Morphology and glycoprotein synthesis of uterine glandular epithelium in human basal plate at termAn ultrastructural and autoradiographic study.код для вставкиСкачать
THE ANATOMICAL RECORD 216:146-153 (1986) Morphology and Glycoprotein Synthesis of Uterine Glandular Epithelium in Human Basal Plate at Term: An Ultrastructural and Autoradiographic Study D. MICHAEL NELSON, JUDITH ORTMAN-NABI, AND EDWARD M. CURRAN Departments of Obstetrics and Gynecology, Washington University School of Medicine, St. Louis, MO 63110, D. M. N. , E.M.C.);Department of Obstetrics and Gynecology, School of Medicine, The Ohio State University, Columbus, OH 43210 (J.0.-N.) ABSTRACT This study describes the morphologic features of uterine glandular epithelium in human basal plate at term and identifies this epithelium as an active site of glycoprotein synthesis. Wedge biopsies were obtained from the basal plate at the time of repeat cesarean section from 11 normal pregnant patients at term. Biopsy specimens were either processed immediately for microscopic examination or incubated in vitro with 25pCi/cc of 3H-galactose or 3H-leucine. Tissues incubated with tritiated compounds (2-hour pulse f %hour chase in nonradioactive medium) were either processed for light microscopic autoradiographic analysis or extracted for determination of trichloroacetic-acid-precipitable tritiated macromolecules in tissues and medium, Profiles of cuboidal-columnarglandular epithelium were identified in the decidual component of the basal plate region adjacent to spiral arterioles and perpendicular to the inner layers of myometrial muscle. Autoradiographic and biochemical studies identified the glandular epithelium, as well as large decidual cells, to be major sites of incorporation of both 3H-galactose and 3H-leucine and to be prime sources for secretion of tritiated macromolecules that appeared in the medium during in vitro incubation of basal plate tissue. Ultrastructurally, the glandular epithelium was noted to rest on a basal lamina, t o have lateral cell membranes with numerous desmosomes, and to exhibit an apical surface with microvilli projecting into a luminal space. Cytologic features of the cells included abundant profiles of rough endoplasmic reticulum, multiple mitochondria with lamellar cristae, a well-developed perinuclear but nonpolarized Golgi apparatus, and nuclei containing predominantly euchromatin. Lipid droplets and glycogen deposits were present in some cells. This study indicates that uterine glands persist throughout human gestation in the basal plate and that these glands continue to be active in glycoprotein synthesis and secretion. The endometrium of the human uterus undergoes sequential histological changes in response to changes in ovarian steriod production during the menstrual cycle (Noyes et al., 1950). These changes involve both the stroma and the endometrial glandular epithelium and provide an endometrium primed appropriately to allow establishment of implantation should a fertilized egg appear in the endometrial cavity. During implantation, trophoblast penetrates the uterine epithelium and aggressively invades the stroma and blood vessels in the area in the process of establishing the site of what will ultimately represent the basal plate attachment of the chorionic placenta. Trophoblast has been described to be involved in phagocytosis of cells composing the glands during the course of its invasion into the endometrium (Boyd and Hamilton, 1970). The secretory products of the glandular epithelum may provide a source of nutrition for the blastocyst prior to the establishment of a 0 1986 ALAN R.LISS. INC. maternal blood supply during the implantation process. Other roles for the secretions of the uterine glandular epithelium during early pregnancy have been suggested by recent studies (Pijnenborg et al., 1980;Dobashi et al., 1984; Bulmer et al., 1984) but the presence and role of uterine glands in later stages of pregnancy are unclear. During the course of a study of blood vessels in wedge biopsy specimens of human basal plate of normal and diabetic patients, we identified the presence of multiple profiles of glandular epithelium. The current study describes the histological and ultrastructural appearance Received July 15, 1985; accepted April 16, 1986. Address reprinted requests to Dr. Michael Nelson, M.D., Ph.D., Department of Obstetrics & Gynecology, Washington University School of Medicine, The Jewish Hospital of St. Louis, 216 S.Kingshighway, St. Louis, MO 63110. 