Implantation of the conceptus and development of a placental connection to the maternal circulation are strong evolutionary advantages of eutherian mammals. The placenta (chorion, allantois, and amnion) forms the interface between the microcirculatory systems of the mother and conceptus and functions for efficient, sustained, and high-throughput exchange of nutrients, respiratory gases, and metabolic wastes, and it protects the growing embryo/fetus and is a source of hormones. Due to its recent appearance in the evolutionary record, a considerable variability exists among species relative to histogenesis and organization of the placenta (Carter and Enders 2013). The placental membranes are formed from components of the blastocyst and embryo; however, both the embryo and maternal endometrium begin to form components of the placenta as soon as the conceptus trophectoderm attaches to the endometrium. Indeed, the placental trophoblast interaction with maternal tissues remains extensive in all species. Blastocysts of some species, including primates and rodents, are invasive and penetrate the epithelial layer of the endometrium. Pigs, however, employ a vastly different strategy, and demonstrate a true epitheliochorial placentation in which there is no displacement or invasion of the maternal epithelium and the conceptus remains within the uterine lumen throughout gestation (Burton 1992). The changes that occur at the interface between trophectoderm and uterine LE during the initial stages of epitheliochorial placentation in pigs, day 13–26, have been elegantly described by Dantzer (1985). Throughout implantation, the glycocaylx that extends from the apical surface of the uterine LE is thicker than the glycocaylx at the surface of conceptus trophectoderm. On days 13 and 14, the uterine LE develops protrusions that become enclosed by caps of trophectoderm cells that serve to physically immobilize the conceptus; and by day 14, there is close apposition between the apical plasma membranes of trophectoderm and uterine LE cells. Interdigitating microvilli form between these plasma membranes through days 15 and 16, and then, the interface becomes increasingly complex as it functionally transitions from histotrophic to histotrophic and hemotrophic nutrient transport over day 15–20. This transition is characterized by the development of apical domes on the uterine LE that are closely related to the trophectoderm and provide long cytoplasmic extensions into a luminal space between the apical domes. Finally, adhesion transitions into placentation through ever-increasing development of interdigitating microvilli between trophectoderm and uterine LE that extends into the peripheral zone by day 26 of gestation.

The term “implantation” is somewhat of a misnomer for the pig but is, nevertheless, used to describe the initial stages of placentation in this species. Despite differences in duration of the pre-implantation period, protracted in the pig, and type of implantation, the initial stages of implantation/placentation are common across species and are characterized as the “adhesion cascade for implantation” (Dantzer 1985; Guillomot 1995; Burghardt et al. 2002). The phases of this adhesion cascade in pigs include (1) shedding of the zona pellucida and elongation of the conceptus trophectoderm, (2) precontact and conceptus trophectoderm orientation to the uterine LE, (3) apposition of trophectoderm to uterine LE, (4) adhesion of the apical surface of trophectoderm to the apical surface of uterine LE, and (5) development of interdigitating microvilli between trophectoderm and uterine LE (Fig. 8.3). As this cascade concludes, adhesion seamlessly transitions to the progressive formation of epitheliochorial placentation that supports fetal–placental development throughout pregnancy (Johnson et al. 2014).

