Differential expression of ezrin/radixin/moesin (ERM) and ERM-associated adhesion molecules in the blastocyst and uterus suggests their functions during implantation

Development of the blastocyst to implantation competency, differentiation of the uterus to the receptive state, and a cross talk between the implantation-competent blastocyst and the uterine luminal epithelium are all essential to the process of implantation. In the present investigation, we examined the possibility for a potential cross talk between the blastocyst and uterus involving the ezrin/radixin/moesin (ERM) proteins and ERM-associated cytoskeletal cross-linker proteins CD43, CD44, ICAM-1, and ICAM-2. In normal Day 4 blastocysts and after rendering dormant blastocysts to implantation-competent by estrogen in vivo (activated), the outer surface of mural trophectoderm cells showed much higher levels of radixin as compared to those in the polar trophectoderm cells, inner cell mass (ICM), and primitive endoderm. In contrast, ezrin was present on both the mural and the polar trophectoderm cell surfaces of normal Day 4 and activated blastocysts at higher intensity than dormant blastocysts. A distinct localization was noted in the primitive endoderm of dormant blastocysts that was not apparent in activated or normal Day 4 blastocysts. The expression of moesin was modestly higher at the mural trophectoderm of implantation-competent blastocysts, while the localization appeared to be present primarily on the polar trophectoderm cell surface of Day 4 blastocysts. The localization of ERM-associated adhesion molecules CD43, CD44, and ICAM-2 was more intense in the implantation-competent blastocysts compared with the dormant blastocysts. However, while CD44 was present both in the trophectoderm and in ICM, CD43 and ICAM-2 were localized primarily to the trophectoderm. The signal for ICAM-1 was very intense in the ICM but was modest in the trophectoderm. No significant changes in fluorescence intensity were noted between activated and dormant blastocysts. In the receptive uterus on Day 4 of pregnancy, ERM proteins were localized to the uterine epithelium, while on Day 5 the localization, especially of radixin and moesin, extended to the stroma surrounding the implantation chamber. With respect to ERM-associated adhesion molecules, while CD44 and ICAM-1 were exclusively localized in the stroma on Day 4, CD43 and ICAM-2 were localized to the epithelium. On Day 5, the localization of CD44 and ICAM-1 became highly concentrated in the antimesometrial stroma of the implantation chamber. The localization of CD43 and ICAM-2 remained mostly epithelial, although some stromal localization of CD43 was noted on Day 5. These results suggest that differential expression and distribution of ERM proteins and ERM-associated adhesion molecules are involved in the construction of the cellular architecture necessary for blastocyst activation and uterine receptivity leading to successful implantation.     

Comparison of two cell-adhesion molecules, uvomorulin and cell-CAM 105

Two cell adhesion molecules, cell-CAM 105 and uvomorulin (UM), were compared by analysing their antigenic structures, their activity in cell aggregation assays and their expression in various tissues. Cell-CAM 105 is a membrane glycoprotein which mediates the intercellular adhesion of reaggregating rat hepatocytes, and UM was first described to be involved in the compaction of preimplantation mouse embryos and embryonal carcinoma cells. UM is not only expressed during embryonic development but also in various adult tissues including liver, epithelia of lung, gut, kidney and uterus. A similar distribution for UM was found in rat tissues on cell types where cell-CAM 105 is known to be present. Our studies show that (i) cell-CAM 105 and UM are distinct and different proteins; (ii) uvomorulin is involved in the compaction of rat preimplantation embryos but Fab anti-UM has no effect on reaggregating rat hepatocytes, where Fab anti-cell CAM is effective; (iii) distribution studies show that UM is expressed on a broader range of epithelial cells while cell-CAM 105 is more restricted to hepatocytes and simple epithelia. In cases where both cell adhesion molecules are expressed on the same cell types they can be localized to different parts of the cell surface.     

