| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Fundación Instituto Valenciano de Infertilidad para el Estudio de la Reproducción Humana (FIVIER), 46020 Valencia, Spain
b Department of Obstetrics and Gynecology and
c Department of Biochemistry and Molecular Biology, School of Medicine, University of Valencia, 46010 Valencia, Spain
d Schools of Medicine and Biological Sciences, University of Manchester, St. Mary's Hospital, Manchester M13 0JH, United Kingdom
ABSTRACT
Expression of MUC1 in endometrial epithelium has been suggested to create a barrier to embryo attachment that must be lifted at the time of implantation. In this study, we investigated the hormonal regulation of human endometrial MUC1 in hormone replacement therapy cycles and in the human blastocyst. We also analyzed the embryonic regulation of MUC1 in human endometrial epithelial cells (EECs) during the apposition and adhesion phases of human implantation using two different in vitro models. Our results indicate that endometrial MUC1 mRNA and immunoreactive protein increase in receptive endometrium compared to nonreceptive endometrium. Human blastocysts express MUC1, as demonstrated by reverse transcription-polymerase chain reaction and immunocytochemistry, localized at the trophectoderm. In vitro, MUC1 was present at the surface of primary cultures of human EEC, and presence of a human blastocyst (i.e., apposition phase) increases EEC MUC1 protein and mRNA compared to control EEC lacking embryos. Interestingly, when human blastocysts were allowed to attach to the EEC monolayer (i.e., adhesion phase), MUC1 was locally removed in a paracrine fashion on EEC at the implantation site. These results demonstrate a coordinated hormonal and embryonic regulation of EEC MUC1. Progesterone combined with estradiol priming induces an up-regulation of MUC1 at the receptive endometrium. During the apposition phase, presence of a human embryo increases EEC MUC1. However, at the adhesion phase, the embryo induces a paracrine cleavage of EEC MUC1 at the implantation site. These findings strongly suggest that MUC1 may act as an endometrial antiadhesive molecule that must be locally removed by the human blastocyst during the adhesion phase.
implantation/early development, menstrual cycle, progesterone, uterus
INTRODUCTION
The receptive status of the maternal host in embryonic implantation is a balance between the activation of adhesion molecules and the presence of a natural barrier that the implanting embryo may encounter in the epithelial glycocalyx. This process is controlled by maternal steroid hormones [1], but it may also be influenced by local embryonic paracrine signals [2].
Mucins are a family of highly glycosylated, high-molecular-weight (>250 kDa) glycoproteins present on epithelial surfaces, including human endometrial epithelial cells (EECs) [3, 4]. The MUC1 is an integral membrane glycoprotein with a large ectodomain containing a variable number (
25–120) of 20-amino acid tandem repeats, resulting in a large and highly extended structure that is both immunogenic and extensively glycosylated [4–6]. This mucin has an hydrophobic transmembrane domain (31 residues) and a short, nonglycosylated cytoplasmic region (56 residues); this cytoplasmic region can be phosphorylated [7] and seems to be associated with the cytoskeleton [8]. Additionally, MUC1 is synthesized in the rough endoplasmic reticulum and passes through the smooth endoplasmic reticulum and the Golgi apparatus, where it is glycosylated. Thereafter, it is transferred to the cell surface. A fraction of uterine epithelial cell mucins (30%–50%) appears to be released from the apical surface into the lumen; the remainder is presumably degraded intracellularly following endocytosis [8]. Two other isoforms are known: a secreted form of MUC1 that lacks the transmembrane and cytoplasmic domains and that may originate either from a proteolytic cleavage site in the proximal extracellular domain [9] or by alternative splicing (MUC1/SEC) [8, 10]; and MUC/Y, which also originates by alternative splicing and lacks the tandem repeat domain [7]. The MUC1/SEC binds and specifically interacts with the extracellular domain of MUC1/Y, and MUC1/SEC in the conditioned medium forms an active receptor complex with MUC1/Y, resulting in the phosphorylation of MUC1/Y and concomitant change in cellular morphology [8].
Expressed at high levels at the cell surface, MUC1 has the ability to inhibit cell-cell adhesion. This property depends on the presence of a long ectodomain and probably results from steric hindrance of receptor-ligand interactions mediated by families of adhesion receptors such as the cadherins [11, 12]. More recently, however, the tandem repeat core peptide has been shown to be capable of binding ICAM-1, and this interaction can mediate cell-cell adhesion [13]. Furthermore, MUC1 can carry sialyl Lewisx in endometrial cells [14], and evidence exists for selectin expression by preimplantation embryos [15], suggesting that a selectin-MUC1 interaction could be mediating cell-cell adhesion. Intrinsic heterogeneity in the glycosylation in this molecule may allow a local mechanism to define a receptive site within the receptive endometrium, because it may be possible that MUC1-associated glycans or its protein backbone could be recognized by the embryo [16].
