INTRODUCTION

Top Abstract Introduction Procedures Results Discussion References

Glycoprotein hormone  subunit is common to the heterodimeric hormones chorionic gonadotropin, luteinizing hormone, follicle-stimulating hormone, and thyroid-stimulating hormone. However, in its free form (free  subunit), it is an important placental (1, 2) and pituitary (3) product, and it has been shown to have functions that are independent of the dimeric hormones (4-7). Glycosylation of free  differs from glycosylation of the combined form (8, 9). The combination of  and  for heterodimer formation takes place in the endoplasmic reticulum prior to processing of the immature glycans. In  subunits that have not combined with a  subunit, enzymes from the post-translational glycosylation machinery have access to substrate sites that are normally protected by the  subunit of the heterodimer. As a result, the free form of  subunit generally contains more elaborate oligosaccharide branching as well as higher amounts of core fucosylation than  subunit obtained from dissociated hormone (8, 9). These characteristic glycosylation patterns prevent secreted free  subunits from combining with  subunits that might be encountered extracellularly, thus ensuring a population of free  molecules (9, 10).

The structural diversity of complex-type N-linked glycans is initiated by GlcNAc branching of the trimannosyl core and continues with the action of different glycosyltransferases that further extend these antennae (11). Specifically, the activity of N-acetylglucosaminyltransferase IV initiates the 14/13-branch of complex glycans, whereas the action of N-acetylglucosaminyltransferase V initiates the 16/16-branch. In most human epithelial tissues, expression of 16/16-branching is low, whereas in malignancy, expression of this branch is increased, and the resulting oligosaccharides are considered to be significant markers of carcinoma (12, 13). However, a contrasting pattern seems to exist in normal and malignant pregnancies. A literature survey of pregnancy-related glycoproteins and oligosaccharides (Table I) (14-25) showed that in transformed placental tissues as well as in glycans isolated during the early part of pregnancy, a large amount of 14/13-branching is expressed, whereas 16/16-branching seems to be typical for glycoproteins obtained from the final stages of pregnancy.

It is our hypothesis that the expression of 14/13-branched glycans in early pregnancy reflects the implantation and placentation process during the invasion of trophoblast tissue. Consequently, glycosylation patterns of pregnancy-related glycoproteins should change concurrently with the decline of the invasiveness of trophoblast tissue during the early second trimester of normal pregnancy. This theory is supported by the different glycosylation patterns found on hCG¹ (8, 15, 26-28) and on free  subunit from normal pregnancy (8, 26), choriocarcinoma (22-24), and non-trophoblastic neoplasms (25, 29). In addition, less highly charged isoforms of hCG were found in late pregnancy (30), and further studies of its electrophoretic mobility suggested that its glycosylation patterns change during the early second trimester of pregnancy (31). Previously, we have presented lectin data that implied increased branching and higher incorporation of fucose into carbohydrate moieties in late pregnancy (32). In the present study, we have analyzed the glycan structures of free  from five individuals throughout their normal pregnancies to determine the exact nature of the glycosylation changes and to define the time in pregnancy at which they occur. Structural analysis of these glycans suggests which glycosyltransferases are involved and contributes to the understanding of normal and pathologic glycobiology in pregnancy.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results Discussion References

Materials-- Reference preparations of hCG (CR 125), hCG (CR 119), and hCG (CR 119) were provided by Drs. S. Birken and R. Canfield through the Center for Population Research. Oligosaccharide standards and -fucosidase were obtained from Oxford GlycoSystems Ltd. (Abingdon, United Kingdom). -Fucosidase was used to remove core fucose from glycans to create non-core-fucosylated standards. Tri(6/6)-F oligosaccharide was kindly donated by Dr. Harald S. Conradt (Gesellschaft für Biotechnologische Forschung, Braunschweig, Germany). Neuraminidase (Vibrio cholerae) was obtained from Calbiochem. BSA (Pentex fraction V) was purchased from Miles Inc. (Kankakee, IL). Sephadex G-100 (superfine) was obtained from Amersham Pharmacia Biotech. LcH-agarose and DSA gel were purchased from E-Y Laboratories, Inc. (San Mateo, CA). Centricon-10 and Microcon-10 concentration devices were obtained from Amicon, Inc. (Beverly, MA).

