| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Molecular Embryology
b and Dairy Science, AgResearch, Ruakura Research Center, Hamilton, New Zealand
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONNOTE ADDED IN PROOFREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONNOTE ADDED IN PROOFREFERENCES |
|---|
Proteins in the uterine luminal fluid (ULF) of these mammals, such as sheep [2, 3], cattle [4, 5], and pigs [6], have been the subject of many previous studies that sought to understand how the biochemistry of the uterine environment adapts to the presence, and supports the development, of the conceptus. In this study, we chose to focus our search for pregnancy-associated proteins in the ULF of sheep on the period between blastocyst expansion (Day 12) and the initial attachment of the conceptus to the endometrium (Days 17 and 18). This period of rapid growth is characterized by primary hematopoiesis and vasculogenesis in the yolk sac and the formation of primitive pigmented blood with the ability to transport oxygen by Day 16. In addition, the allantois begins to emerge as a diverticulum of the embryonic hind-gut from Day 16 onwards, growing rapidly and increasing in length from 2.17 ± 0.36 mm at Day 17 to 31.66 ± 3.51 mm by Day 19 [7]. This growth is accompanied by extensive vascularization and hematopoiesis. By the time attachment to the uterine epithelium is initiated, the trophoblast may extend throughout most of both uterine horns, and the embryo has a beating heart, a primitive liver, and a functioning mesonephros [8]. It is around this stage of pregnancy that most embryonic loss occurs [9, 10].
Most of the ULF proteins that have been studied are highly abundant, and this has allowed the use of standard protein purification methods to enrich for and isolate sufficient material for identification by peptide sequencing or for use as immunogens to generate antibodies for the screening of cDNA expression libraries. Some proteins that have been identified by the use of the above approaches are the ovine uterine milk proteins belonging to the serpin superfamily [11], the Kunitz-type proteinase inhibitors of the pig [12], uteroferrin [13] and retinol-binding protein from the pig [14], and interferon-
, a conceptus-secreted cytokine from the sheep [15–17]. However, many more proteins have been detected and described in ULF but remain unidentified.
In this study, we used the excellent resolution afforded by large-format two-dimensional (2D) polyacrylamide gels [18] in separating complex polypeptide mixtures to identify proteins in ULF that are associated with pregnancy. We took advantage of the higher loading capacities of these gels to obtain sufficient material from protein spots for direct peptide microsequencing after proteolytic fragmentation, thus avoiding the need to enrich for and purify the proteins of interest before identification. This approach allowed us to study the qualitative changes in protein composition in ULF during preimplantation development and, in addition, to identify the proteins of interest. Furthermore, in analyzing separately the proteins that were secreted by the trophoblast, yolk sac, allantois, or the endometrium, we were able to determine the contribution of each of these components to the proteins observed in ULF. We report here the identification of six protein spots by peptide sequencing and two others by deduction from published 2D gel maps and verification by Western blotting. The results have provided new insights into the biology of the early sheep conceptus.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONNOTE ADDED IN PROOFREFERENCES |
|---|
Cycling, mixed-age ewes were treated with controlled internal drug-releasing (CIDR-G; InterAg, Hamilton, NZ) breeding devices for 12 days to synchronize estrus. Intact rams, wearing mating harnesses with colored markers, were introduced at device removal, and the ewes were checked for markings twice daily. The first detected marking was designated Day 0 of gestation. To provide nonpregnant control ewes, anestrous sheep were treated similarly except that at device removal, 800 IU eCG (Horizon Animal Reproduction, North Ryde, Australia) was administered i.m. before the introduction of vasectomized rams. This treatment induces estrus only once, and there is no subsequent cycling activity, thus avoiding any contribution of subsequent pre-estrous surges of estradiol-17ß on the uterine environment of control ewes. At different times postmating, ewes were electricaly stunned and killed by exsanguination, the reproductive tracts were removed, and conceptuses were recovered by flushing each uterine horn with 10 ml isotonic saline. The number of animals used in this study were as follows: Days 12 and 14 pregnant (n = 3 each); Days 12 and 14 nonpregnant (n = 3 each); Days 15 and 17 pregnant (n = 15 each); Days 15 and 17 nonpregnant (n = 5 each); Day 18 pregnant (n = 16).