147 UTERINE GLANDS IN HUMAN BASAL PLATE Figure 1.Hemostasis was achieved at the biopsy site by use of a single 2-0 chromic suture ligature. The specimen was labeled with india ink to identify the area closest to the myometrium in situ, and then the specimen was cut with a razor blade into 1-2-mm cubes in anticipation of subsequent fixation or incubation. - Chorioollanloic Cleavage PloneDecidua- Gland- Placenta 3- Spiral Artery %Basal Artery - Rodiol @- Artery -Arcuate Artery Arcuate Vein Fig. 1. Diagrammatic representation of the area where the basal plate biopsy specimens were obtained at the time of cesarean section. After creating a cleavage plane, the chorioallantoic placenta was removed and a wedge biopsy of the uterine wall was obtained near the center of the placenta’s previous attachment site. Incubation Tissue pieces from six patients’ biopsies were distributed to 2cc conical plastic tubes containing 0.5 cc of medium 199, with or without leucine, with Earle’s salts equilibrated with an atmosphere of 95% air-5% COZ, and containing either 25pCVcc of 3H-galactose or 3Hleucine. The alcohol used as a solvent for the stock 3Hgalactose was evaporated with a stream of nitrogen prior to adding the medium to the 25pCi of 3H-galactose. The 0.01 N HCL was used as a solvent for the stock 3Hleucine was neutralized with an equivalent volume of 0.01 N NaOH prior to the preparation of the 3H-leucine incubation medium. After 2 hours’ incubation at room temperature, media containing radioactive tracers were removed, tissue pieces were rinsed with three changes of fresh medium 199, and tissues were either frozen or placed in fixative or incubated an additional 3 hours in 0.5 cc of gassed medium 199 prior to collection of media and fixation or freezing of tissues. Duplicate incubations were used for each tritiated compound for each of the “pulse” and “pulse-chase” incubation. of uterine glandular epithelium present in the basal plate attachment site of the chorioallantoic placenta of normal patients undergoing repeat cesarean section a t term. Autoradiographic and biochemical analyses of the incorporation of 3H-leucine and 3H-galactose by basal plate tissue incubated in vitro were undertaken to determine if the glandular epithelium continued to function Fixation, Embedding, and Microscopic Examination in glycoprotein synthesis and secretion. These studies identify the ‘persistence of uterine glands even at term Tissue blocks were fixed in 3% glutaraldehyde in 0.1 in the basal plate and show that they continue to be M pH 7.4 cacodylate buffer for 1-2 hours at room temactive in glycoprotein synthesis. perature, rinsed twice in the same buffer at 4”C, postfixed in 2% osmium tetroxide in 0.1 M pH 7.4 buffer for MATERIALS AND METHODS 1 hour at 4”C, dehydrated in a graded series of cold Materials ethanols, and embedded in Spurr’s resin. Sections 1Glutaraldehyde, osmium, lead citrate, uranyl acetate, 2pm in thickness were cut with an MT-2 Sorvall Ultraand Spurr’s resin were purchased from Electron Micros- microtome, stained with toluidine blue, and examined copy Sciences, Fort Washington, PA. Medium 199 with and photographed with an Olympus BHT brightfield Earle’s salts and medium 199 with Earle’s salts without microscope. Thin sections with silver-gold interference leucine were purchased from Grand Island Biological colors were stained 7-10 minutes in a saturated solution Go., Grand Island, NY. Protosol tissue di estant, Aqua- of uranyl acetate in 50% ethanol, rinsed with 50% sol scintillation cocktail, leucine, L-[4,5,8H(N)] with a ethanol, and stained an additional 7-10 minutes with specific activity of 59.2 Cilmmole, and galactose, D-[1- lead citrate prior to examination in a Philips 201 elec3H(N)] with a specific activity of 11.5 Cdmmole, were tron microscope. purchased from New England