During the peri-implantation period of pregnancy, uterine LE and conceptus trophectoderm develop adhesion competency in synchrony to initiate the adhesion cascade within a restricted period of the uterine cycle termed the “window of receptivity” (Fazleabas et al. 2004; Spencer et al. 2007; Bazer et al. 2011). Similar to other species, this window is orchestrated through the actions of progesterone and estrogen to regulate locally produced cytokines, growth factors, cell surface glycoproteins, cell surface adhesion molecules, and extracellular matrix (ECM) proteins in the pig (Johnson et al. 2009). Conceptus synthesis and release of estrogens modulate uterine gene expression (Johnson et al. 2009). The importance of estrogen to implantation of pig conceptuses is underscored by the fact that premature exposure of the pregnant uterus to estrogen on days 9 and 10 results in degeneration of all pig conceptuses by day 15 (Ashworth et al. 2006). Progesterone has a clear role in initiating the adhesion cascade for implantation in pigs; however, the mechanism by which progesterone regulates uterine epithelial cell functions remains a paradox. As long ago as 1973, studies of endocrine regulation of expression of uteroferrin secreted by the uterine epithelia of pigs revealed a requirement for long-term treatment with progesterone indicating that effects of progesterone are not mediated by a “classical” steroid receptor-type mechanism of action (Knight et al. 1973). Based on current evidence, long-term treatment with progesterone is required in order to downregulate PGR in uterine epithelia as a prerequisite to epithelial cell proliferation, gene expression, and differentiated functions. Similar to all species studied, progesterone downregulates expression of PGR in the uterine epithelia of pigs after day 10 of pregnancy, immediately prior to the time when the endometrium becomes receptive to implantation (Geisert et al. 1994; Bazer et al. 2008; Bailey et al. 2010). As previously indicated, uterine stromal cells express PGR throughout gestation, suggesting that effects of progesterone on PGR-negative uterine epithelia may be mediated through an indirect pathway whereby PGR-positive stromal cells stimulate expression of a progestamedin(s) that, in turn, modulates function of epithelial cells (Bazer et al. 2012). Although proposed progestamedins such as FGF10 and hepatocyte growth factor are expressed by the uterine stromal cells of sheep, no presumptive progestamedin has been detected in the stroma of pigs, and FGF7 is actually expressed by the uterine LE during the peri-implantation period in response to estrogen after progesterone downregulates PGR (Chen et al. 2000a, b; Ka et al. 2000). An alternative possibility is that downregulation of PGR in uterine epithelia removes an unidentified intrinsic “block” to differentiated functions in these epithelia (Spencer et al. 2004a). Regardless of the specific mechanism involved, downregulation of PGR in uterine LE is temporally associated with alterations in the expression of anti-adhesive components, mainly MUC1, an intrinsic transmembrane mucin within the glycocalyx of LE that sterically inhibits attachment of the conceptus (Brayman et al. 2004). In vivo administration of progesterone to cyclic gilts results in the loss of MUC1 from the apical surface of uterine LE (Bowen et al. 1996). It is accepted that in all mammals, initial conceptus attachment requires this loss of MUC1.

Downregulation of MUC1 exposes potential low-affinity carbohydrate ligand binding molecules including selectins and galectins and perhaps heparan sulfate proteoglycan, heparin binding EGF-like growth factors, cadherins, integrins, and CD44, which are proposed to contribute to initial attachment of conceptus trophectoderm to uterine LE (Kimber et al. 1995; Kimber and Spanswick 2000; Spencer et al. 2004b; Johnson et al. 2014). The involvement of carbohydrate–lectin interactions during the adhesion cascade of pigs has not been investigated. However, it is likely that filamentous porcine conceptuses undergo a series of attach-and-release events between carbohydrates and lectin receptors at the apical surfaces of trophectoderm and uterine LE that result in maximal apposition of these tissues, similar to the “rolling and tethering” that occur during leukocyte adhesion to the endothelium for extravasation of leukocytes (Kling et al. 1992), and proposed for the initial attachment of human blastocysts to the uterine wall (Red-Horse et al. 2004). In support of this idea, other domestic farm species, goats and sheep, exhibit prominent expression of H-type 1 antigens and glycosylation-dependent glycam 1, respectively, at the uterine–conceptus interface during implantation (Powell et al. 2000; Gray et al. 2004). These low-affinity contacts are then replaced by a more stable and extensive repertoire of adhesive interactions between integrins and maternal ECM which appear to be the major contributors to stable adhesion at implantation (Hynes 1987; Ruoslahti and Pierschbacher 1987; Aplin et al. 1994; Burghardt et al. 1997, 2002; Johnson et al. 2001; Lessey 2002). Integrins are dominant glycoproteins in many cell adhesion cascades (Kling et al. 1992). They are transmembrane glycoprotein receptors composed of non-covalently bound α and β subunits that promote cell–cell and cell–ECM adhesion, cause cytoskeletal reorganization to stabilize adhesion, and transduction signals through numerous signaling intermediates (Giancotti and Ruoslahti 1999; Albelda and Buck 1990). There are 18 α- and 8 β-subunits capable of dimerizing to form 24 different heterodimer combinations that can bind to numerous extracellular ligands including a variety of ECM proteins (Albelda and Buck 1990; Humphries et al. 2006; Gallant et al. 2005). Integrin receptors expressed at the apical surface of uterine LE cells are capable of binding to Arg-Gly-Asp (RGD) and non RGD amino acid sequence-containing ECM molecules and bridge to another complement of potential integrin receptors expressed at the apical surface of conceptus trophectoderm cells.

