SEARCH

SEARCH BY CITATION

Keywords:

  • endotheliochorial;
  • heterophagous areolas;
  • interhaemal barrier;
  • placenta;
  • Talpidae

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Concluding remarks
  8. Acknowledgements
  9. Authors’ contributions
  10. References

This study provides a contribution to the reconstruction of the eulipotyphlan placental morphotype and also may help resolving a long-standing conflict about the interhaemal barrier in moles. As detailed descriptions of talpid placentation, only available for Talpa europaea and Scalopus aquaticus, led to a controversial debate on the nature of interhaemal barrier, the collection of more placental data of further mole species was strongly desired. Hence, the placentas of six gestational stages of Talpa occidentalis have been studied concerning their morphogenesis and ultrastructure with special focus on the structure of the interhaemal barrier and heterophagous regions. Generally, the mode of placentation in T. occidentalis resembles that of T. europaea, including a broad, discoid, antimesometrial, definitive chorioallantoic placenta of labyrinthine type being still villous in earlier stages. Within the labyrinth, the zona intima shows an endotheliochorial interhaemal barrier with a two-layered trophoblast. This clearly contradicts former statements on the S. aquaticus placenta made by Prasad et al. (1979), although their findings cannot exclude a totally different interpretation. Regardless, the placenta of moles represents the least invasive mode of placentation among Eulipotyphla, which otherwise have highly invasive placentas. Although the phagocytic areolas situated above uterine gland openings are heterophagous, they mainly seem to serve fetal histiotrophic nutrition, at least early in pregnancy. In later stages the number of glands and areolas decreases. This special type of additional phagocytic region is usually most common in species with noninvasive, epitheliochorial placentation, which suggests a correlation between placental invasiveness and the occurrence and type of phagocytic placental structures. The compact and invasive mode of placentation of Talpidae and all other Eulipotyphla seems to be plesiomorphic within Laurasiatheria and is always correlated with an altricial neonate.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Concluding remarks
  8. Acknowledgements
  9. Authors’ contributions
  10. References

The reconstruction of the eutherian and metatherian placental stem species pattern or morphotype is an important approach to a better understanding of their diverse evolution (Zeller & Freyer, 2001). As there is already a sound reconstruction of the metatherian placental morphotype available (Freyer et al. 2003), this study was carried out to provide information useful to achieve the long-term objective of a comprehensive, complementary reconstruction of the eutherian placental morphotype.

Based on their conserved morphological characters, insectivores were largely accepted to be the most primitive group of eutherian mammals for more than 100 years (Huxley, 1880). Recent molecular phylogenetic studies have not only revealed the existence of four eutherian superordinal clades (Laurasiatheria, Euarchontoglires, Xenarthra and Afrotheria), but also found the ‘Insectivora’ to be paraphyletic and, thus, an invalid taxon (Springer et al. 1997; Stanhope et al. 1998). This modern phylogenetic approach assigned representatives of the ‘Insectivora’ to three of these four clades (e.g. Madsen et al. 2001; Murphy et al. 2001a; Douady et al. 2002). The most speciose group of core insectivores, or Eulipotyphla, is formed by shrews, hedgehogs, solenodons and moles (Waddell et al. 1999) and represents the most basal branch of Laurasiatheria (Murphy et al. 2001b; Asher et al. 2009).

As a first step we aim at the laurasiatherian placental morphotype and therefore, focus on the placentas of one eulipotyphlan group, the Talpidae. Their mode of placentation, especially the composition of their interhaemal barrier, became the object of a controversial debate in the past. Strahl, who was the first to give a detailed description of the placenta of the European mole (Talpa europaea) in 1892, postulated an epitheliochorial interhaemal barrier. Only 2 years later, Vernhout (1894) disproved this assumption by reinterpreting the findings of Strahl, and hypothesised a haemochorial interhaemal barrier for the European mole. To address this open question, Malassiné & Leiser (1984) investigated the ultrastructure of one late mole placenta using transmission electron microscopy (TEM) and found the interhaemal barrier to be endotheliochorial in nature. Apart from T. europaea, only the placentation of the American mole (Scalopus aquaticus) has been described in detail so far (Prasad et al. 1979). Prasad’s TEM studies resulted in quite surprising findings indicating the existence of an epitheliochorial definitive chorioallantoic placenta. This seemed untypical of eulipotyphlan mammals at all, as all other studied species showed a more invasive, either haemochorial or endotheliochorial to endothelioendothelial placenta at term (e.g. Hubrecht, 1889; Wislocki, 1940; Wimsatt & Wislocki, 1947; Kiso et al. 1990). Hence, it was desirable to collect more data on the fetal membranes of other talpid species to shed light on that issue.

