Actin-binding proteins undergo major alterations during the plasma membrane transformation in uterine epithelial cells

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), transfer waste from the fetus to the mother and 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 propose directions for future research. We compare and discuss the complex modifications to the ultrastructure of uterine epithelial cells and show that there are similarities in the changes to the cytoskeleton and gross morphology of the uterine epithelial cells, 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.

Alterations in the distribution of actin and its binding proteins in the porcine endometrium during early pregnancy: Possible role in epithelial remodeling and embryo adhesion

During early pregnancy, uterine epithelial cells undergo major transformations in their cytoskeleton that make the endometrium receptive for conceptus attachment. Actin binding proteins (ABPs) such as cofilin, gelsolin, and vinculin are involved in regulating actin polymerization, severing or crosslinking actin to integrins. However, whether ABPs are involved in epithelial remodeling or embryo adhesion in pigs is unknown. Therefore, the expression and distribution of these proteins were investigated in porcine endometrium on Days 10 and 13 (pre-implantation period), and 16 (attachment phase) of the estrous cycle or pregnancy. While day and pregnancy status had no effect on ABP gene expression, the protein abundance of vinculin was significantly higher on Day 13 than on Day 10 (p < 0.05) of the estrous cycle, and its abundance was highest on Day 16 in the pregnant endometrium. Immunofluorescent staining showed alterations in the distribution of these proteins depending on the day of the estrous cycle or early pregnancy examined. Double immunofluorescent staining for the ABPs and actin revealed that while cofilin co-localized with actin in the apical epithelium on Days 13 and 16 of the estrous cycle, in pregnant animals, it was strongly associated with actin in the sub-epithelial stroma of the endometrium. Gelsolin was also co-localized with actin in the apical epithelium on Days 13 and 16 of the estrous cycle, but this association was absent in the pregnant endometrium. Vinculin co-localized with actin in the sub-epithelial stroma on Days 13 and 16 irrespective of the reproductive status, but was additionally associated with actin in the apical epithelium on Day 16 of pregnancy. Vinculin interacted with phosphorylated focal adhesion kinase in the endometrial epithelium, and the interaction was dependent on estradiol-17β, a conceptus-secreted pregnancy-recognition factor in pigs. Furthermore, silencing vinculin in the endometrial epithelial cells negatively affected trophoblast adhesion to them. In conclusion, the influence of stage and reproductive status on the specific localization of actin and its binding proteins in the porcine endometrium suggests that they play a role in regulating the endometrial cytoskeleton. Moreover, vinculin may facilitate conceptus attachment to the epithelium by interacting with focal adhesion kinase.

Actin crosslinking protein filamin A during early pregnancy in the rat uterus

During early pregnancy the endometrium undergoes a major transformation in order for it to become receptive to blastocyst implantation. The actin cytoskeleton and plasma membrane of luminal uterine epithelial cells (UECs) and the underlying stromal cells undergo dramatic remodelling to facilitate these changes. Filamin A (FLNA), a protein that crosslinks actin filaments and also mediates the anchorage of membrane proteins to the actin cytoskeleton, was investigated in the rat uterus at fertilisation (Day 1) and implantation (Day 6) to determine the role of FLNA in actin cytoskeletal remodelling of UECs and decidua during early pregnancy. Localisation of FLNA in UECs at the time of fertilisation was cytoplasmic, whilst at implantation it was distributed apically; its localisation is under the influence of progesterone. FLNA was also concentrated to the first two to three stromal cell layers at the time of fertilisation and shifted to the primary decidualisation zone at the time of implantation. This shift in localisation was found to be dependent on the decidualisation reaction. Protein abundance of the FLNA 280-kDa monomer and calpain-cleaved fragment (240 kDa) did not change during early pregnancy in UECs. Since major actin cytoskeletal remodelling occurs during early pregnancy in UECs and in decidual cells, the changing localisation of FLNA suggests that it may be an important regulator of cytoskeletal remodelling of these cells to allow uterine receptivity and decidualisation necessary for implantation in the rat.