Therefore, careful evaluation of both the distribution and the regulation of MUC1 at the endometrial cell surface is necessary. It is interesting that these phenomena vary among species. In mice, rats, pigs, rabbits, and, to a lesser extent, baboons, MUC1 is down-regulated in the receptive endometrium before implantation, when progesterone (P) levels are high. For each of these species, it has been argued that MUC1 inhibits implantation, and that its down-regulation contributes to the acquisition of maternal receptivity [17–20]. Indeed, in the mouse, experimental evidence suggests that MUC1 prevents embryo adhesion to endometrial epithelial cells [21].
In rabbits, MUC1 is increased during the preimplantation and implantation period by a process that is indirectly mediated by P. However, a striking reduction of MUC1 is induced locally by the rabbit blastocyst in the apical luminal epithelium, specifically in the implantation site [22]. Again, the data support the notion of an inhibitory activity that is lifted to allow implantation to occur.
In humans, endometrial MUC1 is up-regulated during the peri-implantation period in natural cycles [23, 24]. The abundance of MUC1 mRNA is increased from the proliferative to mid-secretory phase and is followed by a decrease during the late secretory phase [25]. Therefore, the role of MUC1 in humans is contradictory, and the possibility that both maternal and embryonic signals may contribute to the regulation of MUC1 in human endometrium must be addressed. Furthermore, to our knowledge, expression of MUC1 in human blastocysts has not been yet examined.
In this study, we investigated the hormonal regulation in vivo of MUC1 during the receptive phase in agonadal women undergoing mock cycles of exogenous hormone replacement therapy (HRT) to prime the uterus in an ovum donation program. We also analyzed MUC1 mRNA expression and protein synthesis in preimplantation blastocysts. In parallel, we used two different in vitro models to analyze the distribution and embryonic regulation of MUC1 in EECs during the apposition and adhesion phases of human implantation [26–28].
MATERIALS AND METHODS
Institutional Approval and Informed Consent
This project was approved by the institutional review board on the use of human subjects in research at the Instituto Valenciano de Infertilidad and complied with the Spanish Law of Assisted Reproductive Technologies (35/1988). Endometrial samples and surplus embryos donated for research were obtained after written consent from patients. The clinical and laboratory work was performed at the Instituto Valenciano de Infertilidad.
HRT Protocol
Serum and endometrial samples for immunohistochemistry and Northern blot analysis were obtained in mock cycles from 10 patients (23–29 yr old) undergoing oocyte donation and receiving HRT as previously described [29]. Briefly, serum samples (S) and uterine biopsies (B) were taken at Days 13 (S1, B1), 18 (S2, B2), and 21 (S3, B3) from each patient. Therefore, at the time that serum samples and biopsy specimens were collected, patients were treated with 6 mg/day of estradiol (E2)-valerate (EV) alone and 6 mg/day of EV plus 800 mg/day of P for 3 and 6 days, respectively (Fig. 1). Biopsies were dated histologically according to the method described by Noyes et al. [30]. The E2 and P were measured in serum using commercially available kits (estradiol reagent pack, Abbot Cientifica, Madrid, Spain, and Vidas progesterone, BioMevieux, France).
|
Endometrial Cultures
The ovarian stimulation protocol using Gn-RH-
and gonadotropins has been described previously [31]. Samples of endometrium were obtained during the luteal phase from fertile patients (23–39 yr old) undergoing endometrial biopsy. A portion of each specimen was stained with hematoxylin-eosin for dating according to the method described by Noyes et al. [30].
Endometrial samples were minced into small pieces (<1 mm) and then subjected to mild collagenase digestion. The EECs were grown from isolated endometrial glands that were purified as previously described [28]. This cell type was cultured and grown to confluence in steroid-depleted medium: 75% Dulbecco's modified Eagle's medium (Sigma, St. Louis, MO) and 25% MCDB-105 (Sigma) containing antibiotics (Fungizone 2 µl/ml, gentamicine 1 µl/ml) and supplemented with 10% charcoal-Dextran-treated fetal bovine serum (Hyclone, Logan, UT) and 0.5 µg/ml of insulin (Sigma) as described previously [32, 33].
The homogeneity of cultures was determined by morphological characteristics and verified by immunocytochemical localization of cytokeratin, vimentin, and CD68 antigen as described previously [28]. The functionality of EEC monolayers was demonstrated [33] and the morphology analyzed by scanning electron microscopy [28]. After the cells reached confluence, growth media were replaced by a 1:1 v:v mixture of IVF:S2 media (Scandinavian IVF Science AB, Gotheburg, Sweden).
Coculture of Human Embryos with Autologous EECs to Study the Apposition Phase
After confluence, EECs were cocultured with single human, two- to four-cell embryos. At this time, embryos were grown in 1 ml of IVF:S2 media (1:1 v:v) until they reached the eight-cell stage, then were further cultured with S2 medium until the blastocyst stage. Embryonic development was checked daily, and conditioned media changed every 24 h. On Day 6, blastocysts were transferred to the maternal uterus using a Frydman catheter. In each experiment, cultured EECs in the same conditions but without human embryos were used as controls.