Immunoassays-- Intact hCG was assayed by RIA using a monoclonal antibody, A03C9 (Monoclonal Antibodies Inc., Sunnyvale, CA), with cross-reactivity for free  and  subunits estimated at 1.3 and 0.3%, respectively. Free  subunit was assayed by RIA using polyclonal antiserum SB6, with <0.6% cross-reactivity for hCG (33). Free  subunit was assayed by RIA using an -specific monoclonal antibody (BioMerica, Newport Beach, CA). The cross-reactivity of hCG with the free  monoclonal antibody was <0.1% (34, 35).

Isolation of Free -- Five healthy pregnant volunteers provided 24-h urine collections throughout their full-term uncomplicated pregnancies. A sample was removed from each 24-h urine collection and was assayed for hCG, free , and creatinine. Free  was isolated as described previously (32). Essentially, each urine specimen was precipitated with 2 volumes of acetone at pH 5.5 and 4 °C, followed by centrifugation. The precipitates were resuspended in distilled water and dialyzed against 50 mM ammonium acetate for 72 h at 4 °C, followed by centrifugation. The supernatants from each sample were lyophilized and redissolved in water. If not all material dissolved, then the pellets were redissolved, dialyzed, and centrifuged as described above. The second set of supernatants of each sample was combined with the first. Free  was isolated from each sample by gel filtration on a Sephadex G-100 superfine column (1.6 × 100 cm) run in 0.2 M ammonium acetate (pH 7.4) at 4 °C at a flow rate of 5 ml/h. Fractions of 2 ml were collected into tubes containing BSA (2 mg/tube) and assayed by RIA for free  subunit, free  subunit, and intact hCG, respectively. Fractions containing free  subunit were pooled and lyophilized.

Lectin Affinity Chromatography-- Samples from different time points throughout the second trimester of each individual pregnancy were subjected to affinity chromatography on LcH and DSA lectin columns. To ascertain that the columns were not initially overloaded with glycoproteins, materials from the unbound fractions were reapplied to fresh columns. Additional tests showed that no matrix effects from the eluent interfered with assays of free  in the various fractions.

Liberation of Desialylated N-Linked Glycans-- Immunopurified free  samples were desialylated with neuraminidase, desalted, and concentrated using Microcon-10 concentrators. Subsequently, the samples were denatured by heating at 100 °C for 3-4 min in 0.2 M sodium phosphate (pH 8.0) containing 1% SDS and 0.1 M -mercaptoethanol, followed by addition of EDTA (final concentration of 10 mM), Nonidet P-40 (final concentration of 5%), and recombinant glycerol-free peptide-N⁴-(N-acetyl--D-glucosaminyl)asparagine amidase F (25-100 units/mg of glycoprotein; Genzyme, Cambridge, MA). The mixture was incubated at 37 °C for 18 h; a second aliquot of enzyme was added, and the incubation was continued for another 20 h. The reaction was stopped by addition of an equal volume of 10% trichloroacetic acid, and the precipitate was eliminated by centrifugation. The pellets were washed with methanol (containing 5% water) to remove residual trichloroacetic acid and detergent and examined for complete release of the carbohydrate chains by acid hydrolysis with trifluoroacetic acid, followed by monosaccharide analysis by HPAEC-PAD (36). The liberated glycans in the supernatant were purified on a column (0.7 × 28 cm) of Bio-Gel P-4 (mesh 200-400) run in water at room temperature at a flow rate of 5 ml/h. Fractions of 0.5 ml were collected, and the released glycans were pooled and lyophilized. Elution positions of the glycans were established with oligosaccharide standards, detected by hexose assay using phenol-sulfuric acid reagents (37).

HPAEC-PAD of Released Glycans-- The glycans were resolved by HPAEC-PAD on a CarboPac PA-100 column (0.4 × 25 cm; Dionex Corp., Sunnyvale, CA) and detected with an electrochemical detector (ED40) set in the "carbohydrate" waveform, controlled by a personal computer using PeakNet chromatography software (38). The column was eluted with 250 mM NaOH with a 10% gradient of 0.5 M sodium acetate at a flow rate of 1 ml/min.

