Communicated by: Eisuke Nishida
Molecular characterization of angiomotin/JEAP family proteins: interaction with MUPP1/Patj and their endogenous properties
Article first published online: 19 MAR 2007
Genes to Cells
Volume 12, Issue 4, pages 473–486, April 2007
How to Cite
Sugihara-Mizuno, Y., Adachi, M., Kobayashi, Y., Hamazaki, Y., Nishimura, M., Imai, T., Furuse, M. and Tsukita, S. (2007), Molecular characterization of angiomotin/JEAP family proteins: interaction with MUPP1/Patj and their endogenous properties. Genes to Cells, 12: 473–486. doi: 10.1111/j.1365-2443.2007.01066.x
- Issue published online: 19 MAR 2007
- Article first published online: 19 MAR 2007
- Received: 5 December 2006 Accepted: 27 December 2006
We have previously shown that MUPP1, which has an MRE domain and 13 PDZ domains, is expressed in epithelial cells and localize at tight junctions (TJs) and apical membranes. Using yeast two-hybrid screening, we found here that MUPP1 interacts with angiomotin (Amot), JEAP/Amot-like 1 and MASCOT/Amot-like 2, which we refer to as Amot/JEAP family proteins. PDZ2 and -3 were responsible for MUPP1's interaction with Amot and MASCOT, whereas only PDZ3 was responsible for its interaction with JEAP. All the Amot/JEAP family proteins also interacted with Patj, a close relative of MUPP1. The C-terminal PDZ-binding motives of the Amot/JEAP family were required for these interactions. We successfully generated specific antibodies for these proteins and analyzed the endogenous molecular properties of the family in parallel. Immunofluorescence microscopy of cultured epithelial cells showed that in subcellular distribution, the Amot/JEAP family proteins were indistinguishable; they were apparent at TJs as well as apical membranes, and mostly co-localized with MUPP1. They were also located at TJs in several mouse tissues, but each protein showed a distinct tissue distribution. In biochemical fractionation assays, the Amot/JEAP family behaved not as transmembrane but as peripheral membrane proteins. Unexpectedly, the PDZ-binding motives were not necessarily required for their localization to TJs, and dominant negative MUPP1 or Patj did not affect the localization of Amot/JEAP family proteins, suggesting that the interaction with MUPP1/Patj is not necessarily responsible for their proper subcellular distribution.
The junctional complexes are located in the lateral membranes of epithelial cells, and are composed of tight junctions (TJs), adherens junctions (AJs) and desmosomes (Farquhar & Palade 1963). The TJs are located most apically, and connect adjacent cells so tightly that they constitute a paracellular barrier that prevents the diffusion of solutes across the epithelial cell sheets. They are also responsible for partitioning the apical and basolateral membranes to prevent lipids and membrane proteins from intermingling (Tsukita et al. 2001; Matter & Balda 2003; Anderson et al. 2004; Schneeberger & Lynch 2004). On freeze-fracture electron microscopy, TJs are regarded as a continuous, anastomosing network of intramembranous particle strands, namely TJ strands (Staehelin 1969). Several molecular constituents of TJs have been identified to date. Among them, claudins are a family of integral membrane proteins with four transmembrane domains that are directly responsible for establishing the structural characteristics of TJs (Furuse et al. 1998). Moreover, occludin and JAM family proteins are also known as membranous components of TJs (Furuse et al. 1993; Martin-Padura et al. 1998). Interestingly, many claudin species and JAM possess PDZ-binding motifs at their C-termini. Through these motifs, several PDZ domain-containing proteins are concentrated at TJs (Pontig et al. 1997; Ranganathan & Ross 1997). For example, ZO-1, -2 and -3 are closely related to MAGUK family proteins which have, starting from the N-terminus, three PDZ domains (PDZ1 to -3), one SH3 domain and one GUK domain in this order, and directly interact with the cytoplasmic C-terminal tail of claudins through their PDZ1 domains (Stevenson et al. 1986; Gumbiner et al. 1991; Balda et al. 1993; Furuse et al. 1994; Haskins et al. 1998; Itoh et al. 1999a, 1999b; Wittchen et al. 1999; González-Mariscal et al. 2000). Several other PDZ domain-containing proteins have also been identified, including MAGI-1 (Ide et al. 1999), although the mechanisms by which they are recruited to TJs are not clear.