Human In Vitro Attachment Assay to Study the Adhesion Phase
Three-dimensional cultures were prepared for the embryo attachment assay. Epithelial cells were obtained from endometrial biopsies and processed as described above. Epithelial cells grew polarized on 24-well plates coated with extracellular matrix (ECM-gel; Sigma). Spare blastocysts were cultured on these EECs and allowed to attach to the epithelial surface. After 48–72 h, cells were washed with PBS, fixed with 4% v:v paraformaldehyde (PFA) in PBS for 30 min at 4°C or fixed and permeabilized with 70% v:v ethanol for 2 h at -20°C (to study an intracellular epitope), and washed again with PBS. Attached blastocysts were localized by phase-contrast microscopy, and different techniques were performed.
Antibodies
Monoclonal mouse anti-human MUC1 antibody human milk fat globulin (HMFG)-1 (immunoglobulin [Ig] G) [34] and polyclonal rabbit anti-human MUC1 antibody cytoplasmic tail (CT)-1 [35] were kindly provided by Dr. Joy Burchell (Imperial Cancer Research Foundation [ICRF], London, UK). Monoclonal mouse anti-human MUC1 breast cancer (BC)-2 (IgG) antibody was obtained from Serotec (Oxford, UK) and BC-3 (IgM) from Medical Innovations Ltd. (Queensland, Australia). Anti-cytokeratin was purchased from Dako (Barcelona, Spain).
The HMFG-1, BC-2, and BC-3 recognize a similar peptide sequence in the tandem repeat extracellular region of MUC1 [36]. Although these monoclonal antibodies (mAbs) are directed against the protein core, attached glycans can decrease epitope accessibility [37, 38]. Binding of BC-2 and BC-3 is inhibited, to a lesser extent than HMFG-1, by glycosylation [39]. Binding of HMFG-1 is significantly enhanced after treatment with sialidase. The BC-2 and BC-3 recognize the same peptide epitope, but they are affected in a different manner by glycosylation. Characterization of mAbs by the binding inhibition pattern reveals that binding of BC-2 is inhibited by glycosylation at Thr/Ser residues in the epitope, suggesting that increasing glycosylation would diminish the sensitivity of BC-2 to the molecule [40].
Immunohistochemistry
Immunohistochemistry was performed on endometrial sections as previously described [41]. Briefly, sections were incubated with mAb BC-2 (1:54) or HMFG-1 (1:10) for 90 min at 37°C. After washing, sections were incubated with rabbit anti-mouse IgG (90 min, 1:300, 37°C; Dako). To amplify the signal, sections were rinsed four times with PBS and 0.05% Tween-20 (PBS-T), then incubated with extravidin-horseradish peroxidase (30 min, 1:40, room temperature; Sigma), washed and incubated for 10 min with working substrate solution (0.2 ml of stock aminoethyl carbazol solution with 3.8 ml of 0.05 M acetate buffer [pH 5.0] and 20 µl of 3% H2O2). Binding was terminated by rinsing the slides gently with distilled water. Finally, slides were counterstained with Mayers hematoxylin, rinsed with distilled water, and mounted in glycerol-gelatin. For mAb BC-3 (1:100), immunohistochemistry was performed as previously described [24]. Methyl nuclear counterstaining was performed on all sections.
Staining was scored independently by three double-blind observers according to a semiquantitative scale.
Immunocytochemistry
Immunostaining of human embryos used an extravidin-peroxidase staining method with BC-3, HMFG-1 (1:10), and CT-1 as primary antibodies. Embryos were previously fixed with freshly prepared, 2% PFA in PBS for 30 min at 4°C in microdrops under oil [26]. After fixation, blastocysts were treated with 0.2% Triton X-100 in PBS for 10 min at 4°C to permeabilize the fixed cells, facilitating access of the antibody to the intracellular domain. Incubation with working substrate solution was for 4 min.
Confocal Microscopy
At the implantation sites, the extracellular domain of MUC1 was labeled by immunofluorescence as described previously [24] using BC-3 as the primary antibody and visualized using a confocal microscope. Cytokeratin is a cytoskeleton protein expressed in EECs, and immunofluorescence was performed as described using anti-cytokeratin as the primary antibody. Confocal analysis was performed with a NRC 1024 instrument (Bio-Rad, Hempstead, UK). The excitation line used was 488 nm (fluorescein isothiocyanate [FITC]). The filter used was HQ515/10 (FITC). Transmitted light images were acquired for every field.