Thus, for the first time we studied the morphology, morphogenesis and ultrastructure of the placenta of the Iberian mole (Talpa occidentalis), a species endemic to West and Central Iberian Peninsula. Here, special attention was paid to the controversially discussed nature of interhaemal barrier, as well as to heterophagous regions, the ‘areolas’. Such or similar accessory absorptive placental structures are of particular interest, for they seem to play an important role in fetal nutrition. They are not only known from other Eulipotyphla, such as the Indian musk shrew (Siniza & Zeller, 2005), but can be found in various other mammalian species, too (Burton, 1982; Enders & Carter, 2006).

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Concluding remarks
  8. Acknowledgements
  9. Authors’ contributions
  10. References

As moles do not breed in captivity, pregnant females of the species T. occidentalis were captured in the wild in order to obtain the necessary embryonic material. The moles were captured in poplar groves near Santa Fe (Granada province, Southern Spain) in two consecutive winters between 2005 and 2007, using specially designed live traps (Sanchez et al. 1996) under annual permission granted by the Andalusian Environmental Council. Potential pregnancies were detected and gestational status was initially estimated by abdominal palpation (Barrionuevo et al. 2004), although accurate staging was done after dissection, based on values of crown–rump length (CRL) and body mass, and the morphology of major external structures, as described in Barrionuevo et al. (2004). Pregnant females were taken to the laboratory, where they were sacrificed and tissue samples were processed and prepared for shipping. All procedures were approved by the ‘Ethical Committee for Animal Experimentation’ of the University of Granada.

Light microscopy

Eighteen placentas of six different gestational stages from about mid-pregnancy on [14–16, 17, 18–19, 19–21, 21–27, 24–27 days post-conception (dpc)] were histologically and ultrastructurally studied. After removing the embryos, the uterine swellings were fixed in a mixture of alcohol, formalin and acetic acid (AFA) for light microscopical investigation. Then, the placental tissue was dehydrated and embedded in paraffin, sectioned at 8 μm and stained with Haematoxylin & Eosin (HE), Masson-Goldner trichrome or periodic acid Schiff (PAS). Additionally, Quincke’s reaction was applied to selected slides to detect iron deposits in different placental structures. Semithin sections (1 μm) of the material embedded in Araldite were stained with Methylene blue/Azure II and were also used for histological studies. Placental tissue of several gestational stages of T. europaea was taken from the Hubrecht – Hill Collection at the Museum of Natural History in Berlin for comparative purposes. These samples were processed for histological investigation using the same techniques described before.

Transmission electron microscopy

Fresh, small tissue samples were fixed in buffered glutaraldehyde/paraformaldehyde (Karnovsky fixative) and postfixed in osmium tetroxide. After dehydration they were embedded in Araldite. Ultrathin sections (70 nm) were counterstained with uranyl acetate and lead citrate before being examined by TEM (LEO 912 Omega, Carl Zeiss NTS GmbH).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Concluding remarks
  8. Acknowledgements
  9. Authors’ contributions
  10. References

Light microscopy (histology, histochemistry)

Talpa occidentalis has a broad, discoid definitive chorioallantoic placenta of labyrinthine type. The placenta develops at the antimesometrial side of the implantation swelling opposite to the mesometrium. The latter defines the mesometrial side of the gestation chamber which is separated from the antimesometrial placenta by a more or less broad lateral region. The placental disc can roughly be divided into a labyrinthine zone on the fetal side, a centrally located fetomaternal junctional zone and a uterine gland zone towards the myometrium (Fig. 1A,B). During progressing pregnancy the gland zone becomes smaller, as the labyrinthine zone considerably extends to the maternal base (Fig. 1B).

image

Figure 1.  Section through the placental disc. (A) At 17 dpc the placenta is divided into a developing labyrinthine zone (lab), which is formed by invading chorioallantoic villi (asterisk), a fetomaternal junctional zone (j) and a secretory active uterine gland zone (gl). (B) The uterine gland zone diminishes in later pregnancy (24–27 dpc), as the labyrinth is extending. Methylene Blue & Azure II. Scale bars: 100 μm.