Nuclear, Boston, MA. Autoradiography NTB-2 emulsion, D-19 developer, and Ektaflo fixer were purchased from Eastman Kodak Co., Rochester, NY. All Four to six sections approximately 1 pm thick were other chemicals were of reagent grade and were pur- taken from at least three tissue blocks previously exchased from standard suppliers. posed to a tritiated precursor of glycoprotein synthesis for one of the designated incubation times. These secBasal Plate Specimens tions were heat-annealed to clean glass slides and coated After informed consent was obtained, wedge biopsies with Kodak NTB-2 emulsion as previously described of the uterine lining at the basal plate were obtained (Nelson et al., 1978).Emulsion was previously tested for from 11patient volunteers undergoing elective cesarean background and sensitivity prior to its use for coating section at 38-41 weeks’ gestation. These patients were slides. Coated slides were allowed to dry slowly in a normotensive throughout pregnancy, were not smokers, high-humidity box prior to storage at 4°C in sealed delivered infants with birth weights that were average Bakelite slide boxes containing dessicant. After 4-28 for gestational age, and had no known medical diseases days, slides were developed for 2 minutes at 16°C in Dcomplicating their pregnancies. After infant delivery 19 developer, rinsed 1minute in distilled water, fixed 6 and manual removal of the placenta, an elliptical wedge minutes in 25% Ektaflo, and rinsed in two changes of of tissue approximately 2 cm in diameter and 1 cm in distilled water, 10 minutes each. The gelatin containing depth was obtained from an area of the basal plate near developed silver grains was then stabilized by exposure the center of the previous attachment site of the cho- to a 5% aqueous solution of glutaraldehyde which then rioallantoic placenta as diagrammatically illustrated in allowed staining of tissue sections with warm toluidine 148 D.M. NELSON, J. ORTMAN-NABI, AND E.M. CURRAN Fig. 2.Photomicrograph of a profile of uterine glandular epithelium Fig. 3.Photomicrograph of glandular epithelium directly adjacent t o at the decidual-myometrial interface of the basal plate. The glandular a spiral arteriole which shows the physiologic changes of pregnancy epithelium consists of a single layer of cuboidal-columnar cells ar- described by Brosens et al. (1967). x 1300. ranged as a tubular structure perpendicular to the underlying rnyometrial muscle. x 1300. UTERINE GLANDS IN HUMAN BASAL PLATE blue without emulsion damage. Nonspecific blue staining of the gelatin that occurred with this process was cleared by rinsing slides with a 1%solution of acetic acid. Developed and stained autoradiograms were examined and photographed by using a n Olympus BHT brightfield microscope. Under oil immersion a t 1,OOOX magnification, autoradiograms were used to count grains overlying grandular epithelial cells and decidual cells in each of three slides prepared from three different blocks from each of the six specimens incubated with either 3H-galactose or 3H-leucine. For uniformity of grain analysis between pulse and pulse-chase specimens, only cells readily identified as decidual cells and only glandular epithelial cells cut in cross section perpendicular to their basement membrane were included in the analysis. Cells tangentially sectioned or whose origin was unclear by light microscopic examination were specifically excluded from the grain analysis. Grains per cell for the glandular epithelial cells and decidual cells in the pulse specimens were compared with the labeling of corresponding cells in the chase specimens by the paired t test. Tissue and Media Analyses Frozen tissue samples previously incubated with tritiated galactose or leucine were lyophilized, weighed, and homogenized in Ten Broeck homogenizers containing 2 cc of ice-cold 20% trichloroacetic acid (TCA). Cold 10% TCA was then used (5 cc) and TCA-precipitable components were quantitatively transferred from the homogenizer tubes to 25-mm Millipore MF filter discs (Millipore Corporation, Bedford, MA) placed over the Fig. 4. Low-magnification photomicrograph of a basal plate biopsy specimen. Lumina of coiled arteries are indicated by asterisks. A profile of glandular epithelium lies adjacent to the vessels. Nearby is an arteriole which has maintained its elastic lamina and circumferential coat of smooth muscle cells (arrow). x 320. 