The anatomy of the nervous system of the hydrozoan jellyfish, Polyorchis penicillatus, as revealed by a monoclonal antibody.

Dissociated cells from the margin and tentacles of the hydromedusa Polyorchis penicillatus were centrifuged in a Percoll gradient to remove cnidocytes. The resulting formaldehyde-fixed cells were used to inoculate mice to produce monoclonal antibodies. One of the hybridomas, which secreted antibodies against all neurons, was cloned and designated as mAb 5C6. Immunohistochemical labelling with mAb 5C6 of whole-mount preparations and paraffin sections provided a far more complete picture of the organisation of the hydromedusan nervous system than was previously available when using neuronal labelling techniques that restrict labelling to certain neuronal types. Besides confirming anatomical features described in earlier studies these techniques allowed us to discover a number of new structures and to determine connections that were only suspected. Such findings included:1. The discovery of an arch-like connection between the swimming motor neuron network at the apices of the subumbrellar muscle sheets 2. An orthogonal network connecting each pair of radial nerves in each radius 3. Continuity of a central branch of the radial nerve with the radial innervation of the manubrium 4. Details of the sensory neuronal contribution to the microanatomy of the ocelli and cnidocyte batteries 5. Presence of specialised receptor cells in the margin at the bases of tentacles 6. Neurons apparently innervating the radial muscles of the velum 7. Isolated neurons in the peduncle and gonads

Genomic organization of a voltage-gated Na+ channel in a hydrozoan jellyfish: insights into the evolution of voltage-gated Na+ channel genes.

Voltage-gated Na+ channels are responsible for fast propagating action potentials. The structurally simplest animals known to contain rapid, transient, voltage-gated currents carried exclusively by Na+ ions are the Cnidaria. The Cnidaria are thought to be close to the origin of the metazoan radiation and thus are pivotal organisms for studying the evolution of the Na+ channel gene. Here we describe the genomic organization of the Na+ channel alpha subunit, PpSCN1, from the hydrozoan jellyfish, Polyorchis penicillatus. We show that most of the 20 intron sites in this diploblast are conserved in mammalian Na+ channel genes, with some even shared by Ca2+ channels. One of these conserved introns is spliced by a rare U 12-type spliceosome. Such conservation places the origin of the primary exon arrangement of Na+ channels and different intron splicing mechanisms to at least the common ancestors of diploblasts and triploblasts, approximately 600 million-1 billion years ago.

Evolution of the "classical" cadherin family of cell adhesion molecules in vertebrates.

The cadherins are major mediators of calcium-dependent cell-cell adhesion and are also involved in cell signaling pathways during development. The classical cadherins, which are the definitive group of the cadherin superfamily, are transmembrane proteins that consist of an extracellular domain of five cadherin repeats, including an HAV tripeptide conserved in one binding surface within the first domain, and a highly conserved cytoplasmic domain that interacts with the actin cytoskeleton via the catenin proteins. These cadherins play major roles in vertebrate morphogenesis; they are expressed widely throughout development, antibodies to specific cadherins perturb a variety of developmental processes, and many gene knockouts are lethal at early stages of development. Phylogenetic analysis of the "classical" cadherins shows that in the vertebrates there are four paralog families. The rate of evolutionary change is radically different between the different paralogs, indicating that there are significantly different selection pressures on the functions of the various cadherins, both between the different paralogs in a single organism lineage and between different organism lineages within a single paralog family. There is also evidence for gene conversion between the E-cadherin and P-cadherin paralogs in Gallus gallus and possibly Xenopus laevis, but not between the same paralogs in the mammalian lineages. A scheme for the origin of the paralogs within the vertebrate lineage based on these analyses indicates that the presence of the four paralog families is a characteristic of vertebrates and that variation of cadherin structure and function is a significant factor in morphological evolution of vertebrates.

L-CAM expression induces fibroblast-epidermoid transition in squamous carcinoma cells and down-regulates the endogenous N-cadherin.