Expression Pattern of E- and P-Cadherin in Mouse Embryos and Uteri during the Periimplantation Period

Periimplantation mouse embryos and uterine tissues were examined by means of immunohistochemistry for their expression of the Ca²⁺ dependent cell-cell adhesion molecules, E- and P-cadherin. E-cadherin was detected in all embryonic cells during periimplantation stages, and also detected in the uterine epithelium. When blastocysts attached to the uterine epithelium, E-cadherin was detected at implantation sites between the mural trophectoderm and the uterine epithelium on 5 day of pregnancy. P-cadherin was first detected in the mural trophectoderm on 4.5-day blastocysts, and then detected in the ectoplacental cone, giant cells and visceral endoderm from 5.5 day.
P-cadherin was also detected in the maternal uterine decidual cells from 5.5 day. After degeneration of uterine epithelial cells, giant cells make direct contact with uterine decidual cells, and P-cadherin was detected at contact sites between these cells.
Thus, the complicated process of implantation seems to be supported by temporal and spatial expression of the multiple classes of cadherins.     

sLeX/L-selectin mediates adhesion in vitro implantation model

The complex implantation process is initiated by the recognition and adhesion between the embryo and uterine endometrial epithelium. The expression and interactions between the adhesive molecules from both fetal and maternal sides are crucial for the successful implantation. In this study, we aimed to investigate the expression and adhesive function of sLeX on the trophoblasts and L-selectin on uterine epithelial cells mediated the adhesion at the fetal-maternal interface, and to further explore whether this adhesion system could induce endometrial apoptosis, using in vitro implantation model consisting of the human trophoblast cell line (JAR) and human uterine epithelial cell line (RL95-2). The results showed that sLeX was expressed on JAR cells by indirect immunofluorescence staining. After transfection of JAR cells with fucosyltransferase VII (FUT7) which is the key enzyme for sLeX synthesis, the expression of FUT7 and sLeX synthesis were increased, and the percent adhesion of trophoblast cells to RL95-2 cell monolayer was significantly increased (P?相似文献   

Electron microscopic studies of chimeric blastocysts experimentally produced by aggregating blastomeres of rat and mouse embryos

Xenogeneic chimeras between rat and mouse were produced by aggregating embryos at the 8- to 12-cell stages. Of 114 combinations we have made so far, 26 chimeric embryos developed into blastocysts. The origin of each cell in the composite embryos can be identified unambiguously by the ultrastructural appearance of the cytoplasmic inclusions. Both the rat- and the mouse-derived cells differentiated equally well into either ICM or trophoblast cells. However, the mouse-derived cells gave rise to ICM cells more frequently than the rat-derived cells. Furthermore, when the ratderived cells formed trophoblast cells, they were predominantly mural trophoblast cells, while the mouse-derived cells differentiated predominantly into the polar trophoblast cells. Cells of the same species tend to remain as a group in the chimeric blastocysts.     

Trophectoderm cell subpopulations in the periimplantation mouse blastocyst

Trophectoderm of the preimplantation mouse blastocyst is composed of two cell subpopulations relative to their proximity to the inner cell mass. The polar trophectoderm overlying the inner cell mass proliferates to form the ectoplacental cone, and the mural trophectoderm endoreplicates and gives rise to giant cells. We examined specific differences in the two trophectoderm cell populations using a lectin (Dolichos biflorus) to detect cell surface characteristics and a simple sugar (D-Gal) to detect differences in incorporation. During the first day of delayed implantation, the mural trophectoderm presented twice as many lectin binding sites as did the polar trophectoderm. The mural trophectoderm of both nondelaying and delayed implantation blastocysts showed a greater rate of incorporation of the tritiated sugar by presenting more reduced silver grains in radioautograms. These results indicate that the mural trophectoderm and polar trophectoderm are two distinct cell types in the periimplantation blastocyst.     

