REVIEW ARTICLE: Research on Blastocyst Implantation Essential Factors (BIEFs)

Authors

  • Koji Yoshinaga

    1. Reproductive Sciences Branch, Center for Population Research, Eunice Kennedy Shriver National Institute of Child Health and Human Development, NIH, DHHS, MD, USA
    Search for more papers by this author

Koji Yoshinaga, Reproductive Sciences Branch, Center for Population Research, Eunice Kennedy Shriver National Institute of Child Health and Human Development, NIH, DHHS, Building 6100, Room 8B01, Bethesda, MD 20892-7510, USA.
E-mail: ky6a@nih.gov

Abstract

Citation Yoshinaga K. Research on Blastocyst Implantation Essential Factors (BIEFs). Am J Reprod Immunol 2010

Blastocyst implantation is a process of interaction between embryo and the uterus. To understand this process, this review tries to summarize what blastocyst implantation essential factors (BIEFs) play what roles, as well as where in the uterus and at what stage of implantation process. Addition of more new data to this kind of compilation of information will help the development of diagnosis and treatment of infertility caused by implantation failure. The major, important cells of the endometrial cells that interact with invading blastocyst (trophoblast) are luminal epithelial cells, stromal cells (decidual cells) and resident immune cells. BIEFs regulate these cells to successfully maintain pregnancy.

Introduction

The blastocyst implantation essential factors (BIEFs) are a collective name given to the molecules that are indispensable for successful implantation of a mature and healthy blastocyst.1 This name was given to conveniently include all molecules that are needed for implantation of a blastocyst into the uterine endometrium, therefore, members of the BIEF family vary depending on the species of animals, on the stage of pregnancy, and on the progress of science as new members will be discovered in the future. In a previous review, BIEFs were analyzed according to the biochemical nature of molecules and most of them were found to modulate the maternal immune system.2 Research on blastocyst implantation has been carried out over one century when we consider Lataste, who discovered implantation is delayed in lactating rodents, is one of the earliest implantation researchers.3 However, our knowledge of implantation is still limited and not sufficient enough to diagnose infertility caused by failure of implantation. Neither do we have a single medication to alleviate infertility because of implantation failure. Because close to 50% of fertilized pre-implantation embryos fail to implant,4 infertility problem because of implantation failure is a serious social problem when the marriage age for women is getting later and later as higher education requires more and more time to achieve. Aging is associated with reduction in fertility in women.5

Need of collaborative research

To alleviate implantation-failure infertility, it is imperative that we gain more information on human implantation and establish diagnostic methods and find treatments of implantation failure. We cannot do research on human implantation directly on human embryo because of ethical considerations. Because we do not have a single animal model that represents the human, we have to obtain pieces of information obtained in other species of animals, non-pregnant humans, and cell-line cells. The obtained pieces of information need to be integrated by expert investigators and human implantation process may be extrapolated from the obtained information. This is the reason why implantation research must be carried out in a collaborative way by means of interdisciplinary approach.6 The National Institutes of Health has been encouraging such a collaborative team work by issuing a program announcement [NIH PA-07-445],1 and a group research by Collaborative Team on Interdisciplinary Research on Blastocyst Implantation was initiated in 2009. The ultimate goal of this collaborative team effort that began in May 2009 is to understand how human implantation takes place so that infertility because of implantation failure may be diagnosed and treated.

