Porosomes in uterine epithelial cells: Ultrastructural identification and characterization during early pregnancy.

Porosomes are plasma membrane structures in secretory cells that allow transient docking and/or partial fusion of vesicles during which they release their content then disengage. This is referred to as "kiss and run" exocytosis. During early pregnancy, at the time of receptivity, there is a high level of vesicle activity in uterine epithelial cells (UECs). One of the secretory pathways for these vesicles could be via porosomes, which have yet to be identified in UECs. This study identified porosomes in the apical plasma membrane of UECs for the first time. These structures were present on days 1, 5.5, and 6 of early pregnancy, where they likely facilitate partial secretion via "kiss and run" exocytosis. The porosomes were measured and quantified on days 1, 5.5, and 6, which showed there are significantly more porosomes on day 5.5 (receptive) compared to day 1 (nonreceptive) of pregnancy. This increase in porosome numbers may reflect major morphological and molecular changes in the apical plasma membrane at this time such as increased cholesterol and soluble NSF attachment protein receptor proteins, as these are structural and functional components of the porosome complex assembly. Porosomes were observed in both resting (inactive) and dilated (active) states on days 1, 5.5, and 6 of early pregnancy. Porosomes on day 5.5 are significantly more active than on day 1 as demonstrated by the dilation of their base diameter. Further two-way ANOVA analysis of base diameter in resting and dilated states found a significant increase in porosome activity in day 5.5 compared to day 1. This study therefore indicates an increase in the number and activity of porosomes at the time of uterine receptivity in the rat, revealing a mechanism by which the UECs modify the uterine luminal environment at this time.

Different Genes are Recruited During Convergent Evolution of Pregnancy and the Placenta.

The repeated evolution of the same traits in distantly related groups (convergent evolution) raises a key question in evolutionary biology: do the same genes underpin convergent phenotypes? Here, we explore one such trait, viviparity (live birth), which, qualitative studies suggest, may indeed have evolved via genetic convergence. There are >150 independent origins of live birth in vertebrates, providing a uniquely powerful system to test the mechanisms underpinning convergence in morphology, physiology, and/or gene recruitment during pregnancy. We compared transcriptomic data from eight vertebrates (lizards, mammals, sharks) that gestate embryos within the uterus. Since many previous studies detected qualitative similarities in gene use during independent origins of pregnancy, we expected to find significant overlap in gene use in viviparous taxa. However, we found no more overlap in uterine gene expression associated with viviparity than we would expect by chance alone. Each viviparous lineage exhibits the same core set of uterine physiological functions. Yet, contrary to prevailing assumptions about this trait, we find that none of the same genes are differentially expressed in all viviparous lineages, or even in all viviparous amniote lineages. Therefore, across distantly related vertebrates, different genes have been recruited to support the morphological and physiological changes required for successful pregnancy. We conclude that redundancies in gene function have enabled the repeated evolution of viviparity through recruitment of different genes from genomic "toolboxes", which are uniquely constrained by the ancestries of each lineage.

Structure and permeability of the egg capsule of the placental Australian sharpnose shark, Rhizoprionodon taylori.

Shark placentae are derived from modifications to the fetal yolk sac and the maternal uterine mucosa. In almost all placental sharks, embryonic development occurs in an egg capsule that remains intact for the entire pregnancy, separating the fetal tissues from the maternal tissues at the placental interface. Here, we investigate the structure and permeability of the egg capsules that surround developing embryos of the placental Australian sharpnose shark (Rhizoprionodon taylori) during late pregnancy. The egg capsule is an acellular fibrous structure that is 0.42 ± 0.04 µm thick at the placental interface between the yolk sac and uterine tissues, and 0.67 ± 0.08 µm thick in the paraplacental regions. This is the thinnest egg capsule of any placental shark measured so far, which may increase the diffusion rate of respiratory gases, fetal wastes, water and nutrients between maternal and fetal tissues. Molecules smaller than or equal to ~ 1000 Da can diffuse through the egg capsule, but larger proteins (~ 3000-26,000 Da) cannot. Similar permeability characteristics between the egg capsule of R. taylori and other placental sharks suggest that molecular size is an important determinant of the molecules that can be exchanged between the mother and her embryos during pregnancy.