Proteomic analysis of porcine endometrial tissue during peri-implantation period reveals altered protein abundance

In mammals, successful pregnancy depends upon the readiness of uterus for implantation, followed by correct communication between the endometrium and the developing conceptus. The objective of this study was to elucidate changes in protein abundance associated with progression of estrous cycle and pregnancy from Day 9 to Day 12. We analyzed porcine endometrial tissue lysates by 2D-DIGE. Abundance of several proteins was altered depending upon the pregnancy status of animals. MALDI-TOF/TOF was used to identify a number of these proteins. Endometrial proteins that increased from Day 9 to Day 12 of cycle included annexin A4, beta-actin, apolipoprotein, ceruloplasmin and afamin. Changes in protein abundances associated with conceptus secreted factors, including haptoglobin, prolyl-4-hydroxylase, aldose-reductase and transthyretin, were also observed. Functional analysis revealed that endometrial proteins with altered abundance on Day 12 irrespective of the reproductive status were related to growth and remodeling, acute phase response and free radical scavenging, whereas transport and small molecule biochemistry were the functions activated in the pregnant endometrium as compared to the cyclic endometrium. These data provide information on dynamic physiological processes associated with uterine endometrial function of the cyclic and pregnant endometrium during period of maternal recognition of pregnancy in pigs and may potentially demonstrate a protein profile associated with successful pregnancy.

BIOLOGICAL SIGNIFICANCE

In pigs, the fertility rates are generally very high but the early embryonic loss that occurs during the second and third weeks of gestation critically affects the potential litter size. Temporal changes that take place in the uterine environment during the period of early pregnancy in pigs and a cross-talk between the uterus and the embryo play an important role in embryonic survival and successful pregnancy. A better understanding of the molecular changes associated with these processes will pave way for understanding of endometrial functions and help towards increasing embryo survival. In this study, we present a 2D-DIGE based analysis of changes in porcine endometrial proteome that are associated with progression of cycle and progression of pregnancy. The network analysis of the results clearly revealed the pathways that are involved in rendering the endometrium receptive to the presence of embryo and also the changes that are result of molecular communication between the endometrium and the conceptuses. This comprehensive identification of proteomic changes in the porcine endometrium could be a foundation for targeted studies of proteins and pathways potentially involved in abnormal endometrial receptivity, placentation and embryo loss.

Cofilin and slingshot localization in the epithelium of uterine endometrium changes during the menstrual cycle and in endometriosis

Regulation of the actin cytoskeleton is essential for epithelial cell polarity and protein trafficking within human uterine epithelium. The actin-binding protein cofilin is involved in regulation of actin dynamics by promoting actin branching and cytoskeleton reorganization. Dual immunohistochemical staining of cofilin and G-actin (represented by DNAse I staining) revealed cofilin-G-actin colocalization in the apical side of luminal epithelial cells of human uterine endometrium during the proliferative phase of the menstrual cycle. Interestingly, during the secretory phase of the menstrual cycle, cofilin was only present on the basolateral side. To determine whether the disease endometriosis causes a different pattern of actin remodeling, we investigated an established baboon model of induced endometriosis. The cofilin pattern in the secretory phase of baboons with endometriosis was similar to the proliferative phase in normal animals; cofilin was observed in the apical parts of luminal and glandular epithelium. A phosphatase regulating the activity of cofilin, slingshot (SSH1), revealed a similar staining pattern within these tissues. These patterns were confirmed through quantitative image analysis. Quantification of messenger RNA (mRNA) detected upregulated SSH1 and suggested a progesterone resistance-related pattern of nuclear steroid hormone receptors, but no change in membrane progesterone receptors (mPR alpha or mPR beta) was observed in endometriosis. Our data indicate that the severe dyssynchrony during menstrual cycle phases in endometriosis is connected with improper cytoskeleton rearrangements. We suggest that cofilin-mediated actin reorganization in uterine epithelial cells might be important in preparation for blastocyst implantation; dysregulation of this reorganization may lead to decreased fertility in endometriosis.