Tissue Culture and Metabolic Labeling
Conceptuses were cultured either intact or after dissection into the trophoblastic, allantoic, and yolk sac membrane components within 1 h of collection from the uterus. In one case, the trophoblast was separated into the trophectodermal and mesodermal layers by gently teasing the two cell layers apart. After initially culturing for 20 min in methionine-free Dulbecco's modified Eagle's medium (Gibco Life Technologies, Gaithersberg, MD), conceptuses were placed in medium containing 250 µCi/ml [35S]methionine (TransLabel; ICN, Irvine, CA) at 37°C in 5% CO2:95% air. After 3 h, culture media were removed, and conceptuses were rinsed with PBS and stored at -70°C. The washings were combined with the culture media and centrifuged at 1200 x g for 10 min, and the supernatant was recovered. Finely minced endometrial tissue was rinsed twice in PBS to remove serum proteins, then placed in culture medium containing 250 µCi/ml [35S]methionine. Endometrial explants were cultured for either 3 or 24 h under the above conditions, and the culture supernatants were treated as described above. Culture supernatants were obtained from the following: Day 17 trophoblast (n = 6); Day 17 allantois (n = 3); Day 17 yolk sac (n = 6); Day 18 whole conceptuses (n = 3); pregnant endometrium (n = 6); and nonpregnant endometrium (n = 4).
Preparation of Protein Samples for Electrophoretic Analyses
Uterine flushings from individual animals were lyophilized, redissolved in 2 ml of water containing 1 mM each of PMSF and EDTA, and then dialyzed against water at 4°C overnight, using 6–8-kDa cutoff dialysis membranes (Spectrum Medical Industries, Houston, TX). Culture supernatants, with PMSF added to 1 mM final concentration, were dialyzed against two changes of distilled water at 4°C for 24 h to remove unincorporated label and then lyophilized. Proteins were prepared from cell lysates by sonicating material from conceptuses in a 2D gel loading buffer containing 1 mM PMSF. Two microliters of each labeled protein sample was applied to 1-cm2 squares of Whatman 3MM filter paper, and the amount of trichloroacetic acid-precipitable label was estimated by scintillation counting. Incorporation of label was determined in duplicate samples. Protein content was estimated using the method of Bradford [19], with BSA as the standard. ULF samples from within each of the groups of animals were combined before 2D electrophoretic analyses. For samples collected from pregnant animals on Days 15, 17, and 18, ULF samples from 3–5 animals from within each group were combined to give several sample pools for each gestation day.
Gel Electrophoresis
One-dimensional (1D) gel electrophoresis was carried out as described by Laemmli [20]. Two-dimensional PAGE was carried out essentially as described by O'Farrell [21], except that large-format gels (35 cm x 35 cm) were used, as previously described [18, 22]. Briefly, up to 1000 µg of lyophilized total protein was dissolved in 2D gel loading buffer [18] and loaded onto each prefocused 35-cm x 25-mm (i.d.) tube gel containing 2% (w:v) ampholytes (pH 3.5–10) and 2% (w:v) ampholytes (pH 5–7) (LKB-Pharmacia, Uppsala, Sweden), and proteins were separated by isoelectric focusing. The gels were extruded from the tubes, equilibrated with 62.5 mM Tris-HCl pH 6.9 containing 10% (w:v) glycerol, 5% (v:v) 2-mercaptoethanol, and 2% (w:v) SDS, then attached to 0.075 x 40 x 40-cm 12% (w:v) acrylamide SDS gels, and subjected to electrophoresis in the second dimension. After electrophoresis, proteins were detected by Coomassie blue staining or, alternatively, the gels were fixed in 50% (v:v) methanol, 5% (v:v) acetic acid and dried, and proteins were detected by autoradiography for samples containing > 1 x 106 cpm. For protein samples with < 1 x 106 cpm, gels were soaked in 1 M sodium salicylate for 1 h [23], then rinsed briefly in water before drying. Gels were then subjected to fluorography, using pre-flashed Kodak X-Omat film [24].