Uterine cellular changes during mammalian pregnancy and the evolution of placentation†.

There are many different forms of nutrient provision in viviparous (live-bearing) species. The formation of a placenta is one method where the placenta functions to transfer nutrients from mother to fetus (placentotrophy), to transfer waste from the fetus to the mother, and to perform respiratory gas exchange. Despite having the same overarching function, there are different types of placentation within placentotrophic vertebrates, and many morphological changes occur in the uterus during pregnancy to facilitate formation of the placenta. These changes are regulated in complex ways but are controlled by similar hormonal mechanisms across species. This review describes current knowledge of the morphological and molecular changes to the uterine epithelium preceding implantation among mammals. Our aim is to identify the commonalities and constraints of these cellular changes to understand the evolution of placentation in mammals and to propose directions for future research. We compare and discuss the complex modifications to the ultrastructure of uterine epithelial cells (UEC) and show that there are similarities in the changes to the cytoskeleton and gross morphology of the UEC, especially of the apical and lateral plasma membrane of the cells during the formation of a placenta in all eutherians and marsupials studied to date. We conclude that further research is needed to understand the evolution of placentation among viviparous mammals, particularly concerning the level of placental invasiveness, hormonal control, and genetic underpinnings of pregnancy in marsupial taxa.

Structure of the paraplacenta and the yolk sac placenta of the viviparous Australian sharpnose shark, Rhizoprionodon taylori.

INTRODUCTION: Viviparity (live-birth) has evolved from oviparity (egg-laying) multiple times in sharks. While most transitions from oviparity to viviparity have resulted in non-placental forms of viviparity, some sharks develop a yolk sac placenta during pregnancy. The Australian sharpnose shark (Rhizoprionodon taylori) is a placental species that suspends embryonic development in a diapause for most of pregnancy. METHODS: To identify structures involved in supporting rapid embryonic growth in late pregnancy, we examined uterine and placental morphology by light and electron microscopy. RESULTS: Paraplacental uterine regions have morphological specialisations consistent with secretion and fluid transport between uterine tissues and the lumen. Uterine secretions in the lumen may be absorbed by the outgrowths on the embryonic umbilical cord ('appendiculae'), which are densely covered by microvilli. The placenta consists of uterine villi that interdigitate with the yolk sac and enhance the surface area available for fetomaternal exchange. The yolk sac does not invade the uterine epithelium, and the egg capsule remains intact at the placental interface, separating maternal and fetal tissues. Some placental uterine epithelial cells are secretory, and endocytic vesicles in the opposing yolk sac ectodermal cells suggest that nutrient transport is by histotrophic uterine secretion followed by fetal absorption. Respiratory gases, water and possibly small nutrients likely diffuse across the placenta, where maternal and fetal blood vessels are ~2 µm apart. DISCUSSION: Placental structure in R. taylori is similar to most other sharks, but there are differences in cellular structures between species that may indicate species-specific placental transport mechanisms.

Rab13 and Desmosome Redistribution in Uterine Epithelial Cells During Early Pregnancy.

The luminal uterine epithelial cells are the first point of contact with the implanting blastocyst. Dramatic changes occur in the structure and function of these cells at the time of receptivity including changes in the lateral junctional complex. While these morphological changes are important for uterine receptivity, currently there is no known mechanism of regulation of the lateral junctional complexes. Rab13, a member of the Rab (Ras-related in the brain) family of GTPases has a critical role in endosomal trafficking to the lateral plasma membrane and is involved in modulation of the tight junction in several cell types. The aim of this study is to investigate the role of Rab13 in changes to the lateral junctional complex at the time of receptivity. Immunofluorescence microscopy demonstrated no association between Rab13 and ZO-1 (a tight junction protein) or Rab13 and E-cadherin (an integral component of adherens junctions). Co-localisation was demonstrated between Rab 13 and desmoglein-2 at the time of fertilization and also at receptivity suggesting involvement of Rab13 in relocalisation of desmoglein-2 and formation of giant desmosomes in the apical part of the lateral plasma membrane at the time of uterine receptivity. We suggest that despite the loss of the adherens junction at the time of receptivity, the presently reported redistribution of desmosomes regulated by Rab13 allows the uterine epithelium to maintain structural integrity.