Dynamic expression of the cell adhesion molecule cell-CAM 105 in fetal and regenerating rat liver

Cell-CAM 105 is an integral cell surface glycoprotein that is involved in cell-cell adhesion of adult rat hepatocytes in vitro. In the present report we used a radio-immunoassay, a quantitative immunoblotting technique and immunofluorescence microscopy to investigate the expression of cell-CAM 105 in fetal and regenerating rat liver. In the fetal liver cell-CAM 105 did not appear until day 16 of the gestation, when it increased rapidly to reach the level found in adult liver, 3 weeks after birth. In liver regenerating after partial hepatectomy a transient decrease in the amount of cell-CAM 105 was observed in the plasma membranes of the hepatocytes. A significant decrease was observed as early as 12 h after partial hepatectomy, reaching a minimum by 3 days after the operation, corresponding to approx. 35% of the amount of cell-CAM 105 in normal liver. The amount then increased slowly and was back to the normal level by about 15 days after partial hepatectomy. The results indicate that cell-CAM 105 exerts its major function in terminally differentiated cells. An excellent correlation was seen between the kinetics of the expression of cell-CAM 105 and of reported changes of both enzymatic and organizational patterns of hepatocytes in regenerating and fetal liver. This suggests that cell-CAM 105 could be important for the development and maintenance of the cell-cell binding and organizational pattern characteristic of terminally differentiated hepatocytes.     

Blastocyst implantation in the Chinese hamster (Cricetulus griseus)

Embryonic development of the Chinese hamster (Cricetulus griseus) was studied from the onset of implantation to the formation of the parietal yolk sac placenta. Implantation began on day 6 of pregnancy, when the embryo became fixed to the uterine luminal epithelium. At this time there was no zona pellucida, and microvilli of the trophoblast and uterine epithelium were closely apposed. Stromal cells immediately adjacent to the implantation chamber began to enlarge and accumulate glycogen. By day 7 the mural trophoblast penetrated the luminal epithelium in discrete area. The trophoblast appeared to phagocytize uterine epithelial cells, although epithelium adjoining the points of penetration was normal. In other areas nascent apical protrusions from the uterine epithelium indented the surface of the trophoblast. The epiblast had enlarged and both visceral and parietal endoderm cells were present. The well-developed decidual cells were epithelioid and completely surrounded the implantation chamber. On day 8 the uterine epithelium had disappeared along the mural surface of the embryo. The embryonic cell mass was elongated and filled the yolk sac cavity. Reichert's membrane was well developed. The uterine epithelial basal lamina was largely disrupted, and the trophoblast was in direct contact with decidual cells. Primary and secondary giant trophoblast cells were present and in contact with extravasated maternal blood. The mural trophoblast formed channels in which blood cells were found in close proximity to Reichert's membrane. Decidual cells were in contact with capillary epithelium and in some cases formed part of the vessel wall. Structural changes occurring in the embryo and endometrium during implantation in the Chinese hamster are described for the first time in this report and are compared to those described for some other myomorph rodents.     

Expression of cellCAM-105 in the apical surface of rat uterine epithelium is controlled by ovarian steroid hormones

Affinity-purified antibodies to cellCAM-105, an adhesive cell surface glycoprotein, were used in immunohistochemical investigations of rat uteri at various functional stages: (i) the oestrous, pro-oestrous, metoestrous, and dioestrous stages of the oestrous cycle, (ii) Days 1-8 of normal pregnancy, (iii) delayed implantation, (iv) 18 h after oestrogen reactivation from delay of implantation, and (v) juvenile rats, and normal ovariectomized adults, respectively, before and after experimental injection of progesterone and/or oestrogen. CellCAM-105 was present in the apical zones of the luminal and glandular epithelium cells in a stage-specific and hormone-dependent manner. The results indicate that: (1) steroid hormones are essential for the expression of cellCAM-105 in the uterine epithelial cells; (2) progesterone induces cellCAM-105 expression in the glandular epithelium, and oestrogen induces cellCAM-105 expression in the luminal epithelium; (3) progesterone induces down-regulation of cellCAM-105 from the surface of the uterine luminal epithelium of juvenile rats; (4) cellCAM-105 is absent in the luminal epithelial cells but present in the glandular epithelial cells of the rat uterus at the time of blastocyst implantation.     