SCC9 cells, derived from a squamous carcinoma of the tongue, were shown to lack E-cadherin but express alpha- and beta-catenins and N-cadherin. These cells also lack plakoglobin expression, do not assemble desmosomes and exhibit the typical morphology and growth properties of transformed cells. The N-cadherin expressed in SCC9 cells has properties similar to other classical cadherins, including interactions with the catenins. We transfected SCC9 cells with a full-length cDNA for L-CAM (liver cell adhesion molecule), the functional chicken homologue of E-cadherin. The exogenously expressed L-CAM formed complexes with catenins and the cytoskeleton and induced a morphological transition from fibroblastoid to epithelioid, conferred density-dependent growth inhibition, increased aggregation ability, and increased synthesis and stability of alpha- and beta-catenins. Coincident with these phenotypic changes, we detected a significant reduction in the level of endogenous N-cadherin, primarily as a result of rapid degradation of this protein in L-CAM-expressing cells. These results show the abnormal expression of N-cadherin in these transformed epidermoid cells, demonstrate the dynamics of the relationship between two cadherins, and provide a model system for the functional analysis of the tumor suppressor activity of E-cadherin in carcinomas.

A putative voltage-gated sodium channel alpha subunit (PpSCN1) from the hydrozoan jellyfish, Polyorchis penicillatus: structural comparisons and evolutionary considerations.

Extant cnidarians are probably the simplest metazoans with discrete nervous systems and rapid, transient voltage-gated currents carried exclusively by Na+ ions. Thus cnidarians are pivotal organisms for studying the evolution of voltage-gated Na+ channels. We have isolated a full-length Na+ channel alpha subunit cDNA (PpSCN1) from the hydrozoan jellyfish, Polyorchis penicillatus, that has one of the smallest known coding regions of a four domain Na+ channel (1695 amino acids). Homologous residues that have a critical bearing on the selectivity filter, voltage-sensor and binding sites for tetrodotoxin and lidocaine in vertebrates and most invertebrates differ in cnidarians. PpSCN1 is not alternatively-spliced and may be the only pore-forming alpha subunit available to account for at least three electrophysiologically distinct Na+ currents that have been studied in P. penicillatus.

Burn healing in organ cultures of embryonic chicken skin: a model system.

Burn wounds cause complex damage to the skin. Unlike simple cuts, burns cause a graded damage at the margin of the wound and leave biochemically complex debris in the wound, two factors that complicate the process of wound healing. To develop an in vitro system for studying cellular responses to burns in whole tissue, we have applied a standardized burn wound to organ cultures of embryonic chicken skin and evaluated cellular changes in response to the burn damage. This simplified system is not subject to uncontrolled infection and does not involve angiogenesis and granulation. Thus, these cultures provide a simplified model of how cells of the dermis and epidermis in and around a burn wound respond to heat damage early in the process of healing.

Development and maintenance of bile canaliculi in vitro and in vivo.

The apical surfaces of hepatocytes are specialized to form the boundaries of the bile canaliculi. The canaliculi function to secrete and concentrate components of the bile and to transport the bile out of the interior of the hepatic parenchymal tissue to the epithelium-lined bile ducts. Failure of the canaliculi to form and function properly can lead to biliary stasis or release of bile components into the bloodstream, both potentially life-threatening situations. Experimental analysis of canaliculus development and function has been undertaken in a number of experimental systems, ranging in complexity from intact animals to isolated hepatocyte cell cultures. These approaches each have inherent advantages and disadvantages for studying the various aspects of canaliculus development and function. This article summarizes what is known about how the functional components of the canaliculus develop and the directions that current experimental approaches are leading in analyzing this process. Studies of model epithelial systems have begun to define how interactions between components of the cytoskeleton and plasma membrane regulate the structure of polarized plasma membranes. These results are also discussed in terms of the bile canaliculus.

Voltage sensing in jellyfish Shaker K+ channels.

The S4 segment of the jellyfish (Polyorchis penicillatus) Shaker channel jShak1 contains only six positively charged motifs. All other Shaker channels, including the jellyfish Shaker channel jShak2, have seven charges in this segment. Despite their charge differences, both these jellyfish channels produce currents with activation and inactivation curves shifted by approximately +40 mV relative to other Shaker currents. Adding charge without changing segment length by mutating the N-terminal side of jShak1 S4 does not have a pronounced effect on channel activation properties. Adding the positively charged motif RIF on the N-terminal side of K294 (the homologue of K374 in Drosophila Shaker, which is a structurally critical residue) produced a large positive shift in both activation and inactivation without altering the slope of the activation curve of the channel. When IFR was added to the other side of K294, there was a small negative shift in activation and fast inactivation of the channel was prevented. Our results demonstrate that K294 divides the S4 segment into functionally different regions and that the voltage threshold for activation and inactivation of the channel is not determined by the total charge on S4.