Systemic versus local BIEFs

This review will describe the process of blastocyst implantation, step by step, using model animals (mainly rodents) and point out information we need. Attention is paid to focus the need in immunological information that is one of the gaps in our knowledge of implantation. As BIEF molecules appear in this text, they are written in bold letters. It is important to differentiate two groups of BIEFs: the ones that work systemically and the others that work locally. Because the local BIEFs act through paracrine, juxtacrine, or autocrine mechanisms, this classification will facilitate understanding of function of BIEFs. Some of the molecules expressed in the luminal epithelium in the entire uterus during the receptive state are systemic BIEFs. On the other hand, local BIEFs are expressed only locally at specific areas in relation to the implanting embryo. Leukemia inhibitory factor (LIF) is an example of systemic BIEF as it is expressed most abundantly in the uterine glands of entire uterus on day 4 of mouse pregnancy and is considered to regulate the growth and implantation of blastocyst.5 In the LIF-null mice, the luminal epithelial surface fails to produce pinopods,6 thus LIF appears to play an important role in the loss of polarity that occurs in the receptive luminal epithelial cells. Estrogen, estrogen receptorα(ERα), progesterone and its receptor (PR), Hoxa-10, and amphiregulin are all systemic BIEFs. Expression of LIF depends on estrogen action, and LIF is expressed in the uterine glands in the entire uterus. Its expression in pseudopregnant mice is similar to that in pregnant mice,7 thus, LIF is a systemic BIEF. Members of the IL-6 family, LIF, and interleukin-11 (IL-11) have been reported to play important role in implantation in human.8–10

On the other hand, the gene for heparin-binding epidermal growth factor-like growth factor (HB-EGF) is expressed in the mouse uterine epithelium surrounding the blastocyst 6–7 hr before the attachment reaction that occurs at 2200–2300 hr on day 4 of pregnancy.11 This HB-EGF expression appears to be a response of the epithelial cells to soluble material(s) emanating from the active blastocyst.12 Blastocysts expand and collapse rhythmically emanating the blastocoele contents during the peri-implantation period.13 Characterization of the blastocoele material(s) that stimulates expression of HB-EGF gene in the luminal epithelium surrounding the active blastocyst would be an interesting project. HB-EGF is not the only local BIEF, but other members of EGF family such as epiregulin and amphiregulin also are local BIEFs. Amphiregulin is, however, expressed as systemic BIEF early on day 4 of pregnancy in the mouse, but it again is expressed as a local BIEF later. Because these EGF family members are not expressed in LIF-null mice, their action mechanism appears to be in the down-stream to LIF of the signal transduction system.14Lysophosphatidic receptor (LAP3) is needed not only for blastocyst spacing, but also for timely implantation. Its expression is limited to endometrial luminal epithelium during early pregnancy.15 Chan et al.16 showed that this molecule mediates chemotaxis of immature dendritic cells (DCs) to unsaturated lysophosphatidic acid. Blois et al.17 examined DCs in uterus at different stages of pregnancy and found that number of DCs significantly increased between days 3.5 and 5.5 and remained at high levels during pregnancy. Krey et al.18 investigated the role of DCs during implantation process using a transgenic mouse system that allows transient deletion of CD11c+ cells in vivo. They observed that a temporal deletion of DCs affected maturity of uterine natural killer (uNK) cells, decidual vascular development, DC-dependent protein levels, implantation process, and placental development. Thus, DCs appear to make major contributions to the implantation process.

Endocrine background of implantation

The ovulatory cycle of small rodents such as mice and rats is named as ‘incomplete cycle’ because the newly formed corpora lutea are not fully functional. The next set of follicles grows and ovulates in a few days. Thus, ovulation takes place every 4–5 days. When mechanical stimulation is applied to the vaginal cervix at the time of mating, the cyclic secretion of pituitary gonadotropins stops and the anterior pituitary starts secretion of prolactin (luteotropin in these small rodents). The post-coital secretion of prolactin occurs twice daily, and prolactin stimulates the newly formed corpora lutea and starts secretion of progesterone for a prolonged period of time (12–13 days), during which no ovulation takes place.