Carbohydrate changes in pre- and peri-implantation mouse embryos as detected by a monoclonal antibody

We have examined the tissue and embryonic distribution of an antigen on a large polysaccharide that is recognized by a monoclonal antibody, IIC3, prepared against F9 teratocarcinoma cells. By immunofluorescence the antigen is first detected on compacted morulae and early blastocysts. It is strongly expressed on the primary endoderm and trophoblast of expanded blastocysts, but then disappears from the trophoblast of attached blastocysts in vitro. The binding of the antibody is completely inhibited by D-galactose and N-acetylgalactosamine. Fluoresceinated lectins were used to study further the changes in cell surface carbohydrates on trophoblast during implantation. Ricinus I, specific for terminal galactose, binds to preimplantation stages but does not bind to the trophoblast of the attached blastocyst. On the other hand, wheat germ agglutinin, specific for N-acetylglucosamine and sialic acid, binds to all preimplantation embryos and also to attached blastocysts (embryo proper and trophoblast). Neuraminidase treatment of blastocyst outgrowths enhances binding of both IIC3 and Ricinus I to the trophoblast; conversely, the binding of wheat germ agglutinin is decreased by this treatment. The results obtained in this study show changes of cell surface carbohydrates during early mouse development and suggest that sialic acid may be masking molecules on the surface of the trophoblast at the time of implantation.     

Alpha5beta1, alphaVbeta3 and the platelet-associated integrin alphaIIbbeta3 coordinately regulate adhesion and migration of differentiating mouse trophoblast cells

During blastocyst implantation, interaction between integrins on the apical surface of the trophoblast and extracellular matrix (ECM) in the endometrium anchors the embryo to the uterine wall. Strong adhesion of the blastocyst to fibronectin (FN) requires integrin signaling initiated by exogenous fibronectin. However, it is not known how integrin signaling enhances blastocyst adhesion. We present new evidence that the integrin, alphaIIbbeta3, plays a key role in trophoblast adhesion to fibronectin during mouse peri-implantation development. Trafficking of alphaIIb to the apical surface of the trophoblast increased dramatically after blastocysts were exposed to fibronectin, whereas other fibronectin-binding integrins, alpha5beta1 and alphaVbeta3, were resident at the apical surface before ligand exposure. Functional comparisons among the three integrins revealed that ligation of alpha5beta1 most efficiently strengthened blastocyst fibronectin-binding activity, while subsequent trophoblast cell migration was dependent primarily on the beta3-class integrins. In vivo, alphaIIb was highly expressed by invasive trophoblast cells in the ectoplacental cone and trophoblast giant cells of the parietal yolk sac. These data demonstrate that trafficking of alphaIIb regulates adhesion between trophoblast cells and fibronectin as invasion of the endometrium commences.     

Mouse blastocyst surface expression of galactose-containing epitopes coinciding with trophoblast differentiation

A galactose-containing cell surface epitope of mouse blastocysts was identified and partially characterized by means of immuno- and lectincytochemistry, using a mouse IgM anti-blastocyst monoclonal antibody (mAb N63) and four different galactose-binding lectins (BSL-1, DBA, PNA and SBA) as molecular probes. The mAb was produced by syngeneic intrasplenic immunization with adhesive mouse blastocysts, obtained 18 h after estrogen reactivation from facultative delay of implantation. Labelling of different mouse embryonic stages collected by uterine flushings revealed that the labelling of the epitope by monoclonal antibodies was restricted to the blastocyst stage. A peak labelling intensity was observed on late blastocysts. When examining blastocyst outgrowths, both trophoblast and embryoblast were weakly stained by mAb N63. Direct antigen characterization performed on blastocysts indicated that the mAb N63 recognized a galactose-containing glycolipid antigen. Immunochemistry of cryosectioned, unfixed mouse tissues including ovary, testis, uterus in delay and at implantation, Day 12 and term placenta, liver, kidney, brain, intestine, heart, striated muscle, and skin was negative. In addition, labelling of rat and hamster blastocysts was negative. In vitro experiments demonstrated that the galactose-containing blastocyst surface epitope was not involved in blastocyst attachment to plastic culture dishes. The appearance of the epitope at the embryonic surface in vivo coincides with the time of trophoblast differentiation and implantation in the mouse.     

