[
International C. elegans Meeting,
2001]
Homologues of claudin, major integral protein of mammalian tight junction with four span transmembrane structure, were found in the genomic data base of C. elegans . Among them, CLC-1(C09F12.1) was discovered by a blast search with mouse claudin-6, it was found later that homology is also present with mouse claudin-7 and human claudin-14. It has in its first loop two Cys residues that are conserved among mammalian claudins. CLC-2 (T05A10.2) was found by a blast search with CLC-1, and has 33% identity (70% homology) with CLC-1, although no significant homology was found with vertebrate claudins so far known. CLC-3 (ZK563.4) has homology with mouse claudin-10 and cattle claudin-16. CLC-4 was found with a blast search with mouse claudin-6, but it was thought to be an eight span transmembrane protein at that time. Recently, it was found that this gene (C01C10.1) encodes two separate proteins, and the down-stream operon (C01C10.1b) was reported to encode a Gas3/PMP-22 homlogue (Agostoni E.et al., Gene, 234, 267-274, 1999). Product of the other operon (C01C10.1a) was not characterized well. Homology with claudin-6, -7, and -9 was found with this product, therefore named as CLC-4. These four claudin homologues have either PMP-22/EMP/MP20 motif or transmembrane four signature or both, and most of them have srg integral membrane protein motif. Furthermore, two conserved Cys residues are present in all C. elegans claudins. Interestingly, many of vertebrate claudins have one or two of these motifs or signature. Although most of vertebrate claudins has CLAUDIN3 signature, no such signature was found in the nematode claudins. All of coding sequences of the nematode claudins were isolated from cDNA libraries, therefore they are all expressed in the worm. Some ESTs were reported for CLC-1 and -3, previously. Expression of these claudins was studied with GFP-tagged molecules. All of claudins is expressed in spermatheca, intestine and hypodermis. CLC-1 and -4 were expressed strongly in pharynx, and sometimes localized at cell-cell junctions. They are also seems to be expressed in excretory-secretory system. CLC-1::GFP is also expressed at cell-cell junction of vulva. Localization of CLC-1 is under study with HA-tagged molecule. Preparation of antibodies against the nematode claudins was very difficult so far. But, affinity purified antibodies raised against loop 1 of CLC-1 seem to be useful for CLC-1 detection in the worm. Results with these antibodies were very similar to those obtaind with GFP-tagged CLC-1. To see if these claudins function as mammalian claudins do, i.e. barrier function, penetration of TRITC-dextran (MW=10,000) was checked after injection of dsRNA's (RNAi). Experiments with full length CLC-1 dsRNA showed that barriers for the high molecular weight dye were damaged by RNAi, in other words, penetration of the dye to pharynx and some other tissues were observed. Similar experiments with other claudins did not detect any barrier damage. This is because the dye only goes into entrance of intestine under normal condition, therefore, we tried weak osmotic shock to deliver the dye to entire intestinal lumen and excretory-secretory duct system. Under this condition, control injection of dsGFP did not result in penetration of the dye to other area of the body. On the other hand, RNAi with the combination of CLC-1 and -4 RNA's resulted in penetration of the dye to 72% of the worm from pharynx, intestine and vulva (or from excertory-secretory system) to the body, whereas barrier was damaged only 40% of worm with a combination of CLC-3 and -4. RNAi effects with other combination of CLC's will be reported, and effects of RNAi to the retention of the sperm in spermatheca will be studied. Accordingly, claudins of C. elegans seems to function as barrier at least partly. Other functions, if any, will be surveyed. We would like to thank excellent technical assistance of Miss. Akiko Kamamoto, without her help this project could not be completed.
[
International C. elegans Meeting,
1999]
Recently, new integral proteins of tight junction were discovered in mouse and human (Furuse et al., J. Cell Biol., 141, 1539, 1998; Morita et al., PNAS, 96, 511, 1999). These claudin family proteins are members of tight junction strands. Although presence of tight junctions in C. elegans is not reported, septate junctions and septate-like junctions seem to play similar functions instead. We searched the gene database of C. elegans , and found two homologues of claudin family proteins (claudin-CE1 and -CE2) with four-transmembrane domains, conserved two Cys in the first loop, and similar molecular weight. Interestingly, a protein (claudinD) was also found that has molecular weight about twice of claudin-CE1, and other characteristic structures are likely to have two claudin molecules tandemly repeated. These 3 proteins are coded from nearby sites on chromosome X. Claudin-CE1::GFP with 1.2kb upstream promoter region was expressed in spermatheca which is known to have septate junctions, and gut. Expression of claudin-CE2::GFP was much less, but tissue distribution was similar. RNAi experiments using dsRNA mixture of claudin-CE1, claudin-CE2 and Exon1-4 fragment of claudinD were performed. About 40% of F1 of the injected worms have decreased F2 production (in average 48% decrease), whereas 22% of F1 have almost normal numbers of F2's. Thus, these proteins seem to be important for reproduction of the worms. When expressed in MDCK-II epithelial cells, Claudin-CE1::GFP was localized at cell-cell junctions. Electron microscopic studies are under way. We are grateful to Miss. Akiko Kamamoto whose technical assistance make this work possible.