Expression of Trophinin, Tastin, and Bystin by Trophoblast and Endometrial Cells in Human Placenta¹

Abstract

Trophinin, tastin, and bystin comprise a complex mediating a unique homophilic cell adhesion between trophoblast and endometrial epithelial cells at their respective apical cell surfaces. In this study, we prepared mouse monoclonal antibodies specific to each of these molecules. The expression of these molecules in the human placenta was examined immunohistochemically using the antibodies. In placenta from the 6th week of pregnancy, trophinin and bystin were found in the cytoplasm of the syncytiotrophoblast in the chorionic villi, and in endometrial decidual cells at the utero placental interface. Tastin was exclusively present on the apical side of the syncytiotrophoblast. Tissue sections were also examined by in situ hybridization using RNA probes specific to each of these molecules. This analysis showed that trophoblast and endometrial epithelial cells at the utero placental interface express trophinin, tastin, and bystin. In wk 10 placenta, trophinin and bystin were found in the intravillous cytotrophoblast, while tastin was not found in the villi. After wk 10, levels of all three proteins decreased and then disappeared from placental villi.

Introduction

Implantation of the developing human conceptus in the maternal uterus occurs by the end of the first week of pregnancy. Extra embryonic trophoblast of the blastocyst attaches to the endometrial epithelium at the embryonic pole [1, 2]. Upon attachment, the trophoblast proliferates rapidly and invades the uterine epithelium and underlying endometrial stroma. In the first trimester, the embryo attaches to the maternal decidua by intermediate (extravillous) trophoblast that streams off from the anchoring villi [3, 4]. The intermediate trophoblast invades maternal endometrial spiral arteries and dilates them in order to achieve a sufficient fetal blood supply. This process peaks during the 12th week of pregnancy and declines rapidly thereafter. The placental villi are organized into two layers: an inner layer of mononuclear cytotrophoblast and an outer layer of multinuclear syncytiotrophoblast. The villi cover the vascular mesoderm, which contains fetal capillaries and mesenchymal cells.

Previously we have shown that trophinin, tastin, and bystin comprise a complex that mediates a unique adhesion between trophoblast and endometrial epithelial cells at their respective apical cell membranes [5, 6]. Trophinin is an intrinsic membrane protein with its amino terminal region localized in the cytoplasm. More than 90% of the trophinin peptide consists of decapeptide repeats. The repeats are predicted to form repeated β-turns thought to be the structural basis for the homophilic binding of this protein. Tastin and bystin are soluble cytoplasmic proteins and are required for trophinin to exhibit cell adhesion activity.

Trophinin, tastin, and bystin are implicated in embryo implantation in vivo for the following reasons. Trophinin is strongly expressed in the trophectoderm of the monkey blastocyst [5]. In human endometrium, strong expression of trophinin is seen in a restricted region of the apical plasma membranes of the surface epithelium at the early secretory phase [5]. In mouse uterus, trophinin expression peaks on Days 4–5 of pregnancy, and both blastocyst and endometrial surface epithelium express trophinin at the time of implantation (unpublished data). The embryos of trophinin-null mutant mice, in which the trophinin gene was inactivated by homologous recombination, do not survive (unpublished results).

In this study, we prepared monoclonal antibodies specific to human trophinin, tastin, and bystin, and used these antibodies to examine the expression patterns of these molecules in the human placenta. This paper describes the results revealed by immunohistochemistry and by in situ hybridization.

Materials and Methods

Monoclonal Antibodies

Monoclonal anti-trophinin antibody (clone 3-11, mouse IgM) was prepared by immunizing mice with a synthetic peptide SIVGFSGGP (prepared by Biosynthesis, Louisville, TX) conjugated to keyhole limpet hemocyanin (Imject carbo-link; Pierce, Rockford, IL). Spleen cells from immunized mice were fused to P3X mouse myeloma cells. Hybridoma clones were selected by ELISA for antibodies reactive to human trophoblastic teratocarcinoma (HT-H) cells [7]. The hybridoma clones were further verified by immunofluorescence microscopy of COS cells transfected with trophinin cDNA and by Western blot analysis using GST-553, which is a GST (glutathione-S-transferase) fused to a cell surface domain of trophinin.

Monoclonal anti-bystin antibody (clone 19, mouse IgM) was prepared by immunizing mice with a synthetic peptide CGFRTEKREL conjugated to keyhole limpet hemocyanin (Imject sulfo-link; Pierce). Spleen cells from immunized mice were fused to P3X cells, and hybridoma clones were selected and purified in the same manner as described above for anti-trophinin antibody.