Implantation in vitro: co-culture of rat blastocyst and epithelial cell vesicles

Implantation of blastocysts involves conversion of maternal and embryonic cell surfaces from a nonadhesive to an adhesive state in response to the internally driven developmental program or to externally generated factors. However, the intricacies of the cellular and subcellular changes that promote the attachment are not known, because these changes are difficult to determine in situ because of the nonaccessibility of the site. To overcome this, an in vitro model of implantation was developed by co-culturing rat blastocysts and uterine epithelial cells of the same gestational age (day 5 postcoitum; plug day as day 1) in drops hanging from the lid of a Petri dish. The system was used to study the changes on the surface membranes of the cells of the trophectoderm and uterine epithelium and to evaluate the antiadhesive activity of the newly designed test substances. The isolated epithelial cell vesicles were co-cultured with zona-free blastocysts in the microdrops (40–50 µl) hanging from the lid of a 60-mm Petri dish. The lid was placed over the lower dish, which was presaturated with the medium. The culture was examined 48 h later to determine the site of adhesion of epithelial cell vesicles with the trophoblasts lining the blastocyst. The cell-cell adhesion was monitored on a computerized image analyzer. To validate the adhesion of blastocysts and epithelial cell vesicles in co-culture, the expression of a cell adhesion molecule, uvomorulin, was studied using immunocytochemical technique after incubating with antiuvomorulin antibody. Intense staining was noted on the membrane surfaces at the site of attachment of the blastocyst and cell vesicles.The authors express their sincere thanks to the Ministry of Health and Family Welfare, Government of India, for their financial support     

Scanning electron microscopy of the surface of normal and implantation-delayed mouse blastocysts during development in vitro

Mouse blastocysts undergo developmental steps in culture analogous to those occurring during implantation in utero. We examined cultured blastocysts by scanning electron microscopy (SEM) as they passed through these stages. From the time of hatching to the acquisition of adhesiveness, most blastocysts were exhanded, with flattened cells possessing relatively small numbers of microvilli, centrally raised areas (presumably reflecting the location of the nuclei) and intercellular ridges often possessing microvilli. At, or shortly before, the trophoblast outgrowth stage, blastocysts appeared to contract; the cells bulged noticeably, microvilli covered the entire surface of most cells and intercellular ridges were no longer observable. Blastocysts removed from uteri on the seventh day of ovariectomy delay possessed a variety of morphologies and shapes. The blastocoel was frequently collapsed and cell outlines were difficult to discern. These blastocysts were initially adhesive in vitro, but subsequently disengaged from the substratum before becoming permanently adherent several hours later. During the initial phase of adhesiveness, blastocysts were elongated and had prominent intercellular ridges, particularly in the equatorial region. Detached blastocysts contained bulging cells with contours which obscured the intercellular ridges. Surface ultrastructure during subsequent phases resembled non-delayed blastocysts during attachment and outgrowth. On the basis of our studies, we propose that intercellular ridges play some role in blastocyst adhesiveness. However, we must conclude that there are other factors involved in the acquisition of adhesiveness by the blastocyst which are at least equally important but of a nature too subtle to be identified by our SEM analyses. Insofar as delayed blastocysts are concerned, we find that, within limits, the surface alterations that take place when blastocysts are activated in culture mirror those observed following reversal of delay in vivo by administration of hormones. Since delayed blastocysts placed in saline also undergo morphological changes resembling those seen at the onset of activation in utero, we suggest that reversal of implantation delay requires initially neither direct contact with steroid or macromolecular inducers nor an exogenous supply of metabolites.     

