Tage4/Nectin-like molecule-5 heterophilically trans-interacts with cell adhesion molecule Nectin-3 and enhances cell migration.

Malignant transformation of cells causes disruption of cell-cell adhesion, enhancement of cell motility, and invasion into surrounding tissues. Nectins have both homophilic and heterophilic cell-cell adhesion activities and organize adherens junctions in cooperation with cadherins. We examined here whether Tage4, which was originally identified to be a gene overexpressed in colon carcinoma and has a domain structure similar to those of nectins, is involved in cell adhesion and/or migration. Tage4 heterophilically trans-interacted with nectin-3, but not homophilically with Tage4. Expression of Tage4 was markedly elevated in NIH3T3 cells transformed by an oncogenic Ki-Ras (V12Ras-NIH3T3 cells) as compared with that of wild-type NIH3T3 cells. trans-Interaction of Tage4 with nectin-3 enhanced motility of V12Ras-NIH3T3 cells. Tage4 did not bind afadin, a nectin- and actin filament-binding protein that connects nectins to the actin cytoskeleton and cadherins through catenins. Thus, Tage4 heterophilically trans-interacts with nectin-3 and regulates cell migration. Tage4 is tentatively re-named here nectin-like molecule-5 (necl-5) on the basis of its function and domain structure similar to those of nectins.

In multicellular organisms, cell adhesion and migration are critical for many events, including tissue patterning, morphogenesis, and maintenance of normal tissues (1)(2)(3). They also play roles in malignant transformation of cells (4). Adhesion and migration of non-transformed normal cells are dynamic and well regulated (2). Cells disrupt cell-cell adhesion and start to migrate in response to extracellular cues, such as growth factors, cytokines, and extracellular matrix molecules (4). When migrating cells contact other cells, they stop migration and proliferation and adhere to each other to become confluent (5,6). This phenomenon is known for a long time as contact inhibition of cell movement and proliferation. Transformation of cells causes disruption of cell-cell adhesion, increase of cell motility, and loss of contact inhibition of cell movement and proliferation, eventually leading the transformed cells to invasion into surrounding tissues and metastasis to other organs (4,7). However, molecular mechanisms underlying these physiological or pathological processes are not fully understood.
Cell-cell adherens junctions (AJs) 1 play major roles in cellcell adhesion in fibroblasts and epithelial cells (1,2). Cadherins are key Ca 2ϩ -dependent cell-cell adhesion molecules at AJs (1,2). Cadherins are associated with the actin cytoskeleton through peripheral membrane proteins, including ␣and ␤-catenins, in fibroblasts and epithelial cells (1). This association strengthens the cell-cell adhesion activity of cadherins (1). Nectins and afadin constitute another cell-cell adhesion unit that localizes at cell-cell AJs and regulates organization of AJs in cooperation with cadherins in fibroblasts and epithelial cells (8). Nectins are Ca 2ϩ -independent Ig-like cell-cell adhesion molecules. Afadin is a nectin-and actin filament-binding protein that connects nectins to the actin cytoskeleton. Nectins comprise a family of four members, nectin-1, -2, -3, and -4, each of which has two or three splicing variants. Nectins have one extracellular region with three Ig-like loops, one transmembrane region, and one cytoplasmic region. All nectins except nectin-4 have a C-terminal conserved motif of four amino acids (aa) residues, which interacts with the PDZ domain of afadin. Nectin-4 does not have this motif but binds afadin. Each nectin forms homo-cis-dimers, followed by the formation of homotrans-dimers, causing cell-cell adhesion. Nectin-3 furthermore heterophilically trans-interacts with nectin-1 or -2 and the adhesion activity of these heterophilic trans-interactions is stronger than that of the homophilic trans-interactions. Nectin-4 also heterophilically trans-interacts with nectin-1.
Five or six molecules having one extracellular region with three Ig-like loops, one transmembrane region, and one cytoplasmic region have thus far been identified (Table I) (9 -19).
* The work at Osaka University was supported by grants-in-aid for Scientific Research and for Cancer Research from the Ministry of Education, Culture, Sports, Science, and Technology, Japan (2001,2002). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
We tentatively name here these molecules nectin-like molecules (necls) on the basis of their domain structures similar to those of nectins (see "Discussion"). Of these necls, Tage4 was originally identified to be a gene overexpressed in rat and mouse colon carcinoma (16,17). Northern blot analysis has revealed that Tage4 is expressed in normal adult rat and mouse tissues to small extents (16,17), but its function remains unknown, except that it mediates entry of porcine pseudorabies virus and bovine herpesvirus 1 (20). We have studied here the function of Tage4 and revealed that Tage4 heterophilically trans-interacts with nectin-3 and regulates cell migration. Tage4 is tentatively re-named here necl-5 on the basis of its function and phylogenetic tree of nectins and necls ( Fig. 1) (see "Discussion").