Monoclonal anti-tastin antibody (clone 38, mouse IgG3) was prepared by immunizing mice with a synthetic peptide CDQENQDPRR conjugated to keyhole limpet hemocyanin as described above. Spleen cells from immunized mice were fused to P3X cells, and hybridoma clones were selected in the manner described above for trophinin and bystin. Anti-tastin antibody was purified by affinity chromatography using protein A-Sepharose beads (Pharmacia, Piscataway, NJ). The investigations using mice were conducted in accordance with the Guiding Principles for the Care and Use of Research Animals promulgated by the Society for the Study of Reproduction.

Human Placental Tissues

Paraffin blocks of human placental tissues were selected from the pathology files of the Department of Gynecology and Obstetrics, Keio University School of Medicine; Department of Pathology, Johns Hopkins Hospital; and Central Clinical Laboratories, Shinshu University Hospital. These tissue specimens were obtained from patients at the 6th, 7th, 8th, 10th, and 38th wk of gestation. Informed consent was obtained from the patients, and the investigations were conducted in accordance with the guidelines in the Declaration of Helsinki, as revised in 1983. No pathological findings, including infection and inflammation, were observed in the specimens. These tissues were fixed for 48 h in 20% formalin buffered with 0.1 M phosphate buffer, pH 7.4, at room temperature, embedded in paraffin, and cut into 4-μm sections for immunohistochemistry or 7-μm sections for in situ hybridization.

Western Blot

GST fusion proteins were expressed in bacteria and were purified by affinity chromatography using glutathione sepharose beads as described previously [5, 6]. Proteins were resolved in SDS-PAGE according to Laemmli [8], and proteins were electrotransferred to a nitrocellulose filter according to Towbin et al. [9]. The filters were soaked with PBS, pH 7.4, containing 1% powdered nonfat milk and 0.02% Tween 20, and reacted first with diluted mouse monoclonal antibodies, and then with peroxidase conjugated goat anti-mouse immunoglobulin. Immunoreactive bands were detected using the chemiluminescent ECL reagent (Amersham, Arlington Heights, IL).

Immunohistochemistry

Paraffin sections of human placenta tissues were analyzed immunohistologically using the monoclonal antibodies described above, followed by biotinylated second antibodies. Staining was performed using the peroxidase-conjugated avidin ABC kit (Vector Laboratories, Burlingame, CA) with peroxidase-labeled anti-mouse IgG (tastin) or anti-mouse IgM (trophinin and bystin). Counterstaining was performed using hematoxylin. Anti-cytokeratin 8 antibody (clone 35bH11, mouse IgM) was purchased from Dako (Carpinteria, CA). Immunohistochemistry of cytokeratin was performed using the ABC kit. Double immunofluorescence microscopy was performed using rabbit polyclonal anti-trophinin antibodies raised against the amino terminal region of human trophinin [5] and mouse monoclonal antibody for human lamp-1 [10] as the primary antibodies, followed by fluorescein-conjugated anti-rabbit IgG and rhodamine-conjugated anti-mouse IgG as the secondary antibodies.

In Situ Hybridization

RNA probes for human trophinin, tastin, and bystin were prepared as follows. On the basis of the published sequence [5], the human trophinin cDNA sequence surrounding the initiation methionine (nucleotide resides −3 to +149) was amplified by polymerase chain reaction (PCR) using a primer pair XbaI-TRO (5-GCTCTAGACATGGATATCGACTGCCT-3) and TRO-Asp718 (5-GGGGTACCAGCCCTGGTACTAGCT-3). This amplified fragment was then subcloned into the XbaI and Asp718 sites of pGEM-3Zf (+) (Promega, Madison, WI), and the resultant vector was used as a template for construction of the RNA probe for human trophinin. Similarly, the human tastin cDNA sequence from nucleotides −26 to +124 and the human bystin cDNA sequence from nucleotides +627 to +796 were amplified by PCR using primer sets XbaI-TAS (5-GCTCTAGAGCCCTGAGGGGCCTCG-3) and TAS-Asp718 (5-GGGGTACCGGTCCACGGCGCACGAT-3) for tastin, or XbaI-BYS (5-GCTCTAGAGCTGGATAAGAAGTATGCAC-3) and BYS-Asp718 (5-GGGGTACCAGAGGGCCTCTTTCTGGC-3) for bystin. These amplified fragments were subcloned into the XbaI and Asp 718 sites of pGEM-3Zf (+).