In an attempt to identify novel molecular components of TJs, we have previously found MUPP1 (multi-PDZ-domain protein 1) as a novel binding partner for the C-terminus of claudin-1 by yeast two-hybrid screening (Hamazaki et al. 2002). MUPP1 was originally identified as a protein that interacts with the C-terminus of the serotonin 5-hydroxytryptamine type 2 receptor (Ullmer et al. 1998). MUPP1 does not have any catalytic domains, but it has multiple protein–protein interaction domains, that is, an MRE domain at its extreme N-terminus and 13 PDZ domains. Through these domains MUPP1 interacts with several molecules, including proto-oncogene c-Kit, the transmembrane proteoglican NG2, an adenovirus E4-ORF1 oncoprotein and a high-risk papillomavirus type 18 E6 oncoprotein (Barritt et al. 2000; Lee et al. 2000; Mancini et al. 2000), suggesting the possible involvement of MUPP1 in growth, proliferation and cell movement. Although its functional characteristics in epithelial cells remain to be addressed, MUPP1 appears to act as a scaffold for several TJ components. Indeed, a number of TJ proteins have been shown to bind with MUPP1, including Pals1, JAM and CAR, as well as claudins (Hamazaki et al. 2002; Coyne et al. 2004; van de Pavert et al. 2004). Considering that MUPP1 has multiple PDZ domains, however, it may well possess several other binding partners.
In this study, we have identified angiomotin (Amot), JEAP/Amot-like 1 and MASCOT/Amot-like 2 as novel binding partners for MUPP1 by yeast two-hybrid screening. We characterized their modes of interaction with MUPP1 using several biochemical assays. Moreover, we successfully generated specific antibodies that clearly recognize the respective endogenous proteins. Utilizing these antibodies, we carefully examined the endogenous behavior of these proteins in cultured epithelial cells as well as in mouse tissues.
Identification of Amot/JEAP family proteins as binding proteins for MUPP1 by yeast two-hybrid assays
In an attempt to identify binding partners for MUPP1 by yeast two-hybrid screening, we generated LexA-binding domain fusion constructs of the first three PDZ domains of MUPP1 (pBTM116-PDZ1 to -3), and used them to screen a mouse embryonic cDNA library (in prey vector pVP16). Approximately 2 × 107 clones were screened under selective culture conditions (His-, Leu- and Trp-), and 500 positive clones were obtained. Among them were two independent clones that encode C-terminal region of Amot (Amot y1; aa982–1127 of mouse Amot, Amot y2; aa986–1127) and two independent clones encoding C-terminus of MASCOT (MASCOT y1; aa688–773, MASCOT y2; aa692–773) (Fig. 1A).
Amot was originally identified as a binding partner for angiostatin, an inhibitor of endothelial cell migration, angiogenesis and growth of primary tumors and metastasis (Troyanovsky et al. 2001). Amot-like 2 has been referred to as a structural homologue of Amot (Bratt et al. 2002). It was then cloned and characterized as a novel interacting partner for MAGI-1, and was named MASCOT (MAGI-1-associated coiled-coil tight junction protein) (Patrie 2005). There is another homologue for Amot, JEAP, which was identified in a screen for novel TJ-associated proteins by the fluorescence localization-based expression cloning method (Nishimura et al. 2002) and was also called Amot-like 1 (Bratt et al. 2002). To simplify matters, Amot, JEAP/Amot-like 1 and MASCOT/Amot-like 2 will be referred to as Amot, JEAP and MASCOT, respectively, from here on. Among these, Amot was the first to be cloned and JEAP was firstly regarded as a TJ-associated protein in epithelial cells. Then, we refer to this family as Amot/JEAP family, so that we can clearly represent their properties in epithelial cells. The three are highly similar in structure, and contain a coiled-coil motif in their middle portion and a PDZ-binding motif at their extreme C-terminus (see Fig. 2). Consequently, all three might have the ability to bind MUPP1 via their C-terminus. To verify this, a β-gal assay was performed. As shown in Fig. 1B, the C-terminal region of Amot and MASCOT interacted specifically with the PDZ2 and -3 of MUPP1. Interestingly, JEAP also interacted with MUPP1, but solely with PDZ3, and the interaction was rather weak.
Characterization of the interaction between MUPP1 and Amot/JEAP family proteins
To further characterize the interaction between MUPP1 and the Amot/JEAP family, we performed in vitro binding assays. HEK293 cells were transfected with HA-Amot and MUPP1-myc, and the lysates were immunoprecipitated with anti-myc pAb. As shown in Fig. 2A, HA-Amot co-precipitated with MUPP1-myc, confirming their interaction. To reveal the MUPP1-binding region of Amot, several deletion constructs of Amot were generated. Then, it was found that Amot aa1–1124, which lacks the C-terminal PDZ-binding motif, as well as Amot aa1–233, which contains merely the N-terminal extension, did not bind to MUPP1. By contrast, Amot aa234–1127, which lacks the N-terminal region but retains the C-terminal region (corresponds to p80-Amot; see Discussion), effectively bound to MUPP1. These results suggest that the C-terminal PDZ-binding motif of Amot is responsible for the interaction with MUPP1. Similar assays were also performed for JEAP (Fig. 2B) and MASCOT (Fig. 2C). In both cases, the respective full-length constructs clearly interacted with MUPP1, whereas those lacking the C-terminal PDZ-binding motives did not. Thus, all Amot/JEAP family proteins interact with MUPP1 through their C-terminal PDZ-binding motives.