At present, it is known that eight integrin subunits, α1, α3, α4, α5, αv, β1, β3, and β5, are expressed at the apical surface of both uterine LE and conceptus trophectoderm. Luminal epithelial expression of α4, α5, and β1 increases during the period of maternal recognition of pregnancy, and similar increases in expression of these integrins can be stimulated in the uteri of cyclic pigs through treatment with progesterone. Further, α4, α5, αv, β1, β3, and β5 have been localized to porcine implantation sites on day 12–15 of gestation in pigs (Bowen et al. 1996). These subunits potentially give rise to the integrin receptors αvβ1, αvβ3, αvβ5, α4β1, and α5β1 that likely function as part of an adhesion cascade that serves to generate both stable adhesion between trophectoderm and LE and activation of outside-in signal transduction (Burghardt et al. 1997). Integrin-mediated adhesion results in dynamic macromolecular complexes, termed focal adhesions (FAs) which are composed of heterodimeric transmembrane integrin receptors that connect ECM proteins to the actin cytoskeleton along with a diverse array of cell signaling intermediates (Sastry and Burridge 2000; Wozniak et al. 2004; Larsen et al.. 2006). It is noteworthy that immunofluorescence staining has revealed that αv and β3 integrin subunits co-localize with altered distribution of talin (an intracellular signaling intermediate within FAs) within large aggregates at the junction between trophectoderm and uterine LE on day 20 of gestation which is approximately midway through the process of attachment for implantation and placentation in pigs (Erikson et al. 2009). The size and nature of these aggregates are reminiscent of the well-characterized FAs that form at the base of cultured cells as they attach to ECM on a rigid substrate at their basal surfaces (Sastry and Burridge 2000; Burghardt et al. 2009). Five ligands capable of engaging integrin receptors to induce assembly of focal adhesions have been characterized at sites of implantation in pigs. The inter-α-trypsin inhibitor heavy chain-like (IαIH4) protein contains a von Willebrand type A domain, a recognition site for the αvβ3 integrin receptor (Geisert et al. 1998). The latency-associated peptide (LAP) of transforming growth factor beta (TGFB) binds to αvβ1, αvβ3, and αvβ5 (Massuto et al. 2009a). Fibronectin is capable of binding α4β1, α5β1, and αvβ3 (Bowen et al. 1996). Vitronectin is a major ligand for αvβ3 (Bowen et al. 1996). Finally, osteopontin [OPN, also known as secreted phosphoprotein 1 (SPP1)] is the most promiscuous of the ligands and it interacts with αvβ1, αvβ3, αvβ5, and α4β1 (Johnson et al. 2003).

