Miglino et al. (2004) have presented a remarkable comparative study of the placentas of agouti, capybara, guninea pig, paca and rock cavy that gives superb details of the vascular organization of these organs.

8) Extraplacental Membranes

In all rodents, the yolk sac placenta is the second most important maternal/fetal transport organ besides the chorioallantoic placenta. In the guinea pig as in many other rodent species, it was found to play a particular role in maternal-to-fetal protein transfer (e.g. immunoglobulins) (Enders & King 1970; King 1972). This is further supported by the fine structural studies of King (1982) and King & Enders (1970).

The yolk sac placenta has its own fetal vascular supply that is provided by vitelline vessels, derived from the upper mesenteric artery and vein that leave the fetal abdominal cavity via the umbilicus. They accompany the umbilical vessels in the umbilical cord for about half of its length until they branch off (Fig. 10). There is no yolk sac-specific maternal vascular supply, but nutrients and metabolites from the maternal circulation reach the yolk sac via diffusion from the maternal endometrial stroma, closely attached to the yolk sac.

Vitelline transport routes: Consequently, to reach the fetal circulation, maternal immunoglobulins have to pass the following tissue layers:

the maternal endometrial capillary endothelium (paracellularly);
the endometrial stroma (by extracellular lymphatic flow);
the endometrial epithelium (mostly paracellularly);
the yolk sac epithelium (partly paracellularly, partly by receptor-mediated vesicular transport);
the vitelline connective tissue (by interstitial fluid flow and by diffusion);
the vitelline capillary endothelium (partly paracellularly, partly be vesicular shuttle).

Despite of this long and seemingly complicated route, this appears to be an effective transfer route for macromolecules according to the data of King & Enders (1970) and King (1972).

Structural organization: The yolk sac placenta can be subdivided into the following sections with respect to structure and location (Figures 1, 2, 13):
The visceral yolk sac, opposed to the uterine epithelium (for its structure see Figure 6): This appears to be the most effective maternal/fetal exchange area within the yolk sac. Being close to the main placenta (Figure 2), this part of the yolk sac placenta is characterized by a highly columnar yolk sac epithelium that is ultrastructurally equipped with all the organelles required for effective macromolecular transport.

The visceral yolk sac folds insert at the fetal surface of the main placenta. As can be seen from Figures 1 and 2, these yolk sac folds are no longer opposed to the endometrial surface but rather to the surface of the chorioallantoic placenta. Structurally, they are similar to the section described above. In the absence of a direct endometrial source of maternal proteins, this section can only take up proteins from the yolk sac lumen (surrounded by visceral yolk sac, parietal yolk sac and endometrial surface).

The parietal yolk sac (Figure 2): This is a layer of yolk sac epithelium which is directly connected to the lateral surfaces of the main placenta by a kind of basement membrane (Reichert's membrane). The functional purpose of the parietal yolk sac is still obscure. This section of the yolk sac has no fetal vascular supply and it can therefore, not contribute to the maternal/fetal exchange. Also, experimental data by King (1972) revealed considerably reduced uptake of tracer proteins when compared with that of the visceral sections of the yolk sac. It more likely acts as a kind of mechanically important surface epithelium of the main placenta.


9) Trophoblast External to Barrier

Caviomorph rodents are examples among mammals of the most intense trophoblast invasion. This invasion is not restricted to the uterus but it goes far beyond the uterine walls, deeply into the peritoneal cavity. In spite of this excessive invasion, the process is still only poorly studied in the guinea pig and other caviomorphs. The invasion appears to start at and around the subplacenta, another structure specific to caviomorph rodents.

Subplacenta: The subplacenta is formed as the basal part of the chorioallantoic placenta at about day 14. Because of its most basal position, it is the first part of the chorioallantoic placenta supplied by maternal blood channels, and is the latest to be reached by fetal vessels. In fact, in the maternal blood lacunae of the future subplacenta maternal blood flow ceases as early as between days 20 and 27 post conception by formation of blood clot, while establishment of a fetal circulatory system in this area is starts at day 23 and is completed only at day 32 (Uhlendorf & Kaufmann, 1979). Consequently, because of its extreme position, this particular basal tip of the early cone-shaped chorioallantoic placenta (Fig. 3a), is never simultaneously supplied by the maternal and the fetal circulation; therefore, it never participates in fetal/maternal exchange.

Structurally it consists of highly folded layers (Figs. 2; 3a,b; 10) of fetal connective tissue (roof of the central excavation) (Fig. 3a, 10); basally followed by one or more layers of cytotrophoblast, the only cytotrophoblast surviving in the chorioallantoic placenta to the second half of gestation (Fig. 10); at its endometrial surface, the cytotrophoblast is covered by a thick, bizarre-shaped layer of syncytiotrophoblast (Figs. 2, 13); the meshwork of maternal blood lacunae of the latter is perfused by maternal blood only until the end of the third week; thereafter blood flow stops by blood clot and, with the latter organization, the lacunae gradually disappear. From the basal tips of syncytiotrophoblastic folds, root-like syncytial extensions (syncytial streamers) deeply penetrate the endometrium.