Interaction with Patj
Patj is a close relative of MUPP1; it has an MRE domain and 10 PDZ domains and is localized at TJs in epithelial cells (Lemmers et al. 2002; Roh et al. 2002). Although MUPP1 has not yet been functionally characterized, Patj is recognized as an important regulator of the formation of TJs and epithelial polarity (Michel et al. 2005; Shin et al. 2005). Then, to examine if Patj also has an ability to interact with Amot/JEAP family proteins, co-immunoprecipitation assays were performed. Significantly, Patj also interacted with all the family, and the interaction was dependent on the C-terminal PDZ-binding motives (Fig. 3A), suggesting that Patj has similar molecular characteristics in terms of its ability to bind with Amot/JEAP family proteins. Interestingly, the affinity of Patj for JEAP and MASCOT was apparently higher than that of MUPP1, although that for Amot was almost indistinguishable between MUPP1 and Patj (Fig. 3B).
Examination of the subcellular distribution of endogenous Amot/JEAP family proteins in cultured epithelial cells using specific antibodies
To characterize the endogenous Amot/JEAP family proteins, we tried to obtain specific antibodies for these proteins. To this end, GST-fused forms of the C-terminal region of Amot and MASCOT (aa848–1127, aa671–773 of Amot and MASCOT, respectively) were produced and used as antigens to generate rabbit pAbs. Rat mAb for JEAP was generated using aa808–883 as an antigen as previously described (Nishimura et al. 2002). To verify the specificity of these antibodies, lysates of native MTD1A cells or MTD1A cells over-expressing HA-tagged Amot/JEAP family proteins were blotted with the antibodies. As shown in Fig. 4A, all the antibodies clearly reacted with the respective endogenous proteins; Amot, JEAP and MASCOT were detected as a single band of approximately 130, 105 and 92 kDa, respectively. Importantly, they also reacted with the respective over-expressed proteins but not with the other family members at all, suggesting that they are specific. Thus, we are able to make use of these antibodies as tools for examining the endogenous behavior of the Amot/JEAP family individually and specifically.
We then examined and compared the distribution of endogenous Amot/JEAP family proteins in cultured epithelial cells by indirect immunofluorescence microscopy. When polarized MTD1A cells cultured on Transwell filters were examined by confocal microscopy, Amot was found at ZO-1-positive TJs. In addition, Amot was also detected at apical membranes and in vesicular compartments in the cytoplasm (Fig. 4B,a). Significantly, the distribution of both JEAP and MASCOT was indistinguishable from that of Amot (Fig. 4B,b,c). Thus, all the Amot/JEAP family proteins have similar molecular properties in terms of their subcellular distribution. Interestingly, the localization of Amot and MUPP1 almost completely overlapped (Fig. 4B,d), which suggests endogenous interaction.
Distribution of Amot/JEAP family proteins in mouse tissues
We then examined several mouse tissues. The Amot/JEAP family proteins clearly co-localized with occludin in all the epithelial tissues investigated, suggesting that they are also molecular components of TJs in individual organisms. Interestingly, however, the proteins differed in their tissue distribution. In the pancreas, for example, all Amot/JEAP family proteins were expressed (Fig. 5A), whereas in the liver, Amot and MASCOT were expressed but JEAP was not (Fig. 5B). On the other hand, Amot and JEAP, but not MASCOT, were detected in the salivary gland (Fig. 5C).
Biochemical characterization of Amot/JEAP family proteins
To verify the interaction between endogenous Amot/JEAP family proteins and MUPP1, Amot and MASCOT were immunoprecipitated from lysates of MTD1A cells with respective pAbs and were blotted with anti-MUPP1 mAb. Significantly, both Amot and MASCOT were co-precipitated with MUPP1 (Fig. 6A). Thus, interaction does occur with the endogenous proteins.
