| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Female Reproductive Tract |
Department of Obstetrics and Gynecology3
Department of Clinical Chemistry,4 Helsinki University Central Hospital, 00029 HUS, Finland
Department of Biological Sciences,5 Imperial College, London SW7 2AZ, United Kingdom
Department of Obstetrics and Gynecology,6 University of Hong Kong, Queen Mary Hospital, Hong Kong, China
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
decidua, ovum, pregnancy, sperm
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Similarity or dissimilarity of glycodelins from female reproductive tissues or pregnancy serum has not been determined with regard to their glycosylation or function. This would be important for understanding the temporal contraceptive and immunosuppressive activities that glycodelin may present at the implantation site and the fetomaternal interface [14, 15].
Changes in glycoprotein glycosylation have been reported to take place during the normal human menstrual cycle and during pregnancy, and they are regulated by steroid hormones [16]. Glycodelin-A isolated from amniotic fluid was the first well-defined glycodelin isoform (so-named because of its unique glycosylation pattern) [11], followed by glycodelin-S (from seminal plasma), with a strikingly different glycosylation pattern and biological activity [12]. Subsequently, there has been confusion about using the name glycodelin-A instead of glycodelin.
Because the glycosylation patterns are important for biological activity and also form the basis for glycodelin nomenclature, we examined the glycosylation patterns and biological activity of the glycodelin isolated in reproductive tissues from pregnant and nonpregnant women. We also examined interindividual differences.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The present study was approved by the Ethics Committee/Institutional Review Board of the Department of Obstetrics and Gynecology, Helsinki University Central Hospital, and the Ethics Committee of the University of Hong Kong. Individual midtrimester amniotic fluid samples were obtained from specimens examined for routine prenatal diagnosis of chromosomal abnormalities at 14–22 wk of gestation (n = 25). Purified glycodelin from pooled midtrimester amniotic fluid (n > 50) was used as a glycodelin-A standard in the present study. First-trimester decidual tissue was obtained from six women undergoing legal termination of pregnancy at 7–9 wk, and term pregnancy decidual tissue was obtained from four women after elective cesarean section. Secretory-phase endometrium samples were available from six women undergoing hysterectomy. The tissues were minced in PBS (BioWhittaker, Walkersville, MD) containing 10 mM bentzamidin, 200 trypsin inhibitory unit/L of aprotinin, and 1 mM EDTA (all from Sigma, St. Louis, MO) and homogenized by Ultra Turrax (Janke & Kunkel KG, Staufen, Germany). Cellular debris was removed by centrifugation (7000 x g, 0.5 h). Serum samples were available from women participating in routine prenatal screening at 15–16 wk of pregnancy. Glycodelin concentration was measured from serum samples by immunofluorometric assay detecting glycodelin-A and glycodelin-S glycoforms with similar efficiency [10], and those containing more than 0.3 µg/ml were pooled to obtain enough material for purification and analysis. For isoelectric focusing (IEF), the glycodelin-containing fraction was precipitated with 35%–60% (v/v) saturated ammonium sulfate before affinity purification. For the other analyses, the precipitation step was omitted. Glycodelins were purified by affinity chromatography using a monoclonal antiglycodelin antibody column [10]. Elution was performed using 0.1% trifluoroacetic acid.
Isoelectric Focusing, SDS-PAGE, Immunoblotting, and Silver Staining
Purified glycodelin from all samples was separated by NOVEX isoelectric focusing gel (pH 3–7, 5% polyacrylamide; Invitrogen Life Technologies, Carlsbad, CA) and SDS-PAGE (4%–20%, NOVEX) and transferred onto the polyvinylidene difluoride membranes (Millipore Corporation, Bedford, MA). The membranes were incubated with rabbit anti-human glycodelin antiserum [10], washed with PBS, and treated with peroxidase-conjugated anti-rabbit antibody (DAKO A/S, Glostrup, Denmark) using 3,3'-diaminobenzidine tetrahydrochloride (0.3 mg/ml) as a substrate for the staining reaction. Alternatively, the detection was performed using enhanced chemiluminescence substrate (Amersham, Buckinghamshire, U.K.). The isoelectric points (pI) were estimated using pI markers from the IEF calibration kit (Pharmacia, Uppsala, Sweden). The densitometric quantitation of relative amounts of five major glycodelin isoforms from midtrimester amniotic fluid (n = 13), first-trimester decidua (n = 5), and secretory endometrium (n = 4) was done by measuring the volume of bands, representing different isoforms, using the InGenius Bio Imaging System (SynGene, Cambridge, U.K.). In each different sample, the relative amount (volume) of each isoform was expressed as a percentage of the total amount of immunoreactive glycodelin, including some minor isoforms in addition to the five major ones. Some of the samples (0.4 µg/lane for SDS-PAGE or 1 µg/lane for isoelectric focusing) were analyzed by silver staining (the membrane was incubated in 70 mM sodium citrate, 30 mM iron sulfate, and 12 mM silver nitrate for 10 min; all from Merck, Darmstadt, Germany).