For Western analysis, proteins were electroblotted from polyacrylamide gels to nitrocellulose (Schleicher and Shuell, Dassel, Germany) in 10 mM 3-(cyclohexyaminol)-1-propanesulfonic acid (pH 11) containing 10% (v:v) methanol. Transferred proteins were stained with 0.1% (w:v) Ponceau S in 1% (v:v) acetic acid. Membranes were blocked by incubation with 4% (w:v) nonfat dried milk in 10 mM Tris-HCl (pH 7.5), 0.15 M NaCl, 0.1% (v:v) Tween-20 (TBS-Tween), followed by incubation for 2 h with primary antibody in the same buffer containing 0.1% (w:v) BSA. After washing with TBS-Tween twice for 5 min, the membranes were then incubated for 2 h with horseradish peroxidase (HRP)-conjugated antibody. After two 5-min washes with TBS-Tween and one 5-min wash with TBS, immunoreactivity was detected using hydrogen peroxide and 3,3'-diaminobenzidine tetrahydrochloride dihydrate (Life Technologies). Antibodies to human transferrin, cytoplasmic actin, and HRP-conjugated goat anti-rabbit immunoglobulins were obtained from Sigma (St. Louis, MO). HRP-conjugated rabbit anti-goat antibody was obtained from Dako (Carpinteria, CA).
Peptide sequences were obtained from protein spots of interest by excision of the spot from the gel, transferring to nitrocellulose membranes, and digestion with 1 µg trypsin (sequencing grade, Promega (Madison, WI), as described by Aebersold [25]. Alternatively, protein was digested in situ in crushed acrylamide gel, using endoproteinase Lys-C (sequencing grade; Promega), and the peptides were eluted from the gel fragments by shaking at 37°C overnight. Peptide mixtures were subjected to reverse-phase HPLC in a narrow-bore column, using the following buffer system: buffer A—0.1% trifluoroacetic acid (TFA) in water; buffer B—0.08% TFA in acetonitrile/water, 80:20 (v:v), essentially as described by Wheeler et al. [26]. Chromatographic peaks were collected, and the peptide fragments were subjected to amino acid sequencing by Edman degradation, using the Applied Biosystems Model 470A amino acid sequencer (Foster City, CA).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONNOTE ADDED IN PROOFREFERENCES |
|---|
|
Serum Proteins
Comparison of 2D gels of ULF proteins with gels of serum proteins enabled us to determine which protein spots were serum proteins and which were derived from other sources. In ULF samples from Days 12 to 14, serum proteins made up at least 50% of the total proteins, and hardly any of them differed in abundance between pregnant and nonpregnant animals. The identities of some of these proteins were deduced by comparison with published 2D electrophoresis maps of plasma proteins [28]. Those proteins determined to be serum-derived are designated "S1-S7" in Figure 1; although many more were detected, not all are marked in the figure. The most abundant serum proteins in ULF were albumin (S1) and the immunoglobulins, the latter observed as a broad diffuse band at approximately 55 kDa. We observed one serum protein in ULF whose abundance was altered in pregnancy. This protein (S2 in Fig. 1) was more abundant in pregnant than in nonpregnant ULF at Day 17 (Fig. 2, a and b, respectively) and was also present in culture supernatants of endometrial tissue explants (Fig. 2c). Amino acid sequencing of two peptide fragments derived from the S2 protein revealed 100% identity with bovine apolipoprotein A1 (Apo A1; see Table 1).
|
|
Four of the major protein spots (S3-S6) in pregnant and nonpregnant ULF were found in culture supernatants of endometrial tissue explants but were absent in serum or culture supernatants from conceptuses. However, these proteins were not labeled when the explants were cultured in the presence of [35S]methionine. The S3 and S4 proteins, after proteolytic fragmentation, produced almost identical peptide profiles on HPLC separation, and amino acid sequencing of three peptides identified S3 and S4 as ovine serum albumin (Table 1). S5 and S6 also gave almost identical peptide peak profiles to each other, but these were different from those from S3 and S4. One peptide from a peak common to both S5 and S6, and another, unique to S5, were sequenced. Again, the sequences showed identity with serum albumin (Table 1). Peptide sequences from S3 and S4 mapped to the C-terminal two-thirds of the albumin molecule, whereas those from S5 and S6 mapped closer to the N-terminal portion. However, one peptide fragment from S4 (C-terminus fragment) mapped immediately upstream to a peptide from S5 (N-terminus fragment), indicating that polypeptides S4 and S5 overlapped and were therefore not complementary fragments of the albumin molecule. These four polypeptides were most likely proteolytic fragments of serum albumin.