When the uterine endometrium is exposed to progesterone for longer than 48 hr, the endometrium is brought to the ‘pre-receptive or neutral state’ of the uterine sensitivity for implantation. In case of rats, the endometrium on day 4 after mating is already in the pre-receptive (neutral) state. In the afternoon of day 4 of pregnancy (day 1 is the sperm positive day), there is a small rise in estrogen secretion (potentially due to the effect of remnant cyclic pituitary gonadotropin secretion pattern). This small rise in estrogen secretion together with continuous secretion of progesterone triggers the essential change in the pre-receptive (neutral) endometrium into the ‘receptive state’ for accepting a healthy, mature blastocyst to implant. Implantation does not take place in the endometrium in the neutral state nor in the refractory state. However, the neutral state is different from the refractory state, the difference being the neutral state has potential to become the receptive state when an appropriate dose of estrogen becomes available, but the refractory state does not have that capacity, unless progesterone effect is removed from the endometrium of the refractory uterus and primed again with progesterone for more than 48 hr.

The function of young corpora lutea that started progesterone secretion under the influence of prolactin is in a specific condition from the endocrinological and immunological points of view. Erlebacher et al.19 ligated (neutralized) CD40 by injecting antibody against CD40 to mice daily between day 4 and 7 of pregnancy. In contrast to control treatment with rat IgG, treatment with CD40 antibody resulted in pregnancy failure. Analysis of the cause of this pregnancy failure revealed that serum levels of progesterone and prolactin were significantly lowered in mice treated with CD40 antibody. Supplement with progesterone or prolactin overcame the ill-effect of CD40 antibody. This experiment clearly demonstrates that the immune system influences systemically the endocrine system at the pituitary–ovarian axis during early stages of pregnancy. This ill-effect of CD40 antibody on pregnancy is not observed in later stages of pregnancy. Implantation takes place during the night of day 4 of pregnancy in the mouse and of day 5 in the rat. So, we have to be aware of the endocrine system that regulates successful implantation under a balanced condition between the endocrine and immune systems where the immune system is suppressed, and this balance appears to be maintained by many factors that include BIEFs.2

Embryo spacing mechanism

Rat embryos develop into the morula stage in the oviduct and enter the uterus in the afternoon of day 4 of pregnancy. By the evening of day 5, these embryos develop to the blastocyst stage. By 17:00 hr on day 5 in the rat,20 blastocysts spaced rather evenly along the entire uterine horn. Because this spacing of blastocysts takes place in rats and mice ovariectomized after mating and supplemented with progesterone alone, the spacing is regulated by progesterone and not by the pre-implantation increase in estrogen. Spacing of blastocysts takes place during the neutral state, and if there is a delay in implantation, the blastocysts remain unattached at spaced positions. The mechanism of blastocyst spacing has been reported to be regulated by a noradrenergic transmission via action on myometrial post-synaptic α1-adrenoceptors is involved as a regulatory mechanism.20 LPA3 appears to be involved in muscle contractions that serve spacing blastocysts.21 There are multiple pieces of evidence that lipid is also involved in blastocyst spacing in the mouse. Genetic mutations of female mice with aberrant gene for lipid-related molecules resulted in disturbance of blastocyst spacing: these molecules being cytosolic phospholipase A2α (cPLA2α),22LPA3,15,21 and adrenomedullin.23 How lipid affects blastocyst spacing? A recent study by Lehman’s group24 revealed that lysophospholipids are involved in germ cell migration in drosophila. These investigators emphasize that lysophospholipids play important roles in the migration of lymphocytes, smooth muscle cells, and germ cells. They showed that germ cells in drosophila migrate because the somatic cell under the germ cell express wun and wun2. These two lipidphosphate phosphatases, wun and wun2, are thought to degrade extracellular substrates and to act redundantly in somatic cells to provide a repulsive environment to steer repelling the migration of germ cells. If free blastocysts are in cluster and each blastocyst would be repelled by underlying epithelial cells, these blastocysts would continue to move back and forth and would settle at the least repulsive position. As the availability of estrogen is gradually increased, the repulsive activity of the luminal epithelial cells may decrease and blastocysts would stop moving. The luminal epithelium of the neutral state contains high levels of LPA3, and it abruptly drops when the endometrium becomes receptive.15 Thus, LPA3 may well be involved in the spacing mechanism. Paria et al.25 examined effects of growth factor-soaked beads on spacing of native blastocysts and their implantation. One interesting observation was that beads containing BMP2 or BMP4 affected spacing of native blastocysts. Although we do not know the mechanism of action of these proteins on spacing, these proteins are phospholipid-binding proteins, and thus, there still remains possibility of relating the spacing mechanism to phospholipid molecules. It is fascinating to speculate whether lysophospholipids are involved in blastocyst spacing in that blastocyst at the immediately preceding implantation is repelled from a ‘refractory’ region of the luminal epithelium caused by the sitting blastocyst by means of lipidphosphate phosphatase action. If it is true, the Mossman theory for blastocyst spacing mechanism may be partially correct in the sense that blastocyst creates the ‘refractory zone’ from the attachment site and the inhibitory effect gradually diminishes away from the attachment site (McLaren & Michie26).

