0013-7227/02/$15.00/0 Printed in U.S.A.

The Journal of Clinical Endocrinology & Metabolism 87(12):56415648 Copyright 2002 by The Endocrine Society doi: 10.1210/jc.2002-020766

Variation in the Amount of T Antigen and N-Acetyllactosamine Oligosaccharides in Human Cervical Mucus Secretions with the Menstrual Cycle
PABLO ARGUESO, SANDRA SPURR-MICHAUD, ANN TISDALE,
AND

ILENE K. GIPSON

Schepens Eye Research Institute and Department of Ophthalmology, Harvard Medical School, Boston, Massachusetts 02114
It has been hypothesized that the carbohydrate portion of mucins present in the endocervical canal plays an important role in conferring specific physicochemical properties (e.g. viscosity and hydration) to the mucus gel through the menstrual cycle. Our recent finding showing an increase in the amount of MUC5B mucin protein at midcycle has raised the question of whether the mucin O-glycan content also varies to confer specific hydrodynamic properties to secreted mucins during ovulation. Using lectins as carbohydrate probes, we have identified two common mucin oligosaccharide structures, T antigen and N-acetyllactosamine, within secretory granules in human endocervical glands during the proliferative phase of the menstrual cycle. Analysis of endocervical secretions by enzyme-linked lectin assay revealed that the amounts of T antigen and N-acetyllactosamine are maximal at midcycle. Lectin blot assay of immunoprecipitated MUC5B demonstrated that the mucin is a carrier of the T antigen and N-acetyllactosamine oligosaccharides in cervical mucus secretions. The amounts of T antigen and N-acetyllactosamine oligosaccharides on MUC5B increased during the first half of the cycle, peaked at midcycle, and dramatically dropped at the end of the cycle. The peak in MUC5B mucin protein and carbohydrate content coincides with the change in mucus character that occurs at midcycle. The role of O-glycans on mucins may be to hold water within the endocervical canal during ovulation to facilitate sperm migration. (J Clin Endocrinol Metab 87: 56415648, 2002)

UMAN CERVICAL MUCUS plays an important role in the protection of the endocervical epithelium against bacterial invasion, fluid loss, and, most importantly, in the regulation of sperm transport to the upper reproductive tract (1). Changes in the physicochemical properties of the cervical mucus during the different phases of the menstrual cycle are known to correlate directly with receptivity to sperm, thus contributing to reproductive success (2, 3). The cervical mucus is a viscoelastic gel composed primarily of water, ions, various secreted substances, and cells. The major structural components ( 85% of the total macromolecules) of the mucus gel in the cervix are a group of extraordinarily long and heavily O-glycosylated proteins, termed mucins (4). To date, 16 mucin genes have been described in humans (57). The endocervical epithelium expresses at least 5 different mucin genes, the secreted MUC5AC, MUC5B, and MUC6 mucins and the membrane-spanning MUC1 and MUC4 mucins (8, 9). Of these, the most predominant mucin transcripts are MUC5B and MUC4 mucins (10). The specific rheological and hydrodynamic properties of mucins can be ascribed to their extensive O-linked glycosylation (11, 12). Most of the carbohydrate chains on cervical mucins, which represent up to 65% of the mass by weight, occur as a microheterogeneous population of neutral, sialylated, and sulfated oligosaccharides varying from two to nine sugar residues in length (13, 14). Compositional analysis of the O-linked carbohydrate chains in preparations of purified human cervical secretions collected at midcycle has revealed
Abbreviations: ECA, Erythrina cristagalli lectin; ELLA, enzyme-linked lectin assay; FITC, fluorescein isothiocyanate; GalNAc, N-acetylgalactosamine; PNA, peanut agglutinin lectin.

high levels of galactose, N-acetylgalactosamine (GalNAc), N-acetylglucosamine, fucose, and neuraminic acid (15). Structural studies of cervical mucins as well as in other epithelial mucins have demonstrated the common presence of the tetrasaccharide core structure Gal 1 4GlcNAc 1 6(Gal 13)GalNAc 1 linked to serine or threonine, suggesting that this is an integral structural unit of human epithelial mucins (14, 16). The Thomsen-Friedenreich antigen or T antigen (Gal 13GalNAc 1Ser/Thr) constitutes the inner core of this structure and is recognized by the Arachis hypogaea or peanut agglutinin lectin (PNA). N-Acetyllactosamine and poly-N-acetyllactosamine [recognized by the Erythrina cristagalli lectin (ECA)] result from the sequential addition of Gal 1 4GlcNAc 13 residues to the terminal Gal 1 4GlcNAc 1 6 motif of the tetrasaccharide (Fig. 1), although they may also be present in other O-linked structures, Nglycans or glycolipids (17, 18). PNA and ECA have been previously used to detect carbohydrate ligands on mucins (19). Varying levels of estrogen and progesterone are responsible for the cyclic changes in the amount of mucus and its properties in the endocervix. We have recently reported a sharp increase in MUC5B protein in cervical secretions at midcycle in normal cycling subjects (20). The amount of both MUC5B mRNA and protein abruptly falls as serum progesterone levels rise during the secretory phase of the menstrual cycle, suggesting an inverse correlation between MUC5B mucin expression in the cervical epithelium and progesterone. Although an increase in MUC5B mucin protein secretion by the endocervical epithelial cells at midcycle may indicate that there is a concomitant increase in the amount of mucin O-linked oligosaccharides, early attempts to quantify the levels of carbohydrates in endocervical secretions during

