Itoh M et al. (1994) Worm Breeder's Gazette "Cloning of the cej-1 Gene Encoding a Novel Cell-Junction Protein in Caenorhabditis elegans M.L.A. Khan, M. Tabish, T. Fukushige, S. Tsukita, S. Tsukita,"

Itoh M, Siddiqui SS, Yasuda H

[Worm Breeder's Gazette, 1994]

Intracellular tight junctions, such as zonula occludentes are known to form regulated barrier zones between both endothelial and the cells of epithelium, and provide a means to separate tissue compartments in higher animals. Recently a number of genes encoding "tight junction" proteins have been characterized from different species, such as the discs large tumor suppressor dlg gene of Drosophila (Woods & Bryant, 1991); human erythrocyte membrane protein p55 (Cho et al., 1992), PSD-95 from the rat brain synapses (Sadler et al, 1992), ZO-1 from human cells (Willot et al., 1993), and 220 kDa protein from mouse (Itoh et al., 1993). Although extremely different in the length of the primary sequence, a hallmark of these proteins is the presence of the SH3 and guanylate kinase domain that may not have the proper catalytic function. Using the monoclonal antibody T4192 raised against the mouse 220 KDa protein, we have screened a C. elegans cDNA expression library and cloned a cDNA that encodes a novel protein with no strong homology to the tight junction proteins, but shows the presence of the conserved amino acid residues in the SH3 and GK domains. The cDNA probe recognizes only one clone on the polytene filter, the Y50C3 ,that places the (cej-1 , cell junction -1)gene in the cluster on chromosome III near the mab-21 locus, in a region covered by three cosmid clones (F35G12 ,T23C9 ,and F1 OG11).Screening a genomic library, we have identified a positive clone corresponding to the cej-1 cDNA probe, and have subcloned a 9 kb genomic fragment that includes the gene. Southern blot analysis suggests that the cej-1 is a single copy gene, and the Northern blot data indicates that it encodes a mRNA of 1.5 kb. Immunocytochemical staining of the nematode on wholemount squash preparations using T4192 antibody which recognizes a 34 kDa protein on immunoblots, reveals that the antigen is present during embryogenesis at cell junctions both during the early and late embryonic development. In the larvae and adults, staining of the neural tissue and the excretory canal (H - Cell) is observed. We thank A. Coulson and other members of the Sanger Centre for their assistance in providing the YAC clones.

Savage C et al. (1994) Worm Breeder's Gazette "More tolloid Homologs in the Worm."

Padgett RW, Finelli AL, Savage C, Cho SH

[Worm Breeder's Gazette, 1994]

More tolloid homologs in the worm Cathy Savage, Seo Hee Cho, Alyce L. Finelli, and Richard W. Padgett. Waksman Institute and Department of Molecular Biology and Biochemistry, Rutgers University, Piscataway, NJ 08855. Members of the tolloid/BMP-1 family of metalloproteases are thought to interact with extracellular growth factors to regulate a variety of developmental events. In particular, the Drosophila gene tolloid interacts with decapentaplegic (dpp), a member of the TGF-,B superfamily. Similarly, vertebrate BMP- 1 co-purifies with other BMPs (bone morphogenetic proteins) that are related TGF-,B-like ligands. Proteins in the tolloid/BMP-l family share a similar structure: they contain a predicted metalloprotease domain; EGF repeats; and CUB repeats, which are also found in blood complement proteins. These proteases are thought to interact with other proteins through the EGF and CUB repeats, and to regulate their activity through proteolysis. We have previously reported sequence analysis of the first C. elegans member of this metalloprotease family (WBG 12(4), p.30), which we have called toh-l (tollisk). Since then, the genome sequencing consortium has identified four more predicted ORFs with sequence similarity to tolloid. We have therefore isolated and sequenced cDNAs for these genes, to determine whether these new genes encode true members of the tolloid/BMP-1 family. One of the putative homologs displayed no protease homology. The predicted structures of the other three homologs are summarized below. Two of these, on the cosmids COSDl 1 and F42A10, contain related metalloprotease domains, but none of the characteristic protein interaction motifs of the tolloid/BMP-l family. Only one homolog, on the cosmid R151, was found to contain an EGF repeat and a CUB repeat. We therefore propose the name toh-2 for this gene. toh-2 represents a novel member of this family, because it contains repeats (TSP repeats) found in thrombospondin, an extracellular matrix protein. The presence of these two tolloid homologs in the worm raises several interesting questions: (1) Do these homologs act together in a TGF-,B-like signal transduction pathway, in separate pathways, or combinatorially with other, as yet unidentified, tolloid-like proteins? (2) What roles do these genes have in nematode development? (3) What is the function of the toh-2 thrombospondin repeats? Are there homologs in other organisms that also contain these repeats?