EXPERIMENTAL PROCEDURES
Molecular Cloning of Mouse Necl-5 cDNA-The cDNA of mouse Tage4/necl-5 was originally isolated by reverse transcriptase-PCR from the C26 mouse colon carcinoma cell line (DDBJ/GenBank TM /EBI accession number MMU35836) (17). This cell line was derived from BALB/c mice. Because we generally use nectins derived from C57BL/6 mice, we re-cloned the Tage4/necl-5 cDNA derived from C57BL/6 mice. We searched in the DNA data base and found one sequence similar to that of Tage4/necl-5 (DDBJ/GenBank TM /EBI accession number BC013673). We performed reverse transcriptase-PCR from mouse brain total RNA of C57BL/6 mice on the basis of BC013673. The new sequence of Tage4/ necl-5 showed 93% nucleotide identity to that of the original one. The C-terminal half was identical, but the N-terminal half was slightly different. The new sequence was identical to BC013673 except for the exchange of a single nucleotide from cytosine to adenine, at position 854 (open reading frame). The reason for this difference is not known, but may be due to the different strains of mice. We confirmed that the isolated cDNA encodes the full-length protein: the protein was expressed in L cells and the molecular mass of the expressed protein was compared with that of the endogenous protein, which was expressed in NIH3T3 cells stably expressing V12Ki-Ras, an oncogenic Ki-Ras, (V12Ras-NIH3T3 cells). The molecular masses of the two proteins were apparently similar as estimated by SDS-PAGE, followed by Western blotting (see Fig. 5).
Surface Plasmon Resonance Analysis-A BIAcore X surface plasmon resonance-based biosensor (BIAcore Inc., Piscataway, NJ) was used to measure kinetic parameters for the interaction between Neap-1, Neap-2, or the extracellular fragment of necl-5 fused to SEAP (Leap-5) and immobilized Nef-3 or Lef-5. The F(abЈ) 2 fragment goat anti-human IgG Fc polyclonal Ab (pAb) was immobilized at a concentration of about 4800 resonance units (4.8 ng/mm 2 ) to the sensor chip surface by the amine-coupling method. Nef-3 or Lef-5 was immobilized at a concentration of about 500 resonance units to the sensor chip via the antihuman IgG Fc Ab. Neap-1, Neap-2, or Leap-5 was then diluted in HBS-EP buffer (10 mM HEPES, pH7.4, 150 mM NaCl, 3 mM EDTA, 0.005% Tween 20; BIAcore) to 40 nM and injected at a flow rate of 20 l/min at 25°C for 210 s. Both an association rate constant k a (M Ϫ1 s Ϫ1 ) and a dissociation rate constant k d (s Ϫ1 ) were obtained using the BIAevaluation software version 3.2 (BIAcore), and the dissociation constant (K D ϭ k d /k a ) was derived from the two deduced rate constants.
Intercellular Motility Assay-Intercellular motility assay was done as described previously (29). Briefly, the cells were labeled with 1,1Јdioctadecyl-3,3,3Ј,3Ј-tetramethylindocarbocyanine perchlorate (DiI), and the 1 ϫ 10 2 labeled cells were seeded on a confluent culture of 2 ϫ 10 5 unlabeled nectin-3-L cells in a 24-well dish. After 36 or 48 h of culture, four sister cells that seemed to be derived from one seeded cell were examined by fluorescence microscopy. In the experiments using the anti-necl-5 mAb, this mAb was added at a final concentration of 50 g/ml in the medium. When a cell line A was seeded on a confluent culture of a cell line B, we designated the experiment as A/B analysis. For quantification of intercellular motility, intercellular distances of all combinations between four sister cells were measured and summed as Dc. As a control experiment, the labeled cells were seeded on dishes in the absence of a cell layer. In this case, the intercellular distances were summed as Dd. The degree of intercellular motility was represented as Dc/Dd. At least 24 independent samples were picked up to determine Dc or Dd for each cell line.
Other Procedures-The cell aggregation assay, chemical cross-linking, and SDS-PAGE were done as described previously (21,26,28,30). The inability of necl-5 to bind afadin was confirmed by yeast two-hybrid assay, co-immunoprecipitation assay, and affinity chromatography as described previously (23,26,28), under the conditions where nectin-2 and -3 bound afadin. Protein concentrations were determined with bovine serum albumin as a reference protein as described previously (31).