Flow Cytometry
The EEC monolayers were detached by a cell scraper, retrieved with the coculture medium, and centrifuged 10 min at 300 x g, and the cell pellet was blocked with 1% BSA in PBS for 120 min at 4°C. To study intracytoplasmic epitopes, the cell pellet was resuspended in 70% ethanol at -20°C for 20 min and recentrifuged. After washing with PBS-T, cells were incubated overnight at 4°C with HMFG-1, BC-2, and CT-1. The EEC suspensions were washed and mixed: HMFG-1- and BC-2-incubated cells with FITC-conjugated goat anti-mouse IgG whole molecule (Sigma), and CT-1-incubated cells with FITC-conjugated goat anti-rabbit IgG whole molecule (Sigma, Madrid, Spain). Cell suspensions that were incubated with BC-2 or HMFG-1 were fixed with 1% PFA for 1 h at 4°C. Negative controls were included in each experiment by omission of the primary antibody. Where indicated, cells were subjected to enzymatic digestion with sialidase (from Vibrio cholerae, 1:100 in 0.1 M sodium acetate and 1 mM CaCl2; Boehringer Mannheim, Bercelona, Spain) overnight at 37°C and washed for 15 min in PBS before applying HMFG-1.
Cells were resuspended in PBS and analyzed in a Epics Elite Flow Cytometer (Coulter Cytçometry, Hialeah, FL) using an argon-ion laser tuned at 488 nm and 15 mW. The FITC fluorescence was collected by 575 DC with 525BP filters (data were collected in a four-decade amplification). Debris was excluded by analysis of scatter properties. At least 10 000 events per sample were stored in list-mode files. Data were expressed as luminosity intensity fluorescence arbitrary units (FAUs). Variability in the intensity of the signal was obtained with full peak coefficient of variation (FPCV).
Northern Blot Analysis
The MUC1 cDNA probe pMUC7 was provided by Dr. J. Burchell (ICRF) [24]. It corresponds to seven conserved tandem repeats of 60 base pairs (bp) in the VNTR region of exon 2 in the MUC1 mRNA. It was subcloned in the EcoRI site of pBluescriptSK (2.96 kilobases [kb]). After transforming DH5 Escherichia coli, cells were grown and plasmid isolated with Qiagen plasmid mega kit (Izasa, Madrid, Spain), digested with EcoRI, and purified by electrophoresis, after which a band of 450 bp was isolated with Pharmacia elution band kit (Amersham, Barcelona, Spain). The cDNA was 32P-radiolabeled (random primer DNA labeling kit; Boehringer Mannheim) as previously described [42]. A cDNA probe for 28S rRNA was provided by Dr. Iris L. Gonzalez (Hahnemann University, Philadelphia, PA). Plasmid was amplified and isolated, and the 1.6-kb insert in pGEMz (2.7 kb) was removed using EcoRI/BamHI double-digestion isolated and labeled in the same way.
Endometrial tissue for RNA isolation was collected, weighed, frozen in liquid nitrogen, and stored at -70°C. Adenoids were also collected and processed as controls. Total RNA was extracted from the tissue homogenate and EECs using the guanidinium isothiocyanate method [43] according to the manufacturer's instructions (RNAzol method; Tel-test, Friendswood, TX), then resuspended in diethylpyrocarbonate-treated water. The RNA concentrations were determined by optical density with a Shimadzu 1201 spectrophotometer (Izasa). The absorbance 260:280 ratio for each sample was between 1.6 and 1.8.
The RNA (10 µg from cultured EECs and 15 µg from endometrial biopsy specimens) stained with ethidium bromide was fractioned on a 1% agarose formaldehyde denaturing gel. The integrity of the RNA was assessed, and samples with evidence of ribosomal RNA degradation were excluded from subsequent analysis. Electrophoresis was performed for 2 h at 76 V in 3-N-morpholino propanesulfonic acid (MOPS) buffer. Transfer to positively charged nylon membrane (Boehringer Mannheim) was by capillary action in 1x SSC (0.15 M sodium chloride and 0.015 M sodium citrate), and RNA was cross-linked to the membranes by ultraviolet radiation. For the MUC1 study, membranes were prehybridized at 42°C overnight in 50% formamide, 5x SSC, 5x Denhardt's solution, 0.2% SDS, and 50 mmol/L of sodium phosphate buffer (pH 7.4) containing 0.05 g/L of dextran sulfate. For 28S, prehybridization was performed as described for MUC1 but at 68°C overnight in 6% SSC, 5x Denhardt's solution, and 0.5% SDS. Hybridization was performed overnight in the same solution containing at least 5 x 106 cpm/ml of cDNA probes at 42°C [23]. Membranes were washed twice in 2x SSC and 0.1% SDS for 25 min at room temperature, with a final, high-stringency wash in 0.1x SSC and 0.1% SDS at 68°C for 20 min. Filters were exposed to Kodak XAR-5 film (Kodak, Marca, Spain) at -70°C for 2–6 days for the MUC1 cDNA hybridization and for 30 min to 2 h for the 28S study. After hybridization and exposure to film, blots were stripped for 10 min at 95°C in 0.1% SDS and rehybridized to another probe [25, 33]. The relative intensities of Northern hybridization signals were determined by analysis of autoradiographs after image capture (Tecnología para Diagnótico e Investigación, Madrid, Spain). The MUC1 mRNA levels were normalized to 28S rRNA hybridization signals.