Apoptosis as the mode of uterine epithelial cell death during embryo implantation in mice and rats

An ultrastructural study of mouse and rat embryo implantation sites was undertaken to determine whether the uterine luminal epithelial cells surrounding the blastocyst exhibited the morphologic characteristics of apoptotic or necrotic cell death. In both species the epithelial cells exhibited all of the characteristics of apoptosis, including surface blebbing, shrinkage and fragmentation of the cells, condensation of chromatin, and indentation and fragmentation of nuclei. Cytoplasmic organelles remained morphologically intact, and the cytoplasm maintained normal or increased staining density. Also, the epithelial cells and cell fragments were phagocytosed by the adjacent trophoblast cells. The epithelial cells did not exhibit the characteristics of necrotic cell death, such as swollen cells and mitochondria, damaged surface membranes, and disintegrated cytoplasmic organelles. We conclude that uterine epithelial cells surrounding mouse and rat embryos during implantation undergo apoptotic cell death leading to their phagocytosis by trophoblast cells.     

Adhesion molecules in gamete transport,fertilization, early embryonic development,and implantation—role in establishing a pregnancy in cattle: A review

Cell–cell adhesion molecules have critically important roles in the early events of reproduction including gamete transport, sperm–oocyte interaction, embryonic development, and implantation. Major adhesion molecules involved in reproduction include cadherins, integrins, and disintegrin and metalloprotease domain‐containing (ADAM) proteins. ADAMs on the surface of sperm adhere to integrins on the oocyte in the initial stages of sperm–oocyte interaction and fusion. Cadherins act in early embryos to organize the inner cell mass and trophectoderm. The trophoblast and uterine endometrial epithelium variously express cadherins, integrins, trophinin, and selectin, which achieve apposition and attachment between the elongating conceptus and uterine epithelium before implantation. An overview of the major cell–cell adhesion molecules is presented and this is followed by examples of how adhesion molecules help shape early reproductive events. The argument is made that a deeper understanding of adhesion molecules and reproduction will inform new strategies that improve embryo survival and increase the efficiency of natural mating and assisted breeding in cattle.     

Loss of polar trophoblast during differentiation of the blastocyst of the horse

Twelve blastocysts, collected 7-12 days after ovulation (Day 0), were examined by light and electron microscopy to investigate the nature of the relationship of the polar trophoblast (Rauber's layer) to the inner cell mass. On Day 7, the polar trophoblast was intact and formed a flattened layer overlying the epiblast cells of the inner cell mass. As blastocysts enlarged to greater than 1 mm in diameter, small discontinuities appeared in the polar trophoblast, where epiblast cells intruded onto the surface. At this time, trophoblast cells adhered closely to adjacent and underlying epiblast cells, forming an irregular layer of cells capping the epiblast. With continued increase in blastocyst size, polar trophoblast cells became isolated but maintained their characteristic apical endocytic structures. By Days 10-12, the scattered trophoblast cells showed evidence of deterioration, and vacuoles containing cell debris were common within the epiblast. It is suggested that polar trophoblast cells become scattered, rather than withdrawing as a unit, because they become more adherent to subjacent epiblast cells than to adjacent trophoblast cells. It is further suggested that most of the isolated cells are eventually phagocytosed by epiblast cells.     