Interestingly, it has been reported by Bratt et al. (2005) that Amot is a type of integral membrane protein. According to their topological model, Amot possesses two transmembrane domains near its C-terminus to form an extracellular loop, and its N- and C-terminal regions are intracellular. Also, several computer algorithms imply the existence of hydrophobic, transmembrane regions in the corresponding domain of Amot (data not shown) (Bratt et al. 2005). However, a biochemical evaluation of this hypothesis utilizing endogenous protein has been missing. Moreover, it remains unclear whether this model is also applicable to other Amot/JEAP family proteins. Indeed, in spite of the overall structural similarities among the family, no possible transmembrane domains have been predicted in JEAP and MASCOT (data not shown). Then, to directly address these issues, we performed a fractionation assay. Thus, we isolated bile canaliculi fractions from the mouse liver according to a method previously described, without the use of any detergents (Tsukita & Tsukita 1989). These fractions are rich in TJs and AJs, and are therefore useful for the characterization of TJ/AJ-associated molecules. Then, to determine whether Amot/JEAP family proteins are integral membrane proteins or undercoat-constitutive peripheral membrane proteins, we examined the extractability of these proteins from the bile canaliculi fractions in three distinct extraction buffers (a guanidine–HCl solution, an alkaline solution and a low-salt alkaline solution). As shown in Fig. 6B, irrespective of the extraction buffer used, occludin and E-cadherin, which are transmembrane proteins at TJs and AJs, respectively, were by no means extracted with these treatments. On the other hand, two peripheral membrane proteins at TJs, ZO-1 and MUPP1, were effectively extracted under all conditions. Significantly, both Amot and MASCOT were extracted to the same extent as ZO-1 and MUPP1. It should be noted that we could not examine JEAP in this experiment because it was scarcely expressed in the liver (Fig. 5B) (Nishimura et al. 2002). Thus, these data support not the transmembrane model for the topology of Amot and MASCOT, but rather the peripheral membrane model.
Examination of the functional significance of the interaction between MUPP1/Patj and Amot/JEAP family proteins
Because MUPP1/Patj and Amot/JEAP family proteins not only interact with each other but also show an overall co-localization in epithelial cells, it is possible that the distribution of Amot/JEAP family proteins is dependent on MUPP1/Patj. To address this point, we stably expressed in MTD1A cells HA-Amot aa1–1124, HA-JEAP aa1–880 and HA-MASCOT aa1–770, whose C-terminal PDZ-binding motives were deleted, as well as full-length proteins. As shown in Fig. 7A, the exogenously expressed full-length Amot/JEAP family proteins showed an indistinguishable subcellular distribution from the respective endogenous proteins; they were detected at TJs as well as in apical membranes and the cytoplasm. Surprisingly, this pattern of expression was not affected by the loss of PDZ-binding motives, suggesting that the interaction with MUPP1/Patj is not necessarily required for their proper distribution.
To further address this issue, we examined the effects of over-expression of dominant negative constructs of MUPP1 and Patj. When over-expressed in cultured epithelial cells, the MRE domain of MUPP1 or Patj perturbed the endogenous distribution and/or expression of both MUPP1 and Patj concomitantly (our unpublished observations) (Roh et al. 2002). Significantly, over-expression of either MRE domain of MUPP1 or Patj did not affect the endogenous localization of the Amot/JEAP family proteins (Fig. 7B and data not shown). We suggest that MUPP1/Patj is dispensable for the localization of Amot/JEAP family proteins.
We here identified all Amot/JEAP family proteins as binding partners for MUPP1. The C-terminal PDZ-binding motives of the Amot/JEAP family and both PDZ2 and -3 (for Amot and MASCOT) or just PDZ3 (for JEAP) of MUPP1 were, respectively, responsible for these interactions. Moreover, Patj, a close relative of MUPP1, was also found to bind with Amot/JEAP family proteins. Actually, interaction between Amot and Patj has been reported in a recent paper (Wells et al. 2006). The present study, however, is the first to show that all Amot/JEAP family proteins are binding partners for both MUPP1 and Patj, and the first to identify the PDZ domains responsible. Interestingly, both JEAP and MASCOT bound more tightly to Patj than to MUPP1, although Amot bound similarly to MUPP1 and Patj. Thus, Patj might be more of a physiological binding partner for JEAP and MASCOT than MUPP1.
