A distinct PAR complex associates physically with VE-cadherin in vertebrate endothelial cells

A cell polarity complex consisting of partitioning defective 3 (PAR-3), atypical protein kinase C (aPKC) and PAR-6 has a central role in the development of cell polarity in epithelial cells. In vertebrate epithelial cells, this complex localizes to tight junctions. Here, we provide evidence for the existence of a distinct PAR protein complex in endothelial cells. Both PAR-3 and PAR-6 associate directly with the adherens junction protein vascular endothelial cadherin (VE-cadherin). This association is direct and mediated through non-overlapping domains in VE-cadherin. PAR-3 and PAR-6 are recruited independently to cell-cell contacts. Surprisingly, the VE-cadherin-associated PAR protein complex lacks aPKC. Ectopic expression of VE-cadherin in epithelial cells affects tight junction formation. Our findings suggest that in endothelial cells, another PAR protein complex exists that localizes to adherens junctions and does not promote cellular polarization through aPKC activity. They also point to a direct role of a cadherin in the regulation of cell polarity in vertebrates.     

细胞极性蛋白复合物PAR/aPKC的研究进展

不对称细胞分裂是动物发育过程中用以调控细胞分化的一种进化上保守的基本模式。极性的祖细胞通过不对称分裂产生两个不同命运的子细胞,这一过程涉及细胞命运决定因子的不对称分布、纺锤体的旋转定位等,而这些过程都必须依赖特定细胞极性的存在才能得以正常进行。简要综述了高度保守的蛋白复合物PAR／aPKC在细胞极性建立和维持中的重要作用,以及它如何调控纺锤体定位和命运决定因子不对称分配,并讨论了在该领域的一些新发现和研究进展。     

Thrombin responses in human endothelial cells. Contributions from receptors other than PAR1 include the transactivation of PAR2 by thrombin-cleaved PAR1

The recent identification of two new thrombin receptors, PAR3 and PAR4, led us to re-examine the basis for endothelial cell responses to thrombin. Human umbilical vein endothelial cells (HUVEC) are known to express PAR1 and the trypsin/tryptase receptor, PAR2. Northern blots detected both of those receptors and, to a lesser extent, PAR3, but PAR4 message was undetectable and there was no response to PAR4 agonist peptides. To determine whether PAR3 or any other receptor contributes to thrombin signaling in HUVEC, PAR1 cleavage was blocked with two selective antibodies and PAR1 activation was inhibited with the antagonist, BMS200261. The antibodies completely inhibited HUVEC responses to thrombin, but BMS200261 was only partly effective, even though separate studies established that the antagonist completely inhibits PAR1 signaling at the concentrations used. Since peptides mimicking the PAR1 tethered ligand domain can also activate PAR2, we asked whether the remaining thrombin response in the presence of the antagonist could be due in part to the intermolecular transactivation of PAR2 by cleaved PAR1. Evidence that transactivation can occur was obtained in COS-7 cells co-expressing PAR2 and a variant of PAR1 that can be cleaved, but not signal. There was a substantial response to thrombin only in cells expressing both receptors. Conversely, in HUVEC, complete blockade of the thrombin response by the PAR1 antagonist occurred only when signaling through PAR2 was also blocked. From these observations we conclude that 1) PAR1 is the predominant thrombin receptor expressed in HUVEC and cleavage of PAR1 is required for endothelial cell responses to thrombin; 2) although PAR3 may be expressed, there is still no evidence that it mediates thrombin responses; 3) PAR4 is not expressed on HUVEC; and 4) transactivation of PAR2 by cleaved PAR1 can contribute to endothelial cell responses to thrombin, particularly when signaling through PAR1 is blocked. Such transactivation may limit the effectiveness of PAR1 antagonists, which compete with the tethered ligand domain rather than preventing PAR1 cleavage.     

aPKC, Crumbs3 and Lgl2 control apicobasal polarity in early vertebrate development