A digoxigenin (DIG)-labeled antisense RNA probe was obtained using an XbaI-cut template and T7 RNA polymerase with a DIG RNA labeling kit (Boehringer Mannheim, Mannheim, Germany). Similarly, a sense probe was prepared for negative control experiments using an Asp718-cut template and SP6 RNA polymerase with the same kit.

Tissue specimens were subjected to in situ hybridization to detect the transcripts of trophinin, tastin, and bystin using a nonradioactive system as described [11]. After the tissue sections were deparaffinized in xylene, hydrated slides were immersed in 0.2 M HCl for 20 min, then digested with 100 μg/ml proteinase K at 37°C for 20 min, and postfixed in 4% paraformaldehyde. The slides were rinsed with 2 mg/ml glycine and acetylated for 10 min in freshly prepared 0.25% acetic anhydride in 0.1 M triethanolamine (pH 8.0). The hydrated slides were then defatted with chloroform and air-dried. After prehybridization in 50% deionized formamide/double-strength SSC (single-strength SSC is 0.15 M sodium chloride, 0.015 M sodium citrate) for 1 h at 45°C, the slides were hybridized with 0.5 μg/ml of the antisense or sense probe in 50% deionized formamide, 2.5 mM EDTA (pH 8.0), 300 mM NaCl, single-strength Denhardt's solution, 10% dextran sulfate, and 1 mg/ml brewers' yeast tRNA at 45°C for 16 h. After hybridization, the slides were washed in 50% formamide/double-strength SSC for 1 h at 45°C and digested with 10 mg/ml RNase A at 37°C for 30 min. The slides were treated with single-strength SSC/50% formamide at room temperature for 30 min and were subjected to immunohistochemistry for detection of the hybridized probes using an alkaline phosphatase-conjugated anti-DIG antibody (Boehringer Mannheim, Indianapolis, IN). The alkaline phosphatase reaction was visualized with 5-bromo-4-chloro-3-indolyl phosphate and nitroblue tetrazolium, and, finally, tissue sections were mounted in Glycergel (Dako).

Results

Specificity of Monoclonal Antibodies Used in This Study

Monoclonal anti-trophinin antibody (clone 3-11, mouse IgM) was raised against the SIVGFSGGP epitope or amino acid residues 681–689 of human trophinin. The antibody reacted with COS cells transfected with the mammalian expression vector pcDNAI containing human trophinin cDNA (Fig. 1a), whereas COS cells transfected with pcDNAI vector alone did not show positive immunostaining (not shown). A Western blot using GST-553, in which GST is fused to the extracellular domain of trophinin (amino acid residues 634–719), showed that the antibody reacted with GST-553 but not with GST (Fig. 2, lanes 1 and 2). Human trophoblastic teratocarcinoma HT-H cells, which constitutively express trophinin, were stained with this antibody without permeabilization [6]. A Western blot of an HT-H cell lysate showed a single band at 70 kDa, which is consistent with the calculated molecular weight of 69 214 for human trophinin.

Fig. 1

Immunofluorescence micrographs of COS cells. COS-1 cells were transfected with pcDNAI mammalian expression vectors, containing trophinin (a), tastin (b), or bystin (c) cDNA. Two days later, the cells were stained with monoclonal antibodies for trophinin (a), tastin (b), or bystin (c), followed by fluorescent second antibodies as described previously [26]. COS-1 cells transfected with pcDNAI vector alone and stained with these antibodies showed no staining (not shown). Bar = 10 μm.

Fig. 2

Western blot of GST fusion proteins using monoclonal antibodies. Western blots using anti-trophinin antibody (lanes 1 and 2), anti-bystin antibody (lanes 3 and 4), and anti-tastin antibody (lanes 5 and 6) are shown. Lanes 1, 3, and 5 contain GST as control. Lane 2 contains GST-553 (GST fused to a part of extra cellular domain of trophinin); lane 4, GST-bystin; and lane 6, GST-tastin. The estimated molecular mass for GST-553 is 34 kDa, for GST-tastin 109 kDa, and for GST-bystin 61 kDa.