At about day 52 degenerative processes start in the subplacental syncytiotrophoblast and later include also cytotrophoblast and connective tissue. On day 60 degeneration of the subplacenta is completed. As a consequence of these processes, the main placenta is connected to the junctional zone only by 3 to 4 bundles of uteroplacental vessels from this date onwards; these are arranged around the large degenerative zone of the former subplacenta. After delivery of the fetus, separation of the placenta occurs across this degenerative zone, so that only the uteroplacental vascular bundles need to be disconnected. Therefore, the resulting postpartum wound area is very small. This fact may support the next conception of guinea pigs as early as within 24 hours after delivery during the so-called postpartum estrus.

The functional importance of the subplacenta is still a mystery. The subplacenta does not serve as a fetal/maternal exchange organ in any stage of pregnancy. By degeneration a few days prior to delivery, it contributes to minimizing the uterine wound area postpartum.

Throughout pregnancy the subplacental syncytiotrophoblast (guinea pig: Kaufmann & Davidoff, 1977; Wolfer & Kaufmann, 1980; chinchilla: King & Tibbitts, 1976) produces large amounts of secretory granules with glycoprotein character which are secreted into the maternal blood lacunae. Because of the absent maternal circulation in the lacunae, the unknown secretory product is accumulated within the subplacenta and set free only during its degeneration. It was speculated that this secretion assists either in separation of the placenta or in postpartum removal of cellular debris and wound healing (Wolfer & Kaufmann, 1980).

As will be described in the following paragraphs, throughout pregnancy the subplacenta is the main, if not the exclusive source of trophoblast invasion. Therefore it may act as the guinea pig equivalent of the cell columns in the villous placentas of primate. Unfortunately, this aspect has not yet been studied in detail since the biology of human cell columns and their role in trophoblast invasion were still largely unknown when most studies on subplacenta and main placenta of the guinea pig were performed.

Interstitial trophoblast invasion: The histological survey picture in Figure 14 illustrates the intensity of trophoblast invasion emerging from the subplacental area. As shown in Figures 2, 3b, 10 and 15, this is the only zone of contact between chorioallantoic placenta and endometrium, and consequently the only potential source for trophoblast invasion. Our own recent re-evaluation of old slides of subplacenta and junctional zone of the guinea pig, the chinchilla, the capybara and the degu suggests the following types and routes of trophoblast invasion in this area (Figure 13):

Invasion of syncytial roots (so-called syncytial streamers) occurs from the subplacenta close toward the myometrium. This process starts concomitant with implantation of the blastocyst. Respective syncytial streamers can be found in the junctional zone until term.

Interstitial invasion of cytotrophoblast begins in the third week of gestation. Its existence requires that the syncytiotrophoblastic shell, separating all cellular components of the chorioallantoic placenta from the endometrium, represented by the syncytial layer of the subplacenta (see above), becomes locally interrupted. This is found most often in the neighborhood of uteroplacental vessels passing the subplacenta and entering the basal surface of the main placenta. At such syncytial discontinuities, plugs of cytotrophoblast protrude into the endometrium. Their cellular phenotype shows certain similarities with cell columns in the human. However, their trophoblast cells as well as the daughter cells that penetrate deeper into the endometrium/junctional have never been studied for molecular signs of invasiveness, to the best of our knowledge.

Abdominal invasion of cytotrophoblast: This is a special feature of caviomorph pregnancies. Trophoblastic cells leave the subplacenta and penetrate the full thickness of the uterine wall, reach the perimetrial surface of the uterine wall and the peritoneal surface of the mesometrium. In this location, invasive trophoblast adopts a mesothelial phenotype and replaces large parts of the original mesothelium as a monolayer of highly flattened cells (Nanaev et al., 1995). Microscopically, the peritoneal cover provided by trophoblast cells can be identified by immunohistochemistry. Macroscopically, it can be identified by naked eye upon opening the peritoneal cavity, since the peritoneal trophoblast surface is less shiny than the original mesothelial surface.

Interestingly, these cells secrete nitric oxide (Nanaev et al., 1995) and thus have a vasodilatory effect on the mesometrial and uteroplacental arteries wherever the trophoblast cells replace the original mesothelium in the surrounding of these arteries (Figs. 13 and 14).