Apposition of blastocyst with luminal epithelium

The morphology of rat endometrial luminal epithelium during delayed implantation (equivalent to the neutral state) is characterized by tall columnar cells with short microvilli and a distinct zonation of organelles.27 The apical surface of the luminal epithelium is covered with mucin, a major protein being mucin1 (Muc1) that prevents direct attachment of the blastocyst to the luminal epithelium in the mouse, and removal of this mucin layer is needed to allow direct blastocyst attachment to the luminal epithelium.28 Enders and Schlafke29 considered that implantation starts when a fixed position of the blastocyst in relation to the uterus has been established. This occurs by the evening of day 5 in the rat. These investigators pointed out that this is the earliest time at which the blue reaction can be elicited. The blue reaction is a simple method to detect the implantation site where systemically injected blue dye such as Pontamin blue that is bound to macromolecule protein such as albumin leaks out through the capillaries where their permeability was increased because of some local factor(s) (Psychoyos30). There is no report whether this capillary permeability change involves any local immune reaction. Here is a point we need to investigate further as to whether any specific immune cells and molecules are involved in the pre-implantation increase in capillary permeability.

During this stage, decidualization of the stromal cells is barely beginning, and the blastocyst is almost entirely encompassed by the luminal epithelium and is situated close to the antimesometrial end of the uterine lumen with the embryonic pole oriented mesometrially. The stroma of the mesometrium is edematous as estrogen secretion increases at this time. The lateral margins of the uterine lumen are apposed. This first stage of implantation was named by these investigators as ‘the apposition stage’.29 Potts31 divided the attachment phase of blastocyst into two phases: the first phase is characterized by microvilli (the neutral state) and the second phase is characterized by flattening of the epithelial cell surface (the receptive state). Another prominent surface change is the abundant formation of surface protrusions (pinopods, uterodomes) (Psychoyos & Mandon32; Murphy33). Although their formation definitely results in direct contact of the epithelial cells with trophoblast at these areas, functional significance of pinopods has not been clarified yet.

Preimplantation embryos render the endometrium receptive for its implantation. For example, the blastocyst secretes chorionic gonadotropin (CG), and this hormone changes the endometrium and form the ‘plaque’ in the baboon. Jones and Fazleabas34 infused hCG into the baboon uterus and induced the plaque formation demonstrating that blastocyst signal CG modulates the uterine environment prior to implantation. It has been also shown that hCG stimulates invasion of JEG-3, trophoblastic choriocarcinoma cell-line cells, through matrigel-coated filter in vitro.35 This latter study provides additional supporting evidence that CG facilitates blastocyst’s own implantation in the endometrium.