Lectin Immunoassays
Sandwich-type lectin immunoassays were carried out using lectins from Sambucus nigra (SNA) and Wisteria floribunda (WFA) and monoclonal anti-human glycodelin antibodies essentially as described previously [10]. Briefly, in the WFA lectin immunoassay, the microtitration plates were coated with WFA lectin, incubated with samples (equal amounts of glycodelin, as determined by immunofluorometric assay [10]), and probed with europium-labeled antiglycodelin antibody (clone F25-9D8). In the SNA lectin immunoassay, the monoclonal antiglycodelin antibody (clone F23-9G2) was biotinylated and allowed to attach to streptavidin-coated microtitration plates (Wallac, Turku, Finland), and after incubation with samples, the microtiter plates were treated with europium-labeled SNA lectin. The fluorescence was measured with a 1234 Delfia Research Fluorometer (Wallac). Some samples were also treated with neuraminidase (5 mU/µg of glycodelin for 2 h at room temperature; Sigma) to confirm the specificity of the lectin assays.
Glycan Analyses
The glycan analyses were done from pooled first-trimester decidua and from 13 individual midtrimester amniotic fluid glycodelin samples. The purity of the samples was confirmed by SDS-PAGE and silver staining. The purified glycodelin preparations were digested with bovine pancreatic trypsin (Sigma) at an enzyme:substrate ratio of 1:50 (w/w) at 37°C for 3 h. The products of trypsin digest were digested with 0.5 U of peptide N-glycosidase F (PNGase F; Roche Applied Science, Indianapolis, IN) in 200 µl at 37°C for 16 h. The released carbohydrates were separated from peptides by C18 Sep-Pak (Waters Corporation, Millford, MA) reversed-phase chromatography, permethylated using the sodium hydroxide procedure, and purified on C18 Sep-Pak cartridges as previously described [17]. These permethylated carbohydrate derivatives were then analyzed by fast-atom bombardment mass spectrometry (FAB-MS) using a ZAB-2SE-2FPD mass spectrometer fitted with a cesium gun operated at 30 kV as described previously [11]. The glycan structures attributed to each of the molecular ion signals were assigned on the basis of the presence of identical masses for the molecular and fragment ions in the original analysis of glycodelin-A [11].
Hemizona-Binding Assay
Hemizona-binding assay (HZA) was performed as described previously [18]. Briefly, unfertilized oocytes from the in vitro fertilization program at Queen Mary Hospital (Hong Kong) were bisected into two identical hemizonae by a micromanipulator. In the assay, the test spermatozoa were incubated with glycodelin and the control spermatozoa with buffer alone for 2 h, followed by washing with Earle balanced salt solution/BSA. Each hemizona was incubated with 2 x 104 spermatozoa in a 100-µl droplet of Earle balanced salt solution/BSA under mineral oil for 3 h at 37°C in an atmosphere of 5% CO2 in air. Loosely attached spermatozoa were removed by pipetting the hemizonae through a micropipette. The numbers of tightly bound spermatozoa were counted, and the hemizona-binding index was calculated by dividing the number of spermatozoa bound to the test hemizonae by the number of spermatozoa bound to the control hemizonae. Four different glycodelin concentrations of each sample were analyzed in quintuplicate.