Endovascular trophoblast invasion in the guinea pig (Nanaev et al., 1995) is structurally very similar to endovascular trophoblast invasion in the human (for review see Benirschke & Kaufmann, 2000). As in the case for the human placentation, the point of entry is not yet certain. We suggest that the trophoblast cells pass the arterial walls and enter the arterial lumens where the arteries pass the subplacenta and enter the placental parenchyma (Fig. 14). This is very likely so because (a) of structural findings and (b) since the subplacenta in most stages of gestation is the only potent source of cytotrophoblast. From the point of entry, the trophoblast cells migrate proximally against the blood flow, replacing the endothelium and also replacing parts of the surrounding arterial wall (Nanaev et al., 1995). Whether the mural infiltration is supported by invasive interstitial trophoblast ('intravasation') or depends only on endovascular trophoblast cells ('extravasation') remains to be clarified. It is interesting to note that, different from the classical views, endovascular trophoblast invasion is peripherally always less expansive than is arterial dilatation. Endovascular invasion of the already maximally dilated arteries seems to be a consequence rather than the reason for dilatation.

Pregnancy-induced modification of uteroplacental arteries and their reconstitution following delivery: Mesometrial and uteroplacental arteries undergo a series of dramatic structural and functional changes during and after pregnancy (Nanaev et al., 1995; 2000):

The mesometrial arteries and their uteroplacental continuations into the pregnant segments of the uterine wall are slender tubes (luminal diameters < 0.3mm) during the first two weeks post conception.

As early as in the third week, first invasive trophoblast cells can be found lining the peritoneal surface attached to the uteroplacental arteries. The corresponding arterial segments dilate maximally and reach luminal diameters of 1.5 to 2.0 mm. In spite of their maximal dilatation, structural changes of the arterial walls cannot be observed at this stage. With advancing gestation, peritoneal trophoblast invasion into the mesometrium continues and is accompanied by the progression of arterial dilatation in proximal directions (towards the arcade artery).

At this stage, trophoblastic infiltration of the already maximally dilated arterial wall can be observed only close to the placenta: The trophoblast cells largely replace the endothelium, basal laminas and all elastic elements. Also, smooth muscle media cells appear to disappear. But, as was shown by Nanaev et al. (1995) in the guinea pig, the loss of smooth muscle media cells is caused by dedifferentiation to undifferentiated myoblasts, rather than by destruction. Due to these processes, the arteries are transformed into rigid, inelastic tubes that no longer are responsive to maternal vasoregulation. Only about the distal (placental) half of the mesometrial arteries becomes transformed in this way; the proximal halves are not reached by trophoblast invasion and remain undilated.

Following delivery, the invasive trophoblast cells die and disappear within about five days thereafter. In analogy to human trophoblast invasion (for review see Kaufmann et al., 2003), it can be assumed that trophoblast cells enter the invasive pathway only after leaving the cell cycle; moreover, their post-mitotic, invasive life-span is limited to a few days and thereafter terminated by apoptosis. Therefore, invasive trophoblast unavoidably disappears within a few days after delivery, since the apoptotic cells can no longer be replaced by new generations of post-mitotic daughter cells derived from the placenta. In parallel with the disappearance of the invasive trophoblast cells, arterial myoblasts start re-differentiation into mature and fully active smooth muscle media cells. The luminal lining of the arteries is reconstituted by proliferation of the remaining endothelial cells which replace the dying endovascular trophoblast cells. Eight days after delivery, the mesometrial arteries are fully reconstituted and look like normal arteries.

These events allow quick and complete adaptation of the uterine arterial blood supply to pregnancy conditions in the guinea pig, and to a complete reversal within a few days after delivery. It is still unknown whether the data can be transferred to the human condition. Observations by Catherine Craven point in this direction. This author found in the human that arterial dilation starts prior to infiltration of the arteries by trophoblast (Craven et al., 1998). Moreover, the same authors showed that even after trophoblastic infiltration of the arteries and apparent loss of smooth muscle media cells, alpha-smooth muscle actin could still be detected in the vessel walls. This suggests that the muscle was not destroyed but rather only de-differentiated is also true in the human (Craven & Ward, 1996; for review see Kaufmann et al., 2003).

15) Pathological Features

Due to the clearly defined labyrinthine system of maternal blood lacunae, the guinea pig placenta has been used as a model to induce placental infarction: Systemic treatment with inhibitors of glycolysis (from hours to days), caused severe focal to generalized edematous changes in the syncytiotrophoblast of the interlobium and thereby blockage of venous drainage of maternal blood. Depending on dose and length of treatment, smaller labyrinthine red infarcts or complete degeneration of the placenta were induced (Kaufmann, 1975). In spite of phenotypic similarities with certain infarctions in the human placenta, it remains speculative whether the data obtained with this model can be transferred to the human condition.

Guinea pigs may suffer a variety of skin conditions, intestinal disorders and other diseases that are easily accessible through web sites such as, http://www.oginet.com/pgurney/.

Wynn, R.M.: Electron microscopy of developing decidua. Fertil. Steril. 16:16-26, 1965.