5641

5642

J Clin Endocrinol Metab, December 2002, 87(12):56415648

Argueso et al. Oligosaccharides in Cervical Mucus during Menstrual Cycle

FIG. 1. Diagram depicting the synthesis of the T antigen and N-acetyllactosamine in mucin O-linked oligosaccharides through the core 2 intermediate.

the ovulatory cycle yielded conflicting results. Evidence both for (21, 22) and against (23, 24) variation in the carbohydrate composition of cervical mucin through the menstrual cycle has been presented. In this study using a highly sensitive enzyme-linked lectin assay (ELLA), we evaluated the contents of T antigen and N-acetyllactosamine, two carbohydrate structures previously demonstrated in endocervical mucus secretions (14, 16), throughout the menstrual cycle of healthy women. Additionally, using lectin blot assay we compared the amount of lectin binding to equivalent amounts of MUC5B purified by immunoprecipitation from the same mucus samples taken during the cycle. The relative amounts of T antigen and N-acetyllactosamine were correlated with blood levels of progesterone and LH surge data for each subject.
Materials and Methods Tissues and secretions
Human endocervical tissue and cervical secretions were collected in accordance with human study guidelines and approval from the Schepens Eye Research Institute and Brigham and Womens Hospital institutional review boards. Informed consent was obtained from all subjects in accordance with the approved protocols. Endocervical tissue was obtained from four females at the time of hysterectomy and frozen within 30 min of surgery for cryostat sectioning as described previously (20). The menstrual cycle phase for each woman (two proliferative and two secretory) was estimated on the basis of histological appearance of the endometrium obtained from that patient as described by Noyes et al. (25). Samples of cervical secretions were obtained from normal cycling females who were not using intrauterine devices or oral contraceptives, had normal cervical cytology, and were free from infection. These subjects were asked to self-report cycle day and to determine the LH surge with a commercially available urinary LH detection kit (Clear Plan Easy, Unipath Ltd., Bedford, UK). Initially, six subjects were enrolled in the study, although only three completed the entire protocol. Samples were collected from this population at four time points within the menstrual cycle as described previously (20). Collection points were chosen to be on d 4 and 7 after the start of menses and on d 1 and 7 after the LH surge at midcycle. When possible, duplicate samples for each time point were taken, but only two collections were made in any cycle, so collection proceeded over four cycles. Cervical mucus was obtained by swabbing the cervix with Wilshire foam swabs (VWR Scientific Products, Bridgeport, NJ), and then the mucus was extracted from the swab as previously described (20). The protein concentration of each sample was determined using the bicinchoninic acid protein assay (Pierce Chemical Co., Rockford, IL). Levels of MUC5B protein in the mucus secretions of these three patients throughout the menstrual cycle were previously determined (20). Blood levels of progesterone for each col-

lection point were determined by the Reproductive Endocrine Sciences Center Assay Core Laboratory, Massachusetts General Hospital (Boston, MA) as previously described (10).

Lectin histochemistry
Cellular localization of lectin binding to sections of human endocervix was performed following methods previously reported (26). Fluorescein isothiocyanate (FITC)-conjugated lectins purified from Arachis hypogaea (PNA) and Erythrina cristagalli (ECA) were purchased from Vector Laboratories, Inc. (Burlingame, CA). PNA is specific for d-galactose residues at the nonreducing terminal position of glycoconjugates and strongly reacts with the T antigen (19). ECA shows strong specificity for Nacetyllactosamine and to a lesser extent for galactose and N-acetylgalactosamine (19). Cryostat sections (6 m thick) were placed on gelatincoated glass slides, air-dried at room temperature for 1 h, and rehydrated in PBS (pH 7.2). Nonspecific background staining was blocked with PBS containing 1% BSA (Sigma, St. Louis, MO). Slides subjected to PNA staining were previously digested with 5 mU neuraminidase from Clostridium perfringens (Calbiochem, San Diego, CA) for 60 min at 37 C to enhance epitope disclosure to the lectin as previously observed (27). FITC-conjugated lectins diluted to 10 g/ml in PBS were applied to sections for 20 min at room temperature. After washing with PBS and coverslipping with Vectashield mounting medium with propidium iodide (Vector Laboratories, Inc.), sections were viewed on a Photomicroscope III (Carl Zeiss, New York, NY). As a lectin specificity control, lectins were preincubated with d( )-galactose at a concentration of 0.2 m for 20 min before application to sections. Immunohistochemical localization of MUC5B was determined as previously described (20).