Reverse Transcription-Nested-Polymerase Chain Reaction of MUC1 on Blastocysts
For each embryo, 17.5 µl of RT-Mastermix (Advantage RT-for-PCR kit; Clontech, Palo Alto, CA) were prepared according to the manufacturer's instructions and kept on ice until RNA extraction. A single embryo was added to the RT-Mastermix using a Pasteur pipette, allowing a carry-over of approximately 1 µl of the culture medium. Samples were immediately heated to 99°C for 1 min in a UnoII Thermocycler (Biometra, Gottingen, Germany) to release the total RNA and to denature the proteins. Samples were cooled to 4°C, and 0.5 µl of RNase-inhibitor was added, followed by 1.0 µl of Maloney murine leukemia virus reverse transcriptase. Once all components were mixed, the reaction was incubated at 42°C for 1 h, then heated at 94°C for 5 min to stop cDNA synthesis and destroy DNase activity as described previously [44]. Total RNA from a receptive endometrium and cultured EEC were used as positive controls. First-strand cDNA was reverse transcribed from 1 µg of RNA as described previously [44]. The cDNA solution was diluted with sterile water to 100 µl and stored at -20°C. The product was stored at -20°C until the polymerase chain reaction (PCR).
Sequences for human MUC1 and ß-actin mRNA were obtained from GenBank. Primers (Table 1) were from Gibco (Barcelona, Spain) and designed to cross intron/exon boundaries to distinguish products amplified from cDNAs and those from contaminating genomic DNA (Fig. 2).
|
|
Two microliters of reverse transcription (RT) products were mixed with PCR mastermix containing 5 µl of 10x buffer, 2.5 µl of 50 mM MgCl2, 0.5 µl of Taq Polymerase (all from Bioline, London, UK), 1 µl of deoxy nucleic triphosphate mix (Sigma), 1 µl of 5' and 3' outer primers (10 pmol; Gibco) (Table 1) for MUC1, and sterile distilled water to a final volume of 50 µl. The negative control included in each reaction consisted of H2O.
First-round PCR reactions were carried out in a UnoII Thermocycler (Biometra) with the following program using a heated lid (100°C): 94°C for 5 min; then 30 cycles of 94°C for 45 sec, 58°C for 45 sec, and 72°C for 45 sec, with a final extension of 72°C for 5 min; and then cooled to 4°C.
Five microliters of first-round PCR products from blastocysts, secretory biopsy specimens (1:10), and EECs (1:10) were added to PCR mastermix with 1 µl from each of the inner primers (10 pmol; Gibco) (Table 1). Second-round PCR reactions were performed as described above. The integrity of each cDNA preparation was confirmed using ß-actin primers (Table 1). Two rounds of ß-actin PCRs were performed using specific primers to ensure cDNA synthesis. The number of cycles was 20 for the first round and 24 for the second. Products were then electrophoresed in a 2% agarose gel containing 0.5 µg/ml of ethidium bromide, and bands were analyzed in an image analysis system. The 100-bp DNA Ladder (Gibco BRL Life Technologies, Barcelona, Spain) was used as molecular weight marker.
Restriction sites in the genomic and cDNA products were determined using Seqaid II software (Molecular Genetics Laboratory, Kansas State University, Manhattan, KS). The Kpn I (Roche Molecular Biochemicals, Madrid, Spain) was used in 1x SuRE/Cut Buffer L (Roche Molecular Biochemicals) and 100 µg/ml of BSA (37°C, 2 h).
Statistical Analysis
The FAUs from flow cytometry and densitometric units from Northern blot analysis are expressed as mean ± SEM. Statistical analysis was performed using ANOVA, and for multiple post-hoc comparison, the dimethyl sulfoxide and Bonferroni's test were performed. The level of P < 0.05 was considered to be significant. Statistical analysis was performed using the Statistical Package for the Social Sciences (SPSS, Chicago, IL).
RESULTS
Hormonal Regulation of Endometrial MUC1 mRNA Expression
The objective of this part of the study was to use HRT-stimulated cycles to analyze the hormonal regulation of MUC1 during the acquisition of endometrial receptivity in vivo. For this purpose, we defined Day 13 as nonreceptive, Day 18 as prereceptive, and Day 21 as receptive (Fig. 1). Serum E2 and P levels were consistent with the receptive status (data not shown).
Northern blot analysis of MUC1 mRNA is shown in Figure 3A. Due to the presence of a VNTR sequence, MUC1 mRNA varies in size. However, under the conditions used in this study, the different bands comigrated in each of the five patients studied. Densitometric analysis of the major hybridization signals (Fig. 3B) showed a twofold increase of MUC1 mRNA levels in receptive compared to nonreceptive endometrium (P < 0.05).