In early vertebrate development, apicobasally polarised blastomeres divide to produce inner non-polarised cells and outer polarised cells that follow different fates. How the polarity of these early blastomeres is established is not known. We have examined the role of Crumbs3, Lgl2 and the apical aPKC in the polarisation of frog blastomeres. Lgl2 localises to the basolateral membrane of blastomeres, while Crumbs3 localises to the apical and basolateral membranes. Overexpression aPKC and Crumbs3 expands the apical domain at the expense of the basolateral and repositions tight junctions in the new apical-basolateral interface. Loss of aPKC function causes loss of apical markers and redirects basolateral markers ectopically to the apical membrane. Cell polarity and tight junctions, but not cell adhesion, are lost and outer polarised cells become inner-like apolar cells. Overexpression of Xenopus Lgl2 phenocopies the aPKC knockout, suggesting that Lgl2 and aPKC act antagonistically. This was confirmed by showing that aPKC and Lgl2 can inhibit the localisation of each other and that Lgl2 rescues the apicalisation caused by aPKC. We conclude that an instrumental antagonistic interaction between aPKC and Lgl2 defines apicobasal polarity in early vertebrate development.     

New functions for a vertebrate Rho guanine nucleotide exchange factor in ciliated epithelia

Human ARHGEF11, a PDZ-domain-containing Rho guanine nucleotide exchange factor (RhoGEF), has been studied primarily in tissue culture, where it exhibits transforming ability, associates with and modulates the actin cytoskeleton, regulates neurite outgrowth, and mediates activation of Rho in response to stimulation by activated Galpha12/13 or Plexin B1. The fruit fly homolog, RhoGEF2, interacts with heterotrimeric G protein subunits to activate Rho, associates with microtubules, and is required during gastrulation for cell shape changes that mediate epithelial folding. Here, we report functional characterization of a zebrafish homolog of ARHGEF11 that is expressed ubiquitously at blastula and gastrula stages and is enriched in neural tissues and the pronephros during later embryogenesis. Similar to its human homolog, zebrafish Arhgef11 stimulated actin stress fiber formation in cultured cells, whereas overexpression in the embryo of either the zebrafish or human protein impaired gastrulation movements. Loss-of-function experiments utilizing a chromosomal deletion that encompasses the arhgef11 locus, and antisense morpholino oligonucleotides designed to block either translation or splicing, produced embryos with ventrally-curved axes and a number of other phenotypes associated with ciliated epithelia. Arhgef11-deficient embryos often exhibited altered expression of laterality markers, enlarged brain ventricles, kidney cysts, and an excess number of otoliths in the otic vesicles. Although cilia formed and were motile in these embryos, polarized distribution of F-actin and Na(+)/K(+)-ATPase in the pronephric ducts was disturbed. Our studies in zebrafish embryos have identified new, essential roles for this RhoGEF in ciliated epithelia during vertebrate development.     

Regulated shedding of PAR1 N-terminal exodomain from endothelial cells

G protein-coupled receptors can trigger metalloproteinase-dependent shedding of proteins from the cell surface. We now report that G protein-coupled receptors can themselves undergo regulated metalloproteinase-dependent shedding. The N-terminal exodomain of protease-activated receptor-1 (PAR1), a G protein-coupled receptor for thrombin, displayed regulated shedding in endothelial cells, which normally express this receptor. Cleavage occurred at a site predicted to render the receptor unresponsive to thrombin. A chimeric protein in which the N-terminal exodomain of PAR1 was fused to an unrelated transmembrane segment was shed as efficiently as PAR1, shedding of both proteins was stimulated by phorbol ester and by a PAR1 agonist. TNFalpha protease inhibitor-2 (TAPI-2), phenanthroline, and tissue inhibitor of metalloproteinase-3 (TIMP-3) but not TIMP-1 or -2 inhibited such shedding. These and other data suggest that the information that specifies PAR1 shedding resides within its N-terminal exodomain rather than its heptahelical segment, that activation of protein kinase C or of PAR1 itself can stimulate PAR1 shedding in trans, and that ADAM17/TACE or a metalloproteinase with similar properties mediates PAR1 shedding. Regulated shedding reduced the amount of cell surface PAR1 available for productive cleavage by thrombin by half or more, but thus far we have been unable to demonstrate an effect of PAR1 shedding on cellular responsiveness to thrombin. Nonetheless, regulated shedding of G protein-coupled receptors represents a new mechanism by which signaling by this important class of receptors might be modulated.     