149 ports of a Millipore sampling manifold (Millipore Corporation, Bedford, MA). After applying negative pressure to the filtration apparatus to adsorb the TCAprecipitable components to the filter discs, two additional 5-cc washes with 10% TCA and one 5-cc wash with 95% ethanol were placed sequentially over the filter discs and aspirated by application of negative pressure to the filtration apparatus. The filter discs were then transferred to glass scintillation vials containing 1 cc Protosol. After overnight Protosol digestions (at 60°C) of the precipitates from the filter discs, 15-cc of Aquasol scintillation cocktail was added to each vial and samples were counted for 10 minutes each in a Nuclear Chicago scintillation counter. After subtracting background counts and taking into account the efficiency of counting, results were expressed as c p d m g dry weight tissue. This value then allowed calculation of the micromoles of tritiated compound incorporated per milligram dry weight tissue, thus taking into account the differences in specific activity of the two tritiated precursors used in the study. Frozen medium samples were brought to room temperature, and a 100-p1 aliquot of medium was adsorbed to a filter disc placed on the Millipore filtration apparatus. Each disc was then washed twice with cold 20% TCA, once with cold 95%ethanol, digested with Protosol, and counted as above. To control for nonspecific adsorption of 3H-galactose or 3H-leucine with this analysis for TCAprecipitable macromolecules in medium, media not previously incubated with tissue were handled in a manner similar to that described above. This nonspecific adsorption control had less than 150 cpm per disc for each of the two tritiated labels, and this amount was subtracted TABLE 1. Quantitative grain analysis of glandular and decidual cells Cell type (n) Galactose Grainskell' (n) Leucine Grains/cell' Figs. 5-7. Photomicrographs of glandular epithelial cells (Figs. 5 , 6) and decidual cells (arrow; Fig. 7) in autoradiographs of basal plate tissue incubated for a 2-hour pulse in medium containing either 3Hleucine (3H-Leu P; Fig. 5 ) or 3H-galactose ?H-Gal P; Figs 6, 7). Qualitatively, autoradiographs prepared by using either tritiated compound showed labeling of the same cell types as described in the Results section; 2-week exposure. x 3200. Gland' Pulse Chase Decidua13 Pulse Chase Figs. 8,9.Photomicrograph of glandular epithelium (Fig. 8 ) and decidual cells (Fig. 9)from autoradiographs of tissue incubated for a 2hour pulse in 'H-galactosecontaining medium and an additional 3hour chase in nonradioactive medium (3H-Gal C). Silver grains overlying the glandular epithelium in such specimens were present in fewer numbers than in specimens incubated for the 2-hour pulse only. In contrast, decidual cells showed no difference in their labeling pattern, with some cells showing no grains or a few grains (Fig. 9, large arrow) while others had multiple grains (Fig. 9, small arrow); 2-week exposure. x 3,200. 'Grains per cell is reported as mean k standard deviation with n indicating number of cells fulfilling the criteria for analysis, as described in Materials and Methods. 2By paired t test, gland cells from chase specimens had significantly fewer (P = .001) grains per cell than pulse specimens for each tritiated compound. 3By paired t test, grains per cell for decidual stromal cells from chase specimens were not significantly different from pulse specimens for either tritiated compound (P = .35 for 3H-leucine; P = .15 for 3Hgalactose). (116) (155) 8.6 k 4.0 4.3 f 2.7 (179) (140) 15.4 f 8.7 5.2 f 3.7 (120) (161) 12.9 f 5.7 14.7 f 13.7 (135) (134) 13.9 f 12.1 15.1 f 10.3 TABLE 2. Trichloroacetic-acid-DreciDitabletritiated macromolecules Galactose Sample Tissue after pulse3 Tissue after chase3 Medium after chase3 cpml 25,316 f 15,918 11,223 f 5,790 8,057 f 5,576 Leucine PM2 2.98 & 1.87 1.32 & 0.68 0.95 -t 0.66 cpml 36,523 t 22,278 24,221 f 10,904 15,927 k 12,332 PM2 0.84 t 0.51 0.55 f 0.25 0.36 0.28 'Expressed as counts per minute per milligram dry weight (tissue) or 0.5 cc (medium). 