Statistical Analyses
The Mann-Whitney U-test was used for comparison of differences between different groups/individuals. Spearman rank correlation was used for determining linear relationship between two variables.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The glycodelin preparations purified from secretory-phase endometrium, midpregnancy amniotic fluid (glycodelin-A), first-trimester pregnancy decidua, term pregnancy decidua, and pooled pregnancy serum all migrated on SDS-PAGE as one major band at 28 kDa (not shown). Some samples, especially serum glycodelin, also contained small amounts of immunoreactive dimers and larger aggregates. The same bands were positive both by immunoblotting and by silver staining. In isoelectric focusing, all these glycodelins were similar, having several isoelectric points between 4.1 and 5.2 (Fig. 1A). However, some minor differences were observed in the relative abundance of different isoforms between amniotic fluid, first-trimester decidual, and secretory-phase endometrial glycodelins (Fig. 1B). Glycodelin-S [10] was used as a differentially glycosylated control (pI 5.2). After isoelectric focusing, silver staining of different isoforms of glycodelin-A showed the same intensities as those seen by immunoblotting, confirming that affinity of the antibody used is not dependent on glycosylation of the samples (not shown).
|
Lectin Immunoassays
Purified glycodelins from first-trimester and term pregnancy decidua, secretory endometrium, pregnancy serum, and amniotic fluid reacted in the same way in SNA and WFA lectin immunoassays (Fig. 2A). However, equal amounts (1 µg/ml for SNA assay and 2 µg/ml for WFA assay) of individual glycodelin samples showed quantitative differences in their lectin reactivity (Fig. 2B). A negative correlation was observed between SNA and WFA reactivities (n = 32, r = -0.47, P < 0.01). The treatment with neuraminidase decreased SNA reactivity and increased WFA reactivity, confirming specificity of the lectin assays (results not shown).
|
Glycan Analyses
Glycodelin-A samples from 13 women were analyzed individually by FAB-MS. Representative data are shown in Figure 3, and the relative molecular ion abundances for all 13 samples are indicated by the data in Table 1. The glycan compositions shown in Table 1 are derived from the molecular ion masses (Table 1, column 2). The major structures corresponding to these compositions are indicated in Figure 3B. The major structures were consistent with those in the glycodelin-A standard [11], but each sample exhibited some variation in its specific N-glycan profile (Table 1).
|
|
The complex-type N-glycans detected in the mass spectra of glycodelin-A samples yield molecular ions (Fig. 3B and Table 1) and fragment ions (Fig. 3A) consistent with the presence of lacNAc (Galß1-4GlcNAc) and lacdiNAc (GalNAcß1-4GlcNAc) antennae that can be either sialylated or fucosylated. The antennae sequences are defined by characteristic fragment ions in the mass spectra that result from cleavage of each antenna at the GlcNAc residue (Fig. 3A). Semiquantitative data (Table 1) clearly show that the most abundant glycans differ between samples and that some of those, like high mannose structure (Hex5HexNAc2), which is one of the most abundant glycans in the majority of samples, are totally absent in some samples. Interestingly, the women whose glycodelin-A contained little or no high mannose structures (m/z 1580) were older than those exhibiting this structure in high abundance (P < 0.05), whereas glycodelin-A samples having weaker signals at m/z 2431, 2472, and 3008 were from younger women (P < 0.05 for all). All these structures, excluding high mannose, contain sialylated lacdiNAc and/or sialylated lacNAc. Samples GdA11, GdA12, and GdA13 were each examined in duplicate, and in all cases, both sets of data were almost identical (not shown).
Glycodelin from pooled first-trimester pregnancy decidua samples was found to contain the same glycans as glycodelin-A. The major nonreducing epitopes in the complex-type glycans were Galß1-4GlcNAc (lacNAc), GalNAcß1-4GlcNAc (lacdiNAc), NeuAc
2-6Galß1-4GlcNAc (sialylated lacNAc), NeuAc2-6GalNAcß1-4GlcNAc (sialylated lacdiNAc), Galß1-4(Fuc1-3)GlcNAc (Lewisx), and GalNAcß1-4(Fuc1-3)GlcNAc (lacdiNAc analogue of Lewisx) (Fig. 3).