During the first 3.5 days of mouse pregnancy, subapical E-cadherin in the epithelium gradually increases, but there is an abrupt loss of lateral cell adhesion between days 3.5 and 4.5 in uterine luminal epithelium, which is dependent on the pre-implantation estrogen increase.36 Lymphocytes are commonly found between the bases of luminal epithelial cells in the pre-implantation stage in the rat.29 These intraepithelial lymphocytes seem to disappear from the epithelium as the epithelial cells lose their polarity and become receptive. Their observation suggests that the epithelial cell polarity change and associated cell condition regulate migration of lymphocytes from the epithelium at this stage of pregnancy. Expression of mRNA for LPA3 in the luminal epithelium peaks on day 3.5 and abruptly falls on the following day.15 Although LPA3 is a chemoattractant to immature DCs, we do not have information on distribution of DCs in the uterus at this stage of pregnancy. However, in vitro studies using peripheral blood suggest that immune cells may regulate human uterine receptivity,37 and splenic cells have been shown to advance implantation in mice.38 Thus, it is very interesting to consider that there may be interactive relationship between the receptive luminal epithelium and the immune cells in the endometrium.

Adhesion of blastocyst to the luminal epithelium

Uterine receptivity

The apposition stage of the implantation is followed by the adhesion stage. As stated in the foregoing section, there are several morphological changes in the apical surface of the endometrial luminal epithelium when the neutral state becomes the receptive state. Together with the flatting of the apical surface of the luminal epithelium and pinopods formation, the epithelial cells show the signs of polarity loss. Once the epithelial cells become receptive, they cannot go back to the neutral state, without depleting progesterone influence from these cells. The receptive state lasts for a short period of time and these cells automatically become refractory. While the receptive period lasts <24 hr in small rodents, it lasts for 3 days in humans as the conceptus implants 8–10 days after ovulation (Wilcox et al.39), this period is named the ‘Window of implantation’.40 Global gene profiling in human endometrium during the window of implantation41 showed that there are a number of genes up-regulated during the window of implantation and some others down-regulated. Among those genes up-regulated were important for cholesterol transport, PLA2, enzymes that catalyze the hydrolysis of membrane glycol-phospholipids important for generation of lipid signals, etc. It is worth to notice that a number of genes involved in immune modulation increased during the window of implantation. Compared with humans, the window of implantation is much shorter in rodents: it is <24 hr in the evening of day 4 in mice and in the evening of day 5 in the rat. Markers of the receptive endometrium in women and mice have been well summarized by Sharkey and Smith.42 Many of the molecules involved in the receptive endometrium are produced by its constituent cells, i.e., the luminal and glandular epithelial cells, stromal cells, and at the site of implantation by embryonic (trophoblast) cells. Some molecules contributing to the receptivity appear to be produced by resident immune cells, which have waited for establishing pregnancy.43 Although these authors found the endometrial natural killer (NK) cells differentiate when the stromal cells decidualize, they did not see difference in NK cells between proliferative phase and secretory phase. Lobo et al.44 demonstrated up-regulation of immune-related genes such as decay accelerating factor (DAF), indoleamine 2,3 dioxygenase (IDO), interleukin-15 (IL-15), IL-15 receptor alpha subunit (IL-15Rα), interferon regulatory factor-1(IRF-1) and natural killer-associated transcript 2 (NKAT2) in secretory endometrium. Their function is considered to be related to immune tolerance. These workers localized uNK cells scattered throughout the stroma in secretory endometrium, and expression of NKG5 (a transcript of cytotoxic protein of NK cells) was found strong in NK cells surrounding the glandular epithelium. At present, we have no information as to whether or not any specific types of immune cells localized near the luminal epithelium participate in transforming the luminal epithelium from the neutral state to the receptive state.