ELLA
ELLA was used to determine the relative amounts of T antigen and N-acetyllactosamine in crude preparations of human cervical mucus throughout the menstrual cycle. Samples tested for PNA binding were previously digested with 10 mU neuraminidase from C. perfringens for 60 min at 37 C. Preliminary serial dilution experiments using decreasing amounts of crude cervical mucus collected at midcycle were carried out to determine the range of linear response of the lectins in the assay. Cervical secretions (7.5 g when testing PNA or 0.4 g when testing ECA) corresponding to the proliferative, ovulatory, and secretory phases of normal cycling females were coated in triplicate onto microtiter plates (Costar, Cambridge, MA) and kept overnight at 4 C in 0.05 m carbonate/bicarbonate buffer (pH 9.6). The plates were washed four times with PBS and blocked for 2 h with PBS containing 1% BSA. After three washes with PBS, the plates were incubated with peroxidaselabeled PNA or ECA (Sigma) diluted to 0.5 g/ml (PNA) or 4.0 g/ml (ECA) in PBS for 60 min at 37 C. After washing, the plates were incubated with the peroxidase substrate, tetramethylbenzidine (Sigma), at room temperature. The reaction was stopped after 30 min by the addition of 0.5 m H2SO4. The OD of each well was read at 450 nm in a SpectraMax microplate spectrophotometer system using SoftMax Pro version 2.1

Argueso et al. Oligosaccharides in Cervical Mucus during Menstrual Cycle

J Clin Endocrinol Metab, December 2002, 87(12):56415648 5643

software (Molecular Devices, Sunnyvale, CA). All samples from a single individual were run in the same ELLA plate. Secretions were tested and were found to be negative for endogenous peroxidase activity by incubation with peroxidase substrate.

Electrophoresis and Western blot

Glycoproteins in human cervical secretions were separated by SDSPAGE and blotted onto nitrocellulose membranes as described previously (26). Briefly, SDS-PAGE was performed using the method of Laemmli (28). Electrophoresis was carried out under reducing conditions on 6% separating and 4% stacking polyacrylamide gels. Proteins in gels were transferred to nitrocellulose membranes (Millipore Corp., Bedford, MA) according to the method of Towbin et al. (29). After blotting, membranes were incubated with blocking solution containing 10% normal horse serum in Tris-buffered saline (pH 7.5) for 30 min and with peroxidase-labeled PNA or ECA (20 g/ml) for 60 min at room temperature. Membranes were then washed well with Tris-buffered saline, followed by colorimetric detection of positive binding with diaminobenzidine peroxidase substrate (Bio-Rad Laboratories, Inc., Richmond, CA). MUC5B protein on nitrocellulose membranes was detected as described previously (20). Prestained molecular mass markers included myosin (207 kDa), -galactosidase (120 kDa), and BSA (78 kDa).

nuclear stain propidium iodide (data not shown). The intense reactivity was observed in most, but not all, endocervical cells. Binding of PNA and ECA to endocervical epithelial cells was less intense or negative in tissue taken from the secretory phase of the menstrual cycle (Fig. 2, D and E), indicating that the mucus stored in the endocervical glands during the proliferative phase has been secreted into the cervical canal during the secretory phase. Similarly, binding of the MUC5B antibody to sections of endocervix was greatly diminished in the late secretory phase (Fig. 2F) compared with the proliferative phase (Fig. 2C), thus correlating to the reduction in PNA and ECA binding. No binding was observed in sections from the proliferative phase in which PNA and ECA were previously incubated with d( )-galactose at a concentration of 0.2 m, indicating the lack of nonspecific binding (Fig. 2, G and H). All sections probed with PNA were pretreated with neuraminidase, because initial experiments showed increased intensity in lectin binding compared with untreated sections (data not shown).
Lectin binding to cervical secretions

MUC5B mucin immunoprecipitation

Immunoprecipitation of MUC5B mucin from individual samples taken throughout the menstrual cycle was carried out using a polyclonal antibody raised against a synthetic peptide from the deduced amino acid sequence of a unique region of the D4 domain (nontandem repeat region) of MUC5B (20). The specificity of this antibody toward MUC5B mucin has been previously demonstrated by preadsorption experiments on Western blot and ELISA as well as immunohistochemistry on tissues expressing other membrane-bound and gel-forming mucins (20). To compare amounts of T antigen and N-acetyllactosamine per mucin molecule throughout the cycle, identical units of MUC5B protein were immunoprecipitated at each time point analyzed in this study. An MUC5B unit was defined as the picogram amount of purified cervical mucin standard that corresponds to the OD reading obtained per sample in a quantitative MUC5B ELISA (20). The anti-MUC5B antibody was covalently attached to amine-terminated magnetic particles (Sigma) using the glutaraldehyde procedure as described in the manufacturers protocol. Conjugated beads were transferred to a sterile tube and washed five times with a buffer containing 10 mm Tris-HCl, 2 mm EDTA, 0.1% Triton X-100, and 0.1% SDS (pH 7.4). Samples to be assayed for PNA binding were digested with 250 mU neuraminidase from C. perfringens for 60 min at 37 C before incubation with beads. MUC5B protein units (5.7 106) of cervical mucus in mucin isolation buffer (0.1 m NH4HCO3, 2.0 mm phenylmethylsulfonylfluoride, 0.5 m NaCl, 5 mm EDTA, 2 mm N-ethylmaleimide, and 0.02% NaN3) were added to the conjugated beads and incubated for 2 h on a rocker at room temperature. After three washes, the beads were boiled for 2 min with 30 l 2 Laemmli buffer (28) and centrifuged, and the supernatant was loaded onto a gel for SDS-PAGE and Western blot as described above. As a control, cervical mucus was immunoprecipitated with magnetic particles conjugated to an isotype-matched Ig of irrelevant specificity.