We then examined whether necl-5 has heterophilic cell-cell adhesion activity with other nectins. Necl-5-L cells were mixed with nectin-1-L, -2-L, or -3-L cells followed by the aggregation assay. Nectin-3-L cells formed small aggregates in the absence of necl-5-L cells as described previously (26) (Fig. 2C), but formed relatively big aggregates with necl-5-L cells (Fig. 2D, Da-Dc). This aggregate was similarly formed even in the absence of Ca 2ϩ (data not shown). The size of the aggregates formed between necl-5-L and nectin-3-L cells were about 20% that of the aggregates formed between nectin-1-L and -3-L cells, which formed the biggest aggregates among various combinations of nectins thus far examined (32) (see Fig. 7, Aa and Ba). Necl-5-L cells did not form mixed aggregates with nectin-1-L or -2-L cells (Fig. 2, Ea-Ec and Fa-Fc). Small aggregates observed were formed by nectin-1-L or nectin-2-L cells by themselves as described (28) (Fig. 2, Ea, Ec, Fa, and Fc). These results indicate that necl-5 heterophilically trans-interacts selectively with nectin-3 in a Ca 2ϩ -independent manner, causing To confirm that necl-5 directly interacts with nectin-3, we performed surface plasmon resonance analysis using Nef-3 and Leap-5. Neap-1 and -2 were used as controls. Nef-3 is the extracellular fragment of nectin-3 fused to the human IgG Fc; Leap-5 is the extracellular fragment of necl-5 fused to SEAP; Neap-1 is the extracellular fragment of nectin-1 fused to SEAP; and Neap-2 is the extracellular fragment of nectin-2 fused to SEAP. Nef-3 bound all of these molecules, and the K d value of Nef-3 for Leap-5 was about 17 nM, whereas the K d values of Nef-3 for Neap-1 and -2 were about 2.3 and 360 nM, respectively (Fig. 3). Lef-5 did not bind Leap-5 (data not shown). Lef-5 is the extracellular fragment of necl-5 fused to the human IgG Fc.
Inability of Necl-5 to Bind Afadin-The extracellular region of necl-5 showed 28 -42% aa identity to that of nectins, but necl-5 does not have a C-terminal consensus motif with four aa for binding to the PDZ domain (data not shown). We examined whether necl-5 binds afadin. Necl-5 did not bind afadin as estimated by yeast two-hybrid assay, co-immunoprecipitation assay, and affinity chromatography under the conditions where nectin-2 or -3 bound it (data not shown).
Elevated Expression of Necl-5 in V12Ras-NIH3T3 Cells-We then examined the tissue distribution of necl-5 in mouse by Western blotting, but the significant immunoreactive band was not detected in any normal tissue examined, including heart, brain, spleen, lung, liver, kidney, skeletal muscle, and testis (data not shown), consistent with the earlier observation (17). Any band was not detected in cultured cell lines, including L (Fig. 5), MTD-1A (data not shown), and MDCK cells (data not shown), but two faint bands were detected in NIH3T3 cells (Fig. 5). Tage4 was originally isolated from rat and mouse colon carcinoma (16,17). We therefore examined the expression of necl-5 in V12Ras-NIH3T3 cells. Expression of necl-5 was markedly elevated in the transformed cells as compared with that of the wild-type cells (Fig. 5).