'Expressed as micromoles of tritiated compound in trichloracetic-acid-precipitablemacromolecules per milligram dry weight (tissue) or 0.5 cc (medium). 3Results expressed as mean k standard deviation. 150 D.M. NELSON, J. ORTMAN-NABI, AND E.M. CURRAN from the incubation medium cpms prior to calculating the cpms in the total 0.5-cc sample of medium used for the chase incubation. The total micromoles of tritiated compound present as TCA-precipitable macromolecules in the medium at the end of the incubation was calculated by using the specific activities of 3H-leucine and 3H-galactose. RESULTS Light Microscopic Morphology Profiles of glandular epithelium were located in the decidua of the biopsy specimens; often they were observed to be adjacent to myometrial muscle cells (Fig. 2). The epithelium outlined a tubular gland structure composed of a single layer of cuboidal-to-columnarcells surrounding a luminal space. Most of the cells had a basophilic cytoplasm. A single euchromatic nucleus was centrally located. The profiles of glands were often adjacent to lumina of vessels exhibiting a thin endothelial lining with stromal cells embedded in a noncellular meshwork that composed the vessel wall (Fig. 3). Other adjacent vessels were observed to retain their internal elastic lamina and circumferential coat of smooth muscle cells (Fig. 4). Lumina of both glands and vessels were usually free of contents following the fixation and embedding procedures. Tritiated Leucine and Galactose Incorporation for the 2-hour pulse period only, with the appearance of TCA-precipitable tritiated macromolecules in the medium (Table 2). Ultrastructural Morphology The cytologic features of the epithelial cells composing the glandular profiles are illustrated in Figures 10 and 11. All epithelial cells rested on a basal lamina. Their luminal surface was composed of multiple microvilli. Few pinocytotic vesicles were apparent in the apical cytoplasm. Numerous profiles of rough endoplasmic reticulum without dilated cisternae and mitochondria with lamellar cristae were distributed throughout the cytoplasm. The ground plasm was less electron dense in some cells (Fig. 10) than others (Fig. 111,which is possibly related to fixation of ground plasm components. A well-developed Golgi apparatus was present in the perinuclear region of the cell but no polarity was noted in its distribution relative to the cells microvillous and basal surfaces. Glycogen deposits and multivesicular bodies were present in the cytoplasm of some cells. Lipid droplets were identified in the basal cytoplasm of numerous cells but secretory granules were not a characteristic feature of the cells. The euchromatic nucleus was centrally located within the cell. Close approximation of the apical areas of adjacent cells was typical of the epithelium with desmosomes frequently identifiable in this region (Fig. 12) as well as in other regions of the cells' lateral surface membranes (Fig. 11). The most apically located surface membranes of adjacent cells were noted to come into very close approximation (Fig. 12). However, fusion of these membranes to form the ultrastructrural appearance identifiable as a zonulae occludens was not observed. Autoradiographic analysis of tissue blocks previously incubated for 2 hours with either 3H-leucine (Fig. 5) or 3H-galactose (Fig. 6) revealed silver grains associated glandular epithelial cells (Figs. 5, 6; Table 11, some but not all decidual cells (Fig. 7, Table l),and a few trophoblast cells. Labeline of the latter two comDonents showed great variability from cell to cell in a given specimen. Grains overlying lumina of glands were occasionally DISCUSSION observed (Fig. 6) but no specific labeling of the connective tissue surrounding and enmeshing the decidual cells The basal Plate is the area where the most intimate interaction of maternal and fetal tissues occurs during was observed. By grain analysis, both glandular epithelial cells and gestation. Early in Placental development, troPhoblast