Ezrin and EBP50 redistribute apically in rat uterine epithelial cells at the time of implantation and in response to cell contact

Uterine epithelial cells (UECs) undergo extensive morphological remodelling in preparation for an implanting blastocyst. This remodelling involves changes in the actin cytoskeleton and surface structures including microvilli. Ezrin and ezrin-radixin-moesin-binding protein-50-kDa (EBP50) link actin filaments to intra-membranous adhesion molecules and are important molecules in polarised epithelia. The current study is the first to describe the colocalisation and molecular association of ezrin and EBP50 in rat UECs by using immunofluorescence microscopy and immunoprecipitation techniques. These proteins have also been localised in relation to uterine epithelial cytoskeletal rearrangement during early pregnancy in the rat and to the effect of apical surface contact between opposing epithelial cells, blastocyst contact and contact with a silicon filament. Immunofluorescence microscopy has revealed that ezrin and EBP50 respond to contact between opposing epithelial cells and increase apically on day 6 of pregnancy. This apical distribution is also observed in UECs in contact with a silicon filament. Ezrin and EBP50 are however absent within the implantation chamber itself, seemingly mimicking the events that take place in leucocyte-endothelium binding. Thus, ezrin and EBP50 occur apically in UECs at the time of implantation in the rat and in response to a substitute blastocyst (filament) suggesting a role for these proteins in the cytoskeletal rearrangements that facilitate uterine receptivity and blastocyst-epithelial adhesion. Their loss within the implantation chamber possibly allows the subsequent invasion of the embryo.

Focal adhesions disassemble during early pregnancy in rat uterine epithelial cells

During early pregnancy in rodents, invasion of the blastocyst into the endometrial decidual cells is accompanied by the removal of uterine epithelial cells around the implantation sites. The present study investigated the distribution and expression of two focal adhesion proteins, namely talin and paxillin, in rat uterine epithelial cells during early pregnancy and their role in the loss of these cells at the time of implantation. A major distributional change of talin and paxillin was demonstrated in uterine epithelial cells during early pregnancy. From a highly concentrated expression along the basal cell surface on Day 1 of pregnancy, talin and paxillin were lost from the basal cell surface at the time of implantation. There was also a corresponding statistically significant decrease in paxillin seen through western blotting analysis. Together, these observations suggest that uterine epithelial cells are less adherent to the underlying basal lamina due to the disassembly of talin and paxillin from focal adhesions, facilitating removal of these cells at the time of implantation. This phenomenon was restricted to the period of receptivity because talin and paxillin reappeared along the basal cell surface soon after implantation.

Moesin is involved in the cytoskeletal remodelling of rat decidual cells

The extensive actin cytoskeletal remodelling of the uterine stroma during early pregnancy involves changes in actin-binding proteins. This study provides the first detailed localisation of the actin-binding protein, moesin, in rat uterine endometrium during this period. Western blot analysis and immunofluorescence microscopy showed that the amount of moesin in the uterus peaked at the time of implantation, corresponding to the presence of intensely immunolabelling decidual cells. Furthermore, moesin increased in active membrane/cytoskeleton bound protein at the time of implantation, concomitantly decreasing in cytosolic protein. The increase of moesin at decidualisation corresponds with the appearance of alpha smooth muscle actin (alphaSMA) suggesting that decidual cells have contractile abilities that may aid in containing an invasive trophoblast. The results of this study suggest that moesin is important in developing a specialised cytoskeleton and increased adhesiveness of decidual cells, possibly functioning to bridge adhesion molecules to the underlying cytoskeleton.

Actin binding protein expression is altered in uterine luminal epithelium by clomiphene citrate, a synthetic estrogen receptor modulator