Developmental potency of mouse primitive ectoderm cells, embryonic ectoderm cells and primordial germ cells after blastocyst injection

Developmental potency of primitive and embryonic ectoderm cells from 4.50-day to 6.25-day post-coitum (p.c.)mouse embryos and primordial germ cells from 12.50-day p.c. male genital ridges of fetal mice were studied by direct introducing them into 3.50-day p.c. blastocysts. Sixteen (61.5%) overt chimaeras out of 26(50%) offsprings were obtained after transfer of 52 blastocysts injected with 4.50-day primitive ectoderm cells; four (16.0%) overt chimaeras were obtained out of 25 (51.0%) offsprings with 4.75-day primitive ectoderm cells from 49 transferred blastocysts. However, no overt chimaera was obtained with either 5.25-day or 6.25day embryonic ectoderm cells or 12.50-day male primordial germ cells. GPI analysis of mid-gestation conceptuses developed from injected blastocysts showed that 5.25-day embryonic ectoderm cells could only contributed to yolk sac of conceptus. Results suggested that implantation acts as a trigger for the determination of primitive ectoderm cells, and their developmental potency becomes limited within a short period of time in normal development.     

Immunodetection of annexins 1 and 2 in ciliated cells from quail oviduct.

Annexins 1 and 2 are Ca(2+)-binding proteins related to the cytoskeletal proteins which have been reported to bind in a calcium-dependent manner of F-actin and phospholipids in vitro. Proteins immunologically related to the brain 37-kDa annexin 1 and 36-kDa annexin 2 were characterized by immunoblotting epithelial ciliated cells from quail oviduct. They were detected by immunofluorescence in ciliated as well as glandular cells, using antisera and purified antibodies directed against pig brain annexins. The pattern of labeling was found in the apical part of both cell types, with close membrane association. However, a wider distribution was observed in mature ciliated cells: annexins were localized in the well developed cytoskeletal meshwork in which the ciliary apparatus is tightly anchored. After immunogold labeling, annexins 1 and 2 were located in the same area as spectrin 240/235 and at the connection sites of F-actin; both these cytoskeletals proteins were associated with the appendages of the basal body. In contrast, annexins were not detected in immature epithelial cells, while actin and spectrin were present. During ciliogenesis, the staining gradually appeared associated with the lateral and apical membranes. In this cellular model, the annexins may function during exocytosis in gland epithelial cells, where a close cytoskeleton-membrane association is observed; moreover, in ciliated cells, a relationship between cytoskeletal elements of the terminal web and annexins may exist.     

Protease-activated receptor-1 (PAR1) and PAR2 but not PAR4 mediate relaxations in lower esophageal sphincter

Protease-activated receptor-1 (PAR1), PAR2 and PAR4 activation can alter the gastrointestinal motility. To investigate effects mediated by PARs in the lower esophageal sphincter, we measured contraction or relaxation of transverse strips from the guinea-pig lower esophageal sphincter caused by PAR1 (TFLLR-NH2 and SFLLRN-NH2), PAR2 (SLIGKV-NH2 and SLIGRL-NH2) and PAR4 peptide agonists (GYPGKF-NH2, GYPGQV-NH2 and AYPGKF-NH2) as well as PAR protease activators (thrombin and trypsin). In resting lower esophageal sphincter strips, TFLLR-NH2 and SFLLRN-NH2 caused moderate concentration-dependent relaxation whereas thrombin did not cause any relaxation or contraction. Furthermore, in carbachol-contracted strips, TFLLR-NH2 and SFLLRN-NH2 caused marked whereas thrombin caused mild concentration-dependent relaxation. These indicate the existence of PAR1 mediating relaxation. Similarly, in resting lower esophageal sphincter strips, trypsin caused moderate concentration-dependent relaxation whereas SLIGRL-NH2 and SLIGKV-NH2 did not cause any relaxation or contraction. In addition, in carbachol-contracted strips, trypsin caused marked whereas SLIGRL-NH2 and SLIGKV-NH2 caused mild concentration-dependent relaxation. These indicate the existence of PAR2 mediating relaxation. The relaxant response of thrombin, TFLLR-NH2, trypsin and SLIGKV-NH2 was insensitive to atropine or tetrodotoxin, suggesting a direct effect. The relaxant response of trypsin was not affected by apamin, charybdotoxin, indomethacin and capsaicin but was attenuated by NG-nitro-L-arginine methyl ester, indicating involvement of NO. FSLLR-NH2, a PAR1 control peptide, and VKGILS-NH2, a PAR2 control peptide, as well as all three PAR4 peptide agonists, GYPGKF-NH2, GYPGQV-NH2 and AYPGKF-NH2, did not cause any relaxation or contraction. Taken together, these results demonstrate that PAR1 and PAR2 but not PAR4 mediate relaxations in the guinea-pig lower esophageal sphincter.     