Download figure to PowerPoint

At about mid-pregnancy the placenta still shows a rather villous appearance (Fig. 1A). In the course of gestation the chorioallantoic villi, growing deep into the endometrium, strongly ramify and form a network of fetal and maternal tissues and blood vessels. Hence, the near-term placenta has a zona intima of labyrinthine type (Figs 1B and 2A). The definitive labyrinth is composed of fetal mesenchyme carrying allantoic vessels, maternal vessels and an intermediate bilayered trophoblast. The maternal epithelium seems to become totally destroyed during pregnancy and is absent from the late labyrinth. Cross-sectioned fetal villi often exhibit a star-like pattern with mostly one central arteriole, one to two venules and numerous small capillaries. The latter are located at the villus periphery, where they directly abut on the trophoblast without intermediate fetal connective tissue (Fig. 2A). The fetal capillaries are much smaller in diameter than the maternal ones and can be identified by their noticeably thinner endothelial lining and bigger erythrocytes (Fig. 2A). In contrast to maternal blood cells, fetal erythrocytes are still nucleated at mid term. In most light microscopy slides the two different trophoblast layers are clearly distinguishable. The darkly stained cellular trophoblast, which borders the fetal mesenchyme and capillaries, is covered by a rather sieve-like and therefore paler syncytiotrophoblastic layer (Fig. 2A).

image

Figure 2.  (A) Longitudinal section through the late labyrinth (24–27 dpc). Heavily branching, mesenchyme-rich (mes) chorioallantoic villi with fetal vessels (fv) are embedded in a network of maternal vessels (mv). (B) Section through the deeper portion of placenta (17 dpc) showing the transitional junctional zone between labyrinth (upper left) and gland zone (lower right). Note the fetal villus tip (asterisk) amidst multinucleate, vacuolised decidual cells (dc) reaching to uterine glands (ug). Connective tissue (mct). Methylene Blue & Azure II. Scale bars: 50 μm.

Download figure to PowerPoint

The endometrial connective tissue undergoes a decidualisation process and forms big, multinucleate decidual cells, which often contain numerous vacuoles (Fig. 2B). The junctional zone is an area of close contact between decidua and the trophoblast of fetal villi (Fig. 2B). In the course of gestation the uterine epithelium and underlying decidual cells increasingly degenerate and disintegrate due to the invading trophoblast. Even at mid-gestation there is no more luminal uterine epithelium left in the junctional zone, whereas remnants of decidua can be found until term. As the fetal tissue progressively invades and replaces the decidua, the whole junctional zone shifts towards basal uterine glands (Fig. 1A,B). The glands, which are often arranged in clusters, are embedded in maternal connective tissue (Figs 1A, 2B and 5A). In early pregnancy the uterine gland zone is distinctly broader and shows many more secretory active glands than in later stages (Fig. 1A,B). Although the total number of glands decreases during gestation, the remaining seem to continue to secrete until parturition. They probably add to the fetal histiotrophic nutrition throughout the whole pregnancy, which is indicated by the positive PAS reaction of the gland lumina in all stages. Moreover, from about 20 dpc on, a strong positive Quincke’s reaction of the gland cell apices and lumina supports the formation of a large amount of iron-containing secretion in the remaining glands.