The tight junction protein ZO-2 contains three PDZ (PSD-95/Discs-Large/ZO-1) domains and an alternatively spliced region.

The complete cDNA sequence for canine ZO-2, a tight junction-specific protein, is presented. A single open reading frame encodes a polypeptide of 1,174 amino acids with a predicted molecular mass of 132,085 daltons. As noted previously (), ZO-2 is a member of the membrane-associated guanylate kinase-containing (MAGUK) protein family, a family which includes an additional tight junction-associated protein, ZO-1. These proteins contain a region homologous to guanylate kinase, an SH3 domain, and variable numbers of PSD-95/discs-large/ZO-1 (PDZ) domains, shown to be involved in protein-protein interactions. ZO-2 and ZO-1 contain three PDZ domains in the N-terminal half of the molecule. Between the first and second PDZ domains, ZO-2 displays a basic region (pI = 10.27) containing 22% arginine residues. Both ZO-1 and ZO-2 have proline-rich C-terminal regions that are not homologous to other MAGUK family members. Sequence analysis of multiple ZO-2 cDNAs reveals a 36-amino acid domain in this C-terminal region present in only some of the cDNAs. Overall, ZO-2 is highly homologous to ZO-1, showing 51% amino acid identity; however, the C-terminal ends of the molecules show only 25% amino acid identity. This suggests that the C-terminal ends of ZO-1 and ZO-2 have different functions.

Multiple Shaker potassium channels in a primitive metazoan.

Voltage-gated potassium channels are critical elements in providing functional diversity in nervous systems. The diversity of voltage-gated K+ channels in modern triploblastic metazoans (such as mollusks, arthropods and vertebrates) is provided primarily by four gene subfamilies (Shaker, Shal, Shab, and Shaw), but there has been no data from the ancient diploblastic metazoans until now. Diploblasts, represented by jellyfish and other coelenterates, arose during the first major metazoan radiation and are the most structurally primitive animals to have true nervous systems. By comparing the K+ channels of diploblasts and triploblasts, we may determine the fundamental set of K+ channels present in the first nervous systems. We now report the isolation of two Shaker subfamily cDNA clones, jShak1 and jShak2, from the hydrozoan jellyfish Polyorchis penicillatus (Phylum Cnidaria). JShak1 and jShak2 express transient outward currents in Xenopus oocytes most similar to Shaker currents from Drosophila in their rates of inactivation and recovery from inactivation. The finding of multiple Shaker subfamily genes is significant in that multiple Shaker genes also exist in mammals. In Drosophila, multiple Shaker channels are also produced, but by a mechanism of alternative splicing. Thus, the Shaker K+ channel subfamily had an established functional identity prior to the first major radiation of metazoans, and multiple forms of Shaker channels have been independently selected for in a wide range of metazoans.

Calcium regulation of gene expression in neurons: the mode of entry matters.

Ca2+ entry into neurons is one of the major effectors of stimulus-induced physiological change. Ca2+ can enter neurons through a number of different voltage-gated and ligand-gated channels. Depending on the route of entry, Ca2+ stimulates distinct intracellular signaling pathways, which activate different sets of genes, resulting in alternative physiological outcomes for the cell. These recent results suggest that the specific effect of a single biochemical second messenger can vary as a consequence of its route of entry into the cell.

Effects of fetal calf serum and disruption of cadherin function on the formation of bile canaliculi between hepatocytes.