Anti-tastin antibody (clone 38, mouse IgG3) was raised against the DQENQDPRR epitope or amino acid residues 41–49 of human tastin. The antibody reacted with COS cells transfected with human tastin cDNA and showed a cytoskeleton-like pattern (Fig. 1b). Association of tastin with the cytoskeleton in the trophoblastic cells and endometrial epithelial cells has been shown in a previous study using polyclonal anti-tastin antibodies [5]. A Western blot using GST-tastin fusion protein, in which GST is fused to a full-length tastin, showed that the antibody reacted with GST-tastin, but not with GST (Fig. 2, lanes 5 and 6). HT-H cells stained with this monoclonal antibody after permeabilization with detergent showed cytoskeletal patterns as well as punctate cytoplasmic profiles [6]. A Western blot of HT-H cell lysate showed a single band at 85 kDa, which is consistent with the calculated molecular weight of 83 754 for human tastin.

Anti-bystin antibody (clone 19, mouse IgM) was raised against the GFRTEKREL epitope or amino acid residues 215–223 of human bystin. The antibody reacted with COS cells transfected with pcDNAI containing human bystin cDNA (Fig. 1c). A Western blot using GST-bystin, in which GST is fused to full-length bystin, showed that the antibody reacted with GST-bystin but not with GST (Fig. 2, lanes 3 and 4). HT-H cells that stained with this antibody after permeabilization with detergent exhibited punctate granules in the cytoplasm as well as cytoskeletal profiles [6]. A Western blot of HT-H cell lysate showed a single band at 35 kDa, which is consistent with the calculated molecular weight of 35 169 for human bystin.

Immunohistochemistry

Paraffin-embedded tissue sections of human placenta from the first to the third trimesters were subjected to immunohistochemical analysis using the monoclonal antibodies described above. The results are presented in Figures 3–5.

Fig. 3

Immunohistochemistry of wk 6 placenta. Each panel shows serial tissue sections stained with antibodies for trophinin (a), tastin (b), or bystin (c). d) A control without the first antibody. a–c) Large arrows: syncytiotrophoblast; b) arrow: apical microvilli of syncytiotrophoblast; a and c) arrowheads: cytotrophoblast. CV, chorionic villi; D, decidua. Bar = 10 μm.

Fig. 4

Immunofluorescence micrographs of the chorionic villi from wk 6 placenta. Part of the villus subjected to reaction with anti-trophinin antibody shows trophinin staining at apical plasma membranes (a). The tissue was double-stained with polyclonal anti-trophinin antibody and monoclonal anti-human lamp-1 antibody followed by fluorescein-conjugated goat anti-rabbit IgG (b) and rhodamine-conjugated goat anti-mouse IgG (b′) antibodies. Note significant overlap between trophinin (b) and lamp-1 staining (lysosomal membrane marker; b′). Arrows indicate apical surfaces of syncytiotrophoblast. VC, Villous core. Bar = 10 μm.

Fig. 5

Immunohistochemistry of wk 10 placenta showing staining for trophinin (A), tastin (B), or bystin (C). D) A control without the first antibody. A–C) Small arrows: lysosome-like vesicles in syncytiotrophoblast; B) large arrow: apical microvilli of syncytiotrophoblast. CV, chorionic villi. Bar = 10 μm.

Trophinin

Strong immunostaining for trophinin was detected in the cytoplasm of syncytiotrophoblast of chorionic villi in placenta from the 6th week of pregnancy (Figs. 3a and 4, a and b). Cytotrophoblast in the villi were largely negative for trophinin. Some of the stromal cells within the villi were weakly positive. In the endometrium adjacent to the chorionic villi, staining of stromal cells or decidual cells (marked D, in Fig. 3a), although weak, was also observed but was not detectable in controls (Fig. 3d). In wk 6 placenta, trophinin was seen in the cytoplasm of the syncytiotrophoblast (Fig. 3a). Immunofluorescence microscopy showed that the staining was restricted to lysosome-like vesicles (Fig. 4, a and b). These vesicles were most likely lysosomes and endosomes, as double immunostaining for trophinin and a lysosomal membrane marker, lamp-1, overlapped (Fig. 4, b and b′).

In wk 10 villi, trophinin was more abundant in cytotrophoblast than in syncytiotrophoblast (Fig. 5A), the reverse of the situation seen in villi of wk 6 placenta; at that stage, trophinin was found mostly in syncytiotrophoblast (Fig. 3a). The lysosome-like vesicles were also seen in syncytiotrophoblast (Fig. 5A, long arrows). In the second and third trimesters, trophinin was barely detectable in the placenta tissues (data not shown).