TRPC1 forms the stretch-activated cation channel in vertebrate cells

The mechanosensitive cation channel (MscCa) transduces membrane stretch into cation (Na(+), K(+), Ca(2+) and Mg(2+)) flux across the cell membrane, and is implicated in cell-volume regulation, cell locomotion, muscle dystrophy and cardiac arrhythmias. However, the membrane protein(s) that form the MscCa in vertebrates remain unknown. Here, we use an identification strategy that is based on detergent solubilization of frog oocyte membrane proteins, followed by liposome reconstitution and evaluation by patch-clamp. The oocyte was chosen because it expresses the prototypical MscCa (>or=10(7)MscCa/oocyte) that is preserved in cytoskeleton-deficient membrane vesicles. We identified a membrane-protein fraction that reconstituted high MscCa activity and showed an abundance of a protein that had a relative molecular mass of 80,000 (M(r) 80K). This protein was identified, by immunological techniques, as the canonical transient receptor potential channel 1 (TRPC1). Heterologous expression of the human TRPC1 resulted in a >1,000% increase in MscCa patch density, whereas injection of a TRPC1-specific antisense RNA abolished endogenous MscCa activity. Transfection of human TRPC1 into CHO-K1 cells also significantly increased MscCa expression. These observations indicate that TRPC1 is a component of the vertebrate MscCa, which is gated by tension developed in the lipid bilayer, as is the case in various prokaryotic mechanosensitive (Ms) channels.     

Expression of ASIC2 in ciliated cells and stereociliated cells

Acid-sensing ion channel 2 (ASIC2) plays a role as a mechanorecptor and acid receptor in the peripheral and central nervous systems. However, several recent studies have suggested that ASIC2 is expressed in several organs, in addition to the nervous system. We have examined the expression and distribution of ASIC2 in rat ciliated cells (trachea and oviduct) and stereociliated cells (epididymis, Corti organ, and ampullary crest) by immunohistochemistry and transmission electron microscopy (TEM). Immunohistochemistry revealed that ASIC2 was expressed in both ciliated cells and stereociliated cells, but the localization differed between these cell types. In ciliated cells, ASIC2 was coexpressed with a cilial marker (acetylated tubulin). In stereociliated cells stained with a stereocilial marker (phalloidin), ASIC2 was observed in the cell body. Observation by TEM suggested that ASIC2 expression was present at the apical side of the cilial membrane in ciliated cells and at the apical side of the cell body in stereociliated cells. This study thus indicates that the proton receptor ASIC2 is expressed in both ciliated and stereociliated cells.     

Immuno-electron-microscopic localization of a centriole-related antigen in ciliated cells

Summary An antigen common to purported centriolar and basal body regions of a variety of cell types was previously visualized by immuno-fluorescence microscopy. The present study demonstrates the localization of the antigen relative to the defined basal body structures of ciliated tracheal cells at the electron-microscopic level. After ethyldimethylaminopropyl carbodiimide-glutaraldehyde-saponin (EGS) fixation and permeabilization, immunoferritin labeling is consistently found associated with amorphous electron-opaque material in proximity to basal bodies and their ciliary rootlets, but not with basal body microtubules themselves. This distribution pattern is distinct from that of other proteins found in the apical region of ciliated cells, such as calmodulin. It is proposed that the dense material may be analogous to pericentriolar material of centrosomes.     

Early morphogenesis of ciliated cells in human oral cavity

Ciliated cells were found in the epithelium of the oral cavity of human embryos and fetuses starting from the seventh week of prenatal development. At the early stages of prenatal development (until the 13th week), cells with cilia cover most of the dorsal surface of the tongue and the soft palate, whereas they are found only near the gland ducts in the circumvallate and foliate lingual papillae after 17 weeks of development. The ultrastructure of the axoneme of cilia corresponds to the structure of motile cilia and is represented by nine microtubule doublets that surround the central pair of microtubule singlets. An immunohistochemical study performed on weeks 10–12 of development identified nerve endings associated with the ciliated cells. Until the 14th week of development, the cytoplasm of ciliated cells is immunopositive for NSE. The spatial distribution of ciliated cells in the tongue epithelium until the 13th week of development is not related to the morphogenesis of lingual papillae, and their role in the human oral cavity during the first trimester of pregnancy is unclear and requires further study.