Hemizona-Binding Assay
Glycodelin-A standard, glycodelin from secretory-phase endometrium (pool of two samples), first-trimester pregnancy decidua (pool of two samples), and four individual glycodelin-A samples (GdA11, GdA12, GdA13, and GdA14) were analyzed by HZA (Fig. 4). All glycodelin samples dose-dependently inhibited sperm binding to the hemizonae, but some were different from the glycodelin-A standard (Fig. 4) or from each other (GdA13 and GdA12 were different from GdA14, P < 0.05 for both). The hemizona-binding index correlated with SNA reactivity (Spearman r = -0.77, P = 0.048, n = 7) but not with WFA or IEF.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Physicochemical properties and biological activity of glycodelin from secretory-phase endometrium, first-trimester decidua, and midtrimester amniotic fluid were analyzed. Purified glycodelins from all these sources were similar regarding isoelectric points, reactions with WFA and SNA lectins, and sperm-egg binding inhibitory activity. Given the tissue-dependent heterogeneity of the characterized glycodelin isoforms from amniotic fluid and seminal plasma [11, 12] and from the follicular fluid and the cumulus matrix [20], it would seem to be irrational that biologically irrelevant sources of these proteins, such as pregnancy specimens, were chosen for the comparison of sperm-egg binding activity. The reason for choosing these sources was simply unproven assumptions. More precisely, it is believed that because amniotic fluid glycodelin-A inhibits human sperm-egg binding [21] and amniotic fluid proteins are probably derived from the maternal decidua, glycodelin from nonpregnancy and pregnancy endometrium would be expected have glycosylation pattern and biological activity similar to those of the amniotic fluid glycodelin-A. The present study provides evidence, to our knowledge for the first time, showing sufficient structural and biological similarity of the glycodelin isoforms from pregnancy and nonpregnancy endometrium and from the amniotic fluid, legitimizing (in retrospect) the use of the term glycodelin-A for the isoform from these sources. In addition to the endometrium, glycodelin has also been found in follicular fluid. It is synthesized in luteinized granulosa cells, and it is partially deglycosylated by the cumulus cells that do not synthesize but take up glycodelin [20]. Because of their differences in glycosylation and sperm-egg binding capacity as well as their different effects on acrosome reaction, glycodelin from follicular fluid has been named glycodelin-F to distinguish it from glycodelin-A and glycodelin-S [22, 23]. Differences in glycosylation may also have other biological consequences, because glycosylation may affect the proteins in many ways, ranging from folding and sorting of proteins to regulation of biological activity, stability, and half-life [24]. Interestingly, defective biosynthesis and distribution of glycoconjugates is associated with unexplained infertility [25, 26].
Pregnancy serum and term pregnancy decidua did not contain enough glycodelin for the sperm-egg binding assay. We analyzed glycodelin from these sources by isoelectric focusing and lectin assays. These analyses showed that glycodelin from term pregnancy decidua and pregnancy serum is similar to the other female reproductive tract glycodelin forms, with pregnancy serum having the weakest SNA reactivity. Our result of similar isoelectric points in the first-trimester and term pregnancy decidua and the midtrimester amniotic fluid glycodelins have two important implications. First, it is compatible with a decidual origin of these glycodelins [1]. Second, it suggests that the degree of glycodelin sialylation is not altered with advancing pregnancy. In this respect, decidual glycodelin is different from human chorionic gonadotropin, which becomes noticeably less acidic as pregnancy progresses [27]. Lectin-binding assays suggest that all female reproductive tract glycodelins contain terminal GalNAc that reacts with the lectin from WFA and NeuAc2-6Gal(NAc) that reacts with the lectin from SNA. This result is of interest, because these glycans may affect the biological activity of glycodelin by inhibiting E-selectin-mediated adhesion and binding to the B cell-associated receptor CD22 [15, 28, 29]. Recently, Jeschke et al. [30] showed that glycodelin from amniotic fluid and pregnancy serum inhibits E-selectin-mediated adhesion very efficiently. Interestingly, the glycodelin from amniotic fluid was a slightly better inhibitor than the glycodelin from pregnancy serum, and deglycosylation led to dramatically reduced activity [30]. Of note, human sperm-zona pellucida binding appears to require selectin-like interaction between human sperm and zona pellucida [15, 31].