Results Lectin binding to proliferative and secretory endocervical tissue

The two lectins used in this study, Arachis hypogaea (PNA) and Erythrina cristagalli (ECA), bound to sections of human endocervical epithelium (Fig. 2). Strong binding was observed in endocervical epithelia from biopsies taken during the proliferative phase of the menstrual cycle (Fig. 2, A and B), demonstrating the presence of T antigen and N-acetyllactosamine in intracellular compartments in these specimens. Both lectins bound to cytoplasmic granules in the supranuclear region of the epithelial cells, as visualized with the

The presence of the T antigen and N-acetyllactosamine was initially demonstrated in midcycle cervical secretions using the ELLA method. Peroxidase-labeled PNA and ECA consistently bound to microtiter plates coated with decreasing amounts of cervical mucus, producing linear responses starting with nanogram quantities of total protein (Fig. 3A). ECA binding was observed when plates were coated with at least 6 ng total protein and was maximal when they were coated with 400 ng. Binding of PNA was only detectable in cervical secretions after previous treatment of the samples with neuraminidase (Fig. 3A). The linear response of PNA binding to cervical mucus was ranged between 500 ng and 7.5 g total protein (Fig. 3A). Nonspecific binding of peroxidase-conjugated lectins to microtiter wells as well as endogenous peroxidase activity in mucus secretions was tested and found to be negative. Western blot analysis further confirmed binding of PNA and ECA to human cervical secretions (Fig. 3B). After neuraminidase treatment of the sample, PNA bound most intensely to a band in the upper region ( 207 kDa) of the separating gel (Fig. 3B, lane 1). The electrophoretic mobility of this band was similar to those usually obtained when analyzing secreted mucins (20). Two additional bands of high molecular mass ( 100 and 200 kDa) were detected in the separating gel, indicating the presence of the T antigen in additional glycoproteins. ECA exclusively bound to a high molecular mass band ( 207 kDa) within the separating gel (Fig. 3B, lane 2). Western blot analysis of cervical secretions with the anti-MUC5B antibody (Fig. 3B, lane 3) confirmed previous results showing strong binding to a high molecular mass band (20).
T antigen and N-acetyllactosamine content in cervical secretions taken through the menstrual cycle

The relative amounts of the T antigen and N-acetyllactosamine in cervical secretions of the three normal cycling females varied among the proliferative, ovulatory, and secretory phases, as determined by ELLA (Fig. 4A). During the

5644

J Clin Endocrinol Metab, December 2002, 87(12):56415648

Argueso et al. Oligosaccharides in Cervical Mucus during Menstrual Cycle

FIG. 2. Lectin and MUC5B antibody binding to sections of proliferative (AC) and secretory (DF) human endocervix. FITC-conjugated PNA and ECA bound strongly to intracellular granules within cervical epithelium taken during the proliferative phase of the menstrual cycle (A and B). In secretory phase tissues, binding of the lectins was considerably reduced or abolished (D and E). A decrease in MUC5B mucin content during the late secretory phase (F) compared with the proliferative phase (C) was also observed. Lack of nonspecific binding of PNA and ECA was demonstrated in control sections from the proliferative phase and probed with lectins previously incubated with D( )-galactose at a concentration of 0.2 M (G and H). The MUC5B antibody was omitted in a consecutive section of C to demonstrate the lack of nonspecific binding of the secondary antibody (I). Sections tested for PNA binding were previously digested with neuraminidase as described in Materials and Methods. Scale bar, 50 m.

FIG. 3. Detection of the T antigen and N-acetyllactosamine in human cervical secretions taken at midcycle. A, ELLA experiments using decreasing amounts of total protein, demonstrating binding of peroxidase-conjugated PNA and ECA to midcycle cervical mucus. Protein probed with PNA was analyzed before and after digestion with neuraminidase, as described in Materials and Methods. Wells filled with PBS were used as negative controls to test for nonspecific lectin binding. B, Western blot of SDS-PAGE containing 10 g (PNA) and 25 g (ECA, MUC5B) total protein from human midcycle cervical secretions probed with peroxidase-labeled PNA, ECA, and anti-MUC5B. Protein tested for PNA binding was previously digested with neuraminidase. The molecular masses of the markers are indicated in kilodaltons.

proliferative phase, binding of PNA and ECA to cervical secretions was low or undetectable. At the time of ovulation, binding of PNA to desialylated cervical mucus increased 3-

to 17-fold, whereas ECA binding to cervical mucus increased 3- to 14-fold. During the secretory phase, binding of PNA to the mucus secretions still remained elevated compared with

Argueso et al. Oligosaccharides in Cervical Mucus during Menstrual Cycle

J Clin Endocrinol Metab, December 2002, 87(12):56415648 5645

FIG. 4. Change in the amount of the T antigen and N-acetyllactosamine in cervical secretions from three normal cycling females taken throughout the menstrual cycle, as determined by ELLA. A, OD values for PNA () and ECA (f) binding were obtained after coating wells in triplicate with samples from each individual taken at different time points over four consecutive hormone cycles. The values represent the mean SEM for each time point and are plotted against the cycle day, relative to LH surge (d 0, determined by urinary LH detection kit). B, For each subject in A, the correlative amounts of MUC5B mucin in cervical secretions and blood levels of progesterone (nanograms per milliliter) on the collection day relative to LH surge are shown below [data obtained from Gipson et al. (20)].