|
Hormonal Regulation of Immunoreactive Endometrial MUC1 Protein
In epithelial cells, MUC1 immunoreactivity could be detected during all three phases of the cycle. However, after estrogen alone (Day 13, or nonreceptive phase), reactivity was weak. The staining was homogeneously distributed at the apical surface of both glandular and luminal epithelium, as observed with BC-2 (Fig. 4a) and BC-3 antibodies (Fig. 5a). The BC-3 antibody was most effective in demonstrating luminal epithelial reactivity (Fig. 5a). At Day 18, in the prereceptive endometrium, staining was perceptibly increased in glands with mAb BC-2 (Fig. 4b), and this was even more evident when BC-3 (Fig. 5b) was used. In luminal epithelium, apical reactivity was evident only with BC-3. In the receptive endometrium, a further increase in MUC1 reactivity in the glandular epithelium was observed at Day 21. In luminal epithelium, up-regulation was observed with BC-3 (Fig. 5c), whereas BC-2 was unable to detect the protein (Fig. 4c). No staining was observed in negative controls lacking the primary antibody (Fig. 4d). Semiquantitative assessment of the staining confirmed that MUC1 immunoreactive protein is more abundant in receptive endometrium compared to prereceptive or nonreceptive endometrium (data not shown).
|
|
RT-Nested-PCR in Human Blastocysts
To investigate the expression of MUC1 mRNA in human blastocysts, RT-nested-PCR was performed. A product of between 400 and 300 bp corresponds to an amplification of MUC1 cDNA (Fig. 2) and was observed in blastocysts (n = 3) as well as in positive controls (secretory endometrium and EEC) (Fig. 6). Another clearly defined band was observed in blastocysts near 700 bp due to the presence of genomic DNA (Fig. 6), as explained in Figure 2. No significant bands were observed in the negative control lanes. Integrity of cDNA was confirmed in all samples with nested-PCR for ß-actin (data not shown). Restriction analysis was performed using Kpn I, which was predicted to cut at a single site in both products. The genomic product of 707 bp yielded two fragments in the predicted size range of 383 and 324 bp. The cDNA product also yielded two bands in the predicted range of 180 and 135 bp. These data (not shown) confirm the authenticity of the products.
|
Immunolocalization of MUC1 in Blastocysts
We investigated the presence and localization of MUC1 mucin in human embryos by immunocytochemistry. These experiments were performed using eight blastocysts (two per antibody and two for negative controls). Immunostaining with BC-3, HMFG-1, and CT-1 confirmed the presence of MUC1 protein in human blastocysts. The signal with BC-3 was higher than that with CT-1 or HMFG-1. The MUC1 was present in all blastocyst cells, although the trophectoderm showed the most intense staining (Fig. 7, B–D). Figure 7A shows a negative control.
|
Embryonic Regulation of Endometrial Epithelial MUC1 mRNA Expression During the Apposition Phase
To investigate the possibility that maternal MUC1 expression is subject to embryonic regulation during the apposition phase (without direct embryonic-EEC contact), EEC cultures were maintained with or without developing human embryos from the cleavage to the blastocyst stage. Hybridization detected two transcripts in some cases, whereas in others, the two alleles comigrated. Bands were observed in the range of 4–7 kb (Fig. 8A). Quantitative densitometric analysis (Fig. 8B) showed that MUC1 mRNA levels increased more than twofold in EECs cultured with embryos compared to controls (P < 0.05).
|
To quantify the embryonic regulation of immunoreactive endometrial MUC1, flow cytometric analysis was performed on EECs cultured with embryos that developed normally to the blastocyst stage, with embryos that showed arrested development, or without embryos. Because glycosylation may affect antibody binding to the ectodomain, flow cytometry was performed with two different mAbs to the VNTR (BC-2 and HMFG-1); sialidase unmasking treatment was also used. In addition, the cytoplasmic C-tail antibody CT-1 was used. Figure 9A shows a representative cytofluorometric analysis of MUC1 in EECs with CT-1. Data for the experiments with each antibody were combined and expressed as the intensity of stained cells in FAUs (Fig. 9B). The CT-1 fluorescence levels were less than those recorded after staining with antibodies against the TR domain because of epitope repetition in the latter cases. Heterogeneity (measured as FPCV) was higher when extracellular epitopes were studied compared to the intracellular epitope detected with CT-1 (data not shown), probably because glycosylation leads to more variation in binding in the former case. Digestion with sialidase increased the binding of HMFG-1 antibody to the core protein, as expected. The results indicate that human preimplantation embryos up-regulate endometrial epithelial MUC1 compared to controls without embryos. This effect was observed with three different antibodies and was statistically significant with BC-2 and CT-1 (P < 0.05). Arrested embryos showed an endometrial effect that was intermediate between those of controls and embryos that developed to blastocysts.
|
Embryonic Regulation of Endometrial Epithelial MUC1 During the Adhesion Phase
The implantation sites (n = 3) (Fig. 10, B and D) of human blastocysts on EECs were fixed after initial attachment (i.e. before significant trophoblast migration occurred from the embryo). The extracellular domain of MUC1 was immunolocalized with BC-3 (Fig. 10, C and E), and cytokeratin staining was performed to check the presence of epithelial cells and antibody accessibility beneath the attached blastocysts (Fig. 10A).
|
Despite MUC1 expression in preadhesion blastocysts, we were unable to detect MUC1 in attached embryos (data not shown). Similarly, MUC1 was absent from EECs subjacent and adjacent to the attached embryo (Fig. 10D). Expression in EECs gradually increased with distance from the implantation site (Fig. 10, E and F). Strong and uniform staining was observed away from the blastocyst (Fig. 10E). No staining was observed in negative controls (data not shown).