Proteins designated as S7 in Figure 1, also shown in Figure 3a, migrated in a position similar to that of a group of proteins identified as transferrin in 2D gels of plasma proteins [28]. This same group of proteins in sheep serum was immunoreactive with antibodies against human transferrin (Fig. 3b), strongly suggesting that S7 in Figures 1 and 3a was transferrin. The S7 group of proteins was also detected in culture supernatants of yolk sac membranes (Fig. 3c); this finding is described in more detail below. Unfortunately, we were not able to perform Western blot analysis on these samples because of lack of material for further analysis.
|
Conceptus-Secreted Proteins
Conceptus-secreted proteins in ULF were identified by comparing 2D gels of proteins from pregnant ULF with [35S]methionine-labeled proteins secreted by whole conceptuses or isolated extraembryonic membrane components. We cultured the yolk sac in groups of three because of the small amount of tissue, whereas trophoblast, allantois, and whole conceptuses were cultured individually. Many proteins were secreted by whole Day 18 conceptuses after 3 h in culture, as detected by autoradiography of the 2D gels (Fig. 4). The majority of the proteins in Figure 4 were different from proteins of whole lysates of the conceptus (results not shown). The yolk sac was the most metabolically active of the extraembryonic membranes, incorporating more [35S]methionine when adjusted for relative size, than did the other extraembryonic membranes. In this study, the average amount of labeled proteins secreted by each yolk sac was 0.58 x 106 cpm after 3 h of incubation. In contrast, the Day 17 trophoblast, which was visually estimated to be several hundred-fold larger in surface area, secreted 1.48 ± 0.29 x 106 cpm labeled proteins in the culture medium, only 2- to 3-fold higher than the yolk sac. The yolk sac was the source of all the groups of proteins within boxes in Figure 4. Interestingly, most of these were probably identical to glycoproteins found in the serum as they migrated to very similar positions on the 2D gels. One of these proteins was transferrin (Fig. 3c and protein S7 in Fig. 4).
|
The Day 17 allantois secreted 0.65 ± 0.07 x 106 cpm labeled proteins after 3 h in culture; examples of these are shown in Figure 4. Although composed of the same primary germ layers as the yolk sac, the proteins secreted were significantly different from those produced by the yolk sac, e.g., the serum proteins of the yolk sac were absent from allantoic cultures. These allantoic proteins remain to be identified.
The two most abundant pregnancy-associated proteins, C1 and C2 (Figs. 1 and 4), were found among proteins secreted by the trophectoderm, as was interferon (IFN), the antiluteolytic factor in ruminants [15–17]. The C1 and C2 proteins were sufficiently abundant in ULF to be detectable by Coomassie blue staining in 1D SDS polyacrylamide gels (not shown). Other examples of proteins secreted by the trophectodermal layer are indicated with arrowheads in Figure 4. The two cell layers of the trophoblast secreted quite distinct complements of proteins; examples of those secreted by the mesodermal layer are shown enclosed by dotted boxes in Figure 4. C1 protein migrated in 2D gels at a position very similar to cytoplasmic actin and was immunoreactive with an antibody against cytoplasmic actin (not shown). After 3 h, C1 in the culture supernatant of trophoblast was metabolically labeled. No C1 equivalent was detected in 2D autoradiographs of proteins from the culture medium of allantois or yolk sac membranes treated under identical culture conditions at the same time. C1 was as abundant in the culture supernatant as in whole lysate proteins of trophoblast when equivalent amounts of total protein were subjected to 2D PAGE (results not shown). Using immunoblotting from SDS-PAGE, actin was detected in pregnant ULF from Day 15 onwards (Fig. 5a), but not in Day 10–14 pregnant samples nor in the equivalent nonpregnant samples. Allantoic fluid from in vivo-derived conceptuses that were never subjected to culture also contained large amounts of actin, even though no actin was detected in the culture medium of the allantois itself. The amount of actin in allantoic fluid was equivalent to that found in Day 18 pregnant ULF and approximately half that found in whole homogenates of mammary tissue (Fig. 5b). Like the yolk sac and allantois culture medium, the culture medium of mammary tissue explants did not contain actin.