Several independent reports have indicated that Amot/JEAP family proteins are localized at TJs in epithelial cells (Nishimura et al. 2002; Patrie 2005; Wells et al. 2006). Because of the use of different experimental conditions, however, the endogenous behavior of these proteins had not been closely examined. Moreover, in spite of the molecular similarity among the Amot/JEAP family, the antibodies used in these studies have not necessarily been examined for cross-reactivity among the family. Here, we successfully generated specific antibodies that do not cross-react and examined the endogenous behavior of the Amot/JEAP family in parallel. In cultured epithelial cells, the family members were expressed at TJs and concomitantly in the apical membranes and the cytoplasmic vesicular compartments. This pattern of localization was quite similar to that of MUPP1/Patj, supporting endogenous interaction. Also in mouse tissues, the Amot/JEAP family were all expressed at occludin-positive TJs, but differed in their tissue distribution. Thus, they might have tissue-specific functions at TJs. Our antibodies are powerful tools to further clarify their molecular functions.
Although Amot interacts with MUPP1/Patj in the cytoplasm, it was originally identified as a binding partner for angiostatin, a circulating extracellular inhibitor of angiogenesis (Troyanovsky et al. 2001). Indeed, the inhibitory effect of angiostatin on stimulus-dependent migration or tube formation of cultured aortic endothelial cells (MAE cells) was dependent on the exogenous expression of Amot (Troyanovsky et al. 2001; Bratt et al. 2005). Then, Amot might well be a transmembrane protein, and some in vitro experiments support this possibility (Bratt et al. 2005). In the present analysis, we examined the extractability of Amot/JEAP family proteins from bile canaliculi fractions of mouse liver, and showed that they were significantly extracted from the fractions to the same extent as known peripheral membrane proteins, although no known transmembrane proteins were extracted in these conditions. This result suggests that Amot, as well as MASCOT, is not a transmembrane protein, but rather a peripheral membrane protein. Of course we cannot completely exclude the possibility that some sort of transmembrane protein might be somehow extracted in these conditions. It should be noted, however, that we did not apply any manipulations that could impair the membrane structure itself, and that MASCOT, which does not have any hydrophobic transmembrane regions, was indistinguishable from Amot in terms of its subcellular localization and extractability. Importantly, the transmembrane model was proposed based on experiments with cultured endothelial cells. However, because the bile canaliculi fractions in our analysis were derived from the liver, that is, an epithelial tissue, it is possible that Amot has different molecular properties in endothelial and epithelial cells. Thus, Amot might be inserted into the membrane in endothelial cells but only attached to the membrane in epithelial cells. Further studies will be needed to address this issue.
What is the physiological meaning of the interaction between the Amot/JEAP family and MUPP1/Patj? When exogenously expressed in cultured epithelial cells, all Amot/JEAP family proteins localized to TJs, and the lack of C-terminal PDZ-binding motives did not alter their subcellular distribution. Moreover, over-expression of a dominant negative MUPP1 did not affect the localization of Amot/JEAP family proteins. Thus, it seems that MUPP1 and/or Patj is not necessarily required for the localization of Amot/JEAP family proteins to TJs. It should be noted, however, that an alternatively spliced form of Amot (p80-Amot), which does not possess the N-terminal region of the full-length Amot (p130-Amot), was still present at TJs but was expressed diffusely in the cytoplasm when its C-terminal PDZ-binding motif was deleted (data not shown) (Wells et al. 2006), suggesting that the N-terminal region, together with the C-terminal PDZ-binding motif, is involved in the localization of Amot to TJs. In this regard, it has been reported that p130-Amot, JEAP and MASCOT, but not p80-Amot, can interact with the TJ-associated protein MAGI-1 through their N-terminal regions (Bratt et al. 2005; Patrie 2005). Thus, MAGI-1 and MUPP1/Patj might have redundant roles in the localization of Amot/JEAP family proteins at TJs.
As described above, Amot has been reported to be responsible for endothelial cell migration; over-expression of Amot in cultured endothelial cells promotes cell migration and tube formation, and Amot knockout mice showed severe abnormalities in the visceral endoderm's movements during embryogenesis (Troyanovsky et al. 2001; Levchenko et al. 2003, 2004; Shimono & Behringer 2003; Ernkvist et al. 2006). Significantly, when the C-terminal PDZ-binding motif was deleted, Amot did not promote but rather inhibited cell migration or tube formation (Levchenko et al. 2003, 2004). Although the endothelial functions of MUPP1/Patj have not been addressed to date, it would be important to examine whether MUPP1/Patj are involved in these processes.