Periodate Oxidation and Reduction-- Periodate oxidation of the glycans (39, 40) was performed by incubation in 9 mM NaIO₄, buffered with 0.1 M sodium acetate at pH 5.5, for 3 days at 4 °C in the dark. The reaction was quenched with 3 µl of ethylene glycol and incubated overnight under the same conditions. The product was neutralized with 0.1 M NaOH, reduced by the direct addition of 5 mg of solid NaBD₄, and kept at room temperature for an additional 16 h. Excess reducing agent was destroyed by the addition of 5 µl of acetic acid, and the solutions were dried in a vacuum centrifuge. Borate was removed by repeated addition and drying with methanol. The samples were vacuum-desiccated overnight prior to methylation.

Methylation-- Each preparation (1-2 µg) was dissolved in 200 µl of a NaOH/Me₂SO suspension (41). After 1 h at room temperature, 50 µl of methyl iodide was added, and the suspensions were left for 1 h at room temperature with occasional vortexing (39). The methylated product was extracted by adding 1 ml of chloroform, and the suspensions were backwashed four times with 2-3 ml of 30% acetic acid. The chloroform layer was dried down and stored at 20 °C.

Electrospray Ionization Mass Spectrometry-- ESI-MS was performed on a TSQ 700 triple quadrupole mass spectrometer (Finnigan MAT, San Jose, CA) equipped with an electrospray ion source (Analytica Inc., Branford, CT). Samples were dissolved in methanol/water solutions (6:4, v/v) containing 0.25 mM NaOH and analyzed by syringe pump flow injection directly into the electrospray chamber through a stainless steel hypodermic needle at a flow rate of 0.85 µl/min. The voltage difference between the needle tip and the source electrode was 3.5 kV.

	RESULTS

Top Abstract Introduction Procedures Results Discussion References

Lectin Affinity Chromatography-- Free  subunit was purified from 24-h urine collections throughout the pregnancies of five healthy individuals. In all individuals, the total amount of free  production increased as pregnancy progressed. Typical recovery achieved was 95% of the initial free  immunoreactivity. In all procedures, loss of material was minimized by adding small amounts of BSA to the eluent. BSA also prevented loss of highly purified free  on membranes of the devices used for desalting.

View larger version (14K): [in this window] [in a new window]   Fig. 1.   LcH lectin affinity chromatography of early (upper panel) and late (lower panel) pregnancy free  subunits. Purified free  samples were subjected to affinity chromatography on LcH-agarose. Fractions of 2 ml were collected and assayed by RIA. The profiles shown represent material from the early (upper panel) and late (lower panel) second trimester of the pregnancy of one individual. The arrow indicates addition of buffer containing 0.1 M -methyl-D-mannoside (MM) as a competitive sugar.

View larger version (12K): [in this window] [in a new window]   Fig. 2.   DSA affinity chromatography of dissociated hCG subunit and of early and late pregnancy free  subunits. Immunopurified desialylated  samples were subjected to affinity chromatography on DSA gel. Unbound and retarded material was collected in fractions of 2.3 ml and assayed by RIA. The profiles shown represent analysis of dissociated hCG subunit and of free  preparations from the early and late second trimester of the pregnancy of one individual.

View larger version (13K): [in this window] [in a new window]   Fig. 3.   LcH and DSA lectin affinity chromatography of free  samples from five volunteers (A-E) throughout the second trimester of their pregnancies. Samples of purified free  subunit obtained throughout the second trimester of pregnancy from five healthy individuals were subjected to LcH and DSA lectin affinity chromatography to analyze the changes in their glycosylation. Shaded boxes represent pooled samples from the indicated time periods.