|
The identity of C2 was determined by amino acid sequencing of two tryptic fragments of the protein. These yielded sequences that were 100% homologous (Table 1) to both human and bovine aldose reductase (alditol:NAD(P)+ 1-oxidoreductase, EC 1.1.1.21), a member of the superfamily of monomeric NADPH-dependent oxidoreductases [29]. Like C1, labeled C2 was found only in the culture supernatants of trophoblast but not in allantois or yolk sac cultures.
The majority of conceptus-secreted proteins in Figure 4 did not stain with Coomassie blue dye in 2D gels. We noted, however, several conceptus-associated proteins that stained with Coomassie blue dye but did not incorporate [35S]methionine. These included a group of 14-kDa proteins (E1, Fig. 1), a 39-kDa protein, and two proteins of 60 kDa (indicated by arrowheads in Fig. 1). The source of these latter three proteins was not determined.
Endometrial-Secreted Proteins
To determine the maternal contribution to proteins in ULF, endometrial explants were cultured in the presence of [35S]methionine for 3 or 24 h. Compared to conceptuses, endometrial tissue was far less active in protein synthesis and secretion, so that after 3 h, there were insufficient labeled proteins for analysis by 2D PAGE and fluorography. Even culture supernatants after 24-h incubation yielded lower amounts of labeled proteins than supernatants from conceptuses, and the number of labeled polypeptide species secreted were few in comparison (not shown).
One of the most abundant proteins found in pregnant ULF from Day 14 onwards was the 14-kDa group (E1, Figs. 1 and 6a). We also observed much lower levels of E1 in both pregnant and nonpregnant ULF at Day 12. Analysis of proteins in the culture supernatants of endometrial explants at Day 17 of gestation revealed a group of labeled proteins in only the pregnant samples, migrating in a position in 2D gels identical to that of the E1 from ULF (Fig. 6b). These proteins were labeled after 24 h in culture, demonstrating that they were synthesized in endometrial tissue. Large amounts of the E1 proteins were found to be secreted by the trophoblast during 3-h in vitro culture. In addition, E1 was also found in trophoblast lysates (not shown) and allantoic fluid at Day 18 (Fig. 6c), but was never labeled when conceptuses were cultured with [35S]methionine. These results suggest that the E1 group of proteins is synthesized by endometrial tissue in pregnancy, secreted into the uterine lumen, and then taken up and sequestered by the conceptus.
|
Proteins in the Allantoic Fluid
The proteins in allantoic fluid pooled from two whole Day 18 conceptuses after 3-h incubation in [35S]-methionine were analyzed; over 300 proteins were detected by autoradiography (Fig. 7). Proteins secreted by the trophoblast, yolk sac, and allantois all contributed to the labeled proteins. The serum glycoproteins produced by the yolk sac (see Fig. 4) were highly represented in the allantoic fluid. Other proteins found in the allantoic fluid included aldose reductase (C2), cytoplasmic actin (C1), IFN, transferrin (S7), and E1 (Fig. 6c). The latter protein was not metabolically labeled and so is not visible in the autoradiograph shown in Figure 7. Thus, many but not all conceptus-secreted proteins as well as the endometrial-secreted protein E1 appeared to accumulate in the allantoic fluid.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONNOTE ADDED IN PROOFREFERENCES |
|---|
The majority of pregnancy-associated proteins detected in Day 17 and 18 ULF were derived from the conceptus. At this stage of development, both the trophoblast and allantois are undergoing extensive growth, whereas the yolk sac is rapidly regressing and is only a very small fraction of the mass of the trophoblast. Despite this, the yolk sac continues to synthesize large amounts of serum proteins, and one identified in this study is transferrin, an iron-binding and transport protein, which is also an important factor for cell function, differentiation, and proliferation. Our finding that the yolk sac synthesizes transferrin from Day 15, just before vasculogenesis and hematopoiesis, and before the differentiation and growth of the allantois, demonstrates its importance as a supplementary source for transferrin at a time when the demand for iron is thought to be high. It is likely that the transferrin in pregnant ULF is derived from both maternal serum and the conceptus. The extrahepatic synthesis of transferrin during embryonic development has also been shown to occur in the mouse visceral yolk sac [33] as well as in the human yolk sac [34]. Although both yolk sac and allantois are composed of mesodermal and endodermal germ cell layers, and both are hematopoietic, it is interesting that only the yolk sac synthesizes transferrin.
