring blood to the developing embryo. The decidual reaction of the uterus to the implanting embryo shares many characteristics with acute inflammatory responses. The permeability of the uterine blood vessels increases, and new blood vessels are made (angiogenesis). In addition, the trophoblast starts secreting digstive enzymes--collagenases (for the collagens), stromelysin (for fibronectin, laminin, and proteoglycans) and plasmin (to activate the above proteases). Once within the uterus, the integrin population changes again. The human cytotrophoblast cells within the chorionic villi produce the a6b4 laminin-binding integrin. Moreover, as they migrate through the epithelium, they downregulate the a6b4 integrin and upregulate a5b1, which binds to the fibronectin of the stromal matrix. Once within the uterine wall, they produce another integrin, a1b1, a receptor for both laminin and type IV collagen. In human pre-eclampsia, the blastocyst does not fully move into the uterus. Here, the a1b1 integrin fails to be upregulated (Cross et al., 1994). The uterus does not make its normal amount of blood vessels, and collagenases are not secreted by the trophoblast. It is difficult to determine whether the mother or the embryo is the deficient partner in this interaction. It appears that the lack of trophoblast penetration prevents the decidual reaction from forming uterine blood vessels. In the absence of these vessels, the uterus is hypoxic. Under hypoxic conditions, the integrin population doesn't change and the collagenases are not secreted.

When mouse embryos are transplanted to non-uterine sites within the mouse, the trophoblast invades the tissue and does not stop (see Alpin and Glasser, 1994; Alpin, 2000). In the uterus, the growth of the trophoblast is tightly regulated. It is probable that TGFb1, expressed in the uterus, blocks the secretion of proteases, and protease inhibitors are also present in the uterus.

Trophoblast control of maternal endocrine functions

Progesterone is required throughout pregnancy to maintain a uterine environment that will continue the pregnancy. Different species induce progesterone in different ways, but in primates and horses, chorionic gonadotropins from the trophoblast maintain the ovarian corpus luteum in an active state, allowing it to make progesterone. Later in pregnancy, this function is taken over by the placenta, itself. In many mammals, the trophoblast also produces somatomammotrophin (placental lactogen), a hormone that causes the growth of the mammary glands and the initiation of milk production.

Placental control of the maternal immune system

Somehow, the fetus must survive as an allograft--that is as a genetically foreign tissue within the body of another. The fetus (and its trophoblast) contain a genome that encodes proteins from both the mother and the father. The most critical of these antigens are the histocompatability antigens. These proteins distinguish self from non-self. They are critical in allowing each person to mount an immune response against just about any foreign substance, and they are also the antigens that prevent grafts from being accepted by genetically non-identical people or animals.

The potency of this block against immunological rejection can be seen when embryos are transplanted from one species into another (zebra embryos into horses, for instance). The embryo will do fine, even though it shares none of the histocompatability antigens with the mother. There are several ways that the fetus can cause such immunosuppression. The first is non-selective. The trophoblast appears to secrete several molecules that suppress the immune system. These include factors such as progesterone, a-fetoprotein, placental lactogen, activin, inhibin, and interleukins 1b, 6 and 10. Interleukin 10 is extremely interesting in this respect, since it blocks the proliferation of those helper T-cells associated with stimulating allograft rejection (Cadet et al., 1995). These factors would give a non-specific immunosuppression, especially in the vicinity of the trophoblast.

here is also evidence for specific immunosuppression directed towards the particular paternally encoded histocompatability antigens. Here, the maternal T cells (those lymphocytes capable of rejecting allografts) that recognize paternal antigens on the trophoblast are selectively abrogated. The pregnant mouse had reduced its number of T cells, recognizing the paternally encoded histocompatability antigen. Moreover, if tumor cells bearing that foreign antigen were inoculated into such pregnant mice, the mice would not reject the tumor as well as it would normally. After delivery, the immune system returns to normal. Thus, it appears that during pregnancy, the mother's T cells acquire a reversible "tolerant" state whereby they no longer recognize paternal histocompatability antigens (Tafuri et al, 1995).

t is not known how the trophoblast mediates this specific immunosuppression. However, there are some tantalizing clues. One is that the cytotrophoblast bears a trophoblast-specific histocompatability protein called HLA-G. HLA-G is known to protect cells against being killed by natural killer lymphocytes, and it may also create tolerance to the T cells that cause allograft rejection (Schmidt and Orr, 1993). One other possibility is that HLA-G or some other protein acts on the T cell precursors to become tolerant of the paternal antigens. Heyborne and coworkers (1994) show that there is a T-cell recognition of trophoblast cells which is unique in the body. The T-cells interacting with the trophoblasts are not instructed to divide or to destroy tissue. It is not yet known if this special trophoblast-T cell interaction protects the trophoblast and fetus from the mother's immune response.