Irrespective of the overall similarity, some minor quantitative physicochemical and functional differences were found between glycodelins from different sources as well as within the same tissue/fluid between different individuals. Glycodelin from secretory-phase endometrium especially was different from first-trimester decidual and amniotic fluid glycodelin regarding SNA and WFA reactivities and from first-trimester decidual glycodelin regarding inhibition in sperm-egg binding. The SNA reactivity of glycodelin was correlated with its sperm-egg binding inhibitory activity.
Amniotic fluid and first-trimester decidua were the only sources that contained sufficiently high glycodelin concentrations to allow definitive carbohydrate sequence analysis. These analyses revealed that decidual glycodelin contains the same complex-type glycans as amniotic fluid glycodelin-A. Most glycans of decidual and amniotic fluid have antennae composed of sialylated or fucosylated lacdiNAc sequences, which are rare in higher animals [11]. The N-linked oligosaccharides of this type, especially the fucosylated lacdiNAc-type antenna, have been shown to potently block selectin-mediated adhesions [28].
In addition to extensive similarities in the major glycans, we found minor interindividual differences in the biological activity and in the relative amounts of the glycans isolated from individual glycodelin samples. Interestingly, the amount of high mannose structures in individual glycodelin-A samples was smaller in older women. Whether this variation is caused by differences in the expression of glycosyl transferases and/or glycosidases in the endometrium cannot be determined by the present results. Because carbohydrate interactions play an important role in blastocyst implantation [32], further studies will be required to learn how common the age-related decline in high mannose structures is and whether it has any bearing on endometrial function and age-related decrease of implantation [33]. Unfortunately, the amount of samples from which both hemizona assay and glycan analyses were done was too low for meaningful statistical analyses aimed at showing correlation between individual glycans and sperm-egg binding inhibition.
The reason for interindividual and between-tissue differences might be variations in hormonal environment, because several terminal carbohydrate structures are regulated by ovarian steroids in mouse uterine luminal epithelium during pregnancy [16, 34].
It is concluded that despite substantial similarities between glycodelins from endometrium/decidua, pregnancy serum, and amniotic fluid, some minor quantitative differences were found in glycosylation and inhibition of sperm-egg binding. Differences in glycosylation might have a role in regulation of the biological activity of glycodelin.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Riitta Koistinen, Department of Obstetrics and Gynecology, Helsinki University Central Hospital, Biomedicum Helsinki, P.O. Box 700, Haartmaninkatu 8, 00029 HUS, Finland. FAX: 358 9 47171731; riitta.koistinen{at}hus.fi
Received: 2 April 2003.
First decision: 1 May 2003.
Accepted: 9 June 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2-globulin, the major secretory protein of the luteal phase and first trimester pregnancy endometrium, is not glycosylated prolactin but related to ß-lactoglobulins. J Clin Endocrinol Metab 1987 65:1067-1071[Abstract]This article has been cited by other articles:
|
|
 |
|
  B. Toth, K. Roth, C. Kunert-Keil, C. Scholz, S. Schulze, I. Mylonas, K. Friese, and U. Jeschke Glycodelin Protein and mRNA Is Downregulated in Human First Trimester Abortion and Partially Upregulated in Mole Pregnancy J. Histochem. Cytochem., May 1, 2008; 56(5): 477 - 485. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Seppala, H. Koistinen, R. Koistinen, P.C.N. Chiu, and W.S.B. Yeung Glycosylation related actions of glycodelin: gamete, cumulus cell, immune cell and clinical associations Hum. Reprod. Update, May 1, 2007; 13(3): 275 - 287. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Lapid and N. Sharon Meet the multifunctional and sexy glycoforms of glycodelin Glycobiology, March 1, 2006; 16(3): 39R - 45R. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. A. Horcajadas, A. Riesewijk, J. Polman, R. van Os, A. Pellicer, S. Mosselman, and C. Simon Effect of controlled ovarian hyperstimulation in IVF on endometrial gene expression profiles Mol. Hum. Reprod., March 1, 2005; 11(3): 195 - 205. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