The polarization of hepatocytes to form a connected network of bile canaliculi (BC) is necessary for the function of the liver. Hepatocyte polarization may be controlled by soluble factors and/or physical interactions between cells. Monolayer cultures of embryonic chicken hepatocytes in DMEM supplemented with ornithine, dexamethasone, and insulin express BC-specific antigens for at least 7 days. However, BC-specific antigen expression is lost within 3 days of culture initiation in DMEM containing 10% fetal calf serum. The dedifferentiating effects of fetal calf serum (FCS) can be reversed. Furthermore, cultures in medium containing ornithine, dexamethasone, insulin, and 10% FCS appear identical to cultures grown in 10% FCS alone. Thus FCS contains a soluble inhibitor of hepatocyte polarization. Aggregate cultures grown in suspension maintain hepatocyte polarization for 10-12 days. This may be due to the increased cell-cell contact between hepatocytes in aggregate culture or to more normal contact with the extracellular matrix. We have evaluated the role of cadherin-mediated interactions on hepatocyte polarization. Anti-E-cadherin Fab' fragments disrupted the formation of long networks of BC in monolayer cultures but did not stop polarized expression of BC-specific antigens. The BC antigens in anti-E-cadherin-treated cells were concentrated in small areas between cells and were present at lower levels uniformly on the cell surface. These results indicate that E-cadherin is required for the formation of extended BC networks, but that other factors are responsible for maintaining the synthesis and localization of BC-specific antigens.

Toxic effects of beta-aminopropionitrile treatment on developing chicken skin.

beta-aminopropionitrile (BAPN), a compound that inhibits crosslinking of collagen, has been widely used to examine the function of collagen in developing tissues. We used BAPN to examine the function of collagen during the formation and patterning of feather rudiments in embryonic chicken skin. Organ cultures of skin were treated with different concentrations of BAPN and examined for changes in morphology, histology, and extent of collagen crosslinkage. Increasing concentrations of BAPN led to striped patterns of feather rudiments, increases in feather rudiment diameter, and increases in skin thickness in cultured skin. These changes in morphology were paralleled by a progressive decrease in collagen crosslinkage as the concentration of BAPN was increased. However, these changes were also paralleled by an increase in cell death in the epidermis. Culture of monolayers of separated epidermal and dermal cells in the same concentration range of BAPN that altered feather development in the organ cultures showed that BAPN killed epidermal cells but not dermal cells. Our results suggest that developmental alterations caused by BAPN may, in some cases, be due to acute toxicity to specific cell types, and not simply to inhibition of collagen crosslinkage.

Cloning and sequencing of goldfish activin subunit genes: strong structural conservation during vertebrate evolution.

Inhibin and activin are structurally related dimeric peptides sharing two distinct but related subunits, beta A and beta B. By using polymerase chain reaction (PCR), we have cloned and sequenced activin beta A and beta B genes encoding the mature region of the peptides from the goldfish genomic DNA. A single form of beta A and two forms of beta B subunits were found. There is high identity with mammalian counterparts; the goldfish beta A subunit has 78% protein sequence identity with human beta A subunit, and the two beta B subunits have more than 94% sequence identity with human beta B subunit. This is the first demonstration of the activin family of peptides in the fishes.

Development of bile canaliculi between chicken embryo liver cells in vivo and in vitro.

The development of an organized network of bile canaliculi is essential for the normal functioning of the liver. We have characterized bile canaliculus development in situ from Days 3-19 and in vitro in cultured hepatocyte monolayers using electron microscopical and immunofluorescent staining with antibodies that specifically recognize antigens of the bile canaliculus. Although the liver first forms as a discrete epithelial bud of endodermal tissue at stage 12-14 (45-53 h after laying), canaliculi were first detected by our antibodies at low levels in 4-day embryos and at high levels in stage 27 (5 days after laying) and later embryos. During Days 4, 5, and 6 the canaliculi near the periphery of the rudiment do not stain while canaliculi in central areas, closer to the gut, are strongly stained. During this transition period the ultrastructure of the canaliculi in the peripheral regions is also less developed than the central canaliculi where the antigens appear. By 7 days post laying, canaliculi throughout the entire liver rudiment express the marker antigens equally and have the ultrastructural characteristics of mature, functional canaliculi. Cells prepared from liver of embryos of 11 days incubation and grown in monolayer culture reformed discernible canalicular specializations, as determined by immunofluorescent staining and electron microscopy, but only transiently (for 1 to 3 days after plating). Not all of the antigens were expressed or polarized in these cultures. The capacity of the embryonic parenchymal cells to develop and maintain polarity appears to depend on factors possibly including age-dependent changes in the cells themselves, interactions with other cell types or extracellular matrix, or the shape of the cells.

Structural variants of the neural cell adhesion molecule (N-CAM) in developing feathers.