HPAEC-PAD of Released Glycans-- Glycans were liberated with peptide-N⁴-(N-acetyl--D-glucosaminyl)asparagine amidase F from immunopurified desialylated free  from volunteer A during the early (weeks 13-15 from LMP) and late (week 26 from LMP) second trimester of pregnancy (Fig. 3A). Using HPAEC-PAD, the glycans were resolved on a CarboPac PA-100 column (Fig. 4). Elution conditions were established by which base-line separation of the oligosaccharide standards was obtained (Fig. 4C). Analysis of early pregnancy free  glycans (Fig. 4A) showed the presence of Di-F (22%) and Di (31%), with retention times of 11.8 and 13.1 min, respectively, and of triantennary (13%, 14.5-15.5 min) and tetraantennary (12%, 17.5 min) glycans (Fig. 5). In addition, minor amounts of hybrid-type glycans (5%, 9.0 min) were detected. Late second trimester free  glycans (Fig. 4B) contained 9% less Di (22%), about equal amounts of Di-F (22%), and 7% more triantennary structures (20%) as compared with the early free  sample. Oligosaccharide standards suggested that the twin peaks at 14.7 and 15.2 min represent Tri(6/6) and Tri(4/3), respectively (data not shown), implying increased relative amounts of Tri(6/6) in the late free  sample. The relative amounts of tetraantennary (10%) and hybrid-type (3%) glycans were both ~2% lower in late free . In addition, an unidentified peak eluting at 16.0 min was more prominent (+3%) in the late pregnancy free  sample. Taken together, these data show increased branching, specifically by generation of Tri(6/6) in the late free  sample. However, definitive conclusions could not be made due to coelution of several specific glycans. Particularly, some fucosylated triantennary glycans were found to coelute with Di-F at 11.8 min. Therefore, further structural analysis of the glycans was performed.

View larger version (23K): [in this window] [in a new window]   Fig. 4.   HPAEC-PAD oligosaccharide profiles of early and late pregnancy free  subunits. Glycans were enzymatically released and isolated from purified free  subunit during the early (A) and late (B) second trimester of the pregnancy of volunteer A. The chromatogram in C shows the elution profile of oligosaccharide standards: peak 1, Di-F; peak 2, Di; peak 3, Di-bis-F (bis indicates the presence of a bisecting GlcNAc); peak 4, Tri(4/3); peak 5, Tetra (for structures, see Fig. 5). The separation was performed on a CarboPac PA-100 column eluted with 250 mM NaOH with a sodium acetate gradient as described under "Experimental Procedures." nC, nanocoulomb.

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Structures of the major glycans of free  subunit (desialylated) and their abbreviations as used in the text.

Electrospray Ionization Mass Spectrometry-- Enzymatically released and methylated glycans from immunopurified desialylated free  samples from volunteers A and B and from a hCG reference preparation were analyzed by ESI-MS (Table II). Profiles from early (weeks 13-15 from LMP) and late (week 26 from LMP) second trimester free  glycans from volunteer A are shown in Fig. 6. A unique feature of electrospray ionization is the generation of multiply charged ions (z) from a single molecular species. These provide, together with the detected ion mass to charge ratio (m/z), a direct indication of the relative molecular mass through the following relationship: relative molecular mass = z(m/z  23), where 23 is the mass of the adherent sodium cation. The percentages shown in Table II represent molar ratios of the individual glycans as a summation of all of their detected charge states (z). Branching of the glycans increased from early to late pregnancy, as is evidenced by a decrease in the relative amounts of Mono m/z 822 (2+) and Di m/z 1047 (2+)/705 (3+) glycans and a concurrent increase in the relative amounts of Tri m/z 1272 (2+)/855 (3+) and Tri-F m/z 1359 (2+)/913 (3+) glycans in the late free  samples (Fig. 6 and Table II). Furthermore, complete disappearance of all hybrid-type and monoantennary glycans was observed in late pregnancy free  from both subjects (Table II). Additionally, core fucosylation increased as pregnancy progressed, as evidenced by increased relative amounts of core-fucosylated glycans (Di-F, m/z 1134 (2+)/763 (3+) and Tri-F, m/z 1359 (2+)/ 913 (3+)), concurring with lower amounts of non-fucosylated glycans (e.g. Di, m/z 1047 (2+)/705 (3+)). The total increase in core-fucosylated glycans from early to late pregnancy was 57.6% in volunteer A and 25.2% in volunteer B. Nearly 100% of the glycans were core-fucosylated in both subjects at the end of pregnancy (Table II). The hCG reference preparation contained mainly hybrid-type (31.9%), monoantennary (42.5%), and some diantennary (24.5%) glycans (Table II); only a small amount of core fucosylation was detected (0.6%), and no tri- or tetraantennary glycans were found.

View this table: [in this window] [in a new window]   Table II Molar ratios of individual glycans of free  isolated at different time points of the pregnancies of two volunteers and from dissociated hCG The glycans were desialylated, liberated with peptide-N4-(N-acetyl--D-glucosaminyl)asparagine amidase F, methylated, and subjected to ESI-MS analysis as described under "Experimental Procedures." The designation of the two volunteers corresponds with that in Fig. 3. wk, week.