The open reading frames (ORFs) of mouse Amot and MASCOT were determined based on sequences obtained from a public database (GenBank accession number NM_153319 and NM_019764, respectively). Because these sequences lacked an N-terminal coding region as compared to their human orthologues, we performed a search of the EST database as well as 5′ RACE, and collectively revealed their ORFs to be 3381 and 2319 bp in length, respectively. The ORF of mouse JEAP (2649 bp) was reported previously (GenBank accession number AX763684) (Nishimura et al. 2002). To yield prey vectors for the yeast two-hybrid analysis, the C-termini of Amot (aa1028–1127), JEAP (aa784–883) and MASCOT (aa674–773) were amplified by PCR and subcloned into pVP16. Sequences for PDZ domains of MUPP1 in bait vectors (pBTM116) were previously described (Hamazaki et al. 2002). Mammalian expression vectors for full-length or the respective deletion constructs of Amot/JEAP family proteins were constructed by inserting these fragments into the multiple cloning site of pCAGGS-HA. The pCAGGS-based expression vectors for MUPP1-myc, Patj-myc, myc-MUPP1-MRE (aa1–142) and myc-Patj-MRE (aa1–138) will be described elsewhere.
Yeast two-hybrid analysis
The methodology of yeast two-hybrid screening was previously described (Hamazaki et al. 2002). LexA-binding domain fusion vectors (pBTM116) of PDZ1, -2 and -3 of MUPP1 were used as bait, and the screening of approximately 2 × 107 clones from a mouse embryonic cDNA library (cloned in pVP16; containing the VP16 transactivating domain, generously provided by Dr J. Behrens (Max-Delbruck Center for Molecular Biology, Berlin)) gave approximately 500 positive colonies. Among them, 10 clones encoded the C-terminal coding region of mouse Amot (prey#95, #132, #146, #199, #281, #452 and #480, aa982–1127 (Amot y1); prey#28, #304 and #429, aa986–1127 (Amot y2)), and 3 clones encoded the C-terminus of mouse MASCOT (prey#104 and #205, aa688–773 (MASCOT y1); prey#481, aa692–773 (MASCOT y2)). Their interactions were further examined by conducting β-galactosidase assays on filters with pVP16 constructs for the C-terminus of Amot/JEAP family proteins and pBTM116 constructs for PDZ1 to -3 of MUPP1.
Antibodies and cells
Rabbit anti-Amot pAb and rabbit anti-MASCOT pAb were raised using GST-fused forms of the C-terminal region of the respective proteins (aa848–1127 of Amot and aa671–773 of MASCOT) as antigens. They were used after affinity purification with the maltose-binding protein fusion construct containing the same antigenic region. Rat anti-JEAP mAb raised against the C-terminus (aa808–882) of JEAP (2E5-1) was previously described (Nishimura et al. 2002). Mouse anti-ZO-1 mAb (T8-754) and rat anti-occludin mAb (MOC37) were generated and characterized previously (Itoh et al. 1991; Saitou et al. 1997). Rat anti-E-cadherin mAb (ECCD2) was provided by Dr M. Takeichi (Center for Developmental Biology, Kobe, Japan). Mouse anti-MUPP1 mAb (clone43; BD Transduction Laboratories, Inc.), rabbit anti-ZO-1 pAb and rabbit anti-occludin pAb (Zymed Laboratories, Inc.), mouse anti-HA mAb (16B12; Covance) and mouse anti-myc pAb (A14; Santa Cruz Biotechnologies, Inc.) were obtained commercially. MTD1A cells and HEK293 cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum.
HEK293 cells were transfected with expression vectors tagged with either myc or HA and then lyzed with lysis buffer containing 20 mm HEPES (pH 7.5), 50 mm NaCl, 2 mm EGTA, 2 mm MgCl2, 10% glycerol, 1% TritonX-100, 2 mm DTT, 1 mm PMSF, 2 µg/mL leupeptin and 0.25% aprotinin. The lysates were mixed with protein A sepharose beads preabsorbed with anti-HA mAb or anti-myc pAb. After extensive washing with the lysis buffer, bound proteins were eluted with SDS-sample buffer. For immunoprecipitating the endogenous proteins, MTD1A cells were lyzed with the lysis buffer and the lysates were incubated with protein A sepharose bound with either anti-Amot pAb or anti-MASCOT pAb and processed as above.
Cells plated on glass coverslips or Transwell filters (Coster) were fixed with either 2% formaldehyde for 15 min at room temperature or with methanol for 5 min at −20 °C. When fixed with formaldehyde, cells were then treated with 0.2% Triton X-100 in PBS for 5 min. Cells were washed 3 times with PBS, and soaked in PBS containing 1% bovine serum albumin. Samples were then incubated with primary antibodies for 1 h in a moist chamber. They were washed three times with PBS and incubated for 30 min with secondary antibodies. After three washes with PBS, they were embedded in Mowiol (Calbiochem). Mouse tissues were cut in pieces and embedded in O.C.T. compound using liquid nitrogen. Frozen sections ∼5 µm thick were cut on a cryostat, mounted on glass slides, air-dried and fixed in 95% ethanol at 4 °C for 30 min followed by 100% acetone at room temperature for 1 min. They were then washed with PBS and processed as above. Specimens were examined using an Olympus BX51 photomicroscope (Olympus) or a Zeiss LSM510 confocal laser-scanning microscope (Carl Zeiss).