The neural cell adhesion molecule (N-CAM) is expressed in a specific spatiotemporal pattern during feather development, suggesting that adhesion mediated by this molecule is involved in feather morphogenesis. To begin to investigate N-CAM's function in developing feathers, we determined what forms of N-CAM polypeptide are present and the distribution of polysialic acid (PSA), a carbohydrate moiety that decreases N-CAM-mediated cellular adhesion. N-CAM in skin appears as a Mr 145-kDa polypeptide compared to the 140-kDa brain N-CAM polypeptide, and is encoded by a 6.4-kb mRNA, compared to the 6.1-kb mRNA in brain. Polymerase chain reaction analysis of the exon splicing pattern of skin N-CAM shows that the 6.4-kb mRNA band represents two transcripts, with and without a 93-bp insert between exons 12 and 13. Thus, two N-CAM polypeptides are expressed in skin, but the 93-bp insert does not account for the larger size of the skin mRNAs and polypeptides. We show that the size difference of the polypeptides is instead due to N-linked oligosaccharides attached to the skin N-CAM proteins. The larger size of the skin mRNAs may be due to use of a different transcriptional start site. Staining of skin sections and wholemounts confirms previous descriptions of N-CAM in developing feathers, but reveals that N-CAM is also present at low levels on epidermal cells as early as stage 29 (E6). We find that PSA is expressed only on a subset of the cells that express N-CAM, in particular on dermal cells in the feather rudiments from stage 35-36 (E9-10) and on smooth muscle cells at the base of the filaments from stage 37 (E11) until the latest stage examined (stage 44, E18). The known effects on cell-cell adhesion of amount of N-CAM and PSA suggest that the variations we observe in skin may regulate cell-cell interactions that are important in feather development.

Genes for two calcium-dependent cell adhesion molecules have similar structures and are arranged in tandem in the chicken genome.

Genomic sequences immediately upstream of the translational start site for the chicken liver cell adhesion molecule (L-CAM) gene contain a second closely related gene, which, because of its location, we have designated the K-CAM gene. Less than 700 base pairs separate the presumed poly(A) site in the K-CAM gene from the translation initiation site for L-CAM. The sizes of exons 4-15 of the K-CAM gene are almost identical to those in the L-CAM gene and the exon/intron junctions occur at exactly equivalent positions in both genes. Exon 16, which includes the 3' untranslated region, is much shorter in the K-CAM gene and intron sizes and sequences are not generally conserved between the two genes. Probes from the K-CAM gene hybridized to a 3-kilobase mRNA that was present at high levels in embryonic skin, at lower levels in kidney, heart, and gizzard, and at still lower levels in brain and liver, as determined by Northern blotting. The sequence of the predicted gene product was nearly identical to that of the chicken B-cadherin cDNA, although the distribution of the K-CAM gene transcript differed from that reported for the cadherin. The proximity and identical overall structure of the K-CAM and L-CAM genes strongly suggest that they arose by gene duplication and raise the possibility that genes for other calcium-dependent CAMs may be located in clusters. Moreover, the tandem arrangement of the genes may have important implications for the regulation of their expression.

cDNAs of cell adhesion molecules of different specificity induce changes in cell shape and border formation in cultured S180 cells.

The liver cell adhesion molecule (L-CAM) and N-cadherin or adherens junction-specific CAM (A-CAM) are structurally related cell surface glycoproteins that mediate calcium-dependent adhesion in different tissues. We have isolated and characterized a full-length cDNA clone for chicken N-cadherin and used this clone to transfect S180 mouse sarcoma cells that do not normally express N-cadherin. The transfected cells (S180cadN cells) expressed N-cadherin on their surfaces and resembled S180 cells transfected with L-CAM (S180L cells) in that at confluence they formed an epithelioid sheet and displayed a large increase in the number of adherens and gap junctions. In addition, N-cadherin in S180cadN cells, like L-CAM in S180L cells, accumulated at cellular boundaries where it was colocalized with cortical actin. In S180L cells and S180cadN cells, L-CAM and N-cadherin were seen at sites of adherens junctions but were not restricted to these areas. Adhesion mediated by either CAM was inhibited by treatment with cytochalasin D that disrupted the actin network of the transfected cells. Despite their known structural similarities, there was no evidence of interaction between L-CAM and N-cadherin. Doubly transfected cells (S180L/cadN) also formed epithelioid sheets. In these cells, both N-cadherin and L-CAM colocalized at areas of cell contact and the presence of antibodies to both CAMs was required to disrupt the sheets of cells. Studies using divalent antibodies to localize each CAM at the cell surface or to perturb their distributions indicated that in the same cell there were no interactions between L-CAM and N-cadherin molecules. These data suggest that the Ca(++)-dependent CAMs are likely to play a critical role in the maintenance of epithelial structures and support a model for the segregation of CAM mediated binding. They also provide further support for the so-called precedence hypothesis that proposes that expression and homophilic binding of CAMs are necessary for formation of junctional structures in epithelia.