decidual cells showed incorporation of both 3H-gala&,ose invades into the endometrium and subsequently conand 3 ~ - l(Table ~ ~ 1). Quantitatively, ~ i ~ ~ analysis of the tributes to alterations Of coiled arteries that Will provide tissue for incorporated tritiated precursor showed 3H- nutrition to the fetus throughout gestation. Marked alleucine incorporation was less than 3H-galactose incor- terations of the spiral arteriole morphology are required poration when the different specific activities of the two to accommodate the increased blood flow necessary to compounds were taken into account (Table 2). provide adequate perfusion of the chorioallantoic plaAutoradiographic analysis of tissue previously incu- cents- These changes have been the major focus Of a bated 2 hours in 3H-leucine or 3H-galactosefollowed by number of previous studies of the histology and CfiOlOgy a 3 how incubation in medium without tritiated corn- of human basal plate, as recently reviewed (Pijnenborg pounds showed the number of silver grains overlying et ale, 1981)-The C I J I T ~ ~study ~ describing the Presence glandular epithelial cells to be much less for both tri- and moQholoD Of glandular epithelium in term basal tated precursors of glycoprotein synthesis (Fig. 8, Table 1). Labeling of decidual cells after the longer incubation showed the same variability from cell to cell and the Same average number of grains Per cell as observed to Figs. I O , I I . Electron micrographs illustrating typical cytologic feabe present in cells in the tissue incubated for only 2 tures of the glandular epithelial cells as described in the Results hours. Grains were not concentrated over the lumina of section. Glycogen deposits were absent from many cells fe.g., Fig. 10) but were present in others (e.g., Fig. 11).The large arrows in Figure 11 the glandular In addition, there was no de- indicate desmosomes, which were a characteristic feature ofthe cells' monstrable labeling Of the extracellular matrix, includ- lateral surface membranes. RER, rough endoplasmic reticulum; W B , ing fibrinoid deposits, adjacent to glandular epithelium multivesicular body. Figure 10 x 9700; Figure 11 x 20,200. or labeled decidual cells (Fig. 9). Fig. 12. Electron micrograph of the most apical portion of two adja. Of tissue and medium from the branes cent glandular cells' lateral surface membranes. The surface mempulse-chase incubations &owed a drop in the ~adioactivof adjacent cells come into close approximation, as indicated by ity present in this tissue compared to tissue incubated the arrow, and a desmosome connects the two cells. x 77,800. u I ~~ UTERINE GLANDS IN HUMAN BASAL PLATE 151 152 D.M. NELSON, J. ORTMAN-NABI, AND E.M. CURRAN plate specimens developed as a separate project during the course of a similarly focused study of spiral arteriole morphology, which compared normal and diabetic patients. Wedge biopsy specimens from term basal plate were frequently noted to contain profiles of epitheliallined spaces, with a well-developed apical microvillous surface and not encased with a wall of fibrinoid such as vessels show when they have undergone “physiologic changes” in pregnancy (Brosens et al., 1967).This observation raised the question as to the fate of uterine glands at the placental attachment site during gestation and the realization that little was known about their fate in human gestation. A systematic study of glandular profile morphology and glycoprotein synthetic activity was thus undertaken. Boyd and Hamilton (1970)described the uterine glands at the site of trophoblast invasion to be ingested or digested during the early weeks of placental development. From their extensive series of uterine specimens from various stages of pregnancy, they identified glandular activity to persist through the third month of gestation but no description of glandular presence beyond this point in gestation was provided. Pijnenborg et al. (1980) described only isolated fragments of glands present in specimens obtained at 15-18 weeks’ gestation, and these were predominantly in the marginal areas of the placental attachment site. They noted that areas where glandular tissue was