Membrane trafficking directed by VAMP2 and syntaxin 3 in uterine epithelial cells.

Luminal uterine epithelial cells (UEC) have a surge in vesicular activity during early uterine receptivity. It has been predicted these vesicles exit the UEC via exocytosis resulting in secretion and membrane trafficking. The present study investigated the changes in SNARE proteins VAMP2 (v-SNARE) and syntaxin 3 (t-SNARE) localisation and abundance in UECs during early pregnancy in the rat. We found VAMP2 and syntaxin 3 are significantly higher on day 5.5 compared to day 1 of pregnancy. On day 5.5, VAMP2 is perinuclear and syntaxin 3 is concentrated in the apical cytoplasm compared to a cytoplasmic localisation on day 1. This change in localisation and abundance show VAMP2 and syntaxin 3 are involved in vesicular movement and membrane trafficking in UECs during early pregnancy. This study also investigated the influence of cytoskeletal disruption of microtubules and actin filaments on VAMP2 and syntaxin 3 in UECs grown in vitro, since microtubules and actin influence vesicle trafficking. As expected, this study found disruption to microtubules with colchicine and actin with cytochalasin D impacted VAMP2 and syntaxin 3 localisation. These results suggest VAMP2 and syntaxin 3 are involved in the timely trafficking of vesicular membranes to the apical surface in UECs during early pregnancy, as are of microtubules and actin.

Uterine epithelial remodelling during pregnancy in the marsupial Monodelphis domestica (Didelphidae): Implications for mammalian placental evolution.

Mammalian pregnancy involves remodelling of the uterine epithelium to enable placentation. In marsupials, such remodelling has probably played a key role in the transition from ancestral invasive placentation to non-invasive placentation. Identifying uterine alterations that are unique to marsupials with non-invasive placentation can thus elucidate mechanisms of marsupial placental evolution. We identified apical alterations to uterine epithelial cells prior to implantation in Monodelphis domestica, a member of the least derived living marsupial clade (Didelphidae) with invasive (endotheliochorial) placentation. We then compared these traits with those of Macropus eugenii (Macropodidae) and Trichosurus vulpecula (Phalangeridae), both with non-invasive placentation, to identify which alterations to the uterine epithelium are ancestral and which facilitate secondarily evolved non-invasive placentation. In M. domestica, remodelling of the uterine epithelium involves reduced cellular heterogeneity and development of uterodome-like cells, suggesting that similar alterations may also have occurred in the marsupial common ancestor. These alterations also overlap with those of both T. vulpecula and Ma. eugenii, suggesting that the placental shift from invasive to non-invasive placentation in marsupials involves essential, conserved characteristics, irrespective of placental mode. However, unique apical alterations of both T. vulpecula and Ma. eugenii, relative to M. domestica, imply that lineage-specific alterations underpin the evolutionary shift to non-invasive placentation in marsupials.

Structural changes to the uterus of the dwarf ornate wobbegong shark (Orectolobus ornatus) during pregnancy.

Embryos of the viviparous dwarf ornate wobbegong shark (Orectolobus ornatus) develop without a placenta, unattached to the uterine wall of their mother. Here, we present the first light microscopy study of the uterus of O. ornatus throughout pregnancy. At the beginning of pregnancy, the uterine luminal epithelium and underlying connective tissue become folded to form uterine ridges. By mid to late pregnancy, the luminal surface is extensively folded and long luminal uterine villi are abundant. Compared to the nonpregnant uterus, uterine vasculature is increased during pregnancy. Additionally, as pregnancy progresses the uterine epithelium is attenuated so that there is minimal uterine tissue separating large maternal blood vessels from the fluid that surrounds developing embryos. We conclude that the uterus of O. ornatus undergoes an extensive morphological transformation during pregnancy. These uterine modifications likely support developing embryos via embryonic respiratory gas exchange, waste removal, water balance, and mineral transfer.