View larger version (27K): [in this window] [in a new window]   Fig. 6.   ESI-MS profiles of glycans obtained from free  subunit samples during the early and late second trimester of the pregnancy of one volunteer. The early pregnancy sample was from weeks 13-15 (from LMP), and the late sample was from week 26 of the pregnancy of volunteer A (Fig. 3A). ESI-MS was performed following desialylation, enzymatic release, and methylation of the glycans. The abbreviations are explained in Fig. 5. Tri comprises both Tri(4/3) and Tri(6/6) glycans.

View larger version (22K): [in this window] [in a new window]   Fig. 7.   ESI-MS profiles of ODM glycans from free  subunit samples from the early and late second trimester of the pregnancy of one volunteer. The samples were from the same volunteer and from the same period in pregnancy as described in the legend to Fig. 6. ESI-MS was performed following desialylation, enzymatic release, and ODM of the glycans. The area around m/z 1290, which is indicative for the ratio of the Tri(4/3) and Tri(6/6) isomers, is shown expanded in Fig. 8.

View larger version (20K): [in this window] [in a new window]   Fig. 8.   Expanded and overlaid segments of ESI-MS spectra of ODM glycans (from Fig. 7) from early and late pregnancy free  subunits. The expanded segments around m/z 1290 show the relative amounts of unoxidized Tri(4/3) and oxidized Tri(6/6) isomers of the fucosylated triantennary glycan after oxidation, reduction with NaBD4, and methylation. The carbon 12 isotope peaks at m/z 1290 from both spectra were scaled to 100%.

	DISCUSSION

Top Abstract Introduction Procedures Results Discussion References

Previously, we found evidence that glycosylation of free  from the third trimester of pregnancy is different from that from the first trimester (32). In this study, we have analyzed these changes in five individual pregnancies throughout the second trimester to determine the exact nature of the glycosylation changes and to identify when they occur. Using LcH and DSA lectin analysis, we observed that the binding properties of free  in all five individuals underwent marked changes beginning at around week 14 of pregnancy (Fig. 3).

Binding to LcH requires glycans with a trimannosyl core and fucosylation at the innermost GlcNAc residue (42). We observed increased binding to LcH (mean difference of 35.6 ± 6.0%) as pregnancy progressed, which was initially interpreted as reflecting an increase in core fucosylation. Interestingly, compositional analysis revealed that free  subunits from early pregnancy contained enough fucose to account for at least one core-fucosylated glycan per subunit, yet many of those molecules were unable to bind to LcH. Therefore, it was proposed that 14/13-branching, which prevents binding to LcH, might be more abundant in early pregnancy.

DSA interacts more strongly with 16/16-branched tri- and tetraantennary glycans than with 14/13-branched structures; thus, it is particularly useful in discerning differently branched triantennary glycans (43, 44). Dissociated hCG subunit passed through the DSA column without any interaction, confirming that this subunit, formerly combined with hCG, does not contain any tri- or tetraantennary glycans (8-10, 45). In contrast, some free  interacted with DSA, and the amount increased from a mean of 17.4 ± 4.5% in the early part to 51.3 ± 2.2% in the late part of the second trimester of pregnancy (mean difference of 33.9 ± 5.0%). The major changes in free  interaction with DSA took place during the same period in which the changes in LcH binding were observed (Fig. 3). Taken together, the lectin data suggest that during weeks 14-17, an increase in core fucosylation occurs, and there is a shift in the type of branching of the glycans of free , from the presence of 14/13-branched structures to higher amounts of 16/16-branched structures.

HPAEC-PAD analysis of released glycans from free  of volunteer A during the early and late second trimester of pregnancy indicated a decrease in diantennary and an increase in triantennary structures (Fig. 4). Standards indicated that this was essentially due to generation of more 16/16-branched glycans in the late free  sample. The ESI-MS data provided accurate molar ratios of individual N-linked glycans (Table II) and structurally confirmed the conclusions from the lectin affinity and HPAEC-PAD experiments. Nearly 100% of the glycans isolated from free  from the third trimester were core-fucosylated. In addition, hybrid and monoantennary glycans disappeared in late pregnancy. ESI-MS analysis of ODM glycans from early and late second trimester free  samples demonstrated that virtually all of the increase in branched structures was due to increased 16/16-branched triantennary glycans (Figs. 7 and 8 and Table II).