DISCUSSION
This study demonstrates that endometrial MUC-1 is hormonally regulated during a HRT cycle that induces the acquisition of endometrial receptivity [29]. This study also provides novel evidence for the role of MUC1 in the human implantation process during the apposition and adhesion phases.
The results indicate that P, when applied after a previous estrogen stimulus, up-regulates MUC1 mRNA expression, an effect that is evident in the prereceptive endometrium (P + 3 days) and becomes statistically significant in the receptive endometrium (P + 6 days). The translated product appears specifically at the apical surface of glandular and luminal epithelial cells, where it is detectable as a continuous array after 6 days of P treatment (i.e., when an embryo would be expected to implant). Variations in glycosylation may account for the observed differences in luminal distribution as detected by different antibodies. The data are consistent with previous observations in natural cycles [23, 25] and similar to what has been reported in the rabbit [22].
Heterogeneity of MUC1 expression in cultured epithelial cells was detected with all antibodies used, although the binding of CT-1 was the value with the least variation as shown by the analysis of flow cytometric profiles. This suggests that intercellular variations in the binding of antibodies that recognize VNTR or other ectodomain epitopes arise mainly because of variations in glycosylation and only to a relatively minor extent because of variations in core protein abundance.
In experiments involving coculture of developing preimplantation embryos with EEC monolayers, the human blastocysts induce an up-regulation of the core protein and its corresponding mRNA compared to control EECs lacking embryos. This effect was reduced, but still visible, in the presence of embryos suffering arrested development, providing evidence that at least some of the paracrine influence arises during preblastocyst stages. The effect is quantitative, because the distribution of MUC1 in cells is not perceptibly affected by the presence of the embryo (data not shown). In any case, changes are always at the level of the plasma membrane. To our knowledge, these are the first data reported on MUC1 regulation during the apposition or preadhesion phase in humans.
These data must be interpreted within the limitations of an in vitro model. However, they strongly suggest that paracrine signals may be transmitted from the embryo to the endometrial epithelium, even during the prehatching stages of development. Furthermore, the resulting increase in MUC1 results from an effect on mRNA abundance. Future studies will be directed at identification of the signaling pathways that are responsible for this effect. Finally, the evidence of increased MUC1 abundance argues against the notion that, during the preimplantation stages, embryo signaling tends to diminish the maternal glycocalyx.
In addition to being present at the maternal cell surface, MUC1, as we have shown, is present in human embryos during the blastocyst stage. The light microscopic localization data suggest that MUC1 is present at the outer trophectodermal surface beneath the zona pellucida as well as in the inner cell mass. However, in embryos that had hatched and attached to epithelial monolayers, we were unable to detect MUC1 immunostaining. Using the in vitro model for embryonic adhesion, we showed that cultured epithelial cells both beneath and immediately adjacent to the attached embryo lacked detectable MUC1 reactivity, with a progressive increase in signaling from cells further away in the epithelial layer. Cytokeratin labeling further demonstrated the presence of epithelial cells at the implantation site that did not show MUC-1 expression. The presence of a hole in which cells have been removed is consistent with previous demonstrations of the induction of EEC apoptosis by the human embryo as a mechanism to break the epithelial barrier [45] This information suggests that the antibodies used are able to reach the EECs at the implantation site (i.e., underneath the embryo). Careful examination of Figure 10 reveals that MUC1 staining is absent in EECs at and surrounding the implantation site, where cytokeratin staining is present, suggesting that the embryo is producing a local removal of MUC1. Nevertheless, future experiments should be developed to confirm this process directly.
The MUC1 in the apical luminal glycocalyx is predicted to extend as much as 500 nm above the plasma membrane [9] and, given its abundance in the receptive phase, is expected to be encountered by an embryo that approaches the maternal cell surface. It has been proposed that MUC1 acts as an antiadhesion molecule and may inhibit the interaction between embryo and maternal apical epithelium during implantation, creating a uterine barrier to implantation [16, 46]. Because EECs tend to show increased MUC1 reactivity during the apposition phase, the healthy embryo may act to reinforce the maternal barrier to premature attachment. Similarly, expression of MUC1 in the blastocyst may prevent attachment at an anatomically inappropriate site, such as the tubal epithelium. Another interesting and complementary hypothesis is that its presence could provide protection against enzymatic and microbial attack from the uterus to the embryo or have suppressive activity regarding the maternal immune response to the blastocyst [47, 48]. During the adhesion phase, our studies reinforce this hypothesis by addressing the question of how this barrier may eventually be overcome by the embryo. Human blastocysts produce a specific, highly localized down-regulation/cleavage of MUC1 at attachment sites, as previously seen in rabbits [22].