Interference of lead with implantation in the mouse: a study of the surface ultrastructure of blastocysts and endometrium

Implantation chambers, trophoblast and uterine luminal surfaces were examined on days 5 and 6 of pregnancy by electron microscopy in mice with implantation failure due to an intravenous injection of 75 ppm of lead chloride on day 4. Attachment of the trophoblast cells to the surface of the endometrium and closure of the uterine lumina had failed to occur. Uterine epithelial cells in implantation chambers and along the lumina were covered with abundant microvilli. This appearance is similar to that seen in mice in experimental delay of implantation before the oestrogen-induced attachment of the blastocyst has occurred. It may therefore be assumed that lead has in some way interfered with the activity of ovarian steroid hormones on the endometrium. No significant changes were observed in surface ultrastructure of the blastocysts from the lead-treated and control groups.     

Cell-surface location and molecular properties of cell-CAM 105 in intestinal epithelial cells

Cell-CAM 105 is involved in intercellular adhesion of isolated rat hepatocytes in vitro. In addition to liver, cell-CAM 105 occurs in several different epithelia, in platelets, and in granulocytes. In this paper we present a detailed analysis of cell-CAM 105 in the small intestinal mucosa. Isolated rat intestinal epithelial cells and brush-borders were investigated by indirect immunofluorescence. A strong fluorescence occurred in the brush-border region and a much weaker staining was seen in the lateral cell surfaces. The brush-border staining was heterogeneous and concentrated to the periphery where brush-border microvilli from adjacent cells are in contact with each other. It is suggested that cell-CAM 105 might mediate binding between the outer surfaces of neighboring microvilli. Immunoblotting and electrophoretic analyses of the biochemical properties of intestinal cell-CAM 105 demonstrated significant differences compared with cell-CAM 105 isolated from liver. Intestinal cell-CAM 105 was smaller than liver cell-CAM 105 under reducing conditions, whereas it was larger than liver cell-CAM 105 under non-reducing conditions. Chemical reduction decreased the size of intestinal cell-CAM 105, but increased the size of liver cell-CAM 105. Our interpretation of these data is that intestinal cell-CAM 105 occurs as a part of a macromolecular complex. This interpretation was supported by electrophoretic analyses of intestinal cell-CAM 105 isolated by immunoaffinity chromatography on anti-cell-CAM antibodies. In addition to cell-CAM 105, this material contained several other proteins of lower molecular weight than cell-CAM 105. These data suggest that intestinal cell-CAM 105 participates in cell-surface interactions that may regulate the structure and function of the apical brush-border regions of the intestinal epithelial cells.     

Two different cell adhesion molecules--cell-CAM 105 and a calcium-dependent protein--occur on the surface of rat hepatocytes

We have recently identified a 105000 D plasma membrane glycoprotein, denoted cell-CAM 105 (CAM, cell adhesion molecule), that is involved in intercellular adhesion of reaggregating rat hepatocytes (Ocklind &; Öbrink, J biol chem 257 (1982) 6788 [11]). In this communication we identify another cell surface protein that is also involved in hepatocyte cell-cell adhesion. This protein has an apparent molecular weight (MW) of 70000 and can be released from the surface membrane by chelation of calcium with EGTA. Results are presented indicating that it is identical with a previously discovered protein, CDP-1 (CDP, calcium-dependent protein) (Öbrink, Lindström &; Svennung, FEBS lett 70 (1976) 28 [28]). Antisera produced against either cell-CAM 105 or CDP-1 inhibit hepatocyte aggregation, but not attachment to collagen. Cell-CAM 105 and CDP-1 are present on the cell surface as separate components, as judged by the fact that both EGTA treatment and trypsin treatment of hepatocytes selectively make the cells insensitive to blocking of aggregation by antibodies against CDP-1 but not by antibodies against cell-CAM 105. However, although much less efficiently, the antibodies against CDP-1 can recognize a 105000 D protein which is also bound by the antibodies against cell-CAM 105, and under certain conditions the antibodies against cell-CAM 105 seem to recognize a 70000 D protein. CDP-1 may thus be derived from cell-CAM 105, or the two proteins might have a common precursor.