A special feature of all studied talpid placentas is the existence of phagocytic regions – so-called areolas – which are mainly situated above uterine gland mouths at the fetal surface of the placental disc (Fig. 3A). These heterophagocytic regions can mostly be found in early stages, before 20 dpc. In later gestation the number of uterine glands and areolas noticeably declines, the latter mainly persisting above the lateral uterine wall (Fig. 3B). During pregnancy the upper parts of gland mouth epithelium and underlying connective tissue break down to some extent, leaving cell debris and leaked maternal blood cells in the lumen of the areolas (Fig. 3B). The columnar trophoblast cells of the areolas are arranged in arch-like structures and take up secretion of the associated uterine glands, as well as extravasated erythrocytes and cell debris (Fig. 3A,B). Despite their proven heterophagous nature, the areolas mainly seem to serve for histiotrophic nutrition and possibly have minor significance for the iron supply of the developing fetus. Their supposed major function is also consistent with the histochemical findings including a strong positive PAS reaction of areola lumina, gland necks and mouths throughout gestation, which confirms the presence of histiotrophs. By contrast, only traces of iron can be found in trophoblast cells and the lumen of remaining areolas during last third of pregnancy. This suggests that the areolas are of minor significance for fetal iron supply.

image

Figure 3.  The columnar trophoblast cells of heterophagous areolas (a) above uterine gland openings (ug) take up maternal secretion, cell debris (arrowhead) and erythrocytes. (A) Areolas are mostly found in the early placenta (14–16 dpc). (B) In late pregnancy (24–27 dpc) they become restricted to lateral and mesometrial regions. Methylene Blue & Azure II. Scale bars: 50 μm.

Download figure to PowerPoint

Transmission electron microscopy

Areolas and associated uterine glands

Throughout gestation the tall columnar trophoblast cells of the areolas contain a moderate amount of mitochondria and ribosomes, whereas rough endoplasmatic reticulum (rER) and Golgi complexes are scarce. The organelles are mostly distributed around the nuclei, which are situated at the cell bases (Fig. 4A). The trophoblast cells show lateral finger-like processes, which form irregular cell–cell borders with large intercellular spaces (Fig. 4A–D) being sealed by junctional complexes at the luminal ends (Fig. 4A). Some microvilli project from cell apices into the cavity of the areolas, where fine flocculent or granular material (probably gland secretion), extravasated blood cells and remnants of degenerated maternal tissue can be found (Fig. 4C). All this is taken up and catabolised by the heterophagous trophoblast cells of the areolas.

image

Figure 4.  Electron micrographs of the areola trophoblast at 17 dpc (A), 19–21 dpc (B), 24–27 dpc (C) and 21–27 dpc (D). Throughout gestation the phagocytic trophoblast cells contain different stages of lysosomal breakdown of phagocytised material (asterisk), e.g. erythrocytes, as well as numerous empty or filled vacuoles (long arrow), heterogeneous secondary lysosomes (l) and residual bodies (arrowhead). Note the large intercellular spaces between the cells, sealed by junctional complexes at the luminal ends (short arrow). Scale bars : 3 μm.

Download figure to PowerPoint

Areolas, showing signs of intracellular digestion, can be observed in all investigated stages, although there are conspicuously more of them in early gestation until about 20 dpc. Throughout the second half of pregnancy, phagocytised erythrocytes and cell debris in different stages of lysis are present within the areola trophoblast (Fig. 4A,B,D). In the course of the degenerative process the haemoglobin is released from ingested blood cells and becomes exposed to lysosomal enzymes. These denatured the haemoglobin, which results in an increasing electron-density within the erythrolysosomes (Fig. 4B). As the disintegration progresses, the centres of phagocytised blood cells become less electron-dense and finally become totally lucent, whilst a layer of extremely dense material is deposited on the inner surface of the erythrocyte membrane. Later, this dark rim is interrupted at several points (Fig. 4D). Like the released haemoglobin, the blood cell remnants undergo further degeneration caused by lysosomal enzymes. Finally, this results in the formation of indigestible, often whorled, membrane-like residues of high electron-density, which accumulate within vacuoles and are called residual bodies (Fig. 4A,B). All other phagocytised material, such as histiotrophs, cell debris and even whole cells, seems to undergo a similar lysosomal breakdown. This is indicated by the large quantity of extremely heterogeneous secondary lysosomes (Fig. 4C) and residual bodies scattered in the cytoplasm of areola trophoblast cells, which emphasises the heterophagous character of talpid areolas. The numerous empty or more or less electron-dense vacuoles with unknown content are evidence of the ongoing catabolic processes in the areola trophoblast of all investigated placental stages (Fig. 4A–C).