the sharp reduction to low or undetectable levels observed with ECA binding. These results indicate that the amount of the T antigen in mucus secretions peaks during the ovulatory and secretory phases of the cycle, whereas the level of Nacetyllactosamine peaks exclusively during the ovulatory phase and drops sharply during the secretory phase. The increase in the amounts of the T antigen and N-acetyllactosamine at midcycle correlates with the peak in MUC5B mucin in the same subjects, as determined previously using ELISA techniques (Fig. 4B). In the secretory phase, levels of MUC5B mucin dramatically dropped, following the same pattern as N-acetyllactosamine. However, the content of T

antigen remained elevated compared with the reduced levels of MUC5B. Ovulation was verified for each cycle by measuring the LH surge, followed by an increase in serum progesterone levels. As shown in Fig. 4B, the increase in the amount of progesterone correlated with the decrease in ECA binding during the secretory phase of the cycle.
T antigen and N-acetyllactosamine contents in MUC5B mucin throughout the menstrual cycle

We then determined whether T antigen and N-acetyllactosamine are structural parts of MUC5B and whether they

5646

J Clin Endocrinol Metab, December 2002, 87(12):56415648

Argueso et al. Oligosaccharides in Cervical Mucus during Menstrual Cycle

vary on MUC5B through the cycle. MUC5B was isolated by immunoprecipitation from cervical secretions taken from different time points during the menstrual cycle, and equivalent amounts were electrophoresed, blotted onto nitrocellulose membrane, and probed with peroxidase-labeled PNA and ECA. Sample size limitations restricted the analysis to samples collected exclusively from subject 3. Based on the assay demonstrating the amount of MUC5B in each of the cervical mucus samples taken through the cycle (Fig. 4B), we calculated the amount of total protein needed to obtain equivalent amounts of MUC5B for each time point. Thus, 228, 203, 114, 100, and 285 g from samples taken on cycle d ( LH surge) 11, 8, 5, 1, and 6, respectively, were used to immunoprecipitate equivalent amounts of MUC5B mucin (5.7 106 MUC5B units). Three immunoprecipitations were performed for each sample (time point), and each was electrophoresed and blotted. One was probed for PNA, one for ECA, and one with MUC5B antibody (control) to verify that equivalent amounts of MUC5B were loaded on the gel. As shown in Fig. 5A, the secreted MUC5B mucin in the endocervical epithelium carried both T antigen and Nacetyllactosamine. The amount of oligosaccharides on MUC5B varied throughout the menstrual cycle, reaching maximal values at the time of ovulation (cycle d 1) and dropping dramatically during the secretory phase (cycle d 6). The decrease in T antigen and N-acetyllactosamine in MUC5B after ovulation coincided with an increase in progesterone levels in blood, as shown in Fig. 4B. The amount of MUC5B protein immunoprecipitated at each time point was similar, as seen in the membrane probed with the MUC5B antibody. Abnormal migration of the band corresponding to MUC5B on cycle d 6 may be due to decreased levels of carbohydrates on MUC5B or altered mucin O-glycosylation.

The absence of nonspecific binding to the magnetic particles was demonstrated by incubation of crude preparations of cervical secretions with the particles covalently linked to an isotype-matched Ig of irrelevant specificity (Fig. 5B).
Discussion

This study demonstrates that the amount of mucin-associated oligosaccharides in human cervical secretions varies through the menstrual cycle. Levels of T antigen and Nacetyllactosamine in cervical mucus peak at midcycle and correlate with the results of our previous study, which showed that the gel-forming MUC5B mucin is maximal at the time of ovulation (20). Immunoprecipitation experiments have shown that both the T antigen and N-acetyllactosamine are carried by the MUC5B mucin and suggest that variation in mucin O-glycosylation contributes to the specific biophysical properties (e.g. viscosity and water retention capacity) of cervical mucus throughout the menstrual cycle. It has been hypothesized that cyclic variations in the carbohydrate composition of mucins and glycoproteins play an important role in controlling cervical mucus rheology and hydration (2, 22). Rheology seems to be affected by the presence of negatively charged terminal carbohydrates on mucus glycoproteins, because they contribute to the distension and reduced flexibility of the mucin protein coil and thus to increase mucus gel viscosity (22). Using histochemical and chemical techniques, several researchers have reported a reciprocal relationship between neutral and charged terminal carbohydrates in human cervical mucus during the menstrual cycle (21, 22, 30, 31). Levels of neutral carbohydrates in terminal positions (e.g. fucose) seem to be more abundant during the ovulatory phase, whereas negatively charged residues (e.g. sialic acid) peak during the periovulatory phase.

FIG. 5. Change in the amount of the T antigen and N-acetyllactosamine on MUC5B mucin throughout the menstrual cycle, as determined by specific immunoprecipitation. A, MUC5B protein units (5.7 106) purified from mucus samples taken throughout the menstrual cycle [cycle d ( LH) surge, 11, 8, 5, 1, and 6] in subject 3 were immunoprecipitated with anti-MUC5B-conjugated particles and probed with PNA, ECA, and the MUC5B antibody (control). Protein tested for PNA binding was previously digested with neuraminidase as described in Materials and Methods. B, Magnetic particles conjugated to an isotype-matched Ig of irrelevant specificity were used to demonstrate the lack of nonspecific binding of the beads to the MUC5B mucin. Lane a, Ten micrograms of protein from cervical secretions collected at midcycle; lane b, 100 g protein from cervical secretions collected at midcycle immunoprecipitated with anti-MUC5B; lane c, 100 g protein from cervical secretions collected at midcycle immunoprecipitated with an isotype-matched Ig of irrelevant specificity. The blot was probed with the anti-MUC5B antibody. The arrow indicates the interface between the stacking and separating gels. The molecular masses of the markers are indicated in kilodaltons.