Tastin

In wk 6 placenta, tastin was detected exclusively in the syncytiotrophoblast in close association with apical cell membranes (Fig. 3b). Tastin was not detected in the endometrial stromal cells. In wk 10, tastin was found at the surface membranes and in lysosome-like vesicles of syncytiotrophoblast (Fig. 5B). In the second and third trimesters, tastin was undetectable in the placenta (not shown).

Bystin

In wk 6 placenta, strong immunostaining for bystin was detected in the cytoplasm of syncytiotrophoblast (Fig. 3c). Cytotrophoblast in the chorionic villi was largely negative. In the endometrium adjacent to the chorionic villi, bystin was detected in decidual cells. In wk 10, bystin was present in the cytotrophoblast and in lysosome-like vesicles in the syncytiotrophoblast (Fig. 5C), in a staining pattern similar to that of trophinin (Fig. 5A). In the second and third trimester, bystin was barely detectable in the placenta tissues (not shown).

In Situ Hybridization

Since the results shown above indicated that trophinin, tastin, and bystin are expressed in the trophoblast and endometrium in the early stage of pregnancy, expression profiles of the transcripts of these genes were determined by in situ hybridization (Fig. 6). Tissue specimens subjected to analysis were from two patients who underwent spontaneous abortion, one at wk 7 and the other at wk 8 after conception. Placental tissues from these two patients showed similar expression profiles of transcripts for trophinin, tastin, and bystin.

Fig. 6

In situ hybridization for trophinin, tastin, and bystin transcripts in human placenta and endometrium during wk 7 after conception. Serial sections of the placental tissue were sequentially hybridized with the specific probes. a,d,g,j) Chorionic plate and placental villi. b,e,h,k) High magnification of a placental villus. c,f,i,l) Endometrial gland in the decidua. In situ hybridization with trophinin antisense probe (a–c), tastin antisense probe (d–f), bystin antisense probe (g–i), and bystin sense probe (j–l). cp, Chorionic plate; v, mesenchymal villi; bp, basal plate; g, endometrial gland. Bar = 150 μm.

Trophinin

A specific signal for trophinin was detected in cytotrophoblast covering the chorionic plate and villi (Fig. 6a). The syncytiotrophoblast of the chorionic villi exhibited a moderate trophinin signal (Fig. 6b), and endometrial glands of the uterus expressed weak but definite signals (Fig. 6c). In addition, stromal cells of the chorionic plate as well as decidual cells expressed various levels of the trophinin transcripts.

Tastin

A strong tastin signal was detected in the double-layered trophoblastic epithelia of the chorionic plate as well as in mesenchymal villi (Fig. 6, d and e). An intense signal was also observed in endometrial glands (Fig. 6f). Stromal cells in the chorionic plate and decidual cells were also positive for tastin transcripts.

Bystin

Bystin transcripts were strongly expressed in the cytotrophoblast of the chorionic plate and villi (Fig. 6g). The intensity of the signal in syncytiotrophoblast of the chorionic plate and villi was lower than that seen in cytotrophoblast (Fig. 6h). A moderate signal for bystin was detected in the endometrial glands (Fig. 6i). Stromal cells and decidual cells were also positive for bystin RNA. Control hybridization with sense probes for trophinin, tastin, and bystin was negative (Fig. 6, j–l).

Discussion

This study demonstrates that trophinin, tastin, and bystin are expressed in the human placenta during early pregnancy. It is noteworthy that these molecules are found in the placenta only in the early stage of the first trimester (Figs. 3–6). This study demonstrates the expression of these molecules in both fetal and maternal cells at the utero placental interface (Figs. 3 and 6).

We have previously demonstrated that these three proteins form a complex with a potential role in blastocyst-uterine adhesion during early implantation [5, 6]. In the previous study, we found restricted and unique expression of trophinin in the trophectoderm of the blastocyst, and in trophoblast and endometrial epithelial cells at the attachment site in the monkey [5]. In this study, we found trophinin in the syncytiotrophoblast and cytotrophoblast in the chorionic villi from the human placenta at the 6th wk of pregnancy (Fig. 3a). Expression of trophinin continues after wk 6 in the cytotrophoblast but decreases in the syncytiotrophoblast (Figs. 5A and 6, a and b). Results of in situ hybridization also indicate that trophinin, tastin, and bystin are expressed in the cytotrophoblast at wk 7 (Fig. 6, a, b, d, e, g, h). However at wk 6, trophinin protein may be in the process of being degraded, because trophinin is mainly associated with lysosomes (Fig. 4, b and b′). At wk 10 of pregnancy, trophinin protein is significantly reduced in the syncytiotrophoblast (Fig. 5A). Although the significance of trophinin expression in the cytotrophoblast is unknown, it is possible that trophinin plays a role in trophoblast invasion, since intermediate (extravillous) trophoblast, which is derived from the villous cytotrophoblast [3], expresses trophinin together with tastin and bystin (unpublished results). Intermediate trophoblast also expresses the melanoma cell adhesion molecule, Mel-CAM [12], suggesting a parallel to cell adhesion exhibited by other types of invasive cells. Remarkably, trophinin, tastin, and bystin are expressed in an implantation- site, intermediate trophoblastic cell line (IST-1), established from a placenta at wk 7 of gestation (unpublished results). The availability of such a cell line will increase our understanding of cell adhesion in invasive trophoblastic cells and tumor cells.