lost had trophoblast cells present throughout the whole thickness of the decidua, but where glands were present, trophoblastic invasion was limited to the compact and spongy layers of the decidua overlying these glands. An inference from their study might be that glandular activity contributes to limiting trophoblast invasion into the basal plate. Reference to uterine glands in the basal plate in later gestation is limited. Uterine glands at the placental attachment site of a single specimen from a 32-week gestation, described by Harris and Ramsey (19661, were noted to be compressed. A passing reference to uterine glands was made by Wells et al. (1984) in noting the factor VIII and TA1 antigen negativity of residual glands in their biopsy specimens from term basal plate studied with immuno fluorescent techniques. The current study has identified the persistence of glandular profiles in the decidua of the basal plate at term in the area directly adjacent to the inner layers of myometrial muscle fibers. The cytology of the cells composing the gland suggested they continued to have the capacity for active glycoprotein synthesis, with a well-developed Golgi complex and abundant rough endoplasmic reticulum. Autoradiographic and biochemical analyses of incorportaion of tritiated precursors for glycoprotein synthesis demonstrated that these cells were active in glycoprotein synthesis and were a prime source for the tritiated macromolecules that appeared in the medium during the course of in vitro incubation of basal plate tissue. The large decidual cells were also active in incorporating both tritiated compounds, but we found no evidence to suggest they were actively secreting tritiated macromolecules. However, little is known about the functions of these large decidual cells, which clearly expressed synthetic activity, and future studies to characterize their function are contemplated. The specific functions of the glands and their glycoprotein synthetic products are left to speculation at this time since the 3H-galactose and ‘H-leucine tracers used in this study are nonspecific precursors for glycoproteins. However, immunocytochemical studies of uterine glands from basal plate specimens obtained in early pregnancy have demonstrated a progestagen-associated protein (Mazurkiewicz et al., 1981) and Pregnancy-associated Plasma Protein-A (PAPP-A)(Dobashi et al., 1984) in the cytoplasm of endometrial gland cells. In addition, Bulmer et al. (1984) have described first-trimester endometrial glands that stain with two monoclonal antibodies that react exclusively with fetal trophoblastic tissues of the human placenta and extra-embryonic membranes. They suggested that these antibodies recognize substances similar to pregnancy-associated proteins which could either be produced by glandular tissue and actively taken up by trophoblast or be produced by trophoblast and absorbed by glandular epithelium. The former possibility was dismissed by the authors as unlikely on the assumption that no glands were present in the placental bed at term, yet trophoblast from term placenta continued to show reactivity to the monoclonal antibodies (Johnson et al., 1981; Johnson and Molloy, 1983).The current study’s identification of glands active in glycoprotein synthesis at term again raises the possibility that they could have a role in secreting the antigen recognized by the two monoclonal antibodies. The use of 3H-galactoseas one of the tracers for glycoprotein synthesis leaves open the possibility that conversion to glucose through epimerization and incorporation into glycogen could contribute to a part of the observed cellular incorporation. However, this route is unlikely to have contributed significantly to the observed incorporation of 3H-galactose since all cells of the glands showed incorporation of the compound while glycogen deposits were not present in all cells. In addition, the concentration of glucose in the incubation medium was greater than a million times the concentration of 3Hgalactose; thus making any 3H-glucoseformed through epimerization occurred at a very low concentration compared to the nonradioactive glucose present. A final