Argueso et al. Oligosaccharides in Cervical Mucus during Menstrual Cycle

J Clin Endocrinol Metab, December 2002, 87(12):56415648 5647

Substitution of sialic acid with fucose at midcycle is thought to reduce the total intrinsic negative charge of mucin oligosaccharides and consequently diminish the viscosity of the cervical mucus (22). The data presented in this study provide additional evidence showing that the amount of carbohydrate on cervical mucus varies throughout the menstrual cycle and suggests that they may have a central role in cervical mucus hydration. Due to the hydrophilic character of carbohydrates, we propose that higher amounts of mucin oligosaccharides are required during the ovulatory phase of the cycle to hold water within the endocervical canal and thus increase sperm mobility through the cervical canal. Recently, we reported an increase in MUC5B mucin, the major gel-forming mucin expressed in the reproductive tract epithelia, at midcycle (20). Maximal amounts of T antigen and N-acetyllactosamine oligosaccharides and MUC5B protein during ovulation correlate with earlier reports of increased water content in the cervical mucus at midcycle (97.8 99%) compared with the secretory phase mucus (91%) (2, 32). Taken together, these results indicate that cervical secretions at midcycle contain more mucin and that its carbohydrate content is higher and less acidic than during the periovulatory phase, resulting in a more hydrated and less viscous mucus gel. Other researchers reported no cyclic variations in the carbohydrate composition of purified human cervical mucus collected serially during the menstrual cycle of healthy women (23, 24). In these studies the macromolecular components from samples of cervical secretions were isolated by isopycnic gradient centrifugation and/or size exclusion chromatography. These techniques, however, do not allow the separation of individual mucin gene products and may underestimate the contributions of smaller mucins and/or glycoproteins to the total carbohydrate content in cervical secretions. Several mucin genes are expressed by human endocervical epithelia (8). Our results indicate that the gel-forming MUC5B mucin is differentially glycosylated during the menstrual cycle. After ovulation, the amount of T antigen linked to MUC5B dramatically drops, but remains elevated in cervical secretions, as demonstrated by ELLA, suggesting that other mucins may carry this glycan epitope during the secretory phase. Candidates include MUC5AC and MUC6, which are expressed by endocervical epithelia (8). Semiquantitative analysis of MUC5AC and MUC6 mRNA demonstrate, however, that they are expressed at lower levels (10). Due to its high level of expression in the reproductive tract epithelium (10), the membrane-spanning MUC4 mucin may be considered as contributing to endocervical mucus carbohydrate composition. Although MUC4 contains a hydrophobic domain that anchors the mucin to the cell surface, a soluble form has been found in the rat female reproductive tract during the estrous cycle (33). MUC4 has also been demonstrated to be present in secretory granules of colonic and lacrimal gland epithelia, suggesting that these other epithelia also produce a soluble secreted form of the mucin (34, 35). Determining whether the human MUC4 ectodomain is shed or secreted from the endocervical epithelium and also its glycosylation pattern during the menstrual cycle will help to

evaluate the contributions of individual mucins to the total glycan content in cervical secretions. Two major factors seem to regulate the incorporation and release of carbohydrates in cervical mucus glycoproteins: 1) the hormonally regulated expression of glycosyltransferases genes, and 2) the presence of hydrolytic enzymes associated with the cervical mucus (31, 36). As demonstrated by Chilton et al. (37), a reduction in the titer of serum estradiol in ovariectomized rabbits reduced the specific activity in the endocervix of the polypeptide GalNAc transferase, an enzyme that catalyzes the initial step in the biosynthesis of O-linked glycans on proteins, including the mucins. Supplementation with exogenous estradiol in these rabbits returned GalNAc transferase activity to estrous control levels. In progesteronedominated pseudopregnant rabbits, GalNAc transferase activity is also decreased, suggesting that the estradioldependent GalNAc transferase activity is antagonized by progesterone (37). The data presented in this study reveal that high levels of progesterone in the secretory phase of the cycle are associated with a decrease in the amount of Nacetyllactosamine in cervical mucus secretions and T antigen on MUC5B. Thus, it is possible that low levels of polypeptide GalNAc transferase or other related enzymes are responsible for the reduction in the amount of carbohydrates during the periovulatory phase of the cycle. Additionally, the presence of hydrolytic enzymes in cervical secretions may account for the reduced levels of carbohydrates during the secretory phase. It has been reported that cyclic changes in the amounts of sialidase and -l-fucosidase affect the terminal glycosylation of human cervical mucins throughout the menstrual cycle (31, 38). Furthermore, the enzymatic activity of these enzymes varies depending on the pH gradient within the reproductive tract and may also affect the final carbohydrate content of mucins (36). Taken together, these data suggest that the structure and amount of mucin O-glycans in cervical mucus depend on the ratio between synthesis and degradation of carbohydrates throughout the menstrual cycle and support the role of steroids in regulating this process. This study provides preliminary evidence for the presence of T antigen (Gal 13GalNAc 1Ser/Thr) and N-acetyllactosamine (Gal 1 4GlcNAc 13) on the major gel-forming MUC5B mucin in human cervical secretions. As shown by immunofluorescence microscopy, the pattern of PNA and ECA binding in human endocervical glands was similar to that of MUC5B binding during the proliferative stage (20). Furthermore, specific immunoprecipitation of MUC5B mucin from crude secretions of cervical mucus has shown that purified MUC5B binds to PNA and ECA. Lack of PNA binding to mucus secretions in the absence of neuraminidase treatment indicated that the T antigen might be sialylated (39). These glycans have recently been found as common inner structures on the MUC5B mucin isolated from human submandibular/sublingual saliva of a blood group O individual (40). The core structures on salivary MUC5B mucin are further elongated and terminated in a large and diverse repertoire of oligosaccharides (40), which may be differentially attached to the MUC5B protein backbone to generate different glycoforms (41, 42). Wickstrom et al. (41) suggested that the different MUC5B glycoforms are produced by different cell populations, presumably expressing different rep-