The epithelial cells of uterine glands, which are closely associated with the overlying areolas, have a diverse appearance depending on their position within the placental disc. Cells of the upper gland necks are mostly cuboidal and become squamous towards the gland mouths (Fig. 3B), whereas cells of the gland bodies are rather long and columnar in shape (Figs 2B and 5C,D). Many fundic gland cells have lateral cytoplasmic interdigitation, which leaves broad intercellular spaces sometimes filled with residual bodies (Fig. 5B). During the second half of gestation the fundic gland epithelium contains some macrophages, which are probably responsible for the phagocytosis and digestion of old or dying gland cells (Fig. 5C,D). In late pregnancy even the lumina of some glands contain macrophages (Fig. 5A). Altogether, this shows that intracellular digestive and degenerative processes are very common in the gland epithelium. This might be a consequence of general degeneration of uterine glands and surrounding connective tissue, which finally leads to the observed decline of glands in late pregnancy.

image

Figure 5.  Electron micrographs of uterine gland cells at 24–27 dpc (A,C,D) and 14–16 dpc (B). (A) Lumens of glands (ug), embedded in connective tissue (mct), often contain macrophages (asterisk). (B,C,D) At all stages some gland cells display high secretory activity indicated by the presence of numerous mitochondria (m), secretory vesicles (long arrow), extended Golgi zones (arrowhead) and dilated rER (short arrow). (C,D) Especially in late pregnancy, the glands may show signs of degeneration indicated by the presence of macrophages (asterisk) penetrating the glandular epithelium. Scale bars: 9.4 μm (A), 3 μm (B,C, D).

Download figure to PowerPoint

Intriguingly, at the same time many of the fundic gland cells display a high secretory activity. This is strongly supported by the large amount of mitochondria, ribosomes and widely dilated rER (Fig. 5D) found in their cytoplasm. Furthermore, there are extensive Golgi zones (Fig. 5C) and numerous small vesicles in the apical cell regions (Fig. 5B,C). Interestingly, these regions were positively tested for iron, as were the macrophages in the gland lumina. Hence, the macrophages are possibly involved in the transport of iron-containing secretion from the uterine glands to the embryo in late gestation, as described by Zeller & Kuhn (1994). Many of the fundic gland cells show numerous apical microvilli and kinocilia projecting into the lumen (Fig. 5B). The movement of kinocilia might possibly create a flow, which facilitates the release of secretion into the gland lumina and from here to the areolas, where it can be taken up by fetal trophoblast cells. Some epithelial cells of uterine gland mouths also show signs of degradation and sometimes even leave the cell aggregate. Consequently, the upper gland lumina are filled with cellular debris deriving not only from maternal connective tissue but also from destroyed epithelium of uterine gland mouths (Fig. 3B).

Ultrastructure of interhaemal barrier

At least during the second half of gestation the interhaemal barrier within the labyrinthine zone is formed by the endothelial lining of fetal and maternal vessels and an intermediate bilayered trophoblast (Fig. 6A,B). The syncytiotrophoblast and maternal endothelium are separated by an extremely thin, inhomogeneous interstitial lamina. It contains isolated collagen fibres, which represent remnants of eroded maternal connective tissue (Fig. 6A). However, this layer is often bridged by projections of syncytiotrophoblast (Fig. 6A). In most places the fetal mesenchyme does not contribute to the interhaemal barrier. All in all, this convincingly confirms the endotheliochorial nature of interhaemal barrier in T. occidentalis.

image

Figure 6.  Electron micrographs of the endotheliochorial interhaemal barrier. (A) (17 dpc) The syncytiotrophoblast (st), which is adjacent to the cytotrophoblast (ct), forms irregular processes (long arrow) bridging the interstitial lamina (il). The latter separates the syncytiotrophoblast from the hypertrophied maternal endothelium (me) with its distinct basal lamina (arrowhead). (B) In late pregnancy (24–27 dpc) the syncytiotrophoblast becomes increasingly sponge-like in structure. It is attached to the cytotrophoblast through the formation of junctional complexes (short arrow). Note the partial fusion of basal laminas (arrowhead) of the thin fetal endothelium (fe) and cytotrophoblast. Scale bars : 3 μm.