From Day 12, the trophectodermal cells appear to accumulate or synthesize large amounts of lipid [35, 36]; however, lipid metabolism in the ovine conceptus is still poorly understood. Our finding that Apo A1 was increased in pregnant ULF at Day 17 suggests that this protein, which is a major component of high-density lipoprotein, may play a role in lipid transport and/or accumulation. We were unable to determine whether Apo A1 was synthesized by the yolk sac as we could not be sure of its identity in the autoradiograph in Figure 4 in the absence of characteristic neighboring spots. Synthesis by the yolk sac has been reported for the mouse [33]. Alternatively, increased levels in pregnant ULF could be due to increased levels in maternal serum at pregnancy. In humans, some serum proteins, such as the acute-phase proteins 
1-antitrypsin and ceruloplasmin, are elevated during pregnancy [37]; however, we do not know if this is the case with Apo A1 in the sheep.
Fructose is the predominant hexose in fetal fluids and fetal blood of ungulates [38]; in the late pregnant ewe, conversion of glucose to fructose occurs on passage of glucose through the placenta. Aldose reductase, which catalyzes the conversion of glucose to sorbitol, has been purified from placenta of the late pregnant ewe [39]. Our finding that aldose reductase was synthesized before implantation suggests that the switch to fructose synthesis may occur much earlier than previously thought. Synthesis of sorbitol and fructose in early sheep pregnancy has been implicated in the work of Wales and Murdoch [40], who found both hexoses in Day 22 ovine allantoic fluid; however, the site of synthesis of these hexoses was not determined. We have shown in this study that the trophoblast of the preimplantation conceptus produces one of the enzymes, aldose reductase, required for the synthesis of these hexoses. Although aldose reductase is also present in ULF and allantoic fluid at Days 17 and 18, whether it has a metabolic role there is unknown. As the amount of aldose reductase in ULF appears to be far in excess of that required for hexose metabolism, it is possible that the protein has another unidentified function(s). An example of dual function for some metabolic enzymes is found in the lens of the eye. Several members of the aldose reductase superfamily have been recruited as soluble structural proteins (crystallins) in the ocular lens [41, 42].
Among the pregnancy-associated proteins observed in ULF at Day 17, actin (C1) was one of the most abundant. Although actin is not generally considered a secretory protein, we consider its presence in the culture medium and in ULF to reflect secretion for the following reasons. Its presence in ULF recovered with or without flushing the tracts, and in the allantoic fluid of conceptuses that were never cultured, implies that extracellular actin is normally found in the uterine tract at this stage of pregnancy. In vitro culture of the extraembryonic membranes separately showed that the trophectoderm was the source of this actin, and its labeling with [35S]methionine after 3–6 h in culture suggests that it was actively sequestered extracellularly rather than released as a consequence of cell death and lysis due to poor culture conditions (see Note Added in Proof). Occasional dead and disintegrating cells among apparently healthy and sometimes mitotic cells in the trophectoderm have been reported in ultrastructural studies of Day 12–16 sheep trophoblast [35]. It was suggested that this was due to programmed cell death and appeared to coincide with the initial formation of the binucleate cells in the trophectoderm. We have stained portions of trophoblast with trypan blue dye immediately after flushing the conceptus from the uterus and have observed very few cells that stained positively. The number of dead cells we observed appeared far too low to account for all the actin we detected in all our Day 17 pregnant ULF samples, though it is possible that they contributed to some extent. In our study, we cultured trophoblast in large intact sheets rather than minced, as other investigators have done, and we have never observed the level of cell lysis that may be expected to produce the amount of actin we detected in the culture supernatants after only 3-h culture. It is interesting to note that Wildeman et al. [43] have also reported the presence of a protein with characteristics of actin in the culture medium of a cell line derived from bovine trophoblastic vesicles. Other observations argue against cell death and lysis contributing to the cellular proteins detected in the culture medium in our study. More than 85% of the protein spots in 2D gels of culture supernatant differed from cell lysate proteins (data not shown). This was particularly obvious with known secretory proteins, such as the serum glycoproteins and IFN-, which were either markedly under-represented or virtually absent in cell lysates. Furthermore, in all trophoblast lysates, two isoforms of actin were detected, whereas in culture supernatants, one form was predominant. Clearly, the majority of proteins in our culture medium were not intracellular proteins.