suggested role for the glandular epithelium in the basal plate is that of providing a source of epithelium to proliferate after delivery of the placenta, allowing reestablishment of a uterine epithelial lining in the denuded surface at the previous attachment site of the chorionic placenta. Uterine glandular epithelial proliferation has been described as contributing in a similar manner to the reestablishment of the uterine lining following menstruation (Ferenczy, 1976a;b). In summary, uterine glandular epithelium persists in the basal plate throughout human gestation and continues to be active in glycoprotein synthesis. Though the specific secretory products are yet to be determined, several potential products are suggested from immunocytochemical studies of uterine glands in early pregnancy. Other potential roles for the uterine glands, such as one in limiting trophoblast invasiveness in pregnancy or in contributing to the process of reepithelialization of the uterine cavity after delivery, are interesting possibilities though clearly speculative at this time. ACKNOWLEDGMENTS This investigation was supported in part by Department of Health, Education, and Welfare biomedical research support grant t2S07RR05389-22 given to the UTERINE GLANDS IN HUMAN BASAL PLATE Washington University School of Medicine (D.M.N.). We wish to express our sincere thanks to Dr. Elizabeth M. Ramsey for her helpful discussions and review of material during the course of this study. LITERATURE CITED Boyd, J.D., and W.J. Hamilton (1970) The Human Placenta. W. Heffer and Sons, Ltd., Cambridge, p. 53. Brosens, I., W.B Robertson, and H.G. Dixon (1967) The physiological response of the vessels of the placental bed to normal pregnancy. J. Pathol. Bacteriol., 93:569-579. Bulmer, J.N., W.D. Billington, and P.M. Johnson (1984) Immunohistologic identification of trophoblast populations in early human pregnancy with the use of monoclonal antibodies. Am. J. Obstet. Gynecol., 148:19-26. Dobashi, K., K. Ajika, T. Ohkawa, H. Okano, S. Okinaga, and K. Arai(1984) Immunohistochemical localization of pregnancy-associated plasma protein A (PAPP-A)in placentae from normal and preeclamptic pregnancies. Placenta 5:205-212. Ferenczy, A. (1976a) Studies on the cytodynamics of human endometrial regeneration. I. Scanning electronmicroscopy. Am. J. Obstet. Gynecol., 124:64-74. Ferenczy, A. (197613) Studies on the cytodynamics of human endometrial regeneration. 11. Transmission electron microscopy and histochemistry. Am. J. Obstet. Gynecol., 124t582-595. 153 Harris, J.W.S., and E.M. Ramsey (1966) The morphology of human uteroplacental vasculatue. Contrib. Embryol., 38t43-58. Johnson, P.M., H.M. Cheng, C.M. Molloy, C.M.M. Stern, and M.B. Slade (1981) Human trophoblast-specific surface antigens identified using monoclonal antibodies. Am. J. Reprod. Immunol., 1t246254. Johnson, P.M., and C.M. Molloy (1983) Localization in human term placental bed and amniochorion of cells bearing trophoblast antigens identified by monoclonal antibodies. Am. J. Reprod. Immunol., 4t33-40. Mazurkiewicz, J.E., J.F. Bank, and S.G. Joshi (1981)Immunocytochemical localization of a progestagen associated endometrial protein in the human decidua. J. Clin. Endocrinol. Metab., 521006-1008. Nelson, D.M., A.C. Enders, and B.F. King (1978) Cytological events involved in protein synthesis in cellular and syncytial trophoblast of human placenta. An electron microscope autoradiographic study of 3H-leucineincorporation. J. Cell Biol., 76r400-417. Noyes, R.W., A.T. Hertig, and J. Rock (1950) Dating the endometrial biopsy. Fertil. Steril., 1:3-25. Pijnenborg, R., G. Dixon, W.B. Robertson, and I. Brosens (1980) Trophoblastic invasion of human decidua from 8 to 18 weeks of pregnancy. Placenta, 1:3-19. Pijnenborg, R., W.B. Robertson, I. Brosens, and G. Dixon (1981)Review article: Trophoblast invasion and the establishment of haemochorial placentation in man and laboratory animals. Placenta, 271-92. Wells, M., B.-L. Hsi, C.-J.G. Yeh, and W.P. Faulk (1984) Spiral (uteroplacental) arteries of the human placental bed show the presence of amniotic basement membrane antigens. Am. J. Obstet. Gynecol., 15Ot973-977.