Download figure to PowerPoint

Throughout pregnancy the cellular trophoblast adjacent to the fetal endothelium of capillaries is mostly broader than the irregular, sieve-like syncytiotrophoblast. With progressing gestation the syncytial layer becomes considerably thinner and spongier (Fig. 6A,B). There is a rather uneven border between the two trophoblastic layers, which are interconnected by several junctional complexes (Fig. 6B). Cytotrophoblast cells are mostly cuboidal and show large nuclei of varying shape. They have numerous ribosomes and mitochondria, some Golgi fields, a few vesicles and a high amount of rER in different degrees of dilation. Altogether, this indicates an active cell metabolism. In contrast, the nuclei of the syncytial layer are often lobulated and generally smaller. The syncytiotrophoblast also contains a large quantity of rER with dilated cisterns and many densely packed ribosomes, but conspicuously less mitochondria and more residual bodies and filled or empty vesicles than the cytotrophoblast. Together with these empty vesicles and vacuoles, various cytoplasmic processes towards both sides account for the spongy appearance, especially of the late syncytiotrophoblast (Fig. 6B). Due to these apical appendages and invaginations the syncytium has a more irregular border with the maternal endothelium than the cytotrophoblast has with the fetal endothelium (Fig. 6B).

Particularly in early stages it is easy to identify the basal lamina of maternal endothelium, apparently emerging from the interstitial layer (Fig. 6A). The endothelium lining maternal capillaries is much thicker than that lining fetal vessels and has prominent nuclei. The lining shows signs of active cell metabolism, which is supported by its high amount of ribosomes, mitochondria and dilated rER and the presence of vesicles. The comparatively thin fetal endothelium attenuates even more in late pregnancy (Fig. 6A,B). Its flat cells are interconnected by conspicuous cell–cell junctions and show elongated, discoid nuclei and fewer ribosomes, rER and mitochondria compared with the maternal endothelium (Fig. 6A). In early pregnancy, immature fetal blood cells can be found within the lumen of capillaries. As in most places there is no fetal mesenchyme sheathing the peripheral capillaries, the basal laminas of cytotrophoblast and fetal endothelium are in direct contact with each other and sometimes even fuse into one layer (Fig. 6B).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Concluding remarks
  8. Acknowledgements
  9. Authors’ contributions
  10. References

Considering all our results, we can say that the chorioallantoic placenta of T. occidentalis strongly resembles that of the closely related T. europaea, which has previously been described by different authors, such as Strahl (1892), Vernhout (1894) and Malassiné & Leiser (1984). We could also find a number of similarities to the placenta of S. aquaticus. The striking differences concerning the structure of the interhaemal barrier, allegedly found by Prasad et al. (1979), will be discussed later.

Heterophagous areolas

The chorioallantoic placenta of many Eutheria develops specialised regions for the uptake and digestion of material from maternal sources. So-called haemophagous regions mainly ingest blood cells extravasated from ruptured maternal vessels (Burton, 1982), whereas heterophagous structures are able to take up gland secretions, as well as erythrocytes and maternal cell debris (Wimsatt, 1950). Hence, heterophagous regions are possibly responsible for both histiotrophic nutrition and the iron supply of the fetus. In contrast, only the latter seems to be the main function of haemophagous regions. Therefore, these phagocytic structures are of great nutritive significance for the embryo, after the chorioallantoic placenta replaces the yolk sac placenta and thus, eliminates one important pathway of histiotrophic fetal nutrition (Enders & Carter, 2006). Recently, Enders & Carter (2006) gave a summarising review of the occurrence of haemophagous and heterophagous regions in various species with reference to their different types of interhaemal barrier. They placed emphasis on the fact that despite their variable position and appearance within the placenta these regions universally comprise a columnar cytotrophoblast capable of ingesting and digesting material of maternal origin. Moreover, their synoptic review revealed the existence of such phagocytic areas in the placenta of numerous species covering all types of interhaemal barrier – epitheliochorial, endotheliochorial and even haemochorial. Interestingly, many species with noninvasive epitheliochorial placentas develop regions mainly serving the uptake of histiotrophs, whereas only a few species showing the most invasive haemochorial placentation form accessory placental areas at all. If they do so, these structures merely have a haemophagous function. In species with medium invasive endotheliochorial placentation, haemophagous regions, as found in the cat (Leiser, 1982), are very common. Within Eulipotyphla only moles and shrews, which both show endotheliochorial placentation, have specialised phagocytic regions. Unlike moles, shrews, such as Suncus murinus, show haemophagocytic areas not formed by the chorioallantoic placenta, but part of the bilaminar omphalopleure of the fetal yolk sac (Siniza & Zeller, 2005). The function of these haemophagous regions seems to be confined to the fetal supply with substances extracted from blood cells, such as iron.