Clomiphene citrate (CC), a synthetic oestrogen, is often prescribed as a superovulator in treating infertility. Although CC works efficiently, pregnancy rates following CC treatment are approximately 10 times lower than "natural" rates. This study investigates how a dose of 1.25 mg CC given to ovariectomized rats before the implantation priming hormones (a single dose of progesterone for 3 days and a dose of estradiol-17beta on d3, P-P-PE), alters the expression and distribution of alpha-actinin, gelsolin and vinculin. Actin binding proteins show a specific distribution within the uterine epithelium during implantation, linking the actin cytoskeleton to integrin expression on the uterine surface and in this way aiding "adhesiveness" for blastocyst apposition to the uterine epithelium. In this study, immunocytochemistry on frozen uterine sections using mouse monoclonal antibodies against alpha-actinin, gelsolin and vinculin and peroxidase-conjugated secondary antibodies, show that CC, administered before the P-P-PE regimen, down-regulates the expression of vinculin, does not alter the expression of gelsolin and up-regulates alpha-actinin on the uterine apical surface, when compared to P-P-PE treated animals. All three proteins are down-regulated on the apical surface of the luminal epithelium and glands in all groups when compared to pregnant controls. Vinculin was only localized in the basolateral compartment of the uterine epithelial cells in the CC treated groups. By down-regulating these proteins on the uterine surface and up-regulating vinculin on the basolateral membrane of the epithelium, CC may impede adhesion and invasion of blastocysts at implantation. These results may aid the exogenous manipulation of uterine tissue to control fertility and improve assisted reproductive out-comes.

The cytoskeletal proteins alpha-actinin, Ezrin, and talin are De-expressed in endometriosis and endometrioid carcinoma compared with normal uterine epithelium

In this retrospective study on banked tissue, we found that alpha-actinin and talin were completely de-expressed in both endometriosis and endometrioid carcinoma tissue. Some patchy, depolarized labeling for ezrin was noted in the endometrioid carcinoma but not in endometriosis. The loss of these proteins in both endometriosis and endometrioid carcinoma tissue indicates a significant change in the integrity of these tissues compared with normal and the possibility that individual cells may break away from the parent histology due to loss of cell adhesion. It also indicates a similarity between endometrioid cancer and endometriosis with respect to epithelial cell function and adhesion.

In vitro studies on endometrial adhesiveness for trophoblast: cellular dynamics in uterine epithelial cells

Initiation of embryo implantation involves adhesion of trophoblast cells to the epithelial lining of the endometrium. The mechanisms regulating the adhesive properties of the uterine epithelium for trophoblast during initiation of human embryo implantation, however, are still incompletely understood. We report here on model studies that we have performed in our laboratory, and in particular on certain methodological approaches that seem to yield new insight into basic mechanisms involved. Of central interest is the ability of the uterine epithelium to develop an adhesion competence at its apical cell pole. This confronts us with a cell biological paradox in that adhesion must be established at the pole which in simple epithelia is typically specialized to resist adhesion. Gain of apical adhesion competence by uterine epithelial cells should be related to cellular rearrangements, i.e. a modulation of their apicobasal cell polarity. Here, we used monolayer-cultured uterine epithelial RL95-2 cells as an in vitro model for the human receptive uterine epithelium. We demonstrated that formation of stable cell-to-cell bonds between the free (apical) pole of these cells and attaching trophoblast (modelled by JAr cells) depends on a number of structural and functional peculiarities that RL95-2 cells have in contrast to other uterine epithelial cells (HEC-1-A cells) which resist attachment via this cell pole. RL95-2 cells were shown to lack tight junctions and to exhibit only rudimentary adherens junctions and a non-polar organization of the actin cytoskeleton. Using the atomic force microscope in a force spectroscopy mode, we exactly defined the time dependence of adhesive interactions between RL95-2 cells and trophoblast, measured the pressure force needed to initiate this process, and screened the buildup of the adhesive forces between the binding partners. A dynamic interaction between the actin cytoskeleton and integrins (a prerequisite for functional activity of integrins) was shown to be an important aspect of the adhesive properties of RL95-2 cells. In addition, at least two types of calcium channels in the plasma membrane of RL95-2 cells seem to play a role in activation of a variety of calcium-sensitive response mechanisms including adhesiveness for trophoblast, i.e. diltiazem-sensitive channels seem to contribute to the initiation of JAr cell binding and SKF-96365-sensitive channels to participate in a feedback loop that controls the balance of bonds. By extrapolation, these data suggest an active role of the uterine epithelium in the process of embryo implantation which we are just beginning to understand in terms of its cell biology.