Blaxter ML (1991) Worm Breeder's Gazette "Surface-Directed Biotinylation of C. elegans Foiled: An Endogenous Avidin-Binding Protein"

Blaxter ML

[Worm Breeder's Gazette, 1991]

Specific labelling of (glyco)proteins on the surface of C. elegans will allow their isolation, characterization and, ultimately, cloning. I reported earlier (WBG 11.2) on the use of lodogen-mediated radio- iodination, and presented further data derived from lodogen studies at the Worm meeting in Madison. Both lodogen (a solid-phase reagent which attacks tyrosine residues) and Bolton-Hunter reagent (a lipophilic iodination reagent attacking lysines) label a small set of proteins on the surface of C. elegans. While labelling efficiencies are an order of magnitude less when compared to parasitic species such as Brugia, they do do the job. One drawback is the need for [125I] and its associated hazards. The biotin-avidin system offers an alternative nonradioactive tag which has been widely used in immunochemistry. Reactive biotin derivatives are available which will attack amino groups (N-hydroxysuccinimidobiotins [NHS-B]) and aldehydes (biotin hydrazide [BHz]). Biotinylated proteins can be analyzed by SDS-PAGE and western blotting with an avidin-peroxidase conjugate. NHS-B has been used to label surface antigens of Brugia and Dirofilaria parasitic nematodes, where it displays similar specificity to lodogen-mediated iodination. I have tried NHS-B labelling on larval and adult C. elegans N2 using both NHS-B (dissolved in DMF) and a water-soluble sulfo- derivative, SNHS-B (both from Pierce). Labelings (of ~25 adults) were carried out in 10 l of NaHCO3 buffer, pH 8.5, at biotin ester concentrations of 0.1 to 1.0 mg/ml. At room temperature, biotinylation of a large number of products was seen. The profile of products basically followed the Ponceau S staining pattern of the protein content of the worms, indicating that the reagents had penetrated deeply. Reducing the incubation time reduced the intensity of labelling slightly but did not increase specificity. The worms probably swallowed the NHS-B or SNHS-B. The results from labelings carried out on ice were at first encouraging. At all NHS-B and SNHS-B concentrations and at all time points (5 min to 30 min) only a single molecule of 80 kDa appeared to be labelled. It was present in both adults and L1/L2 larvae. However it was also present in unlabeled worm extract. The reactivity seen is not an endogenous peroxidase (no reaction with substrate alone) or a peroxidase-binding protein (it is not seen in ordinary westerns). The 80 kDa band was also present in other free-living nematodes tested (Rhabditis, Oxycerca, Panagrellus, C. briggsae). Two possibilities are: [1] It is a biotinylated protein. The biotin groups would have to be attached by covalent bonds not reduced by 2-ME. [2] It is an avidin-binding protein. Avidin is a 66 kDa tetramer of identical 128 amino acid subunits, each of which carry a single asparagine-linked carbohydrate sidechain. The 80 kDa could recognize either the CHO (and would thus be a 'lectin') or the protein (or both). I favor the biotinylated protein model. BHz will label glycoproteins if the carbohydrates are first derivatised with periodate to produce exposed aldehyde groups. I have not tried this reagent, because [1] 10 mM NaIO4 doesn't sound too good for worms, [2] the reactions should be carried out at room temperature (when the worms will swallow the NaIO4 and the BHz) and [3] the NaIO4 derivatisation works on mammalian glycoproteins by attacking sialic acid residues, but C. elegans doesn't have any surface sialic acid ( as determined by radiolectin assays). Apparent bottom line: Biotinylation isn't an alternative to [125I]. Shame really. However, the observation that the freeliving nematode surface is refractory to attack by NHS esters is interesting, and distinguishes it from most other biological surfaces. It suggests that [1] there are no exposed amino groups on the surface of C. elegans [or, if there are they aren't on glycoproteins] and [2] the cuticle is highly impermeable. Perhaps worms with more permeable cuticles, a phenotype which might be found in srf mutants or in mutants especially sensitive to SDS, could be extrinsically biotinylated. Such biotinylation might then provide an assay and probe for cuticle permeability and the determinants thereof.