In disagreement with Malassiné & Leiser (1984), who reported on the existence of ‘haemophagous zones’ in the T. europaea placenta, our histological and ultrastructural findings rather point to a heterophagous nature of the areolas in T. occidentalis. This is consistent with the findings in S. aquaticus (Prasad et al. 1979). We observed an immense reduction of areolas in T. occidentalis in the course of gestation, which surprisingly contradicts the results of Strahl (1892), Malassiné & Leiser (1984) and Prasad et al. (1979), who found numerous areolas in T. europaea, as well as in S. aquaticus even in late stages. Nevertheless, in our opinion a noticeable decrease in areolas inevitably has to accompany the reduction of associated uterine glands, which has not only been observed by us in the Iberian mole placenta, but also by Strahl (1892) in the European mole placenta.

Remarkably, the uterine glands and areolas of T. occidentalis show the same histochemical characters as those of the llama, a species with epitheliochorial placentation (Iturrizaga et al. 2007). This indicates a similar function of areolas in two distantly related species with different mode of placentation. Furthermore, our electron microscopic studies give proof of striking ultrastructural similarities between the areolas in T. occidentalis and the hetero- and haemophagous placental regions of other species. Although we generally observed only small amounts of blood in the areola lumina, we were able to detect all typical stages of lysosomal erythrocyte breakdown in phagocytic trophoblast cells of the T. occidentalis placenta. These stages were described for the first time in detail by Burton et al. (1976) in heterophagous regions of the epitheliochorial sheep placenta. However, the presence of these universal degradative stages has also been confirmed for the haemophagous regions of different carnivores, such as the cat (Leiser & Enders, 1980; Malassiné, 1982), the spotted hyena (Enders et al. 2006) and some Eulipotyphla, e.g. Blarina brevicauda (King et al. 1978) and S. murinus (Siniza & Zeller, 2005). Our TEM studies provide convincing support for the similar anatomy and ultrastructure of those widespread absorptive placental structures.

As accessory phagocytic placental structures of various appearances with similar anatomy and function are widespread among Eutheria, it is possible that they independently evolved several times in the different eutherian clades and therefore are unlikely to be part of the eutherian placental morphotype. The fact that Talpidae, which show the least invasive – although endotheliochorial – mode of placentation among Eulipotyphla, developed areolas mainly for the uptake of histiotrophs might possibly suggest a direct association between the invasiveness of placenta and the occurrence and type of phagocytic structures.

Interhaemal barrier

Since Malassiné and Leiser published their ultrastructural findings on the near term placenta of T. europaea in 1984, it has been largely accepted that the European mole placenta has an endotheliochorial interhaemal barrier. Our TEM studies of the chorioallantoic placental labyrinth in T. occidentalis led to the same result. Unlike representatives of the genus Talpa, S. aquaticus was found to have an extremely thin and diffuse chorioallantoic placenta of villous rather than labyrinthine type. The single Scalopus placenta ultrastructurally studied by Prasad et al. (1979) was considered to be definitive and showed a thin continuous uterine epithelial layer which formed part of an epitheliochorial interhaemal barrier within the zona intima. After careful examination of the published TEM micrographs of the interhaemal barrier of S. aquaticus, we tend to challenge this assumption. Obviously, we are not the only ones who did this, as Carter (2005) who reviewed the slides of the Scalopus placenta cautiously expressed his scepticism concerning the hypothesis. The pivotal question to answer is whether the thin, syncytial layer found by Prasad is really uterine epithelium or, probably, syncytiotrophoblast. If there are actually two fused basal laminas belonging to the endothelial lining of maternal vessels and the uterine epithelium, respectively, as stated by Prasad, it would be a strong argument for his theory. Nevertheless, based on the published TEM micrographs we would question this. Particularly, Prasad’s figures 19, 21 and 24 quite clearly demonstrate the existence of a single thin basal lamina which borders the maternal endothelium and separates it from an adjacent interstitial layer filled with several collagen fibres. The adjoining syncytial layer, believed by Prasad to be epithelium, seems to have no basal lamina at all and strongly resembles the syncytiotrophoblast we found in the interhaemal barrier of T. occidentalis. None of our TEM slides of any gestational stage showed more than the one basal lamina of the maternal endothelium. We could not detect a second or even two fused basal laminas. In our opinion this leaves no doubt about our interpretation of the ultrastructural findings in T. occidentalis, which verifies the existence of an endotheliochorial placenta in both studied species of the genus Talpa.