Isolation of bile canaliculi fraction and extraction of undercoat proteins
The isolation of bile canaliculi fractions from mouse liver was conducted as previously described (Tsukita & Tsukita 1989). To extract most of the undercoat-constitutive proteins, the isolated fraction was first diluted threefold with solution A (10 mm Tris–HCl (pH 7.5), 150 mm NaCl, and protease inhibitors) and centrifuged at 100 000 g for 10 min at 4 °C. The pellets were then homogenized with either of the following extraction buffers: a guanidine–HCl solution (4 m guanidine, 10 mm Tris–HCl (pH 7.5), 150 mm NaCl, 1 mm EDTA−3Na, and protease inhibitors), an alkaline solution (1 N NaOH) or a low-salt alkaline solution (2 mm Tris–HCl (pH 9.2), 1 mm EGTA, and protease inhibitors). They were kept on ice for 15 min, and centrifuged at 100 000 g for 10 min at 4 °C. This procedure was repeated a total of three times. The supernatants were removed and stored after every extraction, collected and used as the supernatant fractions. The final remnant precipitates were used as the pellet fractions.
We thank all members of our laboratory (Department of Cell Biology, Faculty of Medicine, Kyoto University) for their helpful comments and suggestions. This study was supported in part by a Grant-in-Aid for Cancer Research (to S.T. and M.F.) and a Grant-in-Aid for Scientific Research (A) (to S.T.) from the Ministry of Education, Culture, Sports, Science and Technology of Japan.
- 2004) Setting up a selective barrier at the apical junction complex. Curr. Opin. Cell Biol. 16, 140–145. , & (
- 1993) Assembly of the tight junction: the role of diacylglycerol. J. Cell Biol. 123, 293–302. , , , & (
- 2000) The multi-PDZ domain protein MUPP1 is a cytoplasmic ligand for the membrane-spanning proteoglycan NG2. J. Cell. Biochem. 79, 213–224. , , , , , & (
- 2005) Angiomotin regulates endothelial cell–cell junctions and cell motility. J. Biol. Chem. 280, 34859–34869. , , , , , & (
- 2002) Angiomotin belongs to a novel protein family with conserved coiled-coil and PDZ binding domains. Gene 298, 69–77. , , , , , & (
- 2004) The coxsackievirus and adenovirus receptor interacts with the multi-PDZ domain protein-1 (MUPP-1) within the tight junction. J. Biol. Chem. 279, 48079–48084. , , & (
- 2006) p130-Angiomotin associates to actin and controls endothelial cell shape. FEBS J. 273, 2000–2011. , , , , , , , & (
- 1963) Junctional complexes in various epithelia. J. Cell Biol. 17, 375–412. & (
- 1998) Claudin−1 and −2: novel integral membrane proteins at tight junctions with no sequence similarity to occludin. J. Cell Biol. 141, 1539–1550. , , , & . (
- 1993) Occludin: a novel integral membrane protein localizing at tight junctions. J. Cell Biol. 123, 1777–1788. , , , , , . & (
- 1994) Direct association of occludin with ZO-1 and its possible involvement in the localization of occludin at tight junctions. J. Cell Biol. 127, 1617–1626. , , , , , . & (
- 2000) MAGUK proteins: structure and role in the tight junction. Semin. Cell Dev. Biol. 11, 315–324. , & (
- 1991) Identification of a 160-kDa polypeptide that binds to the tight junction protein ZO-1. Proc. Natl. Acad. Sci. USA 88, 3460–3466. , & (
- 2002) Multi-PDZ domain protein 1 (MUPP1) is concentrated at tight junctions through its possible interaction with claudin-1 and junctional adhesion molecule. J. Biol. Chem. 277, 455–461. , , , & . (
- 1998) ZO-3, a novel member of the MAGUK protein family found at the tight junction, interacts with ZO-1 and occludin. J. Cell Biol. 141, 199–208. , , , & (
- 1999) Localization of membrane-associated guanylate kinase (MAGI)-1/BAI-associated protein (BAP) 1 at tight junctions of epithelial cells. Oncogene 18, 7810–7815. , , , , , , , , , & (
- 1999a) Direct binding of three tight junction-associated MAGUKs, ZO-1, ZO-2, and ZO-3, with the COOH termini of claudins. J. Cell Biol. 147, 1351–1363. , , , , & . (