Three-dimensional reconstruction of leukocyte internalisation in the luminal uterine epithelium following mating.

Following mating, leukocytes are recruited to the uterine epithelium where they phagocytose spermatozoa and mediate maternal immune tolerance as well as a mild inflammatory response. In this ultrastructural study we utilised array tomography, a high-resolution volume scanning electron microscopy approach to 3D reconstruct the cellular relationships formed by leukocytes recruited to the luminal uterine epithelium 12 h post-mating in the rat. We report that following mating, neutrophils and macrophages are internalised by the luminal uterine epithelium, with multiple leukocytes internalised via contortion through a small tunnel in the apical membrane into a large membrane-bound vacuole within the cytoplasm of luminal uterine epithelial cells (UECs). Once internalised within the UECs, recruited leukocytes appear to phagocytose material within the membrane-bound vacuole and most ultimately undergo a specialised cell death, including vacuolisation and loss of membrane integrity. As these observations involve ultrastructurally normal leukocytic cells internalised within non-phagocytic epithelial cells, these observations are consistent with the formation of cell-in-cell structures via entosis, rather than phagocytic engulfment by UECs. Although cell-in-cell structures have been reported in normal and pathological conditions elsewhere, the data collected herein represents the first evidence of the formation of cell-in-cell structures within the uterine epithelium as a novel component of the maternal inflammatory response to mating.

Dynamic changes to claudins in the uterine epithelial cells of the marsupial Sminthopsis crassicaudata (Dasyuridae) during pregnancy.

The fluid that surrounds the embryo in the uterus contains important nourishing factors and secretions. To maintain the distinct microenvironment in the uterine lumen, the tight junctions between uterine epithelial cells are remodeled to decrease paracellular movement of molecules and solutes. Modifications to tight junctions between uterine epithelial cells is a common feature of pregnancy in eutherian mammals, regardless of placental type. Here we used immunofluorescence microscopy and western blot analysis to describe distributional changes to tight junctional proteins, claudin-1, -3, -4, and -5, in the uterine epithelial cells of a marsupial species, Sminthopsis crassicaudata. Immunofluorescence microscopy revealed claudin-1, -3, and -5 in the tight junctions of the uterine epithelium of S. crassicaudata during pregnancy. These specific claudins are associated with restricting passive movement of fluid between epithelial cells in eutherians. Hence, their function during pregnancy in S. crassicaudata may be to maintain the uterine luminal content surrounding developing embryos. Claudin-4 disappears from all uterine regions of S. crassicaudata at the time of implantation, in contrast with the distribution of this claudin in some eutherian mammals. We conclude that like eutherian mammals, distributional changes to claudins in the uterine epithelial cells of S. crassicaudata are necessary to support pregnancy. However, the combination of individual claudin isoforms in the tight junctions of the uterine epithelium of S. crassicaudata differs from that of eutherian mammals. Our findings suggest that the precise permeability of the paracellular pathway of the uterine epithelium is species-specific.

Ovarian Hyperstimulation Reduces Vascular Endothelial Growth Factor-A During Uterine Receptivity.

The angiogenic factor vascular endothelial growth factor-A (VEGFA) plays a critical role during early pregnancy in many species including the rat, and any alterations in VEGFA levels can severely impact blastocyst implantation rates. The rat ovarian hyperstimulation (OH) model is useful in studying how the induction of superovulation affects VEGFA levels and endometrial receptivity to blastocyst implantation. The present study shows that the major isoform in the rat uterus, Vegf188, is reduced at the time of receptivity in OH compared to normal pregnancy, whereas there is no change in Vegf164 and Vegf120 messenger RNA (mRNA). The VEGFA receptor 2 (VEGFR2) protein levels are also reduced at the time of receptivity in OH. Our ovariectomy studies show that Vegf164, Vegf188, and Vegf120 are significantly decreased by estrogen, and, to a lesser extent progesterone, when compared to control animals. Although no change in the percentage of endometrial blood vessels was seen across all stages of pregnancy, at the time of receptivity in OH pregnancies, blood vessels were typically larger compared to other stages. The altered progesterone-estrogen ratio seen in OH, taken together with our ovariectomy studies, explains the changes to Vegfa mRNA in OH at the time of receptivity. Since VEGFA is important during implantation, the changes to Vegfa and VEGFR2 levels in the endometrium may help explain the observed lower endometrial receptivity following OH. This study aimed to analyse how ovarian hyperstimulation alters the levels of vascular endothleial growth factor and its major receptor, VEGFR2 in the uterus in a rat model.