5648 J Clin Endocrinol Metab, December 2002, 87(12):56415648

Argueso et al. Oligosaccharides in Cervical Mucus during Menstrual Cycle

ertoires of glycosyltransferases. Interestingly, these researchers failed to detect MUC5B glycoforms in pregnancy mucus collected from the cervix during labor (41). Whether there are different MUC5B glycoforms in cervical secretions of normal cycling females through the cycle and the complete identification of terminal carbohydrate structures on the cervical MUC5B mucin remain to be elucidated. In summary, this study demonstrates cyclic changes in the amounts of T antigen and N-acetyllactosamine throughout the menstrual cycle. The increased levels of mucin oligosaccharides in cervical secretions correlate with previous data showing a peak of MUC5B mucin at midcycle and suggest an important role for carbohydrates in conferring specific physicochemical properties to the mucus gel throughout the menstrual cycle. We propose that increased levels of carbohydrates at the time of ovulation enhance mucin hydration by retaining more water and thereby facilitating sperm migration through the endocervical canal.
Acknowledgments
Drs. A. R. Gargiulo and J. A. Hill III (Brigham and Womens Hospital, Boston, MA) collected the human mucus samples. Received May 16, 2002. Accepted September 4, 2002. Address all correspondence and requests for reprints to: Ilene K. Gipson, Ph.D., Schepens Eye Research Institute, Harvard Medical School, 20 Staniford Street, Boston, Massachusetts 02114. E-mail: gipson@vision.eri.harvard.edu. This work was supported by NIH Grant R01-HD-33171 (to I.K.G.).

16. 17. 18.

19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34.

References
1. Hafez ESE 1980 Human reproduction: conception and contraception. Hagerstown, MD: Harper and Row; 221252 2. Morales P, Roco M, Vigil P 1993 Human cervical mucus: relationship between biochemical characteristics and ability to allow migration of spermatozoa. Hum Reprod 8:78 83 3. Wolf DP, Blasco L, Khan MA, Litt M 1977 Human cervical mucus. IIV. Fertil Steril 28:41169 4. Carlstedt I, Lindgren H, Sheehan JK, Ulmsten U, Wingerup L 1983 Isolation and characterization of human cervical-mucus glycoproteins. Biochem J 211: 1322 5. Fowler J, Vinall L, Swallow D 2001 Polymorphism of the human MUC genes. Front Biosci 6:D1207D1215 6. Yin BW, Lloyd KO 2001 Molecular cloning of the CA125 ovarian cancer antigen: identification as a new mucin, MUC16. J Biol Chem 276:2737127375 7. Gum Jr JR, Crawley SC, Hicks JW, Szymkowski DE, Kim YS 2002 MUC17, a novel membrane-tethered mucin. Biochem Biophys Res Commun 291: 466 475 8. Gipson IK, Ho SB, Spurr-Michaud SJ, Tisdale AS, Zhan Q, Torlakovic E, Pudney J, Anderson DJ, Toribara NW, Hill III JA 1997 Mucin genes expressed by human female reproductive tract epithelia. Biol Reprod 56:999 1011 9. Gipson IK 2001 Mucins of the human endocervix. Front Biosci 6:D1245D1255 10. Gipson IK, Spurr-Michaud S, Moccia R, Zhan Q, Toribara N, Ho SB, Gargiulo AR, Hill III JA 1999 MUC4 and MUC5B transcripts are the prevalent mucin messenger ribonucleic acids of the human endocervix. Biol Reprod 60:58 64 11. Strous GJ, Dekker J 1992 Mucin-type glycoproteins. Crit Rev Biochem Mol Biol 27:5792 12. Dekker J, Rossen JW, Buller HA, Einerhand AW 2002 The MUC family: an obituary. Trends Biochem Sci 27:126 131 13. Carlstedt I, Lindgren H, Sheehan JK 1983 The macromolecular structure of human cervical-mucus glycoproteins. Studies on fragments obtained after reduction of disulphide bridges and after subsequent trypsin digestion. Biochem J 213:427 435 14. Yurewicz EC, Matsuura F, Moghissi KS 1987 Structural studies of sialylated oligosaccharides of human midcycle cervical mucin. J Biol Chem 262:4733 4739 15. Yurewicz EC, Moghissi KS 1981 Purification of human midcycle cervical

35.

36. 37. 38. 39. 40. 41.