The IαIH4 is part of the kallikrein–kininogen–kinin protease system. Both IαIH4 protein expression and kallikrein enzymatic activity increase within the uterine environment during the peri-implantation period of pigs (Geisert et al. 1998; Vonnahme et al. 1999). The IαIH4 can directly engage the αvβ3 integrin receptor, as well as interact with hyaluronic acid within the ECM to aid in the initial stages of implantation. However, it is hypothesized that the primary role of IαIH4 during implantation is to act in concert with bikunin to stabilize the surface glycocalyx of the uterine LE during conceptus attachment (Hettinger et al. 2001). The LAP associates with TGFB to form an inactive homodimer called the small latent complex which remains in the cell until it is bound by another protein called latent TGFB-binding protein to form the large latent complex (LLC) that is then secreted into the ECM (Lawrence 1996). After being secreted, LAP is cleaved from the LLC by proteases to release active TGFB (Jenkins et al. 2006). In pigs, TGFB1, TGFB2, and TGFB3, as well as their receptors, TGFBRI and TGFBRII, are expressed by conceptus trophectoderm and by uterine LE between days 10 and 14 of gestation; and acting through LAP, TGFB increases fibronectin synthesis, cell adhesion to fibronectin, and the formation of FAs in a porcine trophectoderm cell line (Jaeger et al. 2005). In addition, when LAP is infused into the uteri of pregnant pigs, conceptuses failed to implant, suggesting that infused LAP competed with the endogenous LLC for binding to integrins expressed on trophectoderm (Massuto et al. 2009b). In support of this idea, aggregates of LAP, β1, β3, and β5, have been detected at the porcine conceptus trophectoderm–uterine LE interface, suggesting functional adhesion complexes that support conceptus attachment during porcine implantation (Massuto et al. 2009a). Both fibronectin and vitronectin have been detected at sites of conceptus attachment in the pig (Bowen et al. 1996), and the glycosylation variant of fibronectin, oncofetal fibronectin (oFN), is constitutively expressed by porcine conceptus trophectoderm as well as uterine LE throughout gestation (Tuo and Bazer 1996). Fibronectin and vitronectin are prototype cell adhesion proteins, and fibronectin recognizes as many as 10 different integrin receptors to generate different signal transduction functions depending upon the specific integrin receptor involved (Johansson et al. 1997; Humphries et al. 2006). It is noteworthy that the human conceptus trophoblast produces oFN so precisely at sites of trophoblast contact with endometrium that it has been referred to as “trophoblast glue” (Feinberg et al. 1994), and fibronectin is the leading candidate for conceptus adhesion molecule in rodents (Armant 2005).

The most extensively studied ECM adhesion protein for implantation in pigs is osteopontin (Johnson et al. 2003, 2014). Osteopontin is a multifunctional secreted acidic member of the small integrin-binding ligand N-linked glycoprotein (SIBLING) family of ECM proteins (Denhardt and Guo 1993; Butler et al. 1996; Sodek et al. 2000). Regarding implantation, studies have focused on the ability of OPN to support integrin-mediated cell adhesion and cell migration (Senger et al. 1994). It is likely that the expression of no other ECM protein is conserved across different species to the degree that has been observed for OPN, which is expressed abundantly within the conceptus–uterine environment of humans, mice, rabbits, sheep, goats, and pigs (Johnson et al. 1999a, b; Garlow et al. 2002; Apparao et al. 2003; Mirkin et al. 2005; Joyce et al. 2005; White et al. 2006). In pigs, expression of OPN is initially induced by conceptus estrogens in discrete regions of the uterine LE juxtaposed to the conceptus trophectoderm just prior to implantation on day 13, expands to the entire uterine LE by day 20 when firm adhesion of conceptus trophectoderm to uterine LE occurs, and remains high at this interface throughout pregnancy (Garlow et al. 2002; White et al. 2005). In vitro affinity chromatography and immunoprecipitation experiments illustrated that OPN directly binds the αvβ6 integrin heterodimer on porcine trophectoderm cells and the αvβ3 integrin heterodimer on uterine LE cells (Erikson et al. 2009). In addition, OPN binding promotes dose- and integrin-dependent attachment of trophectoderm and uterine LE cells and stimulates haptotactic trophectoderm cell migration, meaning that cells migrated directionally along a physical gradient of nonsoluble OPN (Erikson et al. 2009). Interestingly, immunofluorescence staining of tissue sections of porcine implantation sites revealed FAs containing these same integrins distributed in a similar cell-type-specific pattern to that suggested by in vitro binding to OPN. The αv integrin subunit staining revealed large aggregates at the junction between trophectoderm and uterine LE, suggesting the formation of FAs at the apical surfaces of both conceptus trophectoderm and uterine LE that facilitate conceptus attachment to the uterus for implantation. The β3 subunit, however, appeared in aggregates on the apical surface of uterine LE, but not trophectoderm, which supports results from affinity chromatography studies indicating direct in vitro binding of αvβ3 on uterine LE to OPN (Erikson et al. 2009). Finally, OPN-coated microspheres co-localize with the αv integrin subunit and talin at FAs present on the apical domain of porcine trophectoderm cells in vitro (Erikson et al. 2009). Collectively, results indicate that OPN binds integrins to stimulate integrin-mediated FA assembly, attachment, and cytoskeletal force-driven migration of conceptus trophectoderm cells to promote implantation of conceptuses in pigs.