An ectocytotic mechanism could explain the apparent secretion of cellular proteins such as actin and aldose reductase by the trophoblast. Gudeman et al. [44] reported the formation of small vesicles, which they called aposomes, from live choroid plexus explants in culture. Such aposomes continued to incorporate labeled amino acids and contained actin, but not the major secretory protein of this tissue, transthyretin. In another study by Lee et al. [45], fibroblasts undergoing stress-relaxation formed actin clusters along the cell margins, and vesicles were observed to bud off from the plasma membranes. These vesicles had a very specific cytoskeletal composition and contained actin, annexins II and VI, and ß1 integrin receptors but not tubulin, vimentin, or vinculin. Furthermore, the proteins within these vesicles were labeled when the fibroblasts were cultured in the presence of [35S]methionine. It is possible that the extracellular actin (and perhaps the aldose reductase) we observed in pregnant ULF and in culture supernatants was secreted by vesicular budding from the trophectodermal cells. If so, the low-speed centrifugation used to pellet cellular debris may have been insufficient to remove these vesicles from the supernatant. Such a mechanism could constitute a pathway by which cells export cytoplasmic proteins that lack a signal sequence. The appearance of actin in ULF coincided with the period of rapid growth and differentiation of the conceptus and the initiation of attachment of the trophoblast to the uterine epithelium. Actin could form one of the extracellular matrix components involved in this attachment. Its presence in the extracellular matrix of bovine aorta [46] and mouse smooth muscle cells [47] has been reported. At present, the function of secreted actin in the uterine lumen remains to be elucidated.
Similarly, the biological role, if any, for the proteolytic fragments of serum albumin (S3-S6) in ULF is unknown. Whether these fragments had arisen as a result of in situ cleavage by endometrial proteases or during sample collection and treatment was not determined. We did not observe evidence of extensive proteolysis in samples that had been treated with EDTA and PMSF, as the majority of the serum proteins in ULF had similar electrophoretic mobilities in the 2D gels compared to proteins from serum.
The only pregnancy-associated, endometrial secretory proteins detected in this study were the 14-kDa proteins, E1. Continued and increased synthesis of the E1 proteins after the midluteal phase of the estrous cycle appeared to be dependent on the establishment of pregnancy. We believe that the E1 proteins are most likely the progesterone-induced, endometrial proteins reported by Kazemi et al. [48], which were shown by immunochemistry and electron microscopy to be localized to the crystalline inclusion bodies of trophectodermal cells and to the cytoplasm of the uterine luminal epithelia. Although Kazemi et al. [48] were unable to determine the site of synthesis of these proteins, we have shown, by metabolic labeling, that the pregnant endometrium at Day 17 synthesized E1. Currently, we are sequencing E1 so as to elucidate its function in early pregnancy.
Although the allantois is regarded as the storage site of fetal nitrogenous wastes after the mesonephros becomes active at Day 18, we observed that most of the conceptus-secreted proteins were rapidly transported and sequestered in allantoic fluid. It is possible that the serum proteins synthesized by the yolk sac are more highly represented, as they are most likely transported from the yolk sac via its circulatory network to the embryo, where they are then secreted either through the mesonephros or from the vessels on the allantoic membrane. Interestingly, proteins secreted by the trophectoderm, such as C1, C2, and IFN, as well as by the endometrium, such as E1, were also present in allantoic fluid, even though at this stage the allantois and trophoblast are not yet fused and there is no circulatory network connecting the two membranes. This suggests that the yolk sac is capable of absorbing proteins and probably other nutrients from the uterine lumen and, subsequently, transporting it to the embryo and the allantois, again illustrating the important role the yolk sac plays in the nutrition of the preimplantation conceptus.