Enhancement of Motility of V12Ras-NIH3T3 and Necl-5-L Cells by trans-Interaction of Necl-5 with Nectin-3-We finally studied the role of the trans-interaction of necl-5 with nectin-3 on motility of V12Ras-NIH3T3 cells by using intercellular motility assay, because transformed cells show generally enhanced migration activity (4). Necl-5-L and necl-5-⌬EC-L cells were used as control cells. Necl-5-⌬EC-L cells were L cells expressing necl-5, of which extracellular region except the juxtamembrane 13 aa were deleted. In this assay, cell motility in a confluent cell sheet, which is influenced by dynamic cell-cell adhesion, could be measured. V12Ras-NIH3T3 cells labeled with DiI, a fluorescence dye, were seeded on a confluent culture of non-labeled nectin-3-L cells, and after 36 h (twice the doubling time), the cell scatter property was analyzed by measuring the mass distance among four sister labeled cells. As a control experiment, labeled V12Ras-NIH3T3 cells were seeded on the dish in the absence of nectin-3-L cells. V12Ras-NIH3T3 cells scattered on a confluent culture of nectin-3-L cells more actively than on the dish (Fig. 6, Aa1, Aa2, and B). The scattering of V12Ras-NIH3T3 cells on nectin-3-L cells was inhibited by the anti-necl-5 mAb, whereas the scattering on the dish was not affected by this mAb (Fig. 6, Aa3, Aa4, and B). The anti-necl-5 mAb inhibited the interaction of necl-5 with nectin-3 as estimated by the aggregation assay using necl-5-L and nectin-3-L cells (Fig. 7, Aa and Ab). This mAb did not affect the interaction of nectin-1 with nectin-3 (Fig. 7, Ba and Bb). Necl-5-L cells labeled with DiI were seeded on a confluent culture of non-labeled nectin-3-L cells, and after 48 h (twice the doubling time), the cell scatter property was similarly analyzed. Necl-5-L cells scattered on a confluent culture of nectin-3-L cells more actively than on the dish (Fig. 6, Ab1, Ab2, and B). The scattering of necl-5-L cells on nectin-3-L cells was inhibited by the anti-necl-5 mAb, whereas the scattering of necl-5-L cells on the dish was not affected by this mAb (Fig. 6, Ab3, Ab4, and B). Necl-5-⌬EC-L cells labeled with DiI scattered on a confluent culture of nectin-3-L cells less actively than on the dish (Fig. 6,  Ac1, Ac2, and B). Necl-5-⌬EC-L cells did not adhere to nectin-3-L cells as estimated by the aggregation assay (data not shown). The scattering of necl-5-⌬EC-L cells in the presence or absence of nectin-3-L cells was not affected by the anti-necl-5 mAb (Fig. 6, Ac3, Ac4, and B). These results indicate that the trans-interaction of necl-5 with nectin-3 enhances motility of V12Ras-NIH3T3 and necl-5-L cells. DISCUSSION We have shown here that necl-5 does not homophilically trans-interact with necl-5, but heterophilically trans-interacts selectively with nectin-3, causing cell-cell adhesion. This property of necl-5 is quite different from that of nectins which both homophilically and heterophilically trans-interact (8). We have previously proposed that nectins are involved in the formation of AJs in cooperation with E-cadherin, on the basis of the observations that the trans-interaction of nectins recruits Ecadherin to the nectin-based cell-cell adhesion sites, resulting in formation of AJs, and that the disruption of this transinteraction of nectins by their antagonists impairs the formation of E-cadherin-based AJs (8). The association of nectins and E-cadherin at AJs is mediated through afadin and ␣-catenin (8). We have shown here that necl-5 does not bind afadin. The inability of necl-5 to bind afadin suggests that necl-5 has no potency to recruit cadherins to the cell-cell adhesion site formed by the trans-interaction of necl-5 with nectin-3 and is not involved in the formation of AJs.
We have shown here that the heterophilic trans-interaction of necl-5 with nectin-3 rather enhances motility of V12Ras-NIH3T3 and necl-5-L cells. It has previously been reported that L cells stably expressing full-length E-cadherin (EL cells) shows inter-EL-cellular EL cell motility (29,33). The mechanism of this intercellular motility of EL cells is not clear, but it has been suggested that dynamic attachment of EL cells to neighboring EL cells and dynamic detachment of EL cells from neighboring EL cells are necessary for the motility of EL cells (29,33). The mechanism of intercellular motility of V12Ras-NIH3T3 and necl-5-L cells is not known, either, but may be analogous to that of EL cells.
Transformation of cells increases cell motility, causing invasion into surrounding tissues. Since expression of necl-5 is elevated by transformation as shown here and described previously (16,17), this elevation of necl-5 may be at least partly responsible for the enhanced cell motility and invasion of transformed cells. Intercellular motility is observed in vivo in the process of morphogenetic rearrangement of cells in embryonic tissues (2,3). It remains unknown what kind of the cells express necl-5 in embryonic tissues, but if necl-5 is expressed in rapidly migrating cells, such as mesenchymal cells, the dynamic trans-interaction of necl-5 with nectin-3 may also play a role in their intercellular motility. Further studies are necessary for establishing the physiological and pathological roles of necl-5 in these processes.
We lastly discuss about other necls which have thus far been identified in addition to necl-5/Tage4 (Table I). Five or six necls including necl-5 have been identified but have many nomenclatures. We propose here that a group of proteins with structures similar to those of nectins but without ability to directly bind afadin are called nectin-like molecules (necls). NECL1/ TSLL1/SynCAM3, NECL2/IGSF4/RA175/SgIGSF/TSLC1/Syn-