42.

mucin and characterization of its oligosaccharides with respect to size, composition, and microheterogeneity. J Biol Chem 256:1189511904 Yurewicz EC, Matsuura F, Moghissi KS 1982 Structural characterization of neutral oligosaccharides of human midcycle cervical mucin. J Biol Chem 257: 2314 2322 Van den Steen P, Rudd PM, Dwek RA, Opdenakker G 1998 Concepts and principles of O-linked glycosylation. Crit Rev Biochem Mol Biol 33:151208 Ujita M, Misra AK, McAuliffe J, Hindsgaul O, Fukuda M 2000 Poly-Nacetyllactosamine extension in N-glycans and core 2- and core 4-branched O-glycans is differentially controlled by i-extension enzyme and different members of the 1,4-galactosyltransferase gene family. J Biol Chem 275:15868 15875 Peumans WJ, Van Damme EJ 1998 Plant lectins: specific tools for the identification, isolation, and characterization of O-linked glycans. Crit Rev Biochem Mol Biol 33:209 258 Gipson IK, Moccia R, Spurr-Michaud S, Argueso P, Gargiulo AR, Hill III JA, Offner GD, Keutmann HT 2001 The amount of MUC5B mucin in cervical mucus peaks at midcycle. J Clin Endocrinol Metab 86:594 600 Gilks CB, Reid PE, Clement PB, Owen DA 1989 Histochemical changes in cervical mucus-secreting epithelium during the normal menstrual cycle. Fertil Steril 51:286 291 Moghissi KS, Syner FN 1976 Cyclic changes in the amount and sialic acid of cervical mucus. Int J Fertil 21:246 250 Van Kooij RJ, Roelofs HJ, Kathmann GA, Kramer MF 1980 Human cervical mucus and its mucous glycoprotein during the menstrual cycle. Fertil Steril 34:226 233 Wolf DP, Sokoloski JE, Litt M 1980 Composition and function of human cervical mucus. Biochim Biophys Acta 630:545558 Noyes RW, Hertig AT, Rock J 1950 Dating the endometrial biopsy. Fertil Steril 1:325 Watanabe H, Gipson IK 1994 Detection of blood group differences in human corneal epithelium using a monoclonal antibody and lectins. Arch Ophthalmol 112:667 673 Shanker S, Das RH 2001 Identification of a cDNA clone encoding for a galactose-binding lectin from peanut (Arachis hypogaea) seedling roots. Biochim Biophys Acta 1568:105110 Laemmli UK 1970 Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680 685 Towbin H, Staehelin T, Gordon J 1979 Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci USA 76:4350 4354 Iacobelli S, Garcea N, Angeloni C 1971 Biochemistry of cervical mucus: a comparative analysis of the secretion from preovulatory, postovulatory, and pregnancy periods. Fertil Steril 22:727734 Chantler E, Debruyne E 1977 Factors regulating the changes in cervical mucus in different hormonal states. Adv Exp Med Biol 89:131141 Kopito LE, Kosasky HJ, Sturgis SH, Lieberman BL, Shwachman H 1973 Water and electrolytes in human cervical mucus. Fertil Steril 24:499 506 Idris N, Carraway KL 1999 Sialomucin complex (Muc4) expression in the rat female reproductive tract. Biol Reprod 61:14311438 Arango ME, Li P, Komatsu M, Montes C, Carraway CA, Carraway KL 2001 Production and localization of Muc4/sialomucin complex and its receptor tyrosine kinase ErbB2 in the rat lacrimal gland. Invest Ophthalmol Vis Sci 42:2749 2756 Rossi EA, McNeer RR, Price-Schiavi SA, Van den Brande JM, Komatsu M, Thompson JF, Carraway CA, Fregien NL, Carraway KL 1996 Sialomucin complex, a heterodimeric glycoprotein complex. Expression as a soluble, secretable form in lactating mammary gland and colon. J Biol Chem 271: 33476 33485 Chantler EN, Scudder PR 1984 Terminal glycosylation in human cervical mucin. In: Nugent J, OConnor M, eds. Mucus and mucosa. London: Pitman; 180 195 Chilton BS, Kaplan HA, Lennarz WJ 1988 Estrogen regulation of the central enzymes involved in O- and N-linked glycoprotein assembly in the developing and the adult rabbit endocervix. Endocrinology 123:12371244 Paulesu L, Pessina GP 1982 Cyclic changes of sialidase in human cervical mucus. Int J Biochem 14:561563 Boland CR, Chen YF, Rinderle SJ, Resau JH, Luk GD, Lynch HT, Goldstein IJ 1991 Use of the lectin from Amaranthus caudatus as a histochemical probe of proliferating colonic epithelial cells. Cancer Res 51:657 665 Thomsson KA, Prakobphol A, Leffler H, Reddy MS, Levine MJ, Fisher SJ, Hansson GC 2002 The salivary mucin MG1 (MUC5B) carries a repertoire of unique oligosaccharides that is large and diverse. Glycobiology 12:114 Wickstrom C, Davies JR, Eriksen GV, Veerman EC, Carlstedt I 1998 MUC5B is a major gel-forming, oligomeric mucin from human salivary gland, respiratory tract and endocervix: identification of glycoforms and C-terminal cleavage. Biochem J 334:685 693 Thornton DJ, Khan N, Mehrotra R, Howard M, Veerman E, Packer NH, Sheehan JK 1999 Salivary mucin MG1 is comprised almost entirely of different glycosylated forms of the MUC5B gene product. Glycobiology 9:293302