Analyses of ULF proteins by 2D gel electrophoresis have revealed the dynamic changes in protein composition during this period of rapid growth and differentiation and have afforded some understanding of the molecular events occurring at this stage of pregnancy. This study has illustrated some of the interactions between the different extraembryonic membranes of the conceptus, such as the transport of proteins synthesized by the trophoblast and the yolk sac to the fluid of the rapidly growing allantois, as well as the interaction between the endometrium and the conceptus. The discovery of aldose reductase as a major trophoblastic protein raises interesting questions about the carbohydrate metabolism of the periimplantation ovine conceptus. One area of future research will be the developmental regulation and pattern of expression of aldose reductase and its role in sorbitol and fructose synthesis in this early stage of pregnancy.
|
NOTE ADDED IN PROOF |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONNOTE ADDED IN PROOFREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Rita S.F. Lee, AgResearch, Ruakura Research Center, East St., Private Bag 3123, Hamilton, New Zealand. FAX: 64 7 838 5628; leer{at}agresearch.cri.nz
Accepted: May 7, 1998.
Received: November 10, 1997.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONNOTE ADDED IN PROOFREFERENCES |
|---|
-crystallin of the European common frog. Proc Natl Acad Sci USA 1988; 85:11–15.This article has been cited by other articles:
|
|
 |
|
  T. E Spencer, O. Sandra, and E. Wolf Genes involved in conceptus-endometrial interactions in ruminants: insights from reductionism and thoughts on holistic approaches Reproduction, February 1, 2008; 135(2): 165 - 179. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Parmar, G. Sachdeva, L. Savardekar, R.R. Katkam, S. Nimbkar-Joshi, S. Gadkar-Sable, V. Salvi, D.D. Manjramkar, P. Meherji, and C.P. Puri Protein repertoire of human uterine fluid duringthe mid-secretory phase of the menstrual cycle Hum. Reprod., February 1, 2008; 23(2): 379 - 386. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. W. Ross, M. D. Ashworth, F. J. White, G. A. Johnson, P. J. Ayoubi, U. DeSilva, K. M. Whitworth, R. S. Prather, and R. D. Geisert Premature Estrogen Exposure Alters Endometrial Gene Expression to Disrupt Pregnancy in the Pig Endocrinology, October 1, 2007; 148(10): 4761 - 4773. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. C. Satterfield, K. A. Dunlap, K. Hayashi, R. C. Burghardt, T. E. Spencer, and F. W. Bazer Tight and Adherens Junctions in the Ovine Uterus: Differential Regulation by Pregnancy and Progesterone Endocrinology, August 1, 2007; 148(8): 3922 - 3931. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.-P. R Kayser, J. G Kim, R. L Cerny, and J. L Vallet Global characterization of porcine intrauterine proteins during early pregnancy Reproduction, February 1, 2006; 131(2): 379 - 388. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. A. Gray, C. A. Abbey, P. D. Beremand, Y. Choi, J. L. Farmer, D. L. Adelson, T. L. Thomas, F. W. Bazer, and T. E. Spencer Identification of Endometrial Genes Regulated by Early Pregnancy, Progesterone, and Interferon Tau in the Ovine Uterus Biol Reprod, February 1, 2006; 74(2): 383 - 394. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. S. Georgiou, E. Sostaric, C. H. Wong, A. P. L. Snijders, P. C. Wright, H. D. Moore, and A. Fazeli Gametes Alter the Oviductal Secretory Proteome Mol. Cell. Proteomics, November 1, 2005; 4(11): 1785 - 1796. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. K. Dey, H. Lim, S. K. Das, J. Reese, B. C. Paria, T. Daikoku, and H. Wang Molecular Cues to Implantation Endocr. Rev., June 1, 2004; 25(3): 341 - 373. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Girotti and H. H. Zingg Gene Expression Profiling of Rat Uterus at Different Stages of Parturition Endocrinology, June 1, 2003; 144(6): 2254 - 2265. [Abstract] [Full Text] [PDF]
|
|