Endotheliochorial placentas are generally considered to be invasive because in the zona intima there is only trophoblastic tissue between fetal and maternal placental vessels. However, the interhaemal barrier of the European and Iberian mole comprises two trophoblastic layers, a prominent cellular and a thinner syncytial one. Therefore, the diffusion distance between fetal and maternal circulation is noticeably greater than in other Eulipotyphla, such as shrews, solenodons and hedgehogs. Among these core insectivores, solenodons and hedgehogs have the most invasive, haemochorial mode of placentation (Hubrecht, 1889; Wislocki, 1940). Up to now, this has only been ultrastructurally proven in the white-bellied hedgehog (Atelerix albiventris) (S. Siniza, personal observation). A transitional state of placental invasiveness, reached by an endotheliochorial to endothelio-endothelial interhaemal barrier, can be found in shrews, such as B. brevicauda (Wimsatt et al. 1973) and S. murinus (Kiso & Nakagawa, 1994; Bhiwgade et al. 1998). Here, a very thin, in late gestation sometimes sponge-like or discontinuous, syncytiotrophoblastic layer separates fetal from maternal placental vessels. In its sponginess the syncytial layer somehow resembles the syncytiotrophoblast of late T. occidentalis placentas.

Concluding remarks

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Concluding remarks
  8. Acknowledgements
  9. Authors’ contributions
  10. References

A few years ago, a promising attempt of placental morphotype reconstruction for Eutheria was published (Carter & Mess, 2007), which hypothesised a compact and invasive labyrinthine placenta in the eutherian stem species. Although the eulipotyphlan mode of placentation seems to be quite diverse, it is always rather invasive and compact and therefore can be assumed to be plesiomorphic, at least within Laurasiatheria. Within Eulipotyphla the moles with their less invasive type of placentation could possibly represent a more derived state. There seems to be a direct association between compact and invasive placentation in Eulipotyphla and the occurrence of altricial neonates representing the eutherian neonate morphotype, as reconstructed by Szdzuy & Zeller (2009).

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Concluding remarks
  8. Acknowledgements
  9. Authors’ contributions
  10. References

We are especially grateful to Jutta Zeller for helpful advice with histological staining methods and Gabriele Drescher for technical support with electron microscopy and Dr F. J. Carmona, Dr R. D. Dahdich, Dr F. M. Real, and E. Martín for their priceless help in the mole capture work. We also thank Dr Thomas Göttert for helpful comments on an earlier draft. The dissertation project of Swetlana Siniza was supported by the Graduate Research Programme ‘Evolutionary Transformations and Mass Extinctions’ (Museum of Natural History, Berlin). This Programme was funded by the Deutsche Forschungsgemeinschaft (DFG). The contribution of the University of Granada laboratory was supported by the Spanish Ministry of Science and Innovation through grants No. CGL2004-00863/BOS and by the Andalusian Government through grant No. CVI2057.

Authors’ contributions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Concluding remarks
  8. Acknowledgements
  9. Authors’ contributions
  10. References

Dr Rafael Jiménez and Dario Lupiañez collected the moles, dissected them and provided the placental material used in this study. Prof. Ulrich Zeller provided laboratory facilities and revised the manuscript. Swetlana Siniza conducted the histological and ultrastructural investigation of the Talpa placentas including all laboratory work, and wrote the manuscript.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Concluding remarks
  8. Acknowledgements
  9. Authors’ contributions
  10. References