Trophinin is not a typical membrane protein [5]. Its amino terminal region, consisting of about 70 amino acids, is hydrophilic, and this region is predicted to be cytoplasmic. Antibodies raised to this region do not recognize trophinin in cells unless they are permeabilized. The rest of the trophinin polypeptide including its carboxyl terminus consists of decapeptide repeats. Within the repeats are three relatively hydrophobic regions, suggesting that this protein is an intrinsic membrane protein. Also within the repeats, there are three relatively hydrophilic regions likely to be exposed to the cell surface. Trophinin mediates cell adhesion at apical plasma membranes in syncytiotrophoblastic cells. This suggests that, in the earliest stage of pregnancy, trophinin localizes to apical plasma membranes in syncytiotrophoblast. The presence of trophinin in the apical plasma membranes was seen in a part of the villus from wk 6 placenta (Fig. 4a), although majority of trophinin was seen in lysosomes at this stage of placenta (Fig. 4, b and b′). It appears that trophinin moves from the plasma membrane to endosomes and/or lysosomes by endocytosis (Fig. 4, a and b). Mechanisms by which trophinin localizes to plasma membranes or cytoplasmic vesicles remain to be addressed in future studies.

Tastin is a soluble cytoplasmic protein. Previous studies showed that tastin colocalizes to the cytoskeleton in trophoblastic teratocarcinoma HT-H and endometrial adenocarcinoma SNG-M cells [5]. In vitro protein-binding assays showed that tastin binds to cytokeratins through bystin [6]. The present study showed that tastin appears to localize to the apical microvilli in syncytiotrophoblast (Fig. 3b). The core of the apical microvilli is made of bundles of actin, and proteins of the ezrin-radixin-moesin family associate with actin [13, 14]. It is known that the apical microvilli in syncytiotrophoblast are rich in ezrin [15, 16], and that other actin-binding proteins, such as annexin II (p36) and p11, are abundant in placenta syncytiotrophoblast [17]. Although microvilli localization of tastin has not been seen in tissue culture cells, such restricted microvilli localization of tastin in vivo suggests that tastin associates with a protein(s) other than bystin and that localization of this protein is regulated developmentally in the human placenta.

Bystin is a cytoplasmic protein that directly binds to trophinin, tastin, and cytokeratin [6]. The bystin gene is well conserved from yeast to humans [6, 18, 19]. The ENP1 gene, which encodes a bystin homologue in yeast, is essential for the vegetative growth of yeast [19]. As shown in this study, expression of bystin and trophinin is well coordinated in the human placenta (Figs. 3 and 5). A previous study showed that bystin enhances trophinin-mediated cell adhesion in vitro [6]. However, the in vivo role of bystin in implantation and placentation is presently unknown. Generation of bystin gene knock-out mice will provide a clearer evaluation of the role of this protein in mammals.

Trophoblast invasion is linked to trophoblast differentiation, and this process in normal pregnancy is well controlled [20]. However, the processes of trophoblastic invasion are similar to those of malignant tumor metastasis, as both processes are often accompanied by aggressive cell proliferation, host cell destruction, and angiogenesis [21–25]. Understanding the role of trophinin, tastin, and bystin may further our understanding of placental development and also shed light on tumor metastasis.

Acknowledgments

The authors thank Drs. M. Sakamoto and N. Hiraoka, the Burnham Institute, for their helpful suggestions and discussions; M.A. Laurence and Dr. N. Varki, the Tissue Bank, University of California-San Diego, for their assistance in preparation of tissue sections; E.E. Lamar for editing the manuscript; and K. Lowitz for technical assistance.

Author notes

¹

This study was supported by NIH RO1 HD34108, the Mizutani Foundation 960073, and Kyowa Medex.