Adhesion BIEFs

During the adhesion stage of rat implantation, stromal cells surrounding the blastocyst decidualize to form the primary decidual zone that grows from the antimesometrial end toward mesometrial direction, resulting in formation of the implantation chamber (Enders & Schlafke29). The trophoblast cells on the lateral side of the blastocyst adhere to the luminal epithelial cells and junctional complex may be formed in between these cells (Potts31). The trophectoderm of the active blastocyst expresses on its surface specific proteins such as heparan sulfate proteoglycans (HSPG),45L-selectin, and trophinin. Human trophoblasts express L-selectin little before hatching, but it becomes more intense after hatching out of the zona pellucida. Mouse trophectoderm also expresses L-selectin, and the receptive uterus expresses carbohydrate ligand epitopes. Thus, the adhesion of active blastocysts to the receptive endometrial surface uses the same or a similar mechanism as that of blood cells to the vessel wall (Genbacev et al.46). Trophinin is a hemophilic adhesion molecule and is expressed in trophectoderm in monkeys and in extra-embryonic and maternal cells at monkey and human embryo implantation sites. Trophinin-mediated cell adhesion induces tyrosine phosphorylation and actin reorganization. Trophinin-mediated cell adhesion acts as a molecular switch for ErbB4 activation, and thus, involved in HB-EGF action in murine implantation (Sugihara et al.47). In addition to these molecules, integrins (α4β1andαvβ3) have been suggested to have important roles in the process of implantation and decidualization (Lessey et al.48). Colony-stimulating factor-1 (CSF-1) is synthesized under the influence of ovarian hormones, and its mRNA was detectable in the uterine epithelium just before implantation and increased gradually to peak at mid-pregnancy.49

Invasion of trophoblast into the endometrium

When stromal cells decidualize, the overlying luminal epithelial cells are destined to become apoptotic and gradually loosened from the basement membrane by trophoblast cells. The first sign of penetration of luminal epithelium is observed in the morning on day 7 of rat pregnancy. In rats and mice, the luminal epithelium is invaded from the lateral side of the blastocyst and the fragmented epithelial cells are phagocytosed by trophoblast cells (Enders & Schlafke29). Immune cells do not appear to be involved in phagocytosis of the luminal epithelial cells at this stage of trophoblast invasion.

After penetration through the luminal epithelium, the basement membrane is breached. It is decidual cells that penetrate first the basement membrane, not trophoblast, in the rat (Schlafke et al.50). So the break-down of the basement membrane at this stage may be from both sides of this membrane, trophoblast, and decidual sides. The mechanism for penetration of basement membrane by trophoblast is not clear. There are, however, in vitro studies suggesting involvement of ligand–receptor interactions. Sutherland et al.51 showed the surface of trophoblast contains receptors for extracellular matrix components, and ligand-receptor interaction of these components is involved in trophoblast attachment and outgrowth in vitro. Liotta’s hypothesis of tumor metastasis is that penetration of the basement membrane by tumor cells involves laminin receptor production in the invading tumor cells, and tumor cells anchor themselves to the basement membrane by means of ligand–receptor binding mechanism. The anchored tumor cells then secrete hydrolytic enzymes to degrade the basement membrane locally.52

During decidualization, sphingolipid metabolizing enzymes are up-regulated, suggesting sphingolipids are involved in lipid signaling molecules during early pregnancy,53 and disturbance of the pathway by double knock-out of its kinase genes resulted in early pregnancy loss.54 Because sphingosin-1-phosphate (S1P) plays important roles in angiogenesis and also in migration of lymphocytes, detailed studies on the roles of S1P in successful implantation and early placental development are awaited.

Decidua and immune cells

There are important reports that indicate decidualization of endometrial stromal cells has profound influence on immune cells. Erlebacher’s group demonstrated in mice that decidualization of stromal cells entraps DCs and prevents their migration to the lymphatic vessels of the uterus, thus reaching the draining lymph nodes (Collins et al.55). Presence of DCs in the uterus is also essential for decidualization of stromal cells. Depletion of uDCs on day 3.5 resulted in pregnancy failure on day 5.5 because of failure in decidualization.56 Blois et al.17 observed DCs in mice during pregnancy. It is interesting to observe that the ratio of immature/mature CD11 c+ cells changed significantly between gestation day 3.5 and 5.5 and that change is maintained through day 8.5. Decidua from healthy women at 6–9 weeks of gestation contained four different types of mature DCs with morphological functional features of typical mature DCs.57