In conclusion, our data suggest a novel hypothesis of mutual, cooperative paracrine signaling between maternal and embryonic cells at implantation, leading to the loss of an inhibitory barrier to implantation. This leads to a possible mechanism whereby poor-quality embryos, lacking the ability to cause down-regulation/cleavage of MUC1, might fail to implant, because the antiadhesive glycocalyx would remain in place. On the other hand, defective MUC1 endometrium might abnormal embryos to implant, leading to recurrent abortions, as has been suggested [16, 46]. These findings have far-reaching implications for assisted-reproduction programs, in which low implantation success rates remain a major problem. Further studies clearly are required to address the mechanisms of synergistic communication between maternal and embryonic cells.
ACKNOWLEDGMENTS
We would like to thank to Drs. Amparo Mercader and Arancha Galán (FIVIER) for in vitro system development, Alberto Alvarez (Centro de Citometría de Flujo, Universidad Complutense) for confocal microscopy support, and Jan Krüsser (Dusseldorf University) for primer design.
FOOTNOTES
1 Supported by Conselleria dEducació i Ciència (Generalitat Valenciana), Fundación Instituto Valenciano de Infertilidad, and grant FISss 98/0855 from the Spanish Government, Ministerio de Sanidad y Consumo, Madrid. 
2 Correspondence: Carlos Simón, FIVIER, C/Guardia Civil 23, 46020 Valencia, Spain. FAX: 34 963694735; csimon{at}interbook.net
Accepted: September 19, 2000.
Received: April 19, 2000.
REFERENCES
4 and 1 in human endometrial epithelial cells in vitro. J Clin Endocrinol Metab 1997; 82:2607–2616., IL-1ß and IL-1 ra: IL-1 endometrial microenvironment of the human embryo at the apposition phase under physiological and supraphysiological steroid level conditions. J Reprod Immunol 1996; 31:165–184.[CrossRef][Medline]This article has been cited by other articles:
|
|
 |
|
  F. Dominguez, S. Martinez, A. Quinonero, F. Loro, J.A. Horcajadas, A. Pellicer, and C. Simon CXCL10 and IL-6 induce chemotaxis in human trophoblast cell lines Mol. Hum. Reprod., July 1, 2008; 14(7): 423 - 430. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. K Gipson, T. Blalock, A. Tisdale, S. Spurr-Michaud, S. Allcorn, A. Stavreus-Evers, and K. Gemzell MUC16 Is Lost from the Uterodome (Pinopode) Surface of the Receptive Human Endometrium: In Vitro Evidence That MUC16 Is a Barrier to Trophoblast Adherence Biol Reprod, January 1, 2008; 78(1): 134 - 142. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. Brayman, N. Dharmaraj, E. Lagow, and D. D. Carson MUC1 Expression Is Repressed by Protein Inhibitor of Activated Signal Transducer and Activator of Transcription-y Mol. Endocrinol., November 1, 2007; 21(11): 2725 - 2737. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Diedrich, B.C.J.M. Fauser, P. Devroey, G. Griesinger, and on behalf of the Evian Annual Reproduction (EVAR) The role of the endometrium and embryo in human implantation Hum. Reprod. Update, July 1, 2007; 13(4): 365 - 377. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. L. Terry, L. Titus-Ernstoff, J. R. McKolanis, W. R. Welch, O. J. Finn, and D. W. Cramer Incessant Ovulation, Mucin 1 Immunity, and Risk for Ovarian Cancer Cancer Epidemiol. Biomarkers Prev., January 1, 2007; 16(1): 30 - 35. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Hirota, Y. Osuga, K. Koga, O. Yoshino, T. Hirata, C. Morimoto, M. Harada, Y. Takemura, E. Nose, T. Yano, et al. The Expression and Possible Roles of Chemokine CXCL11 and Its Receptor CXCR3 in the Human Endometrium J. Immunol., December 15, 2006; 177(12): 8813 - 8821. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Achache and A. Revel Endometrial receptivity markers, the journey to successful embryo implantation Hum. Reprod. Update, November 1, 2006; 12(6): 731 - 746. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Brayman, J. Julian, B. Mulac-Jericevic, O. M. Conneely, D. P. Edwards, and D. D. Carson Progesterone Receptor Isoforms A and B Differentially Regulate MUC1 Expression in Uterine Epithelial Cells Mol. Endocrinol., October 1, 2006; 20(10): 2278 - 2291. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A W Horne, E-N Lalani, R A Margara, and J O White The effects of sex steroid hormones and interleukin-1-beta on MUC1 expression in endometrial epithelial cell lines. Reproduction, April 1, 2006; 131(4): 733 - 742. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Dimitriadis, C.A. White, R.L. Jones, and L.A. Salamonsen Cytokines, chemokines and growth factors in endometrium related to implantation Hum. Reprod. Update, November 1, 2005; 11(6): 613 - 630. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. C. Enders, S. Meyers, C. A. VandeVoort, and G. C. Douglas Interactions of macaque blastocysts with epithelial cells in vitro Hum. Reprod., November 1, 2005; 20(11): 3026 - 3032. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