Construction of epithelioid sheets by transfection of mouse sarcoma cells with cDNAs for chicken cell adhesion molecules.

Pleiomorphic mouse sarcoma S180 cells were transfected with cDNAs for the liver cell adhesion molecule (L-CAM), the neural cell adhesion molecule (N-CAM), or both CAMs. Transfected cells expressed the appropriate CAMs at their surface and those expressing L-CAM (S180L cells) changed from adjoining spindle or round shapes to a closely linked "epithelioid" sheet when grown to confluence. Cells transfected with cDNA for N-CAM (S180N cells) also expressed this CAM on the cell surfaces and bound brain vesicles containing N-CAM but showed no phenotypic change to an epithelioid state. In S180L cells and doubly transfected (S180L/N) cells, L-CAM was concentrated at regions of cell contact and was codistributed with cortical actin. In S180N cells, N-CAM was uniformly distributed on the cell surface. When S180L cells were cocultured with S180L/N cells, N-CAM was not concentrated at boundaries between the S180L and S180L/N cells but was concentrated at boundaries between pairs of S180L/N cells. Fab' fragments of anti-L-CAM dissociated the epithelioid sheets of S180L or S180L/N cells into cells with shapes resembling those of untransfected cells. Cells in epithelioid sheets were polygonal in shape but, unlike cells in true epithelia, had no basement membrane or polar structure; they also lacked tight junctions and desmosomes. Ultrastructural examination showed that, in contrast to the untransfected phenotype, cells in epithelioid sheets had large increases in adherens junctions and gap junctions. Dye coupling experiments indicated that the gap junctions were functional. The frequency of expression of both kinds of junctions was sharply decreased by treatment with anti-L-CAM Fab' fragments. These experiments provide support for the precedence hypothesis, which proposes that the linkage of cells by means of CAMs is a necessary event for the extensive expression of junctional structures.

Cellular expression of liver and neural cell adhesion molecules after transfection with their cDNAs results in specific cell-cell binding.

Mouse L cells, which do not express the known primary cell adhesion molecules (CAMs), were permanently transfected with vectors containing the simian virus 40 early promoter and cDNA sequences encoding chicken liver CAM (L-CAM) or each of the three major polypeptide forms of chicken neural CAM (N-CAM). Transfected cells in culture expressing the Ca2+-dependent L-CAM showed uniform surface expression of the molecule. Unlike untransfected L cells, these cells aggregated readily; the aggregation was inhibited by Fab' fragments of antibodies to L-CAM but not by fragments of anti-N-CAM. These cells spread more efficiently in culture than did their untransfected counterparts, forming small colonies of flattened cells that gradually assumed morphologies resembling closely packed L cells. Transfected L cells expressing either the small or large intercellular domain polypeptide (sd or ld) chains of N-CAM aggregated specifically with each other or bound membrane vesicles from chick brain. Both types of binding were specifically inhibited by Fab' fragments of anti-N-CAM antibodies. These cells, in contrast to those transfected with vectors for L-CAM, showed rounded morphologies and spread inefficiently in culture. L cells transfected with vectors specifying the small surface domain polypeptide (ssd) chain of N-CAM showed no phenotypic changes and no evidence for linkage of ssd chains to the cell membrane by phosphatidylinositol intermediates. Instead, these cells synthesized the molecule and released it into the medium. These findings complete the demonstration that different CAMs have specific roles in ligating the cells that synthesize them, and they provide further evidence that L-CAM and N-CAM bind by homophilic mechanisms. The different phenotypic changes observed for each specific CAM are consistent with the hypothesis that CAM synthesis or differing associations of CAM carboxyl-terminal domains with the cell surface and cortex may lead directly or indirectly to specific alterations in the cells bound together by that CAM.