Saito et al.58 reviewed immune cells that play important roles in fetoplacental tolerance and concluded that decidual immune cells are different from peripheral immune cells in that transforming growth factor-β (TGF-β) and IL-10 secreting cells must dominate to maintain pregnancy. The dominant lymphocytes in implantation sites are pregnancy-associated uterine NK cells, and decidual cellularity and integrity seem to depend on uNK cells that secrete INF-γ, and this cytokine initiates pregnancy-induced remodeling of decidual blood vessels.59 Aluvihare et al.60 demonstrated that regulatory T cells are essential for suppressing maternal immunological rejection of the conceptus. These data support the notion that immune cells and decidual cells interact with each other to promote successful maintenance of pregnancy. Galectin-1, a member of the family of glycan-binding proteins, is expressed in decidua61 and plays a pivotal role in fetomaternal tolerance by balancing the cytokines in favoring those cytokines for maintenance of pregnancy.62

Trophoblast and Immune cell

Decidual cells or placental tissue cells were thought to regulate NK cells.63 Soares’ group64 studied proteins of the uteroplacental prolactin family. They identified one of this family members, prolactin-like protein A (PLP-A), is expressed by trophoblast cells, and this protein binds one of the NK cell surface proteins, gp42, providing evidence that the embryonic-maternal communication is mediated by PLP-A signaling pathway.65 Further study of PLP-A interactions with NK cells revealed that PLP-A suppresses the ability of NK cells to produce interferon-γ (IFN γ), a key mediator of NK cell function.66 Therefore, there exists fine tuning between trophoblast invasion and maternal decidual cell reaction to the invading trophoblast through adjustment of cytokine/hormone secretions to maintain pregnancy.

In early pregnancy, the decidua is densely infiltrated by maternal NK cells with a distinct phenotype (CD56bright CD16),67 and these NK cells are in direct contact with fetal trophoblast.68 King et al.69 carried out a series of cytotoxicity studies and concluded that trophoblast cells appear to escape decidual NK cell lysis the reason being HLA-E is expressed on trophoblast cells and HLA-E acts as a sentinel molecule on behalf of the other major histocompatibility complex (MHC) class I molecules to protect cells against NK attack. In the rhesus monkey, another non-classical MHC class I molecule, a putative homolog of HLA-G (Mamu AG), plays an important protective role in coordinated interaction between endometrial and fetal tissues.70 In cattle, the majority of embryos created by somatic nuclear transfer are immunologically rejected by the transferred uterus due, in part, to inappropriate expression of trophoblast MHC class I antigens.71 These studies indicate that ‘taming’ of MHC class I molecules on trophoblast cells is important for successful implantation.

Trophoblast-decidual interactions

Once the basement membrane is breached, trophoblast cells will have direct contact with decidual cells and other components of stromal tissue. Decidual cell processes penetrate the residual basement membrane that underlines luminal epithelium. Decidual cells, thus, have direct contact with trophoblast. The decidual cells that juxtaposed with trophoblast undergo apoptotic degeneration (Welsh & Enders72). Because decidual cells produce laminin, trophoblast–decidual cell interaction may involve laminin–laminin receptor binding during the invasion of trophoblast into decidual tissue. Loke et al.73 showed that first trimester human trophoblast attaches to areas of a culture dish overlaid with laminin, in preference to type IV collagen and to BSA. In human implantation, maternal decidua expresses high levels of tissue inhibitors of matrix metalloproteinases (MMPs) and insulin-like growth factor-binding protein (IGFBP)-1. Invading trophoblast expresses matrix MMPs to facilitate degradation of maternal tissues. Thus, tissue inhibitors of MMPs regulates trophoblast invasion by inhibiting these enzymes. IGFBP-1 binds α5β1 integrin in trophoblast. Thus, trophoblast and decidua interact with each other by adjusting the degree of invasiveness by use of these molecules.74

To overview what is the decidual response to trophoblast signals, Hess et al.75 examined the effects of human trophoblast secretions on global gene expression in in vitro decidualized human endometrial stromal cells. The data obtained in this study showed up-regulation of genes encoding cytokines, chemokines, angiogenic factors, and others, and down-regulation of genes regulating stromal miosis and the decidual phenotype among other processes. This type of approach provides us with a general idea as to which molecules to be studied further. Multidisciplinary (morphological, cell biological, biochemical, physiological, immunological, etc.) studies on these molecules are essential to gain enough information to apply these basic pieces of information to develop diagnosis and treatment of infertility because of implantation failure.