Sex steroids influence the plasma membrane transformation in the uterus of the fat-tailed dunnart (Sminthopsis crassicaudata, Marsupialia).

The uterine epithelium undergoes remodelling to become receptive to blastocyst implantation during pregnancy in a process known as the plasma membrane transformation. There are commonalities in ultrastructural changes to the epithelium, which, in eutherian, pregnancies are controlled by maternal hormones, progesterone and oestrogens. The aim of this study was to determine the effects that sex steroids have on the uterine epithelium in the fat-tailed dunnart Sminthopsis crassicaudata, the first such study in a marsupial. Females were exposed to exogenous hormones while they were reproductively quiescent, thus not producing physiological concentrations of ovarian hormones. We found that changes to the protein E-cadherin, which forms part of the adherens junction, are controlled by progesterone and that changes to the desmoglein-2 protein, which forms part of desmosomes, are controlled by 17ß-oestradiol. Exposure to a combination of progesterone and 17ß-oestradiol causes changes to the microvilli on the apical surface and to the ultrastructure of the uterine epithelium. There is a decrease in lateral adhesion when the uterus is exposed to progesterone and 17ß-oestradiol that mimics the hormone environment of uterine receptivity. We conclude that uterine receptivity and the plasma membrane transformation in marsupial and eutherian pregnancies are under the same endocrine control and may be an ancestral feature of therian mammals.

α-Parvin and ß-parvin in the rat uterus during decidualisation and uterine receptivity.

During early pregnancy, the uterine luminal epithelial cells (UECs) and endometrial stromal cells (ESCs) undergo morphological changes to enable blastocyst implantation. The present study investigates, for the first time, the cytoskeletal-associated proteins and α-actinin superfamily members, α-parvin and ß-parvin, during early pregnancy in the rat uterus. These two PARVA proteins are involved in cell adhesion, morphological changes and regulation of other cytoskeletal proteins, through binding with proteins such as actin and integrin-linked kinase. α-parvin is present in UECs at fertilisation and significantly decreases by the time of implantation. ß-parvin acts in opposition; significantly increasing in both UECs and ESCs at the time of implantation, suggesting a role in the process of decidualisation. Additionally, the presence of a serine-8 residue-phosphorylated α-parvin, which is associated with cell morphology changes, was found in the nuclear region of both UECs and ESCs during implantation and decidualisation. We also show that the presence of both ß-parvin and phosphorylated α-parvin in ESCs is dependent on decidualisation occurring. This study demonstrates that the changing balance and localisation of the two PARVA proteins are dependent on the time of uterine receptivity, suggesting a co-dependent role in the cytoskeletal re-organisation crucial to the changing conditions necessary for implantation and decidualisation.

Uterine Receptivity in Merriam's Kangaroo Rat (Dipodomys merriami).