The ESI-MS analysis of volunteers A and B (Table II) correlated well with the DSA binding properties observed for these two individuals (Fig. 3, A and B). There was considerably more heterogeneity between individuals in early pregnancy as evidenced by both DSA binding and structural analysis. This may reflect normal variations between individual pregnancies or lack of precision in dating gestational age based on LMP. As the shift in glycosylation nears completion at the end of the second trimester, there is greater uniformity among pregnancies on the basis of both DSA binding and structural analysis by ESI-MS. It is important to note that when glycans are released from free  and examined individually, observed differences are likely to be smaller than those from experiments involving the intact glycoprotein. Changes in structure of only one of the two N-linked glycans can lead to altered affinity of the entire free  molecule for a specific lectin.

Previous compositional analysis has shown that urinary free , pooled during the second and third trimesters of pregnancy, is fully sialylated and that both glycosylation sites are occupied with intact glycans (8). Furthermore, 97.3% of the Gal residues of early second trimester free  were found to be sialylated. These data indicate that the glycans on free  isolated from pregnancy urine have not been partially degraded. In addition, sialic acid is expected to be predominantly present in 23-linkage on the N-acetyllactosamine antennae of free  since human placenta contains almost exclusively the Gal14GlcNAc-R 23-sialyltransferase variant of the possible sialyltransferases that specifically elongate these branches (46).

The observation that the glycosylation changes occur in all five pregnancies within a narrow window of gestational time suggests that the activities of the enzymes involved are developmentally regulated. Specifically, the activities of N-acetylglucosaminyltransferases IV and V and of 6-fucosyltransferase seem to be affected during weeks 14-17 of pregnancy. Our data, together with previously published structural analyses of pregnancy-related glycoproteins (Table I), imply that during this period in time, there is a large increase in N-acetylglucosaminyltransferase V activity and perhaps a corresponding decrease in N-acetylglucosaminyltransferase IV activity. The activity of 6-fucosyltransferase appears to increase during the same period and stays at a high level throughout the early third trimester of pregnancy, as illustrated by the almost complete core fucosylation of free  glycans in late pregnancy (Table II). Subsequently, 6-fucosyltransferase activity may decline, as aging trophoblast tissue is associated with a decrease in core fucosylation of complex glycans (47).

During pregnancy, free  subunits are secreted by cytotrophoblasts, in which little or no  subunit is expressed, and by syncytiotrophoblasts, in which free  and free  subunits as well as hCG are produced (48). In early pregnancy, part of the cytotrophoblast population becomes invasive, penetrating into the endometrium and eventually into the superficial layers of the myometrium and uterine blood vessels (49). These invasive cells behave much like tumor cells and likely give rise to pathologic conditions such as choriocarcinoma when appropriate regulatory factors are not recognized. The type of triantennary branching observed on early pregnancy free  was similar to glycan structures found on hCG associated with invasive mole and choriocarcinoma, reflecting increased N-acetylglucosaminyltransferase IV activity (Table I) (50). Furthermore, the changes that we observed in the glycosylation of free  occur during the time frame that coincides with the decline in cytotrophoblast invasiveness as normal pregnancy progresses. Thus, changes in the state of trophoblast differentiation appear to be associated with alterations in glycosyltransferase activity.

Changes in glycosylation may have functional significance for free  receptor binding, signal transduction, or circulatory clearance, as has been observed for the heterodimeric glycoprotein hormones (51). Alternatively, the changes in glycan structures on free  may not be directly involved in  function, but rather, could be coincidental to its synthesis within cells in which the general glycosylation machinery has differentiated, reflecting altered functional status of the cell. Increased branching and core fucosylation as well as changes in the relative amounts of triantennary isomers may reflect the change from an invasive state to a more nurturing role for the placenta during the second trimester of pregnancy.

In conclusion, glycosylation of free  changes dramatically during the early part of the second trimester of pregnancy. Similar changes occurred in all five pregnancies examined, suggesting that there is developmental regulation of the placental glycosylation machinery during normal pregnancy. Our findings have important implications for the proposed use of specific forms of glycosylation as markers for cancer in pregnancy (50). Since activities of the same enzymes appear to be altered in certain stages of gestational development and in malignant transformation of the trophoblast, the relative amounts of these glycan markers in normal pregnancy will be determined by gestational age.