Metrial glands [Mesometrial lymphoid aggregates of pregnancy (MLAps)]

In the rodent uterus, there is a specialized region named ‘the mesometrial triangle’ on the mesometrial side of each discoidal placenta. This area is important for supply of rich vascularization for full function of the discoidal placenta during pregnancy. The arteries of the mesometrial triangle branches off the mesometrial arteries, dilate markedly, and follow tortuous course to supply blood to the deciduas basalis. Rodents’ arteries in this region are considered equivalent to the human spiral arteries.76 Because the process of implantation starts at the apposition of the blastocyst with the luminal epithelium and ends with invasion of spiral arteries by the trophoblast cells, discussion of the metrial glands will conclude this review.

Accumulation of lymphocytes in the mesometrial triangle starts on day 6 of pregnancy and continues to reach at maximum at mid-gestation. As pregnancy progresses, the accumulation becomes prominent and lymphocytes are granulated. This lymphocyte accumulation was used to be called ‘the metrial gland.’ However, Parr et al.63 identified mouse granulated metrial gland cells as originated from activated uNK cells. Coincidentally with blastocyst implantation and decidualization, these NK cells become activated, proliferate, and produce IFN γ, perforin, serine esterase and other molecules77,78. Deletion of these cells resulted in severe compromise in pregnancy as proper formation of vascular system was interfered (Guimond et al.79).

Interleukin-15 (IL-15) was found to regulate the differentiation of granulated metrial gland cells in uterus of pregnant mouse.80 In this report, IL-15 is probably produced by macrophages and is capable of inducing the expression of perforin and granzymes in pregnant uterine tissues explanted in vitro. It is difficult to judge these results actually reflect the behavior of these NK cells in normal pregnant uterus in vivo. Ashkar et al.81 studied development of NK cells and their results indicated that uNK cells development and maturation share some aspects with NK cells development in other tissues, but also display distinctive tissue-specific regulation. These authors demonstrated that IL-15 is absolutely essential for the support of NK cell differentiation in the decidualizing uterus. They found that IL-15 null mice revealed a complete absence of uNK cells, poor development of deciduas, failure of spiral artery modification, and absence of metrial glands (MLAps).

In their effort to find the reason why the metrial glands are rich in progesterone receptors, Martel et al.82 reached the conclusion that the metrial glands are strictly progesterone-dependent and the high PR concentration could be because of constitutive expression. Arck et al.83 reviewed progesterone effects on immune cells during pregnancy. Readers are referred to this review for detailed accounts of progesterone actions. The recent research outcome convinces us that progesterone action directly and indirectly via other molecules on the immune system is one of the major mechanisms that maintain pregnancy in many species of animals, and further research in this area is warranted to enable us to devise diagnostic methods and treatment of infertility caused by implantation failure.

Conclusion

In this review, I tried to illustrate the process of blastocyst implantation, step by step, and point out what BIEFs play what roles as well as where at what stage of the process. Although there are a number of important discoveries in this field, the information is not sufficient enough to answer these questions satisfactorily. We should continue to try to accomplish this objective to gain sufficient information that will help alleviate the infertility caused by implantation failure by learning how BIEFs work on target cells during the process of implantation, and, at the same time, we can establish the basic knowledge of implantation. Our goal of implantation research is to create a similar compilation of full data on human implantation that will be useful for development of diagnosis and treatment of infertility caused by implantation failure. Finally, it should be emphasized that the uterine cells, especially endometrial epithelial cells, endometrial stromal (decidual) cells, and resident immune cells, play major important roles in accepting the implanting blastocyst, and BIEFs are the molecules that regulate these cells to successfully maintain pregnancy.

Ancillary