The uterine surface undergoes significant remodeling, termed the "plasma membrane transformation," during pregnancy to allow for implantation of the blastocyst and formation of the placenta in viviparous amniote vertebrates. Unlike other species within the superorder Euarchontoglires, which have a hemochorial (highly invasive) placenta, kangaroo rats (Dipodomys spp.) exhibit a less invasive endotheliochorial placenta. We characterized the changes that occur to membrane molecules and the cellular ultrastructure of the uterine epithelium during early pregnancy in Merriam's kangaroo rat, Dipodomys merriami using electron microscopy and immunofluorescence microscopy. Epithelial cadherin (E-cadherin) is an adhesion protein that forms the adherens junction and is localized to the lateral plasma membrane of uterine epithelium during the nonreproductive state but localizes nonspecifically in the uterine epithelium immediately preceding implantation. Desmosomes are a type of cadherin that form junctional complexes along the lateral plasma membrane of epithelium. Dsg-2, a marker for desmosomes, is localized along the lateral plasma membrane in non-pregnant animals but redistributes to the apical region of the lateral plasma membrane during early pregnancy. The shift in desmosome and cadherin distribution before implantation suggests that there is a reduction in lateral adhesion between epithelial cells to allow for invasion by the blastocyst. Surprisingly, although Kangaroo rats form a less invasive placenta, these same changes occur during pregnancy in species with highly invasive placentation, such as the laboratory rat and human. These commonalities suggest that it is not through the retention of lateral adhesion that the blastocyst is prevented from further invasion in this rodent species. Anat Rec, 301:1928-1935, 2018. © 2018 Wiley Periodicals, Inc.

Uterine Epithelial Cells Undergo a Plasma Membrane Transformation During Early Pregnancy in the Domestic Cat (Felis catus).

Mammals exhibit similar changes in uterine epithelial morphology during early pregnancy despite having a diverse range of placental types. The uterine epithelium undergoes rapid morphological and molecular change ("plasma membrane transformation") during the early stages of pregnancy to allow attachment of the blastocyst. The domestic cat, Felis catus is in the order Carnivora; all species within the Carnivora studied so far develop an endotheliochorial placenta during pregnancy. The endotheliochorial placental type is a common form of placental invasion in mammals. The molecular changes that allow remodeling of the uterine epithelium in preparation for implantation are unknown in most mammals but would provide us with an understanding of what molecules underpin successful implantation and pregnancy among Carnivora. We used immunofluorescence microscopy to localize the key adherens junction proteins desmoglein-2 and E-cadherin in the lateral plasma membrane of the uterine epithelium of F. catus during pregnancy. We show that redistribution of desmoglein-2 and E-cadherin likely facilitates reduction of cell-to-cell adhesion allowing for implantation of the blastocyst and formation of the placenta. The ultrastructural and molecular changes to the uterine epithelium during early pregnancy in F. catus are similar to that in species with other levels of placental invasiveness, suggesting that key molecules such as desmoglein-2 and E-cadherin are crucial to successful pregnancy across all mammals. Anat Rec, 301:1497-1505, 2018. © 2018 Wiley Periodicals, Inc.

Microtubules are reorganised and fragmented for uterine receptivity.

For the development of uterine receptivity, many morphological and molecular changes occur in the apical surface of luminal uterine epithelial cells (UECs) including an increase in vesicular activity. Vesicular movements for exocytosis and endocytosis are dependent on microtubules; however, changes in microtubules in UECs during early pregnancy have received little attention. ß-tubulin, one of the main component of microtubules, is distributed throughout the cytoplasm of UECs on day 1 (non-receptive) of pregnancy in the rat. On day 5.5, ß-tubulin is concentrated above the nuclei and by day 6 (receptive), ß-tubulin is concentrated in a band-like fashion above the nucleus. Western blot analysis of isolated UECs found two bands (50 and 34 kDa) for ß-tubulin in UECs during early pregnancy. The intensity of the 34 kDa band was significantly higher on day 6 compared to day 1. The increase in the 34 kDa band may be due to higher proteolytic activity associated with microtubule polymerisation during the receptive state. Transmission electron microscopy showed fragmented microtubules at the time of receptivity in UECs. This is the first study to show that microtubules are reorganised during uterine receptivity. This re-organisation likely facilitates vesicular movement and promotes the reorganisation of the apical plasma membrane for uterine receptivity.

Change in distribution of cytoskeleton-associated proteins, lasp-1 and palladin, during uterine receptivity in the rat endometrium.