- 1999b) Characterization of ZO-2 as a MAGUK family member associated with tight as well as adherens junctions with a binding affinity to occludin and α catenin. J. Biol. Chem. 274, 5981–5986. , & . (
- 1991) A 220-kD undercoat-constitutive protein: its specific localization at cadherin-based cell-cell adhesion sites. J. Cell Biol. 115, 1449–1462. , , , . & (
- 2000) Multi-PDZ domain protein MUPP1 is a cellular target for both adenovirus ORF1 and high-risk papillomavirus type 18, E6 oncoproteins. J. Virol. 74, 9680–9693. , , , & (
- 2002) hINAD/PATJ, a homolog of Discs Lost, interacts with Crumbs and localizes to tight junction in human epithelial cells. J. Biol. Chem. 277, 25408–25415. , , , , & (
- 2003) Loss of responsiveness to chemotactic factors by deletion of the C–terminal protein interaction site of angiomotin. J. Cell Sci. 116, 3803–3810. , , , & (
- 2004) Angiomotin expression promotes hemangioendothelioma invasion. Oncogene 23, 1469–1473. , , & (
- 2000) The direct association of the multiple PDZ domain containing proteins (MUPP-1) with the human c-Kit C-terminus is regulated by tyrosine kinase activity. FEBS Lett. 482, 54–58. , , , & (
- 1998) Junctional adhesion molecule, a novel member of the immunoglobulin superfamily that distributes at intercellular junctions and modulates monocyte transmigration. J. Cell Biol. 142, 117–127. , , , , , , , , , , & (
- 2003) Signalling to and from tight junctions. Nat. Rev. Mol. Cell Biol. 4, 225–236. & (
- 2005) PATJ connects and stabilizes apical and lateral components of tight junctions in human intestinal cells. J. Cell Sci. 118, 4049–4057. , , , , & (
- 2002) JEAP, a novel component of tight junction in exocrine cells. J. Biol. Chem. 277, 5583–5587. , , , , , , & (
- 2005) Identification and characterization of a novel tight junction-associated family of proteins that interacts with a WW domain of MAGI-1. Biochim. Biophys. Acta 1745, 131–144. (
- 2004) Crumbs homologue 1 is required for maintenance of photoreceptor cell polarization and adhesion during light exposure. J. Cell Sci. 117, 4169–4177. , , , , , , , , , , & (
- 1997) PDZ domains: targeting signaling molecules to sub-membranous sites. Bioessays 19, 469–479. , , & (
- 1997) PDZ domain proteins: scaffolds for signaling complexes. Curr. Biol. 7, R770–R773. & (
- 2002) The MAGUK protein, Pals1, functions as an adapter, linking mammalian homologues of Crumbs and Discs Lost. J. Cell Biol. 157, 161–172. , , , , , , , & (
- 1997) Mammalian occludin in epithelial cells: its expression and subcellular distribution. Eur. J. Cell Biol. 73, 222–231. , , , , , & . (
- 2004) The tight junction: a multifunctional complex. Am. J. Physiol. Cell Physiol. 286, 1213–1228. & (
- 2003) Angiomotin regulates visceral endoderm movements during mouse embryogenesis. Curr. Biol. 13, 613–617. & (
- 2005) PATJ regulates tight junction formation and polarity in mammalian epithelial cells. J. Cell Biol. 168, 705–711. , & (
- 1969) Further observations of the fine structure of freeze-cleaved tight junctions. J. Cell Sci. 13, 763–786. (
- 1986) Identification of ZO-1: a high molecular weight polypeptide associated with the tight junction (zonula occludens) in a variety of epithelia. J. Cell Biol. 103, 755–766. , , & (
- 2001) Angiomotin: an angiostatin binding protein that regulates endothelial cell migration and tube formation. J. Cell Biol. 152, 1247–1254. , , , & (
- 2001) Multifunctional strands in tight junctions. Nat. Rev. Mol. Cell Biol. 2, 285–293. , & (
- 1989) Isolation of cell-to-cell adherens junctions from rat liver. J. Cell Biol. 108, 31–41. . & (
- 1998) Cloning and characterization of MUPP1, a novel PDZ domain protein. FEBS Lett. 424, 63–68. , , & (
- 2006) A Rich/Amot complex regulates the Cdc42 GTPase and apical-polarity proteins in epithelial cells. Cell 125, 535–548. , , , , , , , , , , , , & (
- 1999) Protein interaction at the tight junction. J. Biol. Chem. 274, 35179–35185. , & (