Cytoskeletal proteins in uterine epithelial cells only partially return to the pre-receptive state after the period of receptivity

Immunohistochemical staining of 5 cytoskeletal proteins (actin, alpha-actinin, gelsolin, plectin and plakoglobin) was used to investigate changes in distribution patterns of these proteins after the period of uterine receptivity for blastocyst implantation in the rat. Actin was found throughout the cytoplasm but it was concentrated along the apical plasma membrane on day 1 of pregnancy, decreased by day 6 and then increased again at day 9. Alpha-actinin and gelsolin were localized in distinctive bands along the apical plasma membrane at day 6 of pregnancy but became diffusely distributed at day 9. Plectin was localized along the apical and basal plasma membranes at day 6 but in higher amounts apically and at day 9, it was concentrated in apical and basal zones in the cells. Plakoglobin was found along the lateral and basal membranes with increased intensity along the apical third of the lateral plasma membrane from day 6 to day 9 of pregnancy. These results show that all 5 cytoskeletal proteins redistributed after the period of uterine receptivity: some exhibited a similar pattern of labelling to that found during the prereceptive state, whereas others only partially returned to the pre-receptive state. This change in distribution patterns may reflect differences in the epithelial barrier function before and after the period of receptivity.

Tenascin, E-cadherin and P2X calcium channel receptor expression is increased during rat blastocyst implantation

The calcium-activated cell-adhesion proteins tenascin, E-cadherin and the purinergic (P2X) calcium channel receptors are expressed in an identical spatial and temporal pattern in uterine epithelium in the rat during implantation. On Day 1 of pregnancy (estrous), a diffuse cytoplasmic and specific basement membrane label for each of the proteins was observed throughout the uterine epithelium. On Day 3 of pregnancy, a specific and prominent lateral plasma membrane label for each protein was seen. At the time of implantation on Day 6, an additional and significant increase in the label for each was observed on the apical epithelium. At this time, the label for tenascin in the apical epithelium was increased 2.1-fold (p < 0.0004), that of E-cadherin was increased 2.5-fold (p < 0.0001) and the P2X receptor label was increased 2.0-fold (p < 0.0001). These observations suggest a major role for the calcium-activated adhesion proteins tenascin and E-cadherin in attachment and implantation, with ionic calcium for protein activation possibly provided by the P2X calcium channels. These events occur along the entire length of the uterine epithelium in preparation for blastocyst adhesion.

Thrombospondin is sequentially expressed and then de-expressed during early pregnancy in the rat uterus

The expression of thrombospondin on Day 1, Day 3 and Day 6 of pregnancy has been examined in the rat, using light microscopic immunoperoxidase and electron immunogold techniques. The glycoprotein was expressed in the apical, lateral and basal uterine epithelium on Days 1 and 3 but was then de-expressed at the time of implantation on Day 6. We propose that these data suggest a role for thrombospondin in remodellig the uterine epithelium during the plasma membrane transformation, but that it does not play a part in attachment and implantation.

Junctional plaque proteins shift to the apical surface of uterine epithelial cells during early pregnancy in the rat

Antibodies against the cytoplasmic plaque molecules, plectin and plakoglobin, and cytokeratin, the molecular component of intermediate filaments (IFs), were used to examine the distribution of these molecules in rat uterine epithelial cells during early pregnancy including the period of blastocyst implantation. On day 1 of pregnancy plectin was detected in concentrated bands along the apical and basal plasma membranes, and diffusely throughout the cytoplasm. Plakoglobin was found along the entire lateral plasma membrane on day 1. By day 6, the time of blastocyst implantation, plectin was localised along the apical and basal membranes and reduced in the basal cytoplasm, and plakoglobin was seen exclusively at the apical-most quarter of the lateral plasma membrane. Cytokeratin was detected throughout the cytoplasm on day 1, but by day 6, was localised to the apical region of the cytoplasm only. These results show a redistribution of plectin, plakoglobin and cytokeratin away from the basal region of the uterine epithelial cells. The change in distribution of these molecules may contribute to the adhesion of the blastocyst to the apical and lateral surfaces of uterine epithelial cells and the subsequent detachment of the uterine epithelium from the basal lamina.