The epithelium of the uterine lumen is the first point of contact with the blastocyst before implantation. To facilitate pregnancy, these uterine epithelial cells (UECs) undergo morphological changes specific to the receptive uterus. These changes include basal, lateral and apical alterations in the plasma membrane of UECs. This study looked at the cytoskeletal and focal adhesion-associated proteins, lasp-1 and palladin, in the uterus during early pregnancy in the rat. Two palladin isoforms, 140 kDa and 90 kDa, were analysed, with the migration-associated 140-kDa isoform increasing significantly at the time of implantation when compared with the time of fertilisation. Lasp-1 was similarly increased at this time, whilst also being located predominantly apically and laterally in the UECs, suggesting a role in the initial contact between the UECs and the blastocyst. This is the first study to investigate palladin and lasp-1 in the uterine luminal epithelium and suggests an importance for these cytoskeletal proteins in the morphological changes the UECs undergo for pregnancy to occur.

Uterine and eggshell modifications associated with the evolution of viviparity in South American water snakes (Helicops spp.).

The evolution of viviparity requires eggshell thinning to bring together the maternal uterus and extraembryonic membranes to form placentae for physiological exchanges. Eggshell thinning likely involves reduced activity of the uterine glands that secrete it. We tested these hypotheses by comparing the uterine and eggshell structure and histochemistry among oviparous and viviparous water snakes (Helicops) using phylogenetic methods. Eggshell thinning occurred convergently in all three origins of viviparity in Helicops and was accomplished by the loss of the mineral layer and thinning of the shell membrane. Uterine glands secrete the shell membrane in both oviparous and viviparous Helicops. These glands increase during vitellogenesis regardless of the reproductive mode, but they always reach smaller sizes in viviparous forms. As there is no phylogenetic signal in eggshell thickness and gland dimensions, we conclude that interspecific differences are related to reproductive mode and not phylogeny. Therefore, our results support the hypothesis that eggshell thinning is associated with the evolution of viviparity and that such thinning result from a reduction in gland size in viviparous taxa. Interestingly, the shell membrane thickness of viviparous females of the reproductively bimodal Helicops angulatus is intermediate between their oviparous and viviparous congeners. Thus, although eggshell thinning is required by the evolution of viviparity, a nearly complete loss of this structure is not. However, uterine gland dimensions are similar across viviparous Helicops. Fewer glands or their functional repurposing may explain the thinner shell membrane in viviparous species of Helicops in comparison to viviparous females of the bimodal H. angulatus.

Non-invasive placentation in the marsupials Macropus eugenii (Macropodidae) and Trichosurus vulpecula (Phalangeridae) involves redistribution of uterine Desmoglein-2.

In mammalian pregnancy, the uterus is remodeled to become receptive to embryonic implantation. Since non-invasive placentation in marsupials is likely derived from invasive placentation, and is underpinned by intra-uterine conflict between mother and embryo, species with non-invasive placentation may employ a variety of molecular mechanisms to maintain an intact uterine epithelium and to prevent embryonic invasion. Identifying such modifications to the uterine epithelium of marsupial species with non-invasive placentation is key to understanding how conflict is mediated during pregnancy in different mammalian groups. Desmoglein-2, involved in maintaining lateral cell-cell adhesion of the uterine epithelium, is redistributed before implantation to facilitate embryo invasion in mammals with invasive placentation. We identified localization patterns of this cell adhesion molecule throughout pregnancy in two marsupial species with non-invasive placentation, the tammar wallaby (Macropus eugenii; Macropodidae), and the brushtail possum (Trichosurus vulpecula; Phalangeridae). Interestingly, Desmoglein-2 redistribution also occurs in both M. eugenii and T. vulpecula, suggesting that cell adhesion, and thus integrity of the uterine epithelium, is reduced during implantation regardless of placental type, and may be an important component of uterine remodeling. Desmoglein-2 also localizes to the mesenchymal stromal cells of M. eugenii and to epithelial cell nuclei in T. vulpecula, suggesting its involvement in cellular processes that are independent of adhesion and may compensate for reduced lateral adhesion in the uterine epithelium. We conclude that non-invasive placentation in marsupials involves diverse and